Volume 1 Synthesis and Reactivity
2
1 Main Group Metallocenes
1
Main Group Metallocenes Peter Jutzi and Neil Burford
1.1 Abstract Cyclopentadienyl complexes of the elements of groups 1, 2, 13, 14 and 15 are classified and compared in terms of their synthesis, characteristic properties and structural types, which include a variety of sandwich and half sandwich complexes as well as dimeric, oligomeric and polymeric arrangements. The variety of structure and bonding features reveal exciting new directions in the coordination chemistry of the main group elements, and the factors governing the structures of multicyclopentadienyl complexes are discussed.
1.2 Introduction The metallocene icon of organometallic chemistry originally refers to bis(cyclopentadienyl)metal complexes (‘sandwich’), but wider usage is now accepted to include cyclopentadienyl complexes (‘half sandwich’) and multicyclopentadienyl complexes (‘multidecker sandwich’) as well as complexes with additional substitution at the metal center [1]. Nevertheless, the metallo- prefix is restrictive when one considers the more recent development of complexes involving nonmetallic elements such as boron, silicon or arsenic. In fact, metallocene-like complexes are now known for many elements in the periodic table. This chapter provides an overview of the extensive series of compounds in which a main group element (Groups 1, 2, 13, 14 and 15) is π-bound to one or more cyclopentadienyl ligands. Synthetic procedures, characterization features, structural types and bonding are classified in general terms and then classic examples of complexes involving elements from each of the periodic groups are examined in some detail [2 ]. The discussion is restricted to complexes involving a 6πelectron cyclopentadienyl ligand, however, it is important to recognize that an extensive chemistry is also developing for main group complexes involving cyclopentadienyl ligands with extended π-frameworks [3 ], as well as heterocyclopentadienyls [4 ], and other isolobal analogs of cyclopentadienyl such as arene [3, 5 ], carbollyl [3], and borazolyl [6 ]. Current assessments of main group metallocene chemistry are delayed by about two decades in comparison to the established chemistry of transition metallocenes,
4
1 Main Group Metallocenes
and most conclusions derive from solid state and gas phase structural data. While these compounds have features in common with transition metal analogs, a new avenue of organometallic chemistry is defined by structural novelty and synthetic utility. Most complexes are air sensitive materials and involve standard formal oxidation states, but the cyclopentadienyl ligand imparts unusual stability to the lower oxidation state for the elements of group 13 (+1) and group 14 (+2). Main group elements interact with cyclopentadienyl ligands in a variety of fashions [3, 7 , 8 ] which we specify structurally in terms of the position of the element with respect to the pentagonal cylinder defined by the cyclopentadienyl ring. For compounds in which the element resides inside the cylinder, the element may be closer to one carbon center of the cyclic frame, which is referred to as a monohapto interaction, and is abbreviated as η1. If the element is equidistant from more than one carbon center of the cyclic frame, the complexes are described as having hapticities of η2, η3, η4 (rare) or η5, respectively. Each of these types of complex (η1–η5) are classified here as metallocenes. When the element resides outside the cyclopentadienyl cylinder, the carbon atom to which it is closest usually exhibits significant distortion from trigonal planarity. In addition, the bond lengths and angles within the cyclopentadienyl ring are nonequivalent and tend towards those of cyclopentadiene [9]. Such compounds are considered to involve sp3 hybridization at the element bound carbon center and are classified as σ-complexes. Di- and tri-hapto complexes are sometimes described as involving peripheral coordination [10, 11], which imposes a substantial distortion on the cyclopentadienyl plane for transition metal analogs and a localized double bond, as illustrated in 1. In contrast, main group metallocenes typically exhibit relatively minor perturbation of the C–C bonds within the cyclopentadienyl framework and the planarity is retained as the hapticity is apparently decreased, as depicted in 2. Moreover, the energetic distinctions between σ- and multi-hapto structures are small for main group elements [12]. This accounts for routine observation of haptotropic shifts and consequential fluxionality [13], as well as the observation of a single 13C NMR resonance for the carbon atoms of the cyclopentadienyl cyclic frame in the solid state spectra of complexes that are shown to be less than pentahapto coordinated on the basis of X-ray crystallographic studies [14]. Most convincing are the recent experimental and theoretical studies of tricyclopentadienylaluminum derivatives, which reveal negligible energy differences (4–8 kJ · mol–1) between hapticities [15]. Clear distinction between σ and π interaction is not possible for many cyclopentadienyl derivatives of the main group elements, and there are uncertainties in the classification of hapticity.
1
2
1.3
Sythetic Procedures
5
The definition of a bonding model for cyclopentadienyl derivatives is further complicated by the degree of ionic character in the ligand–element interaction. A much wider range of electronegativities is available in the main group than for the transition metal elements. Therefore, while most p-block metallocenes are considered to involve covalent interactions between the element and the cyclopentadienyl ligand, the alkali and heavier alkaline earth metallocenes are viewed as principally ionic, and yet both have structural features related to the transition metal analogs. This diversity in bonding types has resulted in the evolution of two compound nomenclatures. The more ionic compounds are usually referred to as ‘element cyclopentadienides’ with the formulation ECpn, while the covalent compounds are usually referred to as ‘cyclopentadienylelement’ with the formulation CpnE. We retain both types of nomenclature and formulation in this chapter but do not necessarily infer a type of bonding, as most compounds exhibit both ionic and covalent characteristics. The development of main group metallocenes has been facilitated by the use of substituted cyclopentadienyl ligands. The most common of these is pentamethylcyclopentadienyl (abbreviated as Cp*), which often gives complexes greater thermal stability than the parent cyclopentadienyl analogs [16, 17]. The range of other derivatized ligands are abbreviated here as CpnSUBSTITUTION, where the superscript indicates the number (n = 1–5) and type of the substituents (M = methyl, I = isopropyl, B = tertiary butyl, Ph = phenyl, Bz = benzyl, D = dimethylsilyl, T = trimethylsilyl, P = diphenylphosphino; Cp* = Cp5M; a number of unique and mixed substitution patterns are defined when referred to in the text; Cp' is used as a general abbreviation for Cp and substituted derivatives). The bulky nature of many Cp' ligands can also impose interesting distortion on the complex [18, 19, 20, 21], but less obvious is the fact that they are good leaving groups, allowing for substitution reactions with element–carbon bond cleavage in many situations [16, 22]. The introduction of a side chain functionality (CpN, CpO, N = dimethylaminoethyl, O = methoxyethyl) with the potential for auxiliary coordination to the element center also offers a new dimension to the development of metallocenes with diversified chemistry and novel molecular structure [23, 24].
1.3 Synthetic Procedures Consistent with general synthetic principles of organometallic chemistry, the formation of main group element–Cp' bonds can be most conveniently achieved using metathesis reactions. For this reason, the alkali metal cyclopentadienides, magnesium dicyclopentadienides and thallium cyclopentadienides are important in terms of their role as synthetic sources of Cp' ligands. This and other generally applicable methods, as well as specific reactions (used for particular complexes), are classi-
6
1 Main Group Metallocenes
fied below with key examples, and synthetic procedures for particular complexes can be found in the references given in the discussion of those complexes.
1.3.1 Reactions Involving Cyclopentadienes 2E + 2Cp'H → 2ECp' + H2 or E + 2Cp'H → ECp'2 + H2 ER + Cp'H → ECp' + RH EH + Cp'H → ECp' + H2 ENH2 + Cp'H → ECp' + NH3
(1-1) (1-2) (1-3) (1-4)
Cyclopentadiene carries a relatively acidic proton and reacts smoothly with a number of the more reactive main group elements Eq. (1-1), such as sodium [25] and magnesium [26], and certain metallocenes can be prepared by co-condensation of the cyclopentadiene with the element [27, 28, 29]. LiCp is most conveniently prepared by reaction of a lithium alkyl (usually LinBu) with cyclopentadiene Eq. (1-2) [30]. This procedure has been extrapolated to other Cp' derivatives [31, 32], and applies to other organometallic reagents such as MgMe2 [33, 34, 35] and Al2Me6 [36]. Sodium or potassium cyclopentadienides can be obtained from reaction of the element hydride with cyclopentadiene Eq. (1-3), or reaction of the element amide with the cyclopentadiene in liquid ammonia Eq. (1-4) [32]. A similar amide displacement reaction is used for calcocene [37] and barocene [8] derivatives, although some of these complexes are formed in donor solvents and are consequently isolated with these solvents bound as auxiliary ligands. Cp'Tl derivatives are routinely prepared from the corresponding cyclopentadiene Eq. (1-2) and thallium hydroxide or thallium salts (usually TlOEt) in basic aqueous solution [38, 39].
1.3.2 Salt Metathesis Reactions nMCp' + EXn → Cp'nE + nMX (M = Li, Na) or n/2MgCp'2 + EXn → Cp'nE + n/2MgX m
(1-5)
The most generally applied procedure for the preparation of metallocenes involving the elements of groups 2, 13, 14 and 15, is the appropriate stoichiometric combination of an element halide (X) or polyhalide with an alkali metal cyclopentadienide salt (usually lithium or sodium) or a magnesium dicyclopentadienide salt [3, 18, 40 ]. Complexes of aluminum and gallium involving the element in less familiar low oxidation states are obtained from metastable halides ‘AlCl’ [41 ] and ‘GaCl’ [42 , 43 ], which are formed in situ. In comparison, InCl [44 ] is readily available as a starting material, and a variety of thallium(I) salts can be used [45 ].
1.3
Sythetic Procedures
7
1.3.3 Disproportionation, Comproportionation and Decomposition 2Cp'EX → Cp'2E + EX2 and 2Cp'EX ← Cp'2E + EX2
(1-6)
Magnesocene and its methylated derivatives can be prepared by thermal decomposition of Cp'MgBr [46, 47, 48, 49], and the opposite comproportionation reaction can be used to obtain substituted derivatives of half sandwich complexes for magnesium [50], germanium [51] and tin [52, 53]. Reductive elimination of Cp ligands from Cp3In [54] and of phenyl groups from Cp2PbPh2 [55] is also thermally induced.
1.3.4 Halide Ion Abstraction Cp'nEX + A → [Cp'nE][AX]
(1-7)
Lewis acids such as aluminum halides (A) are well known as effective halide ion abstraction agents, thereby generating a cationic system from a formerly covalent halide. The procedure is responsible for the formation of half sandwich complexes (n = 1) of boron [56, 57], germanium [3] and tin [58, 59, 60, 61] and the sandwich complexes (n = 2) of arsenic [62], antimony [63], and bismuth [64].
1.3.5 Substitution onto the Cp Ligand Frame ECpn + AR → CpRECpn–1 + AH or ECpn + nAR → ECpRn–1 + nAH
(1-8)
Thallium and tin metallocenes have a relatively high stability, which allows for modification of the Cp ligand with retention of the complex. The procedure has contributed to the preparation of a particularly extensive series of stannocenes, including Cp' ligands bearing silyl [65], stannyl [66 ], phosphine [67], or phosphenium [68] moieties, and is responsible for the formation of the novel Cp tricyanovinylTl [69, 70].
1.3.6 Reduction Cp'nEX2 + M → Cp'nE + MX2
(1-9)
Reductive dehalogenation can be achieved for silicon, germanium and tin in Cp*2EX2 [71, 72], for aluminum in [Cp*AlCl2]2 [73] and can be thermally promoted for Cp3In [54] and Cp2PbPh2 [55] (see above).
8
1 Main Group Metallocenes
1.3.7 Addition of Cp Anions Cp'E + [Cation][Cp'] → [Cation][Cp'2E]
(1-10)
Additional cyclopentadienide anions can be introduced onto lithium, sodium, cesium, thallium, tin and lead cyclopentadienyl compounds to give anionic multicyclopentadienyl complexes. This is achieved by reaction of the neutral metallocene with sources of nucleophilic cyclopentadienide in salts with complex cations such as tetraphenylphosphonium [74, 75] and trisdimethylaminosulfonium [76].
1.4 Characterization Features Characteristic infrared and Raman spectroscopic features were pivotal in the early identification and assignment of main group metallocenes, however, as with many areas of modern inorganic chemistry, X-ray crystallography has played the leading role in the discovery and development of these compounds. Consequently, the majority of the discussion in this chapter focuses on structural features in the solid state, and the implications for bonding and reactivity. Nevertheless, there are a number of important spectroscopic and physical properties to make note of. The most enlightening spectroscopic feature of main group metallocenes is the distinctive element NMR chemical shift (for elements with NMR active nuclei), examples of which are listed in Table 1-1 in comparison with typical δ values Table 1-1. Examples of element chemical shifts (ppm) for metallocene derivatives Nucleus 7Li 6Li 27Al 27Al 25Mg 25Mg 29Si 71Ga 119Sn 119Sn 207Pb 207Pb
Metallocene
Ref.
δ
δ [ERn]
Ref.
LiCp3T [Li(12-crown-4)2][LiisoCp2] Cp*Al ({Cp*Al}4) [Cp*2Al][Cp*AlCl3] Cp*Mg·OEt2 Cp*2Mg Cp*2Si Cp*Ga [Cp*Sn][BF4] Cp*2Sn [Cp*Pb][BF4] Cp*2Pb
[82 ] [78] [84 ] [86 ] [85] [85] [87 ] [43] [81] [81] [81] [81]
–11 –13 –150 (–80) –114 –34 –78 –398 –653 –2247 –2129 –5041 –4390
0 to –0.7 [LiMe] –2 to +2 [LiR] 153 [AlMe3] 153 [AlMe3] 99 [MgEt2] 99 [MgEt2] 0 [SiMe4] 730 [GaMe3] 0 [SnMe4] 0 [SnMe4] 0 [PbMe4] 0 [PbMe4]
[83 ] [78] [85 ] [85] [85] [85] [88 ] [89 ] [88] [88] [88] [88]
1.4
Characterization Features
9
(ppm) for alkyl derivatives of those elements. The dramatic upfield shift is generally rationalized in terms of ring current phenomena [77, 78, 79]. Observation of coupling between Cp' nuclei and NMR active elements can be interpreted as an indication of covalent character for the E–Cp' interaction and the JE–H coupling for derivatives of tin and lead represent appropriate examples. The simulated 31 line multiplet for Cp*2Sn and Cp*2Pb in the 119Sn and 207Pb NMR spectra, respectively, are observed (Sn δ, –2129 ppm; 3JSn–H, 3.7 Hz [72, 80 ]; Pb δ, –4384 ppm; 3JPb–H, 11.2 Hz) [81] as 16 lines for intensity reasons, as shown for the stannocene in Fig. 1-1. Melting point trends for the sandwich metallocenes of groups 2 and 14, summarized in Table 1-2, have been analyzed in detail [21]. The parent sandwich metallocenes and symmetrically substituted derivatives (e.g. decamethyl) typically have a relatively high melting point in comparison to derivatives with single or asymmetric substitution patterns around the Cp' ligand. The observations are interpreted
119Sn
Figure 1-1 NMR signal for Cp*2Sn and comparison with the simulated signal
10
1 Main Group Metallocenes
in terms of the disruption of crystal packing patterns with the introduction of partial substitution which is restored as the substitution is completed [21]. Similar trends are evident for volatility and solubility, although mediated by the increase in molecular weight [19]. For example, a decrease in sublimation temperature is observed as the degree of Cp' substitution is increased, consistent with a decrease in the magnitude of intermolecular interactions.
Table 1-2. Melting points (°C) for selected sandwich metallocene complexes, Cp'2E, for elements of groups 2 and 14 (taken from Ref. [21])
Mg Ca Sr Ba Ge Sn Pb
Cp
CpB
Cp2T
Cp3I
Cp4I
Cp*
176 >265 >360 >420 78 105 140
<25 301–304 375–380 320 – <25 –
80–83 189 150 222 72 solid –
92–100 40–45 <25 92–94 – 18 –
229–231 196–200 151–153 149–150 – 162–163 –
289–292 207–210 216–218 265–268 90–94 100–121 110–105
1.5 Structural Types The structural motifs exhibited in main group metallocene chemistry are varied in terms of the position of the element with respect to the Cp' ligand, the number of Cp' ligands associated with a element center and their relative orientation, the involvement of auxiliary ligands (other than Cp' derivatives) with the element, and the association of the complexes into dimers, oligomers or polymers. The general cyclic fluxionality of the Cp' ligand implies that all five of the carbon atoms in the Cp' frame are involved in the Cp'–element π-interaction, even for systems exhibiting low hapticity. This is apparently frozen out in the solid state so that the nonequivalent element–carbon distances obtained from X-ray crystallographic data often illustrate a closer interaction to some framework carbon atoms than others. This can be defined by the similarity or difference between the Cp'centroid–E distance and the Cp'plane–E distance (3).
1.5 Structural Types
3
5
11
4
6
7
1.5.1 Sandwich Complexes Sandwich structures represent the prototypical metallocenes and, as implied by the term, they involve two Cp' ligands, one on either side of the element. Within each periodic group of elements, there are examples of sandwich complexes with the two ligand planes parallel 4 (usually staggered D5d, but eclipsed D5h conformations are often close in energy), as well as examples of nonparallel 5 (often referred to as ‘bent’ sandwich structures) complexes. The presence of one (e.g. Cp2T2Ca·THF) [27] or more (e.g. CpB2Sr·THF2) [90] auxiliary ligands is usually observed with a nonparallel arrangement of Cp' ligands 6, but are not necessarily responsible for the nonparallel structure. Most sandwich complexes involve an equivalent interaction of the element with each Cp' ligand, although the unusual dimeric structure of LiCp5Bz involves a rare example of asymmetric coordination of two Cp′ ligands to one of the lithium centers [91]. For the purpose of structural comparison between all sandwich complexes (7) we define the angle α between the Cp' ligand planes (α = 0° in complexes of type 4), as well as the Cp'centroid–E–Cp'centroid angle β (β = 180° in complexes of type 4). The anionic sandwich complexes of the alkali metals, [LiCp2]– [74] and [NaCp2]– [75], represent examples of D5d arrangements 4, as do magnesocene [92] and the aluminocenium cation [Cp*2Al]+ [86]. Cp*2Si adopts two conformations
12
1 Main Group Metallocenes
(α = 0° and 25°) in the same crystal [87],but most parent and methylated complexes for elements of groups 2, 13, 14 and 15 are typically nonparallel 5, with α angles as large as 47° in Cp2Sn [93 ]. However, these angles decrease as the size and number of substituents around the Cp' frame is increased so that Cp5Ph2Sn has a parallel structure 4 [94 ]. The ions [Cp2Tl]– [95 ], Cp'2As]+ [62], and [Cp*2Sb]+ [63] adopt nonparallel sandwich structures 5, however, bulkier substituents (e.g. [Cp3B2Sb]+ [63] and [Cp3B2Bi]+) [64] again impose a parallel planar arrangement 4.
1.5.2 Half Sandwich Complexes A wider range of element environments are accessible for complexes involving a single π-coordinated Cp' ligand, summarized in the general structural representations 8–12, in which Y corresponds to a nonbonding electron pair, a covalent bond (substituent R), or a dative covalent bond (auxiliary ligand). Univalent group 13 elements engage cyclopentadienyl derivatives in monomeric half sandwich ‘Milking stool’ type structures 8 (Y = lone pair) in the gas phase (e.g. Cp*Al) [96], and the isovalent cations [Cp*E]+ (E = group 14) have been characterized in the solid state (e.g. [Cp*Sn][BF4]) [61]. Molecular half sandwich complexes are readily obtained for most elements with the presence of an auxiliary substituent(s) or ligand(s), depending on the valence and the size of the element. The most extensive series of structural type 8 are known for lithium (e.g. Cp3TLi·THF) [97] and beryllium (e.g. CpBeCl) [98], and cationic systems are accessible for boron (e.g. [Cp*BBr][AlBr4]) [99]. The presence of two auxiliary electron pairs 9 can be achieved for lithium with a bidentate ligand (e.g. CpTLi·TMEDA) [100 ]. However, for elements of group 14, reduction of the Cp' hapticity gives structures of type 10 (e.g. Cp*GeCl, Y = lone pair and Cl [101 ]; [Cp*Sn·pyridine][BF4], Y = lone pair and pyridine) [102 ]. ‘Piano stool’ structures 11 are evident in the unique examples of [CpMg·PMTEDA]+ containing a tridentate ligand [95] and the inverted sandwich cation [TAS–Cp–TAS]+ (TAS = trisdimethylaminosulfonium) [76], as well as in the aluminum dimer {Cp*AlCl2}2 [103 ]. However, the hapticity is reduced, as shown by 12, in alkylated derivatives, such as {Cp*Al(Cl)Me}2 [104].
8
9
10
11
12
1.5 Structural Types
13
1.5.3 µ–π–Cyclopentadienyl Molecular Complexes Association of element centers by π-bound µ-Cp' ligands is a general feature of polymeric structures (see below), but is also evident in a number of molecular anionic systems, representing neutral metallocenes bound together by an additional cyclopentadienide anion. The general formula [E2Cp3]– is observed as a nonparallel multidecker sandwich 13 for cesium [105] and thallium [106 ], and plumbocene sandwich units, Cp2Pb, are associated in the unusual anions [Cp5Pb2]– 14 and [Cp9Pb4]– 15 [107]. Each lead center accommodates three cyclopentadienyl ligands, as in the tris(cyclopentadienyl) anions 16, which are known for both tin [108 ] and lead [109 ] in the salts [Mg(THF)6][Cp3E]2.
13
15
14
16
14
1 Main Group Metallocenes
1.5.4 µ–π–Cyclopentadienyl Super-Sandwich Polymers The half sandwich formulation for alkali metal Cp' derivatives leads to polymeric arrays in the solid state involving alternating element and µ–π–Cp' ligands. The planes of the Cp' ligands may be parallel or almost parallel (referred to as supersandwich), as in the case of LiCp and NaCp 17, or nonparallel, as in the case of KCp 18 [110 ]. The branched polymeric structure observed in the solid state for the sandwich metallocene Cp2Pb 19 [111 ] reveals a tricoordinate environment for each lead center composed of two µ–π–Cp ligands (dashed lines) and one terminal Cp ligand (full line), although a slight distortion from pentahapticity is apparent. The solid state structure of BaCp*2 20 [112 ] is at first glance related, but the monomers are associated by means of relatively weak Ba–methyl interactions.
17
18
1.5 Structural Types
19
15
20
1.5.5 Dimers Unique examples of dimeric arrangements 21 are observed in the solid state for Cp5BzIn [113 ] and Cp5BzTl [114 ] (one of two crystalline forms), whilst more familiar dimeric structures involving halide bridges with terminal half sandwich Cp' complexes 22 (X = halogen; Y = halogen or donor) are adopted by compounds of magnesium [115 ] calcium [116 ], aluminum [117 ], and bismuth [118 ].
16
1 Main Group Metallocenes
21
22
1.5.6 Clusters With the influence of certain types of Cp' ligands, complexes involving monovalent elements of group 13 (Cp'E), which are monomeric in the gas phase, adopt highly symmetric cluster arrangements in the solid state. A tetrahedral structure is observed for Cp*Al 23 [41] while Cp*Ga [41a] and Cp*In [119 ] adopt octahedral hexameric structures 24 (methyl groups have been omitted for clarity). Cp2TTl forms a hexameric cluster structure, but involves only µ–π–Cp interactions between element centers 25 [120 ].
1.5 Structural Types
23
24
25
17
18
1 Main Group Metallocenes
1.6 Cp-Element Orbital Interactions Main group elements encompass a much wider range of electronegativities and hardness than the transition series so that the Cp'–E bond varies from interactions which are almost purely ionic for elements from group 1, to predominantly covalent for elements from groups 14 and 15. Theoretical studies of LiCp at various levels of sophistication promote an ionic classification and describe a structure with C5v symmetry [78, 121, 122, 123]. Nevertheless, some degree of covalent character is apparent in all main group metallocenes, and orbital interactions likely dominate for metallocenes of the p block elements. In this context, the decrease in Cp'centroid–E distances observed for germanium and tin metallocenes within the series Cp'2E, CpECl and CpE+(Ge, 220, 211, 199 pm; Sn, 240, 230, 229 pm) is somewhat consistent with the ‘octet rule’. Although the differences in these distances are small and may be attributed to a number of factors, including the variety of hapticities, the general trends can be rationalized with the Lewis drawings 26, 27 and 28, illustrating the adjustment in Cp'E bond order.
26
27
28
1.6.1 Orbital Interaction Model for Half Sandwich Complexes A general qualitative molecular orbital energy correlation diagram for half sandwich complexes with C5v symmetry is given in Fig. 1-2. The most effective stabilization occurs between the fully bonding (lowest energy a1) occupied π-orbital of the Cp ligand and the s and p(axis) orbitals of E. A secondary interaction involves the two occupied π-orbitals (e1) of Cp with the degenerate p orbitals of E (perpendicular to the E–Cpcentroid axis).
1.6 Cp-Element Orbital Interactions
19
Figure 1-2 General molecular orbital energy diagram for main group half sandwich complexes with C5v symmetry, showing only occupied molecular orbitals
In Fig. 1-2, Y is a coordinate bond if one invokes the model for the elements of group 1, and a covalent bond for elements of group 2 involved in neutral complexes. Cationic complexes of group 13 elements involve a covalent bond (e.g. [CpBBr]+), while neutral group 13 (e.g. CpAl) complexes and cationic complexes of group 14 (e.g. [CpSn]+) carry a nonbonding electron pair for Y. The validity of this model is demonstrated by calculated MO energies for beryllium complexes being in close agreement with measured vertical ionization energies obtained from PE spectra [3, 124] and calculated NMR chemical shifts for aluminum and gallium complexes in good agreement with experimental data [125]. Most of the compounds for which this model is invoked are characteristic of nido clusters with a pentagonal pyramidal framework [126] when the Cp–E interaction is considered in isolation, but the three occupied bonding MOs [123] (a1 and e1) can also be described in terms of a ‘three dimensional aromaticity’ model [122]. Inclusion of the σ-pair on E nicely fulfills the octet model as depicted in 28,
20
1 Main Group Metallocenes
and recognition of the lower hapticity generally observed for CpECl complexes is also consistent in 27 if the Cp ligand is considered as contributing four valence electrons rather than six.
1.6.2 Orbital Interaction Model for Sandwich Complexes Covalent bonding in sandwich complexes can be modeled qualitatively by the general molecular orbital schemes for parallel planar (D5d) and bent (C2v) structures containing 14 valence electrons, shown in Fig. 1-3. The stabilization of three electron pairs (D5d: 1a1g, 1e1u; C2v: 1a1, 2b1, 1b2) is primarily responsible for the Cp– element bond in both structures. The absence of an element orbital of a2 symmetry in the C2v structure and element orbitals of e1g symmetry in the D5d structure distinguishes the models from those for transition metallocenes, imposing ligand based nonbonding orbitals. While the ‘octet rule’ is inappropriate for systems considered as coordination complexes, it generally provides useful simple models to appreciate the bonding in compounds of p block elements and is applied consistently for half sandwich complexes in 27 and 28. A Lewis model for the sandwich structures is less obvious, but the six electrons which are unaccounted for in the resonance drawings 26 for the 14 valence electron system correspond to three nonbonding ligand based molecular orbitals, and the 2a1g orbital (D5d) and 2a1 orbital (C2v) correspond to the respective element based ‘lone pair’. Most discussions of structural preferences for 14 electron sandwich metallocenes focus attention on the energy and occupation of the 2a1g orbital in a D5d structure, which correlates with the 2a1 orbital in a C2v structure. Structural adjustment from D5d to C2v symmetry effects a stabilization of the 2a1g orbital due to loss of partial antibonding character on transformation into the more nonbonding 2a1 orbital (C2v). The observed structures of neutral Cp2E (E = group 14) and cationic [Cp2E]+ (E = pnictogen) sandwich metallocenes are consistent with this model, which can be understood in more simple terms using VSEPR arguments, implicating the 2a1 orbital (C2v) as the so-called ‘lone pair’. However, the energy of the 2a1g orbital in the D5d structure is also sensitive to the separation of the Cp ligands [127 ]. For a hypothetical planar (D5d) germanocene, the 2a1g orbital is calculated to be above the 1e1g level at small Cp–Cp separations (400 pm), it decreases in energy with increasing Cp–Cp separation, and is below the 1e1g level for Cp–Cp separations in excess of 500 pm, consistent with the antibonding character of the 2a1g orbital. Therefore, observed sandwich metallocene structures are likely the result of a number of factors rather than solely, the most obvious, electronic control. In fact, the electronic imposition or consequences of the nonparallel structure may be minimal. Experimental support for the deep energy of the formal lone pair (2a1g or 2a1) on the element center is provided in the photoelectron spectroscopic data for Cp*2E (E = Si, Ge, Sn, Pb), which are compared in Table 1-3 [128 ]. Although the ordering of the orbital energy levels for the heavier elements (Ge, Sn, Pb) is as
Figure 1-3 General molecular orbital diagrams for 14 electron main group sandwich complexes with parallel (D5d) and bent (C2v) structures, showing only occupied molecular orbitals
1.6 Cp-Element Orbital Interactions
21
22
1 Main Group Metallocenes
shown in Fig. 1-3, the 2a1 orbital is anomalously high in relative energy (7.50 eV) in the silicocene, consistent with the relatively close proximity of the Cp* ligands and the more effective overlap of the 3p valence orbitals of silicon in comparison with the heavier elements. Table 1-3. Orbital energies (eV) for derivatives of Cp'2E (E = Si, Ge, Sn, Pb) from photoelectron spectroscopic and theoretical data Orbital
Cp*2Si
Cp*2Ge
Cp*2Sn
Cp*2Pb
1a2 3a1 2b1 1b2 2a1
6.70 6.96 8.06 8.30 7.50
6.60 6.75 7.91 8.05 8.36
6.60 6.60 7.64 7.64 8.40
6.33 6.88 7.38 7.38 8.93
The wide variety of theoretical molecular orbital studies on 14 electron systems has been comprehensively reviewed [19, 20], concluding that although nonparallel (bent) structures represent global minima, this can only be recognized with higher order theory, as very small energy differences (<10 kJ · mol–1) distinguish parallel from nonparallel structures. Although metallocenes of elements such as the alkali metals are predominantly ionic, modification of Fig. 1-3 involving destabilization and vacancy of the 2a1g orbital provides a model for 12 electron systems consistent with the observed structures for anionic (i.e. [ECp2]–, E = alkali metal), neutral (i.e. [Cp'2Mg]) and cationic (i.e. [Cp*2Al]+) sandwich metallocenes. However, observation of nonparallel structures for the neutral 12 electron Cp2E metallocenes involving the heavier alkaline earth elements (E = Ca, Sr, Ba) are inconsistent with this model. As a result, these complexes have attracted intense theoretical molecular orbital (extensively reviewed in Refs 19 and 21) and molecular mechanics [129 , 130 ] studies. An overview assessment of the factors governing the structure of multicyclopentadienyl complexes is presented toward the end of this chapter. The absence of d orbitals impacts both the electronic structure and the reactivity of main group metallocenes. Electron density remains at the ligands in nonbonding molecular orbitals, which directs the majority of reactivity. This contrasts the chemistry of transition metallocenes where the frontier orbitals are more metal based.
1.7 Metallocenes of the Alkali Metals
23
1.7 Metallocenes of the Alkali Metals LiCp [121] as well as the sandwich anion [LiCp2]– [131] have both been termed the ‘simplest metallocenes’. Although the colorless microcrystalline solid formulated as LiCp was first prepared in 1901 [132], conclusive structural characterization was delayed due to the fact that it is essentially insoluble in nonpolar solvents. The structure of the sandwich anion was also only recently determined.
1.7.1 Monomers and Polymers The solubility of alkali metallocenes of the general formula ECp' is enhanced by the introduction of trimethylsilyl or tert-butyl substituents onto the cyclopentadienyl frame, and hexane solutions of these complexes can be prepared in situ. How-
29
30
31
32
24
1 Main Group Metallocenes
ever, when concentrated, a sudden and irreversible increase in viscosity occurs with precipitation of an amorphous powder which is subsequently insoluble in hexane [97], and crystalline solids are inaccessible [133 ]. Nevertheless, the mono-silylated derivative LiCpT was unexpectedly isolated as a crystalline by-product of the reaction between [(η5CpT)Y(OtBu)]2 and LiCH2SiMe3 [134 ]. It crystallizes as a ‘super-sandwich’ of stacked half sandwiches, with the CpT ligands staggered and the positioning of the SiMe3 groups describing a screw trace along the stack, as depicted in 29. A contrasting dimeric structure 30 is observed for the more heavily substituted salt LiCp5Bz, where a benzene solvate molecule essentially fragments the polymer [91]. This prevention of coordinative polymerization is complete with more effective donors, such as nitrogen or oxygen, which enhance the solvation of LiCp' derivatives, and numerous derivatives (e.g. 31 [97] and 32 [100]) have been comprehensively characterized.
33
34
35
The dilithium cationic complex [CpM(Li·TMEDA)2]+ (CpMcentroid–Li, 200 pm) 33 and the related mixed metal complexes Cp[(SiMe3)2N]Sn(µ-Cp)Li·PMDETA 34 (Cpcentroid–Li, 225 pm) and CpTl(µ-Cp)Li·TMEDA 35 (Cpcentroid–Li, 223 pm) [95] involve µ–π–Cp interactions, which are substantially weaker than those in the super-sandwich structure of LiCpT (CpTcentroid–Li 196 and 198 pm), but represent models of the ‘inverted sandwich’ moiety.
1.7 Metallocenes of the Alkali Metals
25
Figure 1-4 The solid state structure of LiCp viewed down the super-sandwich strands
Nearly 100 years after the initial discovery of a parent alkali metal cyclopentadienide salt [132] the structure was ultimately confirmed for LiCp, NaCp and KCp by high resolution powder X-ray diffraction. A super-sandwich structure of type 17 is observed for LiCp and NaCp, with eclipsed and almost parallel–planar Cp rings (LiCp, β = 176°; NaCp, β = 178°) [110] as shown in Fig. 1-4. In contrast, KCp adopts a ‘zig-zag’ pattern of type 18 with a dramatically smaller β angle (138°), and auxiliary interactions to the potassium center from two Cp ligands of neighboring polymer chains [110]. The presence of a silyl substituent in KCpT (cf. LiCpT) imposes a slightly larger β angle at potassium (151°) [135 ]. Nevertheless, the availability of additional coordination sites made possible by the nonparallel arrangement of the CpT ligands essentially modifies the relative stability of the polymeric array with respect to the influence of auxiliary ligands.
26
1 Main Group Metallocenes
36
37
In contrast to derivatives of LiCp, derivatives of NaCp and KCp accommodate donor solvents (auxiliary ligands) with retention of the super-sandwich. The structure of Cp*K·(pyridine)2 36 (β = 138°) [136 ] can be considered a modification of KCp (β = 138°) and KCpT (β = 151°), in which the interchain Cp'–K contacts are replaced by the two auxiliary ligand interactions, but with minimal or no impact on the structure of the polymer. The auxiliary coordination in CpNa·TMEDA 37 (β = 128 and 119°) [137 ] effects a dramatic distortion of the polymeric backbone, relative to the base free NaCp polymer (β = 178°). Nevertheless, NaCp' and KCp' polymers can be disrupted by the introduction of excess donors. For example, NaCp* dissolves in pyridine to give the monomeric piano stool structure 38 [136] which is related to 39 [138 ].
38
39
1.7 Metallocenes of the Alkali Metals
27
1.7.2 Lithocene and Sodocene Anions The exothermicity for the association of alkali metal ions with cyclopentadienide is calculated [75, 139 ] to be substantially greater for lithium than for sodium: [Li]+(g) + [Cp]–(g) → LiCp(g) ∆H = –703 kJ · mol–1 [75] (–761 kJ.mol–1) [139] (1-11) [Na]+(g) + [Cp]–(g) → NaCp(g) ∆H = –531 kJ · mol–1 [75] (–607 kJ.mol–1) [139](1-12) However, formation of the sandwich anions from the corresponding half sandwiches is more exothermic for sodium: LiCp(g) + [Cp]–(g) → [LiCp2]–(g) ∆H = –163 kJ · mol–1 [75] NaCp(g) + [Cp]–(g) → [NaCp2]–(g) ∆H = –184 kJ · mol–1 [75]
(1-13) (1-14)
This is due to stronger repulsive interactions between the [Cp]– anions, which are closer in lithocene, and outweighs the advantage of the stronger [Cp]––[Li]+ interaction. Consistent with the discussion of polymer–monomer relationships above, donor solvents such as THF dramatically influence coordination of the second Cp ligand, and an equilibrium is envisaged: (1-15) Conductivity measurements on liquid ammonia solutions [140 ], as well as IR [141 ] and NMR [78] spectroscopic data on THF solutions, first indicated the existence of lithocene anions. However, the use of complex cations enabled isolation and structural confirmation of [Li(12-crown-4)2][LiisoCp2] (containing anion 40) [142 ] and [Ph2PMe2][LiCpB2] [143 ], as well as tetraphenylphosphonium [74] and trisdimethylaminosulfonium [76] salts of the parent anion 41. The anions are parallel planar and almost perfectly staggered with consistent Cp'centroid–Li distances (200 pm). The introduction of excess NaCp is necessary to effect a shift of the equilibrium in favor of the sodocene 42 salt [PPh4][NaCp2] [74], which has a crystal structure that is remarkably similar to that of the analogous lithocene salt [74]. The equilibrium can also be modified by tethering the Cp ligands together to give the ansa-sodocene compound 43 [74], which has a bent sandwich structure, opening a coordination site for an auxiliary THF molecule. In cesium chemistry, the use of the complex cation [PPh4]+ also facilitates the formation of the multidecker anion [Cs2Cp3]– of the type 13 [105].
28
1 Main Group Metallocenes
40
41
42
43
1.8 Metallocenes of the Alkaline Earth Metals Comparisons are frequently made between transition metallocenes and those of the alkaline earth metals in light of their preferred oxidation state of +2 [144]. The variety of structural arrangements observed for alkaline earth metallocenes include the typical parallel sandwich for magnesium, the surprising nonparallel sandwich for heavier elements, and the rare ‘slip–sandwich’ structure of beryllocene.
44
45
1.8
Metallocenes of the Alkaline Earth Metals
29
1.8.1 Beryllocene The solid state structure of beryllocene has been determined a number of times [10, 145 , 146 ] due to the unique asymmetric arrangement of the parallel planar Cp ligands shown in 44, which is also observed in the gas phase [147 ]. A model 45 involving a σ-bonded ligand has been proposed on the basis of theoretical studies, invoking disruption of the parallel relationship of the ligands [148 ]. However, structure 44 accounts for the microwave spectroscopic data [149 ], the large dipole moment (µ25°C in benzene, 2.46 D; in cyclohexane, 2.24 D) [150 ], and the fluxionality of the complex in solution, which is clearly demonstrated by a sharp singlet observed in the 1H NMR spectrum at –135°C [151 ]. The estimated rate of 1010 s–1 at 300K corresponds to an activation energy of 5.2 kJ · mol–1 [152 ] and the process is proposed to occur via two isomerization mechanisms with barriers of 5 (‘gear wheel’ having a η5/η5 transition state) and 8 kJ · mol–1 (‘molecular inversion’ having a η3/η3 transition state), respectively. Consistent with the fluxional behavior in solution, the solid state structure of beryllocene is disordered with the beryllium atom in two sites of 50% occupancy so that each Cp ligand is engaged in both η5 and η1 bonding, as illustrated in 44. The beryllium atom is 150.5 pm from one Cp plane and 151.3 pm from the other [146], and the structural parameters, Raman data [153 ], and PE data [154 ] imply that it is bound to an sp2 (η1) hybridized carbon center with minor perturbation of the π-bonding in the ligand. A similar structure has been observed for the isovalent Cp*2Zn [155 ]. The preference for the η5/η1 structure of Cp2Be can be understood when one considers the Cpplane–Cpplane distance for the hypothetical D5d structure (302 pm), which is substantially shorter than the Cpplane–Cpplane distance in ferrocene (332 pm) [156 ]. The term ‘steric oversaturation’ has been used to account for systems in which no more than one ligand of a particular type can be accommodated by an element center [18]. The phenomenon can be applied to Cp2Be, as attempts to prepare the substituted derivative Cp*2Be have been unsuccessful giving instead the monosubstituted half sandwich compound Cp*BeCl [157 ].
1.8.2 Alkaline Earth Sandwich Metallocenes The numerous derivatives of sandwich complexes for the heavier alkaline earth metals have been extensively reviewed [19–21,40], as well as their potential utility as MOCVD sources [158 ]. Magnesocene adopts a typical staggered, parallel sandwich arrangement 46 in the solid state [92] but is eclipsed in the gas phase [159 ]. Although a similar structure is observed for Cp'2Mg in the gas phase [47, 48] a slight distortion is imposed in the solid state by the introduction of silyl substituents in Cp3T2Mg (α = 8°) [35]. Calcocene [160 ], strontocene, and barocene [161 ] were amongst the first metallocenes to be identified. Cp2Ca was found to adopt a
30
1 Main Group Metallocenes
46
47
surprising nonparallel polymeric structure in the solid state (α = 57°), in which each calcium atom engages four Cp rings with varying hapticity, and each ring adopts a µ–π–coordination with two calcium atoms 47 [162 ]. It can be considered in terms of optimal packing of two cyclopentadienide anions with Ca2+, nevertheless, the sandwich metallocene molecular ion is observed in the mass spectrum. Moreover, a nonparallel structure is generally observed for the heavier alkaline earth sandwich metallocenes, even in the gas phase (Cp*2Ca [48, 49], Cp*2Sr [163 ], Cp*2Ba [163]). Table 1-4. Comparison of interplane angles (α, °), Cp'centroid–E–Cp'centroid angles (β, °) and Cp'centroid–E distances (pm) for selected alkaline earth sandwich metallocenes
Cp*2Ca Cp*2Ba Cp4I2Ca Cp4I2Ba Cp2T2Ca·THF Cp2T2Sr·THF CpB2Ca·(THF)2 CpB2Sr·(THF)2 Cp3I2Ba·(THF)2
α
β
Cp'centroid–E
Ref.
35 48 11 17 46 48 48 48 56
147 131 162 154 135 134 133 133 132
235 274 235 268 240 255 246 260 279
[164 ] [112] [165 ] [165] [27] [27] [90] [90] [166 ]
In contrast to Cp2Ca, most sandwich derivatives are monomeric in the solid state (Cp*2Ba exhibits weak Ba–Me intermolecular interactions) [112], and the conformation of the Cp' ligands seems dependent upon the degree of substitution around the Cp' frame. Structural features for representative examples are compared in Table 1-4 and show smaller α angles and larger β angles for complexes possessing more sterically imposing substituents on the Cp' ring. Derivatives such as Cp4I2Ca are described as encapsulating the element center [165], which is responsible for the air stability of this complex in contrast to Cp*2Ca, and likely prevents forma-
1.8
Metallocenes of the Alkaline Earth Metals
31
tion of adducts. In addition, the solubility and volatility of these heavily substituted metallocenes is enhanced and DNMR studies indicate hindered rotation of the Cp' rings. Although the derivatives listed carry different substituents, the trends indicate that introduction of auxiliary ligands has only a minor effect on the α angle, and is essentially independent of the element, the number of ligands (one or two) or the Cp'centroid–E distance. A decrease in the β angle describes a withdrawal of the element center from the Cp'centroid–Cp'centroid vector relative to complexes without auxiliary ligands.
48
49
50
The ‘open-calcocene’ 48 is a derivative of these complexes [167 ] and there are rare examples of complexes involving bifunctional ligands as a link between two metallocene units 49 [168 ] as well as a chelate 50 [169 ]. Several derivatives of calcium, strontium and barium metallocenes carry donor functionalized side-chain cyclopentadienyl ligands [23, 170 ] including the first optically active calcocene 51 [171 ]. Organic ligands exhibit more flexible coordination, for example, the solid state structure of the bistrimethylsilylbutadiyne complex of Cp*2Ca contains two slightly different molecules, both with a butadiyne molecule wedged between the Cp* rings of the metallocene 52 [172 ]. In one, the diyne is symmetrically disposed about the Cp*centroid–E–Cp*centroid plane and in the other the ligand is slipped to the
32
1 Main Group Metallocenes
side by about 80 pm, possibly due to crystal packing forces. The Lewis acidity of the calcium center in Cp*2Ca can be further exploited to form a number of mixed metal complexes, such as [Cp*2Ca·MeAlMe2THF]2 [173] and Cp*2Ca·ClHfCp'2Cl [174].
51
52
1.8.3 Half Sandwich Metallocenes of the Group 2 Elements Half sandwich beryllium complexes of the general type 53 are monomeric in solution, in the gas phase and in the solid state, and all exhibit pentahapto coordination with similar Cp'centroid–Be distances (e.g. CpBeCl(g) 148 pm [98]; CpBeCl(s) 145 pm [175 ]; CpBeBr(g) 153 pm [176 ]; CpBeMe(g) 150 pm [177 ]; CpBeCCH(g) 149 pm [176]; Cp*BePBu2(s) 148 pm [178 ]) despite the varying degrees of π-interaction possible between the auxiliary ligand and beryllium. This comparison extends to include the borane complexes (e.g. CpBeB5H8 solid 147 pm) [179 ], and beryllocene (Cpplane–Be 151 pm) [146].
53
54
1.9 Metallocenes of the Group 13 Elements
33
Synthesis of half sandwich complexes for the heavier alkaline earth metals is complicated by disproportionation equilibria, but the use of donor solvents and bulky Cp' units impose a kinetic stability. Monomeric structures are observed for complexes with ‘monodentate’ (e.g. Cp4ICaN(SiMe3)2·THF) [180 ], bidentate (e.g. CpMgBr·TMEDA [181 ], Cp3TMgBr·TMEDA) [35], and tridentate (e.g. 54) [95] ligands, while dimeric arrangements are bound by halide bridges (e.g. {Cp*MgCl·OEt2}2 [115], {Cp*CaI·THF2}2 [116], {Cp4ICaI·THF}2) [182 ] or amino bridges (e.g. {Cp4MECa[SiMe2–CH2CH2SiMe2N]}2) [180], and have the general structure 22.
1.9 Metallocenes of the Group 13 Elements The elements of group 13 provide the most structurally diverse series of metallocene complexes, including examples of monomeric half sandwich molecules, various arrangements of dimeric half sandwich complexes, oligomeric clusters, polymeric chains analogous to those observed for the alkali metals, and rare examples of sandwich complexes. A recent review of the syntheses, structures and chemistry of aluminum(I) and gallium(I) compounds includes derivatives of the general formula Cp'E and their oligomers [125]. A comprehensive discussion of the organometallic chemistry of subvalent thallium is also available, which is dominated by examples of cyclopentadienyl derivatives [183 ].
1.9.1 Half Sandwich Monomers, Dimers, Oligomers and Polymers A monomeric half sandwich η5-structure 8 (Y = lone pair) is consistently observed in the gas phase for derivatives of the general formula Cp'E, including Cp*Al [96], Cp*Ga [184 ], CpIn [185 ], CpMIn [186 ], Cp'In [119], CpTl [187 ], and Cp*Tl [188 ]. Cp*Ga is stable enough to be distilled at ambient pressure (b.p. 150°C) and only decomposes under MOCVD conditions at 600°C.14 NMR spectroscopic data indicate that Cp'Ga derivatives also have monomeric structures in solution [43]. Half sandwich Cp'E complexes of indium and thallium with relatively minimal substitution around the Cp' ligand typically form ‘zig-zag’ polymeric structures of type 18 in the solid state, which are analogous to those observed for alkali metallocenes. The β angles are reasonably consistent (e.g. CpIn,128° [186]; Cp4MIn,134° [119]; CpTl, 137° [189 ]; Cp4MTl, 134°) [190 ], and the respective Cp'centroid–E distances (e.g. CpIn, 269 and 273 pm; Cp4MIn, 250 and 281 pm; CpTl, 319 pm; Cp4MTl, 268 and 271 pm) are considerably larger than in representative gas phase monomers (e.g. Cp*In, 229 pm [119]; Cp*Tl, 237 pm [188]). Monomeric solid state structures are observed for the phosphine substituted derivatives Cp4MPIn [191 ] and Cp4MPTl [192 ].
34
1 Main Group Metallocenes
Nucleophilic addition of the cyclopentadienide anion to CpTl gives either the unique anionic thallate sandwich metallocene of structural type 5 in [CpMg·PMDETA][Cp2Tl] [95], or the multidecker sandwich [Cp3Tl2]– [106], which has structure 13, analogous to [Cs2Cp3]– [105]. However, unlike the cesium derivative, the nonparallel conformations of these sandwich derivatives are consistent with the presence of a lone pair at the thallium center(s), the environment of the element being isovalent with neutral sandwich metallocenes for elements of group 14 and cationic sandwich metallocenes for elements of group 15. The heavily substituted Cp5BzTl adopts two crystalline forms depending upon the rate of crystallization. Rapid precipitation effects an alignment of the monomeric units in a super-sandwich polymeric array 17 [193 ], whilst a dimeric structure 21 is formed upon slow crystallization [114] in which the thallium centers are adjacent and the molecular structure is described as a trans bent geometry. The dimer presumably represents the thermodynamically favored arrangement, and is isostructural with the indium complex {Cp5BzIn}2, which has an identical element–element distance (363 pm) [113]. The possibility of In–In and Tl–Tl bonds has been theoretically assessed at various levels and there is much debate about the strength of such interactions [194 , 195 ]. Although there is indication of element–element interactions, they are weak and the {Cp5Bz2E}2 dimers likely involve both element–element and ligand–ligand attractive interactions [195].
1.9.2 Clusters Cp*Al sublimes at 140°C as the monomer, but in the solid state it forms a unique tetrahedron of aluminum atoms with the apices capped by η5-bound Cp* ligands, as shown in 23 [41]. The Al–Al bond is shorter (277 pm) than that in aluminum metal (9 pm longer) and is consistent with that calculated for the parent compound {CpAl}4 (279 pm) [196 ], but is longer than that calculated for {AlX}4 tetrahedra (X = H, SitBu3, SiH3, Cl, F; 262–264 pm) [196, 197 ]. While Cp*Tl is polymeric in the solid state [198 ], the indium analog Cp*In adopts an unusual hexameric structure with Cp* ligands surrounding an octahedral element cluster 24 [119, 199 ]. Cp*Ga [41a] is surprisingly isostructural with Cp*In including identical cell volumes. The Ga–Ga distances (407–417 pm, cf. atomic radius 130 pm) are longer than the In–In distances in {Cp*In}6 (394–396 pm, cf. atomic radius 155 pm) to accommodate the differences in Cp*centroid–E distances (Ga, 208 pm; In 230 pm). Such comparisons indicate that, in contrast to {Cp*Al}4, the ligand–ligand interactions define the cluster structure, with element–element interactions of low significance. Consistently, the gallium metallocene cluster is colorless, whilst the indium metallocene cluster is yellow implicating more effective element–element interactions made possible by the larger atomic size. Interestingly, the cluster structures 24 are distorted from Oh to S6
1.9 Metallocenes of the Group 13 Elements
35
symmetry by the slight off-setting of the Cp*centroid–normal from the center of the octahedron.
1.9.3 Metallocenes Involving Group 13 Elements in Oxidation State III The availability of vacant valence orbitals on high oxidation state group 13 centers might be expected to support π-coordination of Cp' ligands, however, examples are relatively rare. While CpAlMe2 is monomeric in the gas phase and is interpreted in terms of a η2-interaction [200], alkyl derivatives of CpAR2 [201] are typically polymeric in the solid state with bridging Cp ligands [202], and CpGaMe2 has a similar structure [203]. Haloaluminum complexes involving cyclopentadienyl ligands are dimeric bis-halogen bridged structures of the type 22, with varying hapticity for the Cp' ligand [103, 104, 117]. Dimerization can be prevented with the presence of a side chain donor functionality on the cyclopentadienyl ligand (CpN), which imposes a chelate interaction in place of the bridge, but most derivatives are best considered as σ-complexes [23, 24, 204]. The delicate energetic distinction between σ- and π-coordination is illustrated in the typically σ-coordinated derivatives of Cp'3Al, by the unique π(η5), σ, σ structure 55 observed in the solid state for Cp3M3Al [15].
55
56
While the neutral Cp' substituted boranes are also σ-bonded compounds [205 ], cationic complexes of the form [Cp'BR]+ have definitive half sandwich structures comparable to the isoelectronic Cp'BeR complexes 53. 11B, 1H, and 13C NMR spectral data are unambiguously interpreted in terms of an η5-interaction, and this is confirmed in a crystallographic study on [Cp*BBr][AlBr4] [99]. Cationic sandwich metallocenes of general formula [Cp'2E]+ have only been identified for boron and aluminum. Cp*2BCl (11B, –74 ppm) reacts with BCl3 to
36
1 Main Group Metallocenes
give the tetrachloroborate salt of [Cp*2B]+ as a microcrystalline precipitate (11B, +44 ppm) [57]. 1H NMR spectroscopic studies show the cation to be structurally analogous to the isoelectronic slip–sandwich Cp2Be complex 44. The decamethylaluminocenium cation 56 has been identified in the salts [Cp*2Al][Cp'AlCl3] [86] and [Cp*2Al][LiCp5Bz2] [91], and the crystal structures confirm the staggered parallel planar D5d structure. Although salts of the parent cation [Cp2Al]+ have not yet been isolated, it has been shown to be an effective olefin polymerization initiator [206 ].
1.9.4 Heteroatom Clusters Reaction of {Cp*Al}4 with elemental selenium or tellurium gives heterocubanes of the general formula {Cp*AlZ}4 (Z = Se or Te) [73], where the tetrahedral cluster of aluminum atoms is retained. The higher oxidation state for aluminum effects a significant shortening of the Cp*centroid–Al distances (Se: 195 pm; Te: 197 pm) relative to the starting material {Cp'Al}4 (202 pm) [41], but the gallium analogs [207 , 208 ] involve σ-coordination of the Cp* ligands. Reaction of {Cp*Al}4 with elemental phosphorus (P4) results in cleavage of all Al–Al and P–P bonds and formation of an unusual {(Cp*Al)6P4} cage, which can be described as two face-sharing heterocubanes with opposite corners missing [209 ]. Related clusters {(Cp*AlF)4(SiPh2)2} and {(Cp*Al)3Sb2} are also accessible [210 ], as well as {NiCp2(Cp*Al)2}, which shows the Cp*Al monomer as a bridging ligand [211 ].
1.10 Metallocenes of the Group 14 Elements 1.10.1 Group 14 Sandwich Complexes The group 14 sandwiches represent the most extensive series of isolated main group metallocenes. Silicocene derivatives are less accessible than germanocenes, stannocenes and plumbocenes, perhaps due to the difficulties associated with the stabilization of a divalent silicon center, and Cp*2Si is the only isolated derivative [212 ]. It is most conveniently prepared by the quantitative reduction of dibromobis(pentamethylcyclopentadienyl)silane using potassium anthracenide [87], and melts without decomposition. Two geometrical isomers are observed in the solid state, one with the Cp* ligand planes parallel 57a and the other 57b with α = 25° (β = 155°). The solid state CP-MAS 29Si NMR spectrum contains two signals, –403.2 ppm (nonparallel) and –423.4 ppm (parallel), and the solution shift (–398 ppm) indicates that the nonparallel isomer (α = 25°) predominates [80].
1.10 Metallocenes of the Group 14 Elements
57a
37
57b
Sandwich complexes of group 14 elements are generally monomeric in all phases and in solution, with the parent plumbocene as the only exception, composed of a zig-zag chain of lead atoms linked by π-coordinated µ-Cp ligands 19, and with each lead atom bearing a third terminal Cp ligand [111]. The larger atomic radius of the heavier elements allows for greater structural flexibility, so that interplane angles (α) for Cp*2Ge(g) [101], Cp*2Sn(s) [58], and Cp*2Pb(s) [93], listed for comparison with other derivatives in Table 1-5, are significantly larger than observed for Cp*2Si. Nevertheless, α is sensitive to steric encumbrances, and is smaller for complexes bearing more heavily substituted Cp' derivatives.
Table 1-5. Comparison of interplane angles (α, °) for examples of group 14 sandwich metallocenes (averages where necessary) Ge
α Cp2E Cp'2E(g) Cp*2E Cp5Bz2E Cp4M(SiB2M)2E a
50 34b 22c 31 d –
Polymeric CpM2E c Gas phase d SiB2M = ditbutylmethylsilyl b
Sn
Pb
Ref.
α
Ref.
α
Ref.
[213] [127] [101] [215 , 216] –
46 50b 36 33 0
[93] [127] [58] [216] [217]
60a 45c 37 33 0
[111] [214] [93] [216] [218]
38
1 Main Group Metallocenes
Figure 1-5 Decaphenylstannocene viewed perpendicular to the plane of the Cp5Ph ring
In some cases, the substitution pattern is so sterically demanding that a parallel planar structure is observed, as for Cp5I2Sn [219 ] and Cp5Ph2Sn (Fig. 1-5) [94]. However, the structural impact of the substituents is often difficult to predict. For example, Cp4I2Sn adopts a nonparallel structure (α = 28°; Cp4Icentroid–Sn 242 pm) [220 ] while the apparently less sterically loaded Cp3I2Pb is parallel [20], despite having a larger Cp'centroid–E distance (247 pm). Cp4M(SiB2M)2Sn [217] and Cp4M(SiB2M)2Pb [218] also exhibit parallel structures with less bulky substitution, which has led to the conclusion that the lone pair at the element center is stereochemically inactive. The asymmetric complex Cp5PhSnCp is noticeably nonparallel (α = 43°) [221 ] consistent with the release of steric strain in comparison with Cp5Ph2Sn. The structural flexibility of sandwiches involving heavier elements is further demonstrated by the introduction of auxiliary chelate ligands onto Cp2Pb, for which 58 is an example [222 ] (see also tricyclopentadienyl complexes below).
58
59
1.10 Metallocenes of the Group 14 Elements
60
39
61
1.10.2 Half Sandwich Complexes of Group 14 Elements [CpGe]+ and [CpSn]+ were first observed by mass spectrometry, but use of the Cp* ligand enabled the isolation of salts containing [Cp*Ge]+ [58, 223 ], [Cp*Sn]+ [58, 60, 61], and [Cp*Pb]+ cations [81]. [Cp*Pb][BF4] adopts a dimeric structure in the solid state bound by bridging anions [81], while the crystal structure of [Cp*Sn][BF4] confirms the presence of isolated ions 59 involving pentahapto interaction [61]. Cations of the type [Cp'Sn]+ are also observed as units of more complicated structures including the solid state structure of Cp2Sn·BF3 [224 ] and the novel cluster [Cp4M(SiB2M)6Sn9Cl12] [225 ]. The nucleophilicity of [Cp*E]+ cations is demonstrated by the formation of coordination complexes with neutral ligands such as pyridine to give representatives of ‘slip–sandwich’ structures (e.g. 60) [102]. Similar structural features are observed for the isoelectronic neutral half sandwich derivatives of general formula Cp'EX (E = Sn [52, 53, 226 ], Pb [227 ]; X = Cl, Br, I, CH3COO). The solid state structure of CpSnCl involves a short Sn–Cl bond as well as long Sn–Cl contacts, responsible for a coordination polymeric structure, illustrated in 61 [53, 226]. However, the absence of tin coupling in the 1H and 13C NMR spectra implies disproportionation in solution [228], and Cp*SnCl is described as monomeric in the solid state [226]. Molecular units are observed in the solid state structures of 62 [51] as well as in the chelated derivative 63 [230], but the same ‘slip–sandwich’ interaction is evident and is also apparent for Cp*GeCl in the gas phase [101]. Finally, one of the Cp ligands of Cp2Sn can be substituted by reaction with LiN=C(NMe2)2 to give the unusual cyclic dimeric structure 64 [231] involving pentahapto coordination.
40
1 Main Group Metallocenes
62
63
64
1.10.3 Tricyclopentadienyl Complexes Tin and lead exhibit the unique ability to accommodate three π-bound Cp ligands in molecular species. The fundamental complexes [Cp3E]– are obtained by nucleophilic addition of [Cp]– (NaCp or LiCp, or Cp2Mg) to Cp2Sn or Cp2Pb, and isolation of the salt [Mg(THF)6][Cp3Sn]2 reveals an nonassociated (cation contact free) anion [108] with a ‘Paddle-Wheel’ structure 16. The geometry for tin can be described as a shallow pyramid with symmetrical η3 attachment of the three Cp ligands. Although the same anion is present in the [Na·PMDETA]+ salt [232], one of the Cp ligands behaves as a bridge (µ–π–Cp) between the metals, which distorts the anion. The [Li·PMDETA]+ salt 34 of the aminostannyl anion [Cp2SnN(SiMe3)2]– [233] also shows the presence of both terminal and µ–π–Cp ligands. Reaction of [Cp]– with Cp2Pb also accounts for the two chain-like multinuclear anions [Cp5Pb2]– 14 and [Cp9Pb4]– 15, in the same [Li(12-crown-4)2]+ salt [107]. Although, at first glance, their structures are related to the solid state structure of Cp2Pb 19, they exhibit a disruption of the polymer planarity featured in the neutral parent metallocene. Nevertheless, similar variations in the Cpcentroid–Pb distances (terminal Cpcentroid–Pb, 251–262 pm, are significantly shorter than those involving the µ-Cp ligands, 273–292 pm) prompt models involving the Cp2Pb unit, so that 14 is viewed as two Cp2Pb molecules bound by a [µ-Cp]– and 15 is viewed as 14 bearing two additional Cp2Pb terminal units [107]. 1.10.4 Reaction Chemistry The extensive series of reactions already established for the metallocenes of group 14 can be summarized in five classifications. 1) Cp' ligand exchange is generally observed with the trend Ge<Sn
1.11 Metallocenes of the Pnictogens
41
2) Oxidative addition is a typical reaction for Cp*2Si leading to compounds of the type Cp*2Si(Y)X, or intermediates of the type Cp*2Si=Y [236, 237]. Metallocenes of germanium and tin are also oxidized to give the corresponding E(IV) product [229, 238]. 3) Nucleophilic substitution occurs with reagents such as alkyllithiums, which readily transform Cp*GeR compounds into R2Ge compounds [239] and Cp*2Sn into R2Sn [235], and can be envisaged to involve oxidative addition and reductive elimination steps. 4) Nucleophilic addition or the demonstration of Lewis acidity is a general theme for metallocenes generating tricyclopentadienyl complexes (Sn and Pb) and complexes with a variety of auxiliary ligands. 5) Homolytic bond cleavage occurs on thermal treatment of group 14 decamethyl metallocenes to give the element, tetramethylfulvene and Cp*H, under MOCVD conditions. The temperature of decomposition (Cp*2Si, ~550°C; Cp*2Ge, ~300°C; Cp*2Sn, ~200°C; Cp*2Pb, ~200°C) is lower for the heavier congeners. Cp*2Pb even decomposes in sunlight [240 ].
1.11 Metallocenes of the Pnictogens While X-ray crystallographic studies of Cp3Sb [241 ] and Cp3Bi [242 ] clearly show a σ-interaction, π-coordination is evident in compounds with fewer cyclopentadienyl ligands (Cp'2ECl or Cp'ECl2). The solid state structure of Cp3B2BiCl shows an almost parallel (α = 18°) sandwich arrangement 65, but with the bismuth center withdrawn from the Cp3Bcentroid–Cp3Bcentroid axis (β = 145°) [118].
65
66
42
1 Main Group Metallocenes
67
68
Cationic sandwich metallocenes exhibit varying degrees of distortion from a parallel planar arrangement. Consistent with observations for the more extensive series of neutral group 14 derivatives, heavily substituted Cp' ligands enforce a close to parallel planar structure as in the case of [Cp3B2Sb][AlCl4] (α = 12°) [63] and [Cp3B2Bi][AlCl4] (α = 10°) [64] (cation shown in 66, E = Sb or Bi), whilst the structures of [Cp*2As][BF4] (α = 36°) [62] (cation shown in 67) and [Cp*2Sb][AlI4] (α = 36°) [63] 68 are substantially nonparallel and the element is displaced from the centroid normal, with the latter displaying a significant interaction between antimony and iodine of the anion. Theoretical studies of the hypothetical parent phosphorus complex [Cp2P]+ present a surprising interplane angle (α = 111°) for the lowest energy structure and postulate η2/η1 coordination [243 ]. However, a nonbonding lone pair MO of stereochemical significance could not be identified. The monosubstituted cyclopentadienyl complexes CpSbCl2 [244 ] and CpBiCl2 [245 ] are classified as η3 and η3-η2, respectively. While CpSbCl2 adopts a chiral polymeric arrangement due to long range intermolecular Cp–element π-interactions, chlorine bridges between bismuth centers are responsible for the polymeric solid state structure of CpBiCl2. Thermal stability imposed by the Cp4I ligand enables the isolation and structural characterization of an homologous series, Cp4IECl2 with E = P, As, Sb and Bi, allowing for useful comparisons [246 ]. The solubility in common aprotic solvents decreases from P to Bi, and the color varies from yellow (P) to red–orange (Bi). Fluxional behavior observed in the 1H NMR spectra can be frozen out for P and As, but not for Sb or Bi. In the solid state, the position of the pnictogen center shows a smooth progression of the element from outside the pentagonal cylinder in the case of phosphorus, to a η3 bonded arrangement in the case of bismuth, as illustrated in 69. The halogen bridged dimer structure 70 and the unusual cluster 71 [118] are consistent with other areas of bismuth coordination chemistry.
1.11 Metallocenes of the Pnictogens
69
43
70
71
The only structurally confirmed example of a phosphorus metallocene was obtained by chloride ion abstraction from Cp*P(Cl)NHtBu with either AlCl3 or AgSO3CF3 [247 ]. The aluminate salt of the phosphanylium cation 72 (phosphenium) shows a definitive η2 interaction in the crystal structure, but NMR spectroscopic studies on complexes of this type confirm fluxionality in solution down to – 78°C [248 ].
72
73
44
1 Main Group Metallocenes
Cyclopentadienyl derivatives of sulfur are definitively σ-complexes, with Cp*2S as the classic representative [249 ]. Nevertheless, the solid state structure of the ionic [TAS–Cp–TAS][Cp] [76], containing cation 73, reveals a distinct interaction between the two sulfur centers and the centroid of the cyclopentadienide anion, which may be considered the first example of a π-interaction with sulfur (TAS = trisdimethylaminosulfonium).
1.12 Factors Governing the Structure of Multicyclopentadienyl Complexes The interplane angles (α) for the Cp' ligands in molecular sandwich complexes and polymeric super-sandwich structures range from 0°, for elements such as Li, Na, Mg and Al, to as much as 50° in Cp2Ge [213]. Experimental observations can be rationalized in covalent terms on the basis of the valence electron count, with 12 electron systems typically adopting parallel structures and 14 electron systems having nonparallel arrangements, implicating a stereochemical role for the two extra electrons, as discussed earlier. However, the inability of electronic models to rationalize the nonparallel structures observed for 12 electron sandwich metallocenes of the heavier group 2 elements (without steric loading) despite the absence of such a nonbonding pair [21], highlights the importance of other factors such as the degree of ionic character in the Cp'–E interaction, crystal packing forces and interligand nonbonded interactions. Molecular orbital calculations in general reveal small energy differences between parallel and nonparallel structures. For example, the linearization of Cp2Ba (β = 147°) is calculated at 1.5 kJ.mol–1 [250 ], and the nonparallel structures for the decamethylmetallocenes of Ca, Sr and Ba are 2 to 3.5 kJ · mol–1 more stable than the corresponding parallel structures [163]. While estimates of such energy differences for the group 14 sandwich metallocenes span a wide range, the values obtained from higher level calculations are generally small (>10 kJ · mol–1) and are comparable with those of the alkaline earth derivatives [19, 20], as are the values for the isoelectronic [Cp2Tl]– [95]. Therefore, noncovalent factors which are normally of minor structural consequence are expected to impose considerable conformational control. In this context, molecular mechanics force field calculations [129, 130] reveal that nonbonded attractive interactions between the Cp' ligands are significant in alkaline earth sandwich metallocenes. Crystal packing forces influence the structures to some extent, but are considered less significant [130]. It is interesting to note that a number of complexes of the general formula Cp*2E exhibit a consistent inter-ligand (methyl–methyl) closest contact distance of around 410 pm [48, 129], which is essentially the Van der Waal’s distance for two methyl groups (400 pm) [251 ], although the closest inter-ligand distance in the nonparallel
1.12
Factors Governing the Structure of Multicyclopentadienyl Complexes
45
conformer of Cp*2Si is as low as 380 pm. It is also instructive to consider interligand distance trends for all sandwich complexes. For this purpose, Table 1-6 lists typical Cp'centroid–E distances in sandwich metallocenes or super-sandwich polymeric arrays, representing half the Cp'centroid–Cp'centroid distance and therefore the inter-ligand distance for parallel parent complexes Cp2E. The introduction of substituents around the Cp' frame (such as methyl) will obviously result in the closest inter-ligand distance being less than the Cp'centroid–Cp'centroid distance, as would any distortion from parallel geometry. Table 1-6. Typical Cp'centroid–E distances (pm) in sandwich metallocenes or supersandwich polymeric arrays Group 1
Group 2
Li Na K Cs Fr
Be Mg Ca Sr Ba
199 236 281 310 –
Group 13 151 199 235 260 270
B Al Ga In Tl
– 184 – 255 260
Group 14
Group 15
C Si Ge Sn Pb
N P As Sb Bi
– 211 225 240 250
– – 215 230 250
The Cp'centroid–Cp'centroid distances are relatively large in comparison to transition metallocenes (e.g. ferrocene 332 pm) [156] as would be expected on the basis of a comparison of the orbital overlap model (due to the absence of d orbitals in the main group). In fact, most Cp'centroid–Cp'centroid distances for sandwich main group metallocenes exceed the Van der Waal’s contact for two aromatic systems (340 pm) [251]. An increase in Cp'centroid–E distance down each group, as well as a general decrease across each period, are consistent with the trends in atomic size and effective nuclear charge. The nonsterically loaded parallel sandwich metallocenes of lithium, magnesium and aluminum are predictably the shortest, and a slightly larger separation in Cp*2Si allows for both a parallel and nonparallel (α = 25°) form in the same crystal structure [87], illustrating the minimal energy difference between the two conformations. The very short Cp'centroid–Be distance prevents beryllocene from adopting a typical sandwich arrangement and a distortion to ‘slip–sandwich’ is observed with one of the ligands η1-bound [146]. For heavier elements, the Cp'centroid–Cp'centroid distances are substantially greater than the typical inter-ligand distance (410 pm) [48, 129], so that distortion from the parallel conformation is possible until this distance is achieved, or slightly less as in nonparallel Cp*2Si. Realization of the close structural similarity for identically substituted sandwich metallocenes of elements from groups 2 and 14 (the structure of Cp4I2Ca can be superimposed on Cp4I2Sn, and Cp*2Ca on Cp*2Sn) has also prompted the conclusion that the steric factors play a more dominant role in the control of molecular structure than electronic factors [21]. These arguments can be extended to rational-
46
1 Main Group Metallocenes
ize other metallocene structures. At first glance, the dimers {Cp5BzE}2 21 of indium [113] and thallium [114] are defined by element–element interactions, however, the similarity of the bond lengths and angles in these isostructural complexes implicate the importance of inter-ligand interactions, and a comparison of structural features for the {Cp*E}6 octahedral clusters of gallium [41a] and indium [119] draws similar conclusions.
1.13 Conclusions Metallocene chemistry is no longer the sole preserve of the d- and f-block elements, as an extensive series of main group metallocene complexes have now been identified and comprehensively examined. Most elements of groups 1, 2, 13, 14 and 15 engage Cp' ligands in a π-interaction, and there are rare examples which may be considered metallocenes of sulfur. Although there are obvious structural similarities with transition metallocenes, a wider structural diversity is made possible by a relatively weaker Cp'–E bond, the nature of which ranges from mainly ionic to highly covalent. As a result, although there are rare examples of catalytic behavior analogous to transition metallocenes, the chemistry of the main group metallocenes is highlighted by their role as synthetic reagents, as precursors for the MOCVD process, and a novel reactivity in general. This stems from the coordinatively flexible Cp'–E interaction and the ability of Cp' to behave as a leaving group, allowing for an extensive substitution chemistry. Further diversification involving isolobal analogs of Cp' has already begun and represents the future of main group metallocene chemistry.
1.14 Acknowledgments We thank the Alexander von Humboldt Foundation for a fellowship (NB), Ms. M. Welschof–Hildebrandt for preparing the drawings, Ms. B. Neumann for assistance with analysis of the structural data and a number of colleagues for informing us of recent results and providing preprints of their work.
References
47
References [1] [2] [3] [4]
[5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37]
Elschenbroich, C; Salzer, A. ‘Organometallics’ VCH Verlagsgesellschaft; Weinheim, 1989. Comprehensive account: Jutzi, P.; Burford, N. Chem. Rev. to be submitted. Jutzi, P. Adv. Organomet. Chem. 1986, 26, 217-295. For examples of recent developments, see, West, R.; Sohn, H.; Powell, D. R.; Müller, T.; Apeloig, Y. Angew. Chem. 1996, 108, 1095-1097; Angew. Chem. Int. Ed. Engl. 1996, 35, 1002-1004. Freeman, W. P.; Tilley, T. D.; Liable-Sands, L. M.; Rheingold, A. L. J. Am. Chem. Soc. 1996, 118, 10457-10468. Freeman, W. P.; Tilley, T. D.; Yap, G. P. A.; Rheingold, A. L. Angew. Chem. 1996, 108, 960-961. Hong, J. -H.; Pan, Y.; Boudjouk, P. Angew. Chem. 1996, 108, 213-215; Angew. Chem. Int. Ed. Engl. 1996, 35, 186-188. Schmidbaur, H. Angew. Chem. 1985, 97, 893-904; Angew. Chem. Int. Ed. Engl. 1985, 24, 893-904. Parkin, G. E. Adv. Inorg. Chem. 1995, 42, 291-393. Reger, D. L. Coord. Chem. Rev. 1996, 147, 571-595. Jutzi, P. J. Organomet. Chem. 1990, 400, 1-17. Jutzi, P. Pure & Appl. Chem. 1989, 61, 1731-1736. Daminani, D.; Ferretti, L.; Gallinella, E. Chem. Phys. Lett. 1976, 37, 265-269. Beattie, J.; Nugent, K. W. Inorg. Chim. Acta 1992, 198-200, 309-318. Cradock, S.; Duncan, W. J. Chem. Soc., Faraday Trans. II 1978, 74, 194-202. Anh, N. T.; Elian, M.; Hoffmann, R. J. Am. Chem. Soc. 1978, 100, 110-116. Jutzi, P. Chem. Rev. 1986, 86, 983-996. Jutzi, P.; Reumann, G. unpublished results. Fisher, J. D.; Budzelaar, P. H. M.; Shapiro, P. J.; Staples, R. J.; Yap, G. P. A.; Rheingold, A. L. Organometallics 1997, 16, 871-879. Jutzi, P. Comments Inorg. Chem. 1987, 6, 123-144. King, R. B.; Bisnette, M. B. J. Organomet. Chem. 1967, 8, 287-294. Gassmann, P. G.; Macomber, D. W.; Hershberger, J. W. Organometallics, 1983, 2, 1470-1472. Janiak, C.; Schumann, H. Adv. Organomet. Chem. 1991, 33, 291-393. Hanusa, T. P. Chem. Rev. 1993, 93, 1023-1036. Hays, M. L.; Hanusa, T. P. Adv. Organomet. Chem. 1996, 40, 117-170. Burkey, D. J.; Hanusa, T. P. Comments Inorg. Chem. 1995, 17, 41-77. Jutzi, P. ‘The Chemistry of Organo Silicon Compounds Part 3’, Rappoport, Z. ed. Wiley, 1997, in press. Jutzi, P.; Dahlhaus, J. Phosphorus, Sulfur and Silicon. 1994, 87, 73-82. Jutzi, P.; Siemeling, U. J. Organomet. Chem. 1995, 500, 175-185. Jutzi, P.; Dahlhaus, J. Coord. Chem. Rev. 1994, 137, 179-199. King, R. B.; Stone, F. G. A. Inorg. Synth. 1963, 7, 99-115. Barber, W. A. J. Inorg. Nucl. Chem. 1957, 4, 373-374. Engelhardt, L. M.; Junk, P. C.; Raston, C. L.; White, A. H. J. Chem. Soc., Chem. Commun. 1988, 1500-1501. Kuz’yants, G. M. Izv. Akad. Nauk SSSR, Ser. Khim. 1976, 25, 1785-1786. Tacke, M. Organometallics 1994, 13, 4124-4125. Wagner, B.O.; Ebel, H. F. Tetrahedron, 1970, 26, 5155-5167. Abel, E. W.; Dunster, M. O. J. Organomet. Chem. 1971, 33, 161-167. Kohl, F. X.; Kanne, D.; Jutzi, P. in ‘Organometallic Syntheses’ Vol. 3, p.381-383, King, R. B.; Eisch, J. J. eds. Elsevier, Amsterdam, 1986. Parris, G. E.; Ashby, E. C. J. Organomet. Chem. 1974, 72, 1-10. Strohmeier, W.; Landsfeld, H.; Gernert, F.; Langhäuser, W. Z. Anorg. Allg. Chem. 1960, 307, 120-122. Morley, C. P.; Jutzi, P.; Krüger, C.; Wallis, J. M. Organometallics 1987, 6, 1084-1090. Kroll, W. R.; Hudson, Jr., B. E. J. Organomet. Chem. 1971, 28, 205-210. Kroll, W. R.; Naegele, W. J. Chem. Soc., Chem. Commun. 1969, 246-247. Jutzi, P.; Leffers, W.; Müller, G.; Huber, B. Chem. Ber. 1989, 122, 879-884.
48 [38]
[39]
[40] [41]
[42] [43] [44]
[45]
[46]
[47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61]
1 Main Group Metallocenes Kurosova, H. Thallium in ‘Comprehensive Organometallic Chemistry’ Wilkinson, G.; Stone, F. G. A.; Abel, E. W. eds. Vol. 1, p. 749, Pergamon, 1982. Anderson, G. K.; Cross, R. J.; Phillips, I. G. J. Chem. Soc., Chem. Commun. 1978, 709. Nielson, A. J.; Rickard, C. E. F.; Smith, J. M. Inorg. Synth. 1986, 24, 97. Nielson, A. J.; Rickard, C. E. F.; Smith, J. M. Inorg. Synth. 1990, 28, 315-316. Fritz, H. P.; Köhler, F. H. J. Organomet. Chem. 1971, 30, 177-185. Koridze, A. A.; Ogorodnikova, N. A.; Petrovsky, P. V. J. Organomet. Chem. 1978, 157, 145-151. Rausch, M. D.; Edwards, B. H.; Rogers, R. D.; Atwood, J. L. J. Am. Chem. Soc. 1983, 105, 3882-3886. Conway, B. G.; Rausch, M. D. Organometallics 1985, 4, 688-693. Spink, W. C.; Rausch, M. D. J. Organomet. Chem. 1986, 308, C1-C4. Schumann, H.; Janiak, C.; Khani, H. J. Organomet. Chem. 1987, 330, 347-355. Arthurs, M.; Al-Daffaee, H. K.; Haslop, J.; Kubal, G.; Pearson, M. D.; Thatcher, P.; Curzon, E. J. Chem. Soc., Dalton Trans. 1987, 2615-2619. Singh, P.; Rausch, M. D.; Bitterwolf, T. E. J. Organomet. Chem. 1988, 352, 273-282. Morcos, D.; Tikkanen, W. J. Organomet. Chem. 1989, 371, 15-18. Rausch, M. D.; Spink, W. C. Atwood, J. L.; Baskar, A. J.; Bott, S. G. Organometallics 1989, 8, 2627-2631. Hughes, R. P.; Lomprey, J. R. Inorg. Chim. Acta 1995, 240, 653-656. Hanusa, T. P. Polyhedron 1990, 9, 1345-1352. Dohmeier, C.; Robl, C.; Tacke, M.; Schnöckel, H. Angew. Chem. 1991, 103, 594-595; Angew. Chem. Int. Ed. Engl. 1991, 30, 564-565; Loos, D.; Baum, E.; Ecker, A.; Schnöckel, H.; Downs, A.J. Angew. Chem. 1997, 109, 894; Angew. Chem. Int. Ed. Engl. 1997, 36, 860. Loos, D.; Schnöckel, H.; Gauss, J.; Schneider, U. Angew. Chem. 1992, 104, 1376-1378; Angew. Chem. Int. Ed. Engl. 1992, 31, 1362-1364. Loos, D.; Schnöckel, H. J. Organomet. Chem. 1993, 463, 37-40. Peppe C.; Tuck, D. G.; Victoriano, L. J. Chem. Soc., Dalton Trans. 1981, 2592-2582. Jutzi, P.; Leffers, W.; Müller, G. J. Organomet. Chem. 1987, 334, C24-C25. Schumann, H.; Lentz, A. Z. Naturforsch. 1994, 49B, 1717-1724. Jutzi, P.; Leffers, W. J. Chem. Soc., Chem. Commun. 1985, 1735-1736. Rausch, M. D.; Lewison, J. F.; Hart, W. P. J. Organomet. Chem. 1988, 358, 161-168. Jones, S. S.; Rausch, M. D.; Bitterwolf, T. E. J. Organomet. Chem. 1990, 396, 279-287. Schumann, H.; Kucht, H.; Kucht, A. Z. Naturforsch. 1992, 47B, 1281-1289. Schumann, H.; Kucht, A.; Kucht, H.; Lentz, A., Esser, L. Z. Naturforsch. 1993, 48B, 297-312. Cotton, F. A.; Wilkinson, G.; Birmingham, J. M. J. Inorg. Nucl. Chem. 1956, 2, 95-113. Wilkinson, G.; Cotton, F. A. Chem. Ind. 1954, 11, 307. Fischer, E. O.; Hafner, W. Z. Naturforsch. 1954, B9, 503-504. Robbins, J. L.; Edelstein, N.; Spencer, B.; Smart, J. C. J. Am. Chem. Soc. 1982, 104, 1882-1893. Andersen, R. A.; Blom, R.; Boncella, J. M.; Burns, J. C.; Volden, H. V. Acta Chem. Scand. 1987, A41, 24-35. Andersen, R. A.; Boncella, J. M.; Burns, J. C.; Blom, R.; Haaland, A.; Volden, H. V. J. Organomet. Chem. 1986, 312, C49-C52. House, H. O.; Latham, R. A.; Whitesides, G. M. J. Org. Chem. 1967, 32, 2481-2496. Jutzi, P.; Hampel, B.; Hursthouse, M. B.; Howes, A. J. Organometallics 1986, 5, 1944-1948. Bos, K. D.; Bulten, E. J.; Noltes, J. G. J. Organomet. Chem. 1972, 39, C52-C54. Bos, K. D.; Bulten, E. J.; Noltes, J. G.; Spek, A. L. J. J. Organomet. Chem. 1975, 99, 71-77. Poland, J. S.; Tuck, D. G. J. Organomet. Chem. 1972, 42, 307-314. Gaffney, C.; Harrison, P. G. J. Chem. Soc., Dalton Trans. 1982, 1055-1060. Jutzi, P.; Seufert, A.; Buchner, W. Chem. Ber. 1979, 112, 2488-2493. Jutzi, P.; Seufert, A. J. Organomet. Chem. 1978, 161, C5-C7. Jutzi, P.; Kohl, F.; Hofmann, P.; Krüger, C.; Tsay, Y. -H. Chem. Ber. 1980, 113, 757-769. Jutzi, P.; Dickbreder, R. J. Organomet. Chem. 1989, 373, 301-306. Kohl, F. X.; Jutzi, P. Chem. Ber. 1981, 114, 488-494. Jutzi, P.; Kohl, F.; Krüger, C. Angew. Chem. 1979, 91, 81-82; Angew. Chem. Int. Ed. Engl. 1979, 18, 59-60.
References [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95]
[96]
49
Jutzi, P.; Wippermann, T. Krüger, C.; Kraus, H. -J. Angew. Chem. 1983, 95, 244-245; Angew. Chem. Int. Ed. Engl. 1983, 22, 250. Sitzmann, H.; Ehleiter, Y.; Wolmershäuser, G.; Ecker, A.; Üffing, C.; Schnöckel, H. J. Organomet. Chem. 1997, 527, 209-213. Sitzmann, H.; Wolmershäuser, G. Z. Naturforsch. 1997, 52b, 398. Cowley, A. H.; Jutzi, P.; Kohl, F. X.; Lasch, J. G.; Norman, N. C.; Schlüter, E. Angew. Chem. 1984, 96, 603-604; Angew. Chem. Int. Ed. Engl. 1984, 23, 616-617. Bulten, E. J.; Budding, H. A. J. Organomet. Chem. 1978, 157, C3-C4. Cowley, A. H.; Lasch, J. G.; Norman, N. C.; Stewart, C. A.; Wright, T. C. Organometallics 1983, 2, 1691-1692. Cowley, A. H.; Kemp. R. A.; Stewart, C. A. J. Am. Chem. Soc. 1982, 104, 3239-3240. Freeman, M. B.; Sneddon, L. G.; Huffman, J. C. J. Am. Chem. Soc. 1977, 99, 5194-5196. Freeman, M. B.; Sneddon, L. G. Inorg. Chem. 1980, 19, 1125-1132. Jutzi, P.; Hielscher, B. J. Organomet. Chem. 1985, 291, C25-C27. Jutzi, P.; Hielscher, B. Organometallics 1986, 5, 1201-1204 Schulz, S.; Roesky, H. W.; Koch, H. J.; Sheldrick, G. M.; Stalke, D.; Kuhn, A. Angew. Chem. 1993, 105, 1828-1830; Angew. Chem. Int. Ed. Engl. 1993, 32, 1729-1731. Harder, S.; Prosenc, M. H.; Angew. Chem. 1994, 106, 1830-1832; Angew. Chem. Int. Ed. Engl. 1994, 33, 1744-1746. Harder, S.; Prosenc, M. H.; Rief, U. Organometallics 1996, 15, 118-122. Wessel, J.; Lork, E.; Mews, R. Angew. Chem. 1995, 107, 2565-2567; Angew. Chem. Int. Ed. Engl. 1995, 34, 2376-2377. Exner, M. M.; Waack, R.; Steiner, E. C. J. Am. Chem. Soc. 1973, 95, 7009-7018. Paquette, L. A.; Bauer, W.; Sivik, M. R.; Buhl, M.; Feigel, M.; Schleyer, P. v. R. J. Am. Chem. Soc. 1990, 112, 8776-8788. Schleyer, P. v. R.; Jiao, H. Pure & Appl. Chem. 1996, 68, 209-218. Wrackmeyer, B.; Sebald, A.; Merwin, L. H. Magn. Reson. Chem. 1991, 29, 260-263. Jutzi, P.; Dickbreder, R.; Nöth, H. Chem. Ber. 1989, 122, 865-870. Jutzi, P.; Schlüter, E.; Pohl, S.; Saak, W. Chem. Ber. 1985, 118, 1959-1967. Scherr, P. A.; Hogan, R. J.; Oliver, J. P. J. Am. Chem. Soc. 1974, 96, 6055-6059. Gauss, J.; Schneider, U.; Ahlrichs, R.; Dohmeier, C.; Schnöckel, H. J. Am. Chem. Soc. 1993, 115, 2402-2408. Benn, R.; Rufinska, A. Angew. Chem. 1986, 98, 851-871; Angew. Chem. Int. Ed. Engl. 1986, 25, 861-881. Dohmeier, C.; Schnöckel, H.; Robl, C.; Schneider, U.; Ahlrichs, R. Angew. Chem. 1993, 105, 1714-1716; Angew. Chem. Int. Ed. Engl. 1993, 32, 1655-1657. Jutzi, P.; Holtmann, U.; Kanne, D.; Krüger, C.; Blom, R.; Gleiter, R.; Hyla-Kryspin, I. Chem. Ber. 1989, 122, 1629-1639. Harris, R. K.; Mann, B. E. ’NMR and the Periodic Table’, Academic Press, London, 1978. Cerny, Z.; Machacek, J.; Fusek, J.; Kriz, O.; Casensky, B.; Tuck, D. G. J. Organomet. Chem. 1993, 456, 25-30. Gardiner, M. G.; Raston, C. L.; Kennard, C. H. L. Organometallics 1991, 10, 3680-3686. Dohmeier, C.; Baum, E.; Ecker, A.; Köppe, R.; Schnöckel, H. Organometallics 1996, 15, 4702-4706. Bünder, W.; Weiss, E. J. Organomet. Chem. 1975, 92, 1-6. Atwood, J. L.; Hunter, W. E.; Cowley, A. H.; Jones, R. A.; Stewart, C. A. J. Chem. Soc., Chem. Commun. 1981, 925-927. Heeg, M. J.; Janiak, C.; Zuckerman, J. J. J. Am. Chem. Soc. 1984, 106, 4259-4261. Armstrong, D. R.; Herbst-Irmer, R.; Kuhn, A.; Moncrieff, D.; Paver, M. A.; Russell, C. A.; Stalke, D.; Steiner, A.; Wright, D. S.; Angew. Chem. 1993, 105, 1807-1809; Angew. Chem. Int. Ed. Engl. 1993, 32, 1774-1776. Haaland, A.; Martinsen, K. -G.; Shlykov, S. A.; Volden, H. V.; Dohmeier, C.; Schnöckel, H. Organometallics, 1995, 14, 3116-3119.
50 [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129]
1 Main Group Metallocenes Jutzi, P.; Leffers, W.; Pohl, S.; Saak, W. Chem. Ber. 1989, 122, 1449-1456. Drew, D. A.; Haaland, A. Acta Chem. Scand. 1972, 26, 3351-3356. Dohmeier, C.; Köppe, R.; Robl, C.; Schnöckel, H. J. Organomet. Chem. 1995, 487, 127-130. Lappert, M. F.; Singh, A.; Engelhardt, L. M.; White, A. H. J. Organomet. Chem. 1984, 262, 271-278. Fernholt, L.; Haaland, A.; Jutzi, P.; Kohl, F. X.; Seip, R. Acta. Chem. Scand. 1984, A38, 211-216. Kohl, F. X.; Schlüter, E.; Jutzi, P.; Krüger, C.; Wolmershäuser, G.; Hofmann, P.; Stauffert, P. Chem. Ber. 1984, 117, 1178-1193. Koch, H. -J.; Schulz, S.; Roesky, H. W.; Noltemeyer, M.; Schmidt, H. -G.; Heine, A.; HerbstIrmer, R.; Stalke, D.; Sheldrick, G. M. Chem. Ber, 1992, 125, 1107-1109. Schonberg, P. R.; Paine, R. T.; Campana, C. F. J. Am. Chem. Soc. 1979, 101, 7726-7728. Harder, S.; Prosenc, M. H. Angew. Chem. 1996, 108, 101-103; Angew. Chem. Int. Ed. Engl. 1996, 35, 97-99. Armstrong, D. R.; Edwards, A. J.; Moncrieff, D.; Paver, M. A.; Raithby, P. R.; Rennie, M. A.; Russell, C. A.; Wright, D. S. J. Chem. Soc., Chem. Commun. 1995, 927-928. Duer, M. J.; Page, N. A.; Paver, M. A.; Raithby, P. R.; Rennie, M. -A.; Russell, C. A.; Stourton, C.; Steiner, A.; Wright, D. S. J. Chem. Soc., Chem. Commun. 1995, 1141-1142. Edwards, A. J.; Paver, M. A.; Raithby, P. R.; Russell, C. A.; Stalke, D.; Steiner, A.; Wright. D. S. J. Chem. Soc., Dalton Trans. 1993, 1465-1466. Paver, M. A.; Russell, C. A.; Wright, D. S. Angew. Chem. 1995, 107, 1679-1688; Angew. Chem. Int. Ed. Engl. 1995, 34, 1545-1554. Dinnebier, R. E.; Behrens, U.; Olbrich, F. Organometallics 1997, 16, 3855-3858. Panattoni, C.; Bombieri, G.; Croatto, U. Acta Crystallogr. 1966, 21, 823-826. Williams, R. A.; Hanusa, T. P.; Huffman, J. C. J. Chem. Soc., Chem. Commun. 1988, 10451046. Schumann, H.; Janiak, C.; Görlitz, F.; Loebel, J.; Dietrich, A. J. Organomet. Chem. 1989, 363, 243-251. Schumann, H.; Janiak, C.; Pickardt, J.; Börner, U. Angew. Chem. 1987, 99, 788-789; Angew. Chem. Int. Ed. Engl. 1987, 26, 789-790. Dohmeier, C.; Loos, D.; Robl, C.; Schnöckel, H.; Fenske, D. J. Organomet. Chem. 1993, 448, 5-8. McCormick, M. J.; Sockwell, S. C.; Davies, C. E. H.; Hanusa, T. P.; Huffman, J. C. Organometallics 1989, 8, 2044-2049. Schonberg, P. R.; Paine, R. T.; Campana, C. F.; Duesler, E. N. Organometallics 1982, 1, 799-807. Sitzmann, H.; Wolmershäuser, G. Chem. Ber. 1994, 127, 1335-1342. Beachley, Jr., O. T.; Blom, R.; Churchill, M. R.; Faegri, Jr., K.; Fettinger, J. C.; Pazik, J. C.; Victoriano, L. Organometallics 1989, 8, 346-356. Harvey, S.; Raston, C. L.; Skelton, B. W.; White, A. H.; Lappert, M. F.; Srivastava, G. J. Organomet. Chem. 1987, 328, C1-C6. Alexandratos, S.; Streitwieser, A.; Schaefer, H. F. J. Am. Chem. Soc. 1976, 78, 7959-7962. Watermann, K. C.; Streitwieser, A. J. Am. Chem. Soc. 1984, 106, 3138-3140. Jemmis, E. D.; Schleyer, P. v. R. J. Am. Chem. Soc. 1982, 104, 4781-4788. Lattman, M.; Cowley, A. H. Inorg. Chem. 1984, 23, 241-247. Böhm, M. C.; Gleiter, R.; Morgan, G. L.; Lusztyk, J.; Starowieyski, K. B. J. Organomet. Chem. 1980, 194, 257-268. Dohmeier, C.; Loos, D.; Schnöckel, H. Angew. Chem. 1996, 108, 141-161; Angew. Chem. Int. Ed. Engl. 1996, 35, 129-149. Wade, K. Adv. Inorg. Radiochem. 1976, 18, 1-66. Almlof, J.; Fernholt, L.; Faegri, Jr., K.; Haaland, A.; Schilling, B. E. R.; Seip, R.; Taugbol, K. Acta Chem. Scand. 1983, A37, 131-140. Jutzi, P. in ‘Frontiers of Organogermanium, Tin and Lead Chemistry’, Lukevics, E.; Ignatovich, I. eds. Latvian Institute of Organic Synthesis 1993. Hollis, T. K.; Burdett, J. K.; Bosnich, B. Organometallics 1993, 12, 3385-3386. Bosnich, B. Chem. Soc. Rev. 1994, 23, 387-395.
References
51
[130] Timofeeva, T. V.; Lii, J. -H.; Allinger, N. L. J. Am. Chem. Soc. 1995, 117, 7452-7459. [131] Stalke, D. Angew. Chem. 1994, 106, 2256-2259; Angew. Chem. Int. Ed. Engl. 1994, 33, 21682171. [132] Thiele, J. Ber. Dtsch. Chem. Fes. 1901, 34, 68. [133] Leffers, W. Thesis, University of Bielefeld, 1988. [134] Evans, W. J.; Boyle, T. J.; Ziller, J. W. Organometallics 1992, 11, 3903-3907. [135] Jutzi, P.; Leffers, W.; Hampel, B.; Pohl, S.; Saak, W. Angew. Chem. 1987, 99, 563-564; Angew. Chem. Int. Ed. Engl. 1987, 26, 583-584. [136] Rabe, G.; Roesky, H. W.; Stalke, D.; Pauer, F.; Sheldrick, G. M. J. Organomet. Chem. 1991, 403, 11-19. [137] Aoyagi, T.; Shearer, H. M. M.; Wade, K.; Whitehead, G. J. Organomet. Chem. 1979, 175, 21-31. [138] Lorberth, J.; Shin, S. -H.; Wocadlo, S.; Massa, W. Angew. Chem. 1989, 101, 793-794; Angew. Chem. Int. Ed. Engl. 1989, 28, 735-736. [139] Lambert, C.; Schleyer, P. v. R. Angew. Chem. 1994, 106, 1187-1199; Angew. Chem. Int. Ed. Engl. 1994, 33, 1129-1140. [140] Strohmeier, W.; Landsfeld, H.; Gernert, F. Z. Electrochem. 1962, 66, 823-827. [141] Ford, W. T. J. Organomet. Chem. 1971, 32, 27-33. [142] Zaegel, F.; Gallucci, J. C.; Meunier, P.; Gautheron, B.; Sivik, M. R.; Paquette, L. A. J. Am. Chem. Soc. 1994, 116, 6466-6467. [143] Wong, W. -K.; Zhang, L.; Wong, W. -T.; Xue, F.; Mak, T. C. W. Polyhedron, 1996, 15, 45934597. [144] Faegri, K.; Almlöf, J.; Lüthi, H. P. J. Organomet. Chem. 1983, 249, 303-313. Evans, S.; Green, M. L. H.; Jewitt, B.; Orchard, A. F.; Pygall, C. F. J. Chem. Soc., Faraday Trans. 2 1972, 68, 1847-1864. Cauletti, C.; Green, J. C.; Kelly, M. R.; van Tilborg, J.; Robbins, J.; Smart, J. J. Electro. Spectr. Rel. Phenom. 1980, 19, 327. [145] Wong, C. H.; Lee, T. Y.; Chao, K. J.; Lee, S. Acta Crystallogr. 1972, 28b, 1662-1665. Wong, C. H.; Lee, T. Y.; Lee, T. J.; Chang, T. W.; Liu, C. S. Inorg. Nucl. Chem. Lett. 1973, 9, 667-673. [146] Nugent, K. W.; Beattie, J. K.; Hambley, T. W.; Snow, M. R. Aust. J. Chem. 1984, 37, 1601-1606. [147] Nugent, K. W.; Beattie, J. K. Inorg. Chem. 1988, 27, 4269-4273. Almenningen, A.; Haaland, A.; Lusztyk, J. J. Organomet. Chem. 1979, 170, 271-284. Haaland, A. Acta Chem. Scand. 1968, 22, 3030-3032. [148] Marynick, D. S. J. Am. Chem. Soc. 1977, 99, 1436-1441. Chiu, N. -S.; Schäfer, L. J. Am. Chem. Soc. 1978, 100, 2604-2607. Jemmis, E. D.; Alexandratos, S.; Schleyer, P. v. R.; Streitwieser, A.; Schaefer, H.F. J. Am. Chem. Soc. 1978, 100, 5695-5700. Gleiter, R.; Böhm, M. C.; Haaland, A.; Johansen, R.; Lusztyk, J. J. Organomet. Chem. 1979, 170, 285-292. [149] Pratten, S. J.; Cooper, M. K.; Aroney, M. J.; Filipczuk, S. W. J. Chem. Soc., Dalton Trans. 1985, 1761-1765. [150] Fischer, E. O.; Schreiner, S. Chem. Ber. 1959, 92, 938-948. [151] Wong, C.; Wang, S. Inorg. Nucl. Chem. Lett. 1975, 11, 677-678. [152] Nugent, K. W.; Beattie, J. K.; Field, L. D. J. Phys. Chem. 1989, 93, 5371-5377. Nugent, K. W.; Beattie, J. K. J. Chem. Soc., Chem. Commun. 1986, 186-187. [153] Lusztyk, J.; Starowieyski, K. B. J. Organomet. Chem. 1979, 170, 293-297. [154] Gleiter, R.; Böhm, M. C.; Haaland, A.; Johansen, R.; Lusztyk, J. J. Organomet. Chem. 1979, 170, 271-284. [155] Blom, R.; Haaland, A.; Weidlein, J. J. Chem. Soc., Chem. Commun. 1985, 266-267. [156] Cp2Fe: Seiler, P.; Dunitz, J. D. Acta Crystallogr. 1979, B35, 2020-2032. Cp'2Fe: Freyberg, D. P.; Robbins, J. L.; Raymond, K. N.; Smart, J. C. J. Am. Chem. Soc. 1979, 101, 892-897. [157] Burns, C. J.; Andersen, R. A. J. Organomet. Chem. 1987, 325, 31-37. [158] Wojtczak, W. A.; Fleig, P. F.; Hampden-Smith, M. J. Adv. Organomet. Chem. 1996, 40, 215-340. [159] Haaland, A.; Lusztyk, J.; Brunvoll, J.; Starowieyscki, K. Z. J. Organomet. Chem. 1975, 85, 279-285. [160] Ziegler, K.; Froitzheim-Kuhlhorn, H.; Hafner, H. Chem. Ber. 1956, 89, 434-443. [161] Fischer, E. O.; Stölzle, G. Chem. Ber. 1961, 94, 2187-2193.
52
1 Main Group Metallocenes
[162] Zerger, R.; Stucky, G. J. Organomet. Chem. 1974, 80, 7-17. [163] Andersen, R. A.; Blom, R.; Burns, C. J.; Volden, H. V. J. Chem. Soc., Chem. Commun. 1987, 768-769. Blom, R.; Faegri Jr., K.; Volden, H. V. Organometallics, 1990, 9, 372-379. [164] Williams, R. A.; Hanusa, T. P.; Huffman, J. C. Organometallics, 1990, 9, 1128-1134. [165] Williams, R. A.; Tesh, K. F.; Hanusa, T. P. J. Am. Chem. Soc. 1991, 113, 4843-4851. [166] Burkey, D. J.; Williams, R. A.; Hanusa, T. P. Organometallics, 1993, 12, 1331-1337. [167] Overby, J. S.; Hanusa, T. P. Angew. Chem. 1994, 106, 2300-2302; Angew. Chem. Int. Ed. Engl. 1994, 33, 2191-2193. [168] Williams, R. A.; Hanusa, T. P.; Huffman, J. C. J. Organomet. Chem. 1992, 429, 143-152. [169] Rieckhoff, M.; Pieper, U.; Stalke, D.; Edelmann, F. T. Angew. Chem. 1993, 105, 1102-1104; Angew. Chem. Int. Ed. Engl. 1993, 32, 1079-1081. [170] Jutzi, P.; Dahlhaus, J.; Kristen, M. O. J. Organomet. Chem. 1993, 450, C1-C3. Rees, W. S. Jr.; Lay, U. W.; Dippel, K. A. J. Organomet. Chem. 1994, 483, 27-31. Hays, M. L.; Hanusa, T. P.; Nile, T. A. J. Organomet. Chem. 1996, 514, 73-79. [171] Molander, J. A.; Schumann, H.; Rosenthal, E. C. E.; Demtsdunk, J. Organometallics 1996, 15, 3817-3824. [172] Williams, R. A.; Hanusa, T. P.; Huffman, J. C. J. Am. Chem. Soc. 1990, 112, 2454-2455. [173] Tanner, P. S.; Williams, R. A.; Hanusa, T. P. Inorg. Chem. 1993, 32, 2234-2235. [174] Sockwell, S. C.; Tanner, P. S.; Hanusa, T. P. Organometallics 1992, 11, 2634-2638. [175] Goddard, R.; Akhtar, J.; Starowieyski, K. B. J. Organomet. Chem. 1985, 282, 149-154. [176] Haaland, A.; Novak, D. P. Acta Chem. Scand. 1974, A28, 153-156. [177] Drew, D. A.; Haaland, A. Acta Chem. Scand. 1972, 26, 3079-3084. [178] Atwood, J. L.; Bott, S. G.; Jones, R. A.; Koschmieder, S. U. J. Chem. Soc., Chem. Commun. 1990, 692-693. [179] Gaines, D. F.; Coleson, K. M.; Calabrese, J. C. J. Am. Chem. Soc. 1979, 101, 3979-3980. [180] Sockwell, S. C.; Hanusa, T. P.; Huffman, J. C. J. Am. Chem. Soc. 1992, 114, 3393-3399. [181] Johnson, C.; Toney, J.; Stucky, G. D. J. Organomet. Chem. 1972, 40, C11-C13. [182] Burkey, D. J.; Alexander, E. K.; Hanusa, T. P. Organometallics 1994, 13, 2773-2786. [183] Janiak, C. Coord. Chem. Rev. 1997, 163, 107-216. [184] Haaland, A.; Martinsen, K. -G.; Volden, H. V.; Loos, D.; Schnöckel, H. Acta Chem. Scand. 1994, 48, 172-174. [185] Shibata, S.; Bartell, L. S.; Gavin, Jr., R. M. J. Chem. Phys. 1964, 41, 717-722. [186] Beachley, Jr., O. T.; Pazik, J. C.; Glassman, T. E.; Churchill, M. R.; Fettinger, J. C.; Blom, R. Organometallics 1988, 7, 1051-1059. [187] Tyler, J. K.; Cox, A. P.; Sheridan, J. Nature [London], 1959, 183, 1182-1183. [188] Blom, R.; Werner, H.; Wolf, J. J. Organomet. Chem. 1988, 354, 293-299. [189] Berar, J. F.; Calvarin, G.; Pommier, C.; Weigel, D. J. Appl. Crystallogr. 1975, 8, 386-387. [190] Schumann, H.; Kucht, H.; Kucht, A.; Görlitz, F. H.; Dietrich, A. Z. Naturforsch. 1992, 47b, 1241-1248. [191] Schumann, H.; Ghodsi, T.; Esser, L. Acta Crystallogr. 1992, C48, 618-620. [192] Schumann, H.; Ghodsi, T.; Esser, L. Hahn, E. Chem. Ber. 1993, 126, 591-594. [193] Schumann, H.; Janiak, C.; Khan, M. A.; Zuckerman, J. J. J. Organomet. Chem. 1988, 354, 713. [194] Budzelaar, P. H. M.; Boersma, J. Recl. Trav. Chim. Pays-Bas 1990, 109, 187-189. Schwerdtfeger, P. Inorg. Chem. 1991, 30, 1660-1663. Treboux, G.; Barthelat, J. -C. J. Am. Chem. Soc. 1993, 115, 4870-4878. [195] Janiak, C.; Hoffmann, R. J. Am. Chem. Soc. 1990, 112, 5924-5946. [196] Ahlrichs, R.; Ehrig, M.; Horn, H. Chem. Phys. Lett, 1991, 183, 227-233. [197] Schneider, U.; Ahlrichs, R.; Horn, H.; Schäfer, A. Angew. Chem. 1992, 104, 327-329; Angew. Chem. Int. Ed. Engl. 1992, 31, 353-355. [198] Werner, H.; Otto, H.; Kraus, H. J. J. Organomet. Chem. 1986, 315, C57-C60. [199] Beachley, Jr., O. T.; Churchill, M. R.; Fettinger, J. C.; Pazik, J. C.; Victoriano, L. J. Am. Chem. Soc. 1986, 108, 4666-4668.
References
53
[200] Drew, D. A.; Haaland, A. Acta Chem. Scand. 1973, 27, 3735-3745. [201] Haaland, A.; Weidlein, J. J. Organomet. Chem. 1972, 40, 29-33. Stadelhofer, J.; Weidlein, J.; Haaland, A. J. Organomet. Chem. 1975, 84, C1-C4. Stadelhofer, J.; Weidlein, J.; Fischer, P. J. Organomet. Chem. 1976, 116, 55-63. [202] Tecle, B.; Corfield, W. R.; Oliver, J. P. Inorg. Chem. 1982, 21, 458-461. [203] Mertz, K.; Zettler, F.; Hausen, H. D.; Weidlein, J. J. Organomet. Chem. 1976, 122, 159-170. [204] Jutzi, P.; Dahlhaus, J.; Kristen, M. O. J. Organomet. Chem. 1993, 450, C1-C3. Jutzi, P.; Dahlhaus, J.; Bangel, M. J. Organomet. Chem. 1993, 460, C13-C15. Jutzi, P.; Dahlhaus, J.; Neumann, B.; Stammler, H. -G. Organometallics 1996, 15, 747-752. [205] Einstein, F. W. B.; Gilbert, M. M.; Tuck, D. G. Inorg. Chem. 1972, 11, 2832-2836. Beachley, Jr., O. T.; Getman, T. D.; Kirss, R. U.; Hallock, R. B.; Hunter, W. E.; Atwood, J. L. Organometallics 1985, 4, 751-754. Beachley, Jr., O. T.; Hallock, R. B.; Zhang, H. M.; Atwood, J. L. Organometallics 1985, 4, 1675-1680. Schumann, H.; Nickel, S.; Weimann, R. J. Organomet. Chem. 1994, 468, 43-47. [206] Bochmann, M.; Dawson, D. M. Angew. Chem. 1996, 108, 2371-2372; Angew. Chem. Int. Ed. Engl. 1996, 35, 2226-2228. [207] Schulz, S.; Gillan, E. G.; Ross, J. L.; Rogers, L. M.; Rogers, R. D.; Barron, A. R. Organometallics 1996, 15, 4880-4883. [208] Schulz, S.; Andruh, M.; Pape, T.; Heinze, T.; Roesky, H. W.; Häming, L.; Kuhn, A.; HerbstIrmer, R. Organometallics 1994, 13, 4004-4007. [209] Dohmeier, C.; Schnöckel, H.; Robl, C.; Schneider, U.; Ahlrichs, R. Angew. Chem. 1994, 106, 225-227; Angew. Chem. Int. Ed. Engl. 1994, 33, 199-200. [210] Schulz, S.; Schoop, T.; Roesky, H. W.; Häming, L.; Steiner, A.; Herbst-Irmer, R. Angew. Chem. 1995, 107, 1015-1016; Angew. Chem. Int. Ed. Engl. 1995, 34, 919-920. [211] Dohmeier, C.; Krautscheid, H.; Schnöckel, H. Angew. Chem. 1994, 106, 2570-2571; Angew. Chem. Int. Ed. Engl. 1994, 33, 2482-2483. [212] Jutzi, P.; Kanne, D.; Krüger, C. Angew. Chem. 1986, 98, 163; Angew. Chem. Int. Ed. Engl. 1986, 25, 164 [213] Grenz, M.; Hahn, E.; du Mont, W. -W.; Pickardt, J. Angew. Chem. 1984, 96, 69-70; Angew. Chem. Int. Ed. Engl. 1984, 23, 61-62. [214] Almenningen, A.; Haaland,, A.; Motzfeldt, T. J. Organomet. Chem. 1967, 7, 97-104. [215] Schumann, H.; Janiak, C.; Hahn, E.; Loebel, J.; Zuckerman, J. J. Angew. Chem. 1985, 97, 765; Angew. Chem. Int. Ed. Engl. 1985, 24, 773. [216] Schumann, H.; Janiak, C.; Hahn, E.; Kolax, C.; Loebel, J.; Rausch, M. D.; Zuckerman, J. J.; Heeg, M. J. Chem. Ber. 1986, 119, 2656-2667. [217] Constantine, S. P.; Hitchcock, P. B.; Lawless, G. A.; De Lima, G. M. J. Chem. Soc., Chem. Commun. 1996, 1101-1102. [218] Constantine, S. P.; Hitchcock, P. B.; Lawless, G. A. Organometallics 1996, 15, 3905-3906. [219] Sitzmann, H.; Boese, R.; Stellberg, P. Z. Anorg. Allg. Chem. 1996, 622, 751-755. [220] Burkey, D. J.; Hanusa, T. P. Organometallics 1995, 14, 11-13. [221] Heeg, M. J.; Herber, R. H.; Janiak, C.; Zuckerman, J. J.; Schumann, H.; Manders, W. F. J. Organomet. Chem. 1988, 346, 321-332. [222] Beswick, M. A.; Cromhout, N. L.; Harmer, C. N.; Raithby, P. R.; Russell, C. A.; Smith, J. S. B.; Steiner, A.; Wright, D. S. J. Chem. Soc., Chem. Commun. 1996, 1977-1978. [223] Jutzi, P.; Kohl, F. X.; Schlüter, E.; Hursthouse, M. B.; Walker, N. P. C. J. Organomet. Chem. 1984, 271, 393-402. [224] Dory, T. S.; Zuckerman, J. J.; Barnes, C. L. J. Organomet. Chem. 1985, 281, C1-C7. [225] Constantine, S. P.; De Lima, G. M.; Hitchcock, P. B.; Keates, J. M.; Lawless, G. A. J. Chem. Soc., Chem. Commun. 1996, 2337-2338. [226] Constantine, S. P.; De Lima, G. M.; Hitchcock, P. B.; Keates, J. M.; Lawless, G. A.; Marziano, I. Organometallics 1997, 16, 793-795. [227] Holliday, A. K.; Makin, P. H.; Puddephatt, R. J. J. Chem. Soc., Dalton Trans. 1976, 435438.
54
1 Main Group Metallocenes
[228] Williams, R. E. Adv. Inorg. Radiochem. 1976, 18, 67-142. Onak, T. in ‘Comprehensive Organometallic Chemistry’, Wilkinson, G.; Stone, F. G. A.; Abel, E. W. eds. Pergamon, Oxford, 1982. [229] Harrison, P. G.; Zuckerman, J. J. J. Am. Chem. Soc. 1969, 91, 6885-6886. [230] Jutzi, P.; Schmidt, H.; Neumann, B.; Stammler, H. -G. J. Organomet. Chem. 1995, 499, 7-10. [231] Edwards, A. J.; Paver, M. A.; Raithby, P. R.; Rennie, M. -A.; Russell, C. A.; Wright. D. S. J. Chem. Soc., Dalton Trans. 1995, 1587-1591. Stalke, D.; Paver, M. A.; Wright, D. S. Angew. Chem. 1993, 105, 445-446; Angew. Chem. Int. Ed. Engl. 1993, 32, 428-429. [232] Davidson, M. G.; Stalke, D.; Wright, D. S. Angew. Chem. 1992, 104, 1265-1267; Angew. Chem. Int. Ed. Engl. 1992, 31, 1226-1227. [233] Paver, M. A.; Russell, C. A.; Stalke, D.; Wright. D. S. J. Chem. Soc., Chem. Commun. 1993, 1349-1351. [234] Harris, D. H.; Lappert, M. F. J. Chem. Soc., Chem. Commun. 1974, 895-896. [235] Jutzi, P.; Hielscher, B. Organometallics 1986, 5, 2511-2514. [236] Jutzi, P.; Möhrke, A.; Müller, A.; Bögge, H. Angew. Chem. 1989, 101, 1527-1528; Angew. Chem. Int. Ed. Engl. 1989, 28, 1518-1519. [237] Jutzi, P. ‘Frontiers of Organosilicon Chemistry’, Bassindale, A. R.; Gaspar, P. eds. Royal Society of Chemistry, 1991. [238] Jutzi, P.; Kohl, F. X. J. Organomet. Chem. 1979, 164, 141-152. Jutzi, P.; Hampel, B. Organometallics 1986, 5, 730-734. [239] Jutzi, P.; Becker, A.; Stammler, H. -G.; Neumann, B. Organometallics 1991, 10, 1647-1648. [240] Jutzi, P.; Dahlhaus, J. unpublished results. [241] Birkhahn, M.; Krommes, P.; Massa, W.; Lorberth, J. J. Organomet. Chem. 1981, 208, 161-167. [242] Lorberth, J.; Massa, W.; Wocadlo, S.; Sarraje, I.; Shin, S.-H.; Li, X.-W. J. Organomet. Chem. 1995, 485, 149-152. [243] Lee, T. J.; Schaefer, III, H. F.; Magnusson, E. A. J. Am. Chem. Soc. 1985, 107, 7239-7243. [244] Frank, W. J. Organomet. Chem. 1991, 406, 331-341. [245] Frank, W. J. Organomet. Chem. 1990, 386, 177-186. [246] Ehleiter, Y.; Wolmershäuser, G.; Sitzmann, H.; Boese, R. Z. Anorg. Allg. Chem. 1996, 622, 923-930. [247] Gudat, D.; Nieger, M.; Niecke, E. J. Chem. Soc., Dalton Trans. 1989, 693-700. [248] Baxter, S. G.; Cowley, A. H.; Mehrotra, S. K. J. Am. Chem. Soc. 1981, 103, 5572-5573. Cowley, A. H.; Mehrotra, S. K. J. Am. Chem. Soc. 1983, 105, 2074-2075. [249] Bard, A. J.; Cowley, A. H.; Leland, J. K.; Thomas, G. J. N.; Norman, N. C.; Jutzi, P.; Morley, C. P.; Schlüter, E. J. Chem. Soc., Dalton Trans. 1985, 1303-1307. [250] Kaupp, M.; Schleyer, P. v. R.; Dolg, M.; Stoll, H. J. Am. Chem. Soc. 1992, 114, 8202-8208. [251] Pauling, L. ‘The Nature of the Chemical Bond’, 3rd ed., Cornell University Press, New York 1960, p 261.
2
Lanthanocenes Frank T. Edelmann
2.1 Introduction The beginnings of organolanthanide chemistry date back to the year 1954, when Birmingham and Wilkinson described the tris(cyclopentadienyl)lanthanide complexes, Ln(C5H5)3 [1]. Thus the chemistry of lanthanocenes is not much younger than that of other metallocenes of the d-transition metals. However, the development of this area of organometallic chemistry was relatively slow due to the intrinsic instability of organolanthanide compounds towards moisture and oxygen. The highly oxophilic character of the 4f-elements requires rigorous exclusion of traces of air and moisture during preparation and characterization of organometallic compounds of these metals. In addition, it turned out that simple alkyl and aryl derivatives of the type LnR3 could not be isolated due to severe steric unsaturation. With today’s more sophisticated experimental and analytical techniques combined with the design of new ligand sets organolanthanides and -actinides are routinely synthesized and handled and their unique properties are easily studied. Historically, the early discovery of Ln(C5H5)3 was followed by a period of relative stagnation which lasted about two decades. The past twenty years, however, have witnessed an enormous and exciting development of organolanthanide chemistry [2–7]. This spectacular growth in research activities is mainly because many organo-f-element complexes exhibit unique structural and physical properties and many of them are highly active in various catalytic processes. In fact, the catalytic activity of certain organolanthanide hydrocarbyls and hydrides is often dramatically higher than that of comparable d-transition metal catalysts. Applications of lanthanocenes in homogeneous catalysis have been compiled in a recent review article [8]. Thus the present review will focus mainly on synthetic aspects of lanthanocene chemistry.
2.2 Lanthanide(II) Metallocenes Lanthanocene(II) complexes are among the oldest known rare earth organometallics. Like alkali metals, several rare earth elements can be dissolved in liquid ammonia. Eu(C5H5)2 was first obtained by reacting the blue solution of Eu metal in liquid ammonia with cyclopentadiene [9]. The base-free material was obtained by heating the initially produced ammonia adduct to 200°C under high vacuum fol-
56
2 Lanthanocenes
lowed by vacuum sublimation of the thermally highly stable product at 400°C. The behavior of Eu(C5H5)2 nicely illustrates the typical properties of most organolanthanide complexes: The compound is thermally extremely stable while it is rapidly decomposed in the presence of traces of oxygen. Yb(C5H5)2 too has been prepared by adopting the same ammonia route [2–7].
Ln + 3C5H6
1. NH3(l) → Ln(C5H5)2 + C5H8 2. subl. Ln = Eu, Yb
(2-1)
Subsequently several other synthetic routes leading to lanthanocene(II) complexes have been developed. They are normally carried out in THF solution and initially yield the THF adducts of Ln(C5H5)2. An elegant transmetallation reaction starts with the metal powders of Sm, Eu, or Yb which are treated with Hg(C5H5)2. Similarly, Yb metal reacts with TlC5H5 in DME solution to give (C5H5)2Yb(DME) (Eqs. (2-2) and (2-3)) [10]. THF Ln + Hg(C5H5)2 →
Yb + 2TlC5H5
(C5H5)2Ln(THF)n + Hg Ln = Sm, Eu, Yb
DME →(C5H5)2Yb(DME) + 2Tl
(2-2)
(2-3)
In the solid-state structures of (C5H5)2Yb(DME) [11], (C5H5)2Yb(THF)2 [12, 13], and (MeC5H4)2Yb(DME) [14] the central Yb atom is tetrahedrally surrounded by two cyclopentadienyl ligands and two oxygen atoms of the ether ligands. Perhaps the most straightforward access to Sm(C5H5)2 was reported by Kagan et al. and involves the reaction of SmI2 with two equivalents of sodium cyclopentadienide (Eq. (2-4)) [15–17]. THF SmI2 + 2Na(C5H5) → Sm(C5H5)2 + 2KI
(2-4)
Yet another possible synthesis of (C5H5)2Ln(II) derivatives is the reduction of various lanthanocene(III) complexes [2–7, 13, 18, 19]. Either sodium, sodium naphthalide, potassium cyclooctadienide, or metallic Yb have been used to reduce (C5H5)2YbCl(THF) to the divalent oxidation state: THF (C5H5)2YbCl(THF) + Na → (C5H5)2Yb(THF)n + NaCl
(2-5)
2.2 Lanthanide(II) Metallocenes
THF 3(C5H5)2YbCl(THF) + Yb → 3(C5H5)2Yb(THF)n + YbCl3
57 (2-6)
Tris(cyclopentadienyl)lanthanides can also be used as precursors for lanthanide(II) metallocenes. Purple (C5H5)2Sm(THF)n was obtained by reductive cleavage of Sm(C5H5)3 with potassium naphthalide [2–7]. Deacon et al. have studied the redox chemistry of Yb(C5H5)3 and found that this compound can be reduced by metallic ytterbium to give Yb(C5H5)2. In turn, Yb(C5H5)2 can be oxidized back to Yb(C5H5)3 by treatment with TlC5H5 [15]. Reduction of Yb(C5H5)3 with sodium naphthalide was originally reported to give Yb(C5H5)2 (after sublimation at 400°C). A recent reinvestigation of this reaction revealed that the sublimed green material was polymeric [NaYb(C5H5)3]n [20] instead. Thus the reaction pathway has to be reformulated according to Eq. (2-7). Yb(C5H5)3 + Na → NaYb(C5H5)3
(2-7)
In this unusual ‘ate’ complex of divalent ytterbium the different metal atoms are bridged by µ-η5:η5-cyclopentadienyl ligands to give a polymeric network. Somewhat surprising is the fact that the compound can be sublimed at very high temperatures (ca. 400 °C) without any sign of decomposition. Retention of sodium cyclopentadienide is an explanation for the various color changes associated with Yb(C5H5)2 in solution and in the solid state. For solid Yb(C5H5)2 a green and a red modification had been described in the early literature [2–7]. It has now become clear that the green color is associated with [NaYb(C5H5)3]n while the red material is genuine Yb(C5H5)2. Yb(C5H5)2 itself is also polymeric in the solid state [20]. Ring-substituted ligands such as t-butylcyclopentadienyl or 1, 3-di-t-butylcyclopentadienyl have been successfully employed in the preparation of highly reactive lanthanide(II) organometallics. The resulting products are often more soluble in hydrocarbon solvents than the parent cyclopentadienyl derivatives. Simple bis-substituted lanthanide(II) complexes containing these ligands have been prepared by metathetical reactions between the lanthanide diiodides and the corresponding sodium or potassium cyclopentadienides. Monomeric (tBuC5H4)2Yb(THF)2 was isolated as the disolvate and characterized by X-ray crystallography [21]. Monomeric (tBuC5H4)2Sm(DME) has been prepared analogously [22, 23]. In the case of divalent samarium the use of the sterically more demanding 1,3-di-t-butylcyclopentadienyl ligand leads to the formation of the mono-THF adduct (tBu2C5H3)2Sm(THF) [25]. Trimethylsilyl-substituted cyclopentadienyl ligands are equally useful in the preparation of soluble, highly reactive organolanthanide(II) complexes. For example, dimeric [(Me3SiC5H4)2Yb(µ-Cl)]2 can be reduced by sodium amalgam in THF to give the bis(THF) solvate of Yb(Me3SiC5H4)2. The THF ligands can be replaced by TMEDA, and simple heating of the THF adduct yields the green, base-free Yb(II) derivative (Scheme 2-1) [25].
58
2 Lanthanocenes
Scheme 2-1. Preparation of Yb(Me3SiC5H4)2. SmI2(THF)2 reacts with two equivalents of K[(Me3Si)2C5H3] to give the monosolvate [(Me3Si)2C5H3]2Sm(THF) which can be desolvated to afford base-free Sm[(Me3Si)2C5H3]2 [26]. [(Me3Si)2C5H3]2Yb(THF) was made analogously and structurally characterized [27]. The bent metallocenes Ln[(Me3Si)2C5H3]2 (Ln = Eu, Yb) have been structurally investigated. Both compounds are polymeric and display unusual intermolecular contacts between the metal centers and γ-methyl groups of neighboring (Me3Si)2C5H3 ligands. The europium derivative exhibits an unprecedented structure with a cyclopentadienyl ring bridging η3:η5 between two nonequivalent europium atoms [28]. Various O- or N-donor-substituted cyclopentadienyl ligands have been shown to be very useful in the stabilization of divalent lanthanide metallocenes. A typical example is the methoxyethylcyclopentadienyl ligand, which forms solvated and unsolvated complexes with divalent Sm and Yb (Eq. (2-8)) [29]: THF LnI2 + 2 K(MeOCH2CH2C5H4) → (MeOCH2CH2C5H4)2Ln + 2 KI (2-8) Ln = Sm, Yb Equally useful properties are exhibited by the 1,1'-(3-oxa-pentamethylene)dicyclopentadienyl ligand, which allowed the preparation of the structurally characterized Yb(II) complex [O(CH2CH2C5H4)2]Yb(DME) [30–32]. Intramolecular chelate stabilization has also been achieved with the amino-functionalized ligand C5Me4CH2CH2NMe2. Treatment of SmI2 with two equivalents of Li(C5Me4CH2CH2NMe2) produced dark green, unsolvated Sm(C5Me4CH2CH2NMe2)2 (1) [33, 34]. N Sm N
1
2.2 Lanthanide(II) Metallocenes
59
An interesting novel aspect of divalent lanthanocene chemistry is the successful synthesis of a ‘metalloligand’. The phosphinoytterbocenes (C5H4PPh2)2Yb(THF) and (C5H4PPh2)2Yb(DME) have been prepared using different synthetic approaches (Eqs. (2-9)–(2-11)) [35]. THF Yb(C6F5)2 + 2C5H5PPh2 → (C5H4PPh2)2Yb(THF) + 2C6F5H
(2-9)
THF Yb + Hg(C6F5)2 + 2C5H5PPh2 → (C5H4PPh2)2Yb(THF) + Hg + 2C6F5H (2-10) DME Yb + 2TlC5H4PPh2 → (C5H4PPh2)2Yb(DME) + 2Tl
(2-11)
Divalent lanthanide metallocenes are generally highly reactive. Especially the Sm(II) and Yb(II) derivatives are powerful one-electron reducing agents. They easily undergo redox reactions and have frequently been used as precursors to prepare bis(cyclopentadienyl)lanthanide(III) complexes. For example, (C5H5)2Yb(DME) can be treated with various Hg, Tl, Cu, or Ag salts to give the corresponding trivalent (C5H5)2YbX derivatives. The main advantage of such oxidative transformations is that the products can be isolated free of alkali metal halides, which tend to be readily incorporated in the products (Eqs. (2-12)) [36, 37]. (C5H5)2Yb(DME) + MX → (C5H5)2YbX + M –DME M = ½ Hg, Tl, Cu, Ag X = Cl, Br, I, C≡CPh, C6F5, O2CMe, O2CC6F5, O2CC5H4N, (MeCO)2CH, (PhCO)2CH
(2-12)
Decamethyllanthanocenes(II) of the divalent lanthanides form a truly remarkable class of organolanthanide complexes. The successful preparation and characterization of unsolvated decamethylsamarocene by Evans et al. can be considered a landmark in organo-f-element chemistry. This compound is perhaps the most reactive organometallic compound ever made and has given rise to a large number of unusual reactions and unexpected products. The most straightforward synthesis involves the reaction of SmI2 with two equivalents of K(C5Me5) (Eq. (2-13)) [38]. Unsolvated Sm(C5Me5)2 can be obtained by desolvation and sublimation of the THF adduct under high vacuum [39]. Eu(C5Me5)2 is prepared analogously from EuI2(THF)2 and K(C5Me5). In this case the resulting THF solvate is more difficult to desolvate than the samarium analog and pure Eu(C5Me5)2 was obtained only after repeated vacuum sublimation [40]. Base-free Yb(C5Me5)2 is accessible analogously.
60
2 Lanthanocenes
THF SmI2(THF)2 + 2K(C5Me5) → (C5Me5)2Sm(THF)2 + 2KI ↓ Sm(C5Me5)2
(2-13)
Ether adducts of the type (C5Me5)2LnL (L = Et2O, THF) can also be obtained by reacting LnCl2 with two equivalents of Na(C5Me5) in THF or diethyl ether (Eq. (214)) [41]. In the case of europium EuCl3 may be employed as starting material because reduction to the divalent oxidation state takes place during the course of the reaction with three equivalents of Na(C5Me5) (Eq. (2-15)) [40]. +L LnCl2 + 2Na(C5Me5) → (C5Me5)2LnL + 2NaCl Ln = Sm, Eu, Yb
(2-14)
1. THF EuCl3 + 3Na(C5Me5) → (C5Me5)2Eu(Et2O)(THF) + 3NaCl + ½(C5Me5)2 (2-15) 2. Et2O (C5Me5)2Sm(THF)2 (2) is isolated as a purple solid, while Sm(C5Me5)2 (3) forms dark green crystals. Both complexes are highly air-sensitive, turning yellow in the presence of traces of oxygen. The molecular and crystal structure of (C5Me5)2Sm(THF)2 was determined by X-ray diffraction. The molecule adopts a typical bent metallocene structure with a pseudo-tetrahedral coordination geometry around the divalent samarium [2–7, 39].
O Sm
Sm O
2
3
The unusual molecular structure of Sm(C5Me5)2 consists of bent metallocene molecules with a centroid–Sm–centroid angle of 140.1°. The shortest separation between two monomeric units is 3.22(1) Å [42]. Base-free Eu(C5Me5)2 and Yb(C5Me5)2 have also been structurally characterized by X-ray analyses [42, 43], showing once again a bent metallocene geometry similar to that in the samarium analog or in M(C5Me5)2 (M = Ca, Ba) [44, 45]. Weak van der Waals attractive
2.2 Lanthanide(II) Metallocenes
61
forces between C5Me5 rings are made responsible for the bent structures in the solid state [46, 47]. Very few examples of π-adducts of lanthanocenes(II) have been described. In the case of π-bonded alkynes it was possible to isolate the adduct (C5Me5)2Yb(η2MeC≡CMe), made from Yb(C5Me5)2 and the free alkyne ligand [48]. The structural data indicate a weak coordination of the alkyne ligand and little or negligible π-backbonding (Yb–C(C5Me5) 2.659(9), Yb–C(η2) 2.85(1) Å). The same is true for the first organolanthanide olefin complex. Bimetallic (C5Me5)2Yb(µC2H4)Pt(PPh3)2 was prepared by reacting Yb(C5Me5)2 and (C2H4)Pt(PPh3)2 (Eq. (2-16), Yb–C(C5Me5) 2.67(2), Yb–C(η2) 2.781(6) Å) [49]. Yb(C5Me5)2 + (C2H4)Pt(PPh3)2 → (C5Me5)2Yb(µ-C2H4)Pt(PPh3)2 (2-16) In the platinum complex (C2H4)Pt(PPh3)2 the coordinated ethylene is more electron-rich than the free olefin and can thus interact as a Lewis base with the Lewis acid Yb(C5Me5)2 [49]. A truly remarkable compound is the substituted methane complex (C5Me5)2Yb(µ-Me)Be(C5Me5), which was prepared according to Eq. (217) and structurally characterized. The most unusual structural feature is the nearly linear Yb–Me–Be unit (177.2°) [50]. Yb(C5Me5)2 + MeBe(C5Me5) → (C5Me5)2Yb(µ-Me)Be(C5Me5)
(2-17)
Several examples of ansa-lanthanocenes(II) have been reported in the literature. For example, the trimethylene-bridged bis(cyclopentadienyl) ligand has been successfully employed in the preparation of divalent lanthanide ansa-metallocenes. Treatment of LnCl2 with the disodium salt of the ligand in THF solution afforded the solvated species [C5H4(CH2)3C5H4]Ln(THF)2 (Ln = Sm, Yb). The ytterbium(II) derivative was structurally characterized (Yb–C 2.67(2)–2.72(2) Å, ring centroid– Yb–ring centroid 127(1)°) [51]. Several complexes containing the tetramethylethylene-bridged dicyclopentadienyl ligand have also been reported. Direct reaction of activated (HgCl2) Ln metal powders with 6,6-dimethylfulvene in THF afforded the tetramethylethylene bridged lanthanide(II) metallocenes [Me4C2(C5H4)2]Ln (Ln = Sm, Yb; Eq. (2-18)) [52].
(2-18)
62
2 Lanthanocenes
2.3 Lanthanide(III) Metallocenes 2.3.1 Lanthanide(III) Metallocenes Containing Ln–Group 17 Element Bonds Lanthanocene fluorides are not generally available starting directly from LnF3 due to the insolubility of anhydrous lanthanide trifluorides. Thus the normal metathetical reactions cannot be applied to these systems, although a few fluorides have been synthesized by other routes. For example, Yb(MeC5H4)2 reacts with perfluoroolefins under defluorination to give (MeC5H4)2YbF(THF) [53]. In contrast to the less-investigated fluorides, lanthanocenes of the type (C5H5)2LnCl play a key role as intermediates in the synthesis of other organolanthanide complexes. The synthetic routes leading to lanthanocene chlorides are summarized in Eqs. (2-19)– (2-24) [2–7]. LnCl3 + 2MC5H5 → (C5H5)2LnCl + 2MCl M = Na, Tl
(2-19)
ScCl3 + Mg(C5H5)2 → (C5H5)2ScCl + MgCl2
(2-20)
Ln(C5H5)3 + NH4Cl → (C5H5)2LnCl + C5H6 + NH3
(2-21)
Ln(C5H5)3 + HCl → (C5H5)2LnCl + C5H6
(2-22)
NaC5H5 (C5H9C5H4)ErCl2(THF)n → (C5H5)2ErCl(THF) (C5H9C5H4 = cyclopentylcyclopentadienyl)
(2-23)
LnCl3 + 2Ln(C5H5)3 → 3(C5H5)2LnCl
(2-24)
Due to coordinative unsaturation (C5H5)2LnCl derivatives are often not available for the lighter lanthanide elements La–Nd. In general, the compounds (C5H5)2LnCl are unstable as monomers. In benzene solution as well as in the solid state, these compounds form chloro-bridged dimers of the type [(C5H5)2Ln(µ-Cl)]2 (4).
Cl Ln
Ln Cl 4
2.3
Lanthanide(III) Metallocenes
63
Treatment with THF readily cleaves the chloride bridges in [(C5H5)2Ln(µ-Cl)]2 under formation of the monomeric 1:1 adducts (C5H5)2LnCl(THF) [2–7]. The molecular and crystal structure of (C5H5)2LuCl(THF) was determined by X-ray crystallography [54, 55]. The molecule exhibits a pseudo-tetrahedral coordination geometry. In addition, the THF-solvated dimers [(C5H5)2Ln(µ-Cl)(THF)]2 (Ln = Nd, Er, Yb) were structurally characterized [56–58]. Bis(cyclopentadienyl)lanthanide hydrocarbyls can also serve as starting materials for complexes with classical co-ligands. For example, the heterobimetallic lanthanocene chloride complex (C5H5)2Yb(µ-Cl)2AlMe2 was synthesized by reacting dimeric [(C5H5)2Yb(µ-Me)]2 with MeAlCl2. The reaction proceeds via cleavage of the ytterbium–carbon bonds [2–7]. Mono- and disubstituted cyclopentadienyl ligands have been extensively employed in the preparation of lanthanocene halides. The chloro-bridged dimers [(MeC5H4)2Yb(µ-Cl)]2 [59] and [(MeC5H4)2Sm(µ-Cl)(THF)]2 have been structurally characterized [60]. t-Butyl-substituted cyclopentadienyl ligands are also increasingly utilized in organolanthanide chemistry. Dimeric chloro-bridged molecular structures were established by X-ray crystallography for the t-butylcyclopentadienyl lanthanide chlorides [(tBuC5H4)2Ln(µ-Cl)]2 (Ln = Pr, Gd, Sm, Lu) [61, 62] and [(tBu2C5H3)2Ln(µ-Cl)]2 (Ln = Ce, Nd, Lu) [63, 64]. Disubstituted lanthanide(III) halides containing the trimethylsilylcyclopentadienyl ligand include the base-free dimers [(Me3SiC5H4)2Yb(µ-X)]2 (X = Cl, I) [25] as well as the ‘ate’ complex precursors (Me3SiC5H4)2Y(µ-Cl)2Li(L) (L = DME, 1/2 THF) [65], (Me3SiC5H4)2Yb(µ-Cl)2Li(Et2O)2 [66], and (Me3SiC5H4)2Lu(µCl)2Li(TMEDA) [66]. Especially useful for the synthesis of lanthanocene halides (besides C5Me5) is the very bulky 1,3-bis(trimethylsilyl)cyclopentadienyl anion. With this ligand two series of disubstituted complexes have been prepared for nearly all elements of the lanthanide series. Treatment of lanthanide trichlorides with two equivalents of lithium 1,3-bis(trimethylsilyl)cyclopentadienide initially affords the solvated lithium chloride adducts [(Me3Si)2C5H3]2Ln(µ-Cl)Li(THF)2 which upon heating above 140°C yield the unsolvated dimers [{(Me3Si)2C5H3}2Ln(µ-Cl)]2 (Scheme 2-2) [2–7].
Scheme 2-2. Preparation of [{(Me3Si)2C5H3}2Ln(µ-Cl)]2.
64
2 Lanthanocenes
Dimeric [{MeC5H2(CH2)5}2Nd(µ-Cl)]2 contains fused-ring bicyclic cyclopentadienyl ligands (5) [67].
5 The use of cyclopentadienyl ligands containing ‘pendant-arm’ donor functions, which are able to coordinatively saturate the lanthanide ion, provides an interesting route to base-free lanthanocene halides. A typical example of such a ligand is the 3-methoxypropylcyclopentadienyl anion, C5H4(CH2)3OMe–. The halide bridged dimers [{C5H4(CH2)3OMe}2Ln(µ-Cl)]2 have been prepared by reacting anhydrous lanthanide trichlorides with the sodium salt of the ligand in a 1:2 molar ratio (Eq. (2-25)) [68, 69].
2LnCl3 + 4Na[C5H4(CH2)3OMe]
(2−25)
THF → [{C5H4(CH2)3OMe}2Ln(µ-Cl)]2 + 4NaCl Ln = Y, La, Pr, Nd, Gd, Ho, Er, Yb The molecular and crystal structure of the La derivative has been determined [68]. The C5H4(CH2)3OMe– anions act as chelating ligands with the dangling methoxy functions being coordinated to the central La atoms. Together with the bridging chlorine ligands this coordination mode results in a formal coordination number of 9 around the metal atoms [70]. Various lanthanocene chlorides containing the related 2-methoxyethylcyclopentadienyl ligand have been synthesized. The lanthanum derivative [{C5H4(CH2)2OMe}2La(µ-Cl)]2 is a dimer with the formal coordination number 10 around lanthanum. In contrast, the structurally characterized complexes [C5H4(CH2)3OMe]2LnCl (Ln = Dy, Yb) are unsolvated monomers in which the formal coordination number is 9 [71]. A related tetramethyldisiloxane-bridged ligand provided the lanthanocene chlorides [O(Me2SiC5H4)2]PrCl(THF)n (n = 1, 2) and [{O(Me2SiC5H4)2}Yb(µ-Cl)]2. In
2.3
Lanthanide(III) Metallocenes
65
this case the oxygen atoms in the bridge did not participate in the coordination to ytterbium [72]. Nitrogen-chelation was found in the chloro-bridged dimers of the type [{MeN(CH2CH2C5H4)2}Ln(µ-Cl)]2 (6, Ln = Y, Nd, Sm, Yb) and [{C5H3N(CH2C5H4)2-2,6}Ln(µ-Cl)]2 (Ln = Y, Pr, Nd, Sm, Dy, Er, Yb, Lu) which have been made by using amine-linked bis(cyclopentadienyl) ligands [73, 74].
Cl H3C N
Ln
Ln
N CH3
Cl 6 The use of the donor-substituted (dimethylaminoethyl)cyclopentadienyl ligand has recently allowed the synthesis and structural characterization of the first solventfree, monomeric lanthanocene chloride. (C5H4CH2CH2NMe2)2NdCl 7 was prepared by reacting NdCl3 with two equivalents of K(C5H4CH2CH2NMe2) followed by sublimation [75].
N Nd
Cl N
7 Especially in the area of lanthanocene(III) complexes the use of the bulky pentamethylcyclopentadienyl ligand has had a strong influence on the development of organolanthanide chemistry. This particular ligand provides enhanced stability and better solubility to many such complexes and its use was crucial for the successful synthesis of many highly reactive and catalytically active lanthanocene derivatives. Decamethyllanthanocene fluorides are not generally available via the normal metathetical routes because of the insolubility of the LnF3 precursors. However, defluorination of perfluoroolefins by the divalent organolanthanides (C5Me5)2Ln(L) (Ln = Sm, Eu, Yb; L = Et2O, THF) has been found to be an elegant way to make the corresponding lanthanide(III) fluorides. The compounds (C5Me5)2LnF(L) (Ln = Sm, Eu, Yb; L = Et2O, THF) have been prepared and fully characterized. Crystal structure determinations of (C5Me5)2YbF(Et2O) and (C5Me5)2YbF(THF) revealed the first terminal Ln–F bonds (Yb–F 2.015(4) and 2.026(2), respectively) [53].
66
2 Lanthanocenes
Dimeric mixed-valence (µ-F)[Yb(C5Me5)2]2 has been prepared and structurally characterized. The compound contains an asymmetric Yb2+–F–Yb3+ bridge [76]. Complexes containing a (C5Me5)2LnX (X = Cl, Br, I) unit are easily available by reacting anhydrous lanthanide trihalides with two equivalents of a C5Me5 transfer reagent. These reactions are generally carried out in a coordinating solvent such as THF or DME. Normally a monomeric (C5Me5)2LnX unit is coordinatively unsaturated. As a consequence, (C5Me5)2LnX complexes show a strong tendency to either form adducts with coordinating solvent molecules or to retain one equivalent of alkali metal halide and form ‘ate’ complexes. At least five different types of decamethyllanthanocene halides have been characterized: 1. A monomeric (C5Me5)2LnCl complex is stable only in the case of the small scandium(3+) ion. 2. THF adducts of the type (C5Me5)2LnX(THF) have been prepared and structurally characterized for several lanthanide elements. 3. In most cases (C5Me5)2LnX derivatives are initially isolated as ‘ate’ complexes of the type (C5Me5)2Ln(µ-X)2ML2 (X = Cl, Br, I; M = Li, Na, K; L = THF, Et2O, 1/2DME, 1/2TMEDA). In these materials one equivalent of alkali metal halide is incorporated to increase the formal coordination number around the lanthanide atom. Lanthanide and alkali metal are bridged by two halide ligands to give a central four-membered Ln(µ-X)2M ring. Formally these compounds could also be described as ‘ate’ complexes of the type [ML2][(C5Me5)2LnX2]. One or two solvent molecules are needed to complete the coordination sphere around the alkali metal ion. 4. Unsolvated species normally exist as symmetrical halide-bridged dimers of the type [(C5Me5)2Ln(µ-X)]2. 5. The compound (C5Me5)2Y(µ-Cl)Y(C5Me5)2Cl is a rare example of an asymmetrical dimer, while [(C5Me5)2Sm(µ-Cl)]3 is trimeric in the solid state. A two-step procedure was reported for the preparation of (C5Me5)2YbCl(THF). Unsolvated (C5Me5)2Yb(µ-Cl)2Li reacts with aluminum trichloride to give the blue bimetallic complex (C5Me5)2Yb(µ-Cl)2AlCl2. Subsequent treatment of the intermediate tetrachloroaluminate with THF results in elimination of AlCl3 and formation of (C5Me5)2YbCl(THF) [2–7]. An elegant way of preparing salt-free THF adducts of the type (C5Me5)2LnX(THF) (Ln = Sm, Eu, Yb; X = Cl, I) is the controlled oxidation of lanthanocenes(II) using alkyl or aryl halides [77, 78]. All (C5Me5)2LnX(THF) derivatives were found to be isostructural exhibiting a pseudotetrahedral geometry around the metal atom. Synthesis and characterization of ‘ate’ complexes of the type (C5Me5)2Ln(µX)2ML2 (X = Cl, Br, I; M = Li, Na, K; L = THF, Et2O, 1/2DME) have been described in detail in a number of publications. A typical reaction sequence is shown in Eq. ((2-26)). THF LnCl3 + 2Li(C5Me5) → (C5Me5)2Ln(µ-Cl)2Li(THF)2 + LiCl
(2-26)
2.3
67
O
Cl Ln
Lanthanide(III) Metallocenes
Li Cl
O
8 Most of these reactions have been carried out with the lanthanide trichlorides (cf. 8), while the corresponding bromides and iodides are much less investigated. Li(C5Me5), Na(C5Me5), and K(C5Me5) have been used as pentamethylcyclopentadienyl transfer reagents, and suitable co-ligands are diethyl ether, THF, and DME or TMEDA. Structurally characterized complexes include [(C5Me5)2Ce(µ-Cl)2K(THF)]n [79], (C5Me5)2Pr(µ-Cl)2Na(DME)2 [80], and (C5Me5)2Yb(µ-Cl)2Li(Et2O)2 [2–7]. Ring-bridged bifunctional cyclopentadienyl ligands play an important role in the preparation of lanthanocene derivatives. It was found that complexes containing a (C5H5)2LnCl unit are unstable for the early lanthanide elements La–Nd. This problem can be circumvented with the use of bridged cyclopentadienyl ligands. Several ligand systems have been developed, in which the cyclopentadienyl rings are connected through simple hydrocarbon chains. More sophisticated ligands have donor atoms such as oxygen or nitrogen incorporated in the bridge. These allow the isolation of unsolvated lanthanocene halide derivatives. The first ligand of this type which has been employed in organolanthanide chemistry was the dianion of 1,1'-trimethylene-bis(cyclopentadiene), –C5H4(CH2)3C5H4–. The disodium salt of this ligand was reacted with anhydrous lanthanide trichlorides to give exclusively ansa-lanthanocenes of the type [C5H4(CH2)3C5H4]LnCl(THF) (Eq. (2-27)). The formation of stable products with the lighter lanthanide elements was in marked contrast to the corresponding reactions with NaC5H5 [81–83]. LnCl3 + Na2[C5H4(CH2)3C5H4] THF →
(2-27)
[C5H4(CH2)3C5H4]LnCl(THF) + 2NaCl Ln = La, Ce, Pr, Nd, Gd, Dy, Ho, Er, Yb, Lu
In several recent papers Marks et al. reported the synthesis, structural systematics, absolute configurations, and structural interconversions of a series of C1-symmetric organolanthanide chloro, alkyl, and amide complexes based on chiral chelating Me2Si(C5Me4)(C5H3R*)2– ligands (= [Me2SiCp″(R*Cp)]2–), where R* = (+)-neomenthyl, (–)-menthyl, and (–)-phenylmenthyl [84–86]. These ligands are accessible in three steps starting from known chiral cyclopentadienes (Scheme 2-3). The synthetic procedure involves metallation of the chiral cyclopentadienes followed by condensation with Me4C5HSiMe2Cl and subsequent in situ lithiation to give the corresponding dianions of the chelating ligands.
68
2 Lanthanocenes
Scheme 2-3. Synthesis of chiral chelating cyclopentadienyl ligands.
The lanthanide chloro precursors can be prepared in high yield by transmetallation of the dilithiated ligands with anhydrous lanthanide trichlorides followed by workup with diethyl ether. Alternatively, the mono-DME adducts can be prepared (Eqs. (2-28) and (2-29)) [86].
2.3
Lanthanide(III) Metallocenes
Li2[Me2SiCp″(R*Cp)] + LnCl3
69 (2-28)
1. THF → [Me2SiCp″(R*Cp)]Ln(µ-Cl)2Li(Et2O)2 + LiCl 2. Et2O R* = (+)-neomenthyl, (–)-menthyl, (–)-phenylmenthyl Ln = Y, La, Nd, Sm, Lu
Li2[Me2SiCp″(R*Cp)] + LnCl3
(2-29)
1. THF → [Me2SiCp″(R*Cp)]Ln(µ-Cl)2Li(DME) + LiCl 2. DME R* = (+)-neomenthyl, (–)-menthyl Ln = Sm, Lu As established by NMR spectroscopy and circular dichroism, the diastereomerically pure chloro complexes can be epimerized in the presence of donor solvents to afford mixtures of R- and S-configurational isomers. The isomer ratio depends on solvent, R*, and lanthanide ion. Importantly, selective epimerization allowed enrichment in either antipode with diastereomerically pure complexes obtained in a single recrystallization step. This was a crucial prerequisite for the preparation of σ-alkyl or amide precatalysts of desired configuration. The molecular structure of [Me2SiCp″{(+)-neomenthylCp}]Lu(µ-Cl)2Li(Et2O)2 has been determined by singlecrystal X-ray analysis [86].
2.3.2 Lanthanide(III) Metallocenes Containing Ln–Group 16 Element Bonds Various lanthanocenes containing Ln–O, Ln–S, Ln–Se, and Ln–Te bonds have been reported in the literature, with oxygen-bonded species forming by far the largest class of compounds. In recent years several examples of µ-hydroxo- and µoxo-bridged organolanthanide complexes of the type [Cp2Ln(µ-OH)]2 and (µO)[LnCp2]2 have been prepared and structurally characterized. They are generally made by carefully controlled hydrolysis of the corresponding Cp2LnX precursors (X = halide, cyclopentadienyl). The µ-oxo bridged ytterbium complex (µO)[Yb(MeC5H4)2]2 exhibits a linear central Yb–O–Yb unit (Yb–O 2.915(1), Yb–C 2.667 Å) [87].
70
2 Lanthanocenes
Various dimeric alkoxides of the type [(C5H5)2Ln(µ-OR)]2 (and in some cases monomeric THF solvates (C5H5)2Ln(OR)(THF)n) have been prepared by treatment of the chloro complexes with sodium alkoxides [2–7, 88] or by protonation of Ln(C5H5)3 with stoichiometric amounts of alcohols [89–92]. The latter method has been used to prepare a series of bis(cyclopentadienyl)ytterbium alkoxides containing chiral alkoxide ligands such as (–)-mentholate [92, 93]. A typical example of a monomeric species is (C5H5)2Nd(OC6H3Ph2-2,6)(THF)2 [89]. The alkoxide complexes (C5H5)2Yb(THF)(ORF) [Ln = Nd, Sm, Yb; RF = 2,4,6-tris(trifluoromethyl)phenyl, -C6H2(CF3)3-2,4,6] were obtained by reaction of the appropriate (C5H5)3Ln with 2,4,6-tris(trifluoromethyl)phenol (RFOH) (Eq. (2-30)) [94]. The thiolate complex (C5H5)2Yb(THF)(SRF) was prepared analogously from (C5H5)3Yb and RFSH. According to X-ray diffraction studies the (C5H5)2Yb(THF)(SRF) is a monomer, in which the ytterbium atom is coordinated by two C5H5-rings, one sulfur of the SC6H2(CF3)3-2,4,6 ligand and one oxygen of solvating THF molecule. Thus the ytterbium has a pseudo-tetrahedral arrangement and the formal coordination number 8.
(2-30) Lanthanocene(III) carboxylates are also dimeric in the solid state. They can be obtained by reacting (C5H5)2LnCl(THF) with various sodium carboxylates in THF solution [2–7, 36]:
(2-31)
2.3
Lanthanide(III) Metallocenes
71
In some cases anhydrous lanthanide triflates have been found to be better precursors for lanthanocenes than the previously used lanthanide trichlorides [95]. Dimeric structures have been reported for lanthanocene triflates, which have been prepared according to Eq. (2-32) [(C5H5)2Yb(µ-O3SCF3)]2 was crystallographically characterized [96, 97]. Ln(O3SCF3)3 + 4Na(C5H5) THF →
(2-32) [(C5H5)2Ln(µ-O3SCF3)]2 + 4NaO3SCF3 Ln = Sc, Yb, Lu
An unusual complex containing a tetradentate oxygen ligand was formed by insertion of carbon monoxide into the reactive Lu–C bond in (C5H5)2LutBu(THF). The initial product in this reaction is the yellow acyl complex (C5H5)2Lu[C(O)tBu] which can be described by two resonance forms. Further CO insertion yields the final product via postulated ketene and carbene intermediates (Scheme 2-4) [2–7]. A large number of (C5Me5)2Ln derivatives containing additional Group 16 ligands have been prepared and characterized. Many of them result from transformations of the corresponding divalent samarium or ytterbium precursors. A typical reaction product is easily observed when (C5Me5)2Sm(THF)2 decomposes in the presence of traces of oxygen. The purple color of the samarium(II) complex then changes to orange–yellow, indicating the formation of (µ-O)[Sm(C5Me5)2]2. This oxygen-bridged dimer is also commonly the final product when (C5Me5)2Sm(THF)2 is reacted with various oxygen-containing substrates such as nitrosobenzene, pyridine-N-oxide etc [98]. The binuclear complex (µ-O)[Sm(C5Me5)2]2 has been structurally characterized. As imposed crystallographically, the compound contains an exactly linear Sm–O–Sm bridge. The two (C5Me5)2Sm units are in a perpendicular arrangement relative to each other [99]. Few alkoxide and aryloxide complexes of the type (C5Me5)2LnOR have been reported. An unsolvated samarium complex was prepared by reacting (C5Me5)2Sm(THF)2 with 2,3,5,6-tetramethylphenol in toluene solution (Eq. (2-33)) [100]: (C5Me5)2Sm(THF)2 + HOC6HMe4 → (C5Me5)2Sm(OC6HMe4) + ½H2 (2-33) –2THF A reaction sequence leading to an unusual alkoxide derivative was performed by reacting (C5Me5)2Sm(THF)2 first with diphenylacetylene and subsequently treating the initial product with carbon monoxide [101]. Formal addition of two CO molecules to diphenylacetylene resulted in the formation of an indenoindene ring system. In the final product two (C5Me5)2Sm units were coordinated only to oxygen atoms of the new ligand showing again the strong preference of lanthanides for hard donor ligands (Scheme 2-5) [101].
72
2 Lanthanocenes
Scheme 2-4. Reactions of (C5H5)2LutBu(THF) with carbon monoxide. One of the most unusual organolanthanide complexes results from the reaction of (C5Me5)2Sm(THF)2 with carbon monoxide alone. Treatment of the samarium(II) precursor with CO under pressure gave a black crystalline material which was shown by X-ray structure analysis to be the tetranuclear samarium(III) complex [(C5Me5)4Sm2(µ-O2CCCO)(THF)]2 [102]. The result of this unexpected reaction is a trimerization of carbon monoxide to give a ketenecarboxylate dianion,
2.3
Lanthanide(III) Metallocenes
73
Scheme 2-5. Sequential reaction of (C5Me5)2Sm(THF)2 with PhC≡CPh and CO. O2CCCO2–. In the product, two of these ketenecarboxylate dianions act as bridging ligands between four (C5Me5)2Sm units. (C5Me5)2Ln derivatives containing sulfur ligands can be prepared in a conventional manner from the corresponding chloro complexes by salt-elimination reactions [103]: THF (C5Me5)2LnCl(THF) + NaS2CNEt2 → (C5Me5)2Ln(S2CNEt2) + NaCl (2-34) Ln = Nd, Yb Protolysis of hydrocarbyls by free alcohols or thiols also leads to decamethyllanthanocene alkoxides and thiolates. For example, methane is eliminated upon treatment of (C5Me5)2LuMe with ethanol to afford the monomeric alkoxide (C5Me5)2LuOEt [104]. One of the first organolanthanide complexes containing bridging thiolate ligands was prepared according to Eq. (2-35) [105]. (C5Me5)2Lu(µ-Me)2Li(THF)2 + 2tBuSH
(2-35)
→ (C5Me5)2Lu(µ-StBu)2Li(THF)2 + 2CH4 Other complexes in which (C5Me5)2Ln units are bonded to organic ligands via Group 16 atoms have been obtained starting from the divalent lanthanide precur-
74
2 Lanthanocenes
sors (C5Me5)2Sm(THF)2 or (C5Me5)2Yb(Et2O) via reductive cleavage of E–E bonds (E = S, Se, Te) [106]: 2(C5Me5)2Sm(THF)2 + [Me2NC(S)S]2
(2-36)
→ 2(C5Me5)2Sm(S2CNMe2) + 4THF 2(C5Me5)2Sm(THF)2 + REER
(2-37)
→ 2(C5Me5)2Sm(ER)(THF) + 2THF E = S, Se, Te R = Mes, C6H2(CF3)3-2,4,6 Mono- and dinuclear (C5Me5)2Yb complexes of the type (C5Me5)2Yb(ER)L (E = O, S, Se, Te; L = Lewis base), (µ-E)[Yb(C5Me5)2]2 (E = S, Se, Te) and (µTe2)[Yb(C5Me5)2]2 have been investigated by Andersen et al. [107–109]. Monomeric species are readily formed upon treatment of (C5Me5)2Yb(Et2O) with dichalcogenides. The ytterbium(II) precursor also reacts with elemental sulfur, selenium, or tellurium to afford the µ-E or µ-E2-bridged complexes. The monomeric ytterbium(III) chlacogenolates (C5Me5)2Yb(SPh)(NH3) (Yb–C 2.64(2), Yb–N 2.428(5), Yb–S 2.674(5) Å) and (C5Me5)2Yb(TePh)(NH3) (Yb–C 2.63(3), Yb–N 2.50(1), Yb–Te 3.039(1) Å) were crystallographically characterized [108, 109]. In the binuclear complex (µ-Te2)[Yb(C5Me5)2]2 the two (C5Me5)2Yb fragments are bridged by a side-on coordinated µ-Te2 unit (Yb–C 2.63(2), Yb–Te 3.156(4) Å) [109]. Related scandium tellurides have been prepared either from (C5Me5)2ScR or (C5Me5)2ScD. Elemental tellurium slowly inserts into the Sc–C bond of the scandium hydrocarbyls to afford the corresponding tellurolates. More rapid Te insertion can be achieved when nBu3PTe is used as a soluble tellurium source. The tellurolates decompose upon heating in solution to afford the Te-bridged binuclear complex (µ-Te)[Sc(C5Me5)2]2 (Scheme 2-6). Photochemically induced reversion of the thermal reaction was observed upon UV irradiation of a mixture of (µTe)[Sc(C5Me5)2]2 and TeR2 [110].
Scheme 2-6. Formation of (µ-Te)[Sc(C5Me5)2]2.
2.3
Lanthanide(III) Metallocenes
75
2.3.3 Lanthanide(III) Metallocenes Containing Ln–Group 15 Element Bonds Basically two different synthetic routes have been employed to prepare lanthanocene complexes with Ln–N bonds. One of them is the metathesis of (C5H5)2LnCl derivatives with metallated nitrogen compounds and the other one involves the protolysis of lanthanide–carbon bonds by N–H functions. A typical example for the metathetical preparation is the synthesis of the lanthanide pyrrolyl complexes (C5H5)2Ln(NC4H4) (Ln = Sm, Dy, Yb, Lu) and (C5H5)2Lu(NC4H2Me2)(THF) from (C5H5)2LnCl(THF) and the corresponding sodium pyrrolyls. The dimethylpyrrolyl derivative has been established by X-ray crystallography (Lu–N 2.289(4) Å) [111, 112]. Both oxygen and nitrogen coordination has been observed in the compound bis[µ-acetoneoximato-bis(cyclopentadienyl)gadolinium], [(C5H5)2Gd(µ-η2-ONCMe2)]2, which was prepared by reacting Gd(C5H5)3 with acetone oxime in THF solution (Eq. (2-38)). An X-ray crystal structure analysis revealed an interesting structure with the N–O fragment of the acetone oximato ligand acting as both a bridging and side-on donating group. This bonding situation represents a novel coordination mode of an oximato ligand to a metal [113]. 2Gd(C5H5)3 + 2HON=CMe2
(2-38)
THF → [(C5H5)2Gd(µ-η2-ONCMe2)]2 + 2C5H6 Among all Group 15 elements nitrogen donor ligands are best suited for coordination to lanthanides. In accordance with the hard Lewis acid character of the lanthanide(3+) cations, the number of stable organolanthanide complexes containing Ln–P, Ln–As, Ln–Sb, and Ln–Bi bonds is quite small. The first example of an ‘organolanthanide-phosphine’ was reported by Schumann et al. The phosphide bridged lutetium complex (C5H5)2Lu(µ-PPh2)2Li(TMEDA)(toluene)0.5 was characterized by an X-ray analysis (Lu–P 2.782(1) and 2.813(2) Å) [114]. (C5H5)2Lu(µAsPh2)2Li(TMEDA) was prepared analogously (Lu–C 2.59(1), Lu–As 2.88(1) Å) [114–116]. Other lanthanocene phosphides have been prepared according to Eq. (239) by treatment of halide precursors with lithium diorganophosphides [2–7, 117]. (C5H5)2LnCl(THF) + LiPRR' THF →
(2-39)
(C5H5)2LnPRR'(THF) + LiCl Ln = Tb, Ho, Er, Tm, Yb, Lu R = R' = tBu, c-C6H11, Ph R = tBu, R' = Ph
76
2 Lanthanocenes
The lutetium hydrocarbyl (C5H5)2Lu(CH2SiMe3)(THF) has been used to prepare compounds containing Lu–P and Lu–As bonds. Treatment of the precursor with diphenylphosphine or diphenylarsine resulted in the formation of (C5H5)2LuPPh2(THF) and (C5H5)2LuAsPh2(THF), respectively (Eq. (2-40)). The preparation is very clean because volatile tetramethylsilane is the only by-product [118]. (C5H5)2Lu(CH2SiMe3)(THF) + HEPh2 →
(2-40)
(C5H5)2LuEPh2(THF) + SiMe4 E = P, As
Reactions of (C5Me5)2Sm(THF)2 with Group 15 compounds are often quite unusual and in several cases further transformations of the organic ligands in the coordination sphere of samarium can be carried out. A typical example for such a reaction sequence is the Sm-mediated insertion of carbon monoxide into the N=N double bond of azobenzene [119, 120]. Treatment of (C5Me5)2Sm(THF)2 with PhN=NPh in toluene (molar ratio 2:1) initially affords the binuclear complex (µη1:η1-PhNNPh)[Sm(C5Me5)2]2 in which the two decamethylsamarocene units are in a trans arrangement. The samarium azobenzene complex was subsequently treated with carbon monoxide, which resulted in a double insertion of CO into the nitrogen–nitrogen bond to give the binuclear complex [µ-η4(PhN)OCCO(NPh)][Sm(C5Me5)2]2 (Scheme 2-7) [119–121].
Scheme 2-7. Sequential reaction of (C5Me5)2Sm(THF)2 with PhN=NPh and CO.
2.3
Lanthanide(III) Metallocenes
77
The use of the bulky bis(trimethylsilyl)amide ligands allows the synthesis of unsolvated organolanthanide amides of the type (C5Me5)2LnN(SiMe3)2. The molecular structure of (C5Me5)2SmN(SiMe3)2 has been determined [2–7, 122, 123]. (C5Me5)2Ln(µ-Cl)2Li(Et2O)2 + MN(SiMe3)2
(2-41)
→ (C5Me5)2LnN(SiMe3)2 + MCl + LiCl Ln = Y, Ce, Nd, Sm M = Li, Na, K Similar amidation of the chiral chloro complexes [Me2SiCp″(R*Cp)]Ln(µCl)2Li(Et2O)2 affords the corresponding chiral amides in high yields (Eq. (2-42)) [86]. [Me2SiCp″(R*Cp)]Ln(µ-Cl)2Li(Et2O)2 + MN(SiMe3)2
(2-42)
→ [Me2SiCp″(R*Cp)]LnN(SiMe3)2 + LiCl + MCl R* = (+)-neomenthyl, (–)-menthyl, (–)-phenylmenthyl M = Na, K; Ln = Y, La, Nd, Sm, Lu Single crystal structure analyses have been carried out on the complexes S[Me2SiCp″{(+)-neomenthylCp}]SmN(SiMe3)2, S-[Me2SiCp″{(–)-menthylCp}]SmN(SiMe3)2, and R-[Me2SiCp″{(–)-menthylCp}]YN(SiMe3)2. These amide complexes are configurationally stable in toluene at 60 °C for many hours [86].. Adducts of (C5Me5)2Sm with bipyridine and diazadienes (Eq. (2-43)) are readily formed by treatment of (C5Me5)2Sm(THF)2 with the corresponding nitrogen donor ligands [124, 125]: (C5Me5)2Sm(THF)2 + RN=CH–CH=NR
(2-43)
→(C5Me5)2Sm(η2-RN=CH–CH=NR) + 2 THF R = iPr, tBu, c-C6H11, p-C6H4Me The N-t-butyl derivative (C5Me5)2Sm(η2-tBuN=CH–CH=NtBu) has been structurally characterized (Sm–C 2.773(7), Sm–N 2.480(5) and 2.489(5) Å) [125]. Structural as well as the spectroscopic data are consistent with the presence of Sm(3+) complexes of the bipyridine or diazadiene radical anions. The reaction of (C5Me5)2La(µ-Cl)2K(DME)2 with one equivalent of Na2(DAD) [DAD = (Ph)N=C(Ph)C(Ph)=N(Ph)] in the presence of DAD or with two equivalents of Na(DAD) gave the ionic complex [Na(DAD)][(C5Me5)2La(DAD)] (Scheme 2-8) [126]. The interaction of (C5Me5)2Y(µ-Cl)2Li(OEt2)2 with Na(DAD) [DAD = (pMeC6H4)N=C(Ph)C(Ph)=N(C6H4Me-p)] led to (C5Me5)2Y(DAD), which was structurally characterized. The DAD ligand in the complex is reduced to the radical
78
2 Lanthanocenes
anion. In the structure the yttrium atom has pseudotetrahedral arrangement formed by the two C5Me5 groups and two N atoms of the cis-1,4-diazadiene. The Y– (C5Me5-ring) and Y–N distances are 2.408(4) and 2.362(6) Å, respectively. The metallacycle Y(N–C–C–N) is planar.
Scheme 2-8. Formation of diazadiene complexes.
2.3
Lanthanide(III) Metallocenes
79
Dianions of certain nitrogen heterocycles have been found to act as anti-aromatic bridging ligands between decamethyllanthanocene units. The first complexes of this type were obtained by in situ preparation of the dianion followed by treatment with (C5Me5)2La(µ-Cl)K(DME)2. Eq. (2-44) shows the example of 2,3-dimethylquinoxaline [127].
(2-44) (µ-η2:η2-N2)[Sm(C5Me5)2] 9, the first dinitrogen complex of an f-element, was prepared by slow evaporation of a dark green toluene solution of Sm(C5Me5)2 in a nitrogen atmosphere inside a dry-box. The resulting red–brown crystals were structurally characterized. In the unusual binuclear complex the dinitrogen ligand is side-on coordinated to two Sm(C5Me5)2 units. The central Sm(µ-η2:η2-N2)Sm unit is planar (Sm–C 2.73(2), Sm–N 2.36(1) Å) [128].
N Sm
Sm N 9
80
2 Lanthanocenes
Thus far no polyphosphide complex derived from (C5Me5)2Sm(THF)2 has been reported, although some interesting results have been obtained with antimony and bismuth. Sb(nBu)3 reacts with unsolvated Sm(C5Me5)2 to give the structurally characterized trinuclear product (µ-η2:η2:η1-Sb3)[Sm(C5Me5)2]3(THF) which can be described as a derivative of the triantimony Zintl ion (Sm–C 2.72(2), Sm–Bi 3.29(2) Å) [129]. The analogous reaction with BiPh3 illustrates the fact that it is often difficult if not impossible to predict the outcome of reactions involving (C5Me5)2Sm(THF)2 or Sm(C5Me5)2. In this case the bismuth analog of the abovementioned dinitrogen complex is formed among other products (Eq. (2-45)). (µη2:η2-Bi2)[Sm(C5Me5)2]2 was structurally characterized [130]. 4Sm(C5Me5)2 + 2BiPh3
(2-45)
→ (µ-η2:η2-Bi2)[Sm(C5Me5)2]2 + PhPh + 2(C5Me5)2SmPh More recent studies have revealed that Sm(C5Me5)2 (2 equivalents) reacts with Ph2EEPh2 to produce (C5Me5)2SmEPh2 (E = P, As), while treatment of (C5Me5)2Sm(THF)2 with Ph2EEPh2 affords the THF adducts (C5Me5)2SmEPh2(THF). The mixed-valence dinuclear complexes (C5Me5)2Sm(µ-EPPh2)Sm(C5Me5)2 were prepared by reacting 4 equivalents of Sm(C5Me5)2 with Ph2EEPh2 (E = P, As) [131]. 2.3.4 Lanthanide(III) Metallocenes Containing Ln–Group 14 Element Bonds Lanthanocene hydrocarbyls are normally prepared from the corresponding (C5H5)2LnCl precursors. A generally applicable synthetic route involves reactions of (C5H5)2LnCl derivatives with equimolar amounts of lithium alkyls or aryls in THF solution (Eq. (2-46) [2–7, 132, 133]. Alternative but not so widely used preparations involve treatment of Ln(C5H5)3 with LiR or oxidation of lanthanide(II) metallocenes with HgR2 [134–137]. THF (C5H5)2LnCl(THF) + LiR → –78°C
(C5H5)2LnR(THF) + LiCl
(2-46)
Ln = Y, Nd, Sm, Dy, Er, Tm, Yb, Lu R = Me, Et, iPr, nBu, tBu, CH2tBu, CH2SiMe3, CH2Ph, Ph, C6H4Me-p, C6H4Cl-p, CH2PMe2 (C5H5)2LuCH2SiMe3(THF) has been structurally characterized by X-ray diffraction [88]. The coordination geometry in this complex is pseudo-tetrahedral. A similar structure was found for (C5H5)2YbMe(THF) (Yb–C(σ) 2.36(1)Å) [138]. Monomeric lanthanocene hydrocarbyls in which the lanthanide atoms and lithium are
2.3
Lanthanide(III) Metallocenes
81
bridged by methyl groups are obtained when (C5H5)2LnCl precursors are treated with two equivalents of methyllithium. The resulting products are hydrocarbyl analogs of the well-known lithium chloride adducts (C5H5)2Ln(µ-Cl)2Li(THF)2 [139, 140]. (C5H5)2LnCl(THF) + 2MeLi
(2-47)
THF →(C5H5)2Ln(µ-Me)2Li(THF)2 + LiCl (2-47) Ln = Er, Lu Unsolvated (C5H5)2LnR species have been successfully prepared by a two-step procedure. In the first step, dimeric lanthanocene chlorides are reacted with lithium tetraalkylaluminates, LiAlR4, to give the bimetallic complexes (C5H5)2Ln(µR)2AlR2 (Eq. (2-48)) [2–7]. From these intermediates the trialkylaluminum can be abstracted by addition of pyridine, which forms an insoluble 1:1 adduct of the type R3AlNC5H5. This reaction provides the unsolvated (C5H5)2LnR derivatives in the form of alkyl-bridged dimers (Eq. (2-49)) [2–7]. In some cases simple metathetical reactions also lead to the µ-Me bridged dimers (Eq. (2-50)) [141]. toluene [(C5H5)2Ln(µ-Cl)]2 + 2LiAlR4 → 2(C5H5)2Ln(µ-R)2AlR2 + 2 LiCl (2-48) R = Me; Ln = Sc, Y, Gd, Dy, Ho, Er, Tm, Yb R = Et; Ln = Sc, Y, Ho 2(C5H5)2Ln(µ-Me)2AlMe2 + 2C5H5N
(2-49)
toluene → [(C5H5)2Ln(µ-Me)]2 + Me3AlNC5H5 [(tBuC5H4)2Nd(µ-Cl)]2 + 2MeLi
(2-50)
→ [(tBuC5H4)2Nd(µ-Me)]2 + 2LiCl Lanthanocene derivatives containing neutral phosphoylides as ligands have been prepared by displacement of THF by the ylide. Halides as well as hydrocarbyls can be used as precursors in these ligand exchange reactions [2–7, 142, 143]. (C5H5)2LuCl(THF) + Ph3P=CH2 → (C5H5)2LuCl(CH2PPh3) + THF
(2-51)
82
2 Lanthanocenes
(C5H5)2LuR(THF) + R'3P=CHR″
(2-52)
→(C5H5)2LuR(CHR″PR'3) + THF R = tBu, R' = Ph, R″ = H R = CH2SiMe3, R' = Ph, R″ = H R = tBu, R' = Me, R″ = SiMe3 Intramolecular metallation of a phenyl ring was observed when the phosphoylide adduct (C5H5)2LuCl(CH2PPh3) was treated with strong bases such as MeLi or sodium hydride (Scheme 2-9) [2–7, 143].
Scheme 2-9. Reaction of (C5H5)2LuCl(CH2PPh3) with MeLi. Effective stabilization of lanthanide–carbon bonds can also be affected by chelating anionic phosphoylide complexes of the type R2P(CH2)2–. Several lanthanocene derivatives of these ligands have been reported. The cyclopentadienyl complexes (C5H5)2Ln[(CH2)2PR2] are monomeric in benzene solution. Their preparation is easily achieved by reacting lanthanocene chlorides with lithiated phosphoylides [2– 7, 98, 144]. (C5H5)2LuCl(THF) + Li(CH2)2PtBu2
(2-53)
THF → (C5H5)2Lu[(CH2)2PtBu2] + LiCl Only a very small number of heavier homologs of the lanthanocene hydrocarbyls (i.e. silyls, germyls etc.) have been studied. The anionic silyl complexes [Li(DME)3][(C5H5)2Ln(SiMe3)2] were prepared by Schumann et al. according to Eq. (2-54) and the lutetium derivative was structurally characterized (Lu–Si 2.888(2) Å) [114, 145–147].
2.3
Lanthanide(III) Metallocenes
(C5H5)2LnCl + 2LiSiMe3
83 (2-54)
DME →[Li(DME)3][(C5H5)2Ln(SiMe3)2] + LiCl Ln = Sm, Dy, Ho, Er, Tm, Lu Organolanthanide complexes containing Ln–Ge and Ln–Sn bonds are also accessible. These were obtained by reactions of lanthanocene chloride precursors with MPh3– (Eq. (2-55), M = Ge, Sn) [2–7]. A similar synthetic route was adapted to prepare a series of monomeric bis(cyclopentadienyl)scandium silyl and germyl complexes (Eq. (2-56)) [148]. THF (C5H5)2LnCl(THF) + LiMPh3 → (C5H5)2LnMPh3 + LiCl
(2-55)
[(C5H5)2Sc(µ-Cl)]2 + 2(THF)3LiER3
(2-56)
→2(C5H5)2Sc(ER3)(THF) + 2LiCl + 2THF ER3 = Si(SiMe3)3, Si(SiMe3)2Ph, Si(tBu)Ph2, SiPh3, Ge(SiMe3)3 (C5H5)2Sc[Si(SiMe3)3](THF) was structurally characterized by X-ray crystallography (Sc–Si 2.863(2) Å). Like organolanthanide hydrocarbyls, the scandium silyls and germyls polymerize ethylene and undergo facile migratory insertion reactions with CO and xylylisocyanide [148]. The lanthanocene hydrocarbyls are highly reactive towards a variety of reagents. They easily undergo hydrogenolysis to form the corresponding organolanthanide hydrides. Facile insertion reactions are observed with carbon monoxide and isonitriles. A typical example is the CO insertion into the Lu–C bond in (C5H5)2LutBu. The initial product can be described by a resonance between an acyl complex and the mesomeric alkoxycarbene structure [2–7]:
(2-57)
Especially important as catalytically active species and precursors for catalytically active organolanthanide hydrides are organolanthanide hydrocarbyls of the type (C5Me5)2LnR [2–7, 80, 149, 150]. They are readily prepared from ‘ate’ complexes
84
2 Lanthanocenes
of the type (C5Me5)2Ln(µ-Cl)2ML2 (M = alkali metal, L = Et2O, THF, 1/2DME) by treatment with alkyllithium reagents. In the case of small alkyl groups such as methyl the products are again ‘ate’ complexes in which the lanthanide and alkali metal are bridged by the hydrocarbyl ligands. The stepwise formation of such complexes is outlined in Scheme 2-10:
Scheme 2-10. Stepwise formation of (C5Me5)2Ln(µ-Me)2Li(THF)2. Halide-free (C5Me5)2Sm hydrocarbyls can be obtained by the reaction of [(C5Me5)2Sm(THF)2][BPh4] with lithium alkyls according to Eq. (2-58) [151]: [(C5Me5)2Sm(THF)2][BPh4] + LiR
(2-58)
→ (C5Me5)2SmR(THF) + LiBPh4 Ln = Me, Ph The THF adducts (C5Me5)2LnMe(THF) (Ln = Sc, Y, Sm, Yb, Lu) have been isolated and all are thermally stable. Structurally characterized solvated organolanthanide methyl complexes are (C5Me5)2YMe(THF) (Y–C(Me) 2.55(2) Å) [152], (C5Me5)2SmMe(THF) (Sm–C(Me) 2.48(1) Å) [153], (C5Me5)2YMe(THF) and (C5Me5)2YbMe(Et2O) [154]. Monomeric organoscandium hydrocarbyls have been prepared in an analogous manner from (C5Me5)2ScCl and organolithium reagents (Eq. (2-59)) [155]. (C5Me5)2ScCl + LiR → (C5Me5)2ScR + LiCl R = Me, Ph, C6H4Me(o, m, p), CH2Ph
(2-59)
In some cases solvent-free hydrocarbyls have been prepared by adding stoichiometric amounts of olefins to the corresponding hydrides (Eq. (2-60)) [155]. 1/n [(C5Me5)2ScH]n + H2C=CHR → (C5Me5)2ScCH2CH2R R = H, Me
(2-60)
2.3
Lanthanide(III) Metallocenes
85
The synthesis of the base-free, highly reactive lanthanide methyls (C5Me5)2ScMe [155] and [(C5Me5)2LnMe]2 (Ln = Y, Lu) was a major achievement in organo-f-element chemistry [2–7, 104, 156]. Base-free [(C5Me5)2YbMe]2 was recently prepared starting from Yb(C5Me5)2 and methylcopper. It forms an asymmetrically µ-Me bridged dimer in the solid state [157]. The monomeric 14-electron species (C5Me5)2ScMe was structurally characterized. Unexpectedly, α-agostic C–H–Sc interaction does not play any significant role in this compound [155]. A dimeric structure was found by X-ray crystallography for [(C5Me5)2LuMe]2. One methyl group acts as a bridge between the two lutetium atoms, while the second methyl ligand is terminal. The methyl complexes are active catalysts for olefin polymerizations, and they activate C–H bonds. Hydrogenolysis of (C5Me5)2ScMe and [(C5Me5)2LnMe]2 (Ln = Y, Lu) yields the corresponding hydrides, [(C5Me5)2Sc(µH)]n and [(C5Me5)2Ln(µ-H)]2 (Ln = Y, Lu) [104, 155]. All three methyl derivatives activate the C–H bonds of a range of hydrocarbons including 13CH4, arenes, styrene, and alkynes [145, 146, 156, 158]. Facile elimination of methane is observed when the σ-methyl complexes are treated with benzene, tetramethylsilane, pyridine or methylenetriphenylphosphorane (Scheme 2-11) [104, 156].
Scheme 2-11. Reactivity of (C5Me5)2LuMe (dimer). C–H activation of 13CH4 by (C5Me5)2LnMe species in cyclohexane solution has been studied by a kinetic analysis. The relative rates of this reaction are in the order Y (250) > Lu (5) > Sc (1). A four-center transition state was formulated as an intermediate. Scheme 2-12 shows the proposed scenario. H–H and C–H activation reactions involving (C5Me5)2LnH and (C5Me5)2LnR derivatives have also been the subject of theoretical studies [159, 160].
86
2 Lanthanocenes
Scheme 2-12. C–H activation of methane by (C5Me5)2MMe (M = Sc, Y, Ln). For (C5Me5)2ScR the relative reactivity toward various types of C–H bonds has been studied in detail (Eq. (2-61)). The reaction rates have been found to depend strongly on the s character of the reacting bonds, i.e. it follows the order R, R' = H > sp C > sp2 C > sp3 C [154, 161].
(2-61)
(C5Me5)2ScMe reacts with substituted olefins (e.g. propylene) at the terminal sp2 C–H bond to afford the corresponding vinylic C–H activation products (Eq. (262)). (C5Me5)2ScMe + H2C=CHR →
(C5Me5)2ScCH=CHR + CH4
(2-62)
Decamethyllanthanocene hydrocarbyls 10 containing the bulky bis(trimethylsilyl)methyl ligand are of great importance as precatalysts in organolanthanide-catalyzed processes. They also appear to be the most suitable precursors for the preparation of the dimeric hydrides [(C5Me5)2Ln(µ-H)]2. Due to the in-
2.3
Lanthanide(III) Metallocenes
87
creased steric bulk of the CH(SiMe3)2 ligand, such compounds can be isolated completely free of coordinating solvents and alkali metal halides (Eq. (2-63)) [122, 162–167]. (C5Me5)2Ln(µ-Cl)2Li(Et2O)2 + LiCH(SiMe3)2
(2-63)
→(C5Me5)2LnCH(SiMe3)2 + 2LiCl + 2Et2O Ln = Y, La, Ce, Nd, Sm, Lu
SiMe 3 Ln CH SiMe 3 10 An interesting synthetic difference was reported for the corresponding organocerium hydrocarbyl (C5Me5)2CeCH(SiMe3)2 [163, 168]. The original preparation called for treatment of the ‘ate’ complex (C5Me5)2Ce(µ-Cl)2Li(THF)2 with LiCH(SiMe3)2. However, the yields were low because the LiCl by-product (two equivalents) readily adds to the formed (C5Me5)2CeCH(SiMe3)2. A more reliable preparation of (C5Me5)2CeCH(SiMe3)2 involves the reaction of sublimed [(C5Me5)2CeCl]n with the organolithium reagent (Eq. (2-64)): 1/n [(C5Me5)2CeCl]n + LiCH(SiMe3)2
(2-64)
→(C5Me5)2CeCH(SiMe3)2 + LiCl An X-ray structure determination of (C5Me5)2YCH(SiMe3)2 revealed α-C–H–Y and γ-Me–Y agostic interactions. Both types of secondary interactions relieve the electronic unsaturation in the formally 14-electron complex [122]. The molecular structures of both (C5Me5)2CeCH(SiMe3)2 and (C5Me5)2NdCH(SiMe3)2 [165] have also been determined by X-ray diffraction. Both complexes exhibit an agostic αC–H–Ln interaction as well as a secondary β-Si–Me–Ln interaction. Bifunctional ligands in which tetramethylcyclopentadienyl groups are connected by Me2Si or Me2Ge [169, 170] units have been demonstrated to be very useful in lanthanocene chemistry. Such compounds, especially the hydrocarbyl and hydride derivatives, show interesting catalytic behavior. They can be regarded as ‘tiedback’ analogs of the (C5Me5)2Ln derivatives. Halides, hydrocarbyls, and hydrides containing these ligands can be synthesized by the methods established in pentamethylcyclopentadienyl chemistry (Scheme 2-13) [171].
88
2 Lanthanocenes
Scheme 2-13. Synthesis of [Me2Si(C5Me4)2]LnCH(SiMe3)2. The ‘tied-back’ organolanthanide hydrocarbyls [Me2Si(C5Me4)2]NdCH(SiMe3)2 [164], [Me2Si(C5Me4)(C5H4)]LuCH(SiMe3)2 [172], [Me2Ge(C5Me4)2]HoCH(SiMe3)2 [169] and [Me2Si(C5Me4){C5H3CH2CH2P(tBu)2}]ScCH(SiMe3)2 [173] as well as the halide precursors [Me2Ge(C5Me4)2]Ln(µ-Cl)2Li(THF)2 (Ln = Sm, Lu) [169] and [Me2Si{C5H2(tBu-4)(SiMe3-2}2]Y(µ-Cl)2Li(THF)2 [174] have been structurally characterized, with all hydrocarbyls exhibiting a characteristic γ-C–H–Ln agostic interaction [175]. Alkylation of the chloro precursors [Me2SiCp″(R*Cp)]Ln(µ-Cl)2Li(Et2O)2 affords the corresponding chiral σ-alkyl complexes (Eq. (2-65)) [86]: [Me2SiCp″(R*Cp)]Ln(µ-Cl)2Li(Et2O)2 + LiCH(SiMe3)2 →
(2-65)
[Me2SiCp″(R*Cp)]LnCH(SiMe3)2 + 2LiCl R* = (+)-neomenthyl, (–)-menthyl, (–)-phenylmenthyl Ln = Y, La, Nd, Sm, Lu
Single crystal structure analyses have been carried out on the complexes (R/S)[Me2SiCp″{(+)-neomenthylCp}]YCH(SiMe3)2, R-[Me2SiCp″{(–)-menthylCp}]SmCH(SiMe3)2, and R-[Me2SiCp″{(–)-menthylCp}]YCH(SiMe3)2. These σ-alkyl complexes are configurationally stable in toluene at 60°C for many hours. Rapid hydrogenolysis of the hydrocarbyl complexes at ambient temperature affords the corresponding hydrides with retention of the configuration [86]. C–H activation takes place upon treatment of the decamethyllanthanocene hydrocarbyls (C5Me5)2LnCH(SiMe3)2 (Ln = La, Ce) with acetonitrile. In this case the binuclear µ-cyanomethyl complexes [(C5Me5)2LnCH2CN]2 have been isolated. The La derivative was structurally characterized (La–C(σ) 2.748(4), La–N 2.537(4) Å) [176]. Monomeric (C5Me5)2ScR complexes also metallate pyridine to give the structurally characterized compound (C5Me5)2Sc(C, N-η2-C5H4N) [155]. An exten-
2.3
Lanthanide(III) Metallocenes
89
sive derivative chemistry based on the decamethylyttrocene hydrocarbyls (C5Me5)2YCH(SiMe3)2 and (C5Me5)2YMe(THF) has been developed by Teuben et al [177, 178]. Typical reactions of (C5Me5)2YCH(SiMe3)2 are summarized in Scheme 2-14. As expected, facile cleavage of the Y–C bond occurs upon treatment of the hydrocarbyl with protic reagents such as alcohols or terminal alkynes [177, 179]. Other characteristic reactions include insertions of CO2 or tBuCN into the yttrium–carbon bond.
Scheme 2-14. Derivative chemistry of (C5Me5)2YCH(SiMe3)2. The lanthanocene hydrocarbyls (C5Me5)2LnCH(SiMe3)2 (Ln = La, Ce) display a differentiated reactivity toward ketones [180]. It was proposed that the differences in reactivity are thermodynamic and not kinetic in origin. Both early lanthanide hydrocarbyls did not react with the sterically hindered di-t-butylketone. Hydrogen transfer and formation of the lanthanide aldolates (C5Me5)2LnOCMe2CH2C(=O)Me (Ln = La, Ce) was observed when the hydrocarbyls were treated with acetone (Eq. (2-66)). In contrast, addition of diethylketone to (C5Me5)2LnCH(SiMe3)2 did not result in C–C coupling but afforded the enolate–ketone adducts (C5Me5)2Ln[OC(Et)=C(H)Me](O=CEt2) [180]. (2-66)
90
2 Lanthanocenes
Interesting bimetallic methyl and ethyl species have been prepared by reacting lanthanocenes with alkyl aluminum or gallium reagents. The bimetallic hydrocarbyls (C5Me5)2Ln(µ-Me)2MMe2 (Ln = Y, Lu; M = Al, Ga) are formed when the dimeric (C5Me5)2LnMe species are reacted with MMe3 [153, 181, 182]. In solution there is an equilibrium between the monomeric complexes and cyclic dimers. The latter contain eight-membered ring systems with methyl bridges between the different metal atoms (Eq. (2-67)). Similar bimetallic alkyls can also be prepared by using organolanthanide(II) precursors. For example, excess Al2Me6 reacts with (C5Me5)2Sm(THF)2 to produce the cyclic dimer (C5Me5)2Sm[(µ-Me)2Al(µMe)2]2Sm(C5Me5)2 [153]. Two derivatives of this type, (C5Me5)2Ln[(µ-Me)2Al(µMe)2]2Ln(C5Me5)2 (Ln = Y, Sm) have been structurally characterized [153, 182].
(2-67)
A mononuclear bimetallic compound with bridging µ-Et ligands can be isolated, when unsolvated Sm(C5Me5)2 is reacted with an excess of AlEt3 (Eq. (2-68)). An X-ray crystal structure determination of (C5Me5)2Sm(µ-Et)2AlEt2 revealed a monomeric complex with highly asymmetrical ethyl bridges between samarium and aluminum (Sm–C(C5Me5) 2.712(2), Sm–C(CH2) 2.662(4) Å) [183]. 3Sm(C5Me5)2 + 4AlEt3 3 →
(C5Me5)2Sm(µ-Et)2AlEt2 + Al
(2-68)
Yet another coordination mode of a µ-Et group was found in the structurally characterized bimetallic ytterbium complex (C5Me5)2Yb(µ-Et)AlEt2(THF) (11) (Eq. (269)) [184]. In this case ytterbium and aluminum are bridged by a single ethyl group but both carbon atoms are coordinated to the ytterbium atom (Yb–C(C5Me5) 2.68(2), Yb–C(1) 2.85(2), Yb–C(2) 2.94(2) Å). (C5Me5)2Yb(THF) + AlEt3 → (C5Me5)2Yb(µ-Et)AlEt2(THF)
CH2 AlEt2(THF) Yb CH3 11
(2-69)
2.3
Lanthanide(III) Metallocenes
91
A major achievement in recent organolanthanide chemistry was certainly the successful synthesis of the first rare earth carbene complexes [185–187]. Treatment of either (C5Me4Et)2Yb(THF) or (C5Me5)2Sm(THF) with stable carbenes of the Arduengo-type (imidazol-2-ylidenes) afforded the mono and bis(carbene) complexes 12 and 13. Crystal structure determinations revealed that the carbene-metal distances in theses derivatives are far longer than in related s-, p-, and d-block metal complexes.
N
R N Yb N R
N Sm N N
R = Me, iPr 12
13
The first lanthanocene butadiene complex was synthesized by reacting (C5Me5)2La(µ-Cl)2K(DME) with ‘magnesium butadiene’, Mg(C4H6)(THF)2. The product, binuclear (C5Me5)2(THF)La(µ-η1,η3-CH2CHCHCH2)La(C5Me5)2 was structurally characterized (La–C(η1) 2.633(4), La–C(η3) 2.73(3) Å) [188]. A mononuclear η2-butadienyl complex, (C5Me5)2Sm[η2-tBuCH=CC(tBu)=CH2] was isolated from a reaction of [(C5Me5)2Sm(µ-H)]2 with the terminal alkyne tBuC≡CH and structurally characterized. The ligand can be formally regarded as a 1,3-di-t-butyl1,3-butadiene metallated in the 2-position [123]. Various other hydrocarbyl complexes containing a (C5Me5)2Sm unit have been prepared from decamethylsamarocene or its THF adduct (C5Me5)2Sm(THF)2. For example, treatment of (C5Me5)2Sm(THF)2 with diphenylmercury (Eq. (2-70)) gave the structurally characterized σ-phenyl complex (Sm–C(σ) 2.511(8) Å). (C5Me5)2SmPh(THF) was also formed in a C–H activation reaction between (C5Me5)2SmMe(THF) and benzene [153]. 2(C5Me5)2Sm(THF)2 + HgPh2
(2-70)
→2(C5Me5)2SmPh(THF) + Hg + 2THF Highly reactive decamethylsamarocene also reacts with a variety of hydrocarbons to give samarium(III) hydrocarbyls. A typical example is the reaction of unsolvated (C5Me5)2Sm with styrene, which affords (µ-η2:η4-CH2CHPh)[Sm(C5Me5)2]2. Simi-
92
2 Lanthanocenes
larly, trans-stilbene yields the binuclear complex (µ-η2:η4PhCHCHPh)[Sm(C5Me5)2]2, while cis-Stilbene is isomerized to the trans form in the presence of (C5Me5)2Sm(THF)2 [189]. Aliphatic olefins generally react with Sm(C5Me5)2 under formation of samarium(III) η3-allyl complexes. One equivalent of olefin is hydrogenated during the course of the reaction (Eq. (2-71)). The same complexes are accessible via addition of olefins to the dimeric hydride [(C5Me5)2Sm(µ-H)]2 [189]. Sm(C5Me5)2 + RCH=CH2 R = Me, Et, CH2Ph
(2-71)
→ (C5Me5)2Sm(η3-CH2CHCHR') + CH3CH2CH2R' R' = H, Me, Ph 1,3-Butadiene is dimerized in the presence of (C5Me5)2Sm to form the bis(η3-allyl) complex (µ-η3:η3-CH2CHCHCH2CH2CHCHCH2)[Sm(C5Me5)2]2. In an extension of this work, a detailed study on the reactivity of (C5Me5)2Sm with polycyclic aromatic hydrocarbons has been carried out. Decamethylsamarocene reacts with hydrocarbons which have reduction potentials more positive than –2.22 V vs. SCE to form a series of binuclear complexes in high yield. The hydrocarbons included in this investigation were anthracene, 9-methylanthracene, pyrene, 2,3-benzanthracene, acenaphthylene, and azulene. In all cases the (C5Me5)2Sm fragments are coordinated to η3-allyl units [190]. Diphenylacetylene was found to add two equivalents of (C5Me5)2Sm(THF)2 to give a black compound formulated as (µ-PhC=CPh)[Sm(C5Me5)2]2 [191]. This material should be regarded as a samarium(III) complex of the dianion of trans-stilbene rather than a samarium(II) adduct of neutral PhC≡CPh. In accordance with this description of the bonding the complex yields pure trans-stilbene upon hydrolysis. (µ-PhC=CPh)[Sm(C5Me5)2]2 is a useful starting material for the preparation of the highly reactive hydride species [(C5Me5)2Sm(µ-H)]2, which is obtained in high yield by simple hydrogenolysis with molecular hydrogen under normal pressure [192]. Structurally different acetylides have been prepared by reacting (C5Me5)2Ln(Et2O) (Ln = Eu, Yb) with phenylacetylene. With (C5Me5)2Eu(Et2O) binuclear [(C5Me5)2Eu(µ-C≡CPh)(THF)2]2 was formed, while the ytterbium(II) complex yielded trinuclear Yb[(µ-C≡CPh)2{Yb(C5Me5)2}2]2 [193]. Monomeric acetylides of the type (C5Me5)2Ln(C≡CR)(THF) (R = tBu, Ph) were prepared by reacting the cationic species [(C5Me5)2Sm(THF)2][BPh4] with KC≡CR or by protonation of (C5Me5)2LnN(SiMe3)2 (Ln = Ce, Nd, Sm) with phenylacetylene [109, 151]. Acetylide bridging between yttrium and lithium was found by X-ray analysis in the organoyttrium acetylide (C5Me5)2Y(µ-C≡CtBu)2Li(THF), which was made straightforwardly from (C5Me5)2YCl(THF) and LiC≡CtBu [194]. Mixed lanthanide potassium acetylides were prepared similarly according to Eq. (2-72) [123].
2.3
Lanthanide(III) Metallocenes
(C5Me5)2Ln(µ-Cl)2K(THF)2 + 2KC≡CPh →
93 (2-72)
1/n [(C5Me5)2Ln(µ-C≡CPh)2K]n + 2KCl Ln = Ce, Nd, Sm
Several other preparative routes leading to (µ-η2:η2-PhC4Ph)[Sm(C5Me5)2]2 via coupling of PhC≡C units have been reported. The µ-η2:η2-diacetylene complex was formed during thermolysis of (C5Me5)2Sm(C≡CPh)(THF) as well as by reacting (C5Me5)2SmCH(SiMe3)2 or [(C5Me5)2Sm(µ-H)]2 with phenylacetylene. Later it has been demonstrated that acetylide carbon–carbon coupling in the coordination sphere of lanthanocenes is a very common reactivity pattern for unsolvated acetylides of the type [(C5Me5)2Ln(µ-C≡CR)]n (Ln = La, Ce, Sm; R = Me, tBu, Ph) [123, 195, 196]. Reactions of Sm(C5Me5)2 with terminal alkynes in the absence of THF may even take a different course. The novel trienediyl complexes (µ-η2:η2RCH2CH2C=C=C=CCH2CH2R)[Sm(C5Me5)2]2 (R = iPr, Ph) have been isolated from reactions of Sm(C5Me5)2 with HC≡CCH2CH2R (R = iPr, Ph) [123]. Phosphaalkynes have only rarely been employed in organo-f-element chemistry. An unprecedented reductive dimerization of a stable phosphaalkyne occurred when (C5Me5)2Sm(THF)2 was treated with one equivalent of tBuC≡P (Eq. (2-73)). The resulting binuclear samarium complex was structurally characterized (Sm–C 2.557(6), Sm–P 2.949(2) Å). The structural data suggest, that the new ligand is the dianion of a 2,3-diphosphabutadiene. Formally the reductive dimerization of a phosphaalkyne closely resembles the above-mentioned dimerization of acetylide units in the coordination sphere of lanthanocenes. 2(C5Me5)2Sm(THF)2 + 2tBuC≡P
(2-73)
→(µ-tBuC=PP=CtBu)[Sm(C5Me5)2]2 + 4THF Anionic chelating phosphoylide ligands have frequently been proven to be versatile ligands in lanthanocene chemistry. The combination of these ligands with decamethyllanthanocene fragments results in thermally highly stable rare earth hydrocarbyl complexes. The first complex of this type was synthesized according to Eq. (2-74) [2–7]. (C5Me5)2Lu(µ-Cl)2Na(Et2O)2 + Li(CH2)2PMe2
(2-74)
→(C5Me5)2Lu[(CH2)2PMe2] + LiCl + NaCl LiCl adducts of (C5Me5)2Ln[(CH2)2PRR'] (Ln = Nd, Sm; R = Me, R' = Ph; R = R' = Me, tBu, Ph) have been prepared analogously from the corresponding (C5Me5)2Ln(µ-Cl)2Li(Et2O)2 precursors. The molecular structure of (C5Me5)2Lu[(CH2)2PMe2] was established by X-ray diffraction. It shows a typical
94
2 Lanthanocenes
bent lanthanocene(III) derivative with a pseudo-tetrahedral coordination around the central Lu atom [198]. Even the unusual (trimethylsilyl)diazomethyl substituent has been attached to decamethyllanthanocenes (Eq. (2-75)). The products are thermally quite sensitive [199]. (C5Me5)2LnCl(THF) + LiC(=N2)SiMe3
(2-75)
→ (C5Me5)2Ln[C(=N2)SiMe3](THF) + LiCl Ln = Y, Yb, Lu Recently the first (C5Me5)2Ln derivatives containing Ln–Si bonds have been prepared and characterized. Treatment of (C5Me5)2LnCH(SiMe3)2 with H2Si(SiMe3)2 in the absence of solvent afforded the neutral silyl complexes (C5Me5)2LnSiH(SiMe3)2 (Ln = Y, Sm, Nd) (Eq. (2-76)). An X-ray crystal structure analysis of the Sm derivative revealed the presence of a dimer stabilized by intermolecular Sm•••Me–Si contacts [200, 201]. The complexes (C5Me5)2LnSiH2Ph (Ln = Y, Nd) were postulated to be intermediates in the organolanthanide catalyzed dehydrogenation of PhSiH3 and the hydrosilylation of olefins [202, 203]. (C5Me5)2LnCH(SiMe3)2 + H2Si(SiMe3)2
(2-76)
→ (C5Me5)2LnSiH(SiMe3)2 + CH2(SiMe3)2 2.3.5 Lanthanide(III) Metallocenes Containing Ln–Hydrogen Bonds Lanthanocene hydrides play an important role in organolanthanide chemistry as they are highly reactive and offer a great potential in homogeneous catalysis. Theoretical studies on the bonding in hydride-bridged (C5H5)2Ln complexes have been carried out by Hoffmann et al. [204]. Subsequently several of these hydrides have been synthesized and structurally characterized. Structural types include dimeric and trimeric complexes of the types [(C5H5)2Ln(µ-H)]2 and [(C5H5)2Ln(µ-H)]3, the trimer anions [(µ3-H){(C5H5)2Ln(µ-H)}3]– as well as heterobimetallic hydrides, including tetrahydroborates and tetrahydroaluminates [2–7]. Two main synthetic routes leading to lanthanocene hydrides have been developed. One of them is hydrogenolysis of suitable (C5H5)2Ln hydrocarbyl precursors [2–7, 133] and the other one involves metathetical reactions of the corresponding halide precursors with metal hydrides, including NaBH4 and LiAlH4 [205]. THF solvates of the dimeric hydrides are obtained when the hydrogenolysis of lanthanocene hydrocarbyls is carried out in the presence of THF [2–7, 133, 206, 207].
2.3
Lanthanide(III) Metallocenes
2(C5H5)2LnR(THF) + 2H2
95 (2-77)
toluene →[(C5H5)2Ln(µ-H)(THF)]2 + 2RH (2-77) R = Me, tBu, Ph; Ln = Y, Er, Lu R = tBu, CH2tBu, CH2SiMe3; Ln = Lu The molecular and crystal structures of [(MeOCH2CH2C5H4)2Y(µ-H)]2 [208], [(MeC5H4)2Ln(µ-H)(THF)]2 (Ln = Y, Er) [133, 209], and [(C5H5)2Lu(µ-H)(THF)]2 [210, 211] have been determined. The complexes are dimerized via two hydride ligands and the coordination sphere around each lanthanide atom is completed by an additional THF ligand of the pending methoxy function. Unsolvated dimeric lanthanocene hydrides have been prepared with the use of bulky cyclopentadienyl ligands such as t-butylcyclopentadienyl or 1,3-di-t-butylcyclopentadienyl. Structurally characterized examples are [(tBuC5H4)2Lu(µ-H)]2 [212] and [(tBu2C5H3)2Ln(µH)]2 (Ln = Ce, Sm) [212, 213]. The hydride-bridged dimers have been shown to be highly reactive. For example, [(MeC5H4)2Er(µ-H)(THF)]2 reacts with tBuNC to give the insertion product [(MeC5H4)2Er(µ-HC=NtBu)]2 [191]. Similar treatment of [(C5H5)2Ln(µ-H)]2 with tBuNC afforded the dimeric insertion products [(C H ) Ln(µ-HC=NtBu)] (Ln = Y, 5 5 2 2 Er) with bridging formimidoyl ligands [191, 214]. [(C5H5)2Y(µ-H)(THF)]2 and [(MeC5H4)2Y(µ-H)(THF)]2 have also been found to react with alkenes, alkynes, allenes, nitriles, and pyridine to afford the new compounds (RC5H4)2Y(Et)(THF), (RC5H4)2Y(nPr)(THF), (RC5H4)2Y(η3-allyl)(THF), [(RC5H4)2Y(µ-C≡CtBu)]2, (RC5H4)2Y[C(R')=CHR'](THF) (R' = Et, Ph), [(RC5H4)2Y(µ-NCHtBu)]2, [(RC5H4)2Y(µ-H)(NC5H5)]2, and (RC5H4)2Y(NC5H6)(NC5H5) (R = H, Me) [206]. The dimeric molecular structure of [(RC5H4)2Y(µ-NCHtBu)]2 was determined by X-ray methods. Scheme 2-15 summarizes the reactivity of the bis(cyclopentadienyl)yttrium hydrides [(RC5H4)2Y(µ-H)(THF)]2 (R = H, Me) towards various substrates.
96
2 Lanthanocenes
Scheme 2-15. Reactivity of the bis(cyclopentadienyl)yttrium hydride complexes [(RC5H4)2Y(µ-H)(THF)]2 (R = H, Me).
Several borohydride complexes of the type (C5H5)2Ln(BH4) have been prepared and studied by IR spectroscopy. These compounds are accessible by different synthetic routes (Eqs. (2-78)–(2-80)) [2–7, 69, 88, 215–217].
(C5H5)2LnCl(THF) + NaBH4 THF → (C5H5)2Ln(BH4)(THF) + NaCl Ln = Sm, Er, Yb, Lu
(2-78)
2.3
Lanthanide(III) Metallocenes
[(Me3Si)2C5H3]2LnCl(THF) + NaBH4
97 (2-79)
THF →[(Me3Si)2C5H3]2Ln(BH4)(THF) + NaCl Ln = Sc, Y, La, Pr, Nd, Sm, Yb Sc(BH4)3(THF)2 + 2 Na(RC5H4)
(2-80)
THF →(RC5H4)2Sc(BH4) + 2 NaBH4 R = H, Me In the case of [(Me3Si)2C5H3]2Ln(BH4)(THF) (Ln = Y, Yb) the IR data indicated a bidentate chelating coordination of the borohydride ligand. This was later confirmed by an X-ray structural analysis of unsolvated [(Me3Si)2C5H3]2Sc(BH4) [216]. According to the IR spectra, the coordination of the BH4– ligand is tridentate in all other complexes of this type. Crystal structure determinations have also been reported for (MeOCH2CH2C5H4)2Ln(BH4) (Ln = Y, Yb) [69]. Lanthanocene hydrides of the type (C5Me5)2LnH are important as rare earthbased homogeneous catalysts. Thus this group of compounds has been thoroughly investigated and elaborate methods for their synthesis have been designed [104, 122, 162, 163, 179, 192, 218]. A monomeric (C5Me5)2LnH fragment appears to be coordinatively unsaturated. As a consequence, these hydride species are usually isolated as dimers or solvated species. The decamethyllanthanocene hydrides 14 are generally prepared by hydrogenolysis of the corresponding hydrocarbyls. Unsolvated hydrocarbyls containing the bulky –CH(SiMe3)2 ligand have been shown to be especially useful for the preparation of hydrides. A typical synthesis starting from the THF adduct of yttrium trichloride is outlined below (Eqs. (2-81)–(2-83)) [219]. THF YCl3(THF)3 + 2Na(C5Me5) → (C5Me5)2YCl(THF) + NaCl
(2-81)
(C5Me5)2YCl(THF) + LiCH(SiMe3)2
(2-82)
Et2O →
(C5Me5)2YCH(SiMe3)2 + LiCl + THF
2(C5Me5)2YCH(SiMe3)2 + H2 → [(C5Me5)2Y(µ-H)]2 + 2CH2(SiMe3)2
(2-83)
98
2 Lanthanocenes
H Ln
Ln H
14 The dimeric lanthanocene hydrides [(C5Me5)2Ln(µ-H)]2 (14, Ln = Ce, Nd, Sm) have been prepared analogously by hydrogenation of (C5Me5)2LnCH(SiMe3)2 [151, 162, 165]. Unsolvated [(C5Me5)2Lu(µ-H)]2 was prepared by hydrogenolysis of base-free [(C5Me5)2LuMe]2 in hexane solution. For this complex 1H NMR studies revealed an equilibrium at room temperature between the dimeric form and the monomer (C5Me5)2LuH (Scheme 2-15) [2–7, 104].
Scheme 2-16. Formation and monomer–dimer equilibrium of [(C5Me5)2Lu(µ-H)]2. The hydrides [(C5Me5)2Ln(µ-H)]2 (Ln = Y, Lu) are dimers in which the lanthanide atoms are symmetrically bridged by µ-H ligands [218]. [(C5Me5)2Sc(µ-H)]n is quite unstable in solution unless a hydrogen atmosphere is provided [(C5Me5)2Lu(µ-H)]2 decomposes in diethyl ether solution to give (C5Me5)2LuOEt and ethane. Generally more stable are the THF solvates (C5Me5)2LnH(THF) (Sc, Y, Lu). Reactions leading to hydride species have also been designed for lanthanide ansa-metallocenes. Treatment of the appropriate halide precursors with alkyllithium or aryllithium afforded a series of new lanthanide hydrocarbyls containing the ring-bridged cyclopentadienyl ligands. Subsequent hydrogenolysis produced the dimeric hydrides (Scheme 2-17) [82, 83, 220]:
2.3
Lanthanide(III) Metallocenes
99
Scheme 2-17. Formation of [{C5H4(CH2)3C5H4}Ln(µ-H)(THF)]2. Similar reaction sequences have been reported for the corresponding 1,1'-pentamethylene-bis(cyclopentadienyl) complexes, e.g. [C5H4(CH2)5C5H4]LnCl(THF) (Ln = Y, Sm, Gd, Dy, Er, Lu) [221]. Hydride complexes of the type [(C5Me5)2Ln(µ-H)]2 are exceedingly reactive. A most astonishing reaction is the formation of (µOSiMe2OSiMe2O)[(C5Me5)2Sm(THF)]2. This compound is formed when a THF solution of [(C5Me5)2Sm(µ-H)]2 comes into contact with high-vacuum grease! An alternative preparation involves treatment of the dimeric hydride with hexamethylcyclotrisiloxane [222]. The dimeric hydrides have also been found to polymerize olefins and to activate small molecules such as carbon monoxide as well as sp2 and sp3 C–H bonds [2–7, 104, 156]. [(C5Me5)2Sc(µ-H)]n adds ethylene or propylene to afford the alkyls (C5Me5)2ScEt and (C5Me5)2Sc(nPr), respectively [155]. Treatment of [(C5Me5)2Sc(µ-H)]n with allene produced the addition product (C5Me5)2Sc(η3C3H5), while pyridine was metallated to give (C5Me5)2Sc(η2-C5H4N). The methyl groups of the C5Me5 ligands in [(C5Me5)2Sc(µ-D)]n have been reported to undergo intramolecular H/D exchange [155]. Treatment of other hydrides of the type [(C5Me5)2Ln(µ-H)]2 (Ln = La, Nd, Sm, Lu) with α-olefins normally results in the formation of η3-allyl complexes. As shown in Scheme 2-18 their formation involves insertion of the olefin into the Ln–H bond followed by hydrogen abstraction.
100
2
Lanthanocenes
Scheme 2-18. Formation of η3-allyl complexes from organolanthanide hydrides and α-olefins. Various C–H activation reactions have also been reported for [(C5Me5)2Y(µH)]2. Metallation of benzene and toluene leads to the σ-aryl compounds (C5Me5)2YPh and (C5Me5)2YCH2Ph, respectively. With substituted aromatic molecules RX (X = OMe, SMe, NMe2, CH2NMe2, PMe2, PPh2=CH2, F, Cl, Br), orthometallation is the dominant reaction. The product resulting from ortho-metallation of Ph3P=CH2, (C5Me5)2Y(o-C6H4PPh2CH2), was structurally characterized. An especially remarkable case of C–H activation is the reaction of (C5Me5)2LuH with tetramethylsilane, which yields (C5Me5)2LuCH2SiMe3 [104]: (C5Me5)2LuH + SiMe4 → (C5Me5)2LuCH2SiMe3 + H2
(2-84)
2.3 Lanthanide(III) Metallocenes
101
Scheme 2-19. C–H activation reactions of (C5Me5)2LuH (dimer). The C–H bonds in benzene are also readily attacked by (C5Me5)2LuH (dimer). In the first step the σ-phenyl complex (C5Me5)2LuPh and hydrogen are formed. (C5Me5)2LuPh can further react with (C5Me5)2LuH yielding a phenylene bridged binuclear complex (Scheme 2-19) [13, 104]: Activation of carbon monoxide by [(C5Me5)2Sm(µ-H)]2 was investigated by Evans et al. [223]. [(C5Me5)2Sm(µ-H)]2 reacts with CO in aromatic hydrocarbon solvents to give cis- and trans-(µ-OCH=CHO)[(C5Me5)2Sm(OPPh3)]2 after recrystallization in the presence of triphenylphosphine oxide as auxiliary ligand. The cis-
102
2
Lanthanocenes
isomer was found to isomerize at room temperature to give the trans-complex. The cis-isomer was structurally characterized by X-ray diffraction. Facile addition of Ln–H bonds has also been reported for metal-coordinated carbon monoxide. This reaction was investigated for (C5Me5)2ScH [161]: (C5Me5)2ScH + (C5H5)M(CO)2
(2−85)
→ (C5Me5)2ScOC(H)=M(C5H5)(CO) M = Co, Rh Novel (C5Me5)2Ln borohydride derivatives have been prepared by addition of [Mes2BH]2 to dimeric decamethyllanthanocene hydrides (Eq. (2-86)). Thermolyses of these complexes did not result in isolatable boryl complexes of the type (C5Me5)2LnBMes2 [200]. [(C5Me5)2Ln(µ-H)]2 + [Mes2BH]2 → 2(C5Me5)2Ln(µ-H)2BMes2
(2-86)
The monomeric ‘tied-back’ scandocene hydride [Me2Si(C5H3tBu)2]ScH was synthesized in order to study its reactions with unsaturated hydrocarbons (Scheme 220). This hydride reacts with butadiene and several nonconjugated dienes to give η3-allyl complexes of the type [Me2Si(C5H3tBu)2]Sc(η3-allyl). The same complexes were obtained from reactions of the hydride with methylenecyclopropane or substituted conjugated dienes [224].
2.4 Conclusion and Perspective During the past 15 years the knowledge about lanthanocenes and their chemistry has increased enormously. A large body of synthetic and structural data has been accumulated and various members of this large class of rare earth complexes are now readily available. A large number of unusual molecular structures have been discovered. Especially impressive is the very high reactivity of certain lanthanide(II) metallocenes and lanthanide(III) derivatives containing σ-alkyl and hydride ligands. This includes facile C–H activation reactions with substrates such as benzene, tetramethylsilane, or even methane, which are usually regarded as inert under the given conditions. A major contribution to the rapid development of lanthanocene chemistry can be attributed to the favorable properties of the pentamethylcyclopentadienyl ligand. The coordination environment in a Ln(C5Me5)2 unit allows the preparation of thermally stable though highly reactive organolanthanide hydrocarbyls and hydrides such as (C5Me5)2LnCH(SiMe3)2 and [(C5Me5)2Ln(µH)]2. More recent work has also been focused on the preparation of ansa-lanthanocenes as well as lanthanocenes containing chiral cyclopentadienyl ligands. The
2.4 Conclusion und Perspective
103
Scheme 2-20. Formation of scandium η3-allyl complexes. driving force for much of this work is the expectation that these lanthanocenes may exhibit useful catalytic activities. In fact, the catalytic activity of lanthanocene hydrocarbyls of the type (C5Me5)2LnCH(SiMe3)2 e.g. in olefin hydrogenation reactions is exceedingly high and surpasses by far the activity of commonly used dtransition metal hydrogenation catalysts. However, the high air-sensitivity of such lanthanocene hydrocarbyls has thus far prevented them from coming into practical,
104
2
Lanthanocenes
i.e. industrial use [225]. Similar considerations can be made for the use of lanthanide organometallics in materials science [226, 227]. Thus future research in this area is expected to continue producing unusual molecular structures and unexpected reaction patterns. However, it will be the great challenge in the up-coming years to put lanthanocenes into practical uses, especially in the area of industrial catalysis. Great efforts have to be made to overcome the practical problems associated with the typical high moisture-sensitivity of organolanthanide complexes. This will be mainly a question of sophisticated ligand design. Imagining a water-soluble organolanthanide catalyst is certainly asking too much at the present state, but it should be possible to design highly reactive lanthanocene catalysts which can be handled under the conditions of industrial processes.
2.5 Acknowledgments Our own contributions to organolanthanide chemistry at Göttingen and Magdeburg have been generously supported by the Deutsche Forschungsgemeinschaft, the Fonds der Chemischen Industrie and the Otto-von-Guericke-Universität Magdeburg. I am also indebted to Professor Herbert W. Roesky for continuing support.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17]
G. Wilkinson and J. M. Birmingham, J. Am. Chem. Soc., 1954, 76, 6210. H. Schumann, Angew. Chem., 1984, 96, 475; Angew. Chem. Int. Ed. Engl., 1984, 23, 474. W. J. Evans, Polyhedron, 1987, 6, 803. R. D. Köhn, G. Kociok-Köhn, H. Schumann, Scandium, Yttrium & the Lanthanides: Organometallic Chemistry in: Encyclopedia of Inorganic Cemistry (Ed. R. B. King), Wiley, 1994. C. J. Schaverien, Adv. Organomet. Chem., 1994, 36, 283. H. Schumann, J.A. Meese-Marktscheffel and L. Esser, Chem. Rev., 1995, 95, 865. F. T. Edelmann in: Comprehensive Organometallic Chemistry II (Eds. E. W. Abel, F. G. A. Stone, G. Wilkinson), Pergamon Press, 1995. F. T. Edelmann, Top. Curr. Chem., 1996, 179, 247. E. O. Fischer and H. Fischer, Angew. Chem., 1964, 76, 52; Angew. Chem. Int. Ed. Engl., 1964, 3, 132. G. B. Deacon, P. I. MacKinnon, T. W. Hambley and J. C. Taylor, J. Organomet. Chem., 1983, 259, 91. A. Hammel, W. Schwarz and J. Weidlein, J. Organomet. Chem., 1989, 378, 347. J. Jin, S. Jin and W. Chen, J. Organomet. Chem., 1991, 412, 71. X. Jusong, G. Wei, Z. Jin and W. Chen, Zhongguo Xitu Xuebao, 1992, 3, 203. T. Jiang, Q. Shen, Y. Lin and S. Jin, J. Organomet. Chem., 1993, 450, 121. G. B. Deacon, G. N. Pain and T. D. Tuong, Polyhedron, 1985, 4, 1149. G. Z. Suleimanov and I. P. Beletskaya, Metalloorg. Khim., 1989, 2, 704. J. L. Namy, P. Girard, H. B. Kagan and P. Caro, Nouv. J. Chem., 1981, 5, 479.
References [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58]
105
G. Z. Suleimanov, T. K. Kurbanov, Y. A. Nuriev, L. F. Rybakova and I. P. Beletskaya, Dokl. Akad. Nauk SSSR, 1982, 265, 896; Proc. Acad. Sci. USSR, 1982, 265, 254. G. Z. Suleimanov, L. F. Rybakova, Y. A. Nuriev, T. K. Kurbanov and I. P. Beletskaya, J. Organomet. Chem., 1982, 235, C19. C. Apostolidis, G.B. Deacon, E. Dornberger, F.T. Edelmann, B. Kanellakopulos, P. MacKinnon and D. Stalke, J. Chem. Soc., Chem. Commun., 1997, 1047. V. D. Makhaev, Y. B. Zvedov, N. G. Chernorukov and V. I. Berestenko, Vysokchist. Veshchestva, 1990, 210. A. L. Wayda, J. Organomet. Chem., 1989, 361, 73. Q. Shen, D. Zheng, L. Lin and Y. Lin, J. Organomet. Chem., 1990, 391, 307. V. K. Bel’skii, Y. K. Gun’ko, B. M. Bulychev, A. I. Sizov and G. L. Soloveichik, J. Organomet. Chem., 1990, 390, 35. M. F. Lappert, P. I. W. Yarrow, J. L. Atwood and R. Shakir, J. Chem. Soc., Chem. Commun., 1980, 987. W. J. Evans, R. A. Keyer and J. W. Ziller, J. Organomet. Chem., 1990, 394, 87. R. D. Rogers, J. Organomet. Chem., 1996, 512, 97. P. B. Hitchcock, J. A. K. Howard, M. F. Lappert and S. Prashar, J. Organomet. Chem., 1992, 437, 177. D. Deng, C. Qian, F. Song, Z. Wang, G. Wu and P. Zheng, J. Organomet. Chem., 1992, 443, 79. C. Qian, C. Ye, H. Lu, Y. Li, J. Zhou and Y. Ge, J. Organomet. Chem., 1983, 247, 161. C. Qian and D. Zhu, J. Organomet. Chem., 1993, 445, 75. C. Qian, C. Zhu, Y. Lin and Y. Xing, J. Organomet. Chem., 1996, 507, 41. P. Jutzi, J. Dahlhaus and M. O. Kristen, J. Organomet. Chem., 1993, 450, C1. P. Jutzi and J. Dahlhaus, Phosphorus, Sulfur, and Silicon, 1994, 87, 73. G. B. Deacon, C. M. Forsyth, W. C. Patalinghug, A. H. White, A. Dietrich and H. Schumann, Aust. J. Chem., 1992, 45, 567. G. B. Deacon, G. D. Fallon, P. I. MacKinnon, R. H. Newnham, G. N. Pain, T. D. Tuong and D. L. Wilkinson, J. Organomet. Chem., 1986, 277, C21 G. B. Deacon, G. D. Fallon and C. M. Forsyth, J. Organomet. Chem., 1993, 462, 183. W. J. Evans, Inorg. Chim. Acta, 1987, 139, 169. W. J. Evans, J. W. Grate, H. W. Choi, I. Bloom, W. E. Hunter and J. L. Atwood, J. Am. Chem. Soc., 1985, 107, 941. W. J. Evans, L. A. Hughes and T. P. Hanusa, Organometallics, 1986, 5, 1285. P. L. Watson, J. Chem. Soc., Chem. Commun., 1980, 652. W. J. Evans, L. A. Hughes and T. P. Hanusa, J. Am. Chem. Soc., 1984, 106, 4270. R. A. Andersen, J. M. Boncella, C. J. Burns, J. C. Green, D. Hohl and N. Rösch, J. Chem. Soc., Chem. Commun., 1986, 405. R. A. Williams, T. P. Hanusa and J. C. Huffman, Organometallics, 1990, 9, 1128. R. A. Williams, T. P. Hanusa and J. C. Huffman, J. Chem. Soc., Chem. Commun., 1988, 1045. T. K. Hollis, J. K. Burdett and B. Bosnich, Organometallics, 1993, 12, 3385. M. Kaupp, P. R. von Schleyer, M. Dolg and H. Stoll, J. Am. Chem. Soc., 1992, 114, 8202. C. J. Burns and R. A. Andersen, J. Am. Chem. Soc., 1987, 109, 941. C. J. Burns and R. A. Andersen, J. Am. Chem. Soc., 1987, 109, 915. C. J. Burns and R. A. Andersen, J. Am. Chem. Soc., 1987, 109, 5853. S. J. Swamy, J. Loebel and H. Schumann, J. Organomet. Chem., 1989, 379, 51. A. Recknagel and F. T. Edelmann, Angew. Chem., 1991, 103, 720; Angew. Chem., Int. Ed. Engl., 1991, 30, 693. P. L. Watson, T. H. Tulip and I. Williams, Organometallics, 1990, 9, 1999. C. Qian, D. Deng, Z. Zhang and C. Ni, Youji Huaxue, 1985, 403. C. Ni, Z. Zhang, D. Deng and C. Qian, J. Organomet. Chem., 1986, 306, 209. Z. Lin, Y. Liu and W. Chen, Sci. Sin., Ser. B, 1988, 30, 1136. H. Yasuda, H. Yamamoto, K. Yokota and A. Nakamura, Chem. Lett., 1989, 1309. J. Jizhu, Z. Jin and W. Chen, Jiegou Huaxue, 1992, 11, 204.
106 [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97]
2
Lanthanocenes
W. J. Evans, R. A. Keyer and J. W. Ziller, J. Organomet. Chem., 1993, 450, 115. E. C. Baker, L. D. Brown and K. N. Raymond, Inorg. Chem., 1975, 14, 1376. V. K. Bel’skii, S. Y. Knyazhanskii, Y. K. Gun’ko, B. M. Bulychev and G. L. Soloveichik, Metalloorg. Khim., 1991, 4, 1135. S. Song, Q. Shen, S. Jin, J. Guan and Y. Lin, Polyhedron, 1992, 12, 2857. V. K. Bel’skii, S. Y. Knyazhanskii, B. M. Bulychev and G. L. Soloveichik, Metalloorg. Khim., 1989, 2, 567. A. Recknagel, F. Knösel, H. Gornitzka, M. Noltemeyer and F. T. Edelmann, J. Organomet. Chem., 1991, 417, 363. J. L. Atwood, W. E. Hunter, R. D. Rogers, J. Holton, J. McMeeking, R. Pearce and M. F. Lappert, J. Chem. Soc., Chem. Commun., 1978, 140. P. L. Watson, J. F. Whitney and R. L. Harlow, Inorg. Chem., 1981, 20, 3271. W. A. Herrmann, R. Anwander, H. Riepl, W. Scherer and C. R. Whitaker, Organometallics, 1993, 12, 4342. D. Deng, C. Qian, G. Wu and P. Zheng, J. Chem. Soc., Chem. Commun., 1990, 880. D. Deng, X. Zheng, C. Qian, S. Jin and Y. Lin, Huaxue Xuebao, 1992, 10, 1024. D. Deng, B. Li and C. Qian, Polyhedron, 1990, 9, 1435. C. Qian, B. Wang, D. Deng, J. Hu, J. Chen, G. Wu and P. Zheng, Inorg. Chem., 1994, 33, 3382. J. Gräper, R. D. Fischer and G. Paolucci, J. Organomet. Chem., 1994, 471, 87. C. Qian and D. Zhu, J. Organomet. Chem., 1993, 445, 79. G. Paolucci, R. D’Ippolito, C. Ye, C. Qian, J. Gräper and R. D. Fischer, J. Organomet. Chem., 1994, 471, 97. W. A. Herrmann, R. Anwander, F. C. Munck and W. Scherer, Chem. Ber., 1993, 126, 319. C. J. Burns and R. A. Andersen, J. Chem. Soc., Chem. Commun., 1989, 136. R. G. Finke, S. R. Keenan, D. A. Schiraldi and P. L. Watson, Organometallics, 1987, 6, 1356. R. G. Finke, S. R. Keenan and P. L. Watson, Organometallics, 1989, 8, 263. W. J. Evans, J. M. Olofson, H. Zhang and J. L. Atwood, Organometallics, 1988, 7, 629. H. Schumann, I. Albrecht, J. Loebel, E. Hahn, M. B. Hossain and D. van der Helm, Organometallics, 1986, 5, 1296. C. Qian, C. Ye, H. Lu, Y. Li and Y. Huang, J. Organomet. Chem., 1984, 263, 333. C. Qian, C. Ye and Y. Li, J. Organomet. Chem., 1986, 302, 171. C. Qian, C. Ye and Y. Li, Youji Huaxue, 1986, 130. V. P. Conticello, L. Brard, M. A. Giardello, Y. Tsuji, M. Sabat, C. L. Stern and T. J. Marks, J. Am. Chem. Soc., 1992, 114, 2761. M. R. Gagné, L. Brard, V. P. Conticello, M. A. Giardello, C. L. Stern and T. J. Marks, Organometallics, 1992, 11, 2003. M. A. Giardello, V. P. Conticello, L. Brard, M. Sabat, A. L. Rheingold, C. L. Stern and T. J. Marks, J. Am. Chem. Soc. 1994, 116, 10212. M. Adam, G. Massarweh and R. D. Fischer, J. Organomet. Chem., 1991, 405, C33. H. Schumann, W. Genthe, N. Bruncks and J. Pickardt, Organometallics, 1982, 1, 1194. G. B. Deacon, S. Nickel and E. R. T. Tiekink, J. Organomet. Chem., 1991, 409, C1. Z. Wu, Z. Xu, X. You, X. Zhou and Z. Jin, Polyhedron, 1992, 12, 2673. Z. Wu, Z. Xu, X. You, X. Zhou, Y. Xing and Z. Jin, J. Chem. Soc., Chem. Commun., 1993, 1494. J. Stehr and R. D. Fischer, J. Organomet. Chem., 1993, 459, 79. G. Massarweh and R. D. Fischer, J. Organomet. Chem., 1993, 444, 67. P. Poremba, M. Noltemeyer, H.-G. Schmidt and F.T. Edelmann, J. Organomet. Chem. 1995, 501, 315. H. Schumann, J. A. Meese-Marktscheffel and A. Dietrich, J. Organomet. Chem., 1989, 377, C5. J. Stehr and R. D. Fischer, J. Organomet. Chem., 1992, 430, C1. H. Schumann, J. A. Meese-Marktscheffel, A. Dietrich and F. H. Görlitz, J. Organomet. Chem., 1992, 430, 299.
References [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134]
107
W. J. Evans, D. K. Drummond, L. A. Hughes, H. Zhang and J. L. Atwood, Polyhedron, 1988, 7, 1693. W. J. Evans, J. W. Grate, I. Bloom, W. E. Hunter and J. L. Atwood, J. Am. Chem. Soc., 1985, 107, 405. W. J. Evans, T. P. Hanusa and K. R. Levan, Inorg. Chim. Acta, 1984, 110, 191. W. J. Evans, L. A. Hughes, D. K. Drummond, H. Zhang and J. L. Atwood, J. Am. Chem. Soc., 1986, 108, 1722. W. J. Evans, J. W. Grate, L. A. Hughes, H. Zhang and J. L. Atwood, J. Am. Chem. Soc., 1985, 107, 3728. T. D. Tilley, R. A. Andersen, A. Zalkin and D. H. Templeton, Inorg. Chem., 1982, 21, 2644. P. L. Watson, J. Chem. Soc., Chem. Commun., 1983, 276. H. Schumann, I. Albrecht, M. Gallagher, E. Hahn, C. Muchmore and J. Pickardt, J. Organomet. Chem., 1988, 349, 103. A. Recknagel, M. Noltemeyer, D. Stalke, U. Pieper, H.-G. Schmidt and F. T. Edelmann, J. Organomet. Chem., 1991, 411, 347. D. J. Berg, C. J. Burns, R. A. Andersen and A. Zalkin, Organometallics, 1989, 8, 1865. D. J. Berg, R. A. Andersen and A. Zalkin, Organometallics, 1988, 7, 1858. A. Zalkin and D. J. Berg, Acta Cryst., 1988, C44, 1488. W. E. Piers, J. Chem. Soc., Chem. Commun., 1994, 309. Y. Yu, S. Wang, H. Ma and Z. Ye, Polyhedron, 1992, 12, 265. H. Schumann, P. R. Lee and A. Dietrich, Chem. Ber., 1990, 123, 1331. Z. Wu, X. Zhou, W. Zhang, Z. Xu, X. You and X. Huang, J. Chem. Soc., Chem. Commun., 1994, 813. H. Schumann, I. Albrecht, M. Gallagher, E. Hahn, C. Janiak, C. Kolax, J. Loebel, S. Nickel and E. Palamidis, Polyhedron, 1988, 7, 2307. H. Schumann, E. Palamidis, G. Schmid and R. Boese, Angew. Chem., 1986, 98, 726; Angew. Chem., Int. Ed. Engl., 1986, 25, 718. H. Schumann, E. Palamidis, J. Loebel and J. Pickardt, Organometallics, 1988, 7, 1008. W. J. Evans, I. Bloom, W. E. Hunter and J. L. Atwood, Organometallics, 1983, 2, 709. H. Schumann, E. Palamidis and J. Loebel, J. Organomet. Chem., 1990, 384, C49. W. J. Evans, D. K. Drummond, S. G. Bott and J. L. Atwood, Organometallics, 1986, 5, 2389. W. J. Evans, J. W. Grate, K. R. Levan, I. Bloom, T. T. Peterson, R. J. Doedens, H. Zhang and J. L. Atwood, Inorg. Chem., 1986, 25, 3614. W. J. Evans, D. K. Drummond, L. R. Chamberlain, R. J. Doedens, S. G. Bott, H. Zhang and J. L. Atwood, J. Am. Chem. Soc., 1988, 110, 4983. K. H. den Haan, J. L. De Boer, J. H. Teuben, A. L. Spek, B. Kojic-Prodic, G. R. Hays and R. Huis, Organometallics, 1986, 5, 1726. W. J. Evans, R. A. Keyer and J. W. Ziller, Organometallics, 1993, 12, 2618. W. J. Evans and D. K. Drummond, J. Am. Chem. Soc., 1989, 111, 3329. A. Recknagel, M. Noltemeyer and F. T. Edelmann, J. Organomet. Chem., 1991, 410, 53. A. Scholz, K.-H. Thiele, J. Scholz and R. Weimann, J. Organomet. Chem., 1995, 501, 195. J. Scholz, A. Scholz, R. Weimann, C. Janiak and H. Schumann, Angew. Chem., 1994, 106, 1220; Angew. Chem., Int. Ed. Engl., 1994, 33, 1171. W. J. Evans, T. A. Ulibarri and J. W. Ziller, J. Am. Chem. Soc., 1988, 110, 6877. W. J. Evans, S. L. Gonzales and J. W. Ziller, J. Chem. Soc., Chem. Commun., 1992, 1138. W. J. Evans, S. L. Gonzales and J. W. Ziller, J. Am. Chem. Soc., 1991, 113, 9880. W. J. Evans, J. T. Leman, J. W. Ziller and S. I. Khan, Inorg. Chem., 1996, 35, 4283. C. Qian, C. Ye, H. Lu, Y. Li, J. Zhou, Y. Ge and M. Tsutsui, J. Organomet. Chem., 1983, 247, 161. W. J. Evans, J. H. Meadows, A. L. Wayda, W. E. Hunter and J. L. Atwood, J. Am. Chem. Soc., 1982, 104, 2008. H. Schumann and G. Jeske, Angew. Chem., 1985, 97, 208; Angew. Chem., Int. Ed. Engl., 1985, 24, 255.
108 [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174]
2
Lanthanocenes
H. Schumann and G. Jeske, Z. Naturforsch., 1985, B40, 1490. H. Schumann, F. W. Reier and E. Palamidis, J. Organomet. Chem., 1985, 297, C30. G. B. Deacon and D. L. Wilkinson, Inorg. Chim. Acta, 1988, 142, 155. W. J. Evans, R. Dominguez and T. P. Hanusa, Organometallics, 1986, 5, 263. H. Schumann, H. Lauke, E. Hahn, M. J. Heeg and D. Van der Helm, Organometallics, 1985, 4, 321. H. Schumann, F. W. Reier and E. Hahn, Z. Naturforsch., 1985, B40, 1289. Q. Shen, Y. Cheng and Y. Lin, J. Organomet. Chem., 1991, 419, 293. H. Schumann, F. W. Reier and M. Dettlaff, J. Organomet. Chem., 1983, 255, 305. H. Schumann and F. W. Reier, Inorg. Chim. Acta, 1984, 95, 43. V. I. Bregadze, N. A. Kovalchuk, N. N. Godovikov, G. Z. Suleimanov and I. P. Beletskaya, J. Organomet. Chem., 1983, 241, C13. H. Schumann, S. Nickel, E. Hahn and M. J. Heeg, Organometallics, 1985, 4, 800. H. Schumann, J. A. Meese-Marktscheffel and F. E. Hahn, J. Organomet. Chem., 1990, 390, 301. H. Schumann, S. Nickel, J. Loebel and J. Pickardt, Organometallics, 1988, 7, 2004. B. K. Campion, R. H. Heyn and T. D. Tilley, Organometallics, 1993, 12, 2584. H. Schumann, I. Albrecht, J. Pickardt and E. Hahn, J. Organomet. Chem., 1984, 276, C5. M. L. H. Green, A. K. Hughes, N. A. Popham, A. H. H. Stephens and L.-L. Wong, J. Chem. Soc., Dalton Trans., 1992, 3077. W. J. Evans, T. A. Ulibarri, L. R. Chamberlain, J. W. Ziller and D. Alvarez, Organometallics, 1990, 9, 2124. K. H. den Haan, J. L. de Boer, J. H. Teuben, W. J. J. Smeets and A. L. Spek, J. Organomet. Chem., 1987, 327, 31. W. J. Evans, L. R. Chamberlain, T. A. Ulibarri and J. W. Ziller, J. Am. Chem. Soc., 1988, 110, 6423. P. L. Watson and T. Herskovitz, ACS Symp. Ser (Initiation Polym.), 1983, 212, 459. M. E. Thompson, S. M. Baxter, A. R. Bulls, B. J. Burger, M. C. Nolan, B. D. Santarsiero, W. P. Schaefer and J. E. Bercaw, J. Am. Chem. Soc., 1987, 109, 203. P. L. Watson, J. Am. Chem. Soc., 1983, 105, 6491. P. T. Matsunaga, Energy Res. Abstr., 1992, 17(6). G. M. Forsyth, S. P. Nolan and T. J. Marks, Organometallics, 1991, 10, 2543. T. Ziegler, E. Folga and A. Berces, J. Am. Chem. Soc., 1992, 115, 636. E. Folga, T. Ziegler and L. Fan, New. J. Chem., 1992, 15, 741. W. E. Piers, P. J. Shapiro, E. E. Bunel and J. E. Bercaw, Synlett, 1990, 74. P. J. Fagan, J. M. Manriquez, T. J. Marks, V. W. Day, S. H. Vollmer and C. S. Day, J. Am. Chem. Soc., 1980, 102, 5393. H. J. Heeres, J. Renkema, M. Booij, A. Meetsma and J. H. Teuben, Organometallics, 1988, 7, 2495. G. Jeske, L. E. Schock, P. N. Swepston, H. Schumann and T. J. Marks, J. Am. Chem. Soc., 1985, 107, 8103. H. Mauermann, P. N. Swepston and T. J. Marks, Organometallics, 1985, 4, 200. J. Renkema and J. H. Teuben, Recl. Trav. Chim. Pays-Bas, 1986, 105, 241. G. Jeske, H. Lauke, H. Mauermann, P. N. Swepston, H. Schumann and T. J. Marks, J. Am. Chem. Soc., 1985, 107, 8091. J. Renkema and J. H. Teuben, Recl. Trav. Chim. Pays-Bas, 1988, 105, 241. C. M. Fendrick, L. D. Schertz, V. W. Day and T. J. Marks, Organometallics, 1988, 7, 1828. H. Schumann, L. Esser, J. Loebel, A. Dietrich, D. Van der Helm and X. Ji, Organometallics, 1991, 10, 2585. H. Schumann, J. Loebel, J. Pickardt, C. Qian and Z. Xie, Organometallics, 1991, 10, 215. D. Stern, M. Sabat and T. J. Marks, J. Am. Chem. Soc., 1990, 112, 9558. W. P. Schaefer, R. D. Köhn and J. E. Bercaw, Acta Cryst., 1992, C48, 251. R. E. Marsh, W. P. Schaefer, E. B. Coughlin and J. E. Bercaw, Acta Cryst., 1992, C48, 1773.
References
109
[175] N. Koga and K. Morokuma, J. Am. Chem. Soc., 1988, 110, 108. [176] H. J. Heeres, A. Meetsma and J. H. Teuben, Angew. Chem., 1990, 102, 449; Angew. Chem., Int. Ed. Engl., 1990, 29, 420. [177] J. H. Teuben, in "Fundamental and Technological Aspects of Organo-f-Element Chemistry", NATO ASI Ser., ed. T. J. Marks and I. L. Fragalà, D. Reidel, Boston, 1985, 155, 195. [178] K. H. den Haan, G. A. Luinstra, A. Meetsma and J. H. Teuben, Organometallics, 1987, 6, 1509. [179] K. H. den Haan, Y. Wielstra and J. H. Teuben, Organometallics, 1987, 6, 2053. [180] H. J. Heeres, M. Maters and J. H. Teuben, Organometallics, 1992, 11, 350. [181] P. L. Watson and G. W. Parshall, Acc. Chem. Res., 1985, 18, 51. [182] M. A. Busch, R. Harlow and P. L. Watson, Inorg. Chim. Acta, 1987, 140, 15. [183] W. J. Evans, L. R. Chamberlain and J. W. Ziller, J. Am. Chem. Soc., 1987, 109, 7209. [184] H. Yamamoto, H. Yasuda, K. Yokota, A. Nakamura, Y. Kai and N. Kasai, Chem. Lett., 1988, 1963. [185] H. Schumann, M. Glanz, J. Winterfeld, H. Hemling, N. Kuhn and T. Kratz, Angew. Chem., 1994, 106, 1829; Angew. Chem., Int. Ed. Engl., 1994, [186] A. J. Arduengo, III, M. Tamm, S. J. McLain, J. C. Calabrese, F. Davidson and W. J. Marshall, J. Am. Chem. Soc., 1994, 116, 7927. [187] W. A. Herrmann, F. C. Munck, G. R. J. Artus, O. Runte and R. Anwander, Organometallics, 1997, 16, 682. [188] A. Scholz, A. Smola, J. Scholz, J. Loebel, H. Schumann and K.-H. Thiele, Angew. Chem., 1991, 103, 444; Angew. Chem., Int. Ed. Engl., 1991, 30, 435. [189] W. J. Evans, T. A. Ulibarri and J. W. Ziller, J. Am. Chem. Soc., 1990, 112, 219. [190] W. J. Evans, S. L. Gonzales and J. W. Ziller, J. Am. Chem. Soc., 1994, 116, 2600. [191] W. J. Evans, J. H. Meadows, W. E. Hunter and J. L. Atwood, Organometallics, 1983, 2, 1252. [192] W. J. Evans, I. Bloom, W. E. Hunter and J. L. Atwood, J. Am. Chem. Soc., 1983, 105, 1401. [193] J. M. Boncella, T. D. Tilley and R. A. Andersen, J. Chem. Soc., Chem. Commun., 1984, 710. [194] W. J. Evans, D. K. Drummond, T. P. Hanusa and J. M. Olofson, J. Organomet. Chem., 1989, 376, 311. [195] H. J. Heeres, J. Nijhoff, J. H. Teuben and R. D. Rogers, Organometallics, 1991, 12, 2609. [196] C. M. Forsyth, S. P. Nolan, C. S. Stern, T. J. Marks and A. L. Rheingold, Organometallics, 1993, 12, 3619. [197] A. Recknagel, D. Stalke, H. W. Roesky and F. T. Edelmann, Angew. Chem., 1989, 101, 496; Angew. Chem., Int. Ed. Engl., 1989, 28, 445. [198] W.-K. Wong, H. Chen and F.-L. Chow, Polyhedron, 1990, 9, 875. [199] H. Siebald, M. Dartiguenave and Y. Dartiguenave, J. Organomet. Chem., 1992, 438, 83. [200] N. S. Radu and T. D. Tilley, Phosphorus, Sulfur, and Silicon, 1994, 87, 209. [201] N. S. Radu, T. D. Tilley and A. L. Rheingold, J. Am. Chem. Soc., 1992, 114, 8293. [202] T. Kobayashi, T. Sakakura, T. Hayashi, M. Yumura and M. Tanaka, Chem. Lett., 1992, 1158 [203] T. Sakakura, H.-J. Lautenschlager and M. Tanaka, J. Chem. Soc., Chem. Commun., 1991, 40. [204] J. V. Ortiz and R. Hoffmann, Inorg. Chem., 1985, 24, 2095. [205] E. N. Zavadovskaya, O. K. Sharaev, G. K. Borisov, Y. P. Yampolskii, E. I. Tinyakova and B. A. Dolgoplosk, Dokl. Akad. Nauk SSSR, 1985, 284, 143. [206] W. J. Evans, J. H. Meadows, W. E. Hunter and J. L. Atwood, J. Am. Chem. Soc., 1984, 106, 1291. [207] I. P. Beletskaya, A. Z. Voskoboinikov and G. K.-I. Magomedov, Metalloorg. Khim., 1989, 2, 810. [208] D. Deng, Y. Jiang, C. Qian, G. Wu and P. Zheng, J. Organomet. Chem., 1994, 470, 99. [209] W. J. Evans, D. K. Drummond, T. P. Hanusa and R. J. Doedens, Organometallics, 1987, 6, 2279. [210] H. Schumann, W. Genthe, E. Hahn, M. B. Hossain and D. Van der Helm, J. Organomet. Chem., 1986, 299, 67. [211] C. Qian, D. Deng, C. Ni and Z. Zhang, Inorg. Chim. Acta, 1988, 146, 129.
110
2
Lanthanocenes
[212] S. Y. Knyazhanskii, V. K. Bel’skii, B. M. Bulychev and G. L. Soloveichik, Metalloorg. Khim., 1989, 2, 570. [213] Y. K. Gun’ko, B. M. Bulychev, G. L. Soloveichik and V. K. Bel’skii, J. Organomet. Chem., 1992, 424, 289. [214] W. J. Evans, T. P. Hanusa, J. H. Meadows, W. E. Hunter and J. L. Atwood, Organometallics, 1987, 6, 295. [215] T. J. Marks and G. W. Grynkewich, Inorg. Chem., 1976, 15, 1302. [216] M. F. Lappert, A. Singh, J. L. Atwood and W. E. Hunter, J. Chem. Soc., Chem. Commun., 1983, 206. [217] V. D. Makhaev and A. P. Borisov, Koord. Khim., 1992, 466. [218] M. Booij, B.-J. Deelman, R. Duchateau, D. S. Postma, A. Meetsma and J. H. Teuben, Organometallics, 1993, 12, 3531. [219] K. H. den Haan and J. H. Teuben, Recl. Trav. Chim. Pays-Bas, 1984, 103, 333. [220] C. Ye, C. Qian and X. Yang, J. Organomet. Chem., 1991, 407, 329. [221] C. Qian, Z. Xie and Y. Huang, Inorg. Chim. Acta, 1987, 139, 195. [222] W. J. Evans, T. A. Ulibarri and J. W. Ziller, Organometallics, 1991, 10, 134. [223] W. J. Evans, J. W. Grate and R. J. Doedens, J. Am. Chem. Soc., 1985, 107, 1671. [224] E. Bunel, B. J. Burger and J. E. Bercaw, J. Am. Chem. Soc., 1988, 110, 976. [225] B. Cornils and W. A. Herrmann, "Applied Homogeneous Catalysis with Organometallic Compounds", VCH, Weinheim, 1996. [226] G. B. Deacon, P. MacKinnon, R. S. Dickson, G. N. Pain and B. O. West, Appl. Organomet. Chem., 1990, 4, 439. [227] Y. K. Gun’ko and F. T. Edelmann, Comments Inorg. Chem., 1997, 19, 153.
3
Group 3 Metallocenes Paul J. Chirik and John E. Bercaw
3.1 Introduction Cyclopentadienyl is perhaps the most important and widespread ancillary ligand in organometallic chemistry. Cyclopentadienyl ligands stabilize a wide variety of organometallic complexes with group 3 metals (Sc, Y, La), as they do for most other transition metals. The focus of this review is to present a general synopsis of the chemistry of group 3 complexes based on the bis(cyclopentadienyl) (or substituted cyclopentadienyl) framework. It concentrates primarily on the chemistry of scandocene and yttrocene compounds. Lanthanum metallocenes have thus far received less attention, and thus lanthanocenes will be mentioned only briefly. Traditionally, group 3 metallocene chemistry has been reviewed with organolanthanide chemistry [1 ], however, elements beyond La will not be considered here. The definition of a metallocene for the purpose of this review is a complex containing two η5-coordinated cyclopentadienyl or substituted cyclopentadienyl rings. The scope of this definition is expanded in some cases to include monocyclopentadienyl complexes having alkoxide or amide ligands, in cases where such compounds undergo chemistry similar to that of the bis(cyclopentadienyl) systems. Group 3 metallocene chlorides [Cp2MCl]x (Cp = (η5-C5H5), M = Sc, Y) were prepared by Wilkinson and Birmingham in the mid-1950s, and apart from some important studies by Lappert, Wailes, Manzer and Atwood in the 1970s, most of the chemistry of the group 3 metallocenes has been developed during the past 10 years. This review will focus primarily on these recent developments, especially the use of group 3 metallocene complexes as reagents for activation of the C–H bonds of hydrocarbons, for polymerization of olefins, and more recently, as reagents for organic synthesis. The first sections concern methods for preparing group 3 metallocene complexes and summarize the general reactivity trends observed for these systems. Special attention is devoted to alkyls and hydrides, since these are most often encountered in catalytic applications, most notably polymerization of α–olefins. Later sections review reactions of the group 3 metallocene compounds with acetylenes, with other small molecules and the applications of some of this chemistry to organic synthesis. The final two brief sections are devoted to some studies of the photophysics of group 3 metallocene compounds and to ligands that are designed as alternatives to the cyclopentadienyl ligand array.
112
3 Group 3 Metallocenes
3.2 Synthesis of Group 3 Metallocenes 3.2.1 Scandium Although the organometallic chemistry of first row transition elements (Ti through Cu) has been extensively investigated over the past three decades, the first member of the series, scandium, has received much less attention. The slow development of organoscandium chemistry may be traced to the perception that organometallic compounds of scandium differ little from those of its group 13 counterparts, being plagued by insolubility due to their tendency to form polymeric materials supported by 3–center, 2- or 4-electron (M–(µ-X)–M) bonds (X = halide, hydrocarbyl, etc.) [2 ]. However, the employment of the cyclopentadienyl ligand (and its substituted derivatives) has provided an opportunity to the prepare soluble, well-defined organoscandium compounds. A veritable renaissance in its organometallic chemistry has followed. The first report of a scandocene complex was communicated by Wilkinson and Birmingham [3] with the synthesis of Cp3Sc from the reaction of anhydrous ScCl3 and NaCp. This material is thermally robust, but is polymeric and insoluble and thus a rather intractable material [4 ]. Scandocene chloride, (Cp2ScCl)2, prepared from scandium trichloride and MgCp2, may be isolated as a green solid [5]. Ebulliometric molecular weight measurements are indicative of a chloro-bridged dimeric structure, which has been confirmed by X-ray crystallography [6]. The syntheses of [Cp2Sc(O2CCH3)]2, Cp2Sc(η3-allyl), and [Cp2Sc(C≡CPh)]2 have also been described [5]. Manzer [7] reports a more convenient preparation for Cp2ScCl(THF) from the reaction of two equivalents of TlCp with anhydrous ScCl3 in THF solution (Eq. (3-1)).
(3-1)
Alkylation with Li(CH2)2PPh2 affords the scandium chelate complex, Cp2Sc(µ2CH2)2PPh2; reaction with LiC6H4-o-CH2NMe2 yields Cp2Sc(η2-C6H4-o-CH2NMe2) [8 ,9 ]. The product obtained from scandium fluoride is quite different. Reaction of ScF3 with Cp2Mg results in a mixture of Cp3Sc and ‘Cp2ScF’ [10 ,11]. The same mixture also results from treatment of the trifluoride with NaCp [11]. X-ray crystallographic studies show that ‘Cp2ScF’ is actually a trimer with bridging fluoride ligands. Lappert [12 ] reported the first example of a scandocene having substituted cyclopentadienyl ligands with the preparations of [(η5-C5H3(SiMe3)2)2ScCl]2 and {(η5-C5H3(SiMe3)2}2Sc(µ2-H)2BH2. Despite the presence of two bulky trimethylsi-
3.2
Synthesis of Group Metallocenes
113
lyl substituents, the resulting scandocene chlorides still undergo dimerization. However, use of the sterically very demanding pentamethylcyclopentadienyl ligand (Cp*) affords monomeric scandium chlorides and alkyls. Thus, reaction of ScCl3·(THF)3 with LiCp* produces the monomeric permethylscandocene chloride, Cp*2ScCl [13 ]. Addition of the appropriate lithium reagent, LiR, results in the formation of the corresponding hydrocarbyl derivative, Cp*2ScR (R = CH3, C6H5, CH2C6H5). Hydrogenolysis of the Sc–C bonds of these derivatives produces Cp*2ScH, which is stable only at low temperatures or under a dihydrogen atmosphere. Hydrogenation of the alkyls in a mixture of hexane and THF affords the more stable THF adduct, Cp*2ScH(THF) (Eq. (3-2)) [14 ].
(3-2)
Since the reactivity of Cp*2ScR compounds is limited by the sterically demanding Cp* ligands, two strategies have been devised to relieve the excess crowding at scandium: (1) the synthesis of mixed ring scandocenes where one of the Cp* rings is replaced by a less bulky cyclopentadienyl ligand and (2) synthesis of ansa scandocenes, where the two cyclopentadienyl rings are ‘tied back’ by a dimethylsilylene linker. For the former strategy, the challenge is the synthesis of the requisite mono(pentamethylcyclopentadienyl) scandium dichloride precursor, since the reaction of one equivalent of Cp*Li with ScCl3 results in the formation of ill-defined products [15 ]. Despite these difficulties, high yield routes to [Cp*ScCl2]x have been developed from the reaction of Sc(acac)3 and Cp*MgCl·THF, producing Cp*Sc(acac), which is then converted to the dichloride through reaction with a stoichiometric amount of AlCl3 (Eq. (3-3)) [16 ].
(3-3)
Mixed-ring scandocene chlorides can be obtained from the reaction of [Cp*ScCl2]x and LiCp (Cp = cyclopentadienyl or substituted cyclopentadienyl). Conversion of the mixed-ring chloride to alkyls and hydrides can be achieved through similar synthetic procedures as with the permethylscandocene system, except that the addition of PMe3 is used to stabilize the resulting alkyl complexes (Scheme 3-1) [16]. Mixed ring scandocene hydrides have also been prepared; however, these compounds readily undergo bimolecular decomposition with release of H2 to give the metallated dimeric compound Cp*(η5-C5H5)Sc(µ2-η5,η1-C5H4)(µ2-H)ScCp*[17].
114
3 Group 3 Metallocenes
Scheme 3-1
The second approach to preparing more reactive scandocene complexes is the use of ‘tied-back’ cyclopentadienyl rings, thus producing an ansa-scandocene. A variety of linked cyclopentadienyl ligands has been synthesized [18] and coordinated to scandium, as shown in Scheme 3-2.
Scheme 3-2
3.2
Synthesis of Group Metallocenes
115
OpScCl and meso-DpScCl are prepared via reactions of equimolar mixtures of the dilithium salt of the respective linked bis(cyclopentadienides) and ScCl3·3THF in refluxing toluene. For the OpScCl system, one equivalent of coordinated lithium chloride is inseparable from the scandocene under normal workup. However, meso -DpScCl is an oligomeric (probably dimeric) material that is free of lithium chloride and Lewis bases. Reaction of bulky alkylating agents such as LiCH(SiMe3)2 with the scandocene chlorides results in the formation of the corresponding bis(trimethylsilyl)methyl complex. Treatment with dihydrogen results in the formation of the scandium hydride species. ‘OpScH’ is most conveniently isolated as a PMe3 adduct as identified by X-ray diffraction [18 ], whereas meso-DpSc-H forms a stable dimeric species. Linked cyclopentadienyl–amide ligands have also been developed for use as ancillary ligands for scandium. These ligands were devised as a result of the striking effect on the observed reactivity of the scandium center as the steric properties on the ligand framework are varied. These Cp-amido systems render the metal more electron-deficient and less sterically encumbered than their bis(cyclopentadienyl) counterparts. Reaction of ScCl3(THF)3 with Li2{(C5Me4)Me2Si(NCMe3)} results in (Cp*SiNR)ScCl (Cp*SiNR = {(η5-C5Me4)SiMe2(η1-NCMe3}, as an adduct with LiCl and THF. Removal of THF generates a scandium derivative that can be alkylated with LiCH2(SiMe3)2 to form the alkyl derivative [19 ]. Hydrogenation in the presence of PMe3 forms the double hydrogen bridged dimer of C2 symmetry (Scheme 3-3).
Scheme 3-3
116
3 Group 3 Metallocenes
Piers et. al. [20 ] have developed a more direct route to scandium alkyls that does not require ligand deprotonation or preparation of the requisite scandium chloride species. In situ generation of Sc(CH2SiMe3)3·2THF from ScCl3·(THF)3 followed by treatment with CpHNMeSiN(H)R results in the formation of [(CpNMeSiNR)Sc(CH2SiMe3)] in 55% yield.
(3-4)
Hydrogenolysis of the scandium alkyl results in the formation of two doubly-hydride bridged species. One of the products was characterized by X-ray crystallography and identified as a Ci symmetric diastereomeric scandium hydride dimer. Of the three other diastereomeric products that are possible, the identity of the other product was tentatively assigned as the ‘1R-cis-1'-R’ isomer based on steric arguments and literature precedent.
3.2.2 Yttrium and Lanthanum Cp3Y, the first cyclopentadienyl derivative of yttrium, was first prepared by Birmingham and Wilkinson from the reaction of YCl3 and NaCp [21 ]. The crystal structure of the THF adduct, Cp3Y(THF) has since been reported [22 ]. The lanthanum complex can be synthesized in a similar manner and has been identified as a ‘zigzag’ polymeric material in which each metal ion is coordinated in an η5 fashion to three Cp ligands and η2-coordinated to a fourth Cp ring [23 ]. Substitution of one methyl group on the cyclopentadienyl ligand in La(C5H4Me)3 reduces the polymeric nature to a tetramer, in which each metal is η5-coordinated to three Cp anions and η1-coordinated to a fourth bridging cyclopentadienyl [24 ]. Since the sterically encumbered tris(cyclopentadienyl) complexes do not offer much opportunity for catalytic activity, more attention has been devoted to the preparation of bis(cyclopentadienyl) complexes of yttrium. Alkyl and hydride derivatives have received the most attention, due to their function as homogeneous catalysts. Reaction of [Cp2Y(µ2-Cl)]2 with LiCH2SiMe3 in dimethoxyethane (DME)/dioxane mixture forms [Cp2Y(CH2SiMe3)2]2Li2(DME)2(dioxane) [25 ]. This dimeric alkyl, characterized by X-ray crystallography, reacts with dihydrogen slowly to produce the trimeric yttrium hydride, [Cp2Y(µ2-H)]3(µ3-H)][Li(DME)2]. It was discovered early on that the rate of the hydrogenolysis of an yttrium–carbon bond is highly dependent on the specific alkyl complex involved in the reaction [26 ]. The cyclopentadienyl complexes (RCp)2Y(tert-C4H9)(THF) (R = H, CH3), synthesized from the reaction of [(RCp)2Y(µ2-Cl)]2 with tert-C4H9Li in THF [27 ],
3.2
Synthesis of Group Metallocenes
117
react rapidly with H2 at room temperature and atmospheric pressure, while the dimeric methyl species take weeks for the reaction to reach completion [25]. In both cases the crystallographically characterized yttrocene hydride, [(RCp)2Y(THF)(µ2-H)]2, is generated. Reaction of [Cp2Y(µ2-H)(THF)]2 with LiH, CH3Li, and tert-C4H9Li has been investigated and a variety of products identified [28 ]. For example, reaction of [Cp2Y(µ2-H)(THF)]2 with tert-C4H9Li results in the trimeric hydride, [Cp2Y(µ2H)]3(µ3-H)][Li(THF)4] in 75% yield, with Cp2Y(tert-C4H9)(THF) and Cp3Y(THF) identified as byproducts. The trimer has been characterized by 1H NMR spectroscopy, with diagnostic resonances being observed for the yttrium-coupled hydride signals. The mechanism for trimer formation is believed to be in situ generation of [Cp2YH2]– which then reacts with another equivalent of the original hydride dimer to form the trimeric hydride. In support of this proposal, reaction of [(MeCp)2ZrH2]x with [(MeCp)2Y(µ2-H)(THF)]2 forms the bimetallic trimer. Reaction with CH3Li does not produce the trimeric hydride, but rather results in an oil identified as Cp2Y(CH3)2Li(THF) as well as other unidentified oligomeric products. Unlike for the organoscandium analogs, use of the sterically demanding pentamethylcyclopentadienyl ligand generally does not produce monomeric yttrocenes free of coordinated salts or Lewis bases. Reaction of YCl3 with 2 equivalents of KCp* in THF results in the formation of Cp*2YCl2K(THF)2 [29 ]. Heating of this compound at 285°C under vacuum allows for isolation of the asymmetric, solventfree dimer, Cp*2Y(µ2-Cl)Y(Cl)Cp*2 (Eq. (3-5)).
(3-5)
A byproduct during the synthesis of Cp*2YCl2K(THF)2 is Cp*2YCl(THF) [30 ], which has been identified by X-ray crystallography. Reaction of Cp*2YCl(THF) with LiOCMe3 in refluxing toluene yields a mixture of products. Both reaction products, [Cp*2Y(µ2-Cl)2Li(THF)2] and [Cp*2YCl(µ2-Cl)Li(THF)3], cocrystallize from the reaction mixture and are found in the same unit cell [31 ]. Teuben and coworkers [32 ] have prepared monomeric, base-free derivatives of permethylyttrocene by using bulky alkyl and amide groups. Reaction of Cp*2YCl(THF) with either NaN(SiMe3)2 or LiCH(SiMe3)2 results in the crystallographically characterized Cp*2YN(SiMe3)2 and Cp*2YCH(SiMe3)2 compounds. The monomeric alkyl, being an electron-deficient, 14-electron species displays a δ agostic C–H interaction, whereas the isoelectronic amide contains a short Y–N bond, suggesting double bond character [32].
118
3 Group 3 Metallocenes
Attempts to prepare monomeric yttrium hydrocarbyls of smaller alkyl groups is plagued by preferential yttrate formation with salts such as LiCl and MgX2 and coordination of ethereal solvents (Et2O, THF) [33 ]. For example, reaction of Cp*2Y(µ2-Cl)2Li·2OEt2 with two equivalents of methyllithium at –80°C results in the bridged dimethyl yttrate, Cp*2Y(µ2-CH3)2Li·OEt2, while the same reaction in THF solution produces Cp*2Y(CH3)(THF), whose structure has been confirmed by NMR spectroscopy, elemental analysis and X-ray diffraction (Scheme 3-4) [34 ]. Coupling of the methyl group to a single 89Y (spin 1/2, 100% abundant) nucleus is observed in both 1H and 13C NMR spectra, indicative of a monomer in solution as well.
Scheme 3-4
Attempts to prepare salt- and solvent-free allyls using the chelating nature of this ligand led instead to Cp*2Y(η1-CH2CMe=CH2)·[MgCl2·2THF], from reaction of Cp*2YCl(THF) with 2-methyl-2-propenylmagnesium chloride. Apparently, the strong tendency for yttrates to form effectively blocks η3 coordination of the allyl [33]. Hydrogenation of the monomeric, Cp*2YCH(SiMe3)2 in pentane results in the hydride dimer, [Cp*2Y(µ2-H)]2 [35 ,36 ]. Unlike the analogous scandium system which is unstable unless a dihydrogen atmosphere is maintained, the yttrium hydride can be isolated in the solid state, presumably due to its dimeric nature. Thermolysis of the dimer results in the formation of the red fulvene-bridged dimer, identified as Cp*2Y(η1,η5-CH2C5Me4)(µ2-H)YCp* on the basis of NMR evidence [37 ]. In view of the common use of alkyl aluminum co-catalysts in Ziegler–Natta αolefin polymerization systems, a variety of tetraalkylaluminates of yttrocene have
3.2
Synthesis of Group Metallocenes
119
been prepared. Reaction of [Cp2YCl]2 and Li[AlR4] (R = CH3, CH3CH2) in toluene results in the formation of Cp2YR2AlR2; the ethyl derivative has been structurally characterized by single crystal X-ray diffraction [38 ].
(3-6)
NMR data indicate the complexes are fluxional, but the pseudorotation can be frozen out at –40°C. Similarly, the alane adducts, [Cp2YCl]2·AlH3·OEt2 [39 ] and [Cp2YCl]2·AlH3·NEt3 [40 ] have also been prepared from [Cp2YCl]2 and MAlH4 (M = Li or Na) in the presence of diethyl ether or triethylamine [41 ]. The structures of these compounds have been confirmed by X-ray crystallography [41,42 ,43 ]. Symmetrical, linearly-bridged dimers of [Cp*2Y(µ2-CH3)2Al(CH3)2]2 and [Cp*2Y(µ2-CH3)2Ga(CH3)2]2 have been reported by Watson [44 ]. These products are formed from the reaction of solvent-free Cp*2Y(CH3)2Li with M'(CH3)3 (M' = Al, Ga). NMR spectra indicate two species in solution, which were shown to be in equilibrium (Eq. (3-7)).
(3-7)
Under normal concentrations and at room temperature the monomer is the dominant species in solution. Mixing Cp*2Y(µ2-CH3)2Al(CH3)2 and Cp*2Y(µ2CH3)2Ga(CH3)2 results in the formation of a random distribution of symmetric and asymmetric dimers. The correlation of relative bond strengths, analogous to those for a bimolecular transition state, have been established through kinetic studies which show that Y–(µ2-CH3)–Al bonds are stronger than Y–(µ2-CH3)–Ga. The data do not, however, correlate well with the release of steric strain in the four membered ring as the driving force for the dimerization. The free energies of dimerization are approximately zero with the enthalpy of the reaction favoring the formation of dimers with linear methyl-bridge bonds over monomers with bent methyl-bridge bonds. From these data, it can be concluded that the methyl group is a
120
3
Group 3 Metallocenes
ligand with a high propensity for hypercoordination with the electropositive metal center. Ethyltetramethylcyclopentadienyl (η5-C5Me4Et) derivatives of yttrium have also been prepared and their reactivity explored. Reaction of yttrium trichloride and two equivalents of Na[C5Me4Et] in THF forms [(η5-C5Me4Et)2YCl(THF)] [45 ]. Conversion to monomeric amide or alkyl derivatives is accomplished by addition of NaN(SiMe3)2 or LiCH(SiMe3)2 in toluene at 0°C. It is apparent from the crystal structure of the amide derivative that the amide ligand is moved from the center of the metallocene wedge by a (µ-CH3) bridging yttrium and silicon.
(3-8)
Other alkyl substituted cyclopentadienyl ligands have been synthesized and coordinated to yttrium using analogous procedures. Reaction of YCl3 with 2 equivalents of K(1,3-C5H3Me2) in THF results in the formation of (η5-1,3C5H3Me2)2Y(THF)Cl, which can be converted to [(η5-1,3-C5H3Me2)2YCl]2 by stirring in toluene [46 ]. The methyl complex, prepared by the reaction of (η5-1,3C5H3Me2)2Y(THF)Cl and 2 equivalents of methyllithium, is a dimer, [(η5-1,3C5H3Me2)2Y(µ2-CH3)]2. Hydrogenolysis in THF/hexane results in an yttrium hydride species, whose structure is dependent on how the product is isolated. Removal of solvent results in the formation of the trimeric, [(η5-1,3-C5H3Me2)2Y(µ2-H)]3, but recrystallization of this compound from THF yields the dimeric species [(η51,3-C5H3Me2)2Y(THF)(µ2-H)]2. It was concluded that dimethyl substitution has only a small effect on the yttrium chemistry compared to (η5-C5H5) or (η5C5H4Me) derivatives [46]. Since synthesis of group 3 metallocenes is often plagued by facile coordination of unwanted solvent, halide or alkyl groups, incorporation of Lewis basic sites into the cyclopentadienyl ligand framework has been pursued in an attempt to prepare ‘base-free’ metallocenes. Treatment of YCl3(THF)3.5 with two equivalents of C5H4CH2CH2OMe·Li(TMEDA) in toluene at –80°C results in the formation of [(η5-C5H4CH2CH2OMe)2Y(µ2-Cl)]2 [47 ]. Conversion to the aluminohydride or borohydride adduct, [(η5,η1-C5H4CH2CH2OMe)2Y]EH4 (E = Al, B) can be accomplished through addition of either LiAlH4 or NaBH4 to the chloride dimer. The crystal structure of the borohydride complex indicates that both ether functions of the cyclopentadienyl ligand are coordinated to the yttrium center. Treatment of the aluminohydride complex with excess NEt3 results in the hydride dimer, [(η5,η1C5H4CH2CH2OMe)2Y(µ2-H)]2. This conversion can be carried out in situ from the chloride derivative, resulting in the formation of the hydride dimer in moderate yield [47]. Hydride dimers have also been prepared by reactions of [(η5,η1-
3.2
Synthesis of Group Metallocenes
121
C5H4CH2CH2OMe)2M(µ2-Cl)]2 (M = Y, La) with NaH in THF [48 ]. Tris(2-methoxyethylcyclopentadienyl) sodium reacts with MCl3 (Ln = La, Y) to yield [M(η5,η1-C5H4CH2CH2OMe)3] [49 ]. X-ray crystallography reveals that the complexes have two of the oxygens of the cyclopentadienyl rings coordinated, while the third oxygen is not coordinated. Diphenylphosphino-substituted cyclopentadienyl yttrium complexes have also been prepared [50 ]. Reaction of lithium diphenylphosphinocyclopentadienide with YCl3·(THF)3.5 results in the formation of three species: (η5-Ph2PC5H4)2Y(µ2Cl)2Li(THF)2·0.5LiCl, {(η5-Ph2PC5H4)2Y(µ2-Cl)}2 and [((η5-Ph2PC5H4)2YCl(THF)]. Reaction of the chloro-bridged dimer with [Mo(CO)6] in refluxing toluene illustrates the ability of the phosphines to ligate to another metal center, forming [{CO)4Mo{(Ph2PC5H4)2Y(µ2-Cl)}}2]. Incorporation of a phosphorus into the cyclopentadienyl ring has also been accomplished. The yttrocene (η5C4Me4P)2YCl(THF) has been prepared from the reaction of YCl3·(THF)3.5 and Li(PC4Me4) (Eq. (3-9)) [51 ].
(3-9)
Another strategy for incorporation of heteroatoms into the cyclopentadienyl ligand array has been through the use of a heteroatom-containing cyclopentadienyl linking group. Ether-oxygen bridged cyclopentadienyls such as 1,1'-(3-oxapentamethylene)dicyclopentadienyl have been reported. Reaction of the disodium salt with either YCl3 or LaCl3 produces the unsolvated monomeric chloride containing an intramolecular metal-oxygen dative bond [52 ]. These compounds in the presence of NaH reduce alkenes such as 1-hexene to the corresponding alkane. Shumann et al. [53 ] reported that the resulting chloride complex can be converted to the dimethylpyrazolyl derivative by reaction of (3,5-dimethylpyrazolyl) sodium in THF (Scheme 3-5). Replacement of the oxygen with a carbon linker in 1,1'-pentamethylenedicyclopentadiene forms only THF solvated complexes when coordinated to yttrium [54 ]. The pyrazol complex reacts with water to produce the µ-hydroxide dimer, whose structure has been characterized crystallographically. A shorter carbon linker of 3 atoms has also been employed in the reaction of 1,1'-trimethylenedicyclopentadienyl sodium with LaCl3 or YCl3 forming the bridged, THF-adduct, [(η5C5H4)2(CH2)3]MCl·THF [55 ]. Addition of aryllithium (Aryl = C6H5; p-MeC6H4) or alkyllithium (alkyl = CMe3, (CH3)3CCH2) reagents to the chloride–THF adduct produce THF-solvated alkyl or aryl complexes.
122
3
Group 3 Metallocenes
Scheme 3.5
Furan-linked cyclopentadienyls have also been prepared and coordinated to yttrium [48]. In these chloro-bridged dimeric compounds, the furan rings coordinate to the yttrium centers.
(3-10)
A variety of silicon linked cyclopentadienyl ligands have been employed in organoyttrium chemistry. Bercaw and co-workers [56 ,57 ] have prepared chloride, alkyl and hydride yttrocenes with the ‘Bp’ ligand array {Bp = {(η5-C5H2-2-SiMe3-4CMe3)2SiMe2}). Reaction of YCl3(THF)3.5 and K2Bp forms the THF–chloride adduct, whose synthesis proceeds smoothly with the potassium salt of the ligand, whereas reaction with Li2Bp results in BpY(µ2-Cl)2Li(THF)2. Alkylation with LiCH(SiMe3)2 yields BpY{CH(SiMe3)2}, which can be hydrogenated to form the hydride dimer (Scheme 3-6).
3.2
Synthesis of Group Metallocenes
123
Scheme 3-6) Piers [58 ] has prepared silicon linked cyclopentadienyl-amide yttrocenes via amine elimination from Y[N(SiMe)2]3, to allow for direct isolation of the yttrium amide such that metallation to the yttrium chloride species can be achieved without ligand deprotonation. Reaction of Y[N(SiMe3)2]3 and Cp*HSiNHCMe3 occurs at 110°C in toluene solution, producing (Cp*SiNR)YN(SiMe3)2 in 40% yield. Unfortunately, conversion of the yttrium-amide to the yttrium-alkyl using standard alkylating agents such as LiCH2SiMe3 has not been successful, presumably due to the thermal instability of the resulting alkyl complex.
(3-11)
Indenyl-based group 3 metallocenes have also been reported. Reaction of YCl3(THF)3.5 with two equivalents of Na+Ind– (Ind = indenyl) results in a 2:1 mixture of (η5-C9H7)YCl2(THF)n and (η5-C9H7)2YCl(THF)m [59 ]. These compounds have proven difficult to isolate, requiring several recrystallizations to obtain pure material. On the other hand, {η5-C9H4-2,4,7-(CH3)3}2MX2Li(THF)2 may be prepared by treatment of 2,4,7-trimethylindenide with MX3(THF)n (M = Y, X = Cl; M = La, X = Br). These compounds can be cleanly converted to alkyls and amides by reaction with LiCH(SiMe3)2 or LiN(SiMe3)2, respectively.
124
3
Group 3 Metallocenes
3.3 Group 3 Metallocenes for the Activation of Hydrocarbons The activation of the C–H bonds of saturated and unsaturated hydrocarbons has been actively pursued in organometallic chemistry [60 ]. By virtue of their strong Lewis acidity and high electrophilicity, 14-electron, group 3 metallocene alkyl and hydride complexes are attractive candidates for activating these types of generally unreactive bonds. A variety of scandocenes and yttrocenes have been employed for hydrocarbon activation and considerable mechanistic information has been gained from the studies of their reactivity. Watson [61 ] has reported the reactivity of Cp*2YCH3 towards the unactivated sp3 bonds of methane, representing the first well-characterized example of methane activation by a homogeneous organometallic complex. Exchange between 13CH4 and the yttrium methyl group is observed upon heating the reaction to 70°C in cyclohexane-d12.
(3-12)
Other hydrocarbons such as ethane and propane react, but the corresponding organometallic products decompose via β-hydrogen elimination. Similarly, the monomeric permethylscandocene Cp*2ScR (R = hydride, alkyl, alkenyl, alkynyl, aryl) complexes have been studied for their reactivity towards the C–H bonds of saturated and unsaturated hydrocarbons [62 ]. The hydride derivative, Cp*2ScH, catalyzes H/D exchange reactions for a variety of C–H bonds. For reaction of the aromatic C–H bonds of benzene with Cp*2Sc-H, an upper limit of 100 seconds has been established as the half life of the reaction at 25°C. For toluene, the rate of exchange for aromatic positions is 60–70 times faster than exchange into the methyl group, and the ratio for aryl exchange remains approximately constant at 39:41:20 for ortho:meta:para hydrogens. Deuterium exchange into the C–H bonds of the Cp* ligand is also observed by heating Cp*2Sc-D in benzene-d6 under a D2 atmosphere. Methane and tetramethylsilane also undergo H/ D exchange with the kinetic products having one deuterium incorporated, e.g. CH3D. Hydrogen/deuterium exchange is also observed for 2° C–H bonds (e. g. for cyclopentane), but the rate is much slower. The overall reactivity of R–H bonds with Cp*2Sc–R' bonds is as follows: R = R' = H >> R = H, R' = alkyl >> R–H = sp C–H R' = alkyl > R–H = sp2 C–H, R' = alkyl > R–H = sp C–H, R' = alkyl > R– H = sp2 C–H R' = alkyl > R–H = sp3 C–H, R' = alkyl.
3.3 Group 3 Metallocenes for the Activation of Hydrocarbons
125
From these studies, the mechanistic features of scandocene reactivity toward the C–H bonds of hydrocarbons has been established. A concerted rather than a stepwise process is indicated. The rate of the reaction increases with the amount of s character in the reacting C–H σ bonds of the substrate. A four-center transition state is consistent with these findings.
(3-13)
A nondirectional “s” orbital provides better overlap and hence more bonding and stability in the transition state [62]. This picture has been supported theoretically by Goddard and Steigerwald [63 ], who concluded that the d orbitals of scandium are ideally suited to provide good overlap with all three hydrogen atoms in the transition state. The geometry of the transition state was found to be ‘kite shaped’ rather than square, favoring positioning of the H atom in the β (central) position relative to scandium. Therefore, the mechanistic picture for these concerted σ-bond metathesis reactions is one where the incoming H–H or C–H bond approaches the vacant 1a1 orbital of the scandocene alkyl, forms the aforementioned transition state, from which the newly formed H–H or C–H bond departs from the opposite side of the 1a1 orbital [62]. The relative bond dissociation energies (BDE) have been obtained for Sc–aryl and Sc–alkyl by the equilibration of the metallated complex, Cp*(η5,η1C5Me4CH2CH2CH3)Sc-C6H5 with Cp*(η5,η1-C5Me4CH2CH2CH2)Sc and C6H6 [64 ]. The Sc-C6H5 BDE was estimated as 16.6(3) kcal.mol–1 stronger than the Sc– CH2CH2CH2C5Me4 bond strength. In combination with other experiments [64], a lower limit was placed on scandium alkynyl bonds, such that BDE(Sc–alkynyl) – BDE(Sc–aryl) ≥ 29(5) kcal.mol–1. It was found that the early transition metal relative M–R BDEs correspond to the analogous relative R–H BDEs (M–alkynyl > M–aryl > M–alkyl), although the M–R BDEs increase more rapidly with increasing s character than their R–H counterparts. The polarity of Sc–aryl bonds has been probed through a study of the metallation reactions of two different substituted benzyl groups. A small dependence of Keq on the nature of X or Y suggests that the Sc–aryl bond is essentially covalent with a small ionic contribution (Eq. (3-14)) [64].
(3-14)
126
3
Group 3 Metallocenes
Teuben has found that Cp*2Y(H)(THF), like its scandium counterparts, catalyzes H/D exchange between sp3 C–H bonds and sp2 C–D bonds [34]. The half life for H/D scrambling between benzene-d6 and CH2(SiMe3)2 is about 27 h. It was also found that the methyl group in toluene is deuterated in an exchange reaction with benzene-d6. However, in contrast to the permethylscandocene system, scrambling of the deuterium into the Cp* ligands is not observed under these particular conditions (vide infra). The yttrium alkyl complex Cp*2YCH(SiMe3)2 reacts with mesitylene at elevated temperature to yield monomeric Cp*2Y-CH2(3,5-(CH3)2C6H3) (Eq. (3-15)).
(3-15)
With toluene, a mixture of benzyl, ortho, meta and para-tolylyttrocenes has been identified, indicating that C–H activation may involve both sp2- and sp3-hybridized C–H bonds. No reaction with methane was observed up to 100°C. The dimeric permethylyttrocene hydride, [Cp*2Y(µ2-H)]2 undergoes thermolysis in the presence of n-octane, cyclohexane or benzene to preferentially activate its pentamethylcyclopentadienyl C–H bonds, yielding the fulvene complex, Cp*2Y(µ2H)(µ2-η1,η5-CH2C5Me4)YCp* [65 ]. However, in deuterated aromatic solvent, H/D scrambling between solvent and yttrocene ligand C–H positions is observed, similar to the permethylscandocene system. Rapid removal of H2 gas allows for isolation of Cp*2Y-C6H5 and Cp*2Y-CH2C6H5, which are believed to be formed by σbond metathesis. For substituted aromatics, C6H5X (X = NMe2, OCH3, SCH3, CH2NMe2, PMe2, F, Cl, Br), ortho-metallation predominates [65].
3.4 Group 3 Metallocenes as Homogeneous Olefin Polymerization Catalysts It is now generally accepted that the active catalysts in group 4 metallocene-based Ziegler–Natta olefin polymerization systems are cationic, d0 alkyls, either as free 14-electron cations, weakly solvated or weakly ion paired with a large counter-anion [66 ]. The neutral, 14 electron, group 3 metallocene alkyl and hydride derivatives offer an opportunity to investigate the mechanisms of chain initiation, propagation and transfer without the complication of ill-defined cocatalysts such as methylalumoxane (MAO) and the possible complications of ion pairing effects. As a result, detailed studies into the rates, enantioselectivities and α agostic assistance
3.4 Group 3 Metallocenes as Homogeneous Olefin Polymerization Catalysts
127
for olefin insertion have been performed with group 3 metallocenes, lending considerable insight as to the essential features of a useful Ziegler–Natta olefin polymerization catalyst. Yttrocene complexes were identified as single component homogeneous olefin polymerization catalysts as early as the mid-1970s. The methyl-bridged dimeric yttrocene complexes, [(η5-C5H4R)2Y(µ2-CH3)]2 (R = H, CH3, SiMe3), are active catalysts for ethylene polymerization [67 ]. Although the dimer is the only organometallic species observed during polymerization, monomeric (η5-C5H4R)2YCH3 is believed to be the active species. Rapid scrambling of bridging alkyl groups between yttrium centers is observed when [(η5-C5H4R)2Y(µ2-CH3)]2 and [(η5-C5H4R)2Y(µ2CH2CH2CH2CH3)]2 are mixed, in accord with a monomer/dimer equilibrium. Evans and coworkers have reported the reaction of α-olefins (ethylene, propylene) with [(η5-C5H4R)2YH(THF)]2 (R = H, CH3), producing the corresponding alkyl complexes [68 ]. In the presence of excess ethylene, formation of the dimeric, µ-n-butyl complex is observed, arising from olefin insertion into the yttrium–alkyl bond. Propylene undergoes 1,2-addition of the Y–H moiety to yield the less hindered primary alkyl product. Linked, mixed-ring yttrocene dimeric hydrides, [{(η5C5Me4)(η5-C5H4)SiMe2}Y(µ2-H)]2 have been prepared where the Cp ligand array spans two metal centers [69 ]. Although ethylene is polymerized, reaction with αolefins slowly yields the µ-hydrido, µ-alkyl complex, but no polymer or further insertion products are observed.
(3-16)
Permethylscandocene complexes Cp*2ScR (R = H, alkyl) insert ethylene rapidly at low temperature (–80°C) without competition from chain termination via β-hydrogen elimination; hence, living ethylene polymerization may be achieved [70 ]. The kinetics of ethylene insertion have been measured. The relative rates of ethylene insertion are reflected in part by the strength of the Sc–C bond, where stronger bonds result in slower rates. The order of decreasing rate of olefin insertion is Sc– H >> Sc–CH2CH2CH3 ≥ Sc–CH2(CH2)nCH3 (n ≥ 2) > Sc–CH3 > Sc–CH2CH3 > Sc–C6H5 [70]. The increased rate of insertion for the hydride derivative is likely due to greater orbital overlap (nondirectional 1s valence orbital) in the transition state, while the greater rate for the propyl species relative to the methyl is presumably due to a weaker Sc–CH2CH2CH3 relative to Sc–CH3 [70,71 ]. The anomalously sluggish rate of ethylene insertion into the scandium–ethyl bond is a result of βagostic interaction, which further stabilizes the ground state. Such an interaction is
128
3
Group 3 Metallocenes
not observed for the scandium–propyl complex, presumably since such an interaction is sterically unfavorable, since the methyl group of the β-carbon would be directed toward one of the (η5-C5Me5) rings (Eq. 3-17).
(3-17)
The transition states for both β-hydrogen elimination, and hence also for ethylene insertion have been probed by systematically varying the substituent at the β-carbon. For the scandium phenylethyl series a linear free energy correlation is found [70], indicating that the transition state for these processes is quite polar with the electropositive scandium center abstracting hydride (and by extension alkyl (Rδ-) for insertion/β-alkyl elimination) of the positively charged β−carbon atom. The linked cyclopentadienyl complexes [DpScH]2 and OpSc(H)(PMe3) rapidly polymerize ethylene, whereas reaction with higher α-olefins (propylene, 1-butene) yield head-to-tail dimers [72 ,73 ]. Addition of either cis- or trans-2-butene to either (DpScH)2 or OpSc(H)(PMe3) results in isomerization about the carbon–carbon bond as well as the formation of head-to-tail dimer [74 ]. Moreover, α,ω-dienes are cleanly cyclized to the corresponding methylenecycloalkane [72]. Five, six, seven, eight or nine-membered rings can be synthesized from the appropriate α,ω-diene in 85–99% yields. For the smaller methylenecycloalkanes, methylenecyclobutane and methylenecyclopropane, the reverse reactions, exothermic ring openings to 1,4pentadiene and butadiene, respectively, are catalyzed (Scheme 3-7). The structure of the transition state for chain propagation of scandocene and yttrocene catalysts has been investigated by examining the possibility of an α agostic C– H interaction which assists olefin insertion, thus supporting a ‘modified Green–Rooney’ mechanism [75 ]. [DpScH]2, OpSc(H)(PMe3) and [(Cp*SiNR)Sc(PMe3)]2(µ2-H2) are active catalysts for hydrocyclization of 1,5-hexadiene to methylcyclopentane. By employing ‘deuterium isotopic perturbation of stereochemistry’, originally developed by Grubbs [76 ], this mechanistic probe has been adapted for the scandium systems. Addition of an achiral α,ω-diene, such as trans, trans-1,6-d2-1,5-hexadiene to achiral OpScH yields precisely a 50:50 mixture of R- and S-6-hexenyl-1-d16-scandium complexes. Due to ring strain there should be a preference for the cis ring fusion in the transition state for olefin insertion. Face selection for the insertion of the pendant olefin then depends on whether H or D occupies the α agostic position. The expected preference for H to occupy the bridging position leads to an expected excess of the R,R (trans) (and S,S (trans)) over the R,S (cis) (and S,R cis)) products, if an α agostic interaction assists olefin insertion (see Scheme 3-9, below).
3.4 Group 3 Metallocenes as Homogeneous Olefin Polymerization Catalysts
129
Scheme 3-7
Analysis of the product mixture by 2H{1H} NMR yields an average trans:cis ratio of 1.226(12):1 at 25°C [77 ]. Similarly, isotopic perturbation of stereochemistry is also observed for the hydrocyclization of deuterated 1,5-hexadienes with other scandium catalysts [78 ]. [DpScH]2 hydrocyclizes trans, trans-1,6-d2-1,5-hexadiene with a trans:cis ratio of 1.203(7):1, while [(Cp*SiNR)Sc(PMe3)(µ2-H2)]2 produces d2-methylcyclopentanes with a trans:cis ratio of 1.209(15):1. These results provide evidence for the ‘modified Rooney–Green’ pathway for chain propagation with these Ziegler–Natta systems. Moreover, these results provide an explanation for the requirement that the active bis(cyclopentadienyl)metal catalysts be 14-electron derivatives with two vacant orbitals: one to accommodate the incoming olefin, another for the α agostic interaction [66,76] (Scheme 3-8). The mechanism of olefin insertion and chain transfer processes by both β-hydride and β-methyl elimination for Ziegler–Natta polymerizations has been examined with OpSc(H)(PMe3) [79 ]. Reaction of isobutene with OpSc(H)(PMe3) forms OpSc(CH3)(PMe3) along with isobutene, 2-methylpentane, isobutane, 2-methyl-1pentene, propane, and n-pentane. The organic products are formed from a series of reactions involving olefin insertion, β-methyl elimination and β-hydride elimination. It was established that β-H elimination proceeds faster than β-CH3 elimination, which proceeds until only 2-methyl-1-alkenes and the predominant organometallic product, OpSc(CH3)(PMe3) remain. Alkane products result from σ-bond metathesis reactions of the C–H bonds of PMe3 and the Sc–C bonds of scandium alkyls. However, β-ethyl elimination is not observed for the related 2-ethylbutyl derivative, OpSc{CH2CH(C2H5)CH2CH3}(PMe3), which is obtained from reaction of 2-ethyl-1-butene with OpSc(H)(PMe3).
130
3
Group 3 Metallocenes
Scheme 3-8
Linked cyclopentadienyl-amido scandium systems have proven to be active catalysts for olefin polymerization. Treatment of [(Cp*SiNR)(PMe3)Sc(µ2-H)]2 with two equivalents of propylene at low temperature yields the phosphine-free alkyl, [(Cp*SiNR)Sc(µ-CH2CH2CH3)]2, which is an active catalyst for the polymerization of ethylene and α-olefins [80 ,81 ]. Kinetic analysis of 1-pentene polymerization in addition to 13C and 31P NMR analysis reveals that the active catalyst for this system is the 12-electron scandium alkyl [(Cp*SiNR)ScCH2CHRP] (P = growing pol-
3.4 Group 3 Metallocenes as Homogeneous Olefin Polymerization Catalysts
131
ymer chain). Although polymerization has been achieved, the rates of polymerization and the molecular weights of the polymers are rather low: polypentene (25¡C, neat 1-pentene): Mn = 6,000; PDI = 1.5; polypropylene (25°C, 25 vol. % in toluene): Mn = 9,600; PDI = 1.8; polypropylene (–9°C, 25 vol% toluene): Mn = 16,500; PDI = 1.7. 13C NMR analysis of the resulting polymers indicates a predominantly atactic structure. Isospecific polymerization of α-olefins has been achieved with the single component, [rac-Me2Si(2-SiMe3-4-CMe3C5H2)Y-R] (‘BpY-R’) catalysts [82 ]. The ligand array is designed such that, upon coordination, only the desired racemic metallocene is formed thereby eliminating tedious separation of the unwanted meso isomer. Inspection of the X-ray crystal structure [83 ] of rac-BpYCl2Li(THF)2 confirms expectations that unfavorable steric interactions between the SiMe3 groups in the narrow portion of the Cp–M–Cp wedge are avoided only for the racemic isomer (Scheme 3-9).
Scheme 3-9
Ethylene is rapidly polymerized by BpYCH(SiMe3)2 producing moderately high molecular weight polyethylene (Mn = 720,000, PDI = 2.0, Tm = 137°C). Solutions of [BpY(µ2-H)]2 polymerize α-olefins slowly to form moderately high molecular
132
3
Group 3 Metallocenes
weight polymers. Polymerization is slow, presumably due to inactivity of the 16electron hydride dimer; thus, slow dissociation into the active monomers dictates low activity. For propylene polymerization, 13C NMR pentad analysis reveals that the mmmm resonance accounts for >99% of the methyl resonances, while chain end analysis by 1H NMR shows <0.1% 2,1 mis-insertions and both vinyl and vinylidene end groups, indicative of both β hydrogen and β methyl elimination for chain termination. Similarly, polymers generated from 1-butene, 1-pentene and 1hexene are also highly isotactic and contain very few regiomistakes. Yasuda [84 ] has modified the Bp ligand array by replacement of the tert-butyl groups with Si(CH3)2(CMe3) groups and has prepared the corresponding {CH(SiMe3)2} and hydride dimers. Polymerization of 1-pentene and 1-hexene produced polymers with Mn = 1 x 104 and high degrees of isotacticity. Although coordination of the Bp ligand avoids formation of the meso isomer, the racemo isomer naturally contains both enantiomers. Resolution of enantiomers is necessary to allow for isolation of an enantiopure metallocene. A chiral auxiliary attached to the silylene linking group with groups extending toward the [SiMe3] groups at the 2-positions of the Cp rings forces the formation of only one enantiomer. Accordingly, a new ligand, {[C5H2-2-SiMe3-4-CMe3)2Si(OC10H6C10H6O)}2– [(BnBp)2–], has been prepared for this purpose and coordinated to yttrium [85 ]. Use of optically pure ligand yields optically pure metallocene, where the (R)-(+)1,1'-bi-2-naphthol directs the formation of the (S)-yttrocene and (S)-(-)-1,1'-bi-2naphthol directs formation of the (R)-yttrocene (Scheme 3-10).
Scheme 3-10
The alkyl derivative, (BnBp)YCH(SiMe3)2, is an active catalyst for both olefin polymerization and hydrogenation [86 ]. Hydrogenolysis of (BnBp)YSiCH(SiMe3)2 cleanly yields the hydride dimer [(BnBp)Y(µ2-H)]2 which can be isolated as a white solid. Immediately upon addition of H2 to the alkyl complex the formation of two different yttrium hydride species is apparent. From NMR investigations, it was found that initially both heterochiral and homochiral hydride dimers are formed,
3.5 Reactions of Group 3 Metallocenes with Acetylenes
133
then over time the heterochiral dimers are converted to the more thermodynamically stable homochiral form [85]. The same transformation occurs in the racemic [BpY(µ2-H)]2 system as well. Significantly, no monomeric hydride species (no doublet in 1H NMR spectrum) is observed, suggesting that monomeric hydrides rapidly dimerize to form much more stable dimers. This behavior greatly reduces the activity of these systems for olefin hydrogenation and polymerization. Development of a resolved, enantiopure group 3 metallocene offers an unprecedented opportunity to determine the stereoselectivity of additions to carbon–carbon double bonds. Optically active deutero-1-pentenes have been synthesized and used to evaluate the stereoselectivity of Y–H and Y–pentyl additions to optically pure (BnBp)Y–R complexes [87 ]. Insertion into metal–pentyl bonds proceeds with a much higher facial selectivity (≥40:1) than does insertion into metal–hydride bonds (2:1). Moreover, insertions into metal hydride and metal–n-pentyl bonds proceed with opposite enantioselectivities. Cyclopentadienyl–alkoxide yttrocene systems have also been employed as olefin polymerization catalysts. Terminal alkenes (ethylene, propylene, 1-butene) react with [Cp*(OAr)Y(µ2-H)]2 (Ar = C6H3(CMe3)2) to produce the µ2-n-alkyl species trans-[{Cp*(OAr)Y}2(µ2-H)(µ2-CH2CH2R) [88 ]. Similar alkyls are produced with longer chain alkyls (e. g. from 1-hexene), but no insertion or isomerization occurs with internal olefins (i. e. trans-3-hexene). For the polymerization of ethylene, it is proposed that the active catalyst is the undetectable, monomeric [Cp*(OAr)YX] (X = H, CH2R). More recently, group 3 metallocene catalysts have been investigated for their reactivity towards cyclic functionalized olefins. The catalytic oligomerization of 2cyclopentene-1-ones has also been investigated with the group III hydrides, [Cp*2La(µ2-H)]2 and [Cp*2Y(µ2-H)]2 [89 ]. Isolation of oligomers reveals that propagation occurs via 1,4 addition, a conclusion also supported by the structure of the isolated trimer. Approximately equimolar amounts of cis and trans isomers are formed, indicating that the process is not stereoselective. Attempts to isolate cyclic intermediates have not been successful, therefore mechanistic details of this process have not yet been elucidated.
3.5 Reactions of Group 3 Metallocenes with Acetylenes Catalytic reactions of group 3 metallocenes with alkynes have received much less attention as compared with their reactions with olefins, although several well documented examples have been reported. Reaction of acetylene with Cp*2ScR (R =H, alkyl, aryl, alkenyl, alkynyl, amide) at low temperature (–78°C) yields R–H and Cp*2Sc–C≡CH (Eq. (3-18)) [90 ].
134
3
Group 3 Metallocenes
(3-18)
The scandocene acetylide derivative then reacts with excess acetylene to produce polyacetylene. The acetylide also cleanly decomposes to give Cp*2Sc–C≡C– ScCp*2, a compound that is cleanly generated by addition of 0.5 equivalents of acetylene to Cp*2ScR. Permethylscandocene alkyls promote the rapid dimerization of terminal alkynes selectively to 2,4-disubstituted 1-buten-3-ynes [62]. Likewise, Teuben and coworkers report that Cp*2Y-CH(SiMe3)2 also serves as a precatalyst for selective head-totail dimerization of aliphatic alkynes (Eq. (3-19)) [91 ].
(3-19)
With the permethylyttrocene system mixtures are obtained for phenylacetylene and (trimethylsilyl)acetylene, indicating that both steric and electronic effects are significant in determining the stereochemical outcome of the reaction. For both systems rapid σ-bond metathesis between the metal alkyl (or alkynyl) and the sp-hybridized C–H bonds of the acetylenes appears to occur to the exclusion of insertion, since higher oligomers are not detected. Evans reports that reaction of [(C5H4R)2Y(µ2-H)(THF)]2 (R = H, CH3) with terminal alkynes does not involve insertion into the Y–H bond to yield an alkenyl derivative, but rather σ-bond metathesis at the terminal C–H bond of the acetylene occurs to yield the alkynyl derivative and dihydrogen [92 ]. This reaction is taken as evidence for the hydridic nature of this class of metal–hydrogen bonds. Reaction of [(C5H4R)2Y(µ2-H)(THF)]2 (R = H, CH3) with internal alkynes does however proceed to the corresponding alkenyl complexes, where the addition of the Y–H bond is exclusively cis. The yttrocene alkyls, Cp*2YCH(SiMe3)2 and Cp*2Y(CH3)(THF), react with a variety of terminal acetylenes in Et2O to yield Cp*2Y–C≡CR·solv (solv = Et2O; THF) [93 ]. Complexation of solvent is believed to be a result of the relatively small, rodlike nature of the acetylide ligand, allowing sufficient space around the
3.6 Reactions of Group 3 Metallocenes with Group 6 Compounds
135
metal for solvent coordination. In the absence of ethereal solvents, dimerization of the acetylene is observed at a rate of 5,400 turnovers/hr at 20°C. Cyclodimerization of disubstituted alkynes has been reported [94 ]. The lanthanum alkyl, Cp*2LaCH(SiMe3)2, is an active catalyst for the cyclodimerization of 2alkynes of the form, CH3–C≡C–R (R = CH3, CH2CH3, CH2CH2CH3), to 1,2-disubstituted, 3-alkylidenecyclobutenes (Eq. (3-20)).
(3-20)
NMR investigations indicate that the first step in the reaction is the activation of the C–H bond in the acetylene methyl group, yielding Cp*2La–CH2C≡CCH3. The yttrium analog is unreactive under these conditions, but Cp*2YCH2C≡CCH3 has been prepared from the reaction of 2-butyne with [Cp*2Y(µ2-H)]2.
3.6 Reactions of Group 3 Metallocenes with Group 6 Compounds The combination of well-defined organometallic complexes with the chemistry of heavier group 16 elements (Se, Te) offers the opportunity to develop new procedures for the synthesis of binary materials. It is thus of considerable interest to isolate intermediates in the process of going from the molecular level to the bulk material. Group 3 metallocene alkyl complexes are very promising reagents in this regard. Reaction of the permethylscandocene alkyl, Cp*2ScCH2SiMe3, with elemental tellurium results in the formation of Cp*2ScTeCH2SiMe3 [95 ,96 ]. This compound can be prepared by reaction with TeP(CH2CH2CH2CH3)3, but complete removal of P(CH2CH2CH2CH3)3 is difficult and hence, the elemental tellurium route is preferred. Insertions of tellurium into other scandium alkyls (R = CH3, CH2Ph, CH2CH2CMe3) has also been observed. Heating of the scandium tellurate leads to extrusion of TeR2 and formation of the tellurium-bridged dimer, Cp*2Sc-TeScCp*2. Interestingly, the telluride dimer can be photochemically reverted to the scandium alkyltellurate by addition of TeR2. The corresponding selenides have also been prepared by similar methodology [96]. Through a series of detailed experiments [96], the general mechanistic features of these systems have been explored. Retention of configuration is observed for the
136
3
Group 3 Metallocenes
thermal extrusion reaction, suggesting a concerted reaction pathway [97 ]. Photoaddition of TeR2 to Cp*2Sc-Te-ScCp*2 likely occurs through an excited state with a weaker Sc–Te bond that allows the dimer to bend and become susceptible to attack by TeR2. The trapping agent employed in the photoreaction of the Cp*2Sc-TeScCp*2 process behaves as a Lewis base, since no reaction occurs in the presence of H-Sn(CH2CH2CH2CH3)3 or Me3SnSnMe3 [96]. Reaction of the excited state of Cp*2Sc-Te-ScCp*2 with TeR2 results in formation of one equivalent of the scandocenetellurate in addition to the Cp*2ScTe./R. radical pair, which is evidenced by the cyclization in the case of R = 5-hexenyl alkyl. Other scandocenes behave similarly to the permethylscandocene system. Reaction of elemental tellurium with DpScCH2SiMe3 results in the formation of the scandium tellurate DpScTeR [98]. The dimeric µ2-telluride DpSc-Te-ScDp can be synthesized from either extrusion of TeR2 from DpSc-TeR or from direct reaction of [DpSc(µ2-H)]2 with 0.5 equivalents of TeP(CH2CH2CH2CH3)3. Addition of phosphine to the telluride dimer results in the formation of Dp(PMe3)Sc-TeSc(PMe3)Dp, which has been characterized crystallographically, revealing the linearity of the Sc–Te–Sc linkage [98 ]. In addition to organoscandium–tellurium complexes, several selenide and sulfide derivatives of yttrium have also been reported [99 ]. Reaction of [(η5C5H4CMe3)2Y(µ2-CH3)]2 with two equivalents of either RSSR (R = C6H5, CMe3, CH3CH2CH2CH2, CH2Ph) or PhSeSePh produce dimeric yttrium organosulfides or organoselenides [(η5-C5H4CMe3)2Y(µ2-ER)]2 (E = S, Se). The reactions are quantitative and several of these compounds have been characterized crystallographically.
3.7 Reactions of Group 3 Metallocenes with Other Small Molecules Group 3 metallocenes are coordinatively unsaturated and hence are voracious Lewis acids. As a result, extremely strong bonds between the metal center and hard ligand atoms such as O, N, F and Cl are expected. Thus, the prospect for catalytic reactions with small molecules having heteroatoms are poor. Nonetheless, the reactivity of group 3 metallocenes with heteroatom-containing substrates has been extensively developed and several examples of catalytic chemistry have been reported. The addition of Sc–H and Sc–C bonds to carbon monoxide, nitriles and carbon dioxide have been examined using permethylscandocene derivatives. The chemistry observed in these systems closely parallels the reactivity patterns of group 4 permethylmetallocenes [100 ]. Insertion of CO and CO2 into Sc–aryl bonds occurs readily (Scheme 3-11) [13].
3.7 Reactions of Group 3 Metallocenes with Other Small Molecules
137
Scheme 3-11 The permethylscandocene hydride, Cp*2Sc-H, undergoes facile addition of its Sc– H bond to the carbonyl ligands of transition metal carbonyls to yield scandoxycarbenes (e. g. Cp(CO)M=CHO-ScCp*2 (M = Co, Rh) and Cp2M=CHO-ScCp*2(M = Mo, W) [100]. This reaction is highly dependent on steric factors. Bulkier substituents such as those of Cp*2Sc-CH2CH2C6H5, Cp*2Sc-N(CH3)2 and Cp*2ScC6H5 do not react with Cp2M–C…O (M= Mo, W). However, the less sterically encumbered CpM(CO)2 system, (M = Co, Rh) produces the scandoxycarbene product in good yield. Insertion of nitriles into Sc–R bonds of Cp*2Sc-R (R = H, CH3) proceeds smoothly to form the alkylideneamido complex, which undergoes further additions of the nitrile (Scheme 3-12) [101 ].
Scheme 3-12
138
3
Group 3 Metallocenes
Permethylscandocene hydride also catalyzes the hydrogenation of tert-butyl cyanide, yielding neopentylamine (Eq. (3-21)).
(3-21) Reaction of one equivalent of hydrazine with Cp*2ScCH3 yields the hydrazido(1–) complex, Cp*2ScN(H)NH2 [102 ]. The hydrazido complex reacts with acetonitrile to form scandocyclic Cp*2ScN(H)CCH3NNH2. The mechanism of the cyclization reaction, explored through the use of 15N labeled acetonitrile, likely proceeds through insertion of the acetonitrile into the Sc–N bond of Cp*2ScNHNH2, followed by tautomerization to form the observed product (Scheme 3-13).
Scheme 3-13
Tilley and co-workers [103] have reported the synthesis of silyl derivatives of scandium from reaction of [Cp2ScCl]2 and lithium silyls [103 ]. These react with CO2, rapidly forming dimerization products (Eq. (3-22)).
(3-22)
3.7 Reactions of Group 3 Metallocenes with Other Small Molecules
139
The dimer can be independently prepared from reaction of Cp2Sc(SiR3)(THF) and HO2CSi(SiMe3)3 [104 ]. Recrystallization of the scandium carboxylate leads to extrusion of CO, forming the siloxide, [Cp2ScOSi(SiMe3)3]x. However, upon heating of the dimer, {Cp2Sc[µ-O2CSi(SiMe3)3]}2, decomposition to the siloxide is not observed, and no structure has definitely been assigned. Reaction of the scandocene silyls with CO and isocyanides have also been reported [105 ]. Carbonylation of Cp2Sc[Si(SiMe3)3](THF) forms the ene–diolate, methyltetrahydrofuran adduct (Eq. (3-23)).
(3-23)
The isoelectronic isocyanate, C≡NC6H3Me2, reacts with Cp2Sc[Si(SiMe3)3](THF) to yield the η2-iminosilylacyl, which undergoes further addition of C≡NC6H3Me2 ultimately to yield the bicyclic structure 1:
1 The mechanism for this rearrangement is postulated to involve a silylene intermediate, although attempts to isolate or characterize this species have been unsuccessful. A variety of yttrocene complexes has been studied for their reactivity with pyridines. Reaction of the yttrium alkyls Cp*2YCH(SiMe3)2 and Cp*2Y(CH3)(THF) with pyridine results in metallation at the α position [106 ]. The mechanism for the reaction is proposed to involve precoordination of the pyridine, followed by hydrogen transfer from the ortho position to the alkyl group by σ-bond metathesis reaction. Coordination of THF to the pyridyl complexes is also observed, forming the 18-electron Cp*2Y(η2-NC5H4)(THF) adduct. Permethylyttrocene hydride, [Cp*2Y(µ2-H)]2, also reacts with pyridine selectively at the ortho C–H position to yield Cp*2Y(η2-NC5H4) [107 ]. The pyridyl complex reacts with H2 to give the hydride addition product, Cp*2Y(NC5H6). Addition of ethylene and propylene to the pyridyl complexes yields the monoinsertion yttrocyclic products, while acetylenes
140
3
Group 3 Metallocenes
undergo C–H activation, forming yttrium acetylides. Reaction with carbon monoxide produces [Cp*2Y]2{µ2-η2:η2-OC(2-NC6H4)2}, which has been identified by Xray crystallography (Scheme 3-14).
Scheme 3-14
In the presence of excess ethylene, pyridine can be converted to 2-ethylpyridine using Cp*2Y(η2-NC5H4) as a catalyst. Trace amounts of 2-n-butylpyridine and 2-nhexylpyridine are also formed, indicating that multiple insertions are possible [107]. The yttrium hydrides, [(η5-C5H4R)2Y(µ2-H)(THF)]2 (R = H, CH3), react with tert-butyl isocyanide in THF to yield [(η5-C5H4R)2Y(µ2-CHNCMe3)]2 (Eq. (3-24)) [108 ].
3.7 Reactions of Group 3 Metallocenes with Other Small Molecules
141
(3-24)
Spectroscopic and crystallographic data [109 ] suggest that the structure of the complex is a formimidoyl dimer where nitrogen atoms donate a lone pair of electrons to the yttrium center. Reaction of the chloro-bridged dimer, [Cp2Y(µ2-Cl)]2, with lithium N,N-dimethylbenzylamine produces Cp2Y(C6H4-2-CH2NMe2), which has been characterized crystallographically [110 ]. The crystal structure reveals the yttrium center is σ-bound to the ortho carbon and the nitrogen atoms of the N,Ndimethylbenzylamine ligand (Eq. (3-25)).
(3-25)
Reaction of [Cp2Y(µ2-H)(THF)]2 with CH2O or CH3OH forms the dicyclopentadienylyttrium methoxide complex [111 ]. This compound also can be prepared directly from reaction of [Cp2Y(µ2-H)(THF)]2 with KOCH3 at low temperature (–78°C). At higher temperatures the polymetallic oxide, Cp5Y5(µ-OCH3)4(µ3-OCH3)4(µ5-O), is isolated from the reaction mixture. However, reaction of [Cp2Y(µ2-H)(THF)]2 with NaOCH3 forms [Cp2Y(µ-OCH3)]2 in 80–95% yield [112 ]. Further investigation into the KOCH3 reaction reveals that [Cp2Y(µ2-OCH3)]2 can be formed from [Cp2Y(µ2-H)(THF)]2 when the KOCH3 is prepared from reaction of KN(SiMe3)2 with dry methanol. A byproduct of this reaction is the trimetallic compound, {[Cp2Y(µ2-OCH3)]2[Cp2Y](µ3-O)}, whose formation is strongly dependent on the amount of KOCH3 present. The pentametallic complex is also isolated from this reaction, and its formation is dependent on the amount of solvated CH3OH. The yttrium polyhydride, [[Cp2Y(µ2-H)]3(µ3-H)[Li(THF)4] [113 ] reacts with one equivalent of CH3OH to form [[Cp2Y(µ2-OCH3)][Cp2Y(µ2-H)]2(µ3H)][Li(THF)4] [114 ]. Reaction with two equivalents of CH3OH forms [[Cp2Y(µ2OCH3)]2[Cp2Y(µ2-H)](µ3-H)][Li(THF)4], which can also be formed from reaction of [[Cp2Y(µ2-OCH3)][Cp2Y(µ2-H)]2(µ3-H)][Li(THF)4] with one equivalent of CH3OH. It is noteworthy that in all cases, the central µ3-hydride core does not react. Yttrium–enolate complexes, [(η5-C5H4R)2Y(µ2-OCH=CH2)]2, have been synthesized from the reaction of [(η5-C5H4R)2Y(µ2-Cl)]2 (R = H, CH3) and
142
3
Group 3 Metallocenes
LiOCH=CH2 [115 ]. Crystallographic studies reveal that the complex is a dimer with bridging enolate oxygen atoms spanning the two yttrium centers. Thermolysis of a sealed NMR tube containing the yttrium alkyl, Cp2Y(CH2SiMe3)(THF), in C6D6 forms the yttrium enolate, [Cp2Y(µ2-OCH=CH2)]2, generating an equivalent of SiMe4. However, thermolysis of the methyl derivatives, (η5C5H4R)2Y(CH3)(THF), under the same conditions forms (η5-C5H4R)3Y with no evidence for the yttrium enolates. The difference in reactivity is attributed to steric factors, where more crowded complexes have a higher propensity to undergo thermolysis with coordinated THF to form the corresponding enolate complexes. Teuben [106] has reported the reactivity of Cp*2YCH(SiMe3)2 and Cp*2Y(CH3)(THF) toward a variety of complexes with ‘active’ hydrogen containing molecules. Reaction with HCl produces [Cp*Y(µ2-Cl)]2 and Cp*2Y(Cl)(THF), while reaction with acetylacetone (acacH) yields Cp*2Y(acac). Reaction of Cp*2YCH(SiMe3)2 and Cp*2Y(CH3)(THF) with alcohols forms the yttrocene alkoxides, Cp*2Y(OR)(Et2O). The mechanism for this reaction likely involves protonolysis of the Y–C bonds. Similarly, [Cp*2Y(µ2-H)]2 reacts with 2 equivalents of furan in pentane/THF to form Cp*2Y(2-C4H3O)(THF), which has been characterized crystallographically (Eq. (3-26)) [116 ]. ((Equation 3-26 here)) (3-26) Cp3Y(THF) reacts with NaOH in THF to form NaCp and Cp2Y(OH)(THF) [117 ]. Addition of hexachlorocyclotriphosphazene (NPCl2)x (80%, x = 3; 20%, x = 4) to remove free NaCp from the reaction mixture allows for efficient isolation of Cp2Y(OH)(THF). The bridging hydroxide dimer, [Cp2Y(µ2-OH)]2(C6H5C…CC6H5), is prepared by hydrolysis of Cp2Y(CMe3)(THF) in the presence of diphenylacetylene. It has been characterized crystallographically. Mixed metal yttrium–rhenium polyhydride complexes have been prepared from reaction of Cp2Y(THF)(CH3) and ReH7(PPh3)2 or ReH5(PMe2Ph)3, eliminating CH4 forming Cp2Y(THF)Re(H)6(PPh3)2 and Cp2Y(THF)Re(H)4(PMe2Ph)3, respectively [118 ]. p2Y(THF)Re(H)6(PPh3)2 is unreactive toward a CO, CO2 and PhC≡CPh, but is active for oligomerization of ethylene, propylene and styrene. Trimetallic complexes have been prepared from the reaction of Cp2Y(THF)(CH3) and Re2H8(PMe2Ph)4 yielding Cp2Y(THF)H7Re2(PMe2Ph)4 [118].
3.8 Applications of Group 3 Metallocenes in Organic Synthesis Although group 3 metallocenes have proven to be efficient catalysts for a variety of organic transformations, their application toward organic synthesis has remained somewhat limited. Attempts to employ group 3 metallocenes as selective hydrogenation catalysts have been explored by Molander. Using 1–mol%
3.8 Applications of Group 3 Metallocenes in Organic Synthesis
143
Cp*2Y(CH3)(THF) as a catalyst, a variety of diolefins are hydrogenated to the corresponding alkane in yields of 70% to 99% [119 ]. Disubstituted alkenes such as limonene, 5-methylenebicyclo[2.2.1]hept-2-ene, or bicycloheptenes cannot be reduced. As a result, the yttrocene catalyst offers the opportunity to selectively hydrogenate terminal olefins in the presence of other types of substituted olefins. Interestingly, ether groups are tolerated by the reaction. Allylic ethers which are rapidly hydrogenated using late metal systems remain intact. For α,ω-dienes, in which two olefins are differentiated by allylic substitution on one of the alkenes, good selectivity is observed. Cp*2Y(CH3)(THF) also catalyzes the hydrocyclization of α,ω-dienes diastereoselectively with excellent yields [120 ]. In addition to ethers, acetal and dithioacetal functionalities are tolerated by the catalyst. Hydrogenation and isomerization of 1-hexene with a variety of yttrocene chloride complexes and NaH has also been reported [121 ]. Chiral yttrocenes, such as the chiral yttrocene hydride precursor rac-Me2Si[η5C5H3SiMe3][η5-C5H3)(–)-menthyl]YCH(SiMe3)2, catalyze the asymmetric hydrogenation of unfunctionalized olefins. Olefins such as α-ethylstyrene and 2-ethyl-1hexene are hydrogenated with enantiomeric excess of 25% and <1% respectively [122 ]. Reductive cyclization of 1,5-dienes is also observed with this system, with olefins such as 1,5-hexadiene undergoing quantitative cyclization. Regio- and stereoselective hydroamination/cyclization of amino olefins such as 1-aminopent-4ene, 2-amino-hex-5-ene, and dimethyl-1-aminohex-5-ene have been reported [123 ] with C1 symmetric yttrocene catalyst precursors such as rac-Me2Si[η5C5H3SiMe3][η5-C5H3)R']YCH(SiMe3)2 (Eq. (3-27)) [124 ]. Enantiomeric excess as high as 69% has been observed in these systems.
(3-27)
Organoyttrium catalysts are also efficient for the hydrosilation of olefins. Cp*2YCH(SiMe3)2 is activated by reaction with PhSiH3, ultimately generating an yttrium hydride that catalyzes the hydrosilylation of a variety of substituted and unsubstituted olefins [125 ]. A combination of the cyclooligomerization process and hydrosilation methodology results in the cyclization and silation of 3-phenyl-1,5hexadiene in the presence of 5 mol percent of Cp*2YCH(SiMe3)2 [126 ]. Improved results are obtained when Cp*2Y(CH3)(THF) is substituted for the yttrium alkyl (Eq. (3-28)).
144
3
Group 3 Metallocenes
(3-28)
The yttrocene systems result in high regio- and stereoselectivity in addition to tolerance of certain functional groups such as ethers, thioacetals and tertiary amines. The phenylsilane products of the reaction can conveniently be converted to alcohols. This methodology has been applied in the total synthesis of (±)-epilupidine, in which the key step is the selective preparation of a polycyclic ring system through the use of an yttrocene catalyzed cyclization/silylation (Scheme 3-15) [127 ].
Scheme 3-15
Organolanthanocenes have been employed in a variety of organic transformations as well. Cp*2LaR complexes (R = H, CH(SiMe3)2) are active catalyst precursors for the hydroboration of terminal olefins in the presence of catecholborane [128 ]. The reaction is useful for both terminal or substituted olefins. The regioselectively is greater than 98% for the anti-Markovnikov product (Eq. (3-29)).
(3-29)
3.9 Photophysics of Group 3 Metallocenes
145
3.9 Photophysics of Group 3 Metallocenes Inorganic and organometallic complexes that emit at room temperature have been studied extensively in hopes of understanding the fundamental properties of their excited states. Group 3 metallocenes offer a new type of luminescent complex, since classical inorganic photochemistry has focused mainly on second or third row complexes with heterocyclic amine ligands. For Cp*2ScCl, three UV–vis absorptions are present in solution and a fourth transition can be observed in the excitation spectrum [129 ]. From studies of the excitation spectrum, the lowest energy state is taken to be the emissive state. The luminescent lifetimes for permethylscandocene chloride in the solid state at 298K are 2.0 µs, 15.0 µs and 6.5 µs. The long lifetimes are believed to be a result of phosphorescence which arises from an emission from the triplet excited state. The observed transitions in the optical spectrum are assigned to the excitations from the p-Cl nonbonding HOMO to the 1a1, 1b2 and the 2a1 orbitals on scandium. In addition to the photophysical properties of permethylscandocenes, calculated molecular orbital diagrams of some of the compounds have also been reported [130 ]. Calculations indicate that the long-lived emissions of these complexes are due to LMCT from the Cp* ligands to the scandium center. It appears that the lone pairs of the halogen are required for the emission to occur, since no emission is observed for either Cp*2ScCH3 or Cp*2ScCH2Ph, although the molecular orbital diagrams are similar to the chloride complex. In addition to calculations for observed emission and visible spectra, Group 3 metallocenes have also been the subject of other theoretical calculations. The general features of the bonding in metallocene complexes have been presented [131 ], but details specific to scandium and yttrium have since appeared. Schleyer [132 ] has performed ab initio calculations for the hypothetical [Cp2M]+ (M = Sc, La) and found that these species prefer bent geometries due to a small σ-contribution, even though that represents only a small part of the overall covalent bonding. The insertion of ethylene into Cp2ScCH3 has been studied through the use of nonlocal density functional calculations [133 ]. The geometries of Cp2ScR (R = H, CH3, SiH3) have also been calculated by ab initio methods [134 ]. It has been found that the scandium systems prefer planar configurations (i. e. the hydride or alkyl substituents prefer the central position in the equatorial plane of the bent metallocene), ultimately attributable to the (4s)2(3d)1 ground state of the metal. It was hypothesized that this planarity would prohibit the syndiotactic polymerization of α-olefins by catalysts having a group 3 metal with the syndiospecific ligand system, [Me2C(η5-C5H4)(η5-fluorenyl)].
146
3
Group 3 Metallocenes
3.10 Alternatives of Cyclopentadienyl Ligands Since the reactivity of group 3 metallocenes is profoundly influenced by the ancillary ligand array, several attempts have been made to employ ligands alternative to the cyclopentadienyl environment [135 ]. One approach has been replacement of one of the cyclopentadienyl moieties with ‘hard’ oxygen or nitrogen-containing ligands. Such examples of these complexes are the ‘Cp*SiNR’ metallocenes discussed previously. Other approaches have used dicarboranes as cyclopentadienyl substitutes. Although the detailed chemistry of these types of complexes is beyond the scope of this report, a brief overview of a few representative cyclopentadienyl replacements is presented. Alkoxide ligands in conjunction with the cyclopentadienyl ligand have been successful in allowing for the preparation and isolation group 3 metallocene analogs. Reaction of Y[OC6H3(CMe3)2]3 with KCp* in toluene results in the formation of Cp*Y[OC6H3(CMe3)2]2 in 70–80% yield [136 ]. An alkyl derivative has been prepared by reaction of Cp*Y[OC6H3(CMe3)2]2 with LiCH(SiMe3)2, which can be hydrogenated under 10 bar of H2 to give the hydride dimer, [Cp*Y(OC6H3(CMe3)2(µ2-H)]2 (Scheme 3-16).
Scheme 3-16
Reactivity studies indicate that the isomer with a cis arrangement of Cp* ligands is the only one formed, and that this species is an active catalyst for the polymerization of ethylene. The bridging alkoxide dimers, [CpY(µ2-OCMe3)(OCMe3)]2 (Cp = Cp*, Cp, (η5-C5H4Me), [η5-C5H3(SiMe3)2]) have been prepared from the corresponding NaCp and Y3(OCMe3)7Cl2 [137 ]. The indenyl complex, [(η5-C9H7)Y(µ2OR)(OR)]2 (R = CMe3), has also been synthesized. Attempts at alkylation with LiCH(SiMe3)2 have resulted in loss of LiCp and formation of bridged alkoxide species.
3.11 Concluding Remarks
147
Teuben and coworkers have prepared Cp*/[PhC(NSiMe3)2] yttrium alkyls. Reaction of YCl3(THF)3.5 with Cp*K followed by reaction with [PhC(NSiMe3)2]LiOEt2 in THF yields the chloride dimer [138 ]. Alkylation of the chloride dimer with traditional alkylating agents such as ME(SiMe3)2 (M = Li, Na; E = CH, N) resulted in formation of complex mixtures; however, a well-defined methyl dimer, Cp*[PhC(NSiMe3)2]Y(µ2-CH3)2Li·TMEDA, has been isolated from the reaction of the chloride dimer with CH3Li. The methyl dimer is subject to protolysis under mild conditions by terminal acetylenes and alcohols. Reaction of M[N(SiMe3)2] (M = Y, La) with Cp*H afforded a mixture of both mono and bisCp* complexes [139 ]; however employing bulky chelating alkyl groups such as in the case of Y(o-C6H4CH2NMe2)3 allows for the isolation of the monocyclopentadienyl complex. A variety of monocyclopentadienyl lanthanum complexes has been prepared. By employing larger iodides to assist in stabilization, Cp*LaI2(THF)3 has been prepared from LaI3(THF)3 and KCp* [140 ]. The thermally unstable alkyl complex, Cp*La[CH(SiMe3)2](THF) can be prepared by addition of two equivalents of KCH(SiMe3)2, while the coordinated THF can be removed by reaction with Me3SiI [141 ]. Reaction of the THF-free bis-alkyl with [PhNMe2H][BPh4] yields the first cationic lanthanum alkyl complex [142 ]. Preparation of some mixed pentamethylcylopentadienyl/dicarbollide derivatives of scandium has also been described [143 ]. Reactions of [Cp*ScCl2]x with Na 2[C 2B 9H 11] or [Cp*ScMe2] x with C2B 9H 13, followed by treatment with THF yield Cp*(C 2B 9H 11)Sc(THF) 3. Alkylation of Cp*(C2B 9H 11)Sc(THF) 3 with LiCH(SiMe3)2 yields Cp*(C2B9H11ScCH(SiMe3)2Li(THF)3 and {[Cp*(C2B9H11)ScCH(SiMe3)2]2Li}·Li(THF)3, which has been characterized structurally. This alkyl derivative reacts slowly with H2 to yield [Cp*(C2B9H11)ScH]2[Li(THF)n]2, a surprisingly unreactive scandium hydride dimer. Recrystallization from toluene affords a crystalline form with less coordinated THF, [Cp*(C2B9H11)ScH]2[Li(THF)]2·3/2(C6H5CH3), whose structure reveals that the two anionic [Cp*(C2B9H11)ScH]– fragments are held together by reciprocal B– H dative bonding from the dicarbollide ligand to the electron deficient scandium.
3.11 Concluding Remarks Since the initial reports of Cp3M complexes over 40 years ago, there has been explosive growth in group 3 metallocene chemistry. An enormous number of both scandocene and yttrocene complexes have been prepared and their reaction chemistry has been applied in a variety of areas. Although many group 3 metallocenes have been prepared, their synthesis is by no means trivial. Most syntheses of the metallocene derivatives rely on salt metathesis reactions in a coordinating solvent
148
3
Group 3 Metallocenes
such as THF or diethyl ether. This methodology coupled with the extreme Lewis acidity of the metal center often leads to unwanted coordination of solvent or salt molecules which can interfere with further synthesis or catalytic activity. Despite these limitations, a large number of useful group 3 metallocenes have been synthesized. Such complexes are active catalysts for olefin hydrogenation, silylation, hydroboration, hydroamination, cyclization and polymerization. Incorporation of the properly substituted cyclopentadienyl framework has lead to the development of chiral metallocenes which are successful in catalyzing some of the aforementioned reactions stereospecifically. Group 3 metallocenes have been particularly effective in elucidating the mechanistic features in olefin polymerization. Essential polymerization steps, including olefin insertion and β elimination have been probed with a variety of scandocenes and yttrocenes, revealing the key features of a Ziegler–Natta catalyst. Group 3 metallocenes also exhibit high reactivity toward the C–H bonds of hydrocarbons. Once again, the relatively clean nature of many of these reactions has allowed investigations of the fundamental processes, notably σ-bond metathesis. Additionally, scandocenes and yttrocenes react with a variety of small molecules that include carbon monoxide, carbon dioxide, nitriles and amines. Despite the strong bonds formed between the metal center and the heteroatoms of these substrates, some catalytic transformations have been developed. The catalytic transformations promoted by yttrocenes are now beginning to be applied in organic synthesis. A variety of stereo- and regiospecific reactions have been developed and are being used to prepare natural products. Notwithstanding the synthetic challenges associated with these complexes, particularly their intolerance to heteroatom-containing substrates and their inherent air and water sensitivity, a wealth of chemistry has been uncovered.
References [1]
[2]
[3] [4] [5] [6] [7] [8]
a) Evans, W.J. Adv. Organomet. Chem. 1985, 24, 131. b) Shaverien, C.J. Adv. Organomet. Chem. 1994, 36, 283. c) Rogers, R.D.; Rogers, L.M. J. Organomet. Chem. 1992, 443, 83. d) Evans, W.J. Polyhedron, 1987, 6, 803. e) Nolan, S.P.; Stern, D.; Hedden, D.; Marks, T.J. ACS Symp. Ser. 1990, 428, 159. f) Burns, C.J.; Burstein, B.E. Comments Inorg. Chem. 1989, 9, 61. g) Evans, W.J.; Foster, S.E. J. Organomet. Chem. 1992, 433, 79. a) Lappert, M.F.; Holton, J.; Ballard, D.G.H.; Pearce, R.; Atwood, J.L.; Hunter, W.E. J. Chem. Soc. Dalton Trans. 1979, 49. b) Lappert, M.F.; Holton, J.; Ballard, D.G.H.; Pearce, R.; Atwood, J.L.; Hunter, W.E. J. Chem. Soc. Dalton Trans. 1979, 54. Wilkinson, G.; Birmingham, J.M. J. Am. Chem. Soc. 1954, 76, 6210. Hart, F.A.; Massey, A.G.; Saran, M.S. J. Organomet. Chem. 1970, 21, 147. Coutts, R.S.P.; Wailes, P.C. J. Organomet. Chem. 1970, 25, 117. Atwood, J.L.; Smith, K.D. J. Chem. Soc. Dalton Trans. 1973, 2487. Manzer, L.E. J. Organomet. Chem. 1976, 110, 291. Manzer, L.E. Inorg. Chem. 1976, 15, 2567.
References [9] [10] [11] [12]
[13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44]
149
Manzer, L.E. J. Am. Chem. Soc. 1978, 100, 8068. Reid, A.; Wailes, P.C. Inorg. Chem. 1996, 5, 1213. Bottomly, F.; Paez, D.E.; White, P.S. J. Organomet. Chem. 1985, 291, 35. (a) Lappert, M.F.; Singh, A.; Atwood, J.L.; Hunter, W.E. J. Chem. Soc. Chem. Comm. 1983, 206. (b) Lappert, M.F.; Singh, A.; Atwood, J.L.; Hunter, W.E. J. Chem. Soc. Chem. Comm. 1983, 69. Thompson, M.E.; Bercaw, J.E. Pure Appl. Chem. 1984, 56, 1. Although Cp*2ScH(THF) is moderately stable in solution in the absence of H2, it does decompose in the solid state after several days at room temperature. Watson, P.L.; Herskovitz, T. ACS Symposium Series 1983, 212, 459. Piers, W.E.; Shapiro, P.J.; Bunel, E.E.; Bercaw, J.E. Synlett 1990, 74. Schaefer, W.P.; Köhn, R.D.; Bercaw, J.E. Acta Crystallogr. 1992, C48, 251. Bunel, E.E., Ph. D Thesis, California Institute of Technology, Pasadena, CA (1989). Shapiro, P.J.; Bunel, E.; Schaefer, W.P.; Bercaw, J.E. Organometallics 1990, 9, 867. Mu, Y.; Piers, W.P.; MacQuarrie, D.C.; Zawaorotko, M.J.; Young, V.G. Organometallics 1996, 15, 2720. Birmingham, J.M.; Wilkinson, G. J. Am. Chem. Soc. 1956, 78, 62. Rogers, R.D.; Atwood, J.L.; Emad, A.; Sikora, D.J.; Rausch, M.D.; J. Organomet. Chem. 1981, 216, 383. Eggers, S.H.; Kopf, J.; Fischer, R.D. Organometallics 1986, 5, 383. Xie, Z.; Hahn, F.E.; Qian, C. J. Organomet. Chem. 1991, 414, C12. Evans, W.J.; Dominguez, R.; Levan, K.R.; Doedens, R.J. Organometallics 1985, 4, 1836. Evans, W.J.; Meadows, J.H.; Wayda, A.L.; Hunter, W.E.; Atwood, J.L. J. Am. Chem. Soc. 1982, 104, 2008. Maginn, R.E.; Manastyrskyj, S.; Dubeck, J. J. Am. Chem. Soc.1963, 85, 672. Evans, W.J.; Meadows, J.H.; Hanusa, T.P. J. Am. Chem. Soc. 1984, 106, 4454. Evans, W.J.; Peterson, T.T; Rausch, M.D.; Hunter, W.E.; Zhang, H.; Atwood, J.L. Organometallics 1985, 4, 554. Evans, W.J.; Grate, J.W.; Levan, K.R.; Bloom, I.; Peterson, T.T.; Doedens, R.J.; Zhang, H.; Atwood, J.L. Inorg. Chem. 1986, 25, 3614. Evans, W.J.; Boyle, T.J.; Ziller, J.W. Inorg. Chem. 1992, 31, 1120. den Haan, K.H.; de Boer, J.L.; Teuben, J.H.; Spek, A.L.; Kojic-Prodic, B.; Hays, G.R.; Huis, R. Organometallics 1986, 5, 1726. den Haan, K.H.; Wielstra, Y.; Eshius, J.J.W.; Teuben, J.H. J. Organomet. Chem. 1987, 323, 181. den Haan, K.H.; De Boer, J.L.; Teuben, J.H.; Smeets, W. J. J. J.; Spek, A.L. J. Organomet. Chem. 1987, 327, 31. den Haan, K.H.; Teuben, J.H. Recl. Trav. Chim. Pays-Bas 1984, 103, 333. den Haan, K.H.; Wielstra, Y.; Teuben, J.H. Organometallics 1987, 6, 2053. den Haan, K. H.; Teuben, J.H. J. Chem. Soc. Chem. Comm. 1986, 682. Holton, J.; Lappert, M.F.; Scollary, G.R.; Ballard, D.G.H.; Pearce, R.; Atwood, J.L.; Hunter, W.E. J. Chem. Soc. Chem. Com. 1976, 425. Lobkovskii, E.B.; Soloveichik, G.L.; Erofeev, A.B.; Bulychev, B.M.; Bel’skii, V. J. Organomet. Chem. 1982, 235, 151. Lobkovsky, E.B.; Soloveichik, G.L.; Bulychev, B.M.; Erofeev, A.B.; Gusev, A.I.; Kirillova, N.I. J. Organomet. Chem. 1983, 254, 167. Belsky, V.K.; Erofeev, A.B.; Bulychev, B.M.; Soloveichik, G.L. J. Organomet. Chem. 1984, 265, 123. Erofeev, A.B.; Bulychev, B.M.; Bel’skii, V.K.; Soloveichik, G.L. J. Organomet. Chem. 1987, 335, 189. Erofeev, A.B.; Bulychev, B.M.; Bel’skii, V.K.; Soloveichik, G.L. J. Organomet. Chem. 1987, 335, 189. Busch, M.A.; Harlow, R.; Watson, P.L. Inorg. Chim. Acta 1987, 140, 15.
150 [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60]
[61] [62] [63] [64] [65] [66]
[67] [68] [69] [70] [71]
[72] [73] [74] [75]
[76] [77] [78]
3
Group 3 Metallocenes
Schumann, H.; Rosenthal, E.C.E.; Kociok-Kohn, G.; Molander, G.A.; Winterfeld, J. J. Organomet. Chem. 1995, 496, 233. Evans, W.J.; Drummond, D.K.; Hanusa, T.P.; Doedens, R.J. Organometallics 1987, 6, 2279. Laske, D.A.; Duchateau, R.; Teuben, J.H.; Spek, A.L. J. Organomet. Chem. 1993, 462, 149. Deng, D.; Jiang, Y.; Qian, C.; Wu, G.; Zheng, P. J. Organomet. Chem. 1994, 470, 99. Qian, C.; Wang, B.; Deng, D.; Sun, J.; Hahn, F.E.; Chen, J.; Zheng, P. J. Chem. Soc. Dalton Trans. 1996, 955. Broussier, R.; Delmas, G.; Perron, P.; Gautheron, B.; Peterson, J.L. J. Organomet. Chem. 1996, 511, 185. Nief, F.; Mathey. F. J. Chem. Soc. Chem. Comm. 1989, 800. Qian, Changtao;Xie, Z..; Huang, Y. J. Organomet. Chem. 1987, 323, 285. Schumann, H.; Loebel, J.; Pickardt, J.; Qian, C.; Xie, Z. Organometallics 1991, 10, 215. Qian, C.; Xie, Z. Huang, Y. Inorg. Chim. Acta 1988, 139, 195. Qian, C.; Ye, C.; Li, Y. J. Organomet. Chem. 1986, 302, 171. Coughlin, E.B.; Bercaw, J.E. J. Am. Chem. Soc. 1992, 114, 7606. Chacon, S.T.; Coughlin, E.B.; Henling, L.M.; Bercaw, J.E. J. Organomet. Chem. 1995, 497, 171. Mu, Y.; Piers, W.E.; MacDonald, M.A.; Zaworotko, M.J. Can. J. Chem. 1995, 73, 2233. Kretschmer, W.P. Verslagen Werkgroep Organometaalchemie en Homogene Katalyse 1996, 9, 67. For general reviews see: (a) Crabtree; R.H. Chem. Rev. 1985, 85, 245. (b) Tanaka, M.; Sakakura, T. In Homogeneous Transition Metal Catalyzed Reactions; Moser, W.R., Slocum, D.W. Eds.; American Chemical Society: Washington, DC, 1992; p. 181. (c) Jones, W.D. In Selective Hydrocarbon Activation; Davies, J.A., Watson, P.L., Liebman, J.F.; Greenberg, A., Eds.; VCH Publishers: New York, 1990; p. 113. Watson, P.L. J. Am. Chem. Soc. 1983, 105, 6491. Thompson, M.E.; Baxter, S.M.; Bulls, A.R.; Burger, B.J.; Nolan, M.C.; Santarsiero, B.D.; Schaefer, W.P.; Bercaw, J.E. J. Am. Chem. Soc. 1987, 109, 203. Steigerwald, M.L.; Goddard, W.A. J. Am. Chem. Soc. 1984, 106, 308. Bulls, A.R.; Bercaw, J.E.; Manriquez, J.M.; Thompson, M.E. Polyhedron, 1988, 7, 1409. Booij, M.; Deelman, B.J.; Duchateau, R.; Postma, D.S.; Meetsma, A.; Teuben, J.H. Organometallics 1993, 12, 3531. Burger, B.J.; Cotter, W.D.; Coughlin, E.B.; Chacon, S.T.; Hajela, S.H.; Herzog, T.A.; Kohn, R.; Mitchell, J.; Piers, W.E.; Shapiro, P.J.; Bercaw, J.E. In Ziegler Catalysts; Fink, G., Mülhaupt, R.; Brintzinger, H.H. Eds.; Springer-Verlag: Berlin, 1995; pp 318. Ballard, D.G.H.; Courtis, A.; Holton, J.; McMeeking, J.; Pearce, R. J. Chem. Soc. Chem. Com. 1978, 994. Evans, W.J.; Meadows, J.H.; Hunter, W.E.; Atwood, J.L. J. Am. Chem. Soc. 1984 106, 1291. Stern, D.; Sabat, M.; Marks, T.J. J.Am. Chem. Soc. 1990, 112, 9558. Burger, B.J.; Thompson, M.E.; Cotter, W.D.; Bercaw, J.E. J. Am. Chem. Soc. 1991, 112, 1566. (a) Bryndza, H.E.; Fong, L.K.; Paciello, R.A.; Tam, W.; Bercaw, J.E. J. Am. Chem. Soc. 1987, 107, 1444. (b) Bryndza, H.E.; Domaille, P.J.; Fong; L.K.; Paciello, R.A.; Tam, W.; Bercaw, J.E. Polyhedron 1988, 7, 1441. Bunel, E.E.; Burger, B.J.; Bercaw, J.E. J. Am. Chem. Soc. 1988, 110, 976. Bercaw, J.E. Pure & Appl. Chem. 1990, 62, 1151. Chirik, P.J.; Bercaw, J.E. unpublished results. (a) Brookhart, M.; Green, M.L.H. J. Organomet. Chem. 1983, 250, 395. (b) Brookhart, M.; Green, M.L.H.; Wong, L. Prog. Inorg. Chem. 1988, 36, 1. (c) Grubbs, R.H.; Coates, G. W. Acc. Chem. Res. 1996, 29, 85. Clawson, L.; Soto, J.; Buchwald, S.L.; Steigerwald, M.L.; Grubbs, R.H.J. Am. Chem. Soc. 1985, 107, 3377. Piers, W.E.; Bercaw, J.E. J. Am. Chem. Soc. 1990, 112, 9406. Herzog, T.A. Ph. D. Thesis California Institute of Technology (1996).
References
151
[79] Hajela, S.; Bercaw, J.E. Organometallics 1994, 13, 1147. [80] Shapiro, P.J.; Bunel, E.E.; Schaefer, W.P.; Bercaw, J.E. Organometallics 1990, 9, 867. [81] Shapiro, P.J.; Cotter, W.D.; Schaefer, W.D.; Labinger, J.A.; Bercaw, J.E. J. Am. Chem. Soc. 1994, 116, 4623. [82] Coughlin, E.B.; Bercaw, J.E. J. Am. Chem. Soc. 1992, 114, 7607. [83] Marsh, R.E.; Schaefer, W.P.; Coughlin, E.B.; Bercaw, J.E. Acta Cryst. 1992, C48, 1773. [84] Yasuda, H,.; Ihara, E. Tetrahedron 1995, 51, 4563. [85] Mitchell, J.P.; Hajela, S.; Brookhart, S.K.; Hardcastle, K. I.; Henling, L.M.; Bercaw, J.E. J. Am. Chem. Soc. 1996, 118, 1045. [86] Under an atmosphere of H2, neat 1-pentene is readily polymerized to afford a polymer of relatively narrow molecular weight distribution and modest molecular weight (PDI = 1.44; Mn = 119,000; mmmm > 95%). [87] Gilchrist, J.H.; Bercaw, J.E. J. Am. Chem. Soc. 1996, 118, 12021. [88] Schaverien, C.J. J. Chem Soc. Chem. Comm. 1992, 11. [89] Meijboom, R. Verslagen Werkgroep Organometaalchemie en Homogene Katalyse 1996, 9, 139. [90] St. Clair, M.; Schaefer, W.P.; Bercaw, J.E. Organometallics 1991, 10, 525. [91] Heeres, H.J.; Teuben, J.H. Organometallics 1991, 10, 1980. [92] Evans, W.J.; Meadows, J.H.; Hunter, W.E.; Atwood, J.L. J. Am. Chem. Soc. 1984 , 106, 1291. [93] den Haan, K.H.; Wielstra, Y.; Teuben, J.H. Organometallics 1987, 6, 2053. [94] Heeres, H.J.; Heeres, A.; Teuben, J.H. Organometallics 1990, 10, 1508. [95] Piers, W.E.; MacGillivray, L.R.; Zaworotko, M. Organometallics 1993, 12, 4723. [96] Piers, W.E.; Parks, D.J.; MacGillivray, L.R.; Zaworotko, M.J. Organometallics 1994, 13, 4547. [97] Piers, W.E. J. Chem. Soc. Chem. Com. 1994, 309. [98] Piers, W.E.; Ferguson, G.; Gallagher, J.F. Inorg. Chem. 1994, 3784. [99] Beletskaya, I.P.; Voskoboynikov, A.Z.; Shestakova, A.K.; Yanovsky, A.I.; Fukin, G.K.; Zacharov, L.V.; Struchkov, Y.T.; Schumann, H. J. Organomet. Chem. 1994, 468, 121. [100] St. Clair, M.A.; Santarsiero, B.D.; Bercaw, J.E. Organometallics 1989, 8, 17. [101] Bercaw, J.E.; Davies, D.L.; Wolczanski, P.T. Organometallics 1986, 5, 443. [102] Shapiro, P.J.; Henling, L.M.; Marsh, R.E.; Bercaw, J.E. Inorg. Chem. 1990, 29, 4560. [103] Campion, B.K.; Heyn, R.H.; Tilley, T.D. Inorg. Chem. 1990, 29, 4356. [104] Brook, A.G.; Yau, l. J. Organomet. Chem. 1984, 271, 9. [105] Campion, B.K.; Heyn, R.; Tilley, T.D. J . Am. Chem. Soc. 1990, 112, 2011. [106] den Haan, K. H.; Wielstra, Y.; Teuben, J.H. Organometallics 1987, 6, 2053. [107] Deelman, B.J.; Stevels, W.M.; Teuben, J.H.; Lakin, M.T.; Spek, A.L. Organometallics 1994, 13, 3881. [108] Evans, W.J.; Meadows, J.H.; Hunter, W.E.; Atwood, J.L. Organometallics 1983, 2, 1252. [109] Evans, W.J.; Hanusa, T.P.; Meadows, J.H.; Hunter, W.E.; Atwood, J.L. Organometallics 1987, 6, 295. [110] Rausch, M.D.; Foust, D.F.; Rogers, R.D.; Atwood, J.L. J. Organomet. Chem. 1984, 265, 241. [111] vans, W.J.; Sollberger, M.S. J. Am. Chem. Soc. 1986, 108, 6095. [112] Evans, W.J.; Sollberger, M.S.; Shreeve, J.L.; Olofson, J.M.; Hain, J.H.; Ziller, J.W. Inorg. Chem. 1992, 31, 2492. [113] Evans, W.J.; Meadows, J.H.; Hanusa, T.P. J. Am. Chem. Soc. 1984, 106, 4454. [114] Evans, W.J.; Sollberger, M.S.; Khan, S.I.; Bau, R. J. Am. Chem. Soc. 1988, 110, 439. [115] Evans, W.J.; Dominguez, R.; Hanusa, T. P. Organometallics 1986, 5, 1291. [116] Ringelberg, S.N. Verslagen Werkgroep Organometaalchemie en Homogene Katalyse, 1996, 9, 79. [117] Evans, W.J.; Hozbor, M.A.; Bott, S.G.; Robinson, G.H.; Atwood, J.L. Inorg. Chem. 1988, 27, 1990. [118] Alvarez, D.; Caulton, K.G.; Evans, W.J.; Ziller, J.W. Inorg. Chem. 1992, 31, 5500. [119] Molander, G.A.; Hoberg, J.O. J. Org. Chem. 1992, 57, 3266.
152 [120] [121] [122] [123] [124]
[125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135]
[136] [137] [138] [139] [140] [141] [142] [143]
3
Group 3 Metallocenes
Molander, G.A.; Hoberg, J.O. J. Am. Chem. Soc. 1992, 114, 3123. Qian, C.T.; Zie, Z.; Huang, Y. Chinese Sci. Bull. 1989, 34, 1106. Haar, C.M.; Stern, C.L.; Marks, T.J. Organometallics 1996, 15, 1765. Giardello, M.A.; Conticello, V.P.; Brard, L.; Gagne, M.R.; Marks, T.J. J. Am. Chem. Soc. 1994, 116, 10241. R' = (1S,2S,5R)-trans-5-methyl-cis-2-(2-propyl)cyclohexyl((+)-neomenthyl), (1R, 2S,5R)-cis-5-methyl-trans-2-(2-propyl)-cyclohexyl((-)-menthyl) and (1R,2S,5R)-cis-5-methyl-trans-2-(2-phenyl-2-propyl)-cyclohexyl((-)-phenylmethyl); (E = N, CH). Molander, G.A.; Julius, M. J. Org. Chem. 1992, 57, 6347. Molander, G.A.; Nichols, P.J. J. Am. Chem. Soc. 1995, 117, 4415. Molander, G.A.; Nichols, P.J. J. Org. Chem. 1996, 61, 6040. Harrison, K.N.; Marks, T.J. J. Am. Chem. Soc. 1992, 112, 9220. Pfennig, B.W.; Thompson, M.E.; Bocarsly, A.B. J. Am. Chem. Soc. 1989, 111, 8947. Pfennig, B.W.; Thompson, M.E.; Bocarsly, A.B. Organometallics 1993, 12, 649. Lauher, J.W.; Hoffmann, R. J. Am. Chem Soc. 1976, 98, 1729. Kaupp, M.; Charkin, O.P.; Schleyer, P. v. R. Organometallics 1992, 11, 2765. Woo, T.K.; Fan, L.; Ziegler, T. Organometallics 1994, 13, 432. Bierwagen, E.P.; Bercaw, J.E.; Goddard, W.A. J. Am. Chem. Soc. 1994, 116, 1481. For bis(N,N’-bistrimethylsilyl)benzamidinato)yttrium complexes see the following: a) Duchateau, R.; Wee, C.T. van; Meetsma, A.; Teuben, J.H. J. Am. Chem. Soc. 1993, 115, 4931. b) Bijpost, E.A.; Duchateau, R.; Teuben, J.H. J. Mol. Cat. 1995, 95, 121. For yttracarborane complexes: Oki, A.R.; Zhang, H.; Hosmane, N.S. Organometallics 1991, 10, 3964. For scandium benzamidinate complexes see: Hagadorn, J.R.; Arnold, J. Organometallics 1996, 15, 984. Schaverien, C.J.; Frijins, J.H.G.; Heeres, H.J.; van den Hende, J.R., Teuben, J.H.; Spek, A. J. Chem. Soc.1991, 642. Evans, W.J.; Boyle, T.J.; Ziller, J.W. Organometallics 1993, 12, 3998. Duchateau, R.; Meetsma, A.; Teuben, J.H. Organometallics 1996, 15, 1656. Booij, M.; Kiers, N.H.; Heeres, H.J.; Teuben, J.H. J. Organomet. Chem. 1989, 364, 79. Hazin, P.N.; Huffman, J.C.; Bruno, J.W. Organometallics 1987, 6, 23. van der Heijen, H.; Schaverien, C.J.; Orpen, A.G. Organometallics 1989, 8, 255. Schaverien, C.J. Organometallics 1992, 11, 3476. Marsh, R.E.; Schaefer, W.P.; Bazan, G.C.; Bercaw, J.E. Acta. Cryst., C48, 1416.
4
Titanocenes Rüdiger Beckhaus
4.1 Introduction The bright green, so-called ‘titanocene′, obtained from the reaction of TiCl2 with CpNa and first reported in 1956 by A. K. Fischer and G. Wilkinson [1], marked the discovery of one of the most extensively utilized complex systems in organotransition metal research. In this chapter, complexes exhibiting a [(η5-Cp)2Ti] unit or derivatives with comparable substitution pattern (Cp′, Cp*) will be discussed with the exclusion of complexes containing only one Cp ligand. Typical η5-Cp type ligands are summarized in Scheme 4-1. Free titanocene [Cp2Ti], as a formally 14 electron species, is obviously not isolable. However, a wide range of derivatives with additional neutral or anionic ligands exists, characterized by the typical bentmetallocene structure and chemical reactions which take place on the more open side of the electronically flexible and sterically compact [Cp2Ti] fragment. The chemistry of titanium is characterized by its multiple oxidation states. For such titanocenes the available oxidation states are normally +2, +3, +4 (Scheme 41). In this chapter, the organometallic properties of titanocenes will be discussed from the point of view of coordination chemistry. During the last few years a wide range of titanocene derivatives exhibiting different substitution patterns have been synthesized and characterized. This article focuses on selected topics in the synthesis and reactivity pattern of titanocene derivatives. For further details, excellent and extensive summaries already exist [2–7].
Scheme 4-1. General ligand and structure types of titanocene complexes.
154
4
Titanocenes
4.1.1 Bonding Modes of Cyclopentadienyl Ligands in Titanium Complexes At least 10 distinct bonding modes for the C5H5 ligand have been structurally characterized to date [8]. Of these, the η5-C5H5, η3-C5H5, and η1-C5H5 bonding modes are of particular interest due to their interconversion chemistry (ring slippage) [8] and the consequent implications for the design of catalytic systems. In 1971 Cotton reported the variable-temperature 1H NMR studies of [(η5-C5H5)2Ti(η1-C5H5)2] (1), for which he was able to observe η1 to η5 cyclopentadienyl interconversions [9, 10].
Although no intermediates were observed, Cotton later proposed an [(η1-C5H5)(η4C5H5)2(η5-C5H5)Ti] intermediate for this η1 to η5 interconversion [8]. 4.1.2 Synthesis of Titanocene Complexes Generally, Cp ligands are introduced by using the salt metathesis reaction. The common derivatives e.g. [Cp2TiCl2] are available by this method (Scheme 4-2, path a) [11], but, the preparation of the permethylated [Cp*2TiCl2] starting from TiCl4 and Cp*Li failed. The only product of this reaction is the monocyclopentadienyl derivative [Cp*TiCl3] [12]. Higher yields, up to 65% of purple–brown [Cp*2TiCl2], were obtained when TiCl3 was treated with Cp*Na or Cp*Li in dimethoxyethane solution in a molar ratio of 1:3, and then oxidized by treating with hydrochloric acid [13, 14].
4.1
Introduction
155
Scheme 4-2. Possible synthetic pathways a–d for titanocene complexes.
The use of cyclopentadiene-substituted tin compounds to prepare monocyclopentadienyl transition metal complexes, first reported by Abel and co-workers [15], is now an established method for the synthesis of [CpTiCl3] and related compounds [16, 17]. Selective and high-yielding preparations of mixed ring complexes [Ti(η5C5H5)(η5-C9H7)Cl2] and [Ti(η5-C5H5)(η5-C5H4But)Cl2] are reported by McCamley and co-workers using the corresponding tin reagents (path b) [18]. The reaction between Cp*SiMe3 and group 4 element tetrahalides MX4 (M = Ti, X = Cl, Br, I; M = Zr, Hf, X = Cl) gives the corresponding [(η5-C5Me5)MX3] derivatives in near quantitative yields in a one-step procedure without the need for further purification [19]. Tetravalent monocyclopentadienyl complexes are useful starting materials for lower valent titanium complexes like [CpTiCl2], from which mixed [CpCpRTiCl] complexes become available by halide exchange reaction [20]. The oxidation of this type of complex using PbCl2 as an oxidizing agent is very useful for the preparation of mixed L2TiCl2 complexes or substituted derivatives exhibiting functionalized side arms (path c) [20]. Lead dichloride is a convenient reagent to cleanly oxidize a wide variety of tervalent [Cp*2TiR] complexes to their tetravalent analogs. Monobromide complexes can be obtained by the same procedure using PbBr2 [21]. The oxidation proceeds smoothly at room temperature, but also takes place at a reasonable rate at temperatures as low as –30°C. Amine elimination reactions from titanium amides and Cp-H, first reported by Chandra and Lappert [22] for the synthesis of [CpTi(NMe2)3] from [Ti(NMe2)4] and Cp-H (72%) are now a well developed method for the preparation of a wide
156
4
Titanocenes
range of other metallocene derivatives (path d) [23–25]. However, lower reactivity is observed in titanium complexes than in the corresponding zirconium complexes and only monocyclopentadienyl type complexes are normally formed [26]. The reactivity of the titanium amides can be increased by using the less crowded titanium azetidide complexes instead of the [Ti(NMe2)4], as shown for the formation of 2 [23].
Scheme 4-3. Amine elimination reaction for titanium complexes. However, mixtures of TiCl4, Cp-H and HNEt2 can be used for the preparation of [Cp2TiCl2] without using the highly sensitive NaCp [27]. By ligand exchange reactions with the corresponding chlorides, fluorides become available in reactions with NaF, whereas bromides and iodides can be prepared using the corresponding boron halides (BBr3, BI3) [28, 29]. Table 4-1. Selected titanocene complexes, syntheses and properties
4.1
Introduction
157
4.1.3 Sources of the Titanocene Fragment as a Versatile Building Block In many applications the bent titanocene fragment of the ‘Cp2Ti’ type is useful in coordinating new unsaturated ligands or in finding new types of C–C bond as well as other coupling reactions. This means it becomes necessary to use titanocenegenerating complexes, because a Cp2Ti fragment does not exist without further stabilizing ligands (Scheme 4-4). In practice, ligand exchange reactions have to take place between selected titanocene precursors and substrates to give the required products. Depending on the reaction conditions, the substrates and the final products, different types of starting material are used. Carbonyls and phosphines can be substituted by stronger π-acceptor ligands, but in many cases, only one ligand can be substituted at a time or the leaving ligand itself is involved in unwanted subsequent reactions. The generation of the titanocene fragments from alkyl derivatives is possible in photochemical reactions, but side reactions sometimes occur under these reaction conditions.
Scheme 4-4. Sources for the generation of titanocene building blocks. In organic syntheses, path d is very useful [30]. By this path the reductive cyclization of ester-containing enynes can be achieved in order to synthesize bicyclic cyclopentenones and iminocyclopentenes. Nitrogen complexes of titanocenes can also be used as titanocene sources. But due to the high sensitivity of the nitrogen complexes to air and moisture, the handling of these complexes requires much experience. The bistrimethylsilyl acetylene complexes are excellent reagents in generating Cp2Ti as well as Cp*2Ti building blocks [31]. Generally, the reactivity of these complexes is characterized by the liberation of the inert silylalkyne, which does
158
4
Titanocenes
not react further, as carbon monoxide or phosphines frequently do. In some cases, the olefin complex [Cp*2Ti(η2-C2H4)] can also act as a titanocene generating complex. Generally, Cp* complexes are of higher stability and more soluble than the equivalent nonmethylated Cp complexes. Cp*2Ti complexes are mostly monomers, whereas the normal Cp2Ti complexes react to form dimers or oligomers. As a result of this, some Cp* complexes are available for which the equivalent Cp systems are unknown. Additionally, there are some reaction paths, which are quite different, depending on whether Cp or Cp* complexes are used. Some examples are given in Table 4-2. Table 4-2. Selected complexes and reactions courses of Cp2Ti vs Cp*2Ti complexes
4.2 Electronic Properties of Titanocenes Normal bis(η5-cyclopentadienyl) transition metal complexes such as ferrocene are highly symmetric molecules with parallel cyclopentadiene rings. In the bent bis(η5cyclopentadienyl)transition metal complexes the rings are not parallel i.e. the angle between the normals to the planes of the cyclopentadienyl ligands is less than 180°, and there are other ligands present at the metal center. A bent Cp2M fragment has C2v symmetry if the Cp ligands have an eclipsed geometry, but only Cs symmetry if the rings are staggered. Generally, the chemistry of the bent metallocene fragment is determined by the three low-lying frontier orbitals 1a1, b2 and 2a1 [43]. All three orbitals have significant density in the mirror plane of the metallocene fragment. The b2 orbital is chiefly dyz in character. The two a1 orbitals each contain some contribution from the s and pz orbitals as well as dx2–y2 and
4.2 Electronic Properties of Titanocenes
159
dz2 contributions. The 1a1 orbital can be described as similar to a dy2 orbital. This leads to the well established bent metallocene geometry, for which the electronic constraints are well documented both through their extensive derivative chemistry and through detailed quantum chemical calculations [44]. For a bent-metallocene fragment, such as Cp2Ti, the ancillary ligands interact with the three low lying orbitals which project in the so-called ‘equatorial binding plane’ or ‘binding wedge’ of the metallocene fragment. The orientation of the ligands will then be dominated by the interactions of their p or π-symmetry orbitals with the π-symmetry combination in this plane [45, 46].
Fig. 4-1. Valence orbitals of bent titanocenes.
4.2.1 Titanocene Complexes in Low Oxidation States [Cp*2TiX] complexes (X: monodentate, one-electron ligand, e.g. halide, amide, alkoxide) appear to be ideal for the study of metal-to-ligand π-bonding. They are monomeric, unlike [Cp2TiX]2 in which the electrons are coupled. The trivalent decamethyltitanocenes have an empty b2 orbital available for π-bonding, unlike [Cp2MX2] in which the b2-orbital is used for σ-bonding. From the electronic and the EPR spectra of a series of d1-bent metallocene compounds Cp*2TiX the order of decreasing π-donor ability was found to be N(Me)H ≈ NH2 ≈ OMe > OPh ≈ F > NMePh > Cl > Br > I > H [47]. Among the known alkyls, the π-donor ability was Et > Me > n-Pr ≈ CH2CMe3 > CH2C6H5 which is rationalized in part by a βagostic interaction in the case of Et. This observation is in accordance with the general observation, that the reactivity of bent titanocene complexes is determined by the electron deficient character of
160
4
Titanocenes
the metal center in combination with the orientation of the acceptor orbitals in the equatorial plane of the metallocene fragment. This can also be illustrated by the orientation of the substituents on nitrogen atoms in [Cp*2TiNR1R2] complexes. If there is no steric hindrance, the substituents on the nitrogen atom (R1 = R2 = H [48]; R1 = H, R2 = Me [47]) are rotated out of the equatorial plane of the metallocene to maximize overlap between the nitrogen lone pair (located in the pz orbital) and the lowest unoccupied molecular orbital (LUMO) of the metal. This effect is also found in the case of [Cp*2Hf(H)(NHMe)] [49]. If larger substituents are present, as in [Cp*2TiN(CH3)(Ph)], the methyl and the phenyl group are now located in the equatorial plane, and consequently a dπ–pπ interaction is no longer possible [50].
Scheme 4-5. Schematic drawing of structures of [Cp*2TiNR2] complexes, Ti–N distances depends on the orientation of the lone pair on nitrogen atom.
4.2 Electronic Properties of Titanocenes
161
A similar orientation of the substituents can be observed in carbene complexes and their cumulogous derivatives. The character of the metal fragment MO b2 determines that in the structure of [Cp2Ti=C=CH2] only the C=CH2 rotamer shown allows a Ti–Ca π-bond (‘back bonding’) [38]. In accordance with this, model calculations on [Cp2Ti=C=CH2] yield a rotational barrier of 134 kJ mol–1 [38]. For asymmetrically substituted vinylidenetitanium compounds it is possible to show a potential rotation around the Ti=C bond in solution [51]. For the titanaallenylidene, no structural dynamics are observed, and NMR data indicate that all C atoms of the cumulene lie in a mirror plane of the molecule [52].
Scheme 4-6. Geometries of titanocenemethylene, -vinylidene, and -allenylidene complexes.
4.2.2 Substituents on the Cp Rings Since its first general use as a ligand [12], the pentamethylcyclopentadienyl moiety has attracted considerable attention and widespread use. It is commonly known that replacement of the cyclopentadienyl group (Cp) by the pentamethylcyclopentadienyl ligand (Cp*) in certain bis- and mono(cyclopentadienyl) complexes of transition metals results in significant changes in chemical reactivity, stability, sensitivity to oxidation, and many other properties. ESCA studies of a series of titanocenes, zirconocenes, and hafnocenes have shown that substitution of methyls for hydrogens on the cyclopentadienyl ligand results in a significant electronic effect, as reflected by the binding energies of the inner-shell electrons of the metal. The substitution of two pentamethylcyclopentadienyl groups for cyclopentadienyl ligands is approximately equivalent to a one-electron reduction of the complexed metal [53]. This replacement (Cp→Cp*) in metallocene derivatives leads to increased stability of complexes owing to the inductive and hyperconjugation effects of the methyl group and to the increased steric shielding of the coordination space
162
4
Titanocenes
about the central metal atom. The inductive and hyperconjugative effects influence the chemical and physical properties of the complexes by increasing the electron density at the metal atom, and decreasing the ionization potential of the Cp ligand [54, 55]. In the titanocene dihalide series the substitution of Cp with Cp* lowers the oxidation potential by 0.5–0.6 V and decreases the binding energies of the Ti(2p) inner shell electrons by 0.8 eV [28, 53, 56]. The 49Ti NMR chemical shifts in Cp* derivatives appear well downfield from where they are in the equivalent Cp complexes, thus following the inverse halogen-dependence found for titanocene dihalides [28]. The methyl substituents have a substantial effect on the energy of valence electrons in titanocene dihalides and cyclopentadienyltitanium trihalides as has been observed by photoelectron spectroscopy [57]. An excellent linear correlation of binding energy (Ti2p) with the number of methyl substituents (1–10) was found, indicating that the electronic influence of methyl groups on the titanium center is additive [58]. Furthermore, the trialkylsilyl group was found to be slightly more electron-donating than methyl [59].
4.3 Ligands in the Coordination Sphere of Titanocenes 4.3.1 Olefins Starting from [Cp2Ti(PMe3)2] solutions of [Cp2Ti(η2-C2H4)] can be prepared [34]. The same complex is generated from [Cp2TiCl2] with magnesium/1,2-dibromethane in the presence of ethylene [60]. However, a well characterized olefin complex has been prepared by reduction of [Cp*2TiCl2] with sodium amalgam in toluene at 25°C in the presence of ethylene. The resulting olefin complex is isolated in 80% yield in the form of green crystals. The bonding of the ethylene ligand is characterized by extensive backbonding. The C–C bond (1.438(5) Å) is long compared with that of free ethylene (1.337 Å) and the hydrogen atoms are bent away from the metal so that the CH2 plane forms an angle of 35° with the C–C axis [35, 61]. The metallacyclopropane form was found to be lower in energy than the π-complex form for [Cl2Ti(C2H4)] complexes, using wave function descriptions [62]. A wide range of substitution and metal centered cycloaddition reactions can be carried out using the [Cp*2Ti(η2-C2H4)] complex [35, 63, 64]. 4.3.2 Acetylenes The first titanocene–acetylene complexes were obtained by partial replacement of carbon monoxide [65, 66] or a trialkylphosphine [67–69] by diphenylacetylene or per(pentafluorophenyl)acetylene [70]. The complexes of dialkylacetylenes and diphenylacetylene were prepared either by the displacement of ethylene from
4.3
Ligands in the Coordination Sphere of Titanocenes
163
[Cp*2Ti(η2-C2H4)] [35] or by acetylene coordination to further titanocene generating systems (L2TiCl2/Mg/THF) L: C5HxMe5–x (x = 0–5) [71–74, 31]. As excellent π-acceptors, alkynes can act as a four- as well as a two-electron ligand. While the shift in the νC≡C frequency on coordination of the alkyne proves to be rather insensitive to the coordination mode of the alkyne, the 13C NMR chemical shift is apparently a better guide. As a general rule δC≡C values of 100–150 are taken to indicate that alkyne acts as two-electron donor, for example when coordinated to nickel complexes, whereas alkynes as four-electron ligands give δC≡C values of 190– 270, when coordinated for example to titanium. In agreement with this differing πacceptor behavior, the chemical shift differences between free and coordinated disubstituted alkynes C2R2 are much higher for early transition metals (∆δC≡C = 106– 140) than for nickel (∆δC≡C = 38–55). The complex [Cp2Ti(η2-PhC≡CPh)] has been isolated by reduction of the [Cp2TiCl2] with magnesium in the presence of the alkyne as a brown microcrystalline solid which decompose at 117–119°C. The extensive backbonding contribution to the alkyne in this complexes is confirmed by the low 13C chemical shift (δC≡C = 196) as well as the low νC≡C IR frequency of 1712 cm–1, which has led to the consideration of a metallacyclopropenylium resonance structure.
Scheme 4-7. Titanocene alkyne complexes. The dialkylacetylene and diphenylacetylene complexes are highly reactive towards acetylenes to give titanacyclopentadiene derivatives and low yields of catalytically formed hexasubstituted benzenes. Thus, only the [L2Ti(η2-PhC≡CPh)] (L: Cp, Cp*) complexes were isolated and characterized by spectroscopic and chemical means [35, 72]. In contrast complexes with mono(trimethylsilyl)acetylene ligands RC≡CSiMe3 (R=Ph, tBu,nBu) as well as the bis(trimethylsilyl)acetylene complexes [Cp2Ti(η2-Me3SiC≡CSiMe3)] [73] and [Cp*2Ti(η2-Me3SiC≡CSiMe3)] [74] are more stable and were obtained in high yields. The X-ray crystal data were published for the latter complex and both complexes were characterized by spectro-
164
4
Titanocenes
scopic data [75]. These two complexes do not form titanacyclopentadiene complexes and do not undergo addition reactions. Instead the silylalkyne Me3SiC≡CSiMe3 is displaced by more strongly coordinating reagents [31]. Table 4-3.
Selected analytical data of acetylene complexes [L2Ti(η2-RC≡CR′)] (L = Cp, Cp*)
4.3
Ligands in the Coordination Sphere of Titanocenes
165
The silylalkyne complexes [Cp2Ti(η2-Me3SiC≡CSiMe3)] and [Cp*2Ti(η2Me3SiC≡CSiMe3)] are suitable reagents for the generation of [Cp2Ti] and [Cp*2Ti] species because they are (a) simple prepared (b) thermally stable at ambient temperature and (c) the coordinated Me3SiC≡CSiMe3 does not undergo addition reactions with an excess of substituted acetylenes.
4.3.3 Carbonyls The synthesis and reactivity of carbonyl complexes of titanium has been reviewed extensively [2, 79], and detailed syntheses of [Cp2Ti(CO)2] and [Cp*2Ti(CO)2] have been described [80]. The mixed-ligand complex [CpCp*Ti(CO)2] is similarly prepared by reduction of the corresponding titanocene dichloride with zinc powder. The electronic structure of [Cp2Ti(CO)2] has been researched by gas-phase HeI and HeII photoelectron spectroscopy and Xa molecular orbital calculations. Of high diagnostic value are the νCO frequencies of some carbonyl derivatives as summarized in Table 4-4.
Table 4-4.
Infrared spectroscopic data for bis(cyclopentadienyl)titanium carbonyl complexes
Kinetic studies were performed for CO substitution reactions of [(η5-L)2Ti(CO)2] derivatives, where (η5-L) = Cp, Cp* or Ind [82]. Nucleophiles used for these reactions include PMe2Ph, PMePh2, PPh3, P(OEt)3, P(n-Bu)3, and CO. The reaction rates of the titanium compound were found to be first order in the substrate and zero order in the entering nucleophile at nucleophile concentrations which gave the limiting reaction rate. The results indicate a dissociative mechanism. Activation parameters for these reactions are also in agreement with a dissociative process. In contrast, reaction rates of the zirconium and hafnium compounds are also first order in both substrate and entering nucleophile, indicating an associative mechanism. This mechanistic difference may be attributed to steric considerations caused by the smaller size of titanium [82].
166
4
Titanocenes
4.3.4 Phosphines [Cp2Ti(PMe3)2] is prepared in high yield by reduction of [Cp2TiCl2] with magnesium in THF in the presence of PMe3 [83–85]. It has proved to be a very versatile starting material for a wide range of reactions. In such reactions the PMe3 ligands are easily displaced under milder conditions than from [Cp2Ti(CO)2]. Both ligand exchange and oxidative addition reactions are common.
4.3.5 Hydrides The first structurally characterized monomeric bis(cyclopentadienyl)titanium(III) hydride, [(PhC5Me4)2TiH], was synthesized by hydrogenolysis of [(PhC5Me4)2TiMe]. This hydride was found to be a monomeric bent sandwich complex by X-ray diffraction methods, and it can be concluded that the [Cp*2TiH] analog possesses a similar molecular structure by comparison of the spectroscopic and reactivity data [86]. A red–brown solution of the hydride is formed by exposing a green pentane solution of [(PhC5Me4)2TiMe] to 1 atm H2 at room temperature, from which red–brown crystals can be separated. The Cp–Ti–Cp angle (Cp– Centroid) is found to be very large (150.84°), which apparently occurs because the hydride ligand is very small. The Ti–H distance (1.768(15) Å) is comparable to other monomeric hydrides. Table 4-5. Selected structural data of titanium hydrides
with PbCl2 yields As expected, oxidation of [(PhC5Me4)2TiH] [(PhC5Me4)2Ti(H)Cl] (δ Ti–H = 5.16), while the reaction of [Cp*2TiH] gives [Cp*2Ti(H)Cl] [86].
4.3
Ligands in the Coordination Sphere of Titanocenes
167
4.3.6 Titanium Ligand Multiple Bonds 4.3.6.1 Carbene Complexes The chemistry of transition-metal alkylidenes and metallacycles has attracted much recent attention and has resulted in a number of useful synthetic transformations [39, 93]. Titanocene carbene complexes can be synthesized via routes starting from the Tebbe reagent (b), [94–97] from titanacyclobutanes [98–100] (c) and from alkyl derivatives (a) [101–115], which act as sources of the parent titanium methylidene derivatives (Scheme 4-8) [116, 117]. These carbene complexes normally act as intermediates in subsequent reactions [118], but also a few examples of titanium group metal carbene complexes have been isolated and characterized [119– 121]. The variety of reactions available for the formation of carbene complexes of electron-deficient transition metals indicates a high synthetic potential. Preparative applications of carbene complexes generated by path a, as opposed to those obtained by path b, frequently give higher yields and selectivities, and a wider choice of reaction media and substituents on the carbene ligand as well as more tolerant functional groups, when used in stoichiometric and catalytic reactions.
Scheme 4-8. Possible synthetic pathways (a–f) for complexes with Ti–C double bonds. Even though α-H eliminations have been well-established since the discovery of Schrock carbenes, this method (a) for the generation of [L2Ti=CR2] complexes has only recently been successfully applied to electron-deficient transition metal complexes. This is in contrast to the extensive use of dimetallic species, of which for
168
4
Titanocenes
example the Tebbe reagent (path b) has assumed a prominent role in organic chemistry. Also, dititanacyclobutane a bismethylene complex reacts rapidly and quantitatively with PMe3 to yield [Cp2Ti=CH2(PMe3)] [117]. The cleavage of bis-µmethylene complexes fails, however, for other metal combinations, such as Ti–Zr, Zr–Zr, or Ti–Si [117, 122–125]. Metallacyclobutanes derived from the Tebbe reagent have also enhanced our knowledge of carbenoid transition metal compounds (path c). The use of titanacyclobutanes as the starting material has the advantage over the Tebbe reagent that additional Lewis acidic components such as [Me2AlCl·L] do not have to be separated off after the reaction. For the synthesis of titanacyclobutanes, procedures can be employed that were developed by Grubbs, using the Tebbe reagent [51, 126– 138], by Bickelhaupt [139], preferably using 1,3-di-Grignard species, or by Stryker [140], employing allyl complexes. Titanacyclobutanes can also be used to generate substituted [131] or cumulogous [Cp2Ti=CR2] species [38]. There is a problem, however, namely that singly substituted titanacyclobutanes such as α,β-dimethyltitanacyclobutane react preferably under elimination of butene [141] instead of propene. Hence, ‘productive’ cleavage reactions are required (Scheme 4-9).
Scheme 4-9. ‘Productive′ and ‘unproductive′ cleavage of metallacyclobutanes. A ‘productive′ cleavage occurs when highly strained cyclic olefins like 3,3-dimethylcyclopropenes are used in conjunction with Tebbe′s reagent 3. The resulting titanacycle 4 is capable of transferring α-substituted carbenes in carbonyl olefinations [131]. The reactive titanium carbene can be stabilized with PMe3 or PMe2Ph (5).
Analogous to many applications of salt metathesis reactions for generation of simple organometallic compounds, geminal dimetallic species can be used to prepare carbene complexes (path d). Butyllithium reacts under mild conditions with diphenylcyclopropene to yield 1,1′-dilithio-3,3-diphenylallene derivative, which can react with [Cp2TiCl2] in the presence of PMe3 to yield the first titanabutatriene derivative 6 [52].
4.3
Ligands in the Coordination Sphere of Titanocenes
169
The reaction of [Cp*2Ti(η2-C2H4)] with a diazoalkane leads to a titanium carbene complex that can be subsequently trapped (path e) [142]. The reaction of cyclopropenes with low-valent metal complex fragments leads to vinylcarbene complexes (path f). Fundamental studies were done by Binger et al., who synthesized 7 in 72% yield [143].
If methylenecyclopropane is used instead of cyclopropenes, the bis-titanacumulene [Cp2Ti(PMe3)=C=C=Ti(PMe3)Cp2] 8 is formed. This cumulene is characterized by a linear [Ti=C=C=Ti] chain [144].
170
4
Table 4-6.
Titanocenes
Selected analytical data of complexes with Ti–C multiple bonds (compare Scheme 4-10)
Scheme 4-10. Examples of isolated titanium carbene complexes.
4.3
Ligands in the Coordination Sphere of Titanocenes
171
The M–C bonds of a few structurally characterized [LnTi=CR2] complexes are shortened to different extents depending on the ligand system. The Ti=C bond length (1.911(3) Å) in 10 is significantly shorter than that in the bisneopentyl starting material (2.120(4) Å) [119]. The Ti=C–C angle of 158.7(2)° and the Ti=C–H angle of 85(3)°, in combination with a Ti–H distance of 2.05(5) Å, indicate an αagostic interaction, which is typical for d0 alkylidenes. The 1H resonance signals at low fields (δ = 11–13 ) and the 13C resonance signals of the carbene C atoms (δ = 265–313) can be considered as characteristic signals of an alkylidene group.
4.3.6.2 Complexes Containing Ti–X Double Bonds (X: N, P, Chalcogenides)
α-H transfers can be used to generate multiple bonds between not only transition metals and carbon, but also between metals and other heteroatoms. In this way, synthetic building blocks incorporating Ti=Si [145], and Ti=N [146] can be obtained. Permethyltitanocene hydride [Cp*2TiH] 13 reacts with elemental selenium or tellurium to give the products [(Cp*2Ti)2(µ-E)] (E = Se, Te) (14) [147]. These dinuclear compounds are paramagnetic and have D2d (idealized) structures, as shown by Xray structural analysis. They may be converted to diamagnetic dichalcogenides through further reaction with the appropriate chalcogen. Derivatives 15 are monomeric in the solid state and in solution.
In the context of these investigations it becomes obvious that in the series [Cp*2Ti=E], the singlet triplet gap is substantial. The Ti–O bond is very strong, and because oxygen is much smaller than tellurium, pyridine can get in to trap the terminal oxo derivative effectively [148]. The oxidation of [Cp*2Ti(N2)] with N2O in toluene gives a oxygen-bridged dimer (16), exhibiting a fulvene substructure [149]. On the other hand, the oxidation of [Cp*2Ti(η2-C2H4)] with N2O in the presence of pyridine as a Lewis base give the diamagnetic orange complex [Cp*2TiO(py)] (17). The Ti–O stretching vibration is observed at 852 cm–1 which shifts to 818 cm–1 on labeling with 18O. The crystal structure confirms the presence of a terminal, very short Ti–O bond of 1.665(3) Å with substantial multiple-bond character [148]. This oxo compound 17 undergoes a [2+2]
172
4
Titanocenes
cycloaddition reaction with allene to give [Cp*2TiOC(=CH2)CH2] (18), which contains a puckered TiOC2 ring in the solid state. This titanacycle does not exchange with either free allene or pyridine on the NMR time scale and slowly converts to the ring-metallated product [Cp*(C5Me4CH2)TiOC(=CH2)CH3] (19) in solution and to the enolate complex [Cp*2Ti(H)(OC(=CH2)CH3] (20) in the presence of 1 atm H2 [150].
Scheme 4-11. Oxidation products of [Cp*2Ti] generating complexes [(Cp*2Ti(η2-C2H4)], [Cp*2Ti(N2))]; formation of [Cp*2TiO(py)] and selected subsequent reactions.
4.3
Ligands in the Coordination Sphere of Titanocenes
173
4.3.7 Metallocene Complexes of the Type [Cp2TiL2]2+ 4.3.7.1 Aqua–Titanocene Complexes Cationic titanocene–aqua complexes can be readily prepared from the corresponding bishalides by reaction with silver complexes [151], or as shown earlier, by substitution reactions of weakly coordinated ligands in [Cp2TiX2] complexes (X = NO3– [152, 153], ClO4– [154]). By the latter method, the complexes [Cp2Ti(H2O)2](NO3)2 and [Cp2Ti(H2O)2](ClO4)2·3C4H8O are prepared. [Cp*2Ti(H2O)2](CF3SO3)2 becomes available from Cp*2TiCl2 and AgCF3SO3 in the presence of small amounts of water as dark–violet crystals. This complex can be handled in air and is soluble in moderately polar solvents such as CH2Cl2 and CH3NO2, but is not soluble in nonpolar solvents such as benzene. From solutions of [Cp*2Ti(H2O)2](CF3SO3)2 the two salts [Cp*2Ti(OH)(H2O)](CF3SO3)2 · H2O and [Cp*2Ti(OH)(H2O)](CF3SO3)2 ·2 H2O can be isolated. The [Cp*2Ti(H2O)2]2+ and the [Cp*2Ti(OH)(H2O)]+ cations show the expected, nearly tetrahedral geometry. The Ti–OH2 distances in the diaqua cation are 2.06(1) and 2.09(1) Å. The Ti–OH2 and Ti–OH distances in the aqua–hydroxo cations are 2.11 and 1.87 Å respectively [151]. The diaqua complexes of bent metallocenes tend to be moderately strong acids, and in this respect they are excellent catalysts in Diels–Alder reactions. For example, the reaction of 3-pentenone with isoprene in the presence of [Cp*2Ti(H2O)2](CF3SO3)2 as catalyst (CDCl2, 25°C) proceeds 6 × 103 times faster than the thermal reaction [155].
Scheme 4-12. Diels–Alder catalysis by [Cp*2Ti(H2O)2](CF3SO3)2. The reaction of [Cp2TiCl2] with Na[BPh4] in water leads to the aqua–metallocene derivative [Cp2Ti(H2O)6][BPh4], which undergo a facile reaction with methanol under loss of water. According to the single crystal X-ray diffraction study, one product turns out to be the salt [Ti3(µ3-O)(µ2-OCH3)3(OCH3)3(η5-C5H5)3][BPh4], in which the µ3-O atom is located at the apex of a trigonal Ti3O pyramid [156]. The aqueous chemistry of titanocenes was of interest for a long time, particularly in the field of cancerostatic activities [157–160]. However, complexes of titanocene are hydrolyzed in aqueous solution, leading to insoluble polymeric [(CpTiO)4O2]n at pH 5.5 [161, 162].
174
4
Titanocenes
4.3.7.2 Titanocene Complexes with Highly Fluorinated Ligand Systems The reaction of [Cp2TiCl2] with two equivalents of [AgEF6] (E = As, Sb) in liquid sulfur dioxide led to the quantitative preparation of the hexafluoropnicogenate complexes [Cp2Ti(AsF6)2] and [Cp2Ti(AsF6)2] [163]. By reaction with the neutral Lewis base CH3CN the cationic metallocene derivatives [Cp2Ti(CH3CN)n]2+[AsF6]2– could be obtained [164]. If [Cp2TiF2] is reacted with two equivalents of SbF5 in liquid sulfur dioxide, [Cp2Ti(SbF6)2] is formed quantitatively. The equilibrium in SO2 solution of [Cp2Ti(SbF6)2] with SbF5 is completely shifted to [Cp2Ti(Sb2F11)2] [165].
Scheme 4-13. Synthesis of highly fluorinated titanocene complexes (E = As, Sb, Bi). The preparation of the fluorine-coordinated metallocene hexafluorobismuthate complex [Cp2Ti(BiF6)2] proceeds from the reaction of [Cp2TiF2] with BiF5 in SO2ClF in the presence of anhydrous HF, followed by recrystallization from liquid SO2. IR investigations clearly indicate that the BiF6 unit is not octahedral but is strongly distorted, and that both bridging (Ti···F···Bi) and nonbridging (Bi–F) fluorines are present [166].
4.3.7.3 Further Complexes with [Titanocene]2+ (d0) and [Titanocene]+ (d1) Units The reaction of [Cp2Ti(CF3SO3)2] with 2,2-bipyridine or 1,10-phenanthroline yields the salts [Cp2Ti(bipy)]2+(CF3SO3)–2 and [Cp2Ti(phen)]2+(CF3SO3)–2. X-ray structure determinations show that the Ti atoms, which are to a first approximation tetrahedrally coordinated, and the atoms of the bipy or phen chelate ligands, are nearly coplanar [167]. However, from reaction of [Cp2Ti(η2-Me3SiC≡CSiMe3)] with trimethylammoniumtetraphenylborate cationic d1-titanocene complexes [{Cp2Ti(L)2+}{BPh4–}] with L = THF (21) and pyridine (22) become available via 1e′-oxidation of the 14e′-Cp2Ti-unit to paramagnetic Ti(III) complexes under evolution of hydrogen [168]. Complex 22 is one of the rare examples of such cationiconly neutral-ligand containing complexes without anionic ligands that have been characterized by X-ray structure analysis.
4.4
Reactions in the Coordination Sphere of Titanocenes
175
4.4 Reactions in the Coordination Sphere of Titanocenes 4.4.1 The free Titanocenes For the rearrangement of [(η5-C5H5)2Ti] (23) to the dimer 26, Brintzinger and Bercaw suggested formation of an η1-C5H5 titanium complex (24), followed by α-H abstraction to give the titanafulvalene 26 [169]. Neither 24 nor 25 was observed.
Scheme 4-14. Formation of µ-(η5:η5-fulvalene)-di(µ-hydrido)-bis(η5-cyclopentadienyl)titanium 26. Bright green ‘titanocene′ (26) was first reported in 1956 to be obtained from the reaction of TiCl2 with CpNa [1]. The same bright green material was later obtained by reduction of [Cp2TiCl2] with sodium naphthalene in tetrahydrofuran [170] or with sodium metal in hydrocarbons [171, 172], or by hydrogenolysis of [Cp2TiMe2] [173]. Its chemical properties were, however, different to the expected chemical behavior of monomeric titanocene and the structure of the compound was subject to extensive discussion. Strong evidence for the dimeric structure was given by Brintzinger and Bercaw [169] and the presence of the fulvalene ligand was established by an X-ray crystallographic study of the compound arising from interaction of 26 with Et3Al [174]. The fulvalene structure of the compound was established on the basis of the 13C NMR spectrum [175] and from the X-ray study of its di-(µ-hydroxyl) derivative [176]. Later, the X-ray study of a purple product arising from the reaction 26 with HCl revealed that it is its di(µ-chloro) derivative [177]. The mixed (µ-hydrido)–(µ-chloro) complex was also prepared [178] and its X-ray structure was determined [179]. Attempts to determine the structure of 26 directly by X-ray crystallographic studies have been unsuccessful for some time [174, 176]. In 1992, Mach, Antropiusova and Troyanov were able to solve the X- ray structure of 26 [90]. Selected structural data of fulvalene titanium complexes (‘titanocene’) are given in Table 4-7.
176
4
Titanocenes
Table 4-7. Structural data in dititanium fulvalene (C10H8)[(C5H5)TiX]2 complexes
If the dimeric ‘titanocene’ (26) is reacted with HCl the dichloro compound 27 is formed, from which organyl derivatives (28) are available in reaction with LiR (R = Me [180], R = Ph [181]) . The X-ray structure confirms the presence of the fulvalene bridge, in which the CpTiMe2 fragments are trans with respect to the fulvalene group [180].
Scheme 4-15. Reactions of the green ‘titanocene’. From the dimeric ‘titanocene’ 26 the air-stable compound [CpTi(H2O)](C10H8,O)]2+(SO3CF3)2– as the first example of an ionic titanafulvene complex can be prepared in reaction with trifluoromethanesulfonic acid and small amounts of water [182]. In the chemistry of permethylated titanocenes, intramolecular C–H activation reactions dominate. When permethylated titanocene complexes are thermolyzed, the formation of carbene [39, 183] or aryne [184] intermediates occurs by H-transformation between the R groups, which react intramolecularly to form fulvene complexes e.g. 30, the so called ‘tuck in’ derivatives. Further heating leads to liberation of a further hydrocarbon molecule producing the bis ‘tuck in’ complex 31 [41].
4.4
Reactions in the Coordination Sphere of Titanocenes
177
However, depending on the number of methyl groups (n) on the Cp rings, the formation of both fulvalene and fulvene complexes is observed. Those with n = 0–3 (µ-η5:η5-fulvalene)(di-µ-hydrido)bis(η5-cyclopentadienyltitanium) complexes and its methylated analogs are formed. For C5Me4H (Cp″″) complexes mixtures of fulvalene and fulvene complexes are observed [71]. Table 4-8. NMR data of fulvene titanium complexes
The proton NMR spectra of the fulvene complexes shows two doublets with coupling constants 2JHH = 4 Hz assigned to the diastereotopic hydrogens of the CH2 group. This coupling constant is more consistent with the geminal sp2 hydrogens of structure B, normally 0–3 Hz, than with the geminal sp3 hydrogens of structure A,
178
4
Titanocenes
normally 12–15 Hz [41, 183–188]. The carbon of the =CH2-fulvene group of 30 resonates at δ = 73.9 with 1JCH = 149–160 Hz, values consistent with sp2-hybridized carbon. However, the crystal structure of the paramagnetic [Cp*(η6C5Me4CH2)Ti] confirm the π-η5: σ-η1 bonding mode for the tetramethylfulvene ligand (A) [91b]. A diene like bonding mode is found by X-ray structure determination in the case of bis-fulvene complexes [91a]. 4.4.2 Reactions of Ti–C σ Bonds Our understanding of the chemistry of the transition metal–carbon σ bond is improved by investigations of the chemistry of 1-alkenyl complexes of electron poor transition metals. There is no other system known in which we can easily switch between the possible reaction pathways, depending on the nature of the metal, the ligands L and the 1-alkenylgroup. Only reductive elimination, α- and β-H elimination reactions give high selectivity. α-H Elimination from [Cp*2Ti(CH=CH2)R] derivatives leads to the versatile titana–allene intermediate [Cp*2Ti=C=CH2], from which a wide range of cycloaddition products of high thermal stability can be prepared. By comparison of the thermal stability of the vinyl, phenyl and alkyl complexes of titanium group metals, it becomes obvious that the vinyl derivative is the most reactive [190]. As well as α-elimination processes, β- and reductive elimination reactions are also observed. In each case, only one reaction pathway results in high selectivity (Scheme 4-16).
Scheme 4-16. Reaction pathways of 1-alkenyl transition metal compounds.
Thermodynamic considerations suggest that the reductive elimination of vinyl groups should be preferred, as is shown, for example, by ab initio calculations on Cl2Ti(CH=CH2)2 model complexes (Scheme 4-17) [191].
4.4
Reactions in the Coordination Sphere of Titanocenes
179
Scheme 4-17. Results from ab initio calculations for [Cl2Ti(CH=CH2)2] transformation products (pseudo potential basis set according to Stevens, Basch, and Krauss).
Kinetic factors are able, however, to force the reaction to go in different directions, especially when the metal center is not readily reduced, when hydrocarbon ligands with acidic hydrogen atoms are employed, or when only limited rotation of 1-alkenyl groups is possible [39].
4.4.2.1 Reductive Elimination Generally, reductive elimination reactions are the most preferred reactions of organometallic complexes from the thermodynamic point of view [191]. However, for alkyl derivatives, H-transformation reactions are kinetically preferred . In the case of 1-alkenyl compounds of the titanium group metals we find that there is a correlation between the rotational barriers of the 1-alkenyl ligand about the M–C σ bond and the observed reaction pathways (Scheme 4-18). If free rotation is possible, reductive elimination products dominate. Due to the orientation of the acceptor orbitals of the bent metallocene fragment in the equatorial plane between the Cp-ligands, electron transfer from the C=C double bond to the transition metal center in this 16 electron complex becomes possible if the vinyl group is orientated perpendicular to it (Scheme 4-18, A). This rotameric orientation leads to a suitable transition state for reductive elimination, due to the differences in partial charge of the α-carbon atoms, as calculated by ab initio methods [191]. Generally, for d0 systems, concerted reductive elimination reactions are symmetry forbidden, although reductive elimination might actually be possible, especially if charge transfer processes are involved [192, 193]. On the other hand, if
180
4
Titanocenes
Scheme 4-18. Proposed transition states of reductive elimination A, α-H elimination B and β-H elimination C from Cp2M alkenyl derivatives (M = Ti, Zr, Hf). this rotation is hindered by using bulky ligands (Cp* instead of Cp [194], or substituted 1-alkenyl groups instead of the simple CH=CH2-ligand [195]) C–H bond activation reactions become dominating (Scheme 4-18, B, C). Whereas the α-CH bond is activated in titanium complexes, β-CH activation occurs in zirconium complexes [37].
4.4.2.2 Oxidative Addition A small number of well characterized oxidative addition reactions (by splitting the R–X bond) involving the [Cp2Ti] unit are known. For example the [Cp*2Ti] fragment, generated from [Cp*2Ti(η2-C2H4)], reacts by oxidative addition with CH3I to form [Cp*2TiI(Me)] [61]. Also, in reaction of [Cp*2Ti], generated from [Cp*2Ti(Me3SiC≡CSiMe3)], with methanol the hydride complex [Cp*2Ti(H)OMe] can be isolated (δ Ti–H = 3.33, OMe = 4.01, Cp* = 1.89) [42]. This type of reaction proceeds only in the case of permethylated titanocenes. Using the nonmethylated derivative [Cp2Ti(Me3SiC≡CSiMe3)], the liberation of hydrogen and a binuclear Ti(III) complex [Cp2TiOR]2 is observed [196]. By using alkynes other than the silylalkyne the protonation of the acetylene ligand to 1-alkenylcomplexes is observed [40].
However, not only polar R–X can be added oxidatively to titanocene fragments. In the reaction of ‘Cp2Ti’ with RC≡C–C≡CR′ bimetallic complexes can be formed with or without splitting of the internal C–C single bond [31]. The obtained reaction products contain intact 1,4-disubstituted µ-η(1,3),η(2–4)-trans,trans-butadiene units between the two metal centers (‘zig-zag-butadiynes’) (32) [197], or alternatively, doubly σ,π-acetylide bridged complexes, Cp2Ti(µ-η1:η2-C≡CR)(µ-η1:η2C≡CR′)MCp2 (33) are obtained [198].
4.4
Reactions in the Coordination Sphere of Titanocenes
181
4.4.2.3 α-H Elimination
α-H eliminations proceed especially readily when there is significant steric crowding and where it is possible to transfer an α hydrogen atom to a carbanionic acceptor ligand. The reaction is intramolecular, and first-order reaction kinetics are observed for organometallic complexes of the titanium group metals. The breaking of the C–H bond is therefore shown to be the rate-determining step with either a cyclic transition state (path A) or a hydride intermediate (path B), where a titanium carbene complex, is formed as an intermediate (Scheme 4-19) [39].
Scheme 4-19. The reaction course of α-H eliminations. The activation enthalpy and entropy are characteristic of σ bond metathesis, which is identified by negative values for ∆S≠, that is, by a high degree of order in the rate-determining step. A comparison of activation enthalpies (Table 4-9) shows that, as expected, α-H abstractions occur more readily for neopentyl species [103] than for methyl derivatives [183]. The values of the activation entropies indicate a higher degree of order, corresponding to a lower mobility, for neopentyl, aryl, and
182
4
Titanocenes
vinyl derivatives than for methyl or benzyl compounds. The strong isotope effects kH/D, in addition to the negative ∆S≠ values are in agreement with a transition state (path A) [118, 183, 184, 199–201]. In contrast to this, hydride intermediates (path B) show kH/D values close to 1 [202]. The half-life for the thermolysis of [Cp2Ti(CH2CMe3)2] at 20°C is between 20 and 56 min. The [Cp2Ti=CHMe3] formed can be trapped with phosphanes or used in further reactions [103]. The steric overloading favorable for α-eliminations can be achieved either by organic moieties (e.g. neopentyl groups) or by other bulky ligands [39]. Table 4-9. Selected kinetic parameters of H eliminations
The high selectivity of the methane elimination from the vinyltitanium compound [Cp*2Ti(CH=CH2)CH3] is of particular value in the generation of the vinylidene intermediate [Cp*2Ti=C=CH2] (35) [184]. The spontaneous intramolecular H transfer between the two vinyl groups in [Cp*2Ti(CH=CH2)2] (34) had been discovered earlier [37], but proceeds under conditions that prevent the isolation of the bisvinyl species. The resulting vinylidene–ethylene complex 36 reacts quantitatively to yield the methylenetitanacyclobutane 37 [190, 203].
Scheme 4-20. Synthesis of methylentitanacyclobutane 37 by vinyl–vinylidene rearrangement of 34.
4.4
Reactions in the Coordination Sphere of Titanocenes
183
The cyclic transition state necessary for the generation of 36 from 34 (Scheme 419, path A) can only be formed if the vinyl group is present as the so called C–H inside rotamer (Scheme 4-21, B). Nuclear Overhauser effects (NOE) measurements, however, show the orientation of the vinyl group in [Cp*2Ti(CH=CH2)CH3] to be exclusively that of the C–H outside rotamer (Scheme 4-21, A).
Scheme 4-21. Orientation of 1-alkenyl ligands. The energy necessary for the rotation process in [Cp*2Ti(CH=CH2)2], is found from MMX-force field calculation to be about 52.1 kJ mol–1, compared to [Cp2Ti(CH=CH2)2] for which a value of 18.3 kJ mol–1 is calculated [191]. Therefore it can be concluded that the rotation barrier is the main contribution to the activation energy of C–H elimination as determined by kinetic measurements for the process 38→35→39 (87.9(5) kJ mol–1) [184]. The alternative elimination of ethylene and formation of a [Cp*2Ti=CH2] intermediate is not observed.
Selective α-H eliminations can also be used to generate a wide range of substituted alkenylating reagents [LnTi=CHR] (R = C6H5 [112], tBu [119], SiMe3 [111, 107]). For instance, during the thermolysis of [Cp2Ti(CH2C6H5)2] [204], more toluene (αH elimination product) is formed than dibenzyl (reductive elimination product; toluene:dibenzyl = 9.4:1) [112]. As the reaction of bicyclopropyltitanocene shows [110], even β-H containing compounds with strained ring systems can be used for ‘alkylidencyclopropanation’ in carbonyl olefination reactions.
184
4
Titanocenes
Scheme 4-22. Alkylidenecyclopropanations via Ti=C intermediates.
4.4.2.4 β-H Elimination The major decomposition pathway for transition metal alkyls is the β-H elimination, which converts a metal alkyl into a hydrido metal alkene complex (Scheme 423).
Scheme 4-23. β-H elimination. Such examples are also known in the case of titanocene alkyl complexes. Attempts to prepare [Cp*2Ti(Et)R] derivatives (41) by substitution of chlorine in [Cp*2Ti(Et)Cl] (40) by salt metathesis with alkali-metal reagents RM were not successful. In general, there was no reaction below –10°C, but the products are too thermally unstable to allow isolation at higher temperatures. Reaction of 40 with MeLi, KCH2Ph or LiCH=CH2 resulted in the formation of the permethylated ethene adduct [Cp*2Ti(η2-C2H4)] (42) and RH in essentially quantitative yields [184].
In order to determine whether the α-hydrogen atoms are involved, [(Cp*d15)2Ti(CD2CH3)Cl] was prepared and treated with MeLi. Only CH4 (1mol/Ti) was formed, so that hydrogen from the α position of the ethyl ligand or from the pentamethylcyclopentadienyl ligands can be ruled out as the source of the hydrogen atom used in the formation of methane. Apparently, β-hydrogen transfer from the ethyl ligand to the leaving group R in Cp*2Ti(Et)R has a very low activation energy.
4.4
Reactions in the Coordination Sphere of Titanocenes
185
The thermolysis of [Cp*2Ti(Ph)Me] followed first-order kinetics for at least 3 half-lives, and the reaction rate appeared to be independent of the concentration. The kinetic isotope effect observed (kH/D = 5.1 at 80°C) indicates that a phenylic C–H bond is broken in the rate-determining step. The aryne intermediate [Cp*2Ti(η2-C6H4)] formed can by trapped with carbon dioxide [184]. The intramolecular decomposition via an ordered transition state, i.e., σ-bond metathesis, is in sharp contrast with the radical decomposition of other mixed alkyl/aryl compounds. The formation of the aryne intermediate [Cp*2Ti(η2-C6H4)] is analogous with the reported reactivity of [Cp2Ti(C6H5)2], which leads to the o-phenylene intermediate [Cp2Ti(η2-C6H4)] [205, 206]. 4.4.3 Reactions of Ti=C Double Bonds Kinetic studies and numerous trapping experiments have shown that thermally generated [L2Ti=CH2] and [L2Ti=C=CH2] species represent real intermediates [41, 51, 118]. In particular, the existence of the titana–allene 35 as a real intermediate, generated from 38 by methane elimination (5–20°C) or from 37 by ethylene liberation (70–100°C) [39, 190, 203, 207] can be proved by simple trapping experiments. Substrates such as ketones, alcohols, cumulenes and heterocumulenes do not react directly with the starting materials. Substitution products 45 or ring-opened derivatives 43 could not be detected in reactions of 37 and 38 with acidic substrates. This is an indication that cycloreversion or α-H transfer from 38 take place before electrophilic addition reactions can occur. However, strong acidic substrates like thiophenol lead to formation of products of type 43 [41].
Scheme 4-24. Selective trapping experiments of [Cp*2Ti=C=CH2].
186
4
Titanocenes
4.4.3.1 Reactions with Electrophiles In view of the nucleophilic properties of the α-atoms in [LnTi=C] derivatives, rapid reactions with electrophiles are expected. If 35 is allowed to react with a stoichiometric amount of water, the monomeric vinyltitaniumhydroxide [Cp*2Ti(CH=CH2)OH] (δ OH = 7.10; IR νOH = 3657 cm–1) is obtained in 86% yield as a yellow, microcrystalline product [41]. The selective protonation of the vinylidene species 35 can also be transferred to C–H [208] and N–H [209] acidic compounds. In such a way the reaction of 38 with benzophenonimine in n-hexane as solvent yields the azavinylidene [Cp*2Ti(CH=CH2)(N=CPh)2] 46 in the form of light yellow crystals (55%, mp 190°C (dec.)) [209].
Fig. 4-2. Molecular structure of 46 in the solid state. Selected bond distances (Å) and angles (°): Ti–N 1.901(2), Ti–C3 2.159(2), C2–C31 1.319(3), N–C 1.259(3), Ti–N–C 179.1(2), Ti–C3–C31 131.8(2).
4.4
Reactions in the Coordination Sphere of Titanocenes
187
4.4.3.2 C–H Activation Reactions on [LnTi=C] Fragments In order to coordinatively saturate the [LnTi=C] derivatives, intramolecular C–H activation reactions occur in many cases. For example, the titana–allene 35 spontaneously reacts to yield the dark green vinylfulvene derivative [Cp*(η6C5Me4CH2)TiCH=CH2] 39 [184]. This behavior is similar with related ligands [105, 119, 183]. The signal of the methyne hydrogen atom of the vinyl group in 39 is shifted to higher field ( δ = 5.11) than in other vinyl compounds [210].
Starting from 47, via the titanium carbene [Cp2Ti=CH(CMe3)] (48), C–H bonds of benzene or p-xylene can undergo intermolecular activation and [CpTi(CH2CMe3)R′] complexes are formed [103].
Complex 47 reacts in neat C6D6 to produce quantitatively the [1-2H] neopentyl [2H5] phenyl complex 49 with the selective incorporation of one deuterium on the neopentyl α-carbon. Intermolecular activation of benzylic sp3-CH bonds was also observed. This behavior is comparable to d0 titanium imides [146, 211]. Overall, this reaction type corresponds to the reverse of the original formation reaction by α-H elimination.
4.4.3.3 [2+1] Additions Due to its electronic structure, the vinylidene intermediate 35 is very useful in cycloaddition reactions [38, 191]. By using isonitriles an azabutatriene complex 50 is formed in the first reaction step in a [2+1] cycloaddition [212, 213]. Due to the high reactivity of the intermediate 50 further reaction occurs, forming the five membered metallacycle 51, which exhibits a heteroradialene substructure [213], or by trapping with W(CO)6 the metallacycle 52 [212].
188
4
Titanocenes
Scheme 4-25. [2+1] Addition reactions of isonitriles with [Cp*2Ti=C=CH2]. Reactions of metal–carbon multiple bond systems with isocyanides present a potential synthetic method for the generation of keteneimines [214]. The coupling of a carbene ligand with carbon monoxide can be demonstrated for isolable carbene complexes with bulky, chelating phosphanealkoxides and a Cp ligand. Ketene complexes are formed in this reaction [119].
4.4.3.4 Cycloaddition Reactions with Acetylenes Titanacyclobutenes are found to be the products of [2+2] cycloadditions of metal carbenes with acetylenes [215, 216], as well as the ring-opening of cyclopropenes [217]. The reactivity of the titana–allene 35 with alkynes is particularly important with respect to the mechanisms of polymerizations of acetylenes [208]. Studies of the behavior of thermally generated 35 towards a number of alkynes have shown that the regiochemistry of the titanacyclobutene ring formation 53→54 is kinetically controlled and proceeds according to the polarity of the alkyne employed [208, 218].
4.4
Reactions in the Coordination Sphere of Titanocenes
189
If acidic (terminal) alkynes are used, a competing reaction occurs, namely the formation of vinyl titanium acetylides 55.
The use of unsubstituted acetylene (HC≡CH), however, leads exclusively to the metallacyclobutene 54 (R = H), a thermally stable compound. Competing reactions of acetylenes and olefins in cycloadditions with 35 yield exclusively the metallacyclobutene derivatives. Ab initio calculations for the system [Cl2Ti=C=CH2 + H2C=CH2] and [Cl2Ti=C=CH2 + HC≡CH] indicate a preference of 70.9 kJ mol–1 [191]. Kinetic studies have shown that the primary alkyne insertion into a Ti–CH3 bond with subsequent γ-elimination from 57 to give 58 could be an alternative to the direct [2+2] cycloaddition of the [Ti=C] species to the alkyne [109].
190
4
Titanocenes
Scheme 4-26. Formation of titanacyclobutenes, by [2+2] cycloaddition reaction as well as by γ-elimination reactions from 1-alkenyltitanocene complexes.
The reactivity of titanacyclobutenes is characterized by their ability in principle to undergo cycloreversions A or electrocyclic ring-opening reactions B. Depending on the substitution patterns, the reaction goes by path A or B (Scheme 4-27).
Scheme 4-27. Reactions of titanacyclobutenes. Alkynes with particularly bulky ligands (SiMe3) can be replaced by less sterically demanding ones [215]. This behavior is also observed for titanacyclobutenes generated from 35 [208]. For the titanacyclobutene exhibiting large substituents (α-R = SiMe3, β-R = Ph), a long internal C–C single bond of 1.502(7) Å is found by Xray structural analysis, and the reactivity is characterized by cycloreversion processes at higher temperatures. On the other hand, with smaller substituents (R = R = Me) the internal C–C single bond is found to be shorter, (1.434(4) Å) indicating a transition in the reactivity of the titanacyclobutene. Indeed, electrocyclic ring open-
4.4
Reactions in the Coordination Sphere of Titanocenes
191
ing reactions are only observed for the nonsubstituted titanacyclobutene, leading to the formation of polyacetylene if an excess of acetylene is used [208]. This behavior corresponds to that of tantalacyclobutenes, which can be isolated in the form of vinylcarbene complexes by reaction with pyridine [219]. Reactions of the titana–allene intermediate with 1,3-diynes, yielding titanacyclobutenes in a regioselective manner. Only one regioisomer (59) is formed, containing the acetylide group in the α position of the titanacyclic ring. The regioselectivity is in accordance with the polarities of the diynes and stereochemical conditions in the cyclobutene ring [220].
4.4.3.5 Synthesis and Reactivity of Metallaoxetanes Metallaoxetanes 60 are discussed as essential intermediates in the reactions of carbenoid metal compounds with carbonyl derivatives (60→61). In addition, metallaoxetanes can be considered as intermediates in the oxidation of olefins as well as in the formation and deoxygenation of epoxides [221]. Generally, metallaoxetanes have been prepared by two principal synthetic routes, by reaction of a carbene with a carbonyl functional group [222, 223] or by oxygen insertion into a coordinated olefin [224]. Additionally, the reverse reaction of a terminal metal oxide fragment (Cp*2Ti=O) and allene also leads to the formation of titanaoxetanes [148]. The formation of titanaoxetanes is also possible by the addition of two equivalents of dimethylsulfoxonium methylide to titanocene chloroacyl complexes i.e. the addition of a methylene fragment to a ketene complex [225].
Scheme 4-28. Reactivity pattern of metalloxetanes.
192
4
Titanocenes
The reactions of the vinylidene intermediate 35 with ketenes, isocyanates or carbon dioxide indicate the possibility of the synthesis of novel titanaoxetanes 63 with considerable thermal stability. The X-ray crystal structure analyses of 63b and 63d confirm the presence of planar, monomeric metallaoxetane rings.
Table 4-10. Selected properties of isolated titanaoxetanes 63
These cycloaddition products represent the first structurally characterized titanaoxetanes with a planar ring geometry. Considering that the driving force of carbonyl olefinations stems from the formation of a titanium oxide fragment [191, 227], the fact that planar oxetanes 63 can be isolated must be attributed to the reduced electrophilicity of the transition metal center due to the strongly basic Cp* ligands [53, 56]. This assumption is supported by the observation that titanaoxetanes similar to 63, but with the unsubstituted Cp ligands 65, have never been detected, even by spectroscopic means [228]. This reaction is very useful in the formation of substituted allenes 66.
4.4
Reactions in the Coordination Sphere of Titanocenes
193
Scheme 4-29. Synthesis of substituted allenes from vinylidene intermediates and carbonyl compounds. Depending on the resulting reaction pattern, these metallaoxetanes can be classified as ‘classical’ (60→61) or ‘nonclassical’ (60→62). Mass spectrometry shows the formation of [Cp*Ti=O] fragments for the metallaoxetanes 63, so that their classification as ‘classical’ oxetanes is justified. As well as the effect of the Cp* ligands, the position of the exocyclic double bond in the metallacyclic ring also has implications for the stability of metallaoxetanes 63. If the double bond is in the β position, such as in cycloaddition products of [Cp*2Ti=O] with allene, rapid ring opening occurs in solution, with the subsequent formation of fulvene enolates [150].
4.4.3.6
Synthesis of Heterodinuclear Carbene Complexes with a Titanaoxetane Substructure
The ‘nonclassical’ metallaoxetanes can be obtained by the use of metal carbonyls as heterocumulene units in reactions with 35 [229]. In contrast to metal-mediated cycloadditions of metal carbonyls with arynes, dienes [230], olefins [63] and heteroolefins [231] to metallocene fragments of group 4 metals the corresponding reactions of Schrock carbene fragments with metal carbonyls are little known thus far. In a few cases, corresponding intermediates have either been observed spectroscopically [232] or their existence has been postulated based on the products obtained [233]. However, the four membered heterodimetallic carbene complexes 67 are easily prepared by trapping the vinylidene titanocene fragment 35 with the corresponding metal carbonyl [229]. The spectroscopic data for 67 clearly indicate the titanaoxetane substructure. By using the binuclear transition metal carbonyls Mn2(CO)10 and Re2(CO)10, a selective axial functionalization occurs. The 13C NMR chemical shifts of the β-carbon atoms in the titanacycle are in the expected range of Fischer– Carbene complexes [234].
194
4
Titanocenes
Scheme 4-30. Top: Metal mediated cycloadditions to metal carbonyls; bottom: [2+2] cycloadditions of [LnM=CR2] to metal carbonyls. M = Ti, Zr; M’ = Cr, Mo, W, Mn, Re, Fe, Rh.
4.4
Reactions in the Coordination Sphere of Titanocenes
195
Table 4-11. Selected properties of nonclassical titanaoxetanes [Cp*2TiOC(=MLn)C=CH2], obtained from [Cp*2Ti=C=CH2] and transition metal carbonyls [229]
In the solid state, these ‘nonclassical’ oxetanes are stable up to 130°C. Solutions, however, show a much higher reactivity, which is quite in contrast to the ‘classical’ oxetanes 63. For example, the rapid release of the heterocumulene occurs even at low temperatures, especially when potential π-acceptor ligands (ethylene, isocyanates) or traces of moisture and oxygen are present. This means that the Fischer carbene complexes 67 with a titanaoxetane substructure represent another effective source of the Schrock carbene intermediate 35. A further nonclassical type of reaction behavior is observed for the titanaoxetanes 67, which leads to an unusual destabilization of the titanium vinylidene moiety, causing it to undergo a rapid vinylidene–acetylene rearrangement, leading to the five membered titanacycles 68 [229]. This transformation illustrates the behavior of the vinylidene :C=CH2 molecule in the gas phase. The five membered dinuclear complexes 68 are crystalline materials of high thermal stability (up to 220°C). In the 13C NMR spectra, signals typical of carbene C atoms are observed. The structures of the titanaoxacyclic carbene complexes are confirmed by single crystal X-ray diffraction. A nearly planar five membered titanaoxacyclic ring can be observed. The Ti–O and the Cr=C or Mn=C bond lengths are in the normal range of cyclic Fischer–carbene complexes. This reaction type is only observed in the case of the nonclassical bimetallic oxetanes 67. That means that a metal-mediated 1,2-H-shift must have taken place. In the first step, the vinylidene intermediate 69 is formed by cycloreversion of 67 followed by coordination of the metal–carbonyl to the C=C double bond of the vinylidene (70). This leads to a vinylidene–acetylene transformation (71). The coordinated acetylene then undergoes cycloaddition with the metal carbonyl [229].
196
4
Titanocenes
Scheme 4-31. Isomerization of nonclassical titanaoxetanes 67 to the five membered ring compounds 68—an unexpected vinylidene–acetylene rearrangement.
4.4.3.7 Synthesis and Reactivity of Heterotitanacyclobutenes and -butanes Cycloadditions of nitriles to Ti=C bond systems ought to lead to the formation of azatitanacyclobutenes (73). The electrophilicity of the titanium center, however, also influences the reaction behavior of the primary reaction product in this case [93], leading to products stemming from intermediary vinylimido derivatives 74 [113, 235].
Using [Cp*2Ti=C=CH2] (35) instead of the [Cp2Ti=CH2] intermediate, the preparation of azatitanacyclobutenes 75 is possible [236]. In n-hexane [Cp*2Ti(CH=CH2)CH3] (38) reacts with acetonitrile, pivalonitrile, cinnamonitrile
4.4
Reactions in the Coordination Sphere of Titanocenes
197
or 2-tert-butyl-1-phosphaacetylene at room temperature yielding the products 75 as well as 76 and its isomer 77.
The complexes 75b and 75c were isolated as red ashlars or needles in 53% and 38% yield, respectively. The isomer 76 was obtained as brown ashlars in 35% yield. The metal bonded λ3 phosphorous atom in compound 76 shows a very low field shift in the 31P{1H} NMR spectrum (δ = 549). The structures of 75b and 76, which were determined in solution using NMR spectroscopy, have been confirmed by X-ray crystal structure analyses. The structures show planar heterotitanacyclobutene rings. Alternating bond lengths of the metallacyclic ring-sequences C=C–C=X (X: N, P) are comparable to a conjugated π bond system like free butadiene. Furthermore the atomic distances Ti–X and C–X (X = N, P) and the small angles Ti–X–C prove the existence of an sp2-hybridized X atom with a noncoordinated lone electron pair. This is in contrast to analogous but unstable compounds based on [Cp2Ti=CH2]. Their high reactivity can be regarded as a consequence of the participation of the nitrogen lone pair in bonding to the titanium center, leading to the widening of the Ti–N–C angle and subsequent opening of the azatitanacyclobutene ring, observed for example by Doxsee in the rearrangement via vinylimido intermediates [113]. Because of the higher basicity of the permethylated cyclopentadienyl ligands, there is reduced electrophilicity at the titanium center in 75 and 77. An excess of nitrile and higher temperatures favor the insertion of another nitrile molecule into 75 to give 78.
198
4
Titanocenes
The different regioselectivities in reactions with nitriles and phosphaalkynes arises from the frontier orbital arrangements of the reactants. Whereas end-on coordination dominates in nitrile complexes, phosphaalkyne ligands prefer a side-on geometry (B, C) due to the π character of their HOMO (i.e. the π–n-separation is greater for phosphaalkynes than for nitriles). Starting from the primary end on coordination of the nitrile (A), the cycloaddition takes place forming only one regioisomer. From the experimental data it can be concluded that a primary η2-coordination of the phosphaalkyne showing its alkyne-like behavior is very likely.
Scheme 4-32. End-on (A) and side-on coordination modes (B, C). The behavior of the azatitanacyclobutenes 75 is quite different compared to the reactivity of azatitanacyclobutanes 79, obtained from 38 and carbodiimides [218]. The azatitanacyclobutanes 79 are thermally stable up to 150°C. They are characterized by the ability to undergo cycloreversion reactions instead of the above-mentioned electrocyclic ring-openings.
4.4
Reactions in the Coordination Sphere of Titanocenes
199
The structure of 79b is confirmed by X-ray structure determination. The azatitanacyclobutane ring is shown to be planar. The Ti–N distance is longer than in the azatitanacyclobutene 75b (2.084(2) compared to 2.017(2) [236]).
Fig. 4-3. Molecular structure of 79b . n-C6H14 in the solid state. Selected bond distance [Å] and angles [°]: Ti–N1 2.084(2), Ti–C3 2.129(3), C3–C31 1.329(4), C3– C2 1.505(4), C2–N2 1.287(4), C2–N1 1.374(4), N1–Ti–C3 65.8(1), Ti–C3–C2 91.5(2), C3–C2–N1 105.1(2), C2–N1–Ti 97.4(2), C2–N1–C11 124.6(3), C11–N1– Ti 135.5(2) [218].
200
4
Titanocenes
The only primary regioisomer formed by cycloaddition of 35 with isothiocyanates is 80, which contains the sulfur atom in the α position of the metallacyclic ring. When heated in the presence of pyridine, 80 can be isomerized into the other possible regioisomer 81 [237]. C=N-cycloaddition products are not observed. The quantitative isomerization 80→81 shows that the lower polarity of the C=S unit in the RNCS molecule allows the formation of a second isomer. In this respect, the behavior of titanathietanes is quite different from the behavior of titanaoxetanes, which normally react to give Ti=O fragments (classical behavior). In cycloaddition reactions of 35 with CS2, only the regioisomer with the sulfur atom in the β-ring position is observed [238]. The inverse regiochemistry results from the difference in reactivity of carbonyl and thiocarbonyl groups with strong carbanionic molecules and shows the thiophilic character of the nucleophilic carbene intermediate 35 [239].
Scheme 4-33. Reactivity of the titanatiethan, obtained from [Cp*2Ti=C=CH2] and RNCS. 4.4.3.8 Structure and Reactivity Pattern for small Titanacycles The number of four-membered metallacyclic compounds that can be prepared from 35 allow a comparative discussion of their molecular structures (Table 4-12). All α-methylenetitanacycles are characterized by a planar four-membered ring. The bond angles indicate a stretching of the ring, which is reflected in the relatively short distances c (2.47 (37)–2.80 Å (76)). Particular in 37, this leads to an extreme high-field shift of the corresponding signals in the NMR spectrum [37]. The planarity of the four-membered rings is lost if the exocyclic methylene group is located in the β instead of the α position (18, 82). The automerization equilibria corresponding to the ring folding (22.5° 18; 33.0(9) 82) could not be frozen out in NMR spectra by either 18 [150] or 82 [240]. Neither compound undergoes cycloreversion.
4.4
Reactions in the Coordination Sphere of Titanocenes
201
The Ti–C distances a and b in 37 are different (2.068(6) and 2.137(7) Å, respectively), in agreement with the different hybridizations of the carbon atoms. The equal C–C bond lengths (f, e) (1.520(10) and 1.521(10) Å, respectively) rule out a partially ring-opened structure (→ 35). In the metallacyclobutene 54b (R = Me), inner C–C bond (e), which can be conjugated, shortens significantly. Bulky substituents facilitate cycloreversions, which is reflected in longer distances e and b in 54c (Rα = SiMe3, Rβ = Ph). The exocyclic α-methylene group exhibit characteristic bond lengths of 1.318(8) Å for 63d and 1.377(4) Å for 54b. The increase of the angle γ in 37 to 152.4° is noteworthy. The distances b to the heteroatoms (O, S. P) are in the expected range, while for 75b a slightly shortened Ti–N bond is found (b = 2.017(2) Å), which is in accordance with the observed reactivity in electrocyclic ring-opening, as well as the lengthening of the M–C bond (a = 2.134(2) Å). The azatitanacyclobutan 79b shows a longer Ti–N bond (b = 2.129(3)) [218].
Table 4-12. Selected structural parameters of four-membered metallacyclic compounds
202 4 Titanocenes
4.4
Reactions in the Coordination Sphere of Titanocenes
203
Remarkably, depending on the nature of the heteroatom, different reactions of the titanacycles are observed. In the case of the titanacyclobutane 37, cycloreversion reactions dominate, forming the vinylidene intermediate 35. In the case of the oxetanes 63, metathesis reactions are observed in the mass spectrometer, whereas the bimetallic oxetanes 67 fragment to give the starting materials. Fast ring opening reactions occur in the case of the aza-titanacyclobutenes 75, leading to products of insertion into the Ti–C bond, which exhibit the exo methylene group [236]. This behavior is explained by the orientation of the lone pair at the heteroatom towards the acceptor orbitals in the equatorial plane of the metallocene Cp*2Ti fragment. The best orbital overlap can be expected in the case of the azatitanacyclobutenes 75, because the lone pair at the nitrogen atom and the lateral acceptor orbital of the titanium center are orientated in the same plane, leading to fast ring opening [191].
Scheme 4-34. Reaction behavior of different four membered titanacycles. On the other hand, if the hybridisation of the nitrogen atom is different, as in the azatitanacyclobutanes 79, no electrocyclic ring opening reaction is observed. In the molecule 79 the nitrogen lone pair is orientated perpendicular to the acceptor orbital on titanium [218]. A similar orientation of donor and acceptor orbitals is also ex-
204
4
Titanocenes
pected in the case of titanacyclobutenes, leading to substituent controlled reactions. These reactivity patterns can also be observed in the insertion behavior of the titanacycles (Scheme 4-35). In the case of the titanacyclobutane 37 [241] and cyclobutenes 54 [208], insertion of small molecules like isonitriles are observed only into the Ti–C bond opposite the exo methylene group. For the azatitanacyclobutene 75, a spontaneous ring enlargement by insertion of a further nitrile molecule is observed 75→78, due to the activation of the Ti–C sp2-bond by the lone pair on the nitrogen center. In a similar manner the oxetanes 63 and 67 show insertion reactions into this Ti–C bond, forming mono or double insertion products [212, 242]. On the other hand, azatitanacyclobutanes 79 are inert to ring enlargement reactions with isonitriles even at higher temperatures [218].
Scheme 4-35. Ring enlargement reactions of small titanacycles. In some cases, cycloaddition reactions involving a titana–allene intermediate are useful in organic synthesis. Via complexes of type 83 (Cp instead of Cp*) the preparation of allenylketenimines 84 becomes possible in high yields [243].
4.4
Reactions in the Coordination Sphere of Titanocenes
205
4.4.3.9 Dimetallic Compounds Derived from Ti=C Species Let’s have a first look at the Tebbe reagent, the most important source of a titanium methylene compound. Ab initio calculations (STO-3G) on [H2Ti=CH2]/[ClAlH2] show that for the Tebbe reagent, strong bonding of the Lewis acid [ClAlMe2] must be assumed, in contrast to the much-discussed weak association [244]. The presence of a strong bond is also shown in the course of the reactions of the Tebbe reagent as compared to the Petasis or Grubbs variants [99]. Kinetic studies of the formation of the Tebbe reagent 87 by the reaction of [Me3Al] with [Cp2TiCl2] (85) and its derivatives [97, 245] show an isotope effect kH/D of 2.9 and a large negative entropy value, which is in agreement with a cyclic transition state 86. The polarization of the Ti–Cl bond by the aluminum center increases the acidity of the α-H atoms of the Ti-CH3 group. This rules out the primary formation of a [Cp2Ti=CH2] species with subsequent complexation by [ClAlMe2] during the course of Tebbe reagent formation.
Scheme 4-36. Formation of the Tebbe reagent from [Cp2TiCl2] and [AlMe3]. However, other bimetallic compounds that are particularly interesting from the coordination chemistry point of view are accessible from [LnTi=CR2] generating complexes and suitable metal fragments. Depending on the nature of the interacting metal fragments, a variety of structures can be obtained, which exhibit symmetric or asymmetric bridges 88–90.
Scheme 4-37. Schematic drawings of methylene bridged dinuclear complexes.
206
4
Titanocenes
Distinct carbenoid character is revealed by the reduced Ti–C distances and a notably low field shift of the µ-CH2 signal [97, 116, 126, 246–250]. Table 4-13 summarizes selected examples. A participation of the mesomeric structure 90 is discussed in particular for the Ti–Pt compound [Cp2Ti(µ-CH2)(µ-Cl)Pt(PPhMe2)(Me)] and [Cp2Ti(µ-CH2)(µ-CH3)Pt(PPhMe2)(Me)] [247]. For example, in contrast to dititanacyclobutane [Cp2Ti(µ-CH2)]2 [126, 250], the silatitanacyclobutane [Cp2Ti(µCH2)(µ-CH2)SiMe2] [124, 125] shows no carbenoid activity. Table 4-13. Selected structural and NMR data of dinuclear µ-methylene and related complexes [Cp2Ti(µ-CH2)(µ-X)MLn]
4.4
Reactions in the Coordination Sphere of Titanocenes
207
The µ-methylene titanium–platinum and –rhodium complexes do not only act as activated [Ti=CH2] equivalents, but can undergo a variety of different subsequent reactions. For example, by varying the bridging ligand novel structural types can be obtained such as those with µ-phenyl or µ-methyl groups and thus providing information about the nature of these ligands. Compound 92, the first example of a dinuclear phenyl species that is asymmetrically bridged by the ipso-C atom of the phenyl group [246], can be synthesized by reaction of 91 with phenyllithium derivatives.
A unique feature of the Ti–Rh complex 93 is the bridging methyl group which forms a three-center 2-electron agostic bond with the titanium center [249]. Crystallographic and spectroscopic studies of this molecule have been carried out to show the static and dynamic behavior of this agostic interaction. The methyl group form an agostic interaction with the titanium atom. For the bridging methyl group, the Ti–H bond distance is 2.02(5) Å and the Ti–C bond distance is 2.294(6) Å. At room temperature the µ-CH3 group shows a broad resonance at δ = –3.13. When cooled to –90°C, this resonance is replaced by two resonances at δ = 1.28 (d) and δ = –12.15 (t, J = 12.8 Hz). The observed coalescence temperature is -40°C. In the 13C NMR at room temperature the µ-CH resonance is at δ = 49.7 (dq, J 3 CH = 114, JCRh = 29 Hz). At -92°C, the µ-CH3 resonance was observed at δ = 51.1. The coupling constant for the bridging proton is found to be 87.7 Hz; for the terminal protons 126.7 Hz. As expected, abnormally low coupling constants for the agostic proton are found [249].
The syntheses of vinylidene bridged complexes of the types 94–96 are of interest e.g. in the understanding of the ring expansion reaction 67→68 and the related possibilities of the formation of µ-vinylidene compounds as well as in the study of this bridging ligand. The majority of known µ-vinylidene complexes belong to the structural type 94 [251]. Some homo- and heterodinuclear µ-vinylidene complexes indicate the possibility of side-on bridges (96). A semi-bridge (95), which is char-
208
4
Titanocenes
acteristic for CO- or CS-bridged heterodinuclear complexes with early–late metal combinations [252] has only once been reported for comparable vinylidene complexes [253].
Scheme 4-38. Dimetallaethylene (94) as well as semi- and side-on-bridged structures (95 and 96, respectively) of dinuclear µ-vinylidene complexes. When [Cp*2Ti(CH=CH2)CH3] (38) reacts with complexes of group 11 metals (97a–e), the µ-vinylidene complexes 98 can be isolated (40–50%) [254, 255]. In the complexes 98a–98e, the signal of the α-vinylidene carbon atom is shifted to very low field (δ = 300–330; Table 4-14).
This is in accordance with a semi-bridging structure 95. The possibility of a sideon bridge (96) can be ruled out due to the orientation of the π-bonding planes in 98. In the case of side-on bridged vinylidene complexes (96) equal protons are expected in the 1H NMR spectra.
4.4
Reactions in the Coordination Sphere of Titanocenes
209
Table 4-14. Selected NMR data of heterodinuclear µ-vinylidene complexes 98
4.4.4 Reactions Involving the Cp Ligand Generally, the chemistry of titanocene fragments is characterized by the great stability of the η5-Cp binding mode. In most organometallic reactions of transition metal complexes, the η5-Cp ligand plays the role of spectator, staying tightly bound η5 to the metal center throughout the course of the reaction. However, cyclopentadienyl ligand-transfer reactions have been known since 1956 when Wilkinson, Cotton, and Birmingham reported that the reaction of chromocene and ferrous chloride gave ferrocene [256]. The most detailed work with early transition metal complexes is that of Brubaker and co-workers [257–259]. Photolysis of [(η5C5H5)2MCl2] and [(η5-C5D5)2MCl2] mixtures ( M = Ti, V, Hf) results in intermolecular exchange of the cyclopentadienyl ligands. The fact that [(η5-C5H5)TiCl3] is not found in the absence of a sensitizer led Brubaker to suggest that in some cases a photochemically induced ring slippage may open up the coordination sites necessary for a bimolecular C5H5-bridged mechanism.
Other new investigations show that the Cp ligand is not as inert as many people believe, as cyclopentadienyl activation has been observed by several groups. Cpactivation seems to become possible if the titanium atom is part of a heteroatom containing metallacycle. For example, the formation of complex 103 is observed during the formation of 102. Whereas 102 can be isolated in a pure form [260] 103 rearranges by Cp-activation to form 104 [261].
210
4
Titanocenes
The formation of Cp-activation products from electron rich titanacyclopentadienes [262] like 105 is also observed [263], and further examples involving titanacyclopentadienes are known [262]. These reactions under formation of 106 are promoted by nonpolar solvents.
The formation of fulvenes as the product of reactions of aliphatic and aromatic ketones with [Cp2Ti(PMe3)2] is discussed in terms of Cp-activation instead of a pinacol coupling reaction [264]. Also, examples of ring enlargements of the Cp* ligand are known. Reaction of [Cp*2TiR] (R = CH3, CH2CMe3) with CO gives labile acyl compounds [Cp*2TiC(O)R]. They react with [(CpMo(CO)3)2] to yield complexes [Cp*2Ti(η2-COR)(µ-OC)MoCp(CO)2], which undergo acyl C–O bond breakage to form benzene derivatives at room temperature. The benzene derivatives are formed through ring expansion of the Cp* ligand by incorporation of the R–C≡ fragment of the acyl group. The acyl oxygen is incorporated into a dimeric titanium oxo complex [265].
4.5 Analytical Data 4.5.1 Structural Data The number of Cp2TiL2 compounds known in each oxidation state decreases in the order: four > three > two [266]. Although the comparison is obviously not strictly relevant, the Ti–Cp bond distance appear to show a slight increase with increasing formal oxidation state of the titanium atom. For example, the mean Ti–Cp distances for Ti(II) (2.041 Å), are shorter than those for Ti(III) (2.048 Å), and these are again somewhat smaller than those for Ti(IV), (2.058 Å) [266]. Comparison of the Ti–Cp distances for the mono- and bis-cyclopentadienyl compounds reveals that
4.5 Analytical Data
211
the mean values are similar for both. In general, the Ti–L distances increases with increasing van der Waals radius of the ligated atom: 2.003 Å (range 1.860–2.155 Å) for O (1.52 Å) < 2.007 Å (range 1.922–2.100 Å) for N (1.55 Å) < 2.359 (range 2.310–2.405 Å) for Cl (1.75 Å) < 2.429 Å (range 2.395–2.455 Å) for S (1.80 Å) < 2.497 Å (2.340–2.585 Å) for P (1.80 Å) [266]. Representative examples of titanocene complexes are summarized in Table 4-15. It becomes obvious that introduction of methyl groups on the Cp ligand leads to increasing Z–Ti–Z angle. Additionally, the Z–Ti–Z angle is smaller for Ti(IV) than for Ti(II) complexes. An extremely large Z–Ti–Z angle of 150.84° is found for [(PhC5Me4)2TiH] [86]. Table 4-15. Selected structural data of representative titanocene complexes L2TiX2
212
4
Titanocenes
4.5.2 NMR Data While the rotation barrier about the M–Cp bond is usually small, restricted rotation in complexes containing substituted cyclopentadienyl ligands has been noted [18, 276]. As part of the characterization of substituted biscyclopentadienyl Ti(IV) compounds, 13C NMR and 1H NMR data indicate that the effects of both metal shielding and ring substitution can be taken into account by a simple additive relationship so that it is possible to accurately predict chemical shifts for these systems [277].
4.5.2.1 1H NMR Data 1H
NMR shifts of some titanocene complexes are given in Table 4-16 [277]. The assignment was based on the following considerations. The alkyl protons and cyclopentadienyl protons resonate in quite different regions. The 1H NMR spectrum of the monoalkylated cyclopentadienyl ring shows two triplets. The 1H NMR spectrum of [1,3-(Me2C5H3)2TiCl2] consists of an AB2 pattern while that of 1,3(MeEtC5H3)2TiCl2 is an ABX pattern. The most surprising result is the apparent lack of cyclopentadienyl proton splitting in [1,2-(Me2C5H3)2TiCl2]. A similar observation has been reported for titanium indenyl complexes. For the series [(Ind)2TiX2], an AB2 pattern is observed where X = CH3, but a singlet is observed for X = C6H5 [278]. In contrast, an AB2 pattern was reported for [(Ind)2Zr(C6H5)2]. Table 4-16. 1H NMR Data of selected Titanium (IV) compounds (measured in CDCl3)
4.5 Analytical Data
4.5.2.2
13C
213
NMR Data
Selected 13C NMR data are given in Table 4-17. The alkyl and the cyclopentadienyl carbon atoms again resonate in two different regions of the spectrum. The chemical shifts of the chain carbon atoms were assigned by comparison with chemical shifts in free alkanes. The ring carbon atoms were identified by comparison to specifically deuterated species [277]. Table 4-17.
13C
NMR data of selected titanium(IV) Compounds (CDCl3)
The C–H coupling constants of methyl groups depend on the amount of σ character at the carbon in the C–H bonds. To take a simple approximation the amount of σ character depends on the effective electronegativity of the group to which the methyl group is bonded; the higher the electronegativity of this group, the lower the character of its bond with the methyl carbon. This leaves more s character in the C–H bonds, and hence, increasing the electronegativity of the group attached to the carbons should increase the C–H coupling constant observed for a methyl group. Conversely, decreasing the electronegativity of the group attached to carbon should lower the C–H coupling constant. Increasing the electron density at the carbon to which the methyl group is bonded effectively lowers the electronegativity and should lower the C–H coupling constant. Indeed, increasing the electron density at the titanium center by adding methyl groups to the Cp ligand causes a lower-
214
4
Titanocenes
ing of the JCH of the methyl group bonded to the Ti. Conversely, substitution of electron-withdrawing chlorides or a CF3 group on the ring causes an increase in JCH [129]. Table 4-18. C–H Coupling constants of methyl groups attached to a titanocene center [129]
As well as the solution NMR techniques also solid state NMR measurements were used. Variable-temperature 13C CP/MAS NMR spectroscopy of (η5-Cp)2Ti(η1-Cp)2 has shown that sigmatropic rearrangement of the σ-cyclopentadienyl rings occurs in the molecule in the crystalline state. Magnetization-transfer experiments at 165 K are consistent with [1,2] shifts being the main pathway for the rearrangement, but there is evidence for the possibility of [1,3] shifts as a minor pathway. These measurements, in conjunction with spectral line shape analysis in the exchange broadening regime, have been used to estimate Arrhenius activation parameters for [1,2] shifts of Ea = 33.2 ± 1.0 kJ mol–1 and A = 2.9 × 1010 s–1. The activation barrier is similar to that suggested by solution NMR studies, implying that control of the rearrangement is principally by electronic factors [279]. Also the structures and dynamics of a series of crystalline ansa-metallocenes, [(C5H4)2-C2Me4]TiX2 ( X = F, Cl, Br, I) have been elucidated by joint application of solid state NMR and single crystal X-ray diffraction techniques [280].
4.5.2.3 Ti NMR Data Both 49Ti and 47Ti resonances appear in the same spectrum, with the 49Ti resonance 268.1 ppm downfield of the 47Ti resonance. Due to its higher quadrupole moment the 47Ti resonance is always broader than the 49Ti resonance. Increasing methyl substitution of the rings causes an increasing downfield shift of both the 47Ti and 49Ti resonances [129]. The 49Ti NMR chemical shifts of a series of ti-
4.5 Analytical Data
215
tanocene derivatives have been shown to have an inverse relationship to the Ti(2p3/ 2) binding energies of these same compounds as measured by X-ray photoelectron spectroscopy (ESCA). Replacement of one cyclopentadienyl (Cp) group in the titanocene difluorides, dichlorides, and dibromides by a pentamethylcyclopentadienyl (Cp*) moiety resulted in an average downfield shift in the 49Ti NMR absorption of 148 ppm. Substitution of both Cp groups by Cp* groups resulted in an average downfield shift of 312 ppm [28]. Table 4-19. ergies [28]
49Ti
NMR chemical shifts (vs TiCl4, δ 0.00) and Ti(2p3/2) binding en-
Titanium NMR spectroscopic data for a series of trichlorotitanium complexes bearing a cyclopentadienyl ligand substituted with tert-butyl and/or trimethylsilyl groups have also been collected. Chemical shift values imply that trimethylsilyl groups are weakly electron-donating, whereas the line-widths reflect the symmetry of the substitution pattern of the five-membered rings [281].
4.5.3 UV vis Spectra The UV vis spectra of ring substituted titanocene dichlorides exhibit a weak absorbance between 500 and 560 nm, which in titanocene dichloride has been assigned to a symmetry-forbidden A1→A2 transition. This band moves to lower energy as electron donating substituents are added to the rings and to higher energy as electron-withdrawing substituents are added. The absorption spectra of the com-
216
4
Titanocenes
pounds with one ring substituent are qualitatively similar to the spectrum of Cp2TiCl2; in addition to the band between 500 and 560 nm, they exhibit a broad region of absorption between 260 and 410 nm. The compounds with more highly substituted rings, exhibit not only this broad region of absorption, but also a band between 470 and 490 nm [129]. Table 4-20. Absorption maxima of lowest energy transitions in the UV–vis spectra of ring-substituted titanocene dichlorides (3.3 × 10-3 M solutions in CHCl3)
Using self-consistent-field-Xα-scattered-wave (SCF-Xα-SW) molecular orbital methods to investigate the electronic structures of Cp2TiX2 (X = F, Cl, Br) the lowest energy electronic transitions are predicted to be due to Cp→Ti charge-transfer. The halide→Ti charge-transfer transitions occur at higher energy. For X = I, Cp→Ti and I→Ti charge-transfer transitions are predicted to be very close in energy. That the fluoride, chloride, and bromide complexes have lowest energy exited states different from those of the iodide and methyl complexes is reflected in the photochemistry of the two type of complexes. In the former complexes, Cp–Ti bond cleavage results from low-energy irradiation; in the latter complexes, Ti–X bond cleavage occurs when the complexes are irradiated [282].
4.6 Selected Applications of Titanocene Complexes
217
4.6 Selected Applications of Titanocene Complexes Titanocene complexes have a large number of applications. These range from stoichiometric organic reactions to catalytic reactions in the field of polymerization and fine chemical synthesis [283, 284], which are explained in other chapters of this book. Selected examples, reduction of carbonyl compounds [285, 286], of epoxides [287], of olefins [288], and epoxidations [289–291], isomerizations [292], Diels–Alder reactions [293, 294], C–F activation reactions [295–297] as well as carbomagnezations [298–301], are summarized in Scheme 4-39.
Scheme 4-39. Selected applications of titanocene complexes. The use of allyl complexes of titanocene e.g. in diastereoselective C–C coupling reactions is another field of applications [302]. Some applications of selected titanocene complexes will be discussed more in detail in the next part of this chapter.
4.6.1 C–C Coupling Reactions 4.6.1.1 Dimerization of Ethylene The selective dimerization of Ethylene, first described by Ziegler, uses a mixture of titanium tetrabutyl ester and trialkylaluminum [303].
218
4
Titanocenes
Interestingly, benzene and toluene solutions containing [Cp*2Ti(η2-C2H4)] (42) catalyze the specific conversion of ethylene to 1,3-butadiene and ethane under very mild conditions (25°C, ≤ 4 atm), although turnover numbers are exceedingly small [61]. As shown in Scheme 4-40, ethylene reacts reversibly with 42 in aromatic solvents to yield an equilibrium mixture containing the titanacyclopentan derivative 107. In the combined dimerization / hydrogen-transfer reaction, butadiene and ethane are formed. The formation of butenes is not observed.
Scheme 4-40. Dimerization of ethylene to butadiene-1,3 and ethane.
4.6.1.2 Head to Tail Coupling Reactions of Acetylenes The acetylene complex [Cp*2Ti(η2-Me3SiC≡CSiMe3)] catalyzed the linear head-totail dimerization of 1-alkynes. A selectivity of better than 98% and a turnover number in the range 1200–1500 mmol of 1-alkyne per mmol of titanium was obtained after 14 days at 30°C for 1-pentyne, 1-hexyne, cyclohexylethyne, phenylethyne and trimethylsilylethyne [304].
4.6 Selected Applications of Titanocene Complexes
219
4.6.1.3 Cyclization Reactions The intramolecular cycloaddition of the titanocene vinylidene complex, formed by dechloroalumination of 111 [305], to an alkene (or alkyne) affords bicyclic titanacyclobutanes and 112 (or butenes) [306]. This cyclization has proven to be possible in all such complexes. The remaining carbon–metal bond in the metallacycles 112 offers great potential for further elaboration to give organic products. This is illustrated by complex 112, which reacts with N-bromosuccinimide (NBS), to give the dibromide 115. The insertion reaction of isonitrile affords the imino complex 113 which gives the aldehyde 114 on acidic work-up; alternative reaction with carbon monoxide followed by acidic work-up gave the ketone 117, presumably via the ene–diolate complex 116 [306].
Scheme 4-41. Selected organic syntheses with a titanocene vinylidene intermediate.
220
4
Titanocenes
4.6.2 Stoichiometric and Catalytic Reactions of Carbenoid Metal Complexes of Electron-Deficient Transition Metals The most important applications of carbenoid complexes of electron-deficient transition metals lie in the areas of organic synthesis and catalytic reactions, such as carbonyl olefination (c) [98] the generation of cyclopropane derivatives by the oxidation of titanacyclobutanes (a) [307, 308] the preparation of heterocyclic fourmembered ring compounds (b) [309, 310], and ring-opening polymerization [99, 311–316] to obtain special polymers, such as ‘living’ [132, 317] (d) or ‘conducting’ [318] (e) polymers (Scheme 4-42).
Scheme 4-42. Applications of titanacyclobutane derivatives.
4.6.2.1 Carbonyl Olefinations with Tebbe, Grubbs, and Petasis Reagents Many carbonyl compounds, such as esters, ketones, and amides, can be methylated with various titanium-based reagents. The most important of these are the Tebbe reagent, titanacyclobutanes (Grubbs reagent) and titanocene alkyl complexes (Petasis reagent). The application of the Tebbe reagent in particular has found broad interest in organic chemistry, owing to its early discovery [98, 99]. The use of [Cp2TiMe2] is particularly advantageous, as this compound is not pyrophoric, is relatively stable toward air and water, and no Lewis acidic aluminum by-products have to be separated off later, which enables its use in olefination of silyl esters,
4.6 Selected Applications of Titanocene Complexes
221
anhydrides, carbonates, and acylsilanes [101]. A number of different, sometimes contradicting, mechanisms [99] have been proposed for the reactions that occur, which will be discussed using the carbonyl olefination of esters as an example (Scheme 4-43). In the absence of an additional base, the reaction between the Tebbe reagent and the ester is first-order with respect to the titanium reagent employed.
Scheme 4-43. Mechanisms of the carbonyl olefinations 1. with Tebbe reagent, 2. titanacyclobutanes (Grubbs reagent), 3. with titanocene alkyls (Petasis reagent).
222
4
Titanocenes
The strongly negative activation entropy observed is in agreement with a six-membered intermediate. In the presence of bases (pyridine), the reaction proceeds more rapidly and shows zero-order kinetics with respect to the ester and first-order kinetics with respect to the titanium reagent. The use of titanacyclobutanes in carbonyl olefination is characterized by the primary formation of a metallaoxetane, which spontaneously decomposes to give the desired olefin and a Ti=O compound [99]. Based on the observed H/D exchange, Petasis and Bzowej favored a methyl addition mechanism [115]; however, they reported deuterium scrambling in the reactions of esters with cyclopropyltitanocenes [110]. Detailed kinetic studies show that the olefination of esters with [Cp2TiMe2] proceeds via the titanium carbene [Cp2Ti=CH2] [118]. This is proved by: 1) the absence of H/D exchange or 13C scrambling when labeled esters or titanium compounds are employed; 2) the finding of zero-order kinetics with respect to the ester, and first-order kinetics with respect to the titanium compound; 3) the reaction of ethyl acetate or dodecyl acetate with [Cp2Ti(CD3)2] yielding a kH/D effect of 9– 10; 4) esters with different electronic and steric environments exhibiting comparable reactivities. The stoichiometry required for the Petasis reagent (substrate : Ti = 1:2) can be explained by the necessity of reactions which trap the Ti=O fragment formed. The products were obtained in better yields with the Tebbe reagent than with the Wittig reagent. This is especially significant when sterically hindered ketones are to be olefinated. For the generation of special vinylsilanes, there are effective reagents in the form of [Cp2Ti(CH2SiMe3)2] and [CpTi(CH2SiMe3)3] [111, 102]. The introduction of a benzylidene group can be successfully performed by using [Cp2Ti(CH2C6H5)2] [112]. Substituents in the µ-position of the phenyl ring (Cl or F) increase the yield. For example, in this way aryl substituted enol ethers 118, which also find use as bioactive prostaglandin analogs, can be formed quantitatively [112].
Disubstituted titanacyclobutenes 119 (Rα = Rβ = CH3, CH2CH3) react with two equivalents of nitrile under mild conditions to yield diazacyclooctatetraenes, a result of nitrile insertion into the titanium–alkyl bond as well as into the alkenyl bond. Subsequent hydrochlorination of these products yields tetrasubstituted pyridines 124 [114]. Steric bulk leads to the exclusive formation of monoinsertion products. Insertions into the titanium–vinyl bond are favored over those into the metal–alkyl bond, when titanacyclobutenes with small substituents (methyl) are used. Diphenyltitanacyclobutenes, however, react with nitriles or ketones to yield solely prod-
4.6 Selected Applications of Titanocene Complexes
223
ucts of monoinsertion into the titanium–alkyl bond, which can be worked up to yield unsaturated products 120 and 121, respectively [114, 319]. Reactions of 119 with ketones can be used to generate dienes 122 [320], or after the reaction with RPCl2, the phosphacyclobutenes 123 [309, 321].
Scheme 4-44. Organic Synthesis employing titanacyclobutenes.
4.6.2.2 Ene Reactions Reactions of Ti=C intermediates with carbonyl compounds can lead to carbonyl olefinations, as discussed. Enolizable ketones, on the other hand, can react in a
224
4
Titanocenes
1,5-sigmatropic intramolecular H shift to yield titanium enolates [322–324], a reaction that is particularly noteworthy with respect to stereoselective C–C linkages. The majority of studies carried out to date have used NMR spectroscopy to characterize such substrates, and only recently have studies of the solid-state structures have been performed [207, 210, 325, 326]. When [Cp*2TiMe2] (29), [Cp*2Ti(Me)CH=CH2] (38), or [Cp*2TiCH2CH2C=CH2] (37) reacts with enolizable ketones, the titanocene enolates 125 [323] and 126 [210] are exclusively obtained.
The formation of products exhibiting ∆1- and E-configurations is always preferred for both 125 and 126. The observed high regio- and stereoselectivity results from the geometric conditions at the [Cp*2Ti=C] intermediate 127, which apparently only permit selected deprotonations of the acidic ketones. Theoretical studies show that for [Cp2Ti=CH2] fragments, the end-on coordination 128 is more favorable than side-on mode 129 by 0.96 eV [227]. This in turn apparently leads to the observed preference for formation of the enolate over the oxetane.
The structural data for enolates of type 126 show a distinct alkoxide character (Ti– O: 1.859(2) Å; Ti–O–C 165.9°) [210]. This reduces the nucleophilic properties of the methylene group and the oxygen in the enolate group, so that typical reactions with electrophiles [327, 328], such as methyl iodide or benzaldehyde, do not occur.
4.6.2.3 Catalytic Reactions Catalytic applications of precursors for [Cp2Ti=CR2] intermediates in the context of ring-opening metathesis polymerization (ROMP) have already been summarized and discussed in various articles [314–316, 329, 330]. New studies show the effec-
4.6 Selected Applications of Titanocene Complexes
225
tive use of [Cp2TiMe2], [CpTiMe3], [CpTiClMe2], and [Cp2Ti(CH2SiMe3)2] as carbenoid reagents [107]. The important advantages in these cases are similar to those already discussed for stoichiometric reactions, namely the easier preparation and handling of the corresponding titanium alkyl compounds as compared to Tebbe or Grubbs reagents. For example, the ring-opening polymerization of norbornene can be carried out at high temperatures in the presence of these reagents (Scheme 4-45).
Scheme 4-45. The mechanism of the ring-opening polymerization of norbornene. When THF is used as solvent, the reaction does not take place, even though carbonyl olefination proceeds more rapidly in THF than in toluene or hexane [115]. This effect is attributed to the nucleophilicity of the carbene C atom. Other studies show a slower rate of chain growth with THF or pyridine as solvents. In general, THF has only a small effect when [CpTiMe3] and [CpTiMe2Cl] are employed. In the case of [CpTi(CH2SiMe3)2], though, THF nearly completely inhibits the ring-opening polymerization. The different ROMP activities correspond with the electronic structure of the [Ti=CR2] derivatives according to studies of Cundari and Gordon [331–333]. These indicate, for example, that [Cp2Ti=C(H)SiMe3] is a particularly effective metathesis initiator. Furthermore, when electron-withdrawing substituents are present on the carbene C atom, increased activity can be expected, which is demonstrated by results obtained with titanocene benzyl complexes [112]. While unsubstituted [Cp2Ti(CH2Ph)2] is suitable for carbonyl olefination, it is unable to initiate the ROMP reactions. Titanocene vinylidene derivatives will be discussed as intermediates in catalytic reactions. Alt et al., for example, have suggested a vinylidene intermediate 131 in the formation of trans-polyacetylene 134 from the acetylene complex 130 and excess acetylene [334].
226
4
Titanocenes
This proposed mechanism is supported by the observation that it is possible, starting from methylenetitanancyclobutene [Cp*2TiCH=CHC=CH2] (54a), to achieve the polymerization of acetylene to form pure trans-polyacetylene [208]. The transformation of a metal-coordinated acetylene to a vinylidene (130→131) has not so far been experimentally verified in an electron-deficient transition metal complex.
4.7 Cp-Equivalents In the following chapter some examples of ligands with similar properties to Cp as well as Cp* ligands will be given. Unlike the titanocene fragment, which cannot exist as a stable molecule without further stabilizing ligands, the so called ‘open’ metallocenes can be isolated [335]. For example, titanium dichloride, generated in situ by the reduction of TiCl4 with magnesium, reacts with potassium 2,4-dimethylpentadienide in THF at –78°C to yield the deep–green, diamagnetic 14-electron complex [Ti(η5-C5H5Me2-2,4)2] (135a). The compound is pyrophoric and can be purified by sublimation [336]. The analogous reaction with K[C5H5But2-2,4] gives the more stable complex [Ti(η5C5H5But2-2,4)2] (135b). The NMR spectrum suggests that the two pentadienyl ligands are orientated at an angle of ca 90° with respect to each other [337]. The 14electron bis(dienyl) complexes readily form adducts 136 with carbon monoxide or phosphines (16-electron complexes) [338].
4.7 Cp-Equivalents
227
In mixed CpTi(pentadienyl) complexes [CpTi(η5-C5H5Me2-2,4)PEt3], molecular orbital calculations show that the titanium–pentadienyl bond is stronger than the titanium cyclopentadienyl bond. This is attributed to the fact that the pentadienyl ligand serves as a much better electron acceptor than cyclopentadienyl [339]. To a
228
4
Titanocenes
large extent this may be attributed to significant d bonding interactions between titanium dxy and pentadienyl π4* orbitals [340]. The alkane elimination reaction of C2B9H13 with [Cp*TiMe3] provides a simple route to the bent-metallocene complex [Cp*Ti(η5-C2B9H11)TiMe] (137) in basefree form [187]. Compound 137 is also related to [Cp*2TiMe]+ by replacement of a C5Me5 ligand with a (η5-C2B9H11)2 ligand. EHMO calculations show that d0 [(C5R5)(η5-C2B9H11)M(H)] species contain two empty low-lying metal-centered orbitals similar to those of the d0 [(C5R5)2M(H)]n+ species, and structural comparisons indicate that the steric properties of the Cp* and [(η5-C2B9H11)]2– ligands are similar [341]. Thus, complexes of the type 137 are best described as 14-electron Ti(IV) species, analogous to [Cp*2Ti(Me)]+ and [Cp*2Sc(Me)]. Consistent with this, 137 species display structural features and reactivity patterns characteristic of electrophilic metal complexes, including agostic interactions, ligand coordination (→139), insertion (→140) and intramolecular C–H activation (→138). In addition to the examples mentioned above, bulky alkoxide ligands support organometallic chemistry at early d-block metal centers. In the case of group 4 metals, both complementary and novel stoichiometric reactivity have been developed at bis(aryloxide) metal complexes. The fragment [(ArO)2M] 141 can be considered to be formally isoelectronic with [Cp2M] if the aryl oxide oxygen atom donates a full six electrons to the metal center [342–345]. The spatial resemblance of cyclopentadienyl ligand and tri-tert-butylmethoxide (tritox) 142 [346] and the Si-analog complexes (silox) 143 have also been discussed [347, 348].
Scheme 4-46. Bulky alkoxides as ‘Cp-equivalents’.
In addition to the bulky alkoxides of titanium group metals, bulky amides are being investigated also in the hope that they may provide the next generation of rea-
4.7 Cp-Equivalents
229
gents and catalysts [349]. One class of amido groups is the sterically-demanding N-tert-hydrocarbylanilide ligand, which stabilizes low coordinate transition metal complexes [350]. Chelating amides like η2-coordinated aminopyridinato ligands with a ‘maximal steric angle′ of 144° [351] are similar to tripod ligands [352]. The chelating bis(borylamido) ligand [MesBNCH2CH2NBMes2]2–, or (Ben)2– would be a structurally and electronically unusual bidentate diamido ligand, since the two mesityl groups should sterically protect two or three coordination sites in a plane roughly perpendicular to the NCCN ligand backbone as a consequence of N–B πbonding [353]. Furthermore, the strongly π-accepting boryl groups should attenuate the ability of the nitrogen atoms to donate electron density to the metal center and therefore should yield complexes in which the metal is more electrophilic than in traditional alkyl and arylamide complexes [353]. First examples of (Ben)Ti complexes 144 have been synthesized.
Scheme 4-47. Schematic drawing of titanium complexes containing the [Mes2BNCH2CH2NBMes2]2– and the [o-C6H4(NSiiPr3)2]2– ligand. From the dilithio reagent o-C6H4(NLiSiiPr3)2 new titanium (145) and zirconium complexes have been prepared. X-ray crystallographic and variable-temperature NMR studies show that these complexes possess a η4-C6H4(NSiiPr3)2 ligand (bonded to the metal via two carbon atoms in the phenylene group), which undergoes a η4–η2 fluxional process in solution [354]. A wide range of further examples of nitrogen-based donors can be derived from the N,N’-bis(alkyl)benzamidinato ligand [η-PhC(NR)2] [355–357]. The benzamidinato ligands [η-PhC(NR)2] have been shown to bond to a transition metal through coordination of the central NCN system, which is formally a three-electron donating group. It was noted that the [ηPhC(NSiMe3)2] ligand has some steric similarities to the η-cyclopentadienyl ligand [358].
230
4
Titanocenes
Acknowledgments I wish to express my particular thanks to many very active co-workers, Dr. Jürgen Oster, Dr. Javier Sang, Dr. Isabelle Strauß, Dipl.-Chem. Jürgen Heinrichs, and Dipl.-Chem. Martin Wagner, for their work, which has contributed significantly to this chapter. I am indebted to all my colleagues for constructive discussions and suggestions, in particular to Prof. Dr. Uwe Rosenthal (Rostock) and Dr. Uwe Böhme (Freiberg). I wish to express my gratitude to Mr. Edward Pritchard (York), for checking the English version of the manuscript. I especially wish to thank the Institute für Anorganische Chemie der RWTH Aachen, where I worked as a guest Professor and have carried out research for the last five years. I want to gratefully acknowledge that my research was supported by the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie. The Bayer AG Leverkusen and the Degussa AG are also acknowledged for their generous financial support.
References [1] [2]
[3]
[4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20]
A. K. Fischer, G. Wilkinson, J. Inorg. Nucl. Chem. 1956, 2, 149-152. M. Bochmann, Comprehensive Organometallic Chemistry II; Titanium Complexes in Oxidation States +2 and +3 in Comprehensive Organometallic Chemistry II (Ed.: E. W. Abel, F. G. A. Stone, G. Wilkinson), Pergamon Elsevier Science Ltd, Oxford, 1995, Vol. 4, p. 221. M. Bochmann, Comprehensive Organometallic Chemistry II; Titanium Complexes in Oxidation States +4 in Comprehensive Organometallic Chemistry II (Ed.: E. W. Abel, F. G. A. Stone, G. Wilkinson), Pergamon Elsevier Science Ltd, Oxford, 1995, Vol. 4, p. 273. G. P. Pez, J. N. Armor, Adv. Organomet. Chem. 1981, 19, 1-50. R. F. Jordan, Adv. Organomet. Chem. 1991, 32, 325-387. S. Collins, D. G. Ward, J. Am. Chem. Soc. 1992, 114, 5460-5462. P. C. Wailes, R. S. P. Coutts, H. Weigold, Organometallic Chemistry of Titanium, Zirconium, and Hafnium, Academic Press, New York, 1974. J. M. O’Connor, C. P. Casey, Chem. Rev. 1987, 87, 307-318. J. L. Calderon, F. A. Cotton, J. Takats, J. Am. Chem. Soc. 1971, 93, 3587-3591. J. L. Calderon, F. A. Cotton, B. G. DeBoer, J. Takats, J. Am. Chem. Soc. 1971, 93, 3592-3597. G. Wilkinson, J. M. Birmingham, J. Am. Chem. Soc. 1954, 76, 4281-4284. R. B. King, M. B. Bisnette, J. Organomet. Chem. 1967, 8, 287-297. J. E. Bercaw, R. H. Marvich, L. G. Bell, H. H. Brintzinger, J. Am. Chem. Soc. 1972, 94, 12191238. J. E. Bercaw, J. Am. Chem. Soc. 1974, 96, 5087-5095. E. W. Abel, S. Moorhouse, J. Chem. Soc., Dalton Trans. 1973, 1706-1711. P. Jutzi, M. Kuhn, J. Organomet. Chem. 1979, 173, 221-229. S. Ciruelos, T. Cuenca, P. Gomez-Sal, A. Manzanero, P. Royo, Organometallics 1995, 14, 177-185. S. L. Hart, D. J. Dunclaf, J. J. Hastings, A. McCamley, P. C. Taylor, J. Chem. Soc., Dalton Trans. 1996, 2843-2849. G. H. Llinas, M. Mena, F. Palacios, P. Royo, R. Serrano, J. Organomet. Chem. 1988, 340, 3740. R. Beckhaus, J. Oster, B. Ganter, U. Englert, Organometallics 1997, 16, 3902-3909.
References [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62]
231
G. A. Luinstra, J. H. Teuben, J. Chem. Soc., Chem. Commun. 1990, 1470-1471. G. Chandra, M. F. Lappert, J. Chem. Soc. A 1968, 1940-1945. G. M. Diamond, R. F. Jordan, J. L. Petersen, Organometallics 1996, 15, 4030-4037. J. N. Christopher, G. M. Diamond, R. F. Jordan, J. L. Petersen, Organometallics 1996, 15, 4038-4044. G. M. Diamond, R. F. Jordan, J. L. Petersen, Organometallics 1996, 15, 4045-4053. W. A. Herrmann, M. J. A. Morawietz, J. Organomet. Chem. 1994, 482, 169-181. J. M. Birmingham, Adv. Organomet. Chem. 1964, 2, 365-413. P. G. Gassman, W. H. Campbell, D. W. Macomber, Organometallics 1984, 3, 385-387. K. Chandra, R. K. Sharma, B. S. Garg, R. P. Singh, Inorg. Chim. Acta 1979, 37, 125-127. R. B. Grossman, S. L. Buchwald, J. Org. Chem. 1992, 57, 5803-5805. A. Ohff, S. Pulst, C. Lefeber, N. Peulecke, P. Arndt, V. V. Burlakov, U. Rosenthal, Synlett 1996, 111-118. D. H. Berry, L. J. Procpio, P. J. Carroll, Organometallics 1988, 7, 570-572. J. M. Manriquez, D. R. McAlister, E. Rosenberg, A. M. Shiller, K. L. Williamson, S. I. Chan, J. E. Bercaw, J. Am. Chem. Soc. 1978, 100, 3078-3083. H. G. Alt, K.-H. Schwind, M. D. Rausch, U. Thewalt, J. Organomet. Chem. 1988, 349, C7C10. S. A. Cohen, J. E. Bercaw, Organometallics 1985, 4, 1006-1014. R. Beckhaus, K.-H. Thiele, J. Organomet. Chem. 1986, 317, 23-31. R. Beckhaus, K.-H. Thiele, D. Ströhl, J. Organomet. Chem. 1989, 369, 43-54. R. Beckhaus, S. Flatau, S. I. Troyanov, P. Hofmann, Chem. Ber. 1992, 125, 291-299. R. Beckhaus, Angew. Chem. 1997, 109, 694-722; Angew. Chem. Int. Ed. Engl. 1997, 36, 686713. H. G. Alt, G. S. Herrmann, M. D. Rausch, J. Organomet. Chem. 1988, 356, C50-C52. R. Beckhaus, J. Sang, J. Oster, T. Wagner, J. Organomet. Chem. 1994, 484, 179-190. R. Beckhaus, M. Wagner, V.V. Burlakov, W. Baumann, N. Peulecke, A. Spannenberg, R. Kempe, U. Rosenthal, Z. anorg. allg. Chem. 1998, 624, 129-134. J. W. Lauher, R. Hoffmann, J. Am. Chem. Soc. 1976, 98, 1729-1742. L. Zhu, N. M. Kostic, J. Organomet. Chem. 1987, 335, 395-411. V. C. Gibson, J. Chem. Soc., Dalton Trans. 1994, 1607-1618. V. C. Gibson, Angew. Chem. 1994, 106, 1640-1648; Angew. Chem. Int. Ed. Engl. 1994, 33, 1565-1572. Lukens Jr., W. W., M. R. Smith III, R. A. Andersen, J. Am. Chem. Soc. 1996, 118, 1719-1728. E. Brady, W. Lukens, J. Telford, G. Mitchell, Acta Cryst. 1994, C51, 558-560. G. L. Hillhouse, A. R. Bulls, B. D. Santarsiero, J. E. Bercaw, Organometallics 1988, 7, 13091312. J. Feldman, J. C. Calabrese, J. Chem. Soc., Chem. Commun. 1991, 1042-1044. J. M. Hawkins, R. H. Grubbs, J. Am. Chem. Soc. 1988, 110, 2821-2823. P. Binger, P. Müller, R. Wenz, R. Mynott, Angew. Chem. 1990, 102, 1070-1071; Angew. Chem. Int. Ed. Engl. 1990, 29, 1037-1038. P. G. Gassman, D. J. Macomber, J. W. Hershberger, Organometallics 1983, 2, 1470-1472. J. L. Robbins, N. Edelstein, B. Spencer, J. C. Smart, J. Am. Chem. Soc. 1982, 104, 1882-1893. E. J. Miller, S. J. Landon, T. B. Brill, Organometallics 1985, 4, 534-538. P. G. Gassman, C. H. Winter, Organometallics 1991, 10, 1592-1598. A. Terpstra, J. N. Louwen, A. Oskam, J. H. Teuben, J. Organomet. Chem. 1984, 260, 207-217. B. E. Bursten, M. R. Callstrom, A. J. Cynthia, L. A. Paquette, M. R. Sivik, R. S. Tucker, C. A. Wartchow, Organometallics 1994, 13, 127-133. P. G. Gassman, P. A. Deck, C. H. Winter, D. A. Dobbs, D. H. Cao, Organometallics 1992, 11, 959-960. S. A. Rao, M. Periasamy, J. Organomet. Chem. 1988, 352, 125-131. S. A. Cohen, P. R. Auburn, J. E. Bercaw, J. Am. Chem. Soc. 1983, 105, 1136-1143. M. L. Steigerwald, W. A. Goddard III, J. Am. Chem. Soc. 1985, 107, 5027-5035.
232 [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76]
[77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97]
4
Titanocenes
K. Mashima, K. Jyodoi, A. Ohyshi, H. Takaya, J. Chem. Soc., Chem. Commun. 1986, 1145-1146. K. Mashima, H. Takaya, Organometallics 1985, 4, 1464-1466. G. Fachinetti, C. Floriani, J. Chem. Soc., Chem. Commun. 1974, 66-67. G. Fachinetti, C. Floriani, F. Marchetti, M. Mellini, J. Chem. Soc., Dalton Trans. 1978, 13981403. B. Demerseman, R. Mahe, P. H. Dixneuf, J. Chem. Soc., Chem. Commun. 1984, 1394-1395. B. Demerseman, P. H. Dixneuf, J. Chem. Soc., Chem. Commun. 1981, 665-666. B. Demerseman, P. Le Coupanec, P. H. Dixneuf, J. Organomet. Chem. 1985, 287, C35-C38. B. H. Edwards, R. D. Rogers, D. J. Sikora, J. L. Atwood, M. D. Rausch, J. Am. Chem. Soc. 1983, 105, 416-426. V. Varga, K. Mach, M. Polasek, P. Sedmera, J. Hiller, U. Thewalt, S. I. Troyanov, J. Organomet. Chem. 1996, 506, 241-251. V. B. Shur, V. V. Burlakov, M. E. Volpin, J. Organomet. Chem. 1988, 347, 77-83. V. V. Burlakov, U. Rosenthal, P. V. Petrovsky, V. B. Shur, M. E. Volpin, Organomet. Chem. USSR 1988, 1, 953-954. V. V. Burlakov, U. Rosenthal, R. Beckhaus, A. V. Polyakov, Y. T. Struchkov, G. Oeme, V. B. Shur, M. E. Volpin, Organomet. Chem. USSR 1990, 3, 476-477. V. V. Burlakov, A. V. Polyakov, A. I. Yanovsky, Y. T. Struchkov, V. B. Shur, M. E. Vol’pin, U. Rosenthal, H. Görls, J. Organomet. Chem. 1994, 476, 197-206. [a] F. H. Allen, O. Kennrad, D. G. Watson, L. Brammer, A. G. Orpen, R. Taylor, J. Chem. Soc., Perkin Trans. II 1987, S1-S19. – [b] K. Krüger, Personal Communication. – [c] A.A. Espiritu, J.G. White, Z. Kristall. 1978, 147, 177-186. R. Threlkel, PhD Thesis, California Institute of Technology 1980. U. Rosenthal, G. Oehme, V. V. Burlakov, P. V. Petrovskii, V. B. Shur, M. E. Volpin, J. Organomet. Chem. 1990, 391, 119-122. D. J. Sikora, D. W. Macomber, M. D. Rausch, Adv. Organomet. Chem. 1986, 25, 317-379. D. J. Sikora, K. J. Moriarty, M. D. Rausch, A. R. Bulls, J. E. Bercaw, V. D. Patel, A. J. Carty, Inorg. Synth. 1986, 24, 147-157. L. B. Kool, M. D. Rausch, H. G. Alt, M. Herberhold, B. Wolf, U. Thewalt, J. Organomet. Chem. 1985, 297, 159-169. G. T. Palmer, F. Basolo, L. B. Kool, M. D. Rausch, J. Am. Chem. Soc. 1986, 108, 44174422. L. B. Kool, M. D. Rausch, H. G. Alt, M. Herberhold, U. Thewalt, B. Wolf, Angew. Chem. 1985, 97, 425-426; Angew. Chem. Int. Ed. Engl. 1985, 24, 394-395. L. B. Kool, M. D. Rausch, H. G. Alt, M. Herberhold, B. Honold, U. Thewalt, J. Organomet. Chem. 1987, 320, 37-45. L. B. Kool, M. D. Rausch, H. G. Alt, M. Herberhold in Organometallic Syntheses (Ed.: R. B. King, J. J. Eisch), Elsevier, Amsterdam, 1986, Vol. 3, p. 13. J. M. de Wolf, A. Meetsma, J. H. Teuben, Organometallics 1995, 14, 5466-5468. J. W. Pattiasina, F. van Bolhuis, J. H. Teuben, Angew. Chem. 1987, 99, 342-343; Angew. Chem. Int. Ed. Engl. 1987, 26, 330-331. S. R. Frerichs, B. K. Stein, J. E. Ellis, J. Am. Chem. Soc. 1987, 109, 5558-5560. Y. You, S. R. Wilson, G. S. Girolami, Organometallics 1994, 13, 4655-4657. a) S. I. Troyanov, H. Antropiusova, K. Mach, J. Organomet. Chem. 1992, 427, 49-55. - b) J. M. Fischer, W. E. Piers, V. G. Jr. Young, Organometallics 1996, 15, 2410-2412. S. I. Troyanov, K. Mach, V. Varga, Organometallics 1993, 12, 3387-3389. S. Xin, J. F. Harrod, E. Samuel, J. Am. Chem. Soc. 1994, 116, 11562-11563. K. M. Doxsee, J. K. M. Mouser, J. B. Farahi, Synlett 1992, 13-21. C. Lamberth, J. Prakt. Chem. 1994, 336, 632-633. S. H. Pine, G. S. Shen, H. Hoang, Synthesis 1991, 165-167. S. H. Pine, R. J. Pettit, G. D. Geib, S. G. Cruz, C. H. Gallego, T. Tijerina, R. D. Pine, J. Org. Chem. 1985, 50, 1212-1216. K. C. Ott, E. J. M. deBoer, R. H. Grubbs, Organometallics 1984, 3, 223-230.
References [98] [99]
[100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137]
233
S. H. Pine, Carbonyl Methylenation and Alkylidation using Titanium-Based Reagents in Organic Reactions (Ed.: L. A. Paquette), John Wiley & Sons, 1993, Vol. 43, p. 1. R. H. Grubbs, R. H. Pine, Comprehensive Organic Synthesis; Alkene Metathesis and Related Reactions in Comprehensive Organic Synthesis (Ed.: B. M. Trost), Pergamon, New York, 1991, Vol. 5, p. 1115. K. M. Doxsee, J. B. Farahi, J. Am. Chem. Soc. 1988, 110, 7239-7240. N. A. Petasis, S.-P. Lu, Tetrahedron Lett. 1995, 36, 2393-2396. J.-P. Begue, M. H. Rock, J. Organomet. Chem. 1995, 489, C7-C8. H. van der Heijden, B. Hessen, J. Chem. Soc., Chem. Commun. 1995, 145-146. L. Scoles, R. Minhas, R. Duchateau, J. Jubb, S. Gambarotta, Organometallics 1994, 13, 49784983. J. A. van Doorn, H. van der Heijden, A. G. Orpen, Organometallics 1994, 13, 4271-4277. K. M. Doxsee, J. J. J. Juliette, K. Zientara, G. Nieckarz, J. Am. Chem. Soc. 1994, 116, 2147-2148. N. A. Petasis, D.-K. Fu, J. Am. Chem. Soc. 1993, 115, 7208-7214. K. M. Doxsee, J. J. J. Juliette, J. K. M. Mouser, K. Zientara, Organometallics 1993, 12, 46824686. N. A. Petasis, D.-K. Fu, Organometallics 1993, 12, 3776-3780. N. A. Petasis, E. I. Bzowej, Tetrahedron Lett. 1993, 34, 943-946. N. A. Petasis, I. Akritopoulou, Synlett 1992, 665-667. N. A. Petasis, E. I. Bzowej, J. Org. Chem. 1992, 57, 1327-1330. K. M. Doxsee, J. B. Farahi, H. Hope, J. Am. Chem. Soc. 1991, 113, 8889-8898. K. M. Doxsee, J. K. M. Mouser, Organometallics 1990, 9, 3012-3014. N. A. Petasis, E. I. Bzowej, J. Am. Chem. Soc. 1990, 112, 6392-6394. B. J. J. van de Heisteeg, G. Schat, O. S. Akkerman, F. Bickelhaupt, Tetrahedron Lett. 1987, 28, 6493-6496. B. J. J. van de Heisteeg, G. Schat, O. S. Akkerman, F. Bickelhaupt, J. Organomet. Chem. 1986, 310, C25-C28. D. L. Hughes, J. F. Payack, D. Cai, T. R. Verhoeven, P. J. Reider, Organometallics 1996, 15, 663-667. J. A. van Doorn, H. van der Heijden, A. G. Orpen, Organometallics 1995, 14, 1278-1283. S. De Angelis, E. Solari, C. Floriani, A. Chiesi-Villa, C. Rizzoli, Angew. Chem. 1995, 107, 1200-1202; Angew. Chem. Int. Ed. Engl. 1995, 34, 1092-1094. M. D. Fryzuk, S. S. H. Mao, M. J. Zaworotko, L. R. MacGillivray, J. Am. Chem. Soc. 1993, 115, 5336-5337. A. Kabi-Satpathy, C. S. Bajgur, K. P. Reddy, J. L. Petersen, J. Organomet. Chem. 1989, 364, 105-117. F. J. Berg, J. L. Petersen, Organometallics 1989, 8, 2461-2470. B. J. J. van de Heisteeg, G. Schat, O. S. Akkerman, F. Bickelhaupt, Organometallics 1986, 5, 1749-1750. W. R. Tikkanen, J. Z. Liu, J. W. Egan Jr., J. L. Petersen, Organometallics 1984, 3, 825-830. K. C. Ott, R. H. Grubbs, J. Am. Chem. Soc. 1981, 103, 5922-5923. S. H. Pine, R. Zahler, D. A. Evans, R. H. Grubbs, J. Am. Chem. Soc. 1980, 102, 3270-3272. J. D. Meinhart, E. V. Anslyn, R. H. Grubbs, Organometallics 1989, 8, 583-589. W. C. Finch, E. V. Anslyn, R. H. Grubbs, J. Am. Chem. Soc. 1988, 110, 2406-2413. E. V. Anslyn, R. H. Grubbs, J. Am. Chem. Soc. 1987, 109, 4880-4890. L. R. Gilliom, R. H. Grubbs, Organometallics 1986, 5, 721-724. L. R. Gilliom, R. H. Grubbs, J. Am. Chem. Soc. 1986, 108, 733-742. T. Ikariya, S. C. H. Ho, R. H. Grubbs, Organometallics 1985, 4, 199-200. J. R. Stille, R. H. Grubbs, J. Am. Chem. Soc. 1983, 105, 1664-1665. J. B. Lee, K. C. Ott, R. H. Grubbs, J. Am. Chem. Soc. 1982, 104, 7491-7496. D. A. Straus, R. H. Grubbs, Organometallics 1982, 1, 1658-1661. J. B. Lee, G. J. Gajda, W. P. Schaefer, T. R. Howard, T. Ikariya, D. A. Straus, R. H. Grubbs, J. Am. Chem. Soc. 1981, 103, 7358-7361.
234
4
Titanocenes
[138] T. R. Howard, J. B. Lee, R. H. Grubbs, J. Am. Chem. Soc. 1980, 102, 6876-6878. [139] F. Bickelhaupt, Angew. Chem. 1987, 99, 1020-1035; Angew. Chem. Int. Ed. Engl. 1987, 26, 990-1005. [140] G. L. Casty, J. M. Stryker, J. Am. Chem. Soc. 1995, 117, 7814-7815. [141] D. A. Straus, R. H. Grubbs, J. Mol. Catal. 1985, 28, 9-25. [142] J. L. Polse, R. A. Andersen, R. G. Bergman, J. Am. Chem. Soc. 1996, 118, 8737-8738. [143] P. Binger, P. Müller, R. Benn, R. Mynott, Angew. Chem. 1989, 101, 647-648; Angew. Chem. Int. Ed. Engl. 1989, 28, 610-611. [144] P. Binger, P. Müller, P. Phillipps, B. Gabor, R. Mynott, A. T. Herrmann, F. Langhauser, C. Krüger, Chem. Ber. 1992, 125, 2209-2212. [145] E. Hengge, M. Weinberger, J. Organomet. Chem. 1993, 443, 167-173. [146] J. L. Bennett, P. T. Wolczanski, J. Am. Chem. Soc. 1994, 116, 2179-2180. [147] J. M. Fischer, W. E. Piers, T. Ziegler, L. R. MacGillivray, M. J. Zaworotko, Chem. Eur. J. 1996, 2, 1221-1229. [148] M. R. Smith, P. T. Matsunaga, R. A. Andersen, J. Am. Chem. Soc. 1993, 115, 7049-7050. [149] F. Bottomley, G. O. Egharevba, I. J. B. Lin, P. S. White, Organometallics 1985, 4, 550-553. [150] D. J. Schwartz, M. R. Smith III, R. A. Andersen, Organometallics 1996, 15, 1446-1450. [151] U. Thewalt, B. Honold, J. Organomet. Chem. 1988, 348, 291-303. [152] H.-P. Klein, U. Thewalt, Z. Anorg. Allg. Chem. 1981, 476, 62-68. [153] H.-P. Klein, U. Thewalt, J. Organomet. Chem. 1981, 206, 69-75. [154] U. Thewalt, H.-P. Klein, J. Organomet. Chem. 1980, 194, 297-307. [155] T. K. Hollis, N. P. Robinson, B. Bosnich, J. Am. Chem. Soc. 1992, 114, 5464-5466. [156] H. Aslan, T. Sielisch, R. D. Fischer, J. Organomet. Chem. 1986, 315, C69-C72. [157] M. L. McLaughlin, J. M. Cronan Jr., T. R. Schaller, R. D. Snelling, J. Am. Chem. Soc. 1990, 112, 8949-8952. [158] H. Köpf, P. Köpf-Maier, Nachr. Chem. Tech. Lab. 1981, 29, 154-156. [159] P. Köpf-Maier, H. Köpf, J. Organomet. Chem. 1988, 342, 167-176. [160] J. H. Toney, T. J. Marks, J. Am. Chem. Soc. 1985, 107, 947-953. [161] K. Döppert, J. Organomet. Chem. 1987, 319, 351-354. [162] K. Döppert, U. Thewalt, J. Organomet. Chem. 1986, 301, 41-48. [163] P. Gowik, T. Klapötke, J. Organomet. Chem. 1989, 372, 33-40. [164] T. Klapötke, Polyhedron 1989, 8, 311-315. [165] P. Gowik, T. Klapötke, U. Thewalt, J. Organomet. Chem. 1990, 385, 345-350. [166] P. Gowik, T. Klapötke, J. Organomet. Chem. 1990, 387, C27-C30. [167] U. Thewalt, K. Berhalter, J. Organomet. Chem. 1986, 302, 193-200. [168] A. Ohff, R. Kempe, W. Baumann, U. Rosenthal, J. Organomet. Chem. 1996, 520, 241-244. [169] H. H. Brintzinger, J. E. Bercaw, J. Am. Chem. Soc. 1970, 92, 6182-6185. [170] G. W. Watt, L. J. Baye, F. O. Jr. Drummond, J. Am. Chem. Soc. 1966, 88, 1138-1140. [171] J.-J. Salzmann, P. Mosimann, Helv. Chim. Acta 1967, 7, 1830-1836. [172] E. E. van Tamelen, W. Cretney, N. Klaentsch, J. S. Miller, J. Chem. Soc., Chem. Commun. 1972, 481-482. [173] K. Clauss, H. Bestian, Liebigs Ann. Chem. 1962, 654, 8-19. [174] L. J. Guggenberger, F. N. Tebbe, J. Am. Chem. Soc. 1973, 95, 7870-7872. [175] A. Davison, S. S. Wreford, J. Am. Chem. Soc. 1974, 96, 3017-3018. [176] L. J. Guggenberger, F. N. Tebbe, J. Am. Chem. Soc. 1976, 98, 4137-4243. [177] G. J. Olthof, J. Organomet. Chem. 1977, 128, 367-373. [178] H. Antropiusova, V. Hanus, K. Mach, Transition Met. Chem. 1978, 3, 121-122. [179] E. G. Perevalova, I. F. Uranzowski, D. A. Lemenovskii, Y. L. Slovokhotov, Y. T. Struchkov, J. Organomet. Chem. 1985, 289, 319-329. [180] T. Wöhrle, U. Thewalt, J. Organomet. Chem. 1993, 456, C21-C23. [181] U. Thewalt, T. Wöhrle, Z. Naturforsch. 1993, 48b, 603-607. [182] U. Thewalt, T. Wöhrle, J. Organomet. Chem. 1996, 506, 331-335. [183] C. McDade, J. C. Green, J. E. Bercaw, Organometallics 1982, 1, 1629-1634.
References [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198] [199] [200] [201] [202] [203] [204] [205] [206] [207]
[208] [209] [210] [211] [212] [213] [214] [215] [216] [217] [218] [219] [220] [221] [222] [223]
235
G. A. Luinstra, J. H. Teuben, Organometallics 1992, 11, 1793-1801. K. Mach, V. Varga, V. Hanus, P. Sedmera, J. Organomet. Chem. 1991, 415, 87-95. T. Vondrák, K. Mach, V. Varga, A. Terpstra, J. Organomet. Chem. 1992, 425, 27-39. C. Kreuder, R. F. Jordan, H. Zhang, Organometallics 1995, 14, 2993-3001. J. W. Pattiasina, C. E. Hissink, J. L. de Boer, A. Meetsma, J. H. Teuben, A. L. Spek, J. Am. Chem. Soc. 1985, 107, 7758-7759. G. Erker, U. Korek, Z. Naturforsch. 1989, 44b, 1593-1598. R. Beckhaus, J. Oster, J. Sang, I. Strauß, M. Wagner, Synlett 1997, 241-249. R. Beckhaus, U. Böhme, unpublished results. B. Akermark, A. Ljungqvist, J. Organomet. Chem. 1979, 182, 59-75. J. M. Brown, N. A. Cooley, Chem. Rev. 1988, 88, 1031-1046. P. Courtot, V. Labed, R. Pichon, J. Y. Salaün, J. Organomet. Chem. 1989, 359, C9-C13. S. L. Buchwald, R. B. Nielsen, Chem. Rev. 1988, 88, 1047-1058. C. Lefeber, A. Ohff, A. Tillack, W. Baumann, R. Kempe, V. V. Burlakov, U. Rosenthal, H. Görls, J. Organomet. Chem. 1995, 501, 179-188. U. Rosenthal, A. Ohff, A. Tillack, W. Baumann, H. Goerls, J. Organomet. Chem. 1994, 468, C4-C8. U. Rosenthal, H. Görls, J. Organomet. Chem. 1992, 439, C36-C41. A. R. Bulls, W. P. Schaefer, M. Serfas, J. E. Bercaw, Organometallics 1987, 6, 1219-1226. L. H. Toporcer, R. E. Dessy, S. I. E. Green, J. Am. Chem. Soc. 1965, 87, 1236-1240. E. L. Motell, A. W. Boone, W. H. Fink, Tetrahedron 1978, 34, 1619-1626. N. M. Doherty, J. E. Bercaw, J. Am. Chem. Soc. 1985, 107, 2670-2682. R. Beckhaus, J. Chem. Soc., Dalton Trans. 1997, 1991-2001. C. P. Boekel, J. H. Teuben, H. J. De Liefde Meijer, J. Organomet. Chem. 1974, 81, 371-377. I. S. Kolomnikov, T. S. Lobeeva, V. V. Gorbachevskaya, G. G. Aleksandrov, Y. T. Struchkov, M. E. Volpin, J. Chem. Soc., Chem. Commun. 1971, 972-973. V. B. Shur, E. G. Berkovitch, L. B. Vasiljeva, R. V. Kudryavtsev, M. E. Volpin, J. Organomet. Chem. 1974, 78, 127-132. R. Beckhaus, Organic Synthesis via Organometallics (OSM 4) Methylidentitanacyclobutane vs. Titanocene-Vinylidene - Versatile Building Blocks in Organic Synthesis via Organometallics (OSM 4), Proceedings of the Fourth Symposium in Aachen, July 15 to 18, 1992 (Ed.: D. Enders, H.-J. Gais, W. Keim), Vieweg Verlag, Braunschweig, 1993, p. 131. R. Beckhaus, J. Sang, T. Wagner, B. Ganter, Organometallics 1996, 15, 1176-1187. R. Beckhaus, M. Wagner, R. Wang, Z. Anorg. Allg. Chem. 1998, 624, 277-280. R. Beckhaus, I. Strauß, T. Wagner, J. Organomet. Chem. 1994, 464, 155-161. C. C. Cummins, C. P. Schaller, G. D. Van Duyne, P. T. Wolczanski, A. W. E. Chan, R. Hoffmann, J. Am. Chem. Soc. 1991, 113, 2985-2994. R. Beckhaus, J. Oster, Z. Anorg. Allg. Chem. 1995, 621, 359-364. R. Beckhaus, I. Strauß, unpublished results. R. Aumann, Angew. Chem. 1988, 100, 1512-1524; Angew. Chem. Int. Ed. Engl. 1988, 27, 1456-1467. F. N. Tebbe, R. L. Harlow, J. Am. Chem. Soc. 1980, 102, 6149-6151. R. J. McKinney, T. H. Tulip, D. L. Thorn, T. S. Coolbaugh, F. N. Tebbe, J. Am. Chem. Soc. 1981, 103, 5584-5586. P. Binger, P. Müller, A. T. Herrmann, P. Philipps, B. Gabor, F. Langhauser, C. Krüger, Chem. Ber. 1991, 124, 2165-2170. R. Beckhaus, M. Wagner, R. Wang, Eur. J. Inorg. Chem., 1998, 253-256. K. C. Wallace, A. H. Liu, W. M. Davis, R. R. Schrock, Organometallics 1989, 8, 644-654. R. Beckhaus, J. Sang, U. Englert, U. Böhme, Organometallics 1996, 15, 4731-4736. K. A. Jørgensen, B. Schiøtt, Chem. Rev. 1990, 90, 1483-1506. R. Beckhaus, I. Strauß, T. Wagner, P. Kiprof, Angew. Chem. 1993, 105, 281-283; Angew. Chem. Int. Ed. Engl. 1993, 32, 264-266. R. Beckhaus, I. Strauß, T. Wagner, Z. Anorg. Allg. Chem. 1997, 623, 654-658.
236 [224] [225] [226] [227] [228] [229] [230] [231] [232] [233] [234] [235] [236] [237] [238] [239] [240] [241] [242] [243] [244] [245] [246] [247] [248] [249] [250] [251] [252] [253] [254] [255] [256] [257] [258] [259] [260] [261] [262] [263] [264] [265]
4
Titanocenes
V. W. Day, W. G. Klemperer, S. P. Lockledge, D. J. Main, J. Am. Chem. Soc. 1990, 112, 2031-2033. S. C. Ho, S. Hentges, R. H. Grubbs, Organometallics 1988, 7, 780-782. I. Strauß, Diplomarbeit, RWTH Aachen 1992. B. Schiøtt, K. A. Jørgensen, J. Chem. Soc., Dalton Trans. 1993, 337-344. S. L. Buchwald, R. H. Grubbs, J. Am. Chem. Soc. 1983, 105, 5490-5491. R. Beckhaus, J. Oster, T. Wagner, Chem. Ber. 1994, 127, 1003-1013. G. Erker, Angew. Chem. 1989, 101, 411-426; Angew. Chem. Int. Ed. Engl. 1989, 28, 397-412. G. Erker, M. Mena, U. Hoffmann, B. Minjón, J. L. Petersen, Organometallics 1991, 10, 291298. E. V. Anslyn, B. D. Santarsiero, R. H. Grubbs, Organometallics 1988, 7, 2137-2145. G. Proulx, R. G. Bergman, J. Am. Chem. Soc. 1993, 115, 9802-9803. K. H. Dötz, H. Fischer, P. Hofmann, F. R. Kreissl, U. Schubert, K. Weiss, Transition Metal Carbene Complexes, Verlag Chemie, Weinheim 1983. K. M. Doxsee, J. B. Farahi, J. Chem. Soc., Chem. Commun. 1990, 1452-1454. R. Beckhaus, I. Strauß, T. Wagner, Angew. Chem. 1995, 107, 738-740; Angew. Chem. Int. Ed. Engl. 1995, 34, 688-690. R. Beckhaus, J. Sang, T. Wagner, U. Böhme, J. Chem. Soc., Dalton Trans. 1997, 2249-2255. R. Beckhaus, J. Sang, unpublished results. J. March, Advanced Organic Chemistry, John Wiley & Sons, New York, 1985. G. E. Herberich, C. Kreuder, U. Englert, Angew. Chem. 1994, 106, 2589-2590; Angew. Chem. Int. Ed. Engl. 1994, 33, 2465-2466. R. Beckhaus, C. Zimmermann, T. Wagner, E. Herdtweck, J. Organomet. Chem. 1993, 460, 181-189. R. Beckhaus, I. Strauß, unpublished. R. D. Dennehy, R. J. Whitby, J. Chem. Soc., Chem. Commun. 1992, 35-36. M. M. Francl, W. J. Hehre, Organometallics 1983, 2, 457-459. U. Klabunde, F. N. Tebbe, G. W. Parshall, R. L. Harlow, J. Mol. Catal. 1980, 8, 37-51. J. W. Park, L. M. Henling, W. P. Schaefer, R. H. Grubbs, Organometallics 1991, 10, 171-175. F. Ozawa, J. W. Park, P. B. Mackenzie, W. P. Schaefer, L. M. Henling, R. H. Grubbs, J. Am. Chem. Soc. 1989, 111, 1319-1327. P. B. Mackenzie, R. J. Coots, R. H. Grubbs, Organometallics 1989, 8, 8-14. J. W. Park, P. B. Mackenzie, W. P. Schaefer, R. H. Grubbs, J. Am. Chem. Soc. 1986, 108, 6402-6404. B. J. J. van de Heisteeg, G. Schat, O. S. Akkerman, F. Bickelhaupt, Organometallics 1985, 4, 1141-1142. M. I. Bruce, Chem. Rev. 1991, 91, 197-257. H. P. Kim, S. Kim, R. A. Jacobson, R. J. Angelici, J. Am. Chem. Soc. 1986, 108, 5154-5158. R. Boese, M. A. Huffmann, K. P. C. Vollhardt, Angew. Chem. 1991, 103, 1542-1543; Angew. Chem. Int. Ed. Engl. 1991, 30, 1463-1464. J. Oster, Dissertation, RWTH Aachen 1996. R. Beckhaus, J. Oster, R. Wang, Organometallics, 1998, 17, in press. G. Wilkinson, F. A. Cotton, J. M. Birmingham, J. Inorg. Nucl. Chem. 1956, 2, 95-113. E. Vitz, C. H. Jr. Brubaker, J. Organomet. Chem. 1974, 82, C16-C18. M. H. Peng, C. H. Jr. Brubaker, J. Organomet. Chem. 1977, 135, 333-337. J. G.-S. Lee, C. H. Jr. Brubaker, Inorg. Chim. Acta 1977, 25, 181-184. W. E. Crowe, A. T. Vu, J. Am. Chem. Soc. 1996, 118, 1557-1558. W. E. Crowe, A. T. Vu, J. Am. Chem. Soc. 1996, 118, 5508-5509. U. Rosenthal, C. Lefeber, P. Arndt, A. Tillack, W. Baumann, R. Kempe, V. V. Burlakov, J. Organomet. Chem. 1995, 503, 221-223. A. Tillack, W. Baumann, A. Ohff, C. Lefeber, A. Spannenberg, R. Kempe, U. Rosenthal, J. Organomet. Chem. 1996, 520, 187-193. R. Gleiter, W. Wittwer, Chem. Ber. 1994, 127, 1797-1798. E. J. M. De Boer, J. De With, J. Organomet. Chem. 1987, 320, 289-293.
References
237
[266] D. Cozak, M. Melnik, Coord. Chem. Rev. 1986, 14, 53-99. [267] A. Clearfield, D. K. Warner, C. H. Saldarriaga-Molina, R. Ropal, I. Bernal, Can. J. Chem. 1975, 53, 1622-1629. [268] S. I. Troyanov, V. B. Rybakov, U. Thewalt, V. Varga, K. Mach, J. Organomet. Chem. 1993, 447, 221-225. [269] R. D. Rogers, M. M. Benning, L. K. Kurihara, K. J. Moriarty, M. D. Rausch, J. Organomet. Chem. 1985, 293, 51-60. [270] T. C. McKenzie, R. D. Sanner, J. E. Bercaw, J. Organomet. Chem. 1975, 102, 457-466. [271] U. Thewalt, T. Wöhrle, J. Organomet. Chem. 1994, 464, C17-C19. [272] J. W. Pattiasina, H. J. Heeres, F. Van Bolhuis, A. Meetsma, J. H. Teuben, A. L. Spek, Organometallics 1987, 6, 1004-1010. [273] G. A. Luinstra, L. C. ten Cate, H. J. Heeres, J. W. Pattiasina, A. Meetsma, J. H. Teuben, Organometallics 1991, 10, 3227-3237. [274] D. J. Sikora, M. D. Rausch, R. D. Rogers, J. L. Atwood, J. Am. Chem. Soc. 1981, 103, 1265-1267. [275] J. L. Atwood, K. E. Stone, H. G. Alt, D. C. Hrncir, M. D. Rausch, J. Organomet. Chem. 1977, 132, 367-375. [276] T. E. Bitterwolf, A. C. Ling, J. Organomet. Chem. 1977, 141, 355. [277] J. H. Davis, H.-N. Sun, D. Redfield, G. D. Stucky, J. Magn. Res. 1980, 37, 441-448. [278] E. Samuel, M. D. Rausch, J. Am. Chem. Soc. 1973, 95, 6263-6266. [279] S. J. Heyes, C. M. Dobson, J. Am. Chem. Soc. 1991, 113, 463-469. [280] A. J. Edwards, N. J. Burke, C. M. Dobson, K. Prout, S. J. Heyes, J. Am. Chem. Soc. 1995, 117, 4637-4653. [281] A. Hafner, J. Okuda, Organometallics 1993, 12, 949-950. [282] M. R. M. Bruce, A. Kenter, D. R. Tayler, J. Am. Chem. Soc. 1984, 106, 639-644. [283] M. T. Reetz, Titanium in Organic Synthesis-A Manual in Organometallics in Synthesis (Ed.: M. Schlosser), John Wiley & Sons Ltd, Chichester, 1994, p. 195. [284] F. Küber, Metallocenes as a Source of Fine Chemicals in Applied Homogenous Catalysis with Organometallic Compounds (Ed.: B. Cornils, W. A. Herrmann), VCH, Weinheim, 1996, Vol. 2, p. 893. [285] C. A. Willoughby, S. L. Buchwald, J. Am. Chem. Soc. 1992, 114, 7562-7564. [286] A. Viso, N. E. Lee, S. L. Buchwald, J. Am. Chem. Soc. 1994, 116, 9373-9374. [287] T. V. RajanBau, W. A. Nugent, M. S. Beattie, J. Am. Chem. Soc. 1990, 112, 6408-6409. [288] J. F. Harrod, S. S. Yun, Organometallics 1987, 6, 1381-1387. [289] R. A. Sheldon, Synthesis of Oxiranes in Applied Homogenous Catalysis with Organometallic Compounds (Ed.: B. Cornils, W. A. Herrmann), VCH, Weinheim, 1996, Vol. 1, p. 411. [290] R. L. Halterman, T. M. Ramsey, Organometallics 1993, 12, 2879-2880. [291] S. L. Colletti, R. L. Halterman, J. Organomet. Chem. 1993, 455, 99-106. [292] W. A. Herrmann, Double-Bond Isomerization of Olefins in Applied Homogenous Catalysis with Organometallic Compounds (Ed.: B. Cornils, W. A. Herrmann), VCH, Weinheim, 1996, Vol. 2, p. 980. [293] T. K. Hollis, N. P. Robinson, B. Bosnich, Organometallics 1992, 11, 2747-2748. [294] J. B. Jaquith, J. Guan, S. Collins, S. Wang, Organometallics 1995, 14, 1079-1081. [295] J. L. Kiplinger, T. G. Richmond, C. E. Osterberg, Chem. Rev. 1994, 94, 373-431. [296] J. L. Kiplinger, T. G. Richmond, J. Am. Chem. Soc. 1996, 118, 1805-1806. [297] J. L. Kiplinger, T. G. Richmond, Chem. Commun. 1996, 1115-1116. [298] M. T. Didiuk, C. W. Johannes, J. P. Morken, A. H. Hoveyda, J. Am. Chem. Soc. 1995, 117, 7097-7104. [299] F. Sato, J. Organomet. Chem. 1985, 285, 53-64. [300] Y. Gao, F. Sato, J. Chem. Soc., Chem. Commun. 1995, 659-690. [301] Y. Gao, F. Sato, J. Chem. Soc., Chem. Commun. 1995(6), 659-660. [302] S. Collins, W. P. Dean, D. G. Ward, Organometallics 1988, 7, 2289-2293. [303] Y. Cauvin, H. Oliver, Reactions of unsaturated compounds: Polymerization, Oligomerization, and Coplymerization of Olefins - Dimerization and Codimerization in Applied Homogenous
238
[304] [305] [306] [307] [308] [309] [310] [311] [312] [313] [314] [315] [316] [317] [318] [319] [320] [321] [322] [323] [324] [325] [326] [327] [328] [329] [330] [331] [332] [333] [334] [335] [336] [337] [338] [339] [340] [341] [342] [343] [344]
4
Titanocenes
Catalysis with Organometallic Compounds (Ed.: B. Cornils, W. A. Herrmann), VCH, Weinheim, 1996, Vol. 1, p. 258. V. Varga, L. Petrusova, J. Cejka, V. Hanus, K. Mach, J. Organomet. Chem. 1996, 509, 235240. T. Yoshida, E. Negishi, J. Am. Chem. Soc. 1981, 103, 1276-1277. R. D. Dennehy, R. J. Whitby, J. Chem. Soc., Chem. Commun. 1990, 1060-1062. M. J. Burk, W. Tumas, M. D. Ward, D. R. Wheeler, J. Am. Chem. Soc. 1990, 112, 6133-6135. M. J. Burk, D. L. Staley, W. Tumas, J. Chem. Soc., Chem. Commun. 1990, 809-810. W. Tumas, J. A. Suriano, R. L. Harlow, Angew. Chem. 1990, 102, 89-90; Angew. Chem. Int. Ed. Engl. 1990, 29, 75-76. W. Tumas, J. C. Huang, P. E. Fanwick, C. P. Kubiak, Organometallics 1992, 11, 2944-2947. R. R. Schrock, Pure Appl. Chem. 1994, 66, 1447-1454. R. H. Grubbs, W. Tumas, Science 1989, 243, 907-915. J. Feldman, R. R. Schrock, Prog. Inorg. Chem. 1991, 39, 1-74. A. J. Amass, Compr. Polym. Sci. 1989, 4, 109-134. W. J. Fast, Compr. Polym. Sci. 1989, 4, 135-142. R. H. Grubbs, L. Gilliom, NATO ASI Ser., Ser. C 1987, 215 (Recent Adv. Mech. Synth. Aspects Polym.), 343-352. W. Risse, R. H. Grubbs, J. Mol. Catal. 1991, 65, 211-217. T. M. Swager, R. H. Grubbs, J. Am. Chem. Soc. 1987, 109, 894-896. J. D. Meinhart, R. H. Grubbs, Bull. Chem. Soc. Jpn. 1988, 61, 171-180. K. M. Doxsee, J. K. M. Mouser, Tetrahedron Lett. 1991, 32, 1687-1690. K. M. Doxsee, G. S. Shen, C. B. Knobler, J. Am. Chem. Soc. 1989, 111, 9129-9130. L. Clawson, S. L. Buchwald, R. H. Grubbs, Tetrahedron Lett. 1984, 25, 5733-5736. C. P. Gibson, D. S. Bem, J. Organomet. Chem. 1991, 414, 23-32. S. H. Bertz, G. Dabbagh, C. P. Gibson, Organometallics 1988, 7, 563-565. P. Veya, C. Floriani, A. Chiesi-Villa, C. Rizzoli, Organometallics 1993, 12, 4892-4898. M. D. Curtis, S. Thanedar, W. M. Butler, Organometallics 1984, 3, 1855-1859. P. Berno, C. Floriani, A. Chiesi-Villa, C. Guastini, Organometallics 1990, 9, 1995-1997. D. A. Evans, F. Urpi, T. C. Somers, J. S. Clark, M. T. Bilodeau, J. Am. Chem. Soc. 1990, 112, 8215-8216. K. J. Ivin, Olefin Metathesis, Academic, New York, 1983. B. M. Novak, W. Risse, R. H. Grubbs, Adv. Polym. Sci. 1992, 102, 47-72. T. R. Cundari, M. S. Gordon, Organometallics 1992, 11, 55-63. T. R. Cundari, M. S. Gordon, J. Am. Chem. Soc. 1992, 114, 539-548. T. R. Cundari, M. S. Gordon, J. Am. Chem. Soc. 1991, 113, 5231-5243. H. G. Alt, H. E. Engelhardt, M. D. Rausch, L. B. Kool, J. Organomet. Chem. 1987, 329, 6167. R. D. Ernst, Chem. Rev. 1988, 88, 1255-1291. J.-Z. Liu, R. D. Ernst, J. Am. Chem. Soc. 1982, 104, 3737-3739. R. D. Ernst, J. W. Freeman, P. N. Swepston, D. R. Wilson, J. Organomet. Chem. 1991, 402, 17-25. R. D. Ernst, J. W. Freeman, L. Stahl, D. R. Wilson, A. M. Arif, B. Nuber, M. L. Ziegler, J. Am. Chem. Soc. 1995, 117, 5075-5081. E. Melendez, A. M. Arif, M. L. Ziegler, R. D. Ernst, Angew. Chem. 1988, 100, 1132-1134; Angew. Chem. Int. Ed. Engl. 1988, 27, 1099-1111. I. Hyla-Kryspin, T. E. Waldman, E. Melendez, W. Trakarnpruk, A. M. Arif, M. L. Ziegler, R. D. Ernst, R. Gleiter, Organometallics 1995, 14, 5030-5040. T. P. Hanusa, Polyhedron 1982, 1, 663-665. J. E. Hill, G. Balaich, P. E. Fanwick, I. P. Rothwell, Organometallics 1993, 12, 2911-2924. G. J. Balaich, J. E. Hill, S. A. Waratuke, P. E. Fanwick, I. P. Rothwell, Organometallics 1995, 14, 656-665. G. J. Balaich, P. E. Fanwick, I. P. Rothwell, Organometallics 1994, 13, 4117-4118.
References [345] [346] [347] [348] [349] [350] [351] [352] [353] [354] [355]
239
I. P. Rothwell, Acc. Chem. Res. 1988, 21, 153-159. T. V. Lubben, P. T. Wolczanski, G. D. van Duyne, Organometallics 1984, 3, 977-983. R. E. Lapointe, P. T. Wolczanski, G. D. van Duyne, Organometallics 1985, 4, 1810-1818. K. J. Covert, P. T. Wolczanski, Inorg. Chem. 1989, 28, 4565-4567. R. K. Minhas, L. Scoles, S. Wong, S. Gambarotta, Organometallics 1996, 15, 1113-1121. A. R. Johnson, W. M. Davis, C. C. Cummins, Organometallics 1996, 15, 3825-3835. T. L. Brown, K. J. Lee, Coord. Chem. Rev. 1993, 128, 89-116. R. Kempe, S. Brenner, P. Arndt, Organometallics 1996, 15, 1071-1074. T. H. Warren, R. R. Schrock, W. M. Davis, Organometallics 1996, 15, 562-569. K. Aoyagi, P. K. Gantzel, K. Kalai, T. D. Tilley, Organometallics 1996, 15, 923-927. H. C. S. Clark, G. N. Cloke, P. B. Hitchcock, J. B. Love, J. Organomet. Chem. 1995, 501, 333-340. [356] R. Gomez, R. Duchateau, A. N. Chernega, J. H. Teuben, F. T. Edelmann, M. L. H. Green, J. Organomet. Chem. 1995, 491, 153-158. [357] J. R. Hagedorn, J. Arnold, J. Am. Chem. Soc. 1996, 118, 893-894. [358] M. Wedler, F. Knösel, F. T. Edelmann, U. Behrens, Chem. Ber. 1992, 125, 1313-1318.
240
4
Titanocenes
5
Zirconocenes Ei-ichi Negishi and Jean-Luc Montchamp
5.1 Introduction This chapter primarily deals with the formation and reactions of zirconocene derivatives with emphasis on fundamental reaction patterns and attendant mechanistic aspects of synthetically interesting reactions. This area of chemistry has rapidly and enormously grown over the past two decades, and it is no longer practical to even touch on all important topics. Fortunately, there are a few monographs [1, 2] and a large number of pertinent chapters including Chapters 7–10 of this book and reviews in journals [3–40]. So, the authors’ attention is mainly focused on providing a coherent overview of the subject and pertinent references. For those topics that are not properly represented in this chapter, the readers are referred to pertinent chapters and reviews which are classified and listed at the end of this chapter for this very purpose. Zirconium discovered in 1824 [41] occurs to the extent of 0.022% in the lithosphere and is roughly as abundant as carbon. It is less abundant than titanium (0.63%) but roughly 40 times as abundant as hafnium (0.0053%) [1]. A hundred and thirty years after its discovery, Wilkinson and his coworkers reported zirconocene dihalides, Cp2ZrX2, where Cp = η5-C5H5 and X = Cl or Br, as some of the earliest examples of organozirconium compounds [42]. Despite the fact that Zr is one of the several least expensive transition metals ($11 mol–1 of 98% pure ZrCl4, Strem Chemicals) along with Fe, Cu, Mn, and Ti, the use of Zr and organozirconium compounds in organic synthesis had been virtually unknown until the 1970s, with the exception of occasional uses of Zr salts as somewhat exotic Lewis acid catalysts in the Friedel–Crafts reactions [43] and generally inferior substitutes for Ti salts in the Ziegler–Natta polymerization [44]. Thus, Zr had mainly remained as a basic scientific curiosity. Triggered by some epoch-making discoveries and developments, ten or a dozen of which are listed in Table 5-1, Zr has become over the last two decades one of the most widely used transition metals in organic synthesis. While it still lags behind Cu and Pd in terms of versatility and frequency of use, it now appears to be as widely used as several others, such as Ti, Cr, Fe, Co, Rh, and Ni. It is striking that all of the reactions listed in Table 5-1 involve Cp2Zr derivatives including those containing modified Cp derivatives.
242
5
Zirconocenes
Table 5-1. Some Notable Discoveries and Developments in the Use of Organozirconium Compounds in Organic Synthesis
Since the chemistry of chiral metallocenes and the Ziegler–Natta polymerization are discussed in Chapters 8–10, their discussion in this chapter is kept at a minimum. Limitation of space does not permit discussion of all of the important topics. As a consequence, discussion of many pioneering and basic studies not emphasizing synthetic applications is also kept at a minimum. For extensive discussion of the basic aspects of zirconocene chemistry, the readers are referred to a few monographs and reviews [1–40]. The great majority of the currently known ZrCp2 derivatives are Zr(IV) and Zr(II) compounds. Some Zr(III) species, such as those shown in Scheme 5-1, are known [60, 61], and the number of such compounds is slowly but steadily increasing. Their formation and structure are of obvious scientific interest. At present, however, there is virtually no synthetically useful chemistry of these compounds known. In fact, they often represent unwanted byproducts to be avoided in the reactions of
5.1
243
Introduction
Cp2Zr(IV) and/or Cp2Zr(II) compounds, and they are significant at present primarily within this context, although some useful new chemistry dealing with Zr(III) compounds, such as those shown in Scheme 5-1, is conceivable. Me I Cp2Zr
X X
ZrCp2
Cp2Zr
ZrCp2
Cp2Zr
I
X = Cl, I, PPhH, etc.
Scheme 5-1
Various structural types are conceivable for Cp2Zr(IV) and Cp2Zr(II) compounds, and examples of most of them are known. Some representative structural types are shown in Table 5-2. The following general comments are in order with regards to the structures shown in Table 5-2. First, formal charges are shown in some cases, but they are frequently omitted, in part, for the sake of simplicity. For example, one may argue that Cp2Zr(PMe3)2 might be better represented by Cp2Zr2–(P+Me3)2 without considering backdonation. However, it appears reasonable to omit all formal charges for simplicity in cases where the species shown are overall neutral. Second, some of the structures shown in Table 5-2, such as the structures IV-1 and XV-1 (Scheme 5-2), can represent two resonance structures of the same species. In fact, there have been no cases where two isomeric compounds should be separately represented by IV-1 and XV-1. On the other hand, many experimental results and observations have supported the notion that IV-1 and XV-1 may be considered as two resonance structures of the same species, which may be viewed as either alkene– ZrIICp2 complexes or zirconacyclopropanes containing a ZrIV atom and that either may be interchangeably used. It appears useful and reasonable to compare IV-1 with the classical bent-bond structure of cyclopropane and XV-1 with its Walsh structure [62]. Admittedly, some such species may be better represented by IV-1, while others by XV-1, but no attempts are made to discuss such a subtle issue here. Third, it is potentially very important to realize that any coordinatively unsaturated species, i.e., 16- and 14-electron species in Table 5-2, can exist as dimeric, oligomeric, or polymeric aggregates. Fourth, such compounds can also form complexes with Lewis bases. For the sake of simplicity, however, such structures are not shown in Table 5-2. As stated earlier, the formation and interconversion of these Cp2Zr(IV) and Cp2Zr(II) species as well as the attendant organic transformations are the main topics of this chapter.
244
5
Zirconocenes
Table 5-2. Structural Types of Cp2Zr(IV) and Cp2Zr(II) Compounds Cp2Zr(IV) Compounds Type
IV
I
Examples (Other Structural Types or Comments)
Structure
Cp2Zr 16e
a
Cp2ZrXY, Cp2ZrH2, Cp2Zr(H)X, Cp2ZrRX, Cp2ZrR1R2, etc.
b
IV
II
Cp2Zr=CR1R2, Cp2Zr=NR, Cp2Zr=O, etc.
a
Cp2Zr 16e
III
- IV
-
Cp2Zr
a
Cp2Zr
-
CR, Cp2Zr
N
C
C
-
O,+ etc.
Cp2Zr
18e IV
IV
Cp2Zr 16e IV
V
Cp2Zr 16e IV
VI
a
Cp2Zr
b
a
Cp2Zr
b
C C
, Cp2Zr
C a
a
Cp2Zr
b
Cp2Zr
IV a Cp2Zr 16e d
a
b c
Cp2Zr
C N
- a
d
c b c
,
Cp2Zr
,
Cp2Zr
N
b , Cp2Zr c
16e c
VII
, Cp2Zr
a , Cp2Zr d
C
, etc.
O C O
- a
b , Cp2Zr
+ , etc.
b , etc.
c b c
a
b
d
c
, Cp2Zr
, etc.
IV
VIII
Cp2Zr
( six-membered and larger zirconacycles )
16e
IX
IV + Cp2Zr a
+
+
Cp2Zr C , Cp2Zr N
14e
X
- IV
Cp2Zr 18e
a b c
+
, Cp2Zr O
( acyclic or cyclic zirconates )
, etc.
5.1
IV
Cp2Zr
C
II
Cp2Zr
C IV-I
Introduction
245
C C XV-I
Scheme 5-2 It is striking that, in virtually all transformations discussed in this chapter, the Cp2Zr unit remains intact. It is, however, important to note that the Cp2Zr unit is, in fact, more fragile and chemically labile than might be thought. For example, treatment of Cp2ZrCl2 with 2 equivalents of n-BuLi is known to give Cp2Zr(Bu-n)2 which serves as a convenient precursor to a ‘ZrCp2’ equivalent [56, 57]. On the other hand, the use of 3 or more equivalents of n-BuLi produces a mixture of CpZr(Bu-n)3 and LiCp which is in equilibrium with a minor amount of a mixture of Cp2Zr(Bu-n)2 and LiBu-n [63] (Scheme 5-3). As might be expected, the facile displacement of a Cp group does not occur with n-butylmagnesium halides under comparable conditions. Conversion of Cp2Zr derivatives into CpZr derivatives can provide useful and convenient routes to the latter, since their direct formation is often problematical due to competitive formation of Cp2Zr derivatives. A representative example of such reactions is treatment of Cp2ZrCl2 with Cl2 to give CpZrCl3 [64] (Scheme 5-4).
246
5
Zirconocenes
Scheme 5-3 Cp2ZrCl2
Cl2, hν
CpZrCl3
Scheme 5-4 In cases where the desirable transformations in high yields are observed with conservation of the Cp2Zr unit, they obviously take place faster than any undesirable side reactions. On the other hand, it has become abundantly clear that, in cases where the desired reactions are comparatively slow, various unwanted side reactions, reversible or irreversible, can divert the reaction courses, and our knowledge of such side reactions becomes critically important.
5.2 Preparation of Cp2Zr Compounds As in the synthesis of any other organometals, there are a dozen or so methods [65] to be considered for the synthesis of Cp2Zr compounds [6]. Several of them listed in Table 5-3 are particularly important. Transmetallation and hydrozirconation convert Cp2Zr(IV) compounds into other Cp2Zr(IV) compounds. On the other hand, oxidative addition and oxidative coupling convert Cp2Zr(II) compounds into Cp2Zr(IV) compounds. If the latter reaction is considered as complexation, the products may be viewed as Cp2Zr(II) compounds. Similar Cp2Zr(II)-to-Cp2Zr(II) conversions may be achieved by complexation with n-donors, e.g., PMe3 and CO. ((Table 5-3))
5.2.1 Transmetallation Transmetallation and hydrozirconation are the two most general methods for the synthesis of Cp2Zr(IV) compounds, and they are often complementary with each other. For the preparation of Cp2Zr(IV) compounds by transmetallation, the most frequently used starting materials are Cp2ZrX2 and Cp2Zr(X)R, where X = Cl or Br, and the latter compounds are most frequently prepared by hydrozirconation with Cp2Zr(H)Cl and by transmetallation with Cp2ZrX2. The starting Cp2ZrX2, where X = Cl, Br, and I, can be prepared by the treatment of the respective ZrX4 with 2 equivalents of CpM, where M = MgCl, MgBr, Li, Na, etc. [1, 42]. Currently, Cp2ZrCl2 is commercially available at a reasonable price. Since ZrBr4 and ZrI4 are very much more expensive than ZrCl4, it may be more economical to prepare
5.1
Table 5-3. Methods of Preparation of Cp2Zr Compounds 1. Transmetallation:
Cp2Zr(IV) MR
Cp2ZrX2
2. Zr-H Exchange:
Cp2Zr(X)R
Cp2Zr(IV) HR
Cp2ZrX2
Cp2Zr(IV)
3. Hydrozirconation:
MR
Cp2ZrR2
Cp2Zr(IV) HR
Cp2Zr(X)R
Cp2Zr(IV)
Cp2ZrR2
Cp2Zr(IV)
C C XCp2Zr
C C H
H Cp2Zr X
C C XCp2Zr
4. Oxidative Addition: RX
Cp2Zr(II)
H
Cp2Zr(IV)
IV
Cp2Zr(X)R
II
Cp2ZrLn RH
5. Oxidative Coupling: ( or Complexation:
IV
Cp2Zr(H)R
Cp2Zr(II) Cp2Zr(II)
C C
IV
Cp2Zr
Cp2Zr(IV) Cp2Zr(II) ) C C
II
Cp2Zr _
C C
II
Cp2ZrLn C C
IV
Cp2Zr
C C
6. Complexation with n-Donors: Cp2Zr(II) II
Cp2ZrLn
mL
II
Cp2ZrLn+m
II
Cp2Zr _
C C
Cp2Zr(II)
Introduction
247
248
5
Zirconocenes
Cp2ZrBr2 or Cp2ZrI2 from Cp2ZrCl2 by transhalogenation. Treatment of Cp2ZrCl2 with KI in acetone [66] gives only an 80/20 mixture of the iodide and chloride. The reaction of Cp2ZrCl2 with BI3 gives a purer Cp2ZrI2 [67]. The use of Me3SiI in place of BI3 [68] is a convenient alternative. However, one of the most convenient methods involves treatment of Cp2ZrCl2 with n-BuLi (2 equiv.) at –78°C followed by iodinolysis with I2 [56] (Scheme 5-5). ZrCl4
ZrBr4
ZrI4
2 CpM
2 CpM
2 CpM
Cp2ZrCl2
Cp2ZrBr2
MR1
MR2
R1
HCl
R2
R1ZrCp2Cl
MH HZrCp2Cl
MH
ZrCp2
H2ZrCp2
Cp2ZrI2
Scheme 5-5 Substitution of one or two Cl atoms of Cp2ZrCl2 with carbon groups by transmetallation (Scheme 5-5) is, in principle, favorable in cases where M is more electropositive than Zr, or more specifically ZrCp2Cl or ZrCp2R. Thus, alkali metals, e.g., Li, Na, and K, as well as Group 2, 12, and 13 metals, e.g., Mg, Zn, and Al, participate in this reaction. However, some subtle but critical differences should be noted. For example, organolithiums and organomagnesiums readily dialkylate Cp2ZrCl2 to give Cp2ZrR2 in most cases. In fact, it is often not practical to monoalkylate Cp2ZrCl2 with RLi or RMgX (X = Cl, Br, or I), although some Grignard reagents, e.g., t-BuMgCl, cleanly reacts with Cp2ZrCl2 in a 1:1 manner to give i-BuZrCp2Cl in high yield [69, 70] (Scheme 5-6). One unwanted side reaction with RLi is the formation of CpZrR3 as discussed earlier (Scheme 5-3). This does not appear to be a serious problem with RMgX. Dialkylzirconocenes, or more generally diorganylzirconocenes, are, in most cases, thermally labile, and they undergo synthetically interesting reactions, as discussed later. Selective monoalkylation of Cp2ZrCl2 is still an unsolved problem in many cases. For example, a somewhat tedious and indirect method shown in Scheme 5-6 has remained as one of the standard methods for the preparation of MeZrCp2Cl [71]. Although possible, there does not appear to be a very convenient method for the preparation of ArZrCp2Cl. In such cases, selective C–Cl exchange reaction of Cp2ZrR2 and/or thermal comproportionation between Cp2ZrR2 and Cp2ZrCl2 might prove to be useful. In most of those cases where hydrozirconation is applicable, RZrCp2Cl in which R is alkyl or alkenyl are best prepared by hydrozirconation. The reaction of organoalanes with Cp2ZrCl2 is a delicate equilibrium reaction. For example, treatment of Cp2ZrCl2 with one equivalent of Me3Al proceeds partially to give MeZrCp2Cl and Me2AlCl, and there has been no indication for the formation of Me2ZrCp2 under a variety of conditions [50, 72] (Scheme 5-6). On the other
249
5.2 Preparation of Cp2Zr Compounds
hand, it is feasible to transfer an alkenyl group from Zr to Al [50, 73, 74] (Scheme 5-6). These results indicate that transmetallation between Zr and Al is very much ligand-dependent in a more or less predictable manner. Thus, the alkylating ability of organoaluminums increases in the order: RAlCl2 < R2AlCl < R3Al < R4Al–. For example, transfer of alkenyl groups from Zr to Al discussed above can be reversed by the use of alkenylaluminates [74] (Scheme 5-6). Good understanding of the Zr– Al equilibria is particularly important in dealing with addition and polymerization reactions of alkenes and alkynes catalyzed by Zr–Al reagents.
t-BuMgCl
Cp2ZrCl2
0.5 H2O PhNH2
Cp2ZrCl2
ZrCp2Cl
2 Me3Al
O(Cp2ZrCl)2
Cp2ZrCl2 + Me3Al
R1
R1
R1
H ZrCp2Cl
H + Cl2ZrCp2
X = Cl or alkyl
(H)R2 R1
H
AlX2 H
+ Cl2ZrCp2 (H)R2
MeZrCp2Cl
MeZrCp2Cl + Me2AlCl
+ ClAlX2 (H)R2
ZrCp2Cl
AlR3Li
+ AlR3 + LiCl (H)R2
ZrCp2Cl
Scheme 5-6
5.2.2 Hydrozirconation The preparation of HZrCp2Cl [45] and its reaction with alkenes and alkynes to give the corresponding hydrozirconation products [46] (Table 5-3) were reported first by Wailes et al. However, it was Schwartz [47] who demonstrated its usefulness in organic synthesis by converting organozirconium derivatives into organic products. The reaction shares some synthetically important characteristics with other facile and clean hydrometallation reactions, such as hydroboration [75] and hydroalumination [76, 77]. Thus, the reaction is generally clean and high-yielding, involving essentially 100% cis addition. However, a few noteworthy differences
250
5
Zirconocenes
among them should be noted. The scope of hydrozirconation with respect to both substrate structure and chemoselectivity lies somewhere between those of hydroboration and hydroalumination (Table 5-4 and Scheme 5-7). The facile isomerization of secondary and tertiary alkylzirconium derivatives into primary alkyl derivatives can occur even at or below room temperature [47], and it is far more facile than that of B or Al. Alkenylzirconiums can also isomerize in the presence of a catalytic amount of HZrCp2Cl, and a regioselectivity of ≥90% has been attained without loss of stereoisomeric purity in the reaction of MeC≡CR, where R is primary, secondary, or tertiary alkyl [78]. It is also noteworthy that conjugated dienes can be converted to homoallylzirconium derivatives in which Zr is bonded to the less hindered of the two terminal carbon atoms of the diene unit [79]. Halogens, ethers, and metal-substituted alcohols are tolerated. Carbonyl compounds, such as aldehydes, ketones, carboxylic acids, and esters are reduced to alcohols, while nitriles are reduced to aldehydes [8]. However, the relative reactivity of various functional groups can be significantly modified, as shown in Scheme 5-8.
Table 5-4. Scope of Hydrometallation Reactions with Respect to Substrate Structures
a
Metal
Substrate structure
B
Zr
Al
RC/CH
+
+
+
RC/CR
+
+
+
RCH=CH2
+
+
+a
R2C=CH2
+
+
sluggish
RCH=CHR
+
+a,b
sluggish
R2C= CHR
+
+b
?
R2C=CR2
+
-
?
Sluggish unless catalyzed.
b
Accompanied by Zr migration.
Despite the ease with which alkylzirconium derivatives regiochemically isomerize, alkyl groups, such as Et and n-alkyl, cannot be readily displaced by an external alkene, suggesting that the regioisomerization process must be nondissociative. It has, in fact, been difficult to free terminal alkenes from alkylzirconiums. The use of the stoichiometric amount of Ph3CBF4 was necessary in one successful case [8]. A wide variety of solvents may be used for hydrozirconation. However, ethers, such as THF, morpholine, and especially oxetane, appear to be some of the most
251
5.2 Preparation of Cp2Zr Compounds
HZrCp2Cl ZrCp2Cl
HZrCp2Cl ZrCp2Cl HZrCp2Cl ZrCp2Cl HZrCp2Cl ZrCp2Cl HZrCp2Cl
R
H
H
ZrCp2Cl
R
Me
> 98% terminal > 98% E
RC CH HZrCp2Cl RC CMe
Me
R
+ H
ZrCp2Cl H ZrCp2Cl II I I / II : R = Et (55/45), n-Pr (69/31), i-Pr (84/16), t-Bu (>98/2) HZrCp2Cl ZrCp2Cl
Scheme 5-7 favorable solvents [10], although less polar solvents, such as benzene and toluene, have also been widely used. The preparation of HZrCp2Cl free of any other species has been accomplished by the treatment of Cp2ZrCl2 with various metal hydrides, such as LiAlH4 [45], LiAlH(OBu-t)3 [45, 82], and Red-Al, i.e., NaAlH2(OCH2CH2OCH3)2 [83], fol-
252
5
Zirconocenes
HZrCp2Cl THF, 21oC
TIPSO
TIPSO
ZrCp2Cl
[80]
O
O
1. HZrCp2Cl
H
2. Li2CuMe2(CN)
TIPSO
TIPSO
Cu(CN)MeLi2
[81]
O
O
Et
1. O
H HO
Et O
2. n-Bu4NF
H
H
O
Scheme 5-8 lowed by purification. The low solubility of HZrCp2Cl in common organic solvents permits a simple, if rather tedious, method of its purification by filtration under an inert atmosphere. One common problem has been competitive formation of Cp2ZrH2, which can cause various unwanted side reactions, e.g., those shown in Scheme 5-9. This problem can be largely avoided by treating Cp2ZrH2 with CH2Cl2 to regenerate HZrCp2Cl [84, 85]. Ph Ph
PhC CPh Cp2Zr
Cp2ZrH2
Ph
[86] Ph
(Me5C5)2ZrH2
MeC CMe - 78oC [87]
Me
Me
25oC Me
(C5Me5)2Zr
(C5Me5)2Zr H
H
Scheme 5-9 In cases where there are no complications, HZrCp2Cl equivalents can be in situ generated by the treatment of Cp2ZrCl2 with t-BuMgCl, LiAlH4, Red-Al, and LiBEt3H [69, 70, 88]. Another useful variation of hydrozirconation is a Zr-cata-
5.2 Preparation of Cp2Zr Compounds
253
lyzed hydroalumination with i-Bu3Al and a catalytic amount of Cp2ZrCl2 [89] (Scheme 5-10). While these variations can provide some convenient alternatives to the use of preformed and purified HZrCp2Cl, they must not be considered to be the same as pure HZrCp2Cl. For example, whereas hydrozirconation of 5-decyne with pure HZrCp2Cl gives essentially one isomer, that with i-Bu3Al and a catalytic amount of Cp2ZrCl2 produces two regioisomers in an essentially 1:1 ratio [90] (Scheme 5-10). These and other recent results indicate that the Zr-catalyzed hydroalumination reaction clearly involves hydrozirconation but that it must be promoted by Al, which clearly is also responsible for the regioisomerization only to an adjacent position. This interesting observation can be readily understood by assuming the intermediacy of an allene derivative and an exclusive placement of Zr to its central carbon atom. These aspects of hydrozirconation are becoming clear only within the last few years, and further clarifications are undoubtedly forthcoming. Bu-n
n-Bu
HZrCp2Cl n-BuC CBu-n
H
ZrCp2Cl
i-Bu3Al cat. Cp2ZrCl2
n-Bu
n-Bu C C CHPr-n
H
MLn H
H
Pr-n
Pr-n C C C HMLn
n-BuHC C C H
H MLn H
MLn = ZrCp2Cl.AlR2Cl or AlR2
Scheme 5-10
The lack of reliable information on the structure of HZrCp2Cl including its state of aggregation has hampered the clarification of the mechanism of hydrozirconation. However, its ab initio MO study [91] suggests a concerted four-center mechanism involving a frontal coplanar attack between H and Cl of HZrCp2Cl by alkenes and alkynes, which is very analogous to those suggested for other facile cis hydrometallation reactions, such as hydroboration [75] (Scheme 5-11). An activation energy of 0–15 kcal mol–1 has been suggested for the reaction of ethylene. In addition to an earlier review by Schwartz [8] several other reviews have discussed hydrozirconation [5, 6, 9, 10].
254
5
Zirconocenes
Scheme 5-11 5.2.3 Oxidative Addition Oxidative addition is a potentially important process for the preparation of organozirconium derivatives. One prototypical example of this reaction was accidentally discovered in an unsuccessful attempt to bicyclize diallyl ether with nBu2ZrCp2. The reaction instead produced an oxidative addition product which must have been formed via oxidative coupling–elimination [92] (Scheme 5-12). This reaction has since been developed into a convenient method for converting allyl and propargyl ethers into allylzirconiums [93, 94] and allenylzirconiums [95], respectively (Scheme 5-12). A review of this reaction with pertinent references has recently been published [11].
Scheme 5-12
5.2 Preparation of Cp2Zr Compounds
255
The scope of the oxidative addition reaction of ‘Cp2Zr’ derivatives has been significantly expanded by the recent development of oxidative addition of alkenyl chlorides to give alkenylzirconocene chlorides, which also appears to proceed via oxidative coupling–elimination [96]. Its ability to afford alkenylzirconocene derivatives corresponding to Markovnikov addition to alkynes is particularly noteworthy (Scheme 5-13). n-Bu2ZrCp2
R
R ZrCp2Cl
Cl
Scheme 5-13
It appears promising to explore the oxidative addition reactions of aryl and benzyl derivatives. As yet, no satisfactory procedures involving these derivatives appear to have been developed.
5.2.4 Oxidative Coupling and Complexation Bonding in π-complexes of transition metal species is best described in terms of the Dewar–Chatt–Duncanson model [97, 98] involving π-donation by a π-ligand and backdonation by a metal. According to the theory, a metal-containing species must have at least one filled nonbonding orbital, i.e., d2, d4,... but not d0, to afford stable π-complexes. However, even d0 complexes may serve as Lewis acids to accept an electron pair from a π-ligand to form unstable transient species. If a pair of σ-electrons are available for interaction (or backdonation), stable species may be formed. This indeed amounts to the well-accepted mechanism for concerted addition reactions, such as hydrozirconation. Since Cp2ZrIV derivatives are d0 complexes, they may interact with π-compounds to produce only transient π-complexes that can be further converted to stable addition products (Scheme 5-14). On the other hand, Cp2ZrII derivatives are d2 complexes which may contain one or two nligands or one π-ligand (Scheme 5-15). The parent Cp2Zr is a 14-electron species which can, in principle, exist as either a singlet (Ia) or a triplet (Ib), but it does not appear to have been detected and characterized as such. All four types of 14- and 16-electron Cp2ZrII species shown in Scheme 5-15 are, in principle, capable of reacting with π-donors to produce π-complexes, but the following comments are in order. Complexes represented by Cp2ZrIIL2, where L is an n-donor, are 18-electron species with no low-lying empty orbital. So, their π-complexation must proceed with loss of one or both n-donors (L). More intricate is the reaction of π-complexes containing a Cp2ZrII moiety. Since they are 16-electron species with no n-electron pair available for interaction, their reaction with a π-donor must involve either
256
5
Zirconocenes
π-ligand displacement or addition to produce five-membered zirconacycles (Scheme 5-16). Furthermore, the latter process can serve as a mechanistic path for π-displacement, as detailed later. For the sake of simplicity, alkene–ZrCp2 complexes are shown in Scheme 5-16, but similar processes involving other π-ligands have also been observed. X X
IV
C C
Cp2Zr
IV
Cp2ZrX2
IV
X
Cp2Zr
X C C
C C
Scheme 5-14 II
II
Cp2Zr
X
Cp2Zr
(Ia), 14e
II
X
Cp2Zr
(Ib), 14e
II
Cp2Zr
L
L
(IIa), 16e
II
Cp2Zr
Cp2ZrL2
X
IV
Cp2Zr
X X
16e
X
(IIb), 16e II
X
16e X
II
Cp2Zr
X
L
X
X X
18e
IV X
Cp2Zr
L
X
18e
(III), 18e X
X
π-donors, such as alkenes, alkynes, carbonyl compounds, and nitriles.
Scheme 5-15
II
Cp2Zr
(IVa)
X
X
II
II
Cp2Zr
X
Cp2Zr
X X
X
IV
Cp2Zr
(IVb)
X
X
IV
Cp2Zr
IV
Cp2Zr
X X
X
Scheme 5-16
X
257
5.2 Preparation of Cp2Zr Compounds
The preparation of the required Cp2ZrII reagents has been achieved by either reduction of Cp2ZrX2, where X is Cl, Br, I, etc., with metals [99] or β-H abstraction of organozirconocene derivatives. An early example of the reduction of Cp2ZrCl2 and Cp2HfCl2 with Na and naphthalene produced some reduced products represented by (C10H10M)n, where M is Zr or Hf. Formation of Cp2Zr as a transient species may occur but has never been experimentally demonstrated. In the presence of PhC≡CPh, the reaction produced the corresponding zirconacyclopentadiene [100] (Scheme 5-17). In the presence of n-donors such as CO and PMe3, Cp2ZrCl2 has been converted to Cp2Zr(CO)2 [101] and Cp2Zr(PMe3)2 [102] respectively (Scheme 5-18), and their structures have been fully characterized by spectroscopic methods including X-ray analysis [103, 104]. Cp2MCl2 +
2 Na / naphthalene
THF
1/n (C10H10M)n
+
2 NaCl
M = Zr, Hf Ph
Ph "Cp2M" (?)
PhC CPh
Cp2M
C
Ph
PhC CPh Cp2M
C Ph
Ph Ph
Scheme 5-17
2 CO
Cp2ZrCl2
Mg, HgCl2 THF 2 PMe3 Na - Hg THF
Cp2Zr(CO)2
Cp2Zr(PMe3)2
Scheme 5-18
These Cp2ZrII complexes tend to be thermally and chemical labile. For example, Cp2Zr(PMe3)2 can be converted to a dimeric ZrIII species [104] (Scheme 5-19). For further information of these complexes, pertinent reviews [3, 27, 36, 105] should be consulted.
258
5
Zirconocenes
2 Cp2Zr(PMe3)2
Zr
Zr
PMe3
Me3P
Scheme 5-19 From the viewpoint of organic synthesis, the β-H abstraction reaction of monoand diorganylzirconocenes represents the most convenient route to Cp2ZrII derivatives at present. The formation of Cp2ZrII complexes containing benzynes [51, 58, 106], alkynes [107–109], alkenes [107, 110, 111], as well as various heteroalkynes [112] and heteroalkenes [113, 114] have been generated by this method. There are two convenient protocols. In one, dialkylzirconocenes, such as n-Bu2ZrCp2 [56, 57, 107] and t-Bu(i-Bu)ZrCp2 [115], are in situ generated and decomposed in the presence of π-compounds, such as alkenes and alkynes (Negishi protocol). In this process, alkyl groups serve as precursors to temporary π-ligands which are to be displaced by externally introduced π-compounds [35] (Scheme 5-20). In the other, the
Scheme 5-20
5.2 Preparation of Cp2Zr Compounds
259
final π-ligands are in situ generated by the decomposition of diorganylzirconocenes (Erker–Buchwald protocol). No external π-compound is used for the preparation of Cp2ZrII π-complexes. In the original Erker procedure [51], one of the two organic groups was sacrificed. This practical difficulty can be avoided by the Buchwald procedure in which an inexpensive Me group serves as the β-H abstracter [29, 58, 108] (Scheme 5-21).
PMe3 PhH, reflux, 24h Cp2Zr
Cp2Zr [51, 58]
R1 Cl Cp2Zr
Li
R2 H
Me
[59]
Cl
1. R1C CR2 2. MeLi
H
[108]
Cp2Zr
PMe3
R1 Cp2Zr
R2 H
Me
R1 PMe3
Cp2Zr Me3P
C C R2
Scheme 5-21 In general, unstabilized 16-electron Cp2ZrII complexes with π-donors are thermally and chemically labile, although some, such as Cp2Zr–stilbene [107], have been spectroscopically characterized. However, they can be stabilized by n-donors, such as PMe3, CO, and amines. Alternatively, Cp2ZrII π-complexes containing heteroalkynes and heteroalkenes can stabilize themselves through dimerization, oligomerization, and polymerization. Many of the stabilized complexes have been fully characterized by spectroscopic methods including X-ray crystallography. Some representative examples of Cp2ZrII π-complexes are shown in Table 5-5.
260
5
Zirconocenes
Table 5-5. Representative Examples of Cp2ZrII π-Complexes Characterized by X-Ray Analysis Complex
Cp2Zr _
Complex
Reference
Cp2Zr _
[58]
Me3P Ph
Me3P
[59]
Me3P
C Cp2Zr _ C Me3P Ph
Cp2Zr _
Reference
CH2 CH2
Cp2Zr _
[107]
Ph H C Cp2Zr _ C Me3P H Ph
[110]
[111, 116, 117]
H Ph C Cp2Zr _ C Me3P H H
[111]
Ph H C Cp2Zr _ N SiMe3 Me3P
[114]
[29]
Me3P
Ph Cp2Zr _ Me3P
C P
N Cp2Zr _ N Me3P Ph
[112]
[118]
It should be pointed out here that, in some cases, transmetallation reactions of Cp2ZrIV derivatives provide convenient alternatives to the preparation of Cp2ZrII πcomplexes, as indicated by the examples shown in Scheme 5-22. Yet another alternative for the preparation of Cp2Zr π-complexes or carbonyl compounds is migratory insertion reactions of acylzirconium derivatives [118–122] (Scheme 5-23). Mg Cp2ZrCl2
n
Cp2Zr
Cp2Zr
[53] PhN NPh
Cp2ZrCl2
Li Li pyridine [118]
Cp2Zr Py
Scheme 5-22
NPh NPh
5.2 Preparation of Cp2Zr Compounds
Ph Ph
CO
Cp2Zr
Cp2ZrPh2
O
Ph
Cp2Zr
C Ph
261
O O
ZrCp2
91% Ph
[124]
Ph
Scheme 5-23
5.3 Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives Monoorganylzirconocene derivatives represented by RZrCp2Cl, where R is alkyl, allyl, allenyl (or propargyl), and alkenyl, are now widely available via transmetallation, hydrozirconation, or oxidative addition. Those containing a benzyl or aryl group are currently less readily available, and their chemistry as well as that of RC≡CZrCp2Cl remains largely unexplored. The majority of known monoorganylzirconocene chlorides are thermally stable at ambient temperatures. However, they are sensitive to both air and moisture, and they must be handled under an inert atmosphere of Ar or nitrogen. The structure of RZrCp2Cl is distorted tetrahedral. For example, the Cp–Zr–Cp and Me–Zr–Cl angles of MeZrCp2Cl are 134 and 95°, respectively, and the Me–Zr and Cl–Zr bond lengths are 2.36 and 2.49°, respectively [125] (Scheme 5-24). It is important to note that an empty valence shell orbital which can act as an acidic or electrophilic center may be available in any of the three regions indicated by the orbital lobes lying in the plane defined by C, Zr, and Cl (Scheme 5-24). As will be clear from the discussions that follow, this empty orbital plays a pivotal role in the reactions of Cp2Zr derivatives including RZrCp2Cl. Since the chemistry of RZrCp2Cl has been extensively and repeatedly reviewed [5, 6, 8, 10], only a brief overview is presented below. 2.49 A Cl Cl 134o
95o
Zr
Zr CH3
CH3 2.36 A Side view
Scheme 5-24
Top view
262
5
Zirconocenes
5.3.1 Protonolysis, Halogenolysis, Oxidation, and Related Carbon–Heteroatom Bond Formation Reactions 5.3.1.1 Protonolysis and Deuterolysis Cleavage of σ Zr–C bonds occurs readily upon treatment with H2O or dilute acids, while the Cp–Zr bond usually survives mild protonolysis conditions. The use of D2O or more reliably DCl–D2O permits replacement of Zr with D. Together with iodinolysis discussed below, deuterolysis provides a generally reliable method for establishing the presence of the C–Zr bond. Protonolysis or deuterolysis of Zr–Csp2 bonds proceeds with retention of configuration [126], while the stereochemistry of the corresponding reaction of Zr–Csp3 bonds does not appear to have been well established. In contrast with protonolysis, catalytic hydrogenation of RZrCp2Cl is sluggish and not very clean [127]. (D)HZrCp2Cl
R
H(D)
H+ (or D+)
R
H(D)
(D)H
H(D)
RC CH(D) (D)H
ZrCp2Cl
Scheme 5-25
5.3.1.2 Halogenolysis Iodinolysis of RZrCp2Cl with I2 in THF occurs readily and widely at or below room temperature. And yet, it generally preserves the Cp–Zr bond [47, 78]. Brominolysis with Br2 is also facile, but it is often advantageous to use NBS in place of Br2. Chlorinolysis is more sluggish, but favorable results may be obtained with NCS [78] or PhICl2 [47]. Iodinolysis and brominolysis of both Zr–Csp2 and Zr– Csp3 bonds have proceeded with retention of configuration (Scheme 5-26). Brominolysis of some homoallyl derivatives suffer from skeletal rearrangement [79].
5.3
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
I2, CCl4 RZrCp2Cl
RI
Br2
RBr
PhICl2 I2
R
H
H
ZrCp2Cl
RCl R
H
H
I
R
H
H
Br
R
H
H
Cl
NBS
NCS
t-Bu H
D
D H
Br2
D H
t-Bu H
ZrCp2Cl
263
D
Br
NBS ZrCp2Cl
Br 85% NBS
ZrCp2Cl
Br 49%
37%
NBS ZrCp2Cl
Br
Scheme 5-26
Br
+
264
5
Zirconocenes
5.3.1.3 Oxidation Alkylzirconocene derivatives can be converted to alcohols with H2O2–NaOH, tBuOOH, m-chloroperbenzoic acid, or O2 [128]. On the other hand, conversion of alkenylzirconocenes to aldehydes and ketones cannot be readily achieved [47]. While oxidation reactions in some cases may proceed stereospecifically, those involving radicals are nonspecific. Clearly, additional investigations are necessary for oxidation of RZrCp2Cl to be synthetically useful. Other heteroatom reagents that react with RZrCp2Cl include SO2 and NO [129, 130].
5.3.2 Carbon–Carbon Bond Formation via Polar Reactions of RZrCp2Cl with Carbon Electrophiles 5.3.2.1 Reactions of Allylzirconocenes with Aldehydes Alkyl- and alkenylzirconiums are generally of low nucleophilicity towards common organic electrophiles, such as those shown in Scheme 5-27. Of the six types of electrophiles shown, only acyl halides are sufficiently reactive towards alkyland alkenylzirconocenes, but the reaction is not general and reliable. An improved procedure involving Zr-to-Al transmetallation discussed later is a superior alternative. The only known exception is the reaction of allylzirconocenes with aldehydes [131, 132] (Scheme 5-28). Dienezirconocenes, which may also be viewed as zirconacyclopent-3-enes, readily react with aldehydes, ketones, esters, and even nitriles as well as nonpolar π-compounds, such as alkenes and alkynes [133, 134] (Scheme 5-29 and 30). Some of these reactions hold considerable promise as practically useful synthetic tools, as exemplified by those shown in Scheme 5-31 [11].
R1
R 1X
R1
R1COR2
R
R
OH R2
R1COX
R1
COR
R1CN
R1
O
RZrCp2Cl
R
OH
C C C O R
COR
Scheme 5-27
C CH C O
5.3
265
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
Cp2Zr
H
2
H
RCHO
O
Cp2Zr
Cp2ZrCl2
X ClCp2Zr
R
R
X=
H
HO
H or Cl
Scheme 5-28
1. R3CN R1
O
R1R2CO
OR
O
R2
OH
+ R1
OH R2
1
O
RCOOR1
R1
H+
ZrCp2 R
ZrCp2
2. H
R2
R
H+
ZrCp2
R O
RCN
R H+
N
ZrCp2
R
ZrCp2
O 1.
R1
R1COR2 2. H
+
R
HO O
R2
Scheme 5-29
RCH=CH2
R
1. R1COR2
ZrCp2
R1 OH
2. H+
R2
R
ZrCp2 RC CR
ZrCp2
R
2. H+ R
Scheme 5-30
R
1. R1CN R
R1 O
266
5
Zirconocenes 1. DIBAL-H 2.
MgBr
3. H3O+
H
H
N-Boc L-proline methyl ester
OMe
4. Br
N
ZrCp2
n-Bu2ZrCp2
O
N
OMe
O OMe
OMe 5. TsOH
OH H
H
H
ZrCp2OMe
BF3.OEt2 N
H O
N+ _ BF3
OH
1. O3 N
2. NaBH4
OH
(-)-macronecine
Scheme 5-31
5.3.2.2 Reactions of Cationic Alkyl- and Alkenylzirconocenes with Aldehydes, Epoxides, Ketones, and Nitriles Treatment of organylzirconocene chlorides with a halogen scavenger, such as AgClO4, has been shown to significantly boost their reactivity towards aldehydes [135] (Scheme 5-32). The use of AgAsF6 [136] in place of AgClO4 is not only safer but also more effective in some cases.
n-Bu
ZrCp2 Cl
n-Bu
PhCHO cat. AgClO4 84%
Ph
n-Bu
92%
ZrCp2 Cl
PhCHO cat. AgClO4
OH Ph
n-Bu OH
Scheme 5-32
The same strategy is also effective in the reaction with epoxides [137] (Scheme 533). This reaction has been shown to involve Zr-catalyzed isomerization of epoxides to aldehydes.
5.3
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
R1 O cat. AgClO4
RZrCp2Cl
R
267
R1 OH
R1
O
+
H
R1
R1
O+
ZrCp2
OZrCp2
R RZrCp2Cl
AgClO4 R
R
+
RZrCp2 R
R1 OZrCp2Cl
H
R1 O
+
ZrCp2
R1 O
RZrCp2Cl
+
ZrCp2
R
Scheme 5-33
Cationic monoalkylzirconocenes, such as MeZr+Cp2(THF)–BPh4, react even with ketones and nitriles in similar manners [16, 138] (Scheme 5-34).
Ph2CO +
Ph2MeCO
+
-
+
-
ZrCp2BPh4 O
-
MeZrCp2BPh4 O MeC N
Me2C N
ZrCp2BPh4
O
Scheme 5-34
5.3.3 Carbon–Carbon Bond Formation via Migratory Insertion Reactions of RZrCp2Cl Organometals of low intrinsic nucleophilicity, such as those containing B and transition metals, may resort to other patterns of processes for carbon–carbon bond formation than those shown in Scheme 5-27. Three such processes available to a wide
268
5
Zirconocenes
variety of organotransition metals are shown in Scheme 5-35. It should be pointed out that reductive elimination, in fact, provides a means of achieving cross coupling, acylation, and conjugate addition shown in Scheme 5-27, as discussed later. Reductive Elimination R1 R2
R2
R1
MLn
+
MLn
Migratory Insertion R1
R1 MLn R
R MLn
X
X
Carbometallation
C C R
C C MLn
R
C C MLn
R MLn C C
Scheme 5-35
5.3.3.1 Carbonylation The reaction of RZrCp2Cl with CO, typically 1–3 atm, readily produces the corresponding RCOZrCp2Cl which can be converted to aldehydes and carboxylic acid derivatives [126, 139] (Scheme 5-36). dil. HCl RCHO 5.8 ppm CO RZrCp2Cl
H2O2 NaOH
room temp.
R
C
ZrCp2Cl
O 1550 cm-1
Scheme 5-36
NBS Br2 MeOH
RCOOH
RCOBr RCOOMe
5.3
269
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
Cyclic diorganylzirconocenes [51, 140] and certain dialkylzirconocenes, e.g., Me2ZrCp2 [141], also readily undergo carbonylation to give ketones. In some cases, the organozirconium products before protonolysis have been shown to be zirconium enolates [17, 140].
5.3.3.2 Isonitrile Insertion Isonitriles are useful alternatives to CO for some one-carbon homologation reactions where CO is not very satisfactory. The preparation of α,β-unsaturated aldehydes shown in Scheme 5-37 is one such example [142]. Another interesting reaction is the cyanation of alkenes shown in Scheme 5-38 [143]. HZrCp2Cl
1. BuNC 2. AcOH
R
H
H
ZrCp2Cl
RC CH
R
H
H
CHO
Scheme 5-37 NSiMe3
1. HZrCp2Cl 2. Me3SiCN
R
R
I2
C N
R
ZrCp2 Cl
Scheme 5-38 5.3.3.3 Other Migratory Insertion Reactions Although the potential scope of the migratory insertion reactions of organozirconium compounds appears to be very broad, its current scope is largely limited to carbonylation and related reactions with isonitriles. Nonetheless, there are some scattered reports on other migratory insertion reactions as shown in Scheme 5-39 [144] and Scheme 5-40 [145]. This topic clearly deserves to be further explored. Li
SiMe3 Cl
SMe
Cp2Zr
Cp2ZrCl2
Ph SiMe3
PhMgBr
SMe Ph Cp2Zr
Ph
SiMe3 SMe
Scheme 5-39
SiMe3 SMe
H+
95oC
Cp2Zr
SiMe3
270
5
Zirconocenes
1. LiCH(Cl)SiMe2Ph n-OctCH2SiMe2Ph 2. HCl Li 1.
n-Oct
Br 2. HCl n-OctZrCp2Cl
H
Li
Me
Br
Ph
1. n-Oct
Me
H
Ph
2. HCl
1. LiC CCH2Cl
n-OctCH C CH2
2. HCl
n-Hex H
H ZrCp2Cl
1. LiC CCH2Cl 2. HCl
n-Hex H
H C CMe
Scheme 5-40
5.3.4 Stoichiometric Carbozirconation of Monoorganylzirconocenes Carbozirconation, i.e., addition of the C–Zr bond to alkenes and alkynes, has been implicated in a number of Zr-catalyzed carbometallation reactions, most notably the Ziegler–Natta olefin polymerization [44] and the Zr-catalyzed carboalumination of alkynes [49, 50] as well as stoichiometric and catalytic carbometallation reactions involving cyclic diorganylzirconocenes, such as those shown in Scheme 5-30. And yet, clear-cut examples of the stoichiometric carbozirconation of monoorganylzirconocenes are rather rare. The reaction of n-PrC≡CAlMe2 with MeZrCp2Cl may represent one of the earliest examples [146] (Scheme 5-41). Another significant example is the methylzirconation with a cationic reagent, MeZr+Cp2(THF)– BPh4 [138] (Scheme 5-42). In both of these examples, cationic methylzirconocene species may be implicated as reactive carbometallating agents. Without such acti-
5.3
271
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
vation, monoorganylzirconocenes other than allyl and related derivatives may not be capable of undergoing facile carbometallation. Either catalysis with metal compounds and other Lewis acids or stoichiometric transmetallation would be required, as discussed later. AlMe2
n-Pr
MeZrCp2Cl n-PrC CAlMe2
Cl Me
ZrCp2
Scheme 5-41 Me +
-
MeZrCp2BPh4
Me C C
MeC CMe
+
-
ZrCp2BPh4
Me O
O
Scheme 5-42
5.3.5 Reactions of Monoorganylzirconocenes via Transmetallation One generally applicable and potentially powerful notion for expanding the scope of any organometallic chemistry is to consider various transmetallation processes. In Sect. 5.2.1, conversion of various organometals into organozirconiums is discussed. In this section, the reverse transmetallation processes, i.e., Zr-to-other metals, are discussed. As discussed earlier (Sect. 5.2.1) in the opposite context, conversion of RZrCp2Cl into RMLn via transmetallation with XMLn, where X is Cl, Br, or an O group, should be favored in cases where Zr, or more specifically ZrCp2Cl, is more electropositive than MLn. In view of the electronegativity index of 1.2–1.4 for Zr [65], transfer of a carbon group from Zr to a variety of metals including B (2.0), Si (1.8), Sn (1.7–1.8), Hg (1.4–1.9), Cu (1.75–1.9), Ni (1.75–1.8), and Pd (1.35–2.2), whose electronegativity indexes are shown in parenthesis, is expected to be favorable. In dealing with transmetallation, however, the following points should be kept in mind. First, the direction of transmetallation may depend on both the number and nature of ligands in cases where the electronegativity indexes of the two participating metals are similar. Second, the products of transmetallation may in some cases be better represented as ‘ate’ complexes containing both metals. Third, in cases where the organometallic reagents are generated and consumed in situ, the required transmetallation needs not to be thermodynamically favorable so long as the overall transformation is favorable. There are the following three scenarios for the reactions of monoorganylzirconocenes via transmetallation: (1) stoichiometric conversion into other organometals, (2) reactions of mo-
272
5
Zirconocenes
noorganylzirconocenes catalyzed by other metal compounds, and (3) reactions of other organometals catalyzed by zirconocene derivatives which involve transient formation of monoorganylzirconocenes.
5.3.5.1 Stoichiometric Conversion of Monoorganylzirconocenes into Other Organometals Treatment of RZrCp2Cl with metal halides and alkoxides containing Al [73, 89], B [147–151], Hg [152], and Sn [153, 154] has led to the formation of organometals containing these metals as discrete and Zr-free species (Scheme 5-43). Once these compounds are formed, their subsequent reactions have little or nothing to do with Zr. It nonetheless provides a useful way of expanding the synthetic value of RZrCp2Cl. Conversion of alkenylzirconocene chlorides with AlCl3 produces the corresponding alkenyldichloroalanes which react readily with acyl halides [73]. The parent alkenylzirconocene chlorides do not readily react with acyl chlorides to give α,β-unsaturated ketones. Even dialkylaluminum chlorides react readily with alkenylzirconocene chlorides to produce alkenylalanes in high yields [89]. On the other hand, the use of trialkylalanes lead to an equilibrium reaction [72]. With alkenyltrialkylaluminates, clean transfer of the alkenyl group from Al to Zr has been observed [74]. R1
AlCl3 ZrCp2Cl
R1
R2COCl
[73]
R1
ClAlR2 ZrCp2Cl
O R1
AlR2
[89]
+
Cl2ZrCp2
O ClB R1
O
ZrCp2Cl
O
R1
+
B
Cl2ZrCp2
O
[147-150] D
D HgCl2 HgCl
ZrCp2Cl [152]
D
Me3Si
XSnBu3 ZrCp2Cl
R2
R1
AlCl2
[154]
D
Me3Si
90% retention
SnBu3
5.3
273
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
X = Cl (24%), OEt (85%) ClSnBu3
Me3Si
ZrCp2Cl
Me3Si
SnBu3
[153]
Scheme 5-43
In many other cases, the formation of new organometals and their structure have not been fully established. It is even likely that many of them are bimetallic ‘ate’ complexes. Despite such ambiguities, the synthetic usefulness of the stoichiometric transmetallation involving Cu in conjugate addition [155] (Scheme 5-44) and Zn in aldehyde addition [156–158] (Scheme 5-45) is unquestionable. More recently, both of these reactions have been shown to be catalytic in Cu [159, 160] (Scheme 5-46) and Zn [161], respectively (Scheme 5-47). 1. LiI, R
2. CuOTf
O
ZrCp2Cl
R
[155]
O
Scheme 5-44
R
1. Me2Zn, -65oC Ph 2. CHO ZrCp2Cl [156]
Scheme 5-45
Scheme 5-46
R
Ph OH
274
5
Zirconocenes
ZnBr2 (0.2 eq.) PhCHO
Me3Si
ZrCp2Cl
Me3Si
Ph
[161]
OH
Scheme 5-47
5.3.5.2 Reactions of Monoorganylzirconocenes Catalyzed by Metal Complexes In 1977 alkenylzirconocene chlorides were found to react with aryl iodides and bromides in the presence of a catalytic amount of a Ni–phosphine complex [162], which appears to represent the first example of the reactions of organylzirconocenes catalyzed by other metals. Catalysis by Pd–phosphine complexes was shown to be more general and applicable to alkenyl–alkenyl, alkenyl–alkynyl, and alkenyl–allyl coupling as well [163, 164] (Scheme 5-48). It was also found that, in cases where alkenylzirconocene chlorides are sterically demanding, addition of ZnX2 (X = Cl or Br) can significantly accelerate the desired cross coupling. Dou-
R
H
H
ZrCp2Cl
ArX
R
H
cat. NiLn or PdLn
H
Ar
R
H
X
H
H
R1
cat. PdLn
H H H
X
R1 cat. PdLn
H
R
R1
H
X H R cat. PdLn H
Scheme 5-48
R1
5.3
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
275
ble transmetallation, i.e., Zr→Zn→Pd or Ni, was suggested as a plausible path (Scheme 5-49). The applicability of the Ni catalysis was immediately extended to conjugate additions, for which a one-electron transfer mechanism was proposed [165–167] (Scheme 5-50). Recent examples of catalysis by Cu [159, 160, 168] and Zn [161] salts (Schemes5- 46 and 47) have further expanded the scope of the metal-catalyzed reactions of organylzirconocenes.
R2
R1
R3
ZrCp2Cl
ArI
R2
R1
cat. PdLn, ZnCl2
R3
Ar
X R2
R1
R3
cat. PdLn, ZnCl2
Scheme 5-49
R
H
H
+ ZrCp2Cl
cat. Ni(acac)2 i-Bu2AlH
H
1. ClCp2Zr
OZ2
O
MeSO2Cl Et3N
CH2OH C5H11-n
2. CH2O OZ1
3. H2O
68%
CO2Me
1. ClCp2Zr
O
O
C5H11-n
cat. NiLn
OZ
R
C C C O
O
1
H
OZ2
O
cat. NiLn C5H11-n OZ1
CO2Me C5H11-n
2. H2O OZ1
OZ2 89%
OZ2 49%
Z1 = CH2Ph
Z2 = SiMe2Bu-t
Scheme 5-50
276
5
Zirconocenes
It is noteworthy that, despite the intrinsically low nucleophilicity of monoorganylzirconocenes towards carbon nucleophiles, all of the polar carbon–carbon bond formation processes shown in Scheme 5-27 are now practically useful, and stoichiometric or catalytic transmetallation as well as activation through ionization are largely responsible for the versatility of organylzirconocenes as ‘nucleophilic’ reagents.
5.3.6 Organometallic Reactions Catalyzed by Zirconocene Derivatives In the preceding section, the reactions of the stoichiometric quantity of organylzirconocenes in the presence of metal catalysts are discussed. Conversely, various reactions involving the stoichiometric use of metal compounds have been shown to be catalyzed by zirconocenes. In most cases, Cp2ZrCl2 and its derivatives serve as catalysts or catalyst precursors. Most of the widely used main group metals including Al, B, Mg, Si, and Zn have participated in such reactions, and most of these reactions appear to involve the intermediacy of organylzirconocenes. Some of these reactions that are thought to proceed via monoorganylzirconocene halides are discussed in this section.
5.3.6.1 Zirconium-Catalyzed Carboalumination and Carbozincation of Alkynes Aside from the Ziegler–Natta-type alkene polymerization catalyzed by zirconocene derivatives [44], most notably the Kaminsky protocol [48], the methylalumination of alkynes by Me3Al and a catalytic amount of Cp2ZrCl2 [49, 50] (Scheme 5-51) appears to represent the first example of the organic synthetic reactions catalyzed by a zirconocene derivative. The reaction is generally high yielding, essentially 100% syn-stereoselective, and about 95% regioselective for methylalumination of terminal alkynes. Although some related reactions using Me3Al-Cp2TiCl2 [169] and R2Zn-Cp2ZrI2 [170] were also discovered, they are of much limited synthetic utility. Various heterofunctionalities, such as halogens, alcohols, and amines, can be accommodated [171], but in some cases, they lead to side reactions including synthetically useful cyclobutenation [172–176] and anti-carbometallation [177]. Once alkenylalanes are generated, they undergo essentially all of the known reactions of organoalanes and organoalanates, the latter of which can be readily generated by treating the former with alkyllithiums [77, 178]. It suffices to point out here that various carbon–heteroatom bond formation reactions (Scheme 5-52), conventional polar reactions for carbon–carbon bond formation (Scheme 5-53), and the Ni- or Pd-catalyzed cross-coupling reactions (Scheme 5-54) proceed satisfactorily in many cases. Collectively, this new methodology has provided the most selective, predictable, and hence reliable route to trisubstituted alkenes which represent (E)-
5.3
277
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
trisubstituted isoprene units of a wide variety of terpenoids, and many complex natural products of terpenoid origin, such as those shown in Scheme 5-55, have been synthesized using this reaction in some critical steps. Allylalanes and benzylalanes [179] undergo similar carbometallation reactions. Me3Al cat. Cp2ZrCl2
R
H
RC CH Me
AlMe2
Scheme 5-51 R
H
Me
H2O (D2O)
MeOBR'2 [74]
H(D)
R
H
Me
I
R
H
R
I2 [180]
H
Me
HgCl2 [181]
AlMe2
1. n-BuLi 2. Cl2ZrCp2
PhSO2Cl
SO2Ph
Me
[74]
R Me
R Me
R Me
H BR'2
H HgCl
H ZrCp2Cl
Scheme 5-52
H O
R
[182]
Me
R Me
O R1
H R
R1 OH
R1
H
ClCOOEt [182]
R Me
H COOEt
n-BuLi
R Me
H
CO2
R
[182]
Me
H COOH
AlMe2BuLi
O
R
OH Me
AlMe2
O [183]
Me
H
R1
R2
ClCH2OMe
R2 [182]
[182]
Scheme 5-53
R Me
H OMe
278
5
Zirconocenes
Scheme 5-54
O
O
OH O
O
vitamin A [191]
HO H H
OH
OH H
milbemycin β3 [192]
H
OMe MeO
HO
CH3
H H HO
MeO O
n
OMe
H O
coenzyme Qn [194]
brassinolide [193]
Scheme 5-55
H
5.3
279
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
The reaction of Me3Al and Cp2ZrCl2 unquestionably generates an equilibrium quantity of MeZrCp2Cl but not Me2ZrCp2 under the methylalumination conditions [50, 72]. Taken together with the stoichiometric methylzirconation shown in Scheme 5-41, it appeared plausible to propose a methylzirconation mechanism in which a cationic MeZr+Cp2 derivative is the active reagent (Scheme 5-56). And yet, a mixture of Me2AlCl and Cp2ZrCl2, which does not undergo Me–Cl exchange detectable by NMR spectroscopy, is nearly as effective as Me3Al-Cp2ZrCl2. In fact, the use of dialkylchloroalanes (R2AlCl) in place of R3Al is the key to achieving straightforward noncyclic alkylalumination with ethyl- and higher alkyl-containing alanes containing β-hydrogens. For these cases, direct C–Al bond addition promoted by Zr, as opposed to C–Zr bond addition promoted by Al, appears to be more plausible [50, 72] (Scheme 5-56). Despite these ambiguities, however, it appears certain that these reactions proceed via straightforward, concerted carbon-metal bond addition to alkynes involving some activated ‘cationic’ Zr or Al species. These straightforward addition processes should be distinguished from and contrasted with mechanistically more convoluted and complex cyclic carbozirconation processes discussed later. Me Me3Al
+ Cl2ZrCp2
Me2Al
Cl ZrCp2Cl
Me2Al
ZrCp2
Cl RC CH Me
Cl R
+ ZrCp2 Cl
Me
-
Cl AlMe2
RC CH
RC CH + Me Al Cl Me
ZrCp2Me Me
Cl
H + ZrCp2 Cl Cl
R
-
Me
-
AlMe2
H + Cl2ZrCp2 AlMe2
Scheme 5-56
5.3.6.2 Zirconium-Catalyzed Hydroalumination and Hydroboration of Alkenes and Alkynes Although a mixture of i-Bu2AlH and Cp2ZrCl2 does undergo hydrozirconation, the reaction is only stoichiometric in Zr [89]. On the other hand, the i-Bu3Al-Cp2ZrCl2 reagent system undergoes a Zr-catalyzed hydroalumination of alkenes [89] and alkynes [90], as discussed in Sect. 5.2.2. These reactions clearly involve a version of hydrozirconation and Zr→Al transmetallation to produce net hydroalumination products, and the hydrozirconation step is clearly promoted by Al, which appears to be also responsible for undesirable side reactions, such as regioisomerization (Scheme 5-10). Although the mechanism of the critical Al-promoted hydrozircona-
280
5
Zirconocenes
tion step remains to be further clarified, the overall transformation proceeds as shown in Scheme 5-57.
Scheme 5-57
A Zr-catalyzed hydroboration of alkynes reported recently [161] further extends the scope of the Zr-catalyzed hydrometallation (Scheme 5-58). O HB
O
O n-HexC
CH cat. HZrCp2Cl
n-Hex
B O (98% E)
Scheme 5-58
5.3.6.3 Zirconium-Catalyzed Enantioselective Carboalumination of Alkenes Early attempts at methylalumination of 1-alkenes with Me3Al and Cp2ZrCl2 were largely unsuccessful. Reinvestigation of the reaction of 1-octene with Me3Al (1 equiv.) and 8 mol % of Cp2ZrCl2 has revealed that the two major products obtained after protonolysis are 2-(n-hexyl)-1-decene (59%) and 2-methyl-1-octene (18%), the yield of the desired 2-methyloctane being <2% [195]. Evidently, the desired methylmetallation takes place, but the isoalkylzirconocene product in conjunction with an alane readily hydroaluminate alkenes, as exemplified by the reaction shown in Scheme 5-57. In view of these results, it is rather surprising that the reaction of 1-alkenes with 1 equiv. of Me3Al and 8 mol % of dichlorobis(1neomenthylindenyl)zirconium [196] in (CH2Cl)2 has produced in 70–90% yields the desired methyl-aluminated products readily oxidizable with O2 to give the cor-
5.3
Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives
281
responding alcohols. Moreover, the reaction has proceeded in 65–85% ee (Scheme 5-59). Evidently, the use of certain bulky and/or chiral Cp derivatives as ligands must slow down β-dehydrometallation relative to the desired carbometallation. The selective formation of the R isomer observed with a wide range of 1-alkenes and other experimental results are consistent with a mechanism represented by Scheme 5-60, in which a C2-symmetric, cationic methylzirconium derivative containing two 1-neomenthylindenyl groups interacts with the re face of 1-alkenes [195]. This mechanism clearly favors direct addition of the C–Zr bond to alkenes, which is promoted by Al. However, it should be emphasized that seemingly minor variations in certain reaction parameters can lead to drastic mechanistic changes and that any mechanistic extrapolation can lead to false conclusions. 1. Me3Al cat. Cl2ZrCp*2 (1)
R
R *
2. O2
OH
65-85% ee
Cl2ZrCp*2 (1) =
Zr Cl
Cl
Scheme 5-59
CH3 R* H R H3C *R
+ Zr
R
R* =
H CH(CH3)2
Scheme 5-60
282
5
Zirconocenes
The reaction of trialkylalanes containing Et and high n-alkyl groups with 1-alkenes in the presence of Cp2ZrCl2 in nonpolar solvents is known to produce aluminacyclopentanes rather than the straightforward alkylalumination products [197] (Scheme 5-61). In principle, this cyclic carboalumination can proceed enantioselectively. However, the ee figures have been limited to 30–35% [198]. As suspected, a complete mechanistic change was observed, when nonpolar solvents, e.g., hexanes, were replaced by polar solvents, e.g., CH2Cl2, (CH2Cl)2, and CH3CHCl2. Thus, the reaction of 1-decene with Et3Al in the presence of Cp2ZrCl2 in (CH2Cl)2 was a totally acyclic process with no indication of the formation of aluminacyclopentanes, even though competitive dehydrometallation and hydrometallation were significant side reactions. Here again, the use of bulky and chiral 1-neomenthylindene in place of cyclopentadiene not only suppressed dehydrometallation but also led to the formation of the desired alkylmetallation products in 60–90% yields in 90–95% ee [198] (Scheme 5-61). Some representative results with Me3Al and Et3Al are summarized in Table 5-6. Some other Zr-catalyzed enantioselective alkene carbometallation reactions have also been developed [199, 200]. These reactions must involve cyclic carbometallation process with Cp2ZrII species, as discussed later.
Et3Al, (1)
n-Oct
O2
n-Oct
OH
AlEt
OH
hexanes
(65%, 33% ee)
n-Oct Et3Al, (1) CH3CHCl2
O2
n-Oct H
Et
Al
n-Oct H
Et
OH
(63%, 92% ee)
(1) : See Scheme 59.
Scheme 5-61
5.4 Reactions Involving π-Complexes of Zirconocene
283
Table 5-6. Zirconium-Catalyzed Enantioselective Alkylalumination of Monosubstituted Alkenesa
Alkene
Methylalumination Yield (%) ee (%)
Ethylalumination Yield (%) ee (%)
nHexCH=CH2
88
72
&
&
n-BuCH=CH2
&
&
74
93
i-BuCH=CH2
92
74
77
90
c-HexCH=CH2
80
65
low
&
t-BuCH=CH2
low
&
&
&
PhCH2CH=CH2
77
70
69
93
HO(CH2)4CH=CH2
79
75
88
90
Et2N(CH2)3CH=CH2
68
71
56
95
Me2Si(CH2CH=CH2)2b
81
74
66
96
a
The reactions were run with 8 mol % of (1) in Scheme 59 in (CH2Cl)2, CH3CHCl2, or CH2Cl2, and the yield figures are for the oxidized alcohols which were uniformly (R)-isomers. b The structure of the oxidation product was: Me2Si
OH
R(Me or Et)
5.4 Reactions Involving π-Complexes of Zirconocenes or Three-Membered Zirconacycles The chemistry of Cp2ZrII compounds had not been extensively developed until recently. Thus, for example, a comprehensive review of the organometallic chemistry of Zr published in 1974 [1] allocated less than a page on this subject. Pioneering investigations in the 1970s and early 1980s including noteworthy contributions by Bercaw [87, 201, 202] and Erker [17, 28, 51, 52, 64, 124, 141, 203–209] led to some fundamentally significant findings. However, it was not until the 1985–87 period that the significance of the Cp2ZrII chemistry as a versatile new tool for the synthesis of both simple and complex organic compounds was recognized and exploited. Development of the bicyclization of enynes promoted by ‘Cp2Zr(II)’ to give zirconabicycles convertible to bicyclic enones via carbonylation in 1985 [55] was a breakthrough inducing numerous other related studies discussed in this section. Recognition that dialkylzirconocenes, in particular n-Bu2ZrCp2, can serve as convenient precursors to ‘Cp2Zr(II)’ was a practical, experimental breakthrough [56], since the presence of π-compounds, such as alkenes, alkynes, carbonyl com-
284
5
Zirconocenes
pounds, and nitriles, provides a simple route to their π-complexes with Cp2Zr(II). Alternatively, treatment of alkyl- and alkenylzirconocene halides with an appropriate organometal, e.g., MeLi, and treatment of MeZrCp2Cl with various organolithiums and Grignard reagents provide convenient routes to alkyne–, alkene– and benzyne–ZrCp2 complexes without the use of alkynes, alkenes, and benzynes as starting compounds [58, 59]. These modifications of earlier related processes reported since 1986 also represented an experimental breakthrough. The synthetic significance of the Cp2Zr(II) chemistry mainly developed since 1985 easily rivals that of RZrCp2Cl discussed in Sect. 5.3. 5.4.1 Formation of π-Complexes of Zirconocenes Alkenes including dienes, alkynes including benzynes, carbonyl compounds, and imines are the four most commonly used and synthetically important classes of πcompounds. Their π-complexes with ZrCp2 are accessible by various methods, of which those shown in Table 5-7 are significant. Table 7. Representative Methods for the Formation of π-Complexes of Zirconocene Applicabilityb
Patterna
Method X
β-H Abstraction
Cp2Zr R
Complexation
Y
X
"Cp2Zr"
Cp2Zr
H
Y
Cp2Zr
Alkyne
Carbonyl
+
+
c
+
+
+
+
+
+
+
d
c
c
e
+
+
Alkene X Y
X Y
Imine
Cp2Zr
X
Ring Contraction
Cp2Zr C
Y
X
Transmetallation
Cp2Zr
C
Y
M M Cp2Zr
Cp2ZrCl2
X Y
X Y
R1 R1
Migratory Insertion
R2
Cp2Zr C
Cp2Zr
C
R2
X
X a
X, Y = C, N, and/or O. dienes. e See Scheme
b
+ indicates that the method is applicable.
c
Not clarified.
d
Known for conjugated
5.4 Reactions Involving π-Complexes of Zirconocene
285
β-Hydrogen abstraction of diorganylzirconocenes and related monoorganylzirconocenes is currently by far the most convenient route to π-complexes of zirconocene. Some representative examples of the Negishi protocol [107, 110] (Scheme 5-62) [115] (Scheme 5-63) and the Buchwald protocol [108] (Scheme 564) [59] (Scheme 5-65) are shown below. Et
2 n-BuLi Cp2ZrCl2
Cp2Zr
Cp2Zr(Bu-n)2 RC CR PMe3
R Cp2Zr
[107]
Et
Me3P
R
Cp2Zr
Ph
PhCH CHPh
PMe3
Cp2Zr [107]
[110]
Ph
Ph Cp2Zr Me3P
Ph
Scheme 5-62 Li
Li Cp2Zr
Cp2ZrCl2
Cl R
Cp2Zr
Cp2Zr
R
PMe3
Cp2Zr
[115]
Me3P
Scheme 5-63
RC CR
R
R C C Cp2Zr
H
C C Cp2Zr
H Cl
Cp2Zr
R
R
MeLi
PMe3
H
R Cp2Zr Me3P
Me
Cl C C
MeLi Cp2Zr
H
PMe3 Cp2Zr
Cl
H Me
Scheme 5-64
Cp2Zr Me3P
R
286
5
Zirconocenes
R1 1. R1
Li
1. LiN
R2
R1
R
Cp2Zr Me3P
R2
2. PMe3
2. PMe3
R1 Cp2Zr Me3P
N
R
Me Li 1.
Cp2Zr Cl 1. LiS
Cp2Zr
Cp2Zr Me3P
R
R
2. PMe3
2. PMe3
Me3P
S
Scheme 5-65 In cases where the starting π-compounds are readily available, the Negishi protocol provides the most convenient route. On the other hand, the Buchwald protocol is applicable even to those cases where π-compounds are exceedingly unstable, e.g., cyclohexyne [59], benzynes [58], and cyclic allenes and cumulenes [210–212]. The two protocols are complementary with each other in many other respects as well. The mechanistic details of these reactions have been reasonably well clarified. For example, Cp2Zr(Bu-n)2, generated at low temperature by transmetallation, decomposes smoothly at or above room temperature by a nondissociative, unimolecular process. The presence of a base, such as PMe3, generally has little or no effect on the kinetics. It merely traps an unstable 16e π-complex. A concerted intramolecular mechanism shown in Scheme 5-66 [57, 213, 214] appears to be plausible, although its more elaborate version involving the transient formation of 18e alkene(alkyl)(hydrido)zirconocene and other minor variations may not be rigorously ruled out at this point. As in the β-dehydrozirconation of RZrCp2Cl, β-agostic interaction must be of crucial importance. Interestingly, those dialkylzirconocenes containing a β-CH3 group, such as Et, s-Bu, and t-Bu, decompose most readily, followed by those containing a β-CH2 group, e.g., n-Bu, which, in turn, are much more reactive than those containing only β-CH groups, e.g., i-Bu2ZrCp2 [214]. The observed order of reactivity, i.e., s-Bu ≈ t-Bu ≥ Et > n-Bu > i-Bu, has little to do with the overall steric requirements. A very high reactivity of PhCH2CH2 indicates that some electronic factors are also important. The order presented above does not appear to agree with the relative order of effectiveness of RZrCp2Cl as a hydrozirconating agent. And yet, the same β-agostic interaction must play a crucial role in both reactions. A plausible explanation of the puzzling discrepancy is that, whereas the hydride-transfer hydrozirconation is an equilibrium reaction, the β-H abstraction reaction of dialkylzirconocenes is essentially irreversible, even though the latter reaction formally involves cleavage and formation of one each of C–H and C– Zr bonds and hence should be essentially thermoneutral. This reaction must be
5.4 Reactions Involving π-Complexes of Zirconocene
287
largely entropically driven, and the relative reactivity must be kinetically governed. On the other hand, the reversible hydride-transfer hydrozirconation must be thermodynamically controlled, which makes isoalkyl groups highly effective. The unexpectedly sluggish alkene displacement shown in parentheses must be attributable to the coordinative saturation of chloro(hydrido)alkenezirconocenes. β-H Abstraction acid-base R
Cp2Zr R
1
interaction
C
Cp2Zr
C
(irreversible)
H 16e
R
+
H
1
R
16e
1
R
reductive elimination Cp2Zr R
(irreversible)
H 18e
H-Transfer Hydrozirconation 2
R
2
1
R
R
slow
1
R
Cp2Zr Cl
H 16e
Cp2Zr Cl
H 18e
Cp2Zr Cl
H 18e
Scheme 5-66 The ring contraction route to π-complexes of zirconocene must involve interaction between the Cβ–Cβ′ bond and an empty orbital of Zr [35], as discussed later in detail. In this connection, it now is reasonably clear that π-complexation via alkene displacement in the Negishi protocol must proceed in most cases by a ring expansion-ring contraction mechanism (Scheme 5-67).
Scheme 5-67 The π-complexation method with ‘Cp2Zr(II)’ other than alkene–ZrCp2 discussed above tends to be more complex and generally inferior. Zirconocene reagents may be isolated and purified as Cp2Zr(CO)2, Cp2Zr(PMe3)2, and so on. These complexes are 18e species, and their complexation reactions are generally sluggish. The
288
5
Zirconocenes
current scope of the transmetallation and migratory insertion routes to π-complexes of zirconocene is essentially limited to those discussed in Sect. 5.2.1. However, they may prove to be much more useful and complementary with the other methods. For example, treatment of Cp2ZrCl2 with 3 equiv. of LiC≡CR first generates LiCp2Zr(C≡CR)3 which then undergo an unprecedented migratory insertion to give alkyne–ZrCp2 complexes, displaying a novel transformation of Cp2Zr derivatives which also promises to be synthetically useful [215] (Scheme 5-68). R HCl
R
3 LiC CR LiCp2Zr(C CR)3
Cp2ZrCl2
H
LiCp2Zr RC C
R
R
I2
H
R
R
Scheme 5-68
5.4.2 Stoichiometric Reactions of π-Complexes of Zirconocenes (or Three-Membered Zirconacycles) 5.4.2.1 Protonolysis and Halogenolysis Treatment of π-complexes of zirconocenes with proton acids can be complicated by simultaneous formation of diprotonated compounds and dissociated π-compounds [216]. To achieve dissociation of π-ligands in a cleaner manner, the use of I2 should be considered [92, 216] (Scheme 5-69). The I2 treatment method is also applicable to release carbonyl compounds from zirconaoxiranes [216] (Scheme 570). H HCl
Ph +
Ph Ph
Ph
H
Cp2Zr
H Ph
I2
Ph +
Ph H
Scheme 5-69
Cp2ZrI2
Ph
5.4 Reactions Involving π-Complexes of Zirconocene
H3O+ (D3O+)
289 OH H(D)
CO O ZrCp2
ZrCp2
I2
O
Scheme 5-70
5.4.2.2 Ring Expansion Reactions The synthetically most attractive transformations of three-membered zirconacycles are their ring expansion reactions producing five-membered zirconacycles. Those involving alkenes and alkynes are formally carbozirconation reactions, and those involving carbonyl compounds, imines, and nitriles are formally polar addition reactions. These two classes of reactions may share some common mechanistic features, but they may also display mutually distinguishing features. For the carbometallation processes of 16e species a concerted four-center addition mechanism may be tentatively proposed (Scheme 5-71). Regardless of precise mechanisms, 16e π-complexes with one empty orbital may be expected to be more reactive in most, if not all, of these reactions than their 18e counterparts. Even when we consider only alkenes, alkynes, carbonyl compounds, and imines, there are four homocoupling combinations and six cross-coupling combinations. Other structural variations, such as conjugated dienes in place of alkenes, benzynes in place of alkynes, thiocarbonyl compounds in place of carbonyl compounds and nitriles in place of imines, lead to an even larger number of binary combinations for the synthesis of a wide range of organic compounds. Another attractive synthetic strategy is to link two π-functional groups with tethers of various lengths and structures. In the ZrCp2-promoted π-π coupling reactions, ‘pair’-selectivity and regioselectivity as well as stereoselectivity in some cases provide challenging synthetic problems. In cases where the two π-compounds can be linked with an appropriate tether, the ‘pair’-selectivity and regioselectivity problems may be alleviated. Furthermore, the intramolecular coupling process can be kinetically very favorable. Here again, a large number of binary combinations of π-functional groups are conceivable, but the current scope of the ZrCp2-promoted bicyclization appears to be still largely limited to the three binary combinations of alkenes and alkynes (Scheme 5-72). A notable example of other binary combinations is the ZrCp2-promoted bicyclization of alkenyl and alkynyl hydrazones [217] (Scheme 5-73).
290
5
Zirconocenes
C C Cp2Zr
Cp2Zr
Cp2Zr
C C C C Cp2Zr
Cp2Zr
Cp2Zr
C C
Scheme 5-71
"Cp2Zr"
"Cp2Zr"
"Cp2Zr"
Cp2Zr
[92, 218, 219]
Cp2Zr
[55, 57, 220, 221]
Cp2Zr
[56, 57, 222, 223]
Scheme 5-72
PNB N NMe2
"Cp2Zr"
N NMe2 CH3
71%
TFA N NMe2
"Cp2Zr"
C6H5
N NMe2
C6H5
Scheme 5-73
74%
5.4 Reactions Involving π-Complexes of Zirconocene
291
Discussion of all of the known Cp2Zr-promoted coupling reactions is not practical. Only some seminal and representative references of intermolecular coupling reactions are presented in Table 5-8. Table 5-8. Some Representative References for Intermolecular Coupling Reactions of π-Compounds Promoted by Cp2Zr
In each of the intermolecular coupling reactions, the ‘pair’-selectivity and regioselectivity can be significant issues, as detailed below for the alkenyl–alkenyl, alkenyl–alkynyl, and alkynyl–alkynyl coupling reactions. At least five different types of ‘pair’-selectivity patterns have been observed, as summarized in Table 5-9 [35].
Table 5-9. Some Distinctive Features of Various Types of Ring Expansion Reactions of Alkene-Zirconocene Complexes
292
5
Zirconocenes
Some representative examples of the five different types of reactions involving alkenes and alkynes are shown in Scheme 5-74. The following common characteristics should also be noted. Firstly, alkyl groups prefer to be β to Zr, presumably to R
Type I
Cp2Zr [225] Cp2ZrEt2
R
Cp2Zr RC CR Cp2Zr [228]
R R
Type II Ph Cp2Zr [216]
Et
Et Cp2ZrBu2
Cp2Zr
Cp2Zr [224]
Et
Type III
R Et
Cp2ZrBu2
Cp2Zr
RC CR
Et
R Cp2Zr
+
[57]
(?)
Cp2Zr Et
R R
Type IV
Me3Si
SiMe3
Et Cp2ZrBu2
Cp2Zr
Cp2Zr [56, 57]
Type V Et Cp2ZrBu2
Bu-n
Cp2Zr [216]
Bu Cp2Zr
Et
+ Cp2Zr Et 70% (2:1:1)
Scheme 5-74
Bu + Cp2Zr
Et
Bu
5.4 Reactions Involving π-Complexes of Zirconocene
293
minimize steric hindrance, while aryl, alkenyl, and alkynyl groups favor the α position, presumably due to favorable benzylic, allylic, and propargylic interaction with Zr. The α/β ratio for aryl is typically 90/10, while that for alkenyl is ≥95%. Secondly, two substituents of the two π-compounds prefer to be trans to each other in the zirconacycles, presumably for steric reasons. The ZrCp2-promoted bicyclization (Scheme 5-72) has proved to be highly useful for the construction of complex bicyclic and polycyclic compounds. It shares some common features with the Co-based Pauson–Khand reaction [238, 239]. However, significant differences presented below should be clearly noted. Firstly, the Pauson–Khand reaction appears to be limited to enyne bicyclization to produce cyclic enones, whereas the Zr-based methodology is applicable to the bicyclization of dienes and diynes as well. Furthermore, the organometallic products can be generated as discrete products which can be further converted to various types of compounds including cyclic enones. So, the Zr-based methodology is more versatile. Secondly, the Pauson–Khand reaction is normally run at high temperatures, whereas the Zr-promoted reactions proceed readily at or below room temperature. Thirdly, while the Pauson–Khand reaction has recently been made catalytic [240], the catalytic version of the Zr-based bicyclization largely remains to be developed. Finally, the enantioselective versions of both of these reactions are also to be developed. Fortunately, both can proceed diastereoselectively, placing most of the substituents on the exo side of the bicyclic skeleton [241, 243]. The initial kinetic argument [241] appears to be untenable, and it must be a largely thermodynamic phenomenon. Another striking feature is the high trans selectivity in the bicyclization of 1,6-heptadienes [92, 218, 243], despite the fact that 1,7-octadienes under the same conditions give predominantly the cis-fused products. The product from 1,7-octadiene was shown to be a kinetic product which can be thermally isomerized to the trans-fused isomer [219]. The trans-selectivity is readily rationalized on the basis of the stereochemistry observed in the intermolecular alkene–alkene coupling, and it offers a novel and attractive route to trans-fused bicyclo[3.3.0]octane derivatives which are otherwise difficult to prepare [243] (Scheme 5-75). The Zr-promoted enyne bicyclization has been applied to efficient and selective syntheses of natural products, notably phorbol (1) [244], pentalenic acid (2) [242], iridomyrmecin (3) [245], dendrobine (4) [246], and 7-epi-β-bulnesene (5) [247] (Scheme 5-76).
294
5
Zirconocenes H n-Bu2ZrCp2
H ZrCp2
[92]
1. CO
O
2. I2
H
H
> 95% trans
> 95% trans
H
H
n-Bu2ZrCp2
ZrCp2
[92]
ZrCp2
[219]
H
H
80% cis H Ph
n-Bu2ZrCp2
Ph ZrCp2
[243]
H
slow
H
n-Bu2ZrCp2
H Ph ZrCp2
[243]
O
2. I2
H Ph
Ph
1. CO
H
H 1. CO
Ph O
2. I2 H
Scheme 5-75
Scheme 5-76
5.4.2.3 Regioisomerization, Stereoisomerization, σ-Bond Metathesis, and Other Stoichiometric Reactions of Three-Membered Zirconacycles In addition to synthetically useful reactions discussed above, many other reactions of three-membered zirconacycles are conceivable, and some have been observed. In some cases, they represent unwanted side reactions, but some of these reactions may prove to be useful. In any event, good knowledge of them is essential to successful use of the Cp2ZrII chemistry.
5.4 Reactions Involving π-Complexes of Zirconocene
295
In some diene bicyclization reactions, a multipositional double bond migration to give conjugated diene–ZrCp2 complexes has been observed. A set of representative and contrasting results are shown in Scheme 5-77 [248, 249]. A number of factors including steric requirements of the two alkene groups and tether structure appear to be involved, and the reaction courses are more or less predictable and interpreted in terms of a dynamic scheme shown in Scheme 5-78. Migration of Zr occurs via 1,3-shift rather than 1,2-shift. Pr-n n-Bu2ZrCp2
ZrCp2 (73%)
n-Bu2ZrCp2
BnN
BnN
ZrCp2
1. CO
BnN
2. I2
O (50%)
R1
R1
R2
ZrCp2
ZrCp2 Et
R2
R3
R3 R1
(CH2)3R1
R1 Et +
R2
R3
ZrCp2
ZrCp2
ZrCp2 R2
R2 R3
R3
Scheme 5-78 Even when regioisomerization of alkenes is not possible, their stereoisomerization may occur. Interestingly, benzyne-ZrCp2 reacts stereospecifically with (E)- and (Z)stilbenes [226] and enynes containing (E)- and (Z)-styrene moieties also stereospecifically react with n-Bu2ZrCp2 to give the corresponding zirconabicycles [220]. And yet, (E)- and (Z)-stilbenes, which are reluctant to couple with alkenes, readily undergo ZrCp2-catalyzed Z-to-E isomerization [107], and a slower but similar isomerization was also shown in Scheme 5-75. The isomerization of (Z)-stilbene was shown to proceed by a very intriguing polar mechanism [243], suggesting that alkene–ZrCp2 complexes are more prone to heterolysis than alkyne–ZrCp2 complexes (Scheme 5-79).
296
5
Zirconocenes
Ph
Ph
Ph
n-Bu2ZrCp2
Ph Ph ZrCp2 Ph
Ph Ph
Ph
+
_
Cp2Zr
ZrCp2 Ph
Ph Ph
Scheme 5-79 It has recently become increasingly clear that zirconacyclopropanes readily react with various σ bonds to undergo σ-bond metathesis. For example, the reaction of Cp2ZrEt(CH2=CH2)MgBr with Cp2ZrMe2 produces in 43% yield (Cp2ZrMe)2(CH2CH2), which is thought to proceed as shown in Scheme 5-80 [250]. It appears certain that many more σ-bond metathesis reactions of threemembered zirconacycles will be found in the future. -
Et
Cp2Zr
Me Cp2Zr
+
Me2ZrCp2
Cp2Zr Me
Cp2Zr
ZrCp2
Me ZrCp2Me
Scheme 5-80
5.4.3 Stoichiometric Reactions of Five-Membered Zirconacycles In some respects, five-membered zirconacycles behave as diorganylzirconocenes. Thus, protonolysis and iodinolysis tend to proceed normally. In many other respects, however, five-membered zirconacycles exhibit their own chemical properties. The five-membered cyclic structure effectively blocks β-dehydrozirconation which makes them thermally more stable than many of the acyclic diorganylzirconocenes. On the other hand, the very cyclic structure must be responsible for their strong tendency to undergo ring contraction to generate three-membered zirconacycles, which can then undergo a variety of reactions discussed in the previous section. One of the earliest examples is an unexpected ring-opening reaction of a zirconabicycle with MeCN [57] (Scheme 5-81).
5.4 Reactions Involving π-Complexes of Zirconocene
SiMe3
SiMe3
297 SiMe3
MeCN ZrCp2
ZrCp2
ZrCp2 Me
N
Scheme 5-81 The ring-contraction reaction most certainly is the microscopic reversal of the carbometallative ring-expansion reaction, and the agostic interaction involving the Cβ–Cβ′ bond must be critically responsible. It is also reasonable to envision a ring expansion–ring contraction tandem as a plausible mechanism for π-substitution on Zr (Scheme 5-67). More important from the synthetic viewpoint is that the ring contraction–ring expansion tandem promises to offer a number of synthetically useful routes to three- and five-membered zirconacycles that are otherwise inconvenient, expensive, and/or difficult to prepare. Some representative examples are shown in Scheme 5-82 [228, 251, 252]. R 1
R
1
R C CR
Cp2Zr R1 1
R R
Cp2ZrEt2
R
R1C N Cp2Zr
R Cp2Zr
N
R
RC CR Cp2Zr
R1 R
R
R R
R1CHO
and/or Cp2Zr
Cp2Zr O
R1
O R1
PMe3
R Cp2Zr Me3P
R
Scheme 5-82
A skeletal rearrangement reaction shown in Scheme 5-83 [253] must be mechanistically closely related to the displacement reactions shown in Scheme 5-82, and it must proceed via a series of ring contraction and ring-expansion steps.
298
5
Zirconocenes
Br Br 1. Mg 2. Cp2MCl2
1. Mg 2. Cp2MCl2
Br
3. Br2
Br
M = Zr or Hf Br2
Cp2M
Cp2M
Cp2M
Scheme 5-83 σ-Bond metathesis is another potentially general pattern for various reactions of five-membered zirconacycles, even though its current scope is still very narrow. The reaction of PhC≡CPh with H2ZrCp2 to give a dimeric product shown in Scheme 5-84 represents one of the earliest examples [127] and has been interpreted in terms of σ-bond metathesis [5].
Ph
H2ZrCp2
Ph
Ph
PhC CPh H
ZrCp2
ZrCp2 H
Ph
Ph
Ph
Ph
Ph
PhC CPh
H2ZrCp2 Ph
Ph ZrCp2
Ph
H
Cp2Zr
Ph ZrCp2
H
Scheme 5-84
Opening of five-membered zirconacycles via transmetallation is another potentially general and significant σ-bond metathesis process (Scheme 5-85). The reaction of zirconacyclopentanes with EtMgBr in the presence of PMe3 gives regioselectively 2-ethylalkylmagnesium bromides and Cp2Zr(CH2=CH2)(PMe3) most probably via transmetallation followed by β-H abstraction [225] (Scheme 5-86). As discussed earlier, abstraction of a β-H atom from the Et group should be strongly favored over the Mg-containing alkyl group which is either a β-CH2 or β-CH group. How-
5.4 Reactions Involving π-Complexes of Zirconocene
299
ever, it is not clear what factors favor the observed regioselectivity. Despite this ambiguity, the σ-bond metathesis tandem consisting of transmetallation and β-H abstraction undoubtedly plays a crucial role in the Dzhemilev ethylmagnesiation of alkenes catalyzed by Cp2ZrCl2 [54], as detailed later. A related reaction with alkyllithiums proceeds with displacement of one Cp group [63] (Scheme 5-87). RM Cp2Zr
Cp2Zr
R
M
Cp2Zr
M R
Scheme 5-85 MgBr R Cp2Zr
EtMgBr
Et
R
?
R
Cp2Zr
Cp2Zr
Et MgBr
H
MgBr
Cp2Zr
Et
PMe3 R
R
MgBr + Cp2Zr PMe3
Scheme 5-86
Scheme 5-87
5.4.4 Catalytic Reactions Involving Zirconacycles and Related Derivatives Several catalytic reactions involving monoorganylzirconocene halides are discussed in Sect. 5.3.6. In this section, those catalytic reactions that involve π-complexes of ZrCp2 or three-membered zirconacycles as well as their derivatives are presented. Stereoisomerization of stilbene catalyzed by ‘ZrCp2’ discussed in Sect. 5.4.2.3
300
5
Zirconocenes
(Scheme 5-79) is an example of this class of reactions. However, it is a series of discoveries and recognition of cyclic mechanisms involving C–H activation via βagostic interaction that have opened up an exciting new chapter in the organozirconium chemistry, which not only is of basic scientific interest but also of considerable synthetic potential.
5.4.4.1 Zirconium-Catalyzed Carbomagnesiation Reactions Involving C–H Activation As discussed in Sect. 5.3.6, the Zr-catalyzed methylalumination of alkynes discovered in 1978 [49] is thought to involve a straightforward concerted addition of a Me–metal bond to alkynes, irrespective of its precise mechanism. Several years later, Dzhemilev [54] reported a seemingly related reaction involving Zr-catalyzed ethylmagnesiation of alkenes (Scheme 5-88), which was also thought to involve a straightforward carbometallation, even though the failure in attempts to achieve the corresponding methylmagnesiation was puzzling. Through systematic investigations of three stoichiometric reactions, i.e., (1) β-H abstraction reaction of Et2ZrCp2 to form ethylene-ZrCp2 [110], (2) its reaction with 1-alkenes to produce ‘pair’-selectively and regioselectively 3-alkylzirconacyclopentanes [225], and (3) their reaction with EtMgBr to afford 2-ethylalkylmagnesium bromides and ethylene-ZrCp2 via regioselective transmetallation and -H abstraction discussed in the previous section [225], the mechanism of the Zr-catalyzed ethylmagnesiation was accidentally clarified [225, see also 254] (Scheme 5-89). A few related suggestions were made [255–257], but they did not lead to full mechanistic clarification. Whereas titanocene derivatives readily undergo α C-H activation [258–260], zirconocenes are reluctant to participate in α-H abstraction. This readily explains the failure in attempts to observe methylmagnesiation. Higher alkylmagnesium halides do undergo a similar reaction, but it is less clean [261]. This can also be readily understood if one reviews the contrasting behavior of ethylene-ZrCp2 and 1-alkene-ZrCp2 and the relative order of reactivity of various alkyl groups in β-H abstraction (Sect. 5.4)
R
+
EtMgBr
cat. Cp2ZrCl2
R
MgBr Et
Scheme 5-88
5.4 Reactions Involving π-Complexes of Zirconocene
301
R ZrCp2 R
EtMgBr
R Et2ZrCp2
ZrCp2
MgBr
H ZrCp2 R
MgBr Et
Scheme 5-89
The corresponding Zr-catalyzed reactions of nonconjugated dienes with alkylmagnesium reagents proceed similarly, but the courses of these reactions depend on various reaction parameters. With EtMgBr, dienes act as two monoenes in one, and the Dzhemilev ethylmagnesiation takes place at each vinyl group [262] (Scheme 590). The reaction of 1,6-heptadiene with n-BuMgCl in THF catalyzed by Cp2ZrCl2 predominantly gives 1-magnesiomethyl-2-methylcyclopentane, while the use of nBu2Mg in place of n-BuMgCl and/or Et2O in place of THF leads to 1,2bis(magnesiomethyl)cyclopentanes [256, 262–265, See also 266] (Scheme 5-91), and some other dienes also reacted similarly. The mechanism of the monomagnesium derivative must closely resemble that of the ethylmagnesiation of 1-alkenes (Scheme 5-92), whereas double transmetallation of zirconacyclopentanes with alkylmagnesiums must be involved in the formation of 1,2bis(magnesiomethyl)cyclopentanes. Et EtMgBr cat. Cp2ZrCl2
ZrCp2
MgBr EtMgBr MgBr
ZrCp2 Et
Scheme 5-90
302
5
Zirconocenes n-BuMgCl THF
CH3 (major) MgCl
n-BuMgX cat. Cp2ZrCl2
ZrCp2
n-BuMgCl, Et2O
X = Cl or n-Bu
or n-Bu2Mg Et2O or THF
MgY
(major)
MgY
Y = Cl or C group
Scheme 5-91 Et ZrCp2
H
CH3
n-BuMgCl ZrCp2
MgCl
MgCl
Scheme 5-92 A potentially very useful variation of the ethylmagnesiation reaction is an ethylmagnesiation–elimination tandem process to convert allyl ethers into 3-ethyl-1-alkenes [267, 268] (Scheme 5-93). OPh R R
OPh
ZrCp2
R Et2ZrCp2
ZrCp2
ZrCp2OPh R ZrCp2
R H Et
Scheme 5-93
EtMgBr
5.4 Reactions Involving π-Complexes of Zirconocene
303
Attempts to observe Zr-catalyzed carbomagnesiation of alkynes have been mostly disappointing. A notable exception is the Zr-catalyzed reaction of conjugated diynes which gives the expected ethylmagnesiation products [269] (Scheme 5-94). R
RC C C CR
EtMgBr
+
R
cat. Cp2ZrCl2
Et
MgBr
Scheme 5-94 One very significant variation of the Zr-catalyzed carbomagnesiation of alkenes is its enantioselective version, which may be related but distinct from the Zr-catalyzed enantioselective carboalumination of alkenes involving straightforward, acyclic C–Zr bond addition (Sect. 5.3.6.3). Two notable examples are the Zr-catalyzed ethylmagnesiation followed by β-elimination of allylic ethers and amines [199] (Scheme 5-95) and its more recent modification producing alkylmagnesium derivatives [200] (Scheme 5-96). For further details, the readers are referred to Chapter 10. EtMgCl Et cat. (1) Cn
Cn
O
(1) =
(>90%ee) OH
Zr(BINOL)
Scheme 5-95 Et2Mg cat. (2)
R2 N
(2) = Cl
Zr
Et R2 N
Cl
Scheme 5-96
MgEt
(36-95%ee)
304
5
Zirconocenes
5.4.4.2 Zirconium-Catalyzed Carboalumination of Alkynes via C–H Activation Clarification of the mechanism of the Zr-catalyzed ethylmagnesiation as a cyclic process involving C–H activation was a significant turning point, which necessitated reinvestigation of all Zr-catalyzed carbometallation reactions, as the overall equation showing the addition of alkyl–metal bond to alkenes and alkynes should no longer be taken as an indication of straightforward four-centered addition reactions. Indeed, reinvestigation of the Zr-catalyzed carboalumination of alkynes with ethyl- and higher alkylalanes [49] has revealed a number of unexpected and highly intriguing results, as represented by the following results with ethylalanes. Firstly, the reaction of alkynes with Et2AlCl and Cp2ZrCl2 shows no sign of cyclic processes. In polar solvents, e.g., (CH2Cl)2, which are essential to observing favorable results, it is a satisfactory reaction of synthetic utility, producing cis addition products in high yields. This reaction can also accommodate higher n-alkyl groups [90]. On the other hand, the corresponding reaction of Et3Al and Cp2ZrCl2 is mostly or nearly exclusively cyclic [270] and proceeds better in nonpolar solvents, e.g., hexanes, than in (CH2Cl)2 [90] (Scheme 5-97). Secondly, a diethylation mechanism proceeding through Et2ZrCp2, ethylene-ZrCp2, and a zirconacyclopentene intermediate shown in Scheme 5-98 is not operative, as the reaction of 5-decyne and Et3Al (3 equiv.) is not catalyzed by the zirconacyclopentene derived from 5-decyne and Et2ZrCp2, generated by treating Cl2ZrCp2 with 2 equiv. of EtMgBr. However, addition of ClAlEt2 (1 equiv. relative to Cl2ZrCp2) induces the desired catalytic process. Clearly, an active form of Cl or Br not sequestered as part of Mg salts is necessary. Thirdly, the three stoichiometric reactions summarized in Scheme 5-99 give three discrete major products. In the absence of 5-decyne, these three reactions were reported to give discrete bimetallic complexes [271–275] (Scheme 5-100). Assuming a 1:1 correlation between the results shown in Scheme 5-99 and 100, a carbozirconation mechanism accommodating all three reactions may be proposed (Scheme 5-101). Direct carbozirconation is tentatively favored to accommodate the D distribution pattern for the formation of the trideuterated product. Fourthly, treatment of EtZrCp2Cl, generated by hydrozirconation of ethylene with HZrCp2Cl, with 1 equiv. of Et3Al provides a clean and high-yielding route to the five-membered bimetallic reagent, which was thought to be the carbozirconating agent in the catalytic reaction (Scheme 5-102). Indeed, its reaction with 5-decyne smoothly gives the same organoalane product and EtZrCp2Cl under the stoichiometric conditions (Scheme 5-103), and it also induces the same catalytic reaction. All these results rather strongly support a novel bimetallic and monoalkylative mechanism shown in Scheme 5-104. In the critical step producing the five-membered bimetallic complex, three components consisting of (1) EtZrCp2 for crucial agostic interaction, (2) Et3Al but not Et2AlCl to provide an Et base, and (3) one Cl atom not sequestered by Mg and hence available for linking Zr and Al are thought to be necessary.
5.4 Reactions Involving π-Complexes of Zirconocene
305
Et2AlCl Bu-n
n-Bu
cat. Cp2ZrCl2
(X = Cl and/or Et)
(CH2Cl)2
Et
AlX2 90%
n-BuC CBu-n Bu-n
3 Et3Al cat. Cp2ZrCl2
Bu-n
n-Bu
n-Bu
D3O+ AlEt
hexanes
D D 92%
Scheme 5-97
Cp2Zr
RC CR R R
Et3Al Cp2ZrCl2
Cp2ZrEt2
Cp2Zr R R
EtAl
Scheme 5-98
Et3Al
306
5
Zirconocenes
1. Et3Al (1 equiv)
Bu-n
n-Bu
Cp2ZrCl2 (1 equiv) D 2. DCl, D2O n-BuC CBu-n
1. Et3Al (3 equiv)
CH3 Bu-n
n-Bu
Cp2ZrCl2 (1 equiv) D CH2D
2. DCl, D2O
n-Bu 3 Et3Al + Cp2ZrCl2
1.
23oC,
Bu-n
24 h
2. n-BuC CBu-n 3. DCl, D2O
Scheme 5-99
Scheme 5-100
D CHD2
5.4 Reactions Involving π-Complexes of Zirconocene
Scheme 5-101
Scheme 5-102
307
308
5
Zirconocenes
Scheme 5-103
Scheme 5-104
5.4.4.3 Other Zirconium-Catalyzed Reactions Involving Cp2Zr(II) Complexes Three-membered zirconacycles can react with various σ-bonded species to undergo cleavage of C–Zr bonds. Their protonolysis and iodinolysis (Sect. 5.4.2.1), and transmetallation with organometals containing Mg (Scheme 5-80) and Al (Scheme 5-102) are but a few examples of such reactions. It is therefore not surprising to find that three-membered zirconacycles can serve as catalysts or catalyst precursors in the reactions of σ-bonded compounds. Hydrogenation [127, 276–279] and hydrosilation [280–282] of alkenes (Scheme 5-105) have indeed been shown to be catalyzed by 1-butenezirconocene and related compounds.
5.4 Reactions Involving π-Complexes of Zirconocene
309
Scheme 5-105 Besides being potentially useful, their mechanisms present interesting puzzles of fundamental significance. In one study, for example, the hydrosilation reaction was thought to proceed via (1) oxidative addition of H2SiPh2 to alkene-ZrCp2, (2) hydrozirconation or silylzirconation, and (3) reductive elimination [280] (Scheme 5106). The feasibility of oxidative addition and the ability of monomeric oxidative addition product to undergo hydrosilation with alkenes have been experimentally demonstrated. And yet, some of the other proposed steps, such as silylzirconation and formation of C–Si bonds via reductive elimination, are not experimentally supported. Mechanistic schemes involving σ-bond metathesis have also been suggested [35, 282, See also 283, 284]. Despite firm experimental indications for oxidative addition of Cp2Zr to H–Si bonds, it appears worthwhile to examine the feasibility of σ-bond metathesis mechanisms such as Scheme 5-107.
Scheme 5-106
Clearly, further explorations of fundamental transformations of Cp2Zr complexes, especially their reactions with σ-bonded compounds are highly desirable, and many additional stoichiometric and catalytic reactions involving such transformations will be found.
310
5
Zirconocenes
Scheme 5-107
5.5 Conclusion The chemistry of Cp2Zr derivatives is now almost half a century old. Developments mainly in the areas of hydrozirconation (from 1971), Zr-mediated carbometallation reactions represented by the Kaminsky polymerization (from 1974), and Zr-catalyzed controlled carboalumination (from 1978) have led to explosive investigations of RZrIVCp2 derivatives discussed in Sect. 5.3. The scientific and practical significance of Cp2Zr chemistry has been substantially elevated by similarly explosive investigations and applications of Cp2ZrII chemistry discussed in Sect. 5.4 mainly in the 1980s and 1990s. Undoubtedly, these areas will be further explored and developed. For example, such fundamentally important processes as oxidative addition, reductive elimination, and σ-bond metathesis of Cp2Zr derivatives remain largely unexplored and poorly understood. Their migratory insertion chemistry with the notable exception of carbonylation is still poorly developed. Many of the topics covered in this chapter are also being rewritten, as their enantioselective versions are developed. Most of the structural types shown in Table 5-2 are covered in this chapter. However, several of them, notably Cp2Zr=a (Types II and XIV), Cp2Zr≡a (Type III), four-membered zirconacycles (Type VI), are hardly covered despite some potentially significant investigations including carbonyl–olefination reactions of 1,1-dimetalloalkanes and -alkenes and Cp2Zr–carbene complexes [146, 285–287] and chemistry of imidozirconocene derivatives [288–296]. Admittedly, it is not possible to properly represent all pertinent topics in a single chapter. The readers are referred to a few monographs [1,2] and various reviews listed in Table 5-10 according to their topics.
5.5 Conclusion
Table 5-10. Classified List of Reviews
311
312
5
Zirconocenes
References Monographs on Cp2Zr Chemistry [1] P. C. Wailes, R. S. P. Courtts and H. Weigold, Organometallic Chemistry of Titanium, Zirconium, and Hafnium, Academic Press, New York, 1974, 302 pp. [2] D. J. Cardin, M. F. Lappert, and C. L. Raston Chemistry of Organozirconium and Hafnium Compounds, John Wiley & Sons, New York, 1986, 451 pp. Reviews [3] D. J. Cardin, M. F. Lappert, C. L. Raston, and R. I. Riley in Comprehensive Organometallic Chemistry, Vol. 1, G. Wilkinson, F. G. A. Stone, and E. W. Abel, Eds., Pergamon Press, Oxford, 1982, 549. [4] D. J. Cardin, M. F. Lappert, C. L. Raston, and R. I. Riley, in Comprehensive Organometallic Chemistry, Vol. 1, G. Wilkinson, F. G. A. Stone, and E. W. Abel, Eds., Pergamon Press, Oxford, 1982, 559. [5] E. Negishi and T. Takahashi, Aldrichim. Acta, 1985, 18, 31. [6] E. Negishi and T. Takahashi, Synthesis, 1988, 1. [7] V. Farina in Comprehensive Organometallic Chemistry II, E. W. Abel, F. G. A. Stone, G. Wilkinson, Eds., Pergamon Press, Oxford, 1995, Vol. 12, p. 161. [8] J. Schwartz and J. A. Labinger, Angew. Chem., Int. Ed. Engl., 1976, 15, 333. [9] J. A. Labinger in Comprehensive Organic Synthesis, Vol. 8, B. M. Trost, I. Fleming, Eds., Pergamon Press, Oxford, 1991, 667. [10] P. Wipf and H. Jahn, Tetrahedron, 1996, 52, 12853. [11] Y. Hanzawa, H. Ito, and T. Taguchi, Synlett., 1995, 299. [12] S. L. Buchwald and R. D. Broene, in Comprehensive Organometallic Chemistry II, E. W. Abel, F. G. A. Stone, G. Wilkinson, Eds., Pergamon Press, Oxford, 1995, Vol. 12, p. 771. [13] H. Yasuda, K. Tatsumi, and A. Nakamura, Acc. Chem. Res., 1985, 18, 120. [14] H. Yasuda and A. Nakamura, Rev. Chem. Intermediates, 1986, 6, 365. [15] R. F. Jordan, Adv. Organomet. Chem., 1991, 32, 325. [16] A. S. Guram and R. F. Jordan in Comprehensive Organometallic Chemistry II, E. W. Abel, F. G. A. Stone, G. Wilkinson, Eds., Pergamon Press, Oxford, 1995, Vol. 4, p. 589. [17] G. Erker, Acc. Chem. Res., 1984, 17, 103. [18] J. Jubb, J. Fong, D. Richeson, and S. Gambarotta in Comprehensive Organometallic Chemistry II, E. W. Abel, F. G. A. Stone, G. Wilkinson, Eds., Pergamon Press, Oxford, 1995, Vol. 4, p. 543. [19] E. Negishi, Acc. Chem. Res., 1982, 15, 340. [20] U. M. Dzhemilev, O. S. Vostrikova, and G. A. Tolstikov, Russ. Chem. Rev., 1990, 59, 1157. [21] U. M. Dzhemilev, Tetrahedron, 1995, 51, 4333. [22] E. Negishi, Pure Appl. Chem., 1981, 53, 2333. [23] U. M. Dzhemilev, O. S. Vostrikova, and G. A. Tolstikov, J. Organomet. Chem., 1986, 304, 17. [24] E. Negishi, Acc. Chem. Res., 1987, 20, 65. [25] E. Negishi and D. Y. Kondakov, Chem. Soc. Rev., 1996, 26, 417. [26] G. Erker, Pure Appl. Chem., 1992, 64, 393. [27] G. P. Pez and J. N. Armor, Adv. Organomet. Chem., 1981, 19, 1. [28] G. Erker, C. Krüger, and G. Müller, Adv. Organomet. Chem., 1985, 24, 1. [29] S. L. Buchwald and R. B. Nielsen, Chem. Rev., 1988, 88, 1047. [30] S. L. Buchwald and R. A. Fisher, Chim. Scripta, 1989, 29, 417. [31] W. A. Nugent, T. V. RajanBabu, and D. F. Taber, Chim. Scripta, 1989, 29, 439. [32] E. Negishi, Chim. Scripta, 1989, 29, 457. [33] E. Negishi in Comprehensive Organic Synthesis, Vol. 5, L. A. Paquette, Ed., Pergamon Press, Oxford, 1991, 1163. [34] R. D. Broene and S. L. Buchwald, Science, 1993, 261, 1696.
References [35] [36] [37] [38] [39] [40]
313
E. Negishi and T. Takahashi, Acc. Chem. Res., 1994, 27, 124. P. Binger and S. Podubrin in Comprehensive Organometallic Chemistry II, E. W. Abel, F. G. A. Stone, G. Wilkinson, Eds., Pergamon Press, Oxford, 1995, Vol. 4, p. 439. R. D. Broene in Comprehensive Organometallic Chemistry II, E. W. Abel, F. G. A. Stone, G. Wilkinson, Eds., Pergamon Press, Oxford, 1995, Vol. 12, p. 313. A. Ohff, S. Pulst, C. Lefeber, N. Peulecke, P. Arndt, V. V. Burkalov, and U. Rosenthal, Synlett., 1996, 111. U. M. Dzhemilev, R. M. Sultanov, and R. G. Gaimaldinov, J. Organomet. Chem., 1995, 491, 1. A. H. Hoveyda and J. P. Morken, Angew. Chem., Int. Ed. Engl., 1996, 35, 1262.
Original papers and others [41] J.J. Berzelius, Ann. Chim. Phys. 1824, 26, 43. [42] G. Wilkinson and J. M. Birmingham, J. Am. Chem. Soc., 1954, 76, 4281. [43] G. A. Olah, Friedel-Crafts and Related Reactions, Interscience, New York, 1963, Vol. 1, 1031 pp. [44] J. Boor, Ziegler-Natta Catalysis and Polymerization, Academic Press, New York, 1978, 670 pp. [45] P. C. Wailes and H. Weigold, J. Organomet. Chem., 1970, 24, 405. [46] P. C. Wailes, H. Weigold, and A. P. Bell, J. Organomet. Chem., 1971, 27, 373. [47] D. W. Hart and J. Schwartz, J. Am. Chem. Soc., 1974, 96, 8115. [48] W. Kaminsky, H.-J. Vollmer, E. Heins, and H. Sinn, Makromol. Chem., 1974, 175, 443. [49] D. E. Van Horn and E. Negishi, J. Am. Chem. Soc., 1978, 100, 2252. [50] E. Negishi, D. E. Van Horn, and T. Yoshida, J. Am. Chem. Soc., 1985, 107, 6639. [51] G. Erker and K. Kropp, J. Am. Chem. Soc., 1979, 101, 3659. [52] G. Erker, J. Wicher, K. Engel, F. Rosenfeldt, W. Dietrich, and C. Kruger, J. Am. Chem. Soc., 1980, 102, 6346. [53] H. Yasuda, Y. Kajihara, K. Mashima, K. Lee, and A. Nakamura, Chem. Lett., 1981, 519. [54] U. M. Dzhemilev, O. S. Vostrikova, and R. M. Sultanov, Izv. Akad. Nauk SSSR, Ser. Khim., 1983, 218. [55] E. Negishi, S. J. Holmes, J. M. Tour, and J. A. Miller, J. Am. Chem. Soc., 1985, 107, 2568. [56] E. Negishi, F. E. Cederbaum, and T. Takahashi, Tetrahedron Lett., 1986, 27, 2829. [57] E. Negishi, S. J. Holmes, J. M. Tour, J. A. Miller, F. E. Cederbaum, D. R. Swanson, and T. Takahashi, J. Am. Chem. Soc., 1989, 111, 3336. [58] S. L. Buchwald, B. T. Watson, and J. C. Huffman, J. Am. Chem. Soc., 1986, 108, 7411. [59] S. L. Buchwald, R. T. Lum, and J. C. Dewan, J. Am. Chem. Soc., 1986, 108, 7441. [60] E. J. Ryan in Comprehensive Organometallic Chemistry II, E. W. Abel, F. G. A. Stone, G. Wilkinson, Eds., Pergamon Press, Oxford, 1995, Vol. 4, 465-480. [61] F. Soleil and R. Choukroun, J. Am. Chem. Soc., 1997, 119, 2938. [62] A. D. Walsh, Trans. Faraday, Soc., 1949, 45, 179 [63] D. Y. Kondakov and E. Negishi, J. Chem. Soc., Chem. Comm., 1996, 963. [64] G. Erker, K. Berg, L. Treschanke, and K. Engel, Inorg. Chem., 1982, 21, 1277. [65] E. Negishi, Organometallics in Organic Synthesis, Vol. 1, J. Wiley & Sons, New York, 1980, 532 pp. [66] A. F. Reid and P. C. Wailes, J. Organomet. Chem., 1964, 2, 329. [67] P. M. Druce, B. M. Kingston, M. F. Lappert, T. R. Spalding, and R. C. Sriniastava, J. Chem. Soc., A, 1969, 2106. [68] Unpublished results by E. Negishi and S. J. Holmes. [69] E. Negishi, J. A. Miller, and T. Yoshida, Tetrahedron Lett., 1984, 25, 3407. [70] D. R. Swanson, T. Nguyen, Y. Noda, and E. Negishi, J. Org. Chem., 1991, 56, 2590. [71] J. R. Surtees, J. Chem. Soc., Chem. Comm., 1965, 567. [72] E. Negishi and T. Yoshida, J. Am. Chem. Soc., 1981, 103, 4985. [73] D. B. Carr and J. Schwartz, J. Am. Chem. Soc., 1977, 99, 638. [74] E. Negishi and L. D. Boardman, Tetrahedron Lett., 1982, 23, 3327.
314 [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95]
[96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111]
5
Zirconocenes
A. Pelter, K. Smith, and H. C. Brown, Borane Reagents, Academic Press, New York, 1988, 503 pp. G. Wilke and H. Müller, Ann. Chem., 1960, 629, 222. G. Zweifel and J. A. Miller, Org. React., 1984, 32, 1. D. W. Hart, T. F. Blackburn, and J. Schwartz, J. Am. Chem. Soc., 1975, 97, 679. C. A. Bertelo and J. Schwartz, J. Am. Chem. Soc., 1976, 98, 262. P. Wipf, W. J. Xu, J. H. Smitrovich, R. Lehmann, and L. M. Venanzi, Tetrahedron, 1994, 50, 1935. B. H. Lipshutz, C. Lindsley, and A. Bhandari, Tetrahedron Lett., 1994, 35, 4669. P. C. Wailes and H. Weigold, Inorg. Synth., 1979, 19, 223. D. B. Carr and J. Schwartz, J. Am. Chem. Soc., 1979, 101, 3521. S. L. Buchwald, S. J. La Maire, R. B. Nielsen, B. T. Watson, and S. M. King, Tetrahedron Lett., 1987, 28, 3895. S. L. Buchwald, S. J. La Maire, R. B. Nielsen, B. T. Watson, and S. M. King, Org. Synth., 1992, 71, 77. D. G. Bickley, N. Hao, P. Bougeard, B. G. Sayer, R. C. Burns, and M. J. McGlinchey, J. Organomet. Chem., 1983, 246, 257. C. McDade and J. E. Bercaw, J. Organomet. Chem., 1985, 279, 281. B. H. Lipshutz, R. Keil, and E. L. Ellsworth, Tetrahedron Lett., 1990, 31, 7257. E. Negishi and T. Yoshida, Tetrahedron Lett., 1980, 21, 1501. E. Negishi, D. Y. Kondakov, D. Choueiry, K. Kasai, and T. Takahashi, J. Am. Chem. Soc., 1996, 118, 9577. J. Endo, N. Koga, and K. Morokuma, Organometallics, 1993, 12, 2777. C. J. Rousset, D. R. Swanson, F. Lamaty, and E. Negishi, Tetrahedron Lett., 1989, 30, 5105. H. Ito, T. Taguchi, and Y. Hanzawa, Tetrahedron Lett., 1992, 33, 7873. M. Chino, T. Matsumoto, and K. Suzuki, Synlett., 1994, 359. H. Ito, T. Nakamura, T. Taguchi, and Y. Hanzawa, Tetrahedron Lett., 1992, 33, 3769. See also (a) H. Ito, T. Taguchi, and Y. Hanzawa, Tetrahedron Lett., 1992, 33, 1295. (b) H. Ito, T. Taguchi, and Y. Hanzawa, J. Org. Chem., 1993, 58, 774. (c) H. Ito, Y. Ikeuchi, T. Taguchi, Y. Hanzawa, and M. Shiro, J. Am. Chem. Soc., 1994, 116, 5469. (d) H. Ito, T. Nakamura, T. Taguchi, and Y. Hanzawa, Tetrahedron, 1995, 51, 4507. T. Takahashi, M. Kotora, R. Fischer, Y. Nishihara, and K. Nakajima, J. Am. Chem. Soc., 1995, 117, 11039. M. J. S. Dewar, Bull. Soc. Chim. Fr. 1951, 18, C71. J. Chatt and L. A. Duncanson, J. Chem. Soc., 1953, 2939. G. W. Watt and F. O. Drummond, J. Am. Chem. Soc., 1966, 88, 5926. G. W. Watt and F. O. Drummond, J. Am. Chem. Soc., 1970, 92, 826. D. J. Sikora, K. J. Moriarty, and M. D. Rausch, Inorg. Synth., 1990, 28, 249. M. D. Fryzuk, T. S. Hadad, and D. J. Berg, Coord. Chem. Rev., 1990, 99, 137. J. L. Atwood, R. D. Rogers, W. E. Hunter, C. Floriani, G. Fachinetti, and A. Chiesi-Villa, Inorg. Chem., 1980, 19, 3812. J. B. Kool, M. D. Rausch, H. G. Alt, M. Herberhold, B. Honold, and U. Thewalt, J. Organomet. Chem., 1987, 320, 37. D. J. Sikora, D. W. Macomber, and M. D. Rausch, Adv. Organomet. Chem., 1986, 25, 318. S. L. Buchwald, B. T. Watson, R. T. Lum, and W. A. Nugent, J. Am. Chem. Soc., 1987, 109, 7137. T. Takahashi, D. R. Swanson, and E. Negishi, Chem. Lett., 1987, 623. S. L. Buchwald, B. T. Watson, and J. C. Huffman, J. Am. Chem. Soc., 1987, 109, 2544. S. L. Buchwald, R. T. Lum, R. A. Fisher, W. M. Davis, J. Am. Chem. Soc., 1989, 111, 9113. T. Takahashi, M. Murakami, M. Kunishige, M. Saburi, Y. Uchida, K. Kozawa, T. Uchida, D. R. Swanson, and E. Negishi, Chem. Lett., 1989, 761. P. Binger, P. Müller, R. Benn, A. Rufinska, B. Gabor, C. Krüger, and P. Betz, Chem. Ber., 1989, 122, 1035.
References
315
[112] P. Binger, B. Biedenbach, A. T. Herrmann, F. Langhauser, P. Betz, R. Goddard, and C. Krüger, Chem. Ber., 1990, 123, 1617. [113] S. L. Buchwald and R. B. Nielsen, J. Am. Chem. Soc., 1988, 110, 3171. [114] S. L. Buchwald, B. T. Watson, M. W. Wannamaker, and J. C. Dewar, J. Am. Chem. Soc., 1989, 111, 4486. [115] D. R. Swanson and E. Negishi, Organometallics, 1991, 10, 825. [116] H. G. Alt, C. E. Denner, U. Thewalt, and M. D. Rausch, J. Organomet. Chem., 1988, 356, C83. [117] T. Takahashi, M. Tamura, M. Saburi, Y. Uchida, and E. Negishi, J. Chem. Soc., Chem. Comm., 1989, 852. [118] P. J. Walsh, F. J. Hollander, and R. G. Bergman, J. Organomet. Chem., 1992, 428, 13. [119] D. A. Straus and R. H. Grubbs, J. Am. Chem. Soc., 1982, 104, 5499. [120] E. J. Moore, D. A. Straus, J. Armantrout, B. D. Santarsiero, R. H. Grubbs, and J. E. Bercaw, J. Am. Chem. Soc., 1983, 105, 2068. [121] S. C. H. Ito, D. A. Straus, J Armantrout, W. P. Schaefer, and R. H. Grubbs, J. Am. Chem. Soc., 1984, 106, 2210. [122] R. M. Waymouth, B. D. Samtarsiero, R. J. Coots, M. J. Bronikowski, and R. H. Grubbs, J. Am. Chem. Soc., 1986, 108, 1427. [123] P. J. Walsh, F. J. Hollander, and R. G. Bergman, J. Organomet. Chem., 1992, 428, 13. [124] G. Erker and F. Rosenfeldt, J. Organomet. Chem., 1982, 224, 29. [125] W. E. Hunter, D. C. Hrncir, R. V. Bynum, R. A. Penttila, and J. L. Atwood, Organometallics, 1983, 2, 750. [126] J. A. Labinger, D. W. Hart, W. E. Seibert, III, and J. Schwartz, J. Am. Chem. Soc., 1975, 97, 3851. [127] P. C. Wailes, H. Weigold, and A. P. Bell, J. Organomet. Chem., 1972, 43, C32. [128] T. F. Blackburn, J. A. Labinger, and J. Schwartz, Tetrahedron Lett., 1975, 3041. [129] P. C. Wailes, H. Weigold, and A. P. Bell, J. Organomet. Chem., 1971, 33, 181. [130] P. C. Wailes, H. Weigold, and A. P. Bell, J. Organomet. Chem., 1971, 34, 155. [131] Y. Yamamoto and K. Muruyama, Tetrahedron Lett., 1981, 22, 2895. [132] K. Mashima, H. Yasuda, K. Asami, and A. Nakamura, Chem. Lett., 1983, 219. [133] H. Yasuda, Y. Kajihara, K. Nagasuna, K. Mashima, and A. Nakamura, Chem. Lett., 1981, 719. [134] M. Akita, H. Yasuda, and A. Nakamura, Chem. Lett., 1983, 217. [135] H. Maeta, T. Hashimoto, T. Hasegawa, and K. Suzuki, Tetrahedron Lett., 1992, 33, 5965. [136] K. Suzuki, T. Hasegawa, T. Imai, H. Maeta, and S. Ohba, Tetrahedron, 1995, 51, 4483. [137] P. Wipf and W. Xu, J. Org. Chem., 1993, 58, 825. [138] R. F. Jordan, C. S. Bajgur, R. Willett, and B. Scott, J. Am. Chem. Soc., 1986, 108, 7410. [139] C. A. Bertelo and J. Schwartz, J. Am. Chem. Soc., 1975, 97, 228. [140] J. M. Manriquez, D. R. McAlister, R. D. Sanner, and J. E. Bercaw, J. Am. Chem. Soc., 1978, 100, 2716. [141] G. Erker and F. Rosenfeldt, J. Organomet. Chem., 1980, 188, C1. [142] E. Negishi, D. R. Swanson, and S. R. Miller, Tetrahedron Lett., 1988, 29, 1631. [143] S. L. Buchwald and S. J. La Maire, Tetrahedron Lett., 1987, 28, 295. [144] E. A. Mintz, A. S. Ward, and D. S. Tice, Organometallics, 1985, 4, 1308. [145] E. Negishi, K. Akiyoshi, B. O’Connor, K. Takagi, and G. Wu, J. Am. Chem. Soc., 1989, 111, 3089. [146] T. Yoshida and E. Negishi, J. Am. Chem. Soc., 1981,103, 1276. [147] T. E. Cole, R. Quintanilla, and S. Rodewald, Organometallics, 1991, 10, 3777. [148] T. E. Cole, S. Rodewald, and C. L. Watson, Tetrahedron Lett., 1992, 33, 5295. [149] T. E. Cole, R. Quintanilla, B. M. Smith, and D. Hurst, Tetrahedron Lett., 1992, 33, 2761. See also R. Quintanilla and T. E. Cole, Tetrahedron, 1995, 51, 4297. [150] T. E. Cole and R. Quintanilla, J. Org. Chem., 1992, 57, 7366. [151] S. Y. Lee and R. G. Bergman, Tetrahedron, 1995, 51, 4255. [152] R. A. Budnik and J. K. Kochi, J. Organomet. Chem., 1976, 116, C3.
316 [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195]
5
Zirconocenes
M. D. Fryzuk, G. S. Bates, and C. Stone, Tetrahedron Lett., 1986, 27, 1537. S. Kim and K. H. Kim, Tetrahedron Lett., 1995, 36, 3725. M. Yoshifuji, M. J. Loots, and J. Schwartz, Tetrahedron Lett., 1977, 1303. P. Wipf and W. Xu, Tetrahedron Lett., 1994, 35, 5197. P. Wipf and W. Xu, Org. Synth., 1997, 74, 205. P. Knochel and R. D. Singer, Chem. Rev., 1993, 93, 2117. B. H. Lipshutz and E. L. Ellsworth, J. Am. Chem. Soc., 1990, 112, 7440. B. H. Lipshutz and R. Keil, J. Am. Chem. Soc., 1992, 114, 7919. B. Zheng and M. Srebnik, J. Org. Chem., 1995, 60, 3278. See also S. Pereira and M. Srebnik, Organometallics, 1995, 14, 3127. E. Negishi and D. E. Van Horn, J. Am. Chem. Soc., 1977, 99, 3168. N. Okukado, D. E. Van Horn, W. L. Klima, and E. Negishi, Tetrahedron Lett., 1978, 1027. E. Negishi, N. Okukado, A. O. King, D. E. Van Horn, and B. I. Spiegel, J. Am. Chem. Soc., 1978, 100, 2254. M. J. Loots and J. Schwartz, J. Am. Chem. Soc., 1977, 99, 8045. M. J. Loots and J. Schwartz, Tetrahedron Lett., 1978, 4381. J. Schwartz, M. J. Loots, and H. Kosugi, J. Am. Chem. Soc., 1980, 102, 1333. B. H. Lipshutz and M. Segi, Tetrahedron, 1995, 51, 4407. D. E. Van Horn, L. F. Valente, M. J. Idacavage, and E. Negishi, J. Organomet. Chem., 1978, 156, C20. E. Negishi, D. E. Van Horn, T. Yoshida, and C. L. Rand, Organometallics, 1983, 2, 563. C. L. Rand, D. E. Van Horn, M. W. Moore, and E. Negishi, J. Org. Chem., 1981, 46, 4093. E. Negishi, L. D. Boardman, J. M. Tour, H. Sawada, and C. L. Rand, J. Am. Chem. Soc., 1983, 105, 5344. L. D. Boardman, V. Bagheri, H. Sawada, and E. Negishi, J. Am. Chem. Soc., 1984, 106, 6105. E. Negishi, L. D. Boardman, H. Sawada, V. Bagheri, A. T. Stoll, J. M. Tour, and C. L. Rand, J. Am. Chem. Soc., 1988, 110, 5383. E. Negishi, F. Liu, D. Choueiry, M. M. Mohamud, A. Silveira, Jr., and M. Reeves, J. Org. Chem., 1996, 61, 8325. F. Liu and E. Negishi, Tetrahedron Lett., 1997, 38, 1149. S. Ma and E. Negishi, J. Org. Chem., 1997, 62, 784. T. Mole and E. A. Jeffery, Organoaluminum Compounds, Elsevier, Amsterdam, 1972, 465 pp. J. A. Miller and E. Negishi, Tetrahedron Lett., 1984, 25, 5863. E. Negishi, D. E. Van Horn, A. O. King, and N. Okukado, Synthesis, 1979, 501. E. Negishi, K. P. Jadhav, and N. Daotien, Tetrahedron Lett., 1982, 23, 2085. N. Okukado and E. Negishi, Tetrahedron Lett., 1978, 2357. M. Kobayashi, L. F. Valente, E. Negishi, W. Patterson, and A. Silveira, Jr., Synthesis, 1980, 1034. E. Negishi and S. Baba, J. Chem. Soc., Chem. Comm., 1976, 596. E. Negishi, L. F. Valente, and M. Kobayashi, J. Am. Chem. Soc., 1980, 102, 3298. S. Baba and E. Negishi, J. Am. Chem. Soc., 1976, 98, 6729. E. Negishi, F. T. Luo, and C. L. Rand, Tetrahedron Lett., 1982, 23, 27. H. Matsushita and E. Negishi, J. Am. Chem. Soc., 1981, 103, 2882. E. Negishi, H. Matsushita, and N. Okukado, Tetrahedron Lett., 1981, 32, 2715. E. Negishi, V. Bagheri, S. Chatterjee, F. T. Luo, J. A. Miller, and A. T. Stoll, Tetrahedron Lett., 1983, 24, 5181. E. Negishi and Z. Owczarczyk, Tetrahedron Lett., 1991, 32, 6683. D. R. Williams, B. A. Barner, K. Nishitani, and J. G. Phillips, J. Am. Chem. Soc., 1982, 104, 4708. S. Fung and J. B. Siddall, J. Am. Chem. Soc., 1980, 102, 6580. B. H. Lipshutz, G. Bulow, R. F. Lowe, and K. L. Stevens, J. Am. Chem. Soc., 1996, 118, 5512. D. Y. Kondakov and E. Negishi, J. Am. Chem. Soc., 1995, 117, 10771.
References
317
[196] G. Erker, M. Aulbach, M. Knickmeier, D. Wingbermuhle, C. Kruger, M. Nolte, and S. Werner, J. Am. Chem. Soc., 1993, 115, 4590. [197] V. M. Dzhemilev, A. G. Ibragimov, A. P. Zoltarev, R. R. Muslukhov, and G. A. Tolstikov, Izv. Akad. Nauk SSSR, Ser. Khim., (Engl. Transl.), 1989, 194; 1991, 2570. [198] D. Y. Kondakov and E. Negishi, J. Am. Chem. Soc., 1996, 118, 1577. [199] J. P. Morken, M. T. Didiuk, and A. H. Hoveyda, J. Am. Chem. Soc., 1993, 115, 6997. [200] L. Bell, R. J. Whitby, R. V. H. Jones, and M. C. H. Standen, Tetrahedron Lett., 1996, 37, 7139. [201] D. R. McAlister, D. K. Erwin, and J. E. Bercaw, J. Am. Chem. Soc., 1978, 100, 5966. [202] J. E. Bercaw, D. H. Berry, A. J. Jircitano, and K. B. Mertes, J. Am. Chem. Soc., 1982, 104, 4712. [203] G. Erker, J. Organomet. Chem., 1977, 134, 189. [204] G. Erker and K. Kropp, J. Organomet. Chem., 1980, 194, 45. [205] F. Rosenfeldt and G. Erker, Tetrahedron Lett., 1980, 21, 1637. [206] G. Erker, K. Kropp, J. L. Atwood, and W. E. Hunter, Organometallics, 1983, 2, 1555. [207] G. Erker, K. Engel, U. Dorf, J. L. Atwood, and W. E. Hunter, Angew. Chem., Int. Ed. Engl., 1982, 21, 914. [208] G. Erker, K. Engel, J. L. Atwood, and W. E. Hunter, Angew. Chem., Int. Ed. Engl., 1983, 22, 494. [209] G. Erker and U. Dorf, Angew. Chem., Int. Ed. Engl., 1983, 22, 777. [210] J. Yin, K. Abboud, and W. M. Jones, J. Am. Chem. Soc., 1993, 115, 3810. [211] J. Yin, K. Abboud, and W. M. Jones, J. Am. Chem. Soc., 1993, 115, 8859. [212] J. Yin and W. M. Jones, Tetrahedron, 1995, 51, 4395. [213] E. Negishi, D. R. Swanson, and T. Takahashi, J. Chem. Soc., Chem. Comm., 1990, 18, 1254. [214] E. Negishi, T. Nguyen, J. P. Maye, D. Choueiry, N. Suzuki, and T. Takahashi, Chem. Lett., 1992, 2367. [215] K. Takagi, C. J. Rousset, and E. Negishi, J. Am. Chem. Soc., 1991, 113, 1440. [216] D. R. Swanson, C. J. Rousset, E. Negishi, T. Takahashi, T. Seki, M. Saburi, and Y. Uchida, J. Org. Chem., 1989, 54, 3521. [217] M. Jensen and T. Livinghouse, J. Am. Chem. Soc., 1989, 111, 4495. [218] W. A. Nugent and D. F. Taber, J. Am. Chem. Soc., 1989, 111, 6435. [219] M. Akita, H. Yasuda, H. Yamamoto, and A. Nakamura, Polyhedron, 1991, 10, 1. [220] E. Negishi, D. R. Swanson, F. E. Cederbaum, and T. Takahashi, Tetrahedron Lett., 1987, 28, 917. [221] T. V. RajanBabu, W. A. Nugent, D. F. Taber, and P. J. Fagan, J. Am. Chem. Soc., 1988, 110, 7128. [222] W. A. Nugent, D. L. Thorn, and R. L. Harlow, J. Am. Chem. Soc., 1987, 109, 2788. [223] D. F. Taber and J. P. Louey, Tetrahedron, 1995, 51, 4495. [224] E. Negishi and S. R. Miller, J. Org. Chem., 1989, 54, 6014. [225] T. Takahashi, T. Seki, Y. Nitto, M. Saburi, C. J. Rousset, and E. Negishi, J. Am. Chem. Soc., 1991, 113, 6266. [226] K. Kropp and G. Erker, Organometallics, 1982, 1, 1246. [227] G. Erker, U. Dorf, P. Czisch, and J. L. Petersen, Organometallics, 1986, 5, 668. [228] T. Takahashi, M. Kageyama, V. Denisov, R. Hara, and E. Negishi, Tetrahedron Lett., 1993, 34, 687. [229] T. Takahashi, Z. Xi, C. J. Rousset, and N. Suzuki, Chem. Lett., 1993, 1001. [230] H. G. Alt and M. D. Rausch, J. Am. Chem. Soc., 1974, 96, 5936. [231] S. Thaneder and M. F. Farona, J. Organomet. Chem., 1982, 235, 65. [232] S. L. Buchwald and R. B. Nielsen, J. Am. Chem. Soc., 1989, 111, 2870. [233] R. Hara, Z. Xi, M. Kotora, C. Xi, and T. Takahashi, Chem. Lett., 1996, 1003. [234] U. Rosenthal, A. Ohff, W. Baumann, A. Tillack, H. Görls, V. B. Burlakov, and V. B. Shur, J. Organomet. Chem., 1994, 484, 203. [235] V. Skibbe and G. Erker, J. Organomet. Chem., 1983, 241, 15.
318
5
Zirconocenes
[236] T. Takahashi, N. Suzuki, M. Hasegawa, Y. Nito, K. Aoyagi, and M. Saburi, Chem. Lett., 1992, 331. See also N. Suzuki, D. Y. Kondakov, M. Kageyama, M. Kotora, R. Hara, and T. Takahashi, Tetrahedron, 1995, 51, 4519. [237] C. Lefeber, A. A. Tillack, W. Baumann, R. Kempe, V. V. Burlakov, and U. Rosenthal, Organometallics, 1995, 14, 3090. [238] I. U. Khand, G. R. Knox, P. L. Pauson, W. E. Watts, and M. I. Foreman, J. Chem. Soc., Perkin Trans. 1, 1973, 977. [239] N. E. Schore in Comprehensive Organic Synthesis, Vol. 5, L. A. Paquette, Ed., Pergamon Press, Oxford, 1991, 1037. [240] B. Y. Lee, Y. K. Chung, N. Jeong, Y. Lee, and S. H. Hwang, J. Am. Chem. Soc., 1994, 116, 8793. [241] E. C. Lund and T. Livinghouse, J. Org. Chem., 1989, 54, 4487. See also B. L. Pagenkopf, E. C. Lund, and T. Livinghouse, Tetrahedron, 1995, 51, 4421. [242] G. Agnel and E. Negishi, J. Am. Chem. Soc., 1991, 113, 7424. [243] E. Negishi, D. Choueiry, T. B. Nguyen, D. R. Swanson, N. Suzuki, and T. Takahashi, J. Am. Chem. Soc., 1994, 116, 9751. [244] P. A. Wender and F. E. McDonald, J. Am. Chem. Soc., 1990, 112, 4956. [245] G. Agnel, Z. Owczarczyk, and E. Negishi, Tetrahedron Lett., 1992, 33, 1543. [246] M. Mori, N. Uesaka, and M. Shibasaki, J. Org. Chem., 1992, 57, 3519. See also (a) M. Mori, F. Saitoh, N. Uesake, K. Okamura, and T. Date, J. Org. Chem., 1994, 59, 4993. (b) N. Uesaka, F. Saitoh, M. Mori, M. Shibasaki, K. Okamura, and T. Date, J. Org. Chem., 1994, 59, 5633. (c) M. Mori, N. Uesaka, F. Saitoh, and M. Shibasaki, J. Org. Chem., 1994, 59, 5643. (d) T. Honda, S. Satoh, and M. Mori, Organometallics, 1995, 14, 1548. (e) F. Saitoh, M. Mori, K. Okamura, and T. Date, Tetrahedron, 1995, 51, 4439. [247] E. Negishi, S. Ma, T. Sugihara, and Y. Noda, J. Org. Chem., 1997, 62, 1922. [248] J. P. Maye and E. Negishi, Tetrahedron Lett., 1993, 34, 3359. [249] E. Negishi, J. P. Maye, and D. Choueiry, Tetrahedron, 1995, 51, 4447. [250] T. Takahashi, K. Kasai, N. Suzuki, K. Nakajima, and E. Negishi, Organometallics, 1994, 13, 3413. [251] C. Cop∫ret, E. Negishi, Z. Xi, and T. Takahashi, Tetrahedron Lett., 1994, 35, 695. [252] T. Takahashi, D. Y. Kondakov, Z. Xi, and N. Suzuki, J. Am. Chem. Soc., 1995, 117, 5871. [253] T. Takahashi, T. Fujimori, T. Seki, M. Saburi, Y. Uchida, C. J. Rousset, and E. Negishi, J. Chem. Soc., Chem. Comm., 1990, 18, 182. [254] D. P. Lewis, R. J. Whitby, and R. V. H. Jones, Tetrahedron, 1995, 51, 4541. [255] A. H. Hoveyda and Z. Xu, J. Am. Chem. Soc., 1991, 113, 5079. [256] K. S. Knight and R. M. Waymouth, J. Am. Chem. Soc., 1991, 113, 6268. [257] D. P. Lewis, D. M. Muller, R. J. Whitby, and R. V. H. Jones, Tetrahedron Lett., 1991, 32, 6797. [258] F. N. Tebbe, G. W. Parshall, and G. S. Reddy, J. Am. Chem. Soc., 1978, 100, 3611. [259] K. A. Brown-Wensley, S. L. Buchwald, L. Cannizzo, L. Clawson, S. Ho, D. Meinhardt, J. R. Stille, D. Straus, and R. H. Grubbs, Pure Appl. Chem., 1983, 55, 1733. [260] K. C. Ott, E. J. M. deBoer, and R. H. Grubbs, Organometallics, 1984, 3, 223. [261] C. J. Rousset, E. Negishi, N. Suzuki, and T. Takahashi, Tetrahedron Lett., 1992, 33, 1965. [262] E. Negishi, C. J. Rousset, D. Choueiry, J. P. Maye, N. Suzuki, and T. Takahashi, Inorg. Chem. Acta, submitted. [263] U. Wischmeyer, K. S. Knight, and R. M. Waymouth, Tetrahedron Lett., 1992, 33, 7735. [264] K. S. Knight, D. Wang, R. M. Waymouth, and J. Ziller, J. Am. Chem. Soc., 1994, 116, 1845. [265] N. Uesaka, M. Mori, K. Okamura, and T. Date, J. Org. Chem., 1994, 59, 4542. [266] K. H. Shaugnessy and R. M. Waymouth, J. Am. Chem. Soc., 1995, 117, 5873. [267] N. Suzuki, D. Y. Kondakov, and T. Takahashi, J. Am. Chem. Soc., 1993, 115, 8485. [268] K. S. Knight and R. M. Waymouth, Organometallics, 1994, 13, 2575. [269] T. Takahashi, K. Aoyagi, V. Denisov, N. Suzuki, D. Choueiry, and E. Negishi, Tetrahedron Lett., 1993, 34, 8301.
References
319
[270] U. M. Dzhemilev, A. G. Ibragimov, A. P. Zoltarev, R. R. Muslukov, and G. A. Tolstikov, Izv. Akad. Nauk SSSR, Ser. Khim., 1991, 2570. [271] H. Sinn and G. Oppermann, Angew. Chem., Int. Ed. Engl., 1966, 5, 962. [272] H. Sinn and E. Kolk, J. Organomet. Chem., 1966, 6, 373. [273] W. Kaminsky and H. Sinn, Liebigs Ann. Chem., 1975, 424. [274] W. Kaminsky and H. Sinn, Liebigs Ann. Chem., 1975, 438. [275] W. Kaminsky, J. Kopf, H. Sinn, and H. Vollmer, Angew. Chem., Int. Ed. Engl., 1976, 15, 629. [276] T. Takahashi, N. Suzuki, M. Kageyama, Y. Nitto, M. Saburi, and E. Negishi, Chem. Lett., 1991, 1579. [277] S. Couturier and B. Gautheron, J. Organomet. Chem., 1978, 157, C61. [278] R. Choukroun, M. B. Bert, and D. Gervais, J. Chem. Soc., Chem. Comm., 1986, 1317. [279] R. Waymouth and P. Pino, J. Am. Chem. Soc., 1990, 112, 4911. [280] T. Takahashi, M. Hasegawa, N. Suzuki, M. Saburi, C. J. Rousset, P. E. Fanwick, and E. Negishi, J. Am. Chem. Soc., 1991, 113, 8564. [281] M. R. Kesti, M. Abdulrahman, and R. M. Waymouth, J. Organomet. Chem., 1991, 417, C12. [282] M. R. Kesti and R. M. Waymouth, Organometallics, 1992, 11, 1095. [283] H. G. Woo, R. H. Heyn, and T. D. Tilley, J. Am. Chem. Soc., 1992, 114, 5698. [284] T. D. Tilley, Acc. Chem. Res., 1993, 26, 22. [285] F. W. Hartner, Jr., J. Schwartz, and S. M. Clift, J. Am. Chem. Soc., 1983, 105, 640. [286] F. W. Hartner, Jr., S. M. Clift, J. Schwartz, Organometallics, 1987, 6, 1346. [287] R. Beckhaus, Angew. Chem., Int. Ed. Engl.,1997, 36, 686. [288] P. J. Walsch, F. J. Hollander, and R. G. Bergman, J. Am. Chem. Soc., 1988, 110, 8729. [289] P. J. Walsh, F. J. Hollander, and R. G. Bergman, J. Am. Chem. Soc., 1990, 112, 894. [290] P. J. Walsh, A. M. Baranger, and R. G. Bergman, J. Am. Chem. Soc., 1992, 114, 1708. [291] A. M. Baranger, P. J. Walsh, and R. G. Bergman, J. Am. Chem. Soc., 1993, 115, 2753. [292] P. J. Walsh, F. J. Hollander, and R. G. Bergman, Organometallics, 1993, 12, 3705. [293] K. E. Meyer, P. J. Walsh, and R. G. Bergman, J. Am. Chem. Soc., 1994, 116, 2669. [294] K. E. Meyer, M. J. Walsh, and R. G. Bergman, J. Am. Chem. Soc., 1995, 117, 974. [295] S. Y. Lee and R. G. Bergman, J. Am. Chem. Soc., 1995, 117, 5877. [296] S. Y. Lee and R. G. Bergman, J. Am. Chem. Soc., 1996, 118, 6396.
320
5
Zirconocenes
6
Group 5 and Group 6 Metallocenes Pascual Royo and Evelyn Ryan
6.1 Introduction In this Chapter the chemistry of group 5 and 6 bis(cyclopentadienyl) complexes is presented, summarizing those synthetic, structural and reactivity aspects which illustrate the chemical behavior of these types of compounds. This study does not pretend to be an exhaustive collection of literature references on this subject, although the references given cover the period up to the end of 1996. The reader is advised to consult the excellent reviews published in the first and second editions of Comprehensive Organometallic Chemistry for more detailed lists of references and previous contributions in the field [1 , 2 ]. A more recent review on ‘open shell’ organometallics, also including sandwich and half-sandwich cyclopentadienyl metal derivatives, has been published by Poli [3 ], and other publications [4 ] reviewing particular topics include compounds of the types described in this Chapter. According to the latest IUPAC recommendations [5 ] we use the name ‘metallocene’ only to describe real ‘sandwich’ metal complexes containing two parallel cyclopentadienyl rings coordinated to the central atom, and no other component, although even here the two cyclopentadienyl rings are not exactly parallel but slightly bent due to attractive van der Waals forces between the rings [6 , 7 ]. All those compounds formed by coordination of one or more additional substituents, which show structures with bent cyclopentadienyl rings are called ‘bentbis(cyclopentadienyl)’ or simply ‘bis(cyclopentadienyl)’ metal complexes (the CpM-Cp bend angle, θ, being that formed at the metal centre by the two perpendiculars from the centroids of the cyclopentadienyl rings). The theoretical explanation of the behavior of the bis(cyclopentadienyl) compounds with bend angles between 180° and 120° is based on the Lahuer and Hoffman calculations [8 ], lately extended to compounds with narrower Cp–M–Cp angles down to 110° [9 , 10]. Most of the compounds described in this Chapter have metal centers with 15 to 18 electrons. As shown in Figure 6-1, twelve of these electrons are accommodated in the six low-lying, essentially cyclopentadienide ligand bonding orbitals, therefore the crucial orbitals which determine the structure of these complexes are the three e2 and a1 orbitals of the metallocene compounds, which provide the 1a1, b2 and 2a1 orbitals for the bent-bis(cyclopentadienyl) complexes. Differences may be interpreted in terms of the different occupation of these three mainly metal-based
322
6
Group 5 and 6 Metallocenes
orbitals by the remaining 1 to 6 electrons. We also consider where appropriate some metal systems with a higher total number of electrons, particularly those with 19. Energy differences between these important three orbitals depend not only on the bend angle but also on factors associated with the nature of the metal, its oxidation state or valence number, and the electron-donating or -withdrawing character of the substituents bonded directly to the metal and indirectly to the cyclopentadienyl rings. The spin-pairing energy required to obtain a lower spin-state system, when possible, is therefore also related to these factors. A useful new approach to the formal classification of covalent compounds has been published by M. L. H. Green [11 ]. In this Chapter [MCp2] is a core fragment common to all complexes; the ligand notations defined by Green are generally used and similar methodology is followed to derive anionic and cationic complexes from their neutral precursors. Fundamental differences between compounds are related to the number of substituents bonded to the [MCp2] fragment and to the bonding capacity of the ligands. Compounds discussed here are therefore classified accordingly in four types: Type A compounds are metallocenes with no additional ligand, which are structurally and chemically different from the bent-bis(cyclopentadienyl) complexes; compounds containing mono-functional single-atom donor ligands are classed as Type B, including therefore complexes with: Lewis acids, mono-functional mono-, biand tri-dentate X ligands, and single-atom donor neutral mono-, bi- and tri-dentate L ligands; complexes with bi-functional single-atom donor ligands such as =CR2, =NR (linear and bent), =O, =S, =C=CR2 and related systems are identified as Type C; complexes formed by side-bonded η2 and η3 ligands appear as Type D. Compounds of Types C and D are presented separately although they belong to some of the three groups studied for compounds of Type B and could be treated on the same basis. Formal dn configurations, oxidation states or valence numbers are frequently given, although sometimes these do not provide a complete, unequivocal description, and further discussion is included where necessary to aid good understanding of the real chemical behavior.
6.2 Type A Metallocene Compounds Compounds of this stoichiometry, free from additional substituents bonded to the metal, have structure A (Figure 6-1) and are unlike any other type of complex described in this Chapter. They are structurally classified as true metallocenes containing parallel or almost parallel cyclopentadienyl rings in a typical sandwich disposition. The e2 orbitals are degenerate and lower in energy than the a1 orbital for all the group 5 and 6 metal complexes (Figure 6-1).
6.2 Type A Metallocene Compounds
323
Figure 6-1 Qualitative variation of available molecular orbitals on going from metallocenes to bent-dicyclopentadienyl metal complexes
The neutral group 5 and 6 metallocenes [MCp2] with 15 and 16 electrons respectively are discussed in Sect. 6.2.1. The related anionic and cationic metallocenes, with a total number of electrons between 15 and 17, obtained from redox reactions of the neutral compounds, are discussed in Sect. 6.2.2.
6.2.1 Neutral Metallocenes 6.2.1.1 Group 5 Metallocenes [MCp2] Synthesis Most of the synthetic methods used to obtain the 15-electron vanadocenes [VCp2] (1–3) involve metathesis of vanadium(II) halides using alkali cyclopentadienides to transfer the cyclopentadienyl rings (Scheme 6-1).
324
6
Group 5 and 6 Metallocenes
Vanadium(II) species used in situ were obtained by reducing [VCl3(THF)3] with agents such as Li[AlH4] [12 ], Zn [13 , 14 ] or more frequently the alkali metal cyclopentadienide when used in excess under reflux [15 , 16 ]. Other vanadium(II) compounds such as [VCl2(THF)6] or better [V2Cl3(THF)6]2 [Zn2Cl6] [17 ] have also been used. Reduction of [VCp2X2] and [VCp2X] (X = Cl, Me) with Li[AlH4] [18 ] or aluminum slurries [19 ] prepared by metal vapor condensation with organic solvent has been used to prepare vanadocene. Complexes containing mixed cyclopentadienyl rings were isolated by the metathetical reaction of [CpVCl(THF)] with alkali metal cyclopentadienides [20 ].
Scheme 6-1
The reported vanadocenes include compounds containing variously substituted cyclopentadienyl rings such as C5H5, C5D5, C5H4R (R = Me, Pri, Bun, But, Et), 1Me-3-Ph-C5H3, 1,3-Ph2-C5H3, C5Me5, C5Me4Et, C5H4EMe3 (E = C, Si, Ge, Sn) and η5-indenyl (2) [1,2,16]. The dinuclear bis(fulvalene) derivative
6.2 Type A Metallocene Compounds
325
[V2(fulvalene)2] (3) [21 ], a dark purple pyrophoric solid, and complexes containing mixed cyclopentadienyl rings [VCp(C5H4R)] (R = CMe2Ph, CHMePh) have been isolated. All of these compounds were thermally stable, air sensitive, red to purple oils or crystals; those containing substituted rings had lower melting points and were more soluble and more air sensitive. The niobium and tantalum derivatives were less stable. Niobocene was formed by H abstraction and spontaneous H2 reductive elimination from [NbCp2H3] [22 ] and by reduction of [NbCp2Cl2] with sodium naphthalenide [23 ]. Substituted niobocenes (4A), formed in solution [24 ] by reduction of their dihalides with 2 equiv. of Na/Hg, are thought to exist in equilibrium with their fulvene isomers (4B) (Scheme 6-2). Their structures were assigned on the basis of their EPR spectra, and reaction of the permethylated niobocene with sulfur, trapping both forms in the products, has provided chemical evidence for these structures [25 ].
Scheme 6-2 Structure Spin-restricted and spin-polarized calculations26 with S=3/2 made for [VCp2] by SCF-DV-Xα methods resulted in an e2g-a1g energy separation (∆ = 0.25 eV) in agreement with experimental values determined by optical spectroscopy [27 ] and the observed magnetic moment of 3.78 µB, consistent with the presence of three unpaired electrons. Similarly, three unpaired electrons in [V(C5Me5)2] consistent with a 4A ground state were suggested by magnetic susceptibility measurements and EPR spectroscopy. [15]. Paramagnetic 1H and 13C NMR spectra [28 ] and UV–Vis spectra [14] have been recorded for the EMe3-ring-substituted vanadocenes. However, the dinuclear difulvalene complex 3 was reported as diamagnetic. The nature of the ring disorder causing the disagreement between previous Xray [29 ] and electron diffraction [30 ] studies of [VCp2] was explained by an Xray structural determination in the temperature interval 108–357 K [31 ]. A crystallographic determination of [V(C5Me5)2] has shown a sandwich structure with par-
326
6
Group 5 and 6 Metallocenes
allel planar cyclopentadienyl rings at an average V–Cring distance of 1.91(5)Å [15], and a similar sandwich disposition was found for [V(C9H7)2] [17], although the presence of two different orientations in the crystal lattice prevented its resolution. All these vanadocenes are very reactive as expected for 15-electron metallocene compounds [2]. They behave as carbene-like fragments, which readily undergo oxidative addition to many organic functional groups. They form unstable complexes with N2, O2, H2 and methane which can only be studied at low temperature and which are easily and rapidly oxidized to higher oxidation states [32 ]. In all of these reactions, they lose their sandwich structure through the introduction of additional substituents in the coordination sphere, giving bent bis(cyclopentadienyl) complexes. Niobocene has a simple ten-line EPR spectrum in THF with ANb = 103 G, which excludes the possibility of solvent coordination [22]. The permethylated niobocenes [Nb(C5Me4R)2] (R = Me, Et) have more complex EPR spectra which indicate the presence of a metallocene species (4A) with ANb = 100 G, together with a second isomer (4B) which results from the C–H activation and oxidative addition of one of the methyl cyclopentadienyl groups [24]. Both isomers shown in Scheme 6-2 were characterized by their chemical reactivity, described in later sections [33 ]. Applications Vanadocene and related compounds are catalysts for acetylene [34 ] and 1,3-diene [35 ] polymerization and for deuteration [36 ] of CH4 by D2 or C2D4. 6.2.1.2 Group 6 Metallocenes [MCp2] Synthesis The 16-electron neutral chromocene derivatives, [CrCp2] (5) are usually obtained [1,2] by reaction of a previously reduced chromium(II) species such as CrCl2 or [Cr2(µ-OAc)4] with alkali metal cyclopentadienides (Scheme 6-3). Used in excess under reflux, these cyclopentadienides can also act as reducing agents for [CrCl3(THF)3]. Similar reactions with appropriate alkali cyclopentadienide salts have been used to prepare ring-substituted [37 , 38 ], indenyl (6) [39 ], and fulvalene dinuclear (7) [40 ] derivatives, and related complexes containing bridged bis(cyclopentadienyl) ligands (8) [41 ].
6.2 Type A Metallocene Compounds
327
Scheme 6-3
Reported complexes include those with two equal or mixed ring-substituted ligands such as C5H5, C5H4R, C5H4EMe3 (E = Si, Ge, Sn), C5HPh4, C5Me5, C5Me4Et, η5indenyl, C10H8 (fulvalene) and µ-[(SiMe2)2(C5H3)2].2 The chromium derivatives, obtained as red solids, are more accessible, thermally more stable and less reactive than those of the heavier metals. Photochemically induced elimination of neutral ligands from [MCp2(CO)] (M = Mo, W) and [WCp2(C2H4)], or similar reductive elimination from [MCp2H2] (M = Mo, W), [WCp2(CH3)H] or [WCp2(Bcat)H] (Bcat = C6H4O2), has been used to obtain the heavier metallocenes [MCp2] (M = Mo, W) which have only been studied in low temperature matrices or as transient species. [42 , 43 , 44 , 45 ].
Structure Chromocenes are stable enough to allow their investigation by different structural techniques and chemical reactions. The orbitally degenerate ground state of 16electron chromocenes leads to temperature-dependent magnetic moments greater than the spin-only value, although the orbital angular momentum is partly quenched by Jahn–Teller effect. Unsubstituted chromocene has a magnetic moment [46 ] of 3.27 µB and values of between 2.90 and 3.20 µB have been found [38] for
328
6
Group 5 and 6 Metallocenes
substituted chromocenes. Various studies conclude [46, 47 ] that the triplet ground state for chromocenes is 3E2g and corresponds to a e2g3a1g1 configuration, leading to a significant orbital contribution. Their structural behavior has been studied by IR [48 ], gas phase absorption [49 ] and electron transmission [50 ] spectroscopy. Coupling of the spins of both chromocene halves due to metal–metal interaction rather than superexchange has been demonstrated by variable temperature 1H NMR for the difulvalene dinuclear complex (7) [40]. The 1H and 13C NMR spectra observed for variously substituted mononuclear complexes and the dinuclear complex 8 give broad and paramagnetically shifted resonances [51 , 52 ] and the 29Si paramagnetic shift has been measured for [Cr(C5H4SiMe3)2] [53 ]. The molecular structure of chromocene has been determined by electron diffraction [30] and X-ray diffraction analysis [54 ] and variously substituted chromocenes have also been characterized by X-ray diffraction studies [38, 41]. All of these reported structures show a sandwich disposition with metal–ring-centroid distances of around 1.80 Å. A single crystal X-ray diffraction analysis of [{Cr(Indenyl)2}2] shows it to be a dimer containing η5- and µ-η3-indenyl groups with metal-ring-centroid distances of 1.999 Å and 2.127 Å respectively [55 ]. The molybdenum and tungsten compounds have only been studied when trapped in inert argon matrices at low temperature, in order to prevent their spontaneous transformation into bent bis(cyclopentadienyl) compounds with η1–η5-bridging rings or fulvene structures. The expected two unpaired electrons for the 16electron molybdenum and tungsten complexes [MCp2] (M = Mo, W), corresponding to a 3E2g ground state, have been inferred from their UV and magnetic circular dichroism spectra [42–44]. Chromocene is a very reactive compound which loses its cyclopentadienyl rings easily when treated with diols, carboxylic acids, acetylacetone or ethylenediamine [56 , 57 ] and has a low capacity to coordinate additional ligands [58 ].
Applications The heterogeneous catalyst generated by Union Carbide [59 ] by reacting [CrCp2] with silica requires neither a cocatalyst nor an activator, and displays high activity in ethylene polymerization at low pressure and low temperature. The system is very sensitive to hydrogen transfer to the polymer chain giving polyethylene of specific melt index. Although many analytical and spectroscopic studies have been carried out, the structure of the active species and its oxidation state are not known. It has been demonstrated however that only some mono(cyclopentadienyl) chromium(III) species are catalytically active in solution [60 ]. Recent solid-state NMR studies conclude that the catalyst contains at least one mononuclear and two dinuclear surface-attached chromium complexes with one or more cyclopentadienyl rings [61]; molecular model compounds and NMR spectra have been used to analyze their nature.
6.2 Type A Metallocene Compounds
329
6.2.2 Ionic Metallocenes A summary of the redox behavior of vanadocenes and chromocenes is shown in Scheme 6-4.
Scheme 6-4
6.2.2.1 Group 5 Cationic Metallocenes [MCp2]+ 14-electron [MCp2]+ group 5 metallocene cations do not exist and all attempts made to oxidize the neutral metallocenes failed [14, 62 ]. Similarly, no base-free [MCp2]2+ dicationic 14-electron group 6 metallocenes nor any base-free cationic 13-electron complex of either group 5 or group 6 metals have ever been observed. The formation of the unstable 13-electron dication [Cp2V]2+ was proposed by rapid disproportionation of the cation [VCp2Cl]+ resulting from the electrochemical oxidation (–0.08 V) of [Cp2VCl] [63 ]. Exceptionally, the diamagnetic difulvalene complex [V2(fulvalene)2] is oxidized by one equivalent of [FeCp2][PF6] to yield purple black crystals of the very air sensitive salt [V2(fulvalene)2][PF6], which is paramagnetic with one unpaired electron [64]; its EPR spectrum shows that the free electron is coupled to only one
330
6
Group 5 and 6 Metallocenes
metal atom. Oxidation of this product by a second equivalent gives the diamagnetic, air sensitive purple salt [V2(fulvalene)2][PF6]2. 6.2.2.2 Group 5 Anionic Metallocenes [MCp2]– Reduction of vanadocene by potassium gives K[VCp2] [21], a highly reactive 16electron metallocene which easily loses one cyclopentadienyl ligand when it reacts with CO to give the more stable 18-electron monocyclopentadienyl derivative [VCp(CO)4] [65 ]. K[VCp2] is a paramagnetic compound probably having the same spin triplet ground state 3E2g as the isoelectronic chromocene derivative. The related anionic vanadium(I) mixed ring sandwich compound [V(η5-Ind)(η6-naphthalene)]– was obtained via the reduction of [V(Ind)2] with potassium naphthalenide [66 ]. Similar mixed-ring neutral [VCp(η6-Arene)] complexes were isolated by reaction of [(VCp)2(µ-benzene)] with LiCp followed by addition of alcohol [67 ]. 6.2.2.3 Group 6 Cationic Metallocenes [MCp2]+ A quasi-reversible oxidation to the 15-electron chromocenium cation [CrCp2]+ with a potential of Ep = –0.67 V was observed in an electrochemical study of chromocene [CrCp2] [68 ]. Stable solutions can be prepared by electrochemical oxidation in the presence of PF6–. Salts of solid paramagnetic [CrCp2]+ with three unpaired electrons were also isolated with various anions; stability was dependent on the anion. The iodide obtained by chemical oxidation with allyl iodide is pyrophoric whereas the salts of [CrCpCl3]– and [CrCp(CO3)]– are stable and the molecular structure of the latter was determined by X-ray diffraction [69 ]. Greater stability is conferred by bulky cyclopentadienyl ligands such as C5Me5 and C5Me4Et and the salts can be isolated by oxidation of the appropriate chromocenes by AgPF6 [14]. The chromocenium cation has a spin quartet ground state and its NMR 1H and 13C paramagnetic shifts have been determined [62]. Related molybdenum and tungsten derivatives are less known, a remarkable exception being [Mo(C5Ph5)2]+ , isolated as its Br3- partially spin-paired paramagnetic salt (µeff = 3.5 µB) [70 ]. 6.2.2.4 Group 6 Anionic Metallocenes [MCp2]– Electrochemical reduction of chromocene shows quasi-reversible formation of the anionic compound [CrCp2]– which is persistent on the electrochemical time scale [68]. Reduction and reoxidation potentials of EP = –2.44 V and EP = –2.26 V were measured, however none of these 17-electron, anionic group 6 metallocenes have ever been isolated.
6.3 Type B-Complexes with mono-functional ligands
331
6.3 Type B-Bent bis(Cyclopentadienyl) Metal Complexes with Mono-Functional Single-Atom Donor Monoanionic and Neutral Ligands Bonding of one or more additional substituents to the metal center leads to structures with bent bis(cyclopentadienyl) rings (Figure 6-1B). The bonding capacity of the bent [MCp2] fragment, and therefore the structural and chemical behavior of these compounds, was rationalized by Lahuer and Hoffmann [8] and this theoretical study is used to classify the different types of complexes described in the following sections.
6.3.1 Complexes with One Coordinated Ligand When only one additional monoanionic (X) or neutral (L) ligand is coordinated to the metal center the resulting species for group 5 and 6 metals belong to the following types (Table 6-1):
Table 6-1. Different types of group 5 and 6 [MCp2X] and [MCp2L] metal complexes Total Number of electrons
16
Configuration
d2
d3
d4
[(M5)Cp2X]
[(M5)Cp2L] [(M6)Cp2X]
[(M6)Cp2L]
[(M5)Cp2L]+ [(M6)Cp2X]+
17
18
X]-
[(M5)Cp2 [(M6)Cp2L]+
[(M6)Cp2X][(M5)Cp2L]-
332
6
Group 5 and 6 Metallocenes
18-Electron Compounds When only one additional σ bonding ligand is coordinated to a metal with a C2v symmetry, the 2a1 metal based orbital is used for bonding interaction with the σ orbital of the ligand, whereas the other two orbitals 1a1 and b2 remain essentially non-bonding, occupied by the d4 metal electrons. This has been predicted as the most stable structure for d4 and high-spin d2 metal complexes [8]. However other less symmetrical structures with the ligand located in the equatorial plane but forming an angle α with the twofold axis (9) are favoured for d0, d1, low-spin d2 and d3 metal compounds,8 although it has been calculated for d0 complexes that the distorsion α angle also depends on the snd3-n ground-state configuration of the metal [71].
Figure 6-2 Qualitative diagram for [MCp2L] d4 metal complexes with σ-donor and π-acceptor ligands
6.3 Type B-Complexes with mono-functional ligands
333
Electron-rich d4 metal complexes are stabilized when the coordinated substituent is a π-acceptor ligand, whose empty π* orbital of b2 symmetry can participate in a π-bonding interaction with a metal orbital of the same symmetry, releasing part of the electron density of the metal (Figure 6-2). It should be expected therefore that formation of d4 metal compounds of this type will be favored by πacid ligands which can alleviate the excess electron density on the metal by adopting a symmetrical disposition of the ligand in the equatorial plane of the cyclopentadienyl rings.
6.3.1.1 Neutral Group 6 [MCp2L] Metal Complexes As mentioned in Section 6.2.1, chromocene has a low capacity to coordinate one additional ligand,58 which would lead to an 18 electron bent bis(cyclopentadienyl) system such as [Cp2Cr(CO)] (10). The estimated formation constant of this adduct is K ≈ 50 atm-1 at 0°C and 2% of uncomplexed chromocene is always present. This resistance is mainly due to the energy required to bend the rings (21 kcal · mol-1) [72], therefore factors which reduce ring-ring repulsive forces by distorsion of the ground-state geometry will favour coordination and for this reason the incorporation of a heteroannular ethylene bridge between the rings (11) provides a more stable CO adduct [35]. Complex (11) and the related tertiary-butyl isocyanide adduct were conveniently prepared via the reaction of the calcium ansa-metallocene with chromium dichloride in the presence of the ligand [73].
Related molybdenum and tungsten complexes are not accessible by this direct preparative method due to the thermal instability of their metallocene precursors. The carbonyl adducts [MCp2(CO)] (M = Mo, W) were isolated by trapping the transient metallocene (obtained by the reduction of [MCp2Cl2] with Na/Hg) [74] with CO, or by thermal or photochemical decomposition of the dihydrides [MCp2H2] or alkyl derivatives in the presence of CO. They are resistant to further carbonylation except in the presence of oxidants to give [WCp(η3-Cp)(CO)2], where formation of an otherwise unstable 20 electron system is avoided by slippage of one of the cyclopentadienyl rings [75 ,76]. The same behaviour was also found for [Mo(η5-
334
6
Group 5 and 6 Metallocenes
Indenyl)2(dppe)],37 for which the exchange between the two η5-η3 coordination modes is rapid even at low temperature. Reaction of [WCp2(CO)] with carbon tetrachloride results in electrophilic addition to one of the rings to give [WCp(η4C5H5-exo-CCl3)Cl(CO)] [77]. Related phosphine adducts were obtained by similar reactions and [MoCp2(PR3)] was isolated by deprotonation of the cationic hydride [MoCp2H(PR3)]+ [78]. They are labile compounds in which the ligand is easily displaced when treated with ketenes [79]. The isocyanide complexes [MoCp2(CNR)] (R = Me, Et, But) were isolated by reduction of the cationic metal(IV) isocyanide derivatives [MoCp2X(CNR)]+ (X= halide, H, Me) with Na/ Hg, or by substitution in the side-bonded acetonitrile compound [MoCp2(η2-NCR)] [80 ]. The x-ray molecular structure of the t-butyl derivative reveals that the ligand is symmetrically located in the equatorial plane and the θ bend angle of the cyclopentadienyl rings is very open (147.7°), typical for this type of complex and even larger than values observed for related compounds (142.8° for the corresponding Mo-PMe3 complex). The extensive bending of the isocyanide ligand with a C-N-C angle of 139.5(4)°, the short Mo-C and the long C-N bond distances of 1.997(4) Å and 1.193(4) Å respectively are consistent with a strong π-back bonding interaction.
6.3.1.2 Anionic Group 5 [MCp2L]-and Group 6 [MCp2X]- Metal Complexes Although the coordination of one π-acceptor ligand to an anionic vanadocene [VCp2]- species would presumably lead to a stable 18-electron vanadium(I) [VCp2L]- system, reaction of the electron-rich metallocene with neutral ligands always takes place with loss of one of the cyclopentadienyl ligands. The neutral adducts [VCp2L] (L = CO, CNCy) undergo reversible one-electron reduction to the anionic species [VCp2L]- as evidenced by cyclic voltametry, but they cannot be obtained on bulk electrolytic reduction, which only leads to their decomposition to give vanadium metal [1]. The isoelectronic group 6 anionic species [MCp2X]- have only seldom been observed since they easily lose the X- ligand to give the more stable neutral metallocene. A rare example of this type of species is the molybdenum hydride derivative [K(18-crown-6)(µ-Η)MoCp2], which results from reaction of [MoCp2H2] with KH in the presence of 18-crown-6, and whose x-ray structure shows a bend angle for the cyclopentadienyl rings of 153.26° with a hydride Mo-H-K bridge described as a two-electron three-center covalent bond [81]. Related species with π-acid ligands [MCp2L]- and more than 18-electrons have however been observed. An electrochemical study of chromocene under CO reveals that reduction to the 19-electron anionic species [Cp2Cr(CO)]- is slightly more demanding than that of the sandwich chromocene, with a reduction potential EP = -2.54 V. The resulting anion is short lived and substitution of one ring by CO
6.3 Type B-Complexes with mono-functional ligands
335
[68] occurs rapidly to give [CpCr(CO)3]-. Similar behaviour was also observed for the ansa- complex containing a Me2C-CMe2 bridge between the two cyclopentadienyl rings, which decays giving an irreversible reduction, although none of these compounds were isolated.
17-Electron Compounds 17-electron complexes with only one additional mono-functional substituent coordinated to the metal centre are essentially represented by the group 5 d3 metal derivatives [MCp2L] with neutral ligands L, whereas the related d3 group 6 metal compounds [MCp2X] with one monodentate anionic ligand X are rather rare and scarcely studied. In addition we mention some aspects of the chemistry of related anionic group 5 [MCp2X]- and cationic group 6 [MCp2L]+ derivatives. For d3 metal compounds with one σ-bonding ligand which uses the a1 orbital to accommodate the σ-bonding electron-pair, the three electrons have to occupy the 1a1 and b2 orbitals (Figure 6-2) to give paramagnetic compounds with one unpaired electron and a fundamental spin-doublet state; these are coordinatively unsaturated and electron-rich systems. When a π-acceptor ligand is used the empty orbital of the ligand can interact with the occupied b2 metal orbital providing stronger metal-ligand bonds and equally leaving one unpaired electron (Figure 63).
Figure 6-3 Electron configuration for [MCp2L] d3 metal complexes with π-acceptor ligands
336
6
Group 5 and 6 Metallocenes
6.3.1.3 Neutral Group 5 [MCp2L] and Group 6 [MCp2X] Metal Complexes We have seen that vanadocene has basic character and behaves as a carbene-like unit. It therefore resists coordination of nucleophilic σ-donor ligands, but reacts easily with π-acceptor ligands leading to paramagnetic 17-electron bis(cyclopentadienyl) vanadium complexes [VCp2L]. As a carbene unit it reacts easily with many organic unsaturated and functionalized groups to give the corresponding oxidative addition d1 metal products which are discussed in Section 6.5.2. In this section we are concerned with essentially d3 complexes containing monodentate, usually π-acid, ligands. [VCp2] reacts rapidly with CO at room temperature to give the adduct [VCp2(CO)] (12), a dark brown paramagnetic solid with one unpaired electron (µeff = 1.76 µB) [82]; (12) has a very low ν(C-O) stretching frequency of 1881 cm-1 (Nujol), which is indicative of a strong V-C bond with a high level of π-back donation. The more electron donating pentamethylcyclopentadienyl ligands increase the basicity of the related [VCp*2] complex [15], which undergoes the same reaction with CO leading to a similar paramagnetic red-maroon compound (µeff = 1.71 µB), with an even lower ν(C-O) stretching frequency of 1845 cm-1 (Nujol). At low CO pressure, both compounds dissociate CO in an equilibrium with the sandwich metallocene [83]. The EPR spectra of their 13C-CO derivatives show hyperfine couplings which demonstrate that the unpaired electron is delocalized onto the 13CO ligand. This is consistent with the SCF-Xα-DV calculations which show a 2A1 ground state with an a1 type HOMO [83], the odd electron occupying a hybrid d orbital localized between the cyclopentadienyl rings. Similar carbonylation takes place with the dinuclear bis(fulvalene) complex leading to green crystals of [V2(CO)2(fulvalene)2] (13) whose IR spectrum suggests a cis- dicarbonyl structure [1]. A similar reaction between CO and [V(Indenyl)2] affords a structurally characterized dicarbonyl compound in which ring slippage of one of the indenyl ligands occurs to give η3-coordination, in order to avoid a formally 19-electron configuration [84]. The x-ray molecular structure of [VCp*2(CO)] shows an average V-Cp distance of 1.928 Å with a θ bend angle of 153.6° and a very short V-CO bond distance of 1.879 Å, with the linear carbonyl ligand symmetrically placed in the equatorial plane. Similar reactions with isocyanides have also been studied but do not result in simple coordination of the ligand [15].
6.3 Type B-Complexes with mono-functional ligands
337
The 17-electron cyclohexylisocyanide complex (14) is formed as a yellow-brown paramagnetic (µeff = 1.78 µB) solid by coordination of the ligand as the first step in a reaction which proceeds with coordination of a second molecule of the ligand and further dealkylation of the isocyanide moiety to finally give the stable 18electron [VCp*2(CN)(CNR)] complexes (Scheme 6-5). This transformation is more rapid with CNBut, for which the intermediate adduct could not be isolated. Evidence for the formation of a 19-electron intermediate species [VCp*2L2] is supported by the related reaction of [VCp*2(CO)] with CNBut in which the CO adduct [VCp*2(CN)(CO)] is formed. The greater lability of the carbonyl ligand facilitates its dissociation to give the final 16-electron compound [VCp*2(CN)].
Scheme 6-5
338
6
Group 5 and 6 Metallocenes
Similar complexes [VCp(η5-C5H4R)(CO)] with mixed cyclopentadienyl ligands have been reported as the products of the thermal decomposition of 18-electron monocyclopentadienyl complexes [VCp(η4-C5H5R) (CO)2] (R = allyl, phenyl) [85]. The reactions proceed with dissociation of CO, elimination of dihydrogen and transformation of the η4- into the η5-coordinated ligand. Related niobium and tantalum complexes are not known because the corresponding niobocene and tantalocene are not accessible. Recently, a niobium derivative of this type [NbCp2(CO)] has been reported as an intermediate species observed in the reduction of [NbCp2Cl(CO)] with excess sodium amalgam [86]. Voltametric studies of the 18-electron niobium(III) complexes [NbCp2ClL] (L = CO, PMe3, CNBut) showed irreversible reduction waves with peak potentials of -2.36, 3.02 and -2.69 V respectively. The ESR spectrum of the reduced carbonyl species showed a 10-line signal centred at g = 2.097 with splitting due to 93Nb and
Nb = 1.72 G, consistent with the presence of the radical [NbCp2(CO)] which is persistent for 5-10 min. at ambient temperature. Similar reactions are observed for complexes with L = PMe3 and CNBut; they react finally with Hg to give Nb-Hg complexes. An unusual, crystallographically characterized, molydenum complex (17) obtained from the reaction of (15) with ethylene at 1 atm was reported (Scheme 6-6) [87]. An unexpected feature of the reaction was the formation of the second fulvalene unit which required the removal of one equivalent of dihydrogen, thought to be via hydrogenation of ethylene, and (16) was a proposed intermediate. The MoMo distance was slightly longer than expected for a metal-metal single bond. The same reaction using 10 atm of ethylene gave (18) and no (17) was detected.
Scheme 6-6
6.3 Type B-Complexes with mono-functional ligands
339
6.3.1.4 Anionic Group 5 [MCp2X]- and Cationic Group 6 [MCp2L]+ Metal Complexes Anionic group 5 d3 metal compounds [MCp2X]- do not exist . It has been observed that the electrochemical reduction of [VCp2Cl] (-1.63 V) is followed by rapid loss of chloride and conversion into vanadocene. Although the neutral group 6 d3 metal compounds [MCp2X] could not be isolated, formation of their related cationic species [MCp2L]+ has been observed in electrochemical studies. Oxidation of neutral chromocene under CO leads to the formation of the 17-electron cationic adduct which instantaneously decays to the sandwich cation, losing CO (see section 6.2.2.3) [68]. The measured potential difference implies that the equilibrium constant for the formation of [CrCp2(CO)]+ is about 1000 times smaller than that for the sandwich complex, and therefore only a small fraction of the carbonyl cation remains in solution under 1 atm of CO. For reasons already discussed the ansa- bis(cyclopentadienyl) carbonyl cation [Cr{(C5H4)CMe2CMe2 (C5H4)}(CO)]+ does not lose CO on the electrochemical time scale even in the absence of an excess of CO, although attempts to isolate salts of this cation were unsuccessful [88]. Similar molybdenum and tungsten derivatives have not been isolated. 16-Electron Compounds An important group of compounds of this type are the neutral group 5 d2 metal complexes [MCp2X] with one anionic (X) ligand. Related ionic compounds are the cationic group 5 [MCp2L]+ and group 6 [MCp2X]+ complexes with neutral (L) or anionic (X) ligands respectively. As shown in Figure 6-2, one of the three equatorial metal-based orbitals (2a1) of the [MCp2] fragment becomes a molecular bonding orbital used to coordinate the anionic or neutral substituent. The two (d2) electrons for [MCp2X] and [MCp2L]+ group 5 metal compounds can occupy the other two (1a1 and b2) orbitals in three different ways, leading to two possible spin states: singlet and triplet (Figure 6-4).
Figure 6-4 Electron configurations for [MCp2X] and [MCp2L]+ d2 metal complexes with σand π-donor/acceptor ligands
340
6
Group 5 and 6 Metallocenes
Occupation of these two orbitals depends on their relative energy, which is drastically changed if one of them (b2) is used for π-bonding with π-donor or π-acceptor ligands. The use of a single σ-donor ligand would lead to paramagnetic 16-electron compounds (A) with two unpaired electrons in the 1a1 and b2 orbitals. The use of one σ-donor and strong π-donor (3 electron donor) ligand would give an additional stabilized occupied b2 orbital, leading to 18-electron compounds (B). The use of one σ- donor and strong π-acceptor ligand would stabilize the metal-based b2 orbital which would then be occupied by the two (d2) electrons to give diamagnetic 16electron compounds (C), which could be easily converted into 18-electron systems similar to (B), occupying the vacant 1a1 orbital by coordination of a second preferably π-acceptor ligand. This is the basis of the very different chemical behaviour observed for bis(cyclopentadienyl) vanadium(III) derivatives compared with the related niobium and tantalum(III) compounds. With σ-donors, vanadium shows a marked tendency to form paramagnetic compounds with two unpaired electrons, and strong π-donor ligands are required to form diamagnetic 18-electron systems. However, when a strong π-acceptor ligand L is used, the spin-singlet state is accessible, but then a second ligand coordinates to achieve the more favourable 18 electron system. Similar behaviour is also observed for niobium and tantalum(III) compounds but, in contrast, they show a higher tendency to give diamagnetic spin-singlet compounds even when σ-donors are present. From extended Hückel calculations [8] the most stable minimum energy structure for paramagnetic compounds locates the substituent in the symmetrically central position of the equatorial plane, with an angle α = 0 with the C2 axis of the [MCp2] fragment. The presence of a free pair of electrons in compounds exhibiting a singlet-spin state would tend to distort this geometry and to locate the substituent at an angle α > 0. From a structural point of view, all the vanadium(III) compounds reported exhibit the same bent bis(cyclopentadienyl) disposition with the substituent occupying the central position in the equatorial plane, and this is a very important, abundantly represented, group of vanadium(III) 16-electron compounds. The estimated energy gap between the two orbitals is narrow (0.3 to 0.6 eV), making spin-pairing very unfavourable in these complexes [89]. Furthermore, their tendency to coordinate one additional ligand to give the 18-electron derivatives is less pronounced than for the heavier niobium and tantalum metals, because the required spin-pairing energy is too unfavourable.
6.3.1.5 Neutral d2 Vanadium Compounds [VCp2X] The simplest of this group are the neutral complexes which result from the coordination of only one monodentate anionic ligand.
6.3 Type B-Complexes with mono-functional ligands
341
Vanadium(III) halides are useful starting materials for a wide variety of other derivatives. Most synthetic methods which use these halides are based on the oneelectron oxidation of the neutral sandwich vanadocenes with various oxidizing agents. Reactions of [VCp2] with alkyl halides, methylene dihalides, hydrogen chloride, halogen (Br2, I2), lead dichloride and [VCp2X2] have been extensively used to prepare monohalides or pseudohalides [VCp2X] (19) [1,2,90]. A particularly interesting method discussed in Section 6.3.1.3 is the oxidative dealkylation of the 17-electron isocyanide complexes [VCp*2(CNR)] on addition of CO [15]. More recently, PCl3 has been reported as a convenient chlorinating agent leading to mixtures of alkyl-substituted bis(cyclopentadienyl) vanadium(III) and (IV) halides in ratios which depend on the bulkiness of the alkyl substituent [16]. An alternative method (Scheme 6-7) uses TlCp to transfer two cyclopentadienyl rings to VCl3 [91]. Some particular complexes containing substituted cyclopentadienyl ligands have also been synthesized by reacting VBr3(THF)3 with an appropriate lithium cyclopentadienide [37].
Scheme 6-7
All of these halides are very air sensitive and reactive paramagnetic compounds with two unpaired electrons and little tendency to dimerize; their unfavourable electron-pairing energy hinders the coordination of one additional ligand to give diamagnetic 18-electron compounds. The alkyl complexes [VCp2R] (20) are one of the products of the reaction of vanadocene with alkyl halides, together with the corresponding vanadium(III) halide, but a better preparative method for (20) is alkylation of the monohalides [VCp2Cl], or the dihalides [VCp2Cl2] which are simultaneously reduced (Scheme 6-8).
342
6
Group 5 and 6 Metallocenes
Scheme 6-8 A great variety of alkyl, aryl, alkenyl and alkynyl complexes with unsubstituted, substituted and permethylated cyclopentadienyl rings have been isolated [1,2,3]. The hydride and deuteride complexes (21) of the permethylated rings have also been obtained by hydrogenolysis of the methyl complex [92], whereas its reaction with Li[BEt3H] proceeds with reduction to the vanadocene. All of these compounds are highly coloured, air sensitive paramagnetic solids with magnetic moments consistent with the presence of two unpaired electrons, obeying the Curie-Weiss law. They have the predicted bent cyclopentadienyl symmetrical geometry which has been found for the cyanide derivative [VCp*2(CN)] by X-ray diffraction [15], showing an average V-Cp* distance of 1.97 Å with a Cp*-V-Cp* angle of 151.5 ° and the linear cyanide ligand symmetrically disposed in the equatorial plane; the V-CN bond distance of 2.07 Å indicates a weak double bond and the triple C-N bond distance is 1.17 Å. The chemical behaviour of compounds with typical σ-donor ligands such as alkyls and hydrides is the best illustration of the reluctance of vanadium to give spin-singlet systems by coordinating an additional ligand to afford diamagnetic 18-electron compounds, and is supported by the following observations: a) Alkyl radicals with potential π-systems are always coordinated as η1-ligands occupying only one coordination site; this is the behaviour observed for allyl and cyclopentadienyl ligands [93]. An excellent demonstration of this behaviour was reported by Nieman and Teuben in the study of the reactivity of the allyl [VCp2(η1-allyl)] and alkyl [VCp2R] (R = alkyl, phenyl) derivatives [85]. They all coordinate ligands with good acceptor properties, such as CO, CNR and CO2, but they do not react with nitriles due to their limited acceptor properties. All of these compounds react with their coordinated isocyanides or CO2 with migration of the appropriate allyl, alkyl and phenyl groups to give the η1-coordinated acyl, iminoacyl (22) and carboxylate (23) complexes; the iminoacyl-nitrogen and the carboxylate-oxygen donor atoms lack acceptor properties to induce the spin-pairing which would lead to their η2-coordination. The same reaction also occurs with CO to give the η1-acyl complex, but a second molecule of the strong π-acceptor ligand CO is coordinated to give the 18-electron species [VCp2{η1-C(O)R}(CO)].
6.3 Type B-Complexes with mono-functional ligands
343
b) Alkyl radicals with substituted donor groups able to act as chelate or bridging bidentate ligands, such as o-aminephenyl and o-aminebenzyl, give only mononuclear paramagnetic species (µeff. = 2.6 µB), and their proposed stability cannot be due to any interaction with the amino substituent [94]. c) Alkyls with β-hydrogen containing radicals are thermally stable and do not transform into the olefin hydride by β-elimination [95], and the related permethylated complex [V(C5Me5)2H] does not react with ethylene at pressures up to 17 atm [92]. d) The unique examples of transformation into diamagnetic 18-electron compounds are the reactions with CO discussed in Section 6.3.2.1. e) Without exception, the hydride and all of the alkyl vanadium(III) compounds reported are paramagnetic with magnetic moments of 2.77 µB for [V(C5Me5)2Me] and 2.81 µB for [V(C5Me5)2H] [92]. Vanadium(III) complexes with potential π-donor anionic ligands such as [VCp2(SR)] (24) have also been isolated by one-electron oxidation of vanadocenes with RSSR; they too are paramagnetic compounds [96].
Exceptional behaviour is observed for stannyl and germyl derivatives (25) which are obtained by reaction of vanadocene with an excess of SnHEt3 or [Cd(GeR3)2] (R = Et, Ph) respectively [1]. They can also be prepared by reaction of [VCp2(CH2SiMe3)2] or [VCp2Me2] with HER3 (E = Si, Ge), and by reacting [VCp2Cl] with Li[GePri3]. Unusually, these compounds are monomeric and diamagnetic.
344
6
Group 5 and 6 Metallocenes
Stronger π-donor ligands may force spin-pairing, as observed [1] for the bridging imido and terminal amido complexes. The dinuclear imido complex was isolated by one-electron oxidation of vanadocene with Si(N3)Me3, which takes place with evolution of N2 leading to the dinuclear complex [Cp2V{µ-N(SiMe3)}VCp2] (26), whereas the mononuclear amido derivative (27) is prepared by reaction of [VCp2(CH2SiMe3)] with NH(SiMe3)2 or [VCp2Cl] with Li[N(SiMe3)2]. These seem to be the only π-donor ligands able to give diamagnetic compounds [97]. 6.3.1.6 Neutral d2 Niobium and Tantalum Compounds [(Nb,Ta)Cp2X] In contrast to the vanadium(III) complexes described above, the bis(cyclopentadienyl)niobium and tantalum(III) complexes show a greater tendency to adopt an 18-electron singlet-spin state, not requiring the cooperation of strong acceptor ligands because simple σ-donor ligands can produce the spin-pairing required to give stable diamagnetic complexes. Similar behaviour is observed with π-acceptor ligands, which stabilize the b2 orbital, favouring the 18-electron spinpaired state. Formation of a true 16-electron [NbCp2Me] (28) paramagnetic species occurs by photolysis of a toluene solution of [NbCp2Me(CO)] [98] ; the dark blue solution is fairly stable at room temperature but decomposes on heating, possibly via a rearrangement to a methylidene hydride, behaviour typically observed for [MCp2R] (M = Nb, Ta) complexes.
6.3 Type B-Complexes with mono-functional ligands
345
Preparation of metal(III) halides by reduction of THF suspensions of [Cp2MCl2] with Na/Hg has been reported [99 ,100 ,101 ,102]. In solution they must exist as adducts with one coordinated solvent molecule which is easily lost by evaporation under vacuum, whereas in the solid they are probably dimeric compounds with µhalide bridges [103]. The increased lipophilicity conferred by cyclopentadienyl rings with bulky TMS substituents gave compounds which were soluble in non-donor solvents, where they probably also exist as dimers, although molecular weights were not measured [104 ,105 ,106]. These metal(III) halides have a great tendency to coordinate one additional ligand, and several metal halides can even accomplish this task through the formation of halide bridges [102]. They cannot therefore be considered as 16-electron d2 metal complexes and are better studied in the following sections 6.3.2.1. The niobium and tantalum hydrides [MCp2H] (M = Nb, Ta) have only been generated photochemically and studied by IR and UV spectroscopy in matrices at low temperature, although their multiplicity state could not be determined [107]. A related niobium(III) hydride has been proposed as consistent with the electrochemical behaviour observed for [Nb(C5H4SiMe3)2H3] [108]. Singular examples are the niobium(III/II) complexes [NbCp]2(µ-fulvalene)(µX)] (29) (X = Cl, Br, SCN) for which the x-ray crystal structure of the chloro derivative has been determined [109].
6.3.1.7 Cationic Group 5 d2 Metal Compounds [MCp2L]+ The monocationic species [Cp2ML]+ (M = V, Nb, Ta) are further examples of 16electron group 5 metal complexes containing only one additional neutral (L) ligand. These compounds are very rare due to the even higher tendency of the metal to coordinate one additional ligand to give the more stable 18-electron compounds [Cp2ML2]+ discussed in Section 6.3.2.1. The use of σ-donor ligands such as acetone, ether, pyridine and acetonitrile gives the green, very air sensitive and paramagnetic cationic 16-electron complexes [VCp2L]+ (30) with two unpaired electrons. These were obtained from the reaction of neutral bis(cyclopentadienyl) vanadium(III) chloride [VCp2Cl] with NaBPh4 in the presence of the ligands [110], or from the oxidation of either neutral sandwich vanadocene or decamethylvanadocene by [FeCp2]PF6 [14].
346
6
Group 5 and 6 Metallocenes
Oxidation of the dinuclear bis(fulvalene) complex with [FeCp2][PF6] in acetonitrile gives air sensitive brown crystals of [V2(NCMe)2(fulvalene)2][PF6]2 (31) which is a paramagnetic compound with µeff = 3.0 µB; its x-ray crystal structure shows a VV distance of 3.329 Å, slightly longer than that expected for a single bond.1 Two related carbonyl cations are the brown paramagnetic [V2(CO)2(fulvalene)2]+ (µeff = 1.7 µB) and the orange dication [V2(CO)2(fulvalene)2]2+ which has a low measured magnetic moment (µeff = 0.9-2.1 µB), probably due to decomposition [1]. However, in all other cases the use of π-acceptor ligands such as CO, isocyanides and phosphines always leads to diamagnetic 18-electron derivatives. Similar species are also present in solution when vanadocene is electrochemically oxidized in acetonitrile, as evidenced by a reversible one-electron wave, whereas the same oxidation of decamethylvanadocene gives a cyclic voltammogram with no reversible waves [14].
6.3.1.8 Cationic Group 6 d2 Metal Compounds [MCp2X]+ These cationic species cannot be isolated for the reasons given for the group 5 metal derivatives. In a recent report on 18-electron cationic tungsten complexes, the 16-electron cations were proposed as transient species [111]. The intermediate reversible formation of [WCp2H]+ was demonstrated by the complete exchange of coordinated acetonitrile in [WCp2H(NCMe)]+ when treated with deuterated acetonitrile. Similarly, formation of an intermediate species [WCp2Me]+ was proposed to explain the quantitative transformation of [WCp2Me(NCMe)]+ via C-H bond activation of one of the cyclopentadienyl rings, and [WCp2Et]+ was proposed to be the intermediate insertion product of the reaction between [WCp2H(η2-C2H4)] and I2 to give [WCp2(=CHCH3)I]+ [112] Formation of a similar molybdenum salt [MoCp2H]+ salt has been suggested to explain the deprotonation of [MoCp2H2] by CH2=PMe3, which finally gives [MoCp2(PMe3)] [113]. However, these types of cationic compound have never been isolated for any of the group 6 metals.
6.3.2 Complexes with Two Coordinated Ligands All the possible formulations and electron configurations for group 5 and 6 metal compounds containing two additional anionic (X) or neutral (L) ligands with between 16- and 18-electrons are summarized in Table 6-2. All of the d0 metal compounds are cationic derivatives of the higher valent metals. Neutral complexes fall into three important groups: i) d2 group 5, ii) d2 group 6 and iii) d1 group 5 metal complexes. In this section we are mainly concerned with these complexes and their ionic derivatives, although other 16-electron compounds are also discussed.
6.3 Type B-Complexes with mono-functional ligands
347
Table 6-2. Different types of [MCp2L2], [MCp2XL] and [MCp2X2] groups 5 and 6 metal complexes
Total Number of electrons
16
17
18
Configuration
d0
d1
d2
[(M5)Cp2X2]+ [(M5)Cp2XL]2+ [(M6)Cp2X2]2+
[(M5)Cp2X2]
[(M5)Cp2XL] [(M6)Cp2X2]
[(M5)Cp2XL]+ [(M5)Cp2L2]2+ [(M6)Cp2X2]+
[(M5)Cp2L2]+ [(M5)Cp2X2][(M6)Cp2XL]+ [(M6)Cp2L2]2+
When two additional σ-donor ligands are coordinated to the metal, contributing their a1+ b2 orbitals, two of the metal based orbitals 2a1 and b2 engage in the bonding interaction (Figure 6-5), whereas the nature and energy of the nonbonding a1 metal orbital is very sensitive to the angle ϕ between the two ligands (32) and their π-acceptor capacity [8]. ϕ values in the ranges 94° - 97° for d0, 85° - 88° for d1 and 76° - 82° for d2 metal complexes have been predicted, and are in reasonably good agreement with the observed values.
Figure 6-5 Qualitative diagram for group 5 [MCp2XL] and group 6 [MCp2X2] d2 metal complexes with σ-donor and π-acceptor ligands
348
6
Group 5 and 6 Metallocenes
18-Electron Compounds The simplest neutral complexes with 18 electrons are the d2 vanadium, niobium and tantalum(III) derivatives [Cp2MXL] containing either one anionic (X) and one neutral (L) monodentate ligand or one monoanionic bidentate ligand, and the d2 chromium, molybdenum and tungsten(IV) complexes [Cp2MX2] with either two monodentate anions or one bidentate dianion. In all of these diamagnetic compounds the nonbonding orbital a1 is occupied by one pair of electrons, which is still available for π-bonding when one of the substituents is a π-acceptor ligand, but no interaction can take place with π-donor ligands, which would orientate their occupied π-orbital perpendicular to the equatorial plane. On the other hand the free pair of a1 electrons can also be donated to an appropriate Lewis acid. Consequently these compounds can have a basic character and from an structural point of view the bend angle θ between the two cyclopentadienyl ligands varies between 120° and 150°.
6.3.2.1 Neutral Group 5 d2 Metal Compounds [MCp2XL] As discussed in Section 6.3.1, the central atom of 16-electron [Cp2MX] complexes possesses two available orbitals which can accommodate the remaining two electrons, leading to either a spin-singlet or -triplet system. An examination of the known chemistry of group 5 metals shows that the spin-pairing energy of the two metal-based orbitals is lower for the heavier niobium and tantalum atoms than for vanadium, although estimates of this energy have not been reported. In fact, most of the vanadium complexes show a very low tendency to coordinate one additional ligand, indicating that the difference in energy between the two metal orbitals is clearly smaller than the spin-pairing energy, whereas the reverse is true for the heavier metals for which the formation of [Cp2MXL] species is favoured, with the exception of the niobium and tantalum(III) derivatives described in Section 6.3.1. Only when strong π-acceptor ligands are used it is possible to isolate 18-electron vanadium [Cp2VXL] complexes, which easily revert to their paramagnetic precursors. Examples are the reactions with isocyanides (Scheme 6-5), which final-
6.3 Type B-Complexes with mono-functional ligands
349
ly give the cyanide complexes [Cp2V(CN)(CNR)] (33). However, isocyanides react with vanadium (III) halides [Cp2VX] to give cationic complexes, but neutral complexes cannot be isolated in this way. This is only possible when CO is used: [Cp*2VI] readily reacts with CO in nonpolar solvents to give [Cp*2VI(CO)] (34), which regenerates the iodide on gentle heating of the solution under vacuum [15]. Similarly, reaction of the hydride (21) or methyl (20) derivatives with CO leads respectively to the diamagnetic green hydrido [Cp*2VH(CO)] (34) or acyl carbonyl (35), which loses CO when stored at -40°C over several weeks [92].
The huge number of known niobium and tantalum [Cp2MXL] compounds demonstrates the great variety of substituents (Table 6-3) which include unsubstituted and substituted cyclopentadienyl rings and an extensive range of anionic X and neutral L ligands. In this section we are concerned only with complexes containing monofunctional single atom donor ligands whereas complexes with η2-coordinated ligands will be discussed in Section 6.5.
350
6
Group 5 and 6 Metallocenes
Table 6-3. Different types of substituents for [MCp2XL] group 5 metal complexes Cyclopentadienyl ligand
Anionic ligand
Neutral ligand
C5H5 C5H4Me C5Me5 C5H4SiMe3 C5H3(SiMe3)2 C5H4tBu C5H4CHMePh C5H4iPr Indenyl
H Cl Br NMe2 N(Tol)(CH=NTol) OR SR PR2 P(O)PhH P(X)R2 (X=O,S) µ-(PR2)M(CO)n µ-[P(S)R2]MLn µ-(OC)M(CO)n CN R SiMe2H SiMe3 SnR3 µ-SnCl2
CO CNR PR3 olefines alkynes CO2 CS2 isocyanate ketene ketenimine ketone aldehyde SR P(OR)3 µ-ClMLn
In all of these compounds the metal centre possesses one pair of free electrons which can be used for π-donation and therefore π-acid ligands are particularly effective in providing strong metal-ligand bonds. Inspection of Table 6-3 shows that most of the ligands used to stabilize this type of compound have empty acceptor orbitals and formation of π-bonds plays an important role. Halides When bis(cyclopentadienyl) metal(IV) halides were reduced in donor solvents (THF) in the absence of additional ligands, the resulting species [MCp2X(solvent)] (M = Nb, Ta) formed in solution could not be isolated as solvent removal under vacuum also eliminates the labile σ-coordinated ligand, giving compounds which probably reach the more stable 18-electron system by dimerization, although molecular weight measurements were not reported [100,102]. Evidence for the dimeric nature of such compounds was provided by an X-ray diffraction study of
6.3 Type B-Complexes with mono-functional ligands
351
[Nb(C5H4SiMe3)2Cl]2 (36), which shows the angle between the chlorine atoms ϕ ≈ 75° [114]. The role played by the second molecule coordinated through a µ-chloro bridge in the dimer can also be taken by other metal halides, indicating the high tendency of these compounds to achieve an 18 electron system. For example, reduction of [MCp2Cl2] (M = Nb, Ta) with one half equivalent of Na/Hg affords purple compounds [MCp2Cl(µ-ClMClCp2)] (37) which contain a µ-chloro bridging metal(IV) halide as ligand [99,102], and which can be transformed by reaction with AgClO4 into the cationic complexes proposed to have two µ-chloro bridges [MCp2(µ-Cl)2MCp2)]+ (38). The µ-chloro bridging ligand can be replaced by a heterometal halide which leads to different heterodinuclear complexes [MCp2(µCl)(µ-X)M’Y2)] (39) (X = Cl, Br; M’Y2 = AlCl2, FeCl2, TlCl2, Tl(C6F5)2).
Similarly, reduction of the bis(cyclopentadienyl) metal(IV) and (V) halides with various reducing agents in the presence of π-acceptor ligands is the method of synthesis generally used to obtain stable carbonyl, isocyanide, phosphine and phosphite halocomplexes [MCp2XL] (40) [L = CO, CNR, PR3, P(OR)3)] which are the most extensively studied compounds in this category [100, 106, 115, 116 , 117, 118, 119, 120, 121]; they can also be prepared by metathesis of the hydride via reaction with HCl, CHCl3, RCOCl and MeI (Scheme 6-9) [122 ,123].
352
6
Group 5 and 6 Metallocenes
Scheme 6-9 Hydrides Hydride complexes with η1-coordinated ligands [MCp2HL] (41) are generally obtained on treatment of the reactive trihydrides with the ligand. The mononuclear derivatives [MCp2H{R2P(CH2)nPR2}] (M = Nb or Ta, R = Me or Ph, n = 1 or 2) were obtained in this way whereas reaction of [MCp2H3] with dppp (dppp = Ph2P(CH2)3PPh2) gave the bimetallic dihydride [(NbCp2H)2(µ-dppp)] or [{TaCp2H}2] [124]. The ability of the mononuclear complexes to act as metalloligands which can coordinate to a second metal centre to give heterodinuclear complexes has been explored. They react with [M’(CO)4L2] (M = Cr, Mo, W; L= norbornadiene or piperidine) to give the unsymetrically bridged complexes (42). An alternative method extracts BH3 by reacting the boranate derivatives with NMe3 in the presence of the π-acid ligand (Scheme 6-10) [102,122]. This is also the method used to isolate the bis-indenyl niobium (43) derivatives containing different ligands such as PMe2Ph, pyridine [125], whereas the reaction with CO gives the mono-indenyl tetracarbonyl complex, and with allene gives the corresponding allyl derivative. [Nb{(η5-C5H4SiMe3)2}2H(CO)] forms adducts with coinage metal cations giving heteronuclear cationic complexes with µ-(hydrido-M’-hydrido) bridging the two Nb centres [126 ].
Scheme 6-10
6.3 Type B-Complexes with mono-functional ligands
353
Alkyls Two alternative methods may be used to prepare alkyl complexes [MCp2RL] (44) (Scheme 6-11). The reduction of haloalkylmetal(IV) complexes has been used to prepare some of these compounds [98,127], but has very few applications due to the limited accessibility of the monoalkylated derivatives.
Scheme 6-11
A more useful method is the alkylation of the halometal(III) complexes with various alkylating agents, usually Li and Mg reagents [117]. When CO and CNR are the ligands present in the coordination sphere, further migration of the alkyl group may take place to finally give the corresponding η2-acyl or η2-iminoacyl derivatives. This reaction is unfavourable because it involves substitution of a π-acceptor ligand CO or CNR by a σ-oxygen or nitrogen donor ligand, and has not been observed for carbonyl derivatives [118], however it readily occurs in isocyanide complexes [119].
354
6
Group 5 and 6 Metallocenes
Complexes with Other Anionic Ligands Related compounds with different anionic substituents such as -OR, -SR, -NR2, -NC, -SiR3, -SnR3, -PR2, -P(O)R2 and-P(S)R2 are also prepared by metathesis of the halides with appropriate reagents [2]. The reactivity of the trihydrides and hydrido-silyl and -stannyl derivatives with various ligands has been extensively studied. As shown in Scheme 6-12, hydrido-silyl and -stannyl complexes eliminate SiR3H or SnR3H when treated with ligands to give the corresponding adducts and this was the method used to prepare the silyl complex with PMe3 (45) [125]. The silyl complexes [TaCp2(L)(SiR3)] (L = PMe3, CO) activate the C-H bonds of unhindered arenes, reacting in neat arenes (benzene, toluene or m-xylene) to give equilibrium mixtures of aryl and silyl complexes; however, the sterically hindered complex [TaCp2(PMe3) (SiBut2H)] gave [TaCp2(PMe3)Ar] as the sole product of the reaction [128].
Scheme 6-12
New thiolate derivatives were synthesized in high yield by reaction of the iodo derivatives with alkali thiolates [123], and various phosphido derivatives [NbCp2(PR2)(CO)] were obtained by insertion of PR2Cl into Nb-H bonds of [NbCp2H(CO)] followed by reaction of the resulting cationic species (46) with KOH to give the phosphido complexes (47) (Scheme 6-13) [129, 130, 131, 132, 133, 134]. The phosphido group in these complexes is a σ-donor ligand and the orientation of the phosphorus lone pair is at ca. 74° with respect to the equatorial plane in order to minimise the electron repulsion, defined as the ‘gauche effect’ [135], of the occupied 1a1 metal orbital, which is partially used for π-back bonding to the carbonyl ligand. Related compounds with mixed cyclopentadienyl rings and chiral metal centres [136], and complexes derived from diphosphine precursors have also been reported [137]. Reactions of the phosphido complexes with H2O2 and S8 lead to similar compounds with -P(O)R2 and -P(S)R2 ligands (48) [134].
6.3 Type B-Complexes with mono-functional ligands
355
Scheme 6-13
This type of complex, with anionic ligands such as thiolate and phosphide which have one additional lone pair of electrons still available for bonding, has been extensively explored as a metallo-ligand able to coordinate to a second metal centre to give heterodinuclear complexes with Fe(CO)4, Fe(CO)3(PMe2Ph), M(CO)nL5-n (M = Cr, Mo, W), and MnCp(CO)2 with µ-phosphido and in some cases also with µ-carbonyl bridges [123, 131, 132, 134, 138, 138, 139, 140, 141, 142, 143]. The heterodinuclear complexes [MCp2(CO){M’(CO)n}], in which the role of the anion (X) is played by a metal-metal bonded system, are related compounds which were isolated by reaction of the trihydrides or carbonyl hydrides with the corresponding hetero-metal carbonyls [144, 145, 86]. They were also generated from photochemical reactions of the dihydride in the presence of the hetero-metal species [146]. All the above complexes contain one π-acceptor ligand which is required to withdraw the electron density of the metal lone pair and few examples of complexes without such π-acceptor ligands have been reported. The trimethylacetate complex [NbCp2(OOCCMe3)] was isolated by reaction of the boranate niobium(III) derivative with the acid, and was characterized by IR spectroscopy as a monomeric species containing a terminal η1-coordinated carboxylate group [147]. Its tendency to coordinate π-acid ligands explains its reactivity with diphenylacetylene to give an adduct characterized by X-ray diffraction [148], although it does not react with pyridine or triphenylphosphine. More recently the formate complex (49) was reported as an air sensitive oily product resulting from the reaction of the trihydride with CO2 [149]; similar acetate and acetylacetonate derivatives were prepared by metathesis of the chloride with the corresponding thallium salts and were char-
356
6
Group 5 and 6 Metallocenes
acterized as species containing chelate ligands, however transformation of the chelate into a monodentate ligand (50) readily takes place on addition of π-acceptor ligands such as CO, isocyanides and other η2-ligands (Scheme 6-14).
Scheme 6-14
6.3.2.2 Cationic Group 5 d2 Metal Compounds [MCp2L2]+ Replacement of the anionic ligand in the neutral group 5 metal complexes [MCp2XL] gives access to the related cationic species [MCp2L2]+. For vanadium, coordination of two neutral ligands to give cationic 18 electron systems is only possible when labile σ-donor anions and strong π-acceptor ligands are used [15,17]. [VCp2X] (Cp = C5H5, C5Me5; X = I, CN, SR) partially reacts with CO leading to an equilibrium between the adduct and the starting material. 18-electron compounds are only formed under appropriate conditions, and [V(Indenyl)2X] does not react at all with CO unless NaBPh4 is present. Niobium and tantalum have a higher tendency to coordinate neutral ligands in order to complete their 18-electron system; displacement of halide by CO or phosphine is not easy and cationic compounds are not directly accessible. The first of these compounds was the dicarbonyl complex (51) reported by Brintzinger [150] and several methods have been used to prepare complexes of this type, most of which contain η2-coordinated groups as π-acceptor ligands (Section 6.5). A convenient method involves protonation of the metal(III) phosphides [MCp2(PR2)L] (L = CO, PR3) with HCl, or methylation with MeI, to give cationic complexes [MCp2 (CO)(PHR2)]+ or [MCp2 (CO)(PMeR2)]+ respectively [131]. Alternatively, reduction of [MCp2Cl2] in the presence of dmpe with various reducing agents leads to the [MCp2 (dmpe)]+ complex [151].
6.3 Type B-Complexes with mono-functional ligands
357
An unusual example of this type of compound is the carbonyl vinylidene complex (52), which results from the oxidation of the acetylene niobium(III) derivative [NbCp2{η2-MeC≡C(COOMe)}] by [FeCp2][BPh4] via a mechanism proposed to involve intermediate oxidation to a niobium(V) species [152]. Oxidative loss of a methoxide group from the acetylene methylcarboxylate substituent generates the carbonyl ligand required to stabilize the coordinated vinylidene, which behaves as a π-acceptor ligand in the resulting cationic species. 6.3.2.3 Anionic Group 5 d2 Metal Compounds [MCp2X2]We have observed that strong π-acceptor ligands are generally required to alleviate the high electron density in these niobium and tantalum(III) complexes. Therefore, as expected, derivatives formed by two essentially σ-donor anionic ligands coordinated to the metal centre are unable to afford stable compounds. In fact most of the reported dihalide derivatives [NbCp2Cl2]- have only been observed as transient species resulting from electrochemical reduction of [NbCp2Cl2] or [Nb(C5H4SiMe3)2Cl2] [63,153 ,154 ,155], which finally leads to the anionic dimer or [Nb(C5H4SiMe3)2Cl]2 respectively. Similar electrochemical reduction of several dialkyl complexes leads to [NbCp2R2]- in a reversible process with E = -2.08 V for R = CH2Ph [156], and E = -1.63 V for R = [CH(SiMe3)]2o-C6H4 [157]. Stable anionic dialkyl complexes [MCp2R2]- (R2 =2 CH2PPh2, (2-CH2C6H4)2 (53) have however been isolated as salts with the bulky cation [Na(18-crown-6)(THF)2]+ from the chemical reduction of metal(IV) dihalides with Na[C10H8] [158 ,159]; their X-ray molecular structures show the expected conformation with bend angles θ of 136° for R2 = 2CH2PPh2 and ϕ angles between the alkyl substituents of 74.2(2)° and 80.1(11)° respectively.
358
6
Group 5 and 6 Metallocenes
More extensive studies [14,15,16] have been made with anionic dihydride complexes [NbCp2H2]- (54) obtained by reaction of [NbCp2Cl2], the trihydrides [NbCp2H3] or the niobium(III) boranate [NbCp2(BH4)] with excess of the alkali hydrides in the presence of crown-ethers [160], or by further addition of lanthanide salts [161]. Similar compounds with stannyl substituents can also be obtained by reaction of the stannyl-hydrides with butyllithium, always in the presence of crown-ethers to provide bulky cations [162]. Salts of the [MCp2H2]- anion containing different cations, such as Li+, Na+, K+, Li(12-crown-4)+, Na(15-crown-5)+, Na(benzo-15-crown-5)+, K(18-crown-6)+, Yb(diglyme)2+, Ce(THF)63+ have been isolated and the molecular structure of some characterized by X-ray diffraction. All of them contain hydrogen bridges between the group 5 transition metal and the alkali or lanthanide atom, with ϕ angles between the substituents of 86(1)° (H-Nb-H) to 87.13(3)° (Sn-Nb-Sn) and θ bend angles between 139.4° and 144.0°. The dinuclear anionic hydride complexes containing one bridging η5-η1-cyclopentadienyl ring between the two metal atoms (55) or substituted by a siloxide group (56) are particularly interesting. They were obtained [14,18,19,20,22] by reduction of [NbCp2Cl2] with an excess of Na, or by sodium naphtalenide reduction of the Nb-Nb bonded (η5-η1-cyclopentadienyl) hydride intermediate (obtained from the reduction of [NbCp2Cl2] by NaH in an attempt to prepare the corresponding niobocene, which decomposes spontaneously by activation of the ring C-H bond and hydrogen transfer to the metal).
6.3 Type B-Complexes with mono-functional ligands
359
Addition of the carbonate dianion to the electro-generated species {Nb(η5C5H3RR’)2Cl} (R = H, R’ = SiMe3; R = R’ = SiMe3) gave the anionic carbonato niobium(III) derivative [Nb(η5-C5H3RR’)2{η2-OC(O)O}] - , which underwent electrochemical oxidation to give the corresponding neutral paramagnetic niobium(IV) complex [163].
6.3.2.4 Neutral Group 6 d2 Metal Compounds [MCp2X2] These are the most important type of molybdenum and tungsten derivatives, represented by an enormous number of complexes which are the result of the considerable efforts made by different groups of researchers. They are 18-electron compounds, isoelectronic with the bis(cyclopentadienyl) group 5 metal(III) complexes [MCp2XL], with a free pair of electrons which has a significant effect in their chemical behaviour.
Halides and hydrides Several synthetic methods have been used to prepare particular compounds [1,2], but the first entrance to the dihalide compounds [MCp2Cl2] (57) was the reaction of MoCl5 with excess of NaCp [1]; a more convenient method giving higher yields uses reactions of halogenating reagents such as chloroform with the dihydrides [MCp2H2] (58), which can be isolated in yields of around 40% by reaction of MoCl5 or WCl6 with NaCp.DME and NaBH4 [164 ,165]. Hydrogenation of (15) gave [Mo2(µ-η5:η5-C10H8)Cp2H4] which was structurally characterized [87].
Scheme 6-15 An improved synthesis of [WCp2Cl2] used the reaction of [WCl4(DME)] with LiCp, and [WCp2H2] was obtained by reaction of Li[AlH4] with the dichloride (Scheme 6-12) [166]. The dihydrides were also prepared by co-condensation of the metal atoms and CpH at low temperature, and by reaction of the dihalides with
360
6
Group 5 and 6 Metallocenes
Na[BH4]. The reverse transformation of the dihydrides into the dihalides was effected by reaction with CHCl3 or other appropriate halogenating agents, but not with halogens X2 since they cause oxidation to higher valent compounds. The ansa-metallocene dichlorides (59) [167], (60) and (61) [168] were obtained in a similar way from reactions of the meta(IV) chlorides with alkali metal salts of the appropriate ansa-cyclopentadienyl ligand. Complexes (59), (60) and (61) were used to prepare hydrido, sulfido, alkyl and aryl derivatives. X-Ray molecular structures of (60) and (61) show the expected reduction in the bend angle θ associated with sterically rigid ansa-bridged complexes.
An alternative synthetic route requiring more steps but leading to higher yields involves the preparation of [MCp(CO)3H] (directly accessible in high yield from [M(CO)6] and CpH) followed by oxidation with PCl5 giving [MCpCl4] [169]. Reduction of these metal(V) complexes with magnesium or other reducing agents and simultaneous reaction with alkali cyclopentadienides gives access to the dihalides; this was the method used to prepare [WCp*2Cl2] [170] and is a particularly useful method for the preparation of compounds with mixed cyclopentadienyl rings. The reaction of [MCp2Cl2] with [M’PPh2(CO)5] (M’ =Cr, or W) leads to ring substitution [171] and the formation of several bimetallic chlorohydride derivatives (62) (M = W); the substitution is followed by orthometallation to (63) when M = Mo [172 , 173]. These metal(IV) compounds have one pair of free electrons which are responsible for the Lewis basic properties of the dihydride complexes, although not chemically significant for the dihalides and other dianionic derivatives due to their higher electronegativities. The location of this electron pair and its chemical role have been the subject of many theoretical and structural studies [1,2]. The narrow H-MH ϕ angle of 75.5° in [MoCp2H2], found by neutron diffraction, reveals that it cannot be located between the two hydride ligands but occupies the 1a1 orbital extended along the y axis; this is in good agreement with the results of theoretical calculations.
6.3 Type B-Complexes with mono-functional ligands
361
Dihydride compounds can be protonated to give the cationic trihydride derivatives [MCp2H3]+ (Section 6.3.3) and many of their reactions can be interpreted in terms of interactions due to their Lewis basicity. Several stable adducts [MCp2H2. A] with different Lewis acids can be isolated, such as those with A = BF3, AlR3 and many transition metal fragments in which the donor-acceptor metal-metal bond is frequently acompanied by µ-H and/or µ-CO bridges. In some cases simple hydride transfer occurs to give the corresponding products or the ionic salts of the cationic hydride [MCp2HL]+. Alkyls Dihydride and dihalide complexes are convenient starting materials for the preparation of many related derivatives. Monoalkylated hydrides [MCp2HR] can be isolated by insertion of alkenes (CHR=CH2) or alkynes (CR≡CH) with activating substituents (R = CN, COOMe) into one M-H bond of the dihydrides; insertion into the second M-H bond is possible only in a few limited cases, such as the reaction which gives [MoCp2{C(CN)=CHCN}2]. A particularly interesting group of alkyl hydrides are those which result from the activation of ring or alkyl-ring C-H bonds when methods designed to prepare metallocenes are applied. All of these reactions, which have been extensively studied, proceed with transfer of hydrogen to the metal and formation of η1-η5-cyclopentadienyl bridged compounds (64) or η5η5-fulvalene hydrides (17). Interesting C-H activations occur in reactions of [MCp2H2] with BH3.THF leading to various boranate derivatives [174]. Protonation of the alkyl hydrides with HX affords the corresponding alkyl derivatives [MCp2RX] (X = Cl, OTf, SPh, SH) [175]. Dialkyl complexes can be isolated by alkylation of the dihalides with various alkali, magnesium and aluminium alkylating agents, and the same reactions can be used to obtain metallacyclobutanes, metallacyclobutabenzenes, metallasilacyclobutanes and other related metallacycles. Different behaviour was observed for certain magnesium alkyls such as allyl, 2methylallyl and benzyl [176], which in the presence of MgCl2 attack the cyclopentadienyl ring and, after elimination of chloride and hydrogen transfer to the ring, give ring-substituted chloro hydrides [MCp(C5H4R)ClH]. After protonation they are transformed into the dichlorides which can be treated in the same way to give [M(C5H4R)2ClH] and [M(C5H4R)2Cl2] derivatives. Similar cyclopentadienyl ringsubstituted compounds have been isolated by silylation of the allylic double bond with silylalkyl magnesium reagents [177]. Several differences between the ansabridged compounds and their non-bridged analogues were noted [169]. Photolysis of [M{CMe2(η5-C5H4)2}H2] in benzene gave [M{CMe2(η5-C5H4)2}PhH] for M=Mo but not for M = W, whereas the reverse situation applies in the unbridged complexes. In addition the dihydrides and the alkyl or aryl hydrides were reported to be much more stable towards reductive elimination than their non-bridged analogues; this behavior was thought to reflect the inability of the 16 electron intermediate [M{CMe2(η5-C5H4)2}] to adopt a relatively stable parallel ring structure [178].
362
6
Group 5 and 6 Metallocenes
Other derivatives The dihydride complexes are convenient starting materials for a wide range of derivatives. For example, the diboryl complex [WCp2(Bcat’)2] (65) (Bcat’ = O2C6H34-But, or O2C6H3-3,5-But2) was obtained from the photochemical reaction of the dihydride with cat’BBcat’. The hydrido boryl complex [WCp2H(Bcat)] (66) (Bcat = O2C6H4) was obtained from the reaction of chlorocatecholborane with Li[WCp2H], catecholborane being eliminated from (66) under photochemical conditions. X-Ray molecular structures obtained for (65) and (66) have geometries which are consistent with the presence of a weak dπ-pπ metal-boryl interaction [179 ,180 ,181]. Alkoxide and carboxylate hydrides [MCp2H(OR)] and [MCp2H(OOCR)] are best prepared via reactions of the alkyl hydrides [MCp2HR] with the corresponding alcohol or carboxylic acid [182]. When the same reactions are carried out using the dihydrides, the resulting protonated complex [MCp2H3]+ can only be transformed by exposure to light in the presence of O2 as a hydrogen acceptor, to initially give monosubstituted [MCp2H(OR)] or [MCp2H(OOCR)] and finally disubstituted compounds [MCp2(OR)2] and [MCp2(OOCR)2] under appropriate conditions [183]. Thiolate complexes [MCp2(SR)2] have been isolated by reaction of the dichlorides with thiols [2,184], and the influence of the alkyl chain length and steric effects on the Mo-S bond enthalpies has been studied [185]. Silanes react photochemically with group 6 metal dihydrides to give hydride complexes [MCp2H(SiR3)], and with the silene tungsten complex (67) to give bis(silyl) compounds [WCp2(SiMe3)(SiR3)] [186]. Similarly the germyl silyl complexes [WCp2(SiMe3)(GeR3)] (68) were obtained from the reaction of (67) with germanes [187]. Crystallographic studies show that complexes with a chloride substituent on the YR group (Y = Si or Ge; R = Me2Cl or Pri2Cl) experienced considerable shortening of the M-Y bond accompanied by lengthening of the Y-Cl bond, resulting from π-back-bonding from the metal to the Y-Cl σ-bond. Thermolysis of (68) gave an unusual dinuclear germylidene complex (69) with a three coordinate germanium moiety bridging the two tungsten centres [188]. Reactions of [MCp2H2] with tin alkyl halides afforded disubstituted [MCp2(SnR3)2] and monosubstituted [MCp2H(SnR3)] stannyl complexes [189]. The molecular structure of the
6.3 Type B-Complexes with mono-functional ligands
363
di(chlorodimethylstannyl) complex (70) shows a Sn-Mo-Sn bond angle of 83.2(1)°, in the range observed for Y-W-Y in the silyl (86.0(1)° - 97.3(1)°) and silyl germyl (85.9(1)°) complexes.
Molybdenum-magnesium bonded compounds (71) have been isolated by reaction of the molybdenum dihydride with magnesium dimethyl in the presence of TMEDA [190].
Bidentate dianions An important group of compounds with dianionic bidentate ligands are accessible in several ways. A particularly interesting synthetic method uses the cis- stilbene derivative [MCp2(η2-PhCH=CHPh)], which acts as a precursor of the metallocene fragment which is then transformed into a series of metal(IV) derivatives by oxidative coupling of various unsaturated substrates [191, 192]; it also reacts with nonenolizable ketones giving kelated diolate compounds [MCp2(-OCR2CR2O-)]. Related complexes can also be obtained directly from hydride derivatives with evolution of H2. For example, reaction of [MoCp2H2] with CO2 gives the sidebound η2-CO2 intermediate which gradually disproportionates to give the carbonyl
364
6
Group 5 and 6 Metallocenes
and the η2-coordinated carbonate complexes [MoCp2(η2-CO3)] (72) [193]; the corresponding tungsten complex is obtained by reaction of [WCp2HPh] with CO2 in wet acetone [183]. An unusual reaction takes place when the tungsten dihydride reacts with diphenylketene giving a product which rearranges to complex (73) [194]. Similar reactions of the dihydrides with CS2 and SO2 give the cyclic compounds with bidentate dianions [MCp2(-SCS-)] and [MCp2(-O(SO2)S-)] respectively [195]. Several chelate structures with oxygen donor ligands such as [MoCp2(-O{PO(OH)}O-] were proposed to explain the species observed in aqueous solution during [MoCp2Cl2] promoted phosphoester bond cleavage reactions [196].
An alternative synthetic strategy giving access to compounds of this type uses [2 + 2] cycloadditions to the M=O bond of the oxo metal complexes [MCp2O] discussed in section 6.4. Their reactions with ketenes, ketenediimines and isocyanates give complexes of the general type [WCp2{-O(C=E)E’-}] (E = CPh2, NR, O; E’ = O, NR) [197]. The related molybdenum dimer [MoCp2{-O(MoO2)O-}] results from irradiation of [MCp2(O)] in the presence of oxygen or by reaction of [MCp2Cl2] with [MoO4]2- [198]. Cyclic compounds with sulphur and selenium donors such as [MCp2{-S(E)E’-}] (E = S2, C=NTol) and [MCp2{-Se (o-C6H4)Se-}] have also been reported [199, 200]. Photolysis of [WCp2H2] with thiophene gave mixtures of the CS2 insertion product (74) and the thienyl deriative [WCp2(C4H3S)] in ratios dependent on the irradiation time, (74) being the initial product [201]. The x-ray molecular structure of (74) showed a nonplanar metallathiacyclic ring in which the bonding was thought to be localized. A similar reaction of the molybdenum dihydride gave only the thienyl complex.
6.3 Type B-Complexes with mono-functional ligands
365
A series of metallaheterocycles was obtained from reactions of the tungsten carbonyl derivative [WCp2(CO)] with heteroallenes [202]. Reactions of elemental chalcogens (S, Se or Te), Me3NO or Me3PE (E = S or Se) with the disilenes [MCp2(η2-Me2Si=SiMe2)] afforded symmetrical metallacycles (75) (M = Mo; E = S, NSiMe3, SiMe2: M = W; E = O, S, Se, Te, SiMe2) [203, 204]. Varying amounts of the unsymmetrical insertion products were also obtained from reactions with phosphine chalcogenides.
6.3.2.5 Cationic Group 6 d2 Metal Compounds [MCp2XL]+ This important group of cationic complexes is represented by a large number of molybdenum and tungsten derivatives which are easily accessible because the presence of one pair of free electrons in the neutral [MCp2X2] compounds favours substitution of at least one of the anionic substituents, mainly by π-acceptor ligands. Substitution in metal(IV) complexes [MCp2X2] (X = Cl, Br, I, SR) may be achieved by addition of ligands, halide abstraction with thallium salts and electrochemical or chemical oxidation with FeCp2+ and NO+ salts. It has been demonstrated that electrochemical oxidation of [MCp2X2] (X = Cl, Br, I, SR) in nitrile solvents leads to the metal(V) cationic species [MCp2X2]+ [205], which undergoes oxidatively induced reductive elimination of X2 to give the nitrile-coordinated cationic metal(IV) complexes. On this basis, chemical oxidation in nitrile solvents was the method used to isolate nitrile coordinated cationic species [MCp2X(NCMe)]+ (76) [206 ], which are appropriate precursors for the substitution of different π-acceptor ligands such as CO or phosphines. Direct substitution of iodide by simple addition of isocyanides affords the cationic complexes [MCp2H(CNR)]+ [80, 207], and related halide compounds can also be obtained by abstraction of halide in the presence of the ligand. Photochemical reactions of [MInd(η4- C5H6)(CO)2]+ gave the hydride [MIndCpH(CO)]+ which was easily transformed into the corresponding halides on addition of chloroform or methyl iodide [208]. The high tendency to coordinate π-acceptor ligands is shown by the easy electrophilic attack on the coordinated CO2 in [MoCp2(η2-CO2)] to form the carbonyl derivative [209]; reaction with HCl or SiClMe3 gives [MoCp2Cl(CO)]+; reaction with RCOCl gives [MoCp2X(OOCR)(CO)]+; reaction with [CoH(CO)4] gives the heterodinuclear compound [MoCp2H(CO)]+ [Co(CO)4]- [210], which is further transformed into [MCp2(µ-H)(µ-CO)Co(CO)3] [Co(CO)4].
366
6
Group 5 and 6 Metallocenes
Dissociation of iodide takes place more easily than chloride or bromide, and tungsten complexes are more resistent to substitution. A solution of [MoCp2HI] in acetonitrile rapidly establishes an equilibrium with its solvolysis product [MoCp2H(NCMe)]+ which is completely shifted by addition of TlBF4, as evidenced by NMR spectroscopy; similar displacement of iodide takes place by simple addition of phosphines, whereas this reaction is slower for chloride and bromide and does not take place for related tungsten derivatives [211]. Halide species such as [WCp2Br(SMe2)]+ and the dimer [MoCp2(µ-SR)]22+ are easily accessible from [WCp2Cl2] [212], and have been used to prepare bis(cyclopentadienyl) tungsten complexes with chiral metal centers [WCpCp’BrH] via nucleophilic attack on one of the cyclopentadienyl rings, with migration of hydrogen to the metal and elimination of the sulphide ligand [213, 214]. The dithiocarbamate complex [WCp2(η2-S2CNHMe)]+ was also obtained by a rare hydrogentransfer reaction from [WCp2(SH)2] [215]. Corresponding alkyl complexes [MCp2R(NCMe)]+ (R= Me, CH2Ph, SR) were obtained by elimination of alkane on addition of NH4PF6 to the dialkyl complexes [215], or by migration of phenyl to the methylidene group of the intermediate [WCp2(CH2)Ph]+ obtained by deprotonation of the methyl derivative with the trityl radical dimer [216]. Isocyanide complexes can be obtained by alternative methods (Scheme 6-16). The molybdenum isocyanide complexes [MoCp2X(CNMe)]+ (X = halide, H, Me) (77) were isolated by reaction of the corresponding cyanide complex wih LiMe [217], and the methyl complexes [MoCp2MeL]+ (L = PR3, CNR) were isolated by reaction of the neutral adducts [MoCp2L] with MeI [80]. Irradiation of [WCp2H(C2H4)]+ in acetonitrile yielded the tungsten hydride [WCp2H(CNMe)]+ (78), whereas [WCp2Et(CNMe)]+ was obtained after stirring the reaction mixture in the dark for one week [218]. Reaction of the dihydride with trifluoromethanesulfonate in the presence of ligand gave [WCp2HL]+ (L = NCR or THF ) [219]. A relatively short W-O bond length (2.194 Å) was observed in the THF compound, which was thought to result from strong σ bonding between the cationic metal centre and the THF without any W-O π−bonding.
6.3 Type B-Complexes with mono-functional ligands
367
Scheme 6-16
6.3.2.6 Dicationic Group 6 d2 Metal Compounds [MCp2L2]2+ When the solutions of [MCp2X(NCMe)]+ described above were treated with an appropriate excess of CO and phosphines the dicationic complexes [MCp2L2]2+ (L = CO, PR3, CNR) were obtained [80, 207, 208]. Reaction of [MCp(η4-C5H6)(CO)2]+ with Ph3C+ also gave similar carbonyl adducts [220, 221]. Crystallographic studies show η5-coordination of the Cp moieties in [MoCpInd(CO)(NCMe)]2+ whereas η3coordination of the indenyl groups was observed in analogous neutral carbonyl species [209]. Complex (78), although stable in the solid state or in acetonitrile, reacts in diethyl ether or acetone solution with loss of H2 to give the structurally characterized, dinuclear C-H activation product (79) [219].
17-Electron Compounds The most simple neutral complexes with 17 electrons are the d1 vanadium(IV), niobium(IV) and tantalum(IV) derivatives. The HOMO accommodating the unpaired electron of all these paramagnetic compounds is the a1, predominantly non-
368
6
Group 5 and 6 Metallocenes
bonding, dz2 metal orbital which has a variable contribution from the dx2-y2 atomic orbital. The other two orbitals are used for bonding with the two anionic substituents (Figure 6-5) All of these complexes are paramagnetic with effective magnetic moments corresponding to the presence of one unpaired electron. Delocalization of the unpaired electron on the ligands has been measured by the Aiso and g parameters in the EPR spectra and coupling with the six methyl protons has been demonstrated for the dimethyl-vanadium complex. It is convenient, not only for structural but also for synthetic purposes, to distiguish between two types of compounds: a) those having independent monoanionic ligands and cyclopentadienyl ligands with typical bend angles θ between 130° and 135°, and the additional two ligands located in the equatorial plane at angles between 87° and 95°; b) those better considered as metallacycles where the two metal-bonded atoms are also bonded together, which have θ angles larger than 135° and the additional two ligands located in the equatorial plane at angles between 35° and 37°. For these metallacyclic compounds an increase in dz2 and a decrease in dx2-y2 character has been observed from a reduction of the vanadium hyperfine coupling in their EPR spectra. With this electronic distribution these compounds show neither basic nor acidic behaviour. Compounds of type a) are included here whereas those of type b) are discussed in Section 6.5.
6.3.2.7 Neutral Group 5 d1 Metal Compounds [MCp2X2] Halides Most of the synthetic methods for vanadium(IV) dihalides use metathetical reactions of the tetrahalides VX4 with reagents which can transfer the cyclopentadienyl ligands, such as alkali cyclopentadienides or MgCpCl, or reactions of the dichlorides [VCp2Cl2] which exchange chloride for other anionic ligands. Mixed dihalides have also been obtained by oxidation of [VCp2Br] with ClCPh3 and more recently the use of PCl3 led to a mixture of [V(C5H4R)2Cl] and [V(C5H4R)2Cl2] (R = Me, Pri, But) [16]. An extensive series of vanadium compounds containing variously substituted cyclopentadienyl and bridged bis(cyclopentadienyl) rings has been reported [1, 2, 222], and related pseudohalide complexes can be obtained by metathesis with the corresponding silver salts. Electrochemical studies of the dihalides show that reduction to [VCp2Cl2]- is accompanied by rapid loss of chloride to give the vanadium(III) chloride, whereas oxidation leads to the unstable cationic species [VCp2Cl2]+. The bis(cyclopentadienyl) niobium and tantalum(IV) dihalides are a large group of readily accessible compounds since the reducing properties of alkali cyclopentadienides and also of SnBunCp (reagents frequently used in excess) lead directly to
6.3 Type B-Complexes with mono-functional ligands
369
these d1 metal compounds; mechanistic studies of this reduction have been reported [1]. The formation of difficult to separate mixtures may be avoided by addition of one equivalent of a reducing agent such as Na/Hg and then stoichiometric amounts of the Cp-transfer agent. The following alternative methods are available for particular purposes: a) Simple metathesis using the easy to isolate NbCl4(THF)2 with LiCp is a clean reaction used for many of these compounds. b) Metathesis of monocyclopentadienyltantalum(IV) halides with different alkali cyclopentadienides has been used in particular to prepare compounds containing mixed Cp rings. c) Metathesis of the anions in previously isolated bis(cyclopentadienyl)metal compounds has been used where other methods cannot be employed. d) Reaction of the trihydrides with alkyl halides takes place with elimination of H2 and alkane to give the dihalide. e) Oxidation of metal(III) compounds such as [M(C5H4SiMe3)2X]2 with alkyl halides or [NbCp*2(η2-BH4)] with HCl has been used to prepare less accessible complexes. Method a) using the thallium salt of the ligand has more recently been used to prepare the ansa-niobium complex [NbCl2{SiMe2(C5H4)2}] (80) [223].
These dihalide complexes are frequent starting materials in a wide range of metathetical reactions. Alkyls Alkylation of vanadium dihalides with the usual alkylating agents in polar solvents (THF) generally causes reduction to vanadium(III) alkyls [VCp2R], although vanadium(IV) dialkyls can be easily isolated by alkylation with lithium, magnesium, aluminium or cadmium reagents in toluene or dichloromethane. Metallacyclic dialkyls have also been isolated by similar methods. Haloalkyl complexes are usually prepared by reaction of the vanadium(III) alkyls [VCp2R] with halogenating oxidants such as CuCl2, CuCl and AgCl, and the chloromethyl complex
370
6
Group 5 and 6 Metallocenes
[VCp2MeCl] can be isolated by reaction of the dimethyl with a stoichiometric amount of hydrogen chloride. Alkylation with alkali metal, magnesium, aluminium, zinc and other reagents has also been extensively used to prepare dialkyl and alkyl halide derivatives of niobium and tantalum, although selective monoalkylation requires the use of particular alkylating agents or reverse reactions of the dialkyls with appropriate reagents. Reported alkyl complexes include Me, CH2SiMe3, CH2But, CH2Ph, Ph, η1C5H5, C(CF3)=CH(CF3) amongst others in compounds with both unsubstituted and substituted cyclopentadienyl rings [2]. Phosphinomethyl complexes have been also used as metallo-ligands in the preparation of heterodinuclear compounds [160]. An important group of related derivatives are the metallacyclo-pentane, -pentene and -pentadiene dialkyls [158, 224].
Hydrides The metal dihydrides are not easily synthesized and very few examples of hydride derivatives have been reported. Niobium and tantalum trihydrides are typically easily transformed into metal(III) hydrides losing H2, and the easy reduction of the assumed metal(IV) dihydrides by hydride-transfer reagents makes these compounds very difficult to isolate. [Nb(C5H4R)2H(µ-H)]2 (R = CH2SiMe3, CH2But) has been electrochemically generated and its structure proposed on the basis of ESR studies [108]. The cyclopentadienyl-bridged hydride complex [{Nb(η5-η1-C5H4)(η5C5H5)H}2] is one of the few hydrides which does not spontaneously rearrange into a fulvalene derivative.225
Other derivatives A large number of vanadium-oxygen-, -sulphur- and -selenium-bonded compounds with independent terminal ligands and cyclic dianionic ligands can be isolated by metathetical reactions or oxidative addition of the appropriate reagents to vanadocene. Related niobium and tantalum derivatives with different donor-atom ligands include silyl, alkoxo, thioalkoxo, thiocyanide, dilkylphosphide, amide, carbonate, bridging-imide etc. Most of them can be prepared by appropriate metathetical reactions of the dihalides or from the trihydrides. Metal(IV) bis(cyclopentadienyl) derivatives have been analysed in detail [1,2]. The crystallographically characterized amido niobium complex (81) with a planar metallacyclic ring was obtained from the reaction of Scheme 6-17 [226]. Several metallacyclic complexes with sulfur donor ligands have recently been reported and their electrochemical behaviour investigated [227].
6.3 Type B-Complexes with mono-functional ligands
371
Scheme 6-17
6.3.2.8 Cationic Group 5 d1 Metal Compounds [MCp2XL]+, [MCp2L2]2+ Paramagnetic monocationic niobium halide complexes [MCp2XL]+ were isolated by oxidation of the metal(III) derivatives with Ag[BF4] [116,120] or [FeCp2][PF6] [228] and by metathesis of the metal(IV) derivatives [116, 229]. They have also been generated by electrochemical oxidation of hydrido- [110] and halo- [116] niobium(III) compounds. Related fulvalene dinuclear cationic compounds [{NbCp(NRR’)}2(µ-fulvalene)]2+ (R = H, R’ = Ph, Tol; R = Et, R´= Tol) and the protonated µ-hydride complex [{NbCp(NHR)}2(µ-H) (µ-fulvalene)]3+ have been reported [109]. Dicationic complexes [MCp2LL’]2+ have also been isolated by displacement of the second halide from [MCp2XL]+ with different monodentate ligands [116] and directly from [MCp2Cl2] by reaction with thallium salts (PF6 or BF4) in the presence of methylisocyanide [230] or bidentate nitrogen [231] donor ligands.
6.3.2.9 Cationic Group 6 d1 Metal Compounds [MCp2X2]+ Few examples of this type of paramagnetic metal(V) cationic species have been reported. Oxidation of [MCp2Cl2] with WF6 in liquid SO2 allows the WF6- and W4F18- salts of the [MCp2X2]+ cation to be isolated [232]. Electrochemical oxidation of [MCp2Cl2] in acetonitrile generates the cation [MoCp2Cl2]+ which in the presence of ultrasound promotes the formation of two electron oxidation product [MoCp2X(NCMe)]2+ at a considerably lower voltage than in the absense of ultrasound [233]. Paramagnetic metal(V) dialkyls are an important group of compounds in which the α-elimination processes responsible for the transformation of metal(IV) dialkyls induced by trityl salts have been extensively studied. Electrochemical studies show that the metal (IV) dialkyls [MCp2R2] are easily oxidized to the monocations [MCp2R2]+ with a more negative potential for R = Me (-0.27 V) than for higher alkyls (ca. -0.21 V) [234] and this oxidation can be effected at a preparative
372
6
Group 5 and 6 Metallocenes
level by chemical oxidation. Mechanistic studies using various combinations of alkyls [MCp2MeR’]+ (R = Me, Et, Ph) indicate that reactions of these species with Ph3C-CPh3 take place by preferential α-abstraction from the methyl group, leading to metal(IV) cationic species [MCp2RL]+ ; only when the reaction of [MCp2MePh]+ is carried out in CH2Cl2 is the chloro derivative [MCp2Cl(CH2Ph]+ isolated [217]. Oxidation of [WCp2(OMe)Me] with ferrocenium hexafluorophosphate gave the structurally characterized dicationic, d1-d1 dinuclear complex (82), in which shortening of the W-O bond (W-O = 1.904(1) Å) was observed, suggesting (together with its electronic spectrum and magnetic measurements) a bond order of about 1.5 [235]. Spin pairing of the d1 metal centers through a π−interaction involving the oxo bridge was proposed to explain its diamagnetism in both solution and the solid state. Complex (82) is photosensitive and disproportionates in acetonitrile to give the terminal oxo complex [WCp2(O)Me]+ (83) and the d2 solvent adduct [WCp2Me(NCMe)]+.
16-Electron Compounds These cationic 16-electron complexes have two orbitals 2a1 and b2 involved in bonding with the two anionic substituents and one vacant orbital 1a1 (refer to Figure 6.5)
6.3.2.10 Cationic Group 5 d0 Metal Compounds [MCp2X2]+, [MCp2XL]2+ Most of the reported d0 cationic complexes are 16-electron compounds. The first cationic group 5 metal(V) derivatives were reported as products of the reaction of the disulphur compounds [MCp2(η2-S2) Me] with MeI and [CPh3][BF4] which gives [MCp2MeI]+, and presumably in the presence of H2O also gives the µ-oxo
6.3 Type B-Complexes with mono-functional ligands
373
dinuclear [(MCp2Me)2(µ-O)]2+ complexes [236]. Similar halo complexes [MCp2ClX]+ were isolated as the Br3- and I3- salts [237] by oxidation of [MCp2ClX] with an excess of halogen (X = Br, I), and by oxidation of [MCp2Cl(SiMe3)] [238] or [MCp2X2] with various oxidizing agents such as AsF5 [239] or [FeCp2]+ [240 ,241 ,242]; the oxidations in DMF give the dicationic species [MCp2X(DMF)]2+. They were also generated by electrochemical oxidation of the metal(IV) dihalides [243]. Related alkyl [160,244 ,245 ,246] and phosphide [247] complexes have also been isolated. The alkoxide complexes (84) and (85) were obtained via the carbon-carbon or carbon-hydrogen bond formation reactions of Scheme 6-18; (84) was hydrolised to the Lewis acid adduct (86) [248]. Oxidation of {Nb(η-C5Me4Et)2} (prepared in situ by Na/Hg reduction of the dichloride) with excess HPF6 yielded the structurally characterized difluoride [Nb(η-C5Me4Et)2F2]+ under spontaneous evolution of H2 [249].
Scheme 6-18
Analogous complexes with niobium-gold bonds [NbCp2{Au(PPh3)2]+ have been isolated from the reaction of the trihydrides with gold(I) salts [250].
6.3.2.11 Dicationic Group 6 d0 Metal Compounds [MCp2X2]2+ The oxidation of [MCp2X2] with very strong oxidants such as AsF5, SbF5, or [(FCN)3F]+[BF4]- gives the dicationic species [MCp2X2]2+ (M = Mo, W; X = Cl, Me) [247, 251, 252].
374
6
Group 5 and 6 Metallocenes
6.3.3 Complexes with Three Coordinated Ligands 18-Electron Compounds The only neutral complexes with three additional substituents bonded to the metal centre are the 18 electron species derived from d0 vanadium, niobium and tantalum(V) metals (Table 6-4). All the other d0 group 5 and 6 metal compounds are cationic species. For these compounds all three metal based valence orbitals are used to bond the three σ-donor ligands (Figure 6-6 ). Table 6-4. Different types of [MCp2L3], [MCp2XL2] and [MCp2X2L] groups 5 and 6 metal complexes
Total Number of electrons
18
Configuration
d0
[(M5)Cp2X3] [(M5)Cp2X2L]+ [(M5)Cp2XL2]2+ [(M6)Cp2X3]+ [(M6)Cp2X2L]2+
Figure 6-6 Qualitative diagram for [MCp2X3] d0 metal complexes with σ-donor ligands
6.3 Type B-Complexes with mono-functional ligands
375
6.3.3.1 Neutral Group 5 d0 Metal Compounds [MCp2X3] Halides Almost a complete list of homo- and hetero-trihalide complexes with equal and mixed unsubstituted and substituted cyclopentadienyl rings have been isolated [1,2]. The presence of three rather bulky donor atoms in the equatorial plane between the two rings makes the coordination sphere too crowded and therefore these compounds have a high tendency to lose a halide ligand to give cationic species with fewer than 18 electrons. Isolation of neutral trihalides usually requires the use of non-coordinating solvents. Furthermore, the reducing properties of alkali cyclopentadienides, used to transfer the rings to the metal(V) halides, lead to the metal(IV) dihalides [MCp2Cl2], which are the most convenient complexes to prepare, store and use as starting materials for the preparation of trihalides [238]. These complexes and their related silyl, alkoxides and phosphides are easily hydrolized to the µ-oxo dimers [(MCp2X2)2(µ-O)] (X = halide, SiPh3, PR2, OR). An alternative method of synthesis is the oxidation of the corresponding metal(IV) compounds by oxygen. Hydrides The trihydride is one of the most important and useful group 5 metal derivatives, however its preparation does not follow a simple direct path because the method involving the reaction of MCl5 with NaBH4 always leads to undesirably low yields. Better yields were obtained by using Na[AlH2(OCH2CH2OMe)2] (vitride or red-al) or reacting [MCp2Cl2] with Li[AlH4]; the mechanism of the latter has been investigated [2]. Many structural studies have been carried out on trihydride compounds, including x-ray and neutron diffraction and photoelectron spectroscopy. The NMR behaviour of the niobium derivatives is particularly interesting; short T1 values and large (up to 100 Hz at room temperature) temperature dependent 1JH-H coupling constants have been reported, in contrast to the low temperature independent values (ca. 10 Hz) usually observed for the tantalum derivatives. Extensive NMR experiments, isotopic labeling studies and theoretical calculations have been undertaken to investigate this unusual NMR behaviour [253 ,254 ,255], which has been attributed to either a combination of magnetic and exchange couplings [254] or to quantum mechanical exchange couplings [255] as a simple manifestation of the quantum mechanical motion associated with the hydrogen nuclei in the trihydride structure. These observations indicate exchange with structures containing η2-coordinated dihydrogen, consistent with the easy loss of hydrogen observed on heating or by addition of ligands [256, 257]. Similar elimination reactions take place with trimethylsilane affording the mono and bis(trimethylsilyl) hydrido derivatives [258 ,259 ,260 ,125]. Likewise, mono
376
6
Group 5 and 6 Metallocenes
and bis(trimethylstannyl) hydride complexes were obtained [261]. Phosphido [262] and arsenido [263] derivatives were obtained from reactions with halo-phosphines and -arsenes (Scheme 6-19). A dihydro phosphido niobium complex was however not isolated and the reaction yielded only the phosphine hydride [NbCp2H(PHPH2)] [264]. The structurally characterized boryl complexes (87a) and (87b) were obtained as an easily separated isomeric mixture from the reaction of Li[WCp2H] with chlorocatecholborane [265]. Mixed hydride- halide, silyl, and amide complexes have also been reported [2].
Scheme 6-19
6.3 Type B-Complexes with mono-functional ligands
377
The trihydrides react with Lewis basic coinage metal cations [e.g. Cu(MeCN)4+, Ag(THF)+, Au(THF)+] to give hydrido bridged heterometallic species of type (88), several of which have been crystallographically characterized [266 ,267]. Alkyls Bis(cyclopentadienyl) trialkyl complexes are not known and very few di- and mono-alkyl derivatives have been reported, most of which are methyl compounds. The dimethyl niobium complex [NbCp2Me2(η1-S2CNEt2)] was prepared by reaction of the niobium(IV) dimethyl derivative with R2NC(S)S-SC(S)NR2 [268], whereas the same reaction with (RO)2P(S)S-SP(S)(OR2) gave the paramagnetic niobium(IV) compound with elimination of both methyl groups. Similarly the monoalkyl complexes [NbCp2Bun{CH(SH)(SMe)}(η1-BH4)] [269], [NbCp2Me(η2S2CNMe2)] [S2CNMe2] [269], [NbCp2MeHX] (X = SiH3, PH2, OMe, SMe, NMe2), [245] the metallaoxacyclic compounds [MCp2Me(η2-O,C)-OCHR-CH2-)] [101,270] and the vinyl derivative [TaCp2(CH=CH2)(SBu)H] [271] have been isolated. 6.3.3.2 Cationic Group 5 d0 Metal Compounds [MCp2X2L]+, [MCp2XL2]2+ and [MCp2L3]3+ Few complexes of this type have been reported. Monocationic niobium complexes [NbCp2Cl2(Me2CO)]+ were obtained when acetone solutions of the trihalo complexes [NbCp2Cl2X] (X = Cl, I) were treated with one equivalent of Ag+ salts, and the related dicationic compounds [NbCp2ClL2]2+ resulted from the reaction with two equivalents of Ag+ salts in acetone or in the presence of bidentate nitrogen donor ligands such as bipy and phen. Even the tricationic complex [NbCp2Cl2(dien)]3+ was isolated using the tridentate ligand and a stoichiometric amount of the silver salt [273]. Steric congestion in the coordination sphere, as found in tris-ligand bis(cyclopentadienyl) complexes, probably explains why cationic tantalum compounds have only been isolated as dihydrides with one neutral ligand [TaCp2H2(PR3)]+ [PR3 = PMe2Ph, P(OMe)3], via protonation of the appropriate Ta(III) hydride [274]. Protonation of [TaCp2H(CO)]+ (Cp = Cp or (η-C5H4But)) gave a mixture of isomers: the classical dihydride and an η2-dihydrogen derivative [TaCp2(η2-H2)(CO)]+ [275]. The niobium derivatives were obtained by protonation of [Nb(η-C5H4SiMe2)2HL] ( L = CO) with [NH3R]PF6 (R = H or Ph). Appropriate reaction conditions and HCF3CO2 as protonating agent also gave the side bonded species identified as (89) by NMR spectroscopy [272]. Electrochemical reduction of the trihydrides in THF gave [NbCp2H2(THF)]+ (Cp = C5H4SiMe3, C5H4CMe3) [108].
378
6
Group 5 and 6 Metallocenes
6.3.3.3 Cationic Group 6 d0 Metal Compounds [MCp2X3]+ The best example of this type of compound is the cationic trihydride obtained by protonation of the dihydride. Deuteration of [WCp*2H2] demonstrated that D+ is firstly bonded in the central position and subsequently slowly isomerizes to occupy an external coordination site [273]. The NMR spectra for [MCp2H3]+ (M = Mo, W) correspond to an AB2 spin system with 1JH-H = 8.5 Hz for tungsten and 1JH-H = 450 Hz for molybdenum due to exchange coupling [254]. Exchange couplings in the cationic trihydrides have been theoretically analyzed and the calculated values were in good agreement with the experimental results and are consistent with an η2-dihydrogen transition state [274]. The Lewis basicity of the trihydrides [MCp2H3] + is comparable to that of Et2O and adducts with different metal salts have been isolated.2 Reactions with CuCl, Fe2(CO)9 or Co2(CO)8 give salts of the [CuCl2]-, [HFe3(CO)10]- or [Co(CO)4]- anions respectively, whereas µ-H bridges are formed with [RhH2(acetone)2(PPh3)2]; activation of the ring C-H bond of one of the cyclopentadienyl ligands takes place with [IrH2(acetone)2(PPh3)2]; hydrogen transfer occurs with [ZrCp2X(COR)] to give [Cp2M-CHR-O-ZrXCp2]. The reducing properties of the cationic trihydride compounds have been investigated in reactions with ketones and alcohols.
6.4 Type C - Complexes with Single-Atom Donor Dianionic Ligands 6.4.1 18-Electron Compounds The dianionic ligand in these compounds provides a double bond using the a1 and b2 metal orbitals, giving the same electron configurations discussed in Section 6.3.2 for related [MCp2X2] group 5 and 6 metal complexes (Table 6-2). Accordingly, one or two electrons are located in the third 1a1 orbital of the neutral paramagnetic [MCp2E] group 5 and diamagnetic group 6 metal complexes respectively. When this orbital is used to bond one additional substituent, the resulting 18-electron systems correspond to the neutral [MCp2EX] and cationic [MCp2EL]+ group 5 and [MCp2EX]+ group 6 metal complexes. These are the most important types of compounds described in this section, where paramagnetic [MCp2E] complexes are limited to the neutral d1 vanadium compounds.
6.4
Type C-Complexes with single-atom donor dianionic ligands
379
6.4.1.1 Neutral Group 5 d0 Metal Compounds [MCp2EX] Bis(cyclopentadienyl) complexes with one dianionic and one monoanionic ligand are more stable and more accessible than those with three monoanionic ligands discussed in Section 6.3. Many are formed spontaneously in order to avoid three substituents. These complexes constitute several families comprising a large number of compounds.
Oxo, Sulfido, Selenido and Tellurido Complexes The high oxophilicity of the group 5 metals is apparent from the variety and number of complexes formed. Dinuclear polinuclear and cluster complexes with µoxo bridges are numerous. Although considerably less common, a significant number of compounds with terminal M=O bonds can be obtained, in particular via oxidation of 18-electron, electron rich metal(III) derivatives (40) (Cp variously substituted) [99, 118, 120, 275 , 276]. Both oxo and peroxo products can result from O2 oxidations; [NbCp2X(CO)] gave either an oxo [NbCp2(=O)Cl] (X = Cl) or a peroxo derivative (X = Me). Less electron-accepting ligands (L = PMe3 or CNR) also favoured the peroxo complex [118,120,117]. The oxo complex was however obtained from [NbCp2X(CNR)] using t-butyl hydroperoxide as oxidising agent or by oxygen abstraction from the peroxo complex using PPh3 [120]. Careful control of the amount of oxygen used in the reaction shows it to be a multistep process involving the intermediate formation of both µ-oxo and peroxide species [280]. The µ-oxo species was in fact generated via a conproportionation reaction between [NbCp2(=O)Cl] and [NbCp2X)]. The oxo derivatives [Nb(η-C5H4Me)2(=O)R] (90) (R = CH2But,CH2Ph or Me) were obtained via thermolysis of the appropriate carbon dioxide precursor [277].
380
6
Group 5 and 6 Metallocenes
Likewise the tantalum (III) derivative (91) used in situ was a convenient precursor to (92) and (93) Scheme 6-20 [278]. (92) was also prepared by treatment of tantalum alkylydine complexes with water [279, 280].
Scheme 6-20
The heterobimetallic species (94-96) were conveniently prepared by the reactions shown in Scheme 6-21 [281 ,282]. The alkyl derivatives [MCp2(=O)Me] (M = Nb or Ta) were obtained by reaction of [NbCp2(=O)Cl] with LiR or “TaCp2Me”, generated in situ, with an appropriate oxirane; both have been structurally characterised [283, 284].
6.4
Type C-Complexes with single-atom donor dianionic ligands
381
Scheme 6-21 Insight into the nature of the metal oxygen bond has been gained through ir stretching frequencies and x-ray structural characterisations. Relatively low ν(M=O) values of ca 850 cm-1 are indicative of a bond order of less than 3; a bond order of between 2 and 3 is suggested by the metal-oxygen bond lengths (M=O: [Nb(η-C5H4SiMe3)2(=O)Me] [118] 1.720(2) Å; [NbCp2(=O)Cl] [285] 1.737(6) Å; [TaCp*2(=O)H] [284] 1.72(3) Å; (90) [281] 1.741 Å). The bond lengths in (95) (Ta=O 1.761(11) Å and Re-O 2.147(11) Å) were consistent with a Ta=O double bond with a Re-O donor-acceptor interaction. In common with oxygen, sulfur can adopt a variety of bonding modes including η-S2, µ-S sulfido bridges as well as terminal M=S. Reaction of the niobecene(III) boranate with S8 gave [Nb(η-C5Me4Et)2(=S)SH] (Nb-SH 2.381(1) Å and Nb=S 2.339(1) Å) together with the η-S2 complex [286]. A series of sulfido permethyltantalocene derivatives (97) were reported (Scheme 6-22) and their capacity to effect catalytic hydrodesulfurization was investigated [272, 287]. Hydrogenation of (97) under forcing conditions gave (98); further hydrogenation to (99) did not however take place even at 200°C. In fact reaction of (99) with H2S at room temperature gave (98), making their use in hydrodesulfurization unlikely. [Ta(C5H4But)2(=S)I] or [Ta(C5H4But)2(=S)H] were obtained from reactions of the persulfido hydride with MeI or PPh3 [288].
382
6
Group 5 and 6 Metallocenes
Scheme 6-22 In a reaction analogous to that discussed for oxygen, heteroatom abstraction from thiiranes has provided a convenient route to the sulfido complexes (100) and (101), Scheme 6-23. The structure of (102) was proposed on the basis of spectroscopic and chemical evidence, although the presence of a species with two Ta(IV) centres was not ruled out [287,288]. Irradiation of a mixture of [TaCp*(η6-C5Me4CH2)(η2S2)] (103) and [TaCp*2(η2-S2)H] (104) by sunlight gave [TaCp*(η6-C5Me4CH2S) (=S)] and [TaCp*2(=S)SH] which were separated by chromatography [289].
6.4
Type C-Complexes with single-atom donor dianionic ligands
Scheme 6-23
383
384
6
Group 5 and 6 Metallocenes
A series of selenido and tellurido complexes analogous to those of sulfur have been reported [290]: [TaCp*2(η2-Se2)H] (105) was converted into [TaCp*2(=Se)(SeH)] (106) in the presence of light; reaction of either (105) or (106) with MeI gave [TaCp*2(=Se)I] (107); selenido- alkyl complexes [TaCp*2(=Se)R] (R = H (108), Me or CH2SiMe3 (109)) were obtained via reactions of (107) with either alkyllithium or Grignard reagents; (108) was also obtained upon reaction of either (105) or (106) with PMe3. The tellurido-hydride [TaCp*2(=Te)H] was obtained in a similar way from [TaCp*2(η2-Te2)H] and from heating [TaCp*2(η2-TeCH2)H] [291]. Ta=Se bond lengths of 3.29(2) Å and 3.372(1) Å were reported for (108) and (109). Low field 77Se NMR chemical shifts (δ 1990-2363) ppm were reported for the selenido complex (cf δ +54 to -595 ppm for singly bonded species) [294].
Imides Relatively few complexes containing terminal imido groups have been reported, and most have been prepared using metal(III) derivatives as starting materials. In common with the oxide and sulfide complexes, [TaCp2(=NR)Me] (R = Me (110), Bun or Ph) was obtained via nitrogen abstraction from aziranes [287, 288]. Imido complexes were also obtained via dinitrogen extrusion from azido complexes and (110) was isolated by this route from the phenylazido complexes [TaCp2(η1N3Ar)Me] [292]. Reaction of [TaCp*2Cl(THF)] with N3Ph gave upon thermolysis the imide [TaCp*2Cl(=NPh)] (111) and with LiNHR gave [TaCp*2H(=NR)] (R = Ph (112) or But); complex (112) was also obtained by treatment of [TaCp*2 (=CH2)H] with aniline (P = Ph) [282, 284]. Reaction of [TaCp*2Cl2] with sodium amide gave [TaCp*2H(=NH)].
6.4
Type C-Complexes with single-atom donor dianionic ligands
385
Complexes (111) and (112) have been structurally characterised and in both compounds the Ta-N-C angle [176.4° (111) and 177.8° (112)] is virtually linear, indicating that the nitrogen atom is sp hybridized (often indicative of Ta≡N). The Ta-N distance [1.799(4) Å (111), 1.831(10) Å (112)] is however longer than expected for a triple bond and a bond order of between 2 and 3 is thus proposed. Molecular orbital calculations on (112) show that a triple bond is not required to justify the linearity of the system [293]. Reaction of metallocene trihydrides with diazoalkanes gave the related diamagnetic η1-diazoalkane complexes of scheme 6-24; the reaction was thought to proceed through the intermediate (113) [294, 295, 296]. X-Ray structural determinations of (114), (115) and (116) reveal Nb-N bond lengths of ca 1.85 Å and a slightly bent Nb-N-N angle (ca 168 ° for (114) and (115) and an average 165° for (116)) with bond orders of slightly less than two suggested.
Scheme 6-24 Alkylidenes We have already observed that bis(cyclopentadienyl) trihalides are sterically congested compounds and that the related trialkyl complexes do not exist. The tantalum alkylidene compounds were the first reported examples of ”Schrock carbenes” and since then a huge number of alkylidene compounds have been synthesised. Most bis(cyclopentadienyl) derivatives [TaCp2(=CH2)R] were prepared by intermolecular α-hydrogen abstraction using an external base or intramolecular α-hydrogen transfer from an alkyl substituent. These compounds have been extensively reviewed [1, 2, 297].
386
6
Group 5 and 6 Metallocenes
The silyl substituted alkylidene complex [TaCp2{=C(H)SiH(But)2}H] (117) was obtained from the reaction of [TaCp2(=CH2)Me] with the hindered silane H2SiBut2 whereas its reaction with unhindered silanes gave a mixture of the alkene derivative [TaCp2(η2-CH2=CH2)Me] and the hydrido silyl complex [TaCp2(SiR3)2H]; the former reaction in benzene afforded the analogous phenyl methylene complex [298]. A number of other synthetic strategies have been employed including oxidation of the tantalum(III) complexes [TaCp2XL] (X = Me, L = CO [283]; X = Cl, L = THF [282]) or thermal rearrangement of η3-allyl derivatives to give vinylidine complexes [299]; several of those used to synthesise pentamethylcyclopentadienyl tantalum derivatives are summarized in Scheme 6-25. The η2-CS2 carbon disulfide complexes (Cp = η-C5H4SiMe3) react with activated alkenes RC≡CR (COOMe or COOBut) to give the 1,3-dithiol-2-ylidene derivatives (118) (X = Cl, Et,CH2CH2Ph) [300]. An extensive series of heterobimetallic species with methylene bridges joining tantalum to various late transition metals has been obtained from reactions of [TaCp2(=CH2)Me] with an appropriate late transition metal complex [2, 301, 302].
6.4
Type C-Complexes with single-atom donor dianionic ligands
387
Scheme 6-25
6.4.1.2 Cationic Group 5 d0 Metal Compounds[MCp2EL]+ Oxidation of [Nb(η-C5H4But)2(CH2Ph)2] with AgBPh4 gave the thermally stable cationic niobium benzylidene complex [Nb(η-C5H4But)2(=CHPh)(THF)]+ [303]. 6.4.1.3 Neutral Group 6 d2 Metal Compounds [MCp2E] These group 6 d2 metal compounds are loosely related to the [MCp2X2] derivatives discussed in Section 6.3.2. They too have an 18 electron closed shell metal centre with two orbitals implied in M-E bonding, in this case to give a σ-π double bond with the dianionic ligand and the third a1 orbital occupied by the pair of free electrons. Only strong π-donor ligands can stabilize this type of coordinatively unsaturated compound and not many of them have been reported. The simplest examples which illustrate their behaviour are the oxo derivatives [MCp2(=O)] (119) with various Cp ligands (C5H5, C5Me5, C5H4R), obtained from the reaction of [MCp2Cl2] with aqueous KOH [165, 304]. The molybdenum complex was also afforded by the photochemical reaction of [MoCp2H2] with water [305]. The M=O bond length of 1.72 Å and IR stretching frequencies ν(M=O) of between 800 and 860 cm-1 (cf ν(M≡O) ≈ 930 - 1000 cm-1) measured in (119) indicate a bond order of about two [293, 306]. Molecular orbital calculations are broadly in agreement with this bonding picture [307]. The M=O bond is thus relatively weak and fairly reactive: the
388
6
Group 5 and 6 Metallocenes
oxo ligand of [WCp*2(=O)] exchanges very rapidly with H218O; it reacts with H2 to give [WCp*2 H2] and conversion of one of the rings into an alkoxo and η1-coordinated ligand occurs in its reaction with O2 or H2O2 to give respectively [WCp*(O-Cp*)(O)2] or [WCp*(η1-Cp*)(O)2] [310]. The M=O undergoes [2+2] cycloaddition reactions with a range of electrophilic substrates [308, 198]. Reaction with a series of metal carbonyl complexes (CO)MLn+ (M’ = Mn, Re, Fe, Ru, or Pt) gave the heterodinuclear compounds [MCp2{(η2-O,O)C=M’Ln)}] and cycloadditions of several heterocumulenes gave (120) (M = Mo, E = NR, E’ = O; M = W, E = NR or O, E’ = CPh2, NR or O). Reactions with reagents such as MeOH, HX, Me3SiCl or carboxylic acids gave [WCp2X2)] (X = OMe, Cl, RCOO). Related sulfido derivatives have not been isolated and there is only indirect evidence for their formation, {MoCp2(S)} was thought to be the reactive intermediate which undergoes cycloaddition, analagous to that observed with the oxo derivatives in the reaction of [MoCp2S2] with di-p-tolylcarbodiimide [309]. A series of imidomolybdenum complexes [Mo(η-C5H4R)2(NR’)] (121) and their mixed ring analogues [MoCp(η-C5H4R)(NR’)] have been prepared from the reaction of a suitable mono(cyclopentadienyl) imido complex with an appropriate sodium cyclopentadienide [310, 311]. The molecular structure of (121), (R = But) implies sp hybridisation at the nitrogen atom (M=N 1.738(2) Å; Mo-N-C 177.7(2)°), however as with (112) a triple bond is not necessary to account for the near linearity of the Mo-N-C moiety, and photoelectron spectra and molecular orbital calculations support a bond order of about two with a ligand based lone pair. Reactions of (121) with HCl gas or LiAlH4 gave the corresponding dichlorides and dihydrides (Scheme 6-26).
Scheme 6-26
6.4
Type C-Complexes with single-atom donor dianionic ligands
389
Reaction of [MCp2HLi]4 with PCl2(PR) yielded the terminal phosphinidene complexes [MCp2(PR)] (M = Mo or W, R = 2,4,6-ButC6H2 [312]; M = W, Cp = Cp*, R = NH(2,4,6-ButC6H2) [313]); extremely low-field 31P NMR shifts of between 666 and 799 ppm were reported. The alkylidene tungsten derivative[WCp2(=CHOZrHCp*2)] was isolated from the reaction of the zirconium dihydride derivative [ZrCp*2H2] with [WCp2(CO)] [314].
6.4.1.4 Cationic Group 6 d0 Metal Compounds [MCp2EX]+ Few examples of this type of complex have been reported. Photolysis of [ WCp2(NCMe)Me]+ under O2 yielded the structurally characterized oxo complex [MCp2(=O)Me]+ (Mo=O 1.700(11) Å) [236]; the ethyl analogue was prepared via photolysis of [ WCp2(CH2CH2)H]+ exposed to air [315]. [MCp2(=S)X]+ was obtained by decomposition of the η2-thiosulphate derivative [MCp2(η2-(-S-SO2-O-)] [196]. Methylation of (121) with MeI gave (122) (Scheme 6-26) [314, 315]. For R = But, the molecular structure of (122), in common with (121), shows an almost linear Mo-N-C angle (176.2(7)°) and a Mo=N distance of 1.704(8) Å. Oxidation of [WCp2H(η2-CH2=CH2)] with I2 yielded [WCp2(=CHMe)I]+ [112].
6.4.2 17-Electron Compounds 6.4.2.1 Neutral d1 Vanadium Compounds [VCp2E] The imido derivatives [VCp2(NR)] (R = SiMe3 or SiPh3) and [VCp*2{NR] (R = Ph, 2,6-Me2(C6H3), 2-Ph(C6H3) or (C6F5)) obtained from reactions of organic azides with vanadocene are examples of this type of paramagnetic complex (Scheme 6-27) [316, 317, 318] when bulky azides (R = SiMe3 or SiPh3) were used with permethylated vanadocene, the azido complex was formed.
Scheme 6-27
390
6
Group 5 and 6 Metallocenes
Relatively low ν(MN) stretching frequencies, in the range 923-968cm-1, were reported [319]. X-ray crystal structures show that the V-N-R angle is almost linear (ca 179 °) implying that the nitrogen is sp hybridized, with short V=N distances of between 1.665 Å and 1.730(5) Å indicative of multiple bonding. In addition to the σ-bond, molecular orbital calculations for [VCp2(NPh)] show two V-N π-bonds with a nonbonding HOMO localized on vanadium. Measured µeff values of 1.561.79 µB are consistent with the presence of one unpaired electron. The complexes are thus formally 19e V(IV) paramagnetic species. Cyclic voltammograms show quasi-reversible oxidations to 18e vanadium (V) complexes. One cyclopentadienyl ligand was readily lost on heating.
6.5. Type D - Complexes with η2-Side-Bonded Ligands 6.5.1 18-Electron Compounds 6.5.1.1 Neutral Group 5 d2 Metal Compounds [MCp2(η2-X-L)] Compounds formed by monoanionic bidentate ligands fall into this category, although relatively few of these metal(III) derivatives have been isolated. Alkylation of the chloro isocyanide derivatives [Nb(C5H4SiMe3)2Cl(CNR’)] with LiR gave the corresponding iminoacyl complexes (123) (R = Me, R’ = Cy, Ph; R = CH2SiMe3, R’= Cy) [119]. Insertion of the isocyanide moiety into the NbR bond of the presumed alkyl isocyanide intermediate occurs spontaneously, except in the case of the tert-butyl isocyanide derivative when the alkyl isocyanide complex was isolated. An alternative synthetic route (R = R’ = Ph) involved Na/ Hg reduction of the ketenimine complex [Nb(η5-C5H4SiMe3)2(η2-(C,N)PhN=C=CPh)] followed by protonation [320].
6.5 Type D-Complexes with η2-side-bonded ligands
391
The η2-dithioformate (124) was obtained via CS2 insertion into the M-H bond of [Nb(C5H4SiMe3)2HL)] (L = P(OR)3) with substitution of L by sulfur (Scheme 628) [321]; when L = CO insertion of CS2 occurs, but without the following ligand substitution, to give the η1-dithioformate (125).
Scheme 6-28 One of the tautomeric forms of the heterodinuclear niobium derivative [NbCp2(η2C,O)(CH2-O-ZrHCp*2] belongs to this group and was prepared by reacting [NbCp2R(CO)] with zirconocene dihydride [157]. 6.5.1.2 Neutral Group 5 Metal Complexes [MCp2X(η2-C-C)] or [MCp2X(η2-C-Y)] (Y = N, O, or S) Complexes with ligands which formally contain either a carbon-carbon multiple bond or a carbon- heteroatom double bond interacting with the metal are discussed here. The bonding in this category can be ambiguous and therefore we have not specified a general metal electron configuration. Some complexes are best described as M(III) d2 whilst others sit more comfortably in the M(V) d0 category. The neutral or dianionic nature of the ligand occupying one or two coordination sites respectively depends on the degree of π-back donation from the metal to the empty π* orbitals of the ligand, and is influenced by the other substituents bonded to either the metal or the ligand. As an illustration, structure (126a) shows a formally M(III) alkene adduct whilst structure (126b) shows a complex with considerable M(V) character implying a significant π-bonding contribution from the metal to the ligand. The hydride alkenes and alkynes are important and widely studied groups in this general category.
392
6
Group 5 and 6 Metallocenes
Alkene complexes A convenient preparative method (Scheme 6-29) for hydride alkene complexes is the Grignard reagent MgClR reduction of a metal(IV) dichloride, probably proceeding through the intermediate formation of the metal(III) alkyl compound [Cp2MR] which finally gives the alkene hydride via β-hydrogen elimination [282, 322, 323]. The same compounds were also obtained from the reaction of Cp*2NbH3 with the corresponding alkene. Both endo- and exo- isomers were found in complexes where Cp = Cp, whereas the endo- isomer was favoured for steric reasons in complexes where Cp = Cp* [258]. Spin-saturation transfer NMR experiments have demonstrated that these insertion-elimination processes take place via species containing agostic M-H-C bonds [324, 325]. Ligand promoted alkene insertion into the M-H bond allows the isolation of the intermediate alkyl metal(III) derivatives, and the kinetics of this reaction have been extensively explored [258, 326].
Scheme 6-29
6.5 Type D-Complexes with η2-side-bonded ligands
393
Both methods were used to prepare [Nb(η5-C5H4SiMe3)2(η2-PhCH=CH2)H], with the first producing exclusively the endo- isomer, while the second gave a mixture of endo- and exo- forms [327]. The analogous endo-bis(indenyl) derivative was obtained from the reaction of [Nb(Ind)2(µ-H2BH2)] with styrene in the presence of NMe3; the complex is fluxional with both dissociative and non-dissociative mechanisms proposed to account for the observed NMR exchange [126]. Reaction of a trihydride complex with the activated alkyne RC≡CR’ (R = R’ = CO2Me or CO2But; R = CO2Me, R’ = H) gave not the alkyne hydride as might be expected, but the alkene hydride (128) via the insertion product (127) (Scheme 630) [328].
Scheme 6-30 A series of alkenyl complexes [Nb(η-C5H4SiMe3)2(η1-C(R)=CH(R’)L] (L = CNR or CO) were obtained from the reaction of appropriate metal(III) hydride isocyanide or carbonyl complexes with similar activated alkynes via alkyne insertion into the metal hydrogen bond. Reaction of the appropriate alkene with [NbCp2(H2)(SnMe3)] afforded the alkene stannyl derivatives [NbCp2(η2H2C=CHR)(SnMe3)] (R = Ph or CMe=CH2) [262]. Alkyne derivatives were obtained in a similar way. X-Ray molecular structures of several of these alkene derivatives show lengthening of the ligand C=C bond in the range 1.429(7) Å to 1.4486(6) Å (cf uncoordinated ethene 1.337(2) Å) indicative of metal to alkene back donation introducing significant (126b) character.
Alkyne complexes Several methods for the preparation of alkyne metal(III) complexes have been reported (Scheme 6-31): a) A series of halide derivatives [NbCp2X(RC≡CR’)] (X = Cl, Cp = Cp or ηC5H4SiMe3, R = Ph or H, R’ = Ph or H; X = Br, I, Cp = η-C5H4SiMe3, R = R’ =
394
6
Group 5 and 6 Metallocenes
Ph; X = Cl, Cp = η-C5H4SiMe3, R1 = H, Me or COOMe, R2 =COOMe) have been isolated by reduction of [MCp2Cl2] with various reducing agents, usually Na/Hg, in the presence of the alkyne [106, 117, 329]. Reduction of (80) gave the ansaderivative [Nb{SiMe2(C5H4)2}Cl (RC≡CR)] (R = Me or Ph) [224]. b) Hydride and alkyl derivatives have usually been prepared by reaction of the hydride alkene complexes with an appropriate alkyne. When the alkyne has bulky substituents, a simple substitution takes place to give the hydride alkyne derivatives, whereas less bulky substituents induce hydride migration to give the alkyl insertion products. For example, reactions of [MCp2H(CH2=CHMe)] with RC≡CH (R = But, SiMe3) lead to the simple substitution products [MCp2H(RC≡CH)] [330]. Monosubstituted acetylenes RC≡CH (R = Me, Et, Ph) however react with [MCp2H(CH2=CH2)] to give a mixture of endo- and exo- [MCp2Et(RC≡CH)], or the ethyl complex [MCp2Et(MeC≡CMe)] is obtained when using MeC≡CMe. The same method was used to prepare the benzyne complex [TaCp*2H(η2-C6H4)] [245]. c) The direct thermal reaction of the trihydride with an appropriate alkyne provides a route to [NbCp2H(RC≡CR)] (R = Et, Ph, SiMe3) [331]. This method was unsuitable for the methyl derivative, which was obtained by photolitic exchange of alkyne from the reaction of [NbCp2H(Me3SiC≡CSiMe3)] with MeC≡CMe.
Scheme 6-31 In common with the alkene complexes discussed above, molecular structures show that lengthening of the C≡C bonds (1.230 - 1.288Å) occurs in the coordinated alkyne ligand, together with a significant deviation from linearity in the C≡C-R angles (137.7 - 146.7°), imparting a high degree of metallacyclopropene character [117, 224, 334]. The complexes [Nb(η-C5H4SiMe3)Cl(η2-R1C≡CR2)](R1 = H, Me or COOMe, R2 =COOMe) and [NbCpH(η2-HC≡CR)]334 (R = SiMe3 or But) were reported as Nb(V) species.
6.5 Type D-Complexes with η2-side-bonded ligands
395
Complexes with (η2-CY2) ligands (Y = O or S) Complexes with η2-CO2 were prepared via chemical or electrochemical reduction of a dihalide complex in the presence of CO2 [281, 332, 333, 334]. An alternative approach was the oxidation with O2 of the carbonyl ligand in the [NbCp2R(CO)] (Cp variously substituted, R = CH2Ph, CH2But, CH2SiMe3, Me) derivatives [335]; labelling experiments show the additional oxygen atom in the resulting [NbCp’2R(η2-CO2)] products is derived from O2. Complexes are generally stable but lose CO when heated at 60°C to give the oxo compound (90); all exhibited a strong ν (C=O) band in their IR spectra at ca. 1700 cm-1, indicative of η2-coordinated CO2. The chloride complexes [NbCp2Cl(η2-CO2)] (Cp = Cp’, Cp*), generated electrochemically and identified from their IR spectra (ν (C=O) 1730 cm-1), contain labile CO2 moieties as evidenced by their FD-MS spectra [121]. A range of similar compounds [MCp2R(η2-CS2)] were prepared by a) carbon disulfide promoted hydride-alkene insertion (Cp = Cp or Cp’) [304, 336]; b) photochemical substitution [340]; c) reaction of a metal (III) derivative with carbon disulfide [106], Scheme 6-32. Complex (129) was obtained cleanly when reaction a) was carried out in nonpolar solvents, however the use of polar solvents gave (129) together with the dithioformato deriveative [NbCp’2(CS2H)(η2-CH2=CH2R)] and free alkene. Complex (129) can act as a ligand through its uncoordinated sulfur atom and several heterobimetallic species have been reported [325, 340] it reacts with RI to give [MCp2R(η2-CS-SR)]+ [270].
Scheme 6-32 Complexes with (η2-C,Y) ligands (Y = N, O or S) A wide range of complexes with (η2-C,Y) ligands (130-132) have been reported [2]. The most widely used synthetic route to aldehydes (130) Y = O [337], isocyanates (131) [106], ketenes (132) (Y = O) [101, 338, 339, 340, 341, 342] and ketimines(132 Y = NR) [324, 343, 344] has been the reaction of a metal(III) chloride [MCp2Cl]2 (obtained via the reduction of [MCp2Cl2] [106] or via the reaction
396
6
Group 5 and 6 Metallocenes
of NbCl3(DME) with LiCp [101]) with an appropriate free ligand. The aldehyde (130) (Y = O, R1 = R2 = H) was also obtained from precursors such as [Nb(ηC5H4SiMe3)Cl(L)] (L = CO or phosphine), and its hydrido analogue was obtained by the reduction of the chloro derivative in the presence of ethanol via the d1 Nb(IV) radical intermediate {Nb(η-C5H4SiMe3) (η2-CH2O)} [341]. Hydrido ketone, aldehyde and thioaldehyde complexes were also prepared from the reactions of alkylidenes [TaCp*2H(=CR)] (R = H2 or =CH2) with MeOH or HSCH2R; the reaction was thought to proceed via β−hydrogen elimination from a transient methoxide or sulphide {TaCp*2H(CHR)(YCH2R’)} (Y = O or S) species to give the hydrido aldehyde or thioaldehyde complexes [TaCp*2H(η2-O=CH2)] [283] or [TaCp*2H(η2-S=CHR’)] [272, 345]. The alkyl complexes [TaCp*2Me(η2O=CHR)] were obtained in a similar way [271]. The telluroformaldehyde derivative [TaCp*2H(η2-Te=CHR)] was prepared by reaction of the tantalum alkenyl complex with Te in the presence of PMe3 [295], and the selenoaldehyde [TaCp*2H(η2-Se=CHR)] was prepared by reaction of [TaCp*2I(=Se)] with LiMe.294Reaction of the cationic enacyl complex [Nb(η-C5H4SiMe3)Cl(η2O,C)O=CCR=CMeH)]+ with KOBut gave the E-vinylketene complex (R = Ph or Et) [346 ].
When more than one coordination mode is possible, as in the heterocumulenes, the oxophilicity of the group 5 metals determines the preferential coordination in the order O > N > C, so that the resulting products always contain η2-(O,C)-ketenes, η2-(N,C)-ketenimines or η2-(O,C)-isocyanates. X-Ray molecular structures of the aldehydes, ketenes and ketimines (Table 6-5) show that considerable lengthing of the C-C and C-Y bonds in the π complexed η2-ligand occurs, compared to the free cumulenes (cf C=O 1.18(1) Å, C=C 1.29(1) Å measured for dimesityl ketene [347]), with lengths approaching those of C-Y single bonds (cf C-O 1.333 Å), thus the complexes are best characterized as Nb(V) d0 species. Likewise the telluroformaldehyde complex was thought to have substantial Ta(V) tellurirane character [295].
6.5 Type D-Complexes with η2-side-bonded ligands
397
Table 6-5. Selected X-Ray diffraction data on neutral aldehydes, ketenes and ketimines, and their cationic acyl and iminoacyl analogues.
Allyl Complexes [MCp2(η3-allyl)] The η3-allyl complexes of d2 niobium and tantalum can be viewed as a particular type of [Cp2MXL] complex if the allyl substituent is considered as an η1-η2-ligand. We have already observed in Section 6.3.1.5 that paramagnetic 16-electon vanadium complexes resist the coordination of an additional ligand because of the high spin pairing energy required, and for this reason vanadium only has η1-allyl derivatives. In niobium and tantalum allyl complexes, all three metal based d orbitals are involved in bonding, always giving 18 electron η3-allyl complexes.
Several methods have been used to prepare group 5 metal allyl derivatives (133); several are based on a strategy which involves the insertion of allene or dienes into the Nb-H bond of metal(III) hydrides [MCp2HL] or their precursors, after displacement of the ligand. The reaction of [TaCp2H3] (Cp = Cp or Cp*) with butadiene, for example, takes place with elimination of H2 [303,348]; [NbCp2H(CH2=CH2)] however reacts with dienes eliminating ethene (the tantalum derivatives were not
398
6
Group 5 and 6 Metallocenes
obtained via this route) [349] and [Nb(Indenyl)2(η2-H2BH2)] eliminates BH3 when treated with diene in the presence of NMe3 [126]. Reactions of the dichlorides with allylmagnesium halides or 2-butenyl magnesium halides were used to obtain the derivatives [TaCp*2(η-C3H5)] and [MCp2(η-C3H4Me)] (M = Nb or Ta) [353, 303]. With the exception of allene all the reactions with dienes can lead to the formation of various syn-anti and endo-exo isomers and the regio- and stereo-selectivity of these processes was studied by labelling experiments and NMR spectroscopy [353]. The allyl complexes have been reported as fairly stable, scarcely reactive compounds, as expected for 18 electron metal systems. 6.5.1.3 Anionic Group 5 d2 Metal Compounds [MCp2(η 2-C,Y)]+ and [MCp2X(η 2-C,Y)]Electrochemical reduction of the ketene complexes (132) afforded the radical anions [Nb(η-C5H4SiMe3)2Cl{η2(C,O)-(OCCR1R2)}]- (R1 = R2 = Ph; R1 = Et, R2 = Ph) characterised by their typical 10 line ESR (giso 2.000 or 1.997, a(Nb93) 16 G) spectra and IR spectroscopy. Low giso values (cf. [NbCp2X] ca. 100 G) are often found in η2-π-complexed ligands [345]. Electrochemical two electron reduction of the ketimine (132 R1 = R2 = Ph) afforded the cationic species [Nb(η-C5H4SiMe3)2{η2(C,N)-(NPhCCPh2)}]+ [324]. 6.5.1.4 Neutral Group 6 d2 Metal Compounds [MCp2(η 2-X-X)] Metal complexes with η2-coordinated ligands where the two donor atoms are bound together are considered in this section as d2 metal compounds [MCp2(η2-X-X)], which result from oxidative addition of an unsaturated system to the metallocene fragment, although they could also be viewed as [MCp2 (η2-L)] type d4 metal compounds with unsaturated side-bonded neutral π-complexed ligands. Metal complexes with η2-coordinated acetonitrile [MCp2(η2-N≡CMe)] can be isolated either by reduction of [MoCp2Cl2] or by decomposition of [WCp2HMe], in acetonitrile [350, 351, 352]. In addition to its reaction with ethene to give (18), the bimetallic fulvalene hydride complex (15) when treated with but-2-yne gave a mixture of [Mo2(µ-η:ηC10H8)Cp2(MeC≡CMe)2] and [Mo2(µ-η:η-C10H8)Cp2(µ-MeC≡CMe)]; in the latter the alkyne group is thought to bridge the two metals. In contrast, reaction with hex-3-yne afforded only the unbridged analogue [87]. The synthetically useful Z-stilbene complex [MCp2(-(Z)-(PhCH=CHPh)], obtained via insertion of PhC≡CPh into the Mo-H bonds of the dihydride [MCp2H2], has provided a route to a variety of η2 derivatives [353, 354]. Isomerisation of the Z-isomer to the E-isomer is achieved in the presence of catalytic amounts of maleic anhydride. The E-isomer is significantly less labile and is not readily substi-
6.5 Type D-Complexes with η2-side-bonded ligands
399
tuted [355]. The labile cis-stilbene adduct is readily replaced by alkynes to give [MCp2(η2-RC≡CR)] and by other π-acceptor ligands, such as heteroalkenes [MoCp2(η2-X-X)] (X-X = PhCOCF3, Ph2Cl, PhCOCO2Me, PhCHO, PhCHNPH, Ph2CS H2CO2 [359, 356 , 357], CO2 [361]); it also provides access to a range of metal(IV) compounds by oxidative coupling reactions (Section 6.3). The structurally characterized complex [MoCp2(η2-CO2)] [361] was also observed as an intermediate in the reaction of the dihydride with CO2 (Section 6.3.2.4) [194]. The related [MoCp2(η2-C,S)CS2)] complex was formed in the unusual reaction of its isomer [MoCp2(η2-S,S)CS2)] with alkynes which are simultaneously transformed into the metallacyclic dialkyls [MoCp2(η2-C,C)C(=S)SCR=CR)] [358]. The silene [WCp2(η2-CH2=SiMe2)] [359] and disilene [MCp2(η2-SiMe2=SiMe2)] (M = Mo or W) derivatives [360] important compounds for their synthetic applications, were obtained from [WCp2Cl(RSiMe2Cl)] (R = CH2 or SiMe2) when treated with magnesium. Their Si-Si [2.260(3) Å] or Si-C [1.800(8) Å] bond lengths lie midway between single and double bonds. These compounds are very reactive and useful in the preparation of an extensive range of silyl derivatives and metalla-silaand disila-cyclo-butanes and -pentanes [187, 204, 205]. 6.5.1.5 Cationic Group 6 d2 Metal Compounds [MCp2(η 2-X-L)]+ This category comprises cationic metal(VI) complexes [MCp2(η2-X-L)]+ with one anionic and one neutral ligand linked together, such as acyls. The molybdenum acyl complex [MoCp2(η2-CMe=NR)]+ was obtained on heating the cationic isocyanide [MoCp2Me(CNR)]I derivative [80]. Reactions of [MoCp2Br2] with either a diarylazenine or diarylamidine yielded the triazenido [MoCp2(ArNNNAr)]+ and amidinato [MoCp2(ArNCRNAr)]+ (Ar = Ph, p-MeOPh, p-MePh, p-FPh; R = CH, CMe, CPh) respectively, and their redox chemistry was investigated [361, 362]. Reaction of either the cationic dimethyl complex [WCp2Me2]+ with the Ph3C radical or the neutral dimethyl [WCp2Me2] with Ph3C+ resulted in their conversion to the alkene hydride derivative [WCp2H(η2-H2C=CH2)]; the reaction occurred via initial formation of an alkylidine species (which could be trapped on addition of phosphine as a phosphorus ylide) followed by methylene insertion into the M-Me bond to give the 16 electron alkyl intermediate, and finally β-hydrogen elimination to give the product [363, 364, 365, 366]. The mechanism of the final stage was further studied by deuterium labelling of isopropylmagnesium bromide in its reaction with [WCp2Cl2] [111]. An example of a dicationic group 6 metal complex is the molybdenum derivative [MoCp2(NCMe)(η2-(NH=CHMe)]2+ which resulted from protonation of the side-bound acetonitrile complex [MoCp2(η2-N≡CMe)] with HBF4. Et2O [354]. Its x-ray molecular structure showed bending at the protonated N and C atoms with a very short N=C distance of 1.096(16) Å (cf. N≡C 1.200(10) Å in the acetonitrile
400
6
Group 5 and 6 Metallocenes
complex [355]). Both ligands were replaced on addition of PMe3 to give [MoCp2(PMe3)2]2+. 6.5.1.6 Neutral Group 5 d0 Metal Compounds [MCp2(η 2-X-X)Z] Complexes of the type [MCp2(η2-X-X)Z] with ligands considered to be η2-dianionic species, such as η2-O2, η2-ONMe, η2-S2, η2-Se2, or η2-Te2, are members of this class and can be clearly defined as 18 electron group 5 metal d0 derivatives. Oxidation of metal(III) complexes has been widely used to prepare oxo derivatives with teminal M=O ligands and carbonyl coomplexes are the most convenient substrates to make this transformation. However, oxidation of related metal(III) isocyanides frequently leads to the formation of η2-peroxides and the complexes [NbCp2X(η2-O2)] (X = Cl, Me; Cp = C5H5, η-C5H4SiMe3 or ?-C5H3(SiMe3)2) were conveniently prepared by this method [120]. The methyl derivative was also obtained by oxidation of [NbCp2Me(CO)] (Cp = Cp), whereas oxidation of the chloro carbonyl derivative gave the terminal oxo-species [118]; hydrogen peroxide oxidation of the dichloride complex [NbCp2X2] has also been used [367]. Reaction of [NbCp2Cl]2 (Cp = η-C5H4SiMe3) with excess O2 gave a mixture of the terminal oxide and the peroxide (section 6.4.1.1) [280]. The tantalum derivatives [TaCp*2R(η2-O2)] (R = Me, Et, Prn, Ph or CH2Ph) were synthesized by O2 oxidation of either the alkylidene complexes [TaCp*2H(=CHR)] (R = H or Ph), the benor the alkene hydride complexes zylidine hydride [TaCp*2H(η2-C6H4)] [TaCp*2H(CH2=CHR)] (R = H or Me) [368]. The peroxo complexes react with triphenyl phosphine to give the terminal oxo derivatives [MCp2X(=O)] (M = Nb or Ta, X = Cl or Me) and triphenyl phosphine oxide. An analogous complex [TaCp*2Me(η2-ONMe)] was isolated by oxidation of [TaCp*2Me2] with NO [369]. Analogous h2-disulfido, -selenido or -tellurido derivatives were obtained from reactions of the dialkyl, alkyl hydride or trihydride derivatives with S8, Se or Te; sulfur extraction from thiiranes was also used (Scheme 6-33) [237,287,288,292,294,295]. Similar reactions with either [NbCp*2(BH4)] or [Nb(η2-C5H4But)2H3)] afforded the polysulfide complexes [{NbCp2(η2-S2)}(µ-S5)] (Cp = Cp* or (η2-C5H4But) [290, 370]. The fulvene derivatives [NbCp*(η6-C5Me4R)(η2-S2)] [R = CH2 or CH2-S (134)] together with the hydrido and thio derivatives were observed in reactions of S8 with {NbCp*}. Likewise, the tantalum fulvene derivative [TaCp*(η6C5Me4CH2)H2] reacted with sulfur to give a mixture of (103) and (104); the tantalum analogue of (134) was obtained upon irradiation of this mixture in the presence of sulfur [293, 25]. [M(η-C5H4But)H(η2-S2)] (M = Nb or Ta) reacts with sulfur to give the corresponding thio derivative and with MeI to give the iodo derivatives [371]. Heterobimetallic complexes formed by addition of Cr(CO)5 fragments to one or both of the sulfur atoms have been reported [372].
6.5 Type D-Complexes with η2-side-bonded ligands
401
Scheme 6-33
6.5.1.7 Cationic Group 5 d0 Metal Compounds [MCp2X2L]+ and [MCp2 (η2-X-L)Z]+ Most of the cationic derivatives which fall into this category were prepared from complexes with (η2-C,Y) ligands (Section 6.5.1.2). Cationic acyl and iminoacyl derivatives (135) were obtained by protonation of appropriate ketene complexes with HBF4.OEt, or ketimine complexes with HBF4. Protonation occurs at the βcarbon of (131) to give (135) as stable diamagnetic solids. An extensive series of complexes in each class have been reported: acyl derivatives [343,344,350] [Nb(ηC5H4SiMe3)2(η2-C(O)CHR1R2)X]+ [(X = Cl: R1 = Me, R2 = Ph, Me or CH2But; R1 = Ph, R2 = Ph, Et or CH2But) (X = H: R1 = Ph, R2 = Et or Me; R1 = Me, R2 = CH2But)]; iminoacyl derivatives [347, 348] [Nb(η-C5H4SiMe3)2(η2-C(NR)CHR1R2)X]+ [(X = Cl or Br: R1 = R2 = Ph; R1 = Ph, R2 = Me or Ph) (X = F, R1 = R2 = Ph)]. Selected bond distances and angles for structurally characterized examples are given in Table 6-5.
402
6
Group 5 and 6 Metallocenes
Electrochemical or chemical oxidation of the Nb(IV) ketimine complex (136) in the presence of NCR or CNR gave the air stable cationic ketimine complex (137 R = Me, Et or Ph, L = MeCN; R = Ph, L = ButCN or PhCN) (Scheme 6-34). Complex (137) reacted with water or MeOH to give the corresponding cationic hydrido or methoxy iminoacyl complexes [373]. The C-N (1.33(1) Å) and C-C (1.36(2) Å) bond distances are similar to those found in the neutral ketimines but shorter and longer respectively than those measured for the cationic iminoacyl complexes (Table 6-5).
Scheme 6-34
Oxidation of the d1 metal derivative [Nb(η-C5H4SiMe3)2(η2(C,C)-(RC≡CR’)]+ with [FeCp2][BPh4] yielded the cationic alkyne complexes [Nb(ηC5H4SiMe3)2(NCMe)(η2-(RC≡CR’)]+ (R = R’ = Me, Ph, COOMe; R = COOMe, R’ = Me (138)), in a reaction analogous to that of scheme 6-34 [153]. An X-ray molecular structural determination of (138) shows lengthening of the C-C bond (1.30(1) Å) which is only slightly shorter than a typical carbon-carbon double bond. The observed ν(C=C) stretching frequencies of between 1750 and 1850 cm-1 and the non-linearity in the C2-C1-R angle (ca. 141°) indicate a high degree of sp2 character in the coordinated carbon atoms. The COOMe substituent is located in the expected endo- position and the actonitrile ligand is almost linear with Nb-N-C and N-C-Me angles (ca. 177°). Both the structural and IR evidence strongly support the view that the alkyne behaves as a strong π-acceptor ligand, giving rise to a high degree of metallacyclopropene character in the complex.
6.5 Type D-Complexes with η2-side-bonded ligands
403
Although these complexes can be formally considered as either d2 alkyne or d0 niobacyclopropene derivatives, the latter description is more in accord with the structural and spectroscopic evidence. Its d0 character is further supported by the unusual oxidation of the same starting complex which under appropriate conditions gives the cationic d0 bimetallic derivative (70). Alkylation of [NbCp2R1(η2-CS2)] with R2X (X = Br or I) yielded [MCp2R1(η2(C,S)-CSSR2)]+ (R1 = Bu or Me, R2 = Me, Et, Pri, CH2COOEt, allyl) and with CH2I2 yielded the bimetallic complex [{NbCp2R1(η2(C,S)-CSS)}2(µCH2)]+ [270, 340].
6.5.2 17-Electron Compounds 6.5.2.1 Neutral d1 Vanadium Compounds [VCp2(η2-X-X)] Addition of various unsaturated substrates to vanadocene (a carbene-like complex) yields a diverse group of complexes (see also section 6.2.1.1). Activated alkenes with electron-withdrawing groups such as propenal, fumarates and diethyl maleate react with vanadocene at room temperature affording vanadacyclopropane derivatives [VCp2(η2-R2C-CR2)]; similar reactions occur with alkynes to give vanadacyclopropene complexes [VCp2(η2-RC=CR)] [374]. Their chemical and structural behaviour was as expected for typical vanadium (IV) derivatives [1,2]. They are paramagnetic compounds with magnetic moments of ca. 1.7-1.8 µB consistent with the presence of one unpaired electron and exhibited typical vanadium(IV) EPR spectral parameters. In common with the niobium and tantalum alkene complexes discussed in section 6.5.1.2, molecular structures show that lengthening of the C-C bonds [1.468 Å in the maleate derivative (only slightly shorter than expected for a single bond) and between 1.269 and 1.276 Å in the coordinated alkynes] occurs in
404
6
Group 5 and 6 Metallocenes
the coordinated alkene or alkyne ligand. This, together with a significant deviation from linearity in the alkyne C-C-R angles (142 - 143°), is consistent with the presence of either cyclopropane or cyclopropene ligands; the ν(C=C) stretching vibration of between 1750 and 1800 cm-1 is also consistent with a metallacyclopropene derivative. Similar reactions occured between vanadocene and a range of unsaturated substrates including aldehydes [375], ketenes, ketimines [376], thiobenzophenone [377] thioketenes [378, 379] and isothiocyanate [380] to give the corresponding cyclometallated product. In common with the niobium and tantalum derivatives, magnetic susceptibility measurements (µeff= 1.74 - 1.77 µB), IR, ESR, and crystallographic data were consistent with vanadium(IV) d1 metal derivatives. 6.5.2.2 Niobium Compounds [NbCp2(η2-L-L)] and [NbCp2(η2-C,Y)] Electrochemical or sodium amalgam reduction of the ketene or ketimine complexes (131 R1 = R2 = Ph) afforded the paramagnetic metal(IV) derivatives [Nb(ηC5H4SiMe3)2{η2(C,Y)-(YCCPh2)}] (Y = NPh or O) [324,377]. Their ESR spectra each comprised a ten line signal (giso 1.997 and 1.999, a(93Nb) 18.1 G and 11.0 G respectively). Similar results were obtained for the aldehydes [Nb(η-C5H4SiMe3)2 {η2(C,O)-(OCHR)}] (R = H or CHPh2) (giso 2.003, a(93Nb) 10.3 G and 8.3 G respectively) [341, 381] . No hyperfine splitting was however observed for the CS2 complex [Nb(η-C5H4SiMe3)2{η2(C,S)-(CS2)}] (giso 2.004). Extended Hückel calculations were consistent with the ESR data, and show that in the aldehyde, ketene, ketinimene and alkyne complexes the unpaired spin density is mainly localised on the metal, whereas in the carbon disulfide complex there is a high degree of delocalisation onto the ligand [385]. Likewise, reduction of [Nb(η-C5H4SiMe3)Cl(η2-R1C≡CR2)] afforded the neutral paramagnetic alkyne derivatives [M(η-C5H4SiMe3)2{η2-(CR1≡CR2)}] (R1 = R2 = Ph; R1 = H, R2 = Ph; R1 = H, R2 = COOMe; R1 = Me, R2 = COOMe; R1 = R2 = COOMe) [333,153]. Their ESR spectra showed a characteristic ten line signal with the small hyperfine coupling constants typically found in d1 complexes with η2 πbonded ligands (giso 2.0026 - 20140; a(93Nb) 12.7 G - 16.5 G). Although these compounds could also be considered as d3 metal complexes with one side-bonded neutral ligand [MCp2(η2-L-L)], their chemical and structural behaviour is better understood if they are classified as d1 metal derivatives.
References
405
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [35] [37] [38] [39] [40] [41]
Wilkinson G, Stone F G. A, Abel E W, Comprehensive Organometallic Chemistry, Pergamon, Oxford, 1982. Wilkinson G, Stone F G A, Abel E W, Comprehensive Organometallic Chemistry II, Pergamon, Oxford, 1995. Poli R, Chem. Rev. 1996, 96, 2135. Wigley D E, Prog. Inorg. Chem. 1994, 42, 239. IUPAC Commission (CNIC) meeting, Maryland, USA, 1996. Hollis T K, Burdett J K, Bosnich B, Organometallics, 1993, 12, 3385. Timofeeva T V, Lii J-H, Allinger N L, J. Am. Chem. Soc. 1995, 117, 7452. Lauher J W, Hoffmann R, J. Am. Chem. Soc. 1976, 98, 1729. Bürgi T, Berke H, Wingbermühle D, Psiorz, C, Fox T, Knickmeier M, Berlekamp M, Fröhlich R, Erker G, J. Organomet. Chem. 1995, 497, 149. Fernández F J, Gómez-Sal P, Manzanero A, Royo P, Jacobsen H, Berke H, Organometallics, 1997, 16, 1553. Green M L H, J. Organomet. Chem. 1995, 500, 127. Handir K, Holecek J, Klikorka J, Z. Chem. 1979, 19, 265. Köhler F H, Prössdorf W, Z. Naturforsch. Teil B, 1977, 32, 1026. Robbins J L, Edelstein N, Spencer, B, Smart J C, J. Am. Chem. Soc. 1982, 104, 1882 Gambarotta S, Floriani C, Chiesi-Villa A, Guatini C, Inorg. Chem. 1984, 23, 1739. Belot J A, McCullough R D, Rheingold A L, Yap G P A, Organometallics 1996, 15, 5062. Bocarsley J R, Floriani C, Chiesi-Villa A, Guastini C, Inorg. Chem. 1987, 26, 1871. Birmingham J M, Fischer A K, Wilkinson G, Naturwissenschaften 1955, 42, 96. Lindsell W E, Parr R A, Polyhedron 1986,6, 197. Jonas K, Rüsseler W, Krüger C, Raabe E, Angew. Chem. 1986, 98, 905. Jonas K, Wiskamp V, Z.Naturforsch. 1983, 38b, 1113. Elson I H, Kochi J K, J. Am. Chem. Soc. 1975,97, 1262. Nesmeyanov A N, Lemenovskii D A, Fedin V P, Perevalova E G, Dokl. Akad. Nauk SSSR (english Transl.) 1979, 245, 142. Brunner H, Gehart G, Meier W, Wachter J, Riedel A, Elkrami S, Mugnier Y, Nuber B, Organometallics 1994, 13, 135. Brunner H, Gehart G, Meier W, Wachter J, Burgemeister T, J. Organomet. Chem. 1995, 493, 163. Kowaleski R M, Basolo F, Osborne J H, Trogler W C, Organometallics, 1988, 7, 1425. Warren K D, Struct. Bonding 1976, 127, 45. Köhler F H, Geike W A, J. Organomet. Chem. 1987, 328, 35. Antipin M Y, Lobkovskii E B, Semenenko K N, Soloveichik G L, Struchkov Y T, J. Struct. Chem. (Engl. Transl.) 1979, 20, 810. Gard E, Haaland A, Novak D P, Seip R, J. Organomet. Chem. 1975, 88, 181. Antipin M Y, Boese R, Acta Crystallogr. B Struct. Sci. 1996, 52, 314. Lokshin B V, Greenwald I I, J. Mol. Struct. 1990, 222, 11. Gordetsov A S, Latyaeva V N, Zimina S V, Levakova E Y, Cherkasov V K, Moseeva E M, Skobeleva S E, Russ. Chem. Bull. 1996, 45, 1214. Tsumura R, Hagihara N, Bull. Chem. Soc. Jpn. 1964, 37, 1889. Ricci G, Panagia A, Porri, L, Polymer. 1996, 37, 363. Grigoryan E A, Dyachkovskii F S, Zhuk S Ya, Vyshinskaya L I, Kinet. Catal. (Engl. Transl.) 1978, 19, 1860. Köhler F H, Doll K H, Prössdorf W, J. Organomet. Chem. 1982, 224, 341. Castellani M P, Geib S J, Rheingold A L, Trogler W C, Organometallics 1987,6, 1703. O’Hare D, Murphy V J, Kaltsoyannis N, J. Chem. Soc. Dalton Trans. 1993, 383. Köhler F H, Doll K H, Prössdorf W, Müller J, Angew. Chem. Int. Ed. Engl. 1982, 21, 151. Atzkern H, Hiermeier J, Kanellakopulos B, Köhler F H, Müller G, Steigelmann O, J. Chem. Soc. Chem. Commun. 1991, 997.
406 [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86]
6
Group 5 and 6 Metallocenes
Perutz R N, Scaiano J C, J. Chem. Soc. Chem. Commun. 1984, 457. Chetwynd-Talbot J, Grebenik P, Perutz RN, Inorg. Chem. 1982, 21, 3647. Cox P A, Grebenik P, Perutz R N, Robinson R D, Grinter R, Stern D R, Inorg. Chem. 1983, 22, 3614. Hartwig J F, He X, Angew. Chem. Int. Ed. Engl. 1996, 35, 315 Gordon K R, Warren K D, J Organomet. Chem. 1976, 117, C27. Gordon K R, Warren K D, Inorg. Chem. 1978, 17, 987. Aleksanyan V T, Greenwald I I, J. Mol. Struct. 1982, 90, 35. Ketkov S Y, Domrachev G A, Inorg. Chim. Acta 1990, 178, 233. Giordan J C, Moore J H, Tossell J A, Weber J, J. Am. Chem. Soc. 1983,105, 3431. Köhler F H, Geike W, J. Magn. Reson. 1983, 53, 297. Hebendanz N, Köhler F H, Scherbaum F, Schlesinger B, Magn. Reson. Chem. 1989, 27, 798. Blümel J, Hofmann P, Köhler F H, Magn. Reson. Chem. 1993, 31, 2. Flower K R, Hitchcock P B, J. Organomet. Chem. 1996, 507, 275. Heinemann O, Jolly P W, Krüger C, Verhovnik G P J, Organometallics, 1996, 15, 5462. Cotton F A, Feng X, Kibala P A, Matusz M, J. Am. Chem. Soc. 1988, 110, 2807. Kalousova J, Benes L, Votinsky J, Coll. Czech. Chem. Commun. 1986, 51, 314. Wong K L T, Brintzinger H H, J. Am. Chem. Soc. 1975, 97, 5143. Karol F J, Karapinka G L, Wu C, Dow A. W, Johnson R N,Carrick W L, J. Polym. Sci. A1, 1972, 10, 2621. Thomas B J, Noh S K, Schulte G K, Sendlinger S C, Theopold K H, J. Am. Chem. Soc. 1991, 113, 893. Schnellbach M, Köhler F H, Blümel J, J. Organomet. Chem. 1996, 520, 227. Blümel J, Hebendanz N, Hudeczek P, Köhler F H, Strauss W J. Am. Chem. Soc. 1992, 114, 4223. Mugnier Y, Moise C, Laviron E, Nouv. J. Chim. 1982, 6, 197. Smart J C, Pinsky B L, J. Am. Chem. Soc. 1980, 102, 3663. Jonas K, Wiskamp V, Tsay Y-H, Krüger C, J. Am. Chem. Soc. 1983, 105, 5480. Jonas, K, Rüsseler W, Krüger C, Raabe E, Angew. Chem. Int. Ed. Engl. 1986, 25, 928. Jonas K, Rüsseler W, Angermund K, Krüger C, Angew. Chem. Int. Ed. Engl. 1986, 25, 927. van Raaij E U, Mönkeberg S, Kiesele H, Brintzinger H H, J. Organomet. Chem. 1988, 356, 307. Pasynskii A A, Eremenko I L, Abdullaev A S, Orazsakhatov B, Nefedov S E, Stomakhina E E, Ellert O G, Katser S B, Yanovskii A I, Struchkov Y T, Zh. Neorg. Khim. 1990, 35, 2257. Hübel W, Merényi R, J. Organomet. Chem. 1964, 2, 213. Bierwagen E P, Bercaw J E, Goddard III W A, J. Am. Chem. Soc. 1994, 116, 1481. Simpson K M, Rettig M F, Wing R M, Organometallics 1992,11, 4363. Foo D M J, Shapiro P, Organometallics, 1995, 14, 4957. Thomas J L, J. Am. Chem. Soc. 1973, 95, 1838. Mönkeberg S, van Raaij E, Kiesele H, Brintzinger H H, J. Organomet. Chem. 1989, 365, 285. Poli R, Mattamana . P, Falvello L R, Gazz. Chim. Ital. 1992, 122, 315. Jernakoff P, Fox J R, Cooper N J, J. Organomet. Chem. 1996, 512, 175. Ito T, Tokunaga T, Minato M, Nakamura T, Chem. Lett. 1991, 1893. Galante J M, Bruno J W, Hazin P N, Folting K, Huffman J C, Organometallics, 1988, 7, 1066. Martins A M, Calhorda M J, Romão C C, Völkl C, Kiprof P, Filippou A C, J. Organomet. Chem. 1992, 423, 367. Bandy J A, Berry A, Green M L H, Perutz R N, Prout K, Verpeaux J-N, J. Chem. Soc. Chem. Commun. 1984, 729. Calderazzo F, Fachinetti G, Floriani C, J. Am. Chem. Soc. 1974, 96, 3695. Kowaleski R M, Basolo F, Trogler W C, Gedridge R W, Newbound T D, Ernst R D, J. Am. Chem. Soc. 1987, 109, 4860. Kowaleski R M, Rheingold A L, Trogler W C, Basolo F, J. Am. Chem. Soc. 1986, 108, 2460. Nieman J, Teuben J H, J. Organomet. Chem. 1985, 287, 207. Thiyagarajan B, Michalczyk L, Bollinger J C, Bruno J W, Organometallics, 1996, 15, 2588.
References ]87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128]
407
Green M L H, Mtetwa S B, Sella A, Chernega A N, J. Chem. Soc. Daton Trans. 1994, 201. Schwemlein H, Zsolnai L, Huttner G, Brintzinger H H, J. Organomet. Chem. 1983, 256, 285. Green J C, Payne H. P, Teuben J H, Organometallics 1983, 2, 203. de Liefde Meijer H J, Jellinek F, Inorg. Chim. Acta 1970, 4, 651. Manzer L E, J. Organomet. Chem. 1976, 110, 291. Curtis C J, Smart J C, Robbins J L, Organometallics 1985,4, 1283. Siegert F W, de Liefde Meijer H J, J. Organomet. Chem. 1968, 15, 131. Ytsma D, Hartsuiker J G, Teuben J H, J. Organomet. Chem. 1974, 74, 239. Bouman H, Teuben J H, J. Organomet. Chem. 1976, 110, 327. Fachinetti G, Floriani C, J. Chem. Soc. Dalton Trans. 1974, 2433. Razuvaev G A, Latyaeva V N, Gladyshev E N, Lineva A. N, Krasil’nikova E. V. Dokl. Akad. Nauk. SSSR (Engl. Transl.) 1975, 223, 481. Otto E E H, Brintzinger H H, J. Organomet. Chem. 1979, 170, 209. Antiñolo A, Fajardo M, Otero A, Royo P, J. Organomet. Chem. 1982, 234, 309. Acedo L, Otero A, Royo P, J. Organomet. Chem. 1983, 258, 181. Fermin M C, Hneihen A S, Maas J J, Bruno J W, Organometallics 1993,12, 1845. Antiñolo A, Fajardo M, Otero A, Royo P, J. Organomet. Chem. 1983, 246, 269. Deutsch P P, Maguire J A, Jones W D, Eisenberg R, Inorg. Chem. 1990, 29, 686. Antiñolo A, García-Lledó S, Martínez de Ilarduya J, Otero A, J. Organomet. Chem. 1987, 335, 85. Nabaoui H, Mugnier Y, Fakhr A, Laviron E, Antiñolo A, Jalón F A, Fajardo M, Otero A, J. Organomet. Chem. 1989, 375, 67. Antiñolo A, Fajardo M, Jalón F A, López Mardomingo C, Otero A, Sanz-Bernabé C, J. Organomet. Chem. 1989, 369, 187. Baynham R F G, Chetwynd-Talbot J, Grebenik P, Perutz R N, Powell M H A, J. Organomet. Chem. 1985, 284, 229. Roullier L, Lucas D, Mugnier Y, Antiñolo A, Fajardo M, Otero A, J. Organomet. Chem. 1991, 412, 353. Lemenovskii D A, Tsikalova M V, Konde S A, Perevalova E G, Koord. Khim. 1983, 9, 1060. Fachinetti G, Del Nero S, Floriani C, J. Chem. Soc. Dalton Trans. 1976, 1046. McNally J P, Cooper N J, Organometallics, 1988, 7, 1704. Miller G A, Cooper N J, J. Am. Chem. Soc. 1985, 107, 709. Fiederling K, Grob I, Maalisch W, J. Organomet. Chem. 1983, 255, 299. Antiñolo A, Espinosa P, Fajardo M, Gómez-Sal P, López-Mardomingo C, Martín-Alonso A, Otero A, J. Chem. Soc. Dalton Trans. 1995, 1007. Serrano R , Royo P, J. Organomet. Chem. 1983, 247, 33. Urbanos F A, Mena M, Royo P, Antiñolo A, J. Organomet. Chem. 1984, 276, 185. Antiñolo A, Gómez-Sal P, Martínez de Ilrduya J, Otero A, Royo P, Martínez Carrera S, García Blanco S, J. Chem. Soc. Dalton Trans. 1987, 975. Antiñolo A, Martínez de Ilarduya J M, Otero A, Royo P, Manotti Lanfredi A M, Tiripicchio J, Chem. Soc. Dalton Trans. 1988, 2685. Martínez de Ilarduya J M, Otero A, Royo P, J. Organomet. Chem. 1988, 340, 187. Gómez M, Martínez de Ilarduya J M, Royo P, J. Organomet. Chem. 1989, 369, 197. Elkrami S, Mourad Y, Mugnier Y, Antiñolo A, del Hierro I, Garcia-Yuste S, Otero A, Fajardo M, Brunner H, Gehart G, Wachter J, Amaudrut J, J. Organomet. Chem. 1995, 498, 165. Bell R A, Cohen S A, Doherty N M, Threlkel R S, Bercaw J E, Organometallics 1986, 5, 972. Kubicki M M, Oudet P, Martin C, Barré C, J. Chem. Soc. Dalton Trans. 1995, 3699. Bonnet G, Oudet P, Moïse C, J. Organomet. Chem. 1995, 487, 105. Green M L H, Hughes A K, J. Organomet. Chem. 1996, 506, 221. Green M L H, Hughes A K, J. Chem. Soc. Dalton Trans. 1992, 527. Antiñolo A, Carrillo F, Chaudret B, Fajardo M, García-Yuste S, Lahoz F, Lanfranchi M, Lopez J A, Otero A, Pellinghelli M A, Organometallics 1995, 14, 1297. Hitchcock P B, Lappert M F, Milrre C R C, J. Chem. Soc. Dalton Trans. 1981, 180.
408
6
Group 5 and 6 Metallocenes
[129] Jiang Q, Pestana D C, Carroll P J, Berry D H, , Organometallics 1994, 13, 3679. [130] Bonnet G, Kubicki M M, Moïse C, Lazzaroni R, Salvadori P, Vitulli G, Organometallics 1992, 11, 964. [131] Bonnet G, Lavastre O, Leblanc J C, Moise C, New J. Chem. 1988, 12, 551. [132] Lavastre O, Bonnet G, Leblanc J C, Moïse C, Polyedron 1995, 2, 307. [133] Oudet P, Kubicki M M, Moïsse C, Organometallics 1994, 13, 4278. [134] Challet S, Kubicki M M, Leblanc J C, Moïse C, Nuber B, J. Organomet. Chem. 1994, 483, 47. [135] Challet S, Lavastre O, Moïse C, Leblanc J C, Nuber B, New J. Chem. 1994, 18, 1155. [136] Buhro W E, Zwick B. D, Georgiou S, Hutchinson J P, Gladysz J H, J. Am. Chem. Soc. 1988, 110, 2427. [137] Bonnet G, Lavastre O, Leblanc J C, Moïse C, Vitulli G, J. Organomet. Chem. 1988, 347, C21. [138] Oudet P, Moïse C, Kubicki M M, Inorg. Chim. Acta 1996, 247, 263. [139] Oudet P, Perrey D, Bonnet G, Moïse C, Kubicki M M, Inorg. Chim. Acta 1995, 237, 79. [140] Boni G, Sauvageot P, Moïse C, J. Organomet. Chem. 1995, 489, C32. [141] Boni G, Sauvageot P, Marpeaux E, Moïse C, Organometallics 1995, 14, 5652. [142] Challet S, Leblanc J C, Moïse C, New J. Chem. 1995, 19, 1139. [143] Sauvageot P, Blacque O, Kubicki M M, Juge S, Moïse C, Organometallics 1996, 15, 2399. [144] Barre C, Kubicki M M, Leblanc J C, Moïse C, Inorg. Chem. 1990, 29, 5244. [145] Moise C, Reynoud J F, Leblanc J C, Broussier R, J. Organomet. Chem. 1982, 240, C15. [146] Skripkin Y V, Pasynskii A A, Kalinnikov V T, Porai-KoshitsL K, Minacheva A, Antayshkina A S, Ostrikova V N, J. Organomet. Chem. 1982, 231, 205. [147] Nakajima T, Takaya M, Shimizu I, Wakatsuki Y, Organometallics, 1995, 14, 5598. [148] Pasynskii A A, Skripkin Y V, Kalinnikov V T, J. Organomet. Chem. 1978, 150, 51. [149] Pasynskii A A, Skripkin Y V, Eremenko I L, Kalinnikov V T, Aleksandrov G G, Struchkov Y T, J. Organomet. Chem. 1979, 165, 39. [150] Antiñolo A, Fajardo M, García-Yuste S, del Hierro I, Otero A, Elkrami S, Mourad Y, Mugnier Y, J. Chem. Soc. Dalton Trans. 1995, 3409. [151] Eckart E H O, Brintzinger H H, J. Organomet. Chem. 1978, 148, 29. [152] Bond A M, Bixler J W, Mocellin E, Datta S, James E J, Wreford, S. Inorg. Chem. 1980, 19, 1760. [153] Antiñolo A, Otero A, Fajardo M, García-Yebra G, Gil-Sanz R, López-Mardomingo C, Martín A, Gómez-Sal P, Organometallics 1994, 13, 4679. [154] Kukharenko S V, Soloveichik G L, Strelets V V, Metalloorg. Khim. 1990, 3, 88. [155] Fakhr A, Mugnier Y, Broussier R, Gautheron B, J. Organomet. Chem. 1984, 269, 53. [156] Fakhr A, Mugnier Y, Broussier R, Gautheron B,Laviron E, J. Organomet. Chem. 1986, 317, 201. [157] Threlkel R S, Bercaw J E, J. Am. Chem. Soc. 1981, 103, 2650. [158] Lappert M F, Raston C L, Skelton B W, White A H, J. Chem. Soc. Dalton Trans. 1984, 893. [159] Bailey S I, Engelhardt L M, Leung W P, Raston C L, Ritchie I M, White A H, J. Chem. Soc. Dalton. Trans. 1985, 1747. [160] Tueting D. R, Olmstead M M, Schore N E, Organometallics 1992, 11, 2235. [161] Lemenovskii D A, Nifant’ev I E, Urazowski I F, Perevalova E G, Timofeeva T V, Slovokhotov Yu L, Struchkov Yu T, J. Organomet. Chem. 1988, 342, 31. [162] Green M L H, Hughes A K, Michaelidou D M, Mountford P, J. Chem. Soc. Chem. Commun. 1993, 591 [163] Green M L H, Hughes A K, Mountford P, J. Chem. Soc. Dalton Trans. 1991, 1699. [164] El Krami S, Mourad Y, Lucas D, Mugnier Y, Antiñolo A, Fajardo M, Garcia-Yuste S, Otero A, J. Organomet. Chem. 1996, 525, 125. [165] Silavwe N D, Castellani M P, Tyler D R, Inorg. Synth. 1992, 29, 204. [166] Green M L H, McCleverty J A, Pratt L, Wilkinson G, J. Chem. Soc. 1961, 4854. [167] Persson C, Andersson C, Organometallics 1993, 12, 2370. [168] Mise T, Maeda M, Nakajima T, Kobayashi K, Shimizu I, Yamamoto Y, Wakatsuki Y, J. Organomet. Chem. 1994, 473, 155.
References [169] [170] [171] [172] [173] [174] [175] [176] [177] [178] [179] [180] [181] [182] [183] [184] [185] [186] [187] [188] [189] [190] [191] [192] [193] [194] [195] [196] [197] [198] [199] [200] [201] [202] [203] [204] [205] [206] [207] [208] [209] [210] [211]
409
Labella L, Chernega A, Green M L H, J. Organomet. Chem. 1995, 485, C18. Murray R C, Blum L, Liu A H, Schrock R R, Organometallics 1985, 4, 953. Parking G, Bercaw J E, Polyhedron 1988, 7, 2053. Rigny S, Bakhmutov I, Nuber B, Leblanc J-C, Moïse C, Inorg. Chem. 1996, 35, 3202. Rigny S, Leblanc J-C, Moïse C, Nuber B, J. Chem. Soc. Chem. Commun. 1995, 45. Rigny S, Leblanc J-C, Nuber B, Moïse C, J. Chem. Soc. Dalton Trans. 1997, 1187. Grebenik P D, Green M L H, Kelland M A, Leach J B, Mountford P, J. Chem. Soc. Chem. Commun. 1989, 1397. Cariou M, Kubicki M M, Kergoat R, Gomes de Lima L C, Scordia H, Guerchais J E, Inorg. Chim. Acta 1985, 104, 185. Forschner T C, Cooper N J, J. Am. Chem. Soc. 1989, 111, 7420. Forschner T C, Corella II J A , Cooper N J, Organometallics 1990, 9. 2478. Smith J A, Brintzinger H H, J. Organomet. Chem. 1981, 218, 159. Hartwig J F, He X, J. Am. Chem. Soc. 1994, 116, 3661. Hartwig J F, He X, Organometallics, 1996, 15, 5350. Hartwig J F, He X, Angew. Chem. Int. Ed. Engl. 1996, 35, 315. Ito T, Sugimoto S, Ohki T, Nakano T, Osakada K, J. Organomet. Chem. 1992, 428, 69. Ito T, Sugimoto S, Nakano T, J. Chem. Soc. Dalton Trans. 1987, 1857. Fu E, Granell J, Green M L H, Lowe V J, Marder S R, Saunders G C, Tuddenham M, J. Organomet. Chem. 1988, 355, 205. Calhorda M J, Carrondo M A A F. de C. T, Dias A. R, Domingos A M T S, Martinho Simoes J A, Teixeira C, Organometallics 1986, 5, 660. Koloski T S, Pestana D C, Carroll P J, Berry D H, Organometallics 1994, 13, 489. Figg L K, Carroll P J, Berry D H, Organometallics 1994, 13, 209. Figg L K, Carroll P J, Berry D H, Angew. Chem. Int. Ed. Engl. 1996, 35, 435. Bel’skii V K, Protskii A N, Bulychev B M, Soloveichik G L, J. Organomet. Chem. 1985, 280, 45. Mackey O N D, Morley C P, Polyhedron 1992, 11, 389. Okuda J, Herberich G E, Raabe E, Bernal I, J. Organomet. Chem. 1988, 353, 65. Petersen J L, Egan Jr J W, Inorg. Chem. 1981, 20, 2883. Belmore K A, Vanderpool R A, Tsai J-C, Khan M A, Nicholas K M, J. Am. Chem. Soc. 1988, 110, 2004. Herberich G E, Linn K, J. Organomet. Chem. 1991, 418, 409. Kubas G J, Ryan R R, Inorg. Chem. 1984, 23, 3181. Kuo L Y, Kuhn S, Ly D Inorg. Chem. 1995, 34, 5341. Pilato R S, Housmekerides C E, Jernakoff P, Rubin D, Geoffroy G L, Rheingold A L, Organometallics 1990, 9, 2333. Adam G J S, Green M L H, J. Organomet. Chem. 1981, 208, 299. Bruce A E, Bruce M R M, Sclafani A, Tyler D R, Organometallics 1984, 3, 1610. Köpf H, Klapötke J, Organomet. Chem. 1986, 310, 303. Jones W D, Chin R M, Crane T W, Baruch D M, Organometallics, 1994, 13, 4448. Jernakoff P, Cooper N J, J. Am. Chem. Soc. 1989, 111, 7424. Berry D H, Chey J C, Zipin H S, Carroll P J, Polyhedron, 1991, 10, 1189. Hong P, Damrauer N H, Carroll P J, Berry D H, Organometallics, 1993, 12, 3698. Kotz J C, Vining W, Coco W, Rosen R, Dias A R, Garcia M H, Organometallics 1983, 2, 68. Calhorda M J, Carrondo M A A F de C T, Dias A R, Domingos A M. T, Duarte M T L S, García M H, Romão C C, J. Organomet. Chem. 1987, 320, 63. Calhorda A M, Dias A R, Duarte M T, Martins A M, Matias P M, Romão C C, J. Organomet. Chem. 1992, 440, 119. Ascenso J R, de Azevedo C G, Gonçalves I S, Herdwick E, Moreno D S, Pessanha M, Romão C C, J Organometallics 1995, 14, 3109. Tsai J-C, Khan M, Nicholas K M, Organometallics 1989, 8, 2967. Tsai J-C, Wheeler R A, Khan M, Nicholas K M, Organometallics 1991, 10, 1344.
410 [212] [213] [214] [215] [216] [217] [218] [219] [220] [221] [222] [223] [224] [225] [226] [227] [228] [229] [230] [231] [232] [233] [234] [235] [236] [237] [238] [239] [240] [241] [242] [243] [244] [245] [246] [247] [248] [249] [250] [251]
6
Group 5 and 6 Metallocenes
De Azevedo C G, Dias A R, Martins A M, Romão C C, J. Organomet. Chem. 1989, 368, 57. Crabtree R H, Dias A R, Green M L H, Knowles P J, J. Chem. Soc. A 1971, 1350. McNally J P, Glueck D, Cooper N J, J. Am. Chem. Soc. 1988, 110, 4838. McNally J P, Cooper N J, J. Am. Chem. Soc. 1989, 111, 4500. 183A Rauchfuss T B, Ruffing C J, Organometallics 1982, 1, 400. Jernakoff P, Cooper N J, J. Am. Chem. Soc. 1984, 106, 3026. Calhorda A M, Dias A R, Martins A M, Romão C C, Polyhedron 1989, 8, 1802. Carmichael A J, McCamley A J. Chem. Soc. Dalton Trans. 1995, 3125. Carmichael A J, McCamley A J. Chem. Soc. Dalton Trans. 1997, 93. Ascenso J R, de Azevedo C G, Gonçalves I S, Herdwick E, Moreno D S, Romão C C, Zühkle J, Organometallics 1994, 13, 429. Gonçalves I S, Romão C C, J. Organomet. Chem. 1995, 486, 155. Dorer B, Prosenc M-H, Reif U, Brintzinger H H, Organometallics 1994, 13, 3868. Antiñolo A, Martinez-Ripoll M, Mugnier Y, Otero A, Prashar S, Rodriguez A M, Organometallics 1996, 15, 3241. Bristow G S, Lappert M F, Martin T R, Atwood J L, Hunter W F, J. Chem. Soc. Dalton Trans. 1984, 399. Lemenovskii D A, Urazovskii I. F, Nifant’ev I E, Perevalova E G, J. Organomet. Chem. 1985, 292, 217. Bott S G, Hoffmann D M, Rangarajan S P, J. Chem. Soc. Dalton Trans. 1996, 1979. Gyon F, Amadrut J, Mercier M-F, Shimizu K, J. Organomet. Chem. 1994, 465, 187. Lucas D, Mugnier Y, Antiñolo A, Otero A, Fajardo M, J. Organomet. Chem. 1992, 435, C3. Sánchez C, Vivien D, Sala Pala J, Viard B, Guerchais J E, J. Chem. Soc. Dalton Trans. 1981, 64. Carrondo M A A F de C T, Morais J, Romão C C, Romão M J, Veiros L F, Polyhedron 1993, 12, 765. Calhorda M J, Carrondo M A A F de C T, Bram A, Olsen P N, Dias A. R, Freitas A M, García M H, Piedade M F M, J Organomet. Chem. 1992, 426, 195. Klapötke T M, Schulz A, Cameron T S, Bakshi P K J. Organomet. Chem. 1993, 463, 115. Compton R G, Eklund J C, Page S D, Rebbitt T O J. Chem. Soc. Dalton Trans. 1995, 389. Dias A. R, Garcia M H, Martins A M, Pinheiro C I, Romão C.C, Veiros L F, J. Organomet. Chem. 1987, 327, C59. Jernakoff P, Fox J R, Hayes J C, Lee S, Foxman B M, Cooper N J, Organometallics, 1995, 14, 4493. Migot J L, Sala-Pala J, Guerchais J E, J. Organomet. Chem. 1983, 243, 427. Antiñolo A, Fajardo M, Otero A, Royo P, J. Organomet. Chem. 1984, 265, 35. Arnold J, Don Tilley T, Rheingold A., L, Geib S J, Organometallics 1987, 6, 473. Gowik P, Klapötke T, J. Organomet. Chem. 1989, 368, 35. Antiñolo A, Fajardo M, Otero A, Puerta M C, Mugnier Y, Polyhedron 1989,8, 1848. Antiñolo A, Fajardo M, Otero A, Mugnier Y, Nabaoui H, Mourad H P, J. Organomet. Chem. 1991,414, 155. Hunter J A, Lindsell W E, McCullough K J, Parr R A, Scholes M L, J. Chem. Soc. Dalton Trans. 1990, 2145. Fakhr A, Mugnier Y, Broussier R, Gautheron B, J. Organomet. Chem. 1985, 279, C15. Parkin G, Bunel E, Burger B. J, Trimmer M S, van Asselt A, Bercaw J E, J. Mol. Catal. 1987, 41, 21. Thiele K H, Kubak W, Sieler J, Borrmann H, Simon A, Z. Anorg. Allg. Chem. 1990, 587, 80. Gowik P, Klapötke T, J. Organomet. Chem. 1989, 375, C20. Ol’dekop, Yu A, Knizhnikov V A. Zh. Obshch. Khim. 1981, 51, 1723. Thiyagaragan B, Kerr M E, Bollinger J C, Young Jr V G, Bruno J W, Organometallics 1997, 16, 1331. Brunner H, Gehart G, Meier W, Wachter J, Organometallics, 1994, 13, 134. Fajardo M, Gómez-Sal M P, Royo P, Martínez-Carrera S, García Blanco S, J. Organomet. Chem. 1986, 312, C44.
References [252] [253] [254] [255] [256] [257] [258] [259] [260] [261] [262] [263] [264] [265] [266] [267] [268] [269] [270] [271] [272] [273] [274] [275] [276] [277] [278] [279] [280] [281] [282] [283] [284] [285] [286] [287] [288] [289] [290] [291] [292] [293] [294] [295]
411
Gowik P, Klapötke T, White P, Chem. Ber. 1989, 122, 1649. Schulz A, Klapötke T, J. Organomet. Chem. 1994, 480, 195. Heinekey D M, J. Am. Chem. Soc. 1991, 113, 6074. Zilm K W et al, J. Am. Chem. Soc. 1990, 112, 920. Antiñolo A, Carrillo F, Fernández-Baeza J, Otero A, Fajardo M, Chaudret B, Inorg. Chem. 1992, 31, 5156. Tebbe F N, Parshall G W, J. Am. Chem. Soc. 1971, 93, 3793. Doherty N M, Bercaw J E, J. Am. Chem. Soc. 1985, 107, 2670. Berry D H, Koloski T S, Carroll P J, Organometallics 1990, 9, 2952. Nikonov G I, Kuzmina L, Lemenovskii D A, Kotov V V, J. Am. Chem. Soc. 1995, 117, 10133. Antiñolo A, Carrillo F, Fajardo M, Otero A, Lanfranchi M, Pellinghelli M A, Organometallics 1995, 14, 1518. Green M L H, Hughes A K, Mountford P, J. Chem. Soc. Dalton Trans. 1991, 1407. Nikonov G I, Kuzmina L, Mountford P, Lemenovskii D A, Organometallics 1995, 14, 3588. Nikonov G I, Lorberth J, Harms K, Lemenovskii D A, Inorg. Chem. 1995, 34, 2461. Nikonov G I, Lemenovskii D A, Lorbeth J, Organometallics 1994, 13, 3127. Lantero D R, Motry D H, Ward D L, Smith M RIII, J. Am. Chem. Soc. 1994, 116, 10811. Antiñolo A, Carrillo-Hermosilla F, Chaudret B, Fajardo M, Fernández-Baeza J, Lanfranchi M, Limbach H-H, Maurer M, Otero A, Pellinghelli M A, Inorg. Chem. 1996, 35, 7873. Antiñolo A, Carrillo, Chaudret B, Fajardo M, Fernández-Baeza J, Lanfranchi M, Limbach HH, Maurer M, Otero A, Pellinghelli M A, Inorg. Chem. 1994, 33, 5163. Sala-Pala J, Migot J L, Guerchais J E, Le Gall L, Grosjean F, J. Organomet. Chem. 1983, 248, 299. Amaudrut J, Kadmiri A, Sala-Pala J, Guerchais J E, J. Organomet. Chem. 1984, 266, 53. Whinnery Jr. L L, Henling L M, Bercaw J E, J. Am. Chem. Soc. 1991, 113, 7575. Nelson J E, Parkin G, Bercaw J E, Organometallics 1992, 11, 2181. Royo P, Gòmez-Carrera J.A, An. Quím. 1984, 80B, 428. Leboeuf J-F, Lavastre O, Leblanc J-C, Moise C, J. Organomet. Chem. 1991, 418, 359. Sabo-Etienne S, Chaudret B, el Makarim H A, Barthelat J-C, Daudey J-P, Moïse C, Leblanc JC, J. Am. Chem. Soc. 1994, 116, 9335. Jalón F A, Otero A, Manzano B R, Villaseñor, Chaudret B, J. Am. Chem. Soc. 1995, 117, 10123. Parking G, Bercaw J E, J. Chem. Soc. Chem. Commun. 1989, 255. Camanyes S, Maseras F, Moreno M, Lledos J M, Lluch J M, Bertrán J, J. Am. Chem. Soc. 1996, 118, 4617. Wiberg N, Häring N W, Schubert U, Z. Naturforsch. B 1980, 35, 599. Gambarotta S, Chiesi-Villa A, Guastini C, J. Organomet. Chem. 1984, 270, C49. Osborne J H, Rheingold A L, Trogler W C, J. Am. Chem. Soc. 1985, 107, 7945. Osborne J H, Rheingold A L, Inorg. Chem. 1985, 24, 3098. Broussier R, Olivier J D, Gautheron B, J. Organomet. Chem. 1983, 251, 307. Thiyagaragan B, Kerr M E, Bruno J W, Inorg. Chem. 1995, 34, 3444. Fu P-f, Khan M A, Nicholas K M, J. Organomet. Chem. 1996, 506, 49. Antonelli D M, Schaefer W P, Parkin G, Bercaw J E, J. Organomet. Chem. 1993, 462, 213. van Asselt A, Burger B J, Gibson V C, Bercaw J E, J. Am. Chem. Soc. 1986, 108, 5347 Parkin G, van Asselt A, Leahy D J, Whinnery L, Hua N G, Quan W, Henling L M, Schaefer W P, Santarsiero B D, Bercaw J E, Inorg. Chem. 1992, 31, 82. Proulx G, Bergman RG, J. Am. Chem. Soc. 1996, 118, 1981. Proulx G, Bergman R G, J. Am. Chem. Soc., 1993, 115, 9802. Proulx G, Bergman R G, J. Am. Chem. Soc., 1994, 116, 7953. Prolux G, Bergman R G, Organometallics, 1996, 15, 133. Reingold A L, Strong J B, Acta Crystallogr. Sect. C, 1991, 47, 1963 Brunner H, Gehart G, Meier W, Wachter J, Nuber B, J. Organomet. Chem. 1993, 454, 117. Brunner H, Kubicki M M, Leblanc J C, Moise C, Volpato F, Wachter J, J. Chem. Soc. Chem. Commun. 1993, 851.
412
6
Group 5 and 6 Metallocenes
[296] Bach H-J, Brunner H, Wachter J, Kubicki M M, Leblanc J-C, Moise C, Volpato F, Nuber B, Ziegler M L, Organometallics, 1992, 11, 1403-1407. [297] Brunner H, Wachter J, Gehart G, Leblanc J-C, Moise C, Organometallics, 1996, 15, 1327. [298] Shin J H, Parkin G, Organometallics, 1995, 14, 1104. [299] Shin J H, Parkin G, Organometallics, 1994, 13, 2147. [300] Prolux G, Bergman R G, Organometallics, 1996, 15, 684. [301] Jorgensen K A, Inorg. Chem. 1993, 32, 1521. [302] Lemenovskii D A, Putala M, Nikonov G I, Kazennova N B, Yufit D S , Struchov Y T, J. Organomet. Chem. 1993, 454, 123. [303] Lemenovskii D A, Putala M, Nikonov G I, Zinin N B, Kazennova N B, Struchov Y T, J. Organomet. Chem. 1993, 452, 87. [304] Lemenovskii D A, Nikonov G I, Brusova G P, Kuzmina L G, Stankovie E, Putala M, J. Organomet. Chem. 1995, 496, 227. [305] Schrock R R, ‘Reactions of coordinated ligands’ ed Braterman, Plenum, New York, 1986. [306] Berry D H, Koloski T S, Carroll P J, Organometallics, 1990, 9, 2952. [307] Gibson V C, Parkin G, Bercaw J E, Organometallics, 1991, 10, 220. [308] Antiñolo A, del Hierro I, Fajardo M, García-Yuste S, Otero A, Blaque O, Kubicki M M, Amaudrut J, Organometallics 1996, 15, 1966. [309] Butts M D, Bergman R G, Organometallics, 1994, 13, 2668. [310] Butts M D, Bergman R G, Organometallics, 1994, 13, 1899. [311] Duncalf D J, Harrison R J, McCamley A, Royan B W, J. Chem. Soc. Chem. Commun. 1995, 2421. [312] Silavwe N D, Chang M Y, Tyler D R, Inorg Chem. 1985, 24, 4219. [313] Yoon M, Tyler D R, J. Chem. Soc. Chem. Commun. 1997, 639. [314] Parkin G, Bercaw J E, J. Am. Chem. Soc. 1989, 111, 391. [315] Silavwe N D, Bruce M R M, Philbin C E, Tyler D R, Inorg. Chem. 1988, 27, 4669. [316] a) Jernakoff P, Geoffroy G L, Rheingold A L, Geib S J, J. Chem. Soc. Chem. Commun. 1987, 1610. b) Pilato R S, Geoffroy G L, J. Chem. Soc. Chem. Commun. 1987, 1287. [317] Pilato R S, Eriksen K A, Stiefel E I, Rheingold A L, Inorg. Chem. 1993, 32, 3799. [318] Green J C, Green M L H, James J T, Konidarkis P C, Maunder G H, Mountford P, J. Chem. Soc. Dalton Trans. 1992, 1361. [319] Green M L H, Konidarkis P C, Michaelidou D M, Mountford P, J. Chem. Soc. Dalton Trans. 1995, 155. [320] Hitchcock P B, Lappert M F, Leung W-P, J. Chem. Soc. Chem. Commun. 1987, 1282. [321] Niecke E, Hein J, Nieger M, Organometallics, 1989, 8, 2290. [322] Wolczanski P T, Threlkel R S, Santarsiero B D, Acta Crystallogr., Sect. C, 1983, 39, 1330. [323] Miller G A, Cooper N J, J. Organomet. Chem. 1997, 528, 151. [324] Antiñolo A, Fajardo M, Mardomingo C L, Otero A, Mourad Y, Mugnier Y, Sanz-Aparicio J, Fonseca I, Florencio F, J. Organomet. Chem. 1990, 9, 2919. [325] Antiñolo A, Carillo F, Fajardo M, Garcia-Juste S, Otero A, J. Organomet. Chem. 1994, 482, 93. [326] Klazinga H, Teuben, J H, J. Organomet. Chem. 1980, 194, 309. [327] Gibson V C, Bercaw J E, Bruton W J, Sanner R D, Organometallics, 1986, 5, 976. [328] Bercaw J E, Burger B J, Green M L H, Santarsiero B D, Sella A, Trimmer M S, Wong L-L, J. Chem. Soc. Chem. Commun. 1989, 734. [329] Burger B J, Santarsiero B D, Trimmer M S, Bercaw J E, J. Am. Chem. Soc. 1988, 110, 3134. [330] Klazinga A H, Teuben J H, J. Organomet. Chem. 1980, 192, 75. [331] Antiñolo A, Carrillo F, Garcia-Yuste S, Otero A, Organometallics, 1994, 13, 2761. [332] Antiñolo A, Carrillo-Hermosilla F, Chaudret B, Fajardo M, García-Yuste S, Lanfranchi M, Lopez J A, Otero A, Pellinghelli M A, Prashar S, Villaseñor E, Organometallics 1996, 15, 5507. [333] Antiñolo A, Fajardo M, Galakhov M, Gil-Sanz R, López-Mardomingo C, Otero A, Lucas D, Chollet H, Mugnier Y, J. Organomet. Chem. 1994, 481, 27.
References
413
[334] Yasuda H, Yamamoto H, Arai T, Nakamura A, Chen J, Kai Y, Kasai N, Organometallics 1991, 10, 4058. [335] Herberich G E, Mayer H, Organometallics, 1990, 9, 2655. [336] Fakhr A, Mugnier Y, Rouiller L, Broussier R, Gautheron B, Laviron E, New J. Chem. 1988, 12, 213. [337] Fu P F, Khan M A, Nicholas K M, Organometallics, 1992, 11, 2607. [338] Bristow G S, Hitchcock P B, Lappert M F, J. Chem. Soc. Chem. Commun. 1982, 1145. [339] Fu P F, Khan M A, Nicholas K M, J. Am. Chem. Soc. 1992, 114, 6579. [340] Amaudrut J, Sala-Pala J, Guerchais J E, Mercier R, Douglade J, J. Organomet. Chem. 1982, 235, 301. [341] Thiyagarajan B, Michalczyk L, Bollinger J C, Huffmann J C, Bruno J W, Organometallics, 1996, 15, 1989. [342] Antiñolo A, Otero A, Fajardo M, López, Mardomingo C, Lucas D, Mugnier Y, Lanfranchi M, Pellingheilli MA, J. Organomet. Chem., 1992, 435, 55. [343] Bruno J W, Fermin M C, Halfon S E, Schulte G K, J. Am. Chem. Soc. 1989, 111, 8738. [344] Fermin M C, Thiyagarajan B, Bruno J W, J. Am. Chem. Soc. 1993, 115, 974. [345] Savaranamuthu A, Bruce A E, Bruce M R M, Fermin M C, Hneihen A S, Bruno J W, Organometallics, 1992, 11, 2190 [346] Kerr M, Fermin M C, Bruno J W, J. Chem. Soc. Chem. Commun. 1996, 1221. [347] Antiñolo A, Fajardo M, Gil-Sanz R, Mardomingo C M, Martín-Villa P, Otero A, Kubicki M M, Mugnier Y, El Krami S, Mourad Y, Organometallics, 1993, 12, 381. [348] Antiñolo A, Fajardo M, Mardomingo C M, Martín-Villa P, Otero A, Kubicki M M, Mourad Y, Mugnier Y, Organometallics, 1991, 10, 3435. [349] Nelson J E, Bercaw J E, Marsh R E, Henling L M. Acta Crystallogr., Sect. C, 1992, 48, 1023. [350] Kerr M E, Bruno J W, J. Am. Chem. Soc. 1997, 119, 3183. [351] Biali S E, Gozin M, Rappoport Z, J. Phys. Org. Chem. 1989, 2, 271. [352] Bunker M J, De Cian A, Green M L H, Moreau J J E, Siganporia N, J. Chem. Soc. Dalton Trans. 1980, 2155. [353] Yasuda H, Arai T, Okamoto T, Nakamura A, J. Organomet. Chem. 1989, 361, 161. [354] McGilligan B S, Wright T C, Wilkinson G, Motevalli M, Hursthouse M B, J. Chem. Soc. Dalton Trans. 1988, 1737. [355] Wright T C, Wilkinson G, Motevalli M, Hursthouse M B, J. Chem. Soc. Dalton Trans. 1986, 2017. [356] Chetcuti P A, Knobler C B, Hawthorne M F, Organometallics 1988, 7, 650. [357] Nakamura A, Otsuka S, J. Am. Chem. Soc. 1972, 94, 1886. [358] Herberich G E, Okuda J, Chem. Ber. 1984, 117, 3112. [359] Okuda J, Herberich G E, Organometallics 1987, 6, 2331. [360] Herberich G E, Okuda J, Angew. Chem. Int. Ed. Engl. 1985, 24, 402. [361] Gambarotta S, Floriani C, Chiesi-Villa A, Guastini C, J. Am. Chem. Soc. 1985, 107, 2985. [362] Conan F, Sala-Pala J, Guerchais J E, Li J, Hoffmann R, Mealli C, Mercier R, Toupet L, Organometallics, 1989, 8, 1929. [363] Koloski T S, Carroll P J, Berry D H, J. Am. Chem. Soc. 1990, 112, 6405. [364] Berry D H, Chey J C, Zipin H S, Carroll P J, J. Am. Chem. Soc. 1990, 112, 452. [365] Dias A R, Queirós M A, J. Organomet. Chem. 1990, 390, 193. [366] Queirós M A, Simão J E J, Dias A R, J. Organomet. Chem. 1987, 329, 85. [367] Hayes J C, Jernakoff P, Miller G A, Cooper N J, Pure Appl. Chem. 1984, 56, 25. [368] Hayes J C, Pearson G D N, Cooper N J, J. Am. Chem. Soc. 1981, 103, 4648. [369] Hayes J C, Cooper N J, J. Am. Chem. Soc. 1982, 104, 5570 [370] Jernakoff P, Cooper N J, Organometallics, 1986, 5, 747. [371] Bkouche-Waksman I, Bois C, Sala-Pala J, Guerchais J. E, J. Organomet. Chem. 1980, 195, 307. [372] van Asselt A, Trimmer M S, Henling L M, Bercaw J E, J. Am. Chem. Soc. 1988, 110, 8254. [373] Middleton A R, Wilkinson G, J. Chem. Soc. Dalton Trans. 1980, 1888.
414
6
Group 5 and 6 Metallocenes
[374] Brunner H, Klement U, Wachter J, Tsunoda M, Leblanc J C, Moise C, Inorg. Chem., 1990, 29, 584. [375] Leblanc J-C, Moïse C, Volpato F, Brunner H, Gehart G, Wachter J, Nuber B, J. Organomet. Chem. 1995, 485, 237. [376] Brunner H, Gehart G, Leblanc J-C, Moïse C, Nuber B, Stubenhofer B, Volpato F, Wachter J, J. Organomet. Chem. 1996, 517, 47. [377] Antiñolo A, Fajardo M, Gil-Sanz R, López-Mardomingo C, Otero A, Organometallics, 1994, 13, 1200. [378] Morán M, Santos-Garcia J J, Masaguer J R, Fernández V, J. Organomet. Chem. 1985, 295, 327. [379] Gambarotta S, Floriani C, Chiesi-Villa, Guastini C, Organometallics. 1986, 5, 2425. [380] Sielsisch T, Behrens U, J. Organomet. Chem. 1984, 272, C40 [381] Pasquali M, Leoni P, Floriani C, Chiesi-Villa, Guastini C, Inor. Chem. 1983, 22, 841. [382] Drews R, Wormsbächer D, Behrens U, J. Organomet. Chem. 1984, 272, C40. [383] Benecke J, Drews R, Behrens U, Edelmann F, Keller K, Roesky H W, J. Organomet. Chem. 1987, 320, C31. [384] Gambarotta S, Fiallo M L, Floriani C, Chiesi-Villa, Guastini C, Inorg. Chem. 1984, 23, 3532. [385] Antiñolo A, Fajardo M, de Jesús E, Mugnier Y, J. Organomet. Chem. 1994, 470, 127.
7
Half-Sandwich Complexes as Metallocene Analogs Jun Okuda and Thomas Eberle
7.1 Introduction The introduction of group 3 and 4 metallocenes has revolutionized the area of Ziegler–Natta catalysis [1]. For the first time in the history of this industrially important process, critical polymerization parameters such as activity, molecular weight, polydispersity and microstructure of the resulting polyolefins can be controlled by structurally well-defined and on the molecular level modifiable metal complexes. Moreover, the use of metallocenes as homogeneous polymerization catalysts has dramatically improved the understanding of mechanistic features such as the nature of the active sites and the influence of ligand structure on the regioand stereoselectivity [2]. One of the many advantages generally associated with the bis(cyclopentadienyl) ligand systems, however, occasionally turns into a disadvantage: the characteristic and highly consistent electronic and steric situation within the bent metallocene unit [3] has long been recognized to cause substantial steric blocking of the metal-centered reaction site. Enhancement of reactivity is observed when the two ring ligands are ‘tied back’ by a dimethylsilanediyl link as in many Brintzinger-type ansa-metallocene complexes. But even in such cases the ‘wedge’ of the metallocene moiety still turns out to be too congested to allow, for instance, the efficient polymerization of α-olefins [4]. In order to alleviate this steric constraint of the metallocenes, one could utilize, in place of two cyclopentadienyl ligands, one cyclopentadienyl ligand that contains an additional coordinating site X or L tethered to the periphery of the fivemembered ring via a bridge Z, where X is a one-electron, L is a two-electron ligand (using the neutral ligand formalism) [5], and Z is a covalent bridge of appropriate length (Fig. 7-1). Such bidentate ligands may form chelate complexes in which the cyclopentadienyl group and the additional donor group X or L are both interacting with one metal center [6].
416
7
Half-Sandwich Complexes as Metallocene Analogs
Figure 7-1. General formula for a metal complex containing a bifunctional cyclopentadienyl ligand. X and X′ denote one-electron, L and L′ two-electron ligands, following the neutral ligand formalism.
The replacement of one cyclopentadienyl moiety in a bridged bis(cyclopentadienyl) ligand by an amido ligand NR′, connected via a bridge Z, results in ligand systems that form complexes differing from both ansa-metallocenes and the simple halfsandwich complexes without the link Z (Fig. 7-2). The amido group is a threeelectron ligand of the LX-type (including π-donation from the sp2-hybridized nitrogen atom), in contrast to the a five-electron L2X-type cyclopentadienyl ligand.
Figure 7-2. Relationship between ansa-metallocenes, linked amido–cyclopentadienyl, and half-sandwich amido complexes.
There are many other possibilities with which a cyclopentadienyl ligand may be replaced. An obviously close analog to the amido function would be a bridging alkoxo group (Z)O [7]. The imido group (Z)N=, a ligand isolobal to the cyclopentadienyl ligand, can also be expected to function as the second donor site, but so far only one example for such a bidentate ligand has been reported [8]. Numerous half-sandwich compounds containing neutral donor functions L, especially amine (NR′2) and ether (OR′) groups at the tether (Z) have been described in the literature [9]. Recently, their potential use as volatile precursors for chemical vapor deposition and as catalysts for olefin polymerization has been noticed. The cyclopentadienyl ligand is known to act mostly as an inert supporting ligand for a reactive transition metal center by not actively participating in a given substrate transformation. However, under certain conditions, for example during catalytic cycles, the cyclopentadienyl ligand may be involved in irreversible
7.2
Complexes with Linked Amido-Cyclopentadienyl Ligands
417
chemical reactions or may even dissociate from the metal. If a second donor (preferentially a multiply bonding ligand) tethered to the cyclopentadienyl ligand is bound as firmly as a cyclopentadienyl ligand, it will add to the stability of the entire ligand framework and prevent exchange or decomposition reactions. By independently modifying the nature of each of the fragments C5R4, X, L, and Z in the bifunctional cyclopentadienyl ligand systems, a great potential for imparting novel properties to the resulting chelate complexes and for controlling the metal reactivity seems to emerge. The present review summarizes the coordination chemistry of linked amido–cyclopentadienyl ligands and the application of mainly group 3 and 4 metal complexes containing this type of ligand. Such complexes constitute a novel class of industrially extremely relevant homogeneous catalysts for olefin polymerization. They possess unique properties which differ significantly from the metallocene catalysts, and major attention is currently being paid to them [10]. The complexes containing a linked alkoxo–cyclopentadienyl ligand are similar to such complexes and are included in this review as well.
7.2 Complexes with Linked Amido–Cyclopentadienyl Ligands In the late eighties Bercaw and Shapiro introduced the first complexes of bridged amido–cyclopentadienyl ligands [11] in the context of developing structurally wellcharacterized single-component olefin polymerization catalysts. As electronically more unsaturated and sterically more accessible analogs of ansa-scandocene complexes, precursor for complexes of the type Sc(η5:η1-C5Me4SiMe2NtBu)X (X = H, alkyl) were synthesized and in fact shown to exhibit much higher reactivity towards α-olefins [11, 12]. In order to explore sterically demanding derivatives of this dianionic ligand [13], iron and titanium complexes were synthesized shortly thereafter [14]. Based on these bridged amido–cyclopentadienyl ligands, a flurry of development occurred independently in the research laboratories of Dow Chemical and Exxon Chemical and culminated in their patent applications filed within two weeks of each other in 1990 [10]. A deluge of patents has appeared in the meantime and most of the data on this subject resides in the patent literature.
7.2.1 Synthesis Following the traditional route to metallocene complexes, amido-bridged halfsandwich complexes are most commonly synthesized by first assembling the ligand (C5R4H)Z(NHR′) and then coordinating it at the metal center according to a suitable protocol. As the bridging function Z, mostly SiMe2 is used, but carbon-
418
7
Half-Sandwich Complexes as Metallocene Analogs
based bridges (CH2)n (n = 2, 3) or mixed group SiMe2CH2 have also appeared in the literature. For the introduction of the SiMe2 link, metallated cyclopentadiene or any substituted cyclopentadiene C5R4H2, usually as its lithium salt Li(C5R4H), is reacted with dichlorodimethylsilane to give (C5R4H)SiMe2Cl (R = H, Me). The chlorodimethylsilyl substituted cyclopentadiene is further transformed into the ligand precursor (C5R4H)SiMe2NHR′ using a variety of lithium amides Li(NHR′), or alternatively, using excess amine NH2R′ to remove the hydrogen chloride (Scheme 7-1). Thus the prototypical ligand (C5Me4H)SiMe2NHtBu, containing the dimethylsilanediyl bridge, first described by Bercaw and Shapiro, is obtained as a moisture sensitive, distillable oil in overall yield of 76% based on Li(C5Me4H) [12].
Scheme 7-1 There seems to be no limit as to the choice of the cyclopentadienyl moiety C5R4: it may also consist of annulated ring systems such as indenyl [15] or donor-substituted systems (see Sec. 7.3). In the case of the fluorenyl ligand (C13H9)SiMe2NHtBu, the synthesis of (C13H9)SiMe2Cl is hampered by the substantial formation of (C13H9)2SiMe2 due to the high nucleophilicity of the fluorenyl anion [16]. A series of di- and trimethylene-linked ligands may be prepared by adding [Br(CH2)nN+H2R′]Br– (n = 2, 3; R′ = Me, iPr, tBu) to an excess of Na(C5H5) or Na(C9H7). These functionalized cyclopentadienes C5H5(CH2)nNHR′ exist as several double-bond isomers (Eq. 7-1) [17, 18].
(7-1)
7.2
Complexes with Linked Amido-Cyclopentadienyl Ligands
419
The double metalation is straightforwardly achieved with n-butyllithium and, as is commonly the case, used in situ without isolation. Occasionally isolation is possible: Li2(C5Me4SiMe2NtBu) can be obtained as a tan powder. Based on what is known about both cyclopentadienyl and amido lithium complexes in solution, complicated structures can be presumed for such dianions. Temperature-dependent NMR spectra of Li2{(C13H8)SiMe2NtBu} indicate an unsymmetrical structure devoid of a mirror plane [16]. When ClCH2SiMe2NHtBu is reacted with two equivalents of fluorenyl lithium the structurally characterized monolithium derivative Li(THF){(C13H8)CH2SiMe2NHtBu} is isolated [19]. Interestingly, the reaction of potassium hydride with (C5Me4H)SiMe2NHtBu gives the mono(potassium) derivative K(THF)(C5Me4SiMe2NHtBu) that contains an intact amino function according to a single-crystal structure analysis [20]. The use of the magnesium derivative {MgCl(THF)}2(C5Me4SiMe2NtBu) as a less reducing ligand transfer agent has been reported [10a, 33]. For the complexation of the linked amido–cyclopentadienyl ligand several different synthetic procedures have been developed. The metathetical reaction of the doubly metallated ligand precursor (C5R4ZNR′)2– with appropriate metal halides appears to be the most common. The reaction of Li2(C5Me4SiMe2NtBu) with ScCl3(THF)3 gives non-stoichiometric Sc(η5:η1-C5Me4SiMe2NtBu)Cl•LiCl(THF)x [12] (which probably contains an ate complex [Sc(η5:η1-C5Me4SiMe2NtBu)Cl2]–) from which the THF can be removed by heating to 100°C under vacuum. After Soxhlet extraction, the chloro complex Sc(η5:η1-C5Me4SiMe2NtBu)Cl is obtained as a white powder in 51% yield (Eq. 7-2) [12].
(7-2)
The first example of a titanium complex containing a linked amido– cyclopentadienyl ligand, racemic Ti{η5:η1-(C5H3tBu)SiMe2NtBu}Cl2, was obtained from the reaction of Li2{(C5H3tBu)SiMe2NtBu} with TiCl4(THF)2 as brown crystals in 35% yield and was shown by NMR spectroscopy to be a configurationally stable molecule [14]. The use of the tris(tetrahydrofuran) complex of titanium(III) chloride followed by oxidation of the titanium(III) intermediate instead of the more reduction-sensitive TiCl4(THF)2 leads to more reproducible and higher yields [21]. Apart from the conventional oxidants HCl, CCl4 and AgCl, the best reagent for the chlorination of the titanium(III) intermediate proved to be lead dichloride, introduced by Teuben et al. [22]. Following this method, dichloro(titanium) complexes of the general type Ti(η5:η1-C5R4SiMe2NR′)Cl2,
420
7
Half-Sandwich Complexes as Metallocene Analogs
with various cyclopentadienyl and indenyl moieties, as well as amido substituents R′, including optically active ones, can be synthesized (Scheme 7-2) [15, 23–25].
Scheme 7-2 However, there are some disadvantages with this synthetic methodology: Occasionally only low yields are encountered during the synthesis of titanium complexes, e. g. when ligands containing an unsubstituted cyclopentadienyl ring is used (yield of Ti(η5:η1-C5H4SiMe2NtBu)Cl2 < 5%) [26]. Not unexpectedly, reduction of the metal center is generally a synthetic obstacle, whenever strongly reducing (i.e. more ionic) cyclopentadienyl anions are employed [21]. An elegant solution for this problem is offered by the reaction of C5H5(CH2)nNHR′ with TiCl4 in the presence of triethylamine to give complexes of the type Ti{η5:η1-C5H4(CH2)nNR′}Cl2 in good yields [18]. In certain cases the preferred formation of bis(ligand) complexes of the type Ti(η5:η1-C5R4ZNR′)2 may account for the low yield (see Sec. 7.4). For example, irrespective of the molar ratio, the reaction between Li2(C5H4SiMe2NtBu) and ZrCl4 exclusively leads to the formation of Zr(η5:η1C5H4SiMe2NtBu)2 [27]. According to another report an equimolar mixture of Zr(η5:η1-C5H4SiMe2NtBu)Cl2 and [Li(THF)n][Zr(η5:η1-C5H4SiMe2NtBu)Cl3] is obtained, whereas the reaction of Li2(C5Me4SiMe2NtBu) with ZrCl4 or ZrCl4(THF)2 gives the desired mono(ligand) complex Zr(η5:η1-C5Me4SiMe2NtBu)Cl2 in 85% yields (Scheme 7-3) [26].
Scheme 7-3
7.2
Complexes with Linked Amido-Cyclopentadienyl Ligands
421
As mentioned above for the scandium complex Sc(η5:η1-C5Me4SiMe2NtBu)Cl•LiCl(THF)x, the expected higher Lewis acidity of the complexes of the linked amido–cyclopentadienyl ligand often results in the formation of solvent adducts [12]. Such solvent adducts may interfere with subsequent alkylation reactions or with activation procedures during olefin polymerization. In the case of fluorenylbased systems, the resulting zirconium complexes are found to be only easily isolable as solvent adducts. Mono(tetrahydrofuran) or mono(diethyl ether) adducts Zr(η5:η1-C13H8SiMe2NtBu)Cl2(L) (L = THF or Et2O) are obtained by reacting Li2(C13H8SiMe2NtBu)(THF)n with ZrCl4(THF)2 or ZrCl4(Et2O)2 [16]. Detailed NMR spectroscopic investigation revealed a labile coordination of one THF molecule in Zr(η5:η1-C13H8SiMe2NtBu)Cl2(THF), which is irreversibly lost upon heating to give insoluble oligomers (Scheme 7-4). By changing the size of the bridging group Z from SiMe2 to CH2SiMe2, the resulting fluorenyl complexes M(η5:η1C13H8CH2SiMe2NR′)Cl2 (M = Ti, Zr, Hf; R′ = tBu), synthesized from Li2(C13H8CH2SiMe2NtBu)(THF)n and TiCl4(THF)2, ZrCl4 and HfCl4, are isolated without a solvent molecule [19]. The reaction of Li2(C5Me4SiMe2NtBu) with chromium trichloride in THF affords the dark green chromium(III) complex also as its THF solvate Cr(η5:η1-C5Me4SiMe2NtBu)Cl(THF) [28]. In the absence of any Lewis base, some zirconium complexes tend to dimerize such as {Zr(η5:η1C5H4SiMe2NtBu)Cl}2(µ-Cl)2) [26].
Scheme 7-4 In order to avoid the problems associated with the metathesis method, the reaction of homoleptic metal amides M(NR″2)n with the functionalized cyclopentadienes with amine elimination has been applied. This method introduced many years ago for the synthesis of metallocene amido complexes of the type M(C5R5)x(NR″)y,
422
7
Half-Sandwich Complexes as Metallocene Analogs
was recently expanded to the stereoselective synthesis of Brintzinger-type ansametallocenes [29]. It was first applied for the synthesis of complexes containing a linked amido–cyclopentadienyl ligand with a carbon bridge C5H5(CH2)nNR′ (n = 2, 3; R′ = Me, iPr, tBu) [17, 18, 30]. The complexation of M(NMe2)4 (M = Ti, Zr, Hf) works well for these ligands and produces distillable M{η5:η1C5H4(CH2)nNR′}(NMe2)2 or M{η5:η1-C9H6(CH2)nNR′}(NMe2)2 (M = Zr, Hf; n = 2: R′ = Me, tBu; n = 3: R′ = Me) (Scheme 7-5). The reaction is more effective when less sterically hindered ligands, preferably containing two CH2 units, are used. Whereas C5H5(CH2)2NHtBu gives the desired product with Ti(NMe2)4 or Zr(NMe2)4, the double aminolysis starting from C5H5(CH2)3NHtBu and M(NMe2)4 (M = Ti, Zr) does not occur even under forcing conditions and results in the formation of M{η5-C5H4(CH2)3NHtBu}(NMe2)3 (M = Ti, Zr). Decreased reactivity is observed using the titanium tetraamides Ti(NMe2)4, which do not react with indenyl substituted ligands.
Scheme 7-5 Following analogous procedures, previously described dimethylsilanediyl-bridged amido–cyclopentadienyl ligands C5H5SiMe2NHR′ (R′ = tBu, Ph) react with Ti(NR″)4, Zr(NR″)4, (R″ = Me, Et) and Hf(NMe2)4 to give the corresponding complexes of the general type M(η5:η1-C5H4SiMe2NR′)(NR″)2 [26, 31]. At least in the case of zirconium, the analogous reaction appears to work for the less acidic (C5Me4H)SiMe2NHtBu′ to give Zr(η5:η1-C5Me4SiMe2tBu)(NMe2)2. The driving force for the successful formation of these chelate complexes lies in the generation of the very volatile amines (bp for NMe2H and NEt2H: 7 and 55°C, respectively). Although often liquids, the products can be obtained in nearly quantitative yields and the reaction is accelerated with the more acidic ligands and with sterically less demanding cyclic systems [31]. Zirconium dichloro complexes Zr[η5:η1C5H4(CH2)2NR′]Cl2 (R′ = tBu) are accessible by using Zr(NMe2)2Cl2(THF)2 and the ligand C5H5ZNHR′. The coordinated dimethylamine can be removed by subli-
7.2
Complexes with Linked Amido-Cyclopentadienyl Ligands
423
mation (Eq. 7-3) [32]. Reaction of optically active indenyl ligand (–)-(S)C9H6SiMe2NCHMePh with Ti(NMe2)4 gives a 4:3 diastereomeric mixture of Ti(η5:η1-(–)-(S)-C9H6SiMe2NCHMePh)(NMe2)2 [33]. (7-3)
Analogously, niobium(V) and tantalum(V) complexes, as well as the molybdenum(IV) complex of the type M(η5:η1-C5H4SiMe2NPh)(NMe2)n (n = 3: Nb, Ta, n = 2: Mo), were obtained in moderate yields [34]. Similarly, using Y{N(SiMe3)2}3 the yttrium complex Y(η5:η1-C5Me4SiMe2NtBu){N(SiMe3)2} was synthesized in 40% yield by amine elimination [35]. An alternative approach to amido-bridged cyclopentadienyl complexes consists of introducing the amido linkage within the preformed half-sandwich complex [36]. The functionalized cyclopentadiene C5H4(SiMe2Cl)SiMe3, which may easily be synthesized from LiC5H5(SiMe3) and SiMe2Cl2, is first reacted with TiCl4 to form Ti(η5-C5H4SiMe2Cl)Cl3, with selective elimination of chlorotrimethylsilane. The trichloro complex can be reacted with lithium amides LiNHR′ (R′ = tBu [36a], iPr [25], CH2Ph [25], (±)-, (–)-(S)-, (+)-(R)-CHMePh [24]) in the presence of a base such as triethylamine to scavenge HCl. The resulting titanium complexes Ti(η5:η1-C5H4SiMe2NR′)Cl2 are available in good yields. The mechanism for this reaction implies that the amide anion first attacks titanium and then intramolecularly bridges the silicon which is presumably less electrophilic than titanium (Scheme 7-6) [36].
Scheme 7-6
424
7
Half-Sandwich Complexes as Metallocene Analogs
However, this method is limited so far to the unsubstituted cyclopentadienyl system C5H4 and cannot be extended to substituted or annulated ring systems such as indenyl [15]. Furthermore, it does not apply to metal centers other than titanium. The reaction of Zr(η5-C5H4SiMe2Cl)Cl3 with LiNHtBu in the presence of NEt3 leads to a mixture of two compounds: the desired Zr(η5:η1-C5H4SiMe2NtBu)Cl2 and the triethylamine adduct Zr(η5-C5H4SiMe2NHtBu)Cl3(NEt3). When Li(NHCHMePh) is reacted with Zr(η5-C5H4SiMe2Cl)Cl3 only Zr(η5C5H4SiMe2NHCHMePh)Cl3(NEt3) is isolated. The observed formation of complexes without the link indicates preferential attack of the amide NHR′– at the silicon [36].
7.2.2 Hydrido and Alkyl Complexes The synthesis of alkyl complexes from the chloro or amido derivatives are frequently studied in view of their importance as precursors for α-olefin polymerization catalysts. As suitable alkylating reagents, organolithium or Grignard reagents without β-hydrogens, including methyl, benzyl, trimethylsilylmethyl, neophyl and neopentyl, are employed. Generally, the bulkier alkyl groups provide more thermally and photochemically stable complexes, whereas some dimethyl titanium derivatives cannot be stored for prolonged periods of time at ambient temperatures, reminiscent of the highly sensitive half-sandwich analog (η5-C5H5)TiMe3 [37]. Remarkably, the only isolable monomeric scandium alkyl complex is Sc(η5:η1C5Me4SiMe2NtBu){CH(SiMe3)2}, isolated from the reaction of the chloro complex Sc(η5:η1-C5Me4SiMe2NtBu)Cl with LiCH(SiMe3)2 [12]. Even the reaction with LiCH2SiMe3 does not afford any tractable products. Sc(η5:η1-C5Me4SiMe2NtBu){CH(SiMe3)2} can be hydrogenated in the presence of PMe3 to form the dimeric hydride {Sc(η5:η1-C5Me4SiMe2NtBu)(PMe3)}2(µ-H)2. It reacts with two equivalents of ethylene to give the unusual ethylene-bridged dimer {Sc(η5:η1C5Me4SiMe2NtBu)(PMe3)}2(µ-η2:η2-C2H4) and one equivalent of ethane, whereas the analogous reaction with propylene affords dimeric n-propyl complex {Sc(η5:η1-C5Me4SiMe2NtBu)}2(µ-nPr)2. The latter gives monomeric Sc(η5:η1C5Me4SiMe2NtBu)(PMe3)(nPr) upon addition of PMe3 and undergoes rapid associative exchange via 16-electron Sc(η5:η1-C5Me4SiMe2NtBu)(PMe3)2(nPr) (Scheme 7-7).
7.2
Complexes with Linked Amido-Cyclopentadienyl Ligands
425
Scheme 7-7 A fairly extensive range of dialkyl complexes of group 4 metals, M(η5:η1C5Me4SiMe2NtBu)R″2, have been reported. They are obtained by the reaction of the dichloro complexes with two equivalents of organolithium or Grignard reagents (Eq. 7-4) [19, 25, 36, 38, 39]. In the case of zirconium complexes containing carbon links, the amine adduct Zr{η5:η1-C5H4(CH2)3NMe}Cl2(NMe2H) can be directly converted into the dialkyl [17]. Compared to their metallocene analogs, the more air- and moisture-sensitive dialkyl complexes clearly show a higher Lewis acidity (Eq. 7-5).
(7-4)
(7-5)
426
7
Half-Sandwich Complexes as Metallocene Analogs
It appears that the retention of solvent molecules depends on the steric bulk of the alkyl groups. While zirconium dimethyl derivatives Zr(η5:η1-C13H8-SiMe2NtBu)Me2(THF) [16] or Zr(η5:η1-C5H4(CH2)3NMe)Me2(Et2O)0.5 tend to retain solvents [17], the complexes with larger alkyls can be isolated without any coordinated solvent. Likewise, the chromium complex Cr(η5:η1-C5Me4SiMe2NtBu)R″ retains a THF molecule for R″ = Me and Ph, whereas for R″ = CH2SiMe3 crystal structure analysis shows a pseudotrigonal configuration (Eq. 7-6) [28].
(7-6)
In benzyl complexes of early transition metals the electronic unsaturation is relieved by a distorted coordination of the benzyl ligand. For example, the titanium compound Ti(η5:η1-C5Me4SiMe2NCH2Ph)(CH2Ph)2 exhibits an α-agostic bonding interaction of the CH2-hydrogen atoms of one of the benzyl groups with the titanium (Fig. 7-3) [25]. In contrast to the other benzyl group the phenyl ring of one benzyl group is nearly coplanar to the cyclopentadienyl ring and the amido group’s phenyl system. These different conformations of the benzyl ligands are related to the various metalen–carbon distances (Ti–C1: 2.131(5) Å, Ti–C8: 2.157(5) Å) and in different angles at the benzylic carbon of the two benzyl groups. Compared to Ti–C8–C9 (110.6(3)°), the angle at the other carbon C1 of the other benzyl group is 131.3(3)°. The titanium–hydrogen distances for the former benzyl group are on average slightly shorter than those of the second benzyl group. The values are similar to those found in Ti(η5-C5Me5)(CH2Ph)3 [40]. However, the unsymmetrical distortion cannot be detected by NMR data at room temperature, which reveal a Cs symmetry.
Figure 7-3. Molecular structure of Ti(η5:η1-C5Me4SiMe2NCH2Ph)(CH2Ph)2 [25].
7.2
Complexes with Linked Amido-Cyclopentadienyl Ligands
427
In contrast to the titanium dibenzyl complex, Ti(η5:η1-C5Me4SiMe2NCH2Ph)(CH2Ph)2, the NMR spectroscopic analysis of the zirconium dibenzyl Zr{η5:η1C5H4(CH2)3NMe}(CH2Ph)2 [17] indicates some η2-benzyl interaction [41], although the molecule possesses overall Cs symmetry. Since at –80°C no change is observed, either both benzyl groups are distorted, or only one of them is distorted, the η2-benzyl and η1-benzyl ligand undergoing rapid exchange. The reaction of Zr{η5:η1-C5H4(CH2)3NMe}Cl2(NMe2H) with one equivalent of C6H5CH2MgCl gives the dimeric mono(benzyl) complex {Zr(η5:η1-C5H4(CH2)3NMe)(CH2Ph)(µCl)}2. Instead of distorting the benzyl ligands, the electron deficiency is relieved by forming bridging chloro ligands. The crystal structure exhibits two different Zr– Cl distances due to the trans influence of the benzyl ligand [17]. The reaction of Zr{η5:η1-C5H4(CH2)3NMe}Cl2(NMe2H) with excess of LiBH4 gives the colorless bis(tetrahydroborate), Zr{η5:η1-C5H4(CH2)3NMe}(η3-BH4)2, which contains rapidly exchanging terminal and bridging hydrido ligands. Alkylidene complexes of early transition metals such as Ti(η5-C5R5)2(=CR′2) have played a pivotal role in the development of active initiators for olefin metathesis and ring-opening metathesis polymerization [42]. Thermolysis of Ti{η5:η1C5H4(CH2)2NtBu}(CH2CMe2R″)2 (R″ = Me, Ph) in the presence of PMe3 affords the neopentylidene and neophylidene complexes Ti{η5:η1-C5H4(CH2)2NtBu}(=CHCMe2R″)(PMe3) in excellent yields, which exhibit characteristically low-field 13C signals at δ 251 and 246 ppm, respectively (Eq. 7-7) [18, 43].
(7-7)
In contrast to the bis(cyclopentadienyl) metal carbene complexes, the Ti=C double bond in Ti{η5:η1-C5H4(CH2)2NtBu}(=CHCMe2R″)(PMe3) (R″ = Me, Ph) are assumed to be rotationally fixed because two orientations of the alkylidene ligand are observed. While 1H and 13C NMR spectroscopic data indicate only one rotamer, NOESY experiments reveal an interaction between the α-proton and the tert-butyl group on the amido substituent, rendering the orientation of the large alkyl group towards the cyclopentadienyl ring most likely. Both complexes were reported to be inactive in ring-opening polymerization of strained cyclic olefins. On the other hand, the thermal decomposition of Ti(η5:η1-C5H4SiMe2NCH2Ph)(CH2SiMe3)2 presumably gives an alkylidene complex that polymerizes norbornene with ringopening [44]. Thermolysis of early transition metal dialkyl complexes that contain at least one phenyl group are known to give benzyne complexes [45]. The elimination of benzene to form benzyne intermediates at the zirconocene fragment allows the
428
7
Half-Sandwich Complexes as Metallocene Analogs
oxidative coupling with unsaturated substrates R″2C=X (X = CR″2, NR″, O, S, P) and R″C≡X (X = CR″, N) [45]. Electronically more unsaturated linked amido– cyclopentadienyl complexes behave similarly. The thermolysis of Ti{η5:η1C5H4(CH2)2NtBu}Ph2 in the presence of PMe3 gives a phosphine-stabilized benzyne complex Ti{η5:η1-C5H4(CH2)2NtBu}(η2-C6H4)(PMe3) that readily undergoes insertion reactions with alkynes, alkenes, nitriles or ketones (Scheme 7-8) [18, 46]. In a related β−hydride elimination reaction, the diethyl complex decomposes to give the ethylene complex Ti{η5:η1-C5H4(CH2)2NtBu}(η2-C2H4)(PMe3) [18, 43].
Scheme 7-8 When dichloro complexes Ti(η5:η1-C5Me4SiMe2NR′)Cl2 (R′ = Ph, tBu) are treated with two equivalents of n-butyllithium in the presence of 1,3-dienes such as piperylene and 2,4-hexadiene, diene complexes of the type Ti(η5:η1-C5Me4SiMe2NR′)(1,3-diene) are formed [47]. The diene coordination mode prone-π at formally titanium(II) or supine-σ,π at formally titanium(IV) (titanacyclopentene) is highly sensitive to the identity of R′. While for R′ = tBu the former bonding situation is found, as authenticated by the crystal structure of Ti(η5:η1C5Me4SiMe2NtBu)(η4-MeCH=CHCH=CHMe), for R′ = Ph an inseparable mixture of both noninterconvertible isomers results (Scheme 7-9).
7.2
Complexes with Linked Amido-Cyclopentadienyl Ligands
429
Scheme 7-9
7.2.3 Reactivity Since the dichloro complexes M(η5:η1-C5R4ZNR′)Cl2 (M = Ti, Zr, Hf) are suitable precursors for alkylations, the conversion of the diamido complexes, M(η5:η1C5R4ZNR′)(NMe2)2, obtained by amine elimination, to the corresponding dichloro derivatives is quite important. Complete conversion, in good yields, can be achieved by treatment of the titanium complexes Ti(η5:η1-C5R4ZNR′)(NMe2)2 with an excess of chlorotrimethylsilane or phosphorus pentachloride [26]. The reaction of the zirconium analog, Zr(η5:η1-C5R4SiMe2NtBu)(NMe2)2, with two equivalents of chlorotrimethylsilane gives dimeric {Zr(η5:η1-C5H4SiMe2NtBu)Cl(µ-Cl)}2 [26]. The reaction of two equivalents of protic reagents such as HCl or (NEt3H)Cl with Ti(η5:η1-C5H4SiMe2NtBu)(NMe2)2 results in inseparable mixtures of Ti(η5:η1C5H4SiMe2NtBu)Cl2 and Ti(η5-C5H4SiMe2Cl)Cl2(NMe2)(NMe2H). In contrast, the reaction with the corresponding zirconium complexes provides only one product, Zr(η5:η1-C5H4SiMe2NtBu)Cl2(NMe2H), with the NMe2H ligand coordinated trans to the appended amido group [26]. Treatment of two equivalents of (NMe2H2)Cl or (NMe2H2)I with M{η5:η1-C5H4(CH2)3NMe}(NMe2)2 (M = Zr, Hf) gives M{η5:η1C5H4(CH2)3NMe}X2(NMe2H) (X = Cl, I) in good yields [17]. These results demonstrate the affinity of four-coordinate zirconium complexes for complexing additional L-type ligands (Eq. 7-8). Noteworthy is the reactivity of the bridging Si–N bond, mainly in titanium complexes. The driving force is the stronger Si–Cl bond versus the Si–N bond [48], the electrophilicity of the silicon atom being enhanced by the ring strain of the chelate structure. Extreme moisture-sensitivity is documented in the electron-
430
7
Half-Sandwich Complexes as Metallocene Analogs
rich late metal complex Fe{(η5:η1-C5H3tBu)SiMe2NtBu}(CO)2 that gives the ferrocene derivative Fe{(η5-C5H3tBu)SiMe2NHtBu}2 upon hydrolysis (Eq. 7-9) [14].
(7-8)
(7-9)
Reaction of carbon dioxide with complexes containing the C5R4SiMe2NR′ ligand such as Ti(η5:η1-C5H4SiMe2NR′)Cl2 (R′ = tBu, CHMePh) results in the cleavage of the Si–N bridge with concomitant formation of isocyanate OCNR′ and a dimeric siloxide complex with a silicon–oxygen link {Ti(η5:µ:η1-C5H4SiMe2O)Cl2}2. This unusual reaction implies the insertion of carbon dioxide into the titanium–nitrogen bond to form first a carbamato bridge which subsequently extrudes the isocyanate molecule OCNR′ (Eq. 7-10) [36b]. The reaction of two equivalents of CO2 with Zr(η5:η1-C5Me4SiMe2NtBu)Me2 affords {Zr(η5:η1-C5Me4SiMe2NtBu)(η2-O2CMe)(µ:η2-O2CMe)}2 [38]. Using a 20-fold excess of CO2 a sparingly soluble product is isolated and shown to be the dimeric bridged siloxide {Zr(η5:µ:η1-C5Me4SiMe2O)(η2-O2CMe)(µ:η2-O2CMe)}2. The conversion of the coordinated silylamido group into the bridged siloxide occurs again with concomitant elimination of OCNtBu (Eq. 7-11) [36].
(7-10)
7.2
Complexes with Linked Amido-Cyclopentadienyl Ligands
431 (7-11)
Irradiation of niobium and tantalum amides M(η5:η1-C5H4SiMe2NtBu)(NMe2)3 causes Si–N bond cleavage followed by a peculiar rearrangement to give tertbutylimido complexes containing an η5-C5H4SiMe2NMe2 ligand [34]. Whereas Ti(η5:η1-C5H4SiMe2NCH2C6H3F2-2,5)Cl2 is formed from Ti(η5:η1C5H4SiMe2Cl)Cl3 and Li(NHCH2C6H3F2-2,5) [25], the analogous reaction of the 2,6-difluorobenzyl derivative Li(NHCH2C6H3F2-2,6) gives Ti{η5:η1-C5H4SiMe2NCH2[C6H3F-6-(NHCH2C6H3F2-2,6)-2]}Cl2 (Eq 7-12) [25]. The amido ligand is modified as a result of nucleophilic aromatic substitution at one ortho-fluorine atom by another molecule of 2,6-difluorobenzylamide. This substitution may be facilitated by a titanium–fluorine interaction. The lithiation of 2,5- or 2,6difluorobenzylamime with n-butyllithium proceeds without complications.
(7-12)
7.2.4 Structure In the context of explaining the specific polymerization properties of group 4 metal complexes containing the linked amido–cyclopentadienyl ligand, single crystal structure analyses of numerous dichloro complexes of the type M(η5:η1C5R4ZNR′)Cl2 have been performed. Some representative examples are compiled in Table 7-1. Fig. 7-4 shows the molecular structure of Ti(η5:η1C5H4SiMe2NiPr)Cl2 [25] as a representative example.
432
7
Half-Sandwich Complexes as Metallocene Analogs
Figure 7-4. Molecular structure of Ti(η5:η1-C5H4SiMe2NiPr)Cl2 These complexes basically adopt a three-legged piano-stool or pseudotetrahedral geometry with the bifunctional amido–cyclopentadienyl ligand C5R4ZNR′ in addition to two terminal ligands X. The cyclopentadienyl ligand C5R4 is bonded in the usual η5 fashion and the metal–ligand bond lengths are in the expected range. As a result of a certain constraint of the chelating ligand, the metal center is often unsymmetrically bound to the planar cyclopentadienyl ring. In the case of the dimethylsilanediyl or ethanediyl bridge (Z = SiMe2 or CH2CH2), the bridging atom connected to the cyclopentadienyl periphery is displaced from the plane of the cyclic π system [49]. In the case of the longer bridge, Z = (CH2)3, this carbon atom is displaced away from the metal center and is located above the cyclopentadienyl ring [18, 25, 49]. The distance between the metal and the amido nitrogen is fairly sensitive to the nature of the metal fragment. For dichloro titanium(IV) complexes Ti(η5:η1C5R4ZNR′)Cl2, the bond lengths are around 1.9 Å. The metal–nitrogen distances are shorter than those observed for single bonds. In titanium derivatives containing titanium–nitrogen single bonds, the distance varies between 1.96 and 1.97 Å [50]. In bridged amido half-sandwich complexes of titanium, Ti–N distances shorter than 1.91 Å are sometimes observed. On the other hand these bond lengths are slightly larger than those found in nonbridged monocyclopentadienyl amido complexes. Thus, the Ti–N distance in Ti(η5-C5R5)(NiPr2)Cl2 are 1.865(2) Å (R = H) [51] and 1.865(5) Å (R = Me) [51] compared to 1.901(3) Å (R = H) [26] and 1.908(6) Å (R
7.2
Complexes with Linked Amido-Cyclopentadienyl Ligands
433
= Me) [52] in Ti(η5:η1-C5R4SiMe2NtBu)Cl2. An optimal overlap of the amido nitrogen pπ orbital with the titanium dπ orbital seems to be disturbed by the chelation. In the formally 18-electron diamido complex Ti(η5:η1-C5H4SiMe2NtBu)(NMe2)2, a value of 1.972(4) Å is observed [26], reflecting a less pronounced double bond character based on a dπ–pπ interaction. Also, in the titanium(II) diene complex Ti(η5:η1C5Me4SiMe2NtBu)(η4-MeCH=CHCH=CHMe), a significantly longer metal–nitrogen bond of 2.007(4) Å is observed [47]. The sum of the bond angles around the appended amido nitrogen is practically always 360°. This is clearly due to a sp2 hybridization caused by a dπ–pπ interaction between the amido nitrogen and the metal center. The Cp–M–N angles (Cp = centroid) in complexes M(η5:η1-C5R4ZNR′)X2 are by 25–35° smaller than typical Cp–M–Cp angles in the corresponding 16-electron metallocene complexes M(η5-C5R5)2X2, where angles vary between 125 and 135° [3]. Strain within the ligand system and an openness of the coordination sphere can be deduced from this finding. While there is a clear trend of increasing Cp–M–N angle with increasing length of the bridge Z, one should not overly emphasize a correlation of this geometric parameter with reactivity [32, 52]. The orbital interaction between the metal and a cyclopentadienyl ligand may not be directly comparable to that of an amido ligand. Chelation also results in a slight reduction of the metal–Cp distance as shown by the comparison of Ti(η5:η1-C5H4SiMe2NtBu)Cl2 (Ti–Cp: 2.019 Å) [26] and Ti(η5-C5H5)(NHtBu)Cl2 (Ti–Cp: 2.032 Å) [53]. Table 7-1. Some structural characteristics of half-sandwich complexes containing bridged amido–cyclopentadienyl ligands (Cp denotes the centroid of the cyclopentadienyl, N denotes the appended amido nitrogen) Compound
Ref.
Ti(η5:η1-C5H4SiMe2NtBu)Cl2 [26] Ti(η5:η1-C5Me4SiMe2NtBu)Cl2 [52] Ti{η5:η1-C5Me4(CH2)2NtBu}Cl2 [52] [25] Ti(η5:η1-C5H4SiMe2NiPr)Cl2 Ti{η5:η1-C5H4(CH2)2NiPr}Cl2 [18, 49] Ti{η5:η1-C5H4(CH2)3NiPr}Cl2 [18, 49] Ti(η5:η1-C5H4SiMe2NtBu)(NMe2)2 [26] Zr(η5:η1-C5Me4SiMe2NtBu)Cl2 [52] [26] Zr(η5:η1-C5Me4SiMe2NtBu)(NMe2)2 Y(η5:η1-C5Me4SiMe2NtBu){N(SiMe3)2} [35]
Cp–M–N (°)
M–Cp (Å)
107.0 107.6 107.9 105.5 104.4 112.6 105.5 102.0 100.2 97.7
2.019 2.030 – 2.017 2.008 2.027 2.083 2.163 2.233 2.300
M–N (Å) 1.901(3) 1.908(6) 1.909(5) 1.878(2) 1.864(2) 1.867(2) 1.972(4) 2.056(6) 2.108(4) 2.184(7)
The most notable structural feature of the benzylamido derivative Ti(η5:η1C5H4SiMe2NCH2C6H3F2-2,5)Cl2 in the solid state is the conformation of the benzylic group which is turned away from the metal center [25]. Single X-ray structural analysis of (–)-(S)-Ti(η5:η1-C5R4SiMe2NCHMePh)Cl2 (Fig. 7-5), the first optically active titanium complexes containing an amido–cyclopentadienyl ligand
434
7
Half-Sandwich Complexes as Metallocene Analogs
[24], reveals a structure in which the phenyl group is arranged coplanarly to the C5H4 ring, turned away from the metal center. NOE measurements suggest restricted rotation about the nitrogen–methine-carbon bond based on a larger NOE between the methyl group of the stereogenic center and one of the two diastereotopic methyl groups of the silanediyl link.
Figure 7-5. Molecular structure of (–)-(S)-Ti(η5:η1-C5R4SiMe2NCHMePh)Cl2 [24]. A similar characteristic feature for complexes containing a secondary amido substituent R′, is the tendency of the methine proton to orientate itself towards the metal center. This phenomenon is consistently found in iso-propylamido complexes such as Ti(η5:η1-C5H4SiMe2NiPr)Cl2 [25], Ti{η5:η1-C5H4(CH2)2NiPr}Cl2 [18, 49] and Ti{η5:η1-C5H4(CH2)3NiPr}Cl2 [18, 49]. Table 7-2. Structural characteristics of some complexes containing bridged amido– cyclopentadienyl ligands, where the amido substituent is a secondary alkyl group Compound
Ref.
M–HMethine M–N–CMethine (°) δ(HMethine) (Å)
(–)-(S)-Ti(η5:η1-C5H4SiMe2NCHMePh)Cl2 [24] Ti(η5:η1-C5H4SiMe2NiPr)Cl2 [25] [18, 49] Ti{η5:η1-C5H4(CH2)2NiPr}Cl2 Ti{η5:η1-C5H4(CH2)3NiPr}Cl2 [18, 49]
2.84 2.79 2.67 2.38
(ppm) 120.5(3) 117.4(1) 114.2(2) 104.5(1)
6.54 5.69 5.92 6.57
7.3
Complexes with Polydentate Amido-Cyclopentadienyl Ligands
435
Values for the M–N–CMethine angle and the M–HMethine distance are given in Table 7-2. A β-agostic interaction is not observed by 1H NMR spectroscopy in complexes with Z = SiMe2. However, it is detected in complexes containing less constrained, longer links (Z = (CH2)3). Here the Ti–N–CMethine angle and the Ti–HMethine distance for the β-agostic interaction are similar to those found in the unbridged cyclopentadienyl amido complex Ti(η5-C5H5)(NiPr2)Cl2 [51], where one of the isopropyl groups is directed towards the metal center due to a β-agostic interaction of the methine proton, resulting in a Ti–HMethine distance of 2.25 Å and a Ti–N– CMethine angle of 101.4(2)°. Finally, the upfield shifts for the Cipso atom of the cyclopentadienyl group in the 13C NMR spectra of complexes of the type Ti(η5:η1-C H SiMe NR′)Cl , compared 5 4 2 2 to the values found for the nonlinked precursor Ti(η5-C5H4SiMe2Cl)Cl3, are characteristic for the presence of a chelating amido group (Table 7-3). Table 7-3. 13C NMR chemical shifts for Cipso atom in amido-bridged half-sandwich complexes and Ti(η5-C5H4SiMe2Cl)Cl3 Complex
Ref.
δ(Cipso) (ppm)
Ti(η5-C5H4SiMe2Cl)Cl3 Ti(η5:η1-C5H4SiMe2NtBu)Cl2 Ti η5:η1-C5H4SiMe2NiPr)Cl2 Ti(η5:η1-C5H4SiMe2NCH2Ph)Cl2 Ti(η5:η1-C5H4SiMe2NCH2C6H3F2-2,5)Cl2
[36] [36] [25] [25] [25]
135.1 110.0 108.1 110.1 109.7
7.3 Complexes with Polydentate Amido–Cyclopentadienyl Ligands In order to attenuate the Lewis acidity of early transition metal centers, a new ligand system with a side chain incorporating an additional weak neutral donor site within the chelating amido-cyclopentadienyl ligand framework has been introduced [23]. Donor groups such as OMe or NMe2 attached to the amido functionality offer new possibilities in tailoring the coordination sphere around a reactive transition metal center. The synthesis of group 4 metal complexes of such tridentate ligands can be achieved by following synthetic methodologies analogous to those for simpler amido substituents. Using the metathetical pathway, hexane-soluble tetramethylcyclopentadienyl titanium, zirconium and hafnium complexes M(η5:η1-C5Me4SiMe2NCH2X)Cl2 (M = Ti, Zr, Hf; X = CH=CH2, CH2OMe, CH2NMe2) [23, 55, 56] have been prepared (Scheme 7-10). The unsubstituted cyclopentadienyl-bridged titanium derivatives Ti(η5:η1-C5H4SiMe2NCH2X)Cl2 (X = CH2OMe, CH2NMe2) are accessible from Ti(η5:η1-C5H4SiMe2Cl)Cl3 [54].
436
7
Half-Sandwich Complexes as Metallocene Analogs
Scheme 7-10
Since the molecules contain a mirror plane irrespective of the coordination of the appended donor group in solution, the question whether the additional donor is rigidly bonded or in a fluxional manner cannot be decided by 1H or 13C NMR spectroscopy, including variable-temperature NMR spectra. The corresponding dimethyl complexes offer the possibility to record NOE measurements and to study the coordination mode. Whereas in titanium compounds Ti(η5:η1-C5Me4SiMe2NCH2X)Me2 a spatial relationship between the TiMe2 signal and OMe and NMe2 groups could not be established, homologous zirconium complexes Zr(η5:η1C5Me4SiMe2NCH2X)Me2 exhibit an NOE between the proton signals of the ZrMe2 groups and the additional donor function OMe or NMe2. Thus, the zirconium (and hafnium) complexes retain the intramolecular coordination in solution. This can be ascribed to the higher Lewis acidity of zirconium and hafnium compared to titanium. For example, the crystal structure of the complex Hf(η5:η1:η1C5Me4SiMe2NCH2CH2OMe)Cl2 shows a rather unusual trigonal bipyramidal con-
7.3
Complexes with Polydentate Amido-Cyclopentadienyl Ligands
437
figuration at the hafnium center with the OMe group occupying a site trans to the ring ligand (Fig. 7-6).
Figure 7-6. Molecular structure Hf(η5:η1:η1-C5Me4SiMe2NCH2CH2OMe)Cl2 [55]. Dialkyl complexes M(η5:η1-C5Me4SiMe2NCH2CH2OMe)R″2 (M = Zr, Hf) can be synthesized with the alkyl ligands R″ including Me, Ph, CH2SiMe3 and CH2Ph, but most remarkably, dialkyls with β-hydrogen atoms (R″ = Et, nPr, nBu) can be isolated in this series. The crystal structure of the thermally most stable (dec. > 90°C) complex Hf(η5:η1-C5Me4SiMe2NCH2CH2OMe)(nBu)2 shows two undistorted nbutyl groups [55]. NMR spectra of the planar chiral derivatives M{(η5:η1-C5H3tBu)SiMe2NCH2CH2NMe2}Cl2(M = Ti, Zr) directly reveal the presence or absence of rigid intramolecular coordination of the NMe2 function. A rigid coordination of the NMe2 group prevents nitrogen inversion, giving rise to two nonequivalent methyl signals as in fact observed in the case of the zirconium complex Zr{(η5:η1C5H3tBu)SiMe2NCH2CH2NMe2}Cl2. In the case of the titanium derivative, Ti{(η5:η1-C5H3tBu)SiMe2NCH2CH2NMe2}Cl2, only one signal for the NMe2 group is observed in the temperature range of +80 to –80°C, suggesting that the coordination is not occurring or is fluxional on the NMR time scale [23]. A fluxional bonding mode (Eq. 7-13) is detected for the bulky dialkyl Zr{(η5:η1-C5H3tBu)-SiMe2NCH2CH2NMe2}(CH2SiMe3)2 with an activation energy of ∆G‡(7°C) = 13.1 kcal mol–1 [56].
438
7
Half-Sandwich Complexes as Metallocene Analogs
(7-13)
A tridendate ligand system with the additional donor attached directly to the cyclopentadienyl ring, {C5H4(CH2CH2NMe2)}SiMe2NHtBu, was prepared from 2(N,N-dimethylamino)ethylcyclopentadiene in a one-pot reaction (Scheme 7-11) [57]. Because of the 1,5-silyl migration two principal isomers are obtained in a 7 : 3 ratio for the 1,3- and 1,2-isomer. Treating Zr(NMe2)4 with the diprotonated ligand gives a mixture of the diamido complexes containing both the 1,2- and 1,3substituted ligands (Scheme 7-11). Reaction with (NMe2H2)Cl leads to the dichloro complex as the 1,3-isomer which can easily be methylated [58]. The complexation of this ligand with zirconium alkyls RnZrCl4–n (R = CH3, n = 3; R = CH2SiMe3, n = 2) unexpectedly favors the formation of the 1,2 isomer (Eq. 7-14) [58]. Only one single resonance was observed for the NMe2 group in the NMR spectra, suggesting fluxional coordination. The X-ray structure analysis of Zr{η5:η1:η1C5H3(CH2CH2NMe2)SiMe2NtBu}ClMe reveals a weak donor bond between the zirconium and the amine ligand.
Scheme 7-11
7.3
Complexes with Polydentate Amido-Cyclopentadienyl Ligands
439
(7-14)
Reaction between Sc(CH2SiMe3)3(THF)2, obtained in situ from the reaction of ScCl3(THF)3 with LiCH2SiMe3, and the diprotonated ligand {C5H4(CH2CH2NMe2)}SiMe2NHtBu, gives under alkane elimination the scandium complex Sc{η5:η1:η1C5H3(CH2CH2NMe2)SiMe2NtBu}(CH2SiMe3). In this case the complex is produced as the 1,3-isomer [57]. Although two diastereomeric pairs of enantiomers are possible (Fig. 7-7), NMR data suggest the selective formation of only one. NOE measurements indicate the proximity of two ring protons to the methyl groups of the alkyl ligand, suggesting a molecule in which the alkyl ligand lies below the two adjacent ring protons, as in the enantiomers (1S,RSc) and (1R,SSc). The alkyl complex Sc{η5:η1:η1-C5H3(CH2CH2NMe2)SiMe2NtBu}(CH2SiMe3) undergoes hydrogenolysis to give dimeric hydride {Sc{η5:η1:η1-C5H3(CH2CH2NMe2)SiMe2NtBu}(µ-H)}2 as a mixture of diastereomers. The alkyl loses SiMe4 after heating to 70°C for 6 days, resulting in the formation of a mixture of dimeric products. For each, one diastereomer is characterized by X-ray crystallography [57].
Figure 7-7. Possible diastereomers for Sc{η5:η1-C5H3(CH2CH2NMe2)SiMe2NtBu}(CH2SiMe3) [57].
440
7
Half-Sandwich Complexes as Metallocene Analogs
The hydroboration of the bridged amido half-sandwich complex containing a 2-propenyl group attached to the cyclopentadienyl ring, Zr{η5:η1C5H3(C3H5)SiMe2NtBu}Cl2, with [(C6F5)2BH]n [59], leads to a product of a highly regioselective 1,2 addition (Eq. 7-15). The 11B NMR chemical shift is in agreement with data found for neutral, three-coordinate boron centers, suggesting the absence of any association between the boron center and the ligands on zirconium. Reaction of the prolinol-functionalized ligand (C5Me4H)SiMe2OCH2{(–)-(S)C4H7NH} with Zr(NMe2)4 gives a linked amido complex that, upon treatment with (NMe2H2)Cl, forms a diastereomeric mixture of the N-protonated complexes (Eq. 716) [60].
(7-15)
(7-16)
7.4 Complexes with two Amido–Cyclopentadienyl Ligands During the synthesis of half-sandwich complexes containing the linked amido– cyclopentadienyl ligands, it was occasionally noted that two of the dianionic ligands are coordinated at a metal center to give compounds of the composition [M(η5:η1-C5R4ZNR′)2] [23, 27, 54, 55]. Evidently, the presence of excess ligand results in the double coordination of the linked amido–cyclopentadienyl ligands. Group 4 metals give neutral complexes with a characteristic C2-symmetric metallocene structure, exemplified by the crystal structure of Zr(η5:η1-C5H4SiMe2NPh)2 [31]. Group 3 metals give rise to anionic complexes of the type [M(η5:η1-C5R4ZNR′)2]–. Such complexes are easily isolated when tridentate ligands of the type C5R4SiMe2NCH2CH2X (X = OMe, NMe2) are used [61]. The
7.4
Complexes with Two Amido-Cyclopentadienyl Ligands
441
reaction of anhydrous yttrium or lutetium trichloride with Li2(C5R4SiMe2NCH2CH2X) produces, in high yields, hydrocarbon-soluble heterobimetallic complexes of the composition Li[M(η5:η1-C5R4SiMe2NCH2CH2X)2] (M = Y, Lu) (Eq. 7-17). As shown by the crystal structure of Li[Y(η5:η1-C5Me4SiMe2NCH2CH2OMe)2] (Fig. 7-8), the metallocene unit tightly coordinates the lithium ion with the ligand side chains. The amido nitrogen atoms exhibit a distorted tetrahedral geometry and a significantly longer yttrium–nitrogen bond than in Y(η5:η1-C5Me4SiMe2NtBu)-{N(SiMe3)2} [35]. When a tert-butylcyclopentadienyl ligand is used for C5R4, three diastereomeric pairs of enantiomers are expected (Fig. 7-9). Under kinetic control the sterically more constrained (R,S) isomer is preferentially formed, that converts into the thermodynamically more stable C2-symmetric (R,R) isomer in donor solvents such as THF at room temperature (Eq. 7-18). The interconversion is first-order and activation parameters indicate a fairly ordered transition state. Intriguingly, these complexes are highly efficient polymerization catalysts for ε-caprolactone, producing poly(ε-caprolactone) of high molecular weight (Mn > 30000) and moderate polydispersity (Mw/Mn < 2.0) [62].
(7-17)
(7-18)
442
7
Half-Sandwich Complexes as Metallocene Analogs
Figure 7-8. Molecular structure of Li[Y(η5:η1-C5Me4SiMe2NCH2CH2OMe)2] [61].
Figure 7-9. Possible diastereomers for Li[Y{(η5:η1C5H3tBu)SiMe2NCH2CH2OMe}2]. R′ denotes CH2CH2OMe coordinated to the lithium ion.
7.5 Complexes with Linked Alkoxo-Cyclopentadienyl Ligands
443
7.5 Complexes with Linked Alkoxo–Cyclopentadienyl Ligands Compared to amido-bridged half-sandwich derivatives, complexes containing a cyclopentadienyl with an alkoxo-functionalized side chain (X = O in Fig. 7-1) are relatively scarce. The first example Ti{η5:η1-C5Me4(CH2)3O}Cl2 was formed by thermolysis of the titanium ylide Ti{η5-C5Me4(CH2)3OMe}Cl2(CHPPh3) at 150°C (Eq 7-19) [63].
(7-19)
Scheme 7-12 More rationally, titanium complexes with a bridged alkoxo cyclopentadienyl ligand Ti{η5:η1-C5Me4(CH2)nO}Cl2 are prepared in quantitative yields by reacting the ligands as trimethylsilyl ethers, C5H4(SiMe3)(CH2)nOSiMe3, with TiCl4 (Scheme 712) [64]. The length of the bridge connecting the chelating alkoxo group and the cyclopentadienyl moiety influences the molecular structure dramatically: the (CH2)3 chain gives rise to a monomeric complex, whereas the shorter (CH2)2 chain
444
7
Half-Sandwich Complexes as Metallocene Analogs
affords a centrosymmetric dimer in which the alkoxo functions bridge the two titanium centers [64]. Alkoxo-functionalized fluorenyl complexes of zirconium of the type Zr{η5:η1-C13H8ZO}Cl2(THF)2 (Z = (CH2)2, (CHR)2) have also been reported [65]. The monomeric titanium fulvene complex Ti(η5-C5Me5){η7-C5Me3(CH2)2}, generated by thermolysis of paramagnetic titanium(III) complex Ti(η5-C5Me5)2R (R = Me, Et, Pr), gives an alkoxo bridging titanium complex after reacting with acetophenone (Eq. 7-20) [66].
(7-20) Likewise, the fulvene complex Ti(C5H4Me)(Ph)(η6-C5H4CH2), generated by thermolysis of the diphenyl derivative Ti(η5-C5H4Me)2Ph2, undergoes insertion with various aldehydes and ketones. NOE measurements indicate that acetaldehyde gives one diastereomer (> 95%) in which the methyl substituent and the titaniumbound phenyl ligand are disposed cis at the metallacyclic ring [67]. A remarkable formation of a chelate complex is observed during thermolysis of a titanium complex containing a tetramethylcyclopentadienyl ligand with a 2,6dimethoxyphenyl group (Eq. 7-21). Similarly, an ortho-methoxybenzyl functionalized titanocene Ti(η5-C5H4CEt2C6H4OMe)(η5-C5H5)Cl2 is converted in the presence of LiBr into a benzyloxo bridged complex Ti(η5:η1-C5H4CEt2C6H4O)(η5-C5H5)Cl following cleavage of the methoxy group (Eq. 7-22) [68].
(7-21)
(7-22)
7.6 Catalysis
445
The complexation of two deprotonated ligands Li2{(C5H3tBu)CH2CH2O} with ZrCl4 may result in three enantiomeric pairs of diastereomeric Zr{(η5:η1C5H3tBu)CH2CH2O}2 due to the presence of two chiral planes and the metallocene helicity (cf. Sec. 7.4). However, only one enantiomeric pair of one diastereomer, viz. (P,R,R) and (M,S,S), is formed. The alkylation of Zr{(η5:η1-C5H3tBu)CH2CH2O}2 with an excess of methyl triflate results in the formation of the unsymmetric complex Zr{η5:η1-C5H3tBu)CH2CH2O}{(η5-C5H3tBu)CH2CH2OMe}OTf. The exchange of one alkoxy against a triflato ligand increases the electrophilicity of the zirconium center blocking further electrophilic attack of MeOTf at the remaining bridging alkoxo group. The stereospecific formation of rac-isomer Zr{(η5C5H3tBu)CH2CH2OSiMe3}2Cl2 was achieved by reaction with chlorotrimethylsilane (Scheme 7-13) [69].
Scheme 7-13
7.6 Catalysis 7.6.1 Polymerization of Ethylene, α-Olefins and Dienes As was mentioned in the introduction, the great interest in the group 4 metal complexes containing the linked amido–cyclopentadienyl ligand, sometimes referred to
446
7
Half-Sandwich Complexes as Metallocene Analogs
as ‘constrained geometry catalysts’ [52], stems from their great potential as a new generation of olefin polymerization catalysts. In particular, the possibility of producing polyolefins with new rheological properties and good processability at temperatures as high as 160°C, has stimulated frantic activity in synthesizing and testing such complexes. The dimeric scandium hydride complex {Sc(η5:η1C5Me4SiMe2NtBu)(PMe3)}2(µH)2 was found to be capable of catalyzing the aspecific oligomerization of the αolefins propylene, 1-butene, and 1-pentene [12]. Although this presents an advantage over the scandocenes, the polymerization occurs slowly and to relatively low molecular weights (Mn = 4000 for poly(1-butene), 3000 for poly(1-pentene)). Addition of PMe3 results in slower α-olefin polymerization rates. The dimeric Lewis base free n-alkyl complexes {Sc(η5:η1-C5Me4SiMe2NtBu)}2(µ-CH2CH2R)2 (R = Me, Et) are shown to be more active catalyst precursors than the hydrido complex, and polymers with higher molecular weights (Mn = 6000 for poly(1-pentene), 9600 for polypropylene) are obtained. Low-temperature 13C NMR spectroscopic studies of the model complexes {Sc(η5:η1-C5Me4SiMe2NtBu)(PMe3)(CH2CHMeCH2CH2Me)} and {Sc(η5:η1-C5Me4SiMe2-NtBu)(PMe3){13CH2CH(13CH3)2} indicate that one PMe3 adduct is in equilibrium with only one PMe3-free species. By its symmetry it is concluded to be the monomeric 12-electron alkyl complex Sc(η5:η1C5Me4SiMe2-NtBu)R″ (Scheme 7-14).
Scheme 7-14
7.6 Catalysis
447
Methylaluminoxane-activated titanium complexes Ti(η5:η1-C5Me4ZNR′)Cl2 appear to be commercially utilized catalysts for olefin polymerization. Preliminary activity–structure relationships [52] show that these catalysts form, depending on the nature of the ligand framework, high molecular weight polyethylene with longchain branching, resulting from the incorporation of oligoethylene chains formed by β-hydride elimination. Also, superior properties as copolymerization catalysts were recognized, allowing efficient and uniform incorporation of higher α-olefin such as 1-octene to give low-density polyethylene with thermoelastic properties. This pronounced ability of these catalysts to incorporate higher olefins was ascribed to the more open coordination sphere, compared to conventional metallocene systems [52, 70]. Zirconium systems seem to be less active [52, 71], while the nature of the ligand substituents R in the C5R4 ring, R′ of the amido substituent NR′, and foremost the length of the bridge Z was found to influence the catalytic activity. Since catalyst precursors with the shorter bridge Z = SiMe2 and CH2CH2 were found to give the best polymerization characteristics, the bite angle of the chelating ligand (angle Cp–Ti–N, see Sec. 7.2.4) was proposed to be a crucial geometrical criterion for a catalyst to perform well [32]. It is obvious though that electronic properties imparted by R, R′ and Z are important as well, since a peralkylated cyclopentadienyl ring also appears to be more preferable. Finally, the tert-butyl group as amido-substituent R′ seems to be optimal [52]. In contrast to the 14-electron group 4 metallocenium polymerization catalysts [M(η5-C5R5)2R″]+, but in analogy to the scandium catalysts described above, the 12electron alkyl cation of the type [M(η5:η1-C5R4ZNR′)R″]+ is thought to be the active species [72]. It can be generated either by the reaction with methylaluminoxane, or by reacting a dialkyl complex with a Lewis acid such as B(C6F5)3. Structurally characterized complexes include [Zr(η5:η1-C5Me4-SiMe2NtBu)Me]+[FAl(C12F9)3]– [71b]. Variable-temperature NMR studies have revealed an activation barrier of 19.3(4) kcal mol–1 for the ion pair reorganization and symmetrization (Eq. 7-23), caused by the back-skip of the alkyl group in [Zr(η5:η1-C5Me4SiMe2NtBu)Me]+ [MeB(C6F5)3]– [71]. The position of the equilibrium between solvent-separated ion pair and contact ion pair for the complexes [Zr{η5:η1-C5H4(CH2)nNR′}(CH2Ph)]+[PhCH2B(C6F5)3]– has been found to be strongly dependent on the bridge length n and the amido substituent R′ [32]. Diene complexes Ti(η5:η1-C5Me4SiMe2NR′)(1,3-diene), described in Sec. 7.2, may serve as precursors for cationic species [47], since by adding B(C6F5)3, a zwitterionic species form, as was also observed in a zirconocene catalyst generated from Zr(η5-C5H5)2(η4-1,3-diene) and B(C6F5)3 [73].
(7-23)
448
7
Half-Sandwich Complexes as Metallocene Analogs
Polymerization of propylene by methylaluminoxane-activated Ti(η5:η1-C5Me4SiMe2NtBu)Cl2 results in atactic polypropylene with some slightly syndiotactic preference ([rrrr] = 20%) and 2–5% regioirregularity caused by 2,1-insertion. Chiral amido substituent R′ in one diastereomer of Ti(η5:η1-C9H6SiMe2NCHMePh)Cl2 does not increase the isospecificity [33]. Quite intriguingly, the fluorenyl-based system Zr(η5:η1-C13H8SiMe2NtBu)Cl2 [16] is thought to function as a syndiospecific catalyst when activated with methylaluminoxane and as an isopecific catalyst when activated with a Lewis acid. It was suggested that a tight ion pair may force the back-skip of the growing polypropylene chain that takes place prior to the next insertion, a process which would eventually lead to isospecificity [74]. As is evident from the significant regioirregularity and low stereoselectivity, the replacement of one cyclopentadienyl ligand in an ansametallocene by an amido group opens up the reaction site considerably. Cyclopolymerization of 1,5-hexadiene is also achieved by methylaluminoxane-activated Ti(η5:η1-C5Me4SiMe2NtBu)Cl2. The polymer obtained contains randomly distributed cis- and trans-cyclopentane rings [75]. The chromium(III) complex Cr(η5:η1-C5Me4-SiMe2NtBu)CH2SiMe3 is capable of polymerizing ethylene, but not 1-hexene, which is head-to-tail dimerized with a small amount of internal olefin formed [28]. The efficient copolymerization of ethylene with styrene became possible only with the use of metallocene catalysts. Conventional heterogeneous Ziegler catalysts normally induce homopolymerization of each of the monomers. Metallocene catalysts allowed poly(ethene-co-styrene) to be formed, but still suffer from low incorporation of styrene. One of the best copolymerization catalysts consists of the prototypical titanium complex Ti(η5:η1-C5Me4SiMe2NtBu)Cl2, that along with high activity results in the production of ethylene–styrene copolymer with up to 30 mol% styrene. By studying the influence of various substituents R and R′ in Ti(η5:η1C5R4SiMe2NR′)Cl2, it has been shown that both activity and the incorporation of styrene are sensitive to the nature of R and R′ [76]. Moreover, the maximum amount of styrene incorporated appears to be 66 mol% since, according to 13C NMR spectroscopic microstructure analysis of the copolymer, no more than two styrene units coupled in a tail-to-tail manner are present in the chain [77]. This can be explained by the inability of styrene to be inserted in the polymer chain once a secondary insertion of the ultimate styrene has occurred, following the (regular) primary insertion of the penultimate unit. This mechanistic proposal is supported by the isolation of the scandium complex {Sc(η5:η1-C5Me4SiMe2NtBu)(PMe3)}(CHPhCH2CH2CHDPh), from {Sc(η5:η1-C5Me4SiMe2NtBu)(PMe3)}2(µD)2, that cannot insert further styrene (Eq. 7-24) [12].
7.6 Catalysis
449
(7-24)
7.6.2 Hydrogenation and Hydroboration Eclipsed by the overwhelming interest in utilizing the complexes of linked amido– cyclopentadienyl ligands in olefin polymerizations, only two other applications of this class have so far been reported in the open literature.
(7-25)
(7-26)
450
7
Half-Sandwich Complexes as Metallocene Analogs
The enantioselective catalytic hydrogenation of substituted olefins and in particular of imines have been successfully achieved using reductively activated chiral Brintzinger-type titanocene derivatives [78]. When optically active titanium complexes of the type Ti(η5:η1-C5R4SiMe2NR′)Cl2 (R = H, Me; R′ = CHMePh) are treated with n-butyllithium, similarly active hydrogenation catalysts for imines are generated, albeit with low enantioselectivity (Eq. 7-25) [24]. Analogous to the titanocene system, the catalytically active species is presumed to be a titanium(III) hydrido species, possibly of the type Ti(η5:η1-C5R4SiMe2NR′)H. Hydroboration of alkenes with catecholborane is known to be catalyzed by metallocenes of group 3 metals [78]. The complexes Ti{η5:η1-C5H4(CH2)3NMe}Me2 and Zr{η5:η1-C5H4(CH2)3NMe}X2 (X = BH4, CH2Ph) are moderately active as catalysts for the hydroboration of 1-hexene using catecholborane (Eq. 7-26) [79].
7.7 Conclusion The introduction of the linked amido–cyclopentadienyl ligand as a replacement for the bridged bis(cyclopentadienyl) ligand in ansa-metallocenes has led to a new class of outstanding olefin polymerization catalysts. It is clear that the complexes of the type and M(η5:η1-C5R4ZNR′)L′mX′n exhibit some properties intermediate to those of ansa-metallocenes and half-sandwich complexes. However, many coordination chemical aspects have not been fully explored yet and it remains to be seen whether this ligand framework provides a metal template as versatile as the ubiquitous metallocene unit M(η5-C5R5)2. The linked amido–cyclopentadienyl complexes may be also considered as a hybrid between ansa–metallocenes and complexes containing chelating bis(amido) ligands derived from the diamines R′HN(Z)NHR′. The latter type of complexes has only just begun to emerge as yet another type of replacement for the bis(cyclopentadienyl) ligand and as homogeneous olefin catalyst with novel features complementary to both metallocenes and linked amido– cyclopentadienyl complexes [80].
References [1]
[2]
J. Okuda, Angew. Chem. 1992, 104, 49; Angew. Chem. Int. Ed. Engl. 1992, 31, 47; J. Okuda, Nachr. Chem. Tech. Lab. 1993, 41, 8; R. Mülhaupt, Nachr. Chem. Tech. Lab. 1993, 41, 1341; A. M. Thayer, Chem. Eng. News 1995, Sept. 11, 15; R. Mülhaupt, B. Rieger, Chimia 1996, 50, 10; W. Kaminsky, Macromol. Chem. Phys. 1996, 197, 3907. H.-H. Brintzinger, D. Fischer, R. Mülhaupt, B. Rieger, R. Waymouth, Angew. Chem. 1995, 107, 1255, Angew. Chem. Int. Ed. Engl. 1995, 34, 1652; P. C. Möhring, N. J. Coville, J. Organomet. Chem. 1994, 479, 1.
References [3] [4] [5] [6] [7] [8] [9]
[10]
[11]
[12] [13] [14] [15] [16] [17] [18] [19] [20] [21]
[22] [23] [24] [25] [26] [27] [28] [29]
[30] [31]
[32] [33] [34] [35]
451
J. W. Lauher, R. Hoffmann, J. Am. Chem. Soc. 1976, 98, 1729. W. E. Piers, P. J. Shapiro, E. E. Bunel, J. E. Bercaw, Synlett 1990, 2, 74. M. L. H. Green, J. Organomet. Chem. 1995, 500, 127. J. Okuda, Comments Inorg. Chem. 1994, 16, 185. P. T. Wolczanski, Polyhedron 1995, 14, 3335. D. S. Williams, M. H. Schofield, R. R. Schrock, Organometallics 1993, 12, 4560; D. M. Antonelli, M. L. H. Green, P. Mountford, J. Organomet. Chem. 1992, 438, C4. J. C. Flores, J. C. W. Chien, M. D. Rausch, Organometallics 1994, 13, 4140; P. Jutzi, U. Siemeling, J. Organomet. Chem. 1995, 500, 175; W. A. Herrmann, M. J. Morawietz, T. Priermeier, K. Mashima, J. Organomet. Chem. 1995, 486, 291. (a) J. C. Stevens, F. J. Timmers, G. W. Rosen, G. W. Knight, S. Y. Lai (Dow Chemical Co.), European Patent Application, EP 0 416 815 A2, 1991 (filed August 30, 1990). (b) J. A. Canich (Exxon Chemical Co.), European Patent Application, EP 0 420 436 A1, 1991 (filed September 10, 1990). P. J. Shapiro, J. E. Bercaw, Abstracts of Papers, 195th Meeting of the American Chemical Society, Toronto, Canada, American Chemical Society, Washington, D. C., 1988, INOR 584; P. J. Shapiro, Ph. D. Dissertation, California Institute of Technology, 1990; P. J. Shapiro, E. E. Bunel, W. P. Schaefer, J. E. Bercaw, Organometallics 1990, 9, 867. P. J. Shapiro, W. D. Cotter, W. P. Schaefer, J. A. Labinger, J. E. Bercaw, J. Am. Chem. Soc. 1994, 116, 4623. J. Okuda, Topics Current Chem. 1991, 160, 97. J. Okuda, Chem. Ber. 1990, 123, 1649. F. Amor, J. Okuda, J. Organomet. Chem. 1996, 520, 245. J. Okuda, F. J. Schattenmann, S. Wocadlo, W. Massa, Organometallics 1995, 14, 789. A. K. Hughes, A. Meetsma, J. H. Teuben, Organometallics 1993, 12, 1936. P.-J. Sinnema, L. van der Veen, A. L. Spek, N. Veldman, J. H. Teuben, Organometallics 1997, 16, 4245. H. V. R. Dias, Z. Wang, S. G. Bott, J. Organomet. Chem. 1996, 508, 91. W. P. Schaefer, W. D. Cotter, J. E. Bercaw, Acta Crystallogr. 1993, C49, 1489. This has long been known in titanocene chemistry: J. E. Bercaw, R. H. Marvich, L. G. Bell, H. H. Brintzinger, J. Am. Chem. Soc. 1972, 94, 1219; I. F. Urazowski, V. I. Ponomaryov, O. G. Ellert, I. E. Nifant’ev, D. A. Lemenovski, J. Organomet. Chem. 1988, 356, 181. G. A. Luinstra, J. H. Teuben, J. Chem. Soc., Chem. Commun. 1990, 1470. K. E. du Plooy, U. Moll, S. Wocadlo, W. Massa, J. Okuda, Organometallics 1995, 14, 3129. J. Okuda, S. Verch, T. P. Spaniol, R. Stürmer, Chem. Ber. 1996, 129, 1429. J. Okuda, T. Eberle, T. P. Spaniol, Chem. Ber. 1997, 130, 209. D. W. Carpenetti, L. Kloppenburg, J. T. Kupec, J. L. Petersen, Organometallics 1996, 15, 1572. U. Böhme, K. H. Thiele, J. Organomet. Chem. 1994, 472, 39. Y. Liang, G. P. A. Yap, A. L. Rheingold, K. H. Theopold, Organometallics 1996, 15, 5284. G. Chandra, M. F. Lappert, J. Chem. Soc. A 1968, 1940; M. F. Lappert, P. B. Power, A. R. Sanger, R. C. Srivastava, Metal and Metalloid Amides; Ellis Horwood: Chicester, West Sussex, U. K. 1980; G. M. Diamond, S. Rodewald, R. F. Jordan, Organometallics 1995, 14, 5. A. K. Hughes, S. M. B. Marsh, J. A. K. Howard, P. S. Ford, J. Organomet. Chem. 1997, 528, 195. W. A. Herrmann, M. J. A. Morawietz, J. Organomet. Chem. 1994, 482, 169; W. A. Herrmann, M. J. A. Morawietz, T. Priermeier, Angew. Chem. 1994, 106, 2025; Angew. Chem., Int. Ed. Engl. 1994, 33, 1946; M. J. A. Morawietz, Dissertation, Technische Universität München, 1995. P.-J. Sinnema, K. Liekelema, O. K. B. Stall, B. Hessen, J. H. Teuben, J. Mol. Cat., in press. A. L. McKnight, M. A. Masood, R. M. Waymouth, D. A. Straus, Organometallics 1997, 16, 2879. W. A. Herrmann, W. Baratta, M. J. A. Morawietz, J. Organomet. Chem. 1995, 497, C4; W. A. Herrmann, W. Baratta, J. Organomet. Chem. 1996, 506, 357. Y. Mu, W. E. Piers, M.-A. MacDonald, M. J. Zaworotko, Can. J. Chem. 1995, 73, 2233.
452
7
Half-Sandwich Complexes as Metallocene Analogs
[36] (a) S. Ciruelos, T. Cuenca, P. Gomez-Sal, A. Manzanero, P. Royo, Organometallics 1995, 14, 177; (b) S. Ciruelos, T. Cuenca, R. Gomez, P. Gomez-Sal, A. Manzanero, P. Royo, Organometallics 1996, 15, 5577. [37] M. L. H. Green, C. R. Lucas, J. Organomet. Chem. 1974, 73, 259. [38] L. Kloppenburg, J. L. Petersen, Organometallics 1996, 15, 7. [39] P.-J. Sinnema, Organometaalchemie en Homogene Katalyse, Verslagen Werkgroep, Rijksuniversiteit Groningen, 1995, 7, 23. [40] M. Mena, M. A. Pellinghelli, P. Royo, R. Serrano, A. Tiripicchio, J. Chem. Soc., Chem. Commun. 1986, 1118. [41] D. J. Crowther, R. F. Jordan, N. C. Baenziger, A. Verma, Organometallics 1990, 9, 2574; N. H. Dryden, P. Legzdins, J. Trotter, V. C. Yee, Organometallics 1991, 10, 2857. [42] R. Beckhaus, Angew. Chem. 1997, 109, 694, Angew. Chem. Int. Ed. Engl. 1997, 36, 686. [43] P.-J. Sinnema, Organometaalchemie en Homogene Katalyse, Verslagen Werkgroep, Rijksuniversiteit Groningen, 1995, 8, 93. [44] J. Okuda, T. Eberle, unpublished results. [45] S. L. Buchwald, R. B. Nielsen, Chem. Rev. 1988, 88, 1047. [46] L. van der Veen, Organometaalchemie en Homogene Katalyse, Verslagen Werkgroep, Rijksuniversiteit Groningen, 1996, 9, 87. [47] D. D. Devore, F. J. Timmers, D. L. Hasha, R. K. Rosen, T. J. Marks, P. A. Deck, C. L. Stern, Organometallics 1995, 14, 3132. [48] R. Walsh, Acc. Chem. Res. 1981, 14, 246. [49] P.-J. Sinnema, Organometaalchemie en Homogene Katalyse, Verslagen Werkgroep, Rijksuniversiteit Groningen, 1996, 9, 37. [50] H. Bürger, K. Wiegel, U. Thewalt, D. Schomburg, J. Organomet. Chem. 1975, 87, 301; J. Feldman, J. C. Calabrese, J. Chem. Soc., Chem. Commun. 1991, 1042. [51] R. M. Pulpi, J. N. Coalter, J. L. Petersen, J. Organomet. Chem. 1995, 497, 17. [52] (a) J. C. Stevens, Metcon 93, Houston, 26-28 May, 1993, p. 157. (b) J. C. Stevens, Stud. Surface Sci. Cat. 1994, 89, 277. [53] D. M. Giolando, K. Kirschbaum, L. J. Graves, U. Bolle, Inorg. Chem. 1992, 31, 3887. [54] J. Okuda, K. E. du Plooy, W. Massa, H.-C. Kang, U. Rose, Chem. Ber. 1996, 129, 275. [55] F. Amor, T. P. Spaniol, J. Okuda, Organometallics 1997, 16, 4765. [56] F. Amor, K. E. du Plooy, T. P. Spaniol, J. Okuda, unpublished results. [57] Y. Mu, W. E. Piers, D. C. MacQuarrie, M. J. Zaworotko, V. G. Young, Jr., Organometallics 1996, 15, 2720. [58] Y. Mu, W. E. Piers, D. C. MacQuarrie, M. J. Zaworotko, Can. J. Chem. 1996, 74, 1696. [59] R. E. v. H. Spence, W. E. Piers, Organometallics 1995, 14, 4617. [60] Y. Mu, W. E. Piers, L. R. MacGillivray, M. J. Zaworotko, Polyhedron 1995, 14, 1. [61] K. C. Hultzsch, T. P. Spaniol, J. Okuda, Organometallics 1997, 16, 4845. [62] K. C. Hultzsch, J. Okuda, Macromol. Rapid Commun., 1997, 18, 809. [63] R. Fandos, A. Meetsma, J. H. Teuben, Organometallics 1991, 10, 59. [64] G. Trouvé, D. A. Laske, A. Meetsma, J. H. Teuben, J. Organomet. Chem. 1996, 511, 255. [65] B. Rieger, J. Organomet. Chem. 1991, 420, C17. [66] J. W. Pattiasina, C. E. Hissink, J. L. de Boer, A. Meetsma, J. H. Teuben, J. Am. Chem. Soc. 1985, 107, 7758. [67] G. Erker, U. Korek, Z. Naturforsch. 1989, 44b, 1593. [68] Y. Qian, J. Huang, X. Chen, G. Li, W. Chen, B. Li, X. Jin, Q. Yang, Polyhedron 1994, 13, 1105. [69] J. Christoffers, R. G. Bergman, Angew. Chem. 1995, 107, 2423; Angew. Chem. Int. Ed. Engl. 1995, 34, 2266. [70] K. Soga, T. Uozumi, S. Nakamura, T. Toneri, T. Teranishi, T. Sano, T. Arai, Macromol. Chem. Phys. 1996, 197, 4237. [71] (a) Y.-X. Chen, C. L. Stern, S. Yang, T. J. Marks, J. Am. Chem. Soc. 1996, 118, 12451. (b) L. Jia, X. Yang, C. L. Stern, T. J. Marks, Organometallics 1997, 16, 842.
References
453
[72] (a) T. K. Woo, L. Fan, T. Ziegler, Organometallics 1994, 13, 2252. (b) T. K. Woo, P. M. Margl, J. C. W. Lohrenz, P. E. Blöchl, T. Ziegler, J. Am. Chem. Soc. 1996, 118, 13021. [73] B. Temme, G. Erker, J. Karl, H. Luftmann, R. Fröhlich, S. Kotila, Angew. Chem. 1995, 107, 1867, Angew. Chem. Int. Ed. Engl. 1994, 34, 1755. [74] (a) T. Shiomura, T. Asanuma, N. Inoue, Macromol. Rapid Commun. 1996, 17, 9. (b) T. Shiomura, T. Asanuma, T. Sunaga, Macromol. Rapid Commun. 1997, 18, 169. [75] F. G. Sernetz, R. Mülhaupt, R. M. Waymouth, Polymer Bulletin 1997, 38, 141. [76] F. G. Sernetz, R. Mülhaupt, F. Amor, T. Eberle, J. Okuda, J. Polym. Sci., Part A 1997, 35, 1571. [77] F. G. Sernetz, R. Mülhaupt, R. M. Waymouth, Macromol. Chem. Phys. 1996, 197, 1071; Y. Thomann, F. G. Sernetz, R. Thomann, J. Kressler, R. Mülhaupt, Macromol. Chem. Phys. 1997, 198, 739. [78] A. H. Hoveyda, J. P. Morken, Angew. Chem. 1996, 108, 1378; Angew. Chem. Int. Ed. Engl. 1996, 35, 1262. [79] E. A. Bijpost, R. Duchateau, J. H. Teuben, J. Mol. Cat. A. 1995, 95, 121. [80] J. D. Scollard, D. H. McConville, J. Am. Chem. Soc. 1996, 118, 10008; R. Baumann, W. D. Davis, R. R. Schrock, J. Am. Chem. Soc. 1997, 119, 3830; S. Tinkler, R. J. Deeth, D. J. Duncalf, A. McCamley, J. Chem. Soc., Chem. Comm. 1996, 2623.
454
7
Half-Sandwich Complexes as Metallocene Analogs
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides Ronald L. Halterman
8.1 Introduction The synthesis of Group 4 metallocene dichloride complexes has progressed enormously from the preparation of simple unbridged complexes with two identical unsubstituted or singly substituted cyclopentadienyl or indenyl ligands in the 1950s and 60s. In the 1970s the additional aspect of chirality was introduced by incorporating either unsymmetrically disubstituted cyclopentadienyl ligands (planar chirality) or by incorporating a chiral substituent on the cyclopentadienyl rings. The first application of chiral metallocenes as asymmetric catalysts, the enantioselective hydrogenation of alkenes, appeared in the late 1970s. At the end of the 1970s and beginning of the 1980s bridged ligands were introduced to form ansa-metallocene dichlorides containing either two cyclopentadienyl or two indenyl ligands. The 1980s saw a slowly expanding use of chiral metallocenes as catalysts in organic synthesis, but the event that most energized the field was the discovery that C2symmetric ethylene-bridged ansa-bis(tetrahydroindenyl)zirconium dichloride formed an effective catalyst for the stereoregular isotactic polymerization of propene. New syntheses of substituted bridged bis(indenyl) and bis(tetrahydroindenyl)metal complexes and additional chiral bis(cyclopentadienyl)metal complexes for application as polymerization catalysts rapidly followed. The development of new metallocenes in the 1990s has continued to expand to produce better catalysts for an increased variety of polymerization reactions and a rapidly increasing number of catalytic reactions of usefulness in asymmetric organic synthesis. To meet these demands, new bridged cyclopentadienyl, indenyl, and now fluorenyl complexes, as well as new chiral substituted cyclopentadienyl and indenyl complexes are being produced by an increasingly sophisticated variety of synthetic methodologies. The following chapter is intended to give an overview of the synthetic methodologies used over the past forty years in the synthesis of substituted ligands (unbridged and bridged) and the methods used to metalate these ligands to form titanocene or zirconocene dichlorides. By focusing on the synthetic methods used, the information should better guide practitioners to suitable adaptations for the synthesis of the next generations of desired metallocenes—whatever form they may take. The compounds are grouped according to ligand—the major division being unbridged or bridged ligands, the subdivisions being which of the three ‘elec-
456
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
tronic’ types of ligands (cyclopentadienyl, indenyl, or fluorenyl) are incorporated. Although they are properly considered to contain substituted cyclopentadienyl ligands, tetrahydroindenyl complexes are grouped with the indenyl ligands because of the strong historical co-development of these ligands. The preparation of most chiral or prochiral ligands is covered alongside the achiral versions since the synthetic chemistry is often based on the same chemistry. The preparation of ligands having stereogenic bridging groups is treated separately. The ramifications of ligand chirality on the formation of stereoisomeric metal complexes will be discussed in detail in the next section and will be tied into discussions of the metalation of particular ligands. In deciding on the material to be covered, limitations had to be set. This chapter covers the preparation of the titanocene or zirconocene dichloride complexes. The metallocene dichlorides are defined as complexes have two η5-carbocyclic πligands (cyclopentadienyl, indenyl or fluorenyl). These complexes are the most widely found and enable a reasonable comparison to be made of the synthetic methods used in their preparation. Heteroatom substituted ligands such as phosphacyclopentadienyl are not covered. Although many different metallocene complexes—such as metal dialkyls—are known, these complexes can usually be prepared from the metal dichloride complexes through fairly standard methods. The preparation of the less common hafnocene complexes is also not explicitly covered. Many papers dealing with the formation of zirconocene dichlorides also cover the corresponding hafnium complexes.
8.2 Ligand Chirality and Stereoisomeric Metal Complexes The metalation of cyclopentadienyl, indenyl or fluorenyl ligands can produce achiral or variety of stereoisomeric metallocenes depending on the symmetry of the ligands being metalated [1]. In general, chiral Group 4 metallocenes can have chirality arising in three ways: 1) the chirality may be based on a stereogenic metal atom with four different ligands as in 1; [2] 2) the chirality may be introduced by the coordination of the metal to chiral or prochiral ligands as in 2 [3], 3[4] or 4 [5]; or 3) it may be due to both of these elements. Since this review covers the formation of metallocene dichlorides having nonstereogenic metal centers, only chirality due to the coordination of chiral or prochiral ligands appears.
8.2
Ti
457
Ligand Chirality and Stereoisomeric Metal Complexes
CH3 Ti
Cl
Cl
Zr
Cl Zr
Cl
Cl
H
Cl
Cl
CH3 1 [2]
2 [ 3]
3 [4]
4 [5]
A key issue in determining what type of metal complexes (achiral, enantiomeric or diastereomeric) can form from particular ligands is how the two π-faces of a ligand are related to each other. Depending on the substituents on the ligands, the π-faces of the ligands can be related to one another in three ways: 1) the faces are equivalent or homotopic if the faces can interconvert by ligand rotation or if rotation about a C2 axis can interconvert the two faces; 2) the faces are enantiotopic if only reflection through a mirror plane can interconvert the two faces: or 3) the faces are diastereotopic if a rotation or symmetry operation cannot interconvert the faces. Metalation of ligands with homotopic or equivalent faces will produce only a single isomeric complex while the metalation of ligands with enantiotopic or diastereotopic π-faces can produce enantiomeric and diastereomeric mixtures of metallocene dichlorides. Ligands with Equivalent or Homotopic Faces. Two methods for determining facial equivalence are referred to here. If the π-faces of the ligand can interconvert by a ligand rotation about one of the single bonds between the ligand and a substituent (without the need for invoking a C2-symmetry operation), the faces are termed equivalent. If the ligands faces can only be interconverted by a C2-symmetry operation, the faces are described by a more restricted use of the term homotopic. In naming complexes derived from ligands with equivalent or homotopic πfaces, no stereochemical descriptor can be used for the π-coordination of the ligand since no new stereochemical element is introduced upon metalation. Metalation of ligands with equivalent faces will produce only a single isomeric metallocene dichloride complex. The simplest example of equivalent cyclopentadienyl faces is seen in the metalation of unsubstituted cyclopentadienyl to form the single, achiral titanocene dichloride complex 5 [6] (Scheme 8-1). With a single substituent on the cyclopentadienyl anion—whether achiral as in methylcyclopentadienyl (6) [7] or chiral and enantiomerically pure as in menthylcyclopentadienyl (7) [3]—bis(cyclopentadienyl)metal dichlorides 8 and 9 can be formed only as a single isomer since the ligand π-faces are equivalent by rotation about the cyclopentadienyl–methyl or cyclopentadienyl–menthyl single bond. Polysubstituted cyclopentadienyl ligands will have equivalent faces if each of the substituents is
458
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
equivalent as in, for example, 1,3-dimethylcyclopentadienyl (10) which can give only the single metallocene dichloride 11 [8]. CH 3 CH 3 TiCl4
Ti
–
Cl
TiCl4
Ti
–
Cl
Cl
6
5 [ 6]
Cl
8 [ 7]
CH 3 CH 3
CH 3 CpTiCl3
H
7
–
Ti H
Cl
TiCl4 –
Cl
H 3C H 3C
Ti
Cl Cl
CH 3 9 [ 3]
10
11 [ 8]
CH 3
Figure 8-1. Metalation of cyclopentadienyl ligands with equivalent faces give only a single metallocene. If a racemic mixture of chiral mono-substituted cyclopentadienyl ligands is metalated, a mixture of stereomeric complexes can result even though the ligand faces are equivalent. As illustrated in Scheme 8-2, addition of one equivalent of the racemic ligand 12 to the metal produces a racemic mixture of energetically equivalent enantiomeric cyclopentadienylmetal trichlorides 13. These now chiral metal complexes can choose to react further with either of the enantiomeric ligands in the solution to form a mixture of meso and dl 14 [9]. If both ligands on the metal are the same enantiomer, the chiral complex will have C2-symmetry. Since equal portions of the enantiomeric compounds 12 and 13 are present, the enantiomeric complexes will be formed in equal amounts and can be referred to as the racemic dl pair. If the two ligands on one metal are opposite enantiomers, this achiral complex will be the meso isomer. Since the meso isomer is a diastereomer of the dl pair, it can be energetically more or less stable and can be formed in greater or lower amounts than the dl pair.
8.2 Ligand
459
Ligand Chirality and Stereoisomeric Metal Complexes
Titanocene Dichloride Complexes
Initial complexes
Cl
Ph
Ph
Cl
Ti
Cl
Ph
– TiCl4
+
Ph
+
Ti
Cl
Ph
dl-12
Ph Ph
Ph meso -14
Ph
50:50 Racemic mixture
Cl +
Cl
–
Cl
Cl
Cl
Ti
Cl
Ti
Cl
Ti
Cl
dl-13 50:50 Racemic mixture
dl-14
Ph
Scheme 8-2. Metalation of a racemic substituted chiral cyclopentadienyl ligand [9].
Bridged bis(cyclopentadienyl) ligands will have equivalent faces when they are unsubstituted as in the tetramethylethylene-bridged bis(cyclopentadienyl) 15 [10] (Scheme 8-3). The introduction of chirality in the bridge as in 16 [11] does not effect the equivalence of the cyclopentadienyl π-faces and the metalation of 15 or 16 can produce only the single ansa-metallocene dichlorides 17 or 18, respectively. Di- or tetrasubstituted cyclopentadienyls can still have facial equivalence when the substituents are symmetrically placed about the bridging position as in 19, whose metalation can produce only a single ansa-metallocene as in 20 [12]. Cl –
–
Cl Ti
– TiCl4
Ti
Cl
H
15
17 [ 10]
H
H
Cl
–
H
16 18 [ 11]
–
–
Cl Zr Cl
19
20 [12 ]
Scheme 8-3. Metalation of bridged, symmetrically substituted bis(cyclopentadienyl) ligands.
460
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
TiCl4
–
Ti
Cl
2
–
Ph
ZrCl4
Ph Ph
Cl 22
21 [ 13]
ZrCl4
2
H
Zr
–
Cl
Zr
Cl
24 [14]
Cl
–
TiCl4
Ti
Cl
Cl Cl
H –
23
25 [15] 27 [16]
26 – TiCl4
Ti
Cl Cl
2
–
ZrCl4
Zr
– 28
Cl Cl
29 [17]
30 [18]
Scheme 8-4. Metalation of indenyl ligands with equivalent π-faces.
The simplest example of equivalent indenyl faces is seen in the metalation of the unsubstituted indenyl ligand to form a single, achiral complex 21 [13] (Scheme 84). Monosubstituted indenyl ligands will only have equivalent faces when the substituent, either achiral as in the phenyl-substituted 22 [14] or chiral as in the menthyl-substituted 23, [15] is placed at the 2-position of the indenyl ligand. These 2substituted indenyl ligands will give only the single metallocene dichlorides 24 and 25. Metalation of indenyl ligands bridged at the 2-position with either an achiral tether as 26, [16] or a chiral tether as in 28 [17] will lead to the single metallocene dichlorides 27 and 29. More highly substituted indenyls will still have equivalent faces as long as the substituents are symmetrically placed about the 2-position as in the formation of 30 [18]. Ligands with Homotopic Faces. Chiral cyclopentadienes with homotopic faces (faces which can only interconvert by only a C2-symmetry operation) such as 31 can inherently lead to only a single isomeric product 32 [19] due to the symmetry equivalence of the two ligand π-faces (Scheme 8-5). If the cyclopentadienyl ligand is bridged along the axis of symmetry, the resulting bridged ligand still has C2symmetry and can produce only a single product as with 33 [20]. Chiral C2-symmetrical indenyl ligands would be possible by symmetrical placement of chiral auxiliaries at the 1,3- 4,7- or 5,6-positions—but such ligands have not been reported.
8.2
Ligand Chirality and Stereoisomeric Metal Complexes
Cl
Cl ZrCl4
–
Ti
Zr
Cl
Cl 31
461
32 [19]
33 [20]
Scheme 8-5. Cyclopentadienyl ligands with homotopic faces. Ligands with Enantiotopic Faces. Cyclopentadienyl ligands with at least two different achiral substituents will have enantiotopic faces as in 1-isopropyl-3-methylcyclopentadienyl 34 [21]. The approach of an achiral metal moiety (e.g., CpTiCl3) to either face of the prochiral ligand 34 will be energetically equivalent (enantiomeric transition states) and will result in the production of two enantiomeric complexes 35 in equal amounts as shown in Scheme 8-6. The chirality in complexes of this type has been described as ‘planar’ chirality and the π-coordination of ligand carries an R or S stereochemical descriptor [22].
–
CpTiCl3
Ti
Cl Cl
+
Ti
Cl Cl
34 (p- R)-35
(50:50)
(p- S)-35
Scheme 8-6. Metalation of a cyclopentadienyl ligand having enantiotopic faces [21]. Note on Nomenclature. A note of definition is needed here on the R/S nomenclature used in describing metallocenes. According to rules set forth by Schlögl, the metal is treated as being individually bonded to all five of the π-atoms of the ligand and the configuration of one π-atom is used to describe the entire ligand [22]. The choice of the atom to be assigned has not been uniform in the literature. Schlögl has stated that the highest priority π-atom in unbridged ligands be assigned [22] whereas for ansa-metallocenes Brintzinger has used the site of the bridging group—the 1-indenyl position in (EBTHI)TiCl2—to assign the configuration [23]. The Brintzinger convention has been widely adopted although it is often at odds with a rigorous application of Schögl’s rule. Misconceptions can be avoided if authors make clear which atom is being used for the assignment. The following definition is unambiguous. “The planar chirality in cyclopentadienylmetal or indenylmetal complexes is assigned R or S based on the Cahn–Ingold–Prelog configuration of the 1-position of the ligand where the metal is treated as being individually bonded to all five of the π-atoms. The chiral-
462
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
ity can be described as (1R) or (1S) to avoid ambiguity or as (p-R) or (p-S) to emphasize that this is a description of the planar chirality based on the 1-position.” The 1-position is often the highest priority atom in unbridged metallocenes and has been commonly used in ansa-metallocenes. The 1-position is readily identified and is always stereogenic in planar chiral metallocenes. Even in cases where the bridging group in ansa-metallocenes is not at the 1-position, the 1-position can still be unambiguously used.
2 1
3
Ti
Cl
4
3 1
Cl
Ti
Cl Cl
C3
C
C Ti
C2
H
4 C C
6 7 7a 1 2
C5
H
Ti
C4
H
S
C1
Ti
C2
Cl Cl
Ti (1S)-35 or (p-S)-35
4 2
Cl Cl
3
Ti 1
1
i-Pr C5
3
3a 3
Zr
C1
H Ti
35 5
2
Ti
2 4
5
3. Assign configuration
2. Prioritize 1-position
1. Number ligand
C 3a 2
4 3 1
Zr
Cl Cl
1
CH2
H Zr C 7a C
4 C C
C1
Zr 1
H Zr C 2
H
C3
3
C2 S
C1
1
C 7a
Zr
Cl Cl
Zr (1S, 1 S)-3 or (p-S, p- S)-3
Scheme 8-7. Assignment of configuration in metallocenes having an element of planar chirality. This method is illustrated in Scheme 8-7 for titanocene dichloride 35 and ethylenebis(tetrahydroindenyl)-zirconium dichloride 3. In each case the 1-position is identified in Step 1. The priorities on C-1 are assigned in Step 2. The highest priority atom attached to C-1 is the metal. Of the three remaining carbon atoms attached to C-1 in 35, the exocyclic isopropyl group is not further attached to the metal and receives the lowest priority. Of the remaining atoms, C-2 has the priority over C-5 since C-3 is more highly substituted than C-4. In Step 3 the lowest priority group is placed in back and the decreasing order of priorities indicate that 35 has the S-configuration at C-1. Thus, 35 can be assigned the S-configuration [(1S) or (p-S)]. Applying this same method to the enantiomer of ansa-metallocene 3 depicted in Scheme 8-7 results in the (S,S)-assignment. Complexes with Two Enantiotopic Ligands. The situation where two equivalents of a prochiral cyclopentadienyl ligand add to one metal becomes somewhat more complicated. In this case, two diastereomeric bis(cyclopentadienyl)metal dichlorides (meso and dl) can be produced in different amounts (Scheme 8-8). The approaches of the achiral metal to either π-face of the ligand 34 are energetically equivalent and a racemic mixture of enantiomeric cyclopentadienylmetal trichlo-
8.2
463
Ligand Chirality and Stereoisomeric Metal Complexes
rides 36 is produced. When these now chiral metal complexes approach either prochiral face of the second equivalent of the cyclopentadienyl ligand 34, the pathways are now diastereotopic and one approach may be energetically favored to give a potentially unequal mixture of the meso 37 and dl 38 isomers. In this particular case, no selectivity was seen in the metalation [21].
TiCl4
–
Cl
Ti
Ti
Cl
Cl
Ti
Cl Cl
Cl
34 (p- R, p- R)-37
(p- S, p- S)-37
(p- R, p- S)-37 meso -37
dl-37 via
(diastereotopic metalations)
(enantiotopic metalation)
+
Cl
Ti
Ti
Cl
Cl
Cl
Cl
(p- S)-36
Cl
(p- R)-36
(50:50)
Scheme 8-8. Metalation of cyclopentadienyl ligands having enantiotopic faces [21]. Indenyl ligands with an achiral substituent at the 1-position also have enantiotopic ligand faces and their bis(indenyl)metal dichloride complexes also can exist as dl and meso isomers as seen in Scheme 8-9 for the metalation of 1-cyclohexylindenyl 38 to a mixture of dl-39 and meso-39. In this case, the formation of the dl isomer is favored [24]. R 2
–
ZrCl4
R
Zr
Cl
(p- R, p- S)-39 meso -39
Zr
Cl
R
38
+
Cl
+
Cl
R
Cl Zr R Cl
R (p- S, p- S)-39
(p- R, p- R)-39 dl-39
Scheme 8-9. Metalation of indenyl ligands having enantiotopic faces [24].
464
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
– ZrCl4
Cl –
Zr
+
Cl
Cl
Zr
Cl
+
Cl
Zr Cl
40 (1S,1S)41
(1R, 1S)-41
(1R, 1R)-41
meso -41
dl-41 (diastereotopic metalations)
via
–
– (enantiotopic metalation)
Cl
Cl Cl
Zr
Cl
Cl
(1S)-42
(50:50)
Zr
Cl
(1R)-42
Scheme 8-10. Metalation of bridged bis(indenyl) having enantiotopic π-faces [4]. The metalation of Brintzinger’s ethylene-bridged bis(indenyl) 40 can give dl- and meso-bis(indenyl)metal dichlorides 41 since the π-faces of each 1-substituted indenyl are enantiotopic [4] (Scheme 8-10). The initial metalation of 40 leads to formation of racemic enantiomeric indenylzirconium trichloride 42. The approach of the now chiral indenylmetal fragment to either face of the second indenyl moiety is a diastereomeric process which in this case has a lower energy barrier for the selective formation of the dl isomers. Both of the coordinated π-ligands can be described by a stereochemical descriptor to give the dl pair as (1R,1R)- and (1S,1S)41 and the meso isomer as (1R,1S)-41. Ligands with Diastereotopic Faces. Ligand π-faces are diastereotopic if a ligand rotation or symmetry operation (C2-rotation or reflection through a mirror plane) cannot interconvert the π-faces. This situation arises when an unsymmetrically polysubstituted ligand also contains a chiral substituent. Metal complexes of such ligands will carry at least one R- or S-stereochemical descriptor for the ligand substituent(s) and one p-R- or p-S-stereochemical descriptor for the π-coordination of the ligand. When an enantiomerically pure ligand such as 43 (with one chiral substituent having an R configuration) is metalated with cyclopentadienyltitanium trichloride, one new stereochemical element is introduced and a mixture of two diastereomeric complexes (R, p-R)-44 and (R, p-S)-44 can be produced in unequal amounts (Scheme 8-11). When two equivalents of an enantiomerically pure ligand are combined on one metal, a mixture of three diastereomeric bis(cyclopentadienyl)metal dichlorides can be produced, two C2-symmetric isomers [(R, p-R)- (R, p-R)]-45, [(R, p-S)- (R, p-S)]-45, and one C1-symmetric isomer [(R,
8.2
465
Ligand Chirality and Stereoisomeric Metal Complexes
p-R)- (R, p-S)]-45. Complexes such as the C1-symmetric isomer of 45 are often referred to as ‘meso-like’ in reference to the meso-relationship between the elements of planar chirality without regard for any chiral substituents. R*
R* CpTiCl3
–
Ti
Cl
R* +
Ti
Cl
Cl Cl
(R)-43 (R, p- R)-44
R* –
(R)-43
(R, p- S)-44
R* TiCl4
Ti
Cl Cl R*
[( R, p- R)-(R, p- R)]- 45
R*
R*
+
Ti
Cl Cl
Ti +
Cl R*
R* [( R, p- S)-(R, p- S)]- 45
Cl
[( R, p- R)-(R, p- S)]- 45
Scheme 8-11. Metalation of enantiomerically pure ligands having diastereotopic faces. If two equivalents of a racemic ligand having diasterotopic π-faces are added to one metal, the potential stereoisomeric mixture is now potentially horrendous— eleven stereoisomeric products, seven diastereomers and enantiomers of four these complexes can be produced. Unless the ligand exerts strong π-facial selectivity, a bewildering array of complexes will be formed in the metalation of diastereotopic ligands. The most common examples of cyclopentadienyl ligands with diasteromeric faces are annelated cyclopentadienes derived from the chiral pool. For example, enantiomerically pure camphor-cyclopentadienyl 46 can be metalated with cyclopentadienyltitanium trichloride to form a mixture of the two expected diastereomeric complexes of 47 [25]. When two equivalents of 46 are metalated three diastereomeric complexes can be formed. In this case, a strong π-facial preference is exerted and one of the C2-symmetrical isomers is formed as the major product (Scheme 8-12) [5, 25, 26].
466
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides 7S 4R 2
–
5
4R
7S
CpTiCle
1
7S
6
Ti
Cl
4R
+
Ti
Cl
3
(4R, 7S)-46
(4R, 7S,p- S)-47
(4R, 7S,p- R)-47
ZrCl4
7S
–
Cl
Zr
4R
+
Zr
Cl
Cl Cl
Cl Cl
+
Zr
Cl Cl
(4R, 7S)-46
[ ( 4R, 7S,p- S)(4R, 7S,p- S)]- 48
[( 4R, 7S,p- R)(4R, 7S,p- S)]- 48
[(4 R, 7S,p- R)(4 R, 7S,p- R)]- 48
Scheme 8-12. Metalation of camphor-cyclopentadienyl [5, 25, 26]. Interestingly, not all ligands with diastereotopic π-faces will produce chiral complexes. The well-studied norbornadienyl ligand 49 has diastereotopic π-faces and a mirror plane—but one which does not interconvert the π-faces of the ligand. The ligand fits the definition of diastereotopic faces since it is unsymmetrically substituted (it has an R-carbon at C-3a and S-substituent at C-7a, indenyl numbering) and contains chiratopic atoms (Scheme 8-13). The diastereomeric metal complexes 50 can be described as endo,endo , exo,exo, or endo,exo, or using the stereochemical descriptors, [(R,S,p-R)-49]2TiCl2, [(R,S, p-S)-49]2TiCl2, or [(R,S, p-R)-49, (R,S, p-S)-49]TiCl2 [27]. Each of these complexes is achiral and meso. 4R
7S
6 5
– 4R
7S
ZrCl4
1 2
Cl
Zr
Cl
3
Cl +
Zr
Cl
Cl +
Zr
Cl
S
(4R, 7S)-49 R
[(4 R, 7S,p- R)(4 R, 7S,p- R)] -50
[( 4R, 7S,p- R)(4R, 7S,p- S)]- 50
Scheme 8-13. Metalation of norbornacyclopentadienyl having diastereotopic faces [27].
[ ( 4R, 7S,p- S)(4R, 7S,p- S)]- -50
8.2
467
Ligand Chirality and Stereoisomeric Metal Complexes
An example of an indenyl ligand having diastereotopic faces is the 1-(neomenthyl)indenyl 51. Metalation of this ligand with cyclopentadienylzirconium trichloride produces the expected mixture of two diastereomers of 52 [28] (Scheme 8-14). In this case about a 6:1 ratio of the diastereomers is realized. When two equivalents of 51 are added to ZrCl4, a mixture of three diastereomers ensues—one C1-symmetric and two, C2-symmetric complexes—is formed [29]. Cl
– Zr
CpZrCl3
Cl
Cl
H
1S
Zr
Cl +
[28] H
(1S)-51 (1S,p- S)-52
6:1
(1S,p- R)-52
NM 0.5 ZrCl4
Cl NM
(NM = neomenthyl)
Cl
Cl
Zr
Zr
Cl
NM Cl
NM
NM
[( 1S,p- S)-(1S,p- S)]- 53
[( 1S,p- S)-(1S,p- R)]- 53
[29]
Zr NM
Cl
[( 1S,p- R)-(1S,p- R)]- 53
Scheme 8-14. Metalation of an enantiomerically pure indenyl ligand having diastereotopic π-faces.
Bridged ligands containing a chiral auxiliary either as a substituent on the ligand or in the bridging group will have diastereotopic π-faces. An example of a bis(cyclopentadienyl) ligand with an achiral bridge and a chiral auxiliary on only one of the cyclopentadienyl ligands is the dimethyl(menthylCp)(tetramethylCp)silane 54 (Scheme 8-15). Metalation of 54 should and does produce two diastereomeric complexes (3S,p-R)-55 and (3S,p-S)-55 [30]. If the racemic ligand 54 had been used, only two diastereomers would have been produced along with their enantiomers. An example of ligand in which both cyclopentadienyl groups have a chiral substituent and are bridged by an achiral group is the ethylene-bridged bis(bicyclooctylcyclopentadienyl) 56. Metalation of this ligand could give three diastereomers of 57, but only a single diastereomeric titanocene dichloride was formed due to a high degree of facial selectivity [20].
468
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Me Si
Me
–
Me Si
1
ZrCl4
– 3S
Me
Zr
Si Cl
1
H
[30]
Me
Cl +
Me
Cl Zr Cl
1
3S
3S
H
(3S,p- S)-55
(3S,p- R)-55
(1S)-54 1) n-BuLi, THF 2) TiCl3-(thf)3 Ti 3) HCl, CHCl 3, air
56
Cl
Cl
[20]
2 57
Scheme 8-15. Metalation of bridged ligands having diastereotopic faces. One of the few examples of a bis(cyclopentadienyl) ligand having diastereotopic faces due to a chiral bridging group is the (S,S)-butane-2,3-diyl-bridged bis(3-methylcyclopentadienyl) 58 (Scheme 8-16) [31]. The metalation of such ligands with two diastereotopic faces can produce three diastereomeric complexes. If the racemic ligand is used, the diastereoselectivity is unchanged, but each diastereomer will occur with its enantiomer. In the metalation of 58 a 4.6:2.8:1 ratio of the meso-like (S,S,p-R, p-S)-59 and the dl-like (S,S, p-R, p-R)-59 and (S,S, p-S, p-S)-59 is produced. An early example of a bis(indenyl) ligand having diastereotopic faces due to the presence of a chiral bridge at the 1-indenyl positions is the chiracene ligand 60 [32]. Metalation and reduction of the initial bis(indenyl)metals gave a 4.2:2.5:1 mixture of the three expected diastereomeric ansa-bis(tetrahydroindenyl)titanium dichlorides 61. Metalation of racemic ligands of this type will produce the same diastereoselectivity along with enantiomers of the diastereomers.
469
8.3 Unbridged Ligands
– 1) TiCl3
Cl
Zr
2) [O]
Cl
+
Cl
Zr
+
Cl
Cl
Zr
Cl
– (S,S,p- S,p- S)-59
(S,S,p- R, p- S)-59
(S,S,p- R, p- R)-59
(S,S)-58 dl-like
meso- like
– 1) TiCl3
Zr
Cl 2) [O] 3) H 2/ PtO 2
Cl
+
Cl
Zr
Cl
+
Cl
Zr
Cl
– (S,S,p- R, p- S)-61 (S,S)-60
(S,S,p- R, p- R)-61
(S,S,p- S,p- S)-61
meso- like dl-like
Scheme 8-16. Metalation of chiral bridged bis(cyclopentadienyl) having diastereotopic π-faces. In summary, the symmetry of the ligands play a major role in determining whether one or up to eleven stereoisomeric metal complexes may be formed.
8.3 Unbridged Ligands 8.3.1 Cyclopentadienes The preparation of substituted cyclopentadienes will be discussed according to the number of substituents on the cyclopentadiene. The preparation of any chiral derivatives is included as part of each section. In cyclopentadiene syntheses a mixture of double bond isomers is usually obtained. For simplicity only one of the doublebond isomers will usually be depicted and the numbering will be according to the relationship between the substituents without regard for the particular double bond positioning. Monosubstituted Achiral Cyclopentadienes. Two general, established routes have been used most often to prepare monosubstituted cyclopentadienes, the alkylation of a cyclopentadienyl anion or the formation of a pentafulvene followed by reduc-
470
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
tion or organometal addition. Both of these methods were already established in the 1930s [33] with more general applications appearing by the early 1960s [34, 35]. As shown in Scheme 8-17, a variety of different cyclopentadienylmetals in various solvents have reacted with a range of alkyl halides to give the monosubstituted cyclopentadienes [35–38]. The use of the sodium salt in THF or NH3 is the most common method.
–
M = K,
R–X M+
R
solvent
PhH, R-X = MeI, EtBr,
R
+
n-PrBr, n-BuBr (typical yields 50 to 80%) [36]
M = MgBr, Et 2O, R-X = allylBr, tert-BuBr
(44 to 51% yield) [35]
M = Na,
NH 3, R-X = MeBr, EtBr, i-PrBr, tert-BuBr, PhCH2Cl
M = Na,
THF, R-X = n-CnH2n+1Br(Cl), n = 5, 6, 7, 8, 10, 12, 14, 16, 18
M = Na,
THF, R-X = 3-Bromopentane
(typical yields 80 to 85%) [35,36] (typical yields 40 to 80%) [37]
(50% yield) [38]
Scheme 8-17. Preparation of monosubstituted cyclopentadienes by alkylation.
The facile condensation of aldehydes and ketones with cyclopentadiene to form a variety of fulvenes is the basis for the second common route to form monosubstituted cyclopentadienes. The cyclopentadienyl ligands are obtained by the reduction or deprotonation of the fulvenes or organometal addition to the fulvenes (Scheme 8-18). For example, 6,6-dimethylfulvene 62 could be reduced with LiAlH4 to give isopropylcyclopentadienyl, reacted with phenyllithium or methyllithium to give (α,α-dimethylbenzyl)cyclopentadienyl 63 or tert-butylcyclopentadienyl, or it could be deprotonated with sodium amide to give isopropenylcyclopentadienyl [39]. Each of these cyclopentadienyllithium or sodium complexes could be directly metalated to form group 4 metallocene dichloride complexes. This method has been extended to the conversion of diphenylfulvene 64 to (diphenylmethyl)cyclopentadienyl 65 or (triphenylmethyl)cyclopentadienyl. A number of other dialkyl fulvenes such as fulvenes 66 or 67 have been used to generate tert-alkyl substituted cyclopentadienyls [41,42]. Chiral but racemic ligands are readily prepared by reducing the unsymmetrical fulvene 68 by LiAlH4 [9]. A number of vinyl substituted cyclopentadienyls have been prepared by deprotonating pentamethylenefulvene 67 or the unsymmetrically substituted fulvenes 69 [43].
471
8.3 Unbridged Ligands H3C
CH3
LiAlH4, Et 2O, r.t. or PhLi, Et 2O, r.t. or
CH3
H3C
63 MeLi, Et 2O
66
Ph
R
H3C
Ph
CH3
R = n-C4H9 or n-C10H21
PhLi, Et 2O, r.t.
H3C
67
Ph
Ph
–
64
Ph Ph [40]
–
65 Ph
H3C
LiAlH4 or MeLi
R –
[39]
–
or
LiAlH4 or PhLi or MeLi
or
Ph
CH3
H2C
–
LiAlH4, Et 2O, r.t. or
[41]
–
CH3 CH3
or –
–
62 R
H3C
CH3
or
MeLi, Et 2O or NaNH2, NH3 H3C
Ph
H3C
[41]
H3C H3C
Ph
Ph
or –
[9]
–
68
R = H, Me, Ph H3C R'
H
LDA
CH3
CH3
R LDA
H
H
CH3
Ph or
[43] –
CH3 Ph
H
or –
[43] –
– 67
69
R' = H, R = Ph or Me; R' = Ph, R = H
Scheme 8-18. Preparation of monosubstituted cyclopentadienes by the fulvene route.
An additional, less general method for the preparation of monosubstituted cyclopentadienes is the dehydration of cyclopentenols. Monosubstituted cyclopentenols have been generated by the addition of phenyllithium to cyclopentenone or by the reduction of 3-methyl- or 3-phenylcyclopentenones [44, 45]. Chiral Monosubstituted Cyclopentadienes. Chiral, nonracemic monosubstituted cyclopentadienes have most generally been formed by either the addition of cyclopentadienylmetals to chiral alkylating agents or by the enantioselective reduction of fulvenes. The earliest synthesis of a chiral, nonracemic cyclopentadiene was Leblanc’s 1976 asymmetric reduction of fulvene 68 with a quinine-modified LiAlH4 reagent to give the cyclopentadienyl ligand 12 (Scheme 8-19). The enantioselectivity of the reduction was a modest 17.3% e.e. [46]. Additional examples of fulvenes being used to prepare chiral monosubsituted cyclopentadienes utilized steroidal precursors. In one synthesis steroidal ketone 70 was used to form fulvene 71 which was further converted to the cyclopentadiene 72 [47]. In this case, a 3:1 mixture of epimers were obtained due to incomplete selectivity in the addition of methyllithium to fulvene 71. In a second synthesis 5α-cholestan-3-one 73 was similarly converted to a fulvene which underwent selective methyllithium addition to give only
472
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
the single diastereomer 74 [48]. In both cases, the double bonds in the cyclopentadiene existed as a mixture of regioisomers.
CH3
H
LiAlH4 quinine, Et 2O
Ph
CH3
–
68
[46]
Li+ Ph
12
17.3% optical yield H3C
O
H N
H H
CH3Li Et 2O, -78 °C to r.t.
H
H
H
CH3OH
CH3
H3C
H3C
aq. workup
H
H H H
[47] 72
CH3O
70
71
CH3O
CH3O
3:1 ratio of epimers
H N 1)
O 73
2) CH 3Li, Et 2O, -78 °C to r.t.; aq. workup
CH3OH [48] 74
single diastereomer
Scheme 8-19. Preparation of nonracemic monosubstituted cyclopentadienes from fulvenes.
In 1978, Kagan and Cesarotti reported the first example of alkylating a cyclopentadienyl anion by a chiral nonracemic alkylating agent (Scheme 8-20). They employed the tosylates of menthol and neomenthyl to give natural product-derived neomenthylcyclopentadiene 75 and menthylcyclopentadiene 76 [3]. This menthyl framework was later extended to the (8-phenylmenthyl)cyclopentadiene 77 [5b]. The (2-methylbutyl)cyclopentadiene 78 and (2-phenylbutyl)cyclopentadiene 79 were also prepared using alkylating agents derived from the chiral pool [49]. Erker has demonstrated the use of steroid-derived alkylating agents to form chiral cyclopentadienes 80 and 81 [48].
473
8.3 Unbridged Ligands
NaCp
p-TsCl OH
pyridine
OTs
(–)-menthol
2) p-TsCl, pyridine
OTs
(–)-menthone
THF 0 °C to reflux, 1 h (32%)
76 [3]
NaCp
1) PhMgCl, CuI 2) L-Selectride 3) MsCl, Et 3N
O
75 [3]
NaCp
1) L-Selectride O
THF 0 °C to reflux, 1 h (24%)
OMs
THF r.t. 12 h, reflux 2 h (60%)
O
NaCp
1) LiAlH 4 OH
2) PBr3
R
77 [5b]
Br R
R 78: R = Ph 79: R = Me
R = Ph, Me
[49]
NaCp, THF 0 °C, 4h r.t., 24 h reflux, 2 h
MsO α-mesylate MsO
β-mesylate
[48] 80 β-cyclopentadiene (62%) 81 α-cyclopentadiene (25%)
Scheme 8-20. Preparation of chiral cyclopentadienes by alkylation.
Palladium-catalyzed alkylations of chiral allyl acetates by cyclopentadienylsodium have been used to prepare the nonracemic cyclohexenylcyclopentadiene 82 as well as the terpene-derived cyclopentadienes 83–85 (Scheme 8-21) [50].
474
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
AcO OAc
NaCp
NaCp
Pd(dba)2/dpe
Pd(dba)2/dpe (68%)
82
(40%)
83
OAc NaCp OAc
NaCp
Pd(dba)2/dpe
Pd(dba)2/dpe
(56%)
(40%)
84
85
Scheme 8-21. Chiral cyclopentadienes by palladium-catalyzed alkylation [50].
Chiral Monosubstituted Cyclopentadienes Containing Donor Atoms. With the widespread occurrence of oxygen and nitrogen-containing chiral natural products, the possibilities for incorporating derivatives having donor atoms in chiral cyclopentadienes are numerous. Examples of several donor substituted cyclopentadienes have been prepared and are depicted in Scheme 8-22 [51–54]. OR'
O
Ph
Me
NMe2
OMe
R R = Ph, Me, i-Pr [51,52]
Me [51]
[51]
Me
NMe2
[53]
[54]
Scheme 8-22. Chiral donor-substituted cyclopentadienes. Disubstituted Cyclopentadienes. The major problem in synthesizing disubstituted cyclopentadienes is controlling the selectivity for 1,2-disubstituted versus 1,3-disubstituted cyclopentadienes. In the dialkylation of cyclopentadiene, it is apparent that the 1,2-product is favored by electronic factors and the 1,3-product is favored by steric factors. With small alkyl groups, the 1,2-isomer is favored while with larger groups, the 1,3-product is favored (Scheme 8-23). For example, the dimethylation of cyclopentadiene favors the 1,2-dimethylcyclopentadiene over the 1,3-isomer in a 3.5:1 ratio [55]. The reaction of methylcyclopentadienylsodium with 1,2dibromoethane still favors the 1,2-product 86 [55]. Diethylation moderately favors the 1,2-isomer in a 57:43 ratio [56]. Only with the dialkylation with isopropyl groups does the steric interaction predominate giving the 1,3-isomer in a 93:7 ratio [56]. The alkylation of methylcyclopentadiene with long chain n-alkyl groups or with secondary alkyl bromides gave approximately 1:1 mixtures of the 1,2- and 1,3-isomers [37]. The alkylation of tert-butylcyclopentadienylmagnesium complex-
475
8.3 Unbridged Ligands
es with tert-butylhalides give the 1,3-di-tert-butylcyclopentadiene as the only regioisomer [35, 57]. The tert-butyl group can also direct 1,3-alkylation with smaller reagents as seen in the formation of the ethanol derivatives 87 and 88 [58].
CH3I
CH3I –
+
+
Na
–
Na
NH3 78 : 22
+
+
NH3
86
[55] iso -PrBr
CH3CH2Br
[55]
+
+ NaNH2, NH3
NaNH2, NH3
7 : 93 [56]
57 : 43 [56] t- BuCl
RBr –
– MgX+
74:26
Et 2O
Na+
only regioisomer [35,57]
p -TsOCH 2CH 2OMe
+ THF
R
R all ca. 50 : 50 OR
–
Li+
R= c-C6H11, sec -C4H9, n-C16H33Br
[37] [58]
or oxirane-BF3-OEt2 87: R = Me 88: R = H
Scheme 8-23. Preparation of disubstituted cyclopentadienes by dialkylation.
The condensation of ketones with monosubstituted cyclopentadienes can lead to very selective formation of fulvenes (Scheme 8-24). In an earlier procedure wherein sodium ethoxide is used as a base for the condensation of acetone with methylcyclopentadiene, an 80:20 ratio of 3-methyl and 2-methyl fulvenes 89 and 90 was formed [21]. Stone and Little have established a mild, high yield procedure for the formation of fulvenes using pyridine as the base [59]. This method has been used to selectively generate the 1,3-di-tert-butylcyclopentadiene 92 [60, 61]. Collins has applied this method to give exclusively the bis(3-substituted fulvene) 93 [62]. The addition of organometals to 3-substituted cyclopentenones has been used to selectively prepare 1,3-diphenylcyclopentadiene 94 [63] as well as 1-ethyl-3-methyl-, 1,3-dimethyl- and 3-methyl-1-phenylcyclopentadienes [8, 64, 65] (Scheme 825). 2-Methylcyclopent-2-enone has been converted by a similar method to 1,2dimethylcyclopentadiene 98 [66]. This same product 98 is also available through a Friedel–Crafts cyclization method [67]. The novel synthesis of 1,2-di-tert-butylcyclopentadiene 100 utilizes an intramolecular pinacol coupling of diketone 99 followed by a double dehydration [68].
476
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides O LiAlH 4 NaOEt EtOH
89
80:20
90
–
+
–
+
91 [21]
34
H N
O
1) MeLi
[60,61] 2) H 2O
CH3OH
92 R
H N
O
R
1) MeLi or LiAlH4
[62]
2) H 2O
CH3OH
R = Me or H
93
Scheme 8-24. Fulvene route to disubstituted cyclopentadienes.
O
1) EtMgBr or MeMgBr or PhLi
O
Ph 1) PhMgBr 2) H +
Ph
Ph
2) H +
Me
94 [63]
Et or Me
O 1)
MeMgBr
2) p-TsOH, quinone
Me
O
1) TiCl4 Mg/Hg 2) H 2O
99
97 [65]
1) LiAlH 4 2) H +
Me
Me
98 [66] O
Me
96 [8]
O PPA
O
Me
or
Me
95 [64]
O
Ph
Me
Me
Me
Me
98 [67] HO
OH
POCl3
[67]
pyridine
100
Scheme 8-25. Selective formation of 1,2- or 1,3-disubstituted cyclopentadienes. Annelated Achiral Disubstituted Cyclopentadienes. The direct formation of annelated cyclopentadienes through the reaction with dihaloalkanes generally fails due to the kinetically more favorable generation of spiro products. Thus, the alkylation of cyclopentadienylsodium with 1,4-dibromobutane leads to the spiro prod-
477
8.3 Unbridged Ligands
uct 101 [69] (Scheme 8-26). Fortunately, 101 can undergo a sigmatropic rearrangement to give tetrahydroindene 102 in good yield when the conditions are carefully controlled to prevent the thermodynamically favored rearrangement to 103 which ultimately leads to indane [69, 70]. The Skattebøl rearrangement from 104 has been used in an alternate synthesis of tetrahydroindene 102 [71]. Paquette has nicely extended this method of using dibromocyclopropanes to form intermediate carbenes which can rearrange to cyclopentadienes [72]. Another strategy for the synthesis of annelated cyclopentadienes has been to modify existing ring systems. For example, the Diels–Alder adducts 106 from cyclopentadiene and 109 from cyclopentadiene and cyclohexadiene, can be selectively hydrogenated and oxidized to the allylic alcohols 107 and 110. Dehydration of 107 and 110 leads to the 4,7methanotetrahydroindene 108 and the 4,7-ethanotetrahydroindene 111 [38].
Br(CH 2)4Br
gas pyrolysis +
NaNH2, NH3
[69]
+
+
325 °C
[69,70]
101
102
103
(72%) Br
:CBr2
Br
MeLi
[71]
Et 2O
104
102
105 1) H 2, Pd/C, EtOH
106
2) SeO2, Ac 2O, AcOH
OH
109
1% quinoline 160 to 175 °C
107
108 (27%)
[38]
1) H 2, Pd/C, EtOH 2) SeO2, dioxane, H 2O
Al 2O3
OH
110
Al 2O3 1% quinoline 160 to 180 °C
111 (48%)
Scheme 8-26. Preparation of annelated achiral cyclopentadienes. Annelated Chiral Disubstituted Cyclopentadienes. Reacting chiral dialkylating agents with cyclopentadienes can lead to chiral annelated cyclopentadienes (Scheme 8-27). In one case using the tartrate-derived ditosylate 112 the double alkylation lead directly to the ring fused product 113 due to the strain in the possible spiro product 114 [5]. A second C2-symmetric cyclopentadiene 117 was generated from the racemic ditosylate 115—going through the thermal rearrangement of
478
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
spiro intermediate 116. The enantiomerically enriched ligand could be made by replacing the toluenesulfonate groups with camphorsulfonates and resolving the sulfonates [73]. Related nonracemic bicyclooctane-annelated cyclopentadienes 120 and 121 bearing methyl or isopropyl groups were prepared from diols 118 which could be made in high enantiomeric purity through an asymmetric hydroboration reaction [19]. 2,2′-Di(bromomethyl)-1,1′-binaphthyl 122 could be converted the spiro-annelated cyclopentadiene 123 to the C2-symmetric fused cyclopentadiene 124 [74].
O O
NaCp, NaH, THF
OTs
r.t. 8 h; reflux 2 h
OTs
Ph
1) Ph 2CuLi TsO
OH
2) p-TsCl pyridine
Ph
Ph 115
HO
OH
MeSO2Cl
R
R Et 3N, CH2Cl2 MsO 118 (R = Me or i-Pr)
Br
NaCp, NaH THF
Br
0 °C, 2.5 h to reflux, 4 h
122
119
NaCp, NaH THF
OTs 0 °C, 2.5 h to reflux, 4 h
OMs
NaCp, NaH THF
R
0 °C, 2.5 h to reflux, 4 h
[5]
114 not observed
113
O
R
O
O
(31%)
112
O
O
Ph
220 °C
Ph
Ph
toluene
116
Ph 117
R
220 °C toluene
R
[73] R
R 120 R = Me [19] 121 R = i-Pr
220 °C toluene
123
124 [74]
Scheme 8-27. Double alkylation route to chiral annelated disubstituted cyclopentadienes. A number of related syntheses have appeared in which the cyclopentadienyl ring is generated through a cyclopentenone synthesis (Scheme 8-28). In the first example, camphor (125) was converted to phosphonate ester 126 which underwent an intramolecular Waadsworth–Horner–Emmons reaction to form cyclopentenone 127. Reduction and dehydration of 127 gave cam-CpH 128 [5]. A modified version of this synthesis has also appeared [25, 26]. A similar strategy was used to convert verbenone through cyclopentenone 129 to cyclopentadiene 130 [75] Paquette has used the Skattebøl rearrangement to produce cam-CpH 128 and pin-CpH 134 from camphor [25] or the nopol-derived tosylate 132 [72] (Scheme 829). In the first example, the needed diene 131 was made by a coupling reaction and in the second case, 133 was produced by an elimination. The cyclopropana-
479
8.3 Unbridged Ligands
tion, carbene formation and rearrangement gave the cyclopentadienes in good overall yield. i) LDA ii)
O Br
O
125
OMe P OMe O O
OMe
O
NaH THF
1) LiAlH 4, Et 2O
reflux
O
THF
O
127
126
O
2 equiv. LiCH2P(O)(OMe)2
OMe
O 2) p -TsOH, PhH r.t. 12 h
1) Me 2CuLi, Et 2O 2) LDA, HMPA; BrCH2CO2Me
128 [5]
1) LiAlH 4, Et 2O O
3) LiCH2P(O)(OMe)2 4) NaH, DME, reflux
2) p -TsOH, PhH
130 [75]
129
Scheme 8-28. Cyclopentenone route to annelated disubstituted chiral cyclopentadienes.
O
125
1a) LDA, THF 1b) PhN(Tf)2
PhHgCBr3
CH3Li
2)
PhH, reflux
Et 2O
SnMe3 Pd(PPh3), LiCl, THF OTs
131
Br
128 Br
1) PhHgCBr3
KO-t-Bu
[72]
2) CH 3Li, Et 2O
DMSO
132
[25]
133
134
Scheme 8-29. Skattebøl route to chiral disubstituted annelated cyclopentadienes. Achiral Trisubstituted Cyclopentadienes. The synthesis of trisubstituted cyclopentadienes through the polyalkylation of cyclopentadienes generally suffers from poor regioselectivity (Scheme 8-30). The polymethylation of methylcyclopentadiene gave a mixture of 1,2,3-, 1,2,4-, 1,1,2-timethyl isomers in addition to mono-, di-, tetra-, penta- and hexamethyl derivatives [55]. The alkylation of a (presumably 93:7) mixture of diisopropylcyclopentadienes surprisingly gave a 22:78 mixture of the 1,2,3- and 1,2,4-isomers [76]. Evidently the more hindered site between the isopropyl groups must be unusually reactive for this ratio to form. The more hindered 1,3-di-tert-butylcyclopentadiene reacts with tert-butyl chloride to give only
480
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
the sterically less hindered 1,2,4- isomer [77]. Similar results are obtained with two trimethylsilyl groups and one tert-butyl group 135 [78] or with three trimethylsilyl groups 136 [79].
–
x.s. Me 2SO4 +
Na+
+
NaNH2, NH3 ca. 50 : 22 : 28 [55]
iso -PrBr
t- BuCl
+
– Et 2O
NaNH2 NH3
MgX+
7 : 93 1,2- : 1,3Me3Si
n-BuLi
2) n-BuLi 3) Me 3SiCl
only isomer [77] Me3Si
Me3Si
SiMe3
1) Me 3SiCl –
(37%)
22 : 78 [76]
–
SiMe3
Me3Si
SiMe3
–
SiMe3
n-BuLi Me3Si
Me3Si
136 [79]
135 [78]
Scheme 8-30. Synthesis of trisubstituted cyclopentadienes through alkylations.
1,2,4-Trimethylcyclopentadiene 138 can be selectively prepared through a Grignard addition to enone 137 followed by dehydration (Scheme 8-31) [67]. The 1,2,3-trimethyl derivative 140 can be selectively prepared through reduction and dehydration of the readily available cyclopentenone 139 [80]. O
Me
O 1) MeMgBr
[67] Me
137
1) LiAlH4 Me
Me
O
2) H + Me
O
PbO 2
Me
Me
138
Me
Me
139
2) H +
[80]
Me
Me
140
Scheme 8-31. Selective syntheses of trisubstituted cyclopentadienes.
The Friedel–Crafts cyclization of 141 gives cyclopentenone 142 [81] which can be reduced and dehydrated to give 2-methyltetrahydroindene 143 (Scheme 8-32) [82]. Aryl Grignard reagents can add to cyclopentenone 144 to give after dehydration 2aryltetrahydroindene 145 [83]. An apparent exception to the usual lack of selectivity in the alkylation of sterically unhindered cyclopentadienes is the selective alkylation of norborna-CpH 146 to give the 1,2,4-substition pattern seen in 147 and 148 [84].
481
8.3 Unbridged Ligands Me
O O
PPA
O
[81]
1) LiAlH4
1) ArMgCl
2) H +
2) p -TsOH
[82] 141
142
Br
H
n Br
NaNH2, THF
Me
147
145
Ar = Ph, 4-MePh, 4-MeOPh, 4-BrPh
[84]
2) MeI
[83]
144
143
1) n-BuLi, THF
146
Ar
O
Me
146
148
n = 0, 2
n
Scheme 8-32. Achiral annelated trisubstituted cyclopentadienes.
Chiral Trisubstituted Cyclopentadienes. Alkylation of BCOCp 120 with dibromoethane gave a mixture of 1,2,3- and 1,2,4- isomers 149 and 150 whereas fulvene formation was selective for the 1,2,4- isomer 151 [20] (Scheme 8-33). Camphor-derived trisubstituted cyclopentadienes 154 were regioselectively generated using a Nazarov cyclization of bis(allylic)alcohols 153. These alcohols were readily prepared from the camphor-derived vinyllithium complex 152 [85].
n-BuLi
+
TsO
149
150
1:2
1) n-BuLi, THF 2) HMPA 3) acetone
OTs
120
THF/HMPA
151 [20]
85%
O H+
R Li
152
R = Me, Ph, tert-Bu
R
153
OH
154
R
[85]
Scheme 8-33. Chiral trisubstituted annelated cyclopentadienes. Tetrasubstituted Cyclopentadienes. A thorough review of very hindered cyclopentadienes has been recently published [1c] and only representative cyclopentadienes will be covered here. Tetramethylcyclopentadiene has been generated by a Grignard addition to trimethylcyclopentenone 139 followed by dehydration or by reduction and dehydration of tetramethylcyclopentadiene 156 [80] (Scheme 8-34). The alkylation of an isomeric mixture of triisopropylcyclopentadienes by isopropyl bromide gave the 1,2,3,4-tetrasubstituted cyclopentadiene 157 in good yield [76]. The pinacol coupling of tetraphenyl diketone 158 followed by double dehydration gave the tetraphenylcyclopentadiene 159 [86].
482
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
O
O 1) MeMgI
Me Me
Me
Me
[80]
2) H +
Me
Me
Me
Me
Me 155
139
2) I 2 Me
Me
Me
1) LiAlH4
Me
Me
Me
155
156 1) NaNH2
Na+
Ph
–
(i-Pr)3 2) i-PrBr, NH3 3) NaNH2 (85%)
O O
Ph
Ph
1) Zn, HOAc
Ph
2) H +
158
157 [76]
Ph
Ph
Ph
Ph
(60%)
159 [86]
Scheme 8-34. Achiral tetrasubstituted cyclopentadienes. Chiral Tetrasubstituted Cyclopentadienes. Two examples of tetrasubstituted chiral cyclopentadienes have been reported (Scheme 8-35). In each case a Nazarov ring closure approach was used in their synthesis. Addition of camphor-derived or verbenone-derived vinyl lithium complexes 152 or 160 to ethyl formate generated bis(allylic)alcohols which could be cyclized under acidic conditions to give good yields of the cyclopentadienes 161 [87] or 162 [88]. O H
2
100 - 150 °C
Li 152
HO
2
H Li
160
(80%)
H
O
R
H
KHSO4
OEt
161 [87]
R OEt
R
R CSA
R
PhH, 25 °C or I2 [88b]
R = Me, H
HO
H
(82%)
162 [88]
Scheme 8-35. Chiral tetrasubstituted cyclopentadienes. Pentasubstituted Cyclopentadienes. Since the original publication of a pentamethylcyclopentadiene synthesis by de Vries in 1960 [89], several modified or new routes have appeared (Scheme 8-36). In most cases tetramethylcyclopentenone 156 is prepared and converted through the addition of methyllithium or methyl Grignard followed by dehydration to pentamethylcyclopentadiene 163 [89, 90]. Two related syntheses convert 3-pentanone to pyrone 164 through sequential aldol reactions. Pyrone 164 can then be cyclized under acidic conditions to cyclopentenone 156 [90]. Alternatively, tiglic acid can be esterified to 2-butyl tiglate 165 which is similarly cyclized to give 156 [91]. A variant of this route can also be used to make ethyltetramethylcyclopentadiene [91]. A novel approach was taken by Ber-
483
8.3 Unbridged Ligands
caw in which two equivalents of 2-buten-2-yllithium are added to an ester to give divinylcarbinol 166 which can undergo a cationic π-cyclization to give directly pentaalkylcyclopentadiene 163 [92]. This route has the distinct advantage of readily incorporating various substituents by starting with the appropriate ester. In this way the ethyl-, n-propyl-, n-butyl- and phenyltetramethylcyclopentadienes were prepared. Pentaethylcyclopentadiene has also been prepared from 4-heptanone going through a cyclopentenone route [93]. Me
O Me
Me
2) I 2 Me
Me
Me
Me
163
O
O
156
1) MeMgI 2) H +
Me
163 O Me
H
KOH, MeOH
Me HO
164
H
156
Me
(38%)
O
p -TsOH
H
[90b]
[90]
O
(52%)
O
Me
O
KOH, MeOH
(57%)
Me
Me
(81%)
H2SO4
H
Me
Me
O
O
O
Me
Me
Me
Me
(72%)
156
O Me
Me
1) MeLi
KOH, LiCl (54%) MeOH
PhH (78%)
O
164 OH
O
O
O Me
P2O5 -MeSO3H
O
OH
[91]
25 °C
H2SO4, toluene, heat
165
(93%)
Me
Me
(50%)
Me
156 Me
O
R
OH
Li 2
R
H3O+
Me
Me
OEt Me
166
R = Me, Et, n-Pr, n-Bu, Ph
R
163(R = Me, 75%)
[92]
Scheme 8-36. Syntheses of pentamethylcyclopentadiene. Chiral Pentasubstituted Cyclopentadienes. Bercaw’s method of converting various esters to pentasubstituted cyclopentadienes [92] has been used to incorporate a chiral substituent into a pentasubstituted cyclopentadiene (Scheme 8-37). Addition of 2-buten-2-yllithium to the chiral ester 167 followed by cyclization of the intermediate divinylcarbinol gave (1-phenylpropyl)tetramethylcyclopentadiene 168 in high overall yield [94]. 1,2,3-Trimethylcyclopentadiene has been alkylated with the binaphthyl-based chiral dibromide 122 to give the chiral annelated pentasubstituted cyclopentadiene 169 [95].
484
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides Me O
Li
+
2
1) addition
Me
Me
[94]
OEt Ph
167
2) p -TsOH
(90%)
Me
168
Ph
Me Me Br Br
Me
Me
Me
Me 220 °C toluene
NaH, THF
Me
Me
Me 169 [95]
122
Scheme 8-37. Chiral pentasubstituted cyclopentadienes.
8.3.2 Indenes 1-Substituted Indenes. As with the generation and alkylation of cyclopentadienyl anions, the facile deprotonation of indene (pKa = 22) [96] enables its anion to react with alkylating agents to form a variety of alkylated indenes. Unlike the nonregioselective alkylation of most substituted cyclopentadienyls, the alkylation of the indenyl anion is electronically strongly preferred at the 1-position since alkylation at the 2-position would lead to the electronically less stable isoindene [97]. While the initial product is a 1-substituted indene, typical alkylation conditions generally promote the base-catalyzed double bond migration to form an equilibrium mixture of the 1- and 3-substituted indenes (Scheme 8-38) [97]. With most substituents such as alkyl or aryl groups, the 3-substituted indene is preferred. Since our main interest in substituted indenes is as the anionic ligand, it does not much matter which double bond isomer is present in the indene—the 1-substituted indenyl complex will be formed either from the 1- or 3-substituted indene. The preparation of a number of 1- (or 3-) substituted indenes is summarized in Scheme 8-38. The alkylation yields are generally quite good, with only the alkylation of indenyl anion by tert-butylbromide giving lower yields [97]. Through the careful control of reaction conditions, most notably the use of several equivalents of alkylating reagents, it is also possible to prepare unequilibrated 1-substituted indenes if desired [98]. The introduction of a trimethylsilyl group on the 1-position is readily accomplished in high yield. While the trimethylsilyl group prefers to remain on the allylic 1-position, it undergoes a 1,3-migration at room temperature [99, 100]. 3-(Methylthiol)indene can be prepared in good yield by the reaction of indenyllithium with dimethyldisulfide [101].
8.3 Unbridged Ligands
485
1) n-BuLi, THF 2) RX
H
R
R
MeI (-20 °C, 18 h, 93%), EtBr (5 °C, 18 h, 74%), i-PrBr (25 °C, 11 h, 69%), t-BuBr (35 °C, 8 h, 45%), PhCH 2Cl (25 °C, 4 h, 84%), Ph 2CHCl (25 °C, 4 h, 73%) 15% HCl
RMgX
170
O
THF, HMPA 25 °C, 8-12 h
25 °C, 10 h HO
[R = Me (94%), Et (85%),
R
R
n-Pr (90%), n-Bu (81%)]
.
Scheme 8-38. Preparation of 1- (or 3-)substituted indenes [97, 98].
A second general method for preparing 1-substituted indenes is through the addition of alkylmetals to 1-indanone 170 followed by acid-promoted dehydration. In this manner, simple alkyl groups could be introduced in good yields [102]. The addition of para substituted phenyl Grignard reagents to 1-indanone followed by dehydration was used to prepare a number of 3-arylindenes [96] 2-Substituted Indenes. Several methods for the preparation of 2-substituted indenes have been reported (Scheme 8-39). The alkylation of the enolate of 1-indanone (170) with alkyl halides followed by ketone reduction and dehydration is the least general with only the introduction of methyl being satisfactory [102]. A related method involves reaction of the aldol product 171 with either sodium borohydride or methylmagnesium iodide/Cu2Cl2 to give after ketone reduction and dehydration the 2-isopropylindene or 2-tert-butylindene [103]. The addition of nalkyl or aryl Grignard reagents to 2-indanone (172) followed by dehydration can provide some 2-substituted indenes in good yields [14, 96, 104]. The addition of secondary Grignard reagents to 2-indanone, however, does not proceed well due to competing deprotonation or reduction of the ketone [15]. Perhaps the most general route to 2-substituted indenes is the nickel-catalyzed cross-coupling between a wide variety of alkyl or aryl Grignard reagents with 2-bromoindene (173) [102, 105]. This method even allows for the coupling of secondary and tertiary Grignard reagents [105].
486
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
1) NaBH4
1) LDA R O 170 O R = Me (55%), Et (40%), n-Pr (22%) Me
171 O
[102]
R = Me (84%), Et (70%), n-Pr (70%)
NaBH4 or
Me R Me
Me MeMgI/ Cu2Cl2
CH3
2) 15% HCl
2) RX
O
Me R [103] Me
1) LiAlH 4 2) KHSO4
R = H, Me
R = H, Me
[96] Me [104] (Ar) Me toluene [14] (Ar) (Me, 85%; Ph, 64%)) OH H2SO4
MeMgI or ArMgX O Et 2O/C6H6 172
1.5 eq. RMgX
1) NBS, H2O
Br 5% Ni(dppp)Cl 2
2) p-TsOH 173
[102] R [105]
(R = Me, Et, n-Pr, n-Bu, iso- Pr, tert-Bu, Ph)
Scheme 8-39. Preparation of 2-substituted indenes. Indenes with Substitution on the 6-Membered Ring. Most preparations of indenes bearing substituents on the 4-, 5-, 6- or 7-position utilize Friedel–Crafts chemistry to generate an indanone which can be reduced and dehydrated to the substituted indene (Scheme 8-40). When only one substituent is desired at the 4(or 7-) position, the appropriate 3-(2-substituted phenyl)propionic acid is required. 4-Phenyl substituted indene 177a can be built up through homologation of 2-phenylbenzyl bromide 174 to the propionyl chloride 175 which can be cyclized to the 4-phenylindanone 176 [106]. Reduction and dehydration gave 4-phenylindene 177a. This method has also been applied in the preparation of related 4-substituted indenes. Through an analogous reaction sequence using 3-aryl-2-methylpropionic acid, the 2-methyl-4-phenylindene 177b could be made [106]. The base-catalyzed condensation of ketoaldehyde 178 with cyclopentadiene gives 4-isopropylindene 179 directly [107]. Conversion of 179 to 2-indanone 180 enables the incorporation of an additional substituent at the 2-position through the addition of a Grignard reagent followed by dehydration. For example, 2-methyl-4phenylindene 181 was formed in this manner [107]. Another general route to 4substituted indenes utilizes a nickel-catalyzed cross-coupling reaction between 4bromoindene 183 and a variety of Grignard reagents [102]. The main drawback of this route is the multistep Friedel–Crafts based synthesis of 4-bromoindene from 2bromobenzyl bromide 182.
8.3 Unbridged Ligands
[106]
Br 1) EtONa, RCH(CO 2Et) 2
174
R AlCl 3
175
O
KO-t-Bu
O
1) NaBH4
Cl
R
R
toluene
2) KOH 3) 130 °C 4) SOCl2
+
487
O
2) H 176
+
O
177a: R = H 177b: R = Me
O
[107]
CH3OH
H 178
179 H 1) HCO3H, H 2O 2
1) MeMgBr 2) p -TsOH
2) H 2SO4 179 Br
Me [107]
O 181
180
R
Br RMgX
Br
[102] cat. Ni(dppp)Cl 2
182
183
(96-98% yield) (R = Me, Et,
n-Pr, n-Bu)
Scheme 8-40. Preparation of 4-substituted indenes. As shown in Scheme 8-41, 4,7-disubstituted 1-indanones 185 are formed in good yield through the Friedel–Crafts acylation of electron-rich para substituted benzenes 184 with 3-chloropropionyl chloride followed by acid-promoted cyclization. Reduction and dehydration give the 4,7-disubstituted indenes 186. Electron poor indenes, such as the 4,7-difluoroindene 186c can be formed through conversion of the benzaldehyde 187 to the phenylpropionyl chloride 188 which can undergo a Friedel–Crafts acylation. Reduction and dehydration give the 4,7-difluoroindene 186c. Indenes bearing substituents in the 5,6-positions have also been prepared through Friedel–Crafts cyclization of the appropriate 3-(3,4-disubstiuted phenyl)propionic acids [18b]. These syntheses are less efficient since more than one regioisomer can form. The 5,6-dimethoxy-, 5,6-dimethyl- and 5,6-dichloroindenes have been reported [18b, 108]. A very facile synthesis of 4,7-dimethylindene 186a is available through the condensation of 2,5-hexanedione with cyclopentadiene [109].
488
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
R
O
R
1) Cl
1) NaBH4
Cl
AlCl 3 R 184 F
R
O
2) H +
2) H 2SO4 O
R 185
R
F
F
186a: R = Me 186b: R = OMe
H Cl F 187 R
PPA
R
O F 186c
F 188
HO
O
189
R
1) NaBH4
R
R
2) H +
R
190
O
Me O O Me 192
Me +
191a: R = Me 191b: R = OMe
KO-t-Bu CH3OH Me 186a
Scheme 8-41. Friedel–Crafts routes to 4,7- or 5,6-disubstituted indenes. Polymethylated Indenes. More highly substituted indenes have also been prepared using Friedel–Crafts methodology (Scheme 8-42). The PPA-promoted acylation of para-xylene with methylacylic acid gave indanone 193 directly in fair yield. Reduction and dehydration gave 2,4,7-trimethylindene 194 [110a]. A related reaction sequence using tiglic acid chloride and para-xylene produced 2,3,4,7-tetramethylindene 196 [110a]. Similarly, 1,2,3,4-tetramethylbenzene and tigloyl chloride could be converted to hexamethylindanone 197 which upon addition of methyllithium followed by dehydration gave 1,2,3,4,5,6,7-heptamethylindene 198 [110b]. Benzannelated indenes have been prepared through two related synthetic approaches (Scheme 8-43). 2-(Bromomethyl)naphthalene could be alkylated with malonate diester to give after decarboxylation and activation the acid chloride 200. Intramolecular Friedel–Crafts acylation of 200 followed by reduction and dehydration gave benz[e]indene 201a [106]. Following the same reaction sequence but using methyl malonate diester, 2-methylbenz[e]indene 201b could be formed. In each case both double bond isomers were present. The second approach utilizes the intermolecular acylation of naphthalene followed by cyclization to give either the methylbenzindanone derivative 202 or the benzindanone 203. Regioisomers of
489
8.3 Unbridged Ligands
both of these indanones were also formed. Reduction and dehydration led to the benz[e]indenes 201b and 201a [111]. O
Me
Me
Me
HO
Me
O 1) LiAlH 4 Me
2) p -TsOH
PPA Me
Me 193
Me 194
Me
Me
Me
[110a]
Me
[110a]
O Me
Cl
Me
O
1) AlCl 3 M e
1) LiAlH 4 Me
2) p -TsOH
+
2) H 3O
O Me
Cl
Me Me
Me Me 196
Me Me 195
Me
Me
1) AlCl 3
Me
O
Me
1) MeLi Me
Me
Me
Me [110b]
Me +
Me
2) I 2
Me
2) H 3O
Me
Me
Me
Me Me
Me
197
198
Scheme 8-42. Synthesis of polymethylated indenes [110].
(+ isomer)
1) EtONa, RCH(CO 2Et) 2 2) KOH 3) 130 °C 4) SOCl2
Br 199
Cl
O
1) AlCl 3, Toluene
R
2) NaBH4, THF/MeOH 3) p -TsOH
200
R
[106]
201a: R = H 201b: R = Me
(+ isomer) O +
1) Red-Al, THF
AlCl 3, CH 2Cl2, Br
Br
0 °C to r.t. Me Me 202 O
+
O O
Me [111]
Me 2) (HO C) 2 2 toluene O
201b
(+ isomer) 1) AlCl 3, CH 2Cl2,
1) Red-Al, THF
2) AlCl 3, NaCl, 120 °C 3) 200 °C
2) (HO2C)2 toluene 203
O
Scheme 8-43. Preparation of benzannelated indenes.
(+ isomer) [111] 201a
490
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Indenes with Chiral Substituents. A limited number of indenes bearing chiral substituents have been prepared (Scheme 8-44). The most common method has been to treat indenyllithium with a chiral alkylsulfonate. In this way, 3(neoisopinocamphyl)indene 205 was prepared from isopinocampheol 204 [112], and a series of menthyl 208, neomenthyl 206, neoisomenthyl 207 and isomenthyl 209 derivatives of indene [29]. Palladium-catalyzed cross-coupling reactions between menthylmagnesium chloride 211 and 3-indenyl triflate 210 or 2-bromoindene 173 provide another facile route to 3-menthylindene 208 and 2-menthylindene 212 [15, 114]. An example of an annelated chiral indene is the bicyclooctaneannelated indene 214 which was prepared by the bis(alkylation) of indene by chiral dimesylate 119 followed by a thermally-induced sigmatropic rearrangement of spiroindene 213 [115]. ((Author - I cannot find reference 113 cited in the text))
– HO
TsO
TsCl
Li+ THF, reflux, 72 h
pyridine
(58%)
204
[112] 205
Also used for: R = [29] 206 neomenthyl
R
207 neoisomenthyl
209 isomenthyl
208 menthyl
5% Pd(dppf)Cl 2
+
[114] MgCl
210
OTf
menthyl
208
211 5% Pd(dppf)Cl 2
menthyl
Br + MgCl 173
212
211
1)
NaH
2) MsO
[15]
220 °C Toluene
OMs
119
213
Scheme 8-44. Preparation of chiral indenes.
214
[115]
8.4
491
Bridged Ligands with Nonstereogenic Bridging Groups
8.4 Bridged Ligands with Nonstereogenic Bridging Groups 8.4.1 Bis(cyclopentadienes) The coverage of bis(cyclopentadienes) will be grouped according to an increasing level of complexity in the ligands. The major divisions are between nonprochiral ligands, prochiral ligands and ligands with chiral substituents. Ligands with stereogenic bridges will be covered in Sec. 8.5. Unsubstituted Cyclopentadienes. Cyclopentadienes linked by simple alkyl tethers can be prepared by reacting the appropriate alkyl dihalide with cyclopentadienyl anion (Scheme 8-45). In this way, methylene chloride reacted to give the methylene-bridged bis(cyclopentadiene) 215 (only one of the double bond isomers shown) which could be deprotonated to the more stable dilithio salt 216 [116]. 1,2Dibromoethane could be converted under improved conditions to ethylene-bridged bis(cyclopentadiene) 217 in good yield [62, 117]. The use of the magnesium salt of cyclopentadiene has proven to be crucial in minimizing the spiroalkylation of cyclopentadiene. 1,3-Propanobis(cyclopentadienyl)disodium 218 was similarly prepared [118]. The isopropylidene-bridged bis(cyclopentadiene) 219 could be prepared in a one-pot reaction between cyclopentadiene and acetone [119]. Presumably the intermediate 6,6-dimethylfulvene is formed and attacked by cyclopentadienyl anion. 6,6-Dimethylfulvene 62 can be CH2CH2
2 Li+
n-BuLi
– Na+
50 °C, 2 h
(61%)
216
THF, HMPA +
Br
Br
+
–
[62,117]
-78 °C to 0 °C
2Mg
2
217
(80%) Br
[116]
hexanes/ether 215
–
–
–
Br
1) THF, 0 °C, 1 h –
2 Na +
–
[118]
2) NaH, THF, 45-50 °C
Na+
218 O
NaOH powder, THF
[119]
+
2
phase transfer cat. 0 °C to r.t., 6 h (85%) –
Mg, CCl 4, THF 2
r.t. overnight 62
(25%)
219
[10,40]
–
2 MgCl+ 220
Scheme 8-45. Nonprochiral bis(cyclopentadienes) with simple alkyl tethers.
492
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
reductively coupled to give tetramethylethylene-bridged bis(cyclopentadienyl) salt 220 [10]. This coupling was observed as a side product in the dissolving metal reduction of 6,6-dimethylfulvene [40]. Silicon-containing bridges have become common owing to the facility with which chlorosilanes can react with cyclopentadienyl anions. The simple dimethylsilyl-bridged bis(cyclopentadienyl) ligand 221 (isolated as the more stable salt) was prepared from dichlorodimethylsilane (Scheme 8-46) [120]. Similarly, the disilylbridged complex 222 was produced [121]. The achiral cis-cyclopentane-1,3-diyl dimesylate 223 could be reacted with cyclopentadienylmagnesium bromide to give the achiral cyclopentane-1,3-diyl-bridged –
1) THF, - 60 °C, 1 h to r.t. –
2
+ [ClSi(CH3)2CH2] 2
–
2
+
–
H3C
MsO
–
1) THF, 0 °C - r.t.
OMs
Si
[122] 224 – "Ti(0)"
R
THF, reflux, 12 h
R
225
–
Na+
R
+ –
Na+
O OEt O
230
Na+ [123]
R
–
Na+
R = Me 226 R = Ph 227
EtO
+
(85%)
R
Na+
2 LI
222
223
–
[121]
Si
–
THF, r.t. 30 h
MgBr+
O
2 LI+
CH3 221
2) n-BuLi/hexane, 0 °C – 45 °C
Li+
[120]
Si
2) n-BuLi/hexane, - 80 °C to r.t. 2 h
Na+
2
–
+ (CH 3)2SiCl2
228 229
Et 2O, reflux 2 h + 4 Li
(85%)
231
HO
232
OH [124]
n-BuLi/hexane
conc. H2SO4
–
-10 °C 233
0 °C to reflux, 2 h
2 LI+
(10% from 232)
234
Scheme 8-46. Substituted nonprochiral bis(cyclopentadienes).
–
8.4
493
Bridged Ligands with Nonstereogenic Bridging Groups
bis(cyclopentadiene) 224 [122]. The McMurry coupling of cyclopentadienyl ketone 225 gave a stereoisomeric mixture of the dimethyl- or diphenyl-etheno-bridged bis(cyclopentadienyls) 226–229 [123]. Brintzinger has also converted diester 230 to bis(divinylcarbinol) 232 which can undergo cyclization in acid to form the ethano-bis(pentamethylcyclopentadienyl) 234 [124]. Two syntheses of doubly bridged achiral bis(cyclopentadienes) have appeared (Scheme 8-47). The trimethylsilyl groups in 235 prefer the less hindered 1,3-orientation to the ethano-bridge and can help direct the formation of the novel bis(fulvene) 237, albeit in low yield. Reduction of the fulvenes and desilylation gave the bis(cyclopentadiene) 238 [125]. Brintzinger has used a double Skattebøl approach to convert tetramethylenecyclooctane 239 to the doubly bridged dilithio salt 241 in good yield [126]. SiMe3 Li+
O O – Me3Si
Li+
SiMe3
– O
O
O
O
S
S
O 1) LiAlH 4, THF 0 °C
O 236
THF, -100 °C; Et 3N
235
(4%)
237
Me3Si
[125]
2) H 2O 3) ( n-Bu)4NF-3H2O THF, 0 °C (56%)
238 –
1) MeLi, Et 2O
2 CHBr3, t-BuOK pentane 239
Br
Br
Br
Br
240 (+ regioisomer)
2) aqueous workup 3) n-BuLi, pentane (41% from 239)
Li+ [126] – Li+
241
Scheme 8-47. Preparation of nonprochiral doubly bridged bis(cyclopentadienes). Prochiral Bridged Bis(cyclopentadienes). The preparation of prochiral disubstituted bis(cyclopentadienes) suffer from the same regioselectivity problems seen in the synthesis of disubstituted cyclopentadienes. By concentrating on the more selective synthetic methods developed for the unbridged ligands—the use of hindered tert-butyl or trimethylsilyl groups or the use of fulvenes—good, simple syntheses have been developed (Scheme 8-48). tert-Butylcyclopentadienylsodium reacts with 1,2-dibromoethane at the less hindered 3-position to form selectively the ethano-bis(cyclopentadiene) 245a. This reaction goes through a spiroalkylation followed by ring opening of the cyclopentane by tert butylcyclopentadienylsodium under forcing conditions [127]. The propensity of silyl groups to adopt the less hindered position was utilized in the bridging of substituted cyclopentadienyl derivatives to form 246. The silicon atom in such cyclopentadienes undergoes [1,5]shifts, but prefers to remain at the allylic carbon in a 1,3-relationship to the other substituent [128]. A number of related, silyl-bridged polysubstituted cyclopentadienes have been prepared [129]. 3-Substituted 6,6-dimethylfulvenes 244 undergo reductive coupling to give prochiral tetramethylethylene-bridged cyclopentadienes
494
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
247 [61]. Bis(dimethylfulvenes) can be regioselectively formed from bridged cyclopentadienes and reduced or alkylated to form bis(3-tert-butylcyclopentadienes) or bis(3-isopropylcyclopentadienes). In this way, ethylene-bridged, tetramethylethylene-bridged and cyclopentane-1,3-diyl-bridged bis(cyclopentadienes) 245 [62], 249 [130], and 250 [122] where formed from 217, 248 and 224, respectively.
BrCH2CH2Br, NaH – +
Na 242
[127]
dibenzo[18]crown-6 diglyme, -30 to 160 °C
245a
243 Me
Me2SiCl2
R
R
– Li+ R
THF/pentane, r.t. 4 h
Me Si 246
R
2 Mg, CCl 4 THF (25 - 30%)
244
R = t-Bu, 90% R = TMS R = Ph(Me) 2C R = Ph(CH 2)5C
R
R
[128,129]
R = t-Bu, i -Pr R = TMS R = Ph(Me) 2C , PhCH 2 [61] R = Ph(CH 2)5C
247
1) acetone, pyrollidine MeOH, r.t., 3 h
R
[62]
R 2) MeLi, Et 2O, 0-25 °C, 2 h or 2') LiAlH 4, Et 2O, 0-25 °C, 3 h 3) aqueous workup
217
245a,b
R = Me, H (ca. 95%)
1) acetone, pyrollidine MeOH, r.t., 3 h 2) MeLi, Et 2O 248 acetone pyrollidine MeOH (90%) 224
– Li+
– Li+
249
[130]
R
R
MeLi, Et 2O or LiAlH4, Et 2O; aqueous workup (90 - 93%)
[122] 250a: R =Me 250b: R = H
Scheme 8-48. Preparation of prochiral bis(cyclopentadienes). Collins has published routes for the selective synthesis of prochiral ethanobis(cyclopentadienes) in those cases where the above procedures would not be adequate (Scheme 8-49). Each of the methods relies on the selective generation of bis(cyclopentenones) and their conversion to the desired ligands. Hydroboration of 217 followed by oxidation and double bond migration gave bis(enone) 251. Organometal addition to the carbonyl followed by dehydration gave the methyl and ethyl derivatives 245c and 245d [62]. Bis(enone) 252 could be methylated and dehydrated to give bis(2-methylcyclopentadiene) 253. Fulvene formation at the remote
8.4
495
Bridged Ligands with Nonstereogenic Bridging Groups
position followed by the usual reduction or methyllithium addition gave bis(2-methyl-4-alkylcyclopentadienes) 254a,b [131]. The dimethyl derivative 254c could be formed through sequential methyl additions to bis(enones) 252 and 255 [131]. 1) disiamylborane 2) H 2O2, NaOH 3) TPAP, NMO 4) 1 N HCl dioxane
217
O O
Me 1) MeLa(OTf)2, THF 2) p -TsOH, C6H6 O
Me
R 245c: R = Me 245d: R = Et
R 1) acetone, pyrollidine MeOH
[131] Me 254a,b (R = Me, H)
O O 1) MeLa(OTf)2, THF
O
Me
Me
Me
1) MeLa(OTf)2, THF
2) PCC, pyridine CH 2Cl2
Me
O
R
[131]
2) p -TsOH, C6H6 Me
255
252
[62]
Me
2) MeLi, Et 2O or 2') LiAlH 4, Et 2O 3) aqueous workup
253
252
R
2) p -TsOH, Et2O 25 °C, 5 min
251
O
1) MeLi or EtMgBr THF 0 °C
254b
Me
Scheme 8-49. Controlled synthesis of prochiral bis(cyclopentadienes).
O
TsHNN
Co(CO)8
s-BuLi
H2NNHTs
CH3CN
THF
MeOH, p -TsOH NNHTs
O 258
257
256
259
[132]
I I
3% (Ph 3P)4Pd 15% CuI 260
1) H 2NN(H)Ts MeOH, cat. p- TsOH
Co2(CO)8
O
CH3CN
O
2) s-BuLi, THF
261
262
Scheme 8-50. Controlled synthesis of bis(tetrahydroindenes). Halterman has used a double Pauson–Khand approach to selectively form bis(tetrahydroindenes) (Scheme 8-50). Bis(enyne) 256 underwent the cobalt-promoted cyclization with carbon monoxide to give the bis(cyclopentenone) 257 selectively bridged at the 1-indenyl position [132]. The usual reduction and dehydration of 257 did not produce the desired bis(tetrahydroindene) 259, so an alternative
496
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
dehydration involving the Shapiro elimination of bis(tosylhydrazone) 258 was successfully brought into play. This synthetic approach was readily extended to the cyclization of bis(enyne) 260 to bis(enone) 261. Dehydration of 261 gave 1,2-phenyl-bridged bis(tetrahydroindene) 262 [132]. One report of a prochiral doubly bridged bis(cyclopentadiene has appeared (Scheme 8-51) [133]. Starting from either 1,2-dimethylcyclopentadiene or tetrahydroindene, the intermediate silyl-bridged bis(cyclopentadienyls) 263 could undergo a second reaction with dichlorodimethylsilane to give the doubly bridged complexes 264. R
R
–
Li+
1) (CH 3)2SiCl2, THF/hexane, r.t.
– Me
2) aqueous workup R R 3) n-BuLi, pentane, r.t. Li+ R R, R = CH 3, CH 3 R,R = –CH 2CH2CH2CH2–
R
R (CH3)2SiCl2
+
Li
Si
Me2Si
THF/hexane, r.t.
Me – R
[133]
SiMe2 R
263
264
R
(+ isomers)
Scheme 8-51. Synthesis of prochiral doubly bridged bis(cyclopentadienes).
1) THF, r.t. 10 h +
SiMe2Cl
+
265
– Na
–
K+
2) KH, 60 °C
–
K+
Si [30]
266
+ – Na+
1) THF, r.t. 10 h
SiMe2Cl 265
–
Li+
2) LiCH2SiMe3 pentane
–
Li+
Si
267
1) EtMgBr, THF 2) BrCH2CH2Br 120
+ 268
[20] 269
Scheme 8-52. Synthesis of bis(cyclopentadienes) having chiral substituents. Chiral Bridged Bis(cyclopentadienes). Only a couple reports have appeared dealing with the bridging of chiral cyclopentadienes with achiral bridging groups (Scheme 8-52). As in the earlier preparation of substituted cyclopentadienes, overcoming regioselectivity problems is a concern here. Marks has nicely used the propensity of silyl substituents to adopt the less hindered position to achieve regioselectivity. Chlorodimethyl(tetramethylcyclopentadienyl)silane 265 reacted with neo-
8.4
497
Bridged Ligands with Nonstereogenic Bridging Groups
menthylcyclopentadienylsodium or with menthylcyclopentadienylsodium to give after further deprotonation the silyl bridged chiral bis(cyclopentadienyls) 266 and 267 [30]. The bridging of bicyclooctane-annelated cyclopentadiene 120 with 1,2dibromoethane, on the other hand, gave a regioisomeric mixture of bis(cyclopentadienes) 268 and 269 [20]. One selective synthesis of a bis(cyclopentadiene) has been undertaken (Scheme 8-53). In this sequence camphor is refunctionalized to give the cyanoaldehyde 270 which can undergo a bis(Wittig) reaction to give the tethered 271. Conversion of the aldehyde groups in 271 to alkynes enabled an double Pauson–Khand reaction to convert bis(enyne) 272 to bis(cyclopentenone) 273. Reduction and dehydration gave bis(cyclopentadiene) 274 having the bridging group selectively incorporated at only one position [134]. H 1) i-AmONO 2) NaBH4
O
CN
H2SO4
NOH
CHO
H
1) CBr4, PPh 3 Me H
Me H
PPh3
O
Me H
2 2) n-BuLi
Ph3P
2) DiBAL-H
OH
H
H
1)
270 Co2(CO)8
H
Me
CH3CN
H
H
H
272
271
[134]
O
1) LiAlH 4 O
2) P 2O5, Py, SiO2
273 274
Scheme 8-53. Controlled synthesis of chiral bis(cyclopentadiene).
8.4.2 Bis(indenes) The importance of Brintzinger’s use of the ethanobis(1-indenyl) ligand in metallocene chemistry cannot be overstated. Many derivatives of bis(indenes) have been prepared and will be discussed in the order of unsubstituted indenes followed by substituted bis(1-indenes) and finally bridging at other indenyl positions. Indenes bridged by stereogenic groups will be discussed in Sec. 8.5. In Scheme 8-54, the preparation of unsubstituted bis(indenes) bridged by simple alkyl tethers is depicted. The original preparation of ethylenebis(indene) 275 was published in 1967 [135]. Improved conditions for the preparation of 275 include the convenient use of indenyllithium in THF [136, 137]. Bis(indenes) 276 [135]
498
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
and 277 [138] having longer chain butyl- and dodecyl-tethers have been prepared. cis-Dimesylate 232 has been used to form bis(indene) 278 [122]. 1) n-BuLi, THF, -78 °C [135-137]
2) BrCH2CH2Br, -78 °C to r.t. (54%) Br
Br
– +
MgBr
275
[135] toluene/xylene reflux 276
1) n-BuLi, THF, -78 °C [138] 2) MsOCH2(CH2)10CH2OMs 0 °C to r.t., 12 h (85%)
10 277
1) n-BuLi, THF, -78 °C 2)
[122] MsO
OMs (80%)
278
232
Scheme 8-54. Unsubstituted indenes with alkyl bridges. Silyl-bridged bis(indenes) have also become common (Scheme 8-55). In these ligands, the silicon atom is fluctional and prefers the sp3-1-indenyl position—resulting the generation of meso and dl isomers of the bis(indenes). No stereochemistry will be indicated here and it is inconsequential for metallocene chemistry since the stereocenter is no longer present in the deprotonated ligand. Unsubstituted indene is readily bridged to give the dimethylsilylene-bridged bis(indene) 279 [139, 140] or the disilane 280 [139]. Several substituted indenes whose preparation was described in Sec. 8.3.2 have been bridged by the dimethylsilylene group. Importantly, the silyl groups orient themselves away from the substituted 4-indenyl position as seen in 281 [106, 107] and 282 [106, 111]. The silyl group can still bridge even the more hindered 2,4,7-trimethylindenyl group in 283 [110a].
8.4
1) n-BuLi, Et2O
Si Me
THF
R
279
Si
– Cl
[139,140]
Me
2) Me 2SiCl2
Li+
Si
Si
Cl
[139]
Si
280
4
R 2
R = Et, R
4
2
1) n-BuLi, Et2O
R4
Si
2) Me 2SiCl2
= H R 2 = Me, R
R4
Me
R
R2
= H, i-Pr
R2 281
Me Si
2) Me 2SiCl2, 0 °C to 80 °C, 1 h
[106,111]
Me
R = Me, (70%) R = H, (44%)
201
[107,106]
Me 4
1) n-BuLi/hexane, Toluene:THF 20:1, 80 °C 1 h
R
282
R
Me Me
Me 1) n-BuLi, Et2O
Me
Me
Me 194
Me
[110a]
Si
2) Me 2SiCl2 Me
499
Bridged Ligands with Nonstereogenic Bridging Groups
Me Me
Me 283
Scheme 8-55. Indenes with silicon bridges. Dibromoethane has been used to tether a number of substituted indenes (Scheme 8-56). 4,7-Dimethylindene 186a and 5,6-disubstituted indenes 191 gave bis(indenes) 284 and 285 [18b]. Two equivalents of 2,4,7-trimethylindene can be directly converted to bis(trimethylindene) 286 or one equivalent to give (bromoethyl)indene 287. This monoindene could then be reacted with a different indene such as the 1,2,4,7-tetramethyl derivative to give unsymmetrically substituted bis(indene) 288 [110a]. Two novel bridging groups have been developed (Scheme 8-57). Brintzinger has reacted a titanium complex of dilithioferrocene with 1-indanone to give bis(indanol) 289 Dehydration of 289 gave ferrocene-bridged bis(indene) 290 [141]. Halterman has used the palladium-catalyzed coupling of 1,2-diiodobenzene with a zinc chloride complex of various indenyllithium complexes to give a series of 1,2phenyl-bridged bis(indenes) 291 [134].
500
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Me
Me Me
1) n-BuLi, THF
186a
Me X
2) 0.5 BrCH 2CH2Br, HMPA, -78 °C to r.t. (55%)
Me
1) n-BuLi, THF
X
284
X
2) 0.5 BrCH 2CH2Br, HMPA, -78 °C to r.t. X = Me, (72%) X = OMe, (56%)
X 191
Me
X 285 X
Me
Me
Me
Me
0.5 BrCH 2CH2Br, THF
–
Me -78 °C to r.t., 12 h
Li+ Me
Me
(54%)
286
Me
+
Me
Li
Me
THF, -78 °C to r.t.
1 eq. BrCH2CH2Br Me THF, -78 °C, 5 h to r.t.
Me
Me Me
(48%)
Br
(94%)
Me
[110a]
Me
Me
Me –
[18b]
Me
–
Li+
287 Me
Me
Me
Me Me
Me 288
Me
Scheme 8-56. Substituted indenes with ethano-bridge.
HO Li 1) Ti(i-PrO)3Cl, Et 2O, - 80 °C Fe
2) 2 equiv. 1-indanone - 78 °C to r.t. Li 3) aqueous KF workup
p -TsOH Fe
Et 2O, 2 h (9%)
289
290
R
R
R
I + I
[141]
Fe
HO
–
1.2 %Pd(PPh 3)4 [134]
THF, 95 °C, 36 h +
Li /ZnCl2
(30 - 45%)
291
R-Indene = Indene, 4-Meindene, 4,7-dimethylindene, s-hydroindacene
Scheme 8-57. Other bridging groups at 1-position.
8.4
HO OH
1) NCS, Me2S, CH2Cl2
Li SO2Ph
PhO2S
4) n-Bu4NF
292
OH
1) n-BuLi, THF, DTSCl 2) NBS, Me2S, CH2Cl2 3)
S O2Ph
PhO2S
2) Et 2NLi (8 equiv.) THF, -78 °C to 0 °C
[16]
(84%)
293 Li
294
(33%)
295
HO O Me
cat. SeO2 TBHP, CH2Cl2
Me
(14%)
296
501
Bridged Ligands with Nonstereogenic Bridging Groups
O H
H
1) PhMgCl 2) Swern Oxid.
O
O
(73%)
297
298
[142]
O PPA
1) LiAlH 4, THF
120 °C (54%)
2) p -TsOH, benzene (36%)
O 299
R 1) n-BuLi KO-t -Bu
R
R
Cl
1) AlCl 3
2) I 2
300
O
O
R
Cl
R
1) LiAlH 4
[143] 2) p -TsOH
2) H 2SO4
R
R 301
302
O
303
Scheme 8-58. Bridging at other positions, 2,2′ and 2,1′, 1,2′-doubly bridged.
While the common site for bridging indenes has been the 1-position, other bridging sites have been developed (Scheme 8-58). Nantz has used the dilithiated disulfone 293 to alkylate a benzyl bromide derivative to give 294. Activation of the benzyl alcohol groups and a second sulfone alkylation followed by elimination gave ethylenebis(2-indenyl) 295 [16]. A doubly bridged bis(indene) has been prepared from 1,5-dimethylcyclooctadiene 296 [142] Allylic oxidation of 296 gave bis(formyl) 297 which could be reacted with phenyl Grignard to give after oxidation the bis(phenyl ketone) 298. Treatment of 298 with polyphosphoric acid induced a Nazarov type ring closure to give bis(indanone) 299. Reduction and dehydration of 299 gave the doubly bridged bis(indene) 300 containing one bridged between the 1- and 2′-indenyl positions and the second between the 2- and 1′-positions [142]. The novel 7,7-bridged bis(indenes) 303 has been prepared from para-xylene or cymene through coupling to the biphenyls 301, Friedel–Crafts reactions to bis(indanones) 302 (+ regioisomeric ketones) and finally reduction and dehydration [143].
502
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
8.4.3 Bis(fluorenes) A limited number of bridged bis(fluorenes) have been prepared with achiral bridges, mostly by Alt (Scheme 8-59). The simple bridging of fluorenyllithium with 1,2dibromoethane or dichlorodimethylsilane gave the ethylene-bridged or the dimethylsilylene-bridged bis(fluorene) 304 [144] or 305 [145, 146]. Reaction of di-tertbutylfluorene 306 with dichlorodimethylsilane gave monofluorene 307 which could further react with fluorenyllithium to give the unsymmetrical bis(fluorene) 308 [147]. The benzannelated fluorene 309 could be bridged to give the ethanobis(fluorene) 310 [148]. The biphenyl-bridged bis(fluorenes) 312 could be prepared from 2,2′-dihalo-1,1′-biphenyls 311, but these bis(fluorenes) could not be deprotonated [148].
1) n-BuLi, THF, r.t. 1 h
Me Me Si
1) n-BuLi, Et2O [144]
2) 1,2-Dibromoethane THF, r.t. 3 h 3) aq. workup (70%)
2) Me 2SiCl2, Et 2O, r.t. to reflux, 6 h 3) aq. workup
[145] [146]
(70%)
305
304
1) n-BuLi, Et2O r.t., 2 h
FluorenylLi
2) Me 2SiCl2, pentane r.t. 1 h
Me Si
SiMe2Cl Et 2O, TMEDA 4 h , r.t.
(90%)
[147]
Me
(85%)
306
307
308
R 1) n-BuLi, Et2O, r.t. 3 h 2) 1,2-Dibromoethane r.t. 4 h 3) aq. workup
X X
1) n-BuLi, Et2O 0 °C to r.t. 2 h
R
R
2) Fluorenone - 78 °C to r.t. 2 h 3) HI/HOAc, 2 h
(80%) R
309 310 [148]
311a: R = H, X = Br 311b: R = t-Bu, X = I
312 [148]
Scheme 8-59. Bis(fluorenes)
8.4.4 Mixed Ligands This section will cover bridged ligands between two different ligand types, starting with cyclopentadienyl^indenyl ligands, then covering the more developed
8.4
503
Bridged Ligands with Nonstereogenic Bridging Groups
cyclopentadienyl^fluorenyl ligands and ending with indenyl^fluorenyl ligands. Bridged complexes with chiral bridging groups will be covered in Sec. 8.5. Cyclopentadienyl^Indenyl. Indenyl anions add to fulvenes much more readily than do cyclopentadienyl anions and this method been used in the preparation of a number of cyclopentadienyl, indenyl compounds (Scheme 8-60). Green first applied this method with 6,6-dimethylfulvene and pentamethylenefulvene to give the mixed ligand 314 [149, 150]. The method has been extended to 6,6-diphenylfulvene [151]. A series of prochiral ligands has been prepared from the addition of 3substituted indenes to 3-tert-butyl-6,6-dimethylfulvene to give ligands 315 [152]. Through sequential addition of one equivalent of indenyllithium and cyclopentadienylsodium to dichlorodimethylsilane, cyclopentadienylindenylsilane 316 has been prepared [140]. The substituted (t-butylcyclopentadienyl)(t-butylindenyl)silane 317 has been similarly prepared [152]. R
R
R Et 2O
–
+
–
0 °C to r.t. 22 h
Li+
Li+
62 R, R = Me, Me 67 R + R = –CH 2CH2CH2CH2CH2– 64 R, R = Ph, Ph [151] Me
R
R
n-BuLi, Et 2O 0 °C to r.t. 20 h
– +
Li
t-Bu [152]
aq. workup
R = Me, t-Bu
315
R Me
1) Me 2SiCl2, Et 2O, 0 °C
Me Si [140]
– 2) NaCp, Et 2O, THF 0 °C to r.t. 3) aq. workup (92%)
Li+
+
316 Me Me Si
t-Bu
SIMe2Cl
t-Bu
THF – Li+
t-Bu
Li
Me THF
244
Li+ +
Me
R
t-Bu
– –
314 [149]
313
Me +
R
[152]
aq. workup
t-Bu 317
Scheme 8-60. Bridged cyclopentadiene, indenes. Cyclopentadienyl^Fluorenyl. The most common mixed ligand is the combination of cyclopentadienes and fluorenes. The synthetic methods have been based on ei-
504
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
ther the addition of fluorenyl anions to fulvenes or on the use of silyl bridges. The synthesis of the unsubstituted ligands are shown in Scheme 8-61. A variety of fulvenes have been applied in the fluorenyl addition reaction to yield the one-carbonbridged ligands 318 [153–156]. The silyl-bridged ligand 320 has been best prepared from the initial addition of fluorenyllithium to dichlorodimethylsilane, followed by the addition of cyclopentadienyllithium [145]. R
R
Li+ –
+
R
R
1) THF [153-156]
2) aq. workup R, R = Me, Me; R, R = Ph, Ph R + R = –CH 2CH2CH2CH2CH2– 1) Me 2SiCl2, hexane - 78 °C to r.t. 2 h Li+
318
Me
–
Me Si Cl
Me CpLi, THF
Si
Me [145]
r.t. 4 h
2) crystallization at -20 °C
(71%)
(77%) 319
320
Scheme 8-61. Bridged unsubstituted cyclopentadiene, fluorenes. A number of ligands having substituents either on the cyclopentadienyl or fluorenyl moieties has been prepared (Scheme 8-62). Isopropylidene-bridged 3-substituted cyclopentadiene, fluorene compounds 321 and 322 have been prepared by addition of fluorenyllithium to 3-substituted 6,6-dimethylfulvenes [110a, 157]. A number of substituted fluorenyl anions have been added to fulvenes to produce the substituted compounds 323 [156, 158]. The preparation of a number of methylenebridged cyclopentadienyl, fluorenyl compounds 324 has been accomplished through the addition of the substituted fluorenyl anions to 6-dimethylaminofulvene followed by reduction of the incipient fulvene [159]. A substituted fluorenyl, cyclopentadienyl silyl compound 325 has prepared from intermediate 307 [147].
8.4
R
1) MeLi, THF R
2) M e
321: R = Me [159] 322: R = t-Bu [110a]
Me
3) aq. workup R1
R1
Li+
R1
R1
1) THF
–
+
R2
2) aq. workup R 1 = Me or Ph
H
R
505
Bridged Ligands with Nonstereogenic Bridging Groups
[156,157]
R2
2
R
323
NMe2
2
Li+ 1) THF
–
+
R2
R1
R1
2) LiAlH 4 3) aq. workup
[158] 324
1) n-BuLi, Et2O r.t., 2 h
NaCp
SiMe2Cl
2) Me 2SiCl2, pentane r.t. 1 h
Et 2O, HMPA 4 h , r.t.
R2
Me Si Me [147]
(90%) 307
325
Scheme 8-62. Bridged substituted cyclopentadiene, fluorenes. Indenyl^Fluorenyl. The addition of indenyllithium to fluorenylsilyl chloride 319 has been used to prepare the fluorenyl, indenyl mixed ligand 326 (Scheme 8-63) [145].
506
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
1) Me 2SiCl2, hexane - 78 °C to r.t. 2 h Li+
Me
–
Me Si IndenylLi, Et2O Cl
Me
Si
Me
r.t. to reflux, 6 h
2) crystallization at -20 °C
(81%)
(77%) 319
326 [145]
Scheme 8-63. Bridged indene, fluorene.
8.5 Bridged Ligands with Stereogenic Bridging Groups The first examples of bridged ligands contained achiral bridges at the 1-position of the indenyl ligands. The resulting π-faces of the ligands were rendered enantiotopic and metalation would necessarily lead to an equal mixture of d and l enantiomers and possibly also the meso diastereomer. By incorporating chirality in the bridging group at the 1-position of indene, metalation of the resulting diastereotopic faces could now lead to three diastereomers in differing amounts. A hope was by judicious selection of the bridging group the formation of one of the diastereomeric metallocenes would be energetically favored. Ligands of this type will be discussed first in this section. A second approach to introducing chiral bridges was its placement at the symmetrical 2-position of indene. Metalation would lead inherently to only one stereoisomeric metallocene. The historical development of ligands bridged by a stereogenic group started with an unsymmetrical one carbon bridge by Rausch and Chien. The bridging of bis(indenes) at the 1-position was initiated in 1991 in separate reports by Halterman and Green and has been the most developed type. Two examples of using chiral groups to bridge bis(cyclopentadienes) have been reported. A number of examples of biaryl bridging groups at the 2-position of indene and related ligands have appeared. Most recently one-atom chiral silylene bridges and chiral doubly bridged ligands have been introduced. One-Carbon Chiral Bridge. The first example of a chiral bridging group was Rausch and Chien’s 1990 report of the monocyclopentadienyl monoindenyl ligand dl-327 [160] (Scheme 8-64). The ligand was formed by the addition of indenyllithium to pentamethylpentafulvene. Since the chirality is generated only upon formation of the bridging group from achiral starting materials, the ligand is racemic. Alt has reported two similar syntheses of chiral cyclopentadienyl, fluorenyl ligands. In the first, substituted indenyllithium is added to 6-dimethylaminofulvene and the intermediate fulvene reacted with methyllithium to give the racemic compound 328 [158]. The addition of a fluorenyllithium complex to unsymmetrically substituted
507
8.5 Bridged Ligands with Stereogenic Bridging Groups
fulvenes 329 led to the racemic ligands 330 [156]. In a unique synthesis, methyllithium or an allyl Grignard reagent adds to one fulvene in bis(fulvene) 331 and the resulting cyclopentadienyllithium adds to the second fulvene to form (after deprotonation) bis(cyclopentadienyl) dianions 332 [161]. As in Rausch and Chien’s synthesis, the chirality is introduced only upon the formation of the final bridge and bis(cyclopentadienyls) 332 are therefore also racemic.
1) THF –
+
Li+
2) H 2O dl-327 [160]
H
NMe2
+
Li
1) THF
–
+ R2
R
1
R1
2) MeLi 3) aq. workup 328
R
Ph
Li+ –
Ph
R2 R
1) THF
[158]
Ph Ph
Ph 2) aq. workup
329
330 [156] Ph R
O Me
Me O
Me
2
Me
2 CH3Li, Et 2O Me or alllylMgCl; n-BuLi
H N 331
Me
2Li+
332a: R = Me [161a] 332b: R = allyl [161b]
Scheme 8-64. Racemic one-carbon chiral bridged ligands. Bis(indenes) Bridged at the 1-Position. The first reports of nonracemic chiral groups being used to bridge indenes were Halterman’s 1991 report using the 1,1′binaphthyl-2,2′-dimethyl bridge [162] and Green’s report using a tartrate-derived 1,4-butanediyl bridge [163] (Scheme 8-65). Unlike the previous 1-carbon bridges, these ligands were prepared in nonracemic form. The known 1,1′-binaphthyl-2,2′di(bromomethyl) 122 had been previously used to form a chiral cyclopentadiene. Dibromide 122 could be reacted with indenyllithium to form the bis(indene) in 71% yield. This method was later extended to include to bridged bis(4,7-dimethylindene) [164]. 4,7-Diisopropylindene could not be bridged by dibromide 122. The
508
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
tartrate-derived ditosylate 112 had also been previously used in a chiral cyclopentadiene synthesis. Green was able to react ditosylate with indenylmagnesium bromide to give bis(indene) 334 in 76% yield [163]. R R
R Br
THF +
[162, 163]
–
2
0 °C, 2 h
Li+
Br R
R = H (71%) R = CH 3 (69%)
OTs
O
R
toluene +
O
R 333
122
–
2
+
0 °C to r.t.
MgBr
OTs
O
O
[164]
(76%)
112
334
OMs + 2
[165]
DMF, -50 °C to reflux
– +
Na
OMs
(52%) 336
(±-335)
OMs + 2
Et 2O –
MsO
[115]
+ Li+
0 °C to r.t.
119b
(60%) 337
OMs
338
THF, r.t. to reflux
+ 2Mg
OMs
[32]
(32%) (S,S)-240
(R,R )-339 CO2Men
1) LiAlH 4, Et 2O
OMs OMs
CO2Men
2) MsCl, Et 3N, CH 2Cl2 (82%)
341
(Indenyl)2Mg
[166]
THF, r.t. to reflux (83%) 342
Scheme 8-65. Bis(indenes) containing chiral hydrocarbon bridges. The first 2-carbon chiral bridging group was Rieger’s use of racemic trans-cyclohexane-1,2-dimesylate 335. The displacement of the unactivated secondary mesylates required rather forcing conditions—indenylsodium in refluxing DMF–but produced trans-1,2-bis(indenyl)cyclohexane 336 in good yield [165]. The synthesis of diisopropylcyclohexane-1,4-dimesylate 119b had been developed for a chiral cyclopentadiene synthesis. Dimesylate 119b could also be applied to the preparation
8.5 Bridged Ligands with Stereogenic Bridging Groups
509
of bridged bis(indene) 337 [115]. Interestingly, the alkylation of cyclopentadiene with the larger tethers tended to give spiro-annelated cyclopentadienes, whereas the alkylation of indenes generally led to bis(indenyl) products. Presumably the needed proton transfer for the formation of the spiro dialkylated ligand is slower in the indenes than the cyclopentadienes. In the case of the cyclohexane-1,4-dimesylate 119b, the ratio of bis(indene) 337 to spiro-indene 338 was dependent of the conditions. The best conditions for forming the bis(indene) were to add indenyllithium to the dimesylate in ether which gave bis(indene) 337 in 60% yield. The use of HMPA/THF led completely to the spiro-annelated product 338. Dimesylates 339 and 341 had been previously applied to the synthesis of chiral 1,2-diphosphines and 1,4-diphosphines. Bosnich has extended their use to include the bridging of indenes. Displacement of the secondary mesylates in 339 with bis(indenyl)magnesium gave a low yield of the bis(indene) 240 [32], but the primary mesylates in 341 cleanly displaced to give the 1,4-bridged bis(indene) 342 [166]. Bis(Cyclopentadienes) Bridged by Stereogenic Group. Two nonracemic chiral bridging groups for bis(cyclopentadienes) have been reported (Scheme 8-66). The first was by Nantz who employed a double-Skattebøl reaction of tetraene 345 to give the 2,3-butanediyl-bridged bis(3-methylcyclopentadiene) 346 in good yield [31]. The needed tetraene was efficiently produced by the alkylation of sulfonate 344 followed by elimination. The starting 2,3-dimethyl-1,4-butanediol is available in nonracemic form. The 1,4-bridged bis(cyclopentadiene) 349 was prepared in nonracemic form by Helmchen starting from the resolved Diels–Alder adduct 347 [11]. In this case triflates were used as the leaving groups in 348 for a cyclopentadienylmagnesium bromide displacement to give 349.
OH
1) MsCl, Et 3N 2) PhSK
PhO2S
3) mCPBA
Br
1) Br3CH, KO-t -Bu [31]
2) MeLi
2) KO-t-Bu PhO2S
OH (dl-343)
HO2C
1a) n-BuLi 1b)
345
344
OSO2CF3
CO2H
H H
(R,R )-347
346
F3CO2SO H 1) LiAlH 4
H
CpMgBr
H H
2) Tf2O, DMAP (90%)
THF, -78 °C 15 min 348
(91%)
349
Scheme 8-66. Bis(cyclopentadienes bridged by chiral groups.
[11]
510
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Biaryl-Bridging Groups for Bis(2-indenes) and Related Ligands. Brintzinger was the first of three groups to report the use of biaryl groups to bridge ligands in such a manner as to render the ligand π-faces equivalent (Scheme 8-67). The racemic 2,2′-diiodo-1,1′-biphenyl 350 was lithiated and reacted with dimethylcyclopentenone 137 to give (after dehydration and deprotonation) the bis(cyclopentadienyl) ligand 19 [12]. The π-faces of the cyclopentadienyl groups are equivalent by rotation about the bond to the biphenyl group. Since related 6,6′-dimethylbiphenyls have been resolved, the currently racemic dimethyl ligand 19 could possibly be prepared in nonracemic form. The initial addition produced a monocyclopentenol side product which was carried forward for separation at the metallocene stage. OH R
I
R
I
350 (R = H, CH 3)
1) 2 n-BuLi, Et2O
1) 160 °C
R R
2) 2
O
HO
Br
MeI, DMF heat
O
3) H2O
[12]
(+ monocyclopentadienyl)
OH
1) sec -BuLi, -78 °C 2)
K+ 19
(+ monocyclopentenol)
Br
–
R
2) KH, decalin 351
137
3) H 2O
K+ R –
144
OH
352 Cp
also for:
[167]
354
353 Cp =
Cp
355
CO2Me
ClMg ClMg
CF3COOH 359
CO2Me
CH2Cl2 OH
358
360
CO2Me
R
CO2Me
p-TsOH
ClMg
R (52 to 69%))
CH2Cl2
R OH
361
R = H, CH
(42% from 358)
3
355
[168]
OH
ClMg
R
357
356
OH
R R
(86 to 94%)
362
363
Scheme 8-67. Preparation of biaryl-bridged ligands. A related report by Halterman using resolved binaphthyl 352 and cyclopentenone 144 enabled the preparation of nonracemic bis(tetrahydroindene) 354 [167]. Binaphthyl-bridged bis(indene) 355, bis(hydropentalene) 356 and bis(hydroazulene) 357 were similarly prepared [167b]. Bosnich reported the efficient use of di-Grig-
511
8.5 Bridged Ligands with Stereogenic Bridging Groups
nard 359 to convert biaryl diesters 358 and 361 to indenols 360 and 362 which could be dehydrated to give the binaphthyl-bridged bis(indene) 355 and the biphenyl-bridged ligand 363 [168]. Chiral Silicon-Bridged Ligands. Two reports have appeared in which chiral silyl groups have been used to bridge indenes or cyclopentadienes (Scheme 8-68). Two equivalents of menthol can be used to derivatize tetrachlorosilane to the dichlorodimentholsilane 364. Displacement of the two remaining chlorides with indenyllithium gave the chiral bridged bis(indene) 365 [169]. Binaphthol could react with the bis(cyclopentadienyl)silicon dichloride 366 to give the chiral bridged bis(cyclopentadiene) 367. This ligand could also be further silylated and deprotonated to form a single isomer of 368 [170].
–
Cl Si
SiCl4
2
O
OH
O
Li+ MenO
Cl
Si
OMen [169]
364
365 K+
K+ –
–
OH OH
+
Et 3N,THF
Cl Si Cl
Si
+
Li
O
2) LiN(SiMe3)2 THF, -78 °C to r.t.
+
Li
1) Me 3SiCl, THF -78 °C to r.t.
O 2) KO-t-Bu, THF -78 °C to r.t.
–
– Si Me3Si
O
O
SiMe3
366 367
368 [170]
Scheme 8-68. Chiral silyl-bridged ligands. Doubly Bridged Chiral Ligands. Two reports describe the use of Meerwein’s diketone (369) to form chiral doubly bridged ligands (Scheme 8-69). Addition of 2-propenyl Grignard to 369 followed by dehydration gave tetraene 370. This tetraene was cyclopropanated to enable a Skattebøl rearrangement to the doubly bridged bis(cyclopentadienyl) ligand 371 [171]. A different synthetic strategy was used to prepare bis(indene) 376. In this synthesis, dinitrile 372 was reduced, reacted with phenyl Grignard to give after oxidation diketone 374. Diketone 374 underwent a Nazarov type ring closure to give bis(indanone) 375 which could be reduced and dehydrated to the doubly bridged bis(indene) 376 [142].
512
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides Me
Me
1)
1) CHBr3, KO-t-Bu
MgBr / CeCl3
O
Me H
NC
O 1) PhMgBr 2) MnO 2
(93%) 372
O
Me
371
DIBAL-H
369
[171]
–
370
O
2Li+
2) MeLi
2) Tf2O 2,6- t-Bu24-Mepyridine 369
–
Me
O
CN
O
(68%)
O
373 H
O
374
PPA
O O
1) LiAlH 4 2) p-TsOH
(38%) 375
[142]
(62%)
376
Scheme 8-69. Chiral doubly bridged ligands. Chiral Bis(Fluorene) and Fluorene, Indene Ligands. Nonracemic styrene oxide has been used the preparation of nonracemic ligands containing fluorene and indene (Scheme 8-70). Ring-opening of styrene oxide by fluorenyllithium gave a mixture of regioisomer alcohols 377 and 378. Activating the purified isomer 378 as the triflate enabled the displacement by indenyllithium or fluorenyllithium to give the chiral ligands 379 and 380 [172]. 1) FluLi, i-Pr2O 0 °C, to r.t. 12 h
O Ph
2) aq. workup
Ph
Ph
+ OH
OH
Ph 377
1) (CF 3SO2)2O pyridine, CH 2Cl2
378
Ph
or
[172]
Ph
OH 2) IndLi or FluLi dioxane 378
379
380
Scheme 8-70. Chiral bridged bis(fluorene) and indene^fluorene.
513
8.5 Unbridged Metal Complexes
8.6 Unbridged Metal Complexes 8.6.1 Bis(cyclopentadienyl)metals Far too many group 4 metallocene complexes have been prepared to cover completely in this section which will concentrate on general methods and the stereoselectivity of metallocene synthesis. Good reviews of simple metallocenes [173, 1b] and sterically hindered metallocenes [1c] are available and the references for the preparation of cyclopentadienyl ligands in Sec. 8.3.1 generally include the formation of metallocene complexes. Achiral Bis(cyclopentadienyl)metal Complexes. Examples of common methods for the preparation of metallocene dihalides are shown in Scheme 8-71. The first [6] and most common method for preparing simple titanocene dihalides has been to react the cyclopentadienyl anion with titanium tetrahalide. Early examples of substituted complexes being prepared in this way are (MeCp)2TiCl2 [7], (iPrCp)2TiCl2 and (t-BuCp)2TiCl2 [174]. NMR characterization of (RCp)2TiCl2 R = H, Me, Et, n-Pr, n-Bu, 1,2-Me2, 1,3-Et2, and 1-Et-3-Me have appeared [175] in addition to more complete reviews [173]. Early examples of the analogous preparation of (RCp)2ZrCl2 complexes include R = Me, Et, i-Pr, t-Bu, TMS [176]. The early use of (tetraamido)titanium or zirconium complexes to form metallocene complexes was described in 1968 [177]. Irradiation of solutions of two difference bis(cyclopentadienyl)titanium dichloride complexes was shown to interconvert the cyclopentadienyl ligands [178]. An early report of using titanium trichloride to form titanocene dichlorides was Brintzinger’s addition of pentamethylcyclopentadienylsodium to TiCl3, followed by oxidation using conc. HCl to form bis(pentamethylcyclopentadienyl)titanium dichloride in 40% yield [179]. The use of TiCl4 in this reaction was less satisfactory. The benefits of using TiCl3 presumably arise from improved steric interactions and the elimination of the oxidation of Me5CpNa by the more oxidizing TiCl4. Improved yields using TiCl3 instead of TiCl4 are now fairly commonly observed. Me
D
Me MiCl4 –
M
Cl
Cl
Ti
Cl
Cl
M = Ti, Zr
+
D D DD D
Me Me M(NMe2)4 M = Ti, Zr
M
NMe 2
Me
–
Me
NMe 2 Me
Me
Ti
D D Cl
D 313 nm
Cl
2D D
Ti
D D 1) TiCl3, THF -30 °C to r.t. 2) Conc. HCl CHCl 3, -20 °C
Me
D D Cl Cl
Me Me Cl
Me Ti M eM e M e Cl Me Me Me
Scheme 8-71. Examples of bis(cyclopentadienyl)metal complex formation.
514
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Achiral (Cp)(Cp′)titanium Complexes. Through the controlled use of stoichiometry [180] or insertion into the trimethylsilyl–cyclopentadienyl bond [27c], (RCp)TiCl3 complexes can be prepared and reacted with a second cyclopentadienyl anion to form mixed bis(cyclopentadienyl)titanium dichloride complexes. Some examples are shown in Scheme 8-72. The addition of substituted cyclopentadienyl anions to CpTiCl3 generates (RCp)(Cp)TiCl2 381 [173, 181] whereas the (Me5Cp)(Cp)TiCl2 complex was generated by the reverse sequence of adding CpLi to Cp*TiCl3 [181]. Me
Me CpTiCl3 –
Li+
THF
Me
Ti
Cl Cl
Me 381
Me Me Me Cl
Me Me Ti
Cl
Me CpLi, THF
Me Me
Ti
Me Me Cl Cl
Cl
Scheme 8-72. Mixed bis(cyclopentadienyl)titanium dichlorides [181]. Chiral Bis(cyclopentadienyl)metal Complexes having Equivalent π-Faces. The introduction of a single chiral substituent on the cyclopentadienyl ligand leaves the faces of the ligand still equivalent. If the ligand is enantiomerically pure as in menthylcyclopentadiene 76, only one isomeric (R*Cp)2metal complex 382 can form [2]. Several chiral monosubsituted (R*Cp)2MCl2 complexes have been prepared containing the natural-product derived auxiliaries: neomenthyl 383 [3], 8-phenylmenthyl 384 [5b], carboxylic acid derived 385 [49], an estrone derivative 386 [47], cholestanol 387 [48] and cholestanone 388 [48] (Scheme 8-73). If both enantiomers of the ligand are present as in (1-phenylethyl)cyclopentadienyl 12, then a mixture of dl and meso complexes 14 can form. Although the ratio of these stereoisomers in this particular case is 1:1 for both the titanium and zirconium complexes, selectivity for one isomer could have been possible [9, 182, 183]. The related (1-cyclohexylethyl)cyclopentadienyl ligand has been incorporated into an equimolar mixture of dl and meso isomers of zirconium complex 389 [183]. The dl isomers of both complexes 14 and 389 could be obtained in pure form through repeated recrystallizations of the isomeric mixture [182, 183].
515
8.5 Unbridged Metal Complexes
MCl 4
H
–
M
MCl 2
H
TiCl2
H
2
Cl
H
76
R
H Cl
ZrCl 2
2
2 R = Ph, Me
Ph
382 M = Ti, Zr [3]
383 M = Ti, Zr [3]
385 [49]
384 [5b]
CH 3
H 3C
ZrCl 2
H
2
H
2
ZrCl 2
ZrCl 2
2
H
386 [47]
CH 3O
R
R TiCl4 –
388 [48]
387 [48]
M or ZrCl4
Cl
R
R +
Cl
M
Cl
+
M
Cl
[9,182,183]
Cl
Cl
12
R Racemic
R = Ph: meso -14 (M = Ti, Zr) R = c-C6H11: meso -389 (M = Zr)
R
R
R = Ph: dl-14 (M = Ti, Zr) R = c-C6H11: dl-389 (M = Zr)
Scheme 8-73. Chiral bis(cyclopentadienyl)metal dichlorides. The metalation of one enantiomer of C2-symmetric cyclopentadienes can also inherently lead only to the formation of a single stereoisomeric bis(cyclopentadienyl)metal complex (Scheme 8-74). Examples of in the bis(cyclopentadienyl)metal complexes containing disubstituted, annelated cyclopentadienyl ligands are the bicyclooctane-annelated complexes 390 [73], 391 and 32 [19] and the binaphthyldimethylene-annelated complexes 392 [74]. C2-symmetric tetrasubstituted, diannelated cyclopentadienes have been used to form monocyclocyclopentadienylmetal trichlorides 393 [87, 88] and the mixed metallocene complex 394 [88]. Apparently steric hindrance prevents two equivalents of the tetrasubstituted cyclopentadiene from being incorporated into a metallocene dichloride.
516
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Scheme 8-74. Metal complexes of C2-symmetric cyclopentadienes.
TiCl4
Cl Ti
–
Cl
+
Cl
dl--37
34
Cl
Zr
OMe
Cl
Cl
OMe
+
[21]
Cl
(p- S,p- S)-37
(p- R, p- S)-37
MeO
Zr
Cl
MeO
(p- S,p- S)-395
Ti
meso -37
(p- R, p- R)-37 MeO
Cl
Cl Ti
(p- R, p- R)-395
Cl
Zr
Cl
MeO (p- R, p- S)-395
O
Zr
O
[58]
(P,p- S,p- S)-396 (+ enantiomer)
Scheme 8-75. Metalation of cyclopentadiene having enantiotopic faces.
Bis(cyclopentadienyl)metal Complexes having Enantiotopic Ligands. The preparation of metallocene complexes from cyclopentadienes having enantiotopic ligands is relatively undeveloped, perhaps due to the formation of stereoisomeric mixtures of the complexes (Scheme 8-75). An example is the metalation of the 1isopropyl-3-methylcyclopentadienyl ligand 34 which leads to a mixture of dl and meso isomers of 37 [21]. Bergman has used chelating enantiotopic ligands to form a 3:1 mixture of dl and meso isomers of 395 and in a completely stereoselective metalation, the dl isomer 396 [58].
517
8.5 Unbridged Metal Complexes
Bis(cyclopentadienyl)metal Complexes having Diastereotopic Faces. The metalation of cyclopentadienes containing diastereotopic faces can still lead to isomeric mixtures, but in many cases sufficient differentiation exists between the faces to enable selective metalations (Scheme 8-76). In the first example, the metalation of camphor-derived cyclopentadiene 128 was highly selective for one C2-symmetric isomer 48, with the C1-symmetric isomer forming in trace amounts [5, 25, 26]. Metalation of the trisubstituted camphor-based cyclopentadiene 398 was also selective for the C2-symmetric isomers 399 and 400 [85]. Metalations of verbenone-derived cyclopentadiene 130 [75] and pinanyl-cyclopentadiene 134 [72] were selective for a single isomer of 401 and 402, respectively. The racemic phenylpentalene 403 was also selective for the single metallocene dichloride 404 [184]. 1) n-BuLi 2) TiCl3
1) n-BuLi 2) TiCl3 M
3) air, HCl 128
Cl
Cl +
M
[( 4R, 7S,p- R)(4R, 7S,p- S)]- 48
[(4 R, 7S,p- R)(4 R, 7S,p- R)]- 48
401 M = Ti, Zr [75]
1) n-BuLi 2) TiCl3 M
3) air, HCl 1) n-BuLi 2) TiCl3
Me
134 M
R 398
3) air, HCl or 2) ZrCl4
Cl
Cl
or 2) ZrCl4
130
or 2) ZrCl4 [5,25,26]
M
3) air, HCl
Cl
Cl
or 2) ZrCl4
Cl
Cl
Cl
402 M = Ti, Zr [72]
Cl Me
1) TiCl3 MgX
Ti Ph
Ph 399: M = Ti, R = Me 400: M = Zr, R = Me, t-Bu, Ph
Ph
2) HCl, air Cl
rac-403
Cl
404 [184]
Scheme 8-76. Chiral bis(cyclopentadienyl)metal dichlorides.
A remarkable ability to control the facial selectivity in the metalation of norbornenylcyclopentadienyl ligand 49 has been demonstrated (Scheme 8-77) [27c]. Reaction of 49 with titanium trichloride at low temperature leads to the endo,endo isomer 405 whereas the reaction at room temperature leads to the exo,exo isomer 406. The mixed titanocenes having the endo or exo configuration 407 or 408 could likewise be prepared through the reaction of 49 with CpTiCl3.
518
8
1) TiCl3-3THF -64 °C, 5 h
– 49
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Li+
Cl
–
Ti
2) aq. HCl
Cl
Li+
49
1) TiCl3-3THF 25 °C, 14 h 2) aq. HCl
CpTiCl 3 -78 °C
49
Cl
exo,exo -406
endo,endo-405
–
Cl Ti
Cl Ti
–
Cl
Li+
49
Li+
CpTiCl3
Cl Ti
>25 °C
endo-407
Cl
exo -408
Scheme 8-77. Control of stereochemistry in isodicyclopentadienyltitanium complexes [27c].
8.6.2 Bis(indenes) Achiral Complexes. The preparation of bis(η5-indenyl)zirconium and titanium dichlorides is analogous to the preparation of cyclopentadienylmetal complexes. Two equivalents of the indenyl salt (originally the sodium salt [13a]) react with the metal tetrachloride to provide the metal complexes 409 (Scheme 8-78), whose solid state structures were obtained [13b]. A number of substituted bis(indenyl)metal complexes have been prepared from symmetrically substituted indenes. The lithium salt of several 2-arylindenes have been metalated by Waymouth to give complexes 410 in good yields [14]. 5,6-Disubstituted and 4,7-disubstituted indenes have been used to generate electronically varied bis(indenyl)zirconium dichloride complexes 411 and 412 [18]. These metalations were carried out using ZrCl4 and the lithium salt of the indenes to give 30 to 65% yields of the metallocene dichlorides. Bis(2-aryltetrahydroindenyl)metal dichlorides 413 have been prepared directly by the metalation of the tetrahydroindene ligand 145 [83, 185]. The permethylated indene 198 has been converted into the very air sensitive bis(indenyl)zirconium dichloride 414 [186, 187].
519
8.5 Unbridged Metal Complexes
TiCl4, THF
–
2
M
or ZrCl4, THF
Na+
Cl
2
–
ZrCl4
Ar +
Cl
toluene
Li
Ar Ar
(Ar = Ph, 4- t-BuPh, 3,5-Bis(trifluoromethyl)Ph)
M = Ti, Zr 409 [13]
Cl
Zr
410
R
R
R
R
R ZrCl4, THF
–
2
+ R Li
R Zr Cl Cl R
(R = Me, MeO, F)
–
2
R
ZrCl4, THF
Zr
2
Ar 145
2) TiCl3 3) [O]
412
M
[18]
(R = Me, MeO, Cl)
Me Ar Ar
Cl Cl
Li+
411 R 1) n-BuLi 2) ZrCl4 or
[14]
Cl
Cl Cl
Me Me
(Ar = Ph, 4- MePh, toluene 4-MeOPh, 4-BrPh)
Me
Me
Me
R Me
Me
R
Me
Me
Me Cl 1) n-BuLi Me Zr Me Me Me Cl 2) ZrCl4 Me Me R
Me 413 M = Ti, Zr [83,185]
414 [186]
R
Scheme 8-78. Preparation of achiral bis(indenyl)metal dichloride complexes. Chiral Complexes. A number of chiral unbridged bis(indenyl)zirconium dichloride complexes have been reported by Erker. The metalation of 1-cyclohexylindenyllithium with ZrCl4 gave a 1:1 mixture of rac- and meso-bis(indenyl)zirconium dichloride complexes (Scheme 8-79). Catalytic hydrogenation of these complexes yielded the bis(tetrahydroindenyl)zirconium dichloride complexes. Both the indenyl and tetrahydroindenyl complexes were separable by fractional crystallization [24]. Indenes bearing the chiral substituents: neoisopinocamphyl, menthyl, neomenthyl, isomenthyl or neoisomenthyl have been metalated to give the bis(indenyl)zirconium dichloride complexes as varying mixtures of diastereomers [29, 112]. In none of these cases was the stereoselectivity between the two ‘racemic-like’ and the ‘meso-like’ diastereomers very high. The metalation of 4,7-dimethyl-1-neomenthylindene, however, did lead to a very high stereoselectivity for the formation of one ‘racemic-like’ bis(indenyl)zirconium complex a 98:1:1 ratio [10]. The chiral bis(indenyl)titanium dichloride complex 25 which contains equivalent ligand faces has been prepared from 2-menthylindene 415 [15].
520
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
R 2
–
ZrCl4(thf)2
Li+ R
R
Zr
Cl Cl
R
+
Cl
Zr
Cl
R
+
Cl Zr R Cl
R
(p- R, p -S)
(p- S,p- S)
(p- R,p -R)
meso - or meso -like
dl- or dl-like
R=
neoisopinocamphyl neomenthyl
[24]
isomenthyl
neoisomenthyl
[29]
[112]
dl/meso
menthyl
(p- R,p -R) (or (p- S,p- S)) : (p- S,p- S) (or (p- R,p -R)) : (p- R, p -S)
50:50
52:<1:48
70:6:24
53:11:31
28:6:66
1) n-BuLi
2
H
Zr
2) ZrCl4
62:11:26
Cl Cl
H 415
25 [15]
Scheme 8-79. Preparation of chiral unbridged bis(indenyl)zirconium dichlorides.
8.6.3 Bis(fluorenes) Unbridged bis(fluorenyl)zirconium dichlorides can be prepared by the reaction of fluorenyl anions with zirconium tetrachloride (Scheme 8-80). The first example, the unsubstituted bis(fluorenyl)zirconium dichloride 409 was reported by Setton and Samuel in 1965 [13]. The methyl-substituted complex 410 was isolated by crystallization as only the dl isomer [188]. Alt has prepared a number of addition unbridged bis(fluorenyl)zirconium dichlorides such as the 9-substituted complexes 411 and 412 and the isomeric mixture of 413 [189]. Fluorenyl complexes are often difficult to purify and characterize since they tend to be much less soluble than corresponding cyclopentadienyl or indenyl complexes.
521
8.5 Unbridged Metal Complexes Me ZrCl4
ZrCl4 2
–
dimethoxyethane
Cl
Zr
Cl
2
–
M e Cl
Zr
pentane
Cl M e
Li+
Na+ 409 [13]
R Cl R
Zr
R Cl R
Cl
Zr
R Cl R
Cl
Me
Zr
Cl
[189]
Me meso -413: R = Ph
dl-413: R = Ph
411: R = C 6H11 412: R = Ph
410 [188]
Me
Me
Scheme 8-80. Bis(fluorenyl)zirconium dichlorides.
8.6.4 Mixed Ligands Unbridged metallocene dichloride complexes containing mixed ring ligands are fairly undeveloped. Methods for the preparation of some of the few examples are given in this section. Cyclopentadienyl, Indenyl. The mixed metallocene dichloride 414 has been prepared either by addition of indenyllithium to CpTiCl3 or CpZrCl3 [181,190] or by the addition of cyclopentadienyllithium to indenyltitanium trichloride 415 [191] (Scheme 8-81). Indenyltitanium trichloride 415 has been selectively formed from TiCl4 and (indenyl)tributyltin. Complex 414 has also been prepared earlier using a thallium complex [192].
CpMCl 3
– +
Li
Et 2O
M
M = Ti, Zr [181,191]
Cl Cl
Et 2O, 0 °C [192]
414
TiCl4
CpLi
Ti Cl
Cl
CH2Cl2, 0 °C
Cl
SnBu3
415
Scheme 8-81. Unbridged (cyclopentadienyl)(indenyl)metal dichlorides. Cyclopentadienyl, Fluorenyl. Mixed complexes having cyclopentadienyl and fluorenyl ligands have been prepared by the addition of substituted fluorenyllithium
522
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
complexes to either CpZrCl3 or (Me5Cp)ZrCl3 to give complexes 416 [193] and 417 [194] (Scheme 8-82). R'
R
R Cl
CpZrCl3 –
Zr
Cl
R Cl Me Me
(Me5Cp)ZrCl3
–
R
Et 2O
R'
R'
Et 2O
Li+
Li+
Me
R'
416 [193]
Zr
Cl Me
Me 417 [194]
Scheme 8-82. Unbridged (cyclopentadienyl)(fluorenyl)metal dichlorides.
8.7 Metal Complexes with Nonstereogenic Bridging Groups 8.7.1 Bis(cyclopentadienyl)metals Achiral Complexes. A good number of ansa-metallocene dichloride complexes which contain a nonstereogenic bridging group between two achiral cyclopentadienyl ligands have been prepared. With no stereochemical issues to overcome, the major synthetic hurdle is the differentiation between the formation of ansa-metallocenes versus oligomeric complexes. Examples shown in Scheme 8-83 are: one carbon bridged complexes 418 [116] and 419 [119, 195]; two carbon bridged 420 [117], 421 [10], 422 [124] and 423 [123]; propano-bridged complex 424 [118]; cylopentane-1,3-diyl-bridged 425 [122]; silyl-bridged complexes 426 [120] and 427 [121] and the doubly bridged complex 428 [126,127]. Me
Me
Ti
Cl
M
Ti
Cl
Cl 418 [116]
Cl
Cl M Cl
M
Ti Cl
Cl Cl
Me
Si
M
Me Me Si M
Cl
Me
H 424 [118]
Cl
426 [120]
Ti
Cl Cl
Si Me
425 [122]
Me R Cl
Cl R Me Me Me R = Me, Ph 422 [124] 423 [123]
421 [10]
H Cl
Me M e M
Cl
420 [117]
419 [119,195]
Cl
Cl Cl
Cl M Cl
Me
427 [121]
428 [126]
Scheme 8-83. Achiral bridged bis(cyclopentadienyl)metal complexes (M = Ti, Zr).
8.7
523
Metal Complexes with Non-Stereogenic Bridging Groups
Chiral Complexes from Prochiral Ligands. Brintzinger reported the first chiral ansa-metallocene dichloride, dl-propanobis(3-t-butylcyclopentadienyl)titanium dichloride (429), in 1979 [196] (Scheme 8-84). The prochiral ligand 430 was metalated and the crude product purified by silica gel HPLC to give the dl pair 429. The (1S,1′S) isomer could be resolved by reacting 429 with the sodium salt of (S)binaphthol and purifying the ensuing mixture by HPLC. The single enantiomeric adduct was characterized by X-ray crystallography.
1) TiCl3 –
–
430
Ti 2) 6 N HCl purification by SiO 2 HPLC
Cl Cl
+
Ti
Cl Cl
dl-429
(1S,1' S)- selectively resolved with S-binaphthol
Scheme 8-84. Propanobis(3-t-butylcyclopentadienyl)titanium dichloride. Several other examples of ansa-bis(3-substituted cyclopentadienyl)metal dichlorides have been subsequently prepared having a dimethylsilylene-, an ethylene-, a tetramethylethylene- or a cyclopentane-1,3-diyl-bridging group. In most cases a mixture of three stereoisomers was formed—the dl pair and the meso complex. These metalations are summarized in Scheme 8-85. Only one of the dl complexes is depicted along with the ratios observed for the metalation. The dl/meso selectivity in the silyl-bridged zirconium complexes 431 was uniformly 1:1 [128]. Several other related silyl-bridged complexes have also been reported [129]. The dl isomers of the tetramethylethylene-bridged complexes 432 (M = Ti) and 433 (M = Zr) were moderately preferred over the meso isomers. In most of these cases the isomers could be separated. An additional study of the bis(t-butylcyclopentadienyl)titanium complex has been published which includes the use of Cp2TiCl2 as the metal source through ligand exchange [130]. Ethylene-bridged complexes 434 (M = Ti) were selective for the meso isomer [62]. The isopropyl- and t-butylsubstituted complexes could be separated by fractional crystallization. The analogous zirconium complexes 435 were moderately selective for the dl pair [197]. The metalation of cyclopentane-1,3-diyl-bridged bis(cyclopentadienes) 250 could give four isomers of 436, the dl pair and two meso isomers having the substituents either both over the methylene group or the ethylene group of the cyclopentane. In each case, some selectivity was observed for the dl pair and both the dl pair and the first of the meso isomers could be purified by crystallization [122].
524
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
R
R
R 1) KH, THF, -30 °C to r.t.
H3C Si
H3C Si
Zr
2) ZrCl4, THF, r.t. H3C
H3C
(CH3)3C–
1:1
Cl
(CH3)3Si–
1:1
Cl
Ph(CH3)2C–
1:1
Ph(CH2)5C–
1:1
[128]
R
R
431
246 R
(CH3)3C–
1.5-2:1
Cl
(CH3)3Si–
0.6-1.8:1 (CH3)3Si–
1.5-2:1
Cl
(CH3)2CH–
1:1
PhCH2–
1:1
R
Ph(CH3)2C–
1.2-2.2:1
432 M = Ti 433 M = Zr
Ph(CH2)5C–
1.2-2.2:1
M
2) 6 N HCl or 1) ZrCl4 [61]
247
R
R
R
R 1) n-BuLi, -78 °C to r.t. THF 2) TiCl3-3THF, THF -40 °C to reflux 3) 6 N HCl, air R 245
434 [62] dl / meso
M
434 M = Ti R 435 M = Zr H
1: 1.3
Me
1-2: 1
Et
1 : 1.4
i-Pr
1-2: 1
i-Pr
1: 1.8
t-Bu
1-2: 1
t-Bu
1 : 2..0
1) n-BuLi
R
H
R
H
Cl Ti
2) TiCl3-3THF, THF 3) 6 N HCl, air
H
R 250
[122]
Cl
R
dl-436
i-Pr
8
t-Bu
5
Cl Ti
Cl
Cl
R
H
H
meso(1) -436 : :
R
Cl
Ti R
H
435 [197] dl / meso
R
Me Cl Cl
R
H
dl / meso
2.5:1
+
–
433
R
dl / meso
(CH3)3C– 1) TiCl3, THF -78 °C to r.t.
(MgCl )2
432
R
R
–
dl / meso
3 5
: :
R meso(2) -436 1 1
Scheme 8-85. ansa-bis(3-RCp)MCl2 complexes. Several examples of complexes having an additional substituent adjacent to the bridging group are summarized in Scheme 8-86. The ethylene-bridged titanium complexes 437 were initially only moderately selective for the dl pair, but the product mixture could be photochemically equilibrated to give a significant enhancement of the dl isomers [131]. The silyl-bridged zirconium complexes 438 were selective for the dl pair [128]. Several additional examples have been prepared and enhanced in the proportion of the dl isomers by crystallizations [129]. Another trisubstituted bis(cyclopentadienyl) system is the 1,2-phenylene-bridged
8.7
525
Metal Complexes with Non-Stereogenic Bridging Groups
bis(tetrahydroindenyl)metal dichloride complexes 439 and 440. These tetrahydroindenyl complexes were formed directly from the hydroindene ligand without the usual need for hydrogenation of an initial bis(indenyl)metal complex. Both of these complexes were somewhat selective for the dl isomer and the zirconium complex could be photochemically enhanced in the dl isomer [132].
R
1) n-BuLi, -78 °C to r.t. THF 2) TiCl3-3THF, THF -40 °C to reflux 3) 6 N HCl, air filtration through R1 BioBeads SX-1 with toluene
R 254
dl/meso
R1
R1
R
Ti R
R 437 [131]
initial
after h ν
3.0:1
>10:1
R 1 = Me
2.0 : 1
2.5 : 1
R 1 = i-Pr
2.6 : 1
4:1
R 1 = t-Bu
1.6 : 1
>15 : 1
Cl
R = i-Pr R 1 = t-Bu
Cl
R = Me
1
R R R H3C Si
Me
Si
Me Zr
R
Cl
Cl Me
dl / meso
(CH3)3C–
2:1
(CH3)2CH–
6:1
[128,129] 438
246 1) n-Buli 2) TiCl3, THF -78 °C to reflux 3) air, CHCl 3 or 2) ZrCl4, Et 2O, r.t. 262
H3C
2) ZrCl4, THF, r.t. H3C
H3C Me
1) KH, THF, -30 °C to r.t.
R
Cl M Cl
439 M = Ti ( dl/meso = 4:1) 440 M = Zr ( dl/meso = 60:40) after h ν (dl/meso =93:7)
[132]
Scheme 8-86. ansa-Metallocene dichlorides of trisubstituted cyclopentadienes. Two related complexes 441 and 442 having prochiral cyclopentadienyl ligands bridged by two silylene groups have been prepared by Brintzinger [133] (Scheme 8-87).
526
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
R R
R
R 1) n-BuLi, hexanes
Me2Si
SiMe2 R R
Me2Si Cl R
2) isolate salt 3) ZrCl4, toluene r.t. 6 h
SiMe2 Cl
R
[133]
264
Zr
441 R, R = Me, Me 442 R + R = –CH 2CH2CH2CH2–
Scheme 8-87. Doubly bridged bis(cyclopentadienyl)zirconium dichlorides. Bridged Complexes Containing Chiral Cyclopentadienyl Ligands. Marks has reported the metalation of the neomenthyl- and menthyl-containing bis(cyclopentadienyls) 266 and 267 (Scheme 8-88).
–
K+
K+
Me Si Me Me
ZrCl4
Si
–
[30]
266
–
Li+
Li+
ZrCl4
Si
–
[30]
267
Me Si Me Me
Zr
H Cl
Me + Me
Cl Me Me Me
Zr
Zr
Me 443
H Cl
Me +
Cl Me Me Me
Cl
H Cl Me Me Me
Si
Cl H Cl Me Me Me
Si Me Me
Zr
444
Cl 1) n-Buli 268 +
2) TiCl3, THF -78 °C to reflux 3) aq.HCl, CHCl 3
Ti Cl
+
Cl
Ti
33 [20] 269
Scheme 8-88. Metalation of chiral bis(cyclopentadienyl) ligands.
445
Cl
8.7
527
Metal Complexes with Non-Stereogenic Bridging Groups
These ligands contain one ligand with diastereotopic faces. Both diastereomeric complexes of 443 and 444 were formed and separated by fractional crystallization [30]. Halterman has reported the metalation of a mixture of ethanobis(BCOCpH) ligands 268 and 269. Bis(cyclopentadiene) 268 is bridged along its axis of symmetry and the metalation of its homotopic faces could only lead to the single isomer 33. Bis(cyclopentadiene) 269 contains diastereotopic faces and while three stereoisomeric complexes could have been formed, only C2-symmetric complex 445 was observed [20].
8.7.2 Bis(indenes) The first example of an ansa-bis(indenyl)metal complex was Brintzinger’s 1982 report on an ethylene-bridged titanium complex (Scheme 8-89). The initial bis(indenyl)titanium dichloride complexes 446 were formed in dl to meso ratios of 2–10:1, depending on reaction conditions. These isomers could only be
1) n-BuLi
Cl
M
Cl Cl
M
H2
Cl
M
Cl
Cl Cl
PtO2 dl-446 M =Ti
275
M
Cl
2) TiCl4 or 2) ZrCl4
meso -446 M = Ti
dl-448 M =Ti
meso -448 M = Ti [23]
dl-447 M =Zr
dl-3 M =Zr [4]
(Ti: dl/meso ratio 2-10:1, partially separable) (Zr: only dl-247 obtained)
(Ti: meso -246 converts under irradiation into dl--448, completely separable by chromatography)
-
O O (S)-
Ti
dl-448 M =Ti
O Me
O
ONa
HCl
Ti
O
O
O
[23]
(S,S)-448
Me O
O
HCl
Zr
Zr
Ph Ph O
Et 2O
Cl
[198]
Cl
O O
(S,S)-450 + ( R,R )-450 (separable by crystallization)
Cl Cl
(S,S)-449 + uncomplexed ( R,R )-448 (separable by chromatography)
(R) dl-3 M =Zr
O O
O
Me
(S,S)-3 [and ( R,R )-3]
Scheme 8-89. Ethanobis(indenyl)metal dichlorides.
528
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
partially separated by chromatography. Hydrogenation of the 6-membered rings was possible and the resulting bis(tetrahydroindenyl)titanium dichloride complexes 448 could be separated. It was also found that irradiation converted the meso isomer into the dl pair [23]. An analogous reaction with zirconium gave only the dlbis(indenyl)zirconium complex 447 which could also be hydrogenated to its tetrahydroindeny derivative 3 [4, 198]. The addition of hydrogen was shown to occur on the ligand face opposite the metal [199]. A number of synthetic modifications of the conditions for the formation of these complexes have appeared [136, 137]. The use of tetraamidozirconium and the bis(indene) 275 allows a facile preparation of dl-447 under equilibrating conditions [200]. The dl isomers of both metals can be resolved. With titanium complex 448, only the (S,S) enantiomer coordinates with (S)-binaphtholate to give complex 449 which can be purified by chromatography [23]. The zirconium complex 3 could be resolved via the mandelate complexes 450. In this case both enantiomers form complexes, but these are separable by crystallization [201]. The dichlorides can be recovered by treatment of the alkoxide complexes with HCl. A number of related bis(indenyl)metal dichlorides containing unsubstituted indenyl ligands have appeared and are shown in Scheme 8-90. The propano complex 251 and the cyclopentane complex 453 were selective for the dl isomer [122, 201]. The dodecyl-tethered complex 452 was selective for the dl isomer when the metalation was carried out in THF, but for the meso isomer when carried out in toluene [138]. H Ti
Cl
( ) 10
Zr
Cl
Zr Cl Cl
( ) 10
Cl
Cl
Ti
Cl Cl
H
(only isomer)
(only isomer in THF) Me
Me Me
Si
Zr
Cl
Me Si
Cl
Me Si Me
454 [139,140]
meso -452 [138]
dl--452 [138]
dl--451 [201]
Zr
Cl Cl
455 [139]
(major isomer in toluene)
Me Me Si Zr Me Si Me
Cl Cl
456 [140]
dl--453 [122] (only isomer)
Zr
Cl Cl
Zr
Fe
Cl Cl
457 [134]
458 [141]
(dl major)
(dl only)
Scheme 8-90. ansa-bis(Indenes) containing unsubstituted indenyl ligands. Silyl-containing bridges are seen in complexes 454 [139, 140], 455 [139] and 456 [140]. The 1,2-phenylene-bridged complex 457 was selective for the dl isomer [134] while only the dl pair was observed with the ferrocene-bridged complex 458 [141].
8.7
529
Metal Complexes with Non-Stereogenic Bridging Groups
More highly substituted bis(indenyl)zirconium dichloride complexes have also been prepared (Scheme 8-91). Included in this group are: ethylene-bridged complexes 459 and 460 [18b, 110a]; 3-substituted silyl-bridged complexes 461 [140]; substituted silyl-bridged complexes 462 [106, 107, 110a] and the benzannelated complexes 463 [106, 111]. R R
Me
R4
Me Zr
R3
Cl
R 2R 2Zr
Cl Me R
Cl
Me
Cl
Me
Si
Cl
Zr
Cl R3
R7 Me Me R7
Si
R 2R 2 Zr
Me R
460 [18b,110a]
459 [18b]
Cl
Me
Cl
Me
R 2R 2 Zr
Si
Cl Cl
R4 461 [140]
462 [107,106,110a]
463 [106,111]
Scheme 8-91. Substituted bis(indenyl)zirconium dichloride complexes. Several bis(indenyl)metal complexes having bridging groups other than at the 1-indenyl position have been prepared and are shown in Scheme 8-92. Only a single metal complex 464 could form in the metalation of C2-symmetric bis(2indenyl)ethane 295 [16] and in the metalation of the doubly bridged bis(indene) 300 to 465 and 466 [142]. Two sets of complexes having the indenes tethered through the 7-positions have been prepared. Both the titanium and zirconium complexes 467 and 468 formed as a mixture of stereoisomers in the metalation of 303, but only the titanium complex 470 could be isolated in the metalation of bis(indene) 469 [143].
1. n-BuLi 2. TiCl3 3. [O]
1. n-BuLi Cl Ri Cl
or 1. Zr(NMe2)4 2. HNMe2-HCl
295 300
464 [16]
R
R
R
1. n-BuLi 2. TiCl3 3. [O]
303 R = Me, i-Pr
Zr
Cl Cl
M
Cl
465 M = Ti [142] 466 M = Zr Me
Me R
or 1. Zr(NMe2)4 2. HNMe2-HCl
Cl
2. TiCl3 3. [O]
[143] R
R
467 M = Ti ( dl and meso ) 468 M = Zr ( dl and meso )
1. n-BuLi
Ti
2. TiCl3 3. [O]
Cl Cl
Me
Me 470 M = Ti [143]
469
Scheme 8-92. ansa-bis(Indenyl)metal dichlorides bridged through 2- or 7-position.
530
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
8.7.3 Bis(fluorenes) A limited number of ansa-bis(fluorenyl)zirconium dichloride complexes have been prepared (Scheme 8-93). An ethylene-bridge and a dimethylsilylene-bridge are found in the unsubstituted achiral bis(fluorenyl)zirconium complexes 471 [144] and 472 [146]. Both the meso and dl forms of ethylene-bridged complex 473 were formed [148]. One example of a complex containing two different fluorenyl ligands is 474 [147].
Cl
Zr
Cl
Me Si Me
471 [144]
Zr
Zr
Cl Cl
472 [146]
Cl Cl
Me Si Me
Zr
Cl Cl
474 [147]
473 [148]
Scheme 8-93. ansa-bis(Fluorenyl)zirconium dichlorides.
8.7.4 Mixed Ligands Several examples of ansa-metallocenes containing one indenyl and one cyclopentadienyl moiety bridged by nonstereogenic bridging groups are given in Scheme 894. Green was the first to report the isopropylidene and cyclohexylidene-bridged zirconium complexes 475 [149]. Subsequently the titanium complex 475 was reported [150] as well as the diphenyl-substituted zirconium complex 475 [151]. The parent silyl-bridged complex 476 is also known [140]. Two sets of substituted cyclopentadienyl, indenyl complexes have been prepared having either an isopropylidene bridge in 477 or a silyl bridge in 478 [152]. R R R
M Cl Cl
Me Si Me
M Cl Cl
Me
M
Cl Cl
Me
t-Bu 475 [149-151]
476 [140]
477 [152]
t-Bu Me
Si
M
Cl Cl
Me
t-Bu 478 [152]
Scheme 8-94. ansa-Metallocenes containing indenyl with cyclopentadienyl ligands. Several mixed metallocenes containing a fluorenyl ligand tethered through a nonstereogenic group to a cyclopentadienyl ligand have been prepared (Scheme 8-95).
8.8
531
Metal Complexes with Stereogenic Bridging Groups
The first example was the isopropylidene-bridged complex 479 [153–156]. Additional examples are the silyl-bridged complex 480 [145], the di-(t-butyl) derivative 483 [147] and the unsymmetrically substituted complexes 481 [159, 110a] and 482 [156–158]. An interesting member of this last group is the methylene bridged complex 482 (R = H) [158]. The final example of ligands bridged by nonstereogenic groups is the fluorenyl, indenyl complex 484 [145]. R1
R R
Zr
Cl Cl
Me M e Si Zr
Cl Cl
Me Me
R = Me, t-Bu 479 [153-156]
480 [145]
Me M e Si Zr
483 [147]
Cl Cl
Zr
R R
Zr
Cl Cl
R 482 [156-158]
481 [159, 110a]
Cl Cl
R2
Me M e Si Zr
Cl Cl
484 [145]
Scheme 8-95. ansa-Metallocenes containing fluorenyl with cyclopentadienyl or indenyl ligands.
8.8 Metal Complexes with Stereogenic Bridging Groups One-Carbon Bridged Complexes. In Rausch and Chien’s bridged ligand 327, only the indenyl ligand has diastereotopic π-faces and its metalation could produce two diastereomeric ansa-metallocene dichloride complexes. Only one isomer of the titanium or zirconium complexes 485 was isolated in each case and characterized by X-ray crystallography [160] (Scheme 8-96). Apparently, the methyl group in the bridging group directs the indenyl ligand to the observed less hindered orientation. Based on the low selectivity observed in the metalation of related chiral 1-substituted indenes, it is unlikely that the metal coordinates initially to the indene as that should lead to a diastereomeric mixture. Initial coordination to the equivalent faces of the cyclopentadienyl moiety would then allow the indenyl ligand to adopt the favorable conformation for the subsequent ansa-metalation. Several example of one-carbon bridged complexes in which the bridging carbon is part of a tetrahy-
532
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
droindenyl ring have been reported. Erker has prepared the titanium and zirconium complexes 486 and 487 [161]. The di-(t-butyl) derivatives 489 and 490 have also
1) n-BuLi, THF, 0 °C to 40 °C
Me
2Li
Me M
Me
or 1) ZrCl4, Et 2O, toluene
Me
Me
Me
or 1) ZrCl4, Et 2O, toluene
Me t-Bu Cl [202] M Cl
Me
489 M = Ti
t-Bu 490 M = Zr
t-Bu
488
R1
R2 Ph
Ph R
H Me
[161]
486 M = Ti 487 M = Zr
1) TiCl3, Et 2O, toluene t-Bu 2) HCl, hexane, Et 2O
2Li+
Cl Cl
332
Me
[160]
Cl
1) TiCl3, Et 2O, toluene 2) HCl, hexane, Et 2O +
Cl
485
R
Me
M
H 3C
2) TiCl4-2THF, THF, -78 °C to reflux or 2) ZrCl4
dl-327
R
H
Zr
491 [158]
Cl Cl
Ph
Zr
Cl Cl
492 [156]
Scheme 8-96. One-carbon stereogenic bridging group. been reported [202]. A number of ansa-metallocenes such as 491 and 492 having substituted fluorenyl groups tethered by a stereogenic group to cyclopentadienyl have been recently reported [158, 156]. Bis(indenes) Bridged by Stereogenic Groups at the 1-Position. Both indenyl ligands arising from deprotonation the dimethylbinaphthyl-bridged bis(indene) 333
8.8
533
Metal Complexes with Stereogenic Bridging Groups
have diastereomeric faces and three diastereomeric ansa-metallocene complexes could be formed (Scheme 8-97). Treatment of the indenyllithium complexes of 333 with TiCl3 followed by oxidation, or by ZrCl4 produced in each case a single isolable diastereomeric ansa-metallocene complex 493 or 494 in low yield along with what was likely oligomeric metallocene complexes. Initial complexation of the metals to the first indenyl ligand was likely not selective. Due to interactions with the chiral bridge, only one diastereomeric initial complex could cyclize to an ansametallocene complex. The other initial diastereomeric complex presumably underwent intermolecular metalation to form oligomeric material. In the solid state 494a adopts a C1-symmetric R R 1) n-BuLi, THF, -78 to 0 °C 2) ZrCl4, DME, r.t. to 50 °C
R R
or 2) TiCl3-(THF)3, THF, reflux 3) HCl, CHCl 3, air
R
Cl Cl
R=H 493a M = Zr 494a M = Ti
(20 - 24 %)
R
R M
333
[162,163]
R = CH 3 R 493b M = Zr 494b M = Ti
H2 M
Cl Cl
493a
O
334
Cl
2) ZrCl4, THF, reflux or 2) TiCl3-(THF)3, THF, reflux 3) FeCl 3, THF
Cl
Ti
or
Cl
Cl
495a M = Ti ( 600 psi H 2, 54%) 495b M = Zr ( 900 psi H 2,100%)
1) NaNH2, THF
O
M
PtO2, CH2Cl2
O
496 ( 800 psi H 2, 68%) (SiO2 chromatography)
H2 (100 atm), Cl
O
M
Cl PtO , CH Cl , r.t. 2 2 2
meso -like
O O
Cl
M
Cl
[164]
497 M = Zr (16%) 498 M = Ti (17%)
499 M = Zr (87%) 500M = Ti (77%)
Scheme 8-97. Metalation of bis(1-indenyl) ligands. conformation. In solution the flipping between two equivalent C1-symmetric conformations of the titanium complex was found to have an energy barrier of about 13 kcal mol–1. The zirconium complex 493a was more mobile [162, 164]. The bis(indenyl)metal complexes 493a and 494a could be hydrogenated to give either the dimethylbinaphthyl-bridged bis(tetrahydroindenyl)titanium and -zirconium complexes 495 or the bitetralin-bridged bis(tetrahydroindenyl)titanium complex
534
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
496. The titanium complexes could be purified by silica gel chromatography. The bis(tetrahydroindenyl) complex 495a was conformationally more mobile than the corresponding bis(indenyl)titanium complex 494a. Metalation of Green’s threitolbridged bis(indene) 334 could also give three diastereomeric ansa-metallocenes. Bis(indene) 334 was deprotonated by sodium amide and metalated with either ZrCl4 or with TiCl3 followed by FeCl3 oxidation to give low yields of the mesolike metallocene dichlorides 497 and 498. Catalytic hydrogenation led to the mesolike bis(tetrahydroindenyl)metal complexes 499 and 500. The complexes were purified by crystallization from hot toluene [163]. The dilithio salt of Rieger’s 1,2-cyclohexane-bridged bis(indene) 336 was metalated with ZrCl4, filtered through silica gel with methylene chloride and concentrated to give an undefined mixture of the three diastereomeric complexes 501. These isomers could not be separated [165]. Metalation of the 1,4-cyclohexanebridged bis(indene) 337 with TiCl3 followed by oxidation gave only the one diastereomeric titanocene complex 502 having the indenyl groups oriented away from the isopropyl groups [115]. Metalation of the lithium salt of 337 with ZrCl4 failed, but the authors did not attempt metalation under equilibrating conditions with tetrakis(dimethylamido)zirconium. The solid state structure of 502 indicates approximately C2-symmetry. Bosnich’s bis(indene) 240 was deprotonated with nBuLi and metalated with either ZrCl4-2THF or with TiCl4-2THF [32]. The resulting crude bis(indenyl)metal complexes were hydrogenated to give the crude bis(tetrahydroindenyl)metal complexes 503 and 504. Both these ‘chiracene’ zirconium and titanium complexes were purified by chromatography on silanized silica gel. In each case a mixture of two C2-symmetric and one meso-like diastereomeric metal complex were formed. The meso-like complexes were the major isomers. Through recrystallization, one C2-symmetric isomer of the zirconium and titanium complexes were obtained and their structures confirmed through X-ray crystallography. Both the initial mixtures of the bis(tetrahydroindenyl)titanium and zirconium complexes could be photolyzed with a 450-W mercury lamp to a photostationary state favoring one of the C2-symmetric isomers. Metalation of the dilithio salt of bis(indene) 342 with TiCl4-2TiCl4 followed by catalytic hydrogenation gave only one C2-symmetric stereoisomeric ‘cyclacene’ titanocene dichloride 505 (Scheme 8-98). This titanium complex was purified by silanized silica gel chromatography and recrystallization. Its solid state structure was determined [166].
8.8
535
Metal Complexes with Stereogenic Bridging Groups
1) n-BuLi THF, 0 °C
Cl
2) ZrCl4-(THF)2 THF, reflux
Cl
(83%)
[165]
Cl +
Zr
+ 2 meso-like C1-symmetric diastereomers
Zr Cl
501
C2-symmetric diastereomers
336
"forward" diastereomer
"backward" diastereomer
Cl
Cl
Ti 1) n-BuLi, THF, -78 to 0 °C
[115]
2) TiCl3, THF, -78 °C to reflux 3) HCl, CHCl 3, air 337
502
(80%)
1) n-BuLi, THF, -78 °C to r.t. 2) TiCl4-2THF, THF, 40 °C 3) HCl gas (48%) Cl M
4) H 2 (30 atm.), PtO 2, CH 2Cl2
+
M
Cl
M
+
Cl
Cl
Cl
Cl
(78% after silanized SiO2 chromatography)
(S,S)-240
or 2) ZrCl4-2THF, THF, 40 °C 3) HCl gas (18%) 4) H 2 (20 atm.), PtO 2, CH 2Cl2
(S,S)- ( R,R )
(50% after silanized SiO2 chromatography)
(S,S)- ( R,S)
(S,S)- ( S,S) 503 M = Ti: (2.5 : 1 : 8.1 504 M = Zr: (1
: 4.2) : 18)
[32]
1) n-BuLi, THF, -78 °C to r.t. 2) TiCl4-2THF, THF, 40 °C 3) HCl gas (34%) 4) H 2 (30 atm.), PtO 2, CH 2Cl2
342
Ti
(58% after silanized SiO2 chromatography and recrystallization)
Cl
[166]
Cl
505
Scheme 8-98. Metalation of bis(1-indenyl) ligands. Biaryl-Bridged Complexes. Brintzinger’s biphenyl-bridged bis(cyclopentadienyl) ligand 19 has equivalent π-ligand faces and its metalation with either ZrCl4 or TiCl3 followed by oxidation gave the expected single isomeric metal complexes 506 and 507 (Scheme 8-99). The titanocene complexes were purified by silica gel chromatography, the zirconium complex by recrystallization. Since no synthetic intermediates were
536
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides 1) TiCl3-3THF 2) 6 M HCl, air 3) SiO2 chromat.
K+ –
R
R
–
or 1) ZrCl4-2THF 2) recryst. Et2O/hexane
K+
19 (+ monocyclopentadienyl)
(yields are from the diiodobiphenyl starting material for the ligand synthesis)
R
Cl
[12]
M
R
Cl 12
20 506a: R = H, M = Ti (18%) 507: R = H, M = Zr (3%) 506b: R = CH 3, M = Ti (4%)
1) n-BuLi, THF -78 °C to -20 °C Cl 2) TiCl3-3THF, -50 °C to reflux 3) 0.5 N HCl, air, CH 2Cl2
Ti
508
354 1) 2) 3) 4)
n-BuLi, THF, -70 °C to -25 °C TiCl3-3THF, -40 °C to r.t. PtO 2, CH 2Cl2, H 2 0.5 N HCl, air, CH 2Cl2
R or 2) ZrCl4-2THF, THF, -40 °C to reflux 3) PtO 2, CH 2Cl2, H 2 4) chrom. silanized SiO2/benzene then cryst. from CH2Cl2/C 6H12
R
363
[167] Cl
R
Cl [168]
M
R
Cl 509a: 510a: 509b: 510b:
R = H, M = Ti (40%) R = H, M = Zr (30%) R = CH 3, M = Ti (46%) R = CH 3, M = Zr (16%)
1) n-BuLi, THF -78 °C to -20 °C Cl 2) TiCl3-3THF, -50 °C to reflux 3) PtO 2, CH 2Cl2, H 2 4) 0.5 N HCl, air, CH 2Cl2 355
Ti
[168]
Cl 508
(5%)
Scheme 8-99. Biaryl-bridged bis(cyclopentadienes) metal complexes. isolated in this sequence, the reported yields are for several steps used to convert the 2,2′-diiodo-1,1′-biphenyls to the final metallocene dichlorides. The structures of all three complexes were determined by X-ray crystallography. Halterman’s binaphthyl-bridged bis(tetrahydroindene) 354 was deprotonated by n-BuLi and metalated with TiCl3 followed by oxidation to give the titanocene dichloride 508 in 68% yield or with ZrCl4 to give the zirconium complex in good yield [167]. Although a nice series of ligands were mad, only the bis(hydroazulene) 357 was cleanly converted to a titanocene complex. The bis(indene) 355 and the bis(hydropentalene) 356 did not form ansa-metallocene complexes under the conditions examined. Bosnich was able to prepare the biphenyl bridged bis(tetrahydroindenyl)metal complexes 509 and 510 by hydrogenation of initially
8.8
537
Metal Complexes with Stereogenic Bridging Groups
formed bis(indenyl)metal complexes [168]. Binaphthyl-bridged bis(tetrahydroindene) 508 was similarly prepared from bis(indene) 355 [168]. In all of these cases, only a single stereoisomeric product could form due to the homotopicity of the ligand faces. CH 3 1) n-BuLi
Ti
2) TiCl3 3) 12 M HCl, O 2 4) BioBeads/toluene
Ti
:
CH 3 :
2.8
CH 3
Cl
Cl
+
Ti
Cl CH 3
(98%)
CH 3 Cl Cl
CH 3
CH 3
Ti
Ti
+
511
1
2) TiCl3 3) 12 M HCl, O 2 4) SiO2/CH2Cl2
CH 3 Cl Cl
CH 3
1) n-BuLi
(meso -346)
+
Cl
(83%) (dl-346)
Cl
4.6
Ti
+
[31]
CH 3 Cl Cl
Cl CH 3
512
2.6
:
CH 3
1
1
:
Cl
Cl Ti
1) n-BuLi
H H
2) TiCl4 3) cryst: toluene
H H
[11]
(13%) 349
18
Scheme 8-100. Chiral ansa-bis(cyclopentadienyl)titanium dichloride complexes. Bis(cyclopentadienyl)metal Complexes. Nantz has examined the ability of the 2,3-butane-diyl bridging group to exert its influence on the orientation of 3-methylcyclopentadienyl groups in titanocene complexes 511 and 512 (Scheme 8-100) [31]. In the metalation of C2-symmetric bis(cyclopentadiene) dl-346, the meso-like isomer of 511 was favored, whereas in the metalation of meso-346, the formation of one of the two rac-like isomers of 512 was favored. In no case was the selectivity very pronounced and it is apparent that insufficient directing effects are at work here. The ligand faces in bis(cyclopentadienyl) 349 are equivalent and only the single titanocene 18 was formed upon its metalation [11]. Chiral Silyl-Bridged Complexes. Metalation of the di(menthoxy)silyl-bridged bis(indene) 365 leads to three stereoisomeric bis(indenyl)titanium dichlorides 513 with a small preference for the meso-like isomer (Scheme 8-101) [169].
538
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
O O O
Cl
Si
Ti
O
+
Si
2.8
:
+
O
Cl
Cl
1
O
Cl Ti
Si
Cl Ti Cl
[169]
513 4.6
:
Scheme 8-101. Di(menthoxy)silyl-bridged bis(indenyl)titanium dichlorides. Chiral Doubly Bridged Bis(cyclopentadienyl). One example has been published by Buchwald in which two cyclopentadienyl ligands are doubly bridged by a nonstereogenic group (Scheme 8-102). In this case the Meerwein’s diketone-derived ligand 371 was metalated to give the titanocene complex 514. The zirconium complex could not be isolated from an analogous attempted metalation [171]. Me
Me 1) TiCl3-3THF
2Li+
Me
2) CHCl 3
Cl
Ti
Cl
[171]
Me 371
514
Scheme 8-102. Chiral doubly bridged bis(cyclopentadienyl)titanium complex. Fluorenyl-Fluorenyl and Fluorenyl-Indenyl Complexes. Rieger has metalated the fluorenyl-indenyl ligand 379 to give the nonracemic metal complexes 515 and 516 in a 3:1 ratio (Scheme 8-103). These isomers could be separated or they could be hydrogenated then separated. In a related nonracemic bis(fluorenyl)zirconium complex 512 formed from bisfluorene) 380, hydrogenation also led to the perhydrogenated complex 518 (including hydrogenation of the phenyl group to a cyclohexyl group) [172].
539
References
Ph 1) n-BuLI
H Ph
2) ZrCl4
Zr
Cl Cl
+
H Ph
Zr
Cl Cl
379 516
515
[172]
Ph 1) n-BuLI 2) ZrCl4
H Ph
Zr
Cl Cl
H2 PtO2
H Cy
Zr
Cl Cl
380 517
518
Scheme 8-103. Fluorenyl-containing metallocenes.
References [1] [2] [3] [4] [5] [6]
[7] [8] [9] [10] [11] [12] [13] [14]
[15] [16]
a) Halterman, R. L. Chem.Rev. 1992, 92, 965. b) Okuda, J. Topics Curr. Chem. 1991, 160, 99. c) Janiak, C.; Schumann, H. Adv. Organomet. Chem. 1991, 33, 291. Moise, C.; Leblank, J.-C.; Tirouflet, J. Tetrahedron Lett. 1974, 1723. Cesarotti, E.; Kagan, H. B.; Goddard, R.; Krüger, C.J. Organomet. Chem. 1978, 162, 297. Wild, F. R. P.; Wasiucionek, M.; Huttner, G.; Brintzinger, H. H. J. Organomet. Chem. 1985, 288, 63. a) Halterman, R. L.; Vollhardt, K. P. C. Tetrahedron Lett. 1986, 27, 1461. b) Halterman, R. L.; Vollhardt, K. P. C. Organometallics 1988, 7, 883. a) Wilkinson, G.; Pauson, P. L.; Birmingham, J. M.; Cotton, F. A. J. Am. Chem. Soc. 1953, 75, 1011. b) Wilkinson, G.; Birmingham, J. M. J. Am. Chem. Soc. 1954, 76, 4281-4284. Reynolds, L. T.; Wilkinson, G. J. Inorg. Nucl. Chem. 1959, 9, 86. Varga, V.; Mach, K.; Schmid, G.; Thewalt, U. J. Organomet. Chem. 1994, 475, 127. Leblanc, J. C.; Moise, C. J. Organomet. Chem. 1977, 131, 35. Schwemlein, H.; Brintzinger, H. H. J. Organomet. Chem. 1983, 254, 69. Gibis, K.-L.; Helmchen, G.; Huttner, G.; Zsolnai, L. J. Organomet. Chem. 1993, 445, 181. Huttenloch, M. E.; Diebold, J.; Rief, U.; Brintzinger, H. H.; Gilbert, A. M.; Katz, T. J. Organometallics 1992, 11, 3600. a) Samuel, E.; Setton, R.; J. Organomet. Chem. 1965, 4, 156. b) Atwood, J. L.; Hunter, W.E.; Hrncir, D. C.; Samuel, E.; Alt, H.; Rausch, M. D. Inorg. Chem. 1975, 14, 1757. a) Hauptman, E.; Waymouth, R. M.; Ziller, J. W. J. Am. Chem. Soc. 1995, 117, 11586. b) Waymouth, R. M.; Coates, G. W.; Hauptman, E. M. United States Patent 5,594,080, Jan. 14, 1997 Shipman, J. M.S. Thesis University of Oklahoma, 1997. Nantz, M. H.; Hitchcock, S. R.; Sutton, S. C.; Smith, M.D. Organometallics 1993, 12, 5012.
540 [17] [18]
[19] [20] [21] [22] [23] [24] [25] [26] [27]
[28] [29]
[30]
[31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51]
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Lotman, M. M.S. Dissertation, University of Oklahoma, 1992. a) Piccolrovazzi, N.; Pino, P.; Consiglio, G.; Sironi, A.; Moret, M. Organometallics 1990, 9, 3098. b) Lee, I.-M.; Gauthier, W. J.; Ball, J.M.; Iyengar, B.; Collins, S. Organometallics 1992, 11, 2115. a) Chen, Z.; Halterman, R. L. Synlett 1990, 103. b) Chen, Z.; Eriks, K.; Halterman, R. L. Organometallics 1991, 10, 3449. Halterman, R. L.; Chen, Z.; Khan, M. A. Organometallics 1996, 15, 3957. Besançon, J.; Tirouflet, J.; Top, S.; Ea, B. H. J. Organomet. Chem. 1977, 133, 37. a) Schlögl, K.; Fortschr. Chem.Forsch. 1966, 6, 479. b) Schlögl, K. Top. Stereochem. 1967, 1, 39. Wild, F. R. W. P.; Zsolnai L.; Huttner, G.; Brintzinger, H. H. J. Organomet. Chem. 1982, 232, 233. Krüger, C.; Lutz, F.; Nolte, M.; Erker, G.; Aulbach, M. J.Organomet. Chem. 1993, 452, 79. Paquette, L. A.; Moriarty, K. J.; McKinney, J. A.; Rogers, R. D. Organometallics 1989, 8, 1707 McLaughlin, M. L.; McKinney, J. A.; Paquette, L. A. Tetrahedron Lett. 1986, 27, 5595. a) Gallucci, J. C.; Gautheron, B.; Gugelchuk, M.; Meunier, P.; Paquette, L. A. Organometallics 1987, 6, 15. b) Paquette, L. A.; Moriarty, K. J.; Meunier, P.; Gautheron, B.; Crocq, V. Organometallics 1988, 7, 1873. c) Sornay, C.; Meunier, P.; Gautheron, B.; O’Doherty, G. A.; Paquette, L. A. Organometallics 1991, 10, 2082 Bell, L.; Whitby, R. J.; Jones, R. V. H.; Standen, M. C. H. Tetrahedron Lett. 1996,37, 7139. a) Knickmeier, M.; Erker, G.; Fox, T. J. Am. Chem. Soc. 1996, 118, 9623. b) Erker, G.; Aulbach, M. ; Knickmeier, M.; Wingermühle, D.; Krüger, C.; Nolte, M.; Werner, S. J. Am. Chem. Soc. 1993, 115, 4590. a) Giardello, M. A.; Eisen, M. S.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1993, 115, 3326. b) Giardello, M. A.; Eisen, M. S.; Stern, C. L.; Marks, T. J. J. Am. Chem. Soc. 1995, 117, 12114. Sutton, S. C.; Nantz, M. H.; Parkin, S. R. Organometallics 1993, 12, 2248. Rheingold, A. L.; Robinson, N. P.; Whelan, J.; Bosnich, B. Organometallics 1992, 11, 1869. a) Ziegler, K.; Schafer, W. Justus Liebigs Ann. Chem. 1934, 511, 101. b) Alder, K.; Holzrichter, H.Justus Liebigs Ann. Chem. 1936, 524, 145. Hafner, K.; Häfner, K. H.; König, C.; Kreuder, M.; Ploss, G.; Schulz, G.; Surm, E.; Vöpel, K. H. Angew. Chem., Int. Ed. Engl. 1963, 2, 123 and references therein. Riemschneider, R. Z. Naturforsch. B1963, 18, 641 and references therein. Alder, K.; Ache, H.-J. Chem. Ber. 1962, 95, 503. Hoch, M.; Rehder, D. Chem. Ber. 1988, 121, 1541. Scroggins, W. T.; Rettig, M. F.; Wing, R. M. Inorg. Chem. 1976, 15, 1381. a ) Knox, G. R.; Munro, J. D.; Pauson, P. L.; Smith, G.H.; Watts, W. E. J. Chem. Soc. 1961, 4619. b) Knox, G. R.; Pauson, P. L. J. Chem. Soc. 1961, 4610. Little, W. F.; Koestler, R. C. J. Org. Chem. 1961, 26, 3247. Erker, G.; Nolte, R.; Krüger, C.; Schlund, R.; Benn, R.; Grondey, H.; Mynott, R. J. Organomet. Chem. 1989, 364, 119 Collins, S.; Hong, Y.; Kataoka, M.; Nguyen, T. J. Org. Chem.1990, 55, 3395. Erker, G.; Aul, R. Chem. Ber. 1991, 124, 1301. Pauson, P. L. J. Am. Chem. Soc. 1954, 76, 2187. a) Riemschneider, R.; Nerin, R. Monatsh. Chem. 1960, 91, 829. b) Mironov, V. A.; Sobolev, E. V.; Elizarova, A. N. Tetrahecron 1963, 19, 1939. Leblanc, J. C.; Moise, C. J. Organomet. Chem. 1976, 120, 65. Erker, G.; Mollenkopf, C.; Grehl, M.; Schönecker, B. Chem. Ber. 1994, 127, 2341. Erker, G.; Mollenkopf, C. J. Organomet. Chem. 1994, 483, 173. Courtier, S.; Tainturier,G.; Gautheron, B. J. Organomet. Chem. 1980, 195, 291. Fiaud, J. C.; Malleron, J. L. Tetrahedron Lett. 1980, 21, 4437. a) Huang, Q.; Qian, Y. Synthesis 1987, 910. b) Qian, Y.; Li, G.; Chen, W.; Li, B.; Jin, X. J. Organomet. Chem. 1989, 373, 185.
References [52] [53]
541
Van de Weghe, P.; Bied, C.; Collin, J.; Marcalo, J.; Santos, I. J. Organomet. Chem. 1994, 475, 121. Adams, H.; Bailey, N. A.; Colley, M.; Schofield, P. A.; White, C. J. Chem. Soc., Dalton Trans. 1994, 1445. [54] van der Zeijden, A.A.H. Tetrahedron Asymm. 1995, 6, 913. [55] a) McLean, S.; Haynes, P. Tetrahedron 1965, 21, 2343. b) McLean, S.; Haynes, P. Tetrahedron 1965, 21, 2313. [56] Alder, K.; Ache, H.-J. Chem. Ber. 1962, 95, 503. [57] Knothe, L.; Prinzbach, H.; Hadicke, E.; Chem.Ber. 1981, 114, 1656. [58] Christoffers, J.; Bergman, R. G. Angew. Chem., Int. Ed. Engl. 1995, 34, 2266. [59] Stone, K. J.; Little, R. D. J. Org. Chem. 1984, 49, 1849 [60] Okuda, J. J. Organomet. Chem. 1990, 385, C39 [61] Gutmann, S.; Burger, P.; Hund, H.-U.; Hofmann, J.; Brintzinger, H. H. J. Organomet. Chem. 1989, 369, 343. [62] Collins, S.; Hong, Y.; Taylor, N. J. Organometallics 1990, 9, 2695. [63] Drake, N. L.; Adams, J. R., Jr. J. Chem. Soc 1939, 61, 1326. [64] Taylor, D. A. H. J. Chem. Soc. 1958, 4779. [65] Attig, T. G.; Wojcicki, A. J. Organomet.Chem.1974, 82, 397. [66] Bursten, B. E.; Callstrom, M. R.; Jolly, C. A.; Paquette, L. A.; Sivik, M. R.; Tucker, R. S.; Wartchow, C. A. Organometallics 1994, 13, 127. [67] Davies, A. G.; Lusztyk E.; Lusztyk, J. J. Chem. Soc. Perkin Trans. II 1982, 729. [68] Hughes, R. P.; Kowalski, A. S.; Lomprey, J. R.; Rheingold, A. L. Organometallics 1994, 13, 2691. [69] Mironov, V. A.; Ivanov, A. P.; Kimelfled, Ya. M.; Petrovskaya, L. I.; Akhrem, A. A. Tetrahedron Lett. 1969, 3347. [70] Yang, Q.; Jensen, M. D. Synlett 1996, 563. [71] Reinarz, R. B.; Fonken, G. J. Tetrahedron Lett. 1973, 4591. [72] Paquette, L. A.; Gugelchuk, M.; McLaughlin, M. L. J. Org. Chem. 1987, 52, 4732. [73] Halterman, R. L.; Vollhardt, K. P. C.; Welker, M. E.; Bläser, D.; Boese, R. J. Am. Chem. Soc. 1987, 109, 8105. [74] a) Colletti, S. L.; Halterman, R. L. Tetrahedron Lett. 1990, 30, 3513. b) Colletti, S. L.; Halterman, R. L. Organometallics 1991, 10, 3438. [75] Moriarty, K. J.; Rogers, R. D.; Paquette, L. A. Organometallics 1989, 8, 1512. [76] Sitzman, H. J. Organomet. Chem. 1988, 354, 203. [77] a) Vernier, C. G.; Casserly, E. W. J. Am. Chem. Soc. 1990, 112, 2808. b) Sitzmann, H.; Zhou, P.; Wolmershäuser, G. Chem. Ber. 1994, 127, 3. [78] Okuda, J. Chem. Ber. 1989, 122, 1075. [79] Okuda, J.; Herdtweck, E. Chem. Ber. 1988, 121, 1899. [80] Mach, K.; Varga, V.; Antropiusová, H.; Polácek, J. J. Organomet. Chem. 1987, 333, 205. [81] Conia, J.-M.; Leriverend, M.-L. Bull. Chim. Soc. Fr. 1970, 2981. [82] Dübner, F. Diplomarbeit, TU Berlin, 1997. [83] Halterman, R. L.; Ramsey, T. M. J. Organomet. Chem. 1994, 465, 175. [84] Paquette, L. A.; Charumilind, P.; Kravetz,T. M.; Böhm, M. C.; Gleiter, R. J. Am. Chem. Soc. 1983, 105, 3126. [85] Halterman, R. L.; Tretyakov, A. Tetrahedron (Symposium-in-Print) 1995, 51, 4371. [86] Schumann, H.; Janiak, C.; Zuckerman, J.J. Chem. Ber. 1988, 121, 207. [87] a) Erker, G.; Schamberger, J.; van der Zeijden, A.A.H.; Denicke, S.; Krüger, C.; Goddard, R.; Nolte, M. J. Organomet. Chem. 1993, 459, 5832. b) Erker, G.; van der Zeijden, A. A. H. Angew. Chem., Int. Ed. Engl. 1990, 29, 512. [88] a) Paquette, L. A.; Bzowej, E. I.; Kreuzholz, R. Organometallics 1996, 15, 4857. b) Garner, C. M.; Prince, M. E. Tetrahedron Lett. 1994, 35, 2463. [89] de Vries, L. J. Org. Chem. 1960, 25, 1838. [90] a) Burger, U.; Delay, A.; Mazenod, F. Helv. Chim. Acta 1974, 57, 2106. b) Kohl, F. X.; Jutzi, P. J. Organomet. Chem.1983, 243, 119.
542 [91] [92] [93] [94] [95] [96] [97] [98] [99] [100]
[101] [102] [103] [104] [105] [106] [107] [108] [109] [110]
[111] [112] [113]
[114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127]
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Fietler, D.; Whitesides, G. M. Inorg. Chem. 1976, 15, 466. Threlkel, R. S.; Bercaw, J. E. J. Organomet. Chem. 1977, 136, 1. Stein, D.; Sitzmann, H. J. Organomet. Chem. 1991, 402, 249. Dormond, A.; Bouadili, A. El.; Moise, C. Tetrahedron Lett. 1983, 24, 3087. Colletti, S. L. Ph.D. Dissertation, Boston University, 1993. Greifenstein, L. G.; Lambert, J. B.; Nienhuis, R. J.; Fried, H. E.; Pagani, G. A. J. Am. Chem. Soc. 1981, 46, 5125 Meurling, L. Acta Chem. Scand.B 1974, 28, 295. Cedheim, L.; Eberson, L. Synthesis 1973, 159. a) Rakita, P. E.; Taylor, G. A. Inorg. Chem. 1972, 11, 2136. b) Andrews, M. N.; Rakita, P. E.; Taylor, G. A. Tetrahedron Lett. 1973, 1851. a) Rigby, S. S.; Gupta, H. K.; Werstiuk, N. H.; Bain, A. D.; McGlinchey, M. J. Polyhedron 1995, 14, 2787. b) Luzikov, Yu. N.; Gergeyev, N. M.; Ustynyuk, Yu. A. J. Organomet. Chem. 1974, 65, 303. Ready, T. E.; Chien, J. C. W.; Rausch, M. D. J. Organomet. Chem. 1996, 519, 21. Adamczyk, M.; Watt , D. S.; Netzel, D. A. J. Org. Chem. 1984, 49, 4226. Edlund, U. Organ. Mag. Reson. 1978, 11, 516. Ribakove, E. C.; Kerber, R. C. Organometallics 1990, 9, 531. Seljeseth, C. B.S. Thesis, University of Oklahoma, 1997. Spaleck, W.; Küber, F.; Winter, A.; Rohrmann, J.; Bachmann, B.; Antberg, M.; Dolle, V.; Paulus, E. F. Organometallics 1994, 13, 954. Spaleck, W.; Antberg, M.; Rohrmann, J.; Winter, A.; Bachmann, B.; Kiprof, P.; Behm, J.; Herrmann, W. A. Angew. Chem., Int. Ed. Engl. 1992, 31, 1347. a) Cedheim, L. Eberson, L. Acta. Chem. Scand. B 1976, 30, 527. b) Normant-Chefnay, C. Bull. Soc. Chim. Fr. 1971, 1351. Erker, G.; Psiorz, Fröhlich, R.; Grehl, M.; Krüger, C.; Noe, R.; Nolte, M. Tetrahedron 1995, 51, 4347. a) Kaminsky, W.; Rabe, O.; Schauwienold, A.-M.; Schupfner, G. U.; Hanss, J.; Kopf, J. J. Organomet. Chem. 1995, 497, 181-193. b) O’Hare, D.; Green, J. C.; Marder, T.; Collins, S.; Stringer, G.; Kakkar, A. K.; Kaltsoyannis, N.; Kuhn, A.; Lewis, R.; Mehnert, C.; Scott, P.; Kurmoo, M.; Pugh, S. Organometallics 1992, 11, 48. Stehling, U.; Diebold, J.; Kirsten, R.; Röll, W.; Jüngling, S.; Mülhaupt, R.; Langhauser, F. Organometallics 1994, 13, 964 Erker, G.; Aulbach, M.; Krüger, C.; Werner, S. J. Organomet. Chem. 1993, 450, 1. a) Erker, G.; Aulbach, M.; Knickmeier, M.; Wingbermühle, D.; Krüger, C.; Nolte, M.; Werner, S. J. Am. Chem. Soc. 1993, 115, 4590. b) Knickmeier, M.; Erker, G.; Fox, T. J. Am. Chem. Soc. 1996, 118, 9623. Stenzel, O.; Schumann, H.; Halterman, R. L., unpublished results. Chen, Z.; Halterman, R. L. J. Am. Chem. Soc. 1992, 114, 2276. a) Katz, T. J.; Acton, N. Tetrahedron Lett. 1970, 2497. b) Katz, T. J.; Acton, N.; Martin, G. J. Am. Chem. Soc. 1973, 95, 2934. Smith, J. A.; Brintzinger, H. H. J. Organomet. Chem. 1981, 218, 159. Hillman, M.; Weiss, A. J. J. Organomet. Chem. 1972, 42, 123. Nifant’ev, I.E.; Butakov, K. A.; Aliev, Z. G.; Urazovskii, I. F. Organomet. Chem. USSR 1991, 4, 622. Bajgur, C. S.; Tikkanen, W. R.; Petersen, J. L. Inorg. Chem. 1985, 24, 2539. Lang, H.; Seyferth, D. Organometallics 1991, 10, 347. Chen, Z.; Halterman, R. L. Organometallics 1994, 13, 3932. Burger, P.; Brintzinger, H. H. J. Organomet. Chem. 1991, 407, 207. Wochner, F.; Zsolnai, Huttner, G.; Brintzinger, H. H. J. Organomet. Chem. 1985, 288, 69. Mink, C.; Hafner, K. Tetrahedron Lett. 1994, 35, 4087. Dorer, B.; Prosenc, M.-H.; Rief, U.; Brintzinger, H. H. Organometallics 1994, 13, 3868. Hafner, K.; Thiele, G. F.; Mink, C. Angew. Chem., Int. Ed. Engl. 1988, 27, 1191.
References
543
[128] Wiesenfeldt, H.; Reinmuth, A.; Barsties, E.; Evertz, K.; Brintzinger, H. H. J. Organomet. Chem. 1989, 369, 359. [129] Mise, T.; Miya, S.; Yamazaki, H. Chem. Lett. 1989, 1853. [130] Erickson, M. S.; Fronczek, F. R.; McLaughlin, M. L. J. Organomet. Chem. 1991, 415, 75. [131] Collins, S.; Hong, Y.; Ramachandran, R.; Taylor, N. J. Organometallics 1991, 10, 2349. [132] Halterman, R. L.; Ramsey, T. M.; Pailes, N. A.; Khan, M. A. J. Organomet. Chem. 1995 497, 43. [133] Hüttenhofer, M.; Prosenc, M.-H.; Rief, U.; Schaper, F.; Brintzinger, H.-H. Organometallics 1996, 15, 4816. [134] Tretyakov, A. Ph.D. Dissertation, University of Oklahoma, 1995. [135] Maréchal, E.; Lepert, A. Bull. Soc. Chim. Fr. 1967, 2954. [136] Collins, S.; Kuntz, B. A.; Taylor, N. J.; Ward, D. G. J. Organomet. Chem. 1988, 342, 21. [137] Grossman, R. B.; Doyle, R. A.; Buchwald, S. L. Organometallics 1991, 10, 1501. [138] Erker, G.; Mollenkopf, C.; Grehl, M.; Fröhlich, R.; Krüger, C.; Noe, R.; Riedel, M. Organometallics 1994, 13, 1950. [139] Herrmann, W. A.; Rohrmann, J.; Herdtweck, E.; Spaleck, W.; Winter, A. Angew. Chem., Int. Ed. Engl. 1989, 28, 1511. [140] Spaleck, W.; Antberg, M.; Dolle, V.; Klein, R.; Rohrmann, J.; Winter, A. New J. Chem. 1990, 14, 499. [141] Scott, P.; Rief, U.; Diebold, J.; Brintzinger, H. H. Organometallics 1993, 12, 3094. [142] Halterman, R. L.; Tretyakov, A.; Combs, D.; Chang, J.; Khan, M. Organometallics 1997, 6. [143] Halterman, R. L.; Combs, D.; Khan, M. A., Organometallics, in press. [144] Alt, H. G.; Milius, W.; Palackal, S. J. J. Organomet. Chem. 1994, 472, 113. [145] Chen, Y.-X.; Rausch, M. D.; Chien, J. C. W. J. Organomet. Chem. 1995, 497, 1. [146] Resconi, L.; Jones, R. L.; Rheingold, A. L.; Yap, G. P. A. Organometallics 1996, 15, 998. [147] Patsidis, K.; Alt, H. G.; Hilius, W.; Palackal, S. J. J. Organomet. Chem. 1996, 509, 63. [148] Alt, H. G.; Zenk, R. J. Organomet. Chem. 1996, 512, 51. [149] Green, M. L. H.; Ishihara, N. J. Chem. Soc., Dalton Trnas. 1994, 657. [150] Willoughby, C. A.; Davis, W. M.; Buchwald, S. L. J. Organomet. Chem. 1995, 497, 11. [151] Kaminsky, W.; Engehausen, R.; Kopf, J. Angew. Chem., Int. Ed. Engl. 1995, 34, 2273. [152] Miyake, S.; Okumura, Y.; Inazawa, S. Macromolecules 1995, 28, 3074. [153] Ewen, J. A.; Jones, R. L.; Razavi, A.; Ferrara, J. D. J. Am. Chem. Soc. 1988, 110, 6255. [154] Razavi, A.; Ferrara, J. J. Organomet. Chem. 1992, 435, 299. [155] Razavi, A.; Atwood, J. L. J. Organomet. Chem. 1993, 459, 117. [156] Alt, H. G.; Zenk, R. J. Organomet. Chem. 1996, 518, 7. [157] Razavi, A.; Atwood, J. L. J. Organomet. Chem. 1995, 497, 105. [158] Alt, H. G.; Zenk, R.; Milius, W. J. Organomet. Chem. 1996, 514, 257. [159] Alt, H. G.; Zenk, R. J. Organomet. Chem. 1996, 526, 295. [160] a) Mallin, D. T.; Rausch, M. D.; Lin. Y.-L.; Dong, S.; Chien, J. C. W. J. Am. Chem. Soc. 1990, 112, 2030. b) Llinas, G. H.; Day, R. O.; Rausch, M. D.; Chien, J. C. W. Organometallics 1993, 12, 1283. [161] a) Erker, G.; Psiorz, C.; Krüger, C.; Nolte, M. Chem. Ber. 1994, 127, 1551. b) Psiorz, C.; Erker, G.; Fröhlich, R.; Grehl, M. Chem. Ber. 1995, 128, 357. [162] a) Burk, M. J.; Colletti, S. L., Halterman, R. L. Organometallics 1991, 10, 2998. b) Colletti, S. L.; Halterman, R. L. J. Organomet. Chem. 1993, 455, 99. [163] Bandy, J. A.; Green, M. L. H.; Gardiner, I. M.; Prout, K. J. Chem. Soc., Dalton Trans. 1991, 2207. [164] Halterman, R. L.; Combs, D.;. Kihega, J. G.; Khan, M. A. J. Organomet. Chem. 1996, 520, 163. [165] Rieger, B. J. Organomet. Chem. 1992, 428, C33. [166] Hollis, T. K.; Rheingold, A. L.; Robinson, N. P.; Whelan, J.; Bosnich, B. Organometallics 1992, 11, 2812. [167] a) Halterman, R. L.; Ramsey, T. M. Organometallics 1993, 12, 2879. b) Halterman, R. L.;
544
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
Ramsey, T. M. J. Organomet. Chem. 1997, 530, 225. [168] Ellis, W. W.; Hollis, T. K.; Odenkirk, W.; Whelan, J.; Ostrander, R.; Rheingold, A. L.; Bosnich, B. Organometallics 1993, 12, 4391. [169] Chen, Y.-X.; Rausch, M. D.; Chien, J. C. W. J. Polym. Sci. A: Polym. Chem. 1995, 33, 2093. [170] Mitchell, J. P.; Hajela, S.; Brookhart, S. K.; Hardcastle, K. I.; Henling, L. M.; Bercaw, J. E. J. Am. Chem. Soc. 1996, 118, 1045. [171] Grossman, R. B.; Tsai, J.-C.; Davis, W. M.; Gutiérrez, Buchwald, S.L. Organometallics 1994, 13, 3892. [172] a) Jany, G.; Gawzi, R.; Steimann, M.; Rieger, B. Organometallics 1997, 16, 544. b) Rieger, B.; Jany, G.; Fawzi, R.; Steimann, M. Organometallics 1994, 13, 647. c) Rieger, B.; Jany, G. Chem. Ber. 1994, 127, 2417. [173] a) Möhring, P. C.; Coville, N. J. J. Mol. Catal. 1992, 77, 41. b) Möhring, P. C.; Vlachakis, N.; Grimmer, N. E.; Coville, N. J. J. Organomet. Chem. 1994, 483, 159. c) Polymerization review: Möhring, P. C.; Coville, N. J. J. Organomet. Chem. 1994, 279, 1. [174] Sullivan, M. F.; Little, W. F. J. Organomet. Chem. 1967, 8, 277. [175] Davis, J. H.; Sun, H.; Redfield, D.; Stucky, G. D. J. Magn. Reson. 1980, 37, 441. [176] Lappert, M. F.; Pickett, C. J.; Riley, P. I.; Yarrow, P. I. W. J. Chem. Soc., Dalton Trans. 1981, 805. [177] Chandra, G.; Lappert, M. F. J. Chem. Soc. (A) 1968, 1940. [178] Vitz, E.; Brubaker, C. H. Jr. J. Organomet. Chem. 1974, 84, C16. [179] Bercaw, J. E.; Marvich, R. H.; Bell, L. G.; Brintzinger, H. H. J. Am. Chem. Soc. 1972, 94, 1219. [180] a) Ti: Gorsich, R. D.; J. Am. Chem. Soc. 1960, 82, 4211. b) Zr: Erker, G.; Gerg, K.; Treschanke, L.; Engel, K. Inorg. Chem. 1982, 21, 1277. [181] Ott, K. C.; deBoer, E. J. M.; Grubbs, R. H. Organometallics 1984, 3, 223. [182] Erker, G.; Nolte, R.; Tsay, Y.-H.; Krüger, C. Angew. Chem., Int. Ed. Engl. 1989, 28, 628. [183] Erker, G.; Nolte, R.; Aul, Wilker, S.; Krüger, C.; Noe, R. J. Am. Chem. Soc. 1991, 113, 7594. [184] Burger, P.; Hund, H.-U.; Evertz, K.; Brintzinger, H.H. J. Organomet. Chem. 1989, 378, 153. [185] Ramsey, T. M. Ph.D. Dissertation, Univeristy of Oklahoma, 1994. [186] O’Hare, D.; Murphy, V.; Diamond, G. M.; Arnold, P.; Mountford, P. Organometallics 1994, 13, 4689. [187] Kowala, C.; Wailes, P. C.; Weigold, H.; Wunderlich, J. A. J. Chen. Soc., Chem. Commun. 1974, 993. [188] Razavi, A.; Atwood, J. L. J. Am. Chem. Soc. 1993, 115, 7529. [189] Patsidis, K.; Alt, H. G. J. Organomet. Chem. 1995, 501, 31. [190] Schmid, M. A.; Alt, H. G.; Milius, W. J. Organomet. Chem. 1996, 514, 45. [191] Hart, S. L.; Duncalf, D. J.; Hastings, J. J.; McCamley, A.; Taylor, P. C. J. Chem. Soc., Dalton Trans. 1996, 2843. [192] Sodhi, G. S.; Sharma, A. K.; Kaushik, N. K. J. Organomet. Chem. 1982, 238, 177. [193] Schmid, M. A.; Alt, H. G.; Milius, W. J. Organomet. Chem. 1996, 525, 15. [194] Schmid, M. A.; Alt, H. G.; Milius, W. J. Organomet. Chem. 1996, 525, 9. [195] Nifant’ev, I. E.; Churakov, A. V.; Urazowski, I. F.; Mkoyan, Sh.G.; Atovmyan, L. O. J. Organomet. Chem. 1992, 435, 37. [196] Schnutenhaus, H.; Brintzinger, H. H. Angew. Chem., Int. Ed. Engl. 1979, 18, 777. [197] Collins, S.; Gauthier, W. J.; Holden, D. A.; Kuntz, B. A.; Taylor, N. J.; Ward, D. G. Organometallics 1991, 10, 2061. [198] Schäfer, A.; Karl, E.; Zsolnai, L.; Huttner, G.; Brintzinger, H.-H. J. Organomet. Chem. 1987, 328, 87. [199] Waymouth, R. M.; Bangerter, F.; Pino, P. Inorg. Chem. 1988, 27, 758. [200] a) Diamond, G. M.; Rodewald, S.; Jordan, R. F. Organometallics 1995, 14, 5. b) Diamond, G. M.; Jordan, R. F.; Petersen, J. L. Organometallics 1996, 15, 4030. [201] Röll, W.; Zsolnai, L.; Huttner, G.; Brintzinger, H. H. J. Organomet. Chem. 1987, 322, 65. [202] Atovmyan, L.; Mkoyan, S.; Urazowski, I.; Broussier, R.; Ninoreille, S.; Perron, P.; Gautheron, B. Organometallics 1995, 14, 2601.
Volume 2 Applications
546
9
Metallocene Catalysis for Olefin Polymerization
9
Metallocene Catalysts for Olefin Polymerization Christoph Janiak Dedicated to Prof. Dr. Peter Jutzi on the occasion of his 60th birthday.
9.1 Introduction Bis(cyclopentadienyl)group IV metal complexes are currently introduced in industry as a new generation of Ziegler–Natta catalysts for the polymerization of olefins [1–4].* Ziegler–Natta catalysis means the rapid polymerization of ethene and αolefins with the aid and in the coordination sphere of a metal catalyst, operating at low pressures (up to 30 atm) and low temperatures (less than 120°C). Among the group IV metallocenes of titanium, zirconium, and hafnium, the zirconocene complexes deserve the most interest both academically and industrially as their combination of properties gears them towards application (the titanocene catalysts are unstable at conventional polymerization temperatures, the hafnium systems are too expensive). Hence, the so-called ‘metallocene catalysts’ are mostly zirconocene derivatives, with zirconocene dichloride (1) being the parent system [5]. Their development as practical polymerization catalysts is the first large-scale industrial application for the long known and well developed class of metallocene complexes [6]. About all the fascinating organometallic chemical aspects in metallocene catalysis, the key aspect which earned them their enormous industrial research input is that group IV metallocene catalysts make polymers accessible which cannot be produced by conventional Ziegler–Natta catalysts, the polymerization of cyclic olefins without ring-opening being one remarkable example.**
*
The unspecified term cyclopentadienyl ligands (Cp) or cyclopentadienylmetal complexes is meant here to include substituted derivatives of the C5H5 parent ligand, such as C5R5 and also the indenyl and fluorenyl system, respectively the complexes thereof. If the parent cyclopentadienyl ligand or its metal compound is meant exclusively it is specifically stated by giving the formula C5H5 or (C5H5)mMLn, respectively. ** In this article, we will only cover the actual metallocene or bis(cyclopentadienyl)metall catalysts. The half-sandwich or monocyclopentadienyl group IV complexes used predominantly for syndiotactic styrene polymerization [7] and the “constrained geometry catalyst” or “Dow catalysts”, R2Si(Cp)(NR)MCl2, containing the silyl-bridged cyclopentadienyl-amido ligand [8] are not included here.
548
9
Metallocene Catalysis for Olefin Polymerization
1 The fortuitous discovery of the iron derivative, ferrocene (C5H5)2Fe [9], was a milestone in chemistry. The recognition and explanation of the remarkable structural and bond theoretical properties of ferrocene and the metallocenes at large by Fischer, Wilkinson, Woodward, and Dunitz [10], culminated in the award of the 1973 Nobel prize for Fischer and Wilkinson and established the study of direct metal–carbon bonds as an independent discipline, Organometallic Chemistry, as we know it today. Ever since, the cyclopentadienyl ligand has played a key role in the development of this field [11]. Cp complexes (π or σ bonded to the central atom) are now known for all main group and transition metals and semi-metals, as well as for the lanthanoid and the accessible part of the actinoid series. It is estimated that more than 80% of all known organometallic complexes of the transition metals contain the cyclopentadienyl fragment or a derivative thereof [12]. Yet, it was just recently, since about 1991 that an industrial application of this important class of compounds finally materialized with the zirconocene complexes as (pre-) catalysts for the Ziegler–Natta polymerization of olefins. From 1984/1985, when the stereoselective polymerization of propene with zirconocene catalysts was first reported (see Sec. 9.2), the synthetic, structural, mechanistic and applied research on metallocene catalysts has been highly visible in organometallic chemistry. A central piece of work in zirconocene catalyzed polymerizations of α-olefins was the elucidation of ligand effects, the correlation of the steric situation at the Cp-ligand with the polymer parameters, such as molar mass, molar mass distribution, comonomer insertion and distribution, and especially the tacticity (cf. Sec. 9.7). With metallocene catalysts the polymer parameters can be tailored through a rational ligand design at the transition metal center (cf. Sec. 9.7). Figure 9-1 gives a schematic overview on the most important ligand modifications. The zirconocene complexes are not catalytically active by themselves but require the action of a cocatalyst. The cocatalyst directly associated with their application oriented development is methylalumoxane, abbreviated as MAO, whose composition and function is detailed in Sec. 9.3. The novelty of metallocene versus classical Ziegler–Natta catalysis is best summarized in the word ‘single-site catalyst’. This means that the active catalytic sites in the molecular zirconocene species are almost identical, whereas classical Ziegler–Natta catalysts are heterogeneous not just by the phase (as solid state catalysts) but also by their composition in hav-
9.2 Historical Development
549
Figure 9-1. Schematic representation of possible basic ligand modifications starting from (C5H5)2ZrCl2 to control the polymer parameters. Often these modifications are employed in combination, i.e. the C5H5-ligands are replaced by indenyl ligands, which are tied together by a bridge to give an ansa-metallocene. The remaining hydrogen atoms on either the five- or six-membered ring of the indenyl system can be substituted further by alkyl or aryl groups (cf. section 7).
ing active sites with different environments at corners and edges on the solid surface [13, 14]. As a consequence of a single-site catalytic species, molar mass distributions (Mw/Mn) in Ziegler polymerization down to about 2, the statistical lower limit for a Schulz–Flory distribution, can be reached with metallocene catalysts [15]. Metallocene catalysts are sometimes also termed ‘homogeneous’ referring both to their solubility and to their single-site character. In terms of their solubility, such a description may be slightly misleading, however, since for the industrial applications in slurry or gas phase processes and to gain control over the polymer morphology, the metallocene systems have to be heterogenized on support materials [4, 16, 17]. (In connection with the heterogenation of metallocenes we note that activated zirconocenes were recently intercalated in silicates and tested for olefin polymerization [18].) Other descriptive names found in the literature for the zirconocene catalysts are ‘molecular-defined’ or after the names of the founders and long-time principal investigators ‘Kaminsky–Brintzinger’ systems. The impact of ‘metallocene catalysts’ extends beyond simple olefin polymerization. It has opened the doors to new classes of polyolefins inaccessible by classical Ziegler–Natta polymerization. It has advanced the development of chiral metallocenes. It is promoting the interest to explore further possible applications of metallocenes in general and it is encouraging the search for other single-site catalysts in place of heterogeneous industrial systems. We would like to close this introduction by noting that further information, additional or different views on the various aspects of metallocene catalysis for the polymerization of olefins can be found in the literature [19–24].
550
9
Metallocene Catalysis for Olefin Polymerization
9.2 Historical Development Metallocene catalysts for the polymerization of ethene are almost as old as Ziegler–Natta catalysis itself. In 1957, Natta, Pino et al. [25] and independently Breslow and Newburg [26] reported on a soluble, crystalline and isolable [27] complex from (C5H5)2TiCl2 and AlEt3 which was polymerization active towards ethene but much slower than a comparative heterogeneous Ziegler catalyst, e.g. TiCl4/AlEt3. Breslow and Newburg did, however, note that traces of oxygen in the monomer increased the catalyst activity. (C5H5)2TiCl2 alone is completely inactive towards ethene polymerization. Natta et al. recognized the value of these ‘bimetal complexes’ for the study of the heterogeneous Ziegler catalysts [28]. In the following two decades, (C5H5)2TiCl2 with different alkylaluminum halide cocatalysts (AlMe2Cl, AlEt2Cl, AlEtCl2), because of its well defined character, and the low polymerization activity, was used as a model system to develop an understanding of the classical Ziegler catalysis. Important ideas for the elementary processes, the mechanism of Ziegler-type polymerization reactions were derived therefrom, namely having electron deficient compounds as catalysts [29], mechanistic suggestions for Ti alkylation, activation equilibria, Ti–olefin π complex formation, chain growth at titanium and Ti→Al chain transfer processes were based on kinetic studies (cf. Scheme 9-3, Sec. 9-5) [30]. The very important suggestion that the catalytic activity of metallocenes is due to the presence of free ions, such as (C5H5)2Ti+-Me which are formed in some kind of equilibrium reaction, e.g. as in Eq. (9-1) via methylation and halide abstraction, was based on electrodialysis studies of the (C5H5)2TiCl2/AlMe2Cl system [31]. We would like to note that these findings by the group of Dyachkovskii, Shilova and Shilov were long overlooked or not accepted but later proved ingenious in view of the independent isolation of catalytically active zirconocenium cations with noncoordinating borate anions (see Sec. 9.4). Confirmation was later also provided by Eisch et al. through the observation of an initial insertion of an unsaturated hydrocarbon (PhC≡CSiMe3) into the Ti–C bond of the catalytic system derived from (C5H5)2TiCl2/AlMeCl2. The ion-pair product [(C5H5)2Ti-C(SiMe3)=CMePh]+AlCl4– was characterized by an X-ray structure, and the cation is schematically depicted in 2 [32]. The existence of Ti–Al halide bridged dimetallic species, such as 3 in Eq. (9-1), was supported by an Xray structure of 4 [27] and 5 [33]. NMR spectral evidence for the generation of equilibrating solvent separated and contact ion pairs in polar halocarbons was found in the interaction of (C5H5)2Ti(Cl)CH2SiMe3 with AlCl3, giving a system which was active towards ethene polymerization [34].
9.2 Historical Development
551
2 → (C H ) Ti+-Me + AlCl – (C5H5)2TiCl2·AlMeCl2 ← 5 5 2 4 methylation ↓↑
↑↓ halide abstraction
(9-1)
(C5H5)2TiMeCl→AlCl3 3
4
5
In 1973 and 1975, Reichert and Meyer [35] and Long and Breslow [36] noted a strong polymerization rate increasing influence of small amounts of water when added to (C5H5)2TiEtCl/AlEtCl2 and (C5H5)2TiCl2/AlMe2Cl, respectively. These findings were unusual as water was until then generally viewed as a catalyst poison for Ziegler systems and perhaps only used to lower the molar mass of polymers [36]. Neither of the authors did, however, offer a conclusive explanation for this effect, nor was there a follow-up study by them. Up to this point chlorine containing titanocene and especially aluminum components were always used and the chlorine content was seen as essential because it was proportional to the polymerization activity. In 1976, Sinn Kaminsky et al. could, however, show that a chlorine-free system based on (C5H5)2TiMe2 and AlMe3 became highly active, when water had been added to the aluminum alkyl before reaction with the metallocene [37]. The optimum Al to H2O ratio lay between 2:1 and 5:1 and the formation of alumoxanes (see Sec. 9.3) was assumed; no activity was found at a ratio of 1:3 when all three alkyl groups were hydrolyzed. In the same paper, it was also shown that zirconocene complexes could be activated for ethene polymerization. In a follow-up study by the same principal investigators in 1980, methylalumoxane was synthesized separately and employed as an activa-
552
9
Metallocene Catalysis for Olefin Polymerization
tor for (C5H5)2ZrMe2 to give record high activities of 106 g PE/(g Zr · h · bar) [38]. Lowering the zirconocene concentration and adjusting the Al(MAO):Zr ratio raised the activity further to 3.1.106 g PE/(g Zr · h · bar) [39, 40], the use of (C5H5)2ZrCl2 instead of the zirconocene dimethyl precursor even gave 5.106 g PE/(g Zr · h · bar) [41]. Although an activity comparison between soluble catalysts in homogeneous solution and carrier supported systems should be judged carefully, on a g PE/g transition metal basis this activity was more than a magnitude higher than those of the best known conventional Ziegler catalyst [42]. Hence, industrial interest was awakened, but the simple, achiral metallocenes still had a problem: They did not polymerize propene (at least not under normal conditions, see Sec. 9.7), but gave only atactic polypropene of low molecular weight (oligomers, cf. Sec. 9.9) [40, 43]. In 1982, Brintzinger et al. published the synthesis of the chiral titanocene compounds, rac-C2H4(Ind)2TiCl2 (6) and rac-C2H4(IndH4)2TiCl2 (7) [44], the zirconocene derivatives followed in 1985 [45]. (Scheme 9-1 [46, 47]) The way chosen here (out of other possibilities [12, 48]) to introduce chirality in these metallocenes, is freezing the rotation of the substituted cyclopentadienyl rings by tying them together with a handle, namely the ethanediyl bridge. Using the Latin word for bent handle ‘ansa’, this type of metallocene is also called ansa-metallocene [49]. It may be noted that a C2 axis of rotation is still present in the enantiomeric forms of 6 and 7 (as well as in many other ansa-metallocenes), hence the systems are not asymmetric, only dissymmetric (i.e. lacking mirror symmetry). This is, however, sufficient to induce the existence of enantiomeric forms [50]. Moreover, we will see in Sec. 9.7 that the C2-symmetry of these metallocene catalysts is important in that it provides for the stereoselectivity in propene polymerization. The higher stereoselectivity in transition-metal catalyzed reactions with C2-symmetrical systems over comparable chiral catalysts with a total lack of symmetry is a general phenomenon and is explained by means of a reduced number of possible competing diastereomeric transition states [50].
9.3 Methylalumoxane—Characteristics and Function
553
Scheme 9-1. Synthesis of racemic ethylene-bis(4,5,6,7-tetrahydro-1indenyl)titanium [44] or -zirconium dichloride (7) [45] and its meso form (M = Ti, Zr; Ind = indenyl). For titanium the meso form is the kinetically preferred product and the rac isomer had to be derived by a photoconversion reaction. With zirconium only the rac form appeared to be present. With other bridged ligands, e.g. Me2Si(C5H3-3-R)22– and R = tBu, SiMe3, CMe2Ph, 1-phenylcyclohexyl, rac/meso ratios of close to 1 are, however, obtained with zirconium as a central metal [46] and are also subject to photoinduced interconversions [47]. This will conclude the historical introduction. How the development took place from here is then presented further in the following separate sections.
9.3 Methylalumoxane – Characteristics and Function The development of metallocene catalysis for olefin polymerization is closely tied to the discovery and use of the cocatalyst methylalumoxane (MAO) [51]. αlumoxanes in general are understood as species containing an oxygen bridge binding two aluminum atoms – any alkyl, aryl, halide, alkoxy or other group can then be the pendant ligand bonded to the aluminum atoms. Alumoxanes are an independently studied class of compounds, long known before the interest in the methyl derivative as a cocatalyst for zirconocenes arose [52]. Alumoxanes are obtained from the controlled hydrolysis of organoaluminum compounds. For methylalumoxane the careful hydrolysis of trimethylaluminum (TMA) by crystal water of CuSO4 · 5 H2O, MgCl2 · 6H2O or Al2(SO4)3 · ∼14–18H2O proved to be the most effective laboratory method of obtaining uniform products in high yield [52–55]. Technically MAO is made from the direct reaction of water with TMA.
554
9
Metallocene Catalysis for Olefin Polymerization
While higher-alkyl alumoxanes are defined monomeric or oligomeric compounds, methylalumoxane still remains a ‘black box’ despite its uniqueness as a cocatalyst. (Other alumoxanes, such as ethyl-, iso- or tert-butyl-alumoxane, were tried as cocatalysts but did not reach the activity of MAO [39, 56, 57].) Attempts were of course made to elucidate the MAO structure [55, 58], e.g. with the help of size exclusion chromatography [59], NMR [60], mass balance and phase separation experiments [61]. The methylalumoxane structure can, however, only be postulated as being a mixture of linear (8) or cyclic (9) oligomers of MeAlO units, and is most often represented as such. Cluster structures capable of enclathrating a chloride ion or a metallocene in a host–guest complex (10) have been suggested as well [62, 63].
9.3 Methylalumoxane—Characteristics and Function
555
From its synthesis MAO contains residual trimethylaluminum (about 5% in commercial MAO–toluene solutions is given in company certificates, while other studies suggest up to 25–50% [60, 62, 64]), partly free and in part ‘associated’ [62] to MAO, which is important for the solubility of MAO in aromatic hydrocarbons [59]. The agreement appears to be that the MAO oligomers are fluxional molecules with a dynamic equilibrium among them, which changes their size and structure [16, 62]. The average molecular weight depends on the preparative conditions [55] (about 900–1100 g mol–1 for the commercial 10 wt% toluene solutions). Toluene solutions of MAO obviously ‘age’ within several weeks time (especially when the flask is opened more often under inert gas, so that free TMA evaporates) leading to gel formation and a change (usually an increase) in their cocatalytic activity when used in polymerization catalysis [65]. The full cocatalytic functionality of MAO towards zirconocenes is also unknown [57]. An NMR [56] and a combined IR/NMR study on zirconocene dichloride/MAO systems [67] or an X-ray structure of a neutral dimethylzirconocene/aluminoxane adduct {(C5H5)2ZrMe(µ-OAlMe2)}2 [68] did not shed much light on this question. Aside from alkylating the metallocene dihalide, abstracting a chloride or alkide with the formation of a [Cp2ZrMe]+ cation and being a scavenger for impurities, the cocatalytic function of MAO is probably best described as establishing a stabilizing environment for the metallocene cation or cation–anion pair in the form of a host–guest or ‘crown-alumoxane’ complex [69]. This may appear akin to enzymatic catalysis where the organic hull of the enzyme protects its active center. The formulation of metallocene cations as the active species is based on the work of Dyachkovskii et al. noted above [31] and on the synthesis of isolable and olefin-polymerization active cations by independent routes (cf. Sec. 9.4). The identity of the polymers obtained with the ionic zirconocenium-borate catalysts and with zirconocene/MAO systems was also taken as an indication that essentially the same cationic [Cp2ZrMe]+ species are the active form in the latter. Furthermore, Xray photoelectron spectroscopy studies on Cp2ZrR2/MAO (R = Cl, Me) also suggested the formation of cationic species [70]. While in the heterogeneous Ziegler– Natta systems Ti(III) constitutes the active species, reduction of (C5H5)2ZrCl2 with the formation of Zr(III) does occur with MAO but is negligible during the first few hours, according to an EPR-study [71]. It has been found that very high ratios of MAO to metallocene are required to achieve a good polymerization activity. The use of UV–Vis spectroscopy was suggested to follow the activation with Al/Zr and catalyst stability over time [72]. Furthermore, the activity of the Cp2ZrCl2/MAO catalyst often seems to increase ‘indefinitely’ with the increase of the Al:Zr ratio [39, 40, 43, 53]. This effect is usually discussed on the assumption of an equilibrium between an inactive metallocene precursor and its catalytically active form (Eq. (9-2)) [73, 74]. + – Cp2ZrMe2 + MAO → ← [Cp2ZrMe] + [Me-MAO]
(9-2)
556
9
Metallocene Catalysis for Olefin Polymerization
With the ansa-metallocenes C2H4(IndH4)2ZrCl2 [69], Me2C(C5H4)(Flu)ZrCl2, and Me2Si(Ind)2ZrCl2 [75] a maximum activity for molar Al/Zr ≈ 3500, 1300, and 10900, respectively, was reported in propene polymerization and also in ethene polymerization for the latter two with Al/Zr ≈ 5100 and 26600, respectively. For propene polymerization it was also found that the catalyst activities and polymer molecular masses depended upon the MAO concentration, whereas zirconocene concentration and the Al/Zr molar ratio were less significant. A competition between MAO and propene for the vacant coordination sites was invoked to explain the adverse effect on the catalyst activity. Optimum performance with the rac-Me2Si(2Me-benz[e]indenyl)ZrCl2 complex in terms of activity and molecular mass was obtained at an MAO concentration with [Al] = 5 mM [76] The activation equilibrium sketched in Eq. (9-2) is just a simplified version of a most likely much more complicated kinetic scheme with a combination of at least two dynamic equilibria between temporarily inactive (dormant), active, and deactivated zirconocene dimers [77]. The analysis of data from the ethene polymerization with (C5H5)2ZrCl2/MAO and temperature, zirconocene and MAO concentration as variables suggested the presence of two kinds of active species. One kind of species is produced from the other via a pseudo-first-order reaction [78]. Similarly two successive dynamic equilibria were found in the system (C5H5)2Ti(propyl)Cl/AlEtCl2, where a second equilibrium, which lies mostly on the ‘left side’, leads to the active complex and requires a large aluminum excess to be shifted fully to the right [74]. In line with the formation of ion pairs, the solvent polarity has a considerable effect not only on the activity (increasing with polarity) but also on polymer molar mass and tacticity (for propene) [79]. Many zirconocene systems described in the literature require Al:Zr ratios on the order of 1000:1 to 10 000:1 to achieve a reasonable activity. The necessary Al:Zr ratio also depends on the zirconocene concentration. At very small zirconium concentration the MAO content has to be increased overproportionally to prevent dissociation of the active complex upon dilution according to the law of mass action [73, 80, 81]. For the record high activities reported by Kaminsky et al. [38, 41] extremely high Al:Zr ratios of 150 000:1 or even 500 000:1 were employed for the very small metallocene concentrations of 10–7 and 10–8 M, respectively. Economically such high excess ratios of the expensive MAO are very unfavorable because it makes the cocatalyst the actual cost factor in metallocene/MAO-catalysis and investigations were carried out to reduce the ratio MAO/metallocene. Apparently, up to 90% of the MAO can be replaced by TMA without significant loss in polymerization activity towards ethene [53]. Necessary reactivation reactions from bimetallic deactivation processes and equilibria (cf. Sec. 9.5) are apparently responsible for the high MAO excess, appropriate substitution at the cyclopentadienyl or indenyl ligands can decrease the probability for deactivation resulting in less excess of MAO. The MAO/ metallocene ratio can also be reduced by supporting the metallocene on a solid surface, e.g. silica [82].
9.4 The Borane Activators or Cationic Catalysts
557
In view of the often high TMA content of MAO, the role of trimethylaluminum in the activation process was also tried to address separately. An NMR study on the reactions, equilibria and polymerizations in (C5H5)2TiMeCl/AlMe3 and (C5H5)2Ti(13)MeCl/MAO/13C2H4 indicated that MAO is a better alkylating agent and has a greater capacity for producing and stabilizing cation-like complexes, like [(C5H5)2TiMe+][Cl-Al(Me)O]n–. At low Al/Zr ratio two types of ion pairs have been detected [83]. Zirconocene dichloride and TMA alone were found to be either inactive [61, 83] or to exhibit very low polymerization activity [14, 60]. MAO is recently used also in the activation of other, especially zirconium nitrogen and oxygen coordination complexes for olefin polymerization [84]. We close this section by noting that there are also conflicting reports as to the actual role of MAO in metallocene catalysis (e.g. Ref. [60]) and because of the complexity or illdefinition of this cocatalytic system other activators or MAO substitutes were and are still sought (vide infra). An early yet not economically viable example presented the system AlMe3-AlMe2F in CH2Cl2 demonstrated for the activation of (C5H5)2TiPh2 [85]. Until today the only alternative to MAO are noncoordinating borate counter-ions. These will be the topic of the next chapter.
9.4 The Borane Activators or Cationic Catalysts and other Cocatalyst-free Systems In 1986, Jordan et al. first opened the possibility to synthesize and isolate cationic metallocene complexes of group IV and he showed that they polymerize ethene in the absence of an aluminum cocatalyst [86, 87]. One-electron oxidants were used to generate the zirconocenium cations (Eq. (9-3)), which could then be trapped in the presence of labile ligands L (Eq. (9-4)). In CH2Cl2 solution, the THF partly dissociates from the complex [Cp2ZrR(THF)]+[BPh4]– which catalyzed ethene polymerization in low yield. The THF ligand functions as an inhibitor since no activity is observed in tetrahydrofuran solution and the transient intermediate [Cp2ZrR]+BPh4– when generated in CH2Cl2 in the presence of ethene gave a much more rapid polymerization catalyst (R = Me, CH2Ph). Cp2ZrR2 + ox → ‘[Cp2ZrR2]+ · ’[BPh4]– + red
R = Me, CH2Ph
↓ ox = [(C5H4Me)2Fe]+[BPh4]–, Ag+[BPh4]– ‘[Cp2ZrR]+’[BPh4]– + R.
(9-3)
red = (C5H4Me)2Fe, Ag0
[Cp2ZrR]+[BPh4]– + L → [Cp2ZrR(L)]+[BPh4]–
L = THF, MeCN
(9-4)
558
9
Metallocene Catalysis for Olefin Polymerization
An alternative access to cationic group IV metallocenes was developed by Hlatky, Turner and Bochmann et al. through the protonolysis of metal–alkyl bonds with tertiary ammonium–tetraphenylborate (Eq. (9–5)) [88, 89]. This route leads to ionic, base–free (aside from the amine) zirconocene catalysts, which are active even in solvents such as toluene or hexane. With the amine still present, the formation of weakly bound amine complexes may be assumed (11, Scheme 9-2) [89]. Cp2MMe2 + [R3NH][BAr4] → [Cp2MMe]+[BAr4]– + CH4 + R3N
(9-5)
M = Ti, Zr R3N = Ph(Me2)N, nBu3N Ar = Ph, p-C6H4R′ (R′ = Me, Et) The reaction in Eq. (9-5) or the cation–anion interaction can become more complicated, however, as from the reaction of (C5Me5)2ZrMe2 with [nBu3NH][B(pC6H4Et)4] in toluene, the zwitterionic, yet polymerization-active compound 12 (Scheme 9-2) was obtained. In 12 one of the phenyl rings is metallated in the meta position, probably by an electrophilic attack of the intermediate [(C5Me5)2ZrMe]+ cation and CH4 elimination [88]. Low-temperature proton NMR investigations could prove a fluctuating η2 coordination of a phenyl ring from BPh4– to the cationic zirconium center (13, Scheme 9-2) [90]. Hence, aside from irreversible deactivation reactions, solutions of the cationic metallocenes obtained according to Eq. (9-5) contain several species in equilibrium (Scheme 9-2) [23b] which compete with the olefin complex (14, Scheme 9-2). Subsequently, the polymerization activity was still low when compared to MAO-activated systems. To further enhance the activity of cationic metallocene catalysts it is necessary to increase the concentration of the actual polymerization active complex 14 (Scheme 9-2). If the equilibrium between 13 and 14, i.e. the cation–anion interaction and the anion displacement by the olefin, was of special importance then even weaker coordinating anions had to be introduced [90].
9.4 The Borane Activators or Cationic Catalysts
559
Scheme 9-2. Possible solution equilibria of cationic metallocene catalysts obtained by protolysis with tertiary ammonium tetraarylborate (Eq. 9-5) (adapted from Ref. [23b]). For simplicity, reactands such as NR3, solvent or olefin are not added to the reaction arrows; irreversible deactivation reactions are not included. To achieve this, perfluorinated arylboranes were used. Marks et al. employed tris(pentafluorophenyl)borane, B(C6F5)3, and with Cp2ZrMe2 obtained a crystallographically characterizable, active polymerization catalyst [91]. In a side-product free reaction, the borane abstracts a methyl group from the metallocene (Eq. (9-6)) to give in the solid state a weakly coordinated cation–anion pair with a non–linear Zr(µ-Me)B bridge (15). The same feature was observed in the structure of {C5H31,3-(SiMe3)2}2ZrMe(µ-Me)B(C6F5)3 [92]. Perhaps a similar interaction is present in zirconocene/MAO systems. Hydrogenolysis of 15 or hydride abstraction from
560
9
Metallocene Catalysis for Olefin Polymerization
Cp2ZrH2 by B(C6F5)3 gives a hydride complex 16 (Eq. (9-7)), where the anion is weakly coordinated by Zr...F contacts [93]. The perfluorinated borate anions presented here are also referred to as noncoordinating anions. Cp2ZrMe2 + B(C6F5)3 → [Cp2ZrMe]+[µ-MeB(C6F5)3]– (15)
(9-6)
↓ 2 H2 / – 2 CH4 Cp2ZrH2 + B(C6F5)3 → [Cp2ZrH]+[HB(C6F5)3]– (16)
15
(9-7)
16
To use the tetraarylborate anion and at the same time avoid the presence of amines, triphenylcarbonium (trityl) borate salts can be reacted with Cp2ZrMe2; aryl is usually pentafluorophenyl (Eq. (9-8)) [94]. 2Cp2ZrMe2 + 2[Ph3C]+[B(C6F5)4]– → 2[Cp2ZrMe]+[B(C6F5)4]– + 2Ph3CMe ↓
(9-8)
↑
[Cp2Zr(Me)-µ-Me-Zr(Me)Cp2]+[ B(C6F5)4]– + [Ph3C]+[B(C6F5)4]– + Ph3CMe
(9-9)
The counter-ion, e.g. [B(C6F5)4]– versus [MeB(C6F5)3]–, has an effect on the activity and stereoselectivity, when chiral catalysts are used [95]. The sole description of the zirconocene methyl cations as monomeric species may, however, be too simple. It was shown that dimeric complexes initially form at low temperature when Cp2ZrMe2 is reacted with [Ph3C]+[B(C6F5)4]– (Eq. (9-9)). The stability of the dimer and its tendency to further react with the trityl borate to give the monomeric cation depends on the nature of the Cp-ligand; with bridged bis(indenyl) ligands the reaction proceeds only above 0°C [96]. A dimeric complex, as described in Eq. (9-9), could recently be crystallographically characterized (17) by using the bulky tris(perfluorobiphenylborane) (18) as an abstracting agent. The reduced coordina-
9.4 The Borane Activators or Cationic Catalysts
561
tive tendencies of the resulting bulky borate [Me-18]– stabilize the dimer, even with excess borane and long reaction times [97]. The very weak ion-pairing behavior of this bulky borate versus [Me-B(C6F5)3]– leads to a significant activity enhancement with constrained geometry (‘Dow’) catalysts towards ethene and ethene/hexene copolymerization [97].
In 15 and 16 the cation-anion contacts are apparently this weak that the ‘noncoordinating’ anion can be easily displaced by the olefin. Quantitative thermodynamic and kinetic parameters for ion pair formation, dissociation and reorganization in [(C5H3-1,2-Me2)2MMe]+[MeB(C6F5)3]3– complexes (M = Zr, Hf) [98] and other metallocenium ions [99] in various solvents are available. The activity of these ionic, base-free catalysts is comparable or even surpasses that of zirconocene/MAO systems, yet in the absence of scavenging agents, the cationic catalyst is not very long lived. Polymerization times given are in the range of minutes [94, 100]. Thus, in applications of the borate-activated catalysts, higher aluminum alkyls are favorably added to act as purifying agents and thereby enhance the observed activity [95,101]. Except trimethylaluminum is not suitable as a scavenger here, because of inactive heterodinuclear dimers which are formed in equilibrium with the monomeric cations (Eq. (9-10)) [96]. With AlEt3 related heterodinuclear dimers form but dissociate more readily so that polymerization activities increase in the order Me < Et [102]. + [Cp2ZrMe]+ + AlMe3 → ← [Cp2Zr-(µ-Me)2-AlMe2]
(9-10)
Addition of B(C6F5)3 to zirconocene butadiene yields a zirconium-boron betain complex (19, Eq. (9-11)) [103] which allowed the observation of the first insertion step at a single-site catalyst. The primary insertion products (20) for both ethene and propene (Eq. (9-12)) are stable at –50°C and could be observed by NMR spec-
562
9
Metallocene Catalysis for Olefin Polymerization
troscopy. The complexes 20 are still active polymerization catalysts forming polyethene and -propene upon temperature increase in the presence of excess monomer [104].
(9-11)
(9-12)
Ion cyclotron resonance mass spectrometry was used to study the reaction rates of five different bridged and unbridged [Cp2Zr-CH3]+ cations with H2 and C2H4. After insertion of ethene the resulting propyl complex dehydrogenates to give the final product, presumably an η3-allyl complex. For the reaction ‘[Cp2Zr-CH3]+ + C2H4 → [Cp2Zr-C3H5]+ + H2’ the rates decrease with increasing steric demand of the Cp2-ligands in the order Me2Si(C5H4)2 > (C5H5)2 > (C5H5)(Ind) > (Ind)2 >> (Flu)2 [105]. Group III and Lanthanoide Metallocene Catalysts: Neutral group III (Sc, Y, La) and lanthanoide metallocene complexes of the general type Cp2MR [106] and [Cp2MH]2 [107], which in their monomeric form are iso-d-electronic (d0, not counting the f-electrons) with [Cp2ZrR]+, are also capable of polymerizing ethene in the absence of a cocatalyst and show the same mechanistic features (see below), such as olefin insertion, β-H elimination and M–R bond hydrogenolysis as the zirconocene catalysts. New studies in this direction also involve divalent samarocene, Cp2Sm [108] as well as ansa-Cp2Sm(THF)2 [109]. A single-component chiral organoyttrium ansa-metallocene catalyst [Me2Si(C5H2-2-SiMe3-4-tBu)2YH]2 was found to slowly polymerize propene (97% isotactic), 1-butene, 1-pentene, and 1-hexene [110]. The ring-opening Ziegler polymerization of methylenecyclopropane (cf. Eq. (9-19)) with {(C5Me5)2LnH}2 (Ln = Sm, Lu) gave exo-methylene functionalized polyethenes [111]. Group III and lanthanoide metallocenes, e.g. (C5Me5)2LnMe(THF) (Ln = Y, Sm, Yb), {(C5Me5)2SmH}2, and (C5H4Me)2Yb(DME), are also used as initiators for the polymerization of polar monomers such as alkyl acrylates to give high molecular weight polymers with extremely narrow molecular mass distribution [112].
9.5
General Mechanism of Chain Growth, Chain Transfer and Deactivation
563
9.5 General Mechanism of Chain Growth, Chain Transfer and Deactivation Homogeneous group IV metallocene polymerization catalysts were initially investigated to understand the long-debated mechanism of Ziegler–Natta catalysis (see
Scheme 9-3. General features of the Ziegler-polymerization mechanism. (a) From the inactive precatalyst the cocatalyst creates an active metal alkyl species by alkylation (if necessary) and ligand abstraction to create a vacant coordination site. (b) Repeated monomer coordination and insertion then leads to chain growth/propagation in competition with (c) chain transfer (termination) reactions to aluminum, by β-hydrogen transfer to metal or to monomer or by hydrogenolysis (if hydrogen is added as a molar mass regulator). The metal-alkyl or metal-hydride species formed from chain transfer are still active and can start a new polymerization process again.
564
9
Metallocene Catalysis for Olefin Polymerization
Sec. 9.2). The essential features for the insertion polymerization, derived therefrom, are collected in Scheme 9-3. For the insertion of an olefin into a transition metal alkyl bond in Ziegler–Natta polymerization (part b in Scheme 9-3), three basic mechanism have been suggested, each being supported by experimental evidence: (i) the direct insertion mechanism, proposed by Cossee and Arlman [113], involving a loosely coordinated four-center transition state (Scheme 9-4); (ii) the metathesis mechanism, proposed by Green and Rooney [114], in which an α-hydrogen transfer from the end of the polymer chain and formation of a metal carbene/alkylidene precede the formation of a metallacyclobutane complex (Scheme 9-5) and (iii) the ‘modified Green–Rooney mechanism’ (see Sec. 9.5.1).
Scheme 9-4. Direct-insertion mechanism according to Cossee–Arlman for the olefin insertion into a transition metal alkyl bond. ( = vacant coordination site).
Scheme 9-5. Metathesis-mechanism according to Green–Rooney for the olefin insertion into a transition metal alkyl bond.
9.5
General Mechanism of Chain Growth, Chain Transfer and Deactivation
565
9.5.1 α-Agostic Interactions and the Modified Green–Rooney Mechanism [115] Intermediate between the Cossee–Arlman and the Green–Rooney mechanism (see above), the ‘modified Green–Rooney mechanism’ has been suggested (by Brookhart and Green) for olefin insertion into a transition metal alkyl bond [116]. There, an α-agostic C–H interaction in a transition state was proposed to assist the insertion of an olefin (Scheme 9-6). With respect to the last mechanism, a number of experiments have been designed to explore the suggested role of an α-agostic intermediate in the olefin insertion reaction. Thus, while the product of an induced cyclization in racemic 1-d1-5-hexenyl- and heptenylchlorotitanocene (Eq. (9-13)) argued against an α-CH activation [117], the hydrocyclization of trans,trans-1,6d2-1,5-hexadiene by a scandocene hydride, Me2Si(C5Me4)2Sc(PMe3)H (Eq. (9-14)) [118] as well as the hydrodimerization of (E)- or (Z)-1-deutero-1-hexene by achiral zirconocene dichloride/MAO (Eq. (9-15)) [119] gave results, which were judged to provide experimental evidence for an α-agostic transition state (modified Green– Rooney mechanism) in metal catalyzed olefin polymerizations. The cis:trans or erythro:threo ratio differing from unity in Eqs. (9-14) or (9-15), respectively, is seen to stem from an H/D-isotope effect, whose basis is the expected preference of H over D in an α-agostic interaction as shown in Fig. 9-2 for the transition states. In [(C5Me5)2Hf(CH2CHMe2)(PMe3)]+ NMR data established the distortion of the isobutyl group by an α-agostic interaction [120]. α-Hydrogen and to a lesser extent also β-hydrogen participation was demonstrated in the MgX2 and MAO promoted intramolecular olefin insertion (cyclization of alkenyl ligands) of Cp2TiClR complexes [121].
(9-13)
566
9
Metallocene Catalysis for Olefin Polymerization
(9-14)
(9-15)
Figure 9-2. Proposed transition states for the hydrocyclization of trans,trans-1,6d2-1,5-hexadiene by a scandocene hydride, Me2Si(C5Me4)2Sc(PMe3)H (left, cf. Eq. (9-14)) [118] as well as the hydrodimerization of (E)- or (Z)-1-deutero-1-hexene by achiral zirconocene dichloride/MAO (right, cf. Eq. 9-15) [119] showing the slightly preferential hydrogen instead of deuterium α-agostic interaction which then gives rise to the major products as indicated. Both reactions are viewed as models for olefin insertion into a metal–alkyl bond as encountered e.g. in metal-catalyzed olefin polymerizations, thereby probing the validity of the ‘modified Green–Rooney mechanism’.
9.5
General Mechanism of Chain Growth, Chain Transfer and Deactivation
567
Scheme 9-6. Schematic representation of the suggested features for the modified Green–Rooney mechanism for the insertion of an olefin in a transition metal alkyl bond in Ziegler–Natta polymerization.
From a study on polymerizations of deuterium labeled (E)- and (Z)-propene-1-d by chiral ansa-zirconocene catalysts, direct experimental evidence has now been obtained for the role of α-agostic interactions in the formation of isotactic polypropene. Polymerizing either (E)- or (Z)-propene-1-d with rac-C2H4(IndH4)2ZrCl2/ MAO, rac-Me2Si(C5H2-2,4-Me2)2ZrCl2/MAO or rac-Me2Si(C5H2-2-Me-4tBu)2ZrCl2/MAO only slightly affects the isotacticities of the polymers formed but significant differences in the mean degree of polymerization, the chain length are evident. Polymers derived from (E)-propene-1-d have molecular masses which are higher by a factor of 1.3 than those obtained from (Z)-propene-1-d [122]. Some theoretical studies on the mechanism in Ziegler–Natta polymerizations investigated the possibility of agostic interactions. Based on the model system ‘[Cp2Ti-CH3]+ + C2H4 → [Cp2Ti-C3H7]+’ the direct insertion mechanism (Cossee– Arlman) was found viable without agostic interactions [123], whereas a study on ‘[Cl2Ti-CH3]+ + C2H4 → [Cl2Ti-C3H7]+’ found strong agostic interactions [124]. Two Extended Hückel molecular orbital studies conclude that an α-agostic H–Zr contact plays no role in a ground state species but becomes increasingly important
568
9
Metallocene Catalysis for Olefin Polymerization
along the reaction coordinate for the C–C bond-forming insertion reaction. As the Zr–Calkyl bond weakens and the zirconium center becomes more electron deficient, the α-agostic interaction stabilizes (lowers the energy of) the transition state by relieving part of the electron deficiency of the 14-valence electron species [125].
9.5.2 Theoretical Considerations on Olefin Coordination and Insertion The aforementioned ligand abstraction in a metallocene derivative Cp2MR2 creates the space and metal orbital necessary for the σ interaction with the incoming olefin (Fig. 9-3a) [126]. Because there are no d-electrons available for π backbonding from a d0 system into the empty π* orbitals of the olefin (Fig. 9-3b), the metal–olefin interaction is not stabilized and remains weak. The cis-positioned alkyl ligand (chain end) can then be transferred to the olefin or in other words the olefin inserts into the metal–alkyl bond. This dynamic olefin insertion process has been studied theoretically several times on various levels of sophistication, such as density functional theory [127] or by ab initio methods [123, 128] Figure 9-4 illustrates the ethene insertion into the Zr–Me bond of [Cp2ZrMe]+ according to MO calculations. Table 9-1 summarizes the energetics calculated by different methods and various levels of theory for the binding of ethene to the positively charged [Cp2MMe]+ cation (M = Ti, Zr) forming the π-complex (21), for the transition state (22) and for the γ-agostic primary product (23, cf. Fig. 9-4)
Figure 9-3. Schematic representation of the σ-interaction of a [Cp2MR]+ fragment (viewed from the top) with incoming olefin (a). The stabilizing π-backbonding into the empty π* orbitals of the olefin (b) is not possible due to lack of electrons in a d0 system.
9.5
General Mechanism of Chain Growth, Chain Transfer and Deactivation
569
Figure 9-4. Selected snapshots of the ethene insertion into the Zr–Me bond of [Cp2ZrMe]+ according to MO calculations [127] with bond lengths (in Å) and angles (in degree).
Table 9-1. Energetics of the ethene insertion in the Cp2MMe+-C2H4 system on various levels of theorya Cp2MMe+ cation
Method/level of theory [Ref.]b
π-Complex
Transition state
γ-Agostic product
(C5H5)2ZrMe+ (C5H5)2TiMe+ (C5H5)2TiMe+ (C5H5)2TiMe+ (C5H5)2TiMe+ H2Si(C5H4)2ZrMe+ H2Si(C5H4)2ZrMe+
nonlocal DFc SCF/SCF d,e MP2/SCF d,f MP2/MP2 d,g local DFa RHF/RHFj MP2/RHFd,j
–96 –22.3 –99.0 not existenth not existenth –80 –140
–93 +52.2 –121.7 not existenth not existenth –10 –115
–128 –90.9 –135.2 –188.9i –195.1i –100 –161
aEnergies
[127b] [128c] [128c] [128c] [128c] [128d] [128d]
given relative to Cp2MMe+ and C2H4 in kJ mol–1. further details on optimization, basis sets etc. see the original literature. cDF = density functional. dSCF = self-consistent field, MP2 = second-order Møller–Plesset perturbation. eSCF energy at the SCF equilibrium geometry. fMP2 energy at the SCF equilibrium geometry. gMP2 energy at the MP2 equilibrium geometry. hDoes not exist on the correlated level. iAdditional α- (higher energy) and β-agostic products (lower energy) besides the γ-agostic product have been calculated on the correlated level. jRestricted Hartree–Fock (RHF) energy and MP2 energy at RHF geometry, respectively. bFor
570
9
Metallocene Catalysis for Olefin Polymerization
9.5.3 Models for Cp2Zr(R)(olefin)+ Olefin complexes of the type Cp2Zr(R)(olefin)+ are assumed as intermediates in olefin polymerization processes with metallocene catalysts, but they have never been observed. In view of the weak zirconium–olefin interaction because of lack of d-π* backbonding in the d0 system (cf. Fig. 9-3) and the ready possibility for insertion into the Zr–R bond the elusive nature of such an olefin complex is understandable. On the other hand, the characterization of Cp2Zr(R)(olefin)+ would help in the understanding how a such cationic catalyst coordinates the olefin and activates it for the migratory insertion. Attempts to design model complexes for Cp2Zr(R)(olefin)+ lead to the synthesis and structural characterization of 24 [129]. Group-IV metallocene ethene complexes of the type Cp2M(RHC=CHR)PMe3 (M = Zr, Hf; R = H, Ph) are known and structurally or spectroscopically characterized [130]. Also, the structure of an ethylene-bridged zirconocene complex, {(C5H5)2ZrMe}2(µ-η2,η2-C2H4), with ethene side-on coordinated to each metal center, is available [131]. In each case the olefin is lying in the plane bisecting the angle spanned by the two cyclopentadienyl ring planes.
9.5.4 Regioselectivity in α-Olefin Insertion The mechanical properties of different polypropenes (PP) and other poly-α-olefins strongly depend on their microstructure, which is determined by the regio- and stereoselectivity (tacticity). Regioselectivity concerns which end of the α-olefinic double bond is linked to the metal atom of the catalyst and which end to the growing chain: Scheme 9-7 illustrates that if position 1 (the CH2-end) of propene is connected to the metal center and position 2 (the CHMe unit) to the chain end, then the so-called ‘1–2’ (also 1,2- or primary) insertion or addition results. This is the normal type of insertion preferred by the classical Ziegler–Natta [132] and metallocene catalysts. When position 2 of the olefin is linked to the metal center a ‘2–1’ (also 2,1- or secondary) insertion follows. Consecutive regioregular insertions result in a head-to-tail enchainment. Regioirregularities such as a 2–1-addi-
9.5
General Mechanism of Chain Growth, Chain Transfer and Deactivation
571
tion in a series of 1–2-insertion (and vice-versa) than leads head-to-head and tailto-tail structures if the secondary alkyl ligand does not isomerize to a primary ligand before the next insertion to what then becomes a 1–3 insertion. At higher temperature, isomerization or 1–3 insertion becomes more dominant and can replace all original 2–1 insertions [133]. A 2–1 insertion represents a steric hindrance to further chain growth, often leads to a β-hydride elimination or can be overcome via a β-hydride shift initiated isomerization leading to a 1–3 insertion [134]. (Another way to overcome this is to incorporate ethene as a comonomer to activate the catalysts sites blocked by a 2–1 insertion [133].)
Scheme 9-7. Regioselectivity—primary (1–2) and secondary (2–1) insertion in propene polymerization; P = growing polymer chain. After a regio-irregular 2-1 insertion, polymerization can proceed via a 1–2 addition to give head-to-head insertion and sequence or the secondary alkyl ligand can isomerize to a primary ligand before the next insertion to what then becomes a 1–3 insertion. The polyolefin molecular mass is substantially affected by a variation in transition metal concentration such that a decrease in zirconium concentration gives rise to a large increase in molar mass [39,40,81]. Two possible explanations are discussed for this not fully understood phenomenon [21]: (a) a dilution effect favoring the active complex form over the inactive precursor or dormant species in equilibrium (cf. Eqs. (9-2) and (9-16)) and hence the rate of chain propagation over that of chain termination. It is thereby assumed that chain terminations arise predominantly from dormant, i.e. temporarily inactive zirconocene species. (b) A bimolecular chain-transfer mechanism involving the active complex and a second (active or inactive) zirconocene species is proposed by Kaminsky et al. [39, 40] such that a decrease in zirconocene concentration would then reduce the rate of the termination reaction.
572
9
Metallocene Catalysis for Olefin Polymerization
9.5.5 Chain Transfer Processes, β-Agostic Interactions, β-Hydride, β-Methyl Elimination The four principle chain-transfer processes have already been shown in Scheme 93 with ethene as a monomer. For propene, a β-methyl elimination as a fifth transfer process (see below) and the type of end group formed deserves an additional attention (Scheme 9-8). Two main chain types of chain transfer processes have been observed with (chiral) metallocene catalysts: chain transfer by β-H elimination to the metal center, leading to an olefinic end group, and chain transfer to the aluminum cocatalyst, giving a saturated end group after hydrolytic workup. In low molecular mass polypropenes, an end group analysis by 1H and 13C NMR spectroscopy [238] can be used to determine the type of chain transfer process. Chain transfer by β-H elimination to the metal center can take place after a primary or secondary insertion and gives metal-hydride species which serve then as a new starting point for chain start, giving rise to n-propyl start groups (Scheme 9-8a). This monomolecular type of chain transfer is zero order in monomer concentration. Concerning the last inserted monomer unit, a preceding primary insertion leads to a vinylidene end group while a monomolecular hydrogen transfer after a secondary insertion produces vinyl and 2-butenyl end groups (Scheme 9-8b) [197].
Scheme 9-8. β-Hydrogen chain transfer processes involving polypropene (adapted from Ref. [197]). P, P′ = growing polymer chain; C* = active complex; C3 = propene.
9.5
General Mechanism of Chain Growth, Chain Transfer and Deactivation
573
Furthermore, chain transfer via β-H elimination can also be bimolecular, thereby involving hydride transfer to monomer (Scheme 9-8c and d) so that the reaction becomes first order in monomer concentration. After a 1–2 insertion this bimolecular process will produce identical end groups to the monomolecular hydrogen transfer. In case of a preceding secondary insertion, on the other hand, clearly a methylene hydride transfer to the monomer occurs, thus, giving 2-butenyl end groups. Elimination from a methyl hydrogen is not observed with bis(indenyl)-catalysts and unlikely due to steric hindrance as the growing chain would be forced towards one of the six-membered indenyl rings. The ratio between the rates of H transfer to metal and to monomer can be estimated, at a given polymerization temperature, by evaluating the polymer molecular mass dependence on propene concentration [197]. An additional transfer mechanism possible in propene polymerization is chain transfer via β-CH3 elimination (Scheme 9-9). Oligomeric products obtained with the special decamethylmetallocene catalyst system (C5Me5)2MCl2/MAO (M = Zr, Hf) [238] or [(C5Me5)2ZrMe]+[MeB(C6F5)3]– [91] were found by NMR to be mainly vinyl- and isobutyl-terminated. The vinyl to vinylidene (from β-H transfer) ratio was 92/8 for M = Zr and 98/2 for Hf. The isobutyl (chain start) group structure is produced by monomer insertion into the M–CH3 bond (Scheme 9-9). The β-alkyl elimination appears to be limited to alkyl = methyl, with 1-butene as a monomer no β-ethyl elimination was detected [238]. The preference for β-methyl versus βhydrogen elimination with (C5Me5)2M catalysts can be attributed to a lower steric hindrance in the transition state 25 versus 26, where the β-Me group or β-H have to be oriented towards the metal in the bisecting plane. The additional nonbonded interactions indicated in 26 disfavor this transition state and subsequently β-hydrogen transfer to the metal [238].
574
9
Metallocene Catalysis for Olefin Polymerization
Scheme 9-9. β-Methyl chain transfer with e.g. (C5Me5)2MCl2/MAO (M = Zr, Hf) [238]. The role of β-hydrogen transfer in the polymerization with metallocene catalysts has been studied theoretically by density functional and molecular mechanics method [135]. As a model for β-agostic interactions which may be present in the and Cp2M(R)(olefin)+ ions, the X-ray structure of Cp2ZrR+ (C5H4Me)2Zr(PMe3)CH2CH3+ is suggested, which revealed an Hβ–Zr contact of 2.16 Å [136]. In the case of α-olefins, a regioirregular or 2–1 insertion (see above, Scheme 97) leaves the active site in a deactivated state for further olefin insertion and is the preferred mode for chain termination by β-hydride elimination. One way to overcome the steric hindrance to further chain growth built up by a 2–1 insertion is a β-hydride shift initiated Zr-chain isomerization which leads to a 1–3-insertion [133].
9.5.6 Deactivation The activation of Cp2ZrCl2 by MAO involves a very fast reaction yielding the catalytically active sites. From there, the catalyst activity decays depending on temperature and catalyst characteristics. For the well investigated system (C5H5)2ZrCl2/MAO there is a very rapid initial activity decay above 40°C during the first few minutes followed by a second slow deactivation; at 0°C the activity slowly decreases over hours. It is suggested that the catalytically active complexes C* are deactivated in a two-step deactivation process by a reversible followed by an irreversible process to form inactive species I1 and I2 (Eq. (9-16)) [77b]. k1 k2 → I1 → I 2 2C* ←
(9-16)
9.5
General Mechanism of Chain Growth, Chain Transfer and Deactivation
575
At low temperature k2 is very small and in a first approximation there is only reversible deactivation. Both deactivation processes are second order with respect to the catalytically active sites. The species I1 which can recover and again take part in propene polymerization is also called ‘dormant’. The kinetic schemes did not permit the identification of the actual reaction mechanism [77]. Reduction of Zr(IV) to Zr(III) may play a role in the irreversible deactivation, since this process is accentuated in the presence of ethene, compared to ethene-free Cp2ZrCl2/MAO– system [71]. Kaminsky et al. showed that methane develops when metallocenes are reacted with MAO [137] or other aluminum alkyls [40], due to α-hydrogen transfer from a methyl group and with the formation of an intermittently catalytically inactive MCH2-Al complex. The Al/Zr ratio, temperature and type of catalyst determine the rate of gas evolution. Excess MAO than reactivates the inactive M-CH2-Al structures via ligand exchange. A similar α-hydrogen elimination with formation of methane and still slightly active Zr-µ-CH2-Zr complexes was demonstrated by Bochmann et al. in cationic zirconium alkyl complexes [138]. When one cyclopentadienyl ring is lost, as can be the case in non-ansa-metallocenes at elevated temperature under polymerization conditions, the resulting monocyclopentadienyl or half-sandwich complexes can still be active ethene polymerization catalysts [139].
9.5.7 The Problem of Comparing Activities, ‘A Caveat’ Activity in the context of olefin polymerizations with metallocene/MAO systems should always be judged as an ‘apparent activity’, since differences in dynamic activation equilibria (Eqs. (9-2) and (9-16)) of different complexes with MAO cannot be taken into account [77]. The same activity of two complexes does not necessarily imply that both have the same concentration of active species even if all conditions are identical. There does not seem to be a method available to determine the active center concentration. Radiotagging by quenching a polymerization with tritiated methanol, CH3O3H, which is occasionally advocated [53, 140] cannot exclude the dormant or inactive sites if they carry an organo-group or hydrogen atom. In the literature of metallocene catalysis often activity data for related complexes is taken from the sources of other workers. Such a comparison should be judged very carefully because of the unavoidably different polymerization conditions. The large number of obvious and less obvious parameters often renders such comparisons rather dubious. Obvious parameters which affect the activity are e.g. temperature, Al:Zr concentration, pressure, and solvent, less obvious are probably zirconocene concentration, quality of MAO (degree of oligomerization, age, TMA content) [55] and the monomer purity, pre-activation times between the zirconocene and MAO, reactor preparation, stirrer frequency etc. Carefully carried out comparative polymerization runs where all variables except for the metallocene
576
9
Metallocene Catalysis for Olefin Polymerization
catalyst were kept the same are very seldom to find, especially if they involve more than two or three different compounds [80, 81, 152]. Also, if working catalytically with very small catalyst concentrations meaningful activities are only obtained as an average from more than one experiment where consistency is observed. Not adhering to this practice and the custom to collect activity data from the literature for comparison may be the source of many of the conflicting ideas [21] in metallocene catalysis.
9.5.8 Electronic Effects in Olefin Polymerization The elucidation of ligand effects has been a central piece of work in zirconoceneMAO catalyzed polymerization of α-olefins. Most investigations into ligand effects focus on the influence of the steric environment [141] and comparatively few metallocene-polymerization studies have addressed the question how electronic changes in a ligand affect the metal center and its catalytic properties although ‘electronic’ together with steric effects have often been invoked to explain the influence of different cyclopentadienyl ligand substituents [142], however, without any more detailed explanation or proof. The fundamental problem is to single out the, if any, electronic from the concomitant steric effects [80, 81]. Piccolrovazzi et al. employed unbridged indenyl ligands carrying a hydrogen, methyl, methoxy or fluorine ligand in the 4 and 7 positions (on the annelated six-membered ring (27) [143]. A similar approach was taken by Lee et al. who used bridged and unbridged indenyl ligands on a zirconium center which were similarly substituted at the 4 and 7 or 5 and 6 positions (28) [144]. Substituents on the C6 moiety of these bis(indenyl)zirconium complexes were assumed not to interfere sterically with the incoming monomer, the growing polymer chain or with the insertion reaction at the transition metal center. As a result it was proposed that electron withdrawing groups led to a decrease in catalytic activity and polymer molar mass in the polymerization of ethene and propene, while the effect of electron donors was less clear [144]. Recent results with bis(benz[e]indene) [145] and bis(4-arylindenyl)zirconium complexes [146] (see Sec. 9.7), as well as with rac-C2H4(Ind)ZrCl2 having methyl substituents in the 4 and 7 position of the indenyl ring [147] demonstrated, however, a marked steric influence of the substituents on the catalyst’s propene polymerization behavior. Möhring and Coville recently reported a quantification of steric and electronic parameters by the use of cone angles and Hammett functions in the ethene polymerization with (C5H4R)2ZrCl2-ethylalumoxane catalysts [56]. Cone angles, NMRspectroscopic and structural parameters were also used by Möhring et al. to separate the steric and electronic cyclopentadienyl substituent effects of (C5H4R)2TiCl2 catalysts with Et3Al2Cl3 as an activator in the polymerization of ethene [148].
9.6 Ethene Homopolymerization
577
In accordance with the experimental studies a large number of theoretical calculations have focused on the influence of the steric environment at the metal center (see Sec. 9.7), however, we are only aware of one theoretical investigation dealing with electronic effects in alkene polymerizations. There, an attempt was made to substantiate the electronic effects deduced with the aforementioned substituted bis(indenyl)zirconium complexes by AM1 calculations on the free ligands [149].
9.6 Ethene Homopolymerization In comparison to high-density polyethene (HDPE) from classical Ziegler–Natta catalysts, the so-called metallocene HDPEs are characterized by their low polydispersity and the total absence of branching, which in turn leads to a different rheological behavior [150]. Ethene polymerization studies are often used to investigate basic effects like the influence of catalyst concentration, cocatalyst characteristics and other aluminum alkyls, Al:Zr ratio, temperature, steric effects [151], catalyst formulation on support materials [16], etc. on the catalyst activity and the polymer parameters such as molar mass and molar mass distribution. When comparing different metallocenes ethene may in some cases be better suited as a monomer, since with propene or other α-olefins the stereoselectivity of the metallocene strongly affects the activity [152–154]. The use of ethene and the formation of insoluble polyethene creates, however, problems which one has to be aware of when comparing activities. In an ethene solution polymerization, the C–C bond forming reaction is truly homogeneous only at the very beginning. With the precipitation of polyethene the active complex becomes more and more embedded in the swollen, gel-like polymer matrix, which represents a transfer to the heterogeneous phase [14], i.e. a heterogeneous active complex form and this leads to a diffusion-controlled reaction. The reaction rate is then controlled by the rate of diffusion of the monomer through the polymer matrix to the enclosed active center [35, 74, 155] and can no longer be compared in other terms [80, 81]. In such a diffusion controlled polymerization the
578
9
Metallocene Catalysis for Olefin Polymerization
activities of different catalyst may then appear very similar, although in reality they are not.
Figure 9-5. Activity profiles (based on ethene uptake over time at constant ethene pressure) in the ethene polymerization with three zirconocene complexes activated with MAO at a higher (top) and lower (bottom) zirconium concentration (adapted from Ref. [81]). [1 = (C5H5)2ZrCl2/MAO, 2 = (C5HMe4)(C5H5)ZrCl2/MAO, 3 = (C5HMe4)2ZrCl2/MAO] The overlapping curves of the activity profiles and the resulting similar activities for the complexes at the ‘high’ zirconocene concentration of 10–5 M are due to the rapid polyethene precipitation at these conditions. The corresponding activity profiles at the low concentration of 2.5 × 10–7 M show a more constant ethene consumption. The smaller amount of polyethene produced under these conditions ensures the existence of a homogeneous phase over an extended time period. Note that with the lowering of the zirconocene concentration, the Al/ Zr ratio had to be increased considerably to offset a higher degree of dissociation upon dilution of the catalytically active zirconocene–MAO complex.
9.6 Ethene Homopolymerization
579
A way out of the problem is to compare the time–activity profiles and to decrease the zirconocene concentration so as to lower the absolute mass of polyethene formed and to mostly remain in the homogeneous reaction regime (Fig. 95). This problem of diffusion control adds to the difficulty in comparing polymerization activities from different literature sources. In ethene polymerizations with zirconocene–MAO catalysts the activity under comparable, mostly homogeneous conditions is primarily a function of steric factors. The order of activity follows the order of steric demand. This was demonstrated for a series of methyl-substituted zirconocene pre-catalysts [81] where the steric demand was quantified through the coordination gap aperture [156] (Fig. 9-6). The polyethene molar mass from metallocene catalysis can be controlled by the metallocene type and concentration, the reaction temperature or by addition of hydrogen through the hydrogenolysis reaction (cf. Scheme 9-3) [157, 158].
Figure 9-6. Correlation between ethene polymerization activity and the coordination gap aperture (cga) for different methyl-substituted zirconocenes (adapted from Ref. [81]). The coordination gap aperture (left) is a geometric parameter which is defined as ‘the largest possible angle spanned by those two planes through the metal center which touch the inner van der Waals outline of each of the C5-ring ligands’ [156]. For a practical determination of the coordination gap aperture also for unsymmetrically substituted ligands, such as C5HMe4 see Ref. [81]. [1 = (C5H5)2ZrCl2, 2 = (C5HMe4)(C5H5)ZrCl2, 3 = (C5HMe4)2ZrCl2, 4 = (C4Me4P)(C5H5)ZrCl2, 5 = (C4Me4P)2ZrCl2, 6 = (C5Me5)(C5H5)ZrCl2, 7 = (C5Me5)2ZrCl2, each activated with MAO under identical conditions].
580
9
Metallocene Catalysis for Olefin Polymerization
9.7 Propene Homopolymerization 9.7.1 Polypropene Microstructures The three main stereoselective variations, namely isotactic, syndiotactic and atactic PP, and their usual comb-type presentation are depicted in Fig. 9-7. Achiral metallocenes mostly give atactic polypropene which almost always has a low molar mass and is an oily or waxy material (see Sec. 9.9). Chiral, C2-symmetric metallocene/MAO catalysts afford isotactic polypropene. The first such metallocenes used in propene polymerization were the ethanediyl-bridged bis(indenyl)titanium dichloride [43, 44], and bis(tetrahydroindenyl)zirconium dichloride (6,7) [45, 159]. The titanium compound was, however, a 56/44-mixture of the chiral racemic and achiral meso isomers (6a), hence, the polymer was 63% isotactic and 37% atactic [43]. The zirconium derivative, on the other hand, was applied only in its chiral form and subsequently gave isotactic to highly isotactic PP [159]. These first reports on the stereochemical control in polypropene formation appeared in 1984 and 1985, respectively, and they initiated the massive interest by polymer and organometallic chemists in metallocene catalysis, which is still seen now over 10 years later. Polypropene is technically the more important polymer over polyethene and changes in the steric ligand situation allow for an unprecedented control of the polymer microstructure. Besides the iso- and atactic mode, other polypropene microstructures, such as iso-block, stereo-block, isotactic-atactic-block, hemiisotactic and especially syndiotactic PP can be obtained, as well, upon ligand modification. Figure 9-8 gives a correlation between the pre-catalyst and PP-microstructure, illustrating the different degree of stereochemical control. The microstructure can also be greatly affected by temperature. At low enough temperatures, e.g. –45°C for (C5H5)2TiPh2 [43], even achiral catalysts are capable of polymerizing propene with stereocontrol, to give e.g. a stereo-block PP (cf. Fig. 9-8).
9.7 Propene-Homopolymerization
581
Figure 9-7. Schematic illustration of (a) isotactic polypropene (PP) with all methyl groups on the same side of the carbon chain back bone; (b) syndiotactic PP with the methyl groups alternating and (c) atactic PP with an irregular, statistic methyl distribution. At the right side the usual short-hand notation is given. Note that these variations in microstructure are general for vinylpolymers. Also, for pedagogic reasons the polymer is drawn here as a stretched chain whereas in reality a helical structure exists, with equal amounts of right- and left-handed helices in the lattice [190].
582
9
Metallocene Catalysis for Olefin Polymerization
Figure 9-8. Correlation between the pre-catalyst and PP-microstructure, illustrating the different degree of stereochemical control. Refs.: (a) [191], (b) [163, 164], (c) [43], (d) [179], (e) [167], (f) [180, 182]. The term hemiisotactic means that the insertion of every second (even) propene unit is sterically controlled while the intermediate (odd) insertions occur statistically.
9.7 Propene-Homopolymerization
583
All of the different forms of homopolypropenes sketched in Fig. 9-8 have different mechanical properties (without the addition of comonomer) ranging from plastic to elastic PP with decreasing stereoregularity. Metallocene catalysts opened up the way to influence the polypropene properties in an understood and controlled way by altering the ligand system of the metal complexes. This leads to new products with application properties not available with conventional Ziegler–Natta catalysts [4].
9.7.2 Microstructure Analysis by NMR Microstructure Analysis of vinyl polymers can be done by 13C NMR spectroscopy; for polypropene the methyl region is used. Spectrometers above 200 MHz (1H, respectively 50 MHz for 13C) allow at least for the observation of pentad regions, i.e. can discriminate the chemical shift of a methyl group depending on the relative position of the two methyl neighbors to the right and to the left in the polymer chain. A 150 MHz 13C NMR characterization meanwhile allowed for the expansion from the pentad to the nonad level [160]. A special NMR notation is used here for the relative configuration of neighboring centers in vinyl polymers [161]: If the pseudoasymmetric tertiary methine carbon atoms of two neighboring monomer units have the same configuration it is called a meso (m) diad (29), if they are of opposite configuration it is a racemic (r) diad (30). The NMR notation for ideal isotactic PP would then be ‘... mmmmmm ...’ for ideal syndiotactic PP ‘ ... rrrrrrr ...’. An example of a spectrum of atactic polypropene and the assignment of the NMR bands to the ten unique steric arrangements of five adjacent monomer units is shown in Fig. 9-9. The spectrum of ideal isotactic or syndiotactic PP would show only one band corresponding to the mmmm or rrrr pentad, respectively. Interesting are, of course, spectra with more than one and less than nine lines.
584
9
Metallocene Catalysis for Olefin Polymerization
Figure 9-9. 13C NMR spectrum (50.3 MHz) of the methyl pentad region for atactic polypropene obtained with (C5Me5)(C5H5)ZrCl2/MAO at –30 °C, together with the assignment of the nine bands to the ten unique pentad steric arrangements. Note the three groups of signals belonging to the mm, mr and rr centered pentad (spectrum adapted from Ref. [43]). The following examples illustrate how 13C NMR spectroscopy can be used to draw conclusions on the polymerization, on the structure of the transition state and the mechanism of stereochemical control. An example of a three line NMR spectrum of a polypropene sample is shown in Fig. 9-10. Besides the mmmm pentad, the mmmr and the mmrm (+rmrr) pentad are present in a 1:1 ratio. From this the polymer microstructure of a stereo-block PP can be deduced. It is remarkable that such a stereocontrol was achieved with an achiral metallocene precatalyst. It can be concluded further that the chain-end must be the stereoregulating factor, i.e. the chirality of the methine carbon of the last inserted monomer unit provides stereochemical control during the polymerization. An occasional occurring stereo error (r diad)
9.7 Propene-Homopolymerization
585
in the insertion results in the new absolute configuration of the last inserted monomer unit which then controls the succeeding enchainment (m diads) until the next steric inversion [43].
Figure 9-10. 13C NMR spectrum (50.3 MHz) of the methyl pentad region for atactic PP (top) and for polypropene obtained with (C5H5)2TiPh2/MAO at –45 °C (bottom), together with the assignment (spectra adapted from Ref. [43]).
NMR is useful not only in the determination of steric sequence distributions but also in the detection of end groups and structural defects such as regiochemical insertion errors if they are present in amounts greater than 1%.
9.7.3 Stereochemical Control Stereoregulation in the stereoselective polymerization of α-olefins can either be achieved through interaction between the entering monomer and the growing chain (chain-end control) or through an interaction between the monomer and the metal center and its environment (enantiomorphic-site control) or both [132, 162]. If a chiral metal center provides for the stereocontrol then an occurring stereo error would remain isolated, corrected immediately and result in pentads of the type mmmr/mrrm/mmrr, corresponding to an iso-block polypropene, as obtained with the rac-C2H4(Ind)2ZrCl2/MAO system (cf Fig. 9-8) [132, 163, 164]. An example of
586
9
Metallocene Catalysis for Olefin Polymerization
a twofold stereodifferentiation was given with the unbridged chiral (C5H4CHMePh)2ZrCl2/MAO (31) and with other conformationally unrestricted (C5H4CHR1R2)2ZrCl2/MAO catalysts [165]. The quantitative statistical interpretation of the methyl carbon signal indicates that the isotactic polypropene obtained with 31 consists to 35% of an iso-block PP (enantiomorphic-site control) and to 65% of a stereo-block PP (chain-end control) [164].
The polymerization of propene to an isotactic polymer remains a unique example of stereoselectivity in organic, nonenzymatic reactions [132]. How is the stereochemical control carried out? This will be outlined in the following and with the help of Fig. 9-11. As pointed out above, polymerization of propene (and other αolefins) takes place (predominantly) by a regioselective 1–2 insertion into the metal–carbon bond. Before insertion, the propene molecule forms a weak π-complex in the 1–2 direction with the transition metal center. This coordination can be done in two prochiral positions (Fig. 9-11a and b). If one of these is energetically favored, then the polymerization will be stereoselective. The spatial arrangement of the bridged metallocene is such that position a (Fig. 9-11) is energetically favored because of less steric interaction with the neighboring six-membered ring of the indenyl system and with the chain end. Insertion then takes place and the next insertion step (c) starts with chain and monomer having exchanged position with each other. For the isotactic polymerization such a chain migration is no prerequisite because even without it the meso stereosequences, i.e. isospecific behavior would result, but the exchange of positions follows from the insertion model presented in Fig. 9-4 and strong evidence is especially provided by the syndiotactic polymerization behavior of the Cs-symmetrical metallocene catalysts (see below) [166].
9.7 Propene-Homopolymerization
587
Figure 9-11. Model of the active complex for one enantiomeric form of the isospecific, C2-symmetrical metallocene C2H4(Ind)2ZrCl2 with the two possible prochiral positions (a and b) of the 1,2-coordinated propene molecule. Position (a) is the energetically favored one. The steric demand of the indenyl ligand forces the growing chain and the propene methyl group into an orientation in the fourth and second quadrant, respectively, so that a meso (m) stereosequences results (c). The disfavored insertion from position (b) then leads to a stereoerror, a racemic stereosequence (d). Because of the enantiomorphic site control of the catalyst this error can be corrected upon the next insertion (d→e). Note that the respective next insertion step starts with chain and monomer having exchanged positions with each other. In reality the chain flip between ‘left and right’ involves only a small relative motion between chain end and metallocene. When the insertion occasionally takes place via the other prochiral position (b in Fig. 9-11) a racemic diade is obtained (b→d). If the catalysts shows high isospecific behavior and in case of an enantiomorphic site control this stereoerror is immediately corrected (giving a second racemic diade (d→e) and remains isolated. The resulting mrrm (mmrr) pentad sequences are then typical for enantiomorphic site control (see above). The model for the stereochemical induction presented above and in Fig. 9-11 can explain the relation between structural variation of the chiral metallocenes 32–35 and their polymerization results given in Table 9-2, which in turn is a test of the model.
588
9
Metallocene Catalysis for Olefin Polymerization
Table 9-2. Polymerization results of structurally varied bridged metallocenes (data taken from Ref. [166]) Metallocene
Activity
32 33 34 35
115 27 33 6
Me2Si(Ind)2ZrCl2 Me2C(Ind)2ZrCl2 Me2Si(3-MeInd)2ZrCl2 Me2Si(C5H4)(Ind)ZrCl2
a
Mwb × 103
niso
35 15 28 17
50.1 11.5 2.1 4.6
aActivity
in kg polypropene mmol–1 metallocene. bWeight average molecular mass (in g mol–1) of polymer determined by GPC; n iso = average isotactic block length by 13C NMR spectroscopy.
Taking the dimethylsilyl-bridged bis(indenyl) compound 32 as a reference point, then the shortening of the bridge through the introduction of the isopropyl bridge in 33 widens the angle and hence the space between the indenyl ligands. More space means a reduced steric interaction between the ligands and monomer and the chain end. Thus, the energetic difference between the two prochiral positions of the propene monomer (cf. a and b in Fig. 9-11) is decreased so that more stereoerrors occur and a reduction in stereoselectivity is observed. Having methyl substituents on the 3-position of the five-membered ring as in 34 eliminates the energetic difference between (a) and (b) (Fig. 9-11) completely and therefore leads to a total loss of stereoselectiveity. In the mixed indenyl–cyclopentadienyl metallocene 35, a
9.7 Propene-Homopolymerization
589
larger energetic difference between the prochiral positions only exists upon monomer approach on the side of the six-membered ring. Approach from the other side can then only be controlled through interactions with the chain end [166].
9.7.4 Non-C2 Symmetrical or Unsymmetrical ansa-Metallocene Complexes Non-C2 symmetrical or unsymmetrical ansa-metallocene complexes with the cyclopentadienyl-indenyl backbone (36) have also been advocated as stereoregular propene polymerization catalysts. Most of these precursors when activated with MAO produce homopolypropenes with excellent thermoplastic elastomeric properties attributable to hemiisotactic microstructures (cf. Fig. 9-8) [167]. The term hemiisotactic means that the insertion of every second (even) propene unit is sterically controlled while the intermediate (odd) insertions occur statistically. Based on the asymmetric Me2C(C5H4)(Ind)ZrCl2 complex experimental and modeling studies were able to describe such polymer structures [168]. The microstructure of the elastomeric polypropenes produced with these unsymmetrical catalysts is (as expected) sensitive to changes in catalyst structure and temperature but also to changes in pressure [199]. With the appropriate bulky substitution, as shown in 37, the polymerization of propene was found to afford highly isotactic polymers of medium to high molecular weight with %mm > 99 [169]. Ph2C(C5H4)(Ind)ZrCl2 was a tailored metallocene for the copolymerization of ethene with sterically demanding cycloolefins (see below) [170].
9.7.5 Substituted Unbridged Metallocenes Having bulky alkyl groups attached to the unbridged cyclopentadienyl or indenyl ligands in Cp2ZrCl2 led mostly only to atactic and low molecular mass polypropenes (oligomers; see Sec. 9.9) at conventional polymerization temperatures (above 50°C) even though the ring rotation is probably hindered [154, 171]. Chiral auxiliaries, such as cholestanyl [172], neoisopinocamphyl [173], neomenthyl and neoisomenthyl [174] were attached to unbridged indenyl and tetrahydroindenyl ligands.
590
9
Metallocene Catalysis for Olefin Polymerization
Only at low temperature the resulting zirconocene complexes (activated by MAO) produce high molecular weight polypropene with enantiomorphic site control. Even from a dimethylsilyl-bridged bis(cyclopentadienyl) complex, where one of the cyclopentadienyl rings carries a neomenthyl group as a chiral auxiliary, the products were of low molecular mass and only at low temperature (0 to –45°C) was the isotacticity (by mmmm pentad) greater than 90% [175]. The neoisomenthyl-substituted systems were found to be much more stereoselective (up to 77% mmmm pentad) than the isomeric neomenthyl-substituted systems (maximal 37% mmmm pentad). It was assumed that the latter switch back and forth between C2and C1-symmetric conformations of the metallocene backbone during catalysis, thereby giving rise to alternating isotactic and near atactic sequences along the growing polymer chain (cf. to oscillating catalysts below). The role of torsional isomers in these planarly chiral nonbridged bis(indenyl)metal complexes has been investigated by solid state and solution (NMR) structural studies [174]. With the appropriate ligands unbridged group IV metallocene catalysts could, however, polymerize propene to crystalline isotactic polypropene even at conventional polymerization temperatures. Bis(1-methylfluorenyl)zirconium dichloride (38) after activation with MAO gave at 60°C a polymer which was 83% isotactic (mmmm pentade). The activity was 1200 g PP/g metallocene in liquid propene. An analysis of the polymer microstructure showed mrrm/mmrr sequences which are consistent with an enantiomorphic-site stereochemical control mechanism (see above) [176]. It is noteworthy here, that a bridged bis(fluorenyl)zirconium complex gave predominantly atactic polypropene, albeit of high molecular mass [177]. When a phenyl group was placed on the C5 ring (the 9-position) in 38 no propene and only low ethene polymerization activity was observed due to blocking of the insertion path by the bulky 9-substituent [178].
9.7.6 Isotactic–Atactic Block Polymerization, Oscillating Catalysts The concept is a further development of the use of unbridged bis(indenyl) and (tetrahydroindenyl) ligands by Erker et al. with bulky chiral groups such as
9.7 Propene-Homopolymerization
591
cholestanyl [172], neoisopinocamphyl [173], neomenthyl and neoisomenthyl [174] attached, so that torsional isomers become possible. New efforts by Coates and Waymouth concentrated on phenyl substituted indenyl compounds [179]. Figure 912 shows the essentials of the concept of oscillating catalysts. The phenyl-substituted indenyl rings are free to rotate or oscillate about the metal–ring bond and subsequently lead to the rac and meso rotamers, which consequently exercise a different stereochemical control in propene insertion. The enantiomeric rac rotamer gives the isotactic block sequences, the achiral meso form leads to the atactic polypropene. The block length decreases with increasing temperature and the isotactic block lengthens with increasing pressure.
Figure 9-12. Principle of isotactic–atactic stereoblock polypropene formation by oscillating catalysts.
9.7.7 Syndiotactic Polymerization Cs-symmetric catalysts of the type 39 where the metal carries a bridged cyclopentadienyl and fluorenyl (Flu) ligand, yield syndiotactic PP with some isotactic triads (mm) as defects. The first report on this remarkable new syndiospecific polymerization with hafnium and zirconium metallocenes appeared in 1988, up to this time syndiotactic PP could only be obtained in vanadium-catalyzed polymerizations below –50°C [180]. (For a morphological study on syndiotactic PP prepared by ansametallocenes, see Ref. [181].) Figure 9-13 illustrates the syndiotactic propagation with such a Cs-symmetric cyclopentadienyl–fluorenyl catalyst. Strong evidence for the chain migration with every insertion step (cf. Fig. 9-11) is provided by the syndiotactic polymerization behavior of the Cs-symmetrical catalysts such as 39. For the syndiotactic polymerization an exchange of chain and monomer positions is a
592
9
Metallocene Catalysis for Olefin Polymerization
prerequisite because without it meso stereosequences, i.e. an isospecific behavior would be expected [166]. Seemingly small variations in the ligand structure have a significant effect on the stereoselectiveity [182]: Note that the closely related silylene bridged derivative Me2Si(C5H4)(Flu)ZrCl2 when activated with MAO catalyzed nonstereoselective propene polymerization without any syndiotactic tendency [183, 184]. An additional methyl substituent on the cyclopentadienyl ring in 39 (in β position relative to the bridge) gives a complex that polymerizes propene in a distereoselective manner to give syndiotactic–isotactic stereoblock PP [185], a tertbutyl ligand gives an isotactic-specific metallocene complex [186]. Substituents on the sterically little active positions 2 and 7 of the fluorenyl moiety (cf. numbering in 39) do not affect the syndiospecificity of the catalyst but have large effects on the activity [187]. With no bridge present in 39, or ligand substituted derivatives thereof, only atactic propene is formed, believed to arise because the Cs symmetry of the catalyst, necessary for syndiotactic polymerization, is not present anymore owing to the unhindered rotation or oscillation of the fluorenyl ligand [188]. The influence of substituents on the bridgehead atom in RR′C(C5H4)(Flu)ZrCl2/MAO on the propene polymerization activity has been studied and found to increase in the order MeMe ≈ MePh ≈ (CH2)4 < HH < HPh < (CH2)5 < PhPh. The PP molecular mass is much higher (by a factor of five or six) for RR′ = MePh or PhPh [189].
9.7 Propene-Homopolymerization
593
Figure 9-13. Principle of syndiotactic propagation with a Cs-symmetric metallocene catalyst (here Me2C(Flu)(C5H5)ZrCl2/MAO) [180, 182]. The chain or metal moiety swings back and forth upon olefin insertion and the olefin approaches from alternating sides (as already illustrated for the isotactic insertion, cf. Fig. 9-11), but the steric demand of the fluorenyl ligand forces the growing chain and the propene methyl group into an orientation in the first and second quadrant, respectively, so that a racemic (r) stereosequences result. In reality the chain flip between ‘left and right’ involves only a small relative motion between chain end and metallocene.
594
9
Metallocene Catalysis for Olefin Polymerization
9.7.8 Enantiomer Separation and Enantioselective Polymerization Since the propene polymerization is usually carried out only in a stereoselective and not an enantioselective manner, it suffices to use the racemic mixture of the pre-catalyst. The enantiomers differ only in that they form polypropene chains of different (left or right) helical handedness [190]. Enantiomer separation of the precatalyst was shown to be possible by Brintzinger et al. using the enantioselective reaction of the (S,S) form of racemic C2H4(IndH4)2TiCl2 (cf. 7) with 0.5 equivalents of (S)-(–)-1,1′-bi-2-naphthol in the presence of an excess of sodium metal to give the (S,S)-binaphtholate derivative (cf. 50 in Sec. 9.12). After separation by chromatography the pure diastereomer is cleaved by treatment with gaseous HCl to return to C2H4(IndH4)2TiCl2, yet solely in its (S,S)-form [44]. Similarly the racemic mixture of the zirconium derivative was resolved and the pure (S,S) zirconocene enantiomer was employed in propene polymerization. In the resulting polymer the atactic parts were decreased and the isotactic sequence length raised further compared to the racemic catalyst mixture. Furthermore, optically active polypropene could be obtained for the first time under certain conditions (no dissolution or melting of the sample) [40, 191]. Using this enantiomer for the oligomerization of 1-butene yielded the optically active dimer 3-methylheptane and the trimer 3-methyl-5-ethylnonane in an optical purity of 70–80% [192]. The (S,S) form of C2H4(IndH4)2ZrCl2/MAO was also used in the copolymerization of cyclopentene with allyltrimethylsilane; the copolymer obtained shows a small optical rotation which may result from the cyclopentene units only [193].
9.7.9 Comparative Polymerization Comparative Polymerizations involving propene with a larger series of metallocene derivatives are again rare. An illustrative example to the contrary was provided by Kaminsky, Spaleck et al. who compared the catalytic activity of two unbridged and nine ansa-metallocenes together with the resulting polymer properties using a constant set of polymerization parameters [152]. Selected data of this study is provided in Table 9-3.
9.7 Propene-Homopolymerization
595
Table 9.3. Activities of a series of eleven catalysts in polymerization of propene, statistical isotacticities I, fraction of heterotactic triads H, and viscosity-averaged molecular weight Mη (data taken from Ref. [152]) Catalyst
Activity
A, rac-Me2Si(Ind)2ZrCl2 B, rac-Ph2Si(Ind)2ZrCl2 C, rac-(PhCH2)2Si(Ind)2ZrCl2 D, rac-Me2C(C5H3Me)(Ind)ZrCl2 E, rac-Me2C(C5H4)(Ind)ZrCl2 F, rac-C2H4(Ind)2ZrCl2 G, rac-C2H4(Ind)2HfCl2 H, rac-C2H4(IndH4)2ZrCl2 I, Me2C(C5H4)(Flu)ZrCl2 J, (C5H5)2ZrCl2 K, (C5H4Nm)2ZrCl2 e
1940 2160 270 400 180 1690 610 1220 1550 140 170
a
Mη × 103
I
79 90 72 4 3 32 450 24 159 2 3
0.97 0.97 0.95 0.54 0.49 0.95 0.94 0.98 0.94d 0.56 0.69
b
H
c
0.04 0.05 0.02 0.49 0.54 0.09 0.08 0.02 0.10 0.75 0.52
aActivity
in kg polypropene mol–1 metallocene and per h, standardized to equal monomer concentration. bIsotacticity I = (mm) + 0.5 (mr). cHeterotacticity H = (mr). dSyndiotacticity S = (rr) + 0.5 (mr); with (mm), (rr) and (mr) representing isotactic, syndiotactic and heterotactic triad fractions. eNm = neomenthyl.
The data in Table 9-3 indicates that the chiral bis(indenyl) systems are highly isospecific and the Cs-symmetric fluorenyl–cyclopentadienyl system (I) is highly syndiospecific (see above). The compounds D and E are chiral but lack C2-symmetry (see above), hence are of a reduced stereoselectivity. The unbridged species K with its bulky substituents produces stereo-block PP similar to the system 31 described earlier. The complexes D, E, J, and K, which lack a high degree of stereoselectivity, show poor activity and give only low molar mass or oligomeric materials (see Sec. 9.9). From this it is deduced that with the prochiral monomer propene, the stereoselective polymerization mode is strongly favored over the aspecific one, thereby clearly overcompensating steric influences coming from metallocene structure (cf. also the ‘advanced metallocenes’, discussed below). Hafnocenes are known to be less active than their zirconocene analogs but produce much higher molar mass polymers (G versus F). Similarly, syndiospecific catalysts (I) lead to higher M values compared with isospecific analogs [152]. It is important to note that C2-symmetry is necessary but not sufficient for high isospecificity and that activity and stereoselectivity are directed by the same fac-
596
9
Metallocene Catalysis for Olefin Polymerization
tors in propene polymerization with stereorigid metallocenes [194]. The data in Tables 9-2 and 9-3 illustrate that a reduction of stereoselectivity is accompanied by a reduction of catalyst activity towards propene and lowering of the polymer molecular mass [166].
9.7.10 Stereoregularity – Pressure-Dependence and Chain-End Epimerization For propene polymerization besides the obvious variables such as temperature, type of catalyst, Al/Zr concentration and the like, it was demonstrated that a decreasing monomer concentration (below 2 mol/l) at a given temperature lead to a lower stereoselectivity of the polypropene formed by Cs- and C2-symmetric ansametallocene catalysts [195, 195], eventually yielding atactic propene oligomers when the monomer concentration approaches zero. It is suggested that this effect of monomer concentration on propene polymerization parameters is the primary cause for many of the conflicting literature data on propene polymerization [197]. As a possible explanation a slow reaction of epimerization was suggested to compete with the monomer insertion. The change of configuration (or epimerization) at the last inserted unit might proceeds through a β-hydrogen elimination with the olefin remaining coordinated, olefin rotation and reinsertion into the Zr–H bond to give a secondary alkyl complex, followed by another hydrogen elimination (from one of the methyl groups) and olefin reinsertion to return to the primary alkyl chain but now with a racemized end group [196]. Thus, besides propene insertions with the wrong enantioface, the isomerization of the growing chain end can give rise to stereoerrors. This was recently supported by polymerization experiments with isotope labeled propene-1-d [122] and propene-2-d [198] which found that most stereoirregular monomeric units in the isotactic polypropene obtained by C2symmetric catalysts are deuterated at the methyl group. With unsymmetrical ansa-catalysts (36 and 40) it was, however, observed that an increase in monomer concentration (pressure) mostly led to a decrease in stereoregularity of the polypropene produced [199, 200]. This dependence is discussed on the basis of differently chain-ligated species for the catalyst type 40; if the chain preferentially occupies the open space (as indicated) then insertion of the propene monomer through the aspecific site between the two six-membered rings will be more important at high pressure and lead to more stereoerrors in the chain [200]. (Consequently, it should be noted that in another study the type 36 unsymmetric catalyst did not show a dependence in stereoregularity upon pressure [200].)
9.7 Propene-Homopolymerization
597
9.7.11 Bridge Modifications Studies also involved changes in the bridge, e. g. substituting the hydrogen atoms in the ethanediyl or the methyl groups in the dimethylsilyl bridge. One approach involved the introduction of additional stereocenters in the bridge, either by an unsymmetrical ethanediyl substitution pattern [200] or by converting Me2Si to a di(menthoxy)silylene, a silolene or a (1,3-propanediyl)silylene bridge [201]. The idea behind substituting the bridge moiety is to influence the ethanediyl bridge conformation, allow for the isolation of a single diasteromer and to effect polymerization parameters. The effect on the stereoselectivity is, however, much less pronounced compared to substitutions directly on the cyclopentadienyl or indenyl moiety. Bigger effects for the bridge-modified derivatives were observed on molecular mass, tolerance of catalyst and the deactivation process [201]. The bridge atom was also varied from silicon to germanium [184,202] to tin [203] and the polymerization behavior of the resulting complexes investigated. A decrease in tacticity is observed going form Si to Ge [184]. The activity for the Si and Sn bridged species in ethene polymerization was found comparable, yet with Sn the rate of chain transfer decreased significantly [203].
9.7.12 Molecular Mass Enhancement, Advanced Metallocenes For a long time it was a serious drawback that even chiral zirconocene/MAO catalysts produced only low molar mass oligomers at conventional polymerization temperatures above 60°C, where the activity was high (see also Sec. 9.9, olefin oligomerization) [40, 159, 204]. The reason for obtaining oligomeric or low molar mass polymeric products is due to a larger increase of the rate of chain transfer (kT, mainly by β-hydrogen elimination) versus the rate of chain propagation (kP) with the increase in temperature. The mean degree of polymerization, Pn, which is proportional to the number average molecular mass Mn is given by the ratio of the growth rate to the rate of transfer, Mn ~ Pn ≈ kP/kT. First, this disadvantage was tried to counter by using chiral hafnium metallocene catalysts, such as rac-C2H4(Ind)2HfCl2 and rac-C2H4(IndH4)2HfCl2, activat-
598
9
Metallocene Catalysis for Olefin Polymerization
ed with MAO. These hafnium catalysts provided high molar mass polypropene which was also slightly more stereoregular than those obtained with zirconium [205]. However, their yields remained too low in comparison to zirconium catalysts. A clue for directed modifications to block the reaction path and decrease the rate of chain transfer was seen in comparative propene polymerizations involving chiral, 2,4-alkyl-substituted cyclopentadienyl ansa-zirconocenes (41). With R1 = tBu longer chains were obtained than with R1 = iPr. This was traced to the increased steric repulsion between the more bulky tBu group and the chain end upon formation of the β-H-Zr bridge necessary for chain transfer [208]. It was soon recognized that the additional presence of methyl substituents on the five-membered ring in ansa-bis(indenyl) ligands and their position was highly influential in optimizing not only the tacticity but also the polymer molar mass at elevated temperature. The C2-symmetric metallocenes bearing methyl groups at the 2-(α-)position (adjacent to the bridge) (42) gave polymers in higher yield and of higher stereoselectivity [206]. The same was already found earlier for ansabis(cyclopentadienyl) systems [207, 208]. The remarkable effect of the α-methyl substituents may be traced to the suppression of 2–1- or tail-to-tail (mis)insertions (cf. Scheme 9-7). Without the α-methyl group such misinsertion can be found with a magnitude of 1.5%, whereas in the α-methyl substituted complexes (just as in the classical heterogeneous Ziegler–Natta catalysts) these regio-irregularities are suppressed close to or below the detection limit [208]. On the other hand, the methyl group is turned away from catalytically relevant space, so that a direct steric effect on chain transfer and subsequently molecular mass of the polymer may not necessarily follow. Instead, an electron donating effect was proposed, which decreases the Lewis acidity on the cationic center, thereby reducing the tendency for β-hydrogen abstraction, so that the number of chain-transfer reaction is reduced and the molar mass increases [206]. In addition, the increase in stereoselectivity brought about by the α-methyl substituents in chiral metallocene catalysts is connected to a decreased tendency of these α-substituted complexes to undergo thermally induced excursions from C2axial symmetry [209]. To a degree which depends on their substitution pattern ansa-metallocene complexes are fairly easily deformable. The bridged ligand framework can rotate up to ±20° about the metal–C5–ring–centroid axes between two shallow energy minima leading to off-(C2-)axially distorted conformers. Figure 9-14 sketches these two conformations for which an energy difference of 0.95 kcal/mol was deduced from temperature-variable NMR, with the Y (indenyl-backward) conformation suggested to be the ‘higher’ energy conformer [210]. An unrestricted fluctuation in solution between these two conformations has implications for structure–function correlations in olefin polymerization [209]. These studies apparently have led to the synthesis of doubly bridged zirconocenes, e.g. 43. According to one study, intact 43 does not appear to possess catalytic activity but in combination with MAO is slowly degraded to give a catalytically active species for
9.7 Propene-Homopolymerization
599
propene polymerization [211], while another study describes the ethene and propene polymerization activity (atactic oligomers with propene) and does not mention a degradation [212]. A theoretical study finds doubly-bridged ansa-zirconocenes based on the norbornadiene skeleton with suitable substitution highly stereoselective [213]. Doubly bridged titanocene and zirconocene complexes containing two ethanediyl (–C2H4–) bridges have also been described [214], but we are not aware of a report on their polymerization activity.
600
9
Metallocene Catalysis for Olefin Polymerization
Figure 9-14. Indenyl-torsional conformations for the C2H4(Ind)2ZrCl2 complex (ZrCl2 is omitted for clarity) with the notation used in the literature [210]. Eventually the problem of oligomeric molecular mass of polypropene at conventional reaction temperatures of 50°C and higher seemed to have been finally overcome with the introduction of ansa-zirconocene complexes carrying additional aromatic substituents on the six-membered indenyl moiety, either annelated as in Me2Si(2-Me-benz[e]indenyl)2ZrCl2 (44) [145] or sigma-bonded in appropriate positions as in Me2Si(2-Me-4-phenylindenyl)2ZrCl2 (45) or in Me2Si{2-Me-4-(1naphthyl)indenyl}2ZrCl2 (46) [146]. The complexes 44–46 are examples of the socalled ‘advanced metallocenes’ which give activities, stereoselectivities and polypropene molecular masses much higher than those of any previously described metallocene system [146]. The annelated benz[e]indenyl in 44 compared to the standard ansa-bis(indenyl) system leads to a considerable gain in activity because the 44/MAO catalyst is deactivated at an unusually low rate with an about 3 times longer half-life. Here again, a direct comparison of the complex 44 with and without α-methyl substituents revealed the molecular mass enhancing properties (by a factor of 4) of the methyl groups, although some of the activity was lost again [145, 215].
9.7 Propene-Homopolymerization
601
The polypropene molar mass from polymerization with e.g. C2H4(Ind)2ZrCl2/MAO can be lowered by the addition of hydrogen through the hydrogenolysis reaction. The catalyst activity was also found to improve in the presence of hydrogen, the molecular mass distribution remained unchanged [216].
602
9
Metallocene Catalysis for Olefin Polymerization
9.7.13 Molecular Modeling A variety of theoretical modeling studies are available (in addition to those already mentioned in the text) which in a valuable interplay between experiment and theory substantiated or adjusted the initial hypotheses on the stereocontrol in the α-olefin polymerization with the various stereoselective catalysts [168], e.g. ab initio MO and molecular mechanics (MM) studies on specificity control in propene polymerization with bridged metallocenes [217]; an MM study on the back-skip of the growing chain at model complexes for various ethanediyl-bridged metallocenes [218]; a DFT study on the influence of rotation between agostic structures on ethene interaction with a zirconocene polymerization site [219]; a force-field study on π-stacking as a control element in the oscillating (2-PhInd)2ZrCl2 catalysts [220]; In view of the eminent importance of ansa-metallocenes for α-olefin polymerization, an efficient synthetic route is worthwhile noting: In an amine elimination reaction Zr(NMe2)4 reacts with bridged bis(indenes) to yield the corresponding ansa-bis(indenyl)zirconium dimethylamido complexes (Eq. (9-17)). The reaction not only proceeds in high yield but also gives in high excess the racemic mixture versus the achiral meso form [221]. Zr(NMe2)4 + bridge(Ind-H)2 → rac-bridge(Ind)2Zr(NMe2)2 + 2 NHMe2 bridge = Me2E (E = C, Si); CH2CH2
(9-17)
9.8 Homopolymerization of Other α-Olefins In catalytic alkene polymerization with metallocene catalysts the molar masses of the obtained polyalkenes and the reaction rates decrease in the order ethene > propene > 1-butene > 1-pentene > 1 hexene. 1-Butene: The system rac-C2H4(IndH4)2ZrCl2/MAO polymerizes the monomer with high activity [2640 kg polybutene (mol Zr · h)–1 at 20°C] to an isotactic and waxy-crystalline material of Mη = 50–150 000 [159]. From racMe2Si(IndH4)2ZrCl2/MAO a lower molecular mass oligomer (Mn ≈ 2000) was obtained whose end group structure was analyzed in terms the reaction mechanism of initiation and propagation with various chain-transfer reactions [222]. A 13C NMR study on the regioselectivity of isotactic 1-butene polymerization by C2-symmetric metallocenes provided evidence on the occurrence of 1,4 1-butene enchainments [134, 223]. 1-Pentene: Crystalline, syndiotactic poly(1-pentene) was obtained with good polymerization activity (5 × 103 kg of polymer mol–1 Zr · h–1) using the syndiospecific catalyst Me2C(C5H4)(Flu)ZrCl2/MAO [224].
9.9
Olefin Oligomerization
603
1-Hexene: MAO-activated rac-C2H4(Ind)2ZrCl2 yields highly isotactic polyhexene and also in various solvent but with Mn only around 20 000 even at low temperatures [225]. Poly(1-hexene) with high molar mass (above 106 g mol–1 for Mw) could be obtained through the application of pressures over 100 and up to 1000 MPa with achiral Cp-methylated zirconocene and hafnocene/MAO catalysts [226]. 3-Methyl-1-butene: C2-symmetric metallocene/MAO catalysts give isotactic poly(3-methyl-1-butene) with reasonable productivities, contrary to earlier literature reports. A prevailing isotactic polymer was also obtained by Cs-symmetric metallocenes, which normally are syndiospecific in propene polymerization [227].
9.9 Olefin Oligomerization The oligomerization of α-olefins has also been investigated. Initially the finding of low-molar-mass products from propene polymerization with zirconocene catalysts was generally regarded as unfortunate [40, 43, 159, 163, 228, 229]. Now, it is more and more recognized that metallocene catalysts may be used effectively for the directed oligomerization of α-olefins [154, 230–234] to give oligomers with almost exclusively double-bond end groups, predominantly of the vinylidene type. The use of olefin oligomers as intermediates for specialty chemicals drives the interest in the catalytic oligomerization [235]. A variety of functionalization reactions with this double bond are possible leading to organic specialities with possible applications as adhesives, blend compatilizers [236], fragrances, lubricants, additives for fuels or in the paper and leather industry [230]. Also, α-olefin oligomers or derivatives thereof can be used as (macro)monomeric building blocks for novel graft copolymers containing oligo-olefin side chains [232]. Olefin oligomerization is also used to study mechanistic aspects of metallocene catalysis because of the homogenity of the system (no heterogenation through polymer precipitation) and because the oligomeric products are easier to analyze than high-molar-mass polymers [162, 237–239]. At the limit of oligomerization one encounters the use of metallocene Ziegler systems as carbon–carbon coupling catalysts in organic synthesis, e.g. for the dimerization of olefins [154, 234, 240].
9.10 Copolymerization Comonomer insertion into the polymer chain is solely statistical for polymers obtained with metallocene catalysts (proven by NMR), contrary to conventional Ziegler–Natta catalysis where the comonomer is mainly incorporated into the low molar mass fraction.
604
9
Metallocene Catalysis for Olefin Polymerization
With (C5H5)2ZrCl2/MAO the copolymerization of ethene and propene [241] or 1butene gives elastomers with different properties. Copolymerization of ethene with 1,3-butadiene yields polyethenes which are functionalized by olefinic groups [41]; along the same lines copolymerization of ethene with 5-vinyl-2-norbornene using (C5H5)2ZrCl2/MAO occurs with regioselective insertion of the endocyclic double bond into the polymer chain, leaving the exocyclic vinyl double bond as a pendant unsaturation which can be readily converted to the hydroxy/epoxy group [242]. Copolymerizing ethene with 4-vinylcyclohexene with Ph2C(C5H4)(Flu)ZrCl2/MAO allowed hydroboration of the pendant cyclohexene double bond and oxidative workup yielded polyalcohols [243]. 4-Methyl-1-pentene was copolymerized with ethene using a chiral and an achiral catalyst [191]. Also, in ethene/propene or 1hexene copolymerization, the stereoselective rac-C2H4(Ind)2ZrCl2 and racC2H4(IndH4)2ZrCl2 exhibit much higher polymerization activities than simple unbridged metallocene catalysts as revealed in a comparative study. The former gave copolymers having comonomer compositions similar to the feed composition whereas nonbridged catalysts favored the incorporation of ethene over propene (1hexene) [244]. In another comparative study the activities in ethene/1-hexene and 1-octene copolymerization were found to be similar [245]. At a 1:1 ethene/propene feed ratio (C5H5)2ZrCl2/MAO gave rather low molecular mass copolymers (Mη < 10 000 g mol–1) above 40°C; the molar mass increased with decreasing temperature and with higher ethene feed [246]. Small amounts of ethene comonomer content in propene (homo)polymerization lead to polymers with two times higher molecular weights and only slightly lower melting points. The increase in chain length is due to an activation of the catalytic sites by ethene insertion which were blocked by a preceding 2–1 insertion which normally leads to β-hydride elimination and chain growth termination [133]. The ethene/1-hexene copolymerization of homogeneous zirconocene and a classical supported TiCl3 catalyst were compared to investigate the influence of α-olefin on the enhancement of ethene polymerization rate (the ‘comonomer effect’). The metallocene/MAO systems were both found to show a reduction in ethene activity upon comonomer addition, contrary to the classical TiCl3 catalysts [247] as well as a rate enhancement [248]. The copolymer properties of ethene/2-allylnorbornane prepared by TiCl4/Al iBu3 and (C5H5)2ZrCl2/MAO were compared and the metallocene copolymer found homogeneous and random in nature while the classical Ziegler–Natta catalyst gave a heterogeneous and blocky copolymer [249]. Ethene was copolymerized with 1-dodecene and 1-octadecene to study the influence of the comonomer chain length on the rate enhancement effect with the catalysts Me2C(C5H4)(Flu)ZrCl2 or Me2Si(Ind)2ZrCl2/MAO [250]. A comparative ethene/1-octadecene copolymerization study between (C5H5)2ZrCl2 and (C5H5)2HfCl2/MAO showed the hafnocene to be more reactive towards 1-octadecene and giving higher molecular masses of the polymer but at a lower total activity [251]. With (C5H5)2ZrCl2/MAO ethene has been copolymerized with 1-octene, 1-tetradecene and 1-octadecene for a comparison of the copolymer properties
9.11
Polymerization and Copolymerization of Cyclic Olefins
605
[252]. Me2C(C5H4)(Flu)ZrCl2 has been used to investigate the influence of temperature in the ethene/1-hexene or 1-octadecene copolymerization [253]. Ethene–styrene copolymers have been prepared with a catalytic system based on Me(Ph)C(C5H4)(Flu)ZrCl2/MAO/TMA [254]. A propene/1-octene copolymerization with the syndioselective catalyst Me2C(C5H4)(Flu)ZrCl2/MAO showed again random incorporation of large amounts of 1-octene. The polymer parameters were studied as a function of octene content [255]. In a series of copolymerizations of propene with higher α-olefins including 1-butene, 1-hexene, 1-octene, 1-dodecene, and 1-hexadecene with the isospecific catalyst C2H4(Ind)2HfCl2/MAO, the products obtained are random copolymers and the reactivity of the higher α-olefins in the copolymerization decreases only slightly with increasing length of the olefin [256].
9.11 Polymerization and Copolymerization of Cyclic Olefins The metallocene/MAO catalyst systems allow polymerization of cyclic olefins with no ring opening taking place [257]. With chiral metallocenes, cycloolefines like cyclobutene, cyclopentene, and norbornene can be polymerized to give isotactic polymers. For polycyclopentene, contrary to the initial formulation as a 1–2 insertion product 47 [193], it was later proven that incorporation proceeded in cis- and trans-1-3 manner (Fig. 9-15) to give poly(1,3-cyclopentene) [258] so that four microstructures can be distinguished cis-iso/syndiotactic and trans-iso/syndiotactic (cf. Fig. 9-15). From C2H4(Ind)2ZrCl2 and cyclopentene isotactic polycyclopentene is obtained which is insoluble in aromatic or aliphatic hydrocarbons and highly crystalline with a melting point above the decomposition temperature [192].
606
9
Metallocene Catalysis for Olefin Polymerization
Figure 9-15. Polycyclopentene micro structures.
The polycycloalkenes show extremely high melting points which lie above their decomposition temperatures (in air). Under vacuum the melting points were found (by DSC) to be 485°C for polycyclobutene, 395°C for polycyclopentene, and over 600°C for polynorbornene [193]. Such high melting points make the homopolymers difficult to process. To lower the melting points, the cycloolefin can be copolymerized with ethene or propene, thereby it was possible with metallocene catalysts to solve a decade old problem, namely the technical synthesis of cycloolefincopolymers [2, 257]. As pointed out above for the copolymers there is a random distribution of the cycloolefin units in the polymer chain according to 13C NMR measurements. In comparative cycloolefin/ethene copolymerization experiments the activity and the incorporation of the cyclic olefin is much higher with the chiral catalyst than with the achiral (C5H5)2ZrCl2 [193]. Copolymerization of cyclopentene with allyltrimethylsilane using the (S,S) enantiomeric form of C2H4(IndH4)ZrCl2 gave an optically active copolymer [193]. The rigid tetracyclododecene (‘dimethanooctahydronaphthalene’) (48) was copolymerized with ethene using chiral and achiral catalysts and the amorphous product with up to 12%, later up to 33% [259] of the cycloolefin featured a high glass-transition temperature, excellent transparency, thermal stability and chemical resistance [193]. The same holds for the norbornene/ethene copolymer and it is suggested that these materials could be used for optical discs and fibers. With C2H4(IndH4)ZrCl2/MAO ethene is inserted only 1.5 to 3.2 times faster than norbornene and copolymers containing more than 50 mol% of norbornene units can be
9.12 Diene- and Cyclopolymerization
607
made; if the norbornene concentration is higher than 60 mol% a glass transition point of about 120°C can be reached [260]. On the basis of a norbornene/ethene copolymer, the companies Hoechst and Mitsui Sekka have developed a highly transparent technical plastic, called TOPAS (thermoplastic olefin polymer of amorphous structure). The properties of TOPAS open applications in the market of compacts disks and magneto–optic storage disks [2]. In a comparative study with R2C(C5H4)(Flu)ZrCl2 compounds, the asymmetric complex Ph2C(C5H4)(Ind) proved to be the tailored metallocene for the copolymerization of ethene with sterically demanding cyclic olefins, such as 48 and 49 [170].
Copolymerization of ethene with 5-vinyl-2-norbornene using (C5H5)2ZrCl2/MAO occurs with regioselective insertion of the endocyclic double bond into the polymer chain, leaving the exocyclic vinyl double bond as an unsaturation which can be readily converted to the hydroxy/epoxy group [242].
9.12 Diene- and Cyclopolymerization Copolymerization of ethene with 1,3-butadiene or 5-vinyl-2-norbornene yields polyethenes with pendant olefinic groups, which can be further converted to give functional olefins [41, 242]. rac-C2H4(Ind)2ZrCl2/MAO is an excellent catalyst for the ethene (E), propene (P) and ethylidine norbornene [261] or octa-1,6-dienes (from biomass) [262] (D, as diene) terpolymerization with as much as 20 % D uniformly incorporated into the EPDM-terpolymer. EPDM terpolymers with the same components could also be prepared by means of (C5H5)2ZrMe2/MAO [263]. Cyclopolymerization of 1,5-hexadiene (Eq. (9-18)) with either of the enantiomeric forms of C2H4(IndH4)ZrBINOL (BINOL = 1,1′-bi-2-naphtholate) (50) yields either one of the enantiomeric forms of optically active poly(methylene-1,3-cyclopentane). For poly(methylene-1,3-cyclopentane) four microstructures are possible, which are sketched in 51. Of these, only the trans-isotactic structure contains no mirror plane (perpendicular to the chain) and is thus already chiral because of
608
9
Metallocene Catalysis for Olefin Polymerization
its main-chain stereochemistry. A reversibly decreasing molar optical rotation with increasing temperature indicates, however, that the polymer adopts chiral (helical) conformations which may contribute all or in part to the observed optical activity [264].
(9-18)
A novel polyspirane (52) has been obtained by ring-opening - ‘zipping up’ (= cyclo-) polymerization of methylenecyclopropane with [(C5Me5)2ZrMe]+[MeB(C6F5)3]–. It is assumed that at some point the initial ring-opening polymerization (Eq. (9-19)) is replaced through the competing intramolecular C=C bond insertion (Eq. (9-20)) with the spirocyclization process continuing along the chain [265]. When organolanthanoide catalysts {(C5Me5)2LnH}2 (Ln = Sm, Lu) were employed only the ring-opening to exo-methylene functionalized polyethenes was observed [111].
9.12 Diene- and Cyclopolymerization
609
(9-19)
(9-20)
A further example of cyclopolymerization can be found in Sec. 9.13 with 4-trimethylsiloxy-1,6-heptadiene (56) (Eq. (9-21)) [266].
(9-21)
610
9
Metallocene Catalysis for Olefin Polymerization
9.13 Polymerization of Polar Monomers The use of group IV metallocenes as initiating systems for the synthesis of polar or functional monomers has also been reported. Although, here in most cases it is more reasonable to assume not a Ziegler–Natta type insertion but a cationic polymerization reaction. For methyl methacrylate (MMA) which can be efficiently polymerized stereoselectiveally using a two-component catalyst system comprised of (C5H5)2ZrMe2 and [(C5H5)2ZrMe(THF)][BPh4] the mechanism probably involved enolate complexes in a group-transfer polymerization [267]. On the other hand (C5H5)2ZrMe2, rac-C2H4(Ind)2ZrCl2/Ph3CB(C6F5)4, rac-C2H4(IndH4)2ZrCl2/ Ph3CB(C6F5)4 and rac-Me2Si(Ind)2ZrCl2/Ph3CB(C6F5)4 were found to be inactive towards the polymerization of MMA without the addition of a Lewis acid such as dialkylzinc and alkylaluminum compounds [268]. Me2C(C5H4)(Flu)ZrCl2 failed to initiate the polymerization of MMA even under these conditions [268c]. The polymerization of polar monomers such as alkyl acrylates can also be initiated by group III and lanthanoide metallocenes e.g. (C5Me5)2LnMe(THF) (Ln = Y, Sm, Yb), {(C5Me5)2SmH}2 and (C5H4Me)2Ln(DME) to give high molecular weight polymers with extremely narrow molecular mass distribution [112]. Typical C2-symmtrical ansa-zirconocene catalysts, such as rac-Et(Ind)2ZrMe2/ [Ph3C][B(C6F5)4] together with (bis(2,6-di-tert-butyl-4-methylphenoxy)methyl)aluminum as the activator were reported for the synthesis of predominantly isotactic poly(tert-butylacrylate) (53) and highly isotactic poly(methyl methacrylate) [269]. Cationic group IV metallocene catalysts derived from the reaction of (C5Me5)2ZrMe2 or C2H4(IndH4)2ZrMe2 with B(C6F5)3 or [N,N-dimethylanilinium][B(C6F5)4] are active catalysts for the homopolymerization of α-olefins containing silyl-protected alcohols (54) and tertiary amines (55). With 4-trimethylsiloxy-1,6heptadiene (56) cyclopolymerization occurs (Eq. (9-21)) [266].
9.14
Recent Trends in Ligand and Metallocene Catalyst Design
611
Copolymers of propene and the polar monomer 57 were obtained with racMe2Si(IndH4)2ZrCl2/MAO, containing from 1.3 to 5.5 wt.% of comonomer. The addition of the sterically hindered phenolic monomer (or 2,6-di-tert-butylphenol) even increased the catalytic activity, which was attributed to the ability of the shielded phenol unit to function as a weakly coordinating anion [270]. Ethene could also be copolymerized with 10-undecen-1-ol using (C5H4-nBu)2ZrCl2/MAO; the bimodal molar mass distribution curve of the polymers obtained indicated, however, the presence of two or more active sites in the catalyst [271].
9.14 Recent Trends in Ligand and Metallocene Catalyst Design This section summarizes novel types of cyclopentadienyl ligands or ligands closely related to the Cp group which were recently used in the synthesis of group-IV metallocenes and their catalytic application in olefin polymerization. Cyclopentadienyl systems with an additional donor function in a side chain, e.g. dialkylaminoethylcyclopentadienyl (58) are attracting increased interest [272]. It is suggested inter alia that the donor atom may reversibly coordinate to and block a vacant coordination site at the metal center, thereby modulating the activity and catalyst lifetime. In a similar direction go studies on disiloxane bridged cyclopentadienyl- or indenyl-metallocene catalysts (59) for olefin polymerization. These latter complexes polymerize ethene only with MAO as a cocatalyst, propene with either MAO or [Ph3C][B(C6F5)3]. Towards propene the activity is moderate and increases strongly with temperature when MAO is used as an activator. Even when C2-symmetric catalysts such as 59 are employed only atactic polypropene is obtained. These catalytic effects were interpreted in terms of σ donation from the oxygen atom of the siloxane bridge to the metal center [273].
612
9
Metallocene Catalysis for Olefin Polymerization
The 2-siloxy- or 2-dimethylamino-substituted indenyl ligand (60) in bridged and unbridged zirconocene complexes and their application in polymerization catalysis has been investigated for its changed electronic properties [274]. The zirconium compounds 61 and 62, where the cyclopentadienyl rings have been formally substituted by the isoelectronic (amino)boratabenzene ligand, can also be activated with MAO for olefin polymerization [275]. The use of borate-substituted dianionic cyclopentadienyl ligands gives anionic group IV metallocene complexes (63) from which zwitterionic ethene polymerization catalysts are obtained upon activation with [Ph3C][B(C6F5)4] [276].
9.14
Recent Trends in Ligand and Metallocene Catalyst Design
613
The homo- and heterodinuclear ansa-zirconocene and hafnocene complexes (64) were synthesized and their activity as catalysts (activated by MAO, B(C6F5)3 or [Ph3C]+[B(C6F5)4]–) for the polymerization of ethene and propene has been evaluated. As far as polyethene activity is concerned no advantage rather a negative effect of the incorporation of a second metal is seen. In the case of propene, atactic oligomers are obtained and the activity is higher for the di-zirconocene than for the mononuclear compound [277]. We would suggest that the activity trends are due to an increased steric hindrance for ethene but a higher stereoselective environment for propene (see above). Another study on a possible cooperative effect in binuclear zirconocenes of type 65 in comparison to the mononuclear complex 66 found a higher activity and an increased molecular weight in propene oligomerization for the mononuclear system. An electronic coupling/withdrawing effect through the phenylene bridge was invoked to explain the differences [278]. For a related study involving mono- and dinuclear titanocene dichlorides, see Ref. [279].
614
9
Metallocene Catalysis for Olefin Polymerization
References [1]
[2] [3] [4] [5]
[6]
[7]
[8]
[9] [10]
[11]
[12] [13] [14] [15]
[16]
For notes concerning the industrial application, see e.g. Chemical & Engineering News 1996, December 16, p. 15; August 12, p. 9-10; 1995, October 30, p. 10; September 18, p. 17; September 11, p. 15-20; May 22, p. 34-38. May 1, p. 7-8; Nachr. Chem. Tech. Lab. 1995, 43, p. 701; p. 1086; p. 1208. W. Kaminsky, Spektrum der Wissenschaft February 1997, p. 85-89. M. Aulbach, Chemie heute (Fonds der Chemischen Industrie), 1995/96, p. 70-73. A. Wood, Chemical Week 1992, July 1/8, 42-44. A. Wood, E. Chynoweth, Chemical Week 1992, May 13, 52-53. H. G. Hauthal, Nachr. Chem. Tech. Lab. 1995, 43, 822. H. Vennen, Future (Das Hoechst Magazin) IV/1995, p. 48-53. R. Mülhaupt, B. Rieger, Chimia 1995, 49, 486. F. Langhauser, J. Kerth, M. Kersting, P. Kölle, D. Lilge, P. Müller, Angew. Makromol. Chem. 1994, 223, 155. (C5H5)2NbIVCl2 and (C5H5)2NbVCl3 were tested in combination with MAO towards ethene polymerization but did not exhibit a significant activity. R. Quijada, J. Dupont, D. C. Silveira, M. S. L. Miranda, R. B. Sciponi, Macromol. Rapid Commun. 1995, 16, 357-62. We note that another metallocene, namely chromocene, (C5H5)2Cr, is used widely and reacted with silica (SiO2) to form the Union Carbide catalyst for low pressure ethene polymerization. In contrast to the group-IV metallocene polymerization catalysts the chromocene moiety, however, does not remain intact upon reaction with the silica surface. The nature of the active species is not known but at least one cyclopentadienyl ring is lost to give a surface bound halfsandwich complex. See for example M. Schnellbach, F. H. Köhler, J. Blümel, J. Organomet. Chem. 1996, 520, 227 and references therein. M. P. McDaniel, Adv. Catal. 1985, 33, 48. For recent references concerning the syndiospecific styrene polymerization with half-sandwich complexes, see: T. E. Ready, J. C. W. Chien, M. D. Rausch, J. Organomet. Chem. 1996, 519, 21. A. Grassi, A. Zambelli, F. Laschi, Organometallics 1996, 15, 480. P. Longo, A. Proto, A. Zambelli, Macromol. Chem. Phys. 1995, 196, 3015. H. Kucht, A. Kucht, J. C. W. Chien, M. D. Rausch, Appl. Organomet. Chem. 1994, 8, 893. N. Ishirara, M. Kuramotot, M. Uoi, Macromolecules 1988, 21, 3356. L. Fan, D. Harrison, T. K. Woo, T. Ziegler, Organometallics 1995, 14, 2018. J. Okuda, F. J. Schattenmann, S. Wocadlo, W. Massa, Organometallics 1995, 14, 789. W. A. Herrmann, M. J. A. Morawietz, J. Organomet. Chem. 1994, 482, 169. T. J. Kealy, P. L. Pauson, Nature (London) 1951, 168, 1039. S. A. Miller, J. A. Tebboth, J. F. Tremaine, J. Chem. Soc. 1952, 632. E. O. Fischer, H. P. Fritz, Adv. Inorg. Chem. Radiochem. 1959, 1, 55. G. Wilkinson, M. Rosenblum, M. C. Whiting, R. B. Woodward, J. Am. Chem. Soc. 1952, 74, 2125. R. B. Woodward, M. Rosenblum, M. C. Whiting, J. Am. Chem. Soc. 1952, 74, 3458. J. D. Dunitz and L. E. Orgel, Nature (London) 1953, 171, 121. Comprehensive Organometallic Chemistry I and II, (G. Wilkinson, F. G. A. Stone. W. Abel, eds.) Pergamon Press, Oxford, 1982 and 1995. Review series Adv. Organomet. Chem. 1964, 1 up to now. C. Janiak, H. Schumann, Adv. Organomet. Chem. 1991, 33, 291. J. Boor, Ziegler-Natta Catalysts and Polymerizations, Academic Press, New York, 1979. H. Sinn, W. Kaminsky, Adv. Organomet. Chem. 1980, 18, 99. A Schulz-Flory distribution assumes a constant chain growth rate with a probability α, and a probability 1 - α for chain transfer. From this, statistics give for the molecular mass distribution Mw/Mn = Pw/Pn = 1 + α. If the probability for chain growth is much larger than for chain transfer [α >> (1 - α)] or α ≈ 1, then Mw/Mn ≈ 2. (Mw;Pw = weight average molar mass; degree of polymerization - Mn;Pn = number average molar mass; degree of polymerization) C. Janiak, B. Rieger, R. Voelkel, H.-G. Braun, J. Polym. Sci., Polym. Chem. 1993, 31, 2959. C. Janiak, B. Rieger, Angew. Makromol. Chem. 1994, 215, 47.
References [17]
[18] [19] [20] [21] [22] [23] [24]
[25] [26] [27] [28] [29] [30]
[31] [32] [33] [34] [35] [36]
[37] [38] [39] [40] [41] [42]
615
M. C. Sacchi, D. Zucchi, I. Tritto, P. Locatelli, T. Dall’Occo, Macromol. Rapid Commun. 1995, 16, 581. K. Soga, T. Arai, H. Nozawa, T. Uozumi, Macromol. Symp. 1995, 97, 53. K. Soga, M. Kaminaka, H. J. Kim, T. Shiono, in Ziegler Catalysts, (G. Fink, R. Mülhaupt, H. H. Brintzinger, eds.) Springer Verlag, Berlin, 1995, p. 333-342. K. Soga, H.-J. Kim, T. Shiono, Macromol. Chem. Phys. 1994, 195, 3347. J. C. W. Chien, D. He, J. Polym. Sci., Polym. Chem. 1991, 29, 1603. J. Tudor, L. Willington, D. O’Hare, B. Royan, J. Chem. Soc., Chem. Commun. 1996, 2031. S. I. Woo, Y. S. Ko, T. K. Han, Macromol. Rapid Commun. 1995, 16, 489. W. Kaminsky, M. Arndt, Adv. Polym. Sci. 1997, 127, 144. G. Fink, R. Mülhaupt, H. H. Brintzinger (eds.) Ziegler Catalysts, Springer Verlag, Berlin, 1995. H. H. Brintzinger, D. Fischer, R. Mülhaupt, B. Rieger, R. M. Waymouth, Angew. Chem. Int. Ed. Engl. 1995, 34, 1143. M. Aulbach, F. Küber, Chemie in Unserer Zeit 1994, 28, 197. (a) R. Mülhaupt, Nachr. Chem. Tech. Lab. 1993, 41, 1341. (b) M. Bochmann, Nachr. Chem. Tech. Lab. 1993, 41, 1220. (c) J. Okuda, Nachr. Chem. Tech. Lab. 1993, 41, 8. G. W. Coates, R. M. Waymouth, Comprehensive Organometallic Chemistry II (G. Wilkinson, F. G. A. Stone. W. Abel, eds.) Pergamon Press, Oxford, 1995, Vol. 12, p. 1193-1208. W. Kaminsky, M. Arndt, in Applied Homogeneous Catalysis with Organometallic Compounds (B. Cornils, W. A. Herrmann, eds.) VCH, Weinheim, 1996, Vol. 1, p. 220-236. M. Bochmann, J. Chem. Soc. Dalton Trans. 1996, 255. G. Natta, P. Pino, G. Mazzanti, U. Giannini, J. Am. Chem. Soc. 1957, 79, 2975. G. Natta, P. Pino, G. Mazzanti, U. Giannini, E. Mantica, M. Peraldo, J. Polym. Sci. 1957, 26, 120. D. S. Breslow, N. R. Newburg, J. Am. Chem. Soc. 1957, 79, 5072 G. Natta, P. Corradini, I. W. Bassi, J. Am. Chem. Soc. 1958, 80, 755. G. Natta, G. Mazzanti, Tetrahedron 1960, 8, 86. G. Natta, U. Giannini, G. Mazzanti, P. Pino, Angew. Chem. 1957, 69, 686. F. Patat, Hj. Sinn, Angew. Chem. 1958, 70, 496. G. Fink, D. Schnell, Angew. Makromol. Chem. 1982, 105, 15. G. Fink, D. Schnell, Angew. Makromol. Chem. 1982, 105, 31. G. Fink, D. Schnell, Angew. Makromol. Chem. 1982, 105, 39. K. H. Reichert, Angew. Makromol. Chem. 1981, 94, 1. K. Meyer, K. H. Reichert, Angew. Makromol. Chem. 1970, 12, 175. D. S. Breslow, N. R. Newburg, J. Am. Chem. Soc. 1959, 81, 81. J. C. W. Chien, J. Am. Chem. Soc. 1959, 81, 86. F. S. Dyachkovskii, A. K. Shilova, A. E. Shilov, J. Polym. Sci. Part C 1967, 16, 2333. J. J. Eisch, A. M. Piotrowski, S. K. Brownstein, E. J. Gabe, F. L. Lee, J. Am. Chem. Soc. 1985, 107, 7219. M. V. Gaudet, M. J. Zaworotko, T. S. Cameron, A. Linden, J. Organomet. Chem. 1989, 367, 267. J. J. Eisch, S. I. Pombrik, G.-X. Zheng, Organometallics 1993, 12, 3856. J. J. Eisch, K. R. Caldwell, S. Werner, C. Krüger, Organometallics 1991, 10, 3417. K. H. Reichert, K. R. Meyer, Makromol. Chem. 1973, 169, 163. W. P. Long, D. S. Breslow, Liebigs Ann. Chem. 1975, 463. This article was published in German in an organic chemistry journal and carried the straight title: ‘The effect of water on the catalytic activity of bis(cyclopentadienyl)titanium dichloride-dimethylaluminium chloride for the polymerization of ethylene’. A. Andresen, H.-G. Cordes, J. Herwig, W. Kaminsky, A. Merck, R. Mottweiler, J. Pein, Hj. Sinn, H.-J. Vollmer, Angew. Chem. Int. Ed. Engl. 1976, 15, 630. Hj. Sinn, W. Kaminsky, H.-J. Vollmer, R. Woldt, Angew. Chem. Int. Ed. Engl. 1980, 19, 390. W. Kaminsky, M. Miri, Hj. Sinn, R. Woldt, Makromol. Chem., Rapid Commun. 1983, 4, 417. W. Kaminsky, K. Külper, S. Niedoba, Makromol. Chem., Macromol. Symp. 1986, 3, 377. W. Kaminsky, M. Schlobohm, Makromol. Chem., Macromol. Symp. 1986, 4, 103. S. van der Ven, Polypropylene and other polyolefins: polymerization and characterization, Elsevier Science, Amsterdam, 1990, p. 389. If the activity would have been calculated on a g
616
[43] [44]
[45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64]
[65] [66] [67]
[68] [69] [70] [71] [72] [73] [74] [75] [76] [77]
9
Metallocene Catalysis for Olefin Polymerization
PE/g catalyst basis, with catalyst encompasing not only the transition metal complex but also the support and cocatalyst, then the classical Ziegler-Natta catalysts give the higher activity, because of the enormous excess of MAO (molar ratio Al:Zr = 150-500 000 :1) needed to achieve this activity. J. A. Ewen, J. Am. Chem. Soc. 1984, 106, 6355. F. R. W. P. Wild, L. Zsolnai, G. Huttner, H. H. Brintzinger, J. Organomet. Chem. 1982, 232, 233. For an even earlier description of a chiral ansa-titanocene derivative see H. Schnutenhaus, H. H. Brintzinger, Angew. Chem. Int. Ed. Engl. 1979, 18, 777. F. R. W. P. Wild, M. Wasiucionek, G. Huttner, H. H. Brintzinger, J. Organomet. Chem. 1985, 288, 63. H. Wiesenfeldt, A. Reinmuth, E. Barsties, K. Evertz, H.-H. Brintzinger, J. Organomet. Chem. 1989, 369, 359. W. Kaminsky, A.-M. Schauwienold, F. Freidanck, J. Mol. Catal. A: Chemical 1996, 112, 37. For a review on chiral cyclopentadienyl metal complexes, see R. L. Halterman, Chem. Rev. 1992, 92, 965. J. A. Smith, J. von Seyerl, G. Huttner, H. H. Brintzinger, J. Organomet. Chem. 1979, 173, 175. J. K. Whitesell, Chem. Rev. 1989, 89, 1581. M. J. Burk, R. L. Harlow, Angew. Chem. Int. Ed. Engl. 1990, 29, 1462. Macromol. Symp. 1995, Vol. 97, 1, containing the papers of lectures at the Hamburg macromolecular colloquium on ‘Alumoxanes’ in 1994. Review: S. Pasynkiewicz, Polyhedron 1990, 9, 429. J. C. W. Chien, B.-P. Wang, J. Polym. Sci., Polym. Chem. 1988, 26, 3089. C. J. Harlan, M. R. Mason, A. R. Barron, Organometallics 1994, 13, 1957. K. Peng, S. Xiao, J. Mol. Catal. 1994, 90, 201. S. S. Reddy, K. Radhakrishnan, S. Sivaram, Polym. Bull. 1996, 36, 165. P. C. Möhring, N. J. Coville, J. Mol. Catal. 1992, 77, 41. S. S. Reddy, Polym. Bull. 1996, 36, 317. A. R. Barron, Macromol. Symp. 1995, 97, 15. E. Giannetti, G. M. Nicoletti, R. Mazzocchi, J. Polym. Sci. Polym. Chem. 1985, 23, 2117. D. Cam, E. Albizzati, P. Cinquina, Makromol. Chem. 1990, 191, 1641. L. Resconi, S. Bossi, L. Abis, Macromolecules 1990, 23, 4489. H. Sinn, Macromol. Symp. 1995, 97, 27. I. Tritto, M. C. Sacchi, P. Locatelli, S. X. Li, Macromol. Chem. Phys. 1996, 197, 1537. W. Kaminsky, Angew. Makromol. Chem. 1994, 223, 101. A. R. Barron, Organometallics 1995, 14, 3581. For a different method to determine the content of ‘free’ TMA in MAO upon complexation with WOCl4, see E. Thorn-Csányi, J. Dehmel, O. Halle, W. Sciborski, Macromol. Chem. Phys. 1994, 195, 3017. S. Lasserre, J. Derouault, Nouv. J. Chim. 1983, 7, 659. D. Cam, U. Giannini, Makromol. Chem. 1992, 193, 1049. L. A. Nekhaeva, G. N. Bondarenko, S. V. Rykov, A. I. Nekhaev, B. A. Krentsel, V. P. Mar’in, L. I. Vyshinskaya, I. M. Khrapova, A. V. Polonskii, N. N. Korneev, J. Organomet. Chem. 1991, 406, 139. G. Erker, M. Albrecht, S. Werner, C. Krüger, Z. Naturforsch. B 1990, 45, 1205. J. C. W. Chien, R. Sugimoto, J. Polym. Sci., Polym. Chem. 1991, 29, 459. P. G. Gassman, M. R. Callstrom, J. Am. Chem. Soc. 1987, 109, 7875. D. Cam, F. Sartori, A. Maldotti, Macromol. Chem. Phys. 1994, 195, 2817. P. J. J. Pieters, J. A. M. van Beek, M. F. H. van Tol, Macromol. Rapid Commun. 1995, 16, 463-7. B. Rieger, C. Janiak, Angew. Makromol. Chem. 1994, 215, 35. G. Fink, W. Zoller, Makromol. Chem. 1981, 182, 3265. N. Herfert, G. Fink, Makromol. Chem. 1992, 193, 1359. S. Jüngling, R. Mülhaupt, J. Organomet. Chem. 1995, 497, 27. (a) D. Fischer, S. Jüngling, R. Mülhaupt, Makromol. Chem., Macromol. Symp. 1993, 66, 191. (b) D. Fischer, R. Mülhaupt, J. Organomet. Chem. 1991, 417, C7.
References [78] [79] [80] [81] [82] [83]
[84]
[85] [86] [87]
[88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109]
617
J. M. Vela Estrada, A. E. Hamielec, Polymer 1994, 35, 808. J. C. Vizzini, J. C. W. Chien, G. N. Babu, R. A. Newmark, J. Polym. Sci., Polym. Chem. 1994, 32, 2049. C. Janiak, U. Versteeg, K. C. H. Lange, R. Weimann, E. Hahn, J. Organomet. Chem. 1995, 501, 219. C. Janiak, K. C. H. Lange, U. Versteeg, D. Lentz, P. H. M. Budzelaar, Chem. Ber. 1996, 129, 1517. W. Kaminsky, Macromol. Symp. 1995, 97, 79. I. Tritto, S. X. Li, M. C. Sacchi, P. Locatelli, G. Zannoni, Macromolecules 1995, 28, 5358. I. Tritto, M. C. Sacchi, S. Li, Macromol. Rapid Commun. 1994, 15, 217. For a preceding 13C NMR study involving (C5H5)2TiRCl/AlEtnClm with 13C-labelled ethene, see: G. Fink, R. Rottler, Angew. Makromol. Chem.1981, 94, 25. S. Tinkler, R. J. Deeth, D. J. Duncalf, A. McCamley, J. Chem. Soc., Chem. Commun. 1996, 2623. J. D. Scollard, D. H. McConville, N. C. Payne, J. J. Vittal, Macromolecules 1996, 29, 5241. C. Janiak, T. G. Scharmann, K. C. H. Lange, Macromol. Rapid Commun. 1994, 15, 655.G. Conti, G. Arribas, A. Altomare, F. Ciardelli, J. Mol. Cat. 1994, 89, 41. P. Longo, L. Oliva, A. Grassi, C. Pellecchia, Makromol. Chem. 1989, 190, 2357. R. F. Jordan, C. S. Bajgur, R. Willett, B. Scott, J. Am. Chem. Soc. 1986, 108, 7410. A. S. Guram, R. F. Jordan, Comprehensive Organometallic Chemistry II (G. Wilkinson, F. G. A. Stone. W. Abel, eds.) Pergamon Press, Oxford, 1995, Vol. 4, p. 589-625. R. F. Jordan, Adv. Organomet. Chem. 1991, 32, 325. R. F. Jordan, P. K. Bradley, R. E. LaPointe, D. F. Taylor, New J. Chem. 1990, 14, 505. R. F. Jordan, J. Chem. Ed. 1988, 65, 285. G. G. Hlatky, H. W. Turner, R. R. Eckman, J. Am. Chem. Soc. 1989, 111, 2728. M. Bochmann, A. J. Jaggar, J. C. Nicholls, Angew. Chem. Int. Ed. Engl. 1990, 29, 780. A. D. Horton, J. H. G. Frjins, Angew. Chem. Int. Ed. Engl. 1991, 30, 1152. X. Yang, C. L. Stern, T. J. Marks, J. Am. Chem. Soc. 1994, 116, 10015. X. Yang, C. L. Stern, T. J. Marks, J. Am. Chem. Soc. 1991, 113, 3623. M. Bochmann, S. J. Lancaster, M. B. Hursthouse, K. M. A. Malik, Organometallics 1994, 13, 2235. X. Yang, C. L. Stern, T. J. Marks, Angew. Chem. Int. Ed. Engl. 1992, 31, 1375. J. C. W. Chien, W.-M. Tsai, M. D. Rausch, J. Am. Chem. Soc. 1991, 113, 8570. J. C. W. Chien, W. Song, M. D. Rausch, J. Polym. Sci., Polym. Chem. 1994, 32, 2387. J. C. W. Chien, W. Song, M. D. Rausch, Macromolecules 1993, 26, 3239. M. Bochmann, S. J. Lancaster, Angew. Chem. Int. Ed. Engl. 1994, 33, 1634. Y.-X Chen, C. L. Stern, S. Yang, T. J. Marks, J. Am. Chem. Soc. 1996, 118, 12451. P. A. Deck, T. J. Marks, J. Am. Chem. Soc. 1995, 117, 6128. A. R. Siedel, R. A. Newmark, J. Organomet. Chem. 1995, 497, 119. M. Bochmann, S. J. Lancaster, Organometallics 1993, 12, 633. M. Bochmann, S. J. Lancaster, J. Organomet. Chem. 1991, 434, C1. W. Michiels, A. Muñoz-Escalona, Macromol. Symp. 1995, 97, 171. M. Bochmann, S. J. Lancaster, J. Organomet. Chem. 1995, 497, 55. B. Temme, G. Erker, J. Karl, H. Luftmann, R. Fröhlich, S. Kotila, Angew. Chem. Int. Ed. Engl. 1995, 34, 1755. B. Temme, J. Karl, G. Erker, Chem. Eur. J. 1996, 2, 919. N. G. Alameddin, M. F. Ryan, J. R. Eyler, A. R. Siedle, D. E. Richardson, Organometallics 1995, 14, 5005. P. L. Watson, G. W. Parshall, Acc. Chem. Res. 1985, 18, 51. M. E. Thompson, J. E. Bercaw, Pure Appl. Chem. 1985, 56, 1. G. Jeske, H. Lauke, H. Mauermann, P. N. Swepston, H. Schumann, T. J. Marks, J. Am. Chem. Soc. 1985, 107, 8091. W. J. Evans, D. M. DeCoster, J. Greaves, Organometallics 1996, 15, 3210. E. Ihara, M. Nodomo, H. Yasuda, N. Kanehisa, Y. Kai, Macromol. Chem. Phys. 1996, 197, 1909.
618
9
Metallocene Catalysis for Olefin Polymerization
[110] E. B. Coghlin, J. E. Bercaw, J. Am. Chem. Soc. 1992, 114, 7606. [111] L. Jia, X.. Yang, A. M. Seyam, I. D. L. Albert, P.-F. Fu, S. Yang, T. J. Marks, J. Am. Chem. Soc. 1996, 118, 7900. X. Yang, A. M. Seyam, P.-F. Fu, T. J. Marks, Macromolecules 1994, 27, 4625. [112] H. Yasuda, E. Ihara, Macromol. Chem. Phys. 1995, 196, 2417. E. Ihara, M. Morimoto, H. Yasuda, Macromolecules 1995, 28, 7886. T. Jiang, O. Shen, Y. Lin, S. Jin, J. Organomet. Chem. 1993, 450, 121. H. Yasuda, H. Yamamoto, K. Yokota, S. Miyake, A. Nakamura, J. Am. Chem. Soc. 1992, 114, 4908. [113] P. Cossee, J. Catal. 1964, 3, 80. E. J. Arlman, J. Catal. 1964, 3, 89. E. J. Arlman, P. Cossee, J. Catal. 1964, 3, 99. [114] K. J. Ivin, J. J. Rooney, C. D. Stewart, M. L. H. Green, R. Mahtab, J. Chem. Soc., Chem. Commun. 1978, 604. M. L. H. Green, Pure Appl. Chem. 1978, 50, 27. [115] R. H. Grubbs, G. W. Coates, Acc. Chem. Res. 1996, 29, 85. [116] M. Brookhart, M. L. H. Green, L.-L. Wong, Prog. Inorg. Chem. 1988, 36, 1. M. Brookhart, M. L. H. Green, J. Organomet. Chem. 1983, 250, 395. [117] L. Clawson, J. Soto, S. L. Buchwald, M. L. Steigerwald, R. H. Grubbs, J. Am. Chem. Soc. 1985, 107, 3377. [118] W. E. Piers, J. E. Bercaw, J. Am. Chem. Soc. 1990, 112, 9406. [119] H. Krauledat, H. H. Brintzinger, Angew. Chem. Int. Ed. Engl. 1990, 29, 1412. [120] Z. Guo, D. C. Swenson, R. F. Jordan, Organometallics 1994, 13, 1424. [121] N. S. Barta, B. A. Kirk, J. R. Stille, J. Am. Chem. Soc. 1994, 116, 8912. [122] M. K. Leclerc, H. H. Brintzinger, J. Am. Chem. Soc. 1996, 118, 9024. M. K. Leclerc, H. H. Brintzinger, J. Am. Chem. Soc. 1995, 117, 1651. [123] C. A. Jolly, D. S. Marynick, J. Am. Chem. Soc. 1989, 111, 7968. [124] H. Kawamura-Kuribayashi, N. Koga, K. Morokuma, J. Am. Chem. Soc. 1992, 114, 2359. [125] C. Janiak, J. Organomet. Chem. 1993, 452, 63-73. M.-H. Prosenc, C. Janiak, H. H. Brintzinger, Organometallics 1992, 11, 4036. [126] J. W. Lauher, R. Hoffmann, J. Am. Chem. Soc. 1976, 98, 1729. [127] (a) P. Margl, J. C. W. Lohrenz, T. Ziegler, P. E. Blöchli, J. Am. Chem. Soc. 1996, 118, 4434. (b) J. C. W. Lohrenz, T. K. Woo, L. Fan, T. Ziegler, J. Organomet. Chem. 1995, 497, 91. T. K. Woo, L. Fan, T. Ziegler, Organometallics 1994, 13, 2252. [128] (a) V. L. Cruz, A. Muñoz-Escalona, J. Martinez-Salazar, Polymer, 1996, 37, 1663. (b) R. J. Meier, G. H. J. van Doremaele, S. Iarlori, F. Buda, J. Am. Chem. Soc. 1994, 116, 7274. (c) H. Weiss, M. Ehrig, R. Ahlrichs, J. Am. Chem. Soc. 1994, 116, 4919. (d) H. Kawamura-Kuribayashi, N. Koga, K. Morokuma, J. Am. Chem. Soc. 1992, 114, 8687. [129] Z. Wu, R. F. Jordan, J. L. Petersen, J. Am. Chem. Soc. 1995, 117, 5867. [130] H. G. Alt, R. Zenk, J. Organomet. Chem. 1996, 522, 177. R. Goddard, P. Binger, S. R. Hall, P. Müller, Acta Cryst. 1990, C46, 998. P. Binger, P. Müller, R. Benn, A. Rufinska, B. Gabor, C. Krüger, P. Betz, Chem. Ber. 1989, 122, 1035. H. G. Alt, C. E. Denner, U. Thewalt, M. D. Rausch, J. Organomet. Chem. 1988, 356, C83. T. Takahashi, M. Tamura, M. Saburi, Y. Uchida, E. Negishi, J. Chem. Soc., Chem. Commun. 1989, 852. T. Takahashi, M. Murakami, M. Kunishige, M. Saburi, Y. Uchida, K. Kozawa, T. Uchida, D. R. Swanson, E. Negishi, Chem. Lett. 1989, 761. [131] T. Takahashi, K. Kasai, N. Suzuki, K. Nakajima, E. Negishi, Organometallics 1994, 13, 3413. [132] P. Pino, R. Mülhaupt, Angew. Chem. Int. Ed. Engl. 1980, 19, 857. [133] G. Schupfner, W. Kaminsky, J. Mol. Catal. A: Chem. 1995, 102, 59. [134] V. Busico, R. Cipullo, J. Organomet. Chem. 1995, 497, 113, and references therein. [135] L. Cavallo, G. Guerra, Macromolecules 1996, 29, 2729. [136] R. F. Jordan, P. K. Bradley, N. C. Baenziger, R. E. LaPointe, J. Am. Chem. Soc. 1990, 112, 1289. [137] W. Kaminsky, A. Bark, R. Steiger, J. Mol. Catal. 1992, 74, 109. [138] M. Bochmann, T. Cuenca, D. T. Hardy, J. Organomet. Chem. 1994, 484, C10. [139] Q. Wang, R. Quyoum, D. J. Gillis, M.-J. Tudoret, D. Jeremic, B. K. Hunter, M. C. Baird, Organometallics 1996, 15, 693.
References
619
[140] J. C. W. Chien, B.-P. Wang, J. Polym. Sci., Polym. Chem. 1989, 27, 1539. [141] P. C. Möhring, N. J. Coville, J. Organomet. Chem. 1994, 479, 1. [142] J. A. Ewen, L. Haspeslagh, M. J. Elder, J. L. Atwood, H. Zhang, H. N. Cheng, in Transition Metals and Organometallics as Catalysts for Olefin Polymerization, (W. Kaminsky, Hj. Sinn, eds.) Springer Verlag, Berlin, 1988, p. 281-289. J. C. W. Chien, A. Razavi, J. Polym. Sci., Polym. Chem. 1988, 26, 2369. T. Mise, S. Miya, H. Yamazaki, Chem. Lett. 1989, 1853. [143] N. Piccolrovazzi, P. Pino, G. Consiglio, A. Sironi, M. Moret, Organometallics 1990, 9, 3098. [144] I.-M. Lee, W. J. Gauthier, J. M. Ball, B. Iyengar, S. Collins, Organometallics 1992, 11, 2115. [145] U. Stehling, J. Diebold, R. Kirsten, W. Röll, H. H. Brintzinger, S. Jüngling, R. Mülhaupt, F. Langhauser, Organometallics 1994, 13, 964. [146] W. Spaleck, F. Küber, A. Winter, J. Rohrmann, B. Bachmann, M. Antberg, V. Dolle, E. F. Paulus, Organometallics 1994, 13, 954. [147] L. Resconi, F. Piemontesi, I. Camurati, D. Balboni, A. Sironi, M. Moret, H. Rychlicki, R. Zeigler, Organometallics 1996, 15, 5046. W. Kaminsky, O. Rabe, A.-M. Schauwienold, G. U. Schupfner, J. Hanss, J. Kopf, J. Organomet. Chem. 1995, 495, 181. [148] P. C. Möhring, N. Vlachakis, N. E. Grimmer, N. J. Coville, J. Organomet. Chem. 1994, 483, 159. [149] H. J. R. de Boer, B. W. Royan, J. Mol. Catal. 1994, 90, 171. [150] J. F. Vega, A. Muñoz-Escalona, A. Santamaria, M. E. Muñoz, P. Lafuente, Macromolecules 1996, 29, 960. [151] J. Tian, B. Huang, Macromol. Rapid Commun. 1994, 15, 923. [152] W. Kaminsky, R. Engehausen, K. Zoumis, W. Spaleck, J. Rohrmann, Makromol. Chem. 1992, 193, 1643-1651. [153] K. Uozumi, K. Soga, Makromol. Chem. 1992, 193, 823-31. [154] C. Janiak, K. C. H. Lange, P. Marquardt, Macromol. Rapid Commun. 1995, 16, 643-50. C. Janiak, K. Lange, P. Marquardt, GIT Fachz. Lab. 1996, 40, 716-9. [155] M. Lahti, J. Koivumäki, J. V. Seppäla, Angew. Makromol. Chem. 1996, 236, 139-153. H.-F. Herrmann, L. L. Böhm, Polymer Commun. 1991, 32, 58-61. K. Soga, H. Yanagihara, D. Lee, Makromol. Chem. 1989, 190, 995. [156] P. Burger, K. Hortmann, H.-H. Brintzinger, Macromol. Chem., Macromol. Symp. 1993, 66, 127. K. Hortmann, H.-H. Brintzinger, New J. Chem. 1992, 16, 51. [157] W. Kaminsky, H. Lüker, Makromol. Chem., Rapid Commun. 1984, 5, 225. [158] T. K. Han, H. K. Choi, D. W. Jeung, Y. S. Ko, S. I. Woo, Macromol. Chem. Phys. 1995, 196, 2637. [159] W. Kaminsky, K. Külper, H. H. Brintzinger, F. R. W. P. Wild, Angew. Chem. Int. Ed. Engl. 1985, 24, 507 [160] V. Busico, R. Cipullo, P. Corradini, L. Landriani, M. Vacatello, A. L. Segre, Macromolecules 1995, 28, 1887. [161] H. L. Frisch, C. L. Mallows, F. A. Bovey, J. Chem. Phys. 1966, 45, 1565 [162] W. Kaminsky, M. Arndt, in Catalyst Design for Tailor-made Polyolefins (K. Soga, M. Terano, eds.), Kodansha Ltd., Tokyo, 1994, p. 179. [163] B. Rieger, X. Mu, D. T. Mallin, M. D. Rausch, J. C. W. Chien, Macromolecules 1990, 23, 3559. [164] G. Erker, R. Nolte, Y.-H. Tsay, C. Krüger, Angew. Chem. Int. Ed. Engl. 1989, 28, 628. [165] G. Erker, R. Nolte, R. Aul, S. Wilker, C. Krüger, R. Noe, J. Am. Chem. Soc. 1991, 113, 7594. [166] W. Spaleck, M. Antberg, V. Dolle, R. Klein, J. Rohrmann, A. Winter, New J. Chem. 1990, 14, 499. [167] R. Fierro, J. C. W. Chien, M. D. Rausch, J. Polym. Sci., Polym. Chem. 1994, 32, 2817. [168] P. Montag, Y. v. d. Leek, K. Angermund, G. Fink, J. Organomet. Chem. 1995, 497, 201. [169] S. Miyake, Y. Okumura, S. Inazawa, Macromolecules 1995, 28, 3074. [170] W. Kaminsky, R. Engehausen, J. Kopf, Angew. Chem. Int. Ed. Engl. 1995, 34, 2273. [171] E. Polo, M. L. H. Green, F. Benetello, G. Prini, S. Sostero, O. Traverso, J. Organomet. Chem. 1997, 527, 173. G. Conti, G. Arribas, A. Altomare, B. Mendez, F. Ciardelli, Z. Naturforsch. B 1995, 50, 411.
620
9
Metallocene Catalysis for Olefin Polymerization
[172] M. Knickmeier, G. Erker, T. Fox, J. Am. Chem. Soc. 1996, 118, 9623. G. Erker, C. Mollenkopf, J. Organomet. Chem. 1994, 483, 173. G. Erker, B. Temme, J. Am. Chem. Soc. 1992, 114, 4004. [173] G. Erker, M. Aulbach, C. Krüger, S. Werner, J. Organomet. Chem. 1993, 450, 1. [174] G. Erker, M. Aulbach, M. Knickmeier, D. Wingbermühle, C. Krüger, M. Nolte, S. Werner, J. Am. Chem. Soc. 1993, 115, 4590. [175] M. A. Giardello, M. S. Eisen, C. L. Stern, T. J. Marks, J. Am. Chem. Soc. 1993, 115, 3326. [176] A. Razavi, J. L. Atwood, J. Am. Chem. Soc. 1993, 115, 7529. [177] L. Resconi, R. L. Jones, A. L. Rheingold, G. P. A. Yap, Organometallics 1996, 15, 998. Y.-X. Chen, M. D. Rausch, J. C. W. Chien, Macromolecules 1995, 28, 5399. H. G. Alt, W. Milius, S. J. Palackal, J. Organomet. Chem. 1994, 472, 113. [178] K. Patsidis, H. G. Alt, J. Organomet. Chem. 1995, 501, 31. [179] G. W. Coates R. M. Waymouth, Science 1995, 267, 217-9. See also K. B. Wagener, Science 1995, 267, 191. [180] J. A. Ewen, R. L. Jones, A. Razavi, J. D. Ferrara, J. Am. Chem. Soc. 1988, 110, 6255. A. Razavi, J. Ferrara, J. Organomet. Chem. 1992, 435, 299. For a theoretical study on syndiospecific propene polymerization, see Z. Yu, J. C. W. Chien, J. Polym. Sci., Polym. Chem. 1995, 33, 1085. [181] J. Loos, M. Buhk, J. Petermann, K. Zoumis, W. Kaminsky, Polymer 1996, 37, 387. [182] J. A. Ewen, M. J. Elder, R. L. Jones, L. Haspeslagh, J. L. Atwood, S. G. Bott, K. Robinson, Macromol. Chem., Macromol. Symp. 1991, 48/49, 253. [183] Y.-X. Chen, M. D. Rausch, J. C. W. Chien, J. Organomet. Chem. 1995, 497, 1. [184] K. Patsidis, H. G. Alt, W. Milius, S. J. Palackal, J. Organomet. Chem. 1996, 509, 63. [185] A. Razavi, J. L. Atwood, J. Organomet. Chem. 1995, 497, 105. [186] A. Razavi, J. L. Atwood, J. Organomet. Chem. 1996, 520, 115. [187] H. G. Alt, R. Zenk, J. Organomet. Chem. 1996, 526, 295. H. G. Alt, R. Zenk, J. Organomet. Chem. 1996, 522, 39. [188] M. A. Schmid, H. G. Alt, W. Milius, J. Organomet. Chem. 1995, 501, 101. M. A. Schmidt, H. G. Alt, W. Milius, J. Organomet. Chem. 1996, 525, 9. M. A. Schmidt, H. G. Alt, W. Milius, J. Organomet. Chem. 1996, 525, 15. [189] H. G. Alt, R. Zenk, J. Organomet. Chem. 1996, 518, 7. [190] G. Natta, Angew. Chem. 1964, 76, 553. [191] W. Kaminsky, Angew. Makromol. Chem. 1986, 145/146, 149. [192] W. Kaminsky, R. Steiger, Polyhedron 1988, 7, 2375. [193] W. Kaminsky, A. Bark, I. Däke, in Catalytic Olefin Polymerization (T. Keii, K. Soga, eds.) Kodansha-Elsevier, Tokyo-Amsterdam, 1990, p. 425-438. [194] M. Antberg, V. Dolle, R. Klein, J. Rohrmann, W. Spaleck, A. Winter, in Catalytic Olefin Polymerization (T. Keii, K. Soga, eds.) Kodansha-Elsevier, Tokyo-Amsterdam, 1990, p. 501-515. [195] J. A. Ewen, M. J. Elder, R. L. Jones, S. Curtis, H. N. Cheng, in Catalytic Olefin Polymerization (T. Keii, K. Soga, eds.) Kodansha-Elsevier, Tokyo-Amsterdam, 1990, p. 439-482. [196] V. Busico, R. Cipullo, J. Am. Chem. Soc. 1994, 116, 9329. [197] L. Resconi, A. Fait, F. Piemontesi, M. Colonnesi, H. Rychlicki, R. Zeigler, Macromolecules 1995, 28, 6667. [198] V. Busico, L. Caporaso, R. Cipullo, L. Landriani, G. Angelini, A. Margonelli, A. L. Segre, J. Am. Chem. Soc. 1996, 118, 2105. [199] W. J. Gauthier, J. F. Corrigan, N. J. Taylor, S. Collins, Macromolecules 1995, 28, 3771. See also W. J. Gauthier, S. Collins, Macromolecules 1995, 28, 3779. [200] B. Rieger, G. Jany, R. Fawzi, M. Steimann, Organometallics 1994, 13, 647. [201] Y.-X. Chen, M. D. Rausch, J. C. W. Chien, J. Polym. Sci., Polym. Chem. 1995, 33, 2093, and references therein. Y.-X. Chen, M. D. Rausch, J. C. W. Chien, J. Organomet. Chem. 1995, 487, 29. W.-M. Tsai, J. C. W. Chien, J. Polym. Sci., Polym. Chem. 1994, 32, 149. [202] Y.-X. Chen, M. D. Rausch, J. C. W. Chien, Organometallics 1994, 13, 748. [203] W. A. Herrmann, M. J. A. Morawietz, H.-F. Herrmann, F. Küber, J. Organomet. Chem. 1996, 509, 115.
References
621
[204] B. Rieger, Polym. Bull. 1994, 32, 41. [205] J. A. Ewen, L. Haspeslagh, J. L. Atwood, H. Zhang, J. Am. Chem. Soc. 1987, 109, 6544. [206] W. Spaleck,M. Antberg, J. Rohrmann, A. Winter, B. Bachmann, P. Kiprof, J. Behm, W. A. Herrmann, Angew. Chem. Int. Ed. Engl. 1992, 31, 1347. [207] T. Mise, S. Miya, H. Yamazaki, Chem. Lett. 1989, 1853. [208] W. Röll, H. H. Brintzinger, B. Rieger, R. Zolk, Angew. Chem. Int. Ed. Engl. 1990, 29, 279. [209] P. Burger, J. Diebold, S. Gutmann, H.-U. Hund, H. H. Brintzinger, Organometallics 1992, 11, 1319. [210] F. Piemontesi, I. Camurati, L. Resconi, D. Balboni, A. Sironi, M. Moret, R. Zeigler, N. Piccolrovazzi, Organometallics 1995, 14, 1256. [211] W. Mengele, J. Diebold, C. Troll, W. Röll, H. H. Brintzinger, Organometallics 1993, 12, 1931. [212] K. Weiss, U. Neugebauer, S. Blau, H. Lang, J. Organomet. Chem. 1996, 520, 171. [213] L. Cavallo, P. Corradini, G. Guerra, L. Resconi, Organometallics 1996, 15, 2254. [214] B. Dorer, M.-H. Prosenc, U. Rief, H. H. Brintzinger, Organometallics 1994, 13, 3868. [215] S. Jüngling, R. Mülhaupt, U. Stehling, H. H. Brintzinger, D. Fischer, F. Langhauser, J. Polym. Sci., Polym. Chem. 1995, 33, 1305. [216] T. Tsutsui, N. Kashiwa, A. Mizuno, Makromol. Chem., Rapid Commun. 1990, 11, 565. [217] T. Yoshida, N. Koga, K. Morokuma, Organometallics 1996, 15, 766. G. Guerra, L. Cavallo, G. Moscardi, M. Vacatello, P. Corradini, J. Am. Chem. Soc. 1994, 116, 2988. E. P. Bierwagen, J. E. Bercaw, W. A. Goddard, III, J. Am. Chem. Soc. 1994, 116, 1481. J. R. Hart, A. K. Rappé, J. Am. Chem. Soc. 1993, 115, 6159. L. A. Castonguay, A. K. Rappé, J. Am. Chem. Soc. 1992, 114, 5832. L. Cavallo, P. Corradini, G. Guerra, M. Vacatello, Polymer 1991, 32, 1329. V. Venditto, G. Guerra, P. Corradini, R. Fusco, Polymer 1990, 31, 530. L. Cavallo, G. Guerra, L. Oliva, M. Vacatello, P. Corradini, Polym. Commun. 1989, 30, 16. P. Corradini, G. Guerra, M. Vacatello, V. Villani, Gazz. Chim. Ital. 1988, 118, 173. [218] G. Guerra, L. Cavallo, G. Moscardi, M. Vacatello, P. Corradini, Macromolecules 1996, 29, 4834-45. [219] J. A. Støvneng, E. Rytter, J. Organomet. Chem. 1996, 519, 277-80. [220] M. A. Pietsch, A. K. Rappé, J. Am. Chem. Soc. 1996, 118, 10908-9. [221] G. M. Diamond, R. F. Jordan, J. L Petersen, J. Am. Chem. Soc. 1996, 118, 8024. G. M. Diamond, R. F. Jordan, J. L. Petersen, Organometallics 1996, 15, 4030-7. G. M. Diamond, S. Rodewald, R. F. Jordan, Organometallics 1995, 14, 5-7. A Vogel, T. Priermeier, W. A. Herrmann, J. Organomet. Chem. 1997, 527, 297-300. See also W. A. Herrmann, M. J. A. Morawietz, T. Priermeier, Angew. Chem. Int. Ed. Engl. 1994, 33, 1946. [222] A. Rossi, G. Odian, J. Zhang, Macromolecules 1995, 28, 1739. [223] V. Busico, R. Cipullo, A. Borriello, Macromol. Rapid Commun. 1995, 16, 269. [224] M. Galimberti, G. Balbontin, I. Camurati, G. Paganetto, Macromol. Rapid Commun. 1994, 15, 633. [225] D. Coevoet, H. Cramail, A. Deffieux, Macromol. Chem. Phys. 1996, 197, 855. G. D. Babu, R. A. Newmark, J. C. W. Chien, Macromolecules 1994, 27, 3383. [226] A. Fries, T. Mise, A. Matsumoto, H. Ohmori, Y. Wakatsuki, J. Chem. Soc. Chem. Commun. 1996, 783. [227] A. Borriello, V. Busico, R. Cipullo, J. C. Chadwick, O. Sudmeijer, Macromol. Rapid Commun. 1996, 17, 589. [228] B. Rieger, Polym. Bull. 1994, 32, 41. [229] M. R. Meneghetti, M. C. Forte, J. Dupont, Polym. Bull. 1995, 4, 431. [230] K.-D. Hungenberg, J. Kerth, F. Langhauser, H.-J. Müller, P. Müller, Angew. Makromol. Chem. 1995, 227, 159. [231] W. Kaminsky, A. Ahlers, N. Möller-Lindenhof, Angew. Chem. Int. Ed. Engl. 1989, 28, 1216. [232] Wörner, C.; Rösch, J.; Höhn, A.; Mülhaupt, R. Polym. Bull. 1996, 36, 303. Mülhaupt, R.; Duschek, T.; Rösch, J. Polym. Adv. Technol. 1993, 4, 465. Mülhaupt, R.; Duschek, T.; Fischer, D.; Setz, S. Polym. Adv. Technol. 1993, 4, 439. Duschek, T.; Mülhaupt, R. Am. Chem. Soc., Polym. Chem. Div., Polym. Prepr. 1992, 33, 170.
622
9
Metallocene Catalysis for Olefin Polymerization
[233] Psiorz, C.; Erker, G.; Fröhlich, R.; Grehl, M. Chem. Ber. 1995, 128, 357. Thiele, S.; Erker, G.; Fritze, C.; Psiorz, C.; Fröhlich, R. Z. Naturforsch. B 1995, 50, 982. [234] J. Christoffers, R. G. Bergman, J. Am. Chem. Soc. 1996, 118, 4715. [235] J. Skupinska, Chem. Rev. 1991, 91, 613. [236] R. Mülhaupt, T. Duschek, B. Rieger, Macromol. Chem., Macromol. Symp. 1991, 48/49, 317. [237] Rossi, A.; Zhang, J.; Odian, G. Macromolecules 1996, 29, 2331. Eshuis, J. J. W.; Yan, Y. Y.; Meetsma, A.; Teuben, J. H.; Renkema, J.; Evens, G. G. Organometallics 1992, 11, 362. Eshuis, J. J. W.; Tan, Y. Y.; Teuben, J. H.; Renkema, J. J. Mol. Catal. 1990, 62, 277. [238] Resconi, L.; Piemontesi, F.; Franciscono, G.; Abis, L.; Fiorani, T. J. Am. Chem. Soc. 1992, 114, 1025. [239] Tsutsui, T.; Mizuno, A.; Kashiwa, N. Polymer 1989, 30, 428. Mise, T.; Kageyama, A.; Miya, S.; Yamazaki, H. Chem. Lett. 1991, 1525. [240] S. Thiele, G. Erker, Chem. Ber./Recueil 1997, 130, 201. [241] V. Banzi, L. Angiolini, D. Caretti, C. Carlini, Angew. Makromol. Chem. 1995, 229, 113. [242] S. Marathe, S. Sivaram, Macromolecules 1994, 27, 1083. [243] W. Kaminsky, D. Arrowsmith, H. R. Winkelbach, Polym. Bull. 1996, 36, 577. [244] Propene: J. C. W. Chien, D. He, J. Polym. Sci., Polym. Chem. 1991, 29, 1585. 1-Hexene: R. Quijada, R. B. Scipioni, R. S. Mauler, G. B. Galland, M. S. L. Miranda, Polym. Bull. 1996, 35, 299. [245] R. Quijada, J. Dupont, M. S. L. Miranda, R. B. Sciponi, G. B. Galland, Macromol. Chem. Phys. 1995, 196, 3991. [246] P. Pietikäinen, J. V. Seppälä, Macromolecules 1994, 27, 1325. [247] J. C. W. Chien, T. Nozaki, J. Polym. Sci., Polym. Chem. 1993, 31, 227. [248] J. Koivumäki, M. Lahti, J. V. Seppälä, Angew. Makromol. Chem. 1994, 221, 117. [249] S. Marathe, T. P. Mohandas, S. Sivaram, Macromolecules 1995, 28, 7318. [250] J. Koivumäki, G. Fink, J. V. Seppälä, Macromolecules 1994, 27, 6254. [251] J. Koivumäki, Polym. Bull. 1995, 34, 413. [252] J. Koivumäki, Polym. Bull. 1995, 36, 7. [253] P. H. Mühlenbrock, G. Fink, Z. Naturforsch. B 1995, 50, 423. [254] L. Oliva, L. Caporaso, C. Pellecchia, A. Zambelli, Macromolecules 1995, 28, 4665. [255] S. Jüngling, R. Mülhaupt, D. Fischer, F. Langhauser, Angew. Makromol. Chem. 1995, 229, 93. [256] M. Arnold, O. Henschke, J. Knorr, Macromol. Chem. Phys. 1996, 197, 563. [257] W. Kaminsky, R. Spiehl, Makromol. Chem. 1989, 190, 515. [258] W. M. Kelly, N. J. Taylor, S. Collins, Macromolecules 1994, 27, 4477. [259] G. M. Benedikt, B. L. Goodall, N. S. Marchant, L. F. Rhodes, New J. Chem. 1994, 18, 105. [260] W. Kaminsky, A. Bark, M. Arndt, Makromol. Chem., Macromol. Symp. 1991, 47, 83. [261] J. C. W. Chien, D. He, J. Polym. Sci., Polym. Chem. 1991, 29, 1609. [262] M. Dolatkhani, H. Cramail, A. Deffieux, Macromol. Chem. Phys. 1996, 197, 2481. [263] W. Kaminsky, M. Miri, J. Polym. Sci., Polym. Chem. 1985, 23, 2151. [264] G. W. Coates, R. M. Waymouth, J. Am. Chem. Soc. 1993, 115, 91. [265] L. Jia, X. Yang, S. Yang, T. J. Marks, J. Am. Chem. Soc. 1996, 118, 1547. [266] M. R. Kesti, G. W. Coates, R. M. Waymouth, J. Am. Chem. Soc. 1992, 114, 9679. [267] S. Collins, D. G. Ward, K. H. Suddaby, Macromolecules 1994, 27, 7222. [268] (a) H. Deng, T. Shiono, K. Soga, Macromolecules 1995, 28, 3067. (b) H. Deng, T. Shiono, K. Soga, Macromol. Chem. Phys. 1995, 196, 1971-80. (c) K. Soga, H. Deng, T. Yano, T. Shiono, Macromolecules 1994, 27, 7938. [269] H. Deng, K. Soga, Macromolecules 1996, 29, 1847. [270] C.-E. Wilén, J. H. Näsman, Macromolecules 1994, 27, 4051. [271] P. Aaltonen, B. Löfgren, Macromolecules 1995, 28, 5353. [272] P. Jutzi, T. Redeker, B. Neumann, H.-G. Stammler, Organometallics 1996, 15, 4153. [273] W. Song, K. Shackett, J. C. W. Chien, M. D. Rausch, J. Organomet. Chem. 1995, 501, 375. B. Liang, Y. Li, G. Xie, Macromol. Rapid Commun. 1996, 17, 193. D. Lee, K. Yoon, E. Lee, S. Noh, C. Lee, W. Huh, Macromol. Rapid Commun. 1996, 17, 325. See also: H. Naderer, E.
References
[274]
[275] [276] [277]
[278] [279]
Siebel, R. D. Fischer, J. Organomet. Chem. 1996, 518, 181. J. Gräper, G. Paolucci, R. D. Fischer, J. Organomet. Chem. 1996, 501, 211. R. Leino, H. Luttikhedde, C.-E. Wilén, R. Sillanpää, J. H. Näsman, Organometallics 1996, 15, 2450. E. Barsties, S. Schaible, M.-H. Prosenc, U. Rief, W. Röll, O. Weyand, B. Dorer, H.-H. Brintzinger, J. Organomet. Chem. 1996, 520, 63. H. J. G. Luttikhedde, R. P. Leino, C.-E. Wilén, J. H. Näsman, M. J. Ahlgrén, T. A. Pakkanen, Organometallics 1996, 15, 3092. G. C. Bazan, G. Rodriguez, A. J. Ashe III, S. Al-Ahmad, C. Müller, J. Am. Chem. Soc. 1996, 118, 2291. M. Bochmann, S. J. Lancaster, O. B. Robinson, J. Chem. Soc., Chem. Commun. 1995, 2081. G. M. Diamond, A. N. Chernega, P. Mountford, M. L. H. Green, J. Chem. Soc., Dalton Trans. 1996, 921. T. Ushioda, M. L. H. Green, J. Haggitt, X. Yan, J. Organomet. Chem. 1996, 518, 155. S. Jüngling, R. Mülhaupt, H. Plenio, J. Organomet. Chem. 1993, 460, 191. H. Lang, S. Blau, A. Muth, K. Weiss, U. Neugebauer, J. Organomet. Chem. 1995, 490, C32.
Abbreviations used Bu Cp Et Flu Ind IndH4 MAO Me PE PP THF TMA
623
butyl cyclopentadienyl, general ethyl fluorenyl indenyl tetrahydroindenyl methylalumoxane methyl polyethene polypropene tetrahydrofuran trimethylaluminum
624
9
Metallocene Catalysis for Olefin Polymerization
10 Chiral Titanocenes and Zirconocenes in Synthesis Amir H. Hoveyda and James P. Morken
10.1 Introduction The design and development of catalytic processes that effect bond formation under mild conditions in a highly enantioselective fashion (>90% enantiomeric excess) stands as an important and challenging task in modern chemical synthesis [1]. Organometallic systems that possess well-defined three-dimensional structures—metal complexes whose structural integrity in solution can be predicted— offer an attractive option in enantioselective reaction engineering, since the intricacies of substrate–catalyst association can be studied in appreciable detail. Within this context, chiral C2-symmetric ansa-metallocenes, also referred to as bridged metallocenes, have found extensive use as catalysts that effect enantioselective bond-forming processes [2]. In general, bridged ethylene(bis-tetrahydroindenyl)metallocene dichlorides illustrated in Scheme 10-1 (1–3) put forth attractive options for asymmetric reaction development because of their geometrically-constrained structure and relative ease of preparation. The purpose of this article is to provide a review of the chemistry of these and related chiral metallocene complexes, particularly in relation to their utility in catalytic and enantioselective C–C and C–H bond formation reactions.
M
Cl Cl
1 M = Ti Cl
M
Cl
2 M = Zr 3 M = Hf
Scheme 1. Group IV ethylene-bridged bis(tetrahydroindenyl) systems represent well-defined complexes that may be used in asymmetric catalysis.
626
10
Chiral Titanocenes and Zirconocenes in Synthesis
10.2 Chiral Nonracemic EBTHI Complexes of Group IV Transition Metals. Synthesis and Resolution 10.2.1 A Historical Perspective Titanium-based metallocene 1 was first prepared by Brintzinger and coworkers, as part of their pioneering work in the design and synthesis of configurationally stable group IV complexes [3]. The 1982 account describes the preparation of the racemic chiral titanocene, a procedure for separating the chiral complex from the corresponding meso isomer, and a method for the resolution of the racemate through preparation of the derived binaphtholate. In addition, Brintzinger’s landmark paper offered the first crystal structure of a member of this class of C2-symmetric metallocenes. A report detailing studies on the zirconocene complex 2 was published by the Brintzinger laboratory in 1985; synthesis and crystal structure of the Hf complex 3 was reported in 1987 by Ewen and Atwood [4]. Collins and coworkers in 1988 reported a refinement of the original crystal structures for 1 and 2 [5]. In 1995, Piemontesi and Sironi published crystal structures of rac- and meso-(EBI)ZrCl2 systems [6]; these researchers disclosed a detailed spectroscopic study of the organometallic complexes, unveiling a variety of their subtle conformational preferences. In the above mentioned Collins paper [5], experimental modifications in the synthesis procedure for zirconocene complexes (2) were provided as well; the new protocol allowed for improved yields, rendering these organometallic complexes more accessible. Additional modifications of the synthesis route by Buchwald enhanced the efficiency of the catalyst synthesis; this included use of KH to deprotonate indene (see Scheme 10-2) and utilization of a double-dilution technique to couple the bridged-bis(indene) ligand with ZrCl4•(THF)2 complex [7]. Interestingly, Buchwald’s procedural adjustments also led to the formation of ~30% of the corresponding meso isomer (2:1 rac:meso). Most recently, Jensen has published an attractive practical modification of the original Brintzinger procedure [8]. We find that when the latter procedure is used, rac-(EBTHI)ZrCl2 can be obtained in ~80% yield with <2% of the meso contaminant.
10.2.2 General Procedure for Catalyst Preparation 10.2.2.1 Synthesis of rac-(EBTHI)M Complexes As illustrated in Scheme 10-2, the bis(indene) ligand 4 is produced through alkylation of dibromoethane with the lithium salt derived from indene (nBuLi, –78°C, THF). Coupling of bis(indene) dianion to the solvated metal tetrachloride salt [9] affords rac-(EBI)ZrCl2 (5) and meso-(EBI)ZrCl2 (6). In principle, 5 (after appropri-
627
10.2 Chiral Non-Racemic EBTHI Complexes of Group IV Transition Metals
ate resolution procedure) can be used in asymmetric transformations. However, chiral metallocene 5 has less favorable solubility properties (compared to EBTHI complexes) and tends to be somewhat less stable than the complexes derived from hydrogenation of its pendant cyclohexadiene groups. Accordingly, Pt-catalyzed hydrogenation of a mixture of 5 and 6 is typically carried out to afford metallocenes 1–3 [10].
1. n-BuLi
1. KH
2. BrCH2CH2Br
M
2. MCl4•(THF)2
4
1 M = Ti 2 M = Zr 3 M = Hf
M
Cl Cl
Cl Cl
M
+
5
6
racemic
meso
Cl Cl
H2 PtO2 (5 mol%)
racemic
Scheme 2. General synthesis scheme for the preparation of chiral metallocenes 1-3.
Jordan has reported a convenient method for the synthesis of rac-(EBI)ZrCl2, where the derived meso- and chiral (EBI)Zr(NMe2)2 complexes (prepared from amine elimination reaction of 4 with Zr(NMe2)4) under appropriate reaction conditions undergo rapid equilibration [11]. Whereas the initial ratio (after two hours) is 2:1 rac:meso, after 17 h at 100°C it is increased to 13:1; subsequent recrystallization of the amido complex affords the desired chiral metallocene in 95% yield. Subsequent treatment with Me2NH•HCl affords 5 in 69% yield (the entire operation can be carried out in a single vessel as well). Most likely, rac-(EBI)Zr(NMe2)2 can itself be used for a large number of reactions where the dichloride is employed.
10.2.2.2 Resolution of Racemic EBTHI Complexes Resolution through mandelate complexes. Resolution procedures have so far been developed only for the titanium and zirconium metallocenes 1 and 2. Two methods that allow for the effective separation of the two chiral metallocene enantiomers involve the displacement of the bound chlorides with appropriate chiral, nonracemic ligands. Separation of the resulting diastereomers, followed by reinstallation of the chloride ions, affords the optically pure metallocene dichlorides. As illustrated in Scheme 10-3, one successful procedure calls for the use of O-acetylmandelic acid. Diastereomeric mandelate salts 7 and 8 can be separated by recrystallization
628
10
Chiral Titanocenes and Zirconocenes in Synthesis
[12]. Treatment of the derived nonracemic mandelates with hydrochloric acid, or formation of the dimethyl metallocene (after treatment with MeLi), followed by the addition of HCl, readily converts 7 and 8 to chiral nonracemic 1 and 2. OAc OH O
M
Cl Cl
OAc
(R)-O-acetylmandelic acid
O O
M
M O
Et3N
O
OAc O O O
OAc
O
OAc
(±)-1 or (±)-2 7
M
Cl Cl
8 1. separate diastereomers 2. HCl or MeLi / HCl
(R)-1 or (R)-2 Scheme 3. Metallocene resolution through separation of diastereomeric mandalate salt derivatives.
Resolution through binaphtholate complexes. Whereas mandelic acid (cf. Scheme 10-3) is relatively inexpensive (~$5 g–1), the fractional recrystallization required to separate the diastereomeric derivatives (e.g., 7 and 8) is somewhat tedious. A more popular procedure that is used for metallocene resolution [3a, 11] relies on the fact that, as shown in Scheme 10-4, nonracemic binaphthol salts (Na [3a] or Li [11]) efficiently react with only one enantiomer in the racemic chiral metallocene mixture. Hence, treatment of racemic 1 or 2 with 0.5 equivalent of nonracemic lithium binaphtholate results in the formation of (R)-(EBTHI)M-binol 9 together with 0.5 equivalent of the unreacted nonracemic metallocene dichloride, which may be separated upon treatment with chromatography absorbents (in contrast to the binaphthol derivatives, dichlorides are absorbed onto either silica or alumina). Air stable binaphtholate salts (9) are often used directly as precatalysts to effect various asymmetric processes. Alternatively, these chiral nonracemic metallocene complexes can be converted to their derived dichlorides by treatment with HCl. For zirconocene 1, pretreatment with MeLi to form (EBTHI)ZrMe2 is required to separate the free binaphtholate from the metal complex.
629
10.2 Chiral Non-Racemic EBTHI Complexes of Group IV Transition Metals
O
OLi
(±)-1 or (±)-2 +
Zr
OLi
M
+
O
9
(0.5 equiv.)
Cl Cl
(R)-1 or (R)-2 SiO2 or Al2O3
M
treatment; HCl (or MeLi, HCl)
Cl Cl
(S)-1 or (S)-2 Scheme 4. Metallocene resolution through separation of diastereomeric binaphtholate derivatives.
Buchwald has recently reported a procedure [13], where: (i) use of lithium or sodium binaphthol salt is not necessary (Et3N is used as base), and (ii) both enantiomers of the chiral metallocenes can be easily accessed. As illustrated in Scheme 105, with the presence of p-aminobenzoic acid, the metallocene enantiomer that does not match the binaphthol enantiomer, is readily converted to its derived benzoate (10); the crystalline benzoate is then easily separated from soluble binaphtholate (in hexane) and subsequently converted to the nonracemic dichloride 1 or 2. O OLi
(±)-1 or (±)-2 +
OLi
2.5 equiv Et3N
OH
+ H2 N
(1.0 equiv.)
toluene, reflux, 4h (then add hexane)
M
O O
filtrate
9
(0.6 equiv.)
NH2
separate from 9 9 is obtained in 30-37% yield optically pure
M
Cl Cl
then conc. HCl in CH2Cl2
(R)-1 or (R)-2 27-33% yield optically pure
Scheme 5. Improved metallocene resolution through p-aminobenzoates.
O O
M
O O
10
NH2
solid
630
10
Chiral Titanocenes and Zirconocenes in Synthesis
10.3 Stereoselective Reactions with ansa-Metallocenes as Reagents 10.3.1 Enantioselective, Zirconocene-Mediated Synthesis of Allylic Amines Buchwald and coworkers have reported a procedure that employs metallocene complex 2 in a enantioselective route for the preparation of chiral nonracemic allylic amines [14]. Although this protocol for the zirconium-mediated synthesis of allylic amines involves stoichiometric amounts of the chiral metallocene complex, it offers a reliable and highly enantioselective route towards the preparation of these important chiral materials. As shown in Eq. (10-1), the net transformation that is carried out is an asymmetric reductive coupling of an alkyne to the α carbon of a secondary amine. R
1
R
2
+
NAr
NAr
Zr(IV) R
R3
1
* R
R3
(a)
(10-1)
2
As illustrated in Scheme 10-6, the zirconocene-mediated synthesis of allylic amines begins with (S)-(EBTHI)ZrMe2 11, prepared by treatment of binaphtholate 10 (Scheme 10-4) with two equivalents of MeLi. Reaction of dimethylzirconocene 11 with one equivalent of trifluoromethanesulfonic acid at –78°C affords the zirco-
10.3
631
Stereoselective Reactions with Ansa-Metallocenes as Reagents
nium methyltriflate salt 12, which upon treatment with the appropriate lithium amide is readily converted to 13. At this point, heating of the reaction mixture to 80°C results in the formation of the zirconocene–imine complex 14 and 15. It is not clear whether β-H abstraction furnishes 14 selectively, or if equilibrium between 14 and 15 is rapidly established at elevated temperatures [15]. In the latter case, equilibration should favor 15, due to the minimization of steric interactions between the alkyl substituent (R) and the cyclohexyl moiety of the chiral ligand (these steric repulsions likely exist in 14). Regardless, addition of the alkyne substrate leads to a highly regioselective insertion process to provide the five-membered zirconacycle 18. Subsequent hydrolysis of the metallacyclopentane furnishes the desired allylic amine in excellent enantiopurity (Table 10-1). Table 10-1. Zircocene-mediated enantioselective synthesis of secondary allylic amines entry
alkyne
1
Me Me
ee (%)
H NPh
PhNLi Me
yield (%)
product
lithium amide
Me
72
>95
50
>90
59
>95
Me Me HNPh
PhNLi
2
Ph
Me
Me
Me
Ph Me
OTBS PhNLi
3
Me3Si
Me
HNPh Me
SiMe3 OTBS H NPh
PhNLi
4
Me
54 Ph
5
Me
PhNLi
Me
HNPh
43 H
Ph
high
Ph Me
O
[a]
Ph
[b]
94
Me OH
[a] 8:1 Mixture of diastereomers. [b] 7:1 Mixture of diastereomers. As the examples in Table 10-1 illustrate, the zirconocene-mediated synthesis of allylic amines tolerates a variety of functional groups in the alkyne starting materi-
632
10
Chiral Titanocenes and Zirconocenes in Synthesis
als. Simple alkyl-substituted alkynes, as well as silylacetylenes, terminal alkenes and aldehydes, may be employed as substrates. Unsymmetrical alkynes can be used; in cases where the steric presence of the two acetylene substituents is significantly different, high levels of regiocontrol are observed (entries 2–3 of Table 101). It is worthy of note that reactions with terminal alkenes (entry 4, Table 10-1) and a carboxaldehyde substrate occur with useful levels of diastereoselection (entry 5).
10.3.2 Diastereoselective Titanocene-Mediated Addition of Allylmetals to Aldehydes Stereoselective addition of chiral allyl- and crotyltitanocenes to aldehydes, as shown in Scheme 10-7, has been reported by Collins [16]. Treatment of titanocene dichloride 1 with propylmagnesium chloride, followed by the addition of either allyl- or crotylmagnesium chloride, affords titanium(III)–π-allyl complex 20 or 21. Upon subsequent addition of an aldehyde substrate, C–C bond formation takes place, affording the derived homoallylic alkoxides (Scheme 10-7). Workup with anhydrous HCl, followed by exposure to O2 results in alkoxide protonation and reoxidation of the titanocene reagent (to the derived dichloride complex). Therefore, chiral titanocene 1, although required in stoichiometric quantities, can be recovered from the reaction and reused for subsequent transformations.
Ti
Cl Cl
1. nPrMgCl (1.2 equiv)
R
Ti
2. R
(0.8 equiv)
1 Scheme 7
R'
2. HCl / O2
MgCl
R = H or Me
OH
1. R'CHO
20 (R=H) 21 (R=Me)
R
major diastereomer
Whereas excellent diastereoselectivity (40:1 anti:syn) is achieved in the above allylmetal addition procedure, with chiral nonracemic metallocenes, C–C bond formation is effected with moderate levels of enantioselectivity; reaction of pivaldehyde with the crotyltitanocene 21 provides a 36:1 ratio of anti:syn products, but only in 55% ee (83% yield). As shown in Scheme 10-8, other experiments with the crotyl reagent 21 indicate that decreasing the size of the aldehyde substituent results in diminution of diastereo- and enantiocontrol in the alkylation process. Transition structures I and II, illustrated in Scheme 10-8, have been suggested by Collins to rationalize the observed trends in stereoselectivity in the Ti-catalyzed
10.3
Stereoselective Reactions with Ansa-Metallocenes as Reagents
633
allylmetal additions to aldehydes. In reactions where crotyltitanocene 21 is used (Scheme 10-8), it has been proposed that the transformation proceeds through a six-membered chair-like transition state [17], wherein both the alkyl (R) and methyl groups adopt the energetically more favorable pseudoequatorial position. The anti diastereomer therefore emerges as the major addition product (Schemes 10-7 and 10-8). The above mechanistic hypothesis offers a rationale for higher levels of diastereoselection that are observed when aldehyde substrates that bear bulky alkyl substituents are used. Steric repulsion between the aldehyde substrate and cyclohexyl group of the catalyst disfavor transition structure II (Scheme 10-8). It is therefore tenable that allylmetallation of aldehydes with crotyltitanocene 21 proceeds selectively through transition structure I.
Reactions of allyl- and crotyltitanocene reagents with aldehydes have dissimilar stereochemical consequences. Aldehyde substrates that have larger alkyl substituents afford higher levels of relative and absolute stereoselection when crotylmetallocene 21 is employed (Scheme 10-8). However, when allylmetallocene 20 is used (Scheme 10-9), the more sterically demanding aldehydes provide inferior levels of stereochemical control. Furthermore, alkylations with 21 generally proceed with higher degrees of enantiocontrol than those of its companion allylic reagent 20 (compare Schemes 10-8 and 10-9). The chair-like transition structures presented in Schemes 10-8 and 10-9 have been invoked by Collins to explain the observed trends in diastereo- and enantioselectivity in reactions of titanocenes 20 and 21 to aldehydes. However, the proposed transition structures do not provide a strong basis for the stereochemical outcomes observed in alkylations with allyltitanocene 20. It is plausible that the carbonyl oxygen should bind to the metal in such a way that the alkyl group (R) is oriented
634
10
Chiral Titanocenes and Zirconocenes in Synthesis
pseudoequatorially, allowing the reaction to proceed through III (Scheme 10-9) to furnish the (R) product enantiomer predominantly. Nevertheless, it is not entirely clear why increasing the steric bulk of the aldehyde should decrease stereoselectivity. ≠ O R
H
Ti
Ti
O
OH R
R H
R=CH2Ph 40% ee R=t Bu 4% ee
20
III
Scheme 9. (EBTHI)Ti-allyl addition to aldehydes and the related transition structures. .
Comparison with related enantioselective processes. The chiral zirconocene-promoted allylation reactions described above are important since, although modest degrees of enantioselectivity are obtained, high levels of diastereoselectivity in the C–C bond forming reactions are detected. In addition, these studies shed light on a variety of subtle interactions between the allylmetal and various aldehyde substrates; such principles should prove critical in future developments involving these chiral metallocene complexes. However, as far as stereochemical control in this general class of transformations is concerned, a number of other stoichiometric reactions have been reported that provide outstanding levels of enantioselectivity. Reaction technologies reported by Brown [18], Mukaiyama [19], Riediker [20], Corey [21], Hafner and Duthaler [22], Roush [23], and Hoffman [24] serve as notable examples [25]. More recently, impressive advances in the area of catalytic asymmetric allylmetal additions to carbonyls have been reported as well. As illustrated below, procedures put forth by Yamamoto (Eq. 10-2) [26] and Keck and Tagliavini (Eq. (103)) [27] indicate that both Ti- and B-based chiral Lewis acids can effect C–C bond forming reactions efficiently and with excellent enantioselectivity. (10-2)
CO2H
OiPr O
CF3
O
O
OiPr O
O
B
10-20 mol%
O
OH
Me
CF3
(b)
Me
Yamamoto, 1993
SiMe3 EtCN, -78 °C
99% yield, 88% ee
10.3
635
Stereoselective Reactions with Ansa-Metallocenes as Reagents
(10-3) O O
TiL2 OH
R
(c)
R
O SnBu3
Keck, 1993
R=Ph; L = (OiPr)2 (10 mol%); 88% yield, 95% ee
Tagliavini, 1993
R=n octyl; L=Cl2 (20 mol%); 83% yield, 97% ee
10.3.3 Stereoselective Olefin Insertion Reactions with Cationic Zirconocene Complexes Inspired by the ability of cationic ansa-zirconocene complexes to effect stereocontrolled olefin polymerization reactions (discussed in Sec. 10.4.1.1), Jordan has recently reported the stereoselective insertion of simple alkenes to both the (EBI)Zr(η2-pyrid-2-yl) and (EBTHI)Zr(η2-pyrid-2-yl) systems [28]. As shown in Eqs. (10-4) and (10-5), treatment of rac-(EBI)ZrMe2 22 and rac-(EBTHI)ZrMe2 11 with (nBu3)HN+BPh4– in the presence of 2-picoline affords (EBI)Zr(η2-pyrid-2-yl) complex 23 and (EBTHI)Zr(η2-pyrid-2-yl) complex 24, respectively (the derived B(C6F5) derivatives may also be prepared and are in fact reported to be more convenient to use). These metallocenes exist as mixtures of ‘N-inside’ and ‘N-outside’ isomers, rapidly exchange at 23°C, and are differentiated structurally based on the orientation of the η2-pyridyl ligand. Me
Zr
Me Me
(nBu3)HN BPh4
N Zr N
N
(10-4)
Me
Me
22
(d)
BPh4
23 Me
Zr
Me Me
(nBu3)HN BPh4
N
N
11
N Zr
Me
(e)
Me
24
BPh4
(10-5)
636
10
Chiral Titanocenes and Zirconocenes in Synthesis
When 23 and 24 are treated with a variety of alkene substrates, facile insertion into the azazirconacycle takes place. The derived azazirconacyclopentanes are formed with various, but generally high, levels of diastereoselection, depending on the nature of olefin substituents. The cases shown in Scheme 10-10 are illustrative. Whereas reactions with olefins that bear relatively small substituents (Me or nBu) afford metallacyclopentane isomers 25 and 28 selectively, with sterically more demanding terminal olefins, such as styrene or trimethylsilyl vinylsilane, 27 and 29 are obtained with high selectivity. The observed trends in regio- and stereochemical control have been rationalized by Jordan through arguments based on minimization of steric interactions between olefin substituents and the pyridyl ligand. Scheme 10-11 illustrates three modes of alkene insertion into 24. In all the pyridyl complexes the aromatic ring is tilted to avoid Me
Me
Me
N
N
H
Zr
Me Me
N
N
Zr
25
Me
Me
H
26
BPh4
major isomer
N
83% diastereomeric excess
Me
23
N
Ph
BPh4
Ph
Zr H
27 N
Me
BPh4
>98% diastereomeric excess
Me
nBu
N
H
Zr
nBu Me
N
N
BPh4
>98% diastereomeric excess
N
Me
Me
24
28
Me
Zr
TMS
BPh4
Zr
N TMS H
29 N
Scheme 10
Me
Me
Zr
BPh4
>98% diastereomeric excess
BPh4
10.3
637
Stereoselective Reactions with Ansa-Metallocenes as Reagents
unfavorable steric interactions with the chiral EBTHI group. With smaller alkyl substituents, mode of addition I is preferred: (EBTHI)Zr–alkene association occurs in a manner that results in the formation of a less-substituted C–Zr bond, and pyridyl–olefin interaction leads to the approach of the olefin with the stereochemistry shown (R group towards cyclohexyl group of chiral ligand). In II, steric repulsion between the R group and the pyridyl ligand are proposed to be energetically prohibitive. Accordingly, with larger groups at the ortho position of the pyridyl ligand, I is even more preferred, since the resulting larger tilt of the heterocyclic ligand (to avoid the EBTHI moiety) increases the severe steric interactions between the pyridyl ortho H and the alkene R group in II (it thus follows that with smaller ortho substituents, selectivity decreases). Additionally, when the alkene substituent (R) is a large group, such as a tBu unit, steric interactions with the chiral ligand in I result in diminution of stereoselectivity. With olefin substituents such as a phenyl or a SiMe3 group, substituents that can stabilize adjacent partial negative charge on a carbon (developed in the formation of the C–Zr bond) [29], III becomes the favored mode of reaction. However, whereas electronic effects can explain the turnover in regioselectivity of olefin–metal binding, steric effects between these large alkene substituents and the cyclohexyl unit of the EBTHI group provide a plausible rationale for the stereochemical preference in this metal–olefin coordination.
H
H
H Zr
R
N
H
R Zr
H R
N
Zr
N
H
H
I
H
H
II
III Favored with "large R" (R=Ph, SiMe3)
Favored with "small R" (R=Me, nBu)
Scheme 11. Various modes of olefin-metallocene complexation proposed by Jordan for stereoselective insertion of terminal olefins with chiral zirconocene-pyridyl systems.
As illustrated in Eq. (10-6), the nonracemic (EBTHI)Zr(η2-pyridyl)+ complex has been shown to participate in an asymmetric and catalytic C–C bond-forming reaction. This observation augurs well for future developments involving the cationic metallocene complexes in the area of catalytic enantioselective reaction design.
+ Me
N
nBu
3 mol% 24 23 Me (nBu3)HN BPh4
50 psi H2
N
R nBu
(f)
Me
6.2 turnovers, 58% ee
(10-6)
638
10
Chiral Titanocenes and Zirconocenes in Synthesis
10.4 Catalytic Reactions Effected by ansa-Metallocenes 10.4.1 Catalytic Asymmetric C–C Bond Forming Reactions 10.4.1.1 Asymmetric Propylene Polymerization Ansa-metallocenes were initially used by Ewen and coworkers to control tacticity in the polymerization of propylene and butylene. These studies demonstrate that with titanium complexes 5 and 6 (see Scheme 10-2) product mixtures of isotactic and atactic polymers can be obtained [30]. Kaminsky and Brintzinger reported in 1985 that, as illustrated in Scheme 10-12, Cp2ZrMe2 induces the formation of predominantly atactic polymer structures when methylalumoxane (MAO) is used for catalyst activation [3c]. In contrast, with catalytic (EBTHI)ZrCl2, polymers of 91% isotacticity and an average of 60 monomers in each isotactic sequence are produced. The activity of the bridged metallocene catalyst was thus demonstrated to be remarkably high: 7700 kg of polypropylene per hour may be produced for every mole of zirconocene; such values correspond to approximately 180000 propene insertions per catalyst molecule per hour. Because isotactic polymers do not have inherent chirality associated with the main chain, efforts have been extended to control the absolute stereochemistry in oligomerization reactions. Oligomers (small polymers: dimers, trimers, tetramers, etc.)
Me
Cp2ZrMe2 MAO toluene 60 °C
Me
(EBTHI)ZrCl2 MAO toluene 60 °C
Me
Me
Me
Me
atactic 2730 kg / Zr mol•h Me
Me
Me
Me
91% isotactic 7700 kg / Zr mol•h
Me
Me
n O Me
Al
Me O
Al
n
methylalumoxane (MAO)
n
Scheme 12. Propene is polymerized more efficiently by (EBTHI)ZrCl2 than with the parent Cp2Zr system.
are produced when chain termination steps are facilitated in relation to propagation events [31]; these principles were recently discussed in an excellent review article [31b]. One method of promoting termination steps is to include H2 gas in the polymerization reaction. It has been proposed that hydrogenation of the intermediate metal-alkyls effectively inhibits chain growth at the oligomer stage, thus affording
10.4
639
Catalytic Reactions Effected by Ansa-Metallocenes
a fully saturated product, as well as the corresponding metal hydride [32]. Alternatively, it is possible that addition of hydrogen enhances the relative rate of hydrogenation versus chain growth. If the reaction is conducted at lower alkene concentration, β-hydride elimination may compete with chain propagation as well to lead to the desired short polymers [33]. Indeed, both approaches have been employed in attempts to produce optically pure organic molecules by ansa-zirconocene catalysis. As shown in Scheme 10-13, Pino and coworkers demonstrated that when H2 is introduced into a polymerization reaction with (EBTHI)ZrCl2/MAO as catalyst, propene oligomers (e.g., 30) are produced in moderate enantioselectivity [34]. In the above process, 30 constitutes only 0.5% (by weight) of the total reaction product, which represents a significant amount of material, as the system allows for 64000 propene insertions per molecule of catalyst. It is noteworthy that, although the majority (~80%) of the reaction products constitute high molecular weight polymers, other oligomers obtained (e.g., 31) exhibit appreciable optical activity as well. (EBTHI)ZrCl2 MAO
Me
Me
Me
toluene 60°C H2
Zr
Me
Me
Me
+ Me
Me
30
31
70 % ee
70 % ee
+ oligomers
Me
Scheme 13
As was previously mentioned, an alternate method for producing oligomers is to reduce alkene concentration in a polymerization reaction, thereby slowing chain propagation relative to β-H elimination. Kaminsky and coworkers have found that polymerization of butene at low substrate concentration results in the desired low molecular weight polymers (e.g., 32) [33], albeit with low levels of stereochemical induction (Eq. (10-7)). Nevertheless, small oligomers do constitute a large percentage of the reaction product (~18% by weight), and low catalyst loading (0.01 mol% compared to butene) are sufficient for the reaction to proceed efficiently.
Me
0.01 mol% 8 MAO toluene 60°C H2
Me
Me Me
32 27 % ee
(g)
(10-7)
640
10
Chiral Titanocenes and Zirconocenes in Synthesis
10.4.1.2 Asymmetric Cyclopolymerization of Dienes In contrast to the polymerization of terminal alkenes, the cyclopolymerization reaction, recently developed by Waymouth and Coates, affords isotactic polymers that possess main-chain chirality [35]. The cyclopolymerization of 1,5-hexadiene depicted in Scheme 10-14 begins with the insertion of one of the alkenes of the diene into the metal–carbon bond of the growing polymer. Insertion of the remaining olefin results in cyclization of the monomer to afford the polymer elongated by one five-membered carbocycle; subsequent insertion steps account for the continued chain growth. H Zr
Zr P MAO
Scheme 14
H P
H
Zr
P H
It is worthy of note that of the four types of polymerization products that can in principle be obtained from a cyclopolymerization reaction, only the ‘racemo-diisotactic’ product is chiral by virtue of its main-chain stereochemistry (Scheme 1015). The other three adducts contain mirror planes within the repeating subunits that make up the polymer.
meso-diisotactic
n
racemo-diisotactic
n
meso-disyndiotactic
n
racemo-disyndiotactic
n
Scheme 15. Possible stereoisomeric products that may be obtained from cyclopolymerization of 1,5-hexadiene.
Initial studies with the racemic ansa-metallocene 9 have demonstrated that the predominant product from the cyclopolymerization of 1,5-hexadiene is the chiral racemo-diisotactic polymer [35a]; 68% of the ring fusion stereochemistry in the product contains trans configuration. Furthermore, optically enriched metallocene catalysts effect formation of polymers that possess optical activity (with (R)binaphtholate catalyst 9 (Eq. (10-8)), molar optical rotation (per monomer unit) = [Φ]28405 +50.1 (c 7.8, CHCl3)). Waymouth and Coates therefore concluded that isotactic polymers are the major products in the cyclopolymerization. It is interesting
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
641
that, although cis rings are present in the polymer structure, the corresponding molar optical rotations (per monomer unit) are considerably higher than that of the model system, trans-(1R,3R)-1,3-dimethylcyclopentane. In this connection, the Stanford researchers suggest that the polymer may adopt conformations in solution that contribute to the observed optical rotation.
Zr
O
(10-8)
O
MAO
68% trans
n
9
Comparison with related catalytic enantioselective processes. Stereospecific olefin polymerization reactions that are catalyzed by chiral metallocene catalysts are particularly important and were recently reviewed by Brintzinger, Fischer, Mulhaupt, Rieger and Waymouth [31b]. These homogeneous catalysts, alongside enantioselective methods reported by Natta and Farina for polymerization of pentadienes [36] and benzofurans [37], offer highly enantioselective polymerization and cyclopolymerization processes [38]. More importantly, these homogeneous catalytic systems offer unique opportunities in reaction design and engineering, as the polymerization reactions discussed above are effected by a single and well defined metal complex that has a reasonably well understood coordination sphere. In brief, chiral metallocenes have greatly expanded the scope and versatility of the readily accessible types of polyolefin materials. Nonetheless, the studies summarized above are likely to be the harbinger of other great contributions to the metal-catalyzed enantioselective polymerization reactions.
10.4.1.3 Asymmetric Diels–Alder Cycloaddition Reactions Collins and coworkers have performed studies in the area of catalytic enantioselective Diels–Alder reactions where ansa-metallocenes (33) are utilized as chiral catalysts [39]. Cycloadditions are typically efficient (~90% yield), but proceed with modest stereoselectivity (26–52% ee). Collins’s studies provide important mechanistic insights that may ultimately be utilized in the design and development of more selective metallocene-based chiral catalysts. The group(IV) metal catalyst used in the asymmetric Diels–Alder reaction is the cationic zirconocene complex (EBTHI)Zr(O-tBu)_THF (33, Eq. (10-9)). Treatment of the dimethylzirconocene [40] 11 with one equivalent of tert-butanol, followed by protonation with one equivalent of HEt3N•BPh4 results in the formation of the requisite chiral cationic complex (33), which is used to catalyze the [4+2] cycloaddition of acrylate and acrolein derivatives with cyclopentadiene.
642
10
Chiral Titanocenes and Zirconocenes in Synthesis
Zr
Me Me
1. t BuOH (1 equiv)
Zr
Ot Bu THF
(10-9)
BPh4
2. Et3NH BPh4 (1 equiv) 3. THF
11
33
As illustrated in Scheme 10-16, in the presence of 5 mol% 33, the [4+2] cycloaddition reaction of acrolein and cyclopentadiene proceeds efficiently to afford 34 in 93% yield and 53% ee, as a 5:1 mixture of endo:exo isomers. Spectroscopic (1H NMR) experiments indicate that the preferred orientation for binding between acrolein and the chiral metallocene is as shown. Accordingly, it has been proposed by Collins that, acrolein binds the chiral zirconocene complex so that the vinyl group is oriented away from the sterically hindered alkoxide unit in order to avoid steric interactions with the bulky tbutoxide ligand. It is plausible that, as can be seen in transition structures illustrated in Scheme 10-16, approach of cyclopentadiene in the endo mode leads to unfavorable interactions between the diene and the tbutoxy ligand, thus leading to modest levels of diastereoselectivity (5:1 endo:exo). Molecular models indicate that any significant steric influence by the chiral ligand on the stereochemical outcome (both absolute and relative) of the cycloaddition is unlikely; the observed low levels of enantioselection (53% ee) may thus be the result of inefficient stereochemical induction by the chiral EBTHI moiety [41]. ≠ Zr O
O O
H
Endo Addition CHO
≠
+ 5 mol% 33
+ CHO
Zr
O O
H
34-endo
34-exo
5:1; 53% ee Exo Addition
Scheme 16. Stereoselective (EBTHI)Zr-catalyzed [4+2] cycloaddition and the related transition structures.
In reactions where methylacrylate is used as the dienophile (Scheme 10-17), cycloadditions occur with lower levels of enantioselection (23% ee compared to 53% observed for acrolein; compared to 34, the major enantiomer 35 is opposite in absolute configuration), but with significantly higher degrees of diastereoselectivity (17:1, endo:exo). As before, formation of the endo isomer may occur more appreciably, since the transition structure that leads to the exo isomer should include ener-
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
643
getically unfavorable interactions between the diene and the transition metal tbutoxy ligand. Improved levels of endo selectivity are detected in the case of the methyl ester (Scheme 10-17); this is perhaps because, at least in part, the dienophile π-system is oriented towards the tbutoxy ligand where the steric influence of the bulky substituent is expected to be more pronounced. Collins and coworkers put forth the catalyst–substrate mode of association depicted in Scheme 10-17, based on previous reports regarding the preferred stereochemistry of Lewis acid association with carboxylic esters [42].
Most recently, Collins and coworkers reported on an enantioselective catalytic Diels–Alder cycloaddition, where zirconocene and titanocene bis(triflate) complexes are used as catalysts [43]. Most noteworthy is the influence of the solvent polarity on the observed levels of stereoselectivity. For example, as shown in Scheme 10-18, with 36 as catalyst, whereas in CH2Cl2 (1 mol% catalyst) the endo product 37 is formed in 30% ee (30:1 endo:exo, 88% yield), with CH3NO2 (5 mol% catalyst) as solvent the enantioselectivity is raised to 89% (7:1 endo:exo, 85% yield). Extensive 1H and 19F NMR studies further indicate that a mixture of metallocene– dienophile complexes exist in both solutions (~6:1 in CH2Cl2 and ~2:1 in CH3NO2, as shown in Scheme 10-18), and that most likely it is the minor complex isomer that is more reactive and leads to the observed major enantiomer. For example, whereas nOe experiments led to ca 5% enhancement of the CpH protons of the same ring when Hb in the minor complex is irradiated, no enhancements are observed with irradiation of Ha in the major complex.
644
10
Chiral Titanocenes and Zirconocenes in Synthesis
Zr
OTf OTf
36
O O
O
O
N
OTf
N
5 mol% 36
Zr
Zr O
O
CH3NO2
O
OTf
OTf
Ha
Major (less reactive)
O
O
O N
O
N
O
OTf Hb
Minor (more reactive)
37-endo 7:1; 89% ee
Scheme 18. Stereoselective (EBTHI)Zr(OTf)2-catalyzed [4+2] cycloaddition and the related transition structures proposed by Collins.
These recent results illustrate for the first time that metallocene-based chiral Lewis acids can serve effectively to provide [4+2] cycloaddition products with excellent levels of enatiofacial selectivity. Perhaps more importantly, the reported NMR studies and the observed dramatic solvent effect should pave the way for future endeavors in the rational design of better chiral metallocenes. Comparison with related catalytic enantioselective processes. The cycloaddition reactions catalyzed by titanocene and zirconocene complexes described above provide additional useful methods in enantioselective C–C bond forming reactions and offer deeper insights into the inner workings of the (EBTHI)M class of chiral catalysts. Nonetheless, other catalytic systems, in general, present a more selective method for carrying out these and other similar reactions [44]. Related asymmetric cycloadditions have been reported, where similar to Collins’s work oxazolidinones serve as dienophiles. For example, Narasaka (Eq. (10-10)) [45], Corey (Eq. (1011)) [46], Evans (Eq. (10-12)) [47], and Kobayashi (Eq. (10-13)) [48] have published systems that offer similar, or in general higher, levels of relative and absolute stereoselectivity. As further indicated by papers referenced in Ref. [44], asymmetric catalytic [4+2] cycloadditions continue to be subject of extensive investigations in a number of research groups, a fact that reflects the significance of this class of C–C bond forming transformations. It should be noted that enantioselective Diels–Alder cycloaddition is far from being a ‘generally solved problem’. The high enantioselectivities reported in the literature, although extremely important, only concern a specific, relatively small and highly reactive class of dienes and dienophiles [49]. In brief, considering the importance of the [4+2] cycloadditions in organic chemistry, future research in this area should continue to provide notable contributions to the field of asymmetric synthesis.
10.4
645
Catalytic Reactions Effected by Ansa-Metallocenes Ph
Ph
(10-10 to 10-13)
O
O
TiCl2 Ph
O
O
O Me
N
O Ph
Me
Ph
10 mol%
O
(j)
O N
O
, 0 °C
O
49:1 endo:exo 87% yield , 91% ee
Narasaka, 1986 Ph F3CO2SN
O
O N
Ph NSO2CF3
Al
Me 10 mol%
O
(k)
O
, -90 °C
N
O
O
>50:1 endo:exo 92% yield, 95% ee
Corey, 1989 O
O N
N
• Cu(OTf)2
O
O N
O
10 mol%
(l)
O N
O
, -50 °C
O
33:1 endo:exo 87% yield, 96% ee
Evans, 1993
Me
N
N
OH
Me
Me
• Yb(OTf)3, 4Å MS
O
O
OH Me
O
20 mol%
, 0 °C
(m)
Me O
Kobayashi, 1994
N O
O
7:1 endo:exo 66% yield, 88% ee
10.4.1.4 Asymmetric Carbomagnesation Reactions Asymmetric synthesis through catalytic enantioselective addition reactions Work in our laboratories in connection to the zirconocene-catalyzed carbomagnesation as a convenient method for selective C–C bond formation has resulted in the development of catalytic processes that are carried out at ambient temperature to
646
10
Chiral Titanocenes and Zirconocenes in Synthesis
afford synthetically useful products with excellent enantiomeric purity. As illustrated in Table 10-2, in the presence of 2.5–10 mol% nonracemic (EBTHI)ZrCl2 (or (EBTHI)Zr-binol) and with EtMgCl as the alkylating agent, five-, six-, and sevenmembered unsaturated heterocycles undergo facile asymmetric ethylmagnesation [50]. Table 10-2. (EBTHI)Zr-catalyzed enantioselective ethylmagnesation of unsaturated heterocycles [a] entry
substrate
product H
3 O
O
65
>95
75
95
73
92
75
CH3
2 N
>97 HO H
nnonyl
yield (%)
CH3
1 O
ee (%)
HN nnonyl CH3 H HO
H
CH3
4 OH
[a] Reaction conditions: 10 mol% (R)-2, 5.0 equiv. EtMgCl, THF, 22°C for 6–12 h. Entry 1 with 2.5 mol% (R)-2. The rate of ethylmagnesation in the terminal alkenes of the reaction products is sufficiently slower, so that unsaturated alcohols and amines can be isolated in high yield (the second alkylation is not generally stereoselective). Catalytic asymmetric carbomagnesation thus affords nonracemic reaction products that bear an alkene and a carbinol group, functional groups that are readily amenable to a wide range of subsequent derivatization procedures. We have utilized the stereoselective ethylmagnesation shown in entry 1 of Table 10-2 as a key step in the first enantioselective total synthesis of the antifungal agent Sch 38516 [51]. As illustrated in Scheme 10-19, further functionalization of
10.4
647
Catalytic Reactions Effected by Ansa-Metallocenes
the Zr-catalyzed ethylmagnesation product through three subsequent catalytic procedures yields the requisite carboxylic acid synthon in >99% ee. This synthesis route underlines the utility of the Zr-catalyzed carbomagnesation protocol: the reaction product, since it carries an alkene and an alcohol function, can be readily functionalized to a variety of other nonracemic intermediates. 0.4 mol% (S)-(EBTHI)Zr-binol O
EtMgCl; ~70 turnovers
H2 N
OH
Me Me
HO
Me
6
O O
OH 5
1 O
Me
1. 3 mol% Cp2TiCl2, n-PrMgCl; Me
3 mol% (PPh3)2NiCl2
1 O
Me
HN
Sch 38516
OH
Br ; 72%
2. 5 mol% TPAP; NMO, H2O 65%
Scheme 19. Demonstration of the utility of (EBTHI)Zr-catalyzed ethylmagnesation in the enantioselective synthesis of the macrolactam aglycon of Sch 38516.
The catalytic cycle that we have proposed to account for the enantioselective ethylmagnesations is illustrated in Scheme 10-20. Asymmetric carbomagnesation is initiated by the chiral zirconocene–ethylene complex (R)-38, formed upon reaction of dichloride (R)-2 with EtMgCl (Eq. (10-14); the dichloride salt or the binol complex may be used with equal efficiency] [52]. Coupling of the alkene substrate with (R)38 leads to the formation of the metallacyclopentane intermediate i. In the proposed catalytic cycle, reaction of i with EtMgCl affords zirconate ii, which undergoes Zr–Mg ligand exchange to yield iii. Subsequent β-hydride abstraction, accompanied by intramolecular magnesium alkoxide elimination, leads to the release of the carbomagnesation product and regeneration of (R)-38 [53].
Zr Cl Cl
(R)-2
2 EtMgCl Zr
THF
(R)-38
(n)
(10-14)
648
10
Chiral Titanocenes and Zirconocenes in Synthesis
An important aspect in the carbomagnesation of six-membered and larger heterocycles is the exclusive intermediacy of metallacyclopentanes where C–Zr bond is formed α to the heterocycle C–O bond. Whether the regioselectivity in the zirconacycle formation is kinetically nonselective and rapidly reversible, where it is only one regioisomer that is active in the catalytic cycle, or whether formation of the metallacycle is kinetically selective (stabilization of electron density upon formation of C–Zr bond by the adjacent C–O) [54], has not been rigorously determined. However, as will be discussed below, the regioselectivity with which the intermediate zirconacyclopentane is formed is critical in the (EBTHI)Zr-catalyzed kinetic resolution of heterocyclic alkenes.
H
Zr
Zr
2 equiv Cl EtMgCl
MgCl Et
X
i (R)-38
H
Zr
Zr
Cl
(R)-2
EtMgCl
X
ii
X
MgCl(X) H
Et
CH3 ZrL
H ClMg
H
ZrL H
ClMg
X L = EBTHI X = O, N(H)n-nonyl
X
X
iii
Scheme 20. Catalytic cycle proposed by Hoveyda for the (EBTHI)Zr-catalyzed ethyl nesation of unsaturated heterocycles.
Why does the (EBTHI)Zr system induce such high levels of enantioselectivity in the C–C bond formation process? It is plausible that the observed levels of enantioselection arise from minimization of unfavorable steric and torsional interactions in the complex that is formed between (R)-38 and the heterocycle substrates (Scheme 10-20). The alternative mode of addition, illustrated in Fig. 10-1, would lead to costly steric repulsions between the olefin substituents and the cyclohexyl group of the chiral ligand [53]. Thus, reactions of simple terminal olefins under identical conditions results in little or no enantioselectivity. This is presumably because in the absence of the alkenyl substituent (of the carbon that bonds with Zr in i) the aforementioned steric interactions are ameliorated and the olefin substrate reacts indiscriminately through the two modes of substrate–catalyst binding represented in Fig. 10-1.
10.4
649
Catalytic Reactions Effected by Ansa-Metallocenes
These alkylation processes become particularly attractive when used in conjunction with the powerful catalytic ring-closing metathesis protocols. The requisite starting materials can be readily prepared catalytically and in high yield [55]. The examples shown in Scheme 10-21 demonstrate that synthesis of the heterocyclic alkene and subsequent alkylation can be carried out in a single vessel to afford unsaturated alcohols and amides in good yield and >99% ee (judged by GLC analysis) [56]. PCy3 Cl Cl
Ru PCy3
Ph H
CH3
Ph
2 mol%, THF; O
EtMgCl, 10 mol% (R)-(EBTHI)ZrCl 2 73% yield
OH
>99% ee
PCy3 Cl Cl
Ru
Ph
PCy3
Ph
(R)-(EBTHI)Zr-binol 10 mol%, EtMgCl
2 mol% N Ts
CH 2Cl2, 45 °C
N
68% yield
Ts
THF , 70 °C 77% yield
H
CH3
NHTs
>98% ee One pot process (CH 2Cl 2→ THF; 45→ 70 °C); 54% yield; >98% ee
Scheme 21. Ru-catalyzed ring closing metathesis processes, in conjunction with Zr-catalyz enantioselective alkylation reactions provide a convenient protocol for efficient synthesis o optically pure materials.
Reactions where higher alkyls of magnesium are used (Table 10-3), namely nPrMgCl and nBuMgCl, proceed less efficiently (35–40% isolated yield) but with similarly high levels of enantioselection (>90% ee) [53]. A number of issues in connection to the data illustrated in Table 10-3 merit comment. (1) With 2,5-dihydrofuran as substrate, at 22°C a mixture of branched (39 or 41) and n-alkyl products (40 or 42)
650
10
Chiral Titanocenes and Zirconocenes in Synthesis
are obtained. When the reaction mixture is heated to 70°C, the branched adducts 39 and 41 become the major product isomers. In contrast, with the six-membered heterocycle, reactions at 22°C are highly selective (entries 3 and 6, Table 10-3). (2) With nBuMgCl as the alkylating agent, high levels of diastereoselection is observed as well (entry 6). Table 10-3. (EBTHI)Zr-catalyzed enantioselective carbomagnesation of unsaturated heterocycles with longer chain alkylmagnesium chlorides [a] entry
substrate
major product(s)
temp (°C)
RMgCl
22
regioselectivity
ee, %
diastereoselectivity
CH3
1
CH3
CH3
2
O OH
nPrMgCl
2:1
99 (39), 99 (40)
--
70
nPrMgCl
20 : 1
94 (39)
--
22
nPrMgCl
>25 : 1
98
--
22
OH
39
40 CH3
3
CH3 O
OH CH3
4 5
CH3
nBuMgCl
2:1
70
nBuMgCl
15 : 1
22
nBuMgCl
>25 : 1
>99 (41), >99 (42)
15 : 1
CH3
O OH
90 (41)
13 : 1
OH
41
42 CH3 CH3
6 O
>95
>25 : 1
OH
[a] Reaction conditions: 5 equiv. alkylMgCl, 10 mol% (R)-2, 16 h; all yields: 35– 40% after silica gel chromatography. The aforementioned observations carry significant mechanistic implications [53]. As illustrated in Eqs. (10-15)–(10-17), in the chemistry of zirconocene–alkene complexes that are derived from the longer chain alkylmagnesium halides several additional selectivity issues present themselves: (1) The derived transition metal– alkene complex can exist in two diastereomeric forms, exemplified in Eqs. (1015)–(10-17) with (R)-43 anti and syn; reaction through these stereoisomeric complexes can lead to the formation of different product diastereomers [(compare Eqs. (10-15) and (10-16), or Eqs. (10-16) and (10-17)]. The data in Table 10-3 indicate that the mode of addition shown in Eq. (10-16) is preferred. (2) As illustrated in Eqs. (10-15) and (10-16), the carbomagnesation process can afford either the branched or the n-alkyl product. Alkene substrate insertion from the more substituted front of the zirconocene–alkene system affords the branched isomer (Eq. (1016)), whereas reaction from the less substituted end of the (EBTHI)Zr–olefin system leads to the formation of the straight chain product (Eq. (10-15)). The results shown in Table 10-3 indicate that, depending on the reaction conditions, products derived from the two isomeric metallacyclopentane intermediates can be competitive.
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
651
(10-15 to 10-17)
Detailed mechanistic studies from these laboratories shed light on the mechanistic intricacies of asymmetric catalytic carbomagnesations, allowing for an understanding of the above trends in regio- and stereoselectivity [53]. Importantly, our mechanistic studies clearly indicate that there is no preference for the formation of either the anti or the syn (EBTHI)Zr-olefin isomers (e.g., 43 anti vs 43 syn); it is only that one metallocene–alkene diastereomer (syn) is more reactive. Our mechanistic studies also indicate that zirconacyclopentane intermediates (i in Scheme 1020) do not spontaneously eliminate to the derived zirconocene alkoxide; Zr–Mg ligand exchange is likely a prerequisite for the alkoxide elimination and formation of the terminal alkene.
Kinetic resolution through catalytic carbomagnesation The high levels of enantioselectivity obtained in the asymmetric catalytic carbomagnesation reactions (Tables 10-2 and 10-3) imply highly organized (EBTHI)Zr–alkene complex interaction with the heterocyclic alkene substrates. It therefore follows that if chiral unsaturated pyrans or furans are employed, the resident center of asymmetry may induce differential rates of reaction, such that after ~50% conversion one enantiomer of the chiral alkene can be recovered in high enantiomeric purity. As an example, molecular models indicate that with a 2-substituted pyran, as shown in Fig. 10-2, mode of addition labeled as I should be significantly favored over II or III, where unfavorable steric interactions between the (EBTHI)Zr complex and the olefinic substrate should lead to significant catalyst– substrate complex destabilization.
652
10
Chiral Titanocenes and Zirconocenes in Synthesis
O H Zr R
Zr R1
5
R2 O 6
I R 1 = H, R 2 = alkyl II R 1 = alkyl, R2 = H
III R = alkyl
Figure 2. Preferential association of one pyran enantio with (EBTHI)Zr-ethylene complex.
As the data in Table 10-4 suggest, in the presence of catalytic amounts of nonracemic (EBTHI)ZrCl2, a variety of unsaturated pyrans can be resolved effectively to deliver these synthetically useful heterocycles in excellent enantiomeric purity [57]. A number of important issues in connection to the catalytic kinetic resolution of pyrans are noteworthy: (1) Reactions that are performed at elevated temperatures (70°C) afford recovered starting materials with significantly higher levels of enantiomeric purity, compared to processes carried out at 22°C. For example, the 2-substituted pyran shown in entry 1 of Table 10-4, when subjected to the same reaction conditions but at room temperature, is recovered after 60% conversion in only 88% ee (vs 96% ee at 70°C). Experiments that help establish the mechanistic basis of the above temperature effects are in progress. (2) Consistent with the models illustrated in Fig. 10-2, 6-substituted pyrans (Table 10-4, entry 2) are not resolved effectively (C6 substituent should not strongly interact with the catalyst structure; see Fig. 10-2); however, with a C2 substituent also present, resolution proceeds with excellent efficiency (entry 3). (3) Pyrans that bear a C5 group are resolved with high selectivity as well (entry 4). In this class of substrates, one enantiomer reacts more slowly, presumably because its association with the zirconocene–alkene complex leads to sterically unfavorable interactions between the C5 alkyl unit and the coordinated ethylene ligand.
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
653
Table 10-4. (EBTHI)Zr-catalyzed kinetic resolution of unsaturated pyrans [a]
entry
conversion mol% cat. (%)
substrate
unreacted subs. config., ee(%)
Me
1
60
10
R, 96
60
10
S, 41
a R = MgCl
56
10
R, >99
b R = TBS
60
10
R, >99
58
20
R, 99
a R = MgCl
63
10
R, >99
b R = TBS
61
10
R, 94
2 Me
O
2
6 O
Me
3
2 O OR
4
Me
5 O
Me
OR
5
2 O
[a] Reaction conditions: indicated mol% (R)-2, 5.0 equiv of EtMgCl, 70°C, THF. Mass recovery in all reactions is >85%.
As the representative data in Table 10-5 demonstrate, the Zr-catalyzed resolution technology may be applied to medium ring heterocycles as well [56]; in certain instances (e.g., entries 1–2) the recovered starting material can be obtained with outstanding enantiomeric purity. Comparison of the data shown in entries 1 and 3 of Table 10-5 implies that the presence of an aromatic substituent can have an adverse influence on the outcome of the catalytic resolution. That the eight-membered ring substrate in Table 10-5 (entry 4) is resolved more efficiently suggest that the origin of the adverse influence is more due to conformational preferences of the heterocycle than the attendant electronic factors (α phenoxy group is a better leaving group that an alkoxy unit).
654
10
Chiral Titanocenes and Zirconocenes in Synthesis
Table 10-5. Zirconocene-catalyzed kinetic resolution of 2-substituted medium ring heterocycles [a] entry
substrate
conversion (%)
time
unreacted subs. config., ee (%)
58
30 min
R, >99
63
100 min
R, 96
60
8 hr
R, 60
63
11 hr
R, 79
Me
1 O
Me
2
OTBS
O
3 O
Me
4 O
Me
[a] Reaction conditions: 10 mol% (R)-2, 5.0 equiv of EtMgCl, 70 °C, THF. Mass recovery in all reactions is >85%.
Availability of oxepins that carry a side chain containing a Lewis basic oxygen atom (e.g., entry 2, Table 10-5) has further important implications in enantioselective synthesis: The derived alcohol, benzyl ether or MEM ethers, where resident Lewis basic heteroatoms are less sterically hindered, undergo diastereoselective uncatalyzed alkylation reactions readily when treated to a variety of Grignard reagents [58]. The examples shown below (Scheme 10-22) serve to demonstrate the excellent synthetic potential of these stereoselective alkylation technologies. Thus, resolution of the TBS-protected oxepin 45, conversion to the derived alcohol and diastereoselective alkylation with nBuMgBr affords 46 with >96% ee in 93% yield. Alkylation of (S)-47 with an alkyne-bearing Grignard agent (→(S)-48), allows for a subsequent Pauson–Khand cyclization reaction to provide the corresponding bicycle 49 in the optically pure form. In connection with the facility of these olefin alkylations, it is important to note that the asymmetric Zr-catalyzed alkylations with longer chain alkylmagnesium halides (see Table 10-3) are more sluggish than those involving EtMgCl. Furthermore, when catalytic alkylation does occur, the corresponding branched products are obtained; that is, with nPrMgCl and nBuMgCl, isoPr and secBu addition products are formed, respectively [53]. The uncatalyzed alkylation reaction described here thus complements the enantioselective Zr-catalyzed protocol.
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
655
Zirconocene-catalyzed kinetic resolution of dihydrofurans is also possible, as illustrated in Scheme 10-23 [59]. Unlike their six-membered ring counterparts, both of the heterocycle enantiomers react readily, but through distinctly different reaction pathways, to afford—in high diastereomeric and enantiomeric purity—constitutional isomers which are readily separable. A plausible reason for the difference in the reactivity pattern of pyrans and furans is that, in the latter group of compounds, both olefinic carbons are adjacent to a C–O bond: C–Zr bond formation can take place at either end of the π-system. Thus, the furan substrate and the (EBTHI)Zr– alkene complex (38) approach one another so that unfavorable steric interactions are avoided, leading to the formation of readily separable nonracemic products in the manner illustrated in Scheme 10-23.
656
10
Chiral Titanocenes and Zirconocenes in Synthesis
As depicted in Eqs. (10-18)–(10-20), kinetic resolution of a variety of cyclic allylic ethers is effected by asymmetric Zr-catalyzed carbomagnesation. Importantly, in addition to six-membered ethers, seven- and eight-membered ring systems can be readily resolved by the Zr-catalyzed protocol. It is worthy of note that the powerful Ti-catalyzed asymmetric epoxidation procedure of Sharpless [60] is often used in the preparation of optically pure acyclic allylic alcohols through the catalytic kinetic resolution of easily accessible racemic mixtures [61]. When the catalytic epoxidation is applied to cyclic allylic substrates, reaction rates are retarded and lower levels of enantioselectivity are observed. Ru-catalyzed asymmetric hydrogenation has been employed by Noyori to effect resolution of five- and six-membered allylic carbinols [62]; in this instance, as with the Ti-catalyzed procedure, the presence of an unprotected hydroxyl function is required.
OPh
(R)-(EBTHI)Zr-binol 20 mol%
OPh
(10-18 to 10-20) 98% ee @ 55% conversion
(r)
>99% ee @ 60% conversion
(s)
81% ee @ 59% conversion
(t)
5 equiv EtMgCl Me
70 °C
Me
(racemic) OBn
(R)-(EBTHI)Zr-binol 10 mol%
OBn
5 equiv EtMgCl 70 °C OPh
OPh
(R)-(EBTHI)Zr-binol 15 mol% 5 equiv EtMgCl 70 °C
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
657
Modes of addition shown in Fig. 10-3 are similar to those shown in Fig. 10-2 and are consistent with extant mechanistic work [52, 53]; they accurately predict the identity of the slower reacting enantiomer. It must be noted, however, that variations in the observed levels of selectivity as a function of the steric and electronic nature of substituents and the ring size cannot be predicted based on these models alone; more subtle factors are clearly at work. In spite of such mechanistic questions, the metal-catalyzed resolution protocol provides an attractive option in asymmetric synthesis. This is because, although maximum possible yield is ~40%, it requires easily accessible racemic starting materials and conversion levels can be manipulated so that truly pure samples of substrate enantiomers are obtained.
H
Zr
Zr
H
RO
RO
Addition pathway available to the faster-reacting enantiomer
Zr RO
H
Addition pathways available to the slower-reacting enantiomer
Figure 3. Various modes of addition of cyclic allylic ethers to (EBTHI)Zr-alkene complex.
The synthetic versatility and significance of the Zr-catalyzed kinetic resolution of cyclic allylic ethers is readily demonstrated in the example provided in Scheme 1024. Optically pure starting allylic ether, obtained by the above mentioned catalytic kinetic resolution, undergoes a facile Ru-catalyzed rearrangement to afford chromene in >99% ee [63]. Unlike unsaturated pyrans discussed above, chiral 2substituted chromenes are not readily resolved by the Zr-catalyzed protocol. Optically pure styrenyl ethers, such as that shown in Scheme 10-24, are readily obtained by the Zr-catalyzed kinetic resolution, allowing for the efficient and enantioselective preparation of these important chromene heterocycles by a sequential catalytic protocol.
658
10
Chiral Titanocenes and Zirconocenes in Synthesis Cl Cl
(R)-(EBTHI)Zr-binol (10 mol%) Me
O
O
PCy3 Ph
Ru PCy3
H
Me
6 mol%
Ph
ethylene atm.
EtMgCl
>99% ee
60% conv.
O
H
81%
>99% ee
Scheme 24. Scheme 24. Tandem Zr-catalyzed kinetic resolution and Ru-catalyzed rearrangement affords chiral chromenes in high enantiomeric purity.
Comparison with related catalytic enantioselective processes. The zirconocene-catalyzed enantioselective carbomagnesation accomplishes the addition of an alkylmagnesium halide to an alkene, where the resulting carbometallation product is suitable for a variety of additional functionalization reactions (see Scheme 10-19). Excellent enantioselectivity is obtained in reactions with Et-, nPr- and nBuMgCl, and the catalytic resolution processes allow for preparation of a variety of nonracemic heterocycles. Nonetheless, the development of reaction processes where a larger variety of olefinic substrates and alkylmetals (e.g., Me-, vinyl-, phenylmagnesium halides, etc.) can be added to unfunctionalized alkenes efficiently and enantioselectively stands as a challenging goal in enantioselective reaction design. As illustrated below (Eqs. (10-21) and (10-22)), recent reports by Negishi and coworkers, where Erker’s nonbridged chiral zirconocene [64] is used as catalyst, is an important and impressive step towards this end [65]. (10-21 and 10-22)
Cl
Zr
Cl
8 mol%
Me Me
1 equiv Me3Al, 22 °C; O2
Me
OH Me
92% yield, 74% ee OH
1 equiv Et3Al, 22 °C; O2
(u)
Me
Et
69% yield, 93% ee
(v)
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
659
10.4.1.5 Asymmetric Pauson–Khand Type Reactions One of the most useful reactions in organic chemistry, cyclocarbonylation of enynes to cyclopentenones (commonly known as the Pauson–Khand reaction), now has a Ti-catalyzed enantioselective variant [66]. As illustrated in Table 10-6, Buchwald and coworkers have reported that in the presence of 5–20 mol% (S)(EBTHI)TiMe2 ((S)-53) and 96 kPa (14 psi) CO (toluene, 90°C, 12–16 h), various enynes are converted to the corresponding enones with excellent enantioselectivity. When lower reaction temperatures are used, or when CO pressures below and above 96 kPa are utilized, the catalytic process proves to be less efficient. Noteworthy is the lower enantioselectivity of the amine-containing tether versus the similar ether substrate (compare entries 1–3). As the example in entry 2 shows, enynes with geminal substitution cyclize more readily and require lower catalyst loading. It also merits mention that, as shown in entry 4 of Table 10-6, the 1,1-disubstituted alkene readily undergoes cyclization (unlike other versions of Pauson– Khand reaction), albeit with relatively lower levels of enantioselection. Table 10-6. Ti-catalyzed enantioselective enyne cyclocarbonylation entry
alkene
product
mol% [(S)-53)]
yield [%]
ee [%]
20
85
96
7.5
92
94
10
84
74
20
90
72
Ph Ph
1
O
O
O H Ph
2
Ph
MeO2C
MeO2C O MeO2C
MeO2C
H Ph Ph
3
BOC N
BOC N
O H Ph
4
Ph
MeO2C
MeO2C O MeO2C
MeO2C Me
Me
A tentative mechanism has been proposed that is illustrated in Scheme 10-25. Thus, displacement of the CO ligands by the alkyne and olefin moieties, formation
660
10
Chiral Titanocenes and Zirconocenes in Synthesis
of the titanacyclopentene, CO insertion, and subsequent reductive elimination affords the final product, releasing the active catalyst. Also illustrated in Scheme 1025, are the models that rationalize the observed stereochemical outcome based on steric interactions between the substrate and the chiral ligand during the metallacyclopentane formation; the latter stage is proposed to be the stereochemistry-determining step of the catalytic cycle. (EBTHI)TiMe2 Ph
CO
Ph O
O
O
(EBTHI)Ti(CO)2 + 2 CO
Ph
Ph LTi
LTi
O
CO
O
O Ph
Ph
CO
Ph
LTi
O
H
Ti O
LTi
Ph
O
Ti
CO
O
H
Scheme 25. Catalytic cycle proposed by Buchwald for the Ti-catalyzed enyne carbocyclization reaction.
10.4.2 Catalytic Asymmetric C–H Bond Forming Reactions 10.4.2.1 Asymmetric Hydrogenation of Terminal and 1,1-Disubstituted Alkenes In the last five years, much progress has been made in the ansa-metallocene-catalyzed hydrogenation of alkenes. Both Zr- and Ti-based catalytic systems have been used for this purpose; terminal, 1,1-disubstituted olefins, as well as trisubstituted alkenes have been used as substrates. By far, the most successful hydrogenations
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
661
involve trisubstituted alkenes, where enantioselectivities in the range of 83–99% ee are reported. Initial studies in this area have been carried out by Waymouth and Pino, where the polymerization catalyst, (EBTHI)ZrX2/MAO are used in the presence of H2 to effect alkene hydrogenation [32]. Pino and Waymouth recognized that, in certain instances, olefin polymerization proceeds sluggishly because hydrogenation of the metal–alkyl intermediate takes place significantly faster than alkene insertion, leading to the formation of hydrogenated monomers in large excess (compared to the corresponding alkene oligomers and polymers). Table 10-7 illustrates the marked difference in reaction efficiency which arises when different sources of metallocene are employed in the (EBTHI)Zr-catalyzed hydrogenation. The mechanism by which MAO converts the zirconium (IV) salts to active zirconium hydride species remains unclear [67]. However, it has been proposed that a chlorine atom may form an η2-bridge between aluminum and zirconium when the dichloride salt is used, thus preventing formation of the active cationic metal center [32]. Table 10-7. Hydrogenation of styrene as a function of the zirconium catalyst precursor catalyst precursor
gas
Po (kPa; atm)
yield (%)
(EBTHI)ZrMe2
H2
2.0x103; 20
93
(EBTHI)Zr-binol (EBTHI)ZrCl2
D2 H2
3
89
3
0
1.7x10 ; 17.5 2.0x10 ; 20
As was mentioned before, and as illustrated in Table 10-8, an important prerequisite for efficient cationic (EBTHI)Zr-catalyzed hydrogenation is that the alkene should not be able to polymerize in a facile manner. Whereas with 1-decene as substrate, only 28% of the hydrogenation product is obtained (Table 10-8, entry 1; remainder of the product is the derived polymer), 2-methyl-1-pentene affords 97% of the hydrogenation adduct (Table 10-8, entry 3). Interestingly, in a competition experiment, trans-2-hexene is hydrogenated much faster than its corresponding cis isomer (Table 10-8, entry 4).
662
10
Chiral Titanocenes and Zirconocenes in Synthesis
Table 10-8. Hydrogenation of alkenes with (EBTHI)ZrMe2/MAO as precatalyst entry
Po (kPa; atm)
substrate
1
conv yield (%) (%)
103; 10
octyl
95
entry
28
conv yield Po (kPa; atm) (%) (%)
substrate
Me
96
Me 2
10 ; 1
4 2x103; 20
2
94
Me
93
50 (combined) 4
Me Me
5 3
3
Me
2x10 ; 20
100
Me
2x103; 20
20
20
97
Data in Table 10-9 illustrate that when the nonracemic (EBTHI)Zr system is used to catalyze the hydrogenation of prochiral alkenes, moderate levels of enantiofacial differentiation are observed (23–65% ee). Enantioselective deuteration of pentene occurs in low yield but shows noticeable enantioselection (23% ee). The same reaction with styrene proceeds in 61% yield and with moderate enantioselectivity (65% ee). Hydrogenation of 2-phenyl-1-pentene proceeds in excellent yield but with poor control of stereochemistry (95% yield, 36% ee). Table 10-9. Enantioselective hydrogenation and deuteration of alkenes catalyzed by (R)-(EBTHI)Zr-binol and MAO (at 25°C) substrate Me
gas
cat (mol%)
yield (%)
ee (%)
prodct config
D2
-
10
23
R
D2
6.0x10-4
61
65
R
H2
2.2x10-4
95
36
R
Ph Me
An interesting aspect of the (EBTHI)ZrX2-catalyzed hydrogenation and deuteration of alkenes is that these reactions occur with opposite sense of stereochemistry when compared to similar oligomerization reactions. Thus, as the example in Scheme 10-26 illustrates, the prochiral π-face that is hydrogenated is the opposite face on which carbon–carbon bond formation takes place in the related polymerization process.
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
663
preferred π-face Me
1-pentene
Pr
oligomerization Me
Me D
1-pentene
deuteration
D
Me Pr
Scheme 26. Variations in enantiofacial selectivity in reactions catalyzed by (EBTHI)Zr-binol/MAO.
Waymouth has suggested that (EBTHI)Zr-catalyzed hydrogenation occurs from the opposite π-face that undergoes polymerization because a sterically demanding polymer chain would prefer to associate with the zirconocene catalyst in the manner shown in Scheme 10-27. That is, the pendant groups on the alkene substrate may be oriented such that most steric interactions with the EBTHI ligand are avoided. This mode of catalyst–substrate interaction necessitates that the reacting alkene adopt a side-on approach towards the (EBTHI)Zr system in the oligomerization or polymerization processes. In contrast, in the hydrogenation of alkenes, illustrated in Scheme 10-27, the small hydrogen atom may readily adopt the side position below the tetrahydroindenyl ligand, allowing the larger alkene to approach the metal hydride from the sterically more accessible front face.
Polymerization
R M
Hydrogenation
R M
P
Side Approach Side View
M
M
H
R
P
Side Approach Front View
H
R
Frontal Approach Side View
Frontal Approach Front View
P = polymer chain
Scheme 27. Transition structures proposed by Waymouth for the polymerization and hydrogenation reactions of alkenes catalyzed by (EBTHI)ZrX2/MAO.
10.4.2.2 Asymmetric Hydrogenation of Trisubstituted Alkenes Hydrogenation of unfunctionalized trisubstituted olefins with bridged titanocene complexes as precatalysts has recently been reported by Buchwald and coworkers
664
10
Chiral Titanocenes and Zirconocenes in Synthesis
[68]. Unfunctionalized alkenes are appropriate substrates and excellent levels of enantioselectivity are observed. The active catalyst in these systems is postulated to be chiral titanocene (S)-54, produced by treatment of nonracemic (EBTHI)Ti-binol with two equivalents of butyllithium under H2 atmosphere, followed by the addition of 2.5 equivalents of phenylsilane (Eq. (10-23).
Ti
1. nBuLi (2 equiv) H2 (1 atm), 0 °C
O O
9
Ti
H
(w)
(10-23)
2. PhSiH3 (2.5 equiv)
(S)-54
As depicted in Table 10-10, hydrogenation of trisubstituted olefins catalyzed by 5 mol% 54 occurs in high yields and >90% ee. Aromatic olefins, as well as allylic amines and ethers are suitable substrates (entries 4 and 5). Although it is not a stated prerequisite, all substrates reported are olefins attached to an aromatic ring. Interestingly, E and Z alkenes react with much different efficiency under the reported conditions. E-α,β-Dimethyl-p-methoxystyrene is hydrogenated in 48 h to give a 79% yield of the (R) enantiomer of the product in 95% ee (entry 2, Table 10-10). However, when a 2:1 mixture of E:Z alkenes is used as starting material, the reaction requires 146 h to give an 80% yield of the (R) isomer in only 31% ee [69].
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
665
Table 10-10. Enantioselective hydrogenation of trisubstituted alkenes with (EBTHI)Ti-H [a] entry
product
alkene Me
[b]
ee (%)
91
>99
79
95
77
92
75
95
86
94
Me
1 Me
Me Me
Me
2 MeO
MeO
MeO
Me
MeO
Me
3 Me
Me
4
5 MeO
yield (%)
Me
NBn2
NBn2
OMe
OMe MeO
Me
[a] Conditions: 5 mol% (S)-(EBTHI)TiH, 1.38 × 104 kPa (2000 psi H2), 65°C, THF. [b] Absolute stereochemistry was determined only for entries 1 and 2. Transition state structures have been put forth by Buchwald to rationalize the absolute stereochemistry afforded by the (EBTHI)Ti-catalyzed hydrogenation [68]. In the proposed mechanism, the substituent attached to the monosubstituted terminus of the olefin appears to control which prochiral alkene face reacts preferably. If we assume that the hydride attached to titanium takes the side position beneath the cyclohexyl ring of the EBTHI ligand, the alkene should approach the metal center from the front face of the complex (Ti prefers to bond to the less substituted carbon of the olefin; see Scheme 10-28). Hence, there are two modes of attack for either type of olefin substrate (E or Z) that could lead to the hydrogenation product;
666
10
Chiral Titanocenes and Zirconocenes in Synthesis
in both instances, one mode of reaction involves interaction of two of the alkene substituents with the large cyclohexyl ring of the catalyst (II and IV). Because the lower energy path involves an alkene–cyclohexyl interaction (I and III), there is a reactivity difference between E and Z alkenes in the hydrogenation reaction. RL
H
RS
RL
E-Olefin
H Ti
H
Rs H Ti H
RS
RL
I Favored
Ti
II Disfavored
H
45
54
RS H Ti RS
H
RL
H
RL H Ti H RS
RL
III Favored
IV Disfavored
Z-Olefin
Scheme 28. Transition structure models proposed by Buchwald for (EBTHI)Ti-catalyzed hydrogenation of alkenes.
Comparison with related catalytic enantioselective processes. Significant advances in the area of catalytic enantioselective hydrogenation have been made, where various rhodium, ruthenium and cobalt complexes are used as catalysts [70]. However, similar processes with trisubstituted alkenes as substrates are often nonstereoselective [71]. Therefore, the titanocene-catalyzed hydrogenation of trisubstituted alkenes reported by Buchwald represents the first catalytic transformation that accomplishes this important task with excellent efficiency and enantioselectivity. In spite of the importance of this achievement, additional progress with regard to the development of catalytic methods that are effective in the hydrogenation of a wider range of olefinic substrates will undoubtedly remain a preoccupation of researchers who are active in this area. In this context, the highly enantioselective hydrogenation of α,β-unsaturated carbonyl systems by Pfaltz [exemplified below in Eqs. (1024) and (10-25)] represents a major advance in the field [72].
10.4 Me
O
Ph
Me OEt CN
OEt
N
N TBSO
(10-24 and 10-25)
Me
OTBS
1.2 mol%
1 mol% CoCl2, NaBH4, 25 °C
O
(x)
97% yield, 94% ee
H
Ph
O
Ph
Pfaltz, 1989
667
Catalytic Reactions Effected by Ansa-Metallocenes
Me
OEt
O
Ph
OEt
(y)
95% yield, 94% ee
10.4.2.3 Asymmetric Hydrogenation of Imines Asymmetric synthesis through catalytic (EBTHI)Ti-catalyzed imine reduction. As illustrated in Table 10-11, Buchwald and coworkers have reported that titanocene 54 is an effective catalyst for enantioselective hydrogenation of imines [73]. With 5 mol% of the chiral Ti complex and under 1.38 × 104 kPa (2000 psi) H2 atm at 65°C, imines are reduced to amines in good yield and with 58–98% ee [74a]. The lower selectivity observed in the reduction of acyclic imines, as compared to cyclic substrates, could arise because acyclic imines are often composed of a mixture of cis and trans isomers; similar to the hydrogenation of alkenes, the two imine isomers might be expected to react through opposite enantiofaces [74]. Table 10-11. Enantioselective hydrogenation of imines with (EBTHI)Ti-binol as precatalyst [a] entry
product
substrate Me
1
Me
ee (%)
68
58
82
98
Me N
Ph
MeO
Me
N H
Ph
MeO
2 N
MeO
NH
MeO
Me
3
yield (%)
Me
Me
N
Me
N H
Me
4
98
Me N
O
81
Ph O
N H
Ph
70
53
[a] Reaction conditions: 5 mol% (S)-54, 1.38 × 104 kPa (2000 psi) H2, 65°C, THF. [b] Absolute stereochemistry was determined only for entry 2.
668
10
Chiral Titanocenes and Zirconocenes in Synthesis RNH 'R R"
R'
*
H
R"
RN
L*Ti
i ≠
R" R'
*
≠ H
R L*Ti
N L*Ti
N R
H H
iii
H2
R'
*
R"
turnover limiting step L*Ti
R H N R'
*
R"
ii Scheme 29. The proposed mechanism for the (EBTHI)Ti-catalyzed reduction of imines.
In more recent work, Buchwald has demonstrated that the (EBTHI)Ti-catalyzed reduction of cyclic imines proceeds notably faster than that of related acyclic substrates [74b]. Kinetic studies indicate that the catalytic cycle shown in Scheme 1029 represents a plausible mechanism for the Ti-catalyzed reduction process. Specifically, various experimental evidence indicate that hydrogenolysis of the Ti–N bond may be the turnover-limiting step in the titanium-catalyzed imine reduction (ii→iii→i) [74c]. Buchwald has suggested that when a cyclic amine is employed as substrate, where the two nitrogen substituents are tethered, the Ti–N bond is generally more accessible to the incoming H2 molecule and therefore undergoes hydrogenation (iii→i, Scheme 10-29) faster than a related acyclic system (where Ti–N bonds are effectively more hindered) [75]. Buchwald and coworkers subsequently established that with cyclic imines, 551 kPa (80 psi) H2 pressure (vs. 1.38 × 104 kPa or 2000 psi) is sufficient for the Ticatalyzed reduction to proceed smoothly [74b]. Thus, 55–62 (Scheme 10-30) are reduced with 5 mol% 54 (see Eq. (10-23)) under 551 kPa (80 psi) H2 to yield the desired amines in >70% yield and >98% ee. In addition, these workers reported that with the more reactive cyclic imines (as opposed to acyclic imines), exemplified by 55, substrates are reduced with as little as 1 mol% catalyst. Whereas terminal alkenes present in the substrate may be hydrogenated along with the imine, diand trisubstituted olefins are relatively inert to the reaction conditions (62 contained ~16% of the E olefin and ~17% of the saturated analog). In the case of hydroxyl-substituted imines, phenylsilane is used to bring about in situ protection of the resulting carbinol site; hence hydroxyl imines may be subjected to the catalytic
10.4
669
Catalytic Reactions Effected by Ansa-Metallocenes
reduction conditions as well (e.g., 61 with one equivalent excess phenylsilane). Five-, six- and seven-membered cyclic imines, as well as substrates bearing silyl ethers, acetals and protected pyrroles can be utilized in these transformations. Me N
Me
N
55
N
56
N
57
58
O TBSO
4
HO
O
N
59
N
4
60
Me
N
61
5
N
62
Scheme 30. Substrates suitable for (EBTHI)Ti-catalyzed reduction of imines.
Kinetic resolution through catalytic (EBTHI)Ti-catalyzed imine reduction. The possibility of effecting a catalytic kinetic resolution of chiral imines through the titanium-catalyzed enantioselective hydrogenation reaction has also been investigated [76]. As shown in Table 10-12, subjection of (±)-5-methyl-2-phenyl-1-pyrroline to the hydrogenation conditions provides both the recovered starting material and the product in 99% ee at ~50 % conversion (entry 1). Whereas pyrrolines substituted at C5 are readily resolved, those carrying substituents at C3 and C4 undergo catalytic resolution less efficiently (compare entries 1–2 to entries 3–4 in Table 10-12).
Ti H
N
R H
R H
Ti R' H
R'
A
N
B
Transition State for Matched Enantiomer
Transition State for Mismatched Enantiomer
R' Ti
H
N
C
R
Scheme 31. Transition state models proposed by Buchwald for the (EBTHI)Ti-H-catalyzed reduction of cyclic imines.
670
10
Chiral Titanocenes and Zirconocenes in Synthesis
As depicted in Scheme 10-31, models that account for the reactivity and selectivity patterns observed in (EBTHI)-Ti-catalyzed reduction of imines are similar to those offered for catalytic hydrogenation of alkenes (cf. Scheme 10-28). It has been suggested that in the intermediate, where C–H bond formation occurs, the Ti–H is positioned such that it can overlap with the π* of the imine [77]. That is, bond formation proceeds through a four-centered transition structure where the pyrroline ring is in a position so that steric interactions between the group attached to the imine carbon (R) and the tetrahydroindenyl ligand are avoided (A and B in Scheme 10-31, a complex such as C suffers interactions between R and the cyclohexyl group of the chiral ligand). Furthermore, as shown in A (Scheme 10-31), it should be preferable that the heterocycle is situated in a manner where the substituent on the stereogenic carbon (R′) is oriented away from the catalyst structure. With the mismatched substrate enantiomer (B), the R′ group is positioned directly under the cyclohexyl group of the ligand and suffers from unfavorable steric interactions with the cyclohexyl ring of the chiral catalyst.[78] Table 10-12. (EBTHI)Ti-catalyzed kinetic resolution of cyclic imines [a]
entry
1
2
Me
N
Me
N
Ph
Me
N H
undecyl
Me
N H
Ph
undecyl
Ph
3 N
yield (%) [ee (%)] product
product
substrate
34 [99]
41 [>95]
Me
N
Me
N
N H
Ph
Ph
undecyl
37 [99]
41 [>95]
Ph
Ph
Ph
yield (%) [ee (%)] rec. starting mat.
recovered starting mat.
42 [>95] N
Ph
- [75]
6:1 syn:anti Ph
Ph
Ph
4
44 [99] N
Ph
N H
Ph
33 [49] N
Ph
3:1 syn:anti
[a] Conditions: 5 mol% (S)-(EBTHI)Ti-H, 551 kPa (80 psi) H2, 65°C, THF. According to the models proposed for the (EBTHI)Ti-catalyzed reduction of imines (Scheme 10-31), it may be predicted that the C3- and C4-substituted imines should be resolved under these conditions with equal efficiency. However, as shown in Table 10-12 (entry 4), a bulky phenyl group at C4 is not sufficient to allow effective kinetic resolution, and a benzyl substituent at the C3 site is significantly less selective than the corresponding C5-substituted starting materials. Thus, it is likely that there are additional conformational preferences in the pyrroline ring
10.4
671
Catalytic Reactions Effected by Ansa-Metallocenes
that allow alkyl groups at C4 to escape interaction with the catalyst–ligand complex structure. Comparison with related catalytic enantioselective processes. As illustrated below, synthesis of nonracemic amines through reduction of their unsaturated precursors has been the target of a number of important investigations. Other procedures exemplified by reports by Brunner and Wiegrebe (Eq. (10-26)) [79], Fryzuk and James (Eq. (10-27)) [80], Osborn (Eq. (10-28)) [81], Spindler (Eq. (10-29)) [82], and Burk (Eq. (10-30)) [83] offer alternative enantioselective methods for C=N reduction. Comparison of the previous work to that described above clearly indicates that the (EBTHI)Ti-catalyzed reduction of imines represents an important, valuable and useful addition to the variety of available methods for the reduction of imines. H O
Me N
•
[Rh(cod)Cl]2
Me
PPh2
H
PPh2
O
N
SiHPh2
H
2 mol%
(z)
H2SiPh 2, 0→20 °C Brunner & Wiegrebe, 1985
85% yield, 59% ee
Bn Me
N
PPh2
[Rh(nbd)Cl]2 •
Me
O
N Me
Me
Me
>99% yield, 84% ee
I
I
Ir
(10-26 to 10-29)
H
H Ph2
Ir P Ph2 H
N
(aa)
H2 (6.9x10 3 kPa or 103 psi), KI 20 °C Fryzuk & James, 1988 Ph2 H P
Bn
PPh2
1 mol%
Me
H
P
P H Ph2
O
HN Me
Me
Me
Me
0.001 mol%
(bb)
3
Osborn, 1990
H2 (4.0x10 kPa), 20 °C Ph2P
O
Ph2P
O
Me
H
H O
N Me
>99% conv., 54% ee
Me
Me
Me
[Rh(cod)Cl]2
•
Me
1 mol%
O
PPh2 PPh2
O
O
HN Me
Me
Me
Me
H
(cc)
3
Spindler, 1990
H2 (2.0x10 kPa), 18 °C
77% conv., 56% ee
672
10
Chiral Titanocenes and Zirconocenes in Synthesis OMe
N Ph
Me
OMe +
-
[(cod)Rh((R,R)-Et-duphos)] OTf
H N
N
0.2 mol% O
2
H2 (4.1x10 kPa or 60 psi), 20 °C
Burk, 1992
Ph
H N Me
(dd)
O
95% yield, 91% ee
(10-30)
In the context of related catalytic kinetic resolution procedures, Lensink reported in 1993 an effective method for the catalytic kinetic resolution of chiral acyclic imines [84]; an example of this process is illustrated below (Eq. (10-31)). Thus, the Rh-catalyzed and the (EBTHI)Ti-catalyzed resolution methods complement each other in offering convenient routes for the efficient enantioselective syntheses of a variety of chiral imines in the nonracemic form. PPh2 PPh2 Me N Ph
[Rh(cod)Cl]2 Ph
Me
•
Me Me
Me N
1 mol% 3
3
H2 (6.9x10 kPa or 10 psi), 22 °C
Lensink, 1993
Ph
Ph
(10-31)
Me
83% ee @ 67% conv.
10.4.2.4 Asymmetric Hydrogenation of Enamines Buchwald and Lee have reported the (EBTHI)TiH-catalyzed asymmetric hydrogenation of enamines [85]. The reaction conditions are sufficiently mild that aromatic chlorides, ethers and trisubstituted alkenes can be considered as compatible functional groups. However, use of an arylbromide as substrate leads to the reduction of the C–Br bond, presumably effected by a TiIII–hydride complex. As shown in Table 10-13, α-aryl tertiary amines are produced in good yield and with exceptional enantioselectivity; cyclic tertiary amines are most often employed as substrates. However, acyclic methyl benzyl enamines are suitable starting materials as well (entry 2, Table 10-13). It is worthy of note that the alkene group in cyclohexenyl enamine substrate shown in entry 3 of Table 10-13 is not hydrogenated under the reaction conditions.
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
673
Table 10-13. Catalytic and enantioselective hydrogenation of enamines with (EBTHI)Ti-H [a] entry
1
75
92
Me
83
96
Me
72
95
Me
88
91
Me
Me
N
ee [b] (%)
N
N
Me
Ph
N
2
N
Ph
N
3 O
O
N
N
4 MeO
yield (%)
product
substrate
MeO
[a] Conditions: 5 mol% (S)-(EBTHI)Ti-binol, 103–551 kPa (15–80 psi) H2, 65°C, THF. [b] Stereochemistry was determined only for entry 1.
Transition state models that account for the selectivity observed in the (EBTHI)Ticatalyzed asymmetric hydrogenation of enamines have been proposed by Buchwald and are depicted in Scheme 10-32. It has been suggested that one mode of catalyst–substrate association is significantly more favored if we assume that the phenyl group rotates out of conjugation with the enamine C=C to allow the nonbonding pair of electrons on nitrogen to retain their overlap with the olefin π cloud (severe A(1,2) allylic strain would exist if both the phenyl π system and nitrogen nonbonding electrons were to remain coplanar with the C=C bond). These conformational considerations lead to the preference for the top transition structure shown in Scheme 10-32. Although the exact physical basis for such conformational
674
10
Chiral Titanocenes and Zirconocenes in Synthesis
tendencies has not been rigorously established, the above mechanistic paradigm readily accounts for the observed enantioselectivities.
Comparison with related catalytic enantioselective processes. Catalytic enantioselective methods for the hydrogenation of enamides has been the subject of a number of reports [86]. However, the Buchwald study represents the first example of an asymmetric and catalytic reduction of an enamine. Therefore, the (EBTHI)TiH system in this instance has provided the synthetic chemist with a unique tool that permits the preparation of a variety of nonracemic amines efficiently and in high enantiopurity.
10.4.2.5 Asymmetric Hydrosilation of Ketones Development of methods that lead to the enantioselective reduction of ketones has been the subject of intense study in recent years. In general, reported methods are most effective when the two groups attached to the carbonyl differ significantly in size. In an effort to extend the scope of asymmetric ketone reduction, Buchwald and coworkers have employed the (EBTHI)Ti-H complex in the asymmetric catalytic reduction of ketones [87]. Thus, in the presence of polymethylhydrosilane, (R)-(EBTHI)Ti-H complex (generated from (R)-(EBTHI)Ti-binol and two equivalents of butyllithium) catalyzes the reduction of various ketones in high yield and with excellent enantioselectivity.
10.4
Catalytic Reactions Effected by Ansa-Metallocenes
675
Table 10-14. Catalytic and enantioselective reduction of ketones with (R)(EBTHI)Ti-H [a] entry
product
substrate O
Me
O
Me
O
Me
O
O
67
24
72
90
88
12
88
82
[b]
OH Me
4
97
OH Me
3
73
OH Me
2
ee (%)
OH Me
1
yield (%)
Me
OH [c]
5
[a] Product shown is that obtained after treatment of the hydrosilation products with TBAF or HCl. [b] 9 mol% catalyst used. [c] PhSiH3 used. As illustrated in Table 10-14, aryl-, and vinylmethyl ketones are selectively reduced to the corresponding nonracemic carbinols. Selectivity is markedly reduced however, when both of the ketone substituents are aliphatic (entry 4). The selective reduction of cyclohexyl phenyl ketone (entry 5), where both ketone substituents are of similar size, indicates that electronic effects can play a notable role in controlling stereoselectivity. In connection to the last case, Buchwald suggests that the two substituents may effectively have different sizes, because the phenyl and carbonyl π systems may remain coplanar and hence conformationally constrained, whereas the cyclohexyl ring might rotate to avoid costly steric interactions with the chiral ligand. Structures that account for the enantiofacial selectivity in the ketone
676
10
Chiral Titanocenes and Zirconocenes in Synthesis
hydrosilation are similar to those illustrated above for alkene, imine and enamine hydrogenation (cf. Schemes 10-28, 10-31 and 10-32). Comparison with related catalytic enantioselective processes. Illustrated below are examples of other highly enantioselective carbonyl reduction reported by Corey (Eq. (10-32)) [88], Evans (Eq. (10-33)) [89], Bolm (Eq. (10-34)) [90], Ito (Eq. (1035)) [91], Noyori (Eq. (10-36)) [92], and Mukaiyama (Eq. (10-37)) [93]. The Corey method has been utilized in the synthesis of a variety of important target molecules: prostaglandin synthesis intermediate 63 [94], and (trichloromethyl)-carbinol 64 [95], which is used in the preparation of unnatural amino acids in the nonracemic form, are illustrative [96]. The (EBTHI)Ti-H-catalyzed method of Buchwald presents a useful complement to other available catalytic enantioselective methods for enantioselective reduction of ketones. For example, the hydrosilation procedure reported by Ito (Eq. (10-35)) affords the allylic alcohol with similar levels of enantioselectivity as reported by the MIT group (entry 3, Table 10-14); however, a subsequent desilylation is required for the Ito method. On the other hand the cyclohexyl(methyl)ketone shown in entry 2 of Table 10-14 is reduced in 80% ee by the Ito technology (vs. 24% ee by the Ti-catalyzed protocol). The most recent advance reported by Noyori affords outstanding enantioselectivities in a reaction where the hydride source (formic acid) efficiently reduces the ketone substrate. This is in contrast to the use of the more conventional 2-propanol as the source of hydride, which suffers from the reverse process and thus less than ideal conversions. Most recently, structural studies on the active Ru complexes involved in the Noyori protocol [97], as well as recent advances illustrating that the Ru-catalyzed reaction is effective in the kinetic resolution of a range of secondary alcohols [98] add to the impressive utility of this transformation. H Ph N
O
O OH
B
2.5 mol%
Me
Ph
H
Me
1.2 equiv BH3, 25 °C 92% yield, 95% ee
Corey, 1987
(10-32)
O O
Me
OH
Ph O O
OH
82% ee
63
CCl3 Me
95% ee
64
10.4 Ph
Catalytic Reactions Effected by Ansa-Metallocenes
Ph Ph Me
N O
O
; excess
Sm
O
OH OH
Me
I
5 mol%
Me
677
Me
(gg)
95% yield, 97% ee
Evans, 1993 NH
Ph Ph
S
Ph
O
OH
O
OH
10 mol%
Me
(hh)
Me
NaBH4, TMSCl; 25 °C (BH3•Me2S; 25 °C)
Bolm, 1993
Me
92% yield, 84% ee (76% ee)
PBu2 H
Fe O
Fe
OH
H Bu2P
Me
• [Rh(cod)2]BF4 1 mol% Me
Me
(ii)
Ph2SiH 2, -40 °C; K2CO3, MeOH 92% yield, 91% ee
Ito, 1994 Ts Ph
(10-33 to 10-36)
N Ru
O
Ph
OH
Cl
N
0.5 mol%
Ts
(jj)
HCO2H, Et3N, 28 °C >99% yield, 99% ee
Noyori, 1996
N
N Co
O O
O
OH O
O
5 mol%
(kk)
NaBH4, EtOH, -20 °C Mukaiyama, 1996
>99% yield, 73% ee
678
10
Chiral Titanocenes and Zirconocenes in Synthesis
10.5 Summary and Outlook The chemistry described in this article demonstrates that chiral ansa-metallocene complexes can be used to effect a number of useful and important catalytic asymmetric transformations in excellent yield and with high levels of enantioselectivity. Because of the direct involvement of the metal center in the bond forming process, reactions that involve the insertion of a functional group (i.e., an alkene, imine, enamine, or ketone) into a metal–hydride or metal–carbon bond most often proceed with outstanding regio- and stereoselection. However, when bond formation takes place at a site more remote from the transition metal center (e.g., allylmetallation or Lewis acid-catalyzed Diels–Alder reaction), less efficient chirality transfer is observed. Future studies may nonetheless show that, in the latter class of transformations, alteration of the chiral ligand can allow for a higher degree of chirality transfer in metallocene-mediated or catalyzed reactions. Although EBTHI metallocenes have proven to be effective at promoting a variety of transformations, the equipment required to prepare such catalysts (glovebox and high pressure hydrogenation apparatus), as well as costs associated with the required metallocene resolution (nonracemic binaphthol=$45 g–1), may preclude everyday use of these chiral controlling agents in large scale organic synthesis. In this context, promising advances toward practical syntheses of inexpensive and chiral (EBTHI)MX2 equivalents may eventually provide more practical alternatives to this powerful class of transition metal catalysts. These organometallic complexes should not only allow for widespread utilization of the enantioselective and catalytic transformations discussed in this review, they may also be able to effect selective transformations not readily achieved with nonracemic (EBTHI)M ensembles. Regardless, there is little doubt that the chemistry discussed in this article adumbrates future exciting discoveries in the area of design and development of useful asymmetric catalytic transformations. The National Institutes of Health (GM-47480) and the National Science Foundation (CHE-9257580 and CHE-9632278) provide generous support of our programs. Additional support from Pfizer, Johnson & Johnson, Eli Lilly, Zeneca, Glaxo, Monsanto and Sloan and Dreyfus Foundations are gratefully acknowledged. We are grateful to our co-workers and colleagues Zhongmin Xu, Ahmad F. Houri, Mary T. Didiuk, Michael S. Visser, Joseph P. A. Harrity, Charles W. Johannes, Nicola M. Heron, Jeffrey A. Adams, Daniel C. La, Nina R. Horan, Derek A. Cogan and John D. Gleason for their important contributions to various aspects of the carbomagnesation project. Without these talented and dedicated individuals none of this chemistry would have been possible. We thank Professors Robert Waymouth and Richard Jordan for many useful suggestions and a critical reading of this manuscript.
References
679
References [1] [2]
[3]
[4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
[16] [17]
[18] [19] [20] [21] [22] [23]
[24] [25]
[26]
For a general overview of recent advances in this area, see: Catalytic Asymmetric Synthesis, I. Ojima, Ed.; VCH Publishers: New York, Weinheim, Cambridge 1993. For a recent review on chiral metallocene complexes, see: a) R. L. Halterman, Chem. Rev. 1992, 92, 969. b) J. Okuda, Angew. Chem. 1992, 104, 49; Angew. Chem., Int. Ed. Engl. 1992, 31, 47. a) F. R. W. P. Wild, L. Zsolnai, G. Huttner, H. H. Brintzinger, J. Organomet. Chem. 1982, 232, 233. b) F. W. R. P. Wild, M. Wasiucionek, G. Huttner, H. H. Brintzinger, J. Organomet. Chem. 1985, 288, 63. c) W. Kaminsky, K. Kulper, H. H. Brintzinger, F. W. R. P. Wild, Angew. Chem. 1985, 97, 507; Angew. Chem., Int. Ed. Engl. 1985, 24, 507. J. A. Ewen, L. Haspeslagh, J. L. Atwood, H. Zhang, J. Am. Chem. Soc. 1987, 109, 6544. S. Collins, B. A. Kuntz, N. J. Taylor, D. G. Ward, J. Organomet. Chem. 1988, 342, 21. F. Piemontesi, I. Camurati, L. Resconi, D. Balboni, A. Sironi, M. Moret, R. Zeigler, N. Piccolrovazzi, Organometallics 1995, 14, 1256. R. B. Grossman, R. A. Doyle, S. L. Buchwald, Organometallics 1991, 10, 1501. Q. Yang, M. D. Jensen, SynLett 1996, 147. For preparation of solvated group IV metal tetrachlorides from the anhydrous metal salts, see: L. E. Manzer, Inorg. Synth. 1982, 135. For studies on the stereochemistry of hydrogenation of (EBI)ZrCl2 (5), see: R. M. Waymouth, F. Bangerter, P. Pino, Inorg. Chem. 1988, 27, 758. G. M. Diamond, S. Rodewald, R. F. Jordan, Organometallics 1995, 5. A. Schafer, E. Karl, L. Zsolnai, G. Huttner, H. H. Brintzinger, J. Organomet. Chem. 1987, 328, 87. B. Chin, S. L. Buchwald, J. Org. Chem. 1996, 61, 5650. R. B. Grossman, W. M. Davis, S. L. Buchwald, J. Am. Chem. Soc. 1991, 113, 2321. A similar equilibration has recently been proposed for the zirconium-catalyzed isomerization of cis and trans alkenes that do not bear b-hydrides, see: E. I. Negishi, D. Choueiry, T. B. Nguyen, D. R. Swanson; D. R. Suzuki, T. Takahashi, J. Am. Chem. Soc. 1994, 116, 9751. S. Collins, B. A. Kuntz, Y. Hong, J. Org. Chem. 1989, 54, 4154. A similar transition structure has been proposed to explain the stereochemistry of addition of achiral allyl titanocenes to aldehydes. See: F. Sato, K. Iida, S. Iijima, H. Moriya, M. Sato, J. Chem. Soc., Chem. Commun. 1981, 1140. H. C. Brown, P. K. Jadhav, J. Am. Chem. Soc. 1983, 105, 2092. N. Minowa, T. Mukaiyama, Bull. Chem. Soc. Japan 1987, 60, 3697. M. Riediker, R. O. Duthaler, Angew. Chem. 1989, 101, 488; Angew. Chem., Int. Ed. Engl. 1989, 28, 494. E. J. Corey, C-M. Yu, S. S. Kim, J. Am. Chem. Soc. 1989, 111, 5495. A. Hafner, R. O. Duthaler, R. Marti, G. Rihs, P. Rothe-Streit, F. Scwarzenbach, J. Am. Chem. Soc. 1992, 114, 2321. a) W. R. Roush, L. K. Hoong, M. A. J. Palmer, J. C. Park, J. Org. Chem. 1990, 55, 4109. b) W. R. Roush, P. T. Grover, X. Lin, Tetrahedron Lett. 1990, 31, 7563. c) W. R. Roush, P. T. Grover, Tetrahedron Lett. 1990, 31, 7567. d) W. R. Roush, J. C. Park, Tetrahedron Lett. 1991, 32, 6285. R. Sturmer, R. W. Hoffmann, SynLett. 1990, 759. For similar enantioselective addition of allenic and acetylenic units, respectively, see: a) E. J. Corey, G. B. Jones, Tetrahedron Lett. 1991, 32, 5713. b) E. J. Corey, K. A. Cimprich, J. Am. Chem. Soc. 1994, 116, 3151. c) G. E. Keck, D. Krishnamurthy, X. Chen, Tetrahedron Lett. 1994, 35, 8323. K. Ishihara, M. Mouri, Q. Gao, T. Maruyama, K. Furuta, H. Yamamoto, J. Am. Chem. Soc. 1993, 115, 11490. For a more recent related example, see: D. R. Gauthier, E. M. Carreira, Angew. Chem. 1996, 108, 2521; Angew. Chem., Int. Ed. Engl. 1996, 35, 2363.
680 [27]
[28] [29] [30] [31]
[32] [33] [34] [35]
[36]
[37] [38]
[39] [40] [41] [42] [43] [44]
[45] [46]
10
Chiral Titanocenes and Zirconocenes in Synthesis
a) G. H. Keck, K. H. Tarbet, L. S. Geraci, J. Am. Chem. Soc. 1993, 115, 8467. b) A. L. Costa, M. G. Piazza, E. Tagliavini, C. Trombini, A. Umani-Ronchi, J. Am. Chem. Soc. 1993, 115, 7001. c) P. Bedeschi, S. Casolari, A. L. Costa, E. Tagliavini, A. Umani-Ronchi, Tetrahedron Lett. 1995, 36, 7897. d) P. G. Cozzi, P. Orioli, E. Tagliavini, A. Umani-Ronchi, Tetrahedron Lett. 1997, 38, 145. S. Rodewald, R. F. Jordan, J. Am. Chem. Soc. 1994, 116, 4491. a) A. S. Guram, R. F. Jordan, Organometallics 1990, 9, 2190. b) A. S. Guram, R. F. Jordan, Organometallics, 10, 3470. J. A. Ewen, J. Am. Chem. Soc. 1984, 106, 6355. For a brief introduction to the oligomerization process, see: a) J. P. Collman, L. S. Hegedus, J. R. Norton, R. G. Finke, in Principles and Applications of Organotransition Metal Chemistry; University Science Books: Mill Valley, CA 1987, 593. b) For a recent review, see: H. H. Brintzinger, D. Fischer, R. Mulhaupt, B. Reiger, R. M. Waymouth, Angew. Chem. 1995, 107, 1255; Angew. Chem., Int. Ed. Engl. 1995, 34, 1143. R. M. Waymouth, P. Pino, J. Am. Chem. Soc. 1990, 112, 4911. W. Kaminsky, A. Ahlers, N. Moller-Lindenhof, Angew. Chem. 1989, 101, 1304; Angew. Chem., Int. Ed. Engl. 1989, 28, 1216. P. Pino, P. Cioni, J. Wei, J. Am. Chem. Soc. 1987, 109, 6189. a) G. W. Coates, R. M. Waymouth, J. Am. Chem. Soc. 1991, 113, 6270. b) G. W. Coates, R. M. Waymouth, J. Am. Chem. Soc. 1993, 115, 91. For an additional example of an optically active synthetic polymer with main-chain chirality, see: M. Brookhart, M. J. Wagner, G. G. A. Balavoine, H. A. Haddou, J. Am. Chem. Soc. 1994, 116, 3641. a) G. Natta, L. Porri, S. Valenti, Makromol. Chem. 1963, 67, 225. b) T. Tsunetsugu, T. Fueno, J. Furukawa, Makromol. Chem. 1968, 112, 220. c) M. Farina, G. Audisio, G. Natta, J. Am. Chem. Soc. 1967, 89, 5071. d) G. Costa, P. Locatelli, A. Zambelli, Makromolecules 1973, 6, 653. G. Natta, M. Farina, M. Peraldo, G. Bressan, Makromol. Chem. 1961, 43, 68. For stereoselective polymerization of racemic 4-methyl-1-hexene through heterogeneous metal catalysis, see: P. Pino; Mulhaupt, R. Angew. Chem. 1980, 92, 921; Angew. Chem., Int. Ed. Engl. 1980, 19, 857, and references cited therein. Y. Hong, B. A. Kuntz, S. Collins, Organometallics 1993, 12, 964. (EBTHI)ZrMe2 is prepared by treatment of (EBTHI)Zr-binol with two equivalents of MeLi, see Ref. [12]. With Cp2Zr(OtBu)(THF) as catalyst selectivities are generally ~20:1 endo:exo. See: S. Collins, B. E. Koene, R. Ramachandran, N. J. Taylor, Organometallics 1991, 10, 2092. J. M. Hawkins, S. Loren, J. Am. Chem. Soc. 1991, 113, 7794, and references cited therein. J. B. Jaquith, J. Guan, S. Wang, S. Collins, Organometallics 1995, 14, 1079. For a review of catalytic asymmetric Diels-Alder reactions, see: H. B. Kagan, O. Riant, Chem. 4+2] cycloadditions that have appeared subsequent to this review, see: a) Q. Gao, T. Maruyama, M. Mouri, H. Yamamoto, J. Org. Chem. 1992, 57, 1951. b) K. Hattori, H. Yamamoto, J. Org. Chem. 1992, 57, 3264. c) E. J. Corey, T-P. Loh, Tetrahedron Lett. 1993, 34, 3979. d) J. Bao, W. D. Wulff, A. L. Rheingold, J. Am. Chem. Soc. 1993, 115, 3814. e) I. E. Marko, G. R. Evans, Tetrahedron Lett. 1993, 35, 2771. f) J. Bao, W. D. Wulff, Tetrahedron Lett. 1994, 36, 3321. g) K. Ishihara, H. Yamamoto, J. Am. Chem. Soc. 1994, 116, 1561. h) K. Mikami, Y. Motoyama, M. Terada, J. Am. Chem. Soc. 1994, 116, 2812. i) K. Ishihara, M. Miyata, M., K. Hattori, T. Tada, H. Yamamoto, J. Am. Chem. Soc. 1994 116, 10520. j) E. J. Corey, A. Guzman-Perez, T-P. Loh, J. Am. Chem. Soc. 1994, 116, 3611. k) W. Odenkirk, B. Bosnich, J. Chem. Soc., Chem. Commun. 1995, 1181. l) K. Fuji, T. Kawabata, A. Kuroda, T. Taga, J. Org. Chem. 1995, 60, 1914. K. Narasaka, N. Iwasawa, M. Inoue, T. Yamada, M. Nakashima, J. Sugimori, J. Am. Chem. Soc. 1989, 111, 5340, and references cited therein. a) E. J. Corey, R. Imwinkelried, S. Pikul, Y-B. Xiang, J. Am. Chem. Soc. 1989, 111, 5493. b) E. J. Corey, N. Imai, H-Y. Zhang, J. Am. Chem. Soc. 1991, 113, 728; this paper reports cy-
References
681
cloaddition reactions catalyzed by bis(oxazoline) class of chiral ligands (also utilized by Evans in Ref. [47]). O
O N
• FeCl2I
N
O
O
Ph N
O
10 mol%
Ph
, -50 °C
O O
N
99:1 endo:exo 85% yield, 86% ee
O
Corey, 1991
[47]
[48]
[49]
[50]
[51] [52] [53] [54] [55]
[56] [57] [58] [59] [60]
[61]
[62]
[63]
a) D. A. Evans, S. J. Miller, T. Lectka, J. Am. Chem. Soc. 1993, 115, 6460. b) D. A. Evans, J. A. Murry, P. Von Matt, R. D. Norcros, S. J. Miller, Angew. Chem. 1993, 107, 864; Angew. Chem., Int. Ed. Engl. 1995, 34, 798. S. Kobayashi, H. Ishitani, J. Am. Chem. Soc. 1994, 116, 4083, and references cited therein. For a recent review, where Sc-catalyzed enantioselective cycloadditions are discussed, see: S. Kobayashi, SynLett. 1994, 689. For reports that address the reactivity issue (reactions with less reactive dienes) in the asymmetric Diels-Alder reactions, see: a) E. J. Corey, S. Sarshar, D-H. Lee, J. Am. Chem. Soc. 1994, 116, 12089. b) Ref. [45b]. J. P. Morken, M. T. Didiuk, A. H. Hoveyda, J. Am. Chem. Soc. 1993, 115, 6997. See also: HG. Schmalz, Nachr. Chem. Tech. Lab. 1994, 42, 724. For closely related subsequent studies, see: (a) B. Louise, R. J. Whitby, R. V. H. Jones, M. C. H. Standen, Tetrahedron Lett. 1996, 37, 7139. (b) G. Dawson, C. A. Durrant, G. G. Kirk, R. J. Whitby, R. V. H. Jones, M. C. H. Standen, Tetrahedron Lett. 1997, 38, 2335. A. F. Houri, Z-M. Xu, D. A. Cogan, A. H. Hoveyda, J. Am. Chem. Soc. 1995, 117, 2943. See also: H-G. Schmalz, Angew. Chem. 1995, 107, 1981; Angew. Chem., Int. Ed. Engl. 1995, 34, 1833. A. H. Hoveyda, J. P. Morken, J. Org. Chem. 1993, 58, 4237. M. T. Didiuk, C. W. Johannes, J. P. Morken, A. H. Hoveyda, J. Am. Chem. Soc. 1995, 117, 7097. (a) A. S. Guram, R. F. Jordan, Organometallics 1990, 9, 2190. (b) ibid. 1991, 10, 3470. (a) G. C. Fu, R. H. Grubbs, J. Am. Chem. Soc. 1992, 114, 7324. (b) G. C. Fu, R. H. Grubbs J. Am. Chem. Soc. 1993, 115, 3800. (c) R. H. Grubbs, S. J. Miller, G. C. Fu, Acc. Chem. Res. 1995, 28, 446 and references cited therein. (d) H-G. Schmalz, Angew. Chem. 1995, 107, 1981; Angew. Chem., Int. Ed. Engl. 1995, 34, 1833 and references cited therein. (e) R. R. Schrock, J. S. Murdzek, G. C. Bazan, J. Robbins, M. DiMare, M. O’Regan, J. Am. Chem. Soc. 1990, 112, 3875. (f) G. C. Bazan, R. R. Schrock, H.-N. Cho, V. C. Gibson, Macromolecules 1991, 24, 4495. M. S. Visser, N. M. Heron, M. T. Didiuk, J. F. Sagal, A. H. Hoveyda, J. Am. Chem. Soc. 1996, 118, 4291. J. P. Morken, M. T. Didiuk, M. S. Visser, A. H. Hoveyda, J. Am. Chem. Soc. 1994, 116, 3123. N. M. Heron, J. A. Adams, A. H. Hoveyda, 1997, submitted for publication. M. S. Visser, A. H. Hoveyda, Tetrahedron 1995, 4383. (a) V. S. Martin, S. S. Woodard, T. Katsuki, Y. Yamada, M. Ikeda, K. B. Sharpless, J. Am. Chem. Soc. 1981, 103, 6237. (b) Y. Gao, R. M. Hanson, J. M. Klunder, S. Y. Ko, H. Masamune, K. B. Sharpless, J. Am. Chem. Soc. 1987, 109, 5765. For a review of kinetic resolution, see: (a) H. B. Kagan, J. C. Fiaud, Top. Stereochem. 1988, 18, 249. (b) M. G. Finn, K. B. Sharpless, In Asymmetric Synthesis; Morrison, J. D. Ed.; Academic Press: New York, 1985; p. 247-308. M. Kitamura, I. Kasahara, K. Manabe, R. Noyori, H. Takaya, J. Org. Chem. 1988, 53, 708. For Pd-catalyzed enantioselective synthesis of cyclic allylic esters, see: B. M. Trost, M. G. Organ, J. Am. Chem. Soc. 1994, 116, 10320. J. P. A. Harrity, M. S. Visser, J. D. Gleason, A. H. Hoveyda, J. Am. Chem. Soc 1997, 119, 1488.
682 [64] [65]
[66] [67] [68] [69]
[70]
[71] [72] [73]
[74]
[75] [76] [77]
[78]
[79] [80]
[81] [82]
10
Chiral Titanocenes and Zirconocenes in Synthesis
G. Erker, M. Aulbach, M. Knickmeier, D. Wingbermuhle, C. Kruger, M. Nolte, S. Werner, J. Am. Chem. Soc. 1993, 115, 4590. a) D. Y. Kondakov, E. Negishi, J. Am. Chem. Soc. 1995, 117, 10771. b) D. Y. Kondakov, E. Negishi, J. Am. Chem. Soc. 1996, 118, 1577. See also: c) D. Y. Kondakov, S. Wang, E. Negishi, Tetrahedron Lett. 1996, 37, 3803. d) E. Negishi, D. Y. Kondakov, Chem. Soc. Rev. 1996, 417. F. A. Hicks, S. L. Buchwald, J. Am. Chem. Soc. 1996, 118, 11688. For studies related to the characterization of Ziegler-Natta polymerization intermediates with Cp2ZrCl2 as catalyst, see: P. Gassman, M. R. Callstrom, J. Am. Chem. Soc. 1987, 109, 7875. R. D. Broene, S. L. Buchwald, J. Am. Chem. Soc. 1993, 115, 12569. With a 64:36 mixture of E:Z olefin isomers and 100% conversion, a value of 83% ee can be calculated for the hydrogenation product of the Z alkene: 64%(95% ee R) + 36%(83% ee S) = 100%(31% ee R). a) A. Pfaltz, Acc. Chem. Res. 1993, 26, 339. b) R. Noyori, Science 1990, 248, 1194. c) H. Takaya, T. Ohta, R. Noyori, in Catalytic Asymmetric Synthesis; I. Ojima, Ed.; VCH: New York, 1993; p. 1-39 and references cited therein. d) I. Ojima, N. Clos, C. Bastos, Tetrahedron 1989, 45, 6901, and references cited therein. M. Tanaka, I. Ogata, J. Chem. Soc., Chem. Commun. 1975, 735. U. Leutenegger, A. Madin, A. Pfaltz, Angew. Chem. 1989, 101, 61; Angew. Chem., Int. Ed. Engl. 1989, 28, 60. For other examples of metal-catalyzed asymmetric hydrogenation of imines, see: a) Rhodium catalysis: J. Bakos, I. Toth, B. Heil, G. Szalontai, L. Parkanyi, V. Fulop, J. Organomet. Chem. 1989, 370, 263 b) Iridium catalysis: Y. N. C. Chan, J. A. Osborn, J. Am. Chem. Soc. 1990, 112, 9400. a) C. A. Willoughby, S. L. Buchwald, J. Am. Chem. Soc. 1992, 114, 7562. b) C. A. Willoughby, S. L. Buchwald, J. Am. Chem. Soc. 1994, 116, 8952. c) C. A. Willoughby, S. L. Buchwald, J. Am. Chem. Soc. 1994, 116, 11703. C. A. Willoughby, S. L. Buchwald, J. Org. Chem. 1993, 58, 7627. A. Viso, N. E. Lee, S. L. Buchwald, J. Am. Chem. Soc. 1994, 116, 9373. Alkene insertion reactions generally involve a transition state where the alkene and metal-hydride bond are coplanar. See: R. H. Crabtree in The Organometallic Chemistry of the Transition Metals; Wiley: New York 1988, p. 148-153. Buchwald has shown that in the (EBTHI)Ti-H-catalyzed reduction of achiral imines, the turnover-limiting step is hydrogenolysis of the Ti-N bond (Scheme 10-29). Therefore, insertion of an imine into the Ti-H bond is not necessarily a kinetically significant process. However, the proposed mechanistic arguments, where the major product stereoisomer is predicted based on the minimization of steric interactions within the (EBTHI)Ti-H-p complex, provide a convenient predictive tool that allow for the reliable prediction of stereochemical outcome of the reaction. It is possible that, in fact, enamine or alkene association with the chiral Ti-H is the stereochemistry-determining step. In other words, although the Ti-N hydrogenation may be turnover-limiting, there is little energy difference between the slow steps of the various pathways that lead to enantiomeric products: the predominant product isomer arises from the most abundant transition metal-substrate complex. R. Becker, H. Brunner, S. Mahboobi, W. Wiegrebe, Angew. Chem. 1985, 97, 993; Angew. Che.m., Int. Ed. Engl. 1985, 24, 995. G-J. Kang, W. R. Cullen, M. D. Fryzuk, B. R. James, J. P. Kutney, J. Chem. Soc., Chem. Commun. 1988, 1466. See also: a) Ref. [73a]. b) J. Bakos, A. Orosz, B. Heil, M. Laghmari, P. Lhoste, D. Sinou, J. Chem. Soc., Chem. Commun. 1991, 1684. c) A. G. Becalski, W. R. Cullen, M. D. Fryzuk, B. R. James, G-J. Kang, S. J. Rettig, Inorg. Chem. 1991, 30, 5002. Y. N. C. Chan, J. A. Osborn, J. Am. Chem. Soc. 1990, 112, 9400. F. Spindler, B. Pugin, H-U. Blaser, Angew. Chem. 1990, 102, 561; Angew. Chem., Int. Ed. Engl. 1990, 29, 558.
References [83]
683
M. J. Burk, J. E. Feaster, J. Am. Chem. Soc. 1992, 114, 6266. The structure of Et-duphos is shown below. For details of the synthesis of this class of ligands, see: M. J. Burk, J. E. Feaster, W. A. Nugent, R. L. Harlow, J. Am. Chem. Soc. 1993, 115, 10125. Et
P Et Et
P Et
[84] [85] [86]
[87] [88] [89] [90] [91] [92] [93] [94] [95] [96]
[97] [98]
C. Lensink, J. G. de Vries, Tetrahedron Asymmetry 1993, 4, 215. N. E. Lee, S. L. Buchwald, J. Am. Chem. Soc. 1994, 116, 5985. For recent review of enantioselective hydrogenation of enamides, see: a) R. Noyori, Asymmetric Catalysis in Organic Synthesis J. Wiley Interscience: New York, 1994; p. 33-38. b) Refs. [60b] (p. 9-10) and [60c]. M. B. Carter, B. Schiott, A. Gutierrez, S. L. Buchwald, J. Am. Chem. Soc. 1994, 116, 11667. E. J. Corey, R. K. Bakshi, S. Shibata, J. Am. Chem. Soc. 1987, 109, 5551. D. A. Evans, S. G. Nelson, M. R. Gagne, A. R. Muci, J. Am. Chem. Soc. 1993, 115, 9800. a) C. Bolm, A. Seger, M. Felder, Tetrahedron Lett. 1993, 34, 8079. b) C. Bolm, M. Felder, Tetrahedron Lett. 1993, 34, 6041. M. Sawamura, R. Kuwano, Y. Ito, J. Am. Chem. Soc. 1994, 33, 111. A. Fujii, S. Hashiguchi, N. Uematsu, T. Ikariya, R. Noyori, J. Am. Chem. Soc. 1996, 118, 2521. T. Nagata, K. Yorozu, T. Yamada, T. Mukaiyama Angew. Chem. 1995, 107, 2309; Angew. Chem., Int. Ed. Engl. 1996, 34, 2145. E. J. Corey, R. K. Bakshi, S. Shibata, C-P. Chen, V. K. Singh, J. Am. Chem. Soc. 1987, 109, 7925. E. J. Corey, J. O. Link, J. Am. Chem. Soc. 1992, 114, 1906. For other less enantioselective catalytic ketones, see: a) K. Yamamoto, T. Hayashi, M. Kumada, J. Organomet. Chem. 1972, 46, C65. b) K. Tamao, H. Yamamoto, H. Matsumoto, N. Miyake, T. Hayashi, M. Kumada, Tetrahedron Lett. 1977, 1389. c) R. L. Halterman, T. M. Ramsay, Z. Chen, J. Org. Chem. 1994, 59, 2642. K-J. Haack, S. Hashiguchi, A. Fujii, T. Ikariya, R. Noyori, Angew. Chem. 1997, 109, 297; Angew. Chem., Int. Ed. Engl. 1997, 36, 285. S. Hashiguchi, A. Fujii, K-J. Haack, K. Matsumura, T. Ikariya, R. Noyori, Angew. Chem. 1997, 109, 300; Angew. Chem., Int. Ed. Engl. 1997, 36, 288.
684
10
Chiral Titanocenes and Zirconocenes in Synthesis
11 New Chiral Ferrocenyl Ligands for Asymmetric Catalysis Antonio Togni
11.1 Introduction Among the many different types of chiral ligands used in asymmetric homogeneous catalysis [1], ferrocenyl ligands represent one of the most important classes of auxiliaries. Indeed, they find application in a variety of transition-metal catalyzed reactions, from, e.g., Rh-catalyzed hydrogenations, to Ni- and Pd-catalyzed carbon–carbon bond-forming reactions, to the unique Au(I)-catalyzed aldol-type formation of oxazolines from aldehydes and isocyanoacetates, and several others. This field has been pioneered by Hayashi and Kumada who first prepared planar chiral ferrocenyl phosphines of types 1 and 2, shown below [2].
PPh2 NMe2 Fe
Me PPh2
(S)-(R)-1
Me PPh2 N Fe
Me
OH OH OH
PPh2 (S)-(R)-2
These specific ligands, containing either one phosphino group attached to the position adjacent to the stereogenic side-chain, or, in addition, a second one in position 1′ of the ferrocene core have been recently reviewed by Hayashi [3], and will not be dealt with in the present contribution. Let us merely note that the particularity of, e.g., ligand 2 is the functionalized side-chain attached to the pseudo benzylic stereogenic center. This substituent is the most prominent feature, because it may be readily varied, in order to fulfil the purpose of a secondary interaction with the substrate in the course of the catalytic reaction [4]. This is one of the most important aspects explaining the success of these ligands in asymmetric catalysis. It is furthermore to be noted that virtually all known derivatives of type 1 and 2 almost exclusively contain the diphenylphosphino group(s). We shall therefore concentrate on developments occurred mainly during the Nineties. From a conceptual and structural point of view, ferrocenyl ligands of the new generation are represented by way of example by the specific compounds 3–6.
686
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
Me PPh2 PCy2
PPh2 N N
Me
Fe
Me
Me
Fe
(S)-(R)-3 (Josiphos)
(S)-(R)-4
Ph2P Me Me
PPh2
O
Fe
N
Fe PPh2
Fe
Ph
(S)-(R)-6
(R,R)-(S,S)-5
From a synthetic point of view all compounds of type 1–5 have in common the chiral, enantiopure ferrocenyl amine 7 as a starting material. As illustrated in Scheme 11-1, ortho lithiation of 5 occurs diastereoselectively [5], leading, after reaction with a suited electrophile, to a planar chiral derivative [6], that can usually be isolated as single stereoisomer after recrystallization. The second important synthetic step in this chemistry is the SN1-type substitution reaction of the dimethylamino group, affected with a variety of heteroatom nucleophiles in polar solvents. It is important to note that this nucleophilic substitution occurs with retention of configuration, by virtue of the participation of the ferrocene moiety in stabilizing the intermediate carbocationic species [7]. On the other hand, ligands of type 6 are prepared from achiral ferrocene carboxylic acid derivatives.
NMe2 Me
Fe
NMe2
Fe
Me
1) BuLi 2)
(S)-7
E
E NuFe
E+ (S)-(R)-8 (E=e.g., PPh2 )
Scheme 1. The fundamental reactions involving Ugi’s amine 7 used for the synthesis of ferrocenyl ligands.
Nu Me
11.2
Synthetic and Structural Aspects of New-Generation Ferrocenyl Ligands
687
The bidentate, asymmetric ligands of type 3 and 4 are characterized by the presence of two ligand atoms (P,P or P,N) imbedded in different and easily modifiable steric and electronic environments. The C2-symmetric biferrocene 6 and its derivative containing alkyl phosphine substituents have the peculiarity to span a trans coordination geometry, whereas compound 6 leans upon the successful use of chiral oxazolines introduced by Pfaltz [8].
11.2 Synthetic and Structural Aspects of New-Generation Ferrocenyl Ligands 11.2.1 Synthesis of Ligands Involving Diastereoselective Lithiation of Chiral Ferrocenes The preparation of ligands of type 3 and 4 follows the general strategy of Figure 11-1. A substituted ferrocene bearing a stereogenic group (typically the amine 7) is used as the common starting material. The two ligand fragments L1 and L2 are introduced consecutively, whereby L1 is usually introduced first, by lithiation followed by the reaction with a suitable electrophile. On the other hand, L2 is connected to the ligand system by nucleophilic displacement of a heteroatomic leaving group at the stereogenic center (typically dimethylamino). Thereby is to note that the reaction of, e.g., the intermediate amine 8 with a) a secondary phosphine, and b) a pyrazole occurs cleanly in acetic acid at temperatures between 70° and 100°, typically. Only the use of this solvent allows to circumvent the conversion of amine 8 to the corresponding 1-ferrocenylethyl acetate which is known to undergo substitution reactions. Ligand fragment 1 (achiral electrophilic synthon) Step 1/2
L2
L1
Step 2/1
Fe
Me
Ligand fragment 2 (achiral nucleophilic synthon)
Common ferrocene fragment (carrier of the chiral information)
Figure 1. General strategy for the synthesis of new ferrocenyl ligands bearing two different ligand fragments L1 and L2.
688
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
Thus, a series of diphosphines have been prepared in high yields, as shown in Scheme 11-2. These compounds are usually crystalline, air-stable materials [9]. The preparation of Phobyphos is worth noting, since this derivative is an example of a ligand containing the rare fragment [3.3.1]-9-phosphabicyclonon-9-yl. The reaction of an excess of ‘Phobane’ with the amine 8 leads to a kinetic separation of the two ‘Phobane’ isomers [10] otherwise not directly available in pure form. Ligands containing a pyrazole instead of a phosphino group attached to the stereogenic center, are prepared in a much similar way, as compared to their diphosphine congeners [11]. Pyrazoles have been chosen because of the opportunity to easily modulate their steric and electronic properties, by variation of the substituents, in particular at the positions 3 and 5, adjacent to the nitrogen atoms (see Scheme 11-3). The reaction of the pyrazoles with ferrocenylamine 8 shows a clean regioselectivity, in case of different substituents R1 and R3. The sterically more bulky and/or more electron-withdrawing substituent always takes the position of R1, i.e. remote from the ferrocene. An interpretation of this observation, as well as the discussion of few exceptions to this rule, have been reported previously [11]. Pyrazoles are prepared by reacting hydrazine with either a 1,3-diketone or, in the case of 3(5)monosubstituted derivatives, a ketoenamine [12], as illustrated by the particular example of Scheme 11-4 [13]. PR2
NMe2 1) BuLi
Me
Me
Fe
PR2
NMe2
PR'2
HPR'2 Fe
Fe
Me
AcOH
2) ClPR2 (S)-(R)-8
(S)-7
PPh2
The prototype:
Fe
High yields of crystalline products PCy2
Many combinations of PR2 and PR' 2
Me
(S)-(R)-Josiphos
PPh2
NMe2 Me
PH 3:1
PH
PPh2
10-fold xs of "Phobane" mixture
Fe
P Me
Fe AcOH, 80°C
(S)-(R)-8
"Phobyphos", 68%
Scheme 2. Preparation of new ferrocenyl diphosphines (“normal” route).
11.2
689
Synthetic and Structural Aspects of New-Generation Ferrocenyl Ligands
Scheme 3. Synthesis of pyrazole-containing ferrocenyl ligands.
O O
CH(OMe)2NMe2 Ru
Ru
O
NMe2
H2NNH2
N
N H
Ru
N
NH
O NMe2 (R)-(S)-8 Me
Fe N
N
PPh2
Ru
NN
Fe
PPh2 Me
(R,R)-(S,S)-9
Scheme 4. Synthesis of a bis(ferrocenyphosphine/pyrazole) ligand: the ketoenamine route.
690
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
The biferrocenyl ligands of type 5 are prepared by a slightly different route [14], as illustrated in Scheme 11-5. The amine 7 is first converted to the iodo derivative 10, in which the dimethylamino group is then substituted by the lithium salt of a secondary phosphine oxide, leading to 11. Ullmann coupling and reduction afford the desired compounds 5, the so-called TRAP-ligands. NMe2
1) sec-BuLi 2) I2
Fe
(R)-7
Me
I
Fe
(R)-(S)-10
I
Me
(R)-(S)-11
R2P Me Me
P(O)R2
1) MeI 2) LiP(O)R2
1) Cu 2) HSiCl3 / NEt3
Fe Fe
PR2
(R,R)-(S,S)-5
Scheme 5. Preparation of TRAP-ligands.
The key feature of this synthesis is the diastereoselective halogenation of amine 7, constituting a further activation of the cyclopentadienyl position adjacent to the side chain. A similar strategy has recently allowed the preparation of Josiphos-type and pyrazole containing ligands for which the sequence of introduction of the two ligating fragment has been reversed [15]. As depicted in Scheme 11-6, the bromo derivative 12 is afforded by lithiation followed by bromination with C2F4Br2 [16]. Substitution of the dimethylamino group by a secondary phosphine or a pyrazole in acetic acid gives the corresponding monodentate ligands 13 and 14, respectively. The phosphonite 15 and the bis(N-indolyl)phosphine 16 are obtained after treatment with BuLi and a suitable chlorophosphine. The advantage of this route is the opportunity to introduce in the last step ligand units that would be otherwise incompatible with the reaction conditions (acetic acid solvent) of the substitution reaction at the stereogenic center. This is the case for the two phosphorus moieties in 15 and 16. Because of their enhanced π-accepting properties with respect to regular phosphines [17], these two fragments offer the opportunity to extend the study of electronic effects on enantioselectivity in a number of reactions (vide infra).
11.2
Br
Br
PPh2
NMe 2
HPPh2
Br
N N
Fe
Me
HN N
Me
Fe
691
Synthetic and Structural Aspects of New-Generation Ferrocenyl Ligands
Me
Fe AcOH
AcOH
(S)-(R)-14
(S)-(R)-12
(S)-(R)-13
1) BuLi
1) BuLi 2) (RO)2PCl
1) BuLi
2) C2 F4Br2
2) (N-indolyl) 2PCl
NMe2 N O
Fe
O P
PPh2
Fe
Me
(S)-7
Me
N
P
Fe
N N Me
(S)-(R)-16 (S)-(R)-15
Scheme 6. Synthesis of phosphonite and bis(N-indolyl)phosphine derivatives by the “inverse” sequence.
Ligands of type 6 have been reported recently by different research groups. The chiral oxazolinyl ferrocene is obtained by the addition of a 1,2-aminoalcohol system to either cyanoferrocene [18], or ferrocene carboxylic acid [19, 20] and its esters [21], as illustrated in Scheme 11-7. Upon lithiation, the stereogenic substituent is an effective diastereoselective ortho-directing group. It has been shown that the nitrogen atom of the oxazoline, by virtue of its coordination to lithium, fulfils the directing role [22, 23]. By analogous procedures, the C2-symmetric 1,1′-bis(oxazolinyl)derivatives 17 may also be obtained [24, 25].
11.2.2 Coordination-Chemical and Conformational Characteristics Several transition-metal complexes containing chelating diphosphines of the Josiphos-type (3) have been characterized by X-ray crystallography [26, 27]. The generally observed P–M–P bite-angle of the ligands is typically in the range between 93° and 97°, although values slightly below 90° have already been detected [28]. An important feature of these compounds is their ability to maintain a well-defined conformation, no matter which metal they are coordinated to and despite the vary-
692
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
O CN
NH
Fe
Fe
Fe NH2
R OH
NH2 OH R Cat. ZnCl2
COOH
1) (COCl)2
OH
R
PPh3 / CCl4 / NEt3
O
PPh2
O
Fe
N
1) BuLi Fe
N
R
2) PPh 2Cl
(S)-(R)-6 (R=Me, i-Pr, Ph, CH2Ph, t-Bu, s-Bu)
O N Fe PPh2 PPh2 N
R
R 17 (R=i-Pr, t-Bu) R
O
Scheme 7. Synthesis of oxazoline-containing ferrocenyl ligands.
ing nature of the ancillary ligands. In the case of Josiphos, the schematic superposition of several structures depicted in Figure 11-2, illustrates this behavior. Thus, one recognizes the constant orientation of the substituents on the P-atoms. The diphenylphosphino fragments displays an endo–axial and an exo–equatorial orientation of the phenyl groups (endo and exo refer to the position relative to the ferrocene core), whereas the PCy2 fragment always adopts an axial orientation, thus minimizing steric interactions with the ferrocene. These conformational characteristics lead also to a maximization of the metal–ferrocene distance.
11.2
Synthetic and Structural Aspects of New-Generation Ferrocenyl Ligands
693
MLn P2 P1
Fe
Figure 2. Superposition of the X-ray structures of Josiphos (3, dashed lines) and four of its complexes (M = Ru, Rh, Pd, and Pt; ancillary ligands omitted for clarity).
The pyrazole-containing ligands of type 4 also are structurally well behaved and show a typical conformation in the solid state in which the heterocycle and one phosphine phenyl group occupy pseudo axial positions with respect to the ferrocene moiety. Figure 11-3 shows a superposition of ten different derivatives. From it one recognizes that the overall conformation of the common backbone is essentially not influenced by substituents on the pyrazole, the phenyl groups, or the ‘lower’ Cp ring. This is an important structural aspect, since it allows to use the peripheral substituents 1) for controlling the electronic properties of the ligand, and 2) for introducing specific steric interactions within the corresponding metal complexes, as it will be discussed below. The pyrazole ligands coordinate in a chelating fashion to transition metals, maintaining the general conformational features observed in their free state. This is particularly the case for Rh(I) complexes. Despite the relatively large seven-membered chelate ring, the bite angle P–Rh–N is in the range 85°–87°. In the case of Pd(II)–π-allyl complexes, however, the corresponding bite angle increases to 92°– 95°. A consequence of this is the less pronounced differentiation between axially and equatorially oriented phenyl groups on phosphorus. Examples of structures of Rh(I) and Pd(II) complexes are given in Figures 11-4 and 11-5, respectively.
694
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
R
R
ax
R
ax N
N
P R
R Me
Fe
eq
Rn
Figure 3. Superposition of the X-ray structures of ten different pyrazole-containing ferrocenyl ligands.
R
R
R R
N2 P
Rh
ax
ax
N1
N N P
Rh Me
R
Fe
eq
R
R
Fe
R
Figure 4. Superposition of the X-ray structures of four different Rh(I) complexes containing pyrazole ferrocenyl ligands. Ancillary ligands are omitted for clarity.
11.2
Synthetic and Structural Aspects of New-Generation Ferrocenyl Ligands
695
R
ax N
N2
Pd
N
R
Pd N1
P
Me Fe
P
Fe
Figure 5. Superposition of the X-ray structures of four different Pd(II)-π−allyl complexes containing pyrazole ferrocenyl ligands. Allyl fragments are omitted for clarity.
It is interesting to note that the main coordination plane of Rh in the complexes of Figure 11-3 forms a tilting angle of 51°–56° with the ‘upper’ Cp ring and is endooriented with respect to the ferrocene, i.e., the metal center is located slightly below the Cp ring. On the other hand, for the Pd derivatives of Figure 11-5 the corresponding angle is 34°–37°, and the orientation is rather to be defined as exo with the Pd center located above the ‘upper’ Cp ring), a situation similar to that observed for Josiphos complexes (vide supra). Finally, a common feature to all ligands discussed so far is the ‘lifting’ of the PPh2 phosphorus atom out of coplanarity with the Cp ring in the complexes that have been structurally characterized. This distortion is typically in the range of 0.1 Å to 0.4 Å. TRAP-ligands (5) have been shown by Ito and Sawamura to span a trans coordination geometry [29]. The X-ray crystal structure of the specific complexes 18 and 19 schematically depicted below has been determined. It is worth noting that the combination of C2-symmetry and trans coordination geometry is conceptually new in the field and bears promises for future development.
696
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
Br Pd
Ph2P
PPh2
Br Me
O
O
Me
Fe
C
O
Rh
P O Me
Cl Me
Fe
Fe
18
O
P
Fe
19
11.3 Applications of New Ferrocenyl Ligands in Asymmetric Catalysis 11.3.1 Industrial Applications in Hydrogenation Reactions Two ferrocenyl diphosphine of the Josiphos-type have already found applications in production scale hydrogenation processes. Chronologically the first one relates to a new industrial synthesis of the vitamin biotin, developed at Lonza Ltd [30]. The catalytic process and the specific ligand, however, have been developed first at CIBA, in a nice example of industrial collaboration. As illustrated in Scheme 11-8, the Josiphos analog containing a bulky (t-butyl)2P group is used in the Rh-catalyzed, highly diastereoselective hydrogenation of a fully substituted olefin. This reaction introduces the two new stereogenic centers of the bicyclic system of biotin, and constitutes a rare example of the successful reduction of this kind of substrates, although other ferrocenyl ligands [14, 31, 32], as well as DuPhos ligands [33] have successfully been used in catalytic hydrogenations of tetrasubstituted C=C double bonds. A second, and probably more impressive application is shown in Scheme 11-9. The Ir-catalyzed ketimine hydrogenation depicted constitutes the so far largest asymmetric catalytic process generating more than 10,000 tons of the herbicide (S)-metolachlor per year [34]. Although the ees obtained do not go beyond ca 80%, the efficiency of the catalyst plays a much more decisive role. With substrate to catalyst ratios of up to one million and full conversions within hours, this is also one of the fastest catalysts known. The development of this process also indicates the eminent role of empiricism in asymmetric catalysis. Indeed, the Ir-systems needs co-catalytic amounts of iodide and acid, in order to attain full efficiency. However, the exact function of these components is not fully understood [35].
11.3
Applications of New Ferrocenyl Ligands in Asymmetric Catalysis O
O R
N
697
0.2 mol % Rh Ligand*
NH
R
N
NH
H
(+)-Biotin
H
H2 O
O
O
PPh2
P(t-Bu)2
O
90 % ee (R = CH2Ph) 99 % ds (R = CH(Me)Ph)
Me Fe
Ligand* =
Scheme 8. Industrial scale Rh-catalyzed hydrogenation in a new synthesis of (+)-biotin.
O O
N
O HN
Cl
O N
Ir-Cat. H2
80 % ee
PPh2
Catalyst system: Ir(I) / Iodide / H 2SO4
P 2 Me
Herbicide S-Metolachlor
Substrate/Ir up to 1'000'000 !
Fe
Scheme 9. Ir-catalyzed imine hydrogenation in the production of the herbicide (S)-Metolachlor.
698
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
11.3.2 Rhodium-Catalyzed Hydroboration of Olefins Electronic Effects on Enantioselectivity The diphosphine Josiphos has been found to provide relatively high ees in the hydroboration of styrene with catecholborane [9]. However, enantioselectivities slightly exceeding 90% could only be obtained when the reaction was conducted at low temperature. Much better results are given by the pyrazole-containing ligands. Indeed, this system affords the highest known enantioselectivity in this particular reaction at room temperature [36]. An important drawback, however, is the observed relatively low regioselectivity of the reaction, usually generating up to 40% of the undesired primary achiral product. Ligand electronic effects on enantioselectivity are very important in this reaction. Indeed, it has been found that the full range of ees from almost 0% to 98.5% is covered. The ligands used differ from one another essentially in their electronic properties, by virtue of different peripheral substituent, while maintaining practically identical steric characteristics. It is important to note that no ligand was found so far able to switch enantioselectivity based upon an alteration of its electronics only, i.e., the sense of asymmetric induction is still determined by the absolute configuration of the ferrocenyl ligand backbone. Representative examples are depicted in Scheme 11-10. In order to obtain optimum enantioselectivities, it is CF3
F3 C MeO Me Me P
Me
P
Fe
Fe
Me
F3 C
NN
NN
MeO
Me Me P N N
Me
Me
MeO Fe
Me
5% ee
P
...and the best
Fe
F3 C
CF3
P
NN Me
33% ee
Me
Me 2 N
F3 C F3 C
Fe
98.5% ee
98% ee
MeO F3C CF3 P NN
Me Me P N N
F3 C
Fe
Me
95% ee
90% ee
F3 C F3 C
F 3C
CF3 N N Me
Fe
40% ee
Me P
Me2 N Fe
Me N N Me
72% ee
the worst... Scheme 10. Influence of ligand peripheral substituents on the enantioselectivity of the Rh-catalyzed hydroboration of styrene with catecholborane.
11.3
Applications of New Ferrocenyl Ligands in Asymmetric Catalysis
699
clear that the pyrazole nitrogen atom coordinating to the metal has to be a good σdonor, and significant π-accepting properties at phosphorus are also required. This is easily obtained using electron-donating and -withdrawing substituent, respectively. The inverse combination is very much deleterious and leads to very low selectivities. A more extended range of substituents on the pyrazole seems not to be applicable, since stereoselectivity is very sensitive to the size of the group at position 3 of the heterocycle. In fact, already on going beyond the bulk of an ethyl or isopropyl group the ees drop significantly [36]. The origin of such large electronic effects are not fully understood. In view of the poor ligand quality of the nitrogen atom in those systems possessing trifluoromethyl groups on the pyrazole (ligands leading to low ees), one must ask the question whether or not a possible dissociation of the pyrazole from the first coordination sphere of Rh could be responsible for the bad stereochemical outcome. In other words, if such a dissociation were to take place, the monodentate behavior of the ferrocenyl ligand would possibly explain the poor inductions observed. This question was addressed by examining the relative thermodynamic stability of the model cationic complexes [Rh(P,N-ligand)(COD)]+ [36]. It was found that the more electron rich ligands are able to displace the electron poor ones from the Rh center in the course of fast equilibrium reactions. Thus, a qualitative correlation exists between complex thermodynamic stability and the enantioselectivity afforded by the corresponding ligand. However, no species containing any of the P,Nligands coordinated in a monodentate fashion could be identified. Strongly coordinating monodentate ligands such as PEt3 or PPh3 were found to liberate the P,Nligand when added to the Rh complex, even in understoichiometric amounts. On the other hand, excess styrene did not undergo any observable reaction with the model complexes. We therefore conclude that, if any dissociation of the electronpoor ligands is relevant to catalysis, then a highly active species not containing the chiral ligand must be formed, however, in quantities not detectable by common spectroscopic techniques. Although improbable, this event cannot be excluded. Assuming that the observed electronic effects are indeed due to an alteration of the electronic properties of the catalytically active species containing the P,Nligand, we prepared another series of neutral Rh(I) model complexes containing a carbonyl ligand in trans position to the pyrazole moiety, from the reaction of [Rh2(CO)4Cl2] and the P,N-system [37]. The CO ligand may thus be used as diagnostic tool for examining the electronic environment induced at the metal center by the chiral ligand, since the CO-stretching frequency is sensitive to the donor/acceptor properties of the latter. Of course the observed ν(CO) reflects the combined influence of both the pyrazole and the phosphine, thus care must be exercised while interpreting the results and only trend indications may be derived. A representative selection of the data so far collected is given in Table11-1.
700
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
Table 1. IR-data for a selection of [RhCl(CO)(P,N)] complexes and the enantioselectivity obtained in the hydroboration of styrene utilizing the corresponding ligand. R1
R3
Ar
CH3
CH3
Ph
1989
95
CF3
CH3
Ph
1994
44
CF3
CF3
Ph
2000
33
CH3
CH3
p-MeO-Ph
1987
87
CH3
CH3
p-Me-Ph
1991
94
CH3
CH3
4-(CF3)-Ph
CH3
CH3 3,5-(CF3)2-Ph 2008
ν(CO) %ee
(cm-1)
1
R3
R N
Cl
Ar
N
Rh P C
2006
98 98.5
O
Fe
Ar
From the data of Table 11-1 it clearly results that the alteration of the donor capacity of the pyrazole nitrogen, while maintaining constant the nature of the phosphine, has a significant effect on ν(CO). Thus, upon decreasing the electron density at nitrogen and hence its donor ability, an increase of the CO bond order is observed, indicating a less important π -backbonding contribution from Rh, as one would expect. A similar trend, albeit not surprisingly slightly more pronounced, is observed when gradually enhancing the π-acceptor qualities of the phosphine in 3,5-dimethyl pyrazole derivatives. The best ligand for the Rh-catalyzed hydroboration of styrene that have been identified by the systematic study described above, turns out to afford best enantioselectivities also in the corresponding reaction of 2,5-dihydrofuran. As shown in Scheme 11-11, 3-hydroxytetrahydrofuran is obtained after oxidative workup in
Scheme 11. The highly enantioselective hydroboration of 2,5-dihydrofuran.
11.3
Applications of New Ferrocenyl Ligands in Asymmetric Catalysis
701
98.5% ee [38]. This kind of substrate also avoids the still unsolved problem of the poor regioselectivity observed for unsymmetrically substituted olefins, using P,N ferrocenyl ligands. In conclusion, and although the strong electronic effects observed in the hydroboration reaction with these ligands are poorly understood on a mechanistic level, it is clear that one is dealing with subtle changes of the electronic properties of the metal center induced by the substituents attached to the periphery of the chiral ligands. Our observations indicate more generally that the empirical tuning of the ligands in asymmetric catalysis should always include the possibility of influencing stereoselectivity by variations of the donor properties of the inducing agent.
11.3.3
Palladium-Catalyzed Allylic Amination Using Pyrazole Containing Ligands
Steric Effects and Mechanistic Studies Whereas asymmetric Pd-catalyzed allylic substitution reactions involving soft carbon nucleophiles have been the subject of extensive studies in recent years [39], the corresponding transformation utilizing amine derivatives (C–N bond forming reaction) has received much less attention [40–46]. We recently found that pyrazole-containing ligands afford very high enantioselectivities in the amination of 1,3-diphenylallyl acetate (or the corresponding ethyl carbonate) with benzylamine, as illustrated in Scheme 11-12 [47]. Thus, the allylamine product is obtained in 94–97% optical purity utilizing a variety of ligands, containing a fairly large substituent at position 3 of the pyrazole (the methyl group in position 5 is much less relevant). Up to the bulk of a 1- or 2naphthyl group the enantioselectivity is not influenced to any very significant extent. The R enantiomer of the allylamine is formed preferentially when ferrocenyl ligands possessing the (S)–(R) absolute configuration are used. However, the introduction of a 9-anthryl group leads to an unexpected reversal of enantioselectivity. The S form is now produced with an ee of 40%. In order to understand this drastic and, as we will see, purely steric effect on the stereoselectivity, we performed a 2D NMR study of the intermediate Pd–π-allyl complexes formed during the catalytic reaction [47, 48]. Thus, it was found that the anthryl-containing complex exists as mixture of two diastereoisomeric compounds, differing in the absolute configuration at the allyl terminus trans to phosphorus, as shown in Scheme 11-13. The configuration of these two stereoisomers, occurring in relative amounts of 29% and 71%, may be defined as exo–syn–syn and exo–syn–anti, respectively [47]. Assuming that nucleophilic attack takes place at the allylic carbon trans to P exclusively, then the ratio of these two diastereoisomers reflects the enantiomer ratio of the
702
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
3 mol % Pd(dba) 2 / Ligand
OCOR
NHCH2Ph
40 °C / THF Ph
Ph
Ph
rac; R = Me, OEt
H 3C PPh 2
Ph 85-95 % yield
NH 2
R H 3C
N N
PPh 2
N N
Me Me
Fe Fe
R = e.g., Ph, 1-Np, 2-Np, 4-Py, Cy, 3-NO2-Ph 94 - 97% ee R-enantiomer
40.0% ee S-enantiomer
Scheme 12. Asymmetric Pd-catalyzed allylic amination with different pyrazole-containing ferrocenyl ligands.
product within 1–2%. This would indicate that the reaction of benzylamine with these two isomers takes place with equal rates.
+
S
Ph
N N Ph Pd Ph P Ph
exo-syn-syn (29%)
+ N N Ph Pd Ph P
R Fe
Ph
Fe
Ph
exo-syn-anti (71%)
Scheme 13. Configuration of Pd-π−allyl complexes containing a 9-anthryl substituted pyrazole, as found in solution by 2D NMR methods.
11.3
Applications of New Ferrocenyl Ligands in Asymmetric Catalysis
703
The configuration inversion leading to the major exo–syn–anti isomer can only be explained in terms of steric repulsion between the 9-anthryl fragment and the phenyl attached to the allyl carbon trans to P. Configurational analysis of several other Pd–π-allyl complexes in solution utilizing 2D NMR methods revealed the exo– syn–syn form to be present in at least 80% for those ligands affording high ees [48]. In these cases, however, no clear relationship between isomers and product enantiomer ratio is apparent. A 3-(1-adamantyl) substituted pyrazole ligand was found to give the highest ees for this particular reaction (> 99%). For the corresponding π-allyl complex only the exo–syn–syn form could be identified, indicating that the nucleophilic attack indeed takes place at the position trans to phosphorus exclusively (see Scheme 11-14). The hexafluorophosphate salts of several cationic Pd–π-allyl complexes were characterized by X-ray crystallography. The configuration found in the solid state corresponds to that of the major isomer in solution. For all complexes analyzed [47, 48] a deviation of the π-allyl geometry from its expected orientation was observed. The allyl ligand turns out to be ‘rotated’ in such a way that two carbon atoms are almost coplanar with the Pd, P, and N atoms, whereas the position of the allyl terminus trans to phosphorus is significantly out of plane (average distance 0.53 Å for four different complexes). This distortion appears to be dictated by the steric influence of both the diphenylphosphino group and the substituent in position 3 of the pyrazole fragment. The effect of the latter seems to be more subtle (and probably more important) than the one due to the phosphine. Indeed, as described above, the size of this substituent is very crucial, as it may induce a configurational change. On the other hand, we think that the major role of the PPh2 group mainly consists in stabilizing the exo form, by virtue of an optimum ‘interphenyl’ steric interaction. Single exo-syn-syn isomer
Allyl "rotation" (X-ray studies)
+
N
N N Pd
Ph
a)
P
Ph P
Fe
b)
Nu
Increased electrophilicity (ab-initio studies)
Nu
a) b)
R-product
R
S-product
S
> 200 (> 99 % ee)
∆∆G≠ > 3 kcal.mol-1
Scheme 14. Configuration and distortion of the Pd-π-allyl complex containing the 1-adamantyl substituted ferrocenyl ligand.
704
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
Searching for the ultimate reason leading to the pronounced site-selectivity of the nucleophilic attack, we performed first-principles calculations on a model, nondistorted Pd(P,N) system [49, 50]. It clearly turns out that the nucleophilic attack trans to phosphorus is preferred. However, the theoretical difference in activation energy between the two possible pathways—trans to P and trans to N—is not sufficiently high to account for the observed enantioselectivities. The calculated 8 kJ mol–1 difference in ∆∆G≠ would lead to an ee of less than 95%, whereas the best experimental values go beyond 99%. We therefore included the experimentally observed distortion (‘allyl rotation’) into our calculations and found that the preference for the nucleophilic attack at the center out of plane is very pronounced, whether or not this center is trans to phosphorus. In other words, according to theory, the main factor governing site-selectivity is indeed the distortion of the allyl ligand, with a drastically enhanced electrophilicity for the carbon atom out of plane, as schematically shown in Figure 11-6. Nucleophilic attack is favored by P trans influence ∆∆G‡t/c = 8 kJ/mol
+ N
P Pd
N
P
+
+
N
P
Pd
Pd Enhanced electrophilicity for C out of plane
∆∆G‡t/c = 38 kJ/mol
∆∆G‡c/t = 29 kJ/mol
Figure 6. Schematic representation of the effect of the sterically induced allyl rotation on allyl LUMO and hence on the preferred site of nucleophilic attack, as resulting from first-principles calculations.
Theory thus predicts that it should be possible for the nucleophile to attack at the position trans to nitrogen in such systems, provided that the necessary distortion is present. However, no ligand could be found so far for which this could be verified. The experimentally observed allyl rotation of ca 15°–20° around the Pd–allyl axis takes the two carbon atoms forming the olefinic double bond in the final product into an orientation much similar to that they are likely to assume in the Pd(0) complex generated by the nucleophilic attack, for which one would expect a planar geometry. Thus, allyl distortion anticipates an important structural feature of the product.
11.3
705
Applications of New Ferrocenyl Ligands in Asymmetric Catalysis
As discussed above, the reversal of enantioselectivity observed with the 9-anthryl substituted ligand is due to a sterically induced inversion of configuration at the allyl center trans to phosphorus. However, the reversal of selectivity is not complete, since a significant amount of the exo–syn–syn diastereoisomer is still present. With the aim of completely suppressing the formation of this configurational isomer, in favor of the exo–syn–anti form, we prepared a ligand containing a triptycyl substituent at position 3 of the pyrazole [47]. In fact, it turns out that the corresponding Pd–(1,3-diphenylallyl) complex displays the expected exo–syn–anti configuration, both in solution and in the solid state (see Figure 11-7).
+
N Pd Ph
N Ph P
Fe
Figure 7. Molecular structure and schematic representation of a triptycyl substituted cationic Pd-π-allyl complex, showing the enforced exo-syn-anti configuration of the allyl.
With this information in hand, we were expecting the triptycyl ligand to induce the formation of the opposite enantiomer of the allylamine in high ees (S enantiomer with ligand absolute configuration (S)–(R)). However, the use of catalysts containing this ligand did not afford any product at all, even when the reaction was conducted at reflux temperature in THF for prolonged periods of time. An explanation of this, at first sight disappointing result, is given in the cartoon of Scheme 11-15.
706
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
P Ph
Pd
N
P Ph
N Pd Nu
Nu
Ph
Pd(II)-(η3-Allyl)
Ph
Pd(0)-(η2-Olefin)
Scheme 15. Intramolecular steric constraints connected with the nucleophilic attack onto the Pd-allyl complex containing the triptycyl substituted pyrazole ligand.
Upon nucleophilic attack at the carbon atom trans to phosphorus, and in order to form a trigonal planar Pd(0) olefin complex, the allyl fragment will have to perform a side-on movement, towards the bulky triptycyl fragment. However, the steric fitting of the anti- configured half of the allyl ligand between two blades of the propeller-shaped triptycyl is almost ideal, such that the necessary sliding of the allyl becomes impossible. Furthermore, attack at the opposite allyl terminus is hampered because of the allyl rotation, enormously decreasing the electrophilicity at that particular center. The detailed structural analyzes of cationic Pd–π-allyl complexes described above have been performed utilizing the corresponding PF6– salts. Since it is generally accepted that such cationic species form during the catalytic reaction, one would expect that their use as catalyst precursors should give identical results, as compared to reactions carried out with the catalysts formed in situ from the common Pd(0) source [Pd2(dba)3.CHCl3] and the chiral ligand. However, it turns out that, when the well-characterized hexafluorophosphate salts of the intermediate Pd–allyl complexes were used, very low enantioselectivities were obtained. This surprising observation could be traced back to a counter-ion effect [40, 51–58]. In particular, we found that several anions added to the catalyst system, typically in form of tetraalkylammonium salts, were able to influence stereoselectivity. Our findings are summarized in Table 11-2, by the crucial experiments carried out utilizing the particular ligand 20 containing a 3-(1-adamantyl)pyrazole (similar trends are observed for other ligands). Extremely negative effects are exerted by relatively large, virtually noncoordinating anions such as PF6–, whereas a positive influence may be obtained when small, hard and potentially coordinating ions such as fluoride or BH4– are added, other anions displaying intermediate effects. A tentative and simple interpretation of the anion effect, taking into particular consideration the differences between fluoride and hexafluorophosphate, is given in
11.3
Applications of New Ferrocenyl Ligands in Asymmetric Catalysis
707
Scheme 11-16. Having verified that no kinetic resolution of the racemic substrate takes place, it must be assumed that upon oxidative addition of the allylic carbonate diastereoisomeric π-allyl complexes are formed. Table 2.Anion effects on enantioselectivity in the Pd-catalyzed allylic amination using the adamantyl substituted P,N-ferrocenyl ligand 20. PPh2 OCO2Et Ph
Ph
NHCH2Ph
3 mol % Pd 40 °C / THF / PhCH2NH2
Catalyst precursor
Ph
Ph
N N Me
Fe
20
salt added
S/R (%ee)
(equiv./Pd)
[Pd2 (dba)3 ] + 20
–
221 (99.1)
[Pd(η3 -PhCHCHCHPh)(20)]PF6
–
1.14 (6.7)
[Pd2 (dba)3 ] + 20
[NBu4]F (2)
>400 (>99.5)
[Pd(η3 -PhCHCHCHPh)(20)]PF6
[NBu4]F (2)
153 (98.7)
[Pd2 (dba)3 ] + 20
[NBu4][BH4] (1)
>400 (>99.5)
[Pd(η3 -PhCHCHCHPh)(20)]PF6
[NBu4][BH4] (1)
>400 (>99.5)
[Pd2 (dba)3 ] + 20
[NBu4]Cl (2)
221 (99.1)
[Pd2 (dba)3 ] + 20
[NBu4]I (2)
124 (98.4)
[Pd2 (dba)3 ] + 20
[NBu4]Br (2)
73 (97.3)
[Pd2 (dba)3 ] + 20
[NBu4][BF4] (2)
19 (90)
[Pd2 (dba)3 ] + 20
[NBu4](OTf) (2)
15.7(88)
[Pd2 (dba)3 ] + 20
[NBu4][ClO4] (2)
15.7 (88)
[Pd2 (dba)3 ] + 20
[NBu4][PF6] (2)
5.25 (68)
[Pd2 (dba)3 ] + 20
[NBu4][PF6] (10)
2 (33)
708
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis OCOOEt
Ph
Ph
Similar rates of oxidative addition (no kinetic resolution)
S Ph
N Pd Ph exo-syn-syn
+
OCOOEt Ph
R Ph
Slow with PF6(tight ion pairs ?)
N Pd
Ph
P
P -
Fast with F (Coordination?)
+
Ph endo-syn-syn
more stable NHCH2Ph
NHCH 2Ph Ph
R Ph
Ph
S Ph
Scheme 16. A tentative explanation of the anion effect in Pd-catalyzed allylic amination.
As illustrated, these could be represented by the observed exo–syn–syn form and, e.g., the corresponding endo–syn–syn isomer, although the presence of other configurational isomer cannot be excluded. Upon reaction with the amine at the allylic carbon trans to phosphorus, the different diastereoisomers would lead to opposite product enantiomers, unless fast interconversion is operating. Exactly in determining this equilibration, we see the crucial role of the anion. Hexafluorophosphate does not accelerate equilibration, as it is confirmed by most of the NMR studies performed on PF6– salts [47, 48]. Fluoride, however, probably by virtue of reversible coordination to Pd, is enhancing the dynamic behavior of the system, thus allowing the rapid interconversion of the diastereoisomers to give the most stable exo–syn–syn form. This means that fluoride ensures that the system is operating under Curtin–Hammett regime [59–61]. In summary, the studies described above have shown how subtle steric interactions between peripheral parts of the chiral ligand and the π-allyl fragment within the cationic Pd–allyl intermediate may induce 1) a drastic change of the electronic properties of the allyl, leading to a pronounced site-selectivity for nucleophilic attack, and 2) a change in configuration at one allyl carbon inducing a reversal of enantioselectivity of the catalytic reaction. Our system is thus a prominent example where the interplay between steric and electronic effects on stereoselectivity of a catalytic reaction can be unraveled.
11.3
Applications of New Ferrocenyl Ligands in Asymmetric Catalysis
709
11.3.4 Applications of TRAP Ligands The TRAP ligands 5 were first reported in 1991 [62], and have been shown in recent years to be successfully applicable to a variety of transition-metal-catalyzed asymmetric reactions. Thus, alkyl-TRAPs (i.e. such derivatives containing alkyl substituents on the two phosphorus atoms) afford high enantioselectivities in the hydrogenation of dimethyl itaconate [63], as well as in the hydrosilylation of simple ketones [64] and keto esters [65], as illustrated by the selected examples of Scheme 11-17. A further most interesting use of TRAP ligands is concerned with the asymmetric Michael addition of α-cyano carboxylates with vinyl ketones or acrolein catalyzed by Rh complexes [66, 67]. The highest enantioselectivity of 93% ee is obtained for the substrate combination shown in Scheme 11-18. Little is known about the nature of the catalyst which is formed in situ from the precursor [RhH(CO)(PPh3)3] and TRAP. The formation of an intermediate enolate coordinated to Rh via the cyano nitrogen atom has been postulated, based on an X-ray crystal structure of an analogous Ru model system [68]. The products of this Michael addition have been shown to be suitable starting materials for the preparation of αmethyl-α-amino acids [67]. Finally, a recent application of TRAP ligands involves [Rh(COD)2]BF4 Me
(R,R)-(S,S)-EtTRAP MeOOC
COOMe
96% ee
MeOOC
COOMe
1 atm H2 / CH2Cl2
OH
O
95% ee
1) [Rh(COD)2]BF4 (R,R)-(S,S)-nBuTRAP O
Ph 2SiH2 2) H +
O
OH
OEt
O OEt
93% ee
R 2P Me Me
Fe Fe
PR2
(R,R)-(S,S)-5 (R=Et, EtTRAP; R=nBu, nBuTRAP)
Scheme 17. Selected applications of TRAP ligands in Rh-catalyzed asymmetric hydrogenation and hydrosilylation reactions.
710
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
the Pd-catalyzed cycloisomerization of 1,6-enynes [69]. As shown in Scheme 1118, 3-methylene-pyrrolidines are formed in up to 95% ee, using a ligand derivative bearing electron-withdrawing phosphorus substituents. [RhH(CO)(PPh3)3] (S,S)-(R,R)-PhTRAP
O
O
+
Me
OCH(iPr)2 CN
O
O OCH(iPr)2
C6H6
Me
CN
93% ee SiMe3 SiMe3
4.5 mol% [Pd 2(dba)3]CHCl3
SiMe3
(S,S)-(R,R)-(4-CF3)PhTRAP
+ N N SO2Ph
SO2Ph
N
3.5 : 1
SO2Ph
95% ee
Scheme 18. TRAP ligands in the Rh-catalyzed asymmetric Michael addition of α-cyano carboxylates and in the Pd-catalyzed cyclization of 1,6-enynes.
11.4 Additional Ligands In addition to the ligands of type 3–6 described so far, we also prepared the triphosphine ‘Pigiphos’ (21) [70], and its congener ‘S-Pigiphos’ (22) [71], as well as a series of pentamethylated derivatives containing either iron (23) [72] or ruthenium (24) [73], as shown in Scheme 11-19. Pigiphos forms Ni(II) and Pd(II) dicationic monosolvento complexes displaying Lewis-acid catalytic activity in hetero Diels– Alder reactions, however affording poor stereoselectivities. Moderate enantioselectivities (up to 72% ee) have been obtained in the Ru-catalyzed asymmetric transfer-hydrogenation of simple ketones [71].
11.4
Fe
Me
PPh2
Me
P
Ph2P
Fe Fe
PPh2
21 ("Pigiphos")
PR2
PPh2
Me M
Me
S
Ph2P
Fe
22 ("S-Pigiphos") Me
PPh 2
Me
711
Additional Ligands
Me
Me 2N
Me
N N Me
Se Se
Me
Fe NMe2
M Fe
23
M = Fe, Ru
24
25
Scheme 19. Some new bi- and tridentate ferrocenyl ligands.
The Fe-Cp* derivatives 23 and 24 were not found to offer any particular advantages and behaved at best as their nonmethylated congeners in selected catalytic reactions. Furthermore, the corresponding ruthenocenyl ligands appear not to adopt the same conformational constraints as the Fe derivatives, and are in general much poorer chiral auxiliaries. Figure 11-8 shows the very different conformational features of Pd–allyl complexes containing Me5Josiphos and its Ru congeners, respectively. For the latter an endo–equatorial orientation of both phenyl groups at phosphorus becomes possible. The diferrocenyl diselenide 25 [74, 75] has been reported by Uemura as an effective ligand for the asymmetric Rh- and Ir-catalyzed hydrosilylation of simple ketones, affording moderate stereoselectivities [76]. The preparation of the tetradentate diferrocenyl ligands 26 and 27, possessing C2 symmetry, and not containing any stereogenic center (only planar chirality, see Scheme 11-20), has been recently described by Kagan, and Ito and Sawamura, respectively. The P2N2 system 26 easily forms an octahedral RuCl2 complex, upon reaction with [RuCl2(DMSO)4]. Reaction with CuOTf affords a product in which only one phosphine moiety is coordinated to copper [77, 78]. The N2O2 salen-type derivative 27 and its complexes, on the other hand, were found to be rather unstable, in particular against oxidation [79]. No catalytic applications have been reported.
712
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
Pd Pd
Ru
Fe
Common Conformation
Figure 8. The structure of cationic Pd-allyl complexes containing Me5Josiphos and its Ru congener, showing the uncommon conformation adopted by the Ru derivative.
Fe OH
Fe
HO N
PPh2
N
Fe
Ph2P
N
N
Fe 26
27
Scheme 20. Novel tetradentate diferrocenyl ligands based on planar-chirality only.
Weissensteiner, Schlögl and coworkers reported recently a series of homo- and heteroannularly bridged ligands, comprising P,N- (28 and 29) [80–83] and P,O-coordinating units (30) [84, 85]. A selection of these derivatives is shown in Scheme 1121. By virtue of the bridging aliphatic units, these compounds display more rigid conformations, as compared to similar open-chained derivatives. The P,N systems have been used in Pd-catalyzed Grignard cross-coupling reactions, affording moderate to good enantioselectivities. On the other hand, the amino alcohol 30 shown is an average chiral ligand for the asymmetric addition of diethylzinc to aldehydes.
11.4
Ph2P
Me2N
H
PPh2
NMe2
HO Ph Ph
Me2N
Fe
Fe
H
Fe
29
28
713
Additional Ligands
30
Scheme 21. Homo- and heteroannularly bridged ferrocenyl ligands.
Finally, the ferrocenylphosphine-imine derivative 31 has been recently described by Hayashi as an effective ligand for the Rh-catalyzed hydrosilylation of aromatic ketones with Ph2SiH2 [86]. Best enantioselectivities of ca 90% ee were obtained (see Scheme 11-22).
O
1) [Rh2Cl2(NBD)2]
OH
(1 mol%) / L* Ph2SiH2 2) H+
90% ee CF 3
PPh2 N
L* =
Fe Me
31
Scheme 22. A new ferrocenylphosphine-imine ligand in the Rh-catalyzed hydrosilylation of ketones.
714
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
11.5 New Methods for the Synthesis of Planar Chiral Ferrocenes Since amines of type 7 (or the corresponding alcohols as synthetic equivalents) still remain the most versatile starting materials for diastereoselective lithiation processes, and because the amino group can easily be substituted by nucleophiles (vide supra), methods have been developed for their enantioselective preparation. Recent work includes the aminoalcohol-catalyzed addition of dialkylzinc reagents to ferrocene-carbaldehyde [87, 88], and the 1,2-addition of organolithium compounds to the corresponding SAMP-hydrazone [89, 90], as illustrated in Scheme 11-23. Further methods involve the enantioselective reduction of ferrocenylketones using chiral oxazaborolidine borane [91, 92] and lipase-mediated desymmetrization of acetoxymethyl derivatives [93, 94], or ferrocenyl sulfides [95]. All these procedures make use of chiral auxiliaries (or catalysts) for the generation of the α-steR
R
H
H
1) Ac2O / NEt3 OH
NMe2
Fe
2) Me2NH
Fe
MeOH (up to 99% ee) R
5 mol% t-Bu ZnR2
H
N
NHAc
OH
Fe CHO (up to 94% ee)
Fe
1) H2/Raney-Ni 2) AcCl / NEt 3
SAMP Et2O
N
Fe
H
H
N
N
RLi OMe Et2O
Fe
N
R
(>98% de)
Scheme 23. New syntheses of chiral ferrocenyl amines.
OMe
715
11.5 New Methods for the Synthesis of Planar-Chiral Ferrocenes
reogenic center bearing a hetero-atom functionality. Among these methods those starting from ferrocenylketones are more straightforward and do not involve, e.g., cumbersome operations for the removal of chiral auxiliaries, or complex work-up and separation procedures. Therefore, they are anticipated to become even more significant in the near future. For the preparation of the pentamethyl derivatives of type 23 and 24 an alternative stereoselective approach has been reported (see Scheme 11-24).[72, 73] It involves the construction of the heteroleptic metallocenyl amine 34 starting from the chiral cyclopentadienyl salt 33 and a suitable Cp* precursor of iron and ruthenium, respectively. The enantiomerically pure 6-aminofulvene 32 was found to react diastereoselectively with MeLi affording 33 as a major product. However, the method gives high yields of the desired heteroleptic metallocene only with Cp* as the nonstereogenic cyclopentadienyl. The conversion of amine 34 to the ligands 23 and 24 involves then the same reaction conditions, as applied to other nonmethylated compounds of this kind. It is to be noted that the presence of the alkyl substituents leads to a significant acceleration of the substitution reactions at the stereogenic center, due to an increased stabilization of the carbocationic intermediates. Me N
Me + MeLi Cy
32 Me
Me
N Li+
X M
-
Me
N
Me Cy 33
Me
Me Cy M
34 M = Fe ; X = acac M = Ru; X = Cl
PPh 2
Recovery of the chiral auxiliary
Me 2NH, AcOH
NMe2
NMe2
1) BuLi Me M
2) ClPPh2
Me M
Scheme 24. A chiral 6-aminofulvene as starting material for the diasteroselective synthesis of pentamethylated ferrocenyl and ruthenocenyl ligands.
716
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
The strategy involving stereogenic ortho-directing groups for the preparation of planar chiral derivatives, so successfully applied for the ligands discussed so far, has been further developed in more recent years by Kagan and coworkers. Thus, the sulfoxide 35 [78, 96] and the acetal 36 [77, 97] have been found to afford high diastereoselectivities. These functional groups, however, are not necessarily those one may want to incorporate into the final planar chiral product, i.e., they may need removal or modification. t-Bu :
O S O
O
Fe
Fe
35
MeO
36
A completely different approach for the preparation of planar chiral ferrocenes, representing an important development in this area, was reported recently by Snieckus and coworkers [98]. Indeed it was shown that the enantioselective ortholithiation of ferrocenylamides is effectively achieved utilizing the alkaloid (–)sparteine as a chiral inducing agent (see the specific examples in Scheme 11-25). The novelty of the method is constituted by the fact that for the first time it is possible to generate planar chiral ferrocenes highly selectively and without the need of introducing a stereogenic directing group. The application of the reagent BuLi/ sparteine in ferrocene chemistry is a welcome extension of the concept of enanti-
H 1) O Fe
Ni-Pr2
N E
N H BuLi / Et2O / -78°C
Fe
O Ni-Pr2
2) E+ E=PPh2, 90%ee E=SiMe3, 98% ee E=Ph2C(OH), 99% ee
Scheme 25. Synthesis of planar-chiral ferrocene derivatives using sparteine-mediated enantioselective ortho-lithiation of ferrocenyl amides.
11.6
717
Planar-Chiral Ferrocenes as Catalysts
oselective deprotonation developed inter alia by Hoppe and coworkers in recent years [99]. Despite this impressive result, it remains to be seen whether or not the method may be applied so successfully also to ferrocenes bearing substituents other than amides. Furthermore, a major disadvantage of the method is represented the single enantiomeric form of (–)-sparteine [100, 101].
11.6 Planar Chiral Ferrocenes as Catalysts Finally, a completely new use of planar chiral ferrocenes has been recently disclosed by Fu and coworkers [102–104]. Compounds of type 37 and 38 were prepared as racemic mixtures and obtained as pure enantiomers via semi-preparative HPLC. Derivatives 37, analogs of 4-(dimethylamino)pyridine, were used as nucleophilic catalysts in the kinetic resolution of chiral secondary alcohols (see example in Scheme 11-26). The pentaphenyl substitution pattern in compound 37 appears to be very important, since the corresponding pentamethyl derivative is a poor catalyst [104]. The aminoalcohol system 38, on the other hand, is an effective chiral ligand for the asymmetric addition of dialkylzinc reagents to aldehydes (up to 90% ee) [103].
Me2N N Ph
Fe
Ph Ph
Ph
Me Me
Ph
Me
37
38
racemic
O OH
O O
O
Me
2 mol% 37
O
Me
+ Me
F
OH
Me
OH Me
Ph Ph
O N Fe Me
Me
NEt3 / Et2O 64% conversion
F
+
Me
F 99.2% ee
Scheme 26. New planar-chiral ferrocenes used as nucleophilic catalysts in the kinetic resolution of racemic alcohols, and as aminoalcohol ligands.
718
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
11.7 Conclusions The ferrocenyl ligands described above have been reported in the course of the last five to seven years, and constitute an important advance for the field of asymmetric catalysis. In this relatively short time the first two industrial applications of such ligands have been realized, clearly showing their potential. In particular, ferrocenyl ligands of the type 3 (some of which are now commercially available in research quantities) and 4 appear to offer several advantages, as compared to other classes of chiral auxiliaries for transition metal catalysis. As shown above, they are easily accessible from a common chiral precursor and the two different ligand moieties may be introduced in a modular manner using standard reaction conditions. By virtue of their well defined conformational properties, they may be modified by altering the nature of the peripheral substituents, without changing very significantly their steric properties. This gives the opportunity of exploiting subtle electronic effects affecting the stereoselectivity of the catalytic reaction. New synthetic methods for the preparation of planar chiral ferrocenes have been developed in very recent years, thus offering alternative protocols to this kind of derivatives. The research activities involving chiral ferrocenes all over the world show the increasing fundamental interest of the synthesis-oriented community, and are anticipated to generate interesting and relevant new structural combinations and reactivity patterns in the near future.
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15]
H. Brunner, W. Zettlmeier, Handbook of Enantioselective Catalysis, Vols. 1 and 2, VCH, Weinheim 1993. T. Hayashi, K. Yamamoto, K. Kumada, Tetrahedron Lett. 1974, 4405-4408. A. Togni, T. Hayashi (Eds.), Ferrocenes: Homogeneous Catalysis, Organic Synthesis, Materials Science, VCH, Weinheim 1995. M. Sawamura, Y. Ito, Chem. Rev. 1992, 92, 857-871. G. Gokel, D. Marquarding, I. Ugi, J. Org. Chem. 1972, 37, 3052-3058. K. Schlögl, Top. Stereochem. 1967, 1 , 39-92. D. Marquarding, H. Klusacek, G. Gokel, P. Hoffmann, I. Ugi, J. Am. Chem. Soc. 1970, 92 , 5389-5393. A. Pfaltz, Acta Chem. Scand. 1996, 50 , 189-194. A. Togni, C. Breutel, A. Schnyder, F. Spindler, H. Landert, A. Tijani, J. Am. Chem. Soc. 1994, 116 , 4062-4066. H. C. L. Abbenhuis, U. Burckhardt, V. Gramlich, C. Köllner, R. Salzmann, P. S. Pregosin, A. Togni, Organometallics 1995, 14 , 759-766. U. Burckhardt, L. Hintermann, A. Schnyder, A. Togni, Organometallics 1995, 14, 5415-5425. H. Brunner, T. Scheck, Chem. Ber. 1992, 125 , 701-709. U. Burckhardt, G. Trabesinger, A. Togni, V. Gramlich, Organometallics 1997, 16, in print. M. Sawamura, R. Kuwano, Y. Ito, J. Am. Chem. Soc. 1995, 117, 9602-9603. G. Pioda, A. Togni, unpublished results.
References [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56]
719
T.-Y. Dong, L.-L. Lai, J. Organomet. Chem. 1996, 509 , 131-134. K. G. Moloy, J. L. Petersen, J. Am. Chem. Soc. 1995, 117 , 7696-7710. Y. Nishibayashi, S. Uemura, Synlett 1995, 79-81. C. J. Richards, T. Damalidis, D. E. Hibbs, M. B. Hursthouse, Synlett 1995, 74-76. C. J. Richards, A. W. Mulvaney, Tetrahedron: Asymmetry 1996, 7, 1419-1430. K. H. Ahn, C.-W. Cho, H.-H. Baek, J. Park, S. Lee, J. Org. Chem. 1996, 61, 4937-4943. T. Sammakia, H. A. Latham, D. R. Schaad, J. Org. Chem. 1995, 60, 10-11. T. Sammakia, H. A. Latham, J. Org. Chem. 1996, 61, 1629-1635. W. Zhang, Y. Adachi, T. Hirao, I. Ikeda, Tetrahedron: Asymmetry 1996, 7, 451-460. W. Zhang, T. Hirao, I. Ikeda, Tetrahedron Lett. 1996, 37, 4545-4548. A. Togni, F. Spindler, G. Rihs, N. Zanetti, M. C. Soares, T. Gerfin, V. Gramlich, Inorg. Chim. Acta 1994, 222, 213-224. N. C. Zanetti, F. Spindler, J. Spencer, A. Togni, G. Rihs, Organometallics 1996, 15, 860-866. M. C. Soares, A. Togni, unpublished results. M. Sawamura, H. Hamashima, M. Sugawara, R. Kuwano, Y. Ito, Organometallics 1995, 14, 4549-4558. J. McGarrity, F. Spindler, R. Fuchs, M. Eyer, Eur. Pat. Appl.EP 624 587 A2, (LONZA AG), Chem. Abstr. 1995, 122, P81369q. T. Hayashi, N. Kawamura, Y. Ito, J. Am. Chem. Soc. 1987, 109, 7876-7878. T. Hayashi, N. Kawamura, Y. Ito,Tetrahedron Lett. 1988, 29, 5969-5972. M. J. Burk, M. F. Gross, J. P. Martinez, J. Am. Chem. Soc. 1995, 117, 9375-9376. S. Borman, Chem. & Eng. News 1996, July 12, pp. 38-40. F. Spindler, B. Pugin, H.-P. Jalett, H.-P. Buser, U. Pittelkow, H.-U. Blaser, in J. R. E. Malz (Ed.): Catalysis of Organic Reactions, Dekker, New York 1996, p. 153-166. A. Schnyder, L. Hintermann, A. Togni, Angew. Chem. 1995, 107, 996-998 (Angew. Chem. Int. Ed. Engl. 1995, 34, 931-933). A. Schnyder, A. Togni, U. Wiesli, Organometallics 1996, 15, 255-260. A. Schnyder, S. Chaloupka, O. Werbitzky, A. Togni, unpublished results. B. M. Trost, D. L. V. Vranken, Chem. Rev. 1996, 96, 395-422. T. Hayashi, A. Yamamoto, Y. Ito, E. Nishioka, H. Miura, K. Yanagi, J. Am. Chem. Soc. 1989, 111, 6301-6311. T. Hayashi, K. Kishi, A. Yamamoto, Y. Ito, Tetrahedron Lett. 1990, 31, 1743-1746. T. Hayashi, A. Yamamoto, Y. Ito, Tetrahedron Lett. 1988, 29, 99-102. P. v. Matt, O. Loizeleur, G. Koch, A. Pfaltz, C. Lefeber, T. Feucht, G. Helmchen, Tetrahedron: Asymmetry 1994, 5, 573-584. B. M. Trost, D. L. V. Vranken, C. Bingel, J. Am. Chem. Soc. 1992, 114, 9327-9343. B. M. Trost, D. L. V. Vranken, C. Bingel, Angew. Chem. 1992, 104, 194-196 (Angew. Chem. Int. Ed. Engl. 1992, 31, 228-230). B. M. Trost, D. L. V. Vranken, J. Am. Chem. Soc. 1993, 115, 444-458. A. Togni, U. Burckhardt, V. Gramlich, P. S. Pregosin, R. Salzmann, J. Am. Chem. Soc. 1996, 118, 1031-1037. U. Burckhardt, V. Gramlich, P. Hofmann, R. Nesper, P. S. Pregosin, R. Salzmann, A. Togni, Organometallics 1996, 15, 3496-3503. P. E. Blöchl, A. Togni, Organometallics 1996, 15, 4125-4132. T. R. Ward, Organometallics 1996, 15, 2836-2838. L. E. Overman, D. J. Poon, Angew. Chem. 1997, 109, 536-538 (Angew. Chem. Int. Ed. Engl. 1997, 36, 518-521). M. Bovens, A. Togni, L. M. Venanzi, J. Organomet. Chem. 1993, 451, C28-C31. B. M. Trost, M. G. Organ, G. A. O’Doherty, J. Am. Chem. Soc. 1995, 117, 9662-9670. B. M. Trost, R. C. Bunt, J. Am. Chem. Soc. 1996, 118, 235-236. S. Hansson, P.-O. Norrby, M. P. T. Sjögren, B. Åkermark, M. E. Cucciolito, F. Giordano, A. Vitagliano, Organometallics 1993, 12, 4940-4948. A. Gogoll, J. Örnebro, H. Grennberg, J. E. Bäckvall, J. Am. Chem. Soc. 1994, 116, 3631-3632.
720 [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81]
[82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92]
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
P. G. Andersson, A. Harden, D. Tanner, P.-O. Norrby, Chem. Eur. J. 1995, 1, 12-16. A. Loupy, B. Tchoubar, Salt Effects in Organic and Organometallic Chemistry, VCH, Weinheim 1992. P. R. Auburn, P. B. Mackenzie, B. Bosnich, J. Am. Chem. Soc. 1985, 107, 2033-2046. P. B. Mackenzie, J. Whelan, B. Bosnich, J. Am. Chem. Soc. 1985, 107, 2046-2054. J. I. Seeman, Chem. Rev. 1983, 83, 83-134. M. Sawamura, H. Hamashima, Y. Ito, Tetrahedron: Asymmetry 1991, 2, 593-596. R. Kuwano, M. Sawamura, Y. Ito, Tetrahedron: Asymmetry 1995, 6, 2521-2526. M. Sawamura, R. Kuwano, Y. Ito, Angew. Chem. Int. Ed. Engl. 1994, 33, 111-113. M. Sawamura, R. Kuwano, J. Shirai, Y. Ito, Synlett 1995, 347-348. M. Sawamura, H. Hamashima, Y. Ito, J. Am. Chem. Soc. 1992, 114, 8295-8296. M. Sawamura, H. Hamashima, Y. Ito, Tetrahedron 1994, 50, 4439-4454. Y. Mizuho, N. Kasuga, S. Komiya, Chem. Lett. 1991, 2127-2130. A. Goeke, M. Sawamura, R. Kuwano, Y. Ito, Angew. Chem. 1996, 108, 686-687 (Angew. Chem. Int. Ed. Engl. 1996, 35, 662-663). P. Barbaro, A. Togni, Organometallics 1995, 14, 3570-3573. P. Barbaro, C. Bianchini, A. Togni, Organometallics 1997, 16, 3004-3014. H. C. L. Abbenhuis, A. Togni, B. Müller, A. Albinati, U. Burckhardt, V. Gramlich, Organometallics 1994, 13, 4481-4493. H. C. L. Abbenhuis, U. Burckhardt, V. Gramlich, A. Martelletti, J. Spencer, I. Steiner, A. Togni, Organometallics 1996, 15, 1614-1621. Y. Nishibayashi, J. D. Singh, S.-i. Fukuzawa, S. Uemura, J. Chem. Soc. Perkin Trans. I 1995, 2871-2876. Y. Nishibayashi, J. D. Singh, S.-i. Fukuzawa, S. Uemura, J. Org. Chem. 1995, 60, 4114-4120. Y. Nishibayashi, K. Segawa, J. D. Singh, S.-i. Fukuzawa, K. Ohe, S. Uemura, Organometallics 1996, 15, 370-379. A. Masson-Szymczak, O. Riant, A. Gref, H. B. Kagan, J. Organomet. Chem. 1996, 511, 193197. H. B. Kagan, P. Diter, A. Gref, D. Guillaneux, A. Masson-Szymczak, F. Rebière, O. Riant, O. Samuel, S. Taudien, Pure & Appl. Chem. 1996, 68, 29-36. M. Sawamura, H. Sasaki, T. Nakata, Y. Ito, Bull. Chem. Soc. Jpn. 1993, 66, 2725-2729. B. Jedlicka, C. Kratky, W. Weissensteiner, M. Widhalm, J. Chem. Soc. Chem. Comm. 1993, 1329-1330. B. Jedlicka, R. E. Rülke, W. Weissensteiner, R. Fernández-Galán, F. A. Jalón, B. R. Manzano, J. R.-d. l. Fuente, N. Veldman, H. Kooijman, A. L. Spek, J. Organomet. Chem. 1996, 516, 97110. G. Kutschera, C. Kratky, W. Weissensteiner, M. Widhalm, J. Organomet. Chem. 1996, 508, 195-208. A. Mernyi, C. Kratky, W. Weissensteiner, M. Widhalm, J. Organomet. Chem. 1996, 508, 209218. H. Wally, C. Kratky, W. Weissensteiner, M. Widhalm, K. Schlögl, J. Organomet. Chem. 1993, 450, 185-192. H. Wally, M. Widhalm, W. Weissensteiner, K. Schlögl, Tetrahedron: Asymmetry 1993, 4, 285288. T. Hayashi, C. Hayashi, Y. Uozumi, Tetrahedron: Asymmetry 1995, 6, 2503-2506. Y. Matsumoto, A. Ohno, S.-j. Lu, T. Hayashi, N. Oguni, M. Hayashi, Tetrahedron: Asymmetry 1993, 4, 1763-1766. L. Schwink, S. Vettel, P. Knochel, Organometallics 1995, 14, 5000-5001. D. Enders, R. Lochtman, G. Raabe, Synlett 1996, 126-128. C. Ganter, T. Wagner, Chem. Ber. 1995, 128, 1157-1161. A. Ohno, M. Yamane, T. Hayashi, N. Oguni, M. Hayashi, Tetrahedron: Asymmetry 1995, 6, 2495-2503. L. Schwink, P. Knochel, Tetrahedron Lett. 1996, 37, 25-28.
References [93] [94] [95] [96] [97] [98] [99]
[100] [101] [102] [103] [104]
721
G. Nicolosi, R. Morrone, A. Patti, M. Piattelli, Tetrahedron: Asymmetry 1992, 3, 753-758. G. Nicolosi, A. Patti, M. Piattelli, J. Org. Chem. 1994, 59, 251-254. A. Patti, D. Lambusta, M. Piattelli, G. Nicolosi, P. McArdle, D. Cunningham, M. Walsh, Tetrahedron 1997, 53, 1361-1368. F. Rebière, O. Riant, L. Ricard, H. B. Kagan, Angew. Chem. 1993, 105, 644-646 (Angew. Chem. Int. Ed. Engl. 1993, 32, 568-570). O. Riant, O. Samuel, H. B. Kagan, J. Am. Chem. Soc. 1993, 115, 5835-5836. M. Tsukazaki, M. Tinkl, A. Roglans, B. J. Chapell, N. J. Taylor, V. Snieckus, J. Am. Chem. Soc. 1996, 118, 685-686. D. Hoppe, F. Hintze, P. Tebben, M. Paetow, H. Ahrens, J. Schwerdtfeger, P. Sommerfeld, J. Haller, W. Guarnieri, S. Kolczewski, T. Hense, I. Hoppe, Pure & Appl. Chem. 1994, 66, 14791486. D. Price, N. S. Simpkins, Tetrahedron Lett. 1995, 36, 6135-6138. Y. Nishibayashi, Y. Arikawa, K. Ohe, S. Uemura, J. Org. Chem. 1996, 61, 1172-1174. J. C. Ruble, G. C. Fu, J. Org. Chem. 1996, 61, 7230-7231. P. I. Dosa, J. C. Ruble, G. C. Fu, J. Org. Chem. 1997, 62, 444-445. J. C. Ruble, H. A. Latham, G. C. Fu, J. Am. Chem. Soc. 1997, 119, 1492-1493.
722
11
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis
12 Metallocene-Based Polymers Timothy J. Peckham, Paloma Gómez-Elipe, and Ian Manners*
12.1 Introduction Since the discovery and elucidation of the structure of ferrocene in the early 1950s [1], this fascinating species has been a continual source of intrigue for chemists [2]. In addition to possessing a vast organic chemistry (involving e.g. electrophilic substitution reactions), this prototypical metallocene also undergoes reversible oneelectron oxidation to the blue ferrocenium ion. The oxidation of biferrocenes, biferrocenylenes (e.g. 1) and their derivatives has provided access to mixed-valent species which are delocalized on a variety of time scales depending on substituent and solid state environmental effects. Recently liquid crystalline ferrocenes have been prepared and ferrocene-based charge-transfer materials have attracted considerable attention with respect to their interesting magnetic properties [3–5]. Importantly, ferrocene is easy to prepare, cheap, and commercially available, and air-, moisture-, and thermally-stable. Based on these considerations it is not surprising that the possibility of obtaining ferrocene-based polymeric materials where the unique physical and chemical features of this organometallic species are combined with the facile processability of high molecular weight polymers has attracted intense interest. Indeed, research in this area was initiated shortly after the discovery of this novel organometallic compound and the first ferrocene-containing polymer, poly(vinylferrocene) (2), was reported in 1955 [6].
+ C
CH2
H C
C
Fe
Fe C
C
1
n Fe
2
724
12
Metallocene-Based Polymers
To access the favorable characteristics of macromolecular materials (e.g. mechanical strength, film and fiber formation etc) appreciable molecular weights (Mn > ca 10 000) are necessary [7]. Attempts to incorporate ferrocene into the side chain structure of polymers have been very successful. A variety of high molecular weight materials have been prepared with either organic or inorganic main chains by subtle modifications of existing synthetic methodologies. The incorporation of ferrocene into the main chain of polymers where the organometallic groups are joined by long spacers has also been successful. In such cases advantage can be taken of the facile derivatization of ferrocene to access difunctional monomers with organic functional groups. These species can then be effectively used in polycondensation reactions and high molecular weight polymeric materials result. In contrast, until recently, the synthesis of main chain polymers which contain ferrocene units in close proximity to one another has been of limited success. Indeed, until recently, examples of well-characterized, high molecular weight (Mn > 10 000) polymers of this type were exceedingly rare [8]. However, such materials are particularly desirable as the metal atoms are in close proximity which should allow metal–metal interactions. In such cases, polymers with interesting electrical, magnetic or other physical properties might be envisaged. The problem involving the synthesis of such materials arises from the fact that suitable, well-defined difunctional monomers are difficult to prepare in a high degree of purity. Under such conditions, polycondensation reactions are prone to yield low molecular weight oligomeric products. The discovery of ring-opening polymerization (ROP) routes to such materials in the early 1990s represented a key solution to this problem [8]. Such polymerizations proceed via a chain growth route which tend to yield high molecular weight polymers without the stringent, high molecular weight-inhibiting stoichiometry and conversion requirements characteristic of polycondensations. An analogy for this development involves the important recent advance in polycarbonate chemistry involving the discovery of ROP routes which afford polymers with substantially higher molecular weights than those normally available by polycondensation methods [9]. The incorporation of other metallocenes into polymer structures has also been explored but to a much lesser extent. Excellent, critical reviews of metallocenebased polymers were published as early as 1970 [10]. At this time the authors were more than aware of the limitations in the characterization of the vast majority of the materials reported at that time. Even now, many of the poly(metallocenes) reported in the literature are still of low molecular weight, poor solubility, and are inadequately characterized. This Chapter will focus mainly on materials that have been well-characterized, both in terms of structure and molecular weight and covers the literature up to the end of 1996 [11]. In a growing number of cases detailed studies of poly(metallocenes) are being performed. In addition, possible applications as electrode mediators, components of electronic devices and sensors, charge transport materials, precursors to magnetic solids, and bioactive materials are of considerable interest.
12.2
Side Chain Poly(metallocenes)
725
12.2 Side Chain Poly(metallocenes) 12.2.1 Organic Polymers with Metallocene Side Groups Poly(vinylferrocene) Poly(vinylferrocene) (2) was first synthesized shortly after the discovery of ferrocene by radical-initiated polymerization of vinylferrocene [6] (3) and this method has been reinvestigated in subsequent years [12–17]. The initiator normally used is AIBN as peroxide initiators will oxidize the monomer. Molecular weights are typically less than 10 000 but the synthesis of polymers using this route with Mw = 382 000 (PDI = 1.33) have been reported although such higher molecular weight polymers tended to possess a multimodal weight distribution [15]. Yields can range as high as 73% depending upon the conditions used. However, it has been found that the molecular weight of poly(vinylferrocene) does not increase with a decrease in initiator concentration. This is due to the high tendency for chain transfer with vinylferrocene in comparison to other vinyl monomers such as styrene [15].
Fe
3 Cationic initiation has also been used successfully to obtain poly(vinylferrocene) [18]. Due to the electronic rich nature of the ferrocene unit, it was generally believed that anionic initiators do not lead to the formation of homopolymer. Recently, however, living anionic polymerization of vinylferrocene in solution has been achieved at low temperatures [19]. Poly(vinylferrocene) may also be synthesized via Ziegler–Natta polymerization [18]. A great deal of work has been done on studying the properties of poly(vinylferrocene). The polymer itself is a yellow powder which is stable in air in the solid state. The material is also soluble in most common organic solvents [15]. However, in chlorinated solvents, the polymer is reported to be unstable and a green precipitate is formed after standing for a few weeks. The precipitate was believed to be ionic in nature and to contain ferrocenium units. The photolysis of poly(vinylferrocene) in the presence of CCl4 is believed to proceed as shown in Eq. (12-1) [15].
726
12
CH2
Metallocene-Based Polymers
H C
+
CCl4
hν
CH2
H C
+
n Fe
Fe
+
CCl3
(1)
n Cl-
The λmax for poly(vinylferrocene) in the visible region (in methylene chloride) occurs at 440 nm as determined by UV/visible spectroscopy [15]. The IR spectrum reveals peaks consistent with a monoalkyl ferrocene [15]. The 57Fe Mössbauer spectrum exhibits a doublet with an isomer shift of 0.78 mm s–1 and a quadrupole splitting of 2.44 mm s–1 [15]. The melting point of the polymer was reported to be 281–285°C and the glass transition temperature was assigned as 190°C (by DSC) [15]. Much of the interest in poly(vinylferrocene) is due to the presence of redox active iron sites attached to the polymer chain. Cyclic voltammetry experiments [20, 21] have revealed that the iron sites are non-interacting and only one reversible oxidation wave is observed (Fig. 12-1) for which the intensity of the current is direct-
Figure 12-1. Cyclic voltammogram of 2 in CH2Cl2. Reproduced from Ref. [20] with Permission.
12.2
Side Chain Poly(metallocenes)
727
ly proportional to the molecular weight of the polymer sample [20]. This is in contrast to polymers where the ferrocene units are in close proximity in the polymer backbone, such as poly(ferrocenylsilanes) (see Secs. 12.3.2 and 12.3.3). For these materials there is evidence for communication between iron centers as shown by the presence of two reversible oxidation waves. Poly(vinylferrocene) is an insulator in its pristine state (i.e. when all iron centers present as FeII) with conductivities in the region of 10–14 S cm–1 [22]. Oxidation with e.g. I2 yields mixed valent systems in which both FeII and FeIII sites are present. These materials display much higher conductivities with values in the semiconductor range (10–8–10–7 S cm–1) [22]. Based on studies of ferrocenium and mixed-valent biferrocene compounds [23, 24] and Mössbauer spectroscopic analyses of vinylferrocene-containing copolymers [25, 26], charge has been found to be mainly localized on iron. An electron-hopping model has been suggested as the mode for electron transfer [22] as the all-carbon backbone of the polymer is insulating. The redox behavior of poly(vinylferrocene) (PVFc+/0) has been utilized in the construction of a microelectrochemical diode along with an N,N′-dibenzyl-4,4′bipyridinium-based polymer (BPQ2+/+)n (see Eq. (12-2)) [27]. The polymers were deposited electrochemically upon the microelectrodes. Current passes when the applied potential results in the negative lead being attached to the (BPQ2+/+)n coated electrode and the positive lead being connected to the (PVFc+/0) coated electrode. As the applied potential approaches the difference in electrochemical potentials of the two polymers, current flows as the reaction in Eq. (12-2a) is favorable by ≈ 0.9 V. However, current does not flow if the applied potential is in the opposite sense as the reaction in Eq. (12-2b) is uphill by ≈ 0.9 V. The switching time of this diode, however, was long in comparison with solid-state diodes [27].
+
Fc+
BPQ2+ +
Fc0
(2a)
BPQ2+ +
Fc0
BPQ+
Fc+
(2b)
BPQ+
+
Other Organic Polymers with Metallocene-Containing Side Groups A variety of other organic polymers with metallocene-containing side groups have been synthesized. However, few are as well-studied as poly(vinylferrocene). This section will deal primarily with more recent, well-characterized examples. A random copolymer (4) of methyl methacrylate (95 mol%) and a ferrocenecontaining methacrylate (5 mol%) has been synthesized using free radical polymerization [28]. The molecular weight (as determined by GPC relative to polystyrene)
728
12
Metallocene-Based Polymers
of 4 was Mn = 30 000 and its thermal behavior was nearly identical to that of poly(methyl methacrylate). After poling, the material was analyzed for second-order nonlinear optical properties and was found to possess a second harmonic generation efficiency approximately four times that of a quartz standard.
Me C
Me CH2
C
COOCH3
CH2
COOR
n
R=
Fe
CO2Et CN
4 Ring-opening metathesis polymerization (ROMP) of norbornene with ferrocene side groups has provided another route to organic polymers with metallocene-containing side groups such as 5. Due to the living nature of the polymerization, block copolymers could also be prepared [29, 30].
COOMe
Fe
n 5
12.2
Side Chain Poly(metallocenes)
729
Molecular weights for 5 could be controlled from Mn = 5090–9030 with PDI = 1.13 or less. DSC measurements on the block copolymers revealed thermal behavior consistent with a block rather than a random structure. As was found for poly(vinylferrocene), cyclic voltammetry measurements showed the presence of only one reversible oxidation wave, consistent with noninteracting redox sites. The living end of 5 could also be capped with pyrene. The presence of the neighboring ferrocene moieties led to the emission of the pyrene unit being quenched by a factor of 30 compared with emission from 1-vinylpyrene [30]. The synthesis of poly(ethynylferrocene) (6) has been accomplished by a number of different researchers [31–35]. However, the most recent report has provided more well-defined materials with controlled molecular weights as well as similar examples of the analogous poly(ethynylruthenocene) (7) [35].
CH C
CH C
n
n Fe
Ru
6
7
The molecular weights for 6 could be controlled from Mn = 2300–14 600 (values determined by GPC using a light scattering detector) and PDI = 1.06–1.62. The corresponding values for 7 were 6700–16400 and 1.14–1.24, respectively. The living polymers could be terminated with pyridinecarboxaldehyde and the tertiary nitrogen subsequently quaternized with methyl iodide. The charged systems were found to exhibit a more intense, red-shifted absorption in the UV/visible spectrum than the corresponding uncharged systems. Ruthenium and osmium analogs of poly(vinylferrocene) have been reported in the past but characterization was limited. They were tested as preheat shields for targets in inertial-confinement nuclear fusion [36, 37]. Recently, a conjugated polymer with ferrocene substituents, poly(N-methylferrocenylaniline) (8), was employed as an electrode mediator and could be used as sensor for peroxides in organic media [38].
730
12
Metallocene-Based Polymers
N CH2
Fe
n 8
12.2.2 Inorganic Polymers with Metallocene Side Groups Very few examples exist of inorganic polymers with metallocene-containing side groups. The polymers which have been prepared are polyphosphazenes, polysilanes and polysiloxanes. These are outlined in the following sections.
Polyphosphazenes with Ferrocene- or Ruthenocene-Containing Side Groups The most common route to polyphosphazenes is via the thermal ring-opening polymerization of cyclic phosphazenes with halogen substituents at phosphorus. Using this method, ferrocene and ruthenocene-containing polyphosphazenes (9 and 10) have been synthesized with molecular weights (Mw) in excess of 2 × 106 [39, 40].
M M OR P
N
OR
P
OR N
OR
9a (M = Fe) 9b (M = Ru)
P OR
OR N
P
n
OR
N
P OR
N
P OR
10a (M = Fe) 10b (M = Ru)
N
n
12.2
Side Chain Poly(metallocenes)
731
In general, cyclotriphosphazenes with less that three halogen substituents do not undergo thermal ring-opening polymerization. However, a range of cyclic phosphazenes without any halogen side groups (11) have been found to polymerize provided transannular ferrocenyl substituents are present. ROP of 11 yields polymer 12 (Eq. (12-3)) with molecular weights as high as Mw = 1.8 × 106 (PDI = 6.2) [41, 42]. In most cases, the presence of a small amount of an initiator such as [NPCl2]3 is necessary for polymerization to occur. X-Ray crystallographic studies of several of these strained ferrocenylorganocyclotriphosphazenes have shown that the phosphazene ring is forced into a high energy, nonplanar conformation [43]. By contrast, in most cyclotriphosphazenes the phosphorus–nitrogen ring is virtually planar.
11
12
R = non-halogen substituent
Polymers 9 and 10 (as well as copolymers bearing both ferrocene and ruthenocene moieties) have been partially oxidized with iodine resulting in weakly semiconducting materials. These materials have also been deposited on an electrode surface where the polymers act as electrode mediator coatings which can aid electron transfer between the electrode and redox active species in solution [44]. Polyphosphazenes with ferrocenyl substituents (13) have also been synthesized via functionalization of poly(methylphenylphosphazene) as shown in Eq. (12-4) [45].
Ph P Me
1. 0.5 n-BuLi
N
2. O n
R
Ph
Ph
P
N
P
Me
C
CH2 R
C
OH
n
Fe
3. H+
(4)
N
Fe
13a (R = H) 13b (R = Me)
732
12
Metallocene-Based Polymers
Substitution was not complete with a degree of substitution of 45% and 36% for polymers 13a and 13b, respectively. Molecular weights for these polymers were Mw = 2.0 × 105 and 1.5 × 105 for 13a and 13 b respectively (typically PDI = 1.4– 2.0). The glass transition temperatures increased in comparison with the unsubstituted polymers with values of 92°C and 87°C for 13a and 13b respectively.
Polysilanes and Polysiloxanes with Metallocene-Containing Side Groups The synthesis of polysilanes bearing low loadings of ferrocene substituents (14) has been reported via Wurtz coupling [46]. Random copolymers with molecular weights (Mw) up to 390 000 were isolated from mixtures of high and low molecular weight fractions. The ratio of methylphenylsilane to methylferrocenylsilane segments ranged from 6:1 to 27:1. In common with other polysilanes, these materials were photosensitive and depolymerized upon exposure to UV light. However, the copolymers did display greater stability than a corresponding sample of poly(methylphenylsilane) indicating that ferrocene moieties bound to the polymer chain provided a degree of stabilization for the polysilane backbone towards photodegradation. Cyclic voltammetry experiments [47] revealed two oxidation waves for the polymers. The first wave was reversible due to oxidation of iron sites whereas the second, irreversible oxidation at higher potential was assigned to oxidation of the polysilane backbone. No electrochemical interaction was observed between neighboring iron sites or between iron sites and the polymer backbone. Me
Me
Si
Si Ph
Fe
n 14 Polysiloxane random copolymers 15 (with molecular weights of 5000 to 10 000) with pendant ferrocene groups have recently been synthesized via hydrosilylation of vinylferrocene or 9-ferrocenyl-1-nonene with a poly(methylhydrosiloxane)poly(dimethylsiloxane) random copolymer (Eq. (12-5)). These materials were used as amperometric biosensors for the detection of glucose [48]. The polymers effectively mediated electron transfer between reduced glucose oxidase and a conven-
12.3
Main Chain Poly(metallocenes) with Short Spacers
733
tional carbon paste electrode. The response of the sensor to glucose was dependent upon the nature of the polymeric backbone. The optimal response was a compromise between increased polymer flexibility and decreased spacing between individual relay (i.e. ferrocene) sites.
12.3 Main Chain Poly(metallocenes) with Short Spacers 12.3.1
Poly(metallocenylenes) and Poly(metallocenes) with Short Spacers via Condensation Routes
Poly(metallocenylenes) The earliest routes to poly(ferrocenylene) involved free-radical processes (Eq. (126)). Korshak [49, 50] and Nesmeyanov [51] first reported in 1960 the polyrecom-
734
12
Metallocene-Based Polymers
bination reaction of ferrocene at 200°C in the presence of t-butyl peroxide as the radical source. The yields obtained were low (5–16%) and the molecular weight (Mn) was up to 7000 in the best conditions. This work was critically reviewed in 1970 [10]. The authors reinvestigated the original work and they found that the soluble polymer was composed of a mixture of short poly(ferrocenylene) segments, with a mixed substitution (represented by 16) on the ferrocene of homoannular (1,2- and 1,3-) and heteroannular (1,1′-) units, linked by methylene and oxygen-containing hydrocarbon units [52]. H
n
R Fe
H
(6)
Fe
n 16 Since then, different radical reactions have been reported such as mixed Ullmann reactions of halo- and 1,1′-dihaloferrocene with copper bronze [53, 54], coupling reactions of lithio- and 1,1′-dilithioferrocene with cobalt chloride in the presence of organic halides [55–57], treatment of poly(mercuriferrocenylene) with metallic silver [58] or in molten ferrocene [59], or glow discharge polymerization [60]. All led to oligomers with molecular weights in the range Mn = 1000–3500, with poorly defined structures, low conductivity (σ = 10–7 to 10–9 S cm–1) and irregular magnetic behavior. This may be due to the presence of oxidized iron sites within the oligomer chain [10]. In the case of glow discharge polymerization, cross-linked poly(ferrocenylene) is produced. Other reported methods, like polyrecombination from chloromercuri- and bis(chloromercuri)ferrocene assisted by palladium salts [61] or dilithioferrocene.TMEDA with cobalt chloride [62], resulted in formation of paramagnetic oligomers in low yields together with biferrocenylene resulting from internal cyclization. Pittman et al. [22, 63, 64] studied the conductivity of unoxidized oligo(1,1′-ferrocenylene) prepared by radical recombination of ferrocene [24], concluding that the material was an insulator. However, upon partial oxidation, the mixed-valence states induced increased conductivity values from 10–14 to a maximum of 10–6–10– 8 S cm–1 at room temperature. Similar values have been observed for biferrocenylene [65, 66] and biferrocene [23] salts. The most impressive early results on poly(ferrocenylenes) (in terms of yield, molecular weight and purity) were obtained by Neuse et al. [67] using the stepgrowth polycoupling reaction of equimolecular amounts of dilithioferrocene.TMEDA with diiodoferrocene at temperatures not exceeding 25°C (Eq. (127)). Thus, a total yield of 85% of light tan, soluble, spectroscopically well-defined
12.3
735
Main Chain Poly(metallocenes) with Short Spacers
(all heteroannular), amorphous and diamagnetic poly(ferrocenylene) (17) was produced with Mn < 4000 (determined by VPO), from which subfractions in the range Mn = 1000–10 000 were obtained. Li
n
Fe Li
X Fe
Fe X
n
– MgX2 X = Br, I
(7)
Mg
17
More recently, dehalogenation of dihaloferrocenes with magnesium produced semicrystalline poly(ferrocenylene) in 45–77% yield (Eq. (12.7)) [68, 69]. Soluble fractions of this product (Mn = 4600), when partially oxidized with TCNQ, showed a higher degree of crystallinity and, presumably as a consequence [70], higher electrical conductivity (σ = 10–2 S cm–1) than found previously. Rapid electron transfer between FeII and FeIII sites was also observed for these TCNQ-doped materials on the Mössbauer spectroscopic time scale (10–7 s) at room temperature. This feature was supported by electrochemical studies on poly(ferrocenylene) modified electrodes [71], which indicated hole delocalization among iron atoms. Further studies on the electronic structure, by UV photoelectron spectroscopy [72], attributed the moderate conductivity for these polymers to a limited hole mobility and described them as systems with intermediate delocalization between π-conjugated conducting polymers such as polyacetylene and localized polymers such as polyethylene. Nishihara et al. [73] have recently developed a methodology which permits the synthesis of more soluble substituted poly(ferrocenylenes) by reaction of the dihexylfulvene dianion with [FeCl2(THF)2] (Eq. (12-8)). The presence of the hexyl group increased the solubility of the products. Although this polycondensation method is not appropriate for the preparation of high molecular weight materials, a fraction of poly(1,1′-dihexylferrocenylene) (18) with Mn > 4000 was formed in low yield. In addition, when oxidized with TCNQ or TCNE, this material exhibited photoconductivity and acted as a p-type semiconductor, as was previously noted for nonsubstituted ferrocenylenes [71].
736
12
Metallocene-Based Polymers
R
2–
R
FeCl2(thf)2
Fe
R
(8) R
n
18 R = C6H13 Poly(metallocenes) with Short Spacers via Condensation Routes Poly(metallocenes) with single carbon bridges (i.e. 19) have been the focus of numerous investigations [10]. Condensation routes have involved the reaction of aldehydes with the corresponding metallocene (M = Fe, Ru), as well as the cationic polymerization of carbinols. Both were carried out in the presence of Lewis or protic acid-catalysts and yielded similar products. The materials obtained were oligomeric in nature and consisted of a mixture of heteroannular (1,1′-) and homoannular (1,2- and 1,3-) substituted metallocene units. Crosslinking was noticed at high temperatures resulting in an increase in the number of methylene bridges per repeat unit. A
R CHR Fe
Fe
R = H, CH3, C6H5
Ph
Fe
Ph
n
n 19
P
P
20
n 21 A = O, S
A similar route was reported by Neuse [74] and Pittman [75] for the synthesis of phosphorus-bridge poly(ferrocenes) (20, 21). Reaction of ferrocene with PPhCl2, P(O)PhCl2 or P(S)PhCl2 in the presence of a Lewis acid catalyst in the melt or in solution at 80–170°C yielded low molecular weight materials (Mn < 6000, by VPO) the composition of which depended on the reaction conditions. Again, in these materials a mixture of different ferrocenylene linkages is present. The polymers were characterized by 1H NMR and IR spectroscopy, and elemental analysis. Well-defined materials were obtained by Seyferth [76] from the polycondensation reaction of 1,1′-dilithioferrocene.TMEDA with PPhCl2. The molecular weights
12.3
Main Chain Poly(metallocenes) with Short Spacers
737
of the poly(ferrocenylphosphine) products were found to depend on the reaction conditions. For example, the authors reported products with Mw = 8900 (by light scattering) when the reaction was performed in dimethoxyethane at 25°C, while high molecular weights (Mw = 131 000–161 000) were obtained in ether at 25°C or DME at –40°C. The latter feature is highly unexpected when the highly reactive, impure dilithioferrocene is used in condensation reactions, since exact stoichiometries are required in order to achieve high molecular weight materials. Recent work has shown that these high molecular weight products probably arise from a chain growth reaction rather than a polycondensation (see Sec. 12.3.2). This involves the anionic ring-opening polymerization of an in situ generated [1]phosphaferrocenophane. The air and thermally stable (Tdec > 350°C) poly(ferrocenylphosphines) were shown to react with low concentrations of Co2(CO)8, to chelate cobalt in a tridentate fashion [77]. According to IR and 31P NMR, three Co–P bonds are formed and the cobalt centers present a pseudotrigonal bipyramidal geometry. The catalytic potential of the materials in the hydroformylation of 1-hexene was studied and was shown to be similar to the complex HCo(CO)3PPh3. Poly(ferrocenylsilanes) (22) were first described in two patents by Rosenberg in the 1960s [78, 79]. Low molecular weight polymers were prepared from dilithioferrocene by a reported polycondensation reaction with the appropriate dihaloorganosilanes (Me2SiCl2 or Ph2SiCl2) in polar solvent mixtures at 0–25°C. The resulting partially soluble and reasonably thermally stable (Tdec > 250°C) polymers were characterized by elemental analysis, although their dark color indicated that they were impure. VPO gave molecular weights (Mn) of 1700 (R = Me) and 4000–7000 (R = Ph) which corresponded to 5–20 units.
Fe
R
R
Si
Sn
R
Fe
n 22 R = Me, Ph
R
n 23 R = Me, Et, n–Bu, Ph
An alternative route involved the condensation of FeCl2 with Li2 [Cp2SiMe2] [78, 80], which afforded poly(ferrocenyldimethylsilane) with Mn < 4100. These values for Mn would be expected for condensation processes involving dilithio reagents which are difficult to obtain in high purity. Following a similar reaction pathway, Osborne [81] and Seyferth [82] have reported the synthesis of low molecular weight poly(ferrocenylstannanes) (23). Only when R = Me was soluble material (with Mw = 4600) obtained. More recently, Foucher and co-workers [83] have isolated a soluble low molecular weight (Mw =
738
12
Metallocene-Based Polymers
1.4 × 104) elastomer (R = n-Bu) when using low temperature reaction conditions. The oligomer was characterized by 119Sn, 13C and 1H NMR spectroscopy and GPC. Poly(ferrocenylenevinylene) derivatives (24) with Mn = 3 000–10 000 (by GPC) have recently been synthesized [84] in high yields via the titanium-induced McMurry coupling reaction of the corresponding alkylferrocenyl carbaldehydes monomers (Eq. (12-9)). These soluble polymers were characterized by NMR, and IR which revealed the presence of trans-vinylene units. The UV–vis spectra of the polymers are similar to those of the monomers indicating a fairly localized electronic structure. This is also reflected in the electrical and optical properties with values for conductivity (σ = 10–2 S cm–1) and nonlinear third-order optical susceptiblility (χ(3) = 1–4 × 10–12 esu) lower than those of linear conjugated polymers such as poly(1,4-phenylenevinylene) (σ = 2.5 × 103 S cm–1, χ(3) = 8 × 10–12 esu). R
CHO
R
CH
Zn–TiCl4
Fe
(9)
Fe
R
OHC
CH
R
n
24 R = C2H5, C6H13, C12H25 Disulfide-bridged oligoferrocenes (Mw < 3500) were prepared in an attempt to copolymerize ferrocenedithiol with norbornadiene in the presence of AIBN in toluene [85, 86]. Rather than the expected radical-induced polyaddition [87] (Eq. (12-10a)) a condensation of the ferrocene monomer was observed with norbornene and nortricyclane units as end-groups (Eq. (12-10b)). The oligomers (25) were characterized by NMR and field-desorption mass spectrometry, and have a chain length of 2–12 units as determined by GPC. S S
SH
Fe Fe
Fe
(10a)
S S
SH
n
+ AIBN Toluene
S
S Fe
Fe S
25
S
n
n = 1–11
(10b)
12.3
Main Chain Poly(metallocenes) with Short Spacers
739
Another group of polymers to consider are poly(metalloxanes) where metallocene fragments are connected through the metal via oxygen or other elements. Most of the work done in this area involves titanium derivatives, and was carried out with the aim of exploiting the catalytic properties that the TiCp2 unit can introduce to the polymer. Several attempts have been made to synthesize polytitanoxane (26). Thus, Giddings [88] claimed the synthesis of polytitanoxane by sequential reduction–oxidation of titanocene dichloride in organic solvents, since the use of aqueous basic conditions lead to metal-ring bond cleavage. The polymer obtained, however, is an insoluble crosslinked material with poor thermal and hydrolytic stability in which part of the cyclopentadienyl substituents had been spontaneously eliminated. The characterization was poor and the structure (27) was postulated based on elemental analysis. A linear polymer product was postulated when [Cp2Ti]2 or Cp2Ti(CO)2 were used as starting materials but, also in this case an insoluble material was obtained with a suggested peroxo-bridged structure based on its diamagnetic behavior [89].
Ti
O
Cl
Ti
O
O
n 26
Ti
Cl
n
27
12.3.2 Ring-Opening Polymerization (ROP) of Strained Metallocenophanes Introduction ROP reactions generally occur via a chain growth mechanism and represent a particularly desirable route for the preparation of high molecular weight poly(ferrocenes). As is clear from the polymers described above, such materials are very rare, particularly if the ferrocene groups are in close proximity to one another so as to permit metal–metal interactions. The first syntheses of poly(metallocenes) via ROP were reported in early 1992. The atom-abstraction induced ROP process discovered by Rauchfuss and co-workers was described early in that year and is discussed in Sec. 12.3.3. Thermal ROP of strained metallocenophanes, which is discussed in this section, was reported by Manners et al. a few months thereafter. A key requirement for a classical ROP process is a strained cyclic monomer. Metallocenophanes in which a single atom bridges the two cyclopentadienyl ligands (i.e. [1]metallocenophanes) possess strained ring-tilted structures and have
740
12
Metallocene-Based Polymers
been known since 1975. Such species possess tilt-angles between the planes of the Cp ligands of 6–31° but no successful ROP reactions for these species had been reported until very recently. However, in the early 1980s Seyferth et al. showed that phosphorus-bridged [1]ferrocenophanes undergo anionic ring-opening oligomerization to yield oligo(ferrocenylphosphines) (20) with 2–5 ferrocene units (mainly the dimer and trimer) [76]. Although anionic ROP was also attempted by these researchers using small quantities of anionic initiator, these reactions were unsuccessful and no species longer than a pentamer was prepared [76].
Thermal Ring-Opening Polymerization of Silicon-Bridged [1]Ferrocenophanes The first examples of the use of ROP of strained metallocenophanes to prepare high molecular weight poly(metallocenes) involved organosilane spacers [90]. Specifically poly(ferrocenylsilanes) such as 29 were prepared via the thermal ringopening polymerization of strained, ring-tilted [1]ferrocenophanes (28) (Fig. 12-2). These materials were prepared by heating silicon-bridged [1]ferrocenophanes in the melt, at 130–220°C in evacuated tubes (Eq. (12-11)) [91].
Figure 12-2. Molecular Structure of 28 (R = R’ = Me) (where the thermal ellipsoids are at the 25% probability level). Reproduced from Ref. [93] with Permission.
12.3
Main Chain Poly(metallocenes) with Short Spacers
741
R Si Si
Fe
R R'
heat
Fe
(11)
R'
n 28
29
The thermal ROP route is very general and a wide range of semicrystalline, glassy or elastomeric poly(ferrocenylsilanes) have been prepared from silicon-bridged [1]ferrocenophanes with a range of different substituents at silicon or the cyclopentadienyl rings [92–100]. These polymers have been fully characterized using a range of spectroscopic and analytical methods and their molecular weights (Mw), which have been established by absolute methods such as low angle laser light scattering, lie in the 105–106 range with polydispersity indices of around 2. A key feature that was initially noted concerning these polymers was the presence of two reversible oxidation waves in their cyclic voltammograms (e.g. Fig. 12-3) [90]. The proposed explanation invoked initial oxidation at alternating iron
Figure 12-3. Cyclic voltammogram of 29 (R = R’ = Me) in CH2Cl2 at scan rates of 500 and 1000 mVs-1 Reproduced from Ref. [96] with Permission.
742
12
Metallocene-Based Polymers
sites as a consequence of interactions between the iron atoms. Thus, as one iron center is oxidized, the neighboring sites become more difficult to oxidize and therefore do so at a higher potential and so two oxidation waves result. Recent studies on model oligo(ferrocenylsilanes) fully support this theory [101, 102]. The peak separations for the different polymers, which reflect the degree of interaction between the metal centers, are generally in the range ∆E_ = 0.16–0.29 V depending on the substituents present [90, 95, 96, 98]. When oxidized, poly(ferrocenylsilanes) undergo a color change from amber to blue. The interesting electrochromic behavior has been studied in some detail for several polymers [103, 104]. Pristine high molecular weight poly(ferrocenylsilanes) are insulating (σ = 10–14 S cm–1) and when partially oxidized by I2, the conductivities (σ = 10–8–10–3 S cm–1) lie in the semiconductor range and show electron localization on the Mössbauer spectroscopic timescale. Intriguingly, studies on TCNE-oxidized oligo(ferrocenylsilanes) (Mw = ca 1500) reported by Garnier et al. indicated the presence of electron delocalization and in some cases the presence of ferromagnetic ordering at low temperature [105]. Poly(ferrocenylsilanes) such as 29 (R = R′ = Me) are thermally stable up to 350–400°C with a retention weight of 35–40% at 1000°C yielding ferromagnetic Fe–Si–C ceramic composites [106, 107]. The insoluble, semicrystalline poly(ferrocenyldihydrosilane) 29 (R = R′ = H) gave the highest ceramic yield with a weight retention of 90% at 600°C and 63% at 1000°C [108]. Glass transitions for poly(ferrocenylsilanes), as observed by DSC, cover a wide range of temperatures (–26 to 116°C). Symmetrically substituted derivatives often show a propensity to crystallize and have been studied by DSC, WAXS, and atomic force microscopy [109]. The morphology of many of the materials show thermal history dependence with an increase in crystallinity over time. Molecular mechanics modelling studies on oligo(ferrocenylsilanes) have also been performed and provide excellent insight into the conformational preferences of macromolecules in the solid state [110, 111]. These provided results which are interesting to compare with those from single crystal X-ray diffraction work performed on species such as the linear pentamer. Another, recently studied feature of poly(ferrocenylsilanes) is their tunability. Thus functional materials can be prepared via substitution reactions on polymers with Si–Cl bonds [112]. More recently, hydrosilylation reactions have been used with polymers containing Si–H groups to attach mesogenic groups which allows access to thermotropic side chain liquid crystalline materials [113].
Thermal Ring-Opening Polymerization of Other Strained Metallocenophanes This ROP methodology has been extended to other strained metallocenophanes. Thus, poly(ferrocenylgermanes) (30) have been reported as high molecular weight (Mw = 105–106) materials from the ROP of the corresponding [1]ferrocenophanes
12.3
743
Main Chain Poly(metallocenes) with Short Spacers
at 90°C (Eq. (12-12)) [114]. The resulting golden-yellow polymers exhibit similar electrochemical and thermal behavior as poly(ferrocenylsilanes) with slightly lower Tg values [103, 114–117]. R Ge Fe
Ge
R R
heat
Fe
(12)
R
n 30
Poly(ferrocenylphosphines) (31), similar to those previously synthesized by condensation routes [76], have also been prepared via thermal ROP (Eq. (12-13)) [118]. Solubility increased when one of the cyclopentadienyl rings is substituted with an n-butyl group. Sulfurization of the unsubstituted and n-butyl substituted polymers was carried out in order to facilitate their characterization by GPC. Polymers with trimethylsilyl substituents on the cyclopentadienyl rings, however, could be analyzed by GPC without sulfurization. A comparison between sulfurized and unsulfurized samples of this polymer revealed essentially the same molecular weight. S P
P Fe
P
heat R
Fe
S8
R
Fe
(13) n
n 31
R
32
Polymerization of sulfur-bridged [1]ferrocenophane (33), that holds the largest tilt angle reported to date (31.05°) [119], produces poly(ferrocenylsulfide) (34) [120, 121]. The recent synthesis of a [1]stannaferrocenophane has led, upon heating at 150°C, to the first high molecular weight poly(ferrocenylstannane) (Mw = 1.3 × 105, PDI = 1.6). The polymer 35 also possess significant interactions between iron atoms as shown by cyclic voltammetry (∆E_ = 0.24 V) [122].
744
12
Metallocene-Based Polymers t
Bu
S Fe
S
Sn tBu
Fe
Fe
n 33
n
34
35
Hydrocarbon-bridged [2]ferrocenophanes (36, M = Fe) were found to undergo ROP at 300°C providing the first reported poly(ferrocenylethylenes) (37a) Eq. (12-14) [123]. The polymers were insoluble if R = H, but readily soluble when R = Me. In the latter case a bimodal weight distribution was found, with an oligomeric fraction (Mw = 4800) and a high molecular weight fraction (Mw = 8.1 × 104). Due to the more insulating hydrocarbon bridge only slight electronic communication is observed by cyclic voltammetry (peak separation of ∆E½ = 90 mV). Nevertheless, cooperative magnetic behavior is observed in oxidized materials at low temperature [124]. The analogous [2]ruthenocenophanes, which possess greater ring-tilt angles, undergo ROP at lower temperatures (220°C) to yield poly(ruthenocenylethylenes) (37b) although their electrochemistry is significantly different [125]. R
R
CH2 CH2 M
heat
CH2
(14)
M
CH2 R
R
36 R = H, Me
n
(37a) M = Fe (37b) M = Ru
The use of thermal ROP also allows access to novel random copolymers. The first examples, poly(ferrocenylsilane)–co-poly(silane) materials (38), were reported in early 1995 (Eq. (12-15)) [126]. These materials are particularly interesting since they contain ferrocene moieties linked by σ-delocalized oligosilane segments and these are inaccessible via thermal ROP of ferrocenophanes with oligosilanes bridges, as they are insufficiently strained [93]. The polymers are photosensitive and the polysilane segments can be cleaved with UV light. Interesting charge transport properties are also anticipated for these materials.
12.3
745
Main Chain Poly(metallocenes) with Short Spacers
Me Si
Fe
Me
Me
Me Si
150 °C
+ Me Ph Me
Si Me Ph
Fe
Me
Ph
Me Si
Si
Si Me
Ph
n
Me
Si
Si
Ph
Si
Si
Me
Ph
(15)
Ph
38
Ph
Subsequently, a range of other copolymers derived from different [1]ferrocenophanes [108] as well as [1]ferrocenophanes and silicon-bridged bis(benzene)chromium complexes [127] (39, Eq. (12-16)) have been reported. Me Si
Fe
Me SiMe2
heat
+
Fe
(16)
Cr SiMe2
n Me Cr
39
Si Me
Anionic Ring-Opening Polymerization of Strained Metallocenophanes In 1994, the first anionic ROP reactions for [1]ferrocenophanes were reported. Silicon-bridged [1]ferrocenophanes such as 28 (R = R' = Me) were first shown to oligomerize with lithioferrocene as initiator [101]. By decreasing the initiator concentration further, poly(ferrocenylsilanes) were produced. Moreover, using extremely pure monomer and stringent reaction conditions, the anionic ROP of 28 (R = R' = Me) was shown to be living [128]. Poly(ferrocenylsilanes) 29 (R = R' = Me) with controlled molecular weights up to Mw = ca 80 000 and with very narrow polydis-
746
12
Metallocene-Based Polymers
persities (PDI = 1.05–1.10) were formed depending on the monomer to initiator ratio. As the anionic ROP is living, the synthesis of diblock and multiblock copolymers (such as 40) using [1]silaferrocenophanes and hexamethylcyclotrisiloxane and/or styrene has been possible [129]. These block copolymer materials are of interest for the generation of redox active domains in the solid state (spheres, cylinders, lamellae etc) and novel micellar aggregates in solution [129]. Me
Me
Si Me Me3Si
O
CH CH2
Si
CH2 CH
Fe Me
Si
Me y
x
Me 3z
Me
Fe
y Si O Me
x
SiMe3 3z
40 x = 43, y = 47, z = 37
As mentioned above, oligomeric ferrocenylphosphines with up to 5 ferrocene groups have been prepared from the reaction of [1]phosphaferrocenophane with approximately equimolar amounts of lithio(diphenylphosphino)ferrocene [76]. However, previous attempts to prepare polymers by reducing the initiator concentration were reported to be unsuccessful. Recently, it has been demonstrated that anionic ROP of phosphorus-bridged [1]ferrocenophanes does indeed occur [130]. Poly(ferrocenylphosphines) with controlled molecular weights (up to Mw = ca 36 000) and narrow polydispersities (PDI = 1.08–1.25) were prepared together with block copolymers. The demonstration that anionic ROP of phosphorus-bridged [1]ferrocenophanes is possible suggests that the high molecular weight poly(ferrocenylphosphines) (20a) generated under certain conditions during the condensation polymerization of dilithioferrocene.TMEDA and PhPCl2 reported by Seyferth in the early 1980s almost certainly occurs via a chain growth route, namely the anionic ROP of a [1]ferrocenophane which is generated in situ under the reaction conditions used (Eq. (12-17)). It is plausible that similar types of anionic ROP reactions occur at least to some extent in the original condensation routes to poly(ferrocenylsilanes) described in the 1960s by Rosenberg [91].
P
Li
PhPCl2
RLi Fe
Fe
P
Ph
Fe
Li
Ph
(17) n
20a R = n-Bu, fc, or Fc
12.3
747
Main Chain Poly(metallocenes) with Short Spacers
Transition Metal Catalyzed Ring-Opening Polymerization of Metallocenophanes A very convenient transition metal catalyzed ROP route to poly(metallocenes) from metallocenophanes was reported in 1995. Manners et al. reported that a series of RhI, PdII, and PtII complexes catalyzed the ROP of [1]silaferrocenophanes at room temperature to yield high molecular weight poly(ferrocenylsilanes) (Mn = ca 105) [131]. A few months later, Tanaka et al. independently reported transition metal catalyzed ROP reactions using mainly Pd0 and Pt0 catalysts [132]. A major advantage of these transition metal catalyzed processes is that, unlike anionic polymerization, ambient temperature polymerization is possible without the need for a monomer of extremely high purity. Copolymerization of silicon- and germaniumbridged [1]ferrocenophanes is possible and affords random copolymers [115, 132]. Random copolymerization of silicon-bridged [1]ferrocenophanes with cyclocarbosilanes is also possible to yield novel poly(ferrocenylsilane)–poly(carbosilane) random copolymers [133]. Transition metal-catalyzed ROP is especially advantageous in the case of [1]ferrocenophanes with halogen substituents at silicon where thermal ROP proceeds at higher temperatures than normal (> 250°C) [112]. Moreover recent work [134] has shown that unsymmetrically Cp-methylated species such as 41 yield regioregular poly(ferrocenylsilanes) (42) via transition metal catalyzed ROP as a result of exclusive CpH–Si bond cleavage (Eq. (12-18)). In contrast, thermal ROP yields more random structures arising from cleavage of both CpH–Si and CpMe–Si bonds.
Me Si Me Fe
Si
Me
MLx Fe
Me
Fe
(18)
Me Si Me
41
n
42
Furthermore, molecular weight control in the range (Mn = 103–105) is also possible by the addition of Et3SiH which leads to the isolation of end-capped polymers 43. The use of poly(methylhydrosiloxane) as the source of Si–H bonds allows access to novel graft copolymers 44 [134]. Recent work has also focused on understanding the mechanism of the transition metal-catalyzed ROP reactions. A logical first step in the polymerizations is insertion of the transition metal into the strained Cp–carbon-bridging element bond in the ferrocenophane. Very recently Sheridan, Manners, and co-workers reported the first examples of this type of insertion reaction (Eq. (12-19)) [135].
748
12
Metallocene-Based Polymers Me
Me
Me
Me Si Fe
Si
H
Si
O
Me
Si
O
Me
H
O
m
Et3Si Fe
Me
43 Si Me
H
44
n
The [2]platinasilaferrocenophane 45 does not react further with [1]silaferrocenophane 28 (R = R' = Me). However, if the phosphine groups are replaced by more weakly bound ligands as in 46, synthesized via the reaction of 28 (R = R' = Me) with Pt(1,5-cod)2), the species does function as a ROP catalyst [136]. Moreover, the ROP is inhibited by the presence of added 1,5-cod supporting cycloolefin dissociation from 46 as a key step in the ROP mechanism.
Pt Fe SiMe2
46
12.3
Main Chain Poly(metallocenes) with Short Spacers
749
12.3.3 Atom-Abstraction-Induced Ring-Opening Polymerization of Chalcogenido-Bridged Metallocenophanes In early 1992, a novel atom-abstraction-induced ROP process was reported by Rauchfuss [137]. The synthesis of well-defined poly(ferrocenylene persulfides) (48) was achieved via the desulfurization of [3]trithiaferrocenophane (47) with PBu3 (Eq. (12-20)) and this led to yellow materials that were only soluble when ferrocene is substituted by alkyl groups. The molecular weights (Mw) have been determined to be in the range 12 000–359 000 for 48 (R = n-Bu, R′ = H) and 25 000 in the case of a less soluble copolymer between monomers 47 (R′ = H) with R = H and n-Bu in a 53:47 ratio. These polymers have been fully characterized by NMR and elemental analysis, and they exhibited very interesting properties [137–140], and are air-stable in the solid state but photosensitive in solution. The persulfide bond can be reversibly reductively cleaved with Li[BEt3H] and regenerated again upon reoxidation with I2. Electrochemical studies showed two reversible oxidation waves, as observed for poly(ferrocenylsilanes), that can be attributed to oxidation of alternating iron sites along the polymer chain. A wave separation of 0.32 V suggests that the interaction between iron sites is greater than that observed in diferrocenyl disulfide [141] (Fc-S-S-Fc) or poly(ferrocenylsilanes). More recently mono- and disubstituted t-Bu derivatives have been reported [142]. The polymerization was slower and solvent dependent yielding highly soluble polymers 48 with Mw values of 26 000 (R = t-Bu, R′ = H) and as high as 250 000 (R = R′= tBu), with large polydispersities. R
R
S
S S
Fe R'
47
S
PBu3, 25°C
S
Fe
(20)
– S=PBu3 R'
n
48 R = R' = H, t–Bu R = n–Bu, t–Bu R' = H
This atom-abstraction route was also used by the same authors for the preparation of soluble polymeric networks from alkylated [3,3'] bis(trithia)ferrocenophanes [138]. A bimodal molecular weight distribution, with maxima at Mn ≈ 5000 and 5 × 105, was obtained by GPC for the material from which a polymer fraction was isolated (Mn = 8.5 × 105). Low molecular weight poly(ferrocenylene perselenides) have been also prepared from the selenium analog of 47 [140]. These materials also undergo photodegradation upon exposure to UV light in air.
750
12
Metallocene-Based Polymers
12.3.4 Face-to-Face Poly(metallocenes) Multidecker sandwich structures, apart from those in which the stacking takes place upon crystallization, can be classified in the forms shown in Fig. 12-4 .
M
M M
M M
M M
M
a
b
M
M
c
Figure 12-4. Classes of multidecker sandwich structures.
Several attempts have been made to synthesize these types of polymeric structures, but in most cases dimer complexes, or double or triple decker structures, were produced. Rosenblum et al. have developed the synthesis of novel polymeric metallocenes (type b) held face to face by naphthalene or binaphthyl spacer groups. The condensation route involved palladium-catalyzed crosscoupling of chlorozinc-ferrocenes or ruthenocenes with 1,8-diiodonaphthalene [143, 144]. Since the ferrocene reagent is very reactive and cannot be prepared in pure form, it is difficult to achieve the exact stoichiometry of reactants required for a high degree of polymerization. Also, the rigidity [145] of the polymer framework makes these materials relatively insoluble and, consequently, with this methodology only materials up to Mn = 4000 have been afforded. Crystallographic studies on a trimetal oligomer, as a model for a segment of the polymer showed a cisoid arrangement for the naphthalene rings. Extrapolation would suggest a helical structure for the polymer [146]. An alternative, more successful route, in which a ferrocene monomer (49) is polymerized by treatment with sodium bis(trimethylsilyl)amide and FeCl2 at room temperature (Eq. (12-21)) has been recently reported by the same group [147]. The introduction of long hydrocarbon chains as substituents in the monomer unit clearly improves the solubility; thus, polymers 50 with values of Mn up to 18 000 can be easily prepared and in some cases components of higher molecular weight were also detected.
12.3 C6H13 R
C6H13
C6H13 R
Fe
751
Main Chain Poly(metallocenes) with Short Spacers C6H13 R
R
THF, NaN(SiMe3)2 MX2, 25˚C
(21)
Fe
M
n 49
50 M = Fe, Co X = Cl M = Ni, X = acac R = H, 2–octyl
This is a more general route to these materials and allows access to not only processable high molecular weight iron derivatives but also to nickel and mixed nickel– or cobalt–iron structures. However, to date, only low molecular weight products (Mn < 4000 for homometallic and < 2000 for heterometallic materials) have been described [148]. The electrical and magnetic properties for these polymers were investigated [148]. When the iron-based polymers were partially oxidized with iodine, an increase in the electrical conductivity from less than 10–12 S cm–1 to 6.7 × 10–3 S cm–1 was detected, reflecting semiconductor behavior. This value is almost two orders of magnitude lower than that for the analogous crystalline poly(1,1′-ferrocenylene) [68, 69]. The electrochemistry of oligomers reflects some degree of electronic interaction and charge delocalization between the metal centers of the partially oxidized species (∆E½ = 100–200 mV at 50 mV s–1) [143]. The corresponding I2-doped polymers appear to be weakly interacting mixed-valent systems by Mössbauer spectroscopy [148]. The bulk magnetic susceptibility was determined for the paramagnetic nickel polymer and heterometallic copolymers. In all of the polymers the values were greater than those for the corresponding nickelocene or cobaltocene which was interpreted in terms of cooperative magnetic behavior. Among recent attempts to obtain polymers of type c, where the successive metal-ring bonding is laterally displaced, is the interesting contribution by Katz et al. [149]. From the elaborate synthesis of [9]helicene hydrocarbon dianions they achieved, after treatment with CoBr2.DME and anion metathesis, the synthesis of soluble helical cobaltocenium oligomers (51, up to 6 units) in high yields. These are the first examples in which metallocenes are joined by fused conjugated systems of double bonds. Studies on their properties showed a weak interaction between metals via electrochemistry (∆E½ = 130 mV), similar to the face to face polymers described above, and a high optical activity.
752
12
Metallocene-Based Polymers
12.4 Main Chain Poly(metallocenes) with Long Spacers 12.4.1 Insulating Spacers Hydrocarbon Spacers The thermal stability and interesting physical properties of ferrocene were initially the motivation for the inclusion of ferrocene moieties into polymer chains. Thus, in the 1960s, a large number of attempts to produce ferrocene-containing organic polymers, including polyalkenes, polyketones, polyesters, polyamines, polyamides, polyurethanes, polyureas etc were reported. This work has been excellently reviewed by Neuse and Rosenberg in 1970 [10]. In general these polymers were very poorly characterized and were of low molecular weight. Often, no analytical data or structural characterization was given. The conditions used followed classic condensation methods at high temperatures and frequently involved side reactions [150] which led to impure materials or crosslinking and, when successful, the molecular weights for soluble materials were very low (Mn < 6000). Interfacial polycondensation routes at room temperature were considered as a convenient alternative to the classic condensations. They were first reported in the early 1960s by Knobloch and Rauscher [151], for the preparation of polyamides and polyesters from the reaction of 1,1′-ferrocenyldicarbonyl chloride with several diamines and diols. Okawara [152] reported the synthesis of polyurethanes from the condensation of 1,1′-ferrocenedimethanol and 1,1′-bis(dihydroxyethyl)ferrocene with diisocyanates, although the polymers possessed low molecular weights.
12.4
753
Main Chain Poly(metallocenes) with Long Spacers
In 1984 Rausch et al. [153] briefly reported the synthesis of elastomeric polyamides (52, Mn = 10 000–18 000) in high yields from 1,1′-bis(β-aminoethyl)ferrocene and diacid chlorides (Eq. (12-22a)). Also polyureas (53) were prepared when starting from the same ferrocene monomer and diisocyanates (Eq. (12-22b)) and polyesters and polyurethanes, when starting from 1,1′-bis(β-hydroxyethyl)ferrocene. However, these materials had much lower molecular weights and were characterized by scanning electron microscopy, X-ray and IR analyses. The employment of ferrocenes in which the active center is separated by at least two methylene units was crucial in order to reduce steric effects and the instability found previously in polymers of α-functionalized ferrocene due to the α-ferrocenyl carbonium ion stability [154]. O
O
(CH2)2NH2
O
(CH2)2NH
C
O R
C
Cl C R C Cl Fe
(22a)
Fe NH(CH2)2
(CH2)2NH2
n 52
R = Ph, (CH2)4, (CH2)8
O (CH2)2NH
OCN–R'–NCO
O
C NH-R'-NH C
(22b)
Fe NH(CH2)2
n 53 CH3
R' =
CH3
CH2
This technique was also used to prepare ferrocene-containing arylidene polyesters (54) in good yields, from dicarboxyl ferrocenes and organic diols, that were characterized by elemental analysis, IR spectroscopy, viscometry and WAXS [155]. These polymers were semicrystalline, soluble in polar solvents and conductivity studies showed an n-type semiconductor behavior (σ = 3 × 10–10 S cm–1 at room temperature) that followed a one-term Arrhenius-type equation with increasing conductivity over the range 300–500 K.
754
12
Metallocene-Based Polymers
O
O C
O
CH
Fe
O
CH
O
(CH2)x
C
n 54 x = 0, 1, 2
Wright et al. [156] have utilized Knoevenagel condensations to prepare new nonlinear optical (NLO) ferrocene-containing polymers for second harmonic generation (SHG) applications. The recent use of ferrocenes in this area is based on their large hyperpolarizability values [157] combined with their thermal and photochemical stability which make them desirable candidates for NLO materials. Thus, they synthesized a polyurethane copolymer (55) in moderate yields from a ferrocene derivative in which side reactions are minimized. The polymer was fully characterized and possessed a value of Mn = 7600 although no Tg was found using DSC [156]. Contrary to previous results with perfectly chromophore-oriented organometallic polymers [158], due to the random chromophore orientation in a head to tail sense, this material displays SHG activity at 150°C by corona poling with reasonable stability for the SHG signal.
CONH(CH2)6NHCO
OCH2 Fe CO2(CH2)2O CN
55 Zentel et al. [159] synthesized several ferrocene-containing copolyesters, such as 56, via Ti(O-iPr)4-catalyzed melt polycondensation. The polymers had molecular weights in the range 9000–30 000 (by capillary viscometry) and they were characterized by NMR, IR and DSC. When ferrocene diols are incorporated into polyesters there is an increase of the oxidation potential of ≈ 40–50 mV, presumably due to a charge transfer through space interaction between the electron poor ester group and ferrocene, since the ester group is separated from the ferrocene moiety by 4 to 6 σ-bonds. The authors carried out dynamic mechanical analysis which showed a strong influence of the ferrocene unit on the rheological properties, with an unusual, rubber-like behavior.
12.4
O
Me O
(CH2)2 O
C
755
Main Chain Poly(metallocenes) with Long Spacers
O
(CH2)2 O C
(CH2)3
Me
Me Fe
Me
(CH2)3 C O
n
56
A large variety of group 4 (Ti, Zr, Hf) metallocene-containing polyethers, polythioethers, polyesters, polyamines, polyoximes and polyamidoximes have been also reported [160, 161]. Their synthesis, based on the reaction of Cp2MCl2 with Lewis bases (diol, dithiol, etc.) (Fig. 12-5), was carried out either in aqueous solution or using interfacial condensation techniques, depending upon the solubility of the base. In 1977, Carraher [150] reviewed the preparation of these polymers; the materials are yellow to orange for titanium and white to light gray for zirconium and
Figure 12-5. Synthetic routes to group 4 metallocene-containing polymers.
756
12
Metallocene-Based Polymers
hafnium. Molecular weights ranged from oligomers to 106 although their solubility is limited. Some, such as the titanium polyethers, were insoluble and were characterized by IR, which showed characteristic bands for both metallocene and organic fragments. This allowed a determination of metal content [162], since in most cases elemental analysis for carbon were not consistent with expected values possibly due to high stability of thermal degradation products [163]. An interesting condensation reaction, in terms of applications of the resulting polymers, is that in which dyes were utilized as Lewis bases [164, 165]. Thus, soluble, high molecular weight (e.g. Mw = 5.3 × 105 for bromophenol blue) titanium polydyes were synthesized in yields depending on the reaction conditions and the dye used. Their claimed advantage was their potential permanent, nonleaching nature in comparison with monomeric dyes. These polydyes are fluorescent and, when used as coloring or doping agents in paper, cloth, paint and plastic, materials with this property were obtained. The synthesis of polymers containing cobaltocenium salts has also been studied, since this organometallic moiety is isoelectronic with ferrocene and resistant to strong oxidizing agents. Pittman reported the synthesis of polyesters (57 , R = (CH2)4, CH2ArCH2) [166] by solution polymerization techniques yielding low molecular weight (Mn = ca 3000) products in which exchange of PF6¯ by Cl¯ was observed complicating the characterization, as well as polyamides [167], which led to insoluble materials that were difficult to characterize. Later, Sheats [168] summarized the attempts to achieve soluble high molecular weight polymers by copolymerization of cobaltocenium 1,1′-dicarboxylic acid with organometallic monomers, such as titanocene or zirconocene (57; R = Cp2Ti, Cp2Zr). These materials were characterized by IR, TGA and DSC. The presence of charges along the chain was expected to increase the solubility in polar solvents, but the products showed limited solubility although molecular weights (by intrinsic viscosity in 2-chloroethanol) of approximately 80 000 were reported [169, 170]. O C O O O
Co+
C
R
PF6–
n 57
R = (CH2)4 , CH2ArCH2 , Cp2Ti , Cp2Zr Polyamides on the other hand were prepared by Neuse [171] by reacting aromatic diamines with 1,1′dicarboxylcobaltocenium chloride species in molten antimony trichloride at 150–175°C. The products were isolated as PF6¯ salts upon treatment
12.4
Main Chain Poly(metallocenes) with Long Spacers
757
with NH4PF6, and characterized by IR, NMR and elemental analysis, which confirmed the linear polyamide structure. Unfortunately, viscosity and VPO measurements proved inconclusive because of polydissociation in solution, and the authors based the polymeric nature on their film forming properties and the low concentration of the carboxyl end group with respect to the amide. Well-characterized polymers with ferrocenes joined by sulfur-containing moieties, other than disulfide (see Sec. 12.3.3), have been reported by Nuyken et al. [87] by introducing an alkylthio spacer between the ferrocene and the thiol groups. Thus, polyaddition of 1,1′-bis(2-mercapto-propylthio)ferrocene to divinyl sulfone afforded a soluble, orange, low molecular weight (Mn = ca 3000) material in 90% yield that was characterized by NMR and elemental analysis. These workers also reported recently a series of sulfur-containing ferrocene polymers [172] that were prepared from either 1,1′-bis(2-mercaptoethyl)ferrocene or 1,1′-dimercaptoferrocene by based-catalyzed polyaddition to activated diolefins containing electron-withdrawing substituents (e.g. 58) and by interfacial condensation with acid dichlorides for the synthesis of poly(thioesters) (e.g. 59). This afforded yellow, soluble polymers with low molecular weights (Mn < 5000 by GPC; PDI = 1.6–5) which were characterized by several analytical and spectroscopic methods.
O S
O
C (CH2)2 C
Fe S
n 59
Organosilicon Spacers Modification of silicones has been also an important area of interest and ferrocene has been introduced into organosilicon polymers including siloxanes and carbosilanes. The first approaches to siloxanes were made in the 1960s by Schaaf and Greber’s groups. They applied the principle of controlled hydrolysis of diethoxysilyl- [173] or chlorosiloxanylsilyl-ferrocene [174] derivatives, but the products were oils with low molecular weights due in part (in the former case) to intramolecular condensation side reactions [10].
758
12
Metallocene-Based Polymers
Pittman described the first examples of well-characterized, high molecular weight poly(ferrocenylsiloxanes) (60) from the condensation reaction in melt or solution, at 100–110°C of 1,1′-bis(dimethylaminodimethylsilyl)ferrocene with three different disilanols [175]. Molecular weights (Mn) were in the range 9000–20 000 (by GPC) with melt polymerization giving the highest molecular weights and narrowest distributions. Intramolecular cyclization is favored over intermolecular condensation when using diphenylsilanediol as noticed previously by Schaaf. DSC studies revealed for these soluble, fibrous polymers sharp melt transitions between 40–80°C. The polymer 60b appeared to be as thermally stable (up to 400°C) and as stable to hydrolysis (in THF/H2O) as the corresponding arylene polysiloxane without ferrocene moieties in the backbone. Ph Me Si Fe
Si O
R O
Ph R=
Me
n 60
a
Me
Me
Si
Si
Me
x Me
x = 1, b x = 2, c
Greber and Hallensleben have reported the use of Diels–Alder [176] and platinumcatalyzed [176, 177] polyaddition reactions of vinyl ferrocene derivatives to prepare ferrocene-containing poly(carbosilanes). The latter reactions were also used to prepare materials with additional functional groups leading to ferrocene and silicon-containing polyesters, polyamides and polyurethanes [174]. Molecular weights were in the range 3800–6000 and products were liquids or elastomers. The authors only carried out comparative studies in terms of thermal stability with respect to one another. Novel organometallic accordion-type [178] copolymers were prepared by Wright and Sigman in 1992 through the Knoevenagel polycondensation of a bis(ferrocene aldehyde)silane with several bis(cyanoacetates) [179]. The reaction produced soluble copolymers (61), consisting of biferrocenylsilane units with long organic spacers in isomeric (E/Z) mixtures. They were fully characterized by spectroscopic and analytical methods and possessed significantly high molecular weights (Mn = 9100–26 600).
12.4
Main Chain Poly(metallocenes) with Long Spacers
759
Recently a new approach towards ferrocene-silicon-containing polyamides such as 62 and 63 has been developed [180]. Solution or interfacial polycondensation reactions at room or low temperature yielded well-defined solid materials for which molecular weights (Mn) of 10 600–12 500 were determined by VPO. The iron centers are noninteracting, as shown by a single reversible oxidation wave, and no oxidative precipitation of the polymers were observed. Thus, they can be used for chemical modification of electrodes since vapour-deposited films on different electrodes exhibited the characteristic behavior of surface confined redox couples.
760
12
Metallocene-Based Polymers
Poly(metallosiloxanes) The synthesis of poly(metallosiloxanes) by controlled condensation of Cp2MR2 (M = Ti, Zr, Hf) with silanes or siloxanes has been investigated based on the improvements in heat resistance that might be realised with the introduction of the metal moiety in the polysiloxane. The results obtained in the area were critically reviewed by Neuse and Rosenberg [10] since partial cleavage of the metal–ring bond occurred in all the cases and products were poorly characterized with no data provided. Only when Cp2Ti(SH)2 and siloxanes were used as reagents in the presence of triethylamine were linear oligomers (64, n ≈ 6) described, although these were hydrolytically sensitive.
A new approach in this area has been reported by Lichtenhan and Haddad with the synthesis of novel polymeric zirconocene–silsesquioxanes (65) [181, 182]. The condensation of zirconocene derivatives with polyhedral silsesquioxanes led to amorphous polymers in high yields (≈ 90%) that were characterized by NMR spectroscopy and elemental analysis. These high molecular weight (Mn = 14 000) materials exhibited high thermal stability (Tdec = 474–515°C) and surprisingly 65a is stable to both air and hydrolysis.
Si HO
Cp
R
R
Si
O
O O OO O R Si R Si Si O Si R O R O O O Si
O
Zr Cp
O x
Si
R R
65a x = 1 65b x = 2
n
12.4
Main Chain Poly(metallocenes) with Long Spacers
761
12.4.2 Conjugated Spacers As poly(1,1′-ferrocenylene) exhibits significant electrical conductivity when doped the introduction of ferrocene units into the main chain of polymers with σ, σ–π or π-conjugation along the backbone is an interesting research objective. Conjugated organic polymers such as polyacetylene, poly(p-phenylene) or poly(p-phenylene vinylene) are typically insoluble and rigid materials that are difficult to process. Several attempts have been made to introduce skeletal ferrocene into units into the main chain of these materials [183–189] which has been shown to increase the flexibility and solubility of chains [190]. Unfortunately all the materials reported (e.g. 66) have low molecular weights (Mw < 5000) or poor solubility, and electrochemical studies showed weak interactions (∆E½ = 70–120 mV) between the iron centers, which are widely spaced [188].
I Fe H
n 66
Interesting carbosilanes containing organometallic moieties (e.g. 67) are illustrated by the oligomeric diacetylenes (Mw < 3000) isolated by Corriu et al. [191]. They were prepared in high yields following condensation routes from the di-Grignard reagent of diacetylene and 1,1′-bis(chlorodiorganosilyl)ferrocenes, and were characterized by NMR spectroscopy, IR and elemental analysis. R Si R'
Fe
R Si
C
R'
67 R = R' = Me, Ph R = Me, R' = Ph
C
C
C
n
762
12
Metallocene-Based Polymers
Also poly(ferrocenylhexasilane) (68) has been synthesized in an attempt to obtain polymers where ferrocenes are joined to σ-delocalized polysilane segments in a regular alternating structure [192]. The condensation of dilithioferrocene and 1,6dichlorodecamethylhexasilane gave a soluble, well-characterized polymer showing monomodal molecular weight distribution by GPC with maximum at Mw ≈ 3500. After doping, conductivity values (σ = 3 × 10–5 S cm–1) were three orders of magnitude higher than octadecasilane which seems to indicate that conjugation between ferrocenediyl and hexasilanediyl contributes to the conductivity, although the values are lower than for the corresponding poly(1,1′-ferrocenylenes) [68]. Me Si Me
Fe
6
n 68 Thiophene units have also been employed for bridging ferrocenes in order to obtain conjugated polymers [105]. The reaction of a di-zinc ferrocene derivative with dibromothiophene afforded a poly(ferrocenylene thienylene) (69) with moderate molecular weight (Mn ≈ 4500) which was characterized by NMR spectrometry and elemental analysis. Studies on doped materials with SbCl6¯, BF4¯ and TCNQ¯ counteranions showed the presence of weak antiferromagnetic interactions.
S S
S Fe
C6H13
n 69
12.5
763
Star and Dendritic Poly(metallocenes)
12.5 Star and Dendritic Poly(metallocenes) A particularly novel class of organometallic ferrocene-containing polymers are the dendrimers reported by Cuadrado and co-workers (Eq. (12-23))[193]. From tetraallylsilane as the initiator core, the researchers constructed novel silane dendrimers (i.e. 70) in which reactive SiCl end groups are present. The reaction with lithioferrocene or β-aminoethylferrocene provided the corresponding tetra- (71) and octa(72) ferrocenyl dendrimers that represent the first examples of well-defined and well-characterized macromolecules of this type. Cyclic voltammetric studies of these compounds [194] showed one reversible oxidation wave characteristic of independent, noninteracting redox centers. Also the use of these materials for modification of electrode surfaces was explored. The researchers found that platinum, glass and carbon-disk electrodes modified by electrodeposited films of these dendrimers are extremely durable and reproducible with no loss of electroactivity or material with their use in different electrolyte solutions or after standing for long periods in air. Me
Me
Si Cl Si
Si Me
Si
8 FcLi
Me
Si
Me Si Cl
Fe
Me
Si
(23a)
Me
Me
Si
4
Me
Fe
4
Me
70
71 Me Si NHCH2CH2
8 Fc–CH2CH2NH2 Si
Si Me
Fe
Me
(23b)
Me Si NHCH2CH2 Me
Fe
4
72
Astruc et al. [195] have reported the synthesis of star-shaped macromolecules with [FeCp(C6Me6)]+ as the starting core. The hexaferrocenylalkylation of the C6Me6 ligand yielded an heptanuclear complex (73) in which the ferrocene moieties are electrochemically equivalent with no influence of the central cationic unit on the oxidation process, leading to a fully reversible six-electron redox system.
764
12
Metallocene-Based Polymers
Fe
Fe
Fe Fe Fe
Fe
Fe
73
Related to this system is the decaferrocenyl ferrocene derivative synthesized recently by Jutzi et al. [196] by hydrosilylation of a decaallylferrocene with dimethylsilylferrocene. In these materials, the outer ferrocenyl moieties are also noninteracting. Cuadrado et al. have also reported ferrocenyl and permethylferrocenyl star polymers 74 containing a cyclotetrasiloxane as framework [197]. Their synthesis was achieved via hydrosilylation reactions using vinylferrocene monomers and tetramethylcyclotetrasiloxane. Soluble polymers with Mn = 1.9–2.3 × 104 (by VPO) were obtained in high yields and were characterized by NMR, elemental analysis, TGA and cyclic voltammetry. As with the ferrocenyl–dendritic structures, these materials exhibited only one oxidation wave and modification of platinum electrodes showed that, while the nonmethylated ferrocenyl polymer behaved as a surface-confined redox couple, the permethylated material possessed diffusional features that limit the charge transport through the film.
12.6
Summary
765
73 Fe
Me
Si
Me Me
Si
O O
Si
Fe
O
η5–C5H4
Si =
O Me
74a
η5–C5Me4 74b
Fe
Fe
12.6 Summary Metallocene-based polymers have come a long way since the first report of poly(vinylferrocene) in 1955. The development in the last few years of ring-opening polymerization routes and well-planned condensation strategies has provided access to a range of polymers with metallocene units in the main chain and also dendritic structures. This has opened the door for a thorough exploration of the physical characteristics of metallocene-based polymeric materials. Nevertheless, much additional work also needs to be done concerning the development of routes to polymers containing metallocenes other than ferrocene. The polymer chemistry of Group 4 metallocenes and cobaltocene, for example, is still very much undeveloped. Overall, the picture emerging is one of the existence of a range of novel organometallic materials with a variety of interesting and potentially useful properties. In the years ahead it will be interesting to see if any of these metallocenebased polymers fulfil the requirements of niche-markets and find their way into the real world.
766
12
Metallocene-Based Polymers
References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11]
[12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26]
[27] [28] [29] [30] [31] [32] [33] [34]
Wilkinson, G. J. Organomet. Chem. 1975, 100, 273. Ferrocenes; Togni, A.; Hayashi, T., Eds.; VCH Publishers: New York, 1995. Togni, A. In Ferrocenes; A. Togni and T. Hayashi, Eds.; VCH Publishers: New York, 1995; p. 433-469. Deschenaux, R.; Goodby, J. W. In Ferrocenes; A. Togni and T. Hayashi, Eds.; VCH Publishers: New York, 1995; p. 471-495. Inorganic Materials; Bruce, D. W.; O’Hare, D., Eds.; John Wiley & Sons: Toronto, 1992. Arimoto, F. S.; Haven, A. C. J. Am. Chem. Soc. 1955, 77, 6295. Manners, I. Angew. Chem. Int. Ed. Engl. 1996, 35, 1602. For an overview, see: Manners, I. Chem. Br., 1996, 32, 46. G. Odian, Principles of Polymerization, p. 573. 3rd Ed. 1991. John Wiley, New York and refs therein. Neuse, E. W.; Rosenberg, H. J. Macromol. Sci.-Revs. Macromol. Chem. 1970, C4, 1. The following abbreviations will used throughout the text: differential scanning calorimetry (DSC); glass transition temperature (Tg); gel permeation chromatography (GPC); weight average molecular weight (Mw); number average molecular weight (Mn); polydispersity index (PDI); vapor pressure osmometry (VPO); wide angle X-ray scattering (WAXS); ring-opening metathesis polymerization (ROMP); ring-opening polymerization (ROP); (η5-C5H4)(η5C5H5)Fe (Fc); (η5-C5H4)2Fe (fc); η5-C5H4 (Cp, CpH); η5-C5Me4 (CpMe); azobisisobutyronitrile (AIBN); cyclooctadiene (cod); 1,2-dimethoxyethane (DME); N,N,N,N’-tetramethylethylenediamine (TMEDA); tetracyanoethylene (TCNE); 7,7,8,8-tetracyanoquinodimethane (TCNQ). Pittmann, C. U., Jr.; Voges, R. L.; Elder, J. Polym. Lett. 1971, 9, 191. Pittmann, C. U., Jr. J. Polym. Sci. Part A-1 1971, 9, 3175. Lai, J. C.; Rounsefell, T.; Pittmann, C. U., Jr. J. Polym. Sci. Part A-1 1971, 9, 651. Sasaki, Y.; Walker, L. L.; Hurst, E. L.; Pittmann, C. U., Jr. J. Polym. Sci. Polym. Chem. Ed. 1973, 11, 1213. George, M. J.; Hayes, G. F. J. Polym. Sci. Polym. Chem. Ed. 1975, 13, 1049. George, M. H.; Hayes, G. F. J. Polym. Sci. Polym. Chem. Ed. 1976, 14, 475. Aso, C.; Kunitake, T.; Nakashima, T. Makromol. Chem. 1969, 124, 232. Nuyken, O.; Burkhardt, V.; Pöhlmann, T.; Herberhold, M. Makromol. Chem., Macromol. Symp. 1991, 44, 195. Smith, T. W.; Kuder, J. E.; Wychick, D. J. Polym. Sci. Polym. Chem. Ed. 1976, 14, 2433. Flanagan, J. B.; Margel, S.; Bard, A. J.; Anson, F. C. J. Am. Chem. Soc. 1978, 100, 4248. Cowan, D. O.; Park, J.; Pittman, C. U., Jr.; Sasaki, Y.; Mukherjee, T. K.; Diamond, N. A. J. Am. Chem. Soc. 1972, 94, 5110. Cowan, D. O.; Kaufman, F. J. Am. Chem. Soc. 1970, 92, 219. Bilow, N.; Landis, A. L.; Rosenberg, H. J. Polym. Sci. Part A-1 1969, 7, 2719. Pittmann, C. U., Jr.; Lai, J. C.; Vanderpool, D. P.; Good, M.; Prados, R. Macromolecules 1970, 3, 105. Pittmann, C. U., Jr.; Lai, J. C.; Vanderpool, D. P.; Good, M.; Prados, R. In Polymer Characterization: Interdisciplinary Approaches; C. D. Craver, Ed.; Plenum Publishing Corp.: New York, 1971; p. 97-124. Kittlesen, G. P.; White, H. S.; Wrighton, M. S. J. Am. Chem. Soc. 1985, 107, 7373. Wright, M. E.; Toplikar, E. G.; Kubin, R. F.; Seltzer, M. D. Macromolecules 1992, 25, 1838. Albagli, D.; Bazan, G.; Wrighton, M. S.; Schrock, R. R. J. Am. Chem. Soc. 1992, 114, 4150. Albagli, D.; Bazan, G. C.; Schrock, R. R.; Wrighton, M. S. J. Phys. Chem. 1993, 97, 10211. Nakashima, T.; Kunitake, T. Makromol. Chem. 1971, 157, 73. Pittmann, C. U., Jr.; Sasaki, Y.; Grube, P. L. J. Macromol. Sci., Chem. 1974, A8 (5), 923. Simionescu, C.; Lixandru, T.; Negulescu, I.; Mazilu, I.; Tataru, L. Makromol. Chem. 1973, 163, 59. Simionescu, C.; Lixandru, T.; Mazilu, I.; Tataru, L. Makromol. Chem. 1971, 147, 69.
References [35] [36]
[37]
[38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51]
[52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73]
767
Buchmeiser, M.; Schrock, R. R. Macromolecules 1995, 28, 6642. Sheats, J. E.; Hessel, F.; Tsarouhas, L.; Podejko, K. G.; Porter, T.; Kool, L. B.; Nolan, R. L., Jr. In Metal-containing Polymer Systems; J. E. Sheats, C. E. Carraher, Jr. and C. U. Pittmann, Jr., Eds.; Plenum Publishing Corp.: New York, 1985; p. 83-98. Sheats, J. E.; Hessel, F.; Tsarouhas, L.; Podejko, K. G.; Porter, T.; Kool, L. B.; Nolan, R. L. In New Monomers and Polymers; B. M. Culbertson and C. U. Pittmann, Jr., Eds.; Plenum Publishing Corp.: New York, 1984; p. 269-284. Wang, C.-L.; Mulchandani, A. Anal. Chem. 1995, 67, 1109. Allcock, H. R.; Riding, G. H.; Lavin, K. D. Macromolecules 1985, 18, 1340. Allcock, H. R.; Riding, G. H.; Lavin, K. D. Macromolecules 1987, 20, 6. Manners, I.; Riding, G. H.; Dodge, J. A.; Allcock, H. R. J. Am. Chem. Soc. 1989, 111, 3067. Allcock, H. R.; Dodge, J. A.; Manners, I.; Riding, G. H. J. Am. Chem Soc. 1991, 113, 9596. Allcock, H. R.; Dodge, J. A.; Manners, I.; Parvez, M.; Riding, G. H.; Visscher, K. B. Organometallics 1991, 10, 3098. Saraceno, R. A.; Riding, G. H.; Allcock, H. R.; Ewing, A. G. J. Am. Chem. Soc. 1988, 110, 7254. Wisian-Neilson, P.; Ford, R. R. Macromolecules 1989, 22, 72. Pannell, K. H.; Rozell, J. M.; Ziegler, J. M. Macromolecules 1988, 21, 276. Diaz, A.; Seymour, M.; Pannell, K. H.; Rozell, J. M. J. Electrochem. Soc. 1990, 137, 503. Hale, P. D.; Boguslavsky, L. I.; Inagaki, T.; Karan, H. I.; Lee, H. S.; Skotheim, T. A. Anal. Chem. 1991, 63, 677. Korshak, V. V.; Sosin, S. L.; Alekseeva, V. P. Akad. Nauk SSR 1960, 132, 360. Korshak, V. V.; Sosin, S. L.; Alekseeva, V. P. Vysokomol. Soedin. 1961, 3, 1332. Nesmenayov, A. N.; Korshak, V. V.; Voevodskii, N. S.; Kochetkova, S. L.; Sosin, S. L.; Materikova, R. B.; Bolotnikova, T. N.; M., C. V.; Bazhin, N. M. Dokl. Akad. Nauk SSSR 1961, 137, 1370. Rosenberg, H.; Neuse, E. W. J. Organomet. Chem. 1966, 6, 76. Nesmenayov, A. N.; Drozd, V. N.; Sazonova, V. A.; Romanenko, V. I.; Prokofiev, A. K.; Nikonova, L. A. Izv. Akad. Nauk SSSR 1963, 667. Roling, P. V.; Rausch, M. D. J. Org. Chem. 1972, 37, 729. Watanabe, N.; Motoyama, I.; Hata, K. Bull. Chem. Soc. Japan 1966, 39, 790. Spilner, I. J.; Pelligrini, J. P. J. Org. Chem. 1965, 30, 3800. Hedberg, F. L.; Rosenberg, H. Tetrahedron Lett. 1969, 4011. Rausch, M. D. J. Org. Chem. 1963, 28, 3337. Neuse, E. W.; Crossland, R. K. J. Organometal. Chem. 1967, 7, 344. Bradley, A.; Hammes, J. P. J. Electrochem. Soc. 1963, 110, 15. Izumi, T.; Kasahara, A. Bull. Chem. Soc. Japan 1975, 48, 1955. Bednarik, L.; Gohdes, R. C.; Neuse, E. W. Transition Met. Chem. 1977, 2, 212. Pittman, C. U., Jr.; Lai, J. C.; Vanderpool, D. P.; Good, M.; Prados, R. Macromolecules 1970, 3, 746. Pittman, C. U., Jr.; Sasaki, Y.; Mukherjee, T. K. Chem. Lett. 1975, 383. Cowan, D. O.; LeVanda, C. J. Am. Chem. Soc. 1972, 94, 9271. Mueller-Westerhoff, U. T.; Eilbracht, P. J. Am. Chem. Soc. 1972, 94, 9272. Neuse, E. W.; Bednarik, L. Transition Met. Chem. 1979, 4, 104. Sanechika, K.; Yakamoto, T.; Yamamoto, A. Polymer J. 1981, 13, 255. Yakamoto, T.; Sanechika, K.; Yakamoto, A.; Katado, M.; Motoyama, I.; Sano, H. Inorg. Chimica Acta 1983, 73, 75. Meier, H. Organic Semiconductors; Verlag Chemie: Weinheim FGR, 1974, 8. Oyama, N.; Takizawa, Y.; Matsuda, H.; Yakamoto, T.; Sanechika, K. Denkikagaku 1988, 56, 781. Seki, K.; Tanaka, H.; Ohta, T.; Sanechika, K.; Yakamoto, T. Chem. Phys. Lett. 1991, 178, 311. Hirao, T.; Kurashina, M.; Aramaki, K.; Nishihara, H. J. Chem. Soc., Dalton Trans. 1996, 2929.
768
12
Metallocene-Based Polymers
[74] Neuse, E. W.; Chris, G. J. J. Macromol. Sci., Chem. 1967, A-1, 371. [75] Pittman, C. U., Jr. J. Poly. Sci., Part A1, 1967, 5, 2927. [76] Withers, H. P.; Seyferth, D.; Fellman, J. D.; Garrou, P. E.; Martin, S. Organometallics 1982, 1, 1283. [77] Fellman, J. D.; Garrou, P. E.; Withers, H. P.; Seyferth, D.; Traficante, D. D. Organometallics 1983, 2, 818. [78] Rosenberg, H.; Rausch, M. D. U.S. Patent 3,060,215 1962 [79] Rosenberg, H. U.S. Patent 3,426,053 1969 [80] Park, J.; Seo, Y.; Cho, S.; Whang, D.; Kim, K.; Chang, T. J. Organomet. Chem. 1995, 489, 23. [81] Osborne, A. G.; Whitely, R. H.; Meads, R. E. J. Orgnomet. Chem. 1980, 193, 345. [82] Seyferth, D.; Withers, H. P. Organometallics 1982, 1, 1275. [83] Foucher, D. A. Ph. D. Thesis, University of Toronto, 1993. [84] Itoh, T.; Saitoh, H.; Iwatsuki, S. J. Polym. Sci. Polym. Chem. 1995, 33, 1589. [85] Nuyken, O.; Burkhardt, V.; Pöhlmann, T.; Herberhold, M. Makromol. Chem., Macromol. Symp. 1991, 44, 195. [86] Herberhold, M.; Brendel, H. D.; Nuyken, O.; Pöhlmann, T. J. Organomet. Chem. 1991, 413, 65. [87] Nuyken, O.; Pöhlmann, T.; Herberhold, M. Macromolecular Rep. 1992, A29, 211. [88] Giddings, S. A. Inorg. Chem. 1964, 3, 684. [89] Salzmann, J. J. Helv. Chim. Acta 1968, 51, 903. [90] Foucher, D. A.; Tang, B.-Z.; Manners, I. J. Am. Chem. Soc. 1992, 114, 6246. [91] Manners, I. Polyhedron 1996, 15, 4311. [92] Pudelski, J. K.; Manners, I. J. Am. Chem. Soc. 1995, 117, 7265. [93] Finckh, W.; Tang, B.-Z.; Foucher, D. A.; Zamble, D. B.; Ziembienski, R.; Lough, A.; Manners, I. Organometallics 1993, 12, 823. [94] Foucher, D. A.; Ziembinski, R.; Tang, B.-Z.; Macdonald, P. M.; Massey, J.; Jaeger, C. R.; Vancso, G. J.; Manners, I. Macromolecules 1993, 26, 2878. [95] Foucher, D. A.; Ziembinski, R.; Petersen, R.; Pudelski, J.; Edwards, M.; Ni, Y.; Massey, J.; Jaeger, C. R.; Vancso, G. J.; Manners, I. Macromolecules 1994, 27, 3992. [96] Foucher, D. A.; Honeyman, C. H.; Nelson, J. M.; Tang, B.-Z.; Manners, I. Angew. Chem. Int. Ed. Engl. 1993, 32, 1709. [97] Manners, I. J. Inorg. Organomet. Polym. 1993, 3, 185. [98] Nguyen, M. T.; Diaz, A. F.; Dement’ev, V. V.; Pannell, K. H. Chem. Mater. 1993, 5, 1389. [99] Manners, I. Adv. Mater. 1994, 6, 68. [100] Manners, I. Adv. Organomet. Chem. 1995, 37, 131. [101] Rulkens, R.; Lough, A. J.; Manners, I. J. Am. Chem. Soc. 1994, 116, 797. [102] Rulkens, R.; Lough, A. J.; Manners, I.; Lovelace, S. R.; Grant, C.; Geiger, W. E. J. Am. Chem. Soc. 1996, 118, 12683. [103] Foucher, D. A.; Ziembinski, R.; Rulkens, R.; Nelson, J. M.; Manners, I. In Inorganic and Organometallic Polymers II; P. A. Wisian-Nielson, H. R.; Wynne, K. J., Eds, American Chemical Society: Washington, DC, 1994, p. 442. [104] Nguyen, M. T.; Diaz, A. F.; Dement’ev, V. V.; Pannell, K. H. Chem. Mater. 1994, 6, 952. [105] Hmyene, M.; Yassar, A.; Escorne, M.; Percheron-Guegan, A.; Garnier, F. Adv. Mater. 1994, 6, 564. [106] Tang, B.-Z.; Petersen, R.; Foucher, D. A.; Lough, A. J.; Coombs, N.; Sodhi, R.; Manners, I. J. Chem. Soc., Chem. Commun. 1993, 523. [107] Petersen, R.; Foucher, D. A.; Tang, B.-Z.; Lough, A. J.; Raju, N. P.; Greedan, J. E.; Manners, I. Chem. Mater. 1995, 7, 2045. [108] Pudelski, J. K.; Rulkens, R.; Foucher, D. A.; Lough, A. J.; Macdonald, P. M.; Manners, I. Macromolecules 1995, 28, 7301. [109] Rasburn, J.; Petersen, R.; Jahr, T.; Rulkens, R.; Manners, I.; Vancso, G. J. Chem. Mater. 1995, 7, 871. [110] Barlow, S.; Rohl, A. L.; O’Hare, D. J. Chem. Soc., Chem. Commun. 1996, 257.
References
769
[111] Barlow, S.; Rohl, A. L.; Shi, S.; Freeman, C. M.; O’Hare, D. J. Am. Chem. Soc. 1996, 118, 7578. [112] Zechel, D. L.; Hultzsch, K. C.; Rulkens, R.; Balaishis, D.; Ni, Y.; Pudelski, J. K.; Lough, A. J.; Manners, I. Organometallics 1996, 15, 1972. [113] Liu, X.-H.; Bruce, D. W.; Manners, I. J. Chem. Soc, Chem. Commun. 1997, 289. [114] Foucher, D. A.; Manners, I. Makromol. Chem., Rapid Commun. 1993, 14, 63. [115] Peckham, T. J.; Massey, J. A.; Edwards, M.; Manners, I.; Foucher, D. A. Macromolecules 1996, 17, 2396. [116] Foucher, D. A.; Edwards, M.; Burrow, R. A.; Lough, A. J.; Manners, I. Organometallics 1994, 13, 4959. [117] Kapoor, R. N.; Crawford, G. M.; Mahmoud, J.; Dement’ev, V. V.; Nguyen, M. T.; Diaz, A. F.; Pannell, K. H. Organometallics 1995, 14, 4944. [118] Honeyman, C. H.; Foucher, D. A.; Dahmen, F. Y.; Rulkens, R.; Lough, A. J.; Manners, I. Organometallics 1995, 14, 5503. [119] Herberhold, M. Angew. Chem., Int. Ed. Engl. 1995, 34, 1837. [120] Pudelski, J. K.; Gates, D. P.; Rulkens, R.; Lough, A. J.; Manners, I. Angew. Chem., Int. Ed. Engl. 1995, 34, 1506. [121] Insoluble, uncharacterized poly(ferrocenylene sulfide) was reported earlier by a condensation route. See: Chien, J. C. W.; Gooding, R. D.; Lillya, C. P. Polym. Mater. Sci. Eng. 1983, 49, 107. [122] Rulkens, R.; Lough, A. J.; Manners, I. Angew. Chem., Int. Ed. Engl. 1996, 35, 1805. [123] Nelson, J. M.; Rengel, H.; Manners, I. J. Am. Chem. Soc. 1993, 115, 7035. [124] Nelson, J. M.; Rengel, H.; Petersen, R.; Nguyen, P.; Manners, I.; Raju, N. P.; Greedan, J. E.; Barlow, S.; O’Hare, D. Chem. Eur. J. 1997, 3, 573. [125] Nelson, J. M.; Lough, A. J.; Manners, I. Angew. Chem., Int. Ed. Engl. 1994, 33, 989. [126] Fossum, E.; Matyjaszewski, K.; Rulkens, R.; Manners, I. Macromolecules 1995, 28, 401. [127] Hultzsch, K. C.; Nelson, J. M.; Lough, A. J.; Manners, I. Organometallics 1995, 14, 5496. [128] Rulkens, R.; Ni, Y. Z.; Manners, I. J. Am. Chem. Soc. 1994, 116, 12121. [129] Ni, Y. Z.; Rulkens, R.; Manners, I. J. Am. Chem. Soc. 1996, 118, 4102. [130] Honeyman, C. H.; Peckham, T. J.; Massey, J. A.; Manners, I. J. Chem. Soc., Chem. Commun. 1996, 2589. [131] Ni, Y. Z.; Rulkens, R.; Pudelski, J. K.; Manners, I. Makromol. Chem. Rapid Communications 1995, 16, 637. [132] Reddy, N. P.; Yamasita, H.; Tanaka, M. J. Chem. Soc. Chem. Commun. 1995, 2263. [133] Sheridan, J. B.; Gómez Elipe, P.; Manners, I. Makromol. Chem. Rapid Commun. 1996, 17, 319. [134] Gómez-Elipe, P.; Macdonald, P. M.; Manners, I. Angew. Chem. Int. Ed. Engl. 1997, 36, 762. [135] Sheridan, J. B.; Lough, A. J.; Manners, I. Organometallics 1996, 15, 2195. [136] Sheridan, J. B.; Temple, K.; Lough, A. J.; Manners, I. J. Chem. Soc., Dalton Commun. 1997, 711. [137] Brandt, P. F.; Rauchfuss, T. B. J. Am. Chem. Soc. 1992, 114, 1926. [138] Galloway, C. P.; Rauchfuss, T. B. Angew. Chem. Int. Ed. Engl. 1993, 32, 1319. [139] Compton, D. L.; Rauchfuss, T. B. Organometallics 1994, 13, 4367. [140] Compton, D. L.; Brandt, P. F.; Rauchfuss, T. B.; Rosenbaum, D. F.; Zukoski, C. F. Chem. Mater. 1995, 7, 2342. [141] Scholl, H.; Sochaj, K. Electrochimica Acta 1991, 36, 689. [142] Compton, D. L.; Rauchfuss, T. B. Am. Chem. Soc. Polym. Prep. 1993, 34, 351. [143] Arnold, R.; Matchett, S. A.; Rosenblum, M. Organometallics 1988, 7, 2261. [144] Foxman, B. M.; Rosenblum, M.; Sokolov, V.; Khrushchova, N. Organometallics 1993, 12, 4805. [145] Ballauf, M. Angew. Chem., Int. Ed. Engl. 1989, 28, 253. [146] Foxman, B. M.; Gronbeck, D. A.; Rosenblum, M. J. Organomet. Chem. 1991, 413, 287. [147] Nugent, H. M.; Rosenblum, M.; Klemarczyk, P. J. Am. Chem. Soc. 1993, 115, 3848.
770
12
Metallocene-Based Polymers
[148] Rosenblum, M.; Nugent, H. M.; Jang, K.-S.; Labes, M. M.; Cahalane, W.; Klemarczyk, P.; Reiff, W. M. Macromolecules 1995, 28, 6330. [149] Katz, J. T.; Sudhakar, A.; Teasley, M. F.; Gilbert, A. M.; Geiger, W. E.; Robben, M. P.; Wuesnsh, M.; Ward, M. D. J. Am. Chem. Soc. 1993, 115, 3182. [150] Carraher, C. E., Jr. In Interfacial Synthesis; P. Millich and C. E. Carraher, Jr., Eds.; Marcel Dekker, Inc.: New York, 1977; Vol. II; p. 367-417. [151] Knobloch, F.; Rauscher, W. J. Polym. Sci., Part. A1 1961, 10, 651. [152] Okawara, M.; Takemoto, Y.; Kitaoka, H.; Haruki, E.; Imoto, E. Kogyo Kagazu Zasshi 1962, 65, 685. [153] Gonsalves, K.; Zhan-ru, L.; Rausch, M. D. J. Am. Chem. Soc. 1984, 106, 3862. [154] Pittman, C. U.; Rausch, M. D. Pure Appli. Chem. 1986, 58, 617. [155] Abd-Alla, M. M.; El-Zohry, M. F.; Aly, K. I.; Abd-El-Wahab, M. M. J. Appl. Polym. Sci. 1993, 47, 323. [156] Wright, M. E.; Toplikar, E. G.; Lackritz, H. S.; Kerney, J. T. Macromolecules 1994, 27, 3016. [157] Marder, S. R.; Perry, J. W.; Tiemann, B. G.; Schaefer, W. P. Organometallics 1991, 10, 1896. [158] Wright, M. E.; Toplikar, E. G. Macromolecules 1992, 25, 6050. [159] Wilbert, G.; Wiesemann, A.; Zentel, R. Makromol. Chem. 1995, 196, 3771. [160] Carraher, C. E., Jr. In Organometallic Polymers; C. E. Carraher, Jr., J. E. Sheats and C. U. Pittman, Jr., Eds.; Academic Press: New York, 1978; p. 79-86. [161] Carraher, C. E., Jr.; Pittman, C. U., Jr. In Metal-Containing Polymeric Systems; J. E. Sheats, C. E. Carraher, Jr. and C. U. Pittman, Jr., Eds.; Plenum Press: New York, 1985; p. 1-42. [162] Carraher, C. E., Jr.; Bahaj, S. T. Polym. J. 1974, 14, 42. [163] Carraher, C. E., Jr. et al. In Interfacial Synthesis; C. E. Carraher, Jr. and J. Preston, Eds.; Marcel Dekker, Inc.: New York, 1982; Vol. III; p. 77-91. [164] Carraher, C. E., Jr. et al. In Interfacial Synthesis; C. E. Carraher, Jr. and J. Preston, Eds.; Marcel Dekker, Inc.: New York, 1982; Vol. III; p. 93-105. [165] Carraher, C. E., Jr.; Schwarz, R. A.; Schroeder, J. A.; Schwarz, M.; Molloy, H. M. Org. Coat Plast. Chem. 1980, 43, 798. [166] Pittman, C. U., Jr.; Ayers, O. E.; McManus, S. P.; Sheats, J. E.; Whitten, C. Macromolecules 1971, 4, 360. [167] Pittman, C. U., Jr.; Ayers, O. E.; Suryanarayanan, B.; McManus, S. P.; Sheats, J. E. Makromol. Chem. 1974, 175, 1427. [168] Sheats, J. E. In Organometallic Polymers; C. E. Carraher, Jr., J. E. Sheats and C. U. Pittman, Jr., Eds.; Academic Press: New York, 1978; p. 87-94. [169] Carraher, C. E., Jr.; Sheats, J. E. Makromol. Chem. 1973, 166, 23. [170] Sheats, J. E.; Carraher, C. E., Jr.; Bruyer, D.; Cole, M. Am. Chem. Soc. Org. Coatings and Plastics Chem. 1974, 34, 474. [171] Neuse, J. E. In Organometallic Polymers; C. E. Carraher, Jr., J. E. Sheats and C. U. Pittman, Jr., Eds.; Academic Press: New York, 1978; p. 95-100. [172] Nuyken, O.; Pöhlmann, T.; Herberhold, M. Macromol. Chem. Phys. 1996, 197, 3343. [173] Schaaf, R. L.; Kan, P. T.; Lenk, C. T. J. Org. Chem. 1961, 26, 1790. [174] Greber, G.; Hallensleben, M. L. Makromol. Chem. 1966, 92, 137. [175] Patterson, W. J.; McManus, S. P.; Pittman, C. U., Jr. J. Polym. Sci. Polym. Chem. 1974, 12, 837. [176] Greber, G.; Hallensleben, M. L. Makromol. Chem. 1967, 104, 90. [177] Greber, G.; Hallensleben, M. L. Makromol. Chem. 1965, 83, 148. [178] Moore, J. S.; Stupp, S. I. Macromolecules 1990, 23, 65. [179] Wright, M. E.; Sigman, M. S. Macromolecules 1992, 25, 6055. [180] Casado, C. M.; Morán, M.; Losada, J.; Cuadrado, I. Inorg. Chem. 1995, 34, 1668. [181] Lichtenhan, J. D. Comments Inorg. Chem. 1995, 17, 115. [182] Haddad, T. S.; Lichtenhan, J. D. J. Inorg. Organomet. Polym. 1995, 5, 237. [183] Kasahara, A.; Izumi, T. Chem. Lett. 1978, 21. [184] Gamble, A. S.; Pattin, J. T.; Boncella, J. M. Makromol. Chem. Rapid Commun. 1992, 13, 109.
References
771
[185] Bayer, R.; Pöhlman, T.; Nuyken, O. Makromol. Chem. Rapid Commun. 1993, 14, 359. [186] Knapp, R.; Rehahn, M. Makromol. Chem. Rapid Commun. 1993, 14, 451. [187] Ingham, S. L.; Khan, M. S.; Lewis, J.; Long, N. J.; Raithby, P. R. J. Organomet. Chem. 1994, 470, 153. [188] Bochmann, M.; Lu, J.; Cannon, R. D. J. Organomet. Chem. 1996, 518, 97. [189] Stanton, C. E.; Lee, T. R.; Grubbs, R. H.; Lewis, N. S.; Pudelski, J. K.; Callstrom, M. R.; Erickson, M. S.; McLaughlin, M. L. Macromolecules 1995, 28, 8713. [190] Chien, J. C. W.; Gooding, R. D.; Lillya, C. P. Polym. Mater. Sci. Eng. 1983, 49, 107. [191] Corriu, R. J. P.; Devylder, N.; Guérin, C.; Henner, B.; Jean, A. Organometallics 1994, 13, 3194. [192] Tanaka, M.; Hayashi, T. Bull. Chem. Soc. Jpn. 1993, 66, 334. [193] Alonso, B.; Cuadrado, I.; Morán, M.; Losada, J. J. Chem. Soc., Chem. Commun. 1994, 2575. [194] Alonso, B.; Morán, M.; Casado, C. M.; Lobete, F.; Losada, J.; Cuadrado, I. Chem. Mater. 1995, 7, 1440. [195] Fillaut, J.-L.; Linares, J.; Astruc, D. Angew. Chem. Int. Ed. Engl. 1994, 33, 2460. [196] Jutzi, P.; Batz, C.; Neumann, B.; Stammler, H.-G. Angew. Chem., Int. Ed. Engl. 1996, 35, 2118. [197] Casado, C. M.; Cuadrado, I.; Morán, M.; Alonso, B.; Lobete, F.; Losada, J. Organometallics 1995, 14, 2618.
772
12
Metallocene-Based Polymers
Index
773
Index
ab initio calculations 178 f absence of d orbitals — impact on electronic structure and reactivity 22 acetylene 133f, 139 acetylide carbon–carbon coupling 93 acetylides 92 O-acetylmandelic acid 627 achiral cyclopentadienes — annelated 477 — annelated disubstituted 476 — trisubstituted 479 acrolein 642 activation of carbon monoxide 101 activation reactions 100 acyl complex 71 acyclic substrates 668 addition of allylmetals to aldehydes 632 addition of Cp anions 8 additional selectivity issues 650 π-adducts of lanthanocenes(II) 61 α-agostic interactions 565 agostic interactions 87, 88, 117, 126, 128, 207, 297 β-agostic 286 air-sensitivity 103 alkali metal cyclopentadienides 5 alkaline earth sandwich metallocenes 29 — Cp*2Ba 30 — Cp*2Ca 30 — Cp'2Mg 29 — Cp*2Sr 30 — nonparallel structure 30 — parallel sandwich arrangement 29 alkene-zirconocene complexes 243, 291, 295 alkene-zirconocenes 309 1-alkene-ZrCp2 300 alkenyl–alkenyl coupling reactions 291 alkenyl complexes 393 alkenylzirconocene chlorides 274 — conversion 272 alkoxide and aryloxide complexes 71 alkoxide elimination 651 alkoxides 70, 73
— bulky 228 alkoxo ligand — see alkoxo-cyclopentadienyl complexes alkoxo-cyclopentadienyl complexes 416, 443ff — of fluorenyl 444 n-alkyl product 650 alkyl 111, 124, 126, 135 — scandium 113, 115f, 124, 134 — yttrium 116f, 123, 132 alkyl [VCp2R] 342 alkyl- and alkenylzirconocenes 266 alkyl complexes 396 alkyl halide derivatives of niobium 370 alkyl halide derivatives of tantalum 370 σ-alkyl and hydride ligands 102 alkyl transition metal complexes 424ff alkyl vanadium(III) compounds 343 alkylaluminum halide — cocatalysts 550 alkylidene transition metal complex 427 alkylidenes 396 alkyne-zirconocenes 242 alkyne–ZrCp2 complexes 288, 295 alkynyl–alkynyl coupling reactions 291 allenylzirconiums 254 allyl [VCp2(η1-allyl)] 342 allyl absolute configuration 701 η3-allyl complexes 99, 102 allyl- and crotyltitanocene reagents 633 allyl- and crotyltitanocenes 632 allylic amines 664 allylic ethers 664 A(1.2) allylic strain 673 allylic substitution 701 allylzirconiums 254 allylzirconocenes — reactions with aldehydes 264 alternative enantioselective methods for C=N reduction 671 alternative methods for ketone reduction 676 aluminum 90 aluminum(I) 33 alumoxanes 553 amides 649
774
Index
amines 667 π-aminobenzoic acid 629 amido ligand 416 — see also amido-cyclopentadienyl complexes amido-cyclopentadienyl complexes 417ff — active species in catalysis 447 — agostic interactions 425, 435 — alkali- and alkaline-earth metal complexes 419 — alkylidene transition metal complex 427 — amido ligand, reactivity of 431 — benzyl complexes 425ff — benzyne transition metal complex 427 — boron-functionalized amido-cyclopentadienyl ligands 440 — bridging groups 417ff — carbon dioxide, reaction 430 — coordination of donor groups to transition metals 436ff — cyclopentadienyl moiety, choice 418 — diene complexes of titanium 428 — donor ligands joined to amido-cyclopentadienyl ligands 436ff — ethylene transition metal complex 428 — fluorenyl-based 418f, 421, 448 — hydrides and alkyls 424ff — indenyl-based 418, 420, 422 — ligand complexation 419ff — metathesis method (transition metal chlorides) 419ff — reaction with homoleptic transition metal amides 421ff — base-assisted complexation of neutral ligand 420 — ligand synthesis 418ff — planar chiral complexes 437 — polydentate amido-cyclopentadienyl ligands 435ff — reactivity 429ff — diamido complexes 429 — bridging Si-N bond 429ff — towards carbon dioxide 430 — amido ligand 431 — sandwich complexes 440ff — Si-N bond, reactivity 429ff — solvent adducts 421, 425 — structure 431ff — synthesis of transition metal complexes 417 ff, 423ff
— tridentate amido-cyclopentadienyl ligands 435ff — two amido-cyclopentadienyl bonded to one transition metal 440ff — X-ray crystal structures 431ff amination 701 amine-linked bis(cyclopentadienyl) ligands 65 amino-cyclopentadienyl complexes 416 6-aminofulvene 715 anion effect 706 anionic phosphoylide complexes 82 anionic thallate sandwich metallocene 34 annelated cyclopentadienes, chiral pool 465 ansa-bis(cyclopentadienyl)titanium dichloride complex, chiral 537 ansa-bis(fluorenyl)zirconium dichlorides 530 ansa-bis(indenes) 528 — ferrocene-bridged 528 — 1,2-phenylene-bridged 528 — silyl-containing bridges 528 ansa-bis(indenyl)metal complex 527 ansa-bis(indenyl)metal dichlorides 529 ansa-bis(3-substituted cyclopentadienyl)metal dichlorides 523 ansa-lanthanocenes 61, 102 ansa-metallocene dichlorides — trisubstituted cyclopentadienes 525 ansa-metallocene-catalyzed hydrogenation of alkene 660 ansa-metallocene complexes 589 — non-C2 symmetrical or unsymmetrical 589 ansa-metallocenes 415, 549, 552, 562, 625 — containing fluorenyl 531 — containing indenyl with cyclopentadienyl 530 — ligands 531 — with cyclopentadienyl 531 ansa-niobium complexes 369 anti and syn diastereomeric complexes 650 anti-carbometallation 276 aqua complexes 173 aromatic olefins 664 aromatic substituent 653 arylmethyl ketones 675 ArZrCp2Cl, preparation 248 asymmetric allylmetal additions to carbonyls 634 asymmetric carbomagnesation reactions 645
1
Index
asymmetric cyclopolymerization of dienes 640 asymmetric diels–alder cycloaddition reactions 641 asymmetric epoxidation 656 asymmetric homogeneous catalysis 685 asymmetric hydrogenation — of enamines 672 — of imines 667 — of trisubstituted alkenes 663 asymmetric ketone reduction 674 asymmetric propylene polymerization 638 atactic 638 ate complexes 57, 66, 83, 84, 87 azazirconacycle 636 azazirconacyclopentanes 636 π-backbonding 700 base-free lanthanocene halides 64 bent metallocenes 58, 60 benz[e]indene 488 benzyl transition metal complexes 425ff benzyne transition metal complex 427 benzyne-zirconocenes 242, 295 Bercaw, J.E. 122 beryllocene 29 — asymmetric arrangement 29 — fluxional behavior 29 bicyclization — of dienes and diynes 293 — of enynes 283 bicyclo[3.3.0]octane 293 bicyclooctane-annelated indene 490 bidentate ligands 415ff bifunctional cyclopentadienyl ligands 416f bifunctional ligands 87 bimetallic complexes 66, 81 bimetallic hydrocarbyls 90 bimetallic ytterbium complex 90 binaphtholate salts 628 biotin 696 bis(η3-allyl) complex 92 bis(2-aryltetrahydroindenyl)metal dichlorides 518 bis(3-tert-butylcyclopentadienes) 494 bis(t-butylcyclopentadienyl)titanium complex — ethylene-bridged 523 bis(cyclopentadienes) 491, 493 — bicyclooctane-annelated 497 — bridged by chiral groups 509
775
— bridged by stereogenic group 509 — chiral bridged 496 — cyclopentane-1,3-diyl-bridged 494 — ethylene-bridged 494 — having chiral substituents 496 — isopropylidene-bridged 491 — nonprochiral doubly bridged 493 — preparation of prochiral 494 — prochiral bridged 493 — prochiral doubly bridged 496 — substituted nonprochiral 492 — synthesis of chiral 497 — synthesis of prochiral 495 — tetramethylethylene-bridged 494 bis(cyclopentadienes) metal complexes — biaryl-bridged 536 bis(cyclopentadienyl) — biphenyl-bridged 535 — chiral doubly bridged 538 — dimethylsilyl-bridged 492 — silyl bridged chiral 497 — tetramethylethylene-bridged 492 bis(cyclopentadienyl) derivatives 385 bis(cyclopentadienyl)lanthanide(III) complexes 59 bis(cyclopentadienyl)lanthanide hydrocarbyls 63 bis(cyclopentadienyl)metal complexes — achiral 513 — achiral bridged 522 — 2,3-butane-diyl bridging group 537 — chiral having equivalent p-faces 514 — chiral silyl-bridged 537 — having diastereotopic faces 517 — having enantiotopic ligands 516 bis(cyclopentadienyl)metal dichlorides — chiral 515, 517 bis(cyclopentadienyl) metal(IV) halides 350 bis(cyclopentadienyl) niobium complexes 344 bis(cyclopentadienyl) niobium dihalides 368 bis(cyclopentadienyl) tantalum(III) complexes 344 bis(cyclopentadienyl) tantalum(IV) dihalides 368 bis(cyclopentadienyl)titanium complex — chiral doubly bridged 538 bis(cyclopentadienyl)titanium dichloride — irradiation of 513 — mixed 514
776
Index
bis(cyclopentadienyl)scandium silyl and germyl complexes 83 bis(cyclopentadienyl) vanadium(III) derivatives 340 bis(cyclopentadienyl) vanadium(III) and (IV) halides 341 bis(cyclopentadienyl)ytterbium alkoxides 70 bis(cyclopentadienyl)yttrium hydride complexes 96 bis(cyclopentadienyl)yttrium hydrides 95f. bis(cyclopentadienyl)zirconium dichlorides — doubly bridged 526 bis(fluorene) — biphenyl-bridged 502 — bridged 502 — chiral bridged 512 — dimethylsilylene-bridged 502 — ethylene-bridged 502 bis(fluorenyl)zirconium complexes, achiral 530 bis(fluorenyl)zirconium dichlorides 521 — unbridged 520 bis(indene) — achiral complexes 518 — 7,7-bridged 501 — bridged by stereogenic groups 532 — chiral bridged 511 — chiral bridged at the 1-position 507 — containing chiral hydrocarbon bridges 508 — 1,2-cyclohexane-bridged 534 — 1,4-cyclohexane-bridged 534 — dimethylsilylene-bridged 498 — doubly bridged 501, 511, 529 — ferrocene-bridged 499 — 1,2-phenyl-bridged 499 — silyl-bridged 498 — threitol-bridged 534 bis(indene) ligand 626 bis(2-indenes), biaryl-bridging groups 510 bis(indenyl)metal dichloride complexes — preparation of achiral 519 bis(indenyl)niobium derivatives 352 bis(indenyl)titanium dichlorides — di(menthoxy)silyl-bridged 538 bis(η5-indenyl)zirconium 518 bis(indenyl)zirconium dichloride — chiral unbridged 520 — permethylated 518 bis(indenyl)zirconium dichloride complexes
— substituted 529 bis(3-isopropylcyclopentadienes) 494 bismuth 80 bis(silyl) compounds 362 bis(tetrahydroindene) — binaphthyl-bridged 537 — synthesis of 495 bis(tetrahydroindenyl)titanium — dimethylbinaphthyl-bridged 533 bis(tetrahydroindenyl)titanium complex — bitetralin-bridged 533 bis(tetrahydroindenyl)zirconium dichloride complex — isomenthyl 519 — menthyl 519 — neoisomenthyl 519 — neoisopinocamphyl 519 — neomenthyl 519 bis(trimethylsilyl)amide ligands 77 1,3-bis(trimethylsilyl)cyclopentadienyl 63 bis(trimethylsilyl)methyl ligand 86 bite-angle 69 σ-bond metathesis 125f, 134, 148, 294, 296, 298f, 309 borohydride complexes 96 borohydride derivatives 102 boron-fuctionalized amido-cyclopentadienyl ligands 440 boryl complex 362 branched 650 bridged alkoxo-cyclopentadienyl complexes — see alkoxo-cyclopentadienyl complexes bridged amido-cyclopentadienyl complexes — see amido-cyclopentadienyl complexes bridged ligands — with nonstereogenic bridging groups 491 — with stereogenic bridging groups 506 bridging ligand, Cp*Al 36 Brintzinger 626 Brintzinger’s complex 527 bromides 67 Brominolysis 262 i-Bu3Al-Cp2ZrCl2 279 t-Bu(i-Bu)ZrCp2 258 Buchwald 626, 629, 659, 663, 667, 674 Buchwald and Lee 672 Buchwald protocol 285, 286 n-BuMgCl 649 η2-butadienyl complex 91 1-butenezirconocene 308
Index
S,S-butane-2,3-diyl-bridged bis(3-methylcyclopent 468 (t-butylcyclopentadienyl)(t-butylindenyl)silane 503 t-butylcyclopentadienyl 57, 95, 470 t-butylcyclopentadienyl lanthanide chlorides 63 2-tert-butylindene 485 n-Bu2ZrCp2 258, 295 C–Al bond addition 279 C2-symmetry 695 camphor-cyclopentadienyl 465, 466 ε-caprolactone polymerization, with amidocyclopentadienyl transition metal catalysts 441 carbaborane complexes 227 carbene complexes 167 ff, 185 ff — Fischer-Carbene complexes 193 ff — rotational barriers 161 — see alkylidene complexes carbenes of the arduengo-type (imidazol-2ylidenes 91 carboalumination — Zr-catalyzed of alkynes 242, 310 carbomagnesiation — Zr-catalyzed 300, 303 carbomagnezation reactions 217 carbometallation — Zr-catalyzed 304 carbon dioxide 136, 138, 140, 142, 148 carbon monoxide 72, 83, 99, 102, 136, 139, 142, 148 carbon–heteroatom bond formation reactions 276 carbonate complexes 364 carbonyl complexes 699 carbonylation 268 carbonylolefination 184, 220 ff carbozincation 276 carbozirconation 270, 304 catalysis, with amido-cyclopentadienyl transition metal catalysts 445ff catalytic Diels–Alder cycloaddition 643 catalytic kinetic resolution 652 catalytic hydrogenation of RZrCp2Cl 262 catecholborane 698 cationic metallocenes 558 cationic sandwich metallocenes, boron and aluminium 35
777
cationic zirconocene complexes 635 C-F bond activation 217 C-H activation 85, 88, 300 C-H activation reactions 102 C-H bond activation 187, 207 chain propagation 639 chair-like transition state 633 characterization features 8 — melting point trends 9 — NMR coupling 9 — solubility 10 — volatility 10 — X-ray crystallography 8 chelate complexes 415ff chelate stabilization 58 chelating alkoxo-cyclopentadienyl complexes — see alkoxo-cyclopentadienyl complexes chelating amido-cyclopentadienyl complexes — see amido-cyclopentadienyl complexes chemical vapor deposition 416 chiracene 534 chiracene ligand 468 chiral s-alkyl complexes 88 chiral ansa-metallocene complexes 678 chiral auxiliaries 589 chiral chelating cyclopentadienyl ligands 68 chiral chloro complexes 77 chiral cyclopentadienes 67 — annelated disubstituted 477ff — donor-substituted 474 — monosubstituted 471, 474 — by palladium-catalyzed alkylation 474 — pentasubstituted 483f — preparation of 473 — tetrasubstituted 482 — trisubstituted 481 — annelated 481 chiral cyclopentadienyl ligands 102 chiral indenes — preparation of 490 chiral ligands 685 — doubly bridged 511 chiral unsaturated pyrans or furans 651 chirality transfer 678 (C5H5)2Ln(II) derivatives 56 (C5H5)2LnCl 62 (C5H5)2LnCl derivatives 62 (C5H5)2LnR derivatives 81 chlorinolysis 262 chloro(hydrido)alkenezirconocenes 287
778
Index
chloro precursors 88 chromene 657 chromium [Cp2MX2] complexes 348 chromium [MCp2L] complexes 333 chromium ionic metallocenes [MCp2]+, [MCp2]- 330 chromium metallocenes [MCp2] 326 cis and trans isomers 667 clusters 16, 34 — Cp*Al 16, 34 — Cp*Ga 16, 34 — Cp*In 16, 34 — Cp2TTl 16 (C5Me5)2Ln derivatives 71 (C5Me5)2LnX 66 (C5Me5)2LuH 101 (C5Me5)2Sm 77, 91 (C5Me5)2Sm hydrocarbyls 84 (C5Me5)2Sm(THF)2 60, 71, 72, 76, 77, 92, 93 C–O bond 648 CO insertion 71, 83, 660 Collins 632, 633, 641 complexation 246 σ-complexes 4 complexes, chiral 519 π-complexes of zirconocene 285f., 287f., 299 comproportionation 7 conformation 691 conformational preferences and electronic factors in heterocycle resolution 653 conjugate additions 275 conjugated diene-zirconocenes 242 conjugated diene–ZrCp2 complexes 295 constitutional isomers 655 constrained geometry catalyst 547 ~50% conversion 651 coordination gap aperture 578 coordination geometry 687 coordination plane 695 copolymerization 603 — of cyclic olefins 605 copolymerization, with amido-cyclopentadienyl transition metal catalysts 447, 448 Corey 676 Cossee–Arlman insertion mechanism 564 Cossee–Arlman metathesis mechanism 564 CpnSUBSTITUTION 5 Cp' ligands 5 — general abbreviation 5 Cp'centroid–E distances 45
Cp'-element π-interaction 10 Cp-element orbital interactions 18 [Cp*Ge]+, [Cp*Sn]+, [Cp*Pb]+ 39 Cp2HfCl2 257 Cp*2Si 36 — Cp*2Ge(g) 37 — Cp*2Pb(s) 37 — Cp*2Sn(s) 37 Cp2Zr(II) complexes 258, 308 Cp2Zr(II) π-complexes 259 Cp2Zr(II) compounds 243, 246, 283 Cp2Zr(IV) compounds 243, 246 Cp2Zr compounds, preparation 246 Cp2Zr derivatives 255, 310 — σ-bond metathesis 310 — oxidative addition 310 — reaction 255 — reductive elimination 310 Cp2Zr(II) reagents preparation 257 Cp2Zr–stilbene 259 Cp2ZrBr2 from Cp2ZrCl2 by transhalogenation 248 Cp2Zr(Bu-n) 2 245, 286 CpZr(Bu-n)3 245 Cp2Zr–carbene complexes 310 Cp2Zr(CH2=CH2)(PMe3) 298 Cp2ZrCl2 257 — monoalkylation 248 — reaction with BI3 248 — treatment 245 — with n-BuLi 248 — with t-BuMgCl, LiAlH4, Red-Al 252 — with 2 equivalents of n-BuLi 245 — with KI 248 — with metal hydrides 251 CpZrCl3 245 Cp2Zr(CO)2 257, 287 Cp2ZrH2 252 Cp2Zr(H)Cl — preparation 242 Cp2ZrI2 from Cp2ZrCl 2 by transhalogenation 248 Cp2Zr(PMe3)2 257, 287 Cp2ZrX2 — preparation 242 — reduction 257 crystal structure 626 Curtin–Hammett regime 708 cyanation of alkenes 269 µ-cyanomethyl complexes 88
Index
cyclacene 534 cyclic allylic ethers 656 cyclic imines 668 cyclobutenation 276 cyclopentadiene 642 cyclopentenones 659 crystal packing forces 44 cyclic fluxionality 10 cycloaddition reactions — [2+1] addition 187 f — [2+2] addition 188 f, 191 ff, 219 ff, 224 (1-cyclohexylethyl)cyclopentadienyl 514 1-cyclohexylindenyl 463 cycloisomerization 710 cyclopentadiene 469 — annelated 477 — bridged 504 — bridged substituted 505 — C2-symmetric fused 478 — C2-symmetric tetrasubstituted, diannelated 515 — chiral silicon-bridged ligands 511 — disubstituted 474 — fulvene route to disubstituted 476 — monosubstituted achiral 469f — pentasubstituted 482 — pinanyl 517 — spiro-annelated 509 — steroidal 471 — tetrasubstituted 481 — trisubstituted 480 — unsubstituted 491 — verbenone-derived 517 cyclopentadienyl complexes 415ff cyclopentadienyl fluorenyl 504 (cyclopentadienyl)(fluorenyl)metal dichlorides — unbridged 522 cyclopentadienyl indenyl 503 (cyclopentadienyl)(indenyl)metal dichlorides — unbridged 521 cyclopentadienyl ligands 161 — Cp-activation 209 f — Cp-equivalents 226 — ESCA studies 161 — substituents 161 µ–π–cyclopentadienyl molecular complexes 13 — [Cp3E]2 13 — [Cp5Pb2]- 13 — [Cp9Pb4]- 13 — [E2Cp3]- 13
779
µ–π–cyclopentadienyl super-sandwich polymers 14 — BaCp*2 14 — Cp2Pb 14 — KCp 14 — LiCp 14 — NaCp 14 cyclopentenols, dehydration 471 cyclopolymerization 607 — with amido-cyclopentadienyl transition metal catalysts 448 C–Zr bond 648 decamethylaluminocenium cation 36 decamethyllanthanocene 59, 93f decamethyllanthanocene alkoxides and thiolates 73 decamethyllanthanocene fluorides 65 decamethyllanthanocene halides 66 decamethyllanthanocene hydrides 97, 102 decamethyllanthanocene hydrocarbyls 86, 88 decamethylsamarocene 59, 91f. decamethylyttrocene hydrocarbyls 89 decaphenylstannocene 38 decomposition 7 β-dehydrozirconation 296 dendritic poly(metallocenes) 763f deuterolysis 262 Dewar–Chatt–Duncanson model 255 diacetylene complex 93 dialkyl halide derivatives of niobium 370 dialkyl halide derivatives of tantalum 370 dialkylzirconocenes 242, 269, 286 diastereomerically pure complexes 69 diastereoselective halogenation 690 diastereoselective uncatalyzed alkylation reaction 654 diastereoselectively 686 diastereotopic 457 diastereotopic faces 464, 468 diastereotopic π-faces 469 diazadiene complexes 78 diazadiene radical anions 77 diazadienes 77 1,2-di-tert-butylcyclopentadiene 475 1,3-di-tert-butylcyclopentadiene 475, 479 1,3-di-t-butylcyclopentadienyl ligand 57, 95 dicarboranes 146ff dichlorobis(1-neomenthylindenyl)zirconium 280
780
Index
5,6-dichloroindenes 487 Diels-Alder catalysis 173, 217 diene bicyclization reactions 295 diene complexes of titanium 428 dienezirconocenes 264 dienophile 642 diethylzinc 712 difference in the reactivity pattern of pyrans and furans 655 4,7-difluoroindene 487 dihalide compounds [MCp2Cl2] 359 dihydride complexes 358 2,5-dihydrofuran 700 dimers 15 — Cp5BzIn 15 — Cp5BzTl 15 — involving halide bridges 15 1,1-dimetalloalkanes and -alkenes 310 (dimethylaminoethyl)cyclopentadienyl ligand 65 (α,α-dimethylbenzyl)cyclopentadienyl 470 1,2-dimethylcyclopentadiene 474, 475 1,3-dimethylcyclopentadienyl 458 trans-(1R,3R)-1,3-dimethylcyclopentane 641 6,6-dimethylfulvene — organometal addition 470 — reduction 470 4,7-dimethylindene 487 dimethyl(menthylCp)-(tetramethylCp)silane 467 dinitrogen complex of an f-element 79 diorganylzirconocenes 269, 285 diphenylfulvene 470 1,3-disubstituted cyclopentadienes 476 2,3-diphosphabutadiene 93 diphosphines 688 disproportionation 7 (diphenylmethyl)cyclopentadienyl 470 dithiocarbamate complex 366 donor ligands connected to amido-cyclopentadienyl ligands 435ff double bond migration, multipositional 295 dow catalysts 547 dπ-pπ interactions 160 [DpScH]2 115, 128, 136 Dzhemilev ethylmagnesiation 299, 301 — of alkenes 246 — catalyzed by Cp2ZrCl2 242 E and Z alkenes 664
(EBI)Zr(η2-pyrid-2-yl) 635 (EBTHI) metallocene dichlorides 625 (EBTHI)Ti-binol 664, 667 (R)-(EBTHI)Ti-binol 674 (S)-(EBTHI)Ti-binol 673 (EBTHI)Ti-catalyzed hydrogenation 665 (EBTHI)Ti-H 673 (R)-(EBTHI)Ti-H 674 (S)-(EBTHI)Ti-H 665, 670 (EBTHI)Zr-binol 646 (R)-(EBTHI)Zr-binol 662 (EBTHI)ZrCl2 646 (S)-(EBTHI)ZrMe2 630 (EBTHI)Zr(O-tBu)_THF 641 (EBTHI)Zr(η2-pyrid-2-yl) 635 (EBTHI)ZrX2/MAO 661 14 electron sandwich metallocenes 20 electronic control 20 electronic effects 637, 675, 690 electronic properties, titanocenes 158 element–element interactions 34 — In–In and Tl–Tl bonds 34 elevated temperatures 652 enantioselective carboalumination 280 — Zr-catalyzed 303 enantioselective C–C and C–H bond formation 625 enantioselective deuteration 662 enantioselective Diels–Alder cycloaddition 644 enantioselective hydrogenation of a,b-unsaturated 666 enantioselective hydrogenation and deuteration 662 enantioselective ortho-lithiation 716 enantioselective polymerization 594 enantiotopic 457 enantiotopic faces 461, 463 enantiotopic ligands 462 endo:exo isomers 642 energetic distinction 4, 35 — solid state for Cp3M3Al 35 enyne bicyclization 293 — Cp2Zr-promoted 242 EPDM terpolymers 607 epilupidine 144 equatorial bonding plane — bent titanocenes 159 — dπ-pπ interactions 160 equilibrium 631
Index
equipment required to prepare EBTHI mtallocenes 678 Erker–Buchwald protocol 259 Erker’s nonbridged chiral zirconocene 658 ethanobis(cyclopentadienes), prochiral 494 ethanobis(indenyl)metal dichlorides 527 4,7-ethanotetrahydroindene 477 ethene homopolymerization 577 ether-cyclopentadienyl complexes 416 ethylene complex of titanium 428 ethylene polymerization, with amido-cyclopentadienyl transition metal catalysts 447ff ethylenebis(indene) 497 — improved conditions for 497 — original preparation of 497 ethylene-bridged 491 — bis(bicyclooctylcyclopentadienyl) 467 — bis(indenyl) 464 ethylene-styrene copolymerization, with amido-cyclopentadienyl transition metal catalysts 448 ethylene-ZrCp2 300, 304 ethylmagnesiation — Zr-catalyzed 300 ethyltetramethylcyclopentadiene 482 EtMgCl 646 EtZrCp2Cl 304 Et2ZrCp2 300, 304 Eu(C5H5)2 56 Eu(C5Me5)2 59 faces 457 face-to-face poly(metallocenes) 750–751 factors governing the structure of multicyclopentadionyl complexes 44 (1)ferrocenophanes 740 first-principles calculations 704 fluorenes, bridged substituted 505 fluoride effect 706 fluorenyl complexes 418f, 421, 444, 448 fluxionality 4 four-center addition mechanism 289 four-centered addition reactions 304 Friedel–Crafts 486 — cyclization 480 fulvalene complexes, titanium 175 f fulvene 118, 126 — asymmetric reduction 471
781
— complexes, titanium 172, 176 ff, 227 — unsymmetrical 470 furan 122 gallium(I) 33, 90, 119 Goddard, W.A. 125 Green–Rooney metathesis mechanism 565 Grignard cross-coupling 712 group 6 anionic species [MCp2X]- 334 group 5 d2 metal compounds 339 group 6 d3 metal compounds [MCp2X] 335 group 5 d3 metal derivatives [MCp2L] 335 group 14 sandwich complexes — Cp*2Si 36 — germanocenes 36 — plumbocenes 36 — silicocenes 36 — stannocenes 36 β-H abstraction 286, 298, 299, 300, 631 — organozirconocene derivatives 257 — reaction of mono- and diorganylzirconocenes 258 α-H elimination 167, 175, 178, 181 ff β-H elimination 178, 184 half sandwich complexes 12, 415ff — comparison with ansa-metallocenes 415 — of group 14 elements 39 half sandwich metallocenes of the group 2 elements 32 halide ion abstraction 7 halogenolysis 262, 288 hapticity 4 haptotropic shifts 4 head-to-head structure 571 head-to-tail enchainment 570 1,2,3,4,5,6,7-heptamethylindene 488 heterobimetallic hydrides 94 heterocubanes, {Cp*AlZ}4 36 heterocycles 652 heterodinuclear complexes 355 heteroleptic metallocene 715 H–H and C–H activation 85 hindered rotation, DNMR studies 31 homoallylic alkoxides 632 homogeneous catalysis 55 homogeneous catalysts 97 homoleptic metal amides, use in synthesis of cyclopentadienyl metal complexes 421ff homotopic 457
782
Index
homotopic faces 457, 460 hydride 85f, 88, 92, 111, 124, 126 — scandium 113, 115ff. — yttrium 116ff., 120, 122, 132, 140, 143, 146 — lanthanum 147 β-hydride abstraction 647 hydride complexes 362 — titanocenes 166 β-hydride elimination 571, 639 hydride-bridged dimers 95 hydrido transition metal complexes 424ff hydroalumination 253, 279 — of alkenes 279 — Zr-catalyzed 253 hydroboration 279, 698 — with amido-cyclopentadienyl transition metal catalysts 449ff hydrocarbon activation 124, 135, 148 hydrocarbyls 97 β-hydrogen 285 hydrogen atom 663 β-hydrogen elimination 127ff, 596f hydrogen H2 639 β-hydrogen transfer 564 hydrogenation 142ff, 308, 627, 696 — of enamides 674 — with amido-cyclopentadienyl transition metal catalysts 449ff hydrogenolysis 83, 85, 88, 92, 94, 97, 98 hydrosilation 143, 308f, 709 hydrozirconation 242, 246, 249, 286, 287, 304, 309, 310 — mechanism 253 µ-hydroxo 69 HZrCp2Cl 252, 304 — preparation 249, 251 — reaction with alkenes and alkynes 249 — in situ generation 252 — structure 253 imido-cyclopentadienyl complexes 416 imidomolybdenum complexes 388 imidozirconocene derivatives 310 π* of the imine 670 imines 667 indene fluorene — chiral bridged 512 indenes 484, 504 — with alkyl bridges 498
— — — —
benzannelated 488 chiral substituents 490 with ethano-bridge 500 Friedel–Crafts routes to 4,7- or 5,6-disubstitut 488 — by nickel-catalyzed cross-coupling 485 — preparation of 484 — preparation of benzannelated 489 — preparation of 2-substituted 485f — preparation of 4-substituted 487 — 1-substituted 484 — 2-substituted 485 — substitution on the 6-membered ring 486 — synthesis of polymethylated 489 indenyl 123 indenyl fluorenyl 505 industrial applications 718 indium 33 1–3 insertion 571, 574 2–1 insertion 574 insertion of carbon monoxide 71, 76 insertions 89, 95 π-interaction with sulfur 44 inter-ligand distance 45 inter-ligand nonbonded interactions 44 interplane angles 37, 44 — group 14 sandwich metallocenes 37 — selected alkaline earth sandwich metallocenes 30 inverted sandwich, [CpM(Li·TMEDA)2]+ 24 iodides 67 iodinolysis 296, 308 — of RZrCp2Cl 262 Ir-catalyst 696 1,3-isomer 474 isonitrile insertion 269 isonitriles 83 isopropenylcyclopentadienyl 470 2-isopropylindene 485 isopropylindene-bridged — cyclopentadiene, fluorene 504 1-isopropyl-3-methylcyclopentadienyl 461 isopropyl-3-methylcyclopentadienyl ligand 516 isotactic 638 isotactic polymerization 586 isotope effect 565 isotopic perturbation of stereochemistry 128f Ito 676
Index
Josiphos 690 Jordan 627, 635, 636 Kaminsky 639 Kaminsky–Brintzinger systems 549 Kaminsky polymerization 310 Keck and Tagliavini 634 ketimine hydrogenation 696 kinetic resolution 656, 707 — through catalytic (EBTHI)Ti-cat 669 — of dihydrofurans 655 kinetic studies 668 lanthanide(III) 75 lanthanide aldolates 89 lanthanide ansa-metallocenes 98 lanthanide diiodides 57 lanthanide hydrocarbyls 98 lanthanide(II) metallocenes 57, 80, 102 lanthanide(III) metallocenes 69, 80 lantlanthanide methyls 85 lanthanide organometallics in materials science 104 lanthanide pyrrolyl complexes 75 lanthanide triflates 71 lanthanocene 111, 144 lanthanocene (II) 66 lanthanocene butadiene complex 91 lanthanocene(III) carboxylates 70 lanthanocene chlorides 62, 64, 65, 81 lanthanocene(II) complexes 55f. lanthanocene fluorides 62 lanthanocene halides 63 lanthanocene hydrides 94, 97f. lanthanocene hydrocarbyls 80, 82, 89, 94, 103 lanthanocene phosphides 75 lanthanocene triflates 71 lanthanum 111, 116, 121, 133, 135, 147 leaving groups 5 Lensink 672 less substituted carbon of the olefin 665 less-substituted C–Zr bond 637 Lewis acid association with carboxylic esters 643 Lewis model for the sandwich structures 20 LiCp, solid state structure 25 LiCp2Zr(C≡CR)3 288 ligand based nonbonding orbitals 20
783
ligand chirality 456 ligand tuning 701 ligand-element interaction, degree of ionic character 5 linked alkoxo-cyclopentadienyl complexes — see alkoxo-cyclopentadienyl complexes linked amido cyclopentadienyl complexes — see amido-cyclopentadienyl complexes lithiated phosphoylides 82 lithium binaphtholate 628 lithocene and sodocene anions 27 — [LiisoCp2] 27 — [NaCp2] 27 Ln–Ge and Ln–Sn bonds 83 Ln–H bonds 99 Ln–N bonds 75 Ln–O, Ln–S, Ln–Se, and Ln–Te bonds 69 Ln–P, Ln–As, Ln–Sb, and Ln–Bi bonds 75 Ln–Si bonds 94 lone pair 20 — stereochemically inactive 38 Lu–C bonds 71 Lu–P and Lu–As bonds 76 magnesium alkoxide elimination 647 magnesium dicyclopentadienides 5 manipulation of conversion levels 657 main group metallocenes 4 main-chain chirality 640 Manzer, L. 112 McMurry coupling 493 mechanism for the Ti-catalyzed reduction process 668 medium ring heterocycles 653 Meerwein’s diketone 511 Me-, vinyl-, phenylmagnesium halides 658 menthylcyclopentadiene 472 menthylcyclopentadienyl 457 2-menthylindene 490 3-menthylindene 490 2-menthylindenyl 460 metal amides — see homoleptic metal amides metal complexes with nonstereogenic bridging group 522 metal complexes with stereogenic bridging groups 531 metal–hydride or metal–carbon 678 metallacyclopentane 631, 647 metallaoxetanes 191 ff
784
Index
metallocene analogs 415ff metallocene catalysts 572 — β-agostic interactions 572 — borane activators 557 — bridge modifications 597 — C2-symmetry 552 — cationic catalysts 557 — cocatalyst-free systems 557 — Cs-symmetrical 586 — deactivation 574 — enantiomer separation 594 — historical development 550 — β-hydride 572 — ligand modifications 548 — β-methyl elimination 572 — molecular modeling 602 — for olefin polymerization 547 — unbridged chiral 586 metallocenes 4, 5 — alkaline earth sandwich metallocenes 29 — simple 513 — sterically hindered 513 — substituted cyclopentadienyl ligands 5 metallocenes of the alkali metals 23, 27ff — monomers and polymers 23 — lithocene and sodocene anions 27 metallocenes of the alkaline earth metals 28ff — beryllocene 29 — half sandwich metallocenes of the group 2 elements 32 metallocenes of the group 13 elements 34ff. — clusters 34 — half sandwich dimers 33 — monomers 33 — oligomers 33 — polymers 33 — heteroatom clusters 36 — in oxidation state III 35 metallocenes of the group 14 elements 36, 39ff. — group 14 sandwich complexes 36 — half sandwich complexes of group 14 elements 39 — reaction chemistry 40 — tricyclopentadienyl complexes 40 metallocenes of the pnictogens 41 — cationic sandwich metallocenes 42 — Cp'2ECl 41 — Cp'ECl2 41 metalloligand 59
metathesis reactions 5 metathesis-mechanism 564 methane complex 61 methoxyethylcyclopentadienyl ligand 58, 64 σ-methyl complexes 85 β-methyl elimination 129 methyl methacrylate (MMA) 610 methylacrylate 642 methylalumination 276, 280 — Zr-catalyzed 300 methylalumoxane (MAO) 242, 638 — structure 554 — synthesis 553 2-methylbenz[e]indene 488 methylmagnesiation 300 2-methyl-4-phenylindene 486 3-(methylthiol)indene 484 methylzirconation 279 — with MeZr+Cp2(THF)–BPh4 270 (S)-metolachlor 696 Me2ZrCp2 269 MeZrCp2Cl preparation 248 MeZr+Cp2(THF)–BPh4 267 Michael addition 709 migratory insertion 288 — acylzirconium derivatives 260 — reactions of RZrCp2Cl 267 mirror planes 640 MOCVD sources 29 models 670 Molander, G. 142 molecular mechanics force field calculations 44 molecular mechanics studies 22 molecular orbital diagram — 14 electron main sandwich complexes 21 — 12 electron systems 22 molecular orbital energy diagram 19 — half sandwich complexes 19 molybdenum [Cp2MX2] complexes 348 molybdenum [MCp2(η 2-X-X)] complexes 398 molybdenum [MCp2E] complexes 387 molybdenum [MCp2L] complexes 333 molybdenum [MCp2X] complexes 338 molybdenum [MCp2X2] complexes 359 molybdenum ionic [MCp2(η 2-X-L)]+ complexes 399 molybdenum ionic [MCp2EX]+ complexes 389
Index
molybdenum ionic [MCp2L2]2+ complexes 367 molybdenum ionic [MCp2X2]+, [MCp2X2]2+ complexes 371,373 molybdenum ionic [MCp2X3]+ complexes 378 molybdenum ionic [MCp2XL]+ complexes 365 molybdenum isocyanide complexes 366 molybdenum metallocenes [MCp2] 328 molybdenum-magnesium bonded compounds 363 monoalkylzirconocenes 267 monomeric alkali metallocenes 23ff. — LiCp' derivatives 24 monoorganylzirconocenes 285 — derivatives 268 — stoichiometric conversion into other organometals 272 — reactions 271, 274 multidecker anion [Cs2Cp3]- 27 multidecker sandwich 34 — [Cp3Tl2]- 34 — [Cs2Cp3]- 34 natural products 293 Nazarov 511 Nazarov cyclization 481 Negishi 658 Negishi protocol 258, 285, 286, 287 neomenthylcyclopentadiene 472 1-(neomenthyl)indenyl 467 Ni- or Pd-catalyzed cross-coupling reactions 276 nido clusters 19 — pentagonal pyramidal framework 19 N-inside and N-outside isomers 635 niobecene(III) boramate 381 niobium alkene complexes 392 niobium alkyne complexes 393f niobium allyl complexes 397 niobium [MCp2L] complexes 338 niobium [MCp2(η2-X-L)] complexes 390 niobium [MCp2(η 2-X-X)Z] complexes 400 niobium [MCp2X] complexes 344 niobium [MCp2X2] complexes 367f niobium [MCp2X3] complexes 375 niobium [MCp2XL] complexes 348f niobium [NbCp2(η2-L-L)], [NbCp2(η2-C,Y)] complexes 403
785
niobium complexes with (η2-C,Y) ligands (Y = N, O or S) 395f niobium complexes with (η2-CY2) ligands (Y = O or S) 395 niobium hydrides [MCp2H] 345 niobium [MCp2EX] imido complexes 384 niobium ionic [MCp2EL]+ complexes 387 niobium ionic [MCp2(η 2-C,Y)]+, [MCp2X(η 2-C,Y)]- complexes 398 niobium ionic [MCp2L2]+ complexes 356 niobium ionic [MCp2X2]+, [MCp2XL]2+ complexes 372f niobium ionic [MCp2X2]- complexes 357 niobium ionic [MCp2XL]+, [MCp2L2]2+ complexes 371 niobium ionic [MCp2X2L]+, [MCp2 (η2-X-L)Z]+ complexes 401 niobium ionic [MCp2X3L]+, [MCp2XL3]2+, [MCp2L3]3+ complexes 377 niobium(IV) ketamine 402 niobium metallocenes [MCp2] 325 niobium [MCp2EX] oxo, sulphido and tellurido complexes 379 nitriles 136ff. nitrogen heterocycles 79 nOe experiments 643 nomenclatures 5 — CpnE 5 — ECpn 5 nonbonded attractive interactions 44 noncoordinating anions 560 nonracemic carbinols 675 norbornacyclopentadienyl 466 Noyori 656, 676 nucleophilic substitution 686 nucleophilicity of [Cp*E]+ cations 39 octet rule 18, 20ff α-olefin insertion — regioselectivity 570 olefin hydrogenation 103 olefin metathesis 427 olefin polymerization reactions 85, 635 — electronic effects 576 olefin oligomerization 603 oligomerization reactions 638 oligomerization, with amido-cyclopentadienyl scandium catalysts 445ff open-calcocene 31 OpSc(PMe3)(H) 115, 128, 129
786
Index
opposite face 662 optically active transition metal complexes 420, 422f, 433, 437, 445 orbital energies (eV) for — Cp*2Si, Cp*2Ge, Cp*2Sn, Cp*2Pb 22 orbital interaction model for sandwich complexes 18, 20 organocerium hydrocarbyl 87 organolanthanide catalyst 104 organolanthanide hydrides 55, 83, 102 organolanthanide hydrocarbyls 55, 83, 102 organolanthanide olefin complex 61 organolanthanide-phosphine 75 organoscandium hydrocarbyls 84 organoyttrium acetylide 93 ortho lithiation 686 ortho-metallation 100 oscillating catalysts 590 1,1'-(3-oxa-pentamethylene)dicyclopentadienyl ligand 58 oxazaborolidine borane 714 oxazolidinones 644 oxidation of RZrCp2Cl 264 oxidative addition 180f, 246, 254f., 309 — alkenyl chlorides 255 — scope 255 oxidative coupling 246f. oxidative workup 700 oximato ligand 75 µ-oxo bridged organolanthanide complexes 69 µ-oxo bridged ytterbium complex 69 oxo derivatives 379 Paddle-Wheel structure 40 pair-selectivity 289, 291 partial negative charge 637 Pauson–Khand approach 495 Pauson–Khand cyclization 654 Pauson–Khand reaction 293, 659 Pd–π-allyl complexes 706 pending methoxy function 95 pentadienyl complexes 227 pentaethylcyclopentadiene 483 pentamethylcyclopentadiene 482, 483 pentamethylcyclopentadienyl ligand 65, 102, 113, 117 1,1'-pentamethylene-bis(cyclopentadienyl) complexe 99 pentamethylenefulvene 470 Pfaltz 666
phenylazido complexes 384 (1-phenylethyl)cyclopentadienyl 514 4-phenylindene 486 (8-phenylmenthyl)cyclopentadiene 472 phenyltetramethylcyclopentadienes 483 phobane 688 Phobyphos 688 phosphaalkynes 93 phosphinoytterbocenes 59 phosphorescence 145 phosphorus metallocene 43 phosphoylide ligands 81, 93 Piers, W. 116, 123 Pigiphos 710 Pino 639 planar chiral 686 planar chiral complexes 437 planar chirality 461 — cyclopentadienylmetal 461 — indenylmetal 461 polar monomers 610 polar reactions for carbon–carbon bond formation 276 polydentate amido-cyclopentadienyl ligands 435ff poly(vinylferrocene) 723, 725–726 poly(ferrocene) block copolymers 746 poly(ferrocenylene) 733–735 poly(ferrocenylenepersulfides) 749 poly(ferrocenylethylenes) 744 poly(ferrocenylsilanes) 740–742 polymerization — amido-cyclopentadienyl transition metal catalysts 417, 424, 441, 445ff, 450 — isotactic 455 polymerization of cyclic olefins 605 polymerize olefins 99 polymethylhydrosilane 674 polypropene — microstructure analysis by NMR 583 primary insertion 570 nPrMgCl 649 propene homopolymerization 580 propylene polymerization, with amido-cyclopentadienyl transition metal catalysts 448 protonolysis 288, 296, 308 pseudoequatorial position 633 pyrazol 121, 687 pyridyl ligand 636 pyrrolines 669
Index
rare earth carbene complexes 91 RCOZrCp2Cl 268 reaction chemistry 40ff — Cp' ligand exchange 40 — homolytic bond cleavage 41 — nucleophilic addition 41 — nucleophilic substitution 41 — oxidative addition 41 reactions involving cyclopentadienes 6 reduction 7 reductive coupling 630 reductive elimination 178ff, 309, 660 regioisomerization 294f. regioirregular insertion 574 regioregular insertions 570 regioselective insertion 631 regioselectivity 289, 291, 299, 570 — in the zirconacycle formation 648 resolution of racemic EBTHI complexes 627 Rh-catalyzed imine reduction 672 rhodium, ruthenium and cobalt complexes 666 ring closure 511 ring contraction 296 ring-bridged cyclopentadienyl ligands 67, 98 ring-closing metathesis 649 ring-contraction reaction 297 ring-expansion reaction 289, 297 ring-opening metathesis polymerization 427 ring-opening polymerization (ROP) 220, 224ff, 427, 740 rotation barriers — 1-alkenyl complexes 183 — carbene complexes 161 R/S nomenclature used in describing metallocenes — Brintzinger convention 461 — Schlögl 461 — Schögl’s rule 461 Ru-catalyzed reaction 676 Ru-catalyzed rearrangement 657 ruthenocenyl ligands 711 RZrCp2Cl 268 — chemistry 261 — conversion 271 — reaction with carbon electrophil 264 — reaction with CO 268 — structure 261 RZrIVCp2 derivatives 310
787
salt metathesis reactions 6 samarium(III) η3-allyl complexes 92 samarium azobenzene complex 76 samarium(III) hydrocarbyls 91 SAMP-hydrazone 714 sandwich complexes 11 sandwich complexes of amido-cyclopentadienyl ligands 440ff sandwich structures, bent 11 scandium silyls 83 scandium tellurides 74 scavenger 555, 561 Schulz–Flory distribution 549 secondary insertion 570 secondary interactions 87 secondary phosphine 687 Selenides 135 Shapiro elimination 496 sharpless 656 Shumann, H. 121 side chain donor 35 side chain functionality 5 — CpN 5 — CpO 5 side-on approach 663 silyl complexes 82, 94 silylzirconation 309 simplest metallocenes 23 — LiCp and LiCp2- 23 single π-coordinated Cp' ligand 12 single vessel 649 single-site catalyst 548 Skattebøl 479 — double 509 — rearrangement 477, 511 skeletal rearrangement 297 slippage of cyclopentadienyl rings 333 slippage of indenyl ligands 336 slip–sandwich 39 Sm(C5H5)2 56 Sm(C5Me5)2 59f., 79f., 90 solvent adducts 421, 425 sparteine 716 S-Pigiphos 710 stereochemistry-determining step 660 stereogenic metal 456 stereoisomerization 294f., 300 stereoselective polymerization 548 stereoselectivity 289 steric 693
788
Index
steric effects 637 steric encumbrances 37 steric interactions 651, 670 steric repulsions 648 steric unsaturation 55 sterically unfavorable interactions 652 stoichiometric reactions 634 structural comparison, sandwich complexes 11 structural diversity 46 structural flexibility of sandwiches 38 structural types 10ff. — clusters 16 — µ–π–cyclopentadienyl molecular complexes 13 — µ–π–cyclopentadienyl super-sandwich polymers 14 — dimers 15 — half sandwich complexes 12 — sandwich complexes 11 styrene polymerization 547 π-substitution 297 substitution onto the Cp ligand frame 7 sulfur ligands 73 super sandwich alkali metallocenes 25ff. — Cp*K·(pyridine) 26 — CpNa·TMEDA 26 — KCp 25 — KCpT 25 — NaCp 25 — LiCp 25 super-sandwich, LiCpT 24 super-sandwich polymeric array 34 syndiotactic polymerization 586 synthesis of allylic amines 630 synthesis of rac-(EBTHI)M complexes 626 synthetic procedures 5ff — addition of Cp anions 8 — disproportionation, comproportionation and decomposition 7 — halide ion abstraction 7 — reactions involving cyclopentadienes 6 — reduction 7 — salt metathesis reactions 6 — substitution onto the Cp ligand frame 7 synthetic versatility and significance 657 tantalum alkene complexes 392 tantalum [MCp2EX] alkylidene complexes 385
tantalum alkyne complexes 393f tantalum allyl complexes 397 tantalum [MCp2(η 2-X-X)Z] complexes 400 tantalum [MCp2X] complexes 344 tantalum [MCp2X2] complexes 367f tantalum [MCp2X3] complexes 375 tantalum [MCp2XL] complexes 348f tantalum complexes with (η2-C,Y) ligands (Y = N, O or S) 396 tantalum complexes with (η2-CY2) ligands (Y = O or S) 395 tantalum hydrides [MCp2H] 345 tantalum [MCp2EX] imido complexes 384 tantalum ionic [MCp2L2]+ complexes 356 tantalum ionic [MCp2X3L]+, [MCp2XL3]2+, [MCp2L3]3+ complexes 377 tantalum metallocenes [MCp2] 325 tantalum [MCp2EX] oxo, sulphido and tellurido complexes 379 Tebbe reagent 167, 205 tellerium 135 telluroformaldehyde derivatives 396 tellurolates 74 terminal olefins 648 terpolymerization 607 tetrahydroborates and tetrahydroaluminates 94 tetramethylcyclopentadiene 481 tetramethylcyclopentadienyl groups 87 tetramethylcyclopentenone 482 tetramethylethylene-bridged bis(cyclopentadienyl) 459 tetramethylethylene-bridged dicyclopentadienyl ligand 61 2,3,4,7-tetramethylindene 488 tetraphenylcyclopentadiene 481 Teuben, J. 126, 134, 142, 147 thallium 33 thallium cyclopentadienides 5 theoretical molecular orbital studies 22 theoretical studies 567 Ti- and B-based chiral Lewis acids 634 Ti NMR spectroscopy 162, 214f thiolate ligands 73 thiophilic properties 200 three dimensional aromaticity 19 tied-back analogs of the (C5Me5)2Ln derivatives 87 tied-back organolanthanide hydrocarbyls 88 tied-back scandocene hydride 102
Index
Ti–H 670 Tilley, T.D. 138 titanacyclobutanes 167 f, 182f, 196f titanacyclobutenes 188 ff, 196f, 223 titanacyclopentene 660 titanaoxetanes 192 f titanium(III)–π-allyl complex 632 titanium dichlorides 518 titanocenes 153 ff — acetylene complexes 162 ff, 218 — 1-alkenyl complexes 178 ff — amides 159 — applications 217 ff — aqua complexes 173 — aryne complexes 185 — bonding modes 154 — carbene complexes 161, 167 ff, 183, 185 ff — carbonyl complexes 165 — chiral 455 — electronic properties 158 f — fulvalene complexes 175f — hexafluoropnicogenate 174 — hydride complexes 166 — low oxidation states 159 — NMR data 212 ff — olefin complexes 162, 184, 218 — oxidation states 153 — structural data 210 — synthesis 154 ff — UV vis data 215 f Ti-X double bonds 171 ff total synthesis of the antifungal agent Sch 38516 646 trans coordination geometry 695 transfer-hydrogenation 710 transition state models 673 transition state structures 665 transmetallation 246, 260, 288, 298, 299, 300, 308 — involving Cu in conjugate addition 273 transmetallation reaction 56 TRAP-ligands 690 triantimony Zintl ion 80 tricyclopentadienyl complexes 40 — [Cp3E]- 40 — [Cp5Pb2]- 40 — [Cp9Pb4]- 40 tricyclopentadienylaluminum derivatives, differences between hapticities 4 tridentate amido cyclopentadienyl ligands 435ff
789
trienediyl complexes 93 trihydride derivatives [MCp2H3] 361 trihydrides 358 triisopropylcyclopentadienes 481 trimethylacetate complex 355 trimethylaluminum 561 1,2,3-trimethylcyclopentadiene 483 1,2,4-trimethylcyclopentadiene 480 trimethylcyclopentenone 481 1,1'-trimethylene-bis(cyclopentadiene) 67 trimethylene-bridged bis(cyclopentadienyl) ligand 61 2,4,7-trimethylindene 488, 499 trimethylsilylcyclopentadienyl ligand 63 (trimethylsilyl)diazomethyl 94 (triphenylmethyl)cyclopentadienyl 470 triptycyl ligand 705 tris(cyclopentadienyl)lanthanide complexes 55, 57 tris(pentafluorophenyl)borane — metallocene activator 559 trisubstituted alkenes 276 tuck-in derivatives 176 tungsten(IV) [Cp2MX2] complexes 348 tungsten [MCp2(η 2-X-X)] complexes 398 tungsten [MCp2E] complexes 388 tungsten [MCp2L] complexes 333 tungsten [MCp2X2] complexes 359 tungsten ionic [MCp2(η 2-X-L)]+ complexes 399 tungsten ionic [MCp2EX]+ complexes 389 tungsten ionic [MCp2L2]2+ complexes 367 tungsten ionic [MCp2X]+ complexes 346 tungsten ionic [MCp2X2]+ complexes 371 tungsten ionic [MCp2X3]+ complexes 378 tungsten ionic [MCp2XL]+ complexes 365 tungsten metallocenes [MCp2] 328 turnover in regioselectivity 637 Ullmann coupling 690 unfunctionalized alkenes 664 unsaturated alcohols 649 α,β-unsaturated aldehydes — preparation 269 unsaturated pyrans 652 vanadacyclopropene complexes 403 vanadium [MCp2E] complexes 389 vanadium [MCp2L] complexes 335
790
Index
vanadium [MCp2X] complexes 340f vanadium [MCp2X2] complexes 367f vanadium [MCp2XL] complexes 348 vanadium [VCp2(η2-X-X)] complexes 403 vanadium ionic [MCp2L]-, [MCp2L]+ complexes 334f vanadium ionic [MCp2L2]+ complexes 356 vanadium ionic metallocenes [MCp2]+, [MCp2]- 329f vanadium metallocenes [MCp2] 323 vinylic C–H activation 86 vinylidene complexes 161, 178, 182 f, 185 ff, 219 — dinuclear complexes 207 vinylidene-acetylene transformation 195 f vinylmethyl ketones 675 voltametric studies 338 VSEPR arguments 20 Watson, P. 119, 124 Waymouth and Coates 640 Waymouth and Pino 661 Wilkinson, G. 112, 116 X-ray analysis 257 X-ray crystallography 259 X-ray crystal structures of amido-cyclopentadienyl complexes 431ff Yasuda, H. 132 Yb(C5H5)2 56 Yb(C5Me5)2 59, 85 Y–C bond 89 zig-zag polymeric structures 33 Ziegler-Natta catalysis 415, 547 — chain transfer 563
— deactivation 563 — mechanism of chain growth 563 Ziegler–Natta olefin polymerization 270 Ziegler-Natta polymerization 118, 126f., 129, 148 Ziegler–Natta-type alkene polymerization 276 zirconabicycles 295 zirconacycles 293 — five-membered 289, 296f — three-membered 283, 288f, 294, 296f, 299, 308 zirconacyclopentanes 298 zirconacyclopentene 304 zirconacyclopropanes 243, 296 zirconaoxiranes 288 zirconium 241 zirconium hydride 661 zirconocene catalysts see metallocene catalysts zirconocene dihalides 241 — as Lewis acid catalysts — in the Friedel–Crafts reactions 241 — in the Ziegler–Natta polymerization 241 zirconocene and titanocene bis(triflate) 643 zirconocene–ethylene complex 647 zirconocenes 284, 300, 455 — complexes 284 — derivatives 242 — equivalent 245 Zr(III) species 242 Zr-catalyzed carboalumination of alkynes 270 ZrCp2-promoted bicyclization 293 — of alkenyl and alkynyl hydrazones 289 Zr–Mg ligand exchange 647, 651
Preface
The cyclopentadienyl ligand (Cp') is ubiquitous in the organometallic chemistry of the transition elements. It is hard to imagine what this research field would be without this ligand. Soon after the discovery of ferrocene in 1951, the chemistry of so-called metallocenes, bis(cyclopentadienyl) metal derivatives, has been developed for most d-elements. Thus, strictly speaking a metallocene is a “sandwich” compound, being expressed by the formula [M(η5C5R5)2]. However, it is currently accepted to designate as metallocenes also compounds, e.g., containing further ligands, the so-called bent metallocenes such as Cp2WH2, or complexes in which one of the Cp fragments has been replaced by an anionic ligand fulfilling similar coordination-chemical purposes. In this monograph a dozen of internationally recognized experts in the field provide up-to-date and critical overviews concerning metallocenes and their applications. However, the book is not intended to be a comprehensive and systematic treatise of a compound class representing roughly three quarter of the elements. Therefore, the topic of each single chapter has been selected according to its importance, as reflected by current reasearch activities. Part 1 is focussed on synthesis, properties, and reactivity, whereas Part 2 concentrates on applications, in particular homogeneous catalysis and materials science. Metallocenes containing main-group elements have long been considered as the “exotic” representatives of this class of compounds. Jutzi and Burford (Chapter 1) demonstrate the wealth of structural and coordination-chemical features of general theoretical interest of derivatives of this kind that have been reported in recent years. Organolanthanide chemistry is to be identified with the use of cyclopentadienyl ligands, much more than for the transition metals. Chapter 2 (Edelmann) covers mainly synthetic aspects related to the lanthanocenes whose importance also in catalytic chemistry is steadily increasing. The related scandocenes and yttrocenes have received much attention in the last decade because of their role in C-H activation reactions and in catalytic processes involving olefins. Bercaw and Chirik review this area in Chapter 3. Titanocenes (Beckhaus, Chapter 4) and Zirconocenes (Negishi and Montchamp, Chapter 5) are nowaday extremely important compounds with an enormous application potential. Thus, from a variety of stereoselective organic transformations to modern Ziegler-Natta olefin polymerization, these two metallocenes have contributed most significantly to the progress of modern synthetic chemistry. For the organometallic chemist, an important class of metallocenes is the one containing the elements of groups 5 and 6 (Royo and Ryan, Chapter 6). This because of the multifaceted chemistry these compounds display, and the wealth of derivatives and structural types that is not to be found for other metallocenes. Further, Okuda and Eberle discuss in Chapter 7 the chemistry of
VI
Preface
“half-sandwich” complexes, as these are recognized as metallocene analogues. Since titanocene and zirconocene dichlorides are fundamental starting materials for use in organic synthesis, both as reagents and catalysts, Haltermann describes in Chapter 8 recent developments in the synthesis of chiral derivatives for applications in stereoselective reactions. Part 2 of the book starts with a review of what probably is to be seen as the hottest topic in metallocene chemistry, i.e., their application as catalysts for olefin polymerization. This most exciting area is reviewed by Janiak in Chapter 9. Hoveyda and Morken in Chapter 10 present an overview of organic synthesis utilizing enantiomerically pure titanocene and zirconocene derivatives, most prominently C2-symmetric ansa-metallocenes. Ferrocene is the “oldest” prototype metallocene but research in the field of ferrocene chemistry is still very active. This has been reviewed in the companion monograph of this book: Ferrocenes. Homogeneous Catalysis. Organic Synthesis. Materials Science (A. Togni, T. Hayashi, Eds., VCH, 1995). Therefore, the only aspect that has been taken into consideration here is the application of new chiral ferrocenyl ligands in asymmetric catalysis (Togni, Chapter 11). This particular area has recently seen the first large scale industrial application of a metallocene ligand. Metallocene fragments, and most prominently ferrocenes, have also been incorporated into polymeric structures. This kind of materials is of particular appeal because of the interesting electrical, magnetic, and optical properties they exhibit, differentiating them from other, more common polymers. This interdisciplinary area is reviewed in the final Chapter 12 by Manners and coworkers. We anticipate that the monograph will be of utility to many members of the chemical community, from graduate students to practitioners in industry, as well as from organic to inorganic chemists. It is also our hope that it will further stimulate an already extremely active research area. We gratefully acknowledge the excellent work done by all authors, as well as VCH-Wiley for the collaboration at all stages of this book project.
ETH Zürich, Norman, Oklahoma April 1998
Antonio Togni Ronald Halterman
Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . V
Volume 1 Synthesis and Reactivity 1
Main Group Metallocenes P. Jutzi and N. Burford
1.1 1.2 1.3 1.3.1 1.3.2 1.3.3 1.3.4 1.3.5 1.3.6 1.3.7 1.4 1.5 1.5.1 1.5.2 1.5.3 1.5.4 1.5.5 1.5.6 1.6 1.6.1 1.6.2 1.7
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synthetic Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactions Involving Cyclopentadienes . . . . . . . . . . . . . . . . . . . . . . . . . Salt Metathesis Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Disproportionation, Comproportionation and Decomposition . . . . . . . Halide Ion Abstraction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Substitution onto the Cp Ligand Frame . . . . . . . . . . . . . . . . . . . . . . . . . Reduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Addition of Cp Anions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Characterization Features . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structural Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Sandwich Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Half Sandwich Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . µ-π-Cyclopentadienyl Molecular Complexes . . . . . . . . . . . . . . . . . . . . µ-π-Cyclopentadienyl Super-Sandwich Polymers . . . . . . . . . . . . . . . . Dimers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cp-Element Orbital Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Orbital Interaction Model for Half Sandwich Complexes . . . . . . . . . . . Orbital Interaction Model for Sandwich Complexes . . . . . . . . . . . . . . . Metallocenes of the Alkali Metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3 3 5 6 6 7 7 7 7 8 8 10 11 12 13 14 15 16 18 18 20 23
VIII
Contents
1.7.1 1.7.2 1.8 1.8.1 1.8.2 1.8.3 1.9 1.9.1 1.9.2 1.9.3 1.9.4 1.10 1.10.1 1.10.2 1.10.3 1.10.4 1.11 1.12
Monomers and Polymers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lithocene and Sodocene Anions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metallocenes of the Alkaline Earth Metals . . . . . . . . . . . . . . . . . . . . . Beryllocene . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Alkaline Earth Sandwich Metallocenes . . . . . . . . . . . . . . . . . . . . . . . . Half Sandwich Metallocenes of the Group 2 Elements . . . . . . . . . . . . Metallocenes of the Group 13 Elements . . . . . . . . . . . . . . . . . . . . . . . Half Sandwich Monomers, Dimers, Oligomers and Polymers . . . . . . Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metallocenes Involving Group 13 Elements in Oxidation State III . . . Heteroatom Clusters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metallocenes of the Group 14 Elements . . . . . . . . . . . . . . . . . . . . . . . Group 14 Sandwich Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Half Sandwich Complexes of Group 14 Elements . . . . . . . . . . . . . . . Tricyclopentadienyl Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reaction Chemistry . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metallocenes of the Pnictogens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Factors Governing the Structure of Multicyclopentadienyl Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.13 1.14
2
23 27 28 29 29 32 33 33 34 35 36 36 36 39 40 40 41 44 46 46 47
Lanthanocenes F. T. Edelmann
2.1 2.2 2.3 2.3.1 2.3.2 2.3.3 2.3.4 2.3.5 2.4 2.5
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Lanthanide(II) Metallocenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Lanthanide(III) Metallocenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Lanthanide(III) Metallocenes Containing Ln-Group 17 Element Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 Lanthanide(III) Metallocenes Containing Ln-Group 16 Element Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69 Lanthanide(III) Metallocenes Containing Ln-Group 15 Element Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Lanthanide(III) Metallocenes Containing Ln-Group 14 Element Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80 Lanthanide(III) Metallocenes Containing Ln-Hydrogen Bonds . . . . . 94 Conclusion and Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 102 Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
Contents
3
IX
Group 3 Metallocenes P. J. Chirik and J. E. Bercaw
3.1 3.2 3.2.1 3.2.2 3.3 3.4 3.5 3.6 3.7 3.8 3.9 3.10 3.11
4
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synthesis of Group 3 Metallocenes . . . . . . . . . . . . . . . . . . . . . . . . . . . Scandium . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Yttrium and Lanthanum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Group 3 Metallocenes for the Activation of Hydrocarbons . . . . . . . . Group 3 Metallocenes as Homogeneous Olefin Polymerization Catalysts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactions of Group 3 Metallocenes with Acetylenes . . . . . . . . . . . . . Reactions of Group 3 Metallocenes with Group 6 Compounds . . . . . Reactions of Group 3 Metallocenes with Other Small Molecules . . . Applications of Group 3 Metallocenes in Organic Synthesis . . . . . . . Photophysics of Group 3 Metallocenes . . . . . . . . . . . . . . . . . . . . . . . . Alternatives of Cyclopentadienyl Ligands . . . . . . . . . . . . . . . . . . . . . Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
111 112 112 116 124 126 133 135 136 142 145 146 147 148
Titanocenes R. Beckhaus
4.1 4.1.1 4.1.2 4.1.3 4.2 4.2.1 4.2.2 4.3 4.3.1 4.3.2 4.3.3 4.3.4 4.3.5 4.3.6 4.3.6.1 4.3.6.2 4.3.7
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bonding Modes of Cyclopentadienyl Ligands in Titanium Complexes . Synthesis of Titanocene Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . Sources of the Titanocene Fragment as a Versatile Building Block . . Electronic Properties of Titanocenes . . . . . . . . . . . . . . . . . . . . . . . . . . Titanocene Complexes in Low Oxidation States . . . . . . . . . . . . . . . . . Substituents on the Cp Rings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ligands in the Coordination Sphere of Titanocenes . . . . . . . . . . . . . . Olefins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acetylenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carbonyls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Phosphines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hydrides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Titanium Ligand Multiple Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carbene Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Complexes Containing Ti-X Double Bonds (X: N, P, Chalcogenides) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metallocene Complexes of the Type [Cp2TiL2]2+ . . . . . . . . . . . . . . . .
153 154 154 157 158 159 161 162 162 162 165 166 166 167 167 171 173
X 4.3.7.1 4.3.7.2 4.3.7.3 4.4 4.4.1 4.4.2 4.4.2.1 4.4.2.2 4.4.2.3 4.4.2.4 4.4.3 4.4.3.1 4.4.3.2 4.4.3.3 4.4.3.4 4.4.3.5 4.4.3.6 4.4.3.7 4.4.3.8 4.4.3.9 4.4.4 4.5 4.5.1 4.5.2 4.5.2.1 4.5.2.2 4.5.2.3 4.5.3 4.6 4.6.1 4.6.1.1 4.6.1.2 4.6.1.3 4.6.2 4.6.2.1 4.6.2.2 4.6.2.3 4.7 4.8
Contents
Aqua-Titanocene Complexe . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Titanocene Complexes with Highly Fluorinated Ligand Systems . . . Further Complexes with [Titanocene]2+ (d0) and [Titanocene]+ (d1) Units . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactions in the Coordination Sphere of Titanocenes . . . . . . . . . . . . . The free Titanocenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactions of Ti-C σ Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reductive Elimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oxidative Addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . α-H Elimination. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . β-H Elimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactions of Ti=C Double Bonds . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactions with Electrophiles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . C-H Activation Reactions on [LnTi=C] Fragments . . . . . . . . . . . . . . . [2+1] Additions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cycloaddition Reactions with Acetylenes . . . . . . . . . . . . . . . . . . . . . . Synthesis and Reactivity of Metallaoxetanes . . . . . . . . . . . . . . . . . . . Synthesis of Heterodinuclear Carbene Complexes with a Titanaoxetane Substructure . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synthesis and Reactivity of Heterotitanacyclobutenes and -butanes . Structure and Reactivity Pattern for small Titanacycles . . . . . . . . . . . Dimetallic Compounds Derived from Ti=C Species . . . . . . . . . . . . . . Reactions Involving the Cp Ligand . . . . . . . . . . . . . . . . . . . . . . . . . . . Analytical Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structural Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . NMR Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1H NMR Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13C NMR Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ti NMR Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . UV vis Spectra . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Selected Applications of Titanocene Complexes. . . . . . . . . . . . . . . . . C-C Coupling Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dimerization of Ethylene. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Head to Tail Coupling Reactions of Acetylenes . . . . . . . . . . . . . . . . . Cyclization Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stoichiometric and Catalytic Reactions of Carbenoid Metal Complexes of Electron-Deficient Transition Metals . . . . . . . . . Carbonyl Olefinations with Tebbe, Grubbs, and Petasis Reagents . . . Ene Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Catalytic Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cp-Equivalents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
173 174 174 175 175 178 179 180 181 184 185 186 187 187 188 191 193 196 200 205 209 210 210 212 212 213 214 215 217 217 217 218 219 220 220 223 224 226 230 230
Contents
5
XI
Zirconocenes E. Negishi and J.-L. Montchamp
5.1 5.2 5.2.1 5.2.2 5.2.3 5.2.4 5.3 5.3.1 5.3.1.1 5.3.1.2 5.3.1.3 5.3.2 5.3.2.1 5.3.2.2 5.3.3 5.3.3.1 5.3.3.2 5.3.3.3 5.3.4 5.3.5 5.3.5.1 5.3.5.2 5.3.6 5.3.6.1 5.3.6.2 5.3.6.3 5.4 5.4.1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Preparation of Cp2Zr Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . Transmetallation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hydrozirconation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oxidative Addition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oxidative Coupling and Complexation . . . . . . . . . . . . . . . . . . . . . . . . Reactions and Synthetic Applications of Monoorganylzirconocene Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Protonolysis, Halogenolysis, Oxidation, and Related Carbon-Heteroatom Bond Formation Reactions . . . . . . . . . . . . . . . . . Protonolysis and Deuterolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Halogenolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Oxidation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carbon-Carbon Bond Formation via Polar Reactions of RZrCp2Cl with Carbon Electrophiles . . . . . . . . . . . . . . . . . . . . . . . Reactions of Allylzirconocenes with Aldehydes . . . . . . . . . . . . . . . . . Reactions of Cationic Alkyl- and Alkenylzirconocenes with Aldehydes, Epoxides, Ketones, and Nitriles . . . . . . . . . . . . . . . . Carbon-Carbon Bond Formation via Migratory Insertion Reactions of RZrCp2Cl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Carbonylation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Isonitrile Insertion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Migratory Insertion Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . Stoichiometric Carbozirconation of Monoorganylzirconocenes . . . . . Reactions of Monoorganylzirconocenes via Transmetallation . . . . . . Stoichiometric Conversion of Monoorganylzirconocenes into Other Organometals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactions of Monoorganylzirconocenes Catalyzed by Metal Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Organometallic Reactions Catalyzed by Zirconocene Derivatives . . . Zirconium-Catalyzed Carboalumination and Carbozincation of Alkynes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zirconium-Catalyzed Hydroalumination and Hydroboration of Alkenes and Alkynes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zirconium-Catalyzed Enantioselective Carboalumination of Alkenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactions Involving π-Complexes of Zirconocenes or ThreeMembered Zirconacycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Formation of π-Complexes of Zirconocenes . . . . . . . . . . . . . . . . . . . .
241 246 246 249 254 255 261 262 262 262 264 264 264 266 267 268 269 269 270 271 272 274 276 276 279 280 283 284
XII 5.4.2 5.4.2.1 5.4.2.2 5.4.2.3
5.4.3 5.4.4 5.4.4.1 5.4.4.2 5.4.4.3 5.5
6
Contents
Stoichiometric Reactions of π-Complexes of Zirconocenes (or Three-Membered Zirconacycles) . . . . . . . . . . . . . . . . . . . . . . . . . . Protonolysis and Halogenolysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ring Expansion Reactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Regioisomerization, Stereoisomerization, σ-Bond Metathesis, and Other Stoichiometric Reactions of Three-Membered Zirconacycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stoichiometric Reactions of Five-Membered Zirconacycles . . . . . . . Catalytic Reactions Involving Zirconacycles and Related Derivatives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zirconium-Catalyzed Carbomagnesiation Reactions Involving C-H Activation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Zirconium-Catalyzed Carboalumination of Alkynes via C-H Activation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Other Zirconium-Catalyzed Reactions Involving Cp2Zr(II) Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
288 288 289
294 296 299 300 304 308 310 312
Group 5 and Group 6 Metallocenes P. Royo and E. Ryan
6.1 6.2 6.2.1 6.2.1.1 6.2.1.2 6.2.2 6.2.2.1 6.2.2.2 6.2.2.3 6.2.2.4 6.3
6.3.1 6.3.1.1 6.3.1.2 6.3.1.3
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Type A Metallocene Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neutral Metallocenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Group 5 Metallocenes [MCp2] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Group 6 Metallocenes [MCp2] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ionic Mettalocenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Group 5 Cationic Metallocenes [MCp2]+ . . . . . . . . . . . . . . . . . . . . . . Group 5 Anionic Metallocenes [MCp2]- . . . . . . . . . . . . . . . . . . . . . . . Group 6 Cationic Metallocenes [MCp2]+ . . . . . . . . . . . . . . . . . . . . . . Group 6 Anionic Metallocenes [MCp2]- . . . . . . . . . . . . . . . . . . . . . . . Type B-Bent bis(Cyclopentadienyl) Metal Complexes with Mono-Functional Single-Atom Donor Monoanionic and Neutral Ligands. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Complexes with One Coordinated Ligand . . . . . . . . . . . . . . . . . . . . Neutral Group 6 [MCp2L] Metal Complexes . . . . . . . . . . . . . . . . . . . Anionic Group 5 [MCp2L]- and Group 6 [MCp2X]- Metal Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neutral Group 5 [MCp2L] and Group 6 [MCp2X] Metal Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
321 322 323 323 326 329 329 330 330 330
331 331 333 334 336
Contents
6.3.1.4 6.3.1.5 6.3.1.6 6.3.1.7 6.3.1.8 6.3.2 6.3.2.1 6.3.2.2 6.3.2.3 6.3.2.4 6.3.2.5 6.3.2.6 6.3.2.7 6.3.2.8 6.3.2.9 6.3.2.10 6.3.2.11 6.3.3 6.3.3.1 6.3.3.2 6.3.3.3 6.4 6.4.1 6.4.1.1 6.4.1.2 6.4.1.3 6.4.1.4 6.4.2 6.4.2.1 6.5 6.5.1 6.5.1.1 6.5.1.2 6.5.1.3 6.5.1.4 6.5.1.5 6.5.1.6 6.5.1.7
Anionic Group 5 [MCp2X]- and Cationic Group 6 [MCp2L]+ Metal Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neutral d2 Vanadium Compounds [VCp2X] . . . . . . . . . . . . . . . . . . . . Neutral d2 Niobium and Tantalum Compounds [(Nb,Ta)Cp2X] . . . . . Cationic Group 5 d2 Metal Compounds [MCp2L]+ . . . . . . . . . . . . . . . Cationic Group 6 d2 Metal Compounds [MCp2X]+ . . . . . . . . . . . . . . . Complexes with Two Coordinated Ligands . . . . . . . . . . . . . . . . . . . Neutral Group 5 d2 Metal Compounds [MCp2XL] . . . . . . . . . . . . . . . Cationic Group 5 d2 Metal Compounds [MCp2L2]+ . . . . . . . . . . . . . . Anionic Group 5 d2 Metal Compounds [MCp2X2]- . . . . . . . . . . . . . . Neutral Group 6 d2 Metal Compounds [MCp2X2] . . . . . . . . . . . . . . . Cationic Group 6 d2 Metal Compounds [MCp2XL]+ . . . . . . . . . . . . . Dicationic Group 6 d2 Metal Compounds [MCp2L2]2+ . . . . . . . . . . . . Neutral Group 5 d1 Metal Compounds [MCp2X2] . . . . . . . . . . . . . . . Cationic Group 5 d1 Metal Compounds [MCp2XL]+, [MCp2L2]2+ . . . Cationic Group 6 d1 Metal Compounds [MCp2X2]+ . . . . . . . . . . . . . . Cationic Group 5 d0 Metal Compounds [MCp2X2]+, [MCp2XL]2+ . . . Dicationic Group 6 d0 Metal Compounds [MCp2X2]2+ . . . . . . . . . . . . Complexes with Three Coordinated Ligands . . . . . . . . . . . . . . . . . . Neutral Group 5 d0 Metal Compounds [MCp2X3] . . . . . . . . . . . . . . . Cationic Group 5 d0 Metal Compounds [MCp2X2L]+, [MCp2XL2]2+ and [MCp2L3]3+ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cationic Group 6 d0 Metal Compounds [MCp2X3]+ . . . . . . . . . . . . . . Type C - Complexes with Single-Atom Donor Dianionic Ligands . . . 18-Electron Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neutral Group 5 d0 Metal Compounds [MCp2EX] . . . . . . . . . . . . . . . Cationic Group 5 d0 Metal Compounds[MCp2EL]+ . . . . . . . . . . . . . . Neutral Group 6 d2 Metal Compounds [MCp2E] . . . . . . . . . . . . . . . . Cationic Group 6 d0 Metal Compounds [MCp2EX]+ . . . . . . . . . . . . . 17-Electron Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neutral d1 Vanadium Compounds [VCp2E] . . . . . . . . . . . . . . . . . . . . Type D-Complexes with η2-Side-Bonded Ligands . . . . . . . . . . . . . . . 18-Electron Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neutral Group 5 d2 Metal Compounds [MCp2(η2-X-L)] . . . . . . . . . . Neutral Group 5 Metal Complexes [MCp2X(η2-C-C)] or [MCp2X(η2-C-Y)] (Y = N, O, or S) . . . . . . . . . . . . . . . . . . . . . . . . Anionic Group 5 d2 Metal Compounds [MCp2(η2-C,Y)]+ and [MCp2X(η2-C,Y)]- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neutral Group 6 d2 Metal Compounds [MCp2(η2-X-X)] . . . . . . . . . . Cationic Group 6 d2 Metal Compounds [MCp2(η2-X-L)]+ . . . . . . . . . Neutral Group 5 d0 Metal Compounds [MCp2(η2-X-X)Z] . . . . . . . . . Cationic Group 5 d0 Metal Compounds [MCp2X2L]+ and [MCp2 (η2-X-L)Z]+ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
XIII 339 340 344 345 346 346 348 356 357 359 365 367 368 371 371 372 373 374 375 377 378 378 378 379 387 387 389 389 389 390 390 390 391 398 398 399 400 401
XIV
Contents
6.5.2 6.5.2.1 6.5.2.2
17-Electron Compounds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Neutral d1 Vanadium Compounds [VCp2(η2-X-X)] . . . . . . . . . . . . . . Niobium Compounds [NbCp2(η2-L-L)] and [NbCp2(η2-C,Y)] . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7
Half-Sandwich Complexes as Metallocene Analogs
403 403 404 405
J. Okuda and T. Eberle 7.1 7.2 7.2.1 7.2.2 7.2.3 7.2.4 7.3 7.4 7.5 7.6 7.6.1 7.6.2 7.7
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Complexes with Linked Amido-Cyclopentadienyl Ligands . . . . . . . . Synthesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hydrido and Alkyl Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reactivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Complexes with Polydentate Amido-Cyclopentadienyl Ligands . . . . Complexes with two Amido-Cyclopentadienyl Ligands . . . . . . . . . . . Complexes with Linked Alkoxo-Cyclopentadienyl Ligands. . . . . . . . Catalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polymerization of Ethylene, α-Olefins and Dienes . . . . . . . . . . . . . . . Hydrogenation and Hydroboration . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
8
Synthesis of Chiral Titanocene and Zirconocene Dichlorides
415 417 417 424 429 431 435 440 443 445 445 449 450 450
R. L. Halterman 8.1 8.2 8.3 8.3.1 8.3.2 8.4 8.4.1 8.4.2 8.4.3 8.4.4 8.5 8.6 8.6.1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ligand Chirality and Stereoisomeric Metal Complexes . . . . . . . . . . . Unbridged Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cyclopentadienes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Indenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bridged Ligands with Nonstereogenic Bridging Groups . . . . . . . . . . Bis(cyclopentadienes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bis(indenes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bis(fluorenes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mixed Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bridged Ligands with Stereogenic Bridging Groups . . . . . . . . . . . . . Unbridged Metal Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bis(cyclopentadienyl)metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
455 456 469 469 484 491 491 497 502 502 506 513 513
Contents
8.6.2 8.6.3 8.6.4 8.7 8.7.1 8.7.2 8.7.3 8.7.4 8.8
Bis(indenes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bis(fluorenes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mixed Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metal Complexes with Nonstereogenic Bridging Groups . . . . . . . . . . Bis(cyclopentadienyl)metals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bis(indenes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bis(fluorenes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Mixed Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Metal Complexes with Stereogenic Bridging Groups . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
XV 518 520 521 522 522 527 530 530 531 539
XVI
Contents
Volume 2 Applications 9
Metallocene Catalysts for Olefin Polymerization C. Janiak
9.1 9.2 9.3 9.4 9.5 9.5.1 9.5.2 9.5.3 9.5.4 9.5.5 9.5.6 9.5.7 9.5.8 9.6 9.7 9.7.1 9.7.2 9.7.3 9.7.4 9.7.5 9.7.6 9.7.7 9.7.8 9.7.9 9.7.10 9.7.11 9.7.12 9.7.13 9.8 9.9
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Historical Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methylalumoxane - Characteristics and Function . . . . . . . . . . . . . . . . The Borane Activators or Cationic Catalysts and other Cocatalyst-free Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Mechanism of Chain Growth, Chain Transfer and Deactivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . α-Agostic Interactions and the Modified Green-Rooney Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theoretical Considerations on Olefin Coordination and Insertion . . . Models for Cp2Zr(R)(olefin)+ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Regioselectivity in α-Olefin Insertion . . . . . . . . . . . . . . . . . . . . . . . . . Chain Transfer Processes, β-Agostic Interactions, β-Hydride, β-Methyl Elimination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Deactivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The Problem of Comparing Activities, ‘A Caveat’ . . . . . . . . . . . . . . . Electronic Effects in Olefin Polymerization . . . . . . . . . . . . . . . . . . . . Ethene Homopolymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Propene Homopolymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polypropene Microstructures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Microstructure Analysis by NMR . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stereochemical Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Non-C2 Symmetrical or Unsymmetrical ansa-Metallocene Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Substituted Unbridged Metallocenes . . . . . . . . . . . . . . . . . . . . . . . . . . Isotactic-Atactic Block Polymerization, Oscillating Catalysts . . . . . . Syndiotactic Polymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Enantiomer Separation and Enantioselective Polymerization . . . . . . . Comparative Polymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stereoregularity - Pressure-Dependence and Chain-End Epimerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Bridge Modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Molecular Mass Enhancement, Advanced Metallocenes . . . . . . . . . . Molecular Modeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Homopolymerization of Other α-Olefins . . . . . . . . . . . . . . . . . . . . . . Olefin Oligomerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
547 550 553 557 563 565 568 570 570 572 574 575 576 577 580 580 583 585 589 589 590 591 594 594 596 597 597 602 602 603
Contents
XVII
9.10 9.11 9.12 9.13 9.14
Copolymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polymerization and Copolymerization of Cyclic Olefins . . . . . . . . . . Diene- and Cyclopolymerization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Polymerization of Polar Monomers . . . . . . . . . . . . . . . . . . . . . . . . . . . Recent Trends in Ligand and Metallocene Catalyst Design . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
10
Chiral Titanocenes and Zirconocenes in Synthesis
603 605 607 610 611 614
A. H. Hoveyda and J. P. Morken 10.1 10.2 10.2.1 10.2.2 10.2.2.1 10.2.2.2 10.3 10.3.1 10.3.2 10.3.3 10.4 10.4.1 10.4.1.1 10.4.1.2 10.4.1.3 10.4.1.4 10.4.1.5 10.4.2 10.4.2.1 10.4.2.2 10.4.2.3 10.4.2.4 10.4.2.5 10.5
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chiral Nonracemic EBTHI Complexes of Group IV Transition Metals. Synthesis and Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . . A Historical Perspective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General Procedure for Catalyst Preparation . . . . . . . . . . . . . . . . . . . . Synthesis of rac-(EBTHI)M Complexes . . . . . . . . . . . . . . . . . . . . . . . Resolution of Racemic EBTHI Complexes . . . . . . . . . . . . . . . . . . . . . Stereoselective Reactions with ansa-Metallocenes as Reagents . . . . . Enantioselective, Zirconocene-Mediated Synthesis of Allylic Amines . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Diastereoselective Titanocene-Mediated Addition of Allylmetals to Aldehydes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Stereoselective Olefin Insertion Reactions with Cationic Zirconocene Complexes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Catalytic Reactions Effected by ansa-Metallocenes . . . . . . . . . . . . . . Catalytic Asymmetric C-C Bond Forming Reactions . . . . . . . . . . . . . Asymmetric Propylene Polymerization . . . . . . . . . . . . . . . . . . . . . . . . Asymmetric Cyclopolymerization of Dienes . . . . . . . . . . . . . . . . . . . Asymmetric Diels-Alder Cycloaddition Reactions . . . . . . . . . . . . . . . Asymmetric Carbomagnesation Reactions . . . . . . . . . . . . . . . . . . . . . Asymmetric Pauson-Khand Type Reactions . . . . . . . . . . . . . . . . . . . . Catalytic Asymmetric C-H Bond Forming Reactions . . . . . . . . . . . . . Asymmetric Hydrogenation of Terminal and 1,1-Disubstituted Alkenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Asymmetric Hydrogenation of Trisubstituted Alkenes . . . . . . . . . . . . Asymmetric Hydrogenation of Imines . . . . . . . . . . . . . . . . . . . . . . . . Asymmetric Hydrogenation of Enamines . . . . . . . . . . . . . . . . . . . . . . Asymmetric Hydrosilation of Ketones . . . . . . . . . . . . . . . . . . . . . . . . Summary and Outlook . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
625 626 626 626 626 627 630 630 632 635 638 638 638 640 641 645 659 660 660 663 667 672 674 678 679
XVIII 11
Contents
New Chiral Ferrocenyl Ligands for Asymmetric Catalysis A. Togni
11.1 11.2 11.2.1 11.2.2 11.3 11.3.1 11.3.2 11.3.3 11.3.4 11.4 11.5 11.6 11.7
12
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synthetic and Structural Aspects of New-Generation Ferrocenyl Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Synthesis of Ligands Involving Diastereoselective Lithiation of Chiral Ferrocenes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Coordination-Chemical and Conformational Characteristics . . . . . . . Applications of New Ferrocenyl Ligands in Asymmetric Catalysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Industrial Applications in Hydrogenation Reactions . . . . . . . . . . . . . . Rhodium-Catalyzed Hydroboration of Olefins . . . . . . . . . . . . . . . . . . Palladium-Catalyzed Allylic Amination Using Pyrazole Containing Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applications of TRAP Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Additional Ligands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . New Methods for the Synthesis of Planar Chiral Ferrocenes . . . . . . . Planar Chiral Ferrocenes as Catalysts . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
685 687 687 691 696 696 698 701 709 710 714 717 718 718
Metallocene-Based Polymers T. J. Peckham, P. Gómez-Elipe and I. Manners
12.1 12.2 12.2.1 12.2.2 12.3 12.3.1 12.3.2 12.3.3 12.3.4 12.4 12.4.1
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Side Chain Poly(metallocenes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Organic Polymers with Metallocene Side Groups . . . . . . . . . . . . . . . Inorganic Polymers with Metallocene Side Groups . . . . . . . . . . . . . . Main Chain Poly(metallocenes) with Short Spacers . . . . . . . . . . . . . . Poly(metallocenylenes) and Poly(metallocenes) with Short Spacers via Condensation Routes . . . . . . . . . . . . . . . . . . . . . . . . . . . . Ring-Opening Polymerization (ROP) of Strained Metallocenophanes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Atom-Abstraction-Induced Ring-Opening Polymerization of Chalcogenido-Bridged Metallocenophanes . . . . . . . . . . . . . . . . . . Face-to-Face Poly(metallocenes) . . . . . . . . . . . . . . . . . . . . . . . . . . . . Main Chain Poly(metallocenes) with Long Spacers . . . . . . . . . . . . . . Insulating Spacers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
723 725 725 730 733 733 739 749 750 752 752
Contents
12.4.2 12.5 12.6
XIX
Conjugated Spacers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Star and Dendritic Poly(metallocenes) . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
761 763 765 766
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
773
List of Contributors
Rüdiger Beckhaus Department of Inorganic Chemistry Aachen Technical University D-52056 Aachen Germany John E. Bercaw Arnold and Mabel Beckman Laboratories of Chemical Synthesis California Institute of Technology Pasadena, CA 91125 USA
Frank T. Edelmann Department of Chemistry Otto-von-Guericke-University Magdeburg Universitätsplatz 2 D-39106 Magdeburg Germany Paloma Gómez-Elipe Department of Chemistry University of Toronto 80 St. George Street Toronto, ON M5S 3H6 Canada
Neil Burford Department of Chemistry Dalhousie University Halifax Nova Scotia B3H 4J3 Canada
Ronald L. Halterman Department of Chemistry and Biochemistry University of Oklahoma Norman, OK 73019 USA
Paul J. Chirik Arnold and Mabel Beckman Laboratories of Chemical Synthesis California Institute of Technology Pasadena, CA 91125 USA
Amir H. Hoveyda Department of Chemistry Merkert Chemistry Center Boston College Chestnut Hill, Massachusetts USA
Thomas Eberle Departmet of Inorganic Chemistry and Analytical Chemistry Johannes-Gutenberg-University Mainz Johann-Joachim-Becher-Weg 24 D-55099 Mainz Germany
Christoph Janiak Department of Inorganic and Analytical Chemistry Freiburg University Albertstraße 21 D-79104 Freiburg Germany
XXII
List of Contributors
Peter Jutzi Department of Chemistry University of Bielefeld Universitätsstraße 25 D-33615 Bielefeld Germany Ian Manners Department of Chemistry University of Toronto 80 St. George Street Toronto, ON M5S 3H6 Canada Jean-Luc Montchamp Department of Chemistry Purdue University West Lafayette, Indiana 47907 USA James P. Morken Department of Chemistry University of North Carolina Chapel Hill North Carolina USA Ei-ichi Negishi Department of Chemistry Purdue University West Lafayette, Indiana 47907 USA
Jun Okuda Department of Inorganic Chemistry and Analytical Chemistry Johannes-Gutenberg-University Mainz Johann-Joachim-Becher-Weg 24 D-55099 Mainz Germany Timothy J. Peckham Department of Chemistry University of Toronto 80 St. George Street Toronto, ON M5S 3H6 Canada Pascual Royo Department of Inorganic Chemistry Alcalá University 28871 Alcalá de Henares Spain Evelyn Ryan Department of Inorganic Chemistry Alcalá University 28871 Alcalá de Henares Spain Antonio Togni Laboratory of Inorganic Chemistry Swiss Federal Institute of Technology ETH-Zentrum CH-8092 Zürich Switzerland