Advances in
Physical Organic Chemistry
This Page Intentionally Left Blank
Advances in
Physical Organic Chemistry ...
118 downloads
1304 Views
13MB Size
Report
This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form
Advances in
Physical Organic Chemistry
This Page Intentionally Left Blank
Advances in
Physical Organic Chemistry Edited by
V. GOLD Department of Chemistry King's College, University of London
VOLUME 9
1971
Academic Press, London and New York
ACADEMIC PRESS INC. (LONDON) LTD 24-28 Oval Road, London NW17DX
U.S. Edition published by ACADEMIC PRESS INC. 111 Fifth Avenue, New York, New York 10003
Copyright
01971 By Academic Press Inc. (London) Ltd
All Rights Reserved
No part of this book may be reproduced in any form by photostat, microfilm, or any other means, without written permission from the publishers
Library of Congress Catalog Card Number: 62-22125 ISBN 0-12-033509-3
PRINTED IN GREAT BRITAIN BY WILLIAM CLOWES & SONS LIMITED LONDON, COLCHESTER AND BECCLES
CONTRIBUTORS TO VOLUME 9 R. J. GILLESPIE,Department of Chemistry, McJfaster University, Hamilton, Ontario, Canada. G. MODENA,Centro C.N.R. “Meccanismi d i Reazimi Organiche”, Istituto di Chimica Organica, Universitd d i Padova 35100, Padova, Italy. T. E. PEEL,Department of Chemistry, McMaster University, Hamilton, Ontario, Canada. F. RAMIREZ,Department of Chemistry, State University of New York at Stony Brook, Stony Brook, New York 11790. M. SIMONYI,Central Research Institute for Chemistry of the Hungarian Academy of Sciences, Budapest, Hungary. U. TONELLATO, Centro C.N. R. ‘ ‘Meccanismi d i Reazioni Organische’ ’, Istituto d i Chimica Organica, Universitd di Padova 36100, Padova, Italy. F. TUDOS, Central Research Institute for Chemistry of the Hungarian Academy of Sciences, Budapest, Hungary. I. U ~ I Department , of Chemistry, University of Southern California, Los Angeles, California 90007.
This Page Intentionally Left Blank
CONTENTS CONTRIBUTORSTO VOLUME 9
.
.
v
Superadd Systems R. J. GILLESPIEand T. E. PEEL I. Introduction . 11. H2S04Systems . A. 100%H2S04 B. H,S04--SO3 . C. H2S04-HB(HS04)4 . . 111. HS03F and HS03C1Systems A. HZS04-HSO3F B. HS03F-SOs C. HS03F-MFb and HS03F-MF,-S03 . D. H~S04-HSO3Cl IV. HFSystems . A. H20-HF . B. Lewis Acids in H F V. Applications . A. Protonation Studies . B. New Cations References .
.
.
.
I
. .
4 4
.
6
. .
9 9
.
9 10 11 15 15 15 16 17 17 19 23
. . . . . . . . . .
Turnstyle Rearrangement and Pseudorotation in the Permutational Isomerization of Pentavalent Ph0 sphorus Compounds FAUSTO RAMIREZ and IVAR U ~ I I. Introduction . . 11. The Pentavalent State in Phosphorus Stereochemistry . A. General Considerations . . B. Molecular Structure of Oxyphosphoranes from X-Ray Analysis . , 111. Formal Analysis of Permutational Isomerizations of Penta. . valent Phosphorus A. Notation of Isomers . . B. Formal Mechanisms . . vii
26 27 27 29 35 35 38
viii
CONTENTS
. IV. Berry Pseudorotation Mechanism (BPR) . V. Turnstile Rotation Mechanism (TR) A. Schematic Representation of the Single TR . B. Description of the Single TR Mechanism . C. Itinerary for Permutational Isomerizations by Single TRandbyBPR . D. The Multiple TR E. TheTRSwitches . VI. Calculation of Binding Energies of Model Situations in Turn. stile Rotation and in Pseudorotation A. General Considerations . B. Calculations for PF5 . C. Calculations for Other Systems . VII. Comparison Between TR and BPR . A. Common Features. B. Differences . VIII. Survey of Experimental Data . A. Caged Polycyclic Oxyphosphoranes B. Variable Temperature Proton N.M.R. Spectra of Phosphoranes . IX. Irregular Permutational Isomerizations of Compounds with . Pentavalent Phosphorus A. Irregular Processes with Decrease in Coordination Number . B. Irregular Process with Increase in Coordination . Number X. Conclusions
42 44 44 47 50 52 55
58 58 60 63 72 72 72 73 73 82 116 115 118 120
The Hydrogen Atom Abstraction Reaction From
0-H Bonds M. SIMONYI and F. T i i ~ d s I. Introduction . A. Characterization of the Hydrogen Atom Abstraction . Reaction B. Limitations of the Scope . C. Possible Experimental Techniques for Kinetic Studies . 11. Kinetic Studies on 0-H Bond Pission . A. Studies on Phenols : The Kinetic Isotope Effect . B. Studies on Phenols : The Substituent Effect . C. Studies on Phenols : The Steric Effect . D. Studies on Other Hydroxylic Compounds
127 128 130 131 135 136 144 150 164
CONTENTS
111. The Role of Hydrogen Bonding . A. Influence of the Medium on the Rate . B. The Hydrogen-Bonding Equilibrium . C. Consideration of Hydrogen Bonding in the Interpretation of Certain Kinetic Anomalies . IV. Arrhenius Parameters . A. Interrelation Between the Arrhenius Parameters . B. Formal Interpretation of the Compensation Phenomenon . V. Conclusion . References .
ix 157 158 160 164 167 167 172 173 174
Vinyl Cations GIORGIOMODENAand UMBERTO TONELLATO I . Introduction . . 11. Sources of Vinyl Cations A. Electrophilic Addition to Acetylene Derivatives . B. Electrophilic Addition to Allene Derivatives . C. Heterolytic Fission of Bonds Attached to a Vinyl . Carbon Atom D. Electron Removal from Neutral Species . . 111. General Properties of Vinyl Cations A. Geometry and Stability . B. Reactions of Vinyl Cations IV. Related Species . A. Propargyl Cations B. NitriliumIons . C. ImminiumIons . D. Acyl Cations. References .
185 186 187 215 231 253 254 254 265 267 267 270 272 273 274
This Page Intentionally Left Blank
SUPERACID SYSTEMS R. J. GILLESPIE
AND
T. E. PEEL
Department of Chemistry, HcHmter University, Hamilton, Ontario, Canada I. Introduction II. HaSO4Syetems
. .
. .
.
A. 100~oHaS04 B. H s S 0 4 4 O s C. HaS04-HB(HS04)4 111. HSOaF and HSOsCl Systems A. HsSO4-HSOsF B. HSOsF-SOa C. HSOsF-MF5 and HSOSF-MF~-SOS D. HaS04-HSOsC1 IV. HFSystems A. HsO-HF B. LewisAcidsinHF v. Applicstions A. Protonetion Studies B. Newcations References
.
.
. . . . .
I
.
.
.
.
. .
.
. .
.
. .
. .
.
. .
.
. . . . . .
.
.
.
.
.
.
.
.
.
. . .
. .
. .
1 4
.
4
. . . .
6 9 9 9 10 11 16 16 16 16 17 17 19
. . . .
. . . .. . . . . .
.
.
23
I. INTRODUCTION INrecent years a considerable amount of fascinating new chemistry has been carried out in several highly acidic systems. This is exemplified by the preparation of stable solutions of aliphatic mrbonium ions (Olah et d.,1965), the formation of a variety of new inorganic cations, e.g., the 12 and Se:+ cations (Gillespie, 1968), and the protonation of some unusual very weak bases such as carbonic acid to give C(OH)$ (Olah and White, 1968). There has, however, been rather little fundamental information available on these systems and in particular on one of their most important and fundamental properties, namely their acidity. In this review we discuss the properties of systems which have acidities comparable to, or greater than, that of 100% H,S04. We may define the acidity of a medium as its tendency to donate a proton to a base. I n dilute aqueous solutions acidity is measured precisely and quantitatively by the pH but in concentrated aqueous solutions or in non-aqueous systems the concept of pH is not applicable and no alternative precise and quantitative measure of acidity is generally 1
2
R . J . GILLESPIE AND T . E . PEEL
available. The most useful measure of the acidity of such a medium has proved to be the Hammett acidity function 11, which may be defined by the equation (1) Ho = #BE+
CBH+l -log -
(1)
PI where pKBH+ is the acid dissociation constant of the conjugate acid of a base B and [BH+]/[B] is the ionization ratio which has generally been measured spectrophotometrically. By commencing with a base, of known pKBH+whose ionization ratio could be measured in dilute aqueous sdphuric acid, and by using a set of successively weaker bases whose ionization ranges overlapped each other, Hammett and Deyrup (1932) were able to obtain H , values over the whole composition range for the system H20-H2S04. The establishment of an acidity scale in this manner implies the assumption that the activity coefficient ratio (fcfBa+)/(fcH+fB) for two successive bases B and C has the value of unity. This approximation is likely to be more valid if all the bases used are of the same structural type. This was not the case for some of the bases used by Hammett and Deyrup, and recently their early data have been revised and improved by use of an extended and more closely related set of bases consisting only of primary anilines (Jorgenson and Hartter, 1963; and Johnsonetal., 1969). Thelatest H,valuesfortheH20-H2S0, system are given in Table 1. Unfortunately other types of bases, e.g., TABLE1
Ho Values for HzO-HZSO~ System at 26°C (Johnsonet al., 1969)
2 4
6 8 10 12 14 16 18 20 22 24 26 28 30 32 34
0.53 0.20 0.02 0.20 -0.36 0.50 -0.65 -0.78 0.92 1-06 1.20 1-34 1.47 1.60 - 1.73 1.85 - 1.99
-
-
36 38 40 42 44 46 48 60 62 54 66 58 60 62 64 66 68
- 2-12 - 2.27 - 2.42
- 2.60
- 2.77 -2-98 -3.12 3.30 - 3.48 - 3-88 3.90 4.13 4.37 4.62 -4.91 - 5.20 - 5.52
-
-
70 72 74 76 78 80 82 84 86 88 90 92 94 96 98
- 6.82
- 8-13 - 6.47 -6.81
- 7.13 - 7.46
- 7.80 - 8.13 - 8.42
-8.71 9.01 - 9.34 - 9-66 9.98 - 10.43
-
-
SUPERAOID SYSTEMS
3
amides, do not always give the same acidity function values and indeed a number of other acidity function scales have been proposed for such bases. We do not intend to discuss the relative merits of these scales here; we only wish to emphasize that, although the H,, values cannot be regarded as having any absolute significance, they are nevertheless extremely useful as they give at least a semi-quantitative measure of the relative acidities of different highly acidic media. H , values for various H20-HX systems are given in Fig. 1. The data for these systems have been reviewed by Paul and Long (1957). The relative acidities of the concentrated solutions of these acids are, in general, qualitatively the same as the relative strengths of the acids in dilute aqueous solution, although HF, which is rather weak in dilute
//
6t
H20
0.2
0.4 0.6 Mole fraction HA
FIQ.1. Mole fraotion HA.
aqueous solution, approaches the acidity of 100% H2S04in very concentrated solutions. Perchloric acid gives the highest acidities at low concentrations but for obvious reasons more concentrated solutions have not been studied. In dilute aqueous acid systems the acidic species is H,O+, and the acidity is given by the H,O+ concentration or the pH. In sufficiently dilute solution the HsO+ion is hydrated by at least three water molecules, i.e., H,O(H,O)$ or H+(H20)4,but as the concentration of the acid HX is increased the number of water molecules available to hydrate the H80+decreases and the extent of hydration of H,O+eventually decreases accordingly. As the extent of hydration of H,O+ decreases the hydrated complex becomes smaller, its positive charge therefore becomes more concentrated, and its acidify is therefore expected to increase. Consequently the acidity of an HX-H,O system increases more rapidly than
4
R . J . OILLESPIE A N D T . E . PEEL
the HaO+concentration. At still higher H X concentrations there is not enough water to ionize all the acid and molecular HX is then present in the system. As this is a stronger proton donor than H30+the acidity of the system increases further and this continues up to lOOyoHX.
11. H&04 SYSTEMS A. lOOyoH,S04 The most acidic system investigated over the whole compositionrange H20-HX is the sulphuric acid system. Following the work of Hammett and Deyrup (1933) and Treffers and Hammett (1937) and the earlier pioneer work of Hantzsch (1907-1909), concentrated sulphuric acid and 100% H2S04have become the most extensively studied and wellunderstood acidic media and thus have received rather wide application in both preparative and mechanistic studies (Gillespie and Robinson, 1966; Gillespie, 1968). Some of the properties of 1 0 0 ~ oH2S04are summarized in Table 2. Although it has been the best known and most widely used of the superTABLE 2 Some Physical Constants of Sulphuric Acid Freezing point Boiling point Density (26°C) Viscosity (26°C) Dielectric constant (26%) Specific conductance (26OC)
10.371"C 290-317°C 1.8269 g cm-3 24.64 centipoise 100 1.0439 x 10-2 ohm-1 cm-1
acid media some of its properties as a solvent and reaction medium are less than ideal. I n particular its high viscosity is inconvenient in preparative reactions as solid products are difficult to flter from the acid solvent although adhering sulphuric acid solvent can often be conveniently removed by washing with liquid sulphur dioxide. Another disadvantage of lOOyoH2S04is its relatively high freezing point of 10.37' which precludes the study of solutions at low temperatures. Thus for the study of protonations a t temperatures at which proton exchange is relatively slow, fluorosulphuric acid, with a freezing point of - 89", has proved much superior. Hammett and Deyrup (1932) gave H , values up to 100% HzS04 but their values at the highest concentrations were based on the use of 2,4,6-trinitroaniline as an indicator. The basicity of this indicator is
5
SUPERACID SYSTEMS
really too great to enable it to be used with any accuracy beyond approximately 99-5% HzS04. As no significantlyweaker primary aniline base can probably be obtained it is necessary to turn to another type of indicator in order to extend the acidity scale to lOOyo H2S04 and to media with still higher acidities. The indicators which have been used for this purpose are a set of aromatic nitrocompounds which, when suitably substituted, have basicities that enable a rather wide range of acidities to be measured. Almost all the H o values quoted in this review are based on the use of these nitro compound indicators. The ionization ratio data for the most basic nitrocompound indicator used, p-nitrotoluene, appears to overlap the data for the least basic aniline, 2,4,6-trinitroaniline, in a satisfactory manner. Thus the plots of ionization ratios against solvent composition for the two indicators are parallel and lead to a constant value for the pKBH+ of p-nitrotoluene. The pKBH+ values along with the extinction coefficients of the protonated ( eBH+)and unprotonated ( cB) forms of a number of weak nitrocompound bases are shown in Table 3. The measured ionization ratios for these TABLE3 ~ K B HValues + of Aromatic Nitrocompound Indicators Indicator 2,4,6-Trinitroaniline p-Nitrotoluene rn-Nitrotoluene Nitrobenzene p-Nitrofluorobenzene p-Nitrochlorobenzene m-Nitrochlorobenzene 2,4-Dinitrotoluene 2,4-DiNtrofluorobenzene 2,4,6-Trinitrotoluene 1,3,6-Trinitrohnzene (2,4-Dinitrofluorobenzene)H+
PKBH+
- 10.10 11.35 - 11.99 - 13.14 - 12.44 - 12.70 - 13.16 - 13.76 - 14.52 - 15.60 - 16.04 - 17*57*
-
~BH+
CB
200 19,200 16,100 16,150 16,500 24,350 14,380 13,900 12,100 10,600 10,700 20,000*
8,500 2,100 1,200 900 600 400 440 1,700 920 960 800 1,000
* Estimated. bases over a wide range of acidity indicate that they form a satisfactory consistent set of indicators. The new indicators have been used to obtain a value of €i0for 100% H,S04 which is -11.93 (Gillespie et al., 1971). Although the acidity function scale based on the use of the nitrocompounds may not form a completely satisfactory extension of the scale based upon the anilines, the use of these indicators does give a t least a semi-quantitative picture of the acidity of various superacid systems. Moreover the H o values for
6
R . J . BILLESPIE AND T. E . PEEL
superacid media obtained from the nitrocompound indicators are in satisfactory qualitative agreement with such other methods as are available for determining the acidity of these media. Although 100% H2S04,according to its Ha value, is approximately 1013 times as acidic a8 a 0 . 1 aqueous ~ HzS04solution, it is by no means the most acidic medium that can be obtained. One method of obtaining a more acidic medium, at least in principle, is to dissolve in H2S04a substance HA that is capable of protonating HzS04,i.e., an acid of the sulphuric acid solvent system, HA+HaS04
z? HaSOt+A-
(2)
Some of these systems plus other equally acidic systems are described in this review. We may somewhat arbitrarily classify 100% HzS04and other non-aqueous systems containing considerably more acidic species than the hydrated proton as superacid systems as their acidities are of a different order of magnitude from that encountered in the more familiar aqueous systems.
B . H2SO4-SO
3
One of the earliest of the superacid systems to be recognized was oleum, i.e., sulphuric acid containing an exce8s of sulphur trioxide; this has often been used when a very powerful acid catalyst was required. A number of fundamental studies have established that it contains a series of polysulphuric acids H2Sn0312+1 which increase in chain length with increasing sulphur trioxide concentration (Gillespie and Robinson, 1962). The main component of dilute oleums is disulphuric acid, HZS2O7, and it has been shown that this ionizes as a moderately strong acid of the HzS04solvent system (Bass et al., 1960). HzSz07-I-H~zS04
HaSO:+HSzOT,
Ka = 0.014
(3)
There is therefore a substantial increase in the concentration of the H3S02 ion in such a medium and hence a corresponding increase in acidity. On increasing the concentration of SO3in an oleum, the amount of HzS3Olo and still higher polysulphuric acids increases significantly. At the composition H2SzO7 there are extensive amounts of HZS3O10 and higher polysulphuric acids and of course a corresponding amount of H2S04 (Gillespie and Malhotra, 1967). The extent of ionization of a number of weakly basic nitrocompounds has been determined by cryoscopic and conductimetric measurements in oleum having the composition H2S2O7(Gillespie and Malhotra, 1968). The ionization constants determined by these methods are compared with those found for the
SUPERACID SYSTEMS
7
same bases in H2S04in Table 4. It may be seen that these nitrocompounds are much more extensively ionized in H2S2O7 than in H2S04 and it may therefore be concluded that H2S2O7 is a oonsiderably more acidic medium than H2S04. TABLE4 Comparison of K , Values for Some Weak Bases in HzS04, HSOsF end HzSz07 (Gillespie and Melhotre, 1968)
p-Nitrotoluene m-Nitrotoluene Dimethylsulphone Nitrobenzene p-Nitrochlorobenzene m-Nitrochlorobenzene Nitromethane 2,4-Dinitrotoluene m-Dinitrobenzene 2,4-Dinitrochlorobenzene 2,4-Dinitrofluorobenzene
9.5 2.3 1.5 1.0 0.4 0.25
0.06 0-03 Non-electrolyte Non-electrolyte
Fully ionized Fully ionized 8 Fully ionized 76 7.9 2.7 1-4 0.26 0.16 0.16
Fully ionized Fully ionized Fully ionized 66 Fully ionized Fully ionized 26 10 10
Values of the Hammett acidity function for these systems were first given by Lewis and Bigeleisen (1943) who proposed that for fuming sulphuric acid the acidity function could be determined from the vapour pressure of SO3above the solution. However, as pointed out by Paul and Long (1957), there is no particular reason to expect the acidity to be related to the SO3 pressure. The acidity function measurements of Brand (1950) and Brand et al. (1952) confirmed, by the use of aromatic nitrocompound indicators, that the acidity does indeed increase with increasing concentration of SO3 but Brand's H , values do not appear to be related to the vapour pressure of SO3. The H,, values determined by Brand were based on an earlier value of H , for 100% H2S04which has now been superseded. Also he was, in several cases, unable to obtain the nitrocompound indicators in the fully ionized form, which led to errors in the determination of the ionization ratios. Consequently his values have been revised and the latest values are listed in Table 5 (Gillespie et at., 1971) and are also shown in Fig. 2. Although an increase in acid strength would be expected with increasing chain length in the polysulphuric acids, this would be most marked
8
R . J. OILLESPIE A N D T. E. PEEL
TABLE5 Ho Valuea for the Systems HzO-HzSOa HzS04-HB(HS04)4,
Mole% HzO
HaS04-HS03C1,
(nesr lOOyo HzSO~),HzS04-S08, HzSO~-HSO~F,HzSO4-HaSzO7
HzO-HaS04 Ho Mole% HzO
wt%
10.00 9.07 8-13 7-17 6.20 5.21
98.0 98.2 98.4 98.6 98.8 99.0
1.00 2.00 5-00 10.00 15.00 20.00
- 12.24 - 12.42 - 12.73 13.03 - 13.23 - 13.41
-
- 10.44 - 10.50 - 10.56 - 10.62 - 10.71 - 10.84
4.20 3.18 2.14 1.08 0.54
25.00 30.00 35.00 40.00 46.00 50.00
- 13.58 - 13.76 - 13.94
- 12.35 - 12.60 - 12.87 - 13.03 - 13.15
1.00
2.00 4-00 6.00 8.00
1.00 2-00 5.00
10.00 20.00 30-00 40.00 50.00 60.00 70.00 80.00 90.00 95.00 97.50 99.00 100~00
- 12.23 - 12.41 - 12.71 - 12.98 - 13.28 - 13.52 - 13.71 - 13.88 - 14.03 - 14-17 - 14.26 - 14.36 - 14.40 - 14.42 - 14.43 - 14.44
0.0
-14.11
- 14.28 - 14-44
wt%&so4 99.2 99.4 99.6 99.8 99.9 100.0
55.00
60.00 65.00 70.00 75.00
- 13.23
10.00 15-00 20.00
- 13.39 - 13-49 - 13.56 - 13.62
25.00
30.00
- 12.02 - 12.09 - 12.25 - 12.41 - 12.60 - 12.75 - 12.86 - 13.02 - 13-14 - 13.28 - 13.47 - 13.73 - 13.95 - 14.14 - 14.21 - 15.07
- 11.98
- 12.03 - 12.15 - 12.28 - 12-47 - 12.61
- 12.72
- 12.84 - 12-94 - 13.06 - 13.19 - 13.38 - 13.54 - 13.66 - 13.74 - 13.80
Ho
- 10.93
- 11.05 - 11.15
- 11.40
- 11.64 -11.93
- 14.59 - 14.74 - 14.84 - 14.92 - 14-96
9 between the first few members of the series and probably becomes negligible from H284013 onwards. It seems reasonable to suppose that in very concentrated oleums, as the concentration of protons decreases, the acidity must again fall off to low values. However, no data are available for this region of the system. SUPERACID SYSTEMS
C. H,S04--HB( HSO 4) 4 The only known strong acid of the sulphuric acid system is tetra(hydr0gensu1phato)boric acid, HB(HS04)4,which may be obtained in solution in sulphuric acid by dissolving boric acid in oleum in the amount required by equation (4) (Flowers et al., 1956). H3BOa+3HzSz07 -+ HB(HS04)4+2HzS04
(4)
This acid is a strong acid that is extensivelyalthoughnot quite completely ionized HB(HS04)4+HzS04 Z? HzSO: +B(HSO4),
(5)
Table 5 and Fig. 2 show that H o for HB(HS04), solutions increases more rapidly than for solutions of either HS03For H2S2O7, which is consistent with this acid being a stronger acid than either HS03F or HzSzO7 (Gillespie et al., 1971). The acidity that can be obtained in solutions of HB(HS04)4is however limited by the separation of insoluble complex polysulphatoboric acids from more concentrated solutions.
111. HSOsF AND HSOsCl SYSTEMS A. H2SO4-HSOsF Cryoscopic and conductimetric measurements on dilute solutions of HSOsF in H2SO4 have shown that HSOsF behaves as a rather weak acid of the H2SO4 solvent system (Ban et al., 1961) ionizing to a small extent according to equation (6) : z?HaSO:+S03F-, K. = 3 x 10-3 (6) Hence one may reasonably conclude that addition of fluorosulphuric acid to sulphuric acid increasesthe acidity of the medium and this increase might be expected to continue throughout the H2S04-HSOsF system up to the composition 100% HS03F. Cryoscopic and conductimetric studies (Barr et al., 1964; Gillespie et al., 1966)in 100% HSOsF as solvent have confirmed that this is the case as a aeries of nitrocompound weak bases are more extensively ionized in HSOsF than in H2SO4 (Table 4). Values of H o for the H2SO4-HSOsP system have been obtained (Gillespie et al., 1971) using the bases nitrobenzene, p-chloronitrobenzene, mnitrobenzene and 2,4-dinitrotoluene and these are summarized in Table HSOaF+HzS04
10
R . J . CIILLESPIE A N D T . E . P E E L
6 and Fig. 2. They show that addition of HSOsF to H2S04produces a fairly rapid increase in H o initially, although this is less than that produced by H2S2O7, which is consistent with the fact that H2S2O7 is a stronger acid. This is followed by a slower steady increase in acidity and another very rapid increase in the region of 100% HSOsF. The rapid increase in the vicinity of 100% HS03Fis presumably to be associated with the self-ionization of HS03F. 2HSOsF
Z?
HzSOsF+ + SOIF-
(7)
15.014.5
-
-
14.0 13.5
- Ho
-
13.012.5
-
12.0
-
11.5-
I YOE%
HzO
00%
Hz-4
I
I
10
20
1
30
I
I
I
40 50 60 MOLE % H A
I
70
I
80
I
90
I 0
FIQ.2. Acidity function values for the systems R~S04--HB(HS04)4 (A), Hss04SO3 (0), HsS04-HS03F (01,HzSO~-HSO~C~ (V)-
As the concentration of the very weakly basic H28O4decreases to a very low value and the concentration of S03F- decreases accordingly, so the concentration of the highly acidic H2SOsE'+ ion increases rapidly up to the value that it has in 100% HSOsF. Some of the properties of fluorosulphuric acid are summarized in Table 6. Fluorosulphuric acid has a boiling point of 163.7"C and it is therefore easily removed by distillation from a reaction mixture ; it is also readily purified by distil-
SUPERACID SYSTEMS
11
TABLE6 Physical Properties of Fluorosulphuric Acid Freezing point Boiling point Density (25°C) Viscosity (25OC) Specific conductance (25OC)
- 88.98"C 182.7"C 1.728 g cm-3 1.58 oentipoise 1.085 x 10-4 ohm-1 cm-1
lation although the last traces of sulphur trioxide are difficult to remove completely. When free from HF, fluorosulphuric acid does not attack glass and is thus easily and conveniently handled in conventional glass apparatus. Although the acidity of HSOsF is not much greater than that of HzS2O7it is generally a much more useful highly acidic medium (Gillespie, 1968) and there are two main reasons for this. Firstly, its freezing point is low (-89'C). This has proved to be particularly important in the study of the protonation of weak bases by proton n.m.r., as many proton exchange reactions with the solvent are very slow at this temperature. Secondly, it is a considerably poorer sulphonating agent towards organic compounds than is HzSzO,, particularly at low temperatures, and, in general, stable solutions of many more substances can be obtained in HS03F than in H2SzO7. It is clear that by increasing the concentration of HzSOsF+ by the addition of an acid of the HSOSFsolvent system, which ionizes according to equation (€9, HA +HSOsF
a HaSOaF++A-
(8)
the acidity of the system could be further increased. No simple acids are known as all the common protonic acids are non-electrolytes or bases in solution in HS03F. However some Lewis acids act as fluorosulphate ion acceptors and hence increase the HzSOSF+concentration.
B. HSOJi'-SOs Sulphur trioxide causes a marked increase in the acidity of H2S04 because of its ionization as a moderately strong acid of the sulphuric acid solvent system. Freezing point measurements (Gillespie et al., 1966) show, however, that SO3 behaves as a non-electrolyte in solution in HSOSF, and this is confirmed by the observation that SO3 causes a negligible increase in the conductivity. On the other hand, the Raman spectra of solutions of SO3in HS03Fcontain new lines that are not due to either SO3 or HS03F and it has been shown by Gillespie and Robinson (1962) that they can be attributed to the acid HSZO6F(1). It must be concluded that, in contrast to the behaviour of HzSzO7 in H2S04,
12
R . J . G I L L E S P I E A N D T . 1. P E E L
the acid HSzOBFis not a s&ciently strong acid of the HSOSF solvent system to ionize to an appreciable extent. This is confirmed by the very small increase in - H o produced by S 0 3 in HSOaF (Fig. 4).
C. HS03F-XF, and HS03F-MFb-S03 Conductivity and cryoscopic measurements (Thompson et al., 1965 ; Gillespie et al., 1969) on solutions of several pentafluorides in HSOsF have shown that PF, and NbF, are non-electrolytes, AsF, and BiF, are very slightly ionized and SbF, is a moderately strong acid SbFs +2HSOsF
+
HzSOsF+ SbFs(S0aF)-
2%I
(9)
I
MOLE RATIO
FIO.3. Titration of SbFs with KSOaF
KSOsF/SbF,m SbFZ(SO8h ( 0 ) and
of SbFs 3so3 with KS03F (17).
SUPERACID SYSTEMS
13
The ions SbF6SOsF- and AsF6SOsF- have the octahedral structure (2). The fact that SbF6behaves as an acid rather than as a base has been confirmed by conductimetric titrations with a fluorosulphate (Fig. 3) (Thompson et al., 1965; Gillespie et al., 1969). The concentrations of H2SOsF+and SOSF- are easily followed in such titrations because these two ions have considerably greater mobilities than other ions (Barr et al., 1964). Thus the conductivity decreases on addition of SOSF- to an SbF, solution because the highly conducting H2SOSF+ion is replaced by the considerably less mobile K+ ion. Since H[SbF,(SO,F)] is not a strong acid of the HSOsF solvent system, the conductivity passes through a minimum before the end point corresponding to the mole ratio KSOsF/HISbF,(SO,F)]= 1. A slow increase is followed by a sharp break a t the 111 end point, as the excess SOSF- ion produces a marked increase in the conductivity. H[SbFe(SOsF)]+KSOsF + K[SbFs(SOsF)]+HSOsF
(10)
Values for the acidity function H , of the HSOBF-SbF6 and HS0,FAsF6 systems up to the mole ratio 110.25 are given in Table 7 and Fig. 4 (Gillespie and Peel, 1971). We see that there is a very marked increase in the acidity on the addition of SbF, to HSOsF which corresponds to the TABLE 7 HSOsF-MF5, Ho Values Ho
Ho
Mole% MFs
ASF5
SbF5
0.00
- 16.07 - 16.22
- 16.07 - 16.98 - 16.66 - 16.89 - 17.04 - 17.26 - 17.37 - 17.46 - 17-68 - 17.80
0.06 0.10 0.16
0.20 0.30 0.40
0-60 0.76 1.00 1.60 2-00
- 16.36 - 16.48 - 16.66 - 16.70
- 16.80 - 16-88
- 16.03 - 16.12 - 16.24
- 16.31
- 17.98
- 18.10
Mole%MF5 2.60 3.00 3.50 4-00
6-00 8.00 10.00 12.00 14.00 16.00 18-00 20-00
AsF5
- 16.37 - 16.42 - 16.48
- 16.62
SbF5
- 18.19 - 18.31 - 18.36 - 18.43 - 18.67 - 18.60 - 81.26
- 18.63 - 18.66
- 18.69 - 18.62 - 18.66
very rapid increase in the concentration of the very highly acidic H2SOsF+ion. AsF6 gives a much smaller increase in acidity consistent with its behaviour as a very weak acid. At a concentration of SbF6 of approximately 5 mole% the acidity function curve flattens out and higher concentrations of SbF, produce only a very small further increase in the acidity. This is consistent with the conclusion from cryoscopic,
14
R . J . Q I L L E S P I E A N D T. E . P E E L
MOLE % MF, or SO, FIG.4. Acidity function values for the systems. HS03F-SbFb ( 0), HSOsF-AoFs (O), HSOsF-SOs ( A ) .
conductimetric, and n.m.r. measurements (Thompson et al., 1965) that with increasing concentration much of the SbF, forms a dimeric anion according to equation (1 1). SbFa + SbFs(S0sF)- + (SbFs)zSO&’-
(11)
Although H[(SbF,),SO,F] is probably a somewhat stronger acid than H[SbF,(SO,F)] there is no direct increase in the concentration of H2S03F+.The anion (L‘~~F,)~SO,Fhas the structure (3) involving a bridging fluorosulphate group. r-
1
SUPERACID SYSTEMS
16
When sulphur trioxide is added to a solution of SbF, in HS03F there is a marked increase in the conductivity which continues until approximately three moles of SO3 have been added per mole of SbF, originally present (Thompson et al., 1965). This increase in conductivity must be due to an increase in the concentration of the highly conducting H2S03F+ ion arising from the formation of a stronger acid than H[SbF6(S03F)] in the system. From the results of these measurements, combined with cryoscopic and I9Fn.m.r. studies, it has been concluded that a series of acids H[SbF,(SO 3F)2], H [SbF 3( HSO3F)3], and H(SbF2(SOSF) are formed which increase in acidic strength through the series. Dimeric ions such as (4) are also formed at high concentrations. The acid H[SbF2(S03F)4]appears to be a fully ionized strong acid of the HS03F solvent system. This can be shown for example by a conductimetric titration of a sohtion of the acid with a base such as KS03F. The conductivity decreases on addition of KS03Fand passes through a minimum at the mole ratio KS03F/H[SbF2(S03F)4]=1 as expected for a strong acid (Fig. 3). No H o measurements have yet been made on these systems ; however it is expected that the acidity will be somewhat greater than that found in the HS03F-SbF6 system. O N S I F FOzSO-+.O-
F
\O+OBOzF
/ FI --rF F s b /
F
F
(4)
Sulphur trioxide similarly causes an increase in the conductivity of solutions of Asp, in HSOyF (Gillespie et al., 1969). Presumably acids such as H[AsF2(SO3F)l4are formed but they appear to be somewhat weaker than the corresponding antimony acids, just as AsF, forms a weaker acid in HS03F than SbF5. D. H2S04-HS03Cl This system was originally studied by Palm (1956) but his results were based on the older value of the H o of H2S04which has now been revised. New measurements have also been made on this system (Gillespie et al., 1971) and these, together with a recalculation of Palm’s data, lead to the values listed in Table 5. The acidity of this system is very similar to that of the HS03F system with the acidity of HSO3C1being close to that of HS03F.
16
R . J . G I L L E S P I E A N D T . 1. P E E L
IV. HF SYSTEMS A. HZO-HP Hydrogen fluoride is a weak acid in dilute solution in water but H o measurements (Bellet al., 1956) show that in more concentrated solutions the acidity rises rapidly so that 100% HF has an H o value only slightly less than that of 100% HzS04.Confirmationofthis comes from cryoscopic measurements which show that the weak bases water, acetonitrile and p-nitrotoluene are somewhat less ionized in 100% HF than in 100% HZSO4(Gillespie and Humphreys, 1970). The exact value of H o for pure 100% HF is uncertain as it appears to depend strongly on the presence of small amounts of impurities which are very difficult to remove : the most important of these is water which exhibits weak basic behaviour. Thus the measured values of - H o increase from 9.7 t o 11-0 with decreasing impurity concentration, as indicated by the electrical conductivity (Hyman and Katz, 1965). The value of - 11.0 was obtained with an acid that had a conductivity of 3 x lod5ohm-l cm-l at 0'. Presumably, if the last traces of water and other impurities could be removed from HF, the value of - H o would be greater than 11.0, perhaps as high as 11.5. A few measurements have been made on solutions of sodium fluoride, which causes a marked decrease in - H o t and on a 0.02~solution of NbF5, which had H02: - 12.5 (Hyman et al., 1961). TABLE8 Physical Properties of Anhydrous Hydrogen Fluoride Freezing point Boiling point Deneity (0%) Dielectric constant ( O W ) Specific conductance (O'C)
- 89.37OC 19.61"C 1.002 g om-3 84 N 1 x 10-6 ohm-1 om-1
Some of the physical properties of HF are summarized in Table 8. Hydrogen fluoride is a good solvent for a wide variety of organic and inorganic solutes; it has a high dielectric constant, a low viscosity and its low boiling point is an advantage in its use as a preparative solvent. Its major disadvantage as a solvent is of course the fact that it attacks glass. However with the advent of fluorinated plastic materials such as Kel-F and Teflon the use of apparatus constructed from these materials, in conjunction with the use of metal vacuum lines equipped with needle valves containing Teflon and Kel-F, has made the handling and use of anydrous HF a relatively simple matter.
17
SUPERAUID SYSTEMS
B. Lewis Acids in HF It has been shown from cryoscopic (Dean et al., 1970) and conductimetric (Gillespie and Moss, 1966) measurements that SbF6 is a strong acid in HF ionizing in dilute solutions according to equation (13). SbFs+HF + HaF++SbF,
(13)
At higher concentration of SbF, the anions Sb2Ffi (5), Sb3F, (6), etc. are formed with increasing concentrations of SbF6. Presumably, therefore, there is very little increase in acidity when the formation of these ions predominates over the formation of SbF;.
j
+
F--lF
l
T
/+I
/
F
F T -F + ,---JF -
F--Fl
F
Sb
/ + / / F /
b
(6)
Solutions of arsenic pentafluoride have conductivities and freezing point depressions (Dean et al., 1970) that are only approximately onehalf those of SbFbat the same concentration and it has been shown that even in dilute solutions there is essentially complete formation of the As2F, ion according to equation (14).
+
+
2AsF5 2HF = HzF+
l
FI F
F
(5)
S
(14)
These solutions are thus less acidic than those of antimony pentafluoride bemuse the H2F+concentration is lower. PF6appears to behave as a non-electrolyte and it must be concluded that it is too weak a fluoride ion acceptor to enhance the acid properties of hydrogen fluoride (Dean et al., 1970). The strong proton acid HS03F exhibits no appreciable acid properties in HF (Gillespie and Humphreys, 1970). Some other pentafluorides have been placed in order of their acidity on the basis of their effectiveness in increasing the amount of m- and p-xylenes extracted from n-heptane solution into an excess of hydrogen fluoride (McCauleyet al., 1956). It was found that TaF, > NbF6 > TiF4> PF, > PbF, but all these fluorides appear to be weaker acids than SbF, and Asp,. The low solubility of BF3 and the low electrical conductivity of its solutions also suggest that it is a weaker acid than SbF, and AsF, (Kilpatrick and Luborsky, 1964).
18
R . J . B I L L E S P I E A N D T . E. P E E L
V. APPLICATIONS A. Protonation Studies Considerable use has been made of superacid systems, particularly the HS0,F-SbF6 system, for the preparation and identification of the conjugate acids of some very weak bases. The ternary system HS03F-SbFS-S02 has also often been used for such studies. Although SO2 is completely miscible with HS03F-SbF5 mixtures, it is not protonated to any measurable extent and it does not appear to cause any appreciable reduction of the very high acidity of HS03F-SbF,. Sulphur dioxide also considerably reduces the rather high viscosity of SbF6-HS03F solutions and this gives rise to sharper and better resolved n.m.r. spectra. One of the earliest applications of the HS03F-SbF, medium was the first observation of the proton magnetic resonance spectra of the conjugate acids of acetic, propionic and benzoic acids at temperatures of - 70" and lower (Birchall and Gillespie, 1965). At these temperatures proton exchange between the solvent and the conjugate acidis sufficiently slow that the spectra for acetic and propionic acids showed, in addition to the expected peaks for the alkyl groups, two equally intense signals a t very low field due to protons on oxygen. It was concluded that these arise from structure (7). More recent work (Brookhart et al., 1967) using an HS03F-SbF6-S02 solvent has shown that there are other weak signals in the spectrum which may be attributed to the other isomeric form of the conjugate acid (8) in which both protons on oxygen are equivalent. Isomer (8) is present to the extent of only 3% in the case of acetic acid but the corresponding forms of protonated formic acid are present in approximately equal amounts in HSO3F--SbFS--SO2 at
- 60'.
The spectrum of the conjugate acid of acetone was observed for the fist time in HS03F-SbF,, the =OHf proton giving a signal at - 14.5 ppm downfield from tetramethylsilane (Birchall and Gillespie, 1965). Protonated methyl ethyl ketone shows two resonances for the proton on oxygen at - 14-3and - 13.9 ppm in SbF6-HSO3F-SO2 below - 20°C due to 81% of isomer (9) and 19% of isomer (10) (Olah et al., 1967). H
CHs6/0 ' 0 H
I
H
19
SUPERACID SYSTEMS
H
/H +O
H
\O+
II
II
7\ CHI CHaCHr
C CH; 'CHaCHs
(9)
II
C H/\CH3
(10)
.
/H
+O
' 0 .
(11)
II
H (12)
The n.m.r. spectra of the conjugate acids of aldehydes can similarly be observed in the same medium at - 60". The spectrum of acetaldehyde, for example, clearly indicates the presence of both the cis and trans isomers (11) and (12) of its conjugate acid. Sodium and potassium carbonates were found to dissolve in 1 :1 HS03F-SbF5 a t - 78" without the evolution of carbon dioxide (Olah and White, 1968). A single low field peak in the proton n.m.r. spectrum was assigned to C(OH)$, protonated carbonic acid. On warming the solution the peak disappeared and was replaced by another peak characteristic of water in this medium and carbon dioxide was evolved. The assignment of the low field peak to C(OH)$was confirmed by observation of the C13 spectrum which proved to be the expected quartet due to coupling with the three equivalent protons. Many solutes containing more than one basic group undergo multiple protonation in superacid systems. An interesting example is the biologically important biotin which is triprotonated to give (13) (Olah and White, 1968). +OH
I
HNNChH H-C
I
I
t
C-H
p
~
H2C,S0
I
.(CHa)a.CHs .COsHa+
H (13)
B. New Cations Intrcduction Highly acidic systems in general have very low basicities although the acidity and basicity of a given medium are not directly related. Indeed the fact that superacid solvents such as HS08F undergo a significant autoprotolysis, in which solvent molecules act as both acid and base, shows that the basic properties of these solvents are not completely negligible. Nevertheless, the basicities of these highly acidic solvents are sufficiently low that many highly electrophilic species, particularly 1.
20
R . J . GILLESPIE A N D T . E . P E E L
oations, can be obtained as stable species in these solvents although they do not exist in more basic media such as water. Although there is not necessarily any direct relation between the basicity of a medium towards a proton and towards other cations, the highly acidic systems described have in fact been found to have very low basicities towards many cations as well as towards the proton. One of the early applications of a superacid medium was the preparation and identification of the nitronium ion NO$ from nitric acid dissolved in 100% H2S04. HNOs + 2HzS04
-+
NO:
+ H&++ 2HSO7
(15)
This reaction may be considered to proceed by protonation to give H,NO,f followed by dissociation to H20 and NO;. This is driven to completion by the protonation of H20 by the acid medium, thus shifting the equilibrium to the right. A more basic solvent, such as water, attacks the highly electrophilic NO; cation to re-form nitric acid. Since the nitronium ion is the active reagent in aromatic nitration it was postulated that other electrophilic aromatic substitution reactions proceeded by similar mechanisms involving cation intermediates such as I+,Br+ and S03H+. Although no evidence has been found for these particular cations, the search for them has in fact led to the discovery of other new cations, such as,:I which can conveniently be obtained as stable species in solution in a superacid solvent. 2. Carboniurn ions
A very early application of a superacid solvent was the use of 100% H,S04 for the preparation of stable solutions of carbonium ions, such as the triphenylcarbonium ion (Hantzsch, 1908c) (C6H5)3C+, e.g., (C&j)&OH+ 2HzS04 + (C6H5)&++HsO++2HSO7.
(16)
More recently a large variety of different carbonium ions have been prepared in various highly acidic media. For example t-butanol is completely converted to the trimethylcarbonium ion in HS03FSbF,-S02 at -60' (Olah et al., 1967). Stable solutions of the same carbonium ion can be obtained by protonation of 2-methylpropene in HF-SbF, at low temperature (Brouwer et al., 1968).
+
(CHs)aMHa H+ + (CHs)sC+
(17)
I n the same medium methylcyclopropane is protonated to give the methylethylcarbonium ion
21
SUPERACID SYSTEMS
On protonation, in HS03F-SbF6-S02, cyclopropyl bromide undergoes an interesting rearrangement to give a bromonium ion (Olah and Bollinger, 1968).
Many aromatic hydrocarbons can be protonated in various superacid media. For example m-xylene is protonated in HF-SbF6 at - 45" to give the conjugate acid (14) (Mackor et al., 1965), and fluorobenzene in
Q F
@CHs H H
H H
HS03F-SbF6at low temperature gives the fluorobenzenonium ion (15). Oxocarbonium ions (acyl ions) are readily generated in superacid media. For example, Den0 et al. (1964) found that the n.m.r. chemical shift of the methyl protons in a solution of acetic acid in oleum underwent a marked change at about 20% SO3 which they attributed to the dehydration of the conjugate acid CH3.C02H$ to give the methyl oxocarbonium ion CH,CO+. The same reaction also occurs in HS03F-SbF6-S02 when the temperature is increased from - 40" to - 10" (Olah and White, 1967). 3. Halogen catione There has always been considerable interest in the halogen cations and it has often been postulated that the simple ions I+,Br+, and Cl+ are the reactive intermediates in aromatic halogenation reactions. No convincing evidence has ever been obtained for these cations; however, the search for them has led to the discovery of the I$ cation and a number of other related halogen cations (Gillespie and Morton, 1970) which are stable in superacid media. The I$ cation can be conveniently prepared by the oxidation of I2with S206F2in solution in HSOQF. 212
+SzOaF2 + 2G+ 2SOsF-
(20)
A stable, intensely blue, solution of this cation can also be obtained by oxidizing iodine with 65% oleum. I n a more basic medium such 8s
22
R . J . GILLESPIE AND T . E . PEEL
H2S04the 12 cation is almost completely disproportionated to the more stable
+ 1/3 and + 3 oxidation states. 81:
+ 3HSOh
+
I(HS04)a 61:
(21)
The more electrophilic Br; cation can only be obtained in the extremely weakly basic superacid HS08F-SbF6-S03 and does not appear to exist in any measurable concentration in HSOsF because of disproportionation according to an equation similar to that for iodine above. The presumably still more electrophilic Clg cation has not yet been observed in solution and it would appear to need a medium even less basic than HS08F-SbF6-S03. I n addition to 1; and Br;, the cations It, IQ, Brf and Clt have also been obtained in superacid media. I n accordance with the expectation that the X; cations would be less electrophilic than the Xi cations, due to the greater dispersal of their positive charge, they can be obtained in more basic media than the X$ cations. Thus 1; is stable in H2S04but I$ is not. Brg isstableinHSOsF while Br: is only stable in the superacid HS08F-SbF5-SOs. 4. Cations of other non-metallic elements
It has been known for a very long time that the elements sulphur, selenium and tellurium give coloured solutions when they dissolve in a number of highly acid media. For example sulphur gives red, blue and pale yellow solutions in oleum depending on the concentration of the oleum. Investigations on the nature of these solutions have been made since the early 1800’s but it has only very recently been shown that these colours correspond to the formation of the S;$, Si+ and Sq+ cations respectively (Gillespie and Passmore, 1970). They are formed by oxidation of the element by sulphur trioxide according to equations such as (22). 45+6HaSaO7 -+ S ~ + + 2 H S s O i ~ + 6 H ~ s 0 4 + S O a
(22)
Selenium can be oxidized even by 100% H2S04to a dark green solution and by oleum, to an orange solution which have been shown to contain the Seg+ and Sez+ cations respectively. These sulphur and selenium cations can also be prepared in other very weakly basic media. For example, solutions of these cations in fluorosulphuric acid can be conveniently obtained by oxidizing the element with S206F2.
+
4% SaOsFa = Sei++ 2SOsF-
(23)
The fact that tellurium gives a red solution in concentrated sulphuric acid was reported in the first paper published on tellurium by its discoverer Klaproth in 1798 but the origin of this colour remained a mystery until very recently when it was shown to arise from the Te:+ cation. Tellurium differs from sulphur and seleniumhowever in that the two other
SUPERACID SYSTEMS
23
cations that have been discovered are Tef+ and Teg+. These are not at present known for the other elements. Like the halogen cations the sulphur, selenium and tellurium cations are highly electrophilic and they undergo disproportionation in media with any appreciable basic properties although as would be anticipated the ease of disproportionation increases in the series tellurium < selenium < sulphur. For example none of the sulphur cations can be obtained in 100% HeSOcand they all disproportionated in this medium according to equations such as (24) 2S:+
+ 4HaSO4 = SOB + 75 + 4H++ 2IIaSaO7
(24)
whereas the tellurium cation Teii is quite stable even in 98% H,S04. RE~REXOES Barr, J., Gillespie, R. J., and Robinson, E. A. (1961). Can. J . Chem. 39, 1266. Barr, J., Gillespie, R.J., and Thompson, R. C. (1964). Inorg. Chem. 3, 1149. Bass, S. J., Flowers, R. H., Gillespie, R. J., Robinson, E. A., and Solomons, C. (1960). J . Chem. SOC. 4316. Bell, R. P., Bascombe, K. N., and McCoubrey, J. C. (1956). J . Chem. SOC.1286. Birchall, T., and Gillespie, R. J. (1966). Can. J . Chem. 43, 1045. Brand, J. C. D. (1950). J . C h .SOC. 997. Brand, J. C. D., Horning, W. C., and Thornley, J. D. (1952). J . Chem. SOC.1374. Brookhart, M., Levy,G. C., and Winstein, S. (1967). J . Am. Chem. $oc. 89, 1736. Brouwer, D. M., Maokor, E. L., and MacLean, C. (1965). Rec. Traw. Chem. 84,1564. Dean, P. A. W., Gillespie, R. J., Hulme, R., and Humphreys, D. A. (1970). J . Chem. Soo. to be published. Deno, N. C., Pittman, C. U., and Wisotsky, M. J. (1964). J . Am. Chem. SOC. 86, 4370. Deno, N. C., Richey, H. G., Liu, J. S., Hodge, J. D., Houser, J. J., and Wisotsky, M. J. (1962). J . Am. Chem.Soc. 84,2016. Flowers, R. H., Gillespie, R. J., and Oubridge, J. V. (1966). J . Chem. SOC.1925. Gillespie, R. J. (1968). Account8 of Chemical Reaearch 1, 202. Gillespie, R. J. (1968a). In “Inorganic Sulphur Chemistry” (G. Nickless, Ed.), pp. 563-686. Elsevier, Amsterdam. A , to be published. Gillespie, R. J., and Humphreys, D. A. (1970). J . Chem. SOCI! A , 1994. Gillespie, R. J., and Mrtlhotra, K. C. (1967). J . Chem. SOC. Gillespie, R. J., and Malhotra, K. C. (1968). J . C h m . goo. A , 1933. Gillespie, R. J., Milne, J. B., and Thompson, R. C. (1966). Inorg. Chem. 5, 468. Gillespie, R.J., and Morton, M. J. (1971). Quart. Rev. to be published. Gillespie, R. J., and Moss, K. C. (1966). J . Chem. SOC.1170. Gillespie, R. J., Ouchi, K., and Pez, G. P. (1969). Inorg. Chem. 8, 63. Gillespie, R. J., and Passmore, J. (1971). Accounts of Chemical Research. Gillespie, R. J., and Peel, T. E. (1971).J . Am. Chem.800. to be published. Gillespie, R. J., Peel, T. E., and Robinson, E. A. (1971). J . Am. Chem. SOC.to be published. Gillespie, R. J., and Robinson, E. A. (19624. Can. J . Chem. 40, 668. Gillespie, R. J., and Robinson, E. A. (1962b). Can. J. Chem. 40, 675. Gillespie, R. J., and Robinson, E. A. (1965). In “Non-Aqueous Solvent Systems” (T. C. Waddington, Ed.), pp. 117-210. Academic Press, London. 2
24
R. J. Q I L L E S P I E A N D T . E . P E E L
Hamme+t, L.P., and Deyrup, A. J. (1932). J. Am. Chem. SOC.54,2721. Hctmmett, L. P., and Deyrup, A. J. (1933). J. Am. C h .SOC.55, 1900. Hantzsch, A. (1907). 2.Phys. Chem. 61,267. Hantzsch, A. (1908a). 2.Phy8. Chem. 61,46. Hantzsch, A. (1908b). 2.Phys. Chem. 62,626. Hantzsch, A. (1908~).2.Phys. C b m . 65,41. Hantzsch, A. (1909). 2.Phys. Chem. 68, 204. Hyman, H. H., and Katz, J. J. (1966). I n “Non-Aqueous Solvent Systems” (T. C. Waddington, Ed.), pp. 64-81. Academic Press, London. Hyman, H. H., Quartermain, L. A., Kilpatrick, M., and Katz, J. J. (1961).J.Phys. Chem. 65, 123. Johnson, C. D., Katritzky, A. R., and Shapiro, S. A. (1969). J. Am. Chem. SOC. 91, 6654. Jorgenson, M. J., and Hartter, D. R. (1963). J. Am. Chem. SOC.85, 878. Kilpatrick, M., and Luborsky, F. (1964). J. Am. Chem. SOC.76, 6863. Lewis, G. N., and Bigeleisen, J. (1943). J. Am. Chem. SOC.65,1144. McCauley, D. A., Higley, S. W., and Lien, A. P. (1966). J. Am. Chem. SOC. 78,3009. Olah, G. A., and Bollinger, M. J. (1968). J. Am. Chem. SOC. 90, 6082. Olah, G. A., Calin, M., and O’Brien, D. H. (1967). J. Am. Chem. SOC.89, 3686. Olah, G. A., Comisarow, M. B., and Cupaa, C. A. (1966). J. Am. Chem. SOC. 88,362. Olah, G. A., Comisarow, M. B., Cupas, C. A., and Pittman, C. U. (1966). J. Am. Chem. SOC. 87, 2997. Olah, G. A., and Kiousky, T. E. (1967). J. Am. Chem. SOC.89, 6692. Olah, G. A., Sommer, J., and Namanworth (1967). J. Am. Chem. SOC. 89,3676. Olah, G. A,, and White, A. M. (1967). J. Am. Chem.SOC.89, 7072. Olah, G. A., and White, A. M. (1968). J. Am. Chem. SOC.90, 1884. Olah, G. A., and White, A. M. (1968). J. Am. Chem. SOC.90,6087. Palm, V . (1956). Proc. Russ. Acad. Sci. (Chem.)108, 249. Pad, M. A., and Long, F. A. (1967). Chem. Rev. 57, 1. Thompson, R. C., Barr, J., Gillespie, R. J., Milne, J. B., and Rothenbury, R. A. (1966). Inorg. Chem. 4, 1641. Treffers, H. P., and Hammett, L. P. (1937). J. Am. Chem.SOC. 59, 1708.
TURNSTILE REARRANGEMENT AND PSEUDOROTATION IN THE PERMUTATIONAL ISOMERIZATION OF PENTAVALENT PHOSPHORUS COMPOUNDS FAUSTO RAMIREZ Department of chemistry, State University of New York at Stony Brook, Stony Brook, New York 11790 AND
IVAR UGI Department of Chemietry, University of Southern California, Los Angeles, California 90007
.
I. Introduction 11. The Pentavalent State in Phosphorus Stereochemistry . A. General Considerations B. Molecular Structure of Oxyphosphoranesfrom X-Ray Analysis 111. Form81Analysis of Permutational Isomerizations of Pentavalent Phosphorus A. NotationofIsomers B. FormelMechanisms IV. Berry Pseudorotation Mech8nkm (BPR) . V. Turnstile Rotation Mechanism (TR) A. Schematic Representation of the Single TR . B. Description of the Single TR Mechanism , C. Itinerary for Permutational Isomerizations by Single TR and by BPR D. TheMultipleTR E. The TR Switches VI. Calculation of Binding Energies of Model Situations in Turnstile Rotation and in Pseudorotation A. General Considerations B. Calculations for PFs . C. Cdculatione for Other Systems . W.Comparison Between TR and BPR A. CommonFeatures B. Differences VIII. Survey of Experimental Data A. Caged Polycyclic Oxyphosphoranes B. Vari8ble Temperature Proton n.m.r. Spectra of Phosphoranes
.
.
.
.
.
. . . .
.
.
. . .
.
. .
.
.
.
.
. .
. .
.
.
25
.
.
26 27 27 29 35 35 38 42 44 44 47 50 52 55
58 58 60 63 72 72 72 73 73 82
26
FAUSTO RAMIREZ A N D IVAR U O I
IX. Irregular Permutational Isomerizations of Compounds with Pentavalent
.
Phosphorus A. Irregular Processes with Decrease in Coordination Number B . Irregular Process with Increase in Coordination Number X. Conclusions
.
.
.
.
.
.
.
.
. . .
. .
. . . .
115 115 1 18
120
I. INTRODUCTION THE phosphoranes are derivatives of the pentahydride of phosphorus, PH,, in which the five ligands are covalently bonded to phosphorus. The stereochemistry of pentavalent phosphorus relates to the trigonal bipyramid, just as that of tetravalent carbon relates to the tetrahedron. There are significant differences in the stereochemistry of the compounds of these two elements, other than isomer numbers. Some tetracoordinated phosphorus compounds become pentacoordinated rapidly and reversibly. Most pentacoordinated phosphorus compounds change their ligand distribution on the trigonal bipyramidal skeleton very easily by mechanisms that involve the simple deformation of bonds, rather than the rupture and re-formation of bonds. Permutational isomerization, i.e., the interconversion of permutational isomers, is the redistribution of ligands on any polytopal framework. This redistribution can take place by two different processes : (1) The regular permutational isomerizations (or regular polytopal rearrangements) take place without the rupture of bonds, i.e., with preservation of neighborhood relationships. (2) The irregular permutational isomerizations (or irregular polytopal rearrangements) occur with fission and re-formation of bonds. Examples of regular processes are the interconversion of conformers or rotamers in carbon chemistry, and certain types of positional exchange of ligands in molecules whose geometry is representable by the polyhedrons. Examples of irregular processes are provided by other types of positional exchange of ligands in polyhedral molecules. This review? is concerned with permutational isomerizations of molecules which have pentavalent phosphorus and trigonal bipyramidal symmetry. In Section I11 the formally possible regular mechanisms are examined. I n subsequent sections, the phenomenologically acceptable regular processes are analyzed in detail. Then, theoretical calculations of the energy changes that are associated with transitions between model situations are discussed. Finally, the existing experimental data are surveyed to
t The material presented in this review was discussed in part at the ACS 13th Reaction Mechanisms Conference in Senta Cruz, California, on June 23-26, 1970 (F.R.);at the Euchem Conference on Heterocyclic Chemistry in Conglowes, Ireland, on July 6-10,1970 (I.U.);and at the Chirality SymposiuminElmau/Obb., Germany, on October 18-22,1970.
PENTAVALENT PHOSPHORUS COMPOUNDS
27
assess the relative merits of the alternative mechanisms. The last Section is devoted to the irregular permutational isomerizations.
11. THEPENTAVALENT STATEIN PHOSPHORUS STEREOCHEMISTRY A. General Considerations The realization that compounds with pentavalent phosphorus are readily accessible,? that they are of considerable synthetic value, and that they play an important role in the chemistry of the biologically indispensable phosphate esters, has led to mounting interest in the stereochemistry of the phosphoranes. 1-62 The observations that have been made, and the hypotheses that have been advanced, during the past twenty years can be summarized as follows. (i) The stability of phosphoranes varies widely with the nature of the ligands attached to phosphorus. (ii) Regular permutational isomerizations occur very readily in many phosphoranes. Until 1970, the generally accepted and exclusive interpretation of these regular processes was the mechanism originally proposed by Berryb3under the name pseudorotation.$ b 4 Very recently, a new interpretation of the permutational isomerizations of the phosphoranes has been suggested under the name turnstile rotation. 66-69 (iii) The irregular permutational isomerizations can take place by mechanisms in which the pentacoordination of the phosphorus decreases t o four or increases to six, a t some stage of the process. (iv) Phosphoranes are metastable intermediates, rather than transition states, in the reactions of many compounds with tetracoordinated p h o s p h o r ~ s . ~The ~ - ~ permutational ~ isomerizations of these hypothetical intermediates by the regular process have been assumed to take place by the same mechanisms that operate in the stable phosphoranes.$
t The finding (see Ref. 6) that trivalent phosphorus becomes covalently bonded to oxygen in the reactions of many carbonyl Compounds pointed the way to a general synthesis of cyclic oxyphosphoranes (see Ref. 6). This observation was first made on para-quinones, and waa soon extended to ortho-quinones, a-diketones, a-ketoeatere, and many other carbonyl functions. See Section VIII of this review. $ The Berry mechanism in the caae of PF5 gives the impression that the whole molecule has been rotated, but since in Berry’s definition there is no rotational motion, the process was named peeudorotation. The term “pseudorotation” was originally applied to the rapid “up-and-down” motion of the carbon atoms in cyclopentane, and later extended to the puckering of rings in general; see Ref. 64. Objections can be raised to this duplication in terminology, and in this review we denote the Berry mechanism simply as BPR. 5 The hypothesis that phosphoranes are fleeting intermediates in the hydrolysis of certain phosphate esters (see Refs. 35-46) alternated with interpretations of the experimental data in terms of a transition state with pentacoordinated phosphorus (see Ref. 47).
28
FAUSTO RAMIREZ A N D IVAR UO I
This has meant, in general, an application of the BPR mechanismt to the isomerizations of the hypothetical phosphorane intermediates. (v) Nucleophiles, for example ti(-) in Scheme 1, tend to attack the faces, rather than the edges, of tetrahedral phosphorus.”* 36* 4 9 Face 463
SOHEME 1
3 2-
A1 5
x
5
4
3
2 ‘1
attack results in apical entrance of the nucleophile and seems to be a valid h y p ~ t h e s i s ,except ~~ perhaps when the nucleophile is a bulky electropositive atom like silicon or another phosphorus. (vi) Ligands tend to leave the trigonal bipyramid (or TBP)? from an apical rather than from an equatorial position,36as shown in Scheme 1. (vii) Apical bonds are longer than equatorial bonds of the same kind in the TBP. This has been verified in several stable phosphoranes.26*‘O (viii) The relatively more electronegative ligands tend to occupy the apical and the less electronegative ligands the equatorial positions of the 61-63 This fact, and also the greater TBP skeleton (“polarity rule length of the apical bonds, have been explained by the valence shell electron pair repulsion I n this model, a group in the apical position is under more steric interference from the three equatorial groups than if the same ligand were in the equatorial position. Higher electronegativity is associated with smaller electron pairs, and the corresponding ligands should tend to go to the apex of the TBP where the crowding is more severe. ”).18i
t The following abbreviationswill be used throughout this Review: BPR=Berry pseudorotation, defined in Ref. 53. TR=Turnstile Rotation, defined in Refs. 56-69. TBP =Trigond Bipyramid or Trigonal Bipyrmnidd.
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
29
The greater length of the apical bond and the polarity rule can also be interpreted in terms of d-orbital participation. I n this model there is less d-orbital participation for the apical positions, i.e., less backbonding; apical ligands retain more electronic charge, and their bonds have less double-bond character. This point of view will be presented in Section VI. (ix) X-Ray crystallography reveals that there is considerable intramolecular crowding in the TBP, which stems from the existence of several relatively close non-bonded interactions. 6o The consequence is a dramatic stabilization of the TBP by the introduction of approximately planar rings, like five- and four-membered rings.*-1° Puckered six- and seven-membered rings are less effective in decreasing the intramolecular crowding. This intramolecular crowding effect overcomes in some cases the ring-strain associated with bond-angle deformations. I n other words, ring strain is relatively more important as a factor in decreasing the stability of tetracoordinated phosphorus ;36*46 reduction of intramolecular crowding by the ring is relatively more important as a factor in increasing the stability of pentacoordinated phosphorus. l o The corollary is that the optimum thermodynamic and kinetic tendencies to add a nucleophile to tetrahedral phosphorus is encountered among 36 even unsaturated ones.41 five-membered cyclic phosphate esters,36* These esters are the type of compound of interest as models for sugar phosphate and nucleotide chemistry. Phosphoranes also play an important role in the chemistry of four-membered cyclic phosphorus corn pound^.^^-^^, O0 Spirophosphoranes with two five-membered rings are remarkably stable.2$67 (x) X-Ray data shows that five-membered rings occupy the apicalequatorial position in TBP-oxyphosphoranes. 'O For reasons given in Sections V I and VIII, this appears to be a general phenomenon in phosphorane chemistry. (xi) I n the collapse of a cyclic oxyphosphorane to a structure with tetracoordinated phosphorus there is a tendency for the preservation of a five-membered ring in the kinetically controlled process.3G,37 This concept is fruitful in dealing with the hydrolysis of stable oxyphosphoranes, and in dealing with the hydrolysis of phosphate esters where an oxyphosphorane is invoked as a fleeting intermediate.36-46 B. Molecular Structure of Oxyphosphoranesfrom X - Ray Analysis 1. Five-membered oxyphosphoranes
The reaction of trialkyl phosphites with ortho-quinones produces cyclic pentaoxyphosphoranes suitable for X-ray analysis. One of these
30
FAUSTO RAMIREZ A N D I V A R UGI
phosphoranes ( l ) ,is obtained from triisopropyl phosphite and phenanthrenequinone.sO The molecule is a TBP as shown in Table 1, with a TABLE1 Bond Lengths and Angles in a Five-Membered Cyclic Unsaturated Pentaoxyphosphorane
Equatorial P-O(a) 1.641 P-o(4) 1.601 P-O(5) 1.686
88.6' 91.3" 93.1" 117-2" 117.2" 126.6' 127.6"
31Pn.m.r. signal 6 =49-2 p.p.m. (upfield relative to H3P04). The model constructed from these data is reproduced in Fig. 1. The phospholene ring occupies an apical-equatorial position. The apical bonds are longer than the corresponding equatorial bonds. The apical P-O,,, bond
PENTAVALENT PHOSPHORUS COMPOUNDS
31
which forms part of the ring (endocyclic) is longer than the apical P-O,s) bond outside the ring (exocyclic). The same is true among the equatorial endocyclic and exocyclic P-0 bonds. The TBP is very crowded as a result of the existence of several close distances between relatively bulky non-bonded atoms. The apical ring-oxygen (0(1)) is only 2-63 A from the carbon atom C(4)attached to the oxygen atom 0,1) in one of the equatorial isopropoxy groups. The apical oxygen of an isopropoxy group (0(3)) is 2.70 A from carbon atom C(6) attached to
/ Fro. 1. Molecular structure of a 1,3,2-dioxt~phospholene with pentavalent phosphorus, from X-ray crystallographic analysis.
another equatorial isopropoxy oxygen. I n fact the apical alkoxy group (o(&@)) has little freedom of motion since C(3) is 2.74 A from Ot2)and 2.86 A from O(4). Certain rings should increase the stability of the phosphoranes by reducing this type of intramolecular crowding. The lH n.m.r. spectrum of (1)in solution at + 30' shows one signal for the three isopropoxy groups suggesting that there is a relatively rapid positional exchange of these ligands on the TBP skeleton. This phenomenon will be analyzed in detail in later sections, but it should be noted here that one of the most interesting features of the molecule of (1)in the crystalline state is the compression of the angle formed by the two equatorial isopropoxy groups and the phosphorus. This point is brought
32
FAUSTO RAMIREZ AND IVAR UGI
out in formula (1'); the diequatorial angle O(4)-P-O(6) is about 8" smaller than the angle O(6,-P-O(2, formed by an equatorial isopropoxy group and the equatorid ring-oxygen. It will be shown that the most reasonable mechanism t o achieve the observed permutational isomerization of phosphorane (1')includes as one of its features the compression of two equatorial ligands that do not form part of the ring.
A rather striking example of the stability conferredto TBP-phosphorus by the presence of some rings is provided by the phosphorus heterocycle Br
FIQ.2. Molecular structure of a fused 1,2-0xaphospholene-4 and 1,2,6.-dioxapho~phorinene-3 ring system with pentavalent phosphorus, from X-ray crystdographia dyaie.
(2), where five and six-membered rings are fused with a phosphorus atom common to both rings.6a The model constructed from the X-ray data is shown in Fig. 2. The phosphorus (6 = + 60.2 p.p.m.) is at the center of a nearly regular TBP. The apical-equatorial bond angles range from 87" t o 94", while the diequatorial bond angles range from 116" to 124".
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
33
2. Four-membered oxyphosphoranes
A general route to four-membered cyclic oxyphosphoranes is provided by the reaction of hexafluoroacetone (3) with trivalent phosphorus such as phosphonites compounds containing the grouping P-C-H, [(R0)2PCH2R'], phosphinites [(RO)P(CH,R'),], and phosphines [P(CH,R),]. The reaction of hexafluoroacetone with ethyldiphenylphosphine (4) at - 70" gives a five-membered cyclic dioxyphosphorane (5) [S(s'P) = + 6.1 p.p.m.1 which rearranges in benzene solution at 80" into the four-membered analog ( 6 ) [S(s'P) = + 32-1p.p.m.1. A possible mechanism for this rearrangement has been proposed.66 If this phosphorane has TBP-configuration, and if the four-membered ring can occupy only an apical-equatorial position for reasons of ringstrain, there can be six diastereomers for (6). The X-ray analysiseg
discIoses that the stable isomer has the oxygens of the ring and of the alkoxy group in apical positions and the three carbon atoms in equatorial positions in agreement with the polarity rule. (See Table 2 and Pig. 3.)
34
F A U S T O R A M I R E Z A N D I V A R TJGI TABLE 2
Bond Lengths and Angles in a Four-Membered Cyclic Saturated Dioxyphosphorane
CF8 Bond Lengths, A Apical Endo: P-o(4) EXO: P-o(5)
Equatorial 1.79 1.71 O(4)-c(s) 0(5)-C(e)
1.36 1.44
c(2)-c(3)
C(z)-C(a)
P-Ccz) 1.83 P--C(ie) 1.75 P-c(l8) 1.82 1-62 1.61
Bond Angles 04-P-05
0(4)-P-C(2)
76.5'
P-c(2)-c(3)
88*1°
P-o(5)4(g)
94.8' 100.6° 131.4"
P-C(z)--G(e)
118.3"
0(5)-p-c(2)
91.0"
P-o(4)-c(S)
C(Z)-c(s)-0(4)
165.8' O(&)-P4(12) 0(5)-P-C(18) C(18)-P--G(12) C(IZ)-P--C(Z) C(18)-P-C(2) 0(4)-P4(12) 0(4)-p-c(18)
94'4" 93.6" 112.7' 119.3" 127.3' 96.1" 91.2"
The diequatorial angle C(18)-P-C(lz) formed by the two carbons of the phenyl rings attached to phosphorus [see (6')]is 14' smaller than the diequatorial angle C(le)-P-C(z) formed by one of those carbons and the
FIU.3. Molecular structure of a 1,2-oxaphosphetanewith pentavalent phosphorus, from X-ray crystallographioanalysis.
PENTAVALENT PHOSPHORUS COMPOUNDS
35
ring-carbon. There is a significant distortion of the diapical angle 0(4)-P-0(5) from the expected 180”to about 16S0.-f
111. FORMAL ANALYSISOF PERMUTATIONAL ISOMERIZATIONS OF PENTAVALENTPHOSPHORUS
A. Notation of homers A prerequisite to any discussion of the stereochemistry of pentavalent phosphorus is the development of a method of enumerating, classifying and naming all possible isomers. A logical approach for enumeration was suggested by P 0 1 y a and ~ ~ ~by G 0 1 o m b ~and ~ ~extended by Ruch et al. ;74 a correspondingnomenclature was suggested by Ugi, et al.56We consider the class of molecules in which an atom Z is at the center of s peripheral skeletal positions to which 1 ligands are attached. A certain number of permutational isomers will be possible. For example, six ligands attached to a hexacoordinated atom can give rise to 6! =720 isomers if the skeleton with six skeletal positions is asymmetric. However, if the skeleton has octahedral symmetry, Oh,the number of isomers is reduced to 720/24 =30, where 24 is the order of Polya’s “skeletal-equivalence group” in this case. For nomenclatural classification, a pentacoordinated phosphorus compound, P(L,,L,,L3,L4,L5),with five different ligands, L1 . . . L5, attached to a skeleton of five positions identified by the indices : s = I . . . 5 , and of trigonal bipyramidal symmetry, DSh,can exist as 5!/6=20 isomers, or 10 pairs of enantiomers. To represent these isomers one defines a Reference Molecule ME in which the indices of all the ligands and the indices of all the skeletal positions coincide. Those indices will not coincide in the isomers of ME.The skeletal indices are assigned to the Reference Molecule according to some definition^,^^ for example, as shown in ME. The ligand indices are assigned to the Reference Molecule
t For additionalX-ray data on the structure of phoaphoranessee Refs. 70-72.
36
PAUSTO RAMIRE2 A N D IVAR UGI
as prescribed by some rational system, for example by the Sequence Rules.75 The Reference ME can be abbreviated to E in view of the coincidence in it of the skeletal and the ligand indices. 4 4
56
ME
4
E
An isomer of E is obtained by a permutation of the ligands. Thus, the isomer denoted by “a”, “ 12” or “21” in previous publication^^^^^^^ 77 is obtained from E by the ligand permutation (1 4) (2 6 ) ,which is read as : “ligand 1 replaces 4 and 4 replaces 1, while 2 replaces 6 and 5 replaces 2 ”. The name or code or descriptor of an isomer is the ligand permutation by which it is derived from E, regardless of the mechanism by which the actual isomerization takes place, even if there is no real mechanism to perform that isomerization. The permutation of the ligands relative to the skeletal positions is carried out without any alteration of the skeletal indices, as can be seen from Ma and M,, 4 ) ( 2 6). If one draws the Reference ME and gives the 1-descriptor of an isomer, (1 4) (2 5), one can reproduce that isomer in the same skeletal space orientation as that shown in the Reference. Moreover, a drawing of the Reference ME in the abbreviated form E, also specifies a particular orientation of the Reference skeleton because indices 1and s must coincidein the Reference. Therefore, once E is drawn, the representation of an isomer, M(? 6) can be abbreviated to (1 4)(2 6 ) , because in the isomenzation E -+ (1 4) (2 6) the skeletal indices 9 remain the same while the specified ligand indices 1undergo permutations. The drawing (1 4) (2 6), together with the drawing E, corresponds to the statement: “in this isomer 4 1
1
I
ligand 1 occupies skeletal position 4, ligand 2 occupies skeletal position 5, etc.” The usefulness of this approach will become apparent in Section VIII
PENTAVALENT PHOSPHORUS COMPOUNDS
37
which contains an interpretation of the variable temperature n.m.r. spectra of many phosphoranes, some of which are rather complex and give rise to many permutational isomers. It should be understood that one can always select that particular orientation of the Reference Molecule MEEE that will generate the most convenient drawing of it, or of any desired isomer, after the ligand indices are replaced by the corresponding atoms, and groups of atoms, in the analysis of a complex phosphorane. The enantiomer of (1 4) (2 5) in the Permutational Notation is (1 4 2 5), and is derived from the Reference E by the permutation read as “ligand 1 replaces 4, 4 replaces 2, 2 replaces 5 , and 5 replaces 1 ”. Two further illustrations are isomer (3 5) and its enantiomer (1 3 5), for which only the abbreviated representations are given.
i
52 M (1 4 2 a)
(1 4 2 6) 4
4
3
3
(3 5)
(1 3 5)
Table 3 gives the correspondence between this “Permutational Notation” of isomers and previous notations, which can be called ‘(binary” because they name an isomer from the two ligands in apical positions.49,76, 77 The Binary Notation makes no proviso for naming isomers when two or more ligands are identical. The Permutational Notation, on the other hand, provides a unique descriptor for all isomers in all cases of equivalence among ligands. This is brought out in Table 4, which will be used extensively in Section VIII during the interpretation of the experimental data. The asterisks in Table 4 mark different descriptors (i.e., isomers) for each case of ligand equivalence. The information in the Reference E can be represented by the Reference Matrix
(3, (t),
12345 2 3 4 5)
= (I
FATJSTO R A M I R E Z A N D I V A R U O I
38
TABLE3 Permutational and Binary Notations for Isomers of P(l 2 3 4 5) with Five Different Ligmda 4
6
E Permutation 1 2 3 4
LRS
Mb
GWc
a
12 13 14 16 23 24 26 34 36 46
21 31 41 61 32 42 62 43 63 64
b c
d e f g
6
6
7 8
h
9 10
i
j
Permutation
Lauterbur and Remirez, Ref. 76. Midow, et aZ.,Ref. 49. c Gorenstein and Westheimer, Ref. 77.
LRQ
Mb
GWc
1
J
12 13 14 16 23 24 26 34 36 46
a b
The matrix for any isomer is obtained from
(DE
in the usual way
using the descriptor of the isomer as an 1-permutation :
(d),
(1 4)(2 5 ) 1 2 3 4 5 123451
[
=[Z4
pa(:), = The mathematical properties of these matrices (mappings) and of their permutation transforms reveal a great deal of information on the stereochemistry of TBP-molecules of the class P(12 3 4 5) and, in general, on molecules of the class Z(L1...L,) with a variety of skeletal et al. symmetries, as shown by R ~ c h , Ugi,66 '~ =
B . Formal Mechanisms The occurrence of regular permutational isomerizations will depend on the flexibility of the molecular skeletons. This is true not only for monocentric skeletons involved in isomerizations of pentacoordinated
TABLE 4 Permutational Notation for Isomers of P(1 2 3 4 5) with Some Equivalent Ligands Li=La
6
(25)
(15)
(25).
(35).
7
(14)
(24)
(34).
(24).
(24).
(25) (24)
(14)
(35)'
(34). (35).
(35).
(34) (35)
E*
8
(35).
(26)
(35)
9
(34).
(34)
(24)
10
1 '
~~
(25)
E*
(125)
(12)'
(25).
(24)'
(25) (15) (14) (15) (34).
(35)+
(15)
(26)
(24)
(14) (15) (34) (35)
(14)
E*
(24)
(35).
(34) (35)
(15)
(25)
E*
(35)
(34) (35)
(14)
E
E
cl
z U
cn
40
FAUSTO RAMIREZ A N D IVAR UQ I
phosphorus, but also for the dicentric skeletons involved in ethane-like conformational changes and for polycentric skeletons. I n general, the isomerizations of molecules which have skeletons with s positions can be described by permutations of the symmetrical group S,. Thus, the formal mechanistic alternatives for the interconversions of permutational isomers of five-coordinated molecules are obtained by partitioning the group S , in correspondence to the partitions of the number five into its distinct seven classes of conjugate elements :
The indices of P represent the number of elements of cycles of the permutations ; for example, P 2includes all isomerizations brought about by interchange of two positions, P2+2,those in which there is interchange within each of two pairs of positions. P l is the “identity operation” and TABLE6 Permutations of the Symmetrical Group 8 5 and Formal Mechaniama of Regular Permutational Isomerizatiom of Trigonal Bipyramidal Molecules Permutation class
Ca+3
Permutation subclass”
Mechanism name
TRf
Intermediate skeletal symmetry
p :c z v g t: C3”
Equatorial =e. Apical = a. b The “ identity operation” causes no isomerizationand does not represent a mechanism. c M=Muetterties; 8 e Ref. ~ 78. 6 Alternating cyclic permutations of four elements with one fixed point. eMs=BPR (see Ref. 53). f Turnstile rotation (me Ref. 6 5 ) ; 9 p==pair;t=trio. Q
41
PENTAVALENT PHOSPHORUS C O M P O U N D S
does not change anything, while P4also means P1+4,since a permutation cycle of one element does not affect a given set of numbers. We can now refer to six mechanistic classes of permutational isoand C6,since C1does not represent an merizations :C2,Cs, C2+2,C4,c2+& isomerization. When the five ligands are M e r e n t and the skeleton is devoid of any symmetry, each permutation refers to a different process. However, if some of the ligands are equivalent, or if the skeleton has some symmetry such as D3hfor the TBP, some of the mechanisms S C H E ~2 E
4 3+1
(12) 6
6 Cav
.A'
2 (34)
5
? (1 42)
3+: 5
c1 M4
-* t-
Dsa
C4v
42
BAUSTO R A M I R E Z A N D I V A R U G I
within a class of 8 5 have common features and belong to subclasses within that class;74 these are collectively “one mechanism”. The situation is summarized in Table 5. Two subclasses of class C2 have been called mechanisms “MI” and “M2” by M~etterties,’~ who attempted to derive all possible mechanisms for the regular permutational isomerizations of TBP phosphorus compounds, and to represent the symmetry of the corresponding intermediate states. Muetterties’ conclusions78are shown in Scheme 2 . Mechanisms “M3” and “M4” are subclasses of the class C3, while mechanism “Mb” is one of the subclasses of C,+2. Muettertie~”~ mechanism “Ms” is the same thing as Berry pseudorotation63 (BPR); it can be called subclass Cdaof class C4, since it is represented by alternating permutations that are cycles of four elements in which the indices of the apical and the equatorial ligands alternate and from which the index of the pivot is excluded. The permutations of C4a me mutually conjugate. As stated above, a permutation cycle of one element, e.g., (1) does not affect a given set of numbers, and the permutation (1) (2 3 4 5 ) is equivalent to (2 3 4 5). Among the formal mechanistic classes C2, C3, C2+2,and C4,only the latter in the form of BPR or “M6” is physically reasonable in terms of the symmetry of the skeleton in the intermediate state. However, there is a second formal mechanism that is reasonable by these criteria, namely, a subclass of the class C2+3. This mechanism, called the turnstile rotation was fist recognized in 1970; it is represented by those permutations containing one apical and one equatorial ligand in a cycle of two, and the remaining ligands in a cycle of three elements.bb
IV. BERRYPSEUDOROTATION MECHANISM(BPR) BPR was introduced63to explain the observed equivalence of the five fluorines of PF, (7) in the n.m.r. above -100”. The apical and the equatorial positions of the TBP are not symmetry-equivalent and, as was later shown,26the apical P-F bonds are indeed longer than the equatorial ones in (7). Berry suggestedb3that the magnetic equivalence was due to a relatively rapid positional exchangeofligands, (7) 3 (7”),which resulted
PENTAVALENT PHOSPHORUS COMPOUNDS
43
from the synchronous expansion of an originally diequatorial angle F-P-F from 120" to 180" and contraction of an originally diapical angle F'-P-F' from 180" to 120". I n BPR, pairs of fluorines move in planes perpendicular to each other, as shown in Fig. 4. Those planes are defined by the phosphorus, the corresponding pairs of fluorines and the fixed equatorial fluorine or pivot. I n the ideal situation, the molecule has Clv symmetry (7'), as it crosses the energy barrier that separates the TBP (7) and (7"). There has been considerable discussion centering on the relative contributions of the motions of the apical ligands, F', vs. the equatorial ligands, F, at various stages of BPR.79t BPR accomplishes the pairwise simultaneous exchange of two apical and two equatorial fluorines and retention of the position of the pivot.
B
B FIG.4. Motions of ligands in Berry Pseudorotation. F* =Pivot. F'=Apioal. F = Equatorial.
The new and the original TBP differ in spatial orientation by a 90"rotation about the pivotal axis, but there isno internal rotationof some part of the molecule 08. the other. The BPR mechanism is in agreement with the results of a crucial experiment carried out by Whitesides and Mitchell,*Owho analyzed the temperature dependence of the 31Pn.m.r. spectrum of the dimethylaminotetrafluorophosphorane (8). Below - loo", the TBP is frozen with the nitrogen in an equatorial position. I n the range - 100" to -50" the apical fluorines (F') and the equatorial fluorines (F)exchange their positions in pairwise and concerted manner, (8) -+ (S'), according to the analysis of the n.m.r. signal characteristics. The apical (F') and the equatorial (F)fluorines do not undergo partial exchange with each other and are "inseparable pairs" in the lQF-n.m.r.
t The transition state for BPR was given tetragonal-pyramidalsymmetry, C ~inV mme recent calculations mentioned in footnotes 34 and 36 of Ref. 49.
44
FAUSTO RAMIREZ AND IVAR UGI
As the pivot, the dimethylamino-group remains in the equatorial position. P’
F
The Whitesides-Mitchell experirnent8O can be interpreted by BPR using the nitrogen as pivot. The experiment rules out, for this particular as alternatives molecule, mechanisms M1-M6 discussed by M~etterties’~ to M, or BPR. However, the results of the experiment are also in complete agreement with the TR mechanism as will be seen in the next Section.”
v. TURNSTILEROTATION MECHANISM56 (TR) A. Schematic Representation of the Single T R
It is convenient to introduce the TR mechanism by giving a schematic representation of it as shown in Fig. 5. However, it should be emphasized that the actual TR phenomenon is Merent, and corresponds to certain motions of the atoms which will be described in detail in the next Section. Consider a given permutational isomer, for example that designated’, “ a ” in the Binary Notation and (14) (2 5) in the Permutational Notation, and orient the TBP as shown in Fig. 5. This orientation emphasizes that the apical ligand 1 and the equatorial ligand 6 will constitute the pair in this TR, while the apical ligand 2 and the equatorial ligands 3 and 4 will be the trio in the TR. To perform the TR: (a) Transpose (180” rotation) the pair-ligands. (b)Rotate the trio-ligands 120” in the direction that brings the apical ligand of the trio into the original position of the equatorial ligand that must remain equatorial. I n this example the trio rotates 120” clockwise, which places apical ligand 2 where equatorial ligand 3 originally was. Ligand 3 remains equatorial and is the pivot of the BPR that effects the same isomerization. Figure 5 justifies the name “turnstile rotation” given to this mechanism. In the formal treatment presented in Section 111, the single TR corresponds to a subclass of the class C2+3,in which the permutation is of the type (a’e’)“e’’a), where a =apical and e =equatorial. As a consequence of the skeletal symmetry of the TBP, the following four
PENTAVALENT PHOSPHORUS COMPOUNDS
45
permutations accomplish the same isomerization, and thus can be said to be equivalent or analogous : (a’e’) (ee”a); (ae’)(ee”a’) ;
(a”’‘) (ee‘a) (ae”)(ee‘a’)
On the other hand, the BPR process corresponds to a subclass of the class C4, involving the alternating permutations (ae”a‘e’). Any given
I.
180” 0. 120” 6
2
FIG.6. Convenient representation of the turnstile rotation. The actual phenomenon is the 60’ internal relative rotation of pair w8. trio. The hexagons are steric frame of reference.
isomerization by one BPR can also be effected by four equivalent TR, although the BPR and tho TR mechanisms are, physically, quite different (see below). With reference to Fig. 5 , the isomerization “a + j ” or (1 4) (2 5) + (1 2) is effected by the BPR that uses ligand 3 as pivot and also by the four TR which correspond to the following permutations. Only one of these permutations is shown. (1 4) (2 3 5 ) ; (2 4) (1 3 5);
(1 5) (2 3 4) (2 5) (1 3 4)
46
FAUSTO R A M I R E Z A N D I V A R U O I
If the 120"rotation of the trio in Fig. 5 had been counterclockwise,the isomerization would become: (1 4) (2 5 ) --f (3 4)) and the apical ligand 2 moves to the position originally occupied by the equatorial ligand 4. Ligand 4 is the pivot in the BPR process that accomplishes that same isomerization. The permutations that correspond to the four TR processes by which one can effect this isomerization are as follows, of which (1 5)(2 4 3) is illustrated. (1 3) (2 4 5 ) ; (23)(145);
3'
"I 4' 2'
(1 5 ) (2 4 3) (25)(143)
4'1
5
''2
It is self-evident that an overall 60" and the reverse overall 300" internal rotations of a pair of ligands relative to a trio accomplish exactly the same result. The actual TR-phenomenon that accomplishes the isomerization (1 4) (2 5 ) + (1 2) in Fig. 5 corresponds to a clockwise 60" internal rotation of the pair relative to the trio, in which the pairequatorial ligand 5 "points" to the trio-equatorial ligand 3 that will remain equatorial (see Fig. 5 for the schematic representation). The actual TR-phenomenon for (1 4) (2 5) -+ (3 4) corresponds t o a counterclockwise 60" internal rotation of the pair relative to the same trio, and ligand 5 "points" to the other trio-equatorial ligand, 4, that remains equatorial. I n an idealized situation in which the five ligands are equivalent and in which there is no intermolecular exchange of angular momentum, the actual TR-phenomenon that effects the isomerization ( 1 4) ( 2 5 ) -+(1 2) of Fig. 5 corresponds to the clockwise 36" rotation of the pair vs. the counterclockwise 24" rotation of the trio with conservation of internal angular momentum. The conservation of intramolecular angular momentum is not a necessary condition for TR, but it should be pointed out that only TR and BPR, among all the formal mechanisms discussed above, possess this capability. In general, to describe the actual TR in the isomerizstion (1 4) (2 5 ) -+ (1 2) of Fig. 5 all that needs to be said is that there is a clockwise overall 60" internal rotation of the pair relative to the trio. I n the idealized TR that effects the illustrated isomerization (1 4) (2 5) --f ( 3 4) there is a counterclockwise 36" rotation of the pair vs.
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
47
a clockwise 24"rotation of the trio, and in general, it suffices to state that there is a counterclockwise overall 60' internal rotation of the pair relative to the trio. A given TBP has three equatorial ligands, each of which can be used as pivot in BPR to generate three additional TBP, in the general case. As shown above, the results of one BPR can also be achieved by four equivalent or analogous TR. This situation is summarized in Table 6, where the chosen original TBP is the Reference Molecule E in the Permutational Notation (this corresponds to "j " in the Binary Notation' e ) . TABLE 0 Correspondence Between a Given BPR and Four Equivalent TR 4
6 E
Isomerization
Pivot in BPR
Pair-trio in TR
Descriptor of the product of the isomerization in the Permutational Notation. The Binary Notation is that of Ref. 76.
B. Description of the Single T R Mechanism We turn now to a detailed analysis of the motions of the atoms involved in the TR process. Let p, and pa represent the ligands of the pair, and t,, ti and t, the ligands of the trio. The equatorial ti will remain equatorial as a result of this isomerization.
48
FAUSTO RAMIREZ A N D IVAR U U I
(9; TBP)
(10-1; 0"-TR)
(10; 0"-TR)
(10-11; 0'-TR) (Eclipsed)
The following motions begin at the same time and proceed simultaneously. (1) A compression of the diequatorial angle in the trio, ti-P-t, from 1.20' to 90". (2) A tilt of the ligand-pair, pepa, toward the apical ligand, ta, of the trio. The pair swivels approximately 9" about an axis that passes through the phosphorus and is perpendicular to the plane containing atoms Pp,p,. If nothing else happens, the result of these two motions would give an arrangement depicted by (lo),or by symbolically placing the ligands in a hexagonal prism as frame of reference (10-I), or by using a Newman-type projection (10-11). This "eclipsed" arrangement (10) can be called the 0"-TR model situation because a t that stage there has been no internal relative rotation of the pair vs. the trio about their common axis. However, a third type of motion intrudes upon the other two. (3) The internal rotation of the pair relative to the trio, in opposite directions, which reaches the value of 30" midway in the TR process. The resulting arrangement can be depicted by placing the ligands in the hexagonal frame of reference, (11-I), or by the projection (11-11). The "staggered" arrangement (11) can be called the 30"-TRbarrier model situation; it is characterized by an approximate local C2and C,-skeletal symmetry, respectively, for the pair and the trio with a common axis. t , h
9 e2
:@?
@
LPa@-&-@P3
(11-1; 30°-TR)
(11-11; 30"-TR (Staggered)
a3
t
d-
Le
.-'"21
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
49
Continuation of the TR process by an additional 30" rotation results in the new eclipsed 0"-TR model situation shown in the projection (12-11).This is analogous to the previous models, (10-1)and (10-II), with suitable changes in the nature of the ligands. The reversal of the tilt of the pair and the expansion of a diequatorial angle in the trio generates the new isomeric TBP (12). (In this illustration the pair, pap,, tilts away from t, which is now apical, while angle ti-P-t, expands from the original 90" to 1 2 0 O . )
(12)
(12-11; 0"-TR)(Eclipsed)
These three simultaneous ligand motions result in the exchange of apical and equatorial ligands pairwise and concertedly. The energy
0 '
1 0 '
20'
30'
40" 50"
60'
AV FIG.6. Binding energies of PFs in turnstile rotation model situations, from CNDO/2 cdculetions. Z-axis aa "TR-axis"; reaction coordinate =relative change, A+, in the &nqth coorc&&es of the pair and trio ligands.
50
FAUSTO RAMIREZ AND IVAR UQI
changes associated with the various TR model situations in the case of
PF, were calculateda1 using the CNDO/2 approximation to the LCAO-MO-SCF method.a2-s6 The details are discussed in Section VI. The results are summarized in Fig. 6. Point (A) corresponds to the TBP itself; (B) is the result of the angular-compression from 120" -+ 90° (motion 1); (C) represents that motionplus the tilt of the pair (motion 2). Points (D)and (E)correspond to progressive stages of the three concerted motions until the 30"-TR barrier model, represented by (F),is reached. The TR-energy profile of this model is, thus, represented by the solid curve in Fig. 6, where the "new" TBP is (A')with the same energy as (A) in PF,. Point (G) is of particular interest. It corresponds to a situation in which, in the 0"-TR model situation, one of the ligands of the trio is displaced toward the ligands of the pair (and vice versa) to form a new trio, while the two remaining ligands (formerly of the trio) are displaced so as to form the new pair. This generates a new 0"-TR model situation and can be called a 0"-0"-TR Switch of Type I : TypeISwitch:
Q
Q
0 0 0
TypeUSwitch:
8
Q 0
0
___+
0
0
@
@
0 0
o o 0
@
0
8 = Original Pair-Ligands
0 = Original Trio-Ligands
The Type I1 Switch is defined as shown in the scheme. As will be seen later, it is conceivable that under certain circumstances a switch of the types described above provides the route by which a pair-trio combination is changed into a different pair-trio combination ("crossover ") in a sequence of isomerizations by the TR mechanism. C. Itinerary for Permutational Isomerizations by Single I'R and by B P R When the five ligands L1. . .L, of a TBP pcntacoordinated phosphorus compound are different, there are 20 isomers, or 10 pairs of enantiomors, which were identified by descriptors in the Permutational Notation and by other symbols in the Binary Notations, as shown in Table 3. These isomers are interconvertible by both the single TR and the BPR processes according to the itinerary given in Fig. 7. This graph was first a 6 , 87 to permutational isomerizations of compounds with pentacoordinated phosphorus by BPR exclusively.?
t These graphs are similar to others developed in connection with discussions on carbonium ion transformations;see Ref. 88.
PENTAVALENT PHOSPHORUS COMPOUNDS
61
Figure 7 shows that the configurational inversion of TBP phosphorus with five different ligands, for example (1 4)(2 5 ) to (1 4 2 5), requires a minimum sequence of five BPR with five different pivots. To achieve the same result one must also apply five TR processes, but there are 45 =lo24 different TR pathways to inversion for each BPR pathway.
(145)
a.
(1425)
( I 435)
a
\‘a
0
(15) FIG.7. Itinerary for permutational isomerizations by aiq7Ze turnstile rotation, or by Berry pseudorotation, when the five ligands in trigonal bipyramidal phosphorus are different. Isomers are denoted by Permutational Notation.
When some of the ligands L, . . . L, are equivalent, the number of isomers of TBP phosphorus is reduced as can be seen in Table 4. Some examples of the itineraries that result by the application of both the single TR and the BPR processes are shown in Fig. 8. The descriptors for the isomers in Fig. 8 do not have a counterpart in the Binary Notations. I n Fig. 8 achiral isomers fall on the symmetry axis, and chiral isomers are placed in mirror-image positions, if any.
52
F A U S T O RAMIREZ A N D I V A R U O I
(24)(35)
(14)(35)
(15)
(25)
4
E
L, =L*=L3
FIG.8. Itineraries for permutational iaomerizations by aingZe turnatile rotation, or by Berry pseudorotation, when some of the ligands EIV equivdent.
D. The Multiple T R The TR concept allows the transformation of an isomer, for example, (1 4) (2 5 ) into another (2 4) (3 5 ) by a double turnstile rotation, (TR)2, Double TR, (TR)2
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
53
even if the single TR, or its equivalent BPR, are prohibited by the relatively high energy of a given TBP.
In this example, isomer (1 2) which separates (1 4)(2 5) from (2 4)(3 5) in the itinerary of Fig. 7 is "overrun" or bypassed by the (TR)2process. The representation of the isomerization :
depictedabove is the hypothetical counterclockwise 2 x ( - 120")= + 120" rotation of the trio without any transposition of the pair of ligands 2 and 5 (since 2 x 180" does not affect the placement of those two ligands). Note that the rules for the schematic representation of (TR)2are: (a) Do not change the pair. (b) Rotate the trio in the direction of the equatorial ligand that must remain equatorial. I n this example ligand 4 remains equatorial and the 120°-rotation of the trio is clockwise for the (TR)2. This schematic representation should not obscure the actual phenomenon, which corresponds to the 2 x 60" overall internal relative rotation of the pair of ligands 2, 5 vs. the trio 1, 3, 4. Now, after the first 60" relative rotation of pair vs. trio, it becomes energetically relatively favorable to carry out a second 60"relative rotation of the same pair-trio combination, at which point the appropriate reverse-tilt of a pair and expansion of a bond-angle generates the new TBP, (2 4)(3 5). At the stage of the 0"-TR situation that corresponds to the bypassed isomer (1 2), the reverse-tilt of a pair and the corresponding bond-angle expansion to give (1 2) are energetically less favorable than the other motions which result eventually in the formation of (2 4) (3 5). A single TR is represented by one 2-element cycle and one 3-element cycle permutation, but a (TR)2is just a 3-element cycle permutation, hence the statement: [(2 6)(1 3 4)12=(1 4 3). The itinerary of isomerizations by the double turnstile rotation, or (TR)2is shown in Fig. 9. Figure 9 corresponds to isomerizations when the five ligands are different. The lines connect isomers which fall on the same hemispheres. There are no connections between isomers which fall on different hemispheres; i.e., there can be no inversions. Each isomer is connected to six others by (TR)2. The isomers that were connected by
64
FAUSTO RAMIREZ A N D IVAR UGI
single TR (or by BPR) are not connected by (TR)2,for example (1 4) (2 5) and (1 2) (or its enantiomer E) are not connected by (TR)2.
FIG.9. Itinerary for permutational isomerizations by double turnstile rotation, when the five ligands in trigonal bipyramidal phosphorus are different. Isomers are identitied by both Permutational and Binary Notations. There are no connections between isomers in upper and lower hemispheres.
Other examples which illustrate (TR)2processes are as follows. The conversion of (1 4) (2 5) into (1 4 2) by permutation [(l 3) (2 5 a)]' = ( 2 4 5);thereverseisomerizationiscarriedoutas:[(l 3)(2 4 5)j2 =(2 5 4). The conversion of E into ( 1 4 2) by permutation [(3 5) ( 1 2 4)]' =( 1 4 2 ) .
PENTAVALENT PHOSPHORUS COMPOUNDS
55
None of these isomerizations can be performed by one BPR or by a single TR process as can be seen from Figs. 7 and 9. If for some reason, isomer (2 4)(3 5) shown in the illustration of the (TR)2process is energetically unfavorable, the triple turnstile rotation, or (TR)3 becomes a possibility. This corresponds to 3 x 6 0 " internal relative rotation of the same pair-trio combination. The isomerization :
overruns TBP (I. 2) and (2 4) ( 3 5) ; cf. Fig. 7. Triple TR, (TR)3
The (TR)3 is a 2-element cycle permutation as can be seen in the statement: [(2 5) (1 3 4)13=(2 5 ) . The rules for the schematic representation of (TR)3are: (a) Transpose the pair. (b) Do not change the trio. A repetition of the TR six times with the same pair-trio combination regenerates the original bipyramid : (TR)e=identity. This represents another difference between TR and BPR, since a repetition of BPR twice with the same pivot regenerates the original bipyramid: (BPR)2=identity.
E. The TRSwitches I n the discussion of the single TR mechanism it was pointed out that one of the unique features of this concept is the "TR-switch". I n the idealized model for PF6 shown in Fig. 6, this situation corresponds to point (G). The mutual displacement of certain ligands at the energy level of the 0"-TR situation provides a reasonable mechanism to effect a << crossover" from one pair-trio combination to a different pair-trio combination. This process can be illustrated with the aid of Fig. 10. Figure 10 describes first a single TR to isomerize (1 4)(2 5) into (1 2). An attempt is made in Fig. 10 to approximate the motions corresponding to the 60"internal relative rotation of a pair 1, 5 ws. a trio 2, 3,4. If the five ligands were identical, and if an idealized situation were involved, these motions would correspond to a clockwise 36" rotation of the pair 3
56
F A U S T O R A M I R E L A N D I V A R UOf
vs. the counterclockwise 24" rotation of the trio, i.e., 36"+24" =60° overall rotation of pair 215. trio.
s 3 5
I
FIG.10. Isomerization: (14) (2 6 ) --f (1 2) occurs by single TR,shown as a 60" internal relative rotation of pair 218. trio. Isomerbation: (14) (2 6 ) --f (1 4) (3 6), bypaasing isomer (1 2), OCCUIW by TR-switch of Type 1, in which ligand 2 of original trio and ligands 1 and 6 of orighal pair form the new trio. Hexagons are steric frame of reference.
If a t the energy level of the 0"-TR situation one of the ligends of the trio, for instance ligand 2, is displaced toward the ligands of the pair 1 , 6 , in such a way that a new trio, 1, 2, 5 is created, the remaining two
PENTAVALENT PHOSPHORUS C O M P O U N D S
57
ligands formerly in the trio, ligands 3 and 4, will adopt the role of the new pair. By means of this 0"-0"-TRswitch of Type 1, there is a '' crossover"
5
FIG.11. Inversion of trigonal bipyramidal phosphorus with five different ligands (representation as in Fig. 5 ) . a B is Binary Notation. (1 4) (2 5 ) + (1 4 2 6 ) is Permutational Notation. The inversion involves a triple turnstile rotation, (TR)8, a double turnstile rotation (TR)2 and one switch of Type 1,in which ligand 3 and ligands 2 end 6 form the new trio. Hexagons are steric frame of reference.
from the (1 5 ) (2 3 4) to the (3 4) (1 2 5 ) pair-trio combination, without the formation of the TBP (1 2). A second single TR,this time using the new pair-trio (3 4) (1 2 6 ) , followed by the reverse-tilt of a pair and the appropriate bond-angle expansion, results in isomer (1 4) (3 5 ) . In other words, the conversion of
58
FAUSTO RAMIREZ AND IVAR UQI
(1 4) (2 5) into (1 4) (3 5) has been carried out by two TR and one TRswitch which effected the “crossover” between two pair-trio combinations. The shortest pathway for the inversion of TBP phosphorus with five different ligands consists of a total of five TR processes and one crossover from one pair-trio combination to a different pair-trio combination. This crossover can be realized by one TR-switch of Type I , in which case, the inversion of one T B P into its antipode can take place without the intervention of a third TBP. This is illustrated in Fig. 11. Note that of the five TR, three represent a (TR)3and two represent a (TR)2. The 0’-0”-TR switch of Type I involves ligand 3 of the initial trio and ligands 2 and 5 of the initial pair in the formation of the new trio, in this illustration. If there is no switch in the inversion depicted in Fig. 11, the crossover requires the formation of at least one intermediate TBP, for example (1 4). It was already mentioned that the inversion of TBP-phosphorus with five different ligands by BPR requires a minimum of five BPR with five different pivots via four different TBP, which in the example given are : (1 2), (2 4 ) ( 3 5 ) , (1 4) and (1 3 5), and that there are 46 =lo24 different TR inversion pathways for each BPR inversion pathway by the 4 : 1 correspondence principle of Table 6.
VI. CALCULATION O F BINDINGENERGIES O F TR AND BPR MODEL SITUATIONS A. General Considerations The semiempirical quantum mechanical approximation82-86that has been called CND0/2 (for complete neglect of differential overlap) was used to calculate binding energies in the trigonal bipyramidal configuration and in a series of model situations that are pertinent to the TR and the BPR mechanisms in a variety of molecules.81 This approximation to the solution of the LCAO-SCF equations yields results that provide, at least, a trend in the real systems involved. It has been shown that, with reasonable approximations and parametrization, this method offers an adequate compromise with respect to the more rigorous and accurate but far more laborious and expensive ab initio methods.82-86 I n the CND0/2 approach, only valence electrons are treated explicitly by choosing a minimum set of basis functions (atomic orbitals) for the molecule in its ground state. The main approximations are: (1) Basis functions 4” are treated as an orthonormal set ;i.e., the overlap integrals are put to zero unless c $ ~ =4r, in which case they are unity. (2) All two electron integrals which involve overlapping charge densities between
PENTAVALENT PHOSPHORUS COMPOUNDS
69
differentbasis orbitals are neglected. Thus, electron interaction integrals are approximated to account for the repulsion between electrons on adjacent atoms. (3) Approximations are introduced to account for the effect of adjacent atomic cores upon an electron in any atomic orbital +*. (4) Approximation is made to calculate the effect of an electron in the field of two atoms simultaneously. The latter, termed a resonance integral, is estimated by empirical values chosen to fit either the experimental data or the results obtained from ab initio calculations. The alternatives to be considered in the choice of parametrization (i.e., selection of basis functions or atomic orbitals) are sp or spd. A comparison of these two alternatives exists for systems with P-F and S-F bonds, in each case the inclusion of d-orbitals in the basis set of the ground state P-atom reduces bond polarity between the P and the F Calculations in other atoms, and the same is true for the S-F systems for which the dipole moments are known show that no realistic calculations are obtained without inclusion of d-orbitals for the third row element. Extended Huckel MO calculations on phosphorus (V) halidesg2 have left the impression that little d-orbital participation exists and that this effect can be neglected to a first approximation. However, this type of calculation can be criticized on two counts. It produces a more polar character in the bond because it does not allow for the dependence of the electron-attracting power of the orbital on the number of electrons already in that orbital or on surrounding orbitals. The most serious limitation stems from the selection of an optimum d-orbital exponent (Slater exponents). Slater's rules for calculating screening constants do not cover d-orbitals and therefore an orbitrary choice or an optimization must be made for the d-orbital exponent determination. Boyd and Lipscombg3calculated an orbital constant for phosphorus compounds and concluded that the optimum was 1.4, which results in a more contracted orbital set, predictable in the electrostatic field of a ligand set. The arbitrary selection of 1-00 in the extended Huckel calculations yields a more diffuse orbital, producing smaller ligand pd interactions. I n the CND0/2 calculationsB1of binding energies in TR and BPR model situations, the value 1.60 was chosen for the phosphorus orbital ~ ~ value assigns constant in agreement with Boyd and L i p s ~ o m b .This electron densities properly ;the application of Pauling empirical electronegativity correlationQ4to the data obtained in the case of PF, predicts an electronegativity difference between phosphorus and fluorine of 1-7 in good agreement with the published value of 1.78. I n the following calculations, the bond lengths were optimized to within less than +0.01 A, and the bond angles to within +0-5". Comparisons of the results of the calculations for PF, with the data from
60
FAUSTO RAMIREZ A N D I VAR UQI
electron diffraction measurements indicate that the equilibrium bond lengths for the ground state are overestimated? by about 6%.
B. Calculationsfor PF, 1. T R model situations
The first species to be considered is the TBP itself (7), and this is followed by the model situations which result from the progressive F
F
I
1
F
F
( 7 ) BE = -606.1 kcal mole-1
(13) RE = 603.9 kcal mole-'
(14) BE
(15) BE
=
-596.0 kcal mole-1
=
- 592.0 kcal mole-'
TABLE 7 Binding Energies (BE) of the 30"-TR Barrier Models in PF6, as a Function of FPF Bond Angles in the Pair-Trio. BE= -606.1 kcal mole-1 for the TBP.
FPF Angle Pair
Trio
BE, kcal mole-1
90"
9oo
-696.0
85" 96" 90u 86" 96" goo
85"
95"
85'
96' ~~
-691.3 - 696.9 - 696.6
- 691.4
- 697.0 - 692.6 - 689.6 - 692.8
~
t This is probably due to the magnitude of the /I-valuesassigned to the core hamiltonian end could be altered t o fit more closely, but the gain in precision would not provide new information, since it was noticed that P-H, P-0 and P-C bond distances were qimilarly overe&imated; e.g., P--H bond length ie overeetimated by 6.5%.
TABLE 8 b
Binding Energies (BE) of TR Model Situations in PF5 with 90" FPF Bond Angles in the Pair-Trio, 88 a Function of Relative Pair-Trio 4 Internal Rotation. F1, Fz, F a are Equatorid; Fa,F5 &w Apical in the TBP.
*
Fluorine angular coordinates; all P-F distances, r = 1.73 A
BE, kcal mole-1
$1
el
TBP
-606.1
71' 34'
114' 06'
- 71'
OO-TR
- 592.0
60"
125" 16'
-60"
10"-TR
- 603-5
57" 43'
117" 49'
20"-TR
- 599.1
43" 51'
30"-TR barrier
- 595.0 - 600.1
Model species
e2
43
e3
114' 06'
180'
35' 16'
125' 16'
77" 42'
121' 33'
30"
125'16'
45O
94'52'
4% 34'
04
44
4s
05
0"
54'44'
180'
125'16'
180" 45"
0"
45"
180"
125' 16'
117' 49'
180'
38" 31'
0 '
57'29'
170"
125O16'
-83" 51'
121' 35'
180'
41' 45'
0"
48' 15'
160"
125' 16'
-90"
125' 16'
180'
46"
0" 45"
150'
125" 16'
180'
85'08'
180'
175"OS'
~
0"-0" switch Type I
-67"30'
90"
'd
-135"
4'52' ~~
*tr rn
62
BAUSTO RAMIREZ A N D IVAR U G I
compression of the diequatorial angle ( a 3 )of the trio (13), (14), and from the tilt of the pair (15). The last of these (15), is the 0"-TR model situation. The energies of (7), (14), and (15) are plotted in Fig. 6 as points (A), (B) and (C), respectively. The energy of the 30"-TR model situation in the ideal case of PF, corresponds to point (F) in Fig. 6. For this, there are closely related situations which differ only in the angles a2and asof the pair and the trio. Their calculated energies, in PF,, are given in Table 7. The 85&r-95,",,-30"-TR which has a2 =85" and cc3 =95", represents the most stable of the models for the barrier. However, the energy of the barrier is fairly insensitive to angles azand a3 and the values a2 = a 3 =90" were conveniently used in further model calculations. Table 8 gives the energies and the corresponding spherical coordinates of the fluorines in three representative stages of the TR-process : the 10"-TR, the 20"-TR and the 30"-TR barrier model. The geometric situation in PF5 which corresponds to values of the spherical coordinates half-way between those for two eclipsed 0"-TR species, is energetically quite favorable, and is point (G) in Fig. 5 . This is the Type I 0"-0"-TR switch model situation. I n Fig. 5, the relative change, A$, in the azimuth coordinates of the pair and trio ligands is used as the reaction coordinate, with the Z-axis as the "TR-axis". 2. B P R model situations The CNDO/2 calculations give the following results when applied to several model situations of PF, relevant to the BPR mechanism. The data are plotted in Fig. 12. F
F'
F*4 F
F
(16) BE = -602.1 kcd mole-1
(17) BE = -604.3 kcal mole-1
(18) BE = -6'79.0 kcal mole-1
When BPR is possible, as in the case of PF,, the compression of the diapical angle and the expansion of the diequatorial angle must proceed synchronously (19, 20) rather than in a stepwise manner (16, 17, 18). The square planar arrangement or octahedron minus a vertex (18), is energetically out of the question relative to the tetragonal pyramid (20),
P E N T A V A L E N T P H 0 S P H 0R U S C 0 M P 0 U N D S
(19) BE = -604.1 kcal mole-1
63
(20) BE = -602.1 kcal mol-1 105'-1 50°-C4 y
as BPR barrier. The apical bonds are "stiffer" than the equatorial bonds. The idealized intermediates in the positional exchange of the ligands of PF5by the BPR and the TR mechanisms are different by an amount
-602-
Y Y
W
I 0
I
'6
I
I+
I
34
1 1 1 BPR
FIQ.12. Binding energies of PFs in Berry pseudorotation model situations. From CNDOI2 calculations.
greater than the ttmplitudes of the zero point bending vibrations given in the Literature,26as the F-P-F angles are 86") 105", and 150" for the BPR intermediate state, and about 90" for the TR-intermediate state. Motions other than those presented might also be operating in certain cases, since the nature of the saddle point between the BPR- and TRintermediate states is not known.
C. Calculations for Other Systems
In the previous section, the CND0/2 calculations provided a comparison of the energies associated with a given set of atoms, PF,, in various model situations pertaining to the TR and the BPR mechanisms. This section shows the results of CND0/2 calculationss1 on several 2o compounds (7), (22),95(23),and several hypothetical model
64
FAUSTO RAMIREZ A N D IVAR U G I
compounds (21), (26)-(31), all of which have the TBP skeleton with Dgh symmetry, but in which the ligands are placed in different positions. The results are summarized in Scheme 3. SCHEME 3 Binding Energies (inkcal mole-1) of Trigonal Bipyrsnidal Phosphorus Compounds from CND0/2 Calculations. A =Loss of Stability Among Isomers
F
H
(7):-606.1
(21) :-412.5
F
H
H
(22") : -838.7 A = 6.6
(22'): -538.2 A = 8.1
F A ,.**F F
(23):
(3% (23'): A = 38
- 1224
ms+
F
F+
F
CFs
(24):-1326
(24'): A = 23
a + F
F+F F
(25): -617.7
Ho+) OH (26): -1413
c1 (25'): A a 34.6
P E N T A V A L E N T PHOSPHORUS C O M P O U N D S
*oT-o; 4:: OH
0-
(27'): A = 28
(27):-1307
G
a#
HO+OH
OH
0
OH
OH 0(28'): A = 44.8
(28'): A = 31.8
(28):-1027
q H
OH
CHS (29'): A = 20
(29):-2009
CHs
HO
0OH (30): -1904
65
Ho#
OH
OH
0-
CHa
(30"): A
(30'): A = 22
= 32
CHI (30"):A = 61
H
H (31'): A = 20
(31): -3793
(31'):
A = 37
-
(31"): a = 114'; A = 40
C H + OHA fIH
.*
A 43 a-OOo;A=44
a = 102';
106"-160"-C4,. skeletons from (22).
F
H
-589.0
F
-634.2
H -539.6
66
FAUSTO RAMIREZ AND IVAR UQI
Scheme 3 contains several of the fundamental structures with pentavalent phosphorus, among them phosphorane itself, PH, (21), and pentahydroxyphosphorane, P(OH), (26). The latter is the hydrate of phosphoric acid: H3P04+H20-+P(OH),. A hydrate of methylphosphonic acid (CH,)(HO),P (29), is also included as a model for its esters. Mono- and di-ionized forms, (27), (28), (30), are also given, since their stabilities in various isomeric forms provide important data concerning the role of intermediate oxyphosphoranes in the chemistry of phosphoric acid and its derivatives. A model compound of the 1,2-0xaphospholene ring (31),is provided, since this system, in the form of several derivatives, will be discussed extensively in Section VIII. A comparison of absolute values of the binding energies given in Scheme 3 is significant only for a given set of isomers of the same molecule, i.e., for compounds that have the same number and kind of atoms. This is due to the approximations involving the atomic core which are introduced in the CND0/2 calculations. I n these calculations, bond lengths were optimized to within less than &0.01 8, and bond angles to within & 0.5". No attempt was made to optimize the 0-H or C-H bond distances, since their contributions to the total energy were relatively small. The P-H bond distances were optimized; a variation of &0-05 A about the equilibrium position showed a small energy change. The data in Scheme 3 show the following: (1) The methyl group in an equatorial position consistently gives a more stable isomer than the methyl group in an apical position. The difference in energies between the isomers varies with the system, but is quite significant in all cases; the range is from 20 to 39 kcal mole-l, depending on the other ligand that is being exchanged, and on the system undergoing the isomerization. (2) The transfer of a carbon from an equatorial to an apical position is accompanied by greater loss of isomer stability when it involves the CH,-group than when it involves the CH,-group of a five-membered ring. (3) The placement of the five-membered ring in a diequatorial used. position is quite unsatisfactory, regardless of the angle 0-P-C (4) The oxide anion shows a preference for the equatorial position. ( 5 ) As expected, the preferred isomer in PHZP3 has the two hydrogens in equatorial positions, but it is puzzling that the diapical placement of the hydrogens corresponds to a slightly more stable situation than the equatorial-apical hydrogen placement. (6) The various BPR barrier states corresponding to the isomers, (22), (22') and (22"), of PH2F3Q5 differ considerably in energy, as can be seen by the values given for the skeletons with 105"-150"-C4, symmetry in Scheme 3. The CND0/2 calculationsE1provide data pertaining to the electron
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
67
densities in the atoms of the various isomers of a given TBP structure. Data for three of the cases given in Scheme 3 are presented in Table 9. Significant trends can be discerned in Table 9 concerning the distribution of electronic charge for the phosphorus and for the ligands within TABLE9 Electron Densities of Isomers ofTrigonal Bipyramidal Phosphorus Compounds Calculated by CNDO/2 PHzFs (22) Isomer
F, F-ap
F, H-ap
H, H-ap
P F F F H H
4.2094 7.2249eq 7.2776ap 7.2778ap 1.0063eq 1.0063eq
4.1086 7.2346eq 7.2346eq 7.2612ap 1.029Oeq 1.0703ap
4.2003 7.2671eq 7.2671eq 7.2671eq 1-0142ap 1.0142ap
P(OH)z(OCH:CH*CHz)CHs(31) eq _ aP_ _
Isomer
CH3 OH -CHa CH :CHO-
P OH OH OCH CHa CH3
P OH OH OH 0CH3
-
4.5362 8.2848eq 6.3700ap 6-306lap 4-0883eq 4-1378eq
4.6696 6.3364eq 6.3630ap 6.3630ap 6.4683eq 4-1408eq
4.6660 6.3442eq 6-3487eq 6.3786ap 8-4937eq 4.1489ap
eq _ &P _ CH3 OH -0CH :CH * CHz4.6148 6.2922eq 6.3260ap 6.2668eq 4.1 180ap 4.1336eq
4.6692 6.3386eq 6.3499eq 6.3803ap 6.6626ap 4.1346eq
4-6698 6.3679eq 6.3679eq 6-3679eq 645220ap 4.1818ap
68
BAUSTO RAMIREZ AND IVAR U G I
a set of isomers of a given compound. (1) I n all cases, the electron density on phosphorus is greatest for the isomer of lowest energy. This suggests that there is a gain in stability when the phosphorus receives some electron density from the ligands. (2) I n all cases, the difference between the electron density of the phosphorus and the average electron density of the ligand set is lowest for the isomer of lowest energy. 89-gs, g6-103 on the There is at present considerable occurrence and the extent of stabilization that is achieved by the backdonation of electron density into vacant d-orbitals in phosphorus. The CNDOI2 calculationsE1show some interesting trends in this respect, in particular concerning the effect on the relative stability of the various isomers of a given molecule. The method of the calculation is first illustrated for PF6. Table 10 is an extract from the density matrix of the program. The diagonal elements correspond to electron densities of the described orbital. The sum of these elements, 3-8890, is equal to the total valence shell electron density for the phosphorus. Of this, 1.3671 units of electron charge, or about 35%, reside in the five d-orbitals of the phosphorus. The bond orders and the resonance-integrals of the Core Hamilton Operators for PF6are shown in Table 11. The negative product of these two is the stabilizing contribution of the particular ligand orbital with the phosphorus d-orbitals. The same type of data was obtained for the two isomers of the monoanion derived from pentahydroxyphosphorane (27) and (27’) in Scheme 3. The total phosphorus electron density and the d-orbital electron density for both isomers is given in Table 12. Only the density matrices of the five d-orbitals of the phosphorus are reproduced in Table 12. It can be seen that the difference between the total phosphorus electron density in the two isomers is 0.044 units, of which 0.011 units correspond to the Merence between the electron density in the d-orbitals of the two isomers. Therefore, approximately 25% of the difference in electron charge takes place via the phosphorus d-orbitals. The type of data contained in Table 11 for the case of PF5, were also obtained for both isomers (27) and (27’). The former (27), with equatorial oxide is more stable than the latter (27’), with apical oxide, as indicated in Scheme 3. There is a difference of 0-0175 a.u. (1 a.u. =627-463 kcal mole-l) for the d-orbital contribution t o the binding energy. Hence about 30% of the entire binding energy difference in these two isomers is due to d-orbital contribution. I n summary, the CNDO/2 calculations give the following results: (1) I n PF6,approximately 35% of the total valence shell electron density of the phosphorus, i.e., the electron density associated with all 3s-, 3p-
PENTAVALENT PHOSPHORUS COMPOUNDS 0 0 w 0 0 0 0 0 1 0 0 w 0 o o H o o o o o m o o o o
I
I
I I
I
g0 0g 1z g0 g0 0g 0g 0g mg g0 g0 gw g0 o w o o o o o 1 o ~ m o o o ~ o o o o o m o H m o o
I l l
I I
go 1z go og og og ogmz go zm go go go 0 0 0 0 0 0 0 0 0 0 0 0 0 o ooooomoooooo
I
I
I I
I l l
0 0 0 0 0 0 H 0 0 0 0 0 0
ggggggggggggg
Z888%8888%%88
a o o o m o o o o P - m o o
0090Y0900""00 I
I
I I I I I I
0 0 0 0 0 0 0 0 0 0 0 0 0
ooomoooooooob ooomoooooooom
I
I I I
o o o w o o o o o o o o m ggggggggggggz
0 0 ~ 0 0 0 0 0 w 0 0 m 0 0 0 m 0 0 0 0 0 ~ 0 0 m 0 0 0 w 0 0 0 0 0 d 0 0 m 0
0 0 ~ 0 0 0 0 0 3 0 0 ~ 0 I
I I
l
l
0 0 0 0 0 0 0 0 0 0 0 0 0
k 3 o o o a o o o o ~ w o o
2888Z8888Z%88 I
I
I l l
H 0 0 0 0 0 0 0 0 0 0 0 0
009099900"Y90
69
FAUSTO RAMIREZ A N D IVAR U a I
70
TABLE11 Bond Orders (Upper Number) and Resonsnce-integralsof the Core Hamilton Operators (Lower Number in Italics) for PFs. Total Phosphorus Electron Density= 3.8890. P-F = 1.73 A Phosphorus orbital Fluorine orbital
dz*
dzz
4 2
dz-u
dzu
Contribution
- 0.0668 Pz
- 0.0402
43380 --1190
PZI
p,
- 0.0402
.3380 --.I190
- 0.0660
-*3877 ~1420
Electron density = 7.2396 F(1, 2 , 3 )
Total
8
-.1467 el210
-a1160 *I048
--ZOO9 -1815
p,
-.I166 *0355
- -3475 -1200
-0227 -0018
-2020 -*0615
-*0227 --0018
'3213 --1220
p, PZ
-*I609 *0696
Electron density = 7.2106
-0.2013 -0.0662
- 0.0467 -0.0616
- 0.0382
.2786 *I030 Total
+ 0.2017
and 3d-orbitalsYresides in the five 3d-orbitals. (2) I n general a ligand in the apical position carries more electron charge than its equatorial counterpart. (3) If one compares two stereoisomers of a given molecule with TBP pentavalent phosphorus, the electron density of the phosphorus is greater for the isomer that has the lower binding energy. This is true, in general, for all the isomers of a.ny given phosphorane :the phosphorus electron density is greatest for the isomer with the lowest energy, suggesting that there is a gain in stability when the phosphorus receives some electron density possessed by the ligands. The CND0/2 calculations for the hypothetical mono-ionized form of the hydrate of phosphoric acid :
71
PENTAVALENT PHOSPHORUS COMPOUNDS
HaP04+HaO + (H0)5P (H0)5P + (H0)4PO-+H+
reveals that the molecule with an equatorial oxide anion is 28 kcal mole-' more stable than its isomer with the apical oxide. Approximately 30% of the entire binding energy difference between the two isomers is due to TABLE12 Total Valence Shell Electron Density, and &Orbital Density Matrices of Phosphom in P(OH)40Isomer with equatorial anion (0-). Total phosphorus electron density = 4.4677. &Orbital electron density = 1.7893 ~~
P P P
P P
~
DZ2 DXZ DYZ DX-Y DXY
0.4884 0.0009 0.0058 0-0137 -000215
0.0009 0.3079 0.0164 0-0016 -0.0056
0.0058 0.0164 0.3081 0.0066 -0.0017
0.0137 0.0016 0.0065 0-3652 0.0367
-0.0215
- 0.0056 -0.0017 0.0367 0.3217
Isomer with apical anion (0-).Total phosphorus electron density =4.4236. &Orbital electron density= 1.7781
P P P P P
DZ2 DXZ DYZ DX-Y DXY
0.5017 0.0011 0.0013 0.0054 0.0136
0*0011 0.3468 -0.0043 0.0168 0.0011
0.0013 -0.0043 0.3415 0.0064 -0.0174
0.0054 0.0168 0.0064 0.3121 0.0085
0.0136 0.0011 - 0.0174 0.0085 0.2762
d-orbital contribution. Moreover, about 25% of the difference in the total phosphorus electron density of the two isomers takes place via the five 3d-orbitals of the phosphorus. The bonding model that emerges from the CND0/2 calculations has these features. There is significant phosphorus d-orbital participation in the ground state, and the apical positions produce less d-orbital interaction than the equatorial positions. Since there is less back-donation of electrons from the apical than from the equatorial positions toward the central phosphorus, the most electronegative elements will tend to place themselves a t the apex, a t which position they will support more electronic charge on themselves. There is less double bond character at the apical positions and this should result in longer bonds. Some isomers of a given molecule with TBP phosphorus will be stabilized relative to other isomers, by passing electron density into the d-orbitals of the phosphorus.
72
FAUSTO RAMIREZ AND IVAR UQI
VII. COMPARISONBETWEEN TR AND BPR A. Common Features (1) Both achieve the pairwise exchange of apical and equatorial ligands in a regular process, i.e., by bond-deformations. (2) When the ligands are not connected by rings, the energies of the two types of barriers are similar, according to CNDO/2 (LCAO/SCF) calculations. (3) For every BPR there is at least one equivalent TR. (4)Both can proceed, but need not proceed, with conservation of angular momentum in idealized situations.
B. Differences (1) I n TR, but not in BPR, all ligands participate in the motion. ( 2 ) TR, but not BPR, represents relative internal rotation of ligands. (3) The species associated with the barriers are quite different ; in BPR the skeletal symmetry is CkV,while in TR there is approximate local
skeletal symmetry CzVand C3”. (4) The permissible motions of the ligands in TR, the barriers involved, and the itineraries that result, can be related in a predictable manner to the types and the distribution of the ligands, in particular when the TBP contains rings that prefer (or can exclusively be placed on) apical-equatorial positions. (5) The results of any BPR can be duplicated by one TR, and there are three additional equivalent TR that achieve the same result. Other things being equal, TR is a more probable event than BPR. (6) Unfavorable TBP can be overrun or bypassed in TR sequences but not in BPR sequences. (7) Repetition of a BPR with the same pivot regenerates the TBP: (BPR)2=e (identity). It takes six TR with the same pair-trio combination and direction of motion to regenerate a TBP: TRs =e. (8) BPR and TR correspond to different partitions of the ligand set : (1 + 4) for BPR, (2+3) for TR. Their representations by permutations are mathematically different, since they belong to different classes of permutations of the symmetry group S5. I n certain acyclic derivatives of pentavalent phosphorus with TBPgeometry, e.g., PF6,CH3PF4,and (CH&NPF4, the positional exchange of ligands can occur equally well by the BPR or the TR mechanisms. There may even be some advantage associated with BPR. I n other acyclic systems, e.g., PC12F3and PH2F3,the advantage will be in favor of the TR mechanism. Cyclic derivatives of pentacoordinated phosphorus with a TBP D,, skeleton cannot undergo positional exchange of ligands unless a TR mechanism i s involved.
PENTAVALENT PHOSPHORUS COMPOUNDS
73
The TR mechanism may also explain positional exchange in molecules in which the pentacoordinated atom with the TBP skeleton is an element other than phosphorus. VIII. SURVEY OF EXPERIMENTAL DATA
A. Caged Polgcyclic Oxgphosphoranes The caged pentaoxyphosphorane (33)66-69is obtained from the adamantenoid phosphite 04-1 O6 (32),and hexafluoroacetone (3). CFs
?F3
"Fs--?-r CFs
(33)
(32)
I n the static TBP shown in Fig. 13 there are two types of methine protons, two types of cyclohexane-equatorial and two types of cyclohexane-axial methylene protons, and two types of fluorines. Yet, the n.m.r. spectra of this compound in solution at + 30°,which is summarized in Table 13, exhibits only one signal ( T C H ) for the methine protons, one signal for the equatorial and one signal for the axial methylene protons of the cyclohexane (TOH,), and one signal for the fluorines. These spectra
H
FIO.13. Trigonal bipyramidal representation of adamantenoid oxyphosphorane.55-5s
TABLE13 N.M.R. of Caged Polycyclic Oxyphosphoranes.5s-5Q
N.M.R. at
+ 30"4
Compound No.
8 31P
6 19F
(33P
+41*7
-9.5
5.00; 28
(39)f
+36.1
-
(41P
+ 39.2' + 65.1
- 9.5
T C H ; JHCOP
Variable temperature
WE,; JHCOP
ReIatively sharp signal above :
19F n.m.r.b
Coalescence
Relatively sharp signal below :
- 80"'
ca. -122'
- 140'
Nonee
7-20;None& 8.26 ; None 5.58; 11.0
- 80"
-llOoto -115'
- 130°
None
5.05; 13.5j
- 80'
-95" to -1000
-110"
4 In CDC13 or CH2Clz. 31Pn.m.r. signal i n ppm from H#04= 0, at 40-5 MHz. 19F n.m.r. signal (ppm from CF3COaH = 0) at 96.4 MHz. 1H n.m.r. signal (ppm from (CH3)4Si= 10) at 60 MHz. b In vinyl chloride (sealed tubes), with CF3C02Et as internal reference in Varian HA-100. c From l-phospha-2,8,9-triox~amantane and hexafluoroacetone. J H C H = 10 Hz. C From 30" to - 75", one F-doublet, J~ccop=O*3 He, 649 H Z d o d e l d from CFaC02Et; a t - 150", signal 575 Hz downfield from reference. f From 4-methyl-2,6,7-trioxa-l-Y phosphabicyclo-(2.2.2)-octaneand hexafluoroacetone. ~ T C H , C= 9-20ppm. h From 2,6,7-trioxa-l,4-diphosphsbicyclo(2.2.2)octane and hexafluoroacetone. Upper figuredue to pentacoordinated, lower figure due to tricoordinated P. Jgcp= 8.5 Hz.
+
'
75
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
are the result of a relatively rapid positional exchange of ligands to the phosphorus, which is inhibited only when the temperature reaches approximately - i20°.69 The positional exchange of ligands in phosphorane (33)occurs by a regular, and not by an irregular, permutational isomerization process. This conclusion is supported by the data given in Table 14, and by the following arguments : (1) The n.m.r. shift of oxyphosphoranes is very sensitive to the nature of the solvent when there is rapid equilibrium between the pentacoordinated and the tetracoordinated phosphorus species. l o Yet, neither the adamantenoid oxyphosphorane (33)nor its (CH,O),Panalog,lo7rlo8 (34),show this effect in solvents as different as CDCl,, dimethylformamide and hexafluoroisopropyl alcohol. The latter is quite acidic and causes the opening of the oxyphosphorane ring in the (CH,),PC6Hshexafluoroacetone adduct,66(35);note the negative value of the shift in this case. The same is true for the rather unstable phosphine-penanthrenequinoneB63 lo’ adduct (36) (37). (2) The pentaoxyphosphoranes derived from phosphites, (33)and (34), are remarkably stable to water. The dioxyphosphoranes derived from phosphines, (35)and (36) (37)are instantaneously destroyed by one mole equivalent of water. The reasonable conclusion is that the facile hydrolysis is due to the presence of the open dipolar structure (37),in equilibrium with the phosphorane (36),in this and in the related relatively unstable phosphoranes.1° 93
TABLE14 3lP N.M.R.
Shifts of Stable and Unstable Oxyphosphoranesa Under Various Conditions
6 3lP in anhydrous solvents
Compound No.
Trivalent phosphorus compound
Carbonyl compound
CDC13
DMFb
HFIPb
6 31P in CDCla
+
+
CDCl3 0.6 CDCla mole 1 moleequiv. equiv. HFIP HFIP
+1
mole equiv. water After 0-6hr
After 24hr
(33)
CibO3Pc
Hexduorowetone
+42-4a
+41*2
+44*le
$42.3
+42*4
+42.0d
+42.0
(34)
(CHs0)3P
Hexafluoroacetone
+49*7f
+4&9
+48.3e
+49*4
+49-6
+49*9f
+494
(35)
(CHs)~Pc&s Hexduoroacetone
+ 10.9
-
-96.8e.V
+9.6
$9.0
-41.8”
- 86.3c9k
- 28.7
-53.4
-
~~~~
(36),(37)
~~
~
~
(CH3)2PCeH5 Phenenthrenequinonet
+ 1.5j
-39.9%h
fl The oxyphosphoranes are made from the reaction of a trivalent phoqhmw, compound with a carbmyl unnpound. b DMF= dimethylformamide; HFIP=hexduoroisopropyl alcohol. C l-Phospha-2,8,9-trioxaadamantane.d cu. 0*2Ms o h e ca. IM soh. I cu. 2~ soh.
0 Probable structure: [(CH3)2(C&)h-O-C(CF~)2C(CF&OH][(CFs)~]. h Probable structure: (CH&(CeHs)PO. f The (CH80)sPphenanthrenequinone sdduct ha9 6 a l p = +44.7 ppm (in CDC13). f The malogous (CaHr,)~(C&)P-phenanthrenequinone adduct hae: 6 S I P = - 10.8 ppm (in CHSC12) but - 0.9 ppm (in benzene) showing some S I P n.m.r. solvent dependence. k Acyclic dipolar ion.
PENTAVALENT PHOSPHORUS COMPOUNDS
77
The BPR mechanism cannot account for the facile regular permutational isomerization of phosphorane (33). The equatorial oxygen of the five-membered ring cannot be the pivot in BPR because this would require placing one ring of the adamantenoid cage in a diapical position. Neither of the two equatorial oxygens of the adamantenoid cage can be the pivot in BPR. This is seen best in a molecular model, but it is rather obvious from Fig. 13. Consider the equatorial oxygen in front of the drawing in Fig. 13 as the pivot, and consult also Fig. 4 in Section IV. As the 180' diapical angle closes toward 120°, it carries the apical adamantenoid oxygen backwards. As the 120" diequatorial angle opens toward 180°, it carries the second equatorial adamantenoid oxygen forward. But those two oxygens form part of a six-membered 1,3,2-dioxaphospholane ring which is fixed by the attachment at its C14)and Cts)carbons to a three-carbon chain. The role of this three-carbon chain is t o anchor the oxygen-pivot as part of two 1,3,2-dioxaphospholane,rings, each of which, in turn, contains the adamantenoid apical and equatorial oxygens involved in this implausible BPR. Whatever motion is taking place here, it in no way resembles what has come to be called "Berry pseudorotation",63 no matter how liberally one interprets that concept. The TR mechanism provides a satisfactory explanation for this relatively rapid i s o m e r i z a t i ~ n . ~ ~Before - ~ ~ discussing this interpretation, we should consider two other caged polycyclic oxyphosphoranes. The f i r ~ t , (39), ~ ~ -is~made ~ from the reaction of hexafluoroacetone with the bicyclic phosphiteloQ(38). This phosphorane is also undergoing relatively rapid regular permutational isomerization at + 30" as shown in Table 13. The isomerization of (39)is inhibited66only at low temperature which, however, is somewhat higher than the inhibition temperature for the adamantenoid phosphorane (33);i.e., the barrier for the isomerization is somewhat higher in (391, than in (33). The next caged phosphoraneS6 (41) has two phosphorus atoms, one
78
FAUSTO R A M I R E Z AND IVAR U G I
penta and the other tricoordinated. I n (41), the tricoordinated phosphorus has replaced the tetracoordinated carbon at the bridgehead in (39). The data which show the permutational isomerization of (41) and its inhibition at low temperature are included in Table 13.
+
2
The establishment of the TBP-configuration of the P-valences in the oxyphosphoranes (33), (39) and (41), can be realized only as a result of a certain amount of distortion of the caged moieties, which requires the expenditure of some energy. This distortion should be greatest in the adamantenoid moiety (33), and lowest in the others. The TBP-distortion energy is released as the molecule goes over into the 30"-TR barrier symmetry for the trio (the caged situation with its approximate local CQY moieties); this situation is depicted in Fig. 14 and 15. Consequently, the difference in energies between the TBP-ground state and the 30"-TR barrier situations in these phosphoranes should be in the order corresponding to the data in Table 13. The reactivities of the bicyclic phosphites and phosphines, (32), (38), (40), and those of the monocyclic and acyclic analogs, e.g., methyl ethylenephosphite, trimethyl phosphite and trimethylphosphine, toward carbonyl compound, are significantly different.7-11 The three bicyclic compounds react readily with hexafluoroacetone to give the caged oxyphosphoranes (33), (39), (41), and in this respect they resemble the acyclic trimethyl phosphite which readily gives its phosphorane (34). It is remarkable that the phosphorus of the caged phosphine (40) does not react with the highly reactive carbonyl group of hexafluoroacetone, either in the phosphine-phosphite itself, (40), or in the phosphinephosphorane (41). The caged phosphites are quite unreactive toward biacetyl. There is no detectable reaction of this a-diketone with the two phosphatrioxabicyclooctanes (38), (40), in benzene solution at 30' or at 80" after 24 hr. However, as in the many cases previously de~cribed,~-'Oone can detect
PENTAVALENT PHOSPHORUS COMPOUNDS
79
FIG.14. Idealized model of the 30"-TRbarrier situation in the adamantenoid oxyphosphorane. Phosphorus = 0 . Oxygen= 0 .
the formation of mixtures of diastereomeric oxyphosphoranes, (42), (43), resulting from the condensation of two molecules of biacetyl with one of the phosphites, under suitable conditions and prolonged reaction A-----
-7
FIG.16. Idealized model of the 30"-TR barrier situation in caged pentaoxyphosphoranes. Phosphorus = 6-coordinated 0 . Phosphorus or C(CH3) =3-coordinated 0 . Oxygen =
.
80
FAUSTO R A M I R E Z A N D I V A R UGI
(42: X = C(CH8)) (43: x = P) 110 The adamantenoid phosphite (32), also gives the diasteroemeric 2:l adducts with biacetyl (44) at a slow rate, which seems to be somewhat higher than that of the other two phosphites. 66-6Q
times.66-SQ.
A possible explanation of these differences is that the reactions of the trivalent phosphorus compounds with carbonyl compounds may contain reversible and irreversible steps. The condensation of one mole of biacetyl with one of the phosphite probably contains several reversible steps leading eventually to the 1 :1 oxyphosphorane. If this 1 :1 adduct is of high energy as a result of ring strain and/or intramolecular crowding in the TBP, it might not be observable. The carbon-carbon condensation step in the formation of the 2 :1 adducts is probably essentially irreversible under most conditions ;hence, once that step is achieved, there is an opportunity for isolation of the 2 :1 phosphoranes. There is direct experimental evidence, in the case of the reactions of phosphites with pentafluorobenzaldehyde,lll? (45) + (46), for the reversible attack by the phosphorus of phosphite on the carbonylcarbon.l12 At some latter stage, there is attack of phosphorus on
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
81
carbonyl-oxygen leading to C-C condensations and 2 : 1 phosphoranes, see (47), (48) ws. (49), (50). The two types of phosphoranes, (47), (48) and R R
R R
I 1 XaP-j-CC II II
+ I I
0 0
R
0
R
R
I I c=c I I
-
+
I
R
-I
7-7
O y O
x’ I‘x
x
1t CFs
CFa
I CFa-GI
(CFdiCO _____+
o +
CFa
I
CFFC-CCFa
I
o,+
‘PXa
CFa
I
t
+
0-
PXa
CFa
I 1 CFa-C -C-CFs 1 1 O@ .
X ’ I ‘ X X
(49), (50), can be isolated depending on the reaction temperature.’O The 1,4,2-dioxaphospholanes are slowly transformed into the 1,3,2-dioxaphospholanes in solution: (47), (48) + (49), (50). The caged 2 :1 phosphoranes from biacetyl, (42), (43), (44), have not been studied in sufficient detail. Their relevance to the problem of the mechanism of TBP-permutational isomerizations will depend on the
82
FAUSTO RAMIREZ AND IVAR UQI
OCHa
I
I
1
1
H-C-C--H
csF5 &Fs (49, cia-H/H; 50, trans-H/H)
experimental demonstration of the regular us. the irregular nature of the process. B. Fariable-Temperature N . M .R. Spectra of Phosphoranes 1. Acyclic systems
The fluorine atoms Of PF5 (7) and of (CH,)2NPF4(8)undergo exchange among apical and equatorial positions above - 100°.18-21,26, 63, 8o I n the latter case, the amino group remains equatorial during the exchange ; see Section IV. An analogous situation is observed in (CH,)PF4, (23);19 in this case, CND0/2 calculations show that the TBP with apical-CH,, (23'), is about 39 kcal mole-1 more energetic than the isomer with equatorial-CH,; see Scheme 3 in Section VI.6*The F-exchange in (8)and (23)can occur equally well by TR or BPR, because energy-rich TBP with apical (CH,),N- or CH,-groups can be avoided in both cases. This is accomplished in TR by the use of equatorial and apical fluorines as members of the pair, and by utilizing only those directions of rotation of the trio which move the organic group from one equatorial position to another. I n BPR, the organic group is the pivot. Both TR and BPR are in agreement with the experiment of Whitesides and Mitchell.8o The three fluorines of the phosphoranes (51)and (52)give rise to one 1°F n.m.r. signal above + 25°19-21* 2 4 suggesting permutational isomerization. The evidence favors, but does not demand, a regular process (bond-deformation) rather than an irregular process under the conditions of those observations. If the positional exchange of the ligands involves a regular process, it can be seen that a t least one chlorine
PENTAVALENT PHOSPHORUS C O M P O U N D S
83
F (23)
,?*
F
\
.*'
(7: X = F)
(8: X = N(CHs)z)
(23: X = CH3)
atom or one CF,-group must be transferred to an apical position as a result of a single TR or of one BPR. The multiple TR process can account for the observations without violations of the polarity rule.
F
F
(51)
(52)
&I
The multiple TR processes would be represented as follows.
X
F'
F'
F
&' F
F'
(TR9
_iF
X
X
4: F'
X
F'
F' (51: X
= C1)
(52: X = CF3)
,F
84
FAUSTO R A M I R E Z A N D I V A R U B I
Earlier reportsls-zl indicated that at temperatures below ca. + loo", the lgFn.m.r. spectrum of certain trifluorophosphoranes of the type (53), (54) and (55) had two signals in the intensity ratio 2 :1, suggesting that these compounds did not undergo permutational isomerizations by regular or by irregular processes under those conditions. More recent evidence which discloses the occurrence of irregular processes in phosphorane (55) will be discussed in Section IX. As far as the regular process is concerned, the arguments would be analogous to those The results of the presented in connection with ClzPF3and (CF3)2PF3. single and the multiple TR in these systems can be visualized by the appropriate permutation of the ligands, using as pairs, (X, F'), (Y, F') and (F,F'). No final conclusions can be reached at this time concerning the energy changes associated with regular and irregular processes, or with regular processes by BPR and by single or multiple TR, in all these cases.
A:
X F'
Y F'
(53: x = Y = "(CaHa)a) (54: X = N(CzH&; Y = CHa) (55: X = Y = CHI)
2. Monocyclic systems
(a) Fluorophosphoranes. There seems to be a significant difference in the behavior of five- and six-membered cyclic trifluorophosphoranes with two carbon-ligands as part of the ring. The fluorines undergo'g*2o positional exchange in the five-membered phosphorane (56) above - 70". However, there is no fluorine exchange in the six-membered analog (57) even a t much higher temperatures. These observations are consistent with the hypothesis that the five-membered ring is forced to occupy an apical-equatorial position, in violation of the polarity rule, while the six-membered ring occupies a diequatorial position in accord with the polarity rule. Strain considerations could be partly responsible for the difference. The "envelope" shape of the five-membered ring is capable of accommodating the 90" CPC angle in the apical-equatorial situation, without much strain.68 The expansion of endocyclic bond angles to 120' in the diequatorial situation (56') requires considerable energy. The analogous situation for an OPC angle has been evaluated by CNDOIB calculations68as shown in
PENTAVALENT PHOSPHORUS COMPOUNDS
85
F
CHg-CHz
F
I
F (56)
(57)
Scheme 3, Section VI. The diequatorial placement of the six-membered ring, (57) us. (57'), favored by the polarity rule and for steric reasons by the bulk of the puckered six-membered ring, is not opposed by ringstrain considerations. The positional exchange in the five-membered
iF-
7
F
(56')
(57')
compound can occur by TR using the ring carbon as the pair. In this representation an attempt is made to approximate the changes which result from the overall 60" internal relative rotation of pair us. trio.
F'
l?
(56)
(b) Four-membered oxyphosphoranes. Four-membered cyclic oxyphosphoranes are capable of undergoing permutational isomerizations on the experimental by regular and irregular p r o c e ~ s e 6s6~depending ~~ conditions and on the stability of the pentacoordinated phosphorus. The stability increases with the number of highly electronegative elements, e.g., oxygen, that are attached to the phosphorus. Derivatives of the 1,2-oxaphosphetane ring system (58), which are synthesized by the method described in Section 11, illustrate the behavior of these systems.6e The preferred configuration is that in which the ring occupies
86
FAUSTO RAMIREZ AND IVAR UQI
an apical-equatorial position with the oxygen apical and the carbon equatorial. This follows from the stereopolarity rule1*,62, 63 which also determines the placement of the remaining ligands X, Y and Z. The X-ray crystallographic analysis of one of these compounds was discussed in Section I1;59 here we are concerned with the behavior of the molecules in solution.
Lo
CHs-0
I I ..-*x CHa-P'
x
I b Y 2
The regular permutational isomerization of the oxaphosphetane by the TR mechanism should occur most efficiently when the ligands of the pair, in the pair-trio combination, are provided by the ring. I n this manner, the motion of the ligands is minimally hindered by the presence of the four-membered ring. A single TR will always make the ring-oxygen equatorial and the ring-carbon apical in violation of the polarity rule. The single TR also places either ligand X or Y apical. The (TR)2retains the energetically favorable position of the ring in the TBP, with apical oxygen, but requires the placement of either X or Y in an apical position. These considerations are illustrated by an analysis of the stereochemistry of the oxaphosphetane (59),and by observation of its variabletemperature lH n.m.r. spectrum. Phosphorane (59) has four oxygen ligands which provides considerable stability to the pentavalent state of the phosphorus. OCH2. CeH5 C&0,
,OCH(CF&
0/P\C,CHa
\ / H ' CFa/"CFs (59)
The skeleton of the Reference Molecule ME is drawn as prescribed in Section 111. The ligands are indexed in the numerical order prescribed by the Sequence Rules.7s The result in the case of compound (59) is the TBP (59'-E). This is an awkward drawing that can be made clearer by
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
87
the simple expedient of rotating the Reference Skeleton 180' around the axis that passes through the phosphorus and skeletal position 2. The new drawing of the Reference TBP of compound (59)is (59-E). CFs
?Ha
cFsjqTZC 'OCH(CF3)a H&6H5 4
Reoriented skeleton of Mp
OCHs
(59-E)
The isomer M,, 4) is obtained by the permutation (24)of the ligands of the Reference Molecule, without changes in the indices of the Reference skeleton. Placing the appropriate ligands on the appropriate skeletal positions results in the convenient drawing (59-(24))for this isomer.
55
4 2
The Reference TBP (59-E)is not an energetically favorable isomer because it has equatorial oxygen and apical carbon. The same is true for isomer (59-(24)) which is derived from (59-E)by the exchange and L4 =-OCH,. However, of (ea)-ligands: L2 =-OCH(CF,),, (59-(14 2 5)) is an energetically favorable isomer . (Note the operation of the Permutational Notation where ligand 1 goes to skeletal position 4, 4
88
FAUSTO RAMIREZ A N D I V A R U O I
61 41
Ma 4 2 5)
(59-(14 2 5))
ligand 4 goes to skeletal position 2, etc. as given by the notation of the TBP in question.) I n this example, of the 20 isomers of the TBP, two are forbidden because the ring cannot be diapical, and six are forbidden because the four-membered ring cannot be diequatorial. Of the twelve remaining isomers, six fall in one class shown in Scheme 5 , and six fall in another class shown in Scheme 6. SCHEME5
Compound (59) qF3
qF3
------' cF3Q0cH3 H..
4------
CH3 OCH3
E
OCHaCeHa
4------
---_--
FH3
PENTAVALENT PHOSPHORUS COMPOUNDS
89
SCHEME 6 Compound (59)
Figure 16 shows the isomerization itinerary by the TR process. The energetically favorable isomers, with two apical oxygens are shown by solid circles 0 ,while the unfavorable isomers are shown by open circles 0. The broken lines (-----) in Fig. 16 represent isomerizations by a single TR, using the ring as the pair. The solid lines in Fig. 16 represent isomerizations by (TR)2. The variable-temperature n.m.r. spectrum of (59) discloses the occurrence of some permutational isomerizations, but there is no inversion of the configuration at 113 The data can be explained by the (TR)'-process, which bypasses the unfavorable TBP E, (1 3 4) and (1 2 4) of one class (Scheme 6) and the unfavorable TBP (2 4), (3 4) and (1 2) of the second class (Scheme 6). The (TR)2-process, does not allow the inversion of configuration of the phosphorus, i.e., it
90
FAUSTO RAMIREZ AND IVAR U G I
E
(34)
(134)
(12)
FIa. 16. Itinerary for permutational isomerizations of phoaphorane (59) (Schemes 6 and 6). Dmhed line (---) =single TR. Solid line (-) =double TR.
does not allow the interconversion of (I 4 2 5 ) and (1 4) (2 5 ) ,in agreement with the n.m.r. data; cf. Fig. 9 in Section V. Another oxaphosphetane, this time with only two oxygen ligands to
phosphorus, is (60). This type of phosphorane is considerably less stable than the analog (59). There is no evidence for permutational isomerization of (60), below +looo by regular or by irregular processes,6a provided that the compound is obtained in high degree of purity and is protected from moisture. It will be shown in Section IX that catalytic amounts of the relatively acidic hexafluoroisopropyl alcohol,
PENTAVALENT PHOSPHORUS COMPOUNDS
91
(CF,),CHOH, causes the interconversion of (60-(1 4 2 5)) and (60-(1 4) (2 5)) by an irregular process. The alcohol is generated by the rapid hydrolysis of (60) ;hence, the need for strictly anhydrous conditions to demonstrate its stereochemical stability. QF3
I41 = -0
L4 =
-c
\",CF3 T'CF3
FH3
CFswH
IH '
1
CH3
CHzCHs
(60-E)
CF3
CH3 c:qcH2cH3 ,*c6H5 OCH(CF3)z (60-(I 4 2 6))
c 2 . q c H CeH5 CH3 z c H 3 OCH(CFa)a ( 6 0 4 4) (2 6))
The molecular structure of phosphetane (6) was elucidated by X-ray crystallographic analysis, as discussed in Section II. The preferred isomer is (6-(1 4) (2 5 ) ) ,which is shown in two space orientations in order to illustrate the application of the Permutational Notation. The ligand indices are prescribed by the Sequence Rules.76 The ligand and skeletal indices must coincide in (6-E), which thus gives the orientation of the skeleton. The permutation (1 4)(2 5) gives isomer (6'-(1 4)(2 5)) as a QF3
7H3
CF3wH OCH (CF3)2
c 6 H s q ;
CF3 (6-(1 4)(2 6)
92
FAUSTO RAMIREZ A N D IVAR UGI
rather awkward drawing. A 180O-rotation around the axis that passes through phosphorus and one phenyl gives (6-(1 4) (2 5)). (c) Five-membered oxyphosphoranes. One type of five-membered cyclic oxyphosphorane is represented by derivatives of the 1,2-oxaphospholene-4 ring system (64), (65) and (66). These can be madell* by the reaction of phosphite, phosphonite and phosphinite esters, (45), (61) and (62), with some a$-unsaturated ketones, for example (63).
H
7
R-P-OCHS
I
-I-
7c-c PHs I
I I1
CHa M I I s
0 (45) R = R' = OCHa (61) R = OCHa; R' = CEHS (62) R = R' = CEHS
(63)
R, o , p .
d
R'
I
I ,OCHa
c=c I
9
YxCEE5
CHI A-CHs
II
0 (64) R = R' = OCHa (65) R = OCHs; R'= CsHs (66) R R' = CEHS
-
Tertiary phosphines (67), give dipolar ions (68), and not phosphoranes in this reaction, presumably because the decrease in the number of electronegative oxygens attached to the phosphorus decreases its tendency to become p e n t a ~ o v a l e n t . l ~ ~ -The l l ~ 31P n.m.r. chemical shift in (68) is 6 = - 10-9 ppm.
The phosphorane (65), with three oxygen ligands, has been made from dimethyl phenylphosphonite (61). The stereochemistry of phosphorane (65) can be analyzed in a systematic way by means of the Permutational Notation, as illustrated in Scheme 7. The first step is the ordering of the five ligands as prescribed by the Sequence Rules76(i.e., "first atom" of ligand with highest atomic number comes first, etc.) The second step is to create the Reference Molecule E, utilizing any orientation of the Reference Skeleton that generates a drawing with desirable characteristics; for instance, in this example the five-membered ring has been placed in a way that emphasizes its use as the pair in the pair-trio
PENTAVALENT PHOSPHORUS COMPOUNDS
93
SCHEME 7 Compound (65)
,.OCHa 0 CSHSH 4OCHa CaHs (16)
combination of the TR-process (the trio consists of CH,O, CH,O and CBH6). Phosphorane (65) has two equivalent ligands, Lz=L, =-OCH,,
94
FAUSTO RAMIREZ AND IVAR U G I
therefore the twenty permutational isomers of the general case (cf. Table 3 in Section 111)are reduced to ten (cf. Table 4 in Section 111). Moreover, the five-membered ring reduces this number still further. If, as shown in Scheme 3 of Section VI, the five-membered ring cannot be placed diequatorial, three additienal isomers are excluded. Isomer
CHsO
OCHs
CHsO+;!$--ms
1
CH3 CH3
H
'
CeH5
(65'-(2 6))
0
CHs
'CeHs CsHs (654%6 ) )
(65-(2 5)) is derived from E in the usual way. The derivation of the other two forbidden isomers (65-(3 6)) and (65-(2 4) (3 5 ) ) , should be obvious (they are shown in one common skeletal orientation only). A fourth
permutational isomer, (1 4), is excluded because in it the ring would have to be placed in a diapical position. This can be verified by performing the permutation (1 4) of the ligands of the Reference MB (note that ligand 1 is part of the ring, and must replace ligand 4 which is apical in ME). The four unfavorable TBP do not play any role in the permutational isomerization of compound (65), either as participants in the equilibrium with measurable concentration, or as fleeting intermediates (vide infra). The TBP to be considered are the six isomers shown in Scheme 7,which are all capable of existing as pairs of enantiomers as a result of the chiral
PENTAVALENT PHOSPHORUS COMPOUNDS
95
center at a ring-carbon. Of the six isomers in Scheme 7, four violate the polarity rule, since they have one or more carbons in apical positions ; these are (65-E, (3 4), (1 5) and (2 4)). Only two isomers are favorable according t o the stereopolarity rule: (65-(1 4) (2 5)) and (65-(I 4)(3 5)). Consider now the variable-temperature proton n.m.r.116-118 of phosphorane (65). To simplify the discussion, the signals identified with the isomer that has the ci8-C6H6/C6H6 relationship, (65-(1 4) (2 5 ) ) will be designated with an asterisk (*), if necessary. At - 25", there are two coupling) for the benzylicsignals (two doublets with different H-C-P H, two signals for the acetyl-H, two signals for the CH,C-H (superimposed), four signals (four doublets, each doublet due to H4-0-P coupling) for the (CH,O),, (CH,O),Y, (CH,O), and (CH,O),*. There is little difference in the chemical shifts of (CH,O), and (CH,O),*, and the two doublets are superimposed. At + 25", the difference in the chemical shifts for the benzylic-H, the acetyl-H and the (CH,O), protons in the two diastereomers begin to vanish; i.e., one sees little differentiation between the signals for (CH,O), and (CH,O),*. The 31Pn.m.r. spectrum in this temperature range shows two 31P-signals, one for each diastereomer. At +52", the lH-n.m.r. of (65) shows one signal each for benzylic-H, acetyl-H, and CH,C-H, and the signals for (CH,O) and (CH,O)* begin to coalesce. At +104", there is one sharp doublet for benzylic-H, and sharp singlets for acetyl-H and CH,C-H. This last observation is important because it shows that the positional exchange occurs by a regular process. It can be shown that at higher temperatures, above ca. + 114", the irregular permutational isomerization, involving the rupture of the ring P-0 bond begins to intrude.
At +104" one sees two sharp doublets due to (CH,O) and (CH,O)* groups, which are, therefore, not fully equivalent (the two doublets are superimposed and resemble a pseudotriplet). These data show that at -25", the equilibrium (65-(I 4)(3 5) 65-(I 4) (2 5)) is slow. At + 25", the rate is such as to begin to influence the lH-n.m.r. spectrum, but apical and equatorial methoxy-groups are
96
BAUSTO RAMIREZ A N D I V A R U G I
still distinguishable. Above + 52O, the interconversion of the diastereomers is so fast that there is no longer any distinction between apical and equatorial positions, as far as the protons of the benzylic, the acetyl and the CH,O-groups are concerned. The protons of (CH,O) and (CH30)* are still distinguishable, but there are no distinctions for (CH,O),, (CH,O),, (CH,O): and (CH,O),*. The interconversion of the two diastereomers corresponds to a configurational inversion of the pentacoordinated chiral center, or an exchange of the equatorial phenyl and methoxy-groups. An interpretation of the n.m.r. data for compound (65) by the BPR or the single TR mechanisms requires the formation of three unfavorable TBP, (65-(3 a), (1 5 ) and (2 a)),or of the worse E, as shown in Scheme 7. Note, in particular, that in E and in (1 5) the two methoxy-groups arrange themselves with mirror-image symmetry with respect to both the two equatorial and the two apical skeletal positions. The double-TR, or (TR)2-mechanismleads to a different conclusion, as can be seen in the itinerary of Fig. 17-1, where the solid line represents the (TR)2. This is, in fact, the only interpretation of the experimental data that avoids the need to involve the energetically unfavorable TBP: (1 5) and E, and (3 4) and (2 4). Moreover, the (TR)2does not require any involvement of TBP like (65-(2 5 ) , (3 5 ) and (2 4) (3 5 ) ) ,where the five-memberedring would be diequatorial. The (TR)2shown in Scheme 7 and in Fig. 17-1 utilizes the two ring-ligands as the pair in the pair-trio combination of the TR-mechanism. I n fact, this feature is common to all the TR-processes invoked i n this review to explain the experimental data on positional exchange of ligands in four- and Jive-memberedcyclic phosphoranes. Prior to the conception of the TR-mechanism, the appearance of two methoxy-signals in the proton n.m.r. spectrum of (65) in the temperature range +70° to + looo, where isomers (65-(I 4)(2 5)) and (65-(1 4)(3 5)) are undergoing rapid regular permutational isomerization, was taken as an indication that TBP such as (65-(2 5), (3 5) and (2 4)(3 5)) with a diequatorial five-membered ring were excluded from the isomerization itinerary.1° The same conclusion was reached from an analysis of the n.m.r. ~ i g n a l - f o r m .The ~ ~ new conception of the multiple TR process,66-69applied in Fig. 17-1, avoids the involvement, not only of those TBP, but also of all the other isomers that violate the polarity rule. The phosphorane (70) shown in Scheme 8 is analogous to (65) discussed in Scheme 7, except that the dimethyl-compound (70) lacks the chiral center in one of the ring carbons,77which is present in (65). The n.m.r. spectrum of (70) shows77two methoxy signals at - 8”, but only one such signal a t ca. + 62. The interpretation of the data is quite simple : below certain temperatures, (70) exists as two “frozen” isomers, (70-(1 4) (2 5 ) )
PENTAVALENT PHOSPHORUS COMPOUNDS
m
97
(14)(35)
(14)
Ip
FIQ.17. Itineraries for isomerizations by TR-mechanism. Single TR =---; double TR=-; triple TR=-----. Graph I for (65), Scheme 7. Graph I1 for (79),(80) and (81). Scheme 9. Graph I11 for (84)and (85),Scheme 11. Graph IV for (86), Scheme 12. Graph V for (87), Scheme 13.
98
FAUSTO RAMIREZ AND IVAR UGI
and (70-(14)(3 5)); above certain temperatures the positional exchange takes place by the double TR process. Fig. 17-1is adequate to explain the data and brings out the similarities and the differences between (70) and (65): the pair of enantiomera (70-(14)(2 5)) and (70-(14)(3 6)), correspondsto the pair of diastereomers, (65-(1 4)(2 5))and (65-(1 4)(3 6)). The result of the positional exchange should be one CH30-signal in (70) and two CH,O-signals in (65). A second type of five-membered cyclic oxyphosphorane is represented by derivatives of the 1,4,2-dioxaphospholane ring.lll These are made from the reaction of trivalent phosphorus compounds with certain aldehydes, for example :111
c6HS
I
CHsO-P-OCHa
0
+ 2 CII I
H
CdS
,c-0
H I
Two phosphorsnes (71)and (72),which differ in the configuration of the ring-carbons, are isolatedlO from the reaction of dimethyl phenylphosphonite (61) with pentafluorobenzaldehyde (46). Consider first isomer (71)with the &-arrangement of the hydrogens in the ring. There are two energetically favorable TBP: (71-(14)(3 6))and (71-(14)(2 5)); these can be derived from the Reference (71-E), which is an unfavorable TBP, by the Permutational Notation.? Below - 60" one observes
c6H5
(71-E)
t (1) Amign indices to the ligands by the Sequence Rules. (2) Select the particular orientation of the Reference Skeleton that will provide the desired drawing. (3)Draw the Reference E by placing ligand L1 on skeletal position 1, etc. (4) Draw the isomers by carrying out the corresponding permutation of the l47gartd.9 (name of isomer) without changing the skeleton in any way. ( 5 ) If desirable,reorient the entire TBP which represent8 the new isomer, to relate its drawing to that chosen for the Reference drawing. For complex molecules it is more economical to omit the actual Reference Molecule, e.g., (71-E) and operate with ligand and skeletal indices, L1. Ls, and 1 . 5 , and then make replacements on the desired isomers in the desirable orientations.
..
..
99
PENTAVALENT PHOSPHORUS COMPOUNDS
OCH3 (7141 4) (2 6))
(7141 4 ~ 6)) 3
proton signals for two methoxy groups, probably those of the "frozen" (71-(1 4) (2 6 ) ) isomer with the UT~I%-C,&/C~E'~relationship. At about - 42" those signals coalesce and at higher temperatures ( + 30") one sees a new set of two methoxy signals. The two methoxy groups are not fully equivalent when there is permutational isomerization, just as in the case of (65) (Scheme 7). The trans-H/H isomer (72) is more flexible than the corresponding cis-HIH isomer. At + 30", one observes proton signals for two methoxy groups of (72), which are due to the rapidly interconverting isomers
OCH3
OCH3
TABLE16 Coalescence Temperature of CH3O-Signals in the 1H N.M.R. Spectra, of Five-Membered Cyclic Oxyphosphoranes Compound No. 6 31P (p.p.m.)a (47) (61)
(72) (71) (65)
(66)
Coalescence temperature
Phosphorme from PFB-a. ZI (CH3O)sP&-H/H BPD-C. 6 (CH30)sP PFB-as C (CH&)2PCeH5 trans-H/H PFB-apf ( C H & ) Z P C ~ H cis-H/H ~ BPD-C. ( C H ~ O ) ~ P C E H ~
+34.1 27.8 34.0 28.6 16.7s
+ + + + + 13.3 +26.7
BPD-h (CH.gO)P(CeHs)a
ca. -110" ca. -40"
+ None below + 100"
None above 70" ca. -42' ca. 0 ' and ca. 62'
a PFB =Pentafluorobenzaldehyde. For isomer with trana-H/H, 6 3lP= + 38.2 p.p.m.; all signals V8. &PO aa zero. b I n CFaClz solution from 30 to - 140'. C BPD = 3benzylidene-2,4-pent~e~one, CH.g.CO C(CH .C~HF,) .CO .CH3. In CDCls from 30" to - 90". a In CDCls from 30" to - 70'. f In CDC1.g from + 30" t o - 70". 8 Isomer with 6 = -I-16.3 p.p.m. assumed to have anti-CeH5/C&hiconfiguration. Isomer with 6 31P= 13.3 p.p.m. assumed to have 8yn-C6H5/C~&configuration. I n CDC1.g from 30" to - 70". In C12C=CCla from + 30" to + 110'. Without solvent from + 110" to 151'. I n o-dichlorobenzenefrom +25O to + 120". See Ref. 9 and 10.
+
+ +
+
.
+
+
100
FAUSTO RAMIREZ A N D IVAR U a I
(72-(14)(3 5) and (1 4)(2 5)). Now, no coalescence of the two CH,Osignal is observed down to 70°, where low solubility prevents further measurements. The n.m.r. spectra of (71)and (72)show no changes in the temperature range + 30" to + loo", at which point irreversible decomposition of the molecules become apparent. The n.m.r. data of these and of related phosphoranes are summarized in Table 15. The syntheses of (47)and of (64)were described above.lll3 114 Their variable temperature proton n.m.r. will be discussed in the appropriate sections corresponding to the symmetry properties of the molecule. Note, for further reference, the Permutational Description? of the energetically preferred isomer (47-( 1 5))of the 1,4,2-dioxaphosphol~newith the cis-H/H configuration, made from the reaction of trimethyl phosphite with pentafluorobenzaldehyde.
-
A third type of five-membered cyclic oxyphosphorane is represented by derivatives of the 1,3,2-dio~aphospholene~-~~~ l 3 and the 1,3,2dioxaphospholaneSring systems. These compounds are made from the reactions of trivalent phosphorus compounds and certain mono- and dicarbonyl functions, (73). Some a-dicarbonyl structures furnish R R'
I
R-P-OCHa
(45,61,62)
P P
fC-C
I
I
RU R-
I ,OCHs
R,
* o/p\o I 1 c=c I I R"
(73)
Rw
(74)
stable 1 :1 adducts with pentavalent phosphorus (74).Further reactionS of the phospholene (74),with the same dicarbonyl compound (73),or with another carbonyl function (75),leads to the phospholanes (76).
t See footnote on p. 98. $ See footnote on p, 27.
PENTAVALENT PHOSPHORUS COMPOUNDS
101
R'
R
I ,ocIIs
R ,
0 o/p\o + - I l l I CC-C-R" I I I
R"
R" R"
(74)
(75)
(76)
A variation of the synthesisa is the one-step combination of two molecules of a highly reactive carbonyl compound (77), e.g., p-nitroI o 8 with one of the trivalent benzaldehyde112or hexafluor~acetone,~~'~ phosphorus compound to give the phospholane (78).11s, I2O
R'
0
I
I1 + 2 C-R" I
R-P-OCH3 (45,61, 62)
R"
T
R,
,OCHs
o/p'o __f
I
I
R"4-CR"
I
I
R"
R"
(77)
(78)
These procedures can be used to makelo the three closely related phosphoranes, (79), (80)and (Sl),which can be described in terms of the
.q;:
Ri L1= -0-C-RZII LZ= -o+R4
I
4 4
R3 L3 = L4 = -OCH3 L5 = -c&5
ME (79) R1= Rz = CF3 ; R3 = COCH3; R4 = CH3 (EO), R1= COCH3; Rz = CH3; R3 = p.NOz-CaH4; R4 = H (81, R1= CH3; Rz = COCHs; R3 = p .NOz-CgH4; R4 = H
following assignment of ligand indices.? Scheme 9 depicts the six isomers with apical-equatorial ring. Two of these (1 4) and (2 4) are energetically unfavorable by the polarity rule. The n.m.r. spectra of (79), (80) and (81) show signals for two different methoxy groups in the entire range from - 90' to + 20". The data can be explained by the TR and (TR)2-processes as shown in Fig. 17-11. The ring provides the
t See footnote on p. 98.
102
FAUSTO R A M I R E Z A N D I V A R U B I
pair-ligands, and there is no need to violate the polarity rule or of placing the five-memberedring diequatorial. The situation in these phosphoranes (79), (80), (81), is very similar to that in phosphorane (65), except that, in the case of (79),(80) and (81)the placement of the ring in several apical and equatorial positions results in isomers of comparable stability. SCHEME 9
Compounds (79), (80) and (81)
R4z\0+rH7 OCH3
OCH3
The two tetraoxyphosphoranes (82) and (83) in Scheme 10 have in common a cis relationship of two identical substituents on the fivemembered ring. There are some similarities between (82), (83), and the derivative of the 1,2-oxaphospholene (70) of Scheme 8, since in both cases, the five-membered rings (including its carbon-substituents) have
PENTAVALENT PHOSPHORUS COMPOUNDS
103
a local reflexion symmetry. The n.m.r. spectra of the cis-phospholanes (82) and (83) give only one CH,O-signal at + 20". This is to be expected from the occurrence of TR and (TR)2 processes. Note that in the case of (82) and (83) the graph in Fig. 17-11 (which was applied to compounds (79), (80) and (81)) is simplified because there are only two favorable TBP : (1 4 5) and (1 5). Replacing (1 2 5 ) by (1 4 5 ) and (2 6 ) by (1 5) in Fig. 17-11 produces the graph that is applicable to (82) and (83) (Scheme 10). Now, a single TR causes inversiont of the entire molecule, while the (TR)2 causes the direct isomerization: (82) or (83-(1 4 5) (1 5)). SCHEME 10 Compounds (82) and (83) Ri L1= -0-C-R2
Lz = -0-C-R2
I
L3 = L4 = -OC&
I
LS = -c6H5
I
Ri
(82) R1= COCH3; Rz = CH3 (83) RI = p .NOa-CsHr;
Rz
=H
The two phosphoranes (84) and (85) in Scheme 11 have a transrelationship of two identical substituents on the ring. These phosphoranes, (84), (851, have local Cz symmetry, while the phosphoranes (82) and (83) of Scheme 10 have local reflexion symmetry in the five-membered ring. The n.m.r. spectra of the trans-phospholanes (84) and (85) have two methoxy-signals above + 30". Now TR and (TR)'-processes lead to the In phosphoranes (82) and (83)of Scheme 10, the permutations ( 1 2 6)and (1 4 6)of the ligands in the Reference Molecule E, generate two enantiomers (of the entire molecule), both of which have the descriptor (1 4 6). They have the same descriptor because they represent the same configuration of the TBP-phosphorus. They are enantiomers due to the chiral centers in the ligands. The same situation applies to permutations (2 6)and (16) of the ligande in E ;both result in the enantiomer pair ( 1 6).
104
FAUSTO RAMIREZ AND IVAR UOI
same isomcrization, as can be seenin Fig. 17-111. Here, the permutational isomers (2 5) and (1 5) are diastereomers and not enantiomers and, consequently, the two methoxy-groups remain distinguishable. Note that Fig. 17-111 is actually a simplification of Fig. 17-11) when isomers (1 2 5) and (1 5) correspond, and isomers (1 4 5) and (2 5) also correspond. SCHEME 11 Compounds (84)and (85) R1
I I LZ= -0-C-RZ I LI= -0-C-RZ
R'
The n.m.r. spectrum of the trioxyph~sphorane'~(86) Scheme 12 shows signals for two different benzylic protons at -75") which corresponds to a frozen structure of the pair of enantiomers (86-(2 4)(3 5)) and (86-(1 4) ( 3 5)). The n.m.r. spectrum above 30" shows only one benzylic-protons signal. Note that the inversion of the enantiomers is accompanied by an exchange of the apical and equatorial benzylic protons. Fig. 17-IV and Scheme 12 show quite clearly that the (TR)sprocess with the ring functioning as pair, is the only way of avoiding the placement of a carbon in an apical position. The case77of (86) (Scheme 12) is simply a variation on a previous'l' trioxyphosphorane (87) (Scheme 13). This can be demonstrated by a comparison between Fig. 17-IV) which represents the isomerizations of (86) (Scheme 12),and Fig. 17-V, which represents the positional exchange in (87) (Scheme 13). Fig. 17-V is simpler because ligands L, and L2 are equivalent in (87). Isomers (87-(1 4)) and (87-(2 4)) are energetically unfavorable. To obtain positional exchange of the ligands associated with the ring, avoiding unfavorable isomers, one must invoke a (TR)3-
+
PENTAVALENT PHOSPHORUS COMPOUNDS SCHEME
105
12
Compound (86)
R R
(86) R = p.NOz-CeH4
process (---.- in Fig. 17-V). The n.m.r. spectrum of this phosphorane (87) (Scheme 13) shows only one signal for the two methyl groups attached to the ring-carbons, in the entire temperature range from - 60" to +160°. The tetraoxyphosphorane'O9 11' (88) in Scheme 14 exists as two favorable chiral TBP, (2 5) and (1 5). The TBP (1 4) is unfavorable
according to the polarity rule. The n.m.r. spectrum of (88) has one signal for the two CH,O-groups, and one signal for the two methyl groups attached to the ring-carbons, in the entire temperature range of - 60' to + 160'. The interpretation of the data is apparent from Fig. 18-1. The single TR and the double TR accomplish the same thing. SCHEME 14 Compound (88)
Note that the closely related trioxyphosphorane (87) in Scheme 13, with the isomerization itinerary of Fig. 17-V, has only one favorable uchiral TBP, (87-( 1 4) (3 5)), and two unfavorable TBP, and requires the (TR)S-mechanism if unfavorable isomers are to be avoided in the interpretation of the n.m.r. data. Moreover, the phosphoranes (84) and (85) in Scheme 11, which give two CHsO-signals in the n.m.r. spectrum should be compared with (88) in Scheme 14. Note that Fig. 17-111 (applicable to (84) and (85)) and Fig. 18-1 (applicable t o (88)) take the same notation. However, TBP (84), (85-(2 5) and (1 5)) represent diwtereomers, while TBP (88-(25 ) and (1 6)) represent enantiomers.
PENTAVALENT PHOSPHORUS COMPOUNDS
107
The TBP-phosphorus atom with three equivalent ligands can no longer be a chiral center. As a consequence, the positional exchange of ligands in the l,&oxaphospholene (64) (Scheme 15)reported a few years
m
(14)
1p
(14)
FIQ.18. Itineraries for isomerizationa by TR-mechanism. Single TR=---; double TI&=-- ,triple TR=-----. Graph I for (88), Scheme 14. Graph I1 for (64) and (89), Scheme 16. Graph I11 for (90), Scheme 16. Graph IV for (91). Graph V for (109), Scheme 19.
ag0,114-118and that of the analog (89) reported very recently,77can be explained by exactly the same itinerary shown in Fig. 18-11. The n.m.r. spectrum of (64( 1 5 ) ) (see Table 15) at - 65" shows three CH,O-signals. The n.m.r. spectrum of (89-(1 5)) at -28" shows two CH,O-signals. These are due to the "frozen TBP ". Above + 25" or + 5 7 O , respectively,
108
FAUSTO RAMIREZ A N D IVAR U G I
one sees only one CH,O-signal in both cases. The difference between (64) and (89) in the “frozen” state (1 5) is simply the environment, R1, Rz of the two equatorial CH30-groups. The situations are otherwise analogous. Note the possibility of achieving positional exchange by (TR)2(solid lines in Fig. 18-11) without involving unfavorable TBP, E, or TBP with a diequatorial ring. SCREME15 Compounds (64) and (89)
The n.m.r. spectrum of the cis-l,3,2-dioxaphospholane(90) in Scheme 16 shows one signal each for the CH30-, CH,CO-, and CH,-C-groups, a t
+ 30°,which corresponds to positional exchange of ligands according to the itinerary of Fig. 18-111. SCHEME 16 Compound (90)
4, La in Ring; La = L4 = La = -0CHj
The trans-isomer (91) has local Cz-symmetry of the groups on the ring, instead of the local reflexion-symmetry in the related cis-isomer (90). The n.m.r. spectrum of the trans-isomer (91) shows only one signal each
PENTAVALENT PHOSPHORUS COMPOUNDS
109
for the CH30-, CH3CO-, and CH,-C groups, as expected from the itinerary of Fig. 18-IV.
2. Spirobicyclic systems’
l-l 30
Spirooxyphosphoranes (94), are availabls from the reaction of cyclic phosphites (92), with u-dicarbonyl compounds (93).l2I
(92) n = 2,3
(93)
c=c I
R (94) n
I
R’ = 2,3
A second approach is based on the transph~sphoranylation~~, 129 reaction, (95) + (96), in which the phosphorane function is transferred from one alcohol to another. I n some of these reactions the ring of an oxyphosphorane is preserved.129
When a &functional alcohol is involved, the result is a spirooxyphosphorane (95) --f (94). This method suffers from limitations related to the structure of the diol, but it can be useful at times when the cyclic phosphite required in the previous synthesis (92)+ (93) + (94) is unavailable.
110
FAUSTO RAMIREZ AND IVAR U O I
OCHa CHsO,
p””:: /o
1 ,OCHa
yPLT+
c=c I
R
I
R’
WHah / \ OH OH
0 ,
P-OCHs
__f
ON ‘ 0
I c-c I
1 I
R’
R 195)
+ 2 CHSOH
(94)
Transphosphoranylation, (97) -+ (98), occurs very readily among phosphoranes derived from phenol and catech01.*~
+ 4 CoHsOH
(97)
(98)
The scope of the transphosphoranylation reaction is enlarged by the availability of cyclic oxyphosphoranes (99) from the reaction of phosphites with certain dialkyl peroxides.l* Several cyclic and acyclic
(92) = 2,3
(99) n = 2,s
phosphoranes, e.g., (99), (loo), have been prepared in this manner.l‘ The reaction of alcohols with phosphorus pentachloride is not a practical entry into oxyphosphorane chemistry, although the analogous reaction of phenols can be made to yield oxyphosphoranes, e.g., (97) under carefully controlled c~nditions.~’
111
P E N T A V A L E N T PHOSPHORUS COMPOUNDS
Spiro-oxyphosphoranes (lor), with the highly electropositive ligand hydrogen have been made from the reaction of trialkyl phosphites (or of riaminophosphines) with glycols.lz2
“9; R R
(101) R
5
CHa
een A very closely related type of spiro-oxyarserane, (102synthesizedlZ4by an extension of the known reactions of tetracoordinated arsenic compounds, namely, the reaction of arsenic and arsonic acids and esters with glycols.125 0
It
I
Rt/*’\OH OH R-OH R’ = cfl6
+2
R
R
R OH
W OH
R
,,R’
+ 3 HBO
R=H R = CHa (102) R = H; R’ = OH (103)R = H; R’ = OCHs (104) R = CHs; R‘ = OH
(105) R = C&; R’ OCHs (106) R = CHs; R‘ = CsHa
Static TBP of type (101), and (104-106), should give rise t o four proton signals due to the four CH8-Cgroups. These signals are associated with groups R, --- R4as shown in Scheme 17. This analysis also applies to static TBP : (102) and (103), with hydrogens directly attached to the rings. Phosphorane (101), with four electronegative oxygens and one electropositive hydrogen as ligands to phosphorus, and arseranes (102-105), which have five oxygens as ligands to arsenic, display analogous behavior in the variable-temperature n.m.r. spectrum,
112
FAUSTO RAMIREZ A N D IVAR U O I
although the corresponding temperatures for the analogous phenomena vary from case to case. On the other hand, the arserane (106), withfour oxygens and one relatively electropositive phenyl as ligands, behaves differently from the other five compounds (101), and (102-105). I n the first group of compounds (101), and (102-105), there is a “low temperature range”, in which the n.m.r. spectra show two, rather than four, signals for CH3-C groups, (or for the corresponding H-C). There is also a “high temperature range” in which the spectra show only one signal for all the CH3-Cgroups (or for all the corresponding H-C). I n these cases one can retain the conditions : (a)that no five-membered ring occupies the diequatorial position, and : (b) that the hydrogen always remains equatorial, and yet explain the data at both temperatures, if one assumes that the positional exchange a t lower temperatures occurs by single TR processes, and at higher temperatures by double TR processes. I n these T R , the ring always provides the pair-ligands in the T R process. This is shown in Schemes 17 and 18. There is no equivalent BPR explanation of these data. I n the second class of compound (106), there is only one observable type of n.m.r. spectrum, which corresponds to the “lower temperature range” mentioned above, i.e., there are two CH3-C signals. One can conclude that the presence of a phenyl ligand in (106),in place of oxygen, in the case of arsenic, results in a relatively higher barrier for the (TR)’mechanism. Another type of spirotetraoxyphosphorane (107), with the electropositive hydrogen ligand,122but with different substituents on the two rings, is capable of giving rise to four CH,-C proton n.m.r. signals in the static situation. I n fact, the spectrum of (107), from - 70” to 0”exhibits SCHEME 17 Compounds (101-106)
PENTAVALENT PHOSPHORUS COMPOUNDS
113
SCHEME: 18
Compounds (101-105)
" r J R 3 0-
R4
jrRz
R3
only two CH& signals. The methylene groups give a spectrum of type AA'BB'X in the same temperature range. At temperatures above + 3 7 O , one finds only one CH3-C signal for (107), and the methylene groups now give only one signal corresponding to an A4X-system. These
114
FAUSTO RAMIREZ AND I V A R UUI
observations are quite reasonable if the positional exchange of (107) in the lower temperature range occurs by the single TR process, in which both rings can function as pair. The isomerization in the higher temperature range takes place by the (TR)2-process,in which the two rings, alternatively, function as the pair. No ring ever becomes diequatorial. The spiropentaoxyphosphoranelZ1(108), with a planar unsaturated ring, and an envelope-shaped saturated ring, behaves quite differently. Now, the size and the strain of the 0-P-0 ring angles are affected differently. The actual spectrum shows, in the entire range from - 60" to + 1 lo", a single CH,-C signal but a complex system of signals for the methylene protons. The data are quite understandable, if no fivemembered ring can occupy a diequatorial position, and if only the angle is capable of unsaturated ring, which has the larger 0-P-0 providing the pair in the TR process. I n that case, the TR, (TR)' and (TR)8processes become equivalent as far as the outcome of the process is concerned, and are equally appropriate for the interpretation of the data. Scheme 3 in Section VI provides data on relative stability of phosphorane isomers with a five-membered unsaturated ring.
The spiropentaoxyphosphorane (109) in Scheme 19, with a fivemembered unsaturated and a six-membered saturated ring,lZ1 shows two signals for the CH,-C protons and a complex spectrum for the methylene protons. The single-TR mechanism explains the data, if one allows both apical-equatorial and diequatorial placement of the sixmembered ring, as can be seen in Fig. 18-V and Scheme 19. The (TR)3 (- .-.-) must be invoked to explain the data, if one limits the placement of the six-membered ring to the diequatorial position. There is considerable information on the structure and the n.m.r. spectra of spirophosphoranes which contain carbon ligands to phosphorus, e x c l ~ s i v e l y . ~These - ~ compounds fall outside the scope of the present review,
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
115
scmm 19 Compound (109)
CeHa
CeHaO H H
IX. IRREGULAR PERMUTATIONAL ISOMERIZATIONS OF PENTAVALENT PHOSPHORUS COMPOUNDS A. Irregular Processes with Decrease in Coordination Number In principle,the permutational isomerization of TBP phosphorus can take place via an intermediate in which the phosphorus is tetracoordinated, and which is produced either thermally or under acid catalysis. + (PX5)' s [PX4]X
s
+ + + (PXs)'+H Z? [PX4XH]' Z [PX4]+XH
(PX5)*
+ [PX4XH]"
+ (PXs)"+H
Other types of related irregular processes may be possible e.g., when there is a P-H ligand.122 1. Thermal processes
The occurrence of uncatalyzed irregular permutational isomerizations is inferred from observations which show that the difference in the stability between a molecuIe with pentacoordinated phosphorus and its
116
F A U S T O RAMIREZ A N D IVAR U G I
tetracoordinated analog can be small enough to permit the existence of both forms in solution a t relatively low temperatures.lZe The product of the reaction of b e n d , CeH6.CO.CO.CeH6 with trisdimethylaminophosphine, * [(CH,),N],P is obtained in two crystalline forms, one is a triaminodioxaphosphorane ( l l O ) , and the other is a dipolar ion (111).
Both forms (110) and ( l l l ) ,give rise to the same equilibrium in solution. Some solvents favor the pentacoordinated form, while others favor the tetracoordinated form. As a result, there is a significant solvent dependence of the 31Pn.m.r. shift, since the observed shift corresponds to the average value of the signals of the two types of phosphorus nuclei, 6 =m.+ 40 p.p.m. (ws. H,P04) for coordination number five, and ca. -40 p.p.m. for coordination number four, in this particular type of compound. The establishment of the equilibrium is fast in the time scale of n.m.r. When a cyclic triaminophosphine is used in the above reaction, the resulting phosphorane, (112), increases in stability ws. the dipolar ion.127 This reflects the interplay of steric and electronic factors in determining the relative stability of trigonal bipyramidal ws. tetrahedral phosphorus. A decrease in steric interference among groups results from the introduction of the relatively flat ring and favors the oxyphosphorane structure. An increase in the electronegativity of the ligands favors also the oxyphosphorane structure; thus the oxygen analogs are more stable.lo
PENTAVALENT PHOSPHORUS COMPOUNDS
117
2. Acid catalyzed processes The existence of acid-catalyzed irregular permutational isomerizations with decrease in coordination number has come to light recently.6e The two oxaphosphetanes (60-11 4)(2 5)) and 60-11 4 2 5)), described in Section VIII, and reproduced in Scheme 20, differ in the relative disposition of the ethyl group attached to phosphorus and the methyl group attached to the four-membered ring-carbon. The interconversion of these two diastereomers (1 4)(2 5) and ( 1 4 2 5 ) cannot occur by the single TR-mechanism, or by the BPR mechanism for reasons given in Section V; cf. Figs. 7, 9 and 16. I n spite of this restriction, it was observed that the stereomutation (1 4) (2 5 ) 7t (1 4 2 5 ) did indeed occur when the compounds were kept for some time in aprotic solvents, or when they were heated to certain temperatures below those required to effect a Wittig reaction, i.e., the formation of an olefin (CF,)&= CH (CH,) and a phosphinate ester, (C,H,) (CzH6)P( 0)[OCH(CF,)2]. It SCHEME 20 Compound (60) CaH5.
p"
QH3
+ X
x CT~-C~H~/CHJ Enantiomer of (60-(14) (2 6))
RO
RO
118
F A U S T O RAMIREZ A N D I V A R U U I
was at first suggested that this "forbidden" stereomutation occurred by a bond rupture-recombination mechanism which resulted, after certain proton-transfers, in the inversion of the chiral center at carbon rather than at phosphorus.10.88 Subsequent investigations revealed that the stereomutation anti-(60)T? syn-(60) does not take place at all, unless some hexafluoroisopropylalcohol, (CF&CHOH is present. This alcohol is a relatively strong acid, and is a demonstrably effective catalyst for the stereomutation. The addition of traces of water to the oxyphosphorane (60)causes a very rapid hydrolysis, one of whose products is the fluoroalcohol. Therefore, the stereochemical stability of either one of the oxyphosphoranes in Scheme 20 can be demonstrated only on material of high degree of purity. The mechanism of Scheme 20 accounts adequately for the experimental observations in terms of an acid-catalyzed irregular process.66 B. Irregular Processes with Increase in Coordination Number Ionic products can be formed from the disproportionation of certain 131, 132 carbophosphoranes and fluorooxyphosphorane~.2-4~ 2(PXS)'
* [(PXe)($Xa)l * (PXs)"+PXs)"
Evidently, the interconversion among isomers of a TBP phosphorus compound can occur by this kind of irregular process. A variation of this involves the addition of a suitable anion to the phosphorane. (PX#+Z Tt [PXSl z2 (PXs)"+Z The occurrence of base-catalyzed substitutions of one alkoxy group by another in stable oxyphosphorctnes was discussed in Section VIII. lZ9 The rate of these substitutions is enormously accelerated by pyridine or by 2,4,&trimethylpyridine, and depends on the base concentration. A reasonable interpretation is that the coordination number of the phosphorus increases in the transition state or in the intermediate of this reaction (113) or (114). CeHsCHaOH 0CH.g 8-
+B
W
CeHsCHzo + (BH)+ OCH3 CH30,L
CH* (113)
CH3 (114)
PENTAVALENT PHOSPHORUS COMPOUNDS
119
There is a close relationship between this type of behavior in an isolable TBP phosphorane, and the addition of water to TBP cobalt to form octahedral cobalt which has been postulated in explanations of the base catalyzed hydrolysis of halogenoaminecobalt(1V).133 These data and others suggest that permutational isomerizations via intermediates with hexacoordinated phosphorus is a distinct possibility in certain stable phosphoranes. Moreover, this mechanism must be considered when fleeting phosphorane intermediates are postdated in the hydrolysis of phosphate esters, in particular at high pH. Irregular permutational isomerizations of certain fluorophosphoranes, (55) and (115), were recently reported.132 F
F (55)
The proton n.m.r. spectra of (55) and (115) disclosed the occurrence of a temperature- and concentration-dependent intermolecular process with a free energy of activation dQ* 16 kcal mole-l at 300°K.1328I n the difluorophosphorane (115), the coupling between the protons and the phosphorus is maintained, while the coupling between the protons and the fluorines is lost, a t higher temperatures and concentrations. This suggests that P-I? bonds are broken in the exchange, and was explained132ain terms of the transition state (116), i.e., a fluorine-bridged dimer. N
(116)
The disfavored an impurity-catalyzed exchange process, in particular by F- ion; however, they did not discuss the possibility of contamination by water or of acid catalysis. These results132in phosphoranes of the type (CH&PF2 and (CH3),PF3 contrast with those on (CH3)2NPF,80and PF6,1Q-21* 6 3 in which only regular permutational 5
120
FAUSTO R A M I R E Z A N D I V A R U Q I
isomerizations have, so far, been reported. 13’ Electronic and steric factors that tend to decrease the stability of the pentacoordinated state of the phosphorus, and which tend to increase the activation energy for the regular mechanisms (BPR and single or multiple TR), should favor the irregular over the regular process. An additional complicating factor pertains to solvation effects on compounds in which phosphorus is in different valence states, a subject which has, so far, received very little attention.
X. CONCLUSIONS X-Ray crystallographic analysis provides details of the static molecular structure of cyclic oxyphosphoranes. Group theory and topology provide a formal analysis of the way in which trigonal bipyramidal pentacoordinated phosphorus can undergo permutational isomerizations. Theoretical calculations of the energies associated with model situations in the positional exchange of ligands in phosphorus pentafluoride support two mechanisms for this exchange: Berry pseudorotation (BPR) and turnstile rotation (TR). M7hile in some acyclic compounds like PF6, CH3PF4and (CHS),NPF4both mechanisms may be equally possible, in other acyclic systems like PCl,F, or PH2F3, the TR mechanism may very well represent the preferred mechanism for the positional exchange of ligands. Cyclic pentacoordinated phosphorus compounds with trigonal bipyramidal D3,,symmetry cannot undergo positional exchange of ligands unless a TR mechanism is involved. The TR process may involve singre or multiple TR, or TR with “switches”. The available data on the cyclic compounds, including the variable temperature proton n.m.r. spectra, can be consistently interpreted by the TR-mechanism. I n this interpretation, a five-membered ring need never be placed in the diequatorial position of the trigonal bipyramid, and an element such as carbon or hydrogen, which is relatively low in electronegativity, need never be placed in an apical position. The TR mechanism could also explain the positional exchange of ligands in other pentacoordinated, trigonal bipyramidal molecules where the central atom is an element other than p h o s p h o r u ~ . 13e ’~~~ Regular processes are not the only way by which trigonal bipyramidal phosphorus can undergo permutational isomerization. Irregular processes, involving bond ruptures and reformations, must always be considered as alternatives, and some experimental justification or reasoning-by-analogy must be provided in discussions of permutational isomerization itineraries. This is particularly important when selection
PENTAVALENT PHOSPHORUS COMPOUNDS
121
rules are imposed upon reaction mechanisms which can be regarded, by hypothesis or by experimental demonstration, as proceeding via phosphorane intermediates. Irregular permutational isomerizations which involve a decrease in the pentacoordination of the phosphorus are known, and they are prone to acid catalysis. Irregular permutational isomerizations which involve an increase in the pentacoordination of the phosphorus, and which are susceptible to base catalysis, are quite reasonable from the known substitution reactions at pentacoordinated phosphorus.
ACKNOWLEDGMENTS The approach to phosphorus stereochemistry developed in this review rests on the experimental and the theoretical efforts of our collaborators, notably, S. Pfohl, E. A. Tsolis, J. F. Pilot, C. P. Smith and 0. P. Madan (F.R.) and P. Gillespie, P. Hoffmann, H. Klusacek and D. Marquarding (I.U.). One of us (F.R.) acknowledges the extended support of the Research whose outcome is presented here by the Cancer Institute of the National Institutes of Health, U.S.A.,in the form of Grant CA-04769, and the National Science Foundation, U.S.A., in the form of Grant GP-6690. REBERENCES 1. G.Wittig and M. Rieber, Ann. 562, 187 (1949). 2. G.Wittig, A. Maarcker and D. Hellwinkel, Angew. Chem. 76,756 (1964). 3. G. Wittig, Bull. SOC. Chim. Fr. 1162 (1966). 4. (a)D. Hellwinkel, Chem. Ber. 99,3628,3642,3660 (1966);(b)D. Hellwinkel, Angew. Chem. 78, 749 (1966);(c) D. Hellwinkel, Chimicl 22, 488 (1968); (d)D.Hellwinkel and H. J. Wilfinger, Tetrahedron Letters 3423 (1969). 5. (a) F. Ramirez and S. Dershowitz, J. Amer. Chem. SOC.78, 6614 (1966); (b)F.Ramirez and S. Dershowitz,J. Org. Chem.22,856 (1957); (c)F.Ramirez and S. Dershowitz, ibid. 22, 1282 (1957). 82, 2662 (1960);(b) F. 6. (a)F.Ramirez and N. B.Desai, J. Amer. Chern. SOC. Ramirez and N. Ramanathan, J. Org. Chem. 26,3041 (1961);(c) F.Ramirez, N. Ramanathan and N. B. Demi, J. Amer. Chem. SOC.84,1317(1962). 7. F. Ramirez, Pure Appl. Chem. 9,337 (1964). 8. F.Ramirez, Bull. SOC. Chim. Fr. 2443 (1966). 9. F. Ramirez, Accounts Chem. Rea. 1,168 (1968). Chim. Fr. (in Englieh) 3491 (1970). 10. F.Ramirez, Bull. SOC. 11. (a)F.Ramirez, S. L. Glmer, A. J. Bigler and J. F. Pilot, J. Amer. Chem.SOC. 91,496,6696,(1969);(b)F. Ramirez and C. D. Telefus, J. Org. Chem. 34,376 (1969);(c)F.Ramirez,S.B.Bhatia,C.D.TelefusandC.P. Smith, Tetrahedron 25,771 (1969);(d)F. Ramirez and G. V. Loewengart, J. Amer. Chern.Soo. 91, 2293 (1969); (e)F.Ramirez, J. Bauer and C. D. Telefus,J. Amer. Chern.SOC. 92,6935 (1970). 12. G.H.Birum and J. L. Dever, Abstracts, Div. of Org. Chem., A. C. 8. Meeting, Chicago, Ill., Sept., 1958,p. 101-P.
122
FATJSTO RAMIREZ A N D I V A R U G I
13. V. A. Kukhtin, DoklQdy Akad. Nauk,S.S.S.R. 121,466 (1958). 14. (a)D. B. Denney and H. M. Relles, J . Amer. Chem. SOC.86, 3897 (1964); (b)D. B. Denney and S. T. Cough, ibid. 87, 138 (1965);(c)D. B. Denney, D. Z. Denney, B. C. Chang and K. L. Mmi, ibid. 91,5243 (1970). 15. M. Roualt, Ann. P h p . (New York) 14,78 (1940). 16. G. S.Harris and D. S. Payne, J . Chem. Soo. 4617 (1956). 17. D. S.Payne, in “Topics in Phosphorus Chemistry”, Vol. 4 (M. Grayson and E. J. G f i t h , Ed.). Interscience Publishers, New York, N.Y., 1967. 18. (a) E. L. Muetterties, W. Mahler and R. Schmutzler, Inorg. Chem. 2, 613 (1963); (b)E. L. Muetterties, W. Mahler, K. J. Packer and R. Schmutzler, ibid. 3, 1298 (1964). 19. R. Schmutzler, Angew. Chem. 76,893 (1964);Angew. Chem. Internut. Edit. 3, 753 (1964). 20. R.Schmutzler, Angew. Chem. 77,530 (1965);Angew. Chem. I n t e m t . Edit. 4, 496 (1965). 21. (a)R. Schmutzler,Adwun. Fluorine Chem. 5,31 (1965);(b)R. Schmutzler, in “Halogen Chemistry”, Vol. 2 (V. Gutmann, Ed.). Academic Press, New York, N.Y., 1967. 22. E. L.Muetterties and R. A. Schunn, Quart. Rev. (London)20,245 (1966). 23. E. L.Muetterties, Accounts Chem. Rea. 3, 266 (1970). 24. (a)J. E. Griftiths, R. P. Carter Jr. andR. R. Holmes, J . Chem. Phys. 41, 863 (1964); (b)R. R. Holmes, R. P. Carter Jr. and G. E. Peterson, Inorg.Chem. 3, 1748 (1964). 25. (a)R. R. Holmes and R. M. Deiters, J . Amer. Chem. SOC.90, 5021 (1968); (b) R. R. Holmes and R. M. Deiters, Inorg. Chem. 7, 2229 (1968);(c) R. R. Holmes, R. M. Deiters and J. A. Golen, ibid. 8,2612 (1969). 26. (a)K. W.Hansen and L. S. Bartell, Inorg. Chem. 4, 1775 (1965);(b)L. S. Bartell and K. W. Hansen, ibid. 4,1777 (1965). 27. D. H. Brown, G. W. Fraser and D. W. A. Sharp, J . Chem. SOC.171 (1966). 28. F.Ramirez, C. P. Smith and S. Meyerson, Tetrahedron Letters 3651 (1966). 29. (a)K.F.Kumli, W. E. McEwen and C. A. Vander Werf, J . Amer. Chem. Soc. 81,248,3805,(1959); (b)M. Zanger, C. A. Vander Werf and W. E. McEwen, ibid. 81, 3806 (1969);(c) A. Blade-Font, C. A. Vander Werf and W. E. McEwen, ibid. 82,2396 (1960). 30. L. Horner, H.Winkler, A. Rapp, A. Mentrup and D. Beck, Tetrahedron Letter8 161 (1961). 31. R. F. Hudson and M. Green, Angew. Chem. 75,47 (1963). 32. S.Trippett, Quart. Rev. (London) 17,406 (1963). 33. W.E. McEwen in “Topics in Phosphorus Chemistry”, Vol. 2 (M.Grayson and E. J. Griftith, Ed.). Interscience Publishers, New York, N.Y.,1965. 34. M. J. Gallagher and I. D. Jenkins, Top. i n Stereochem. 3, 1 (1968),N. L. Allinger and E. Eliel, Ed., John W h y and Sons Inc., N.Y. 1968. 35. P. C. Haake and F. H. Westheimer, J . Amer. Chem. SOC.83, 1102 (1961). 36. (a)F. Ramirez, B. Hansen and N. B. Desai, J . Amer. Chem. SOC.84, 4588 (1962);(b)F. Ramirez, 0.P. Madan, N. B. Desai, S. Meyerson and E. M. Banas, ibid. 85,2681 (1963). 37. (a)F. Ramirez, N. B. Desai and N. Ramanathan, J . Amer. Chem. Soo. 85, 1874 (1963);(b)F.Ramirez, A. V. Patwardhan, N. B. Desai, N. Ramanathan and C. V. Greco, ibid. 85,3056 (1963); ( 0 ) F.Ramirez, N. Ramanathan, and N. B. Desai, ibid. 85, 3465 (1963).
P E N T A V A L E N T P H O S P H O R U S COMPOUNDS
123
38. (a) F. Ramirez, A. V. Patwardhen, N. B. Desai and S. R. Heller, J. Amer. Chem.SOC. 87,549 (1965);(b)F.Ramirez, A. V. Patwardhen, N. B. Desai and S. R. Heller, aid.87,549 (1965);(c)F. Ramirez, 0. P. Madan and C. P. Smith, ibid. 87, 670 (1965). 39. E.A.DennisandF. H. Westheimer,J.Amer. Chem.Soc.88,3431,3432(1966). 40. G. Aksnes and K. Bergesen, Acta Chem. Scund. 20, 2508 (1966). 41. D. Swank, C. N. Caughlan, F. Ramirez, 0. P. Madan and C. P. Smith,J.Amer. Chem. SOC. 89,6503 (1967). 42. R. Kluger, F. Kerst, D. Lee and F. H. Westheimer, J. Amer. Chem. SOC.89 3918 (1967). 43. D.5.Frank and D. Usher, J. Amer. Chem. SOC.89,6360 (1967). 44. F.Ramirez, H.J. Kugler, A. V. Patwardhan and C. P. Smith, J. Org. Chem. 33, 1186 (1968). 45. R. Kluger, F. Covitz, E. Dennis, L. D. Williams and F. H. Westheimer, J. Amer. Chem. SOC. 91,6066 (1969). 46. F. H.Westheimer, Accounfa C k m . Res. 1, 70 (1968). 47. (a)J. Kumamoto, J. Cox, Jr. and F. H. Westheimer, J. Amer. Chem.SOC.78, 4858 (1956);(b)J. R. Cox, Jr., R. E. Wall andF. H. Westheimer, Chem. and Ind. (London) 929 (1959);( c )E.T. Kaiser, M. Panar and F. Westheimer, J. Amer. Chem. SOC. 85, 602 (1963);(d) F.Covitz and F. H. Westheimer, ibid 85, 1773 (1963). 91,7012(1969); 48. (a)K. Naumann, G. Zon and K. Misolw, J.Amer. Chem.SOC. (b)G. Zon, K. E. DeBruin, K. Naumann and K. Mislow, ibid. 91,7023 (1969); (c)K. E. DeBruin, G. Zon, K. Naumann and K. Mislow, ibid. 91,7027(1969). 49. K.Mislow, Amounts Chem. Res. 3, 321 (1970). 50. (a)W. Hawes and S. Trippett, J. Chem. SOC. C,1465 (1969);(b)J. R. Corfield, J. R. Schutt and S. Trippett, ibad. 789 (1969). 51. S. E.Cremer, R. J. Chorvat and B. C. Trivedi, Chem. C m m . 769 (1969). 91,4724 (1969). 52. K.L.Marsi, J. Amer. Chem. SOC. 53. R. S. Berry, J. Chern. Phys. 32,933 (1960). 54. J. E.Kilpatrick, K. S. Pitzer and R. Spitzer, J. Amer. Chem. SOC.69,2483 (1947). 55. I.Ugi, F. Ramirez, D. Marquarding, H. Klusacek, G. Gokel and P. Gillespie, Angew. Chem. 82,766 (1970);Angew. Chem. Internat. Edit. 9,725 (1970). 56. F. Ramirez, I. Ugi, S. Pfohl, E. A. Tsolis, J. F. Pilot, C. P. Smith, D. Marquarding, P. Gillespie and P. Hoffmann, P h o q ~ h o w1, 1 (1971). 57. I.Ugi, D. Marquarding, H. Klusacek, P. Gillespie and F. Ramirez, Accounts Chem. Rea. 4,288 (1971). 58. P. Gillespie, P. Hoffmann, H. Klusacek, D. Marquarding, S. Pfohl, F. Ramirez, E. A. Tsolis and I. Ugi, Angew. Chem. 83, September (1971); Angew. Chem. Internat. Edit. 10,October (1971). 59. F. Ramirez, I. Ugi, S. Pfohl, E. A. Tsolis, J. F. Pilot, C. P. Smith, D. Marquarding, P. Gillespieand P. Hoffmann, J. Am. Chem.Soo. submitted for publication. 60. (a)W. C. Hamilton, S. J. LaPlma and F. Ramirez, J. Am. Chem.Soo. 87,127 (1965);(b) W.C. Hamilton, S. J. LaPleoa, F. Ramirez and C. P. Smith, ibid. 89, 2268 (1967); (c)R. D. Spratley, W. C. Hamilton and J. Ladel, ibid. 89, 2272 (1967). 61. H.A. Bent,Chem. Rev. 61,276 (1961). 62. (a)R.E.Rundle, Rec. Chem. Progr. 23, 196 (1962);(b) R. E.Rundle, Acta Cqpt. 14,686 (1962);(c)R. E.Rundle,J. Am. Chem. SOC.85, 112 (1963).
124
FAUSTO RAMIREZ A N D IVAR U a I
63. R.J. Gillespie,J . Chem. Ed. 47, 18 (1970). 64. (a)D. P. Craig, A. Maccoll, R. S. Nyholm, L. E. Orgel and L. E. Sutton, J . Chem. SOC.332,(1954);(b) R.J. Gillespie,J . Chem. SOC.1002 (1964). 65. R. J. Gillespie and R. S. Nyholm, Quart. Rev. (London)11,339 (1957). 66. (a)F.Ramirez, C. P. Smith and J. F. Pilot, J . Amer. Chem. SOC.90, 6726 (1968); (b)F.Ramirez, C. P. Smith, J. F. Pilot and A. S . Gulati, J . Org. Chem. 33, 3787 (1968). 67. ( 8 )F.Ramirez, A. J. Bigler and C. P. Smith, J . Amer. Chem. SOC.90,3507 (1968); (b)F. Ramirez, A. J. Bigler, and C. P. Smith, Tetrahedron 24, 5041 (1968). 68. D.D.Swank, C. N. Caughlan, F. Ramirez and J. F. Pilot, J . Amer. Chem.,Soc. 93,in press (1971). 69. M.U.Haque, C. N. Caughlan, F. Ramirez, J. F. Pilot and C. P. Smith, J . Amer. Chem.Soc. 93,in press (1971). 70. P. J. Wheatley, J . Chem. SOC.2206 (1964). 71. (a)J. J. Daly, J . Chem. Soc. 3799 (1964);(b) J. J. Daly, aid. A, 1020 (1966). 72. D. E. C. Corbridge in “Topics in Phosphorus Chemistry”, Vol. 3 (M. Grayson and E. J. Griffith, Ed.).Interscience Publishers, New York, N.Y., 1966. 73. (a)G. Polya, Acta Math. 68, 145 (1937);(b) S. W. Golomb, Information Theory, 4th London Symposium,Universities Press, Belfast, 1961; (c) B. A. Kennedy, D. A. McQuarrie and C. H. Brubaker, Jr., Morg. Chem. 3, 265 (1964). 74. E. Ruch, W.Hasselbarth and B. Richter, Theor. C h h . Acta 19,288 (1970). 75. R. S.Cahn, C. K. Ingold and V. Prelog, Angew, Chem. 78,413(1966);Angew. Chem. Internut. Edit. 5, 385 (1966). 76. P.C. Lauterbur and F. Ramirez, J . Amer. Chem. SOC.90,6722 (1968). 77. D.Gorenstein and F. H. Westheimer, J . Amer. Chem. SOC.92,634 (1970). 78. (a)E.L. Muetterties, J . Amer. Chem. SOC.91,1636,4115 (1969). 79. (a)R. R. Holmes and J. A. Golen, Inorg. Chem. 9, 1596 (1970);(b) L. S. Bartell, ibid. 9,1594 (1970). 80. G. M. Whitesides and H. G. Mitchell, J . Amer. Chem. SOC.91, 5384 (1969). 81. P. Gillespie and I. Ugi, Unpublished Work. 82. (a)J. A. Pople, D. P. Santry and G. A. Segal, J . Chem. Phys. 43,5129(1965); (b)J. A. Pople and G. A. Segal, ibid. 43,5136(1966); (c)J.A. Pople and G. A. Segal, ibid. 44,3289 (1966);(d) D. P. Santry and G. A. Segal, ibicl. 47, 158 (1967);( e ) J. A.Pople, Accounts Chem. Rm. 3, 217 (1970). 83. J.A. Pople and D. L. Beveridge, “Approximate Molecular Orbital Theory”. McGraw-Hill, New York, N.Y., 1970. 84. H.H. J&6, Accounts Chem. Rm. 2,136 (1969). 85. S.Ehrenson and S. Seltzer, Theoret. Chim. Actu 20, 17 (1971). 86. J. D.Dunitz and V. Prelog, Angew. Chem. 80, 700 (1968). 87. (a)M. Gielen, M. DeClercq and J. Nasielski, J . Orgunomet. Chem. 18, 217 (1969);(b) M. Gielen and N. VanLautem, Bull. SOC.C h h . Belgea 79, 679 (1970). 88. (a)A. T. Balaban, D. Farcasio and R. Banica, Rev. Roum. Chim. 11, 1206 (1966);(b) A. T. Balaban, ibid. 15, 463 (1970);(c) A. T.Balaban, W .15, 1267 (1970);(d) A. T. Balaban, D. Farcasio and F. Harary, J . Labelled Compounds 6 , 211 (1970). 89. E.Clementi and D. L. Reimondi, J . Chem. Phys. 38,2686 (1963). 90. 4. J, Dowps 4nd J3.Scbutzler, Spectrochem. Acta. 23A, 681 (1967).
PENTAVALENT PHOSPHORUS COMPOUNDS
125
91. J. I. Musher, Angew. Chem. 81,68 (1969); Angew. Chem. Internut. Edit. 8,54 (1969). 92. P. C. Van der Voorn and R. S. Drago, J . Amer. Chem. SOC. 88, 3255 (1966). 93. D. B. Boyd and W. N. Lipscomb, J . Chem. Phys. 46, 910 (1967). 94. L. Pauling, “The Nature of the Chemical Bond”, Third Edition. Cornell University Press, 1960, p. 90. 95. S. B. Pierce, P. M. Troichel and R. A. Goodrich, J . Amer. Chem. SOC. 89, 201 7 (1967). 96. A. D. Walsh, D i s c z ~ sFaraday . SOC. 2, 18 (1947). 97. D. P. Craig and E. A. Magnusson, J . Chem. SOC.4895 (1956). 98. J. K. Wilmshurst and A. J. Bernstein, J . Chem. Phys. 27, 661 (1957). 99. D. W. J. Cruickshank, J . Chem. SOC.5486 (1961). 100. M. J. Taylor and L. A. Woodward, J . Chem. SOC. 4670 (1963). 101. R. F. Hudson, “Structure and Mechanism in Organo-Phosphorus Chemistry”. Academic Press, New York, N.Y., 1965, Ch. 3, Ch. 7. 102. J. R. VanWazer, “Phosphorus and its Compounds Interscience Publishers, Inc., New York, N.Y., 1958. 103. M.M.Crutchfield, C. H. Dungan, J. H. Letcher, V. Mark and J. R. VanWazer, in “Topics in Phosphorus Chemistry”, Vol. 5 (M. Grayson and E. J. Griffith, Ed.), Interscience Publishers, New York, N.Y., 1967. 104. H. Stetter and K. Steinacker, Chem. Ber. 85, 461 (1962). 106. K. Dimroth and A. Nurrenboch, Chem. Ber. 93,1649 (1960). 106. (a)J. 0. Verkade and R. W. King, Inorg. Chem. 1, 948 (1962); (b) J. G. Verkade, R. W. King and C. W. Heitsch, ibid. 3,884 (1964) ; (c)J. G. Verkade, J. J. Hutternan, M. K. Fung and R. W. King, ibid. 4, 83 (1966). 107. F. Ramirez, C. P. Smith, A. S. Gulati and A. V. Patwardhan, Tetrahedron Letters 2151 (1966). 108. (a)N. P. Gambaryan, Yu. A. Cheburkov and I. L. Knunyants, Bull. A d . Sci. USSR 8 , 1433 (1964); (b) E. M. Rokhlin, Yu. V. Zeifman, Yu. A.
”.
Cheburkov, N. P. Gambaryan and I. L. Knunyants, Dokl. Akud. Nauk. SSSR, 161 (0), 1365 (1965); Proceedings, p. 393. 109. (a)K. J. Coskran and J. G. Verkade, Inorg. Chem. 4, 1656 (1965); (b) J. W. Rathke, J. W. Guyer and J. G. Verkade, J . Org. Chem. 35, 2310 (1970). 110. D. Bernard and R. Burgada, Compt. Rend. 271,418 (1970). 111. (a)F. Ramirez, A. V. Patwardhan and S. R . Heller, J . Amer. Chem. SOC.86, 514 (1964); (b) F. Ramirez, J. F. Pilot, C. P. Smith, S. B. Bhatia and A. S. Gulati, J . Org. Chem. 34, 3386 (1969). 112. (a)F. Ramirez and N. B. Desai, J . Amer. Chem. SOC.85, 3252 (1963); (b) F. Ramirez, S. B, Bhatia and C. P. Smith, Tetrahedron 23, 2067 (1967). 113. F. Ramirez and E. A. Tsolis, unpublished work. 114. F. Ramirez, 0. P. Madan and S. R. Heller, J . Amer. Chem.SOC. 87, 731 (1965). 90, 115. F. Ramirez, J. F. Pilot, 0. P. Madan and C. P. Smith, J . Amer. Chem.SOC. 1275 (1968). 116. F. Ramirez, J. F. Pilot and C. P. Smith, Tetrahedron 24, 3735 (1968). 117. F. Ramirez, Trans. N . Y . Acad. Sci. 24, 3735 (1968). 118. (a)D. G. Gorenstein and F. H. Westheimer, Proc. Nut. Acad. Sci., U.S. 58, 1747 (1967) ; (b)D. G. Gorenstein and F. H. Westheimer, J . Amer. Chem.SOC. 89, 2762 (1967). 119. F. Ramirez and C. P. Smith, Chem. Comm. 662 (1967). 120. (a)I. J. Borowitz and N. Anschel, Tetrahedron Letters 1517 (1967); (b) I. J. Borowitz, M. Anschel and S. Firstenberg, J . Org. Chem. 32, 1723 (1967).
126
B A U S T O R A M I R E Z A N D I V A R UGI
121. F.Ramirez,M.NagabhusbnamandC.P.Smith, Tetrahedron24,1786(1968). 122. (a)M. Sanchez, R. Wolf, R. Burgada and F. Mathis, Bull. SOC. Chim. Fr. 733 (1968); (b)D.Houalla, R. Wolf, D. Gagnaire and J. B. Robert, Chem. Comm. 443 (1969);( c )D.Bernard and R. Burgada, Compt. Rend. 271(c), 418 (1970). 123. (a)B. Englund, J.prakt. Chem. 120,179 (1928); (b)E. J. Selmi, K. Merivuari and E. Laaksonen, Suomen. Kern. 19B, 102 (1946). 124. H. Goldwhite, Chem. Comm. 651 (1970). 126. G. 0.Doak and R. Schmutzler, Chem. Comm. 476 (1970). 126. (a)F.Ramirez, A. V. Patwardhan, H. J. Kugler and C. P. Smith, Tetrahedron Letter8 3053 (1966);(b) F. Ramirez, A. V. Patwardhan, H. J. Kugler and C. P. Smith, Tetrahedron 24,2276 (1968). 127. (a) F.Ramirez, A. V. Patwardhan and C. P. Smith, J. Amer. Chem. SOC.87, 4973 (1966); (b)F. Ramirez,A. V. Patwardhan, H. J. Kugler and C. P. Smith, ibid. 89, 6276 (1967);(c)F.Ramirez, A. S. Gulati and C. P. Smith, ibid. 89, 6283 (1967); (d)F.Ramirez, A. S. Gulati and C. P. Smith, J. Org. Chem. 33, 13 (1968). 128. R. Burgeda, Bull. SOC.Chim. Fr. 347 (1967). 129. F.Ramirez, K. Tasaka, N. B. Desai and C. P. Smith, J. Amer. Chem.SOC.90, 761 (l968). 130. H. R. Allcock, J. Amer. Chem. SOC.86,2691 (1964). 131. (a) R. Schmutzler, J. Amer. Chem. Soc.86,4600 (1964);(b)G. S. Reddy and R. Schmutzler, Inorg. Chem. 5,164(1966);(c)S.C. Peake and R. Schmutzler, Chem. and Indwtry 1482 (1968). 132. (a)T. A. Furtsch, D. S. Dierdorf and A. H. Cowley, J. Amer. Chem. SOC.92, 5769 (1970);(b)H. Dreeekamp and K. Hildebrand, 2.Naturforsch 26B, 269 (1971). 133. R.C.Pearson and F. Basolo, Xmorg. Chem. 4, 1524 (1966). 134. F.A.Cotton, A. Danti, J. S.Waugh and R. W. Fessenden,J.Chem. Phya. 29, 1427 (1968). 136. (a) M. Gielen, M. DeClercq, G. Mayence, J. Nasillski, J. Topart and H. Vanwuytswinkel, Rec. Trav. Chim. 88, 1337 (1969); (b) G. J. D. Peddle and C. Redl,J.Amer. Chem.SOC. 92,366 (1970);(c)C. A. Udovich and R. J. Clark, &bid. 91,626 (1969);(d) J. D. W m e n and R. J. Clark, Inorg. Chem. 9, 373 (1970). 136. L. H.Sommer, " Stereochemistry, Mechanism and Silicon". McGraw-Hill, New York, N.Y., 1966. 137. J. D.Macomber, J. Magn. Resonance 1, 677 (1969).
THE HYDROGEN ATOM ABSTRACTION REACTION FROM 0-H BONDS F. TUD& Central Research Institute for Chemistry of the Hungarian A d e m y of Sciences, Budapest M. SIMONYI
AND
. . . . . . . . . . . . . . . . . , . . . . . . . . . . .
I. Introduction 127 128 A. Characterization of the Hydrogen Atom Abstraction Reaction 130 B. Limitations of the Scope C. Possible Experimental Techniques for Kinetic Studies 131 11. Kinetic Studies on 0-H Bond Fission 135 A. Studies on Phenols: The Kinetic Isotope Effect . 136 B. Studies on Phenols: The Substituent Effect 144 150 C. Studies on Phenols: The Steric Effect D. Studies on Other Hydroxylic Compounds 154 157 111. The Role of Hydrogen Bonding . . A. Influence of the Medium on the Rate 158 B. The Hydrogen-BondingEquilibrium 160 C. Consideration of Hydrogen Bonding in the Interpretation of Certain 164 Kinetic Anomalies 167 IV. ArrheniusParameters . A. Interrelation Between the Arrhenius Parameters 167 B. Formal Interpretation of the Compensation Phenomenon 172 173 V. Conclueion References 174
.
. .
.
. .
.
. . .
.
.
. .
. .
. . . .
.
.
.
.
.
.
.
.
.
.
. .
. .
.
. . . . .
. . . . .
.
.
.
.
.
.
.
.
.
I. INTRODUCTION RADICALS are chemical species having a free valence in classic chemical terms, or having an electron of uncoupled spin in terms of the quantum theory. The lifetime of radicals in homogeneous systems depends on both the nature of the radical and the other species available. Accordingly, radical lifetimes may vary from picoseconds (Porter, 1970) up to months or years. Stable radicals are unable to undergo self-consumption in the reaction 2R'
--f
non-radical product(s)l
(1)
which is characteristic of, for example, reactive alkyl radicals and responsible for their low concentration in homogeneous systems. A variety of radicals can be used for kinetic studies, but only a few of 1 If radicals terminate in recombination, there is only one product. If the termination proceeds Via disproportionstion, two product species &reformed.
127
128
M. SIMONY1 A N D F . T U D 6 S
these are stable. Any process to be studied between a short-lived radical and a molecule must compete with a radical-consuming process. I n some cases, however, reaction (1) can be suppressed. If R reacts with a molecule, say Z, R ' + Z -+ R-2'
(2)
and the product radical is of much lower reactivity than R', the former soon reaches such a high concentration that reaction (3) R' +R-Z'
+
non-radicalproduct(s)
(3)
becomes the only process in which radicals are consumed (TiidBs et al., 1966; James and Troughton, 1966). Short-lived radicals have to be continuously generated for a steady concentration to be maintained. The common methods for initiating radical processes were reviewed by Bagdasaryan (1966). Radical reactions could be classified in different ways. First of all, they could be selected according to the phase they are studied in. This review deals almost entirely with liquid-phase reactions. Certain important comparisons can actually be made between reactions proceeding in gas and liquid phases. I n discussing Arrhenius parameters, the limited liquid phase data are complemented by those taken from gas phase reactions. On the other hand, special attention is given to solvation effects, in general, and to hydrogen bond formation, in particular. A second possible principle to select radical reactions could be the kind of the radical. It is true that different radicals could show different kinetic behaviour (seethe inverse selectivity of Cl' atom and CH; radical towards the hydrogen atoms in propionic acid; Tedder, 1960). Our discussion deals with organic radicals without making any fundamental selection among them. As the title suggests, we chose a different classifying principle by defining the reaction centre instead of the radical. This limitation allows us to consider the particular problem of homolytic reactivity of the 0-H group in satisfactory depth to discuss some kinetic problems of general interest. The field where the limitation has been felt to cause undesirable restrictions is that of Arrhenius parameters. Since a general trend is suspected, that discussion concerns hydrogen abstraction from other bonds as well.
A. Characterization of the Hydrogen Atom Abstraction Reaction Regardless of initiating and terminating processes that lead to radical production and consumption, respectively, there are radical reactions
H Y D R O Q E N ATOM A B S T R A C T I O N R E A C T I O N
129
that conserve the uncoupled spin. They are either unimolecular (decomposition and isomerizetion) or bimolecular ones. The latter reactions fall into two classes according to their mechanisms. First, there are addition processes when the reacting molecule is unsaturated, as in equation (2). Second, there are homolytic cleavages of single bonds by radicals. I n these processes an atom transfer occurs from the molecule to the radical while forming a new radical from the reactant molecule and producing a product molecule from the reactant radical. The atom most commonly transferred is hydrogen : R'+H-Y
k
R-H+Y'
(4)
The fact that hydrogen atom abstraction is so common is empirically well known (Pryor, 1966; Pryor et al., 1969). There are, however, significant differences in the reactivities of differently bonded hydrogen atoms. It is perhaps the most widely accepted view that the reactivity in reaction (4) is mainly determined by the difference in stability of the reactant ( R ) and product (Y') radicals. As a general trend, it seems valid. If one varies the type of R' in the direction of decreasing stability, while keeping the same Y-H compound as reference, the rate constant of reaction (4) will increase (e.g., hydrogen atom abstraction from toluene by different radicals; Jenkins, 1969). On the other hand, keeping R' constant and altering the reactant molecule so as to produce a more stable product radical, the rate constant is again increased (e.g., Simonyi et al., 1967s). Since the fundamental factor in determining radical stability is the resonance effect resulting in delocalization of the uncoupled spin, it can be expected that resonance factors will be of major importance for the radical reactivity. Thermochemistry is not the only factor, however, that governs radical processes. There are symmetrical thermoneutral reactions of the type R'+H-R
+ R--H+R'
(6)
which proceed with significant rates (R =H, Geib and Harteck, 1931, Farkas and Farkas, 1935; R =D, Benson, 1960a, Le Roy et al., 1967; R =CH,, Dainton et al., 1957, 1959; R =CD,, Creak et al., 1962; R = 2,4,6-tri-t-Bu-phenoxy, Kreilick and Weissman, 1962). Polar effects in radical reactions seem also not quite well understood. There are some attempts to divide radicals into electrophilic and nucleophilic classes (Pryor, 1966;Pryor etal., 1969;Johnston etal., 1966) resembling the pattern of ionio reactants, generally on the basis of the Hammett equation. This classification, however, seems to be alien to the nature of most carbon radicals. I n addition reactions (2) the Hrtmmett p-value is usually positive (radical addition to substituted nitrobenzenes: Bartlett and Kwart, 1960, 1962; Sinitsynrt and
130
M . S I M O N Y I A N D F,
TUDBS
Bagdasaryan, 1958, 1960; TiidBs et al., 1962, 1963; Foldes-Berezhnykh et al., 1969 ;radical addition to substituted azobenzenes and Schiff bases: Foldes-Berezhnykh, unpublished ; radical addition to substituted styrenes : Walling et al., 1948 ;Zaitsev and Shtraichman, 1968), suggesting that radicals are, in varying measure, nucleophilic. Considering the hydrogen atom abstraction from C-H (Kooyman et al., 1953; van Helden and Kooyman, 1954; Bridger and Russell, 1963; Pryor et al., 1966, 1969), N-H (Yates and Ihrig, 1965; Simonyi and TudBs, 1969) or 0-H bonds (Tables 3 and 6 ) , one always finds negative p-values, indicating that the very different radicals involved are uniformly more or less electrophilic. A probable component of the driving force for a reaction of a reactive radical is the coupling of the initially uncoupled spin. Any factor that makes it energetically favourable may enhance the reaction rate. The centres of the positive and negative charges may be slightly separated in the course of the reaction. The actual direction of the charge separation is not solely dependent on the radical structure. The reactant molecule and the mechanism of the reaction must strongly influence the possibility of any charge separation in, or within the reactants. This is in some contrast to ionic reactivity. For example, the NOg ion attacks sites possessing electron excess much more easily. Similarly, the reaction of an alkoxide ion will favour attacks on sites with partial positive charges. It seems probable, however, that the majority of radicals we are concerned with cannot be forced to fall into nucleophilic or electrophilic classes. As mentioned, hydrogen atom abstraction reactions are, in general, accelerated by electron donor substituents of the substrate. I n building a pattern of radical reactivity, the hydrogen atom transfer from toluene was suggested to be free of polar influences and other processes were described by introducing polarity parameters characterizing the substrata. The factor considered for radical polarity operates only if substrate polarity is also effective (Bamford et al., 1959; Bamford and Jenkins, 1961, 1963; Jenkins, 1967). The success of this pattern in correlating radical reactivity does not depend on the arbitrary choice of the reference reaction (Jenkins, 1969), which is actually slightly polar, as shown for CH; radicals by Pryor et al. (1969, cf. Kalatzis and Williams, 1966).
B. Limitations of the Scope The homolytic fission of 0-H bonds is dealt with in this chapter from the kineticist’s point of view. Only second-order, bimolecular reactions
H Y D R O G E N ATOM A B S T R A U T I O N R E A C T I O N
131
are considered. Further limitations are given in connection with the nature of the reactants. 1. Limitation for the reactant radical
Some typical kinds of organic radicals, formed in thermal processes, are considered. Radiolysis studies and reactions involving hot atoms (e.g. energetic tritium) and resulting in simultaneous abstraction and substitution (Chou and Rowland, 1969; Baker et al., 1970) are omitted, as are hydrogen abstraction reactions by electronically excited species (Rosenberg and Servb, 1970; Wagner et aE., 1970) which involve the nature of the triplet state, a central problem of photochemistry (Wagner and Hammond, 1968). Photochemically initiated radical processes (Hirata et al., 1960), however, do not differ from those induced by thermal decomposition of initiators, and hence there is no restriction on the nature of initiation of the processes considered. Since we are mainly interested in an elementary radical reaction, experimental techniques that might complicate the elementary step of interest are excluded. For example, it is important to note that the use of peroxide initiators for kinetic studies on phenols seems unfortunate, since peroxides can also participate in ionic reactions (Behrman and Edwards, 1967), including some with phenols (Denney and Denney, 1960; Okada, 1961). 2. Limitation for the reactant molecule
The commonest compounds containing 0-H bonds are water, alcohols, phenols and carboxylic acids (if inorganic acids and bases are neglected). Hydrogen atom transfer from water to a radical seems to be unknown (Kondratiev, 1970; Bennett et al., 1970; for deuterium atom transfer from D 2 0 to NZ see Felder et al., 1970). Phenols are the most suitable models for our purpose and a main part of the discussion deals with their reactions. Alcohols and carboxylic acids, especially in liquid phase, react with radicals mainly through their C-H bonds. This fact has been utilized in synthetic work (Migita, 1969). Reactions of compounds other than phenols are dealt with in Section 1I.D.
C. Possible Experimental Techniquesfor Kinetic Xtudies Reactions proceeding with stable radicals can easily be followed by either spectrophotometry or e.s.r. spectroscopy. For short-lived radicals, the simplest system is the competition of an initiation (6) and a radical-consuming process (I), i.e., I 2R'
k __f
k8
2R' non-radicalproduct+)
(6)
132
M . S I M O N Y I A N D F. T U D ~ S
where I is the molecule of a thermal initiator decomposing into two radicals. (In the liquid phase there usually is a loss of radicals by their primary recombination. This phenomenon, called cage effect, is not our interest here. For recent reports see Herkes et al., 1969; Chakrovorty et al., 1969. For thermal decomposition of initiators see Tobolsky and Mesrobian, 1954; Gragerov, 1969). The steady concentration of shortlived radicals, T , in this most simple system is r 112
=(?)
(7)
where W,is the rate of producing radicals in reaction (6). If we now introduce only one additional process, e.g., a hydrogen abstraction step from molecule H-Y present in the system (4), two additional terminating steps should appear in the detailed kinetic scheme, i.e., R’+Y’
-+
and Y’+Y +
This complexity is involved in almost any kinetic study on radical processes. The consumption of radicals occur now, in effect, in the competing reactions ( l ) , (8) and (9). Measurable quantities are often the yields of R-H and R-R products of reactions (4) and (l),respectively, (e.g., Majer et al., 1969) and the reactivity of H-Y toward R is often expressed by the ratio, k/ki12. It is extremely important, however, that reactions (4) and (1) are of different orders with respect to R and simple neglect of reaction (8) (which is often the dominating termination process in retarded chain reactions, Semenov, 1958) implies that the concentration of R will not vary throughout the whole reaction (Giles and Whittle, 1966). If all possible steps are considered, the evaluation of k/k:I2 from the yields requires knowledge of two additional parameters, as shown by Bazilevskii and Trosman (1968), and the calculation becomes tedious. A good deal of uncertainty comes from our limited knowledge of radical termination processes. These reactions are, as a rule, very fast ; they need no activation energy in the gas phme and their rates are often close to the diffusion-controlledlimit in the liquid phase (Walling, 1957). Although methods of determining single rate constants (called “absolute rate constants”) for reactions of short-lived radicals have long been established (Chapman et al., 1920) and applied (Melville, 1937), there is still increasing interest in the determination of absolute rate constants for radical-terminating (Stepukhovich et al., 1968; De Mare and Huybrechts, 1908; Umanskii and Stepukhovich, 1969 ; Weiner and
H Y D R O G E N ATOM ABSTRACTION REACTION
133
Hammond, 1969) as well as for other radical processes (Howard et al., 1968; Ingold, 1968; Howard and Ingold, 1970). Radical systems become much more complicated if one of the products is not inert towards the radical employed. Such effects can be entirely obscured if only a narrow temperature range is covered by the study. A dramatic example for that was the reaction between CF, and NH,. Gray et al. (1969) found the process free of complexity from measurements covering a temperature range of "only" 127 degrees. Extending the temperature range to a width of 322") Morris and Thynne (1970) found that processes other than those considered should also operate. In some studies, a hydrogen abstractionreaction is applied as standard process in competition with other radical steps to be studied. The reactivity of the compounds investigated is usually expressed by the rate constant ratio of the competing reactions. The superiority of this method lies in the comparison of elementary steps of the same kinetic order with respect to the radical. Radical addition reactions were studiedin this manner (Heilman et al., 1957; Stefani et al., 1961)and also hydrogen abstractions by using isotopically labelled compounds (Berezin and Dobis, 1962a, 1962b). If the model compounds selected for kinetic studies on radical addition are also able to participate in a hydrogen abstraction step, undesirable complexities emerge (Whittemore et al., 1962). A good possibility for exploiting the advantages of competing radical reactions of the aame kinetic order is offered by chain reactions. Although the concentration of chain carrier radicals is low, owing to their high reactivity and short lifetime, they produce easily measured macroscopic changes in the amount of molecular reactants and/or products of the chain process. Any material able to bring about reaction [of types (2) or (4)] with the chain carrier radical and produce another radical of lower reactivity will affect the propagation step, and hence reduce the rate of the chain process. This material will then be called an inhibitor and this procedure to evaluate rate constants is known as the inhibition method. Chain processes most commonly used for kinetic studies are autoxidation and radical polymerization. I n the autoxidation chain the R radical (produced somehow in an initiation step) is not the only chain carrier since plenty of oxygen is available in the system and the reaction R'+Oa
+ RO;
(10)
is very fast. The peroxy radicals then attack the substrate, R-H RO;+R-H
kii
+ROzHfR'
(11)
134
M . SIMONYI A N D F . T U D ~ S
and regenerate R which enters again into reaction (10). The rate of the chain process can conveniently be measured by determining the amount of oxygen consumed as a function of time. By comparing the inhibited rate with that measured without inhibitor, one can obtain the rate constant ratio, k/kllcharacterizing the reactivity of the inhibitor in the reaction RO;
+H-Y
k __j
ROzH +Y'
(12)
(Howard and Ingold, 1962). The complete mechanism of autoxidation is, however, very complex (see Reich and Stivala, 1969). Radical polymerization also affords good possibilities for kinetic studies. Here the chain of subsequent kinetic steps results in the material chain of the polymer product. The propagation step is R'+M
kn
R--M'
(13)
where M is a monomer molecule and the resulting R-M radical is in its reactivity indistinguishable from R'. If reaction (4) stops the growing polymer chain, the molecular weight of the resulting polymer will be lower in the presence of Y-H molecules than in their absence. This phenomenon was used for quantitative studies of the reactivity of "transfer agents" (Mayo, 1943; Gregg and Mayo, 1947, 1953). It is important to note that Mayo's method for determining transfer constants (i.e., k/kpratios) does not require decreased reactivity of Y' as compared with R . On the contrary, it was tacitly assumed that the reinitiation of the chain by Y'+M -+ Y--M'
( 14)
occurred with each Y' radical. If Y'is, in addition, of reduced activity so that reaction (14) would be unimportant, the determination of the inhibited rate of polymerization (easily measurable by following, for example, the volume contraction of the polymerizing mixture) allows a convenient way to evaluate the k/kpratio. I n some cases the reactivity of Y'is moderately decreased and reaction (14)is still significant. A kinetic treatment for this case was given by Kice (1954). The inhibition method allows the determination of the stoichiometry of the reaction between inhibitor and radical. The stoichiometric anomalies found to be general in inhibition processes induced TiidBs (1964a, 1965a,b, 1968)to describe the polymerization process in terms of the participation of hot radicals. Other problems related to radical polymerization are treated in standard textbooks (Burnett, 1954; Bamford et al., 1958; Bevington, 1961; Bagdasarym, 1966).
H Y D R O Q E N ATOM A B S T R A C T I O N R E A C T I O N
135
11. KINETIC STUDIES ON 0-H BOND FISSION Consider the reaction k
R'+H-OQ
R-H+'OQ
(15)
in which R and Q are not yet specified. The heat of the reaction is given by the difference of actual bond energies, i.e., where D-values are bond dissociation energies of model compounds chosen as references. The D-values differ from the energies of cleaving and forming particular bonds by the increments ( E values) the dominant part of which is the resonance energy of the corresponding radicals (EQo = RQo - RHOg ; E , = R E - RE-=). Ethanol has been selected as reference for 0-H bonds (Do-= = 110 kcal mole-l, can be calculated from the heat of formation of C2H50' radical, Gray and Williams, 1959)on the assumption that the C2H50' radical is free of any significant resonance contribution. Consider now four different types of reactant radicals, viz., alkoxy, alkylperoxy, alkyl and hydrazyl. The corresponding reference bond strengths are as follows, Do-H =110 kcal mole-l; DOO--rr=89 kcal molew1 (from hydrogen peroxide, cf., Foner and Hudson, 1956); DDH =94 kcal mole-1 (from i-Pr-H, Cottrell, 1954); DN-H =76 kcal mole-1 (from hydrazine, Foner and Hudson, 1958). The heats of the four reactions are given in Table 1. TABLE 1 Heats of Hydrogen Atom Abstraction from 0-H Bonds by Different Radicals Type of reactant radical AH (kcal mole-1)
Alkoxy Alkylperoxy Alkyl Hydrazyl
-
EE. EQO. ~~+ER.-EQo. 16+Es.-E~o. 344-E~. -EQo.
The value of A H can be taken as a first (although rough) information on the reactivity. Since the activation energy must be more positive than the reaction heat, the possibility of highly endothermic processes can be excluded. It is important to note that suggested correlations (Evans and Polanyi, 1936,1938; Szab6,1962 ;Kagiya et al., 1969)between activation energies and reaction heats (or bond dissociation energies) do not help us
136
M . SIMONYI AND F . T U D B S
too much. One must always remember the objection of Guggenheim and Weiss (1938). Even the correct mathematical treatment of the potential surface method shows the dependence of the activation energy on the reaction heat to be parabolic and not linear (TudBs, 196413,1969). On the other hand the activation energy is not always the only parameter determining reactivity. However, Arrhenius pre-exponential factors cannot, in our opinion, be reliably calculated from entropies of activation. A. Studies on Phenols : The Kinetic Isotope Effect The early experiments of Goldschmidt clearly indicated that phenols are sensitive to radical attack. Not only were fairly stable radicals found in oxidation processes of phenols (Goldschmidtand Schmidt, 1922 ; Goldschmidt and Stiegerwald, 1924),but the oxidation of hydroquinone to quinone could also be brought about by the stable free radical 2,2-diphenyl-l-picrylhydrazyl(DPPH, Goldschmidt and Renn, 1922). The mechanism of the radical attack remained unknown for a long time. The kinetic isotope effect played a very important role in its elucidation. The first suggestion on the mechanism of the reaction between phenols and peroxy radicals emerged from experiments on the antioxidant effect of phenols, by Bolland and ten Have (1947a, b). These authors found a fair correlation between the increase of chain-terminating efficiency and the decrease of the redox potential of phenols and suggested a mechanism in which the phenolic hydrogen atom is transferred to the peroxy radical, i.e., RO;
+ ArOH
--+
ROzH + ArO'
(17)
Another description of a common mechanism for radical reactions of phenols and aromatic amines was given by Boozer and Hammond (1954). They suggested the reversible formation of a loose molecular complex in the f i s t step of the reaction between a peroxy radical and phenol, i.e., RO; +ArOH
RS
[ROiArOH]
-
which was believed to be followed by a very rapid second step [RO;ArOH]
+RO;
k'b
products
(19)
The mechanism of reaction (19) was assumed to be dependent on the structure of the inhibitor; (19) would involve the abstraction of the phenolic hydrogen atom. The whole idea was put forward, in part, to account for the failure of deuteriated inhibitors to show a kinetic isotope effect (Hammond et al., 1955). Another argument was that the Hammett equation, correlating the reactivities of the antioxidants, suggested the
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
137
importance of electron removal from the molecule to the radical. However, the complete electron transfer to the radical was ruled out as being involved in the structure of the complex (Hammond et al., 1955). The scheme proposed in (18) and (19) resembles one frequently applied to multi-step reactions (Bodenstein, 1913). If a steady-state concentration is assumed for the complex (and for the chain carrier R O ; radicals) the overall rate of the process consisting of steps (1 8) and (19) is
where the equilibrium constant Kcis replaced by k,/k-,. If the inequality
k&@Oil B k-c (21) is realized, equation (20) reduces, giving for the overall second-order rate constant, k;:, kit = k, (22) in contrast to kiil = K c k'In a third-order rate constant which would require the opposite inequality
kin[ROi] < k-,. (24) Only (21) and (22) allow for the absence of the isotope effect, while (24) and (23) would predict full isotope effects. The two cases seem to be confused by Hammond et aE. (1955). An interesting observation was made by Davies et a&.(1956) on the correlation of the fundamental stretching frequencies and the antioxidant efficiencies of phenols. Still, several workers could not find any significant isotope effect (Bickel and Kooyman, 1957; Shelton and McDonel, 1958; Ingold, 1960a) and the complex mechanism was further applied (Caldwell and Ihrig, 1960; Hammond and Nandi, 1961). After the important indication of easy isotope exchange in the 0-H group of phenols (Shelton et a&.,1960), the year 1962 became a milestone in the solution of the problem of mechanism. Large deuterium isotope effects were found not only in oxidation processes (Howard and Ingold, 1962; Ingold and Howard, 1962) but also in other systems involving styryl and DPPH radicals (Bird et al., 1962) suggesting that the mechanism is not peculiar to particular radicals. Previous failure to detect thekinetic isotope effect was proved (Howard and Ingold, 1962) to be the consequence of rapid isotope exchange with traces of moisture present even in non-polar media. A number of workers have succeeded in determining deuterium isotope effects since then and the rate-controlling character of the hydrogen abstraction step is no longer in debate.
138
M . SIMONYI A N D F . T U D B S
The high reactivity of N,N,N',N',-tetramethyl-p-phenylenediamine, despite the absence of labile hydrogens, was claimed (Hammond et al., 1955) as an additional argument for a complex mechanism. Against this, we refer to the ability of this compound to react with peroxides (Homer and Schwenk, 1949,1950), and autocatalytic complications found in the presence of other aromatic amines (Gardner et al., 1964, Brownlie and Ingold, 1966), as well as to experiments by Uneyama et al. (1968) which prove the high reactivity of the methyl hydrogen atoms in N,N-dimethylaniline. The existence of measurable kinetic isotope effects does not, however, disprove the complex mechanism, but the need for its postulation is removed. It has also sometimesbeen claimed that every system had its particular characteristics and that it was diacult to say anything, a priori, about an unstudied system (Lloyd and Lange, 1964; Yates and Ihrig, 1965). This is true for the very different systems the radical reaction of phenols can take place in (Caldwell and Ihrig, 1962; Shelton and Vincent, 1963; Vetchinkina et al., 1966). It is certainly diEcult to give ageneral concept of mechanism from a system containing different radicals produced from a phenol (Coppinger, 1964). If our main interest lies in the mechanism of an elementary step, however, it can be hoped to recognize some trends of general validity. I n Table 2, values of kinetic isotope effects are collected. Early values as well as those having large experimental errors have been omitted. I n the reaction of phenols with DPPH, McGowan et al. (1959) suggested a mechanism involving hydride anion transfer from the phenols to the radical. Later Ayscough and Russell (1965, 1967) were able to study reversible hydrogen transfers between DPPH and hindered phenols. Their results again did not exclude the possibility of hydride ion transfer followed by electron re-transfer, but the assumption of such a complication (McGowan and Powell, 1962) seems superfluous. Singh et al. (1966) assumed that phenol first forms a hydrogen-bonded complex with a nitrogen atom of DPPH. The complete transfer of the hydrogen atom to the nitrogen would then occur in a second step. This suggestion seems to be insufficient. First, all radicals studied so far have been found to prefer a charge separation in which the phenol becomes partially positive as compared with the radical in the course of hydrogen abstraction. The suggested hydrogen bond would cause an opposite effect on charges in the course of the reaction. Second, for electronic and steric reasons, the nitro groups in DPPH are more likely to act as proton acceptors in hydrogen bond formation than the nitrogen atoms. Of the data of Table 2, related to DPPH, some isotope effects are quite large while others are about the magnitude characteristic of secondary
TABLE2 Kinetic Isotope Effecta in Radical Reactions of Phenols kE kD
Reactant
No.
Molecule
Radical
Medium
t"C
References
W
a
b ~
1.
2,6-t-Bu~-4-Me-pheno1
Styrylperoxy
Styrene
65
10.6
2.
4-t-Pentylphenol
Styrylperoxy
Styrene
66
10-6
3.
Phenol
Styrylperoxy
Styrene
66
160
4.
2,6 -t-Bu2-phenol
Styrylperoxy
Styrene
66
5. 6.
2,4,6-t-Bu3-phenol 2,6-t-Buz-4-Me-phenol
Styrylperoxy Cumylperoxy
Styrene bene
65 66
7. 8. 9.
Tetralyl hydroperoxide 2,4,6-t -Bus-phenol Phenol
Cumylperoxy Tetralylperoxy Tetralylperoxy
6.7~ cumene in chlorobenzene Tetralin Tetralin
30 65 65
10.
2,6-t-Buz-4-Me-phenol
Tetralylperoxy
Tetralin
66
11.
4-Methoxyphenol
Tetralylperoxy
Tetralin
65
12. 13. 14.
Cumylhydroperoxide 2,6-t-Bu2-4-Me-phenol 4-Hydroxybiphenyl
Tetra1ylperoxy T-butoxy 2,4,6-t-B1~-phenoxy
1 . 7 tetralin ~ in chlorobenzene
cc14 Benzene
24
16.
2,6-t-Buz-phenol 4-Br-phenol 4-NOz-phenol
DPPH DPPH DPPH
Benzene Benzene Benzene
30 30 30
16. 17.
30 122
10.5
2 15
-
4-2 30b
10.0 9-6 9.7
N
10
17* 6.4 3 7.5
.y
8.1
7-0 1.5
Howard and Ingold (1962) Howard and Ingold (1963e) Howard and Ingold (1963s) Howard and Ingold (1963b) Howard et al. (1968) Ingold and Howard (1962) Howard et al. (1968) Gardner et d.(1964) Howard and Ingold (1964~) Howard and Ingold (1964~) Howard and Ingold (1964~) Howard et al. (1968) Ingold (1963a) DaRooge and Mahoney (1967) Bird et al. (1962) Bird et aZ. (1962) Bird et al. (1962)
w
0 0
Z
5
0
g b-
m 1-3
w
2 0
'4
w
0
0
Z
w
TABLE24ontinued Reactant
No.
Molecule
Medim
Radical
t"C
kH -
kr,
References
12
Ayscough and Russell (1965) Singh et al. (1966) Singh et al. (1966) Ayscough and Russell (1967) Godmy et al. (1962) BirdandRussell (1965) Bird and Russell (1965) Bird and Ruasell (1965) Bird and Russell (1965) Bird and Ruasell (1965) Bird and Russell (1965) Bird and Russell (1965) Bird and Russell (1965) Simonyi et d . (1967~) Simonyi et d. (1967~) Simonyi (1969a)
1s.
2,4,6-t-Bus-phenol
DPPH
Benzene
20
19. 20. 21.
Phenol DPPH Oxalic acid DPPH 2,6,-t-Buz-4-i-Pr-phenolDPPH
CH&N Benzene
ca4
20 20 20
1.4 1.90 8.5
22. 23.
a-Naphtol 3-CN-phenol
Styryl Polyvinyl acetate
styrene Vinyl acetate
60 45
6.4 1.0c
24.
3-Ac-phenol
Polyvinyl acetate
Vinyl acetate
45
1.0c
25.
3-C1-phenol
Polyvinyl acetate
V i y l acetate
45
2.0e
26.
Phenol
Polyvinyl acetate
V i y l acetate
45
4.W
27.
3-Me-phenol
Polyvinyl acetate
Vinyl acetate
45
5.8C
28.
4-C1-phenol
Polyvinyl acetate
Vinyl acetate
45
5.7c
29.
4-Me-phenol
Polyvinyl acetate
Vinyl acetate
45
6.5c
30.
2,3,4,6-Me4-phenol
Polyvinyl acetate
Vinyl acetate
45
15.5c
31. 32. 33.
Hydroquinone Pyrocatechol Pyrogallol
Polyvinyl acetate Polyvinyl acetate Polyvinyl acetate
Vinyl acetate Vinyl acetate Vinyl acetate
50 50 50
9.1 12.3 19.2
* Possible experimental errors indicated by the authors.
* The substrate is not a phenol. c
Values recalculated from rate data.
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
141
effects. As far as our knowledge on radical processes may be predictive at all, the case of oxalic acid cannot be regarded as a hydrogen atom transfer from the carboxyl group. There is no doubt, however, that DPPH decomposed with measurable rates even in the presence of oxalic acid. Proll and Sutcliffe (1963) reported DPPH reactions with ethanol, acetic acid and acetic anhydride. They also indicated the accelerating effect of mineral acids, succinic acid, water and tri3uoroacetic acid on the decomposition of DPPH. Proll and Sutcliffe (1963) concluded the formation of an intermediate between solvent and DPPH to be the first step of the reaction while the complete mechanism was supposed to include hydrogen atom abstraction from methyl groups. One may feel these reactions of DPPH to be worthy of further investigation. As was shown by Kennedy and Ingold (1967),the 2,4,6-trihalogenophenoxy radicals were ineffective even in abstracting benzylic hydrogens. The stability of phenoxy radicals depends on their substituents (Pokhodenko, 1969). I n ortho positions the bulky alkyl substituents were found to possess the greatest stabilizing effect. The strength of the 0-H bond in 2,4,6-tri-t-butylphenol was measured by Kalashnikova et al. (1969) to be 71-9& 3 kcal mole-l. The significantly high strain energy (Benson, 1968) associated with the bulky ortho-substituents (8 & 2 kcal mole-l, Mehoney et al., 1969)is an additional effect in decreasingthe bond strength, besides the stabilizing effect of the aromatic ring. Accordingly, 2-alkoxy groups were found to be less effective in stabilizing phenoxy radicals than the 2-t-butyl group (Petrhnek and Pilai., 1969). There seems no indication of significant stabilization of a phenoxy radical by halogen substituents. On the contrary, Howard and Ingold (19658) reported the reactivity of phenoxy radicals in hydrogen atom abstraction to be increased by electron-attracting substituents on the phenol. The same trend was found for the reactivity of styrylperoxy radicals in hydrogen atom abstraction (Howard and Ingold, 196Sc). Since an upper limit of 70 kcal mole-l could be estimated (Kalashnikova et al., 1969)for the strength of the N-H bond in DPPH-H (diphenyl-picrylradical hydrazine) it may be expected that a 2,4,6-trihalogeno-phenoxy would be of higher reactivity than DPPH in abstracting hydrogen atoms from C-H bonds [cf. equation (16).] Thus, it can be suggested that DPPH is unable to abstract hydrogen from a CHs group and that in the mechanism assumed by Proll and Sutcliffe (1963)not only DPPH but a more reactive radical species has to be involved. This means a modification of the direct abstraction mechanism. A possible change in the mechanism of DPPH reactions could also be conjectured from the work of Hogg et al. (1961) who presented a very sharp break in the Hammett plot correlating the
142
M. S I M O N Y I A N D F. T U D B S
reactivity of substituted phenols (of. Table 3). It seems as if a new reaction path is available to more acidic phenols. The isotope effects of Table 2 agree with such a conclusion although the values for phenol and 4-bromophenol are contradictory in this respect. It is to be mentioned that McGowan and Powell (1962)did actually suggest a change in mechanism for phenols containing electron-withdrawing groups. Reactions of phenols with carbon radicals have long been compared to those of phenols and peroxy radicals (Breitenbach et al., 1938; Walling and Briggs, 1946). On the basis of the slight influence of some phenols on the polymerization of certain monomers, an inability of phenols to react with carbon radicals was suggested (Dolgoplosk and Korotkina, 1957; Dolgoplosk and Parfenova, 1957). Although the rate is low for the reaction between styryl radicals and simple phenols (Minouraet al., 1964), Russell and his coworkers were able to measure the rate constants of these processes (Godsay, et al., 1962; Bird et al., 1962) and to prove the hydrogen atom abstraction mechanism as well. On the other hand, in the case of highly reactive organio radicals, the reaction was believed to be the addition of the radical to the aromatic ring of phenol (Bagdasaryan and Sinitsyna, 1961; Whittemore et al., 1962). Though addition of radicals to aromatic systems is known to occur (Heilman et al., 1957; Abramovitch and Kenaschuk, 1967; Abramovici and Pines, 1969; Golden and Benson, 1969) the kinetic isotope effects measured in the reaction of phenols and polyvinyl acetate radicals (Bird and Russell, 1965; Simonyi et al., 1967c)suggest that even with reactive organic radicals hydrogen abstraction will remain the dominant mechanism. (The HO' radical first adds to the ring, according to Adams et al., 1966.) Let us now return to more recent observations connected with the kinetic isotope effect. Bird and Russell (1965) found in the reaction of substituted phenols and polyvinyl acetate radicals that the kinetic isotope effect varied with the nature of substituents in the phenol. They assumed a competition between addition and hydrogen abstraction in those cases where the effect was smaller than five. A re-evaluation of the problem by Simonyi et al. (1967~)suggested that the kinetic isotope effect might vary without changing mechanism, or shifting importance of simultaneous paths. (The mechanism is obviously altered in the caBe of 3-acetylphenol and 3-cyanophenol, cf., Table 2.) Variation of the isotope effect has also been observed in other types of reactions (Bell, 1965; Bell and Goodall, 1966; Jones et al., 1967),without any change in mechanism. Theoretical reasons for varying values of kinetic isotope effects were given (Wiberg, 1955;Brodskii, 1957;Melander, 1960;Westheimer, 1961; Bell, 1961, 1965; Bigeleisen, 1964), suggesting a maximum value for the
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
143
case when hydrogen was in the transition state symmetrically bonded to the atoms between which it was being transferred. The predicted maximum for the kinetic isotope effect has been found experimentally (Bell and Goodall, 1966). There is no need to assume a competition of different mechanisms. An additional remark is necessary, however. Any theoretical consideration of the magnitude of the kinetic isotope effect will only be significant for a single-step reaction (or where essentially one step is reflected in the observed rate; Hanna et al., 1969). I n that case the reactivity can be a good measure of transition state structure (Kresge et al., 1968). The question how far the reaction of phenols with polyvinyl acetate radicals can be regarded as an elementary process will be discussed in detail later in Section 111. For the reaction of polyvinyl acetate radicals and substituted phenols suggested a, Xammett-type correlatian for (Fig. 11, Simonyi et al. (1967~) the isotope effect dependence :
(for u$,, substituent constants see Table 4). Similarly, an approximate Hammett equation was found for primary tritium effects in another system (Jones et at., 1967). An isotope effect dependence of the type suggested in equation (25) implies some inconsistency. If equation (25) holds over a wide range without any change in the mechanism, the isotope effect would either diminish or, on the other side, exceed any reasonable value. On the other hand, any significant deviation from equation (25) would also be difficultto understand, since the “protium-phenols” correlated the Hammett equation fairly well (Simonyi et al., 1967b; cf. also Fig. 2) and a curvature for the isotope effect would mean that deuterio-phenols should not obey the Hammett equation (Simonyi et al., 1969b). A very similar situation was considered by Lewis and Funderburk (1967)in correlating kinetic isotope effects by means of the Brnrnsted relation. This inconsistency emerging from equation (25) will be referred to as “paradox 1” and a possible interpretation w i l l be given in the last paragraph of Section 111. Equation (25) could serve to express an approximate correlation between isotope effects and reactivity. It was shown (Simonyi et al., 1967c)that the isotope effect increased with increasing rate of hydrogen atom tramfer. Although several measured (Bell and Goodall, 1966; Longridge and Long, 1967; Kresge et al., 1968; Hanna et al., 1969) and calculated (More O’Ferrall and Kouba, 1967 and references therein) maxima have been presented for the isotope-effect dependence on
144
M . S I M O N Y I A N D F. T U D B S
transition-state structure, the majority of data on proton transfer reactions simply show increasing isotope effects with decreasing rate of reaction. A similar correlation was found in the chromic acid oxidation of aromatic secondary alcohols (Stewart and Lee, 1964). This reaction was believed to involve a hydride ion transfer. A similar conclusion was drawn from experiments with N-bromosuccinimide by Venkatasubramanian and Thiagarajan (1969), but recent studies suggest that the oxidation of secondary alcohols by Ce(1V) and Cr(V1) involves radical steps (Nave and Trahanovsky, 1968, 1970). A similar, conclusion wa5 drawn by Heckner et al. (1968)from oxidation experiments with alkaline
FIQ.1. Variation of the kinetic isotope effect with substituente of phenol in hydrogen atom abstraction by polyvinyl acetate radicals. The straight line representsequcttion (26). Key: 0 Nos. 26-30 in Table 2. o Nos. 31 and 32 in Table 2. A No. 33 in Table 2.
permanganate. It is therefore not possible to contrast radical and ionic processes according to the type of correlation between rates and isotope effects. At any rate, the correlation between rates and isotope effects of substituted phenols is of the type found less frequently in kinetic studies.
B. Studies 012 Phenols : The Substituent Effect The radical reactivity of phenols is markedly influenced by substitution. I n this paragraph, the substituent effect will be treated in terms of the Hammett equation. All possible substituted phenols are considered except those having t-butyl groups in both ortho positions which are dealt with under IIC. It could be questioned whether phenols containing ortho-substituents other than t-butyl groups would not also cause steric
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
145
effects. The possibility of introducing ortho u+ substituent constants into linear free energy relationship has been demonstrated in gas-phase reactions (Smith et al., 1969; Lum and Smith, 1969)though these results have been attributed to lack of solvation in the transition state. The separation of steric effects from the gross substituent effect has been one of the problems connected with the Hammett equation studied for the last two decades (see Shorter, 1970; Laurence and Wojtkowiak, 1970). The probably easier separation of polar and resonance effects could significantly contribute to our understanding of the nature of radical processes. We must agree, however, with DaRooge and Mahoney (1967)that “there is an incredible lack of appropriate data in the literature”. Consequently, the up treatment has not generally been as careful in radical reactions as in ionic systems. (For a careful treatment in radical reactions, cf. Foldes-Berezhnykh et al., 1969.) I n our opinion, however, such correlations are useful even if drastic approximations are involved. First, despite the enormous amount of work, the Hammett equation still needs some theoretical foundation. As the apt remark of Shorter (1969) expressed: “The empirical success of the Hammett equation is thus to some extent a mystery.” This mystery is connected with the fact that, while the Hammett equation correlates the rate constant with one structural parameter ((T), the rate constant itself is believed to be dependent on at least two parameters and there are only speculations (Ritchie and Sager, 1964) as to what type of correlations between the latter two parameters may exist. On the other hand, even approximate Hammett equations for different radical reactions of phenols permit important comparisons for different systems. I n Table 3 data on Hammett equations of substituted phenols in different systems are collected. Radical reactivities of phenols were shown (Howard and Ingold, 1963a; Simonyi et aZ.,J967b) to correlate better with of constants (Brown and Okamoto, 1958) than with (T constants (McDaniel and Brown, 1958), although in hydrogen abstractions from C-H bonds different authors disagreed in this respect (Gilliom and Ward, 1965; Kennedy and Ingold, 1966; Gilliom and Howles, 1968; Friedrich et aZ., 1970). Further on, we use p values implying correlations with u+ constants. I n spite of the large scatter characteristic of some of the earlier data, it is apparent that electron-donor substituents on the phenol always increase the rate constant of reaction (15). The substituent effect was proved in the same sense by substitution in the radical; electronattracting substituents in the radical increased the rate of hydrogen abstraction (Howard and Ingold, 1964c, 1965a, 1965~).
c rp
TABLE3
0,
Substituent Effects in Radical Reactions of Phenols
No.
Radical
Medium
t°C
1. 2. 3. 4.
Different peroxy radicals Tetralylperoxy Methylmethacrylateperoxy Styrylperoxy
Chlorobenzene Tetralin Methyl methacrylate Styrene
62.5 50 44.4 65
5. 6. 7.
t-Butoxy t-Butow 2,4,6-t-Bu3-phenoxy
ccl4 Chlorobenzene Benzene
DPPH DPPH Polystyryl Polyvinyl acetate Polyvinyl acetate
CCl4 Benzene Styrene Vinyl acetate Vinylacetate
8. 9. 10. 11. 12.
No. of compounds used in calculation
> 30 23 5 11
122 122 24
11 7 5
20 30 60 45 50
14
p
-3.7b
- 3.48d
Pfa
References
E .. Hammond et al. (1955) * ul - 1.47C Davies et al. (1956) H - 2.28d Caldwell and Ihrig (1962) 0 - 1.49 Howard and Ingold (1963e) 2 - 1-19 Ingold (1963b) - 0.74 Ingold (1963b) b - 2.72 DaRooge and Mahoney g (1967) - 4.3~.e McGowan et d.(1959) q - 3.90c Hogg et al. (1961) Y Godsay et aZ. (1959) - 1.68' Bird and Russell (1965) -15.52 Simonyi et al. (1967b) o1
2 u
5
- 6f
4 6 26
- 2.5 -2.70
a Determined from correlationa with u+ constants. " h e correlation coefficient usually given for characterizing the goodness of the correlation is omitted here, since correlation coefficientof the same scatter about a straight line is dependent on the slope. b Determined by Hammond et al. (1955)from a common Hammett equation for phenols and aromatic amines. c Authors did not give the p value. Calculated by Howard and Ingold (1963a). p= - 6.4 and p+= - 4.64 were given by Howard and Ingold (1963a)from six data considered. f Given by Hogg et al. (1961) for the interval: -0.4 < u < 0.2. DaRooge and Mahoney (1967)gave p+ = -2.22 by considering four data; p = - 3.22 and p+= - 2.77 were calculated by Howard and Ingold (1963a) from 13 data considered.
+
2 ra
H Y D R O Q E N ATOM A B S T R A C T I O N R E A C T I O N
147
To interpret the substituent effect, Howard and Ingold (1963s) assumed that the 0-H bond could be at least partially polarized in the transition state, thus giving a partial positive charge on the oxygen and a partial negative charge on the radical, i.e., 8, [ArO
8-
... H . . .R]
(26)
I n another description, the transition state was assumed to involve the contribution of three resonance forms (Howard and Ingold, 1963b; Ingold, 1963b):
+-
[ArO:H.R ++ ArO.H:R ++ ArO.H:R]
(27)
If we consider the hydrogen atom transfer as successive transfer of proton and electron, the kinetic isotope effect requires the proton transfer step to be rate determining.2 This assumption, together with the negative p values found experimentally, implies that the electron moves somewhat ahead of the more slowly moving proton (Ingold, 1967). As a contrast, consider the following kinetic scheme involving total charge separation, i.e., ArOH+R’
ke
k- s
+ -
IArOHRl
k+ ___f
hO’+HR
(28)
By applying steady state to the intermediate, the overall rate constant,
k, will be equal t o
Since electron transfer and re-transfer are much faster than that of the proton, i.e., keBk+,equation (29) reduces to
k =-k+ ke k-,
= Kek+
Equation (30) predicts full isotope effects. The overall Hammett p may be expected to be P = Pe-P-e+P+ (31) if a Hammett equation is valid for each partial step. Since p+ can be quite large (in proton transfer from neutral phenol to an anion, p was found to be + 1.8 by Blake et al., 1966), the sum of \pel and Ip-e] must be even larger. The experimental p values (Table 31,however, do not seem t o be in contradiction with a sum given by equation (31). a The authors’thanks are due to Prof. I(.U. Ingold, who called their attention (Ingold, 1967) to a mistake in their discussion (Simonyi et aZ., 1967b) of the non-simultaneous proton and electron transfer.
148
M . S I M O N Y I A N D F. T U D B S
The experimental indication of a reversible electron transfer (from
N,N,N',N'-tetramethyl-p-phenylenediamineto DPPH, Hausser, 1959) seems to be in accord with the above scheme. On the other hand, a dipolar transition state is implied by scheme (28), which is difficult to accept as a general case. The nature of radicals (Table 3) does not require full charge separation in the transition state. Hence, the type of complex given in (28)seems also of limited applicability in the mechanism of radical reactions of phenols, even if there is no direct kinetic evidence against it. Early expectations that steric hindrance would increase the kinetic isotope effect (Howard and Ingold, 1962; Ingold, 1963a) by analogy to electrophilic substitution (Baciocchi et al., 1960) tacitly referred to a two-step mechanism, however. I n some of the earliest series of Table 3, the scatter of the data is quite large and the magnitude of the p values not certain. One of the least reliable values is that of Hammond et al. (1955)since it was determined by plotting different families of compounds on the same graph. However, aromatic amines present complications : they form nitric oxide radicals in the course of inhibited autoxidation (Thomas and Tolman, 1962b; Adamic et al., 1969; Adamic and Ingold, 1969) and the kinetics are subject to autocatalysis (Gardner et al., 1964; Brownlie and Ingold, L 966). There are also remarkable differences between radical reactivities of substituted anilines and phenols towards carbon radicals (Simonyiand Tudos, 1969). Hence amines and phenols should be treated differently. Two approximations usually applied in plotting Hammett relations for phenols are the assumed equality of ortho and para positions and an expected additivity of single substituent effects for polysubstituted cases. It is clear that neither of these can be exactly valid, although recent evidence shows that both approximations are fairly good for the acidity of phenols (Bolton et al., 1970; &ficheet al., 1970). The effect of a single ortho substituent on the rate of hydrogen abstraction is usually close to that of apara substituent (e.g.Howard and Ingold, 1963b). I n the presence of two ortho substituents, characteristic behaviour can be detected in different types of solvents. I n non-polar media, 2,6-dialkyl-substituted phenols (without bulky alkyl groups) react with DPPH at rates one order of magnitude lower than would have been expected from the effect of single substitution, on the assumption of ortho-para equivalence and additivity (McGowan et al., 1969; Hogg et al., 1961). For styrylperoxy radicals, 2,6-dialkylphenols (even if one of the ortho substituents was t-butyl) gave a distinct Hammett line representing reactivities of about half of simple substitution (Howard and Ingold, 1963b). Some results suggest that the reactivity-decreasing effect can be entirely gradual starting from one o r t b methyl group to
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
149
2,6-di-s-alkyl substitution (Ingold, 1963b; McGowan et al., 1959). On the other hand, studies on polyvinyl acetate radicals, in vinyl acetate solvent, suggested that 2,6-dialkylphenolswere apparently correlated by the Hammett equation for single substitution, the deviation of 2,6dialkylphenols being not significant and always within the average scatter (Bird and Russell, 1965; Simonyi et al., 1967b; cf., Fig. 2). This individual behaviour of 2,6-dialkylphenols is to some extent
FIG.2. Re-establiahment of Hammett correlation for halogen-aubatitutedphenols in the reaction with polyvinyl acetate radical. The straight line repreaents the Hammett equation found for 26 compounds (No. 12 in Table 3). Key: o non-halogen substituted phenols including 2,6-dimethylphenols, excluding 2,6-t-Bua-4-subatituted-phenols; 8 halogen-substitutedphenols.
connected with the reactivity of the radical in the hydrogen atom abstraction process. It is not a full explanation, however. The apparently equivalent reactivity of 2,6-dialkylphenols toward polyvinyl acetate radicals with that of singly substituted phenols is referred to as “paradox 2 ” and a possible explanation will be suggested in Section 111. A change in the mechanism of hydrogen-atom abstraction by DPPH was suggested by McGowan and Powell (1962) when strongly acidic phenols were involved. This assumption agrees with the data of Hogg et al. (1961) indicating that the slope of the Hammett line is almost zero
150
M . SIMONYI A N D F . T U D B S
if u > + 0.2. The induced decomposition of DPPH in acidic media seems to be a possible explanation (see also the preceding paragraph). Somehow similarly, anomalous behaviour of halogen-substituted phenols was presented in studies on polyvinyl acetate radicals. Chloroand bromo-substituents present in ortho or para positions of the phenol increased the rate of hydrogen atom abstraction in spite of the fact that the u+ constants of these substituents were positive. Moreover, other halogenophenols were also found to be more reactive than predicted from the Hammett equation of electron-donor substitution, and the enhancement of the reactivity could be additively calculated by the number of halogen substituents. Intramolecular hydrogen bonds (Baker and Shulgin, 1965),as well as the limited possibility of radical addition to the aromatic ring (Simonyiet ab., 1967b, 1969a)could not account for the additivity, however. New substituent constants (utOrr) were suggested for the particular system, and found to re-establish the Hammett correlation for a number of halogen-substituted-phenols (Fig. 2). The suggestion of constants must not be regarded as an explanation of the anomaly. On the contrary, these constants help somehow to make the problem more explicit. There is still no interpretation for the anomalous behaviour. The suggested atom constants are given in Table 4. TABLE 4 Substituent Constants for Some Halogen Substituents u+a
d m b
Substituent
m-
P-
m-
P-
Fluorine Chlorine Bromine
0.352 0.399 0.405
-0.073 0.114 0.160
0.17
-
-0.14 -0.10 -0.16
0
0.10
Brown and Okamoto (1968).
* Simonyi et al. (1967b).
C. Studies on Phenols : The Steric Euffect If both ortho-positions of phenol are substituted by t-butyl groups, the reactivity of the compound is reduced regardless of the individual features of the system considered. This decrease in the reactivity of 2,6-di-t-butylphenols as compared to other phenols having similar Z u + values may be attributed mainly to steric effects. These compounds,
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
151
usually called hindered phenols, could be correlated with the CJ+ values of their para substituents. The Hammett function of hindered phenols has always been found to be distinct from that of other substituted phenols. Collected data are presented in Table 5. Although it is easy to accept that hindered phenols are of lower reactivity than non-hindered ones, it is not quite straightforward to explain that the slope of Hammett correlation applicable to sterically hindered phenols (pat)is, usually, less negative than that of the other compounds. The two lines must intersect somewhere which implies higher reactivity for sterically hindered compounds in a certain range of substitution. It could be of interest that, the ratio of the slopes of nonhindered and hindered phenols (p/patin Table 5 ) seems to have an average value of about 1.4 for six different systems, with the sole exception of the data taken from a study in chlorobenzene medium. However, in the last case some kind of radical-solvent interaction was assumed (Ingold, 196313) to account for the solvent effect observed. The decreased slope of the Hammett equation for hindered phenols was related to the decreased polar character of the transition state (Howard and Ingold, 1963b; Ingold, 1963b). As an interpretation of the constancy of the PIPstratio, it was suggested that the 0-H bond could be twisted out of the plane of the ring (Simonyi et al., 196713) and that the decreased p value was an apparent consequence of decreased substituent effects. Neighbouring methyl and t-butyl groups were shown to be twisted out of the plane of the aromatic ring (Ryba et al., 1965b)and a good deal of evidence known as steric inhibition of resonance suggested such effects (e.g., Baciocchi and Illuminati, 1967). This suggestion, however, disagreed with an earlier conclusion drawn from the infrared spectra of phenols (Ingold, 1960b). The frequency of the fundamental 0-H stretching vibration of substituted phenols in CC1, was found to be correlated with u values. Since the two lines correlating the frequencies of hindered and non-hindered phenols were found to be parallel, it was concluded that the 0-H group of 2,6-di-t-butylphenols had to be in the plane of the ring. I n gas phase measurements, intensity correlation gave support for this suggestion but not the frequency data (Ingold, 1962). It is widely accepted that the 0-H group of a non-hindered phenol lies in the plane of the ring; cis-trans isomerism has been indicated for orthoalkyl phenols (Ingold and Taylor, 1961a, b), and the barrier height to internal rotation of the 0-H group in phenol has been measured (Ingold and Taylor, 1961a; Lowe, 1968). The earlier frequency correlation was, however, reinvestigated by Leary (1968) who found certain differences in the slopes of correlation lines for different families of phenols. It is also 6
T ~ L 5E Studies on 4-Substituted 2,6-di-t-Butylphenols
No. 1. 2. 3. 4. 5. 6. 7.
Radical Tetralylperoxy Styrylperoxy t-Butoxy t-Butoxy DPPH DPPH POlJWklyl Bo8t8b
Medium Tetrdin Styrene ca4
Chlorobenzene cc14 Benzene Vinyl acetate
t"C 50 65 122 122 20 30 50
No. of compounds usedin calculations 3 10 7 5
3 2
3
Pt
-0.99
- 1.11 - 0.86 - 0.78 -3.14
- 2.75
- 1.10
P m
Reference
1.49 1.34 1.38 0.95 1.37 1.43 1.38
Davies et al. (1956) Howard and Ingold (I963b) Ingold (1963b) Ingold (1963b) McGowen et al. (1959) Hogg el al. (1961) Simonyi el d.(1967b)
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
153
remarkable that the frequency data of Bellamy and Williams (1960) gave good correlations in several solvents but also suggested the lack of correlation in the gas phase. It is perhaps not yet possible to give a final conclusion of this problem. It is, however, undoubtedly proved from infrared measurements that 2,6-d-t-butylphenols differ somehow in structure from the others. Similarly, polarographic half-wave potentials showed anomalous behaviour in c a e of 2,6-di-t-butyl substitution (Ryba et al., 1966a). This substitution causes also effects on the acidity of phenols that are not quite well understood. (Rochester and Rossall, 1969). It still seems most reasonable to assume that the abstraction of the hydrogen atom from hindered phenols occurs in a direction perpendicular to the plane of the ring. The perpendioular direction can be favoured for the transition state of hindered phenols since it is less covered by the bulky ortho substituents than the coplanar conformation. If, at least, one of the ortho-positions is free, the abstraction takes place most probably in the plane of the ring. The lifetime of the coplanar conformation can be much longer than that of the perpendicular state, and more molecules are found at a time in the coplanar form than in perpendicular conformation. Ifthe transition state of the hydrogen abstraction is different for hindered and non-hindered phenols, this difference could ratio is fairly constant. explain why the p/pBDst A predominant factor in determining the distinct behaviour of hindered compounds is most probably the strain between the neighbouring groups. The strain energy in 2,4,6-tri-t-butylphenol was calculated as 8 kcal mole-' while no such energy has been indicated for 2,6-dimethylphenol (Mahoney et al., 1969). The suggestion of lower barrier to rotation of the OH group in the tributylphenol than in nonhindered phenols (Ingold and Taylor, 1961a) is also associated with the strain energy. One may further assume that the volume of the ortho t-Bu-group cannot entirely account for all the distinct behaviour of 2,6-di-t-butylphenols, such as, the infrared frequencies of the 0-H stretching vibration which are the highest of all substituted phenols. I n contrast 2,6-di-t-octylphenols (t-octyl =1,1,3,3-tetramethylbutyl) give frequencies close to those of non-hindered phenols, in spite of the probably greater steric covering associated with the t-octyl group (Ingold, 1962). This difference may speculatively be connected with a higher rotation frequency for the t-Bu-group. Hence the OH group of the 2,6-di-tbutylphenol series could experience more frequent impacts from the rotating alkyl group and hence be found to twist out of the plane more than the OH group of 2,6-di-t-octylphenols.
M . S I M O N Y I A N D F. T U D ~ S
154
D. Studies on Other Hydyoxylic Compounds 1. Hydrogen abstraction from hydroperoxide
This process was assumed to account for the inhibiting effect of tetralin hydroperoxide on the oxidation of cumene (Thomas and Tolman, 1962a) and further confirmation was given from studies on autoxidation inhibited by phenol (Howard and Ingold, 1964c; Thomas, 1964). The hydrogen abstraction from phenol by the tetralylperoxy radical was found to be reversible ;owing to the considerableresonance energy of the latter, the heat of reaction should be very small (of. Table 1). Abstraction from the hydroperoxide by the 4-methoxyphenoxy radical was indicated to be unimportant (Howard and Ingold, 1964~).More direct evidence for the cleavage of the 0-H bond in hydroperoxides has emerged from the extremely large kinetic isotope effects (Table 2) reported by Howard et al. (1968). Finally, Mahoney and DaRooge (1970) were able to study the equilibrium
@f
TOO- -I-
+w 9'
?H
TOOH
t
(32)
t
(where T indicates the tetralyl moiety), and gave d H = - 7 & 1-7 kcal mole-l for the enthalpy change in (32) and 36 kcal mole-l for the difference in the heats of formation of TOO' and TOOH. 2. Hydrogen abstraction from >ATOH groups
The O--H bond of disubstituted hydroxylamine is, in general, easy to cleave homolytically since nitric oxide is produced, a radical well known for its high stability (Hark and Thomas, 1957; Buchachenko, 1962). Preparative evidence was given by Banfield and Kenyon (1926) by the production of a stable free radical in the mild oxidation of the corresponding N-phenyl-N-alkylhydroxylamine. The latter compound and its substituted derivatives readily react with polyvinyl acetate radicals. Substituents in the aromatic ring were found to influence the reactivity of the hydroxylamine slightly ; p = - 0.16 f 0.04 (Simonyi et al., 1967a). I n comparing this value with those of phenols, the decreased substituent effect can be rationalized by considering two factors. First, the O-H bond is more remote from the aromatic ring in hydroxylamine than in phenol. Second, the stability of phenyl nitroxide radicals is higher than that of phenoxy radicals. It is
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
155
interesting, however, that the p value still remained negative. The reactivity of N,N-diphenylhydroxylamine toward polyvinyl acetate radical was found one order of magnitude greater than that of the mono-phenyl derivatives, suggesting that the additional resonance contribution of the second ring was still important. 3 . Hydrogen abstraction from alcohols
The problem of alcohols seems to be more complex. Liquid-phase evidence on 0-H bond fission in methanol was reported by Cher (1963) who found the reactivity of C-H sites in reaction with methyl radicals to be favoured by a factor of 15 over that of the 0-H position. The reaction has also been studied in the gas phase by Shannon and Harrison (1963) and by Shaw and Thynne (1966) who agreed that the 0-H site was about 1.5 times more reactive than a C-H bond a t 164OC. If one accounts for the difference in temperature between the gas and liquid phase data, the disagreement becomes even greater (Shaw and Thynne, 1966). Hydrogen abstraction from methanol by trifluoromethyl radicals have been studied in the gas phase by Carlton et al. (1966)and by Morris and Thynne (1968). The former suggested that the reactivity of a C-H bond is favoured by a factor of about 3 independently of the temperature while the latter found that the reactivity of the 0-H position a t 164°C is about 4.5 times as great as that of a C - H site. Although CfF, radicals are, in general, more reactive than &I, radical8 (e.g.,Tedder and Walton, 1967; Morris and Thynne, 1968; Kerr et al., 1969; Arthur and Gray, 1969a; Ferguson and Pearson, 1970; Bullock and Cooper, 1970; but not always: Arthur and Gray, l969b), data on 0-H bond fission by CF, radicals are still scarce ; no example has been reported in a review by Tedder and Walton (1967). The possibility of the 0-H bond fission is certainly connected with the structure of the resulting alkoxy radica,l. Alkoxy radicals are able to undergo unimolecular decomposition (Quee and Thynne, 1968). The enthalpy requirements for decomposition varies from 22 kcal mole-' for methoxy to 2 kcal mole-' for 1-butoxy (Gray et al., 1967). The structural requirements for rearrangement from an alkoxy radical to a carbon radical have been suggested (Bunn and Das, 1970). It could therefore be expected that the competition between C-H and 0-H positions would be shifted in favour of the former in the case of higher alcohols. Accordingly, Gray and Herod (1968) found the CY-C-H position of ethanol at 15OOC to be twice as reactive as the 0-H site toward methyl radicals. Ethanol was also found to react only through its C-H positions in the liquid phase (Glockling, 1956). Recent direct observations confirm the early statements on the
156
M . S I M O N Y I A N D F. T U D ~ S
preference of the C-H positions in an alcohol to react in liquid phase. In aqueous solution of iso-propanol, two possible mechanisms have been suggested for hydrogen abstraction, resulting in the formation of either and the possibility of forming an (CH&COH or CH,-CHOH-CH,, iso-propoxy radical has been neglected (Burchill and Ginns, 1970). Only the two carbon radicals have been experimentally identified (James and Sicilio, 1970) by use of an electron spin resonance (e.5.r.)flow technique (Norman and Gilbert, 1967). Other direct evidence for the dominant reactivity of the C-H bond in alcohols has been given by Perkins et al. (1970). The high stability of nitroxide radicals allows short-lived radicals to be scavenged by a nitroso compound, and the hyperfine structure of the e.8.r. spectra of the adduct can be studied. (The same 1967. technique had earlier been used by TudCjs et al., 1965; Kende et a!., A new promising method for taking emission 8.s.r. spectra of short-lived radicals has recently been described by Atkins et al., 1970). As shown, a great difference exists in the competition of C-H and 0-H reactivities between gaseous and liquid phases. I n an excellent survey on comparing liquid phase and gas phase radical processes, Mayo (1967) listed rate constant ratios for competing bimolecular hydrogen abstraction from cyclohexane and unimolecular decomposition of t-butoxy radicals a t 4OoC in the gas phase, and in 14 different solvents. (The data were taken from an earlier paper by Walling and Wagner, 1964.) Considering a 200-fold variation of the rate constant ratio from the gas phase value to that in acetic acid solvent, Mayo (1967) concluded that the effect of phase change on the rate constant ratio was the largest one he knew in a free radical reaction. Such a scale for competing C-H and 0-H fissions in alcohols is not yet available. One may suggest, however, that the latter competition could well be as sensitive to phase and/or solvent change as that considered by Mayo. The idea that a free radical in solution would be less free than in the gas phase (Mayo, 1953, 1967 ; Bullock and Cooper, 1970) seems only one factor in this respect. Another important question is how much less free a hydroxyl group will be in a polar solvent than in a non-polar one or in the gas phase (Gray and Herod, 1968). A discussion related to this question is presented in Section 111. 4. Hydrogen abstraction from the carboxyl group
The possibility of abstracting a hydrogen atom from the carboxylic group may be expected to be even more limited than 0-H bond fission of alcohols. The existence of the process was suggested to account for the decarboxylation of acids during the oxidation of hydrocarbons. The reactivity of the acidic hydrogen, however was estimated to be 100-times
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
157
lower than that of a C-H site in the acid (Berezin and Ragimova, 1962). Accordingly, the gross reactivity of acetic acid towards c H 3 radicals in n-heptane solvent originated almost entirely from the methyl group. Tritium labelling in the carboxyl group allowed observation of the partial reactivity of the acidic hydrogcn. However, the tritium isotope effect could not be measured. A further sharp decrease of the rate constant of carboxylic hydrogen abstraction was indicated when the concentration of acetic acid exceeded 0.2 mole% and was attributed to dimerization of the acid (Dobis et al., 1968). The hydrogen atom abstraction from oxaIic acid by DPPH (Table 2, Singh et al., 1966) has already been commented upon. Carboxylic acids seem to be unsuitable models for studying radical reactivities of 0-H bonds.
111. THEROLEO F HYDROGEN BONDINQ’ One may feel qualitatively that the reactivity of an 0-H group would somehow change when this group enters into a hydrogen bond. Free and bonded OH groups generally coexist, however. As a first approximation, one may distinguish free and bonded 0-H groups and attribute different reactivities to them. The characteristic difference of the reactivity of methanol in gas and liquid phases suggests that reactivities of free and bonded groups could differ from each other. In treating the bonded groups as being in a different state from the others, one must remember that “the hydrogen bond is an extremely variable thing’’ (Murrell, 1969). That means not only very different strengths for different types of hydrogen bonds; medium or weak bonds having a strength close to that of solvation will continuously change their lengths and angles and represent some distribution of cleaving and re-forming systems usually treated as equilibria. There can be no doubt about agross equilibrium itself, since the system usually possesses time-independent properties. Defining an equilibrium between a proton donor and an acceptor through hydrogen bonding, however, often implies the same difficulties as specifying the nature of interaction between a solute and the solvent. Although there are several possible experimental techniques to measure some property connected with hydrogen bonding (Pimentel and McClellan, 1960a), a good deal of uncertainty exists, in most cases, in defining a stoichiometric model according to which measured data can be treated and some equilibrium constant evaluated. Accordingly, opposite views coexist in current literature. For a general review see Brat02 (1967).
158
M . S I M O N Y I A N D F. T U D B S
The trailsition from specific (e.g., hydrogen bond) to non-specific interactions between the solvent and a solute is almost continuous (Eastman and Rehfeld, 1970) and it is generally difficult to distinguish between them. Therefore, the discussion of data obtained even by advanced experimental techniques often takes an empirical approach to solvent-solute interactions (Wasylishen et al., 1970). Doubts have recently been cast on attempts to deduce any equilibrium constant for the formation of complexes between a solvent and a solute; the influence of an inert solvent on hydrogen bonding equilibria has also been demonstrated (Christian and Tucker, 1970). Simultaneously, solution models have been suggested and applied (Biais et al., 1970; Dos Santos et al., 1970a; Servanton et al., 1970)to interpret data on self-association by introducing a set of equilibrium constants. A different approach by Huggins ( 1970) involves the fundamental consideration of intermolecularly contacting surface areas distinguished for different segments of which molecules are composed. By introducing equilibrium constants for different types of contact, the theory can predict some physical properties of simple mixtures. It is hoped that further development of this promising theory will reach areas of kinetic interest. Since our interest concerns the reactivity of OH groups, in spite of the complexity of solvent-solute interactions, we agree with Fletcher (1970) that specific and non-specificinteractions of a functional group must be distinguished from those of the molecule as a whole”. Recent kinetic studies on solvent effect concluded that not medium polarity but rather hydrogen bonding between solvent and reactant played the dominant role (Curci et al., 1970; Dearden, 1970). For the same reason, hydrogen bonding has been introduced as an equilibrium process into detailed schemes of radical reactions (Sukhanova and Buchachenko, 1965; Andronov et al., 1967; Zaikov et al., 1967; Simonyi, 1969b; Simonyi et al., 1969b).
A. Influence of the Medium on the Rate The most prominent solvent effects are characteristic of ionic reactions (Parker, 1967, 1969) while rates of homolytic processes are rather independent of the nature of solvents (e.g., Pryor, 1966). I n some radical reactions, however, the effect of solvent is significant. Complex formation between radical and solvent has been suggested sometimes to account for the altered reactivity (Mayo, 1953; Russell, 1958; Ingold, 1963b; Andronov et al., 1967)in particular systems. In chain reactions, additional measurements are needed to decide which of the elementary steps is influenced by the solvent (Howard and Ingold, 1964b, 1965c; Burnett, 1965, 1969; Fischer and Schulz, 1970).
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
159
I n some cases even inert diluents can influence the rate constant of radical processes. This effect has been interpreted in radical polymerization by assuming the existence of hot radicals (TudBs, 1964a, 1965a, b). The rate of oxidation of hydrocarbons was found to be slightly dependent on the solvent (Boozer et al., 1966). The effect was attributed to the slightly dipolar character of the propagation step (Howard and Ingold, 1964a). I n measuring the rate of hydrogen abstraction from substituted phenols by peroxy radicals in different solvents, Howard and Ingold (1964a) inferred a rate-decreasing effect of hydrogen-bond formation between phenols and the solvent. Not only classic (C.K. Ingold, 1953) but also recent solvent effect theories (Parker, 1969)claim that the influence of the solvent on the rate of reaction is a consequence of specific solvation of initial, transition and final states and of any intermediate. If the lifetime of the transition state is too short ( sec), however, the reorientation of the solvent will lag behind (Bell, 1965; Jones, 1969b). Consequently, the solvation of the transition state will resemble that of the initial state. If a hydrogen atom is abstracted from an alkane by an alkyl radical, both the initial and final state of the reaction involve neutral species and it is only the transition state where some limited charge separation can be assumed. I n the case of a homolytic 0-H bond fission, however, the initial state possesses a certain polarity and possible changes in polarity during the reaction depend on both the lifetime of the transition state and the nature of the attacking radical. If the unpaired electron is localized mainly on oxygen in the reactant radical, the polarity of the h a 1 state will be close to that of the initial state and any solvent effect will primarily depend on the solvation of the transition state. Solvent effects can then be expected since the electron and proton transfers are not synchronous. Solvent effects on the reaction of 0-H bonds and carbon radicals could, at least partly, be accounted for by considering initial and final state solvation. Since hydrogen-bond formation with an electron donor (but not proton donor) solvent can exist in the initial state, may exist in the transition state but cannot exist in the final state, a rate-retarding effect of hydrogen bonding is to be expected. In reactions of 0-H bonds with alkoxy or peroxy radicals, however, the solvent effect may be an indication for the existence of a long-lived intermediate. The influence of hydrogen bonding on the rate of proton transfers can be opposite, especially if the proton is bonded to atoms between which it is being transferred. Proton transfers within the hydrogen bond have been suggested to be fast owing to significant tunnelling (Cannon, 1958;
-
160
M . S I M O N Y I A N D F. T U D B S
Bicz6 et al., 1966; Zimmermann, 1969), although recent evidence indicates a much lower rate (Crooks and Robinson, 1970) than is generally supposed.
B. The Hydrogen Bonding Equilibrium Possible methods of detecting hydrogen bonding were surveyed by Pimentel and McClellan (1960a). Here only two particular methods will be considered briefly to represent essential implications in usual equilibrium constant determinations. 1. N.M.R. measurements The n.m.r. position of a hydrogen-bonded proton is shifted downfield
relative to the n.m.r. position of the corresponding proton in a free state (Schneider et al., 1958). The shift of the signal with concentration of the proton donor is interpreted to be the consequence of rapid exchange between different states (e.g. bonded and free) resulting in an average signal the position of which is given by
6,
ZiaiSc (33) where 6, is regarded as a weighted average of the individual ai chemical shifts and ai is the weighting factor of the ith individual state. While hydrogen bonding generally results in a single peak for the proton involved, proton exchange between chemically different positions is often not so fast as to give the average signal (Jentschura and Lippert, 1970), hence the different states can be simultaneously observed (Neszmklyiet al., 1967 ;Gold and Tomlinson, 1970). I n the simplest case of association i.e., for the existence of only two individual states: free (subscript 1)and bonded (subscript 2) =
Sa = a161+(1-q)82
(34)
(Pimento1and McClellan, 1960b). If we know the values of Sl and S2, then equation (34)affords the possibility to determine a1 and further the equilibrium constant from
K =
1-al al(Ca- 1 +all
(35)
where C, is the total concentration of the proton acceptor and only 1: 1 composition of the hydrogen-bonded complex is assumed. Stillassuming that we know the values of s1 and a2 (and they are independent of concentration), equation (34) allows one to study the equilibrium over a wide concentration range of the proton acceptor. (Variation in concentration of the proton donor is strongly limited so as to avoid self-
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
161
association.) This test of the equilibrium is essential if one uses the equilibrium constant to deduce any property of the system. Let us focus attention on the implied difficulties. (a) We very often do not know Sz, the chemical shift of the bonded proton. If K is not too large, a system containing only bonded protons cannot be realized. (b) Both a1 (Gramstad and Becker, 1970) and Sz (Schaefer and Kotowycz, 1968) depend on the nature of “inert” solvent used as third component. Moreover, CCl,, which is frequently used as inert solvent, can itself act as proton acceptor (Fletcher, 1969) or participate in other types of specific interaction (Whetsel and Kagarise, 1962a, b ; Vogel and Drago, 1970). Hence many studies on hydrogen bond equilibria could be seriously in error. Even for known a1 and SZ values, determined in inert, non-polar and polar solvents, respectively, the treatment implies that both a1 and S2 will remain unchanged also in a mixture of polar and non-polar components of variable concentration. Where data allow a test, this assumption is proved to be invalid (Gramstad and Becker, 1970). (c) I n n.m.r. measurements, 6 values are measured relative to some
reference peak. The most frequently used internal reference, tetramethylsilane, has recently been indicated not to be inert (Homer et al., 1970) and only externally referenced chemical shifts have been suggested to be reliable (Becconsall et al., 1970; Malinowski and Weiner, 1970). (d) Most systems are not so simple as to contain the proton in only two different kinds of state. Systems of high proton-donor concentration, when self-association plays a significant role (Goldstein et al., 1970; Limbach et al., 1970), or hydrogen bonding with multifunctionaJ proton donors (de Jeu, 1970) are typical examples. Then the right-hand side of equation (33) contains a t least three terms. However, we have usually no idea on the value of as and hence the equation contains at least four unknown quantities. By taking the spectra in different concentrations we have a set of 8, values; we could further require constancy for the equilibrium constants introduced, thus making the problem tractable. It is entirely arbitrary, however, since the possibility of testing any equilibrium assumption is lost. The system could be made determinate only for a computer while still remaining non-determinate for the scientist. 2. Infrared measurements
The second method considered here is superior to n.m.r. in respect of the more or less distinct appearance of absorption peaks assigned to the stretching vibration of free and bonded species. The direct evidence of
162
M . S I M O N Y I A N D F. T U D G S
hydrogen-bond formation even allows estimation of structural parameters, such as, low bending frequencies of the hydrogen bond from infrared spectra (Thomas and Thompson, 1970). I n equilibrium constant determinations, we need extinction coefficients to evaluate equilibrium concentrations for free and bonded species. The absorption peaks have to be resolved. This is in general attainable by using an inert diluent as third componcnt besides protondonor and proton-acceptor molecules. Let us now consider difficulties emerging from infrared studies. (a) The absorption peak of the bonded species is often unsuitable for determining an equilibrium concentration because of simultaneous contributions of associations of different stoichiometry (House and Cook, 1969),the individual bonded peaks being unresolved. Consequently, one equilibrium constant is often unsatisfactory to describe the system (Dos Santos et al., 1970b). Associations of 2 proton donor molecules to 1 acceptor have several times been suggested (Merrill, 1961; Whetsel and Ragarise, 1962b; Ramaswamy et al., 1967; Detoni et al., 1970),while the stoichiometry of 1 proton donor to 2 acceptor molecules seems to be rarely considered (Gramstad and Van Binst, 1966). The latter stoichiometry suggests, however, that other types of association may perturb measurements on hydrogen bonding. A special feature of association phenomena is the continuum of behaviour between stoichiometric and random associations (Baker and Wilson, 1970) which makes it always difficult to evaluate equilibrium constants for association, regardless of the individual method employed. (b) One often uses the extinction coefficient determined for free molecules (in dilute solution of the proton donor by employing a solvent regarded as inert, see Fletcher and Heller, 1967)to determine equilibrium constants in polar-nonpolar mixtures (e.g., Widom et al., 1957; Kolbe and Pracejus, 1966). The components are, however, not always highly dilute (especially the proton acceptor), and hence even an inert medium could change their optical properties. Moreover, frequency shifts of absorption peaks of free species may be caused by non-specific solvent effects (Huong and Lassegues, 1970). Since there is a quantitative relationship between intensity and frequency of gas-phase infrared peaks (Crawford, 1952) and some relationship can be expected even in liquid phase, extinotion coefficientsdetermined in a nonpolar medium are not exact when applied to a polar-nonpolar mixture. (c) Infrared studies usually ignore spontaneous emission of the sample. Although its effect is small on measurements at room temperature, enthalpy and entropy determinations may be significantly influenced (NeszmClyi and Tmre, 1968 ; Varshyi, 1970).
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
163
3. Structural effects on hydrogen bonding
Owing to the aforementioned difficulties, equilibrium constants of hydrogen bonding determined by different authors are often contradictory. Nevertheless, some general trends are unambiguously established. Electron-donor substituents in the proton acceptor increase the equilibrium constant of hydrogen bonding (Kagarise and Whetsel, 1962 ; Su and Hong, 1968; Engberts and Zuidema, 1970; Bhowmik, 1970) and i t has been suggested that hydrogen-bonding ability is a reliable measure of basicity for similar types of bases (Cook and House, 1969). If the proton donor is substituted, electron-donor substituents decrease the equilibrium constant (Bhowmik and Basu, 1963, 1964; Gragerov and Pogorelyi, 1969; Bauer et al., 1970). An extremely good correlation has been presented between the dissociation constant of substituted phenols and n.m.r. chemical shifts in polar medium, while the same type of correlation was found to be poor in CCll (Socrates, 1970). It may be suggested that the difference between a polar solvent and CC1, in hydrogen bonding is not only quantitative but also qualitative. Similarly, hydrogen bonding to rr-electrons (Ellison and Meyer, 1970) may be regarded as e transition from specific to non-specific interaction. The influence of substitution of either phenol, as proton donor, or aromatic proton acceptors has been indicated to be rtlmost of the same magnitude on hydrogen bonding (Ryebokobylko and Chekunov, 1970). However, both electron-donor and electron-withdrawing substituents have been found to decrease the degree of selfassociation of phenol (Vanderborgh et al., 1970). The influence of steric effects on hydrogen bonding is not a clearly understood problem. Widom et al. (1957) indicated that hydrogen bonding with acetophenone as proton acceptor was hindered as compared to dialkylketones. Considering the gradation of hydrogen-bond energies when the alkyl group of the ketone is varied (Balasubramanian and Rao, 1962) one may assume here the operation of electronic rather than steric effects. However, the same order of influence for the variation of the alkyl group in alcohols has been indicated on hydrogen bonding (Becker, 1961; Reece and Werner, 1968; Cook and House, 1969). The empirical fact that methanol is the alcohol forming the strongest hydrogen bonds, whereas acetone is the most effective dialkylketone proton acceptor suggests that the problem is somehow unclear. Bellamy and Williams (1960) assumed that increasing volume of ortho alkyl substituents in phenols could enhance the lifetime of free 0-H groups. The 0-H group of 2,6-di-t-butylphenols was claimed to be free even in polar solvents. By contrast, Denisov et uZ. (1964) published
164
M . S I M O N Y X A N D B. T U D ~ S
equilibrium constants suggesting the absence of any steric effect on hydrogen bonding with hindered phenols. A similar conclusion was drawn by Singh and Rao (1966a) who found comparable enthalpies of hydrogen bonding for hindered and simple phenols. On the other hand, Zaikov et al. (1968) published equilibrium constants clearly indicating steric effects for hindered phenols on hydrogen bonding with methyl ethyl ketone. Similarly, Kuts et al. (1968) indicated a large steric effect on proton exchange between hindered phenols and ethanol; rates of proton exchange and hydrogen bonding were found to be closely related (Gragerov and Pogorelyi, 1969). Finally, the enthalpy and entropy changes of association seem to be related to each other in series of related compounds (Nelson, 1970; Bhowmik, 1970). Owing mainly to the weak strength of most hydrogen bonds as compared with covalent chemical bonds, the hydrogen bonding equilibrium is, in general, seriously affected by specific and non-specific solvent-solute interactions resulting in a dependence of the equilibrium constant on the kind of solvent used (Scott et al., 1968). The information obtainable from the dipole moment of the solvent is usually insuffcient since the dipole moment corresponds to the molecule as a whole and not to individual groups or bonds of the molecule which are mainly responsible for solvation interactions (Lamotte et al., 1970). If equilibrium studies are performed in dilute solutions involving really inert solvent component, an uncertainty exists in extrapolation to systems containing one of the hydrogen-bonding components as solvent (Christian and Tucker, 1970). Nevertheless, specificsolvation of the reaction center may be expected to exert a greater iduence on the reaction rate than solvation of other parts of the molecule (Fletcher, 1970). Hence, it still seems reasonable to make an approach toward the interpretation of solvent effects on radical processes by considering the specific solvation of the functional group. Since 0-H groups are typical proton donors, the inclusion of equilibrium hydrogen bonding to the kinetic scheme of hydrogen atom abstraction from 0-H bonds could be such an approach, regardless of several difficulties connected with the determination of reliable equilibrium constants. C . Consideration of Hydrogen Bonding to the Interpretation of Certain Kinetic Anomalies
To describe detailed kinetics for systems containing peroxy radicals and phenols in mixed water-methyl ethyl ketone media, six kinds of
H Y D R O G E N ATOM ABSTRACTION REACTION
165
eIementary hydrogen abstraction reactions were considered, each differing either in the nature of peroxy radical (free or bonded to water) or in the nature of phenol (free, bonded to water, or bonded to methyl ethyl ketone). The resulting expression for the overall hydrogen abstraction from phenol became so complex, however, that some of the elementary reactions had to be neglected (Andronov et al., 1967;Zaikov et al., 1967). As a representative example, the reaction of polyvinyl acetate radicals and phenols in vinyl acetate medium seems simpler. For the sake of simplicity we neglect any interaction between the radical and medium and regard all polyvinyl acetate radicals as equally reactive. However, we distinguish free phenol molecules from hydrogen-bonded ones according to the scheme K
ArOH+X R' Jki
ArOH ...X R' Jkr
(36)
For the overall hydrogen abstraction process, still of second order in concentrations of the reactants, we get (Simonyi, 196913)
k =
k,+k,K K+1
a+K
= kz-
K+1
(37)
where K is dependent on the concentration of polar molecules according to the equation K =K'[X] (38) The rate constant corresponding to free phenols ( k , ) has been assumed to be one order of magnitude greater than k2 (Simonyi, 1969b). Howard and Ingold (19644 suggested that Hammett p was independent of the solvent for peroxy radical-phenol reactions. Although hydrogen bonding has been indicated to influence substituent constants (Taft et al., 1963), the interpretation of solvent effects on p values is still a difficult task (e.g., Kondo et al., 1969). The first question emerging from equation (37) is, how the Hammett dependence of k is to be interpreted. The Hammett dependence of K seems to be much more slight (Simonyi and Holly, unpublished; Mukherjee et al., 1970) than that of k (Table 3). Consequently, the Hammett dependence of k is, in effect, that of k2 while the ratio ( a+ K ) / ( K+ 1) vanes little with u+. As a representation, Table 6 gives some typical values for the equilibrium constant of hydrogen bonding between phenols and vinyl acetate as determined by infrared spectroscopy. Let us now try to interpret paradox 1 (see p. 143), i.e., the curved Hammett dependence of the kinetic isotope effect shown by Fig. 1.
M . SIMONYI AND F . T U D ~ S
TABLE 6 Equilibrium Constants for Phenol-Vinyl Hydrogen Rondinga Proton donor Phenol 2.6-Me*-phenol 2,4,6-Mea-phenol 2,6-i-Pra-phenol 2.4-t-Bua-6-Me-phenol 2,6-t-Bua-4-Me-pheno1
Aoetate
K' ( I mole-l)b
Ke
3.0 0.7
32 7.4 8.4 7.4 6.3 2.1
0.8
0.7 0.6 0.2
Simonyi and Holly, to be published. Miiller and Gebler (1970)suggested that mnolality is a much better quantity to determine equilibrium constants than molarity. In our opinion, the w e of molarity is not the main source of errors for equilibriumconstants on hydrogen bonding. c Extrapolated to vinyl acetate medium. a
b
According to equation (37), the measurable value of the kinetic isotope effect is given bv
Of the three factors on the right-hand side of equation (39), the second one could be assumed to give a linear Hammett dependence with small slope, while the first is expected to give extreme (minimum or maximum) dependence by analogy with equilibrium isotope effect dependences (Singh and Rao, 1966b; Kolbe, 1969). Moreover, it has been found that KD > KH in some cases (Simonyi and Holly, unpublished) which could account for some extremely large values. (The possibility of inverse isotope effects for hydrogen bonding was considered by Howard et al., 1968). The third factor on the right-hand side of equation (39)is probably constant and close to unity (Simonyi et al., 1969b). Hence the product of the fist two terms may produce a curved Hammett dependence, as well as some extremely large values that would be diEcult to interpret theoretically. Since the experimental kinetic isotope effect is a composite ratio, it would not be appropriate to discuss any theoretical reason for the magnitude of the isotope effects. Note, however, that the hydrogen atom is attached to oxygen in the reactant and to carbon in the product. Therefore it is not straightforward to assume too much symmetry in the transition state. There is close analogy to the case described by Lewis and Funderburk (1967).
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
167
Let us now turn to paradox 2 (see p. 149); il lack of apparent steric effect for 2,6-dimethylphenols in a polar medium (Fig. 2) while different Hammett equations were found for 2,6-dimethylphenols in non-polar media (e.g., Howard and Ingold, 1963b). The data of Table 6 suggest a gradation in the diminution of K when ortho substituents are introduced to phenol. Accordingly, the ratio (a+ K ) / ( K+ 1) in equation (37) is about 1.3 for simple phenols while it is about 2 for 2,6-dimethylphenols. Thus equation (37) indicates a possible compensation for 2,6-disubstituted phenols ; as K is decreased by ortho substituents, the rate constant k2 has to be multiplied by an increased factor expressing enhanced contribution of free phenol molecules to the overall rate. In case of 2,6-di-t-butylphenols, however, even the partial rate constant of free molecules (k,) is so markedly reduced that the compensation by the factor (a+ K ) / ( K+ 1) is no longer effective. Our data do not allow a more quantitative test of scheme (36). It serves as quaditativc explanation until more experimental information will be obtained.
IV. ARRHENIUS PARAMETERS A. Interrelation Between A rrhenius Parameters The elementary rate constants of the polymerization of vinyl acetate have been determined over a temperature range of 50 degrees (Berezhnykh-Foldes and Tiid&, 1964; Tud6s et al., 1967),and Arrhenius parameter determinations became possible for inhibition processes of radical polymerization of vinyl acetate according to equations (40)-(42) dlnk _- E dT R!P
__
E RT
Ink = 1nA - __
InA
=
d(T In k) dT
(Arrhenius, 1889). It should be remembered that E and A are experimentally determined from the slope and intercept of the tangent to a curve (Christiamen, 1949; Benson, 1960b). In Table 7, data are given for hydrogen abstraotion reactions ; some results obtained by techniques other than polymerization are also included. Some of the data of Table 7 (Nos. 2-9) satisfy fairly well the correlation logA = l-1,E
(43)
TABLE 7 Arrhenius Parameters for Hydrogen Atom Abstraction Reactions
+ Q,
00
E
Reactant No.
1. 2. 3. 4. 6.
6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
Radical Polyvinyl acetate Polyvinyl acetate Polyvinyl acetate Polyvinyl acetate Polyvinyl acetata Polyvinyl acetate Polyvinyl metate Polyvinyl acetate Polyvinyl acetate c -Hexenylperoxy 2,4,6-Me3-heptylperoxy a-Phenethylperoxy a-Phenethylperoxy Styrylperoxy Methyl ethyl ketoneperoxy 2,4,6-t-Bu3-phenoxy 2,4.6-t-Ry-phenoxy 2,6-t-Buz-4-i-Pr-phenoxy 2,6-t-Bu2-4-i-Pr-phenoxy DPPH DPPH DPPH DPPH
A values in litre mole-1 mc-1 units. Simonyi (1967);Tudos (1969). C Robb and Shahin (1959). g Buchachenko et at. (1961). C Tsepalov and Shlyapintokh (1962). f Howard and Ingold (1965b) 17 Andronov et al. (1967). a
b
Moleode
log Am
kcal mole-1
Reference
Ba&eld condenaate Hydroquinone 2,4,6-Mes-phenol 2,4,5-Mea-phenol Phenol 2,6-t-Buz-P-Me-phenol 9,lO-dihydroanthrecene Fluorene Triphenylmethane 2,6-t-Bua-4-Me-phenol Diphenylamine a-Naphthol &Naphthol 2,6-t-Bua-4-Me-phenol a-Naphthol 2,4,6-t-Bu3-phenol 2,2-diphenyl-1-pycrylhydrazine 2,2-diphenyl-1-pycrylhydrazine 2,6-t-Bu2-4-i-Pr-phenoxyk 2,4,6-t-Bu3-phenol 2,4,6-t-B~-phenol-d 2,6-Bu2-4-i-Pr-phenol 2,6-t-Bua-4-i-Pr-phenoxyk
7.9 6-8 6.6 6.4 4.4 4.1 7.4
6.3 f 0.6 6.1f 0-6 5.5f0.6 5.9 f 0.6 4.4 f 1.0 3.4 f 1.0 6.5 f 0.8 4.9 f 1.0 4.3 f 1.0
b b b b b b b b b
6.0
c
3.5 6.8 7.3 5-6 5.5 f 0.5
d
5.7 4.8
7.6r 4.8 9.0 8-9 7.9 6.4 3-21 3-91 4.6Z 4.9’ 3.91 3 8‘ 4.3r 4-51
Kreilick and Weiesmann (1962). Ayeoough and Ruasell(l966). j Ayscough and Russell (1967). Hydrogen abatraction from the tertiary C H bond. I The authors did not give the preexponent. h
1.0 f 0.6 4.8 6.0 f 1-0
6.2 6.5 f 0.8 7-720.7 6.8 f 0.6 6.4 f 1.5
e
e
f
9
h i
i j
i i
i i
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
169
and No. 1 is the only exception among the data for abstractions by polyvinyl acetate radicals. Data from oxidation experiments and those from DPPH studies deviate in opposite directions (although No. 15 correlates well). Correlations of the above type me called compensation effects since the correlated variation of A and E results usually in slight changes of the rate constant, a good part of the individual influences of A and E being cancelled. Equation (43),if strictly valid, implies that rate constants of reactions of the given series depend only on one instead of two parameters. (The strong resemblance of this problem to that of the Hammett equation was considered by Leffler and Grunwrtld, 1963, and by Ritchie and Sager, 1964.) I n a general case, the correlation may be given by logA = logA"+- 2.3 'RTiE
(44)
and the rate constant will be ii = Aexp(
-$] A"exp (E T:(i- -~ RT)) =
(45)
Since A" is constant for a given series satisfying equation (44), any structural change in the reactants must be reflected in the only parameter, E , determining the rate constant. If T =Ti, the rate constants of all reactions in the given set will be identical. For that reason, Ti is called the "isokinetic temperature ". It is a mathematical consequence of equation (44) and has no physical meaning. (The only physically reasonable isokinetic temperature is the absolute zero.) Nevertheless, the value of Ti as compared with a medium value of the temperature range of experiments (Tsxp)can help us to classify possible correlations of the Arrhenius parameters (Simonyi, 1967; Tiidlis, 1969). (i) If Ti > TexP,changes in E are partially compensated by changes in A ("partial compensation") (see Gregg and Mayo, 1947, and Berezin and Dobis, 1962a). (ii) If Ti =Tsw, changes in E are entirely cornpensated by changes in A . This case is the most sensitive t o experimental errors since wrong extrapolations from plots of logk against 1/T may give apparent total compensation (cf. Leyland et aE.,1970). (iii) If Ti > Texp,changes in E are overcompensated by changes in A. This type is represented by equation (43)demonstrating an unusual trend of increasing rate constant with increase of E . The correlation (43) is not restricted to non-hindered phenols, but must nevertheless be of limited validity even for polyvinyl acetate radicals. The Arrhenius A
170
M . S I M O N Y I A N D F. T U D ~ S
cannot exceed some reasonable valuc related to the collision number in gas or liquid phase reactions, and, downward deviations from equation (43)should be expected for higher values of E . To give a more comprehensive picture, the data of Table 7 and those for gas phase reactions are plotted in Fig. 3. For highly polar atoms or radicals, the rate constants of hydrogen abstraction are usually high, due to both high A and low E . For abstractions from hindered sites, mainly from hindered phenols, extremely low A values are obtained and the
F I G . 3. Arrhenius parameters of hydrogen abstraction reactions. The curve drawn suggests the trend of the compensation phenomenon. Key: o abstraction by alkyl radicals from C-H bonds (Kondratiev, 1970); A abstraction by F or C1 atoms and HO' radicals from C-H bonds (Kondratiev, 1970);8 abstractionby CCls radical from methane (Hautecloque, 1970); further data are given in Table 7, as follows: No. 1; Nos. 2-9; 0 NO. 10; W NO.11; 0 NOS.12,13; x NO. 14; 0 No. 15; @ No. 16; v Nos. 17,20,21; Nos. 18, 19, 22, 23.
+
nature of the reactant radical seems to influence primarily the activation energy while affecting A to a smaller extent. (Peroxy radicals seem to be exceptions in this respect.) The occurrence of very low activation energies in abstractions from hindered sites as well as the validity of (43) for hindered compounds is not in accord with the predictions of Leffler (1955) and of Taft (1956). It may be of interest that for carbon radicals (but not for halomethyl ones) a continuous transition appears between liquid and gas phase data suggesting some generality in the compensation phenomenon of homogeneous reactions. Thcre are gas-phase data indicating decreased A values with decrease of E. The line drawn gives the suggested trend. At
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
171
first it follows equation (43), indicating overcompensation, and then reaches the other limiting case where A is almost constant. The rate constant corresponding to a point moving along the line from left to right will first increase, pass through a maximum value and then decrease. Any compensation phenomenon should reach somewhere the second stage where A is approximately constant ; if the first stage corresponds to partial compensation, however, the transition will not be so sharp and the experimentally available range will be insufficient to reach the second stage. Considering the fairly good correlation given by Greiner (1970) between activation energies of OH and CH, radicals in hydrogen abstraction from alkanes, one may assume similar compensation phenomena also for other radicals, as far as the mechanism does not alter (Leffler, 1955; Schaleger and Long, 1963). The mechanism is actually altering when OH radicals attack phenol (Adams et al., 1966). It may be possible that other radicals (e.g., peroxy ones) also give a transition to gas-phase data, though their actual correlation differs from that of carbon radicals. More experimental data are needed for such a test. It could be questioned, whether scheme (36) involving hydrogen bonding of phenols did not imply a change in mechanism as compared with hydrogen atom abstractions from hydrocarbons. Hydrogen bonding does not operate in the latter case. May C-H and 0-H bond fissions be regarded as members of the same series of reactions at all? On the other hand, if the rate constant for phenols are given by the sum of two terms according to equation (37), what do the Arrhenius parameters for phenols represent? The activation energy could be an average of the two values (El-dHK) and E 2 (if K is sufficiently high), where E l and E 2 are activation energies for abstractions from free and bonded phenols, respectively, and A H K is the enthalpy change on hydrogen bonding. Since AHK is negative and E 2 is presumably higher than El,the two values ( E l - d H g ) and E 2 may be quite close to each other and the suggested scheme will not necessarily cause significant modification of E2. The first question, i.e., how far the identity of mechanism has to be restricted, is related to the as yet unknown interpretation of the compensation phenomenon. A detailed survey of compensation effects was given by Leffler and Grunwald (1 963) who also discussed structural and mechanistic implications of the problem. An additional remark on this subject seems relevant. Most discussions have considered compensation effects in both equilibria and rate processes. Accordingly, the effect is often called the enthalpy-entropy relationship. As for equilibria, the enthalpy-entropy relationship “can be considered as part of the axiomatic foundations of
172
M . SIMONYI AND F . T U D ~ S
thermodynamics”, as suggested by Thorn (1969). For rate processes, however, the case could be different. The term “activation entropy” is closely related to the unjustified assumption of thermodynamic equilibrium for the activated complex (Eyring, 1935; Glasstone et al., 1941). The activation entropy refers to a transformed value of the pre-exponential factor but conveys no more information than A. It still appears that even 8r qualitative explanation is lacking for the compensation phenomenon of rate processes in homogeneous systems.
B. Formal Interpretation of the Compensation Phenomenon Arrhenius parameters are, in general, temperature dependent. The physical reason for the dependence is the average character of the activation energy (Tolman, 1920 ; Menzinger and Wolfgang, 1969). Suppose that in a one-step reaction, even in the homogeneous phase, there are simultaneous paths from reactants to products each of which differs in some dynamic parameter from the others (Eliason and Hirschfelder, 1969; Mayer and Simonyi, 1971). Each path has its particular For the sake of activation energy (E,)and pre-exponential factor (Ai). simplicity, consider only two competing paths so that the rate constant of the total (still one-step) reaction will be given by
where E j > E , , and the particular parameters are supposed to be independent of temperature. If Eiand E j are significantly differing the resultant Arrhenius plot will obviously be curved (of., Hulett, 1964), but for a 50°C wide temperature interval (including room temperature), the deviation from linearity is not significant if E j - Ei =5 kcal mole-’. Let us assume that the two competing reaction paths indicated by equation (46) are characteristic for a series of related reactions. Let us further assume that the degree of competition between the two paths varies according to the individual features o f members of the series. The alteration of the weight of a particular path can formally be taken into account by changing one or some of the particular Arrhenius parameters in (46). Some results of these changes are collected in Table 8 together with restricting conditions of different types of correlations realized. The conditions given for cases 2 and 3 in Table 8 are necessary for an important influence of the competition of paths on the value of E . The same conditions could also be expressed in terms of the ratio, A&,
HYDROGEN ATOM ABSTRACTION REACTION
173
TABLE8 Some Types of Correlation Between Arrhenius Parameters &8 a Result of Varying Individual Parmeters in Equation (46)a Variables in No. equation [26] 1. 2. 3. 4. 6. a
Aj Ej Ar EialdEj AiandAj
Restricting conditions None E < Ej- RTexp E > Ei+RT‘exp ( E j - 4 ) = const. Aj/Aa = const.
Type of correlation between A and E
Tc < T e x p ;overcompensation Tc < T e x p ;overcompensation T<> Texp;partial compensation Tr = m; A = const. Ti = 0; E = const.
I. Mayer, unpublished.
for a given value of AE. It should be noted that Table 8 gives conditions for the two limiting cases shown in Fig. 3. The assumption on two distinct reaction paths expressed by equation (46) is a simplification of most real cases. Attempts have been made to examine Arrhenius parameters of reactions proceeding not on two disorete energy levels but in a continuous energy interval. According to preliminary results (Mayer, unpublished), compensation trends similar to those of Table 8 are to be expected if the paths of significantly different energies make an appreciable contribution to the total rate. As for physical reasons of the above assumption, the following possibilities could be considered (Simonyi and Tiidiis, to be published) : (i) The chemical reaction is a non-adiabatic transition (cf. e.g., Kauzmann, 1957) the probability of which is related to the activation energy. (ii) The reaction cross section is related to the total energy of colliding pairs of reactants (Le Roy, 1969). (iii) The tunnel effect makes an important contribution to the rate of reaction. (Cf. Bell, 1959a, b, 1969; Caldin, 1969.)
V. CONCLUSION Some of the particular kinetic problems of hydrogen atom abstractions from 0-H bonds have been discussed in the context of questions of general interest. The aim has been to make unsolved problems explicit rather than to give final solutions. It seems desirable to realize explicitly that many of our rrtheories” and treatments are entirely empirical (Guggenheim, 1938) strongly resembling the entirely empirical origin of the rate constant (Harcourt and Esson, 1866, 1867).4 4 Excerpts from these fundamental papers can be found in a recent collection by Back and Leidler (1967).
174
M . SIMONYI AND F. TUDBS
ACKNOWLEDGMENT The authors are indebted to Prof. G. Schay for his comments on the manuscript of this work, to Dr. G. Horanyi for many stimulating discussions and to dipl. phys. I. Mayer for the use of material prior to publication.
REFERENCES Abramovici, M., and Pines, H. (1969).J. Org. Chem. 34, 266. Abramovitch, R. A., and Kenaachuk, K. (1967).Can. J . Chem. 45, 509. Admnic, K.,and Ingold, K. U. (1969).Can. J . Chem. 47,295. Adamic, K., DUM, M., and Ingold, K. U. (1969).Can. J . Chem. 47, 287. Adams, G. E.,Michael, B. D., and Land, E. J. (1966). Nature 211,293. Albery, W. J., and Robinson, B. H. (1969). Tram. Faraday SOC.65,980. Andronov, L.M., Zaikov, G. E., Maizus, Z. K., and Ernanuel, N.M. (1967). Zhur. Fiz. Khim. 41,2002. Arrhenius, S. (1889). 2.phy8ik. Chem. 4, 226. Arthur, N. L., and Gray, P. (1969a). Trans. Faraday SOC.65,424. Arthur, N.L.,and Gray, P. (1969b). Tram. Faraday Soc. 65,434. Atkins, P. W., Buchanan, I. C., Gurd, R. C., McLauchlan,K. A., and Simpson,A. F. (1970). Chew. Commun. 513. Ayscough, P. B., and Russell, K. E. (1965).Can. J . Chem. 43, 3039. Ayscough, P.B., and Russell, K. E. (1967).Can. J . Chem. 45, 3019. Baciocchi, E.,and Illuminati, G. (1967). Progr. Phys. Org. Chem. 5, 1. Baciocchi, E.,Illuminati, G., and Sleiter, G. (1960). Tetrahedron Letters, 30. Back, M. H.?and Laidler, K. J. (1967).“Selected Readings in Chemical Kinetics”. Pergamon. Bagdasaryan, Kh. S. (1966). “Teoriya Radikalnoi Polimarizatsii”, Izd. Naulra. Moscow. 2nd edition. Begdaaaryan, Kh. S., and Sinitsyna, Z. A. (1961).J. Polymer Sci. 52, 31. Baker, A. W., and Shulgin, A. T. (1965).Can. J . Chem. 43, 650. Baker, K. M.,and Wilson, R. G. (1970). J . Chem. SOC.(B) 236. Baker, R. T.K., Silbert, M., and Wolfgang, R.. (1970).J. Chem. phys. 52, 1120. Belandin, A. A. (1932). 2.phys. Chem. B19,451. Balasubramanian, A.,and Rao, C. N. R. (1962). Spectrochim. Acta 18,1337. Baliga, B. T.,and Whalley, E. (1964).Can. J . Chem. 42, 1835. Bamford, C. H., and Jenkins, A. D. (1961).J. Polymer Sci. 53, 149. Bamford, C. H., and Jenkins, A. D. (1963). Tram. Faraday Soc. 59, 530. Bamford, C. H., Barb, W. G., Jenkins, A. D., and Onyon, P. F. (1958). “The Kinetics of Vinyl Polymerization by Radical Mechanisms ”. Butterworths, London. Bamford, C. H., Jenkins, A. D., and Johnston, R. (1969). Trans.Faraday Soc, 55,
418.
Badield, F.H., and Kenyon, J. (1926). J . Chem. SOC.1612. Bartlett, P. D.,and Kwart, H. (1960).J. Am.Chem. Soc. 72,1606. Bartlett, P. D.,and Kwart, H. (1952).J. Am. Chem. SOC.74,3969. Bauer, R. F.,LaFlair, R. T., and Russell, K. E. (1970).Con. J . Chew&.48,1251.
HYDROGEN ATOM ABSTRACTION REACTION
175
Bazilevskii, M. V., and Trosman, E. A. (1968). Kinetika i Kataliz 9, 516. Becconsall, J. K., Daves, G. D. Jr., and Anderson, W. R. Jr. (1970). J. Am. Chem. SOC. 92, 430. Becker, E. D. (1961). Speclrochim. Acta 17,436. Behrman, E. J., and Edwards, J. 0. (1967). Progr. Phys. Org. Chem. 4, 93. Bell, R. P. (19594. Tram. Faraday SOC. 55, 1. Bell, R. P. (1959b). “The Proton in Chemistry”. Cornell Univ. Press, Ithaca, New York. Bell, R. P. ( 1961). T r m . Faraday Soc. 57,961. Bell, R. P. (1965). Disc. Faraday SOC. 39, 16. Bell, R. P. (1969). Lecture presented a t the Symposium on Isotope Effects, York. Bell, R. P., and Goodall, D. M. (1966). Proc. Roy. SOC.A294, 273. Bell, R. P., Fendley, T. A., and Hulett, J. R. (1956). Proc. Roy. SOC. A235, 453. Bellamy, L. J., and Williams, R. L. (1960). Proc. Roy. SOC. A254, 119. Bennett, J. E., Mile, B., Thomas, A., and Ward, B. (1970). Adv. Phys. Org. Chem. 8, 1. Benson, S. W. (196Oa). “The Foundations of Chemical Kinetics”, Chapter 12. McGraw-Hill. Benson, S. W. (196Ob). “The Foundations of Chemical Kinetics”, Figure IV.3. McGraw-Hill. Benson, S. W. (1968). “Thermochemical Kinetics”, p. 47. Wiley. Berezhnykh-Foldes,T., and Tiid&, F. (1964). Vysokomol. s o d . 6, 1523, 1630. Berezin, I. V., and Dobis, 0. (1962a). Doklady Akad. NaukS.S.S.R. 142,105. Berezin, I. V., and Dobis, 0. (1962b). Doklady Akad. NaukS.S.S.R. 144,374. Berezin, I. V., andRagimova, A. M. (1962). Zhur. Fiz. Rhim. 36, 581. Bevington, J. C. (1981). “Radical Polymerization”. Academic Press, London, New York. Bhowmik, B. B. (1970). J. P h p . Chem. 74,4442. Bhowmik, B. B., andBmu, S. (1963). Tram. Faraday Soc. 59,813. Bhowmik, B. B., and Basu, S. (1964). Trans. Faraday SOC. 60, 1038. Biais, J., Dos Santos, J., and Lemanceau, B. (1970). J. Chim. Phy8.67, 806. Bickel, A. F., and Kooyman, E. C. (1957). J. Chem. SOC.2416. Bicz6, G., Ladik, J., and Gergely, J. (1966). Acta Phys. Acad. Sci.Hung. 20, 11. Bigeleisen, J. (1964). Pure Appl. Chem. 8, 217. Bird, R. A., and Russell, K. E. (1965). Can. J. Chem. 43, 2123. Bird, R. A., Harpell, G. A., and Russell, K. E. (1962). Can. J. Chem. 40, 701. Blake, J. A., Evans, M. J. B., and Russell, K. E. (1966). Can. J. Chem. 44, 119. Bodenstein, M. (1913). 2. physik. Chem. 85, 329. Bolland, J. L., and ten Have, P. (1947a). Tram. Paraday Soc. 43, 201. Bolland, J. L., and ten Have, P. (194713). Disc. Faraday SOC. 2, 252. Bolton, P. D., Ellis, J., and Hall, F. M. (1970). J. Chem. SOC. (B) 1252. Boozer, C. E., Hammond, G. S. (1954). J. Am. Chem. SOC. 76, 3861. Boozer, C. E., Hammond, G. S., Hamilton, C. E., and Peterson, C. (1955). J. Am. Chem. SOC. 77,3380. Bratog, S . (1967). Adv. Qmnt. Chem. 3, 209. Breitenbach, J. W., Springer, A., and Horeischy, K. (1938). Ber. 71B, 1438. Bridger, R. F., and Russell, G. A. (1963). J. Am. Chem. SOC.85, 3754. Brodskii, A. I. (1957). “Khimiya Izotopov”, p. 259. Izd. Akad. Nauk. S.S.S.R. MOBCOW. Brown, H. C., and Okamoto, Y. (1958). J. Am. Chem. SOC.80,4979. Brownlie, I. T., and Ingold, K. U. (1966). Can. J. Chem. 44, 861.
176
M . S I M O N Y I A N D F. T ~ D ~ S
Buchachenko, A. L. (1962). Optika i Spektroskopiya 13, 796. Buchachenko, A. L., Kaganskaya, K. Ya., and Neiman, M. B. (1961).Kinetika i Kataliz 2, 161. Bullock, G., and Cooper, R. (1970). Tram. Faraday Soo. 66,2066. Bunn, D., and Das, R. C. (1970). Chem. Commun. 1466. Burchill, C. E.,and Ginns, I. 8. (1970).Can. J . Chem. 48, 1232. Burnett, G. M. (1964). “Mechanism of Polymer Reactions”, Vol. 3, of High Polymers series. Interscience, New York, London. Burnett, G. M. (1965).Progr. Reaction Kinetim, 3, 449. Burnett, G. M. (1969). Main lecture presented at the IUPAC Symposium on Macromolecular chemistry, Budapest. To be published in “Kinetics and Mechanism of Polyreactions”, Vol. 6,Akademiai Kiadb, Budapest, (1971). Caldin, E. F. (1969). Chern. Revs. 69,136. Caldwell, R.G., and Ihrig, J. L. (1960). J. Polymer Scd. 46, 507. Caldwell, R.G., and Ihrig, J. L.(1962). J . Am. Chem. SOC.84,2878. Cannon, C. G. (1968).Spectrochim. Actol 10,341. Carlton, T. S.,Steeper, J. R., and Christensen, R. L. (1966).J . Phy8. Chem. 70, 3222. Chakrovorty, K., Pearson, J. M., and Szwarc, M. (1969). Intern. J . Chem. Kinetics 1, 367. Chapman, D. L., Briers, F., and Walters, E. (1926).J. Chem. SOC.662. Cher, M. (1963). J. Phys. Chem. 67,605. Chou, C. C., mdRowland, F. S. (1969). J . Chem. Phys. 50, 2763. Christian, S. D., and Tucker, E. E. (1970). J . phys. Chem. 74,214. Chnstiansen, J.A. (1949). Acta Chem. Scand. 3, 61. Cook, R. L.,and House, J. E. Jr. (1969). Tram. Illinois State Acad. Sci. 62, 223. Coppinger, G. M. (1964). J. Am. Chem. SOC.86,4385. Cottrell, T. L. (1954). “The Strength of Chemical Bonds”, p. 187. Butterworths Sci. Publ., London. Crawford, B. Jr. (1962). J. Chem. Phys. 20,977. Creak, C . A,, Dainton, F. S., and Ivin, K. J. (1962). Tram. Faraday SOC.58,326. Crooks, J. E.,and Robinson, B. H. (1970). Chem. Commun. 979. Curci, R.,DiPrete, R. A., Edwards, J. O., and Modena, G. (1970).J.Org. Chem.35, 740. Dainton, F. S., Ivin, K. J., and Wilkinson, F. (1967).Chem. SOC.(London)Spec. Publ. 9, 187. Dainton, F. S., Ivin, K. J., and Wilkinson, F. (1969). Tram. FaradaySoc. 55,929. DaRooge, M. A.,and Mahoney, L. R. (1967).J. Org. Chem. 32,l. Davies, D. S.,Goldsmith, H. L., Gupta, A. K., and Lester, G. R. (1966).J . Chem. SOC. 4926. Dearden, J. C. (1970). Lecture presented at the International Conference on the Mechanisms of Reactions in Solution, Canterbury, Code No. D2. de Jeu, W. H. (1970). J. Phys. Chem. 74,822. De Mare, G. R., and Huybrechts, G. (1968). Tram. Faraday Soc. 64,1311. Denisov, E.T., Aleksandrov, A. L., and Shchedrin, V. P. (1964). Izv. Akad. Nauk S.S.S.R.(seria khimiya) 1683. Denney, D. B., and Denney, D. 2. (1960). J. Am. Chem. SOC.82,1389. Detoni, S., Hadki, D., Smerkolj, R., Hawranek, J. and Sobczyk, L. (1970). J. C h m . SOC.(A) 2861. Dobis, O., Nemea, I., and Kerepes, R. (1968). Acta Chim. A d . Sci. Hung. 55,216. Dolgoplosk, B.A.,and Korotkina, D. Sh. (1957). Zhur. Obshch. Khim. 27,2226.
H Y D R O G E N ATOM A B S T R A C T I O N R E A O T I O N
177
Dolgoplosk, B. A., and Parfenova, G. A. (1957). Zhur. Obshch. Khim. 27, 3083. Dos Santos, J., Biais, J., and Pineau, P. (1970s). J. Chim. Phys. 67, 814. Dos Santos, J., Gruege, F., Pineau, P. (1970b). J. Chim. Phys. 67,826. Eastmen, J . W., and Rehfeld, S. J. (1970). J . Phys. Chem. 74, 1438. Eliason, M. A., and Hirschfelder, J. 0. (1959). J . Chem. Phya 30, 1426. Ellison, H. R., and Meyer, B. W. (1970). J. Phys. Chem. 74, 3861. Engberts, J. B. F. N., and Zuidema, G. (1970). Reo. Trav. Chim. 89, 741, 1202. Evans, M. G., and Polenyi, M. (1936). Trans. Faraday Soo. 32, 1333. Evans, M.G., and Polanyi, M. (1938). Trans. Faraday SOC. 34,22. Eyring, H. (1935). J. Chem. Phy8.3, 107. Farkas, A., and Farkas, L. (1935). Proc. Roy. SOC. A152, 124. Felder, W., Sbar, N., and Dubrin, J. (1970). Chem. Phys. Letters 6, 385. Ferguson, K. C., and Pearson, J. T. (1970). Trans. Fwaday SOC.66,910. Fischer, J. P., and Schulz, G. V. (1970). Ber. Bunsenge8. physik. Chem. 74, 1077. Fletcher, A. N. (1969). J . Phys. Chem. 73, 2217. Fletcher, A. N. (1970). J . Phys. Chem. 74, 216. Fletcher, A. N., and Heller, C. A. (1967). J . Phys. Chem. 71, 3742. Foner, S. N., and Hudson, R. L. (1956). J. Chem. Phys.25, 602. Foner, S . N., and Hudson, R. L. (1958). J. Chem. Phys. 29,442. Foldes-Berezhnykh,T., Tudbs, F., and Szak&cs,S. (1969). KineticsandMechanism of Polyreactiolza, 3, 103. Akaddmiai Kicld6, Budapest. Friedrich, S. S., Andrews, L. J., end Keefer, R. M. (1970). J. Org. Chem. 35, 944. Gardner, D. V., Howard, J. A,, and Ingold, K. U. (1964). Can. J . Chem. 42, 2847. Geib, K. H., and Harteck, P. (1931). 2. physik. Chem. (Bodenstein Volume) 849. Giles, R. D., and Whittle, E. (1966). Trans. Faraday Soo. 62, 128. Gilliom, R. D., and Howles, J. R. (1968). Can. J. Chem. 46, 2752. Gilliom, R. D., and Ward, B. F. Jr. (1965). 6.Am. Chem. SOC.87,3944. Glasstone, S., Laidler, K. J., and Eyring, H. (1941). “The Theory of R a h Processes”. McGraw-Hill, New York. Glockling, F. (1956). J. Chem. Soo. 3640. Godsay, M. P., Lohmann, D. H., and Russell, K. E. (1959). Chem. and Ind. 1603. Godsay, M. P., Harpell, G. A., and Russell, K. E. (1962). J . Polymer Sci. 57, 641. Gold, V., and Tomlinson, C. (1970). Chem. Commun. 472. Golden, D. M., and Benson, S. W. (1969). Chem. Revs. 69, 125. Goldschmidt, S., and Renn, K. (1922). Ber. 55B, 628. Goldschmidt, S., and Schmidt, W. (1922). Ber. 55B, 3197. Goldschmidt, S., and Stiegemald, C. (1924). Ann. 438, 202. Goldstein, M., Mullins, C. B., and Willis, H. A. (1970). J. Chem. SOC.(B) 321. Gragerov, I. P. (1969). Usp. Khim. 38, 1423. Gragerov, I. P., and Pogorelyi, V. K. (1969). Doklady Akud. Nauk. S.S.S.R. 185, 1062.
Gramstad, T., and Becker, E. D. (1970). J . MoZ. Structure 5, 253. Gramstad, T., and Van Binst, G. (1966). Specbrochim. Acta 22,1681. Gray, P., and Herod, A. A. (1968). Trans. Faaraday Soo. 64,1568. Gray, P., and Williams, A. (1959). Chem. Revs. 59, 239. Gray, P., Shaw, R., and Thynne, J. C. J. (1967). Progr. Reaction Kinetics 4, 63. (fray, P., Arthur, N. L., and Lloyd, A. C. (1969). Trans. Faraday SOC.65, 775. (fregg, R. A., and Mayo, F. R. (1947). Disc. Faraday SOC.2,328. Gregg, R. A., and Mayo, F. R. (1953). J . Am. Chem. SOC. 75, 3530. Greiner, N. R. (1970). J. Chem. Phys. 53, 1285. Guggenheim, E. A. (1938). Trans. FarSOC. 34, 27.
178
M . S I M O N Y I A N D F. T U D B S
Guggenheim, E. A., and Weiss, J. (1938). Tram. Faraday Soc. 34, 69. Hammond, G. S., and Nandi, U. S. (1961). J. Am. C h m . SOC.83, 1217. Hammond, G. S., Boozer, C. E., Hamilton, C. E., and Sen, J. N. (1955). J. Am. Chem. SOC.77,3238. Hanna, S. B., Jermini, C., and Zollinger, H. (1969). Tetrahedron Letters, 4415. Harcourt, A. V., and Esson, W. (1866). Philosophical Tranaactwna 156, 193. Harcourt, A. V., and Esson, W. (1867). Philoaophkl Transactions 157, 117. Harle, 0. L., and Thomas, J. R. (1967). J. Am. Chem. SOC.79, 2973. Hausser, K. H. (1959). Naturwisa. 46, 597. Hautecloque, S. (1970). J. Chim. Phya. 67, 771. Heckner, K. H., Landsberg, R., and Delchau, S. (1968). Ber. Bunsengea. physik. Chem. 72, 649. Heilman, W. J., Rembauni, A., end Szwarc, M. (1957). J. Chem. Soe. 1127. Herkes, F. E., Friedmann, J.,and Bartlett, P. D. (1969). Intern. J.Chem. Kinetics 1, 193. Hirata, T., Sujishi, S., and Gunning, H. E. (1960). J. Am. Chem. SOC.82, 5045. Hogg, J. S., Lohmann, D. H., and Russell, K. E. (1961). Can. J. Chem. 39, 1588. Homer, J., Hartland, E. J., and Jackson, C. J. (1970). J. Chem. SOC.(A) 931. Horner, L., and Schwenk, E. (1949). Angew. Chem. 61,411. Horner, L.. and Schwenk, E. (1950). Ann. Chem. (Liebig’s)566, 69. House, J. E. Jr., and Cook, R. L. (1969). Trana. Illinois State Acad. Sci. 62, 154. Howard, J. A., and Ingold, K. U. (1962). Can. J. Chem. 40, 1851. Howard, J. A., and Ingold, K. U. (19634. Can. J. Chem. 41, 1744. Howard, J. A., and Ingold, K. U. (1963b). Can. J. Chem. 41,2800. Howard, J. A., and Ingold, K. U. (1964a). Can. J. Chem. 42,1044. Howard, J. A., andIngold, K. U. (196413). Can. J. Chem. 42,1250. Howard, J. A., and Ingold, K. U. (1964~).Can. J. Chem. 42, 2324. Howard, J. A., and Ingold, K. U. (1965a). Can. J. Chem. 43,2724. Howard, J. A., and Ingold, K. U. (1965b). Can. J. Chem. 43,2729. Howard, J. A., and Ingold, K. U. (1966~).Can. J. Chem. 43, 2737. Howard, J. A., and Ingold, K. U. (1970). Can. J. Chem. 48, 873. Howard, J. A., Schwalm, W. J., and Ingold, K. U. (1968). Adv. Chem. Ser. 75, 6. Huggins, M. L. (1970). J. Phya. Chem. 74, 371. Hulett, J. R. (1964). Quart. Revs. 18, 227. Huong, P. V., and Lassegues, J. C. (1970). Spectrochim. Acta 26A, 269. Ingold, C. K. (1953). “Structure and Mechanism in Organic Chemistry”. Bell, London. Ingold, K. U. (1960a). J. Phys. Chem. 64, 1636. Ingold, K. U. (1960b). Can. J. Chem. 38, 1092. Ingold, K. U. (1962). Can. J. Chem. 40, 111. Ingold, K. U. (1963a). Can. J. Chem. 41, 2807. Ingold, K. U. (196313). Can. J. Chem. 41, 2816. Ingold, K. U. (1967). Private communication. Ingold, K. U. (1968). Lecture presented a t the Symposium on Radical Reactions, Santa Monica. Ingold, K. U., andHoward, J. A. (1962). Nature 195, 280. Ingold, K. U., and Taylor, D. R. (1961a). Can. J. Chem. 39,471. Ingold, K. U., and Taylor, D. R. (1961b). Can. J. Chem. 39,481. James, D. C . L., and Troughton, G. E. (1966). Truns. F a r d a y SOC.62,145. James, R . E., and Sicilio, F. (1970). J. Phys. Chem. 74, 1166. Jenkins, A. D. (1967). Adv. Free Radioat Chern. 2, 139.
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
179
Jenkins, A. D. (1969). Main Lecture presented at the IUPAC Symposium on Macromolecular Chemistry, Budapest. To be published in “Kinetics and Mechanism of Polyreactions ”, Vol. 6. AkadBmiai Kiadb, Budapest, (1971). Jentschura, U., and Lippert, E. (1970). Ber. Bunsenges. phyeik. Chem. 74, 7. Johnston, K. M., Williams, G. H., and Williams, H. J. (1966). J . Chem. SOC.(B) 1114.
Jones, J. R. (1965). Diec. Faraday SOC.39, 58. Jones, J. R. (1969a). Trans. Faraday SOC.65,2138. Jones, J. R. (1969b). Trane. Faraday Soc. 65, 2430. Jones, J. R., Marks, R. E., and Subba Rao, S. C. (1967). Trans. Faraday SOC.63, 111.
Kagarise, R. E., and Whetsel, K. B. (1962). Spectrochim. Acta 18, 341. Kagiya, T., Sumida, Y., and Inoue, T. (1961)). Bull. Chem. SOC. Japan 42,2422. Kalashnikova, L. A., Buchachenko, A. L., Neiman, M. B., and Rosantsev, E. G. (1969). Zhur. Fiz. Khim. 43, 64. Kalatzis, E., and Williams, G. H. (1966). J . Chem. SOC.(B) 1113. Kauzmann, W. (1957). “Quantum Chemistry”, p. 539. Academic Press, New York. Kende, I., Tiid&, F., and Siimegi, L. (1967). Actu Chim. Acad. Sci. Hung. 54, 315. Kennedy, B. R., and Ingold, K. U. (1966). Can. J . Chem. 44, 2381. Kennedy, B. R., andIngold, K. U. (1967). Can. J . Chem. 45, 2632. Kerr, J. A., Stephens, A., and Young, J. C. (1969). Intern. J . Chem. Kinetics 1, 339, 371.
Kice, J. L. (1954). J . A m . Chem. SOC.76, 6274. Kolbe, A. (1969). Tetrahedron Letters, 1049. Kolbe, A., and Pracejus, H. (1966). Ber. Bunaenges. physik. Chem. 70,883. Kondo, Y., Matsui, T., and Tokura, N. (1969). Bull. Chem. SOC. Japan 42, 1037. Kondratiev, V. N. (1970). “Konstanty Skorosti Gazofaznykh Reaktstii”. Izd. Nauka, Moscow. Kooyman, E., van Helden, R., and Bickel, A. (1953). VerhadZ. KoninkZ. Nederland Akad. Wetenshap. 56, 75. Kreilick, R. W., and Weissman, S. I. (1962). J . A m . Chem. SOC.84, 306. Kresge, A. J., Sagatys, D. S., and Chen, H. L. (1968). J . A m . Chem. SOC. 90,4174. Kuts, V. S., Pokhodenko, V. D., and Brodskii, A. I. (1968). Doklady Akad. Nauk S.S.S.R. 180, 1109. Lamotte, M., Gerhold, G. A., and Joussot-Dubien, J. (1970). J . Chim. Phys. 67, 2006.
Laurence, C., and Wojtkowiak, B. (1970). Ann. Chim. aerie [14] 5, 163, 177. Leary, G. (1968). Can. J . Chem. 46, 1831. Leffler, J. E. (1955). J . Org. Chem. 20, 1202. Leffler, J. E., and Grunwald, E. (1963). “Rates and Equilibria of Organic Reactions”, Chapter 9. Wiley, New York. Le Roy, D. J., Ridley, B. A., and Quickert, K. A. (1967). Diec. Faraduy SOC.44, 92.
Le Roy, L. R. (1969). J . Phys. Chem. 73,4338. Lewis, E. S., rtndFunderburk, L. H. (1967). J . A m . Chem. SOC.89, 2322. Lewis, E. S., and Robinson, J. K. (1968). J . Am. Chem. SOC.90, 4337. Leyland, L. M., Majer, J. R., and Robb, J. C. (1970). Trane. Faraday SOC.66,901. Limbach, H. H., Strohbusch, F., and Zirnmermann, H. (1970). Ber. Bunsenges. physdk. Chem. 74, 3. Lloyd, W. G., and Lange, C, E. (1964). J . A m . Chem. SOC.86, 1491.
180
M . SIMONYI A N D F .
TUDBS
Longridge, J. L., and Long, F. A. (1967). J. Am. C h m . SOC.89, 1292. Lowe, J. P. (1968). Progr. Phya. Org. Chem. 6,1. Lum, K.K., and Smith, G. G. (1969). J. Org. Chem. 34, 2095. Mahoney, L. R., and DaRooge, M. A. (1970). J. Am. Chem. SOC.92,4063. Mahoney, L. R., Ferris, F. C., and DaRooge, M. A. (1969).J. Am. C h m . SOC.91, 3883. Majer, J. R., Naman, S.-A. M. A., and Robb, J. C. (1969).Trans. Faraday SOC. 65,3295. Malinowski, E. R., and Weiner, P. H. (1970). J. Am. Chem.SOC.92,4193. Mayer, I., Simonyi, M. (1971). Magy. Kkm. FoZydirat 77, 48. Mayo, F. R. (1943). J. Am. Chem. SOC.65,2324. Mayo, F. R. (1953). Dbc. Faraday Soc. 14,250. Mayo, F. R. (1967). J. Am. Chem. Soc. 89,2655. McDaniel, D. H., and Brown, H. C. (1958). J. Org. Chem. 23, 420. McGowan, J. C., and Powell, T. (1962). J. Appl. Chem. 12, 1. McGowan, J. C., Powell, T., and Raw, R. (1959). J. Chem. SOC.3103. Melander, L. (1960). “Isotope Effects on Reaotion Rates.” Ronald, New York. Melander, L. (1961). Arkiv Kemi 18, 196. Melville, H.W.(1937). Proc. Roy. Soc. A163, 511. Menzinger, M., and Wolfgang, R. L. (1969). Angew. Chem. 81,446. Merrill, J. R. (1961).J. Phys. Chem. 65, 2023. Migita, T. (1969). Yuki Uosey Kagaku (Chemistry of Organic Syntheses)27, 609. Russian translation can be found in Uap. Khim. 39, 1831 (1970). Minoura, Y., Yasumoto, N., and Ishii, T. (1964). Makromol. Chem. 71, 159. More O’Ferrall, R. A., and Kouba, J. (1967). J. Chem. SOC.(B) 985. Morris, E. R., and Thynne, J. C. J. (1968). Tram. Faraday SOC.64,414. Morris, E.R., and Thynne,J. C. J. (1970). Inntern. J . Chem. Kinetics 2,267. Mukherjee, S . , Palit, S. R., and De, 8.K. (1970).J . Phya Chern. 74, 1389. Murrell, J. N. (1969). Chem. Britain 5, 107. Muller, G., and Chiseler, G. (1970). 2.physik. Chem. 243,340. Nave, P. M., and Trahanovsky, W. S. (1968).J . Am. Chem. SOC.90,4755. Nave, P. M., and Trahanovsky, W. S. (1970). J . Am. Chem. SOC.92, 1120. Nelson, J. (1970).Spectrochina. Acfa 26A, 109, 235. NeszmBlyi, A.,and Imre, L. (1968). Spectrochim. Acta 24A,297. NeszmBlyi, A.,Simonyi, M., and Tudgs, F. (1967). Acfa China. A d . Sci. Hung. 53, 369. Norman, R. 0. C., and Gilbert, B. C. (1967). Adv. Phys. Org. Chem. 5, 53. Okada, Y.(1961). Kogyo Kagaku Zasshi 64,309, 1511. Parker, A.J. (1967). Adv. phys. Org. C h m . 5, 173. Parker, A. J. (1969). Chem. Revs. 69, 1. Perkins, M. J.,Ward, P., and Horsfield, A. (1970).J. Chem. SOC.(B) 395. PetrBnek, J.,and Pild, J. (1969). CollectionCzechosbvak Chem. Commun. 34, 79. Pimentel, G.C., and Mcclellan, A. L. (196Oe). “The Hydrogen Bond”, Chapters 2,3 and 4. Freeman, San Frencisco, London. Pimentel, G. C., and Mcclellan, A. L. (1960b). “The Hydrogen Bond”, p. 148. Freeman, San Francisco, London. Pokhodenko, V. D. (1969). “Fenoksilnye radikaly.” Naukova Dumka, Kiev. Porter, G. (1970). Plenary lecture presented a t the International Conference on the Mechanisms of Reactions in Solution, Canterbury, Code No. P3. Proll, P. J., and SutclBe, L. H. (1963). Trans. Faradoy Soo. 59, 2090. Pryor, W.A. (1966). “Free Radicals.” Moartbw-Hill.
H Y D R O G E N ATOM A B S T R A C T I O N R E A C T I O N
181
Pryor, W. A., Echols, J. T., and Smith, K. (1966). J. Am. Chem. SOC.88,1189. Pryor, W. A.,Tonellato, U., Fuller, D. L., and Jumonville, S. (1969).J. Org. Chem. 34,2018. Quee, M. Y., and Thynne, J. C. J. (1968). Tram. Faradag Soc. 64, 1296. Ramaswamy, K., Pickai, R., and Gnanadesian, S. G. (196-7).J . Mol. Spectroscopy 23, 416. Reece, I. H., and Werner, R. L. (1968). Spectrochim. Acta 24A, 1271. Reich, L.,and Stivala, S. S. (1969). “Autoxidation of Hydrocarbons and Polyolefins.” Dekker, New York. Ritchie, C. D., and Sager, W. F. (1904). Progr. Phys. Org. Chem. 2,323. Robb, J. C., and Shahin, M. (1969). Tram. Faraday Soc. 55, 1763. Rochester, C.H., and Rossall, B. (1969). Tram. Faruday SOC.65, 992, 1004. Rosenberg, H.M., and Serv6, P. (1970). J. Am. Chem. SOC.92,4746. Russell, G . A. (1968).J . Am. Chem. SOC.80, 4987. Ryabokobylko, Yu. S., and Chekunov, A. V. (1970). Zhur. Tewet. i Eksperim. Khim. (Kiev)6 , 660. Ryba, O., Petrdnek, J., and Pospigil, J. (1966a). Collection Czechoslovak Chem. Commun. 30,843. Ryba, O., PetrAnek, J., and Pospigil, J. (1966b). Collection Czechodovak Chem. Commun. 30,2167. Schaefer, T., and Kotowycz, G. (1968). Can. J . Chem. 46, 2866. Schaleger, L. L., and Long, F. A. (1063). Adv. Phys. Org. Chem. 1,l. Schneider,W. G.,Bernstein, H. J.,and Pople, J. A. (1968).J.Chem. Phys. 28,601. Scott, R., DePalma, D., and Vinogradov, S. (1968). J. Phy8. Chem. 72, 3192. Semenov, N. N. (1968). “0 Nekotorykh Pmblemakh Khimicheskoi Kinetiki i Reaktsionnoi Sposobnosti”, Part 111, Chapter 4, Section 6. Izd. Akad. Nauk S.S.S.R. Moscow. Servanton, M., Biais, J., and Lemanceau, B. (1970).J. Chirn. Phy8.67,800. Shannon, T. W., and Harrison, A. G. (1963).Can. J . Chem. 41,2466. Shaw, R.,and Thynne, J. C. J. (1966). Trans. Faraduy Soo. 62, 104. Shelton, J. R., and McDonel, E. T. (1968). J . Polymer Sci. 32, 75. Shelton, J. R., and Vincent, D. N. (1963). J. Am. Chem. SOC.85, 2433. Shelton, J. R.,McDonel, E. T., and Crano, J. C. (1960). J. Polymer Sci. 42, 289. Shorter, J. (1969). Chem. Britain 5, 269. Shorter, J. (1970).Quart. Revs. 24,433. Simonyi, M. (1967). Thesis, Budapest. Simonyi, M. (1909a). Lecture presented at the Symposium on Isotope Effects, York. Simonyi, M. (1909b). K h . K6zlemdnyek 32,216. Simonyi,M., and Tudgs, F. (1969). “Kinetics and Mechanism of Polyreactions”, Vol. 3, p. 119. Akad6miai Kiad6, Budapest. Simonyi,M., TiidBs, F., and Heidt, J. (19674.Acta Chim. Acad. Sci. Hung. 53,43. Simonyi, M., Tiid&, F., and Pospigil, J. (1967b). Europ. Polymer J . 3, 101. Simonyi, M.,Tud&, F., Holly, S., and Pospifd, J. (19670).Europ. Polymer J . 3, 669.
Simonyi, M., TudBs, F., and KOVL~CS, L. (1969a). “Kinetics and Mechanism of Polyresctions”, Vol. 3, p. 116. Akademiai Kiad6, Budapest. Simonyi, M., TudBs, F., Pospfil, J., and Holly, S. (1969b). “Kinetics and Mechanism of Polyreactions”, Vol. 3,p. 126. Akad6miai Kiad6, Budapest. Singh, S., and Rao, C. N. R. (1966~).J . Am. Chem.SOC.88,2142. Singh, S . , and Rao, C. N. R. (1966b). Can. J . Chem. 44, 2611.
182
M . S I M O N Y I A N D F. TUDOYS
Singh, S., Bhaskar, K. R.,and Rao, C. N. R. (1966). Can. J . Chem. 44, 2657. Sinitsyna, Z. A., and Bagdasaryan, Kh. S. (1958). Zhur. Fiz. Khim. 32, 2663. Sinitsyna, Z. A., and Bagdasaryan, Kh. S. (1960). Zhur. Fiz. Khim. 34, 1110. Smith, G. G., Lum, K. K., Kirby, J. A., and Pospiiil, J. (1969). J. Org. Chem. 34, 2090. Socrates, G. (1970). Trans. Faraday SOC.66, 1052. Stefani, A. P., Herk, L., and Szwarc, M. (1961). J. Am. Chem. SOC.83, 4732. Stepukhovich, A. D., Ulitskii, V. A,, and Sharaevskii, A. P. (1968). Zhur. Fiz. Khim. 42, 1276. Stewart, R., and Lee, D. F. (1964). Can. J. Chem. 42,439. Su, Y . S., and Hong, H.-K. (1968). Spectrochim. Acta 24A, 1461. h e h a , L., Urner, Z . , and Suchhek, M. (1970). Collection Czechodovak Chem. Commun. 35,3651. Sukhanova, 0. P., and Buchachenko, A. L. (1965). Zhur. Fiz. Khim. 39,2413. Szabb, Z. G. (1962). Chem. SOC.(London)Spec. Publ. 16,113. Taft, R. W. Jr. (1956). In Newman, M. S. “Steric Effects in Organic Chemistry”, p. 653. W h y , New York. Taft, R. W., Price, E., Fox, I. R., Lewis, I. C., Andersen, K. K., and Davis, a. T. (1963). J . Am. Chem. SOC.85, 709. Tedder, J. M. (1960). Quart. Revs. 14, 340. Tedder, J. M., and Walton, J. C. (1967). Progr. Reaction Kinetim, 4, 37. Thomas, J. R. (1964). J . Am. Chem. SOC.86, 4807. Thomas, J. R., and Tolman, C. A. (1962a). J . Am. Chem. SOC.84, 2079. Thomas, J. R., and Tolman, C. A. (196213). J . A m . Chem. SOC.84,2930. Thomas, R.K., and Thompson, H. (1970). Proc. Roy. SOC.316, 303. Thorn, R. J. (1969). J . Chem. Phys. 51, 3582. Tobolsky, A. V., and Mesrobian, R. B. (1954). “Organic Peroxides.” Wiley (Interscience), New York. Tolman, R. C. (1920). J . Am. Chem. SOC.42, 2506. Tsepalov, V. F., and Shlyapintokh, V. Ya. (1962). Kinetika i Katdiz 3, 870. Tiidas, F. (1964a). Makromol. Chem. 79, 8 . Tiidbs, F. (196413). Thesis, Budapest-Leningrad. Tudbs, F. (196th). Acta Chim. Acad. S c i . Hung. 43, 397. Tiidbs, F. (196513). Acta Chim. Aead. Sci. Hung. 44,403. Tudbs, F. (1968). J . Polymer Sci. 16, 3461. Tiidas, F. (1969). Main lecture presented a t the IUPAC Symposium on Macromolecular Chemistry, Budapest. To be published in “ Kinetics and Mechanism of Polyreactions”, Vol. 6, (1971). Tiidbs, F., arid Berezhnykh-Foldes, T. (1966). Europ. Polymer J . 2, 229. Tiidas, F., arid Simhndi, T. L. (1962). Vysokomol. S o d . 4, 1425. Tiidbs, F., Kende, I., and Azori, M. (1962). VysokomoE.Soed. 4, 1262. Tiidas, F., Kende, I., and Azori, M. (1963). J . Polymer Sci. A l , 1353, 1369. Tiidas, F., Kende, I., Berezhnykh-Foldes, T., Solodovnikov, S., and Voevodski, V. V. (1965). Kinetika i Kataliz 6, 203. Tiidas, F., Berezhnykh-Foldes, T., and Simonyi, M. (1967). Vysokomol. Soed. 9, 2284. Umanskii, V. M., Stepukhovich, A. D. (1969). Zhur. Fiz. Khim. 43, 2490. Uneyama, K., Namba, H., and Oae, S. (1968). Bull. Chem. SOC.Japan. 41, 1928. Vanderborgh, N. E., Armstrong, N. R., and Spall, W. D. (1970). J. Phys. Chem. 74, 1734.
H Y D R O U E N ATOM A B S T R A C T I O N R E A C T I O N
183
ven Helden, R.,and Kooyman, E. (1964). Reo. Trav. China. 73, 269. Vars&nyi,G. (1970). Acta China. A d . Sci. Hung. 65, 125. Venkataaubramenian, N., and Thiagarajan, V. (1969). Can. J. Chem. 47, 694. Vetchinkina, V. N., Maizus, Z. K., and Emanuel, N. M. (1966). Z h w . Fiz. Khim. 60, 762. Vogel, G. C., and Drago, R. S. (1970). J. Am. Chem. Soc.92,8347. Wagner, P. J., and Hammond, G. S. (1968). Adv. Photochem. 5,99. Wagner, P. J., Kemppainen, A. E., and Schott, H. N. (1970). J. Am. Chem. Xoc. 92, 6280. Walling, C. (1967). “Free Radicals in Solution.” Wiley, New York. Walling, C., and Briggs, E. R. (1946). J. Am. Chem. Soo. 68, 1141. Walling, C., and Wagner, P. J. (1964). J. Am. Chern. SOC. 86, 3368. Welling, C., Briggs, E. R., Wolfstirn, K. B., and Mayo, F. R. (1948). J . Am. Chem. Soo. 70, 1637. Waaylishen, R., Schaefer, T., and Schwenk, R. (1970). Can. J. Chern. 48,2886. Weiner, S . A., and Hammond, G. 5. (1969). J. Am. Chem. SOC. 91, 986. Wwtheimer, F. H. (1961). Chem. Revs. 61, 266. Whetad, K. B., and Kagarise, R. E. (1962a). Spectrochim. Acta 18, 315. Whetsel, K. B., and Kagarise, R. E. (1962b). Spectrochim. Acta 18, 329. Whitternore, I. M., Stefani, A. P., and Szwarc,M. (1962). J. Am. Chem. SOC.84, 3799. Wiberg, K. B. (1966). Chem. Revs. 55, 713. Widom, J.M., Phillippe, R. J.,and Hobbs, M. E. (1967).J.Am. Chem.Soc. 79,1383. Yates, W. R., and Ihrig, J. L. (1966). J. Am. Chem. Soc. 87,710. Zsikov, G. E., Maims, Z. K., and Emanuel, N. M. (1967). DokZady Akad. Nauk S.S.S.R.173, 869. Zaikov, G . E., Lezina, V. P., and Maims, Z. K. (1968). Zhur. 3%. K h h . 42,1273. Zeiteev, B. A,, and Shtraichman, G. A. (1968). Vysokomo~. Soed. 10A, 438. Zimmermann, H. (1969). C h i m b 23, 363.
7
This Page Intentionally Left Blank
VINYL CATIONS GIORGIO MODENA AND UMBERTO TONELLATO
Centro C.N.R. “Meccanismi di Reazioni Organiche” Istitdo di Chimica Organica, Universita’ di Padova 35100, Padova, Italy
.
. .
.
.
.
.
. . .
.
. .
. .
. . .
.
. . . .
.
.
.
.
. . .. ..
. .
. . .
. . .
. . .
I. Introduction 11. SourcesofVinylCations A. Electrophilic Addition to Acetylene Derivatives B. Electrophilic Addition to Allene Derivatives . C. Heterolytic Fiseion of Bonds Attached to a Vinyl Carbon Atom D. Electron Removal from Neutral Species 111. General Properties of Vinyl Cations A. Geometry and Stability B. Reactions of Vinyl Cations . IV. Related Species A. PropmgylCations B. NitriliumIons . c. ImminillmIons D. AcylCetionn References.
.
.
.
.
.
.
.
. .. .
.
.
. .
.
.
.
. .
.
.
. .
. . . . . . . .
. . .
186 186 187 216 231 263 264 264 266 267 267 270 272 273 274
I. INTRODUCTION VINYL cations 1 are derivatives of the ethenyl cation, H,C:CH+, in which the hydrogen atoms may be replaced by other organic residues. The terms vinylium ion and, less correctly (Bethell and Gold, 1967), vinyl carbonium ion are also often used to indioate (1). The nomen-
>k6(1)
clature practice is based both on the “vinyl cation ’’ system considering the ethenyl cation as the parent “type” of the class and the “-enyl cation ” system based on the-International Pure and Applied Chemistry convention. The acceptance of vinyl cations as reaction intermediates is very recent and this is certainly surprising if one considers the enormous amount of work on saturated carbonium ions published since the beginning of the present century (Baeyer, 1906; Hantzsch, 1921). Vinyl cations have for a long time been thought to be exceedingly unstable and their formation was assumed to be unlikelv. For instance. although it was generally agreed that electrophilic addikons to alkynes 186
186
QIORQIOMODENA A N D UMBERTO TONELLATO
resemble closely those to alkenes (Dewar, 1949; Ingold, 1953), vinyl cations were not explicitly recognized as actual intermediates in these processes. Such a reluctance lasted up to the early sixties and was supported by the fact that vinyl derivatives were reported to be unreactive even under forcing solvolytic conditions (Shriner, et al., 1964). I n general, S,l-type reactions of vinyl compounds were “hardly expected to occur” (Rappoport and Patai, 1964). Relevant mechanistic studies of electrophilic additions to acetylene derivatives and the first demonstration of a unimolecular solvolysis of vinyl halides (Grob and Cseh, 1964) stimulated a still growing interest in the field of vinyl cations. Vinyl cations are more stable “than probably is realized by most chemists” (Richeyand Richey, 1970) but a direct proof of their existence is still awaited. However, strong indirect evidence of their formation has accumulated and from the published material, although still largely fragmentary, a reasonable picture of these intermediates is now being defined. The scope of this review is a detailed survey of reactions proceeding through vinyl cations and an attempt of a systematic definition of the properties of these intermediates with reference to those of saturated carbonium ions. Although attention will be particularly devoted to linear cations, bridged unsaturated species will be considered as alternative structures of vinyl cations rather than as a distinct type of reactive intermediates. The “?r-complex” terminology (Dewar, 1949) widely abused in the past decades to indicate especially cyclic cations and recently reassessed by Banthorpe (1970) will be generally avoided. The most recent studies not covered by published reviews on the subject (Rappoport, 1969; Richey and Richey, 1970; Richey, 1970; Hanack, 1970) are discussed in greater detail than others and data are collected in pertinent Tables.
11. SOURCES OF VINYLCATIONS A growing body of evidence implicates vinylic cations as intermediates in a variety of reactions which may be roughly organized as follows : (a) electrophilic addition to acetylene derivatives, including triplebond participation ; (b) electrophilic addition to allene derivatives, including cumulate double-bond participation ; (c) heterolytic fission of vinyl derivatives ; (d) electron removal from neutral molecules.
187
VINYL CATIONS
Attention will be particularly devoted to recent mechanistic studies which clearly indicate the intermediacy of vinyl cations. No attempt will be made critically to review earlier studies of reactions, which, in the light of recent knowledge, may quite reasonably involve these ions as intermediates. A. Electrophilic Addition to Acetylene Derivatives Vinyl cations can be generated by addition to a carbon-carbon triple bond of a variety of positively charged electrophiles, (equation 1). E\ c,=C&
Ef+-c,&
4.
uproducts
(1)
The most commonly encountered electrophile is the proton, but other species are frequently involved in such additions. Alternative mechanistic routes for electrophilic additions to acetylene derivatives are also possible. According to the principle of microscopic reversibility , the same spectrum of different mechanisms is expected for addition to the triple bond as for elimination from a double bond. The relevant studies of electrophilic addition will be examined according to the nature of the electrophile involved. 1. Proton Addition
(a) Acid-catalyzed Hydration. Kinetic studies have been reported for the acid catalyzed hydration of arylacetylene derivatives (2) (equation 2) and of alkoxy- and thioalkoxyacetylene derivatives (3) (equation 3). Hf
+
RCkC.CaH4.9 +R.HC=C-CeH4X
Ha0 4
[R-HC-C(OH)*CeH4.X]
The specific substrates and reaction conditions are indicated in Table 1 and will be referred to by the numbers in the first column. Hydration of adamantylacetylene (6 of Table 1) has not been investigated from a kinetic point of view and the intermediacy of the adamantylvinyl
188
G I O R G I O MODENA A N D U M B E R T O T O N E L L A T O
cation has been inferred from the product distribution. This c a e will be discussed below. The cleavage of the phenylacetylenegermanyl derivative (5 of Table 1) is not strictly a hydration reaction following equation (2) but still proceeds via rate-determining protonation of the triple bond through a vinyl cation. The kinetic law is generally of the type :rate =k2[acetylene][H+]. I n concentrated acid solutions, the plot of log koba.v8 H , in the case of compounds 2 is linear with essentially unit slope (Noyce et al., 1965, 1967; Noyce and De Bruin, 1968). The study of the reaction in a series of buffers showed that it is subject to general acid catalysis (Noyce and Schiavelli, 1968a; Stamhuis and Drenth, 1961) and the application of the Bunnett, Grunwald and similar treatments in the case of thioalkoxyacetylene derivatives (Hogeveen and Drenth, 1963b) clearly indicate that the addition of a water molecule does not take place in the slow step of the reaction. The kinetic solvent isotope effects, kHaO/kD,O,range between 2 and 4 in the case of compounds 2 (Noyce et al., 1965, 1967; Noyce and Schiavelli 1968b; Noyce and De Bruin, 1968) and between 1.7 and 2.0 in the case of compounds 3 (Stamhuis and Drenth, 1963a; Drenth and Hogeveen, 1960). The observed values accord with the idea of a ratelimiting proton transfer to the triple bond : a small inverse isotope effect would have been observed in the case of specific acid catalysis (Noyce et al., 1967). The substituent effects on the reaction rate strongly support the hypothesis of the formation of a vinyl cation. The p values (see Table 1) in the cme of arylacetylene derivatives 2 are all negative and quite large. The use of u+ constants usually gives the best fit of rate data, with the Hammett relationship and indicates that in the transition state of the reaction the positive charge is on the carbon atom next to the aromatic ring and that it is largely shared by the aromatic 7~ system. The investigation of an unusually large number of substituents allowed Bott et al., (1964, 1965) to conclude that the best fit of data for the hydration of phenylacetylenes and for the cleavage of acetylenegermany1 derivatives (1 and 5 of Table 1) is obtained by the use of the Yukawa-Tsuno (1959, a, b) relationship, log krel= p[u + R (u+ - u)]with R = 0.81 and R = 0.46 in the two cmes. The above authors suggest that the R value of less than unity may indicate that, although gain of resonance stabilization is achieved by overlap of the vacant p , orbital with the aromatic ring, there is some loss of resonance stabilization on going from the initial to the transition state where the conjugation between the ring and the carbon-carbon multiple bond is not possible. I n the case of compounds 3, the intermediate vinyl cation is stabilized
TABLE 1 Acid Catalyzed Hydration of Acetylene Derivatives ~-
No.
6
Substrate
HWAd
Reaction Medium
t, “C
P5
Main Products
-4.3
HzS04aq.-HOAcl: 2 v/v 648%
50.2 25
MeCo. c6H4x
HZ804 48-57% HzSOg 62-70%
25 25
&So4 60-80% MeOH-HC104 5 :2 V/V
25 29
MeCHz. CO. C&t5 (HOzC.CHz .CO . C & X ) COZ MeCO C&X Arc0 CHz CO .C a 4 X EtsCe .OH + HCkC C a 4 X
hexane-90%
HzS04
5
+ .
.
- 3.84 -4.77
.
-4.2
-3.3b
.
References Bott et d.,1965 Noyce et al., 1965; Noyce and Schiavelli, 1968a,b Noyce et al., 1965 Noyce el al., 1967 NoyceandDeBruin, 1968 Bott et al.. 1964
rie
Me *cO*Ad+
F
* Q
EtOH 42*7%-H20 7a H W . O R ’ (R’=Et, n-Bu, Ph) 7b H W . O R ’ HzO (buffem) (R‘=Me, Et, n-Bu, t -Bu) 8 9 10 11
25
MeCO .OR’
25
MeCO OR’
Jacobs and Searles, 1944
% Drenth and Hogeveen, 1960 rn Stamhuis and Drenth, 1961, 1963a, b 2*57*CStmnhuis and Drenth, 1963 Hekkert and Drenth, 1963 Stamhuisand Drenth, l963a - 0.73*C DrenthandHogeveen, 1960; Hogeveen and Drenth, 1963a, b 3*76*cHogeveen and Drenth, 1963~1,b
.
-6.2*c
.
-
RkC.OEt (R=Me, Et, n-Bu, t-Bu) M e w . 0.C H d H z H M . S R (R’=Me, Et, n-Bu, i.Pr)
HzO (buffers)
25
RCHz CO .OEt
Ha0 (buffers) HzO (buffers)
25 25
M-2 .CO. 0.CH-2 Me. CO .S R
RM.SEt
HzO (buffers)
25
R WaCO. SEt
.
~
Hammett’s reaction constants using u+ values, unleas otherwise indicated. b By use of the Yukawa and Tsuno equation (see text). C p* values by use of Taft’s equation (see text).
5
2 o
~~~
w
m
(0
190
Q I O R G I O M O D E N A AND UMBERTO TONELLATO
through allenic structures 4 where the positive charge is on the heteroatom. The effect of alkyl substituents is well correlated by the Taft relationship, log krkl=p*a* +C, and the p* values are shown in Table 1. The p* value is exceptionally large ( - 6-2)for R’ substituents in alkoxyacetylene derivatives (7b of Table 1). This is also confirmed by the observation of a rate factor of 7 x los between ethoxy- and phenoxyacetylene reported by earlier authors (7a of Table 1). The effect of R‘ substituents in the case of thioalkoxyacetylenes (10 of Table 1) is instead much smaller and indicates that resonance structures 4 are much more important when Y is oxygen than when it is sulphur. Accordingly, the rate of hydration of ethoxyacetylene is ca. 3000 times that of ethylthioacetylene. The generally good correlation of rate data for alkyl substituents of Werent steric requirements with the simple form of the Taft relationship shows that the steric influence of both R and R of 3 is almost negligible. This is expected if the rate-limiting step is a proton transfer to the triple bond and does not involve simultaneous addition of a water molecule.
When R = H, the rate of hydration of compounds of type 2 and type 3 is much greater than expected. Phenylacetylene reacts ca. 30 times faster than 1-phenylpropyne in identical conditions (Noyce et al., 1965) and the rate for ethylthioacetylene (and ethoxyacetylene) is some 2000 times faster than calculated on the basis of the Taft relationship on the series of compounds shown under 10 (and 7b) of Table 1. The large rate of hydration of monosubstituted acetylene derivatives is not fully understood. Alhrnative explanations have been offered (Noyoe et al., 1966; Noyce and Schiavelli, 1968b; Hogeveen and Drenth, 1963): (i) Hyperconjugative stabilization of the initial state, important when R of (2) and (3) is methyl and not hydrogen. Indeed, a methyl group stabilizes an alkyne by ca. 5 kcal mol-l more than an alkene. Such stabilization is lost in the transition state when the methyl-substituted carbon becomes sp2 hybridized. (ii) Hyperconjugation in the transition state, which may be important when R = H. The C-H bond is, in fact, very favourably oriented for hyperconjugative overlap of the developing empty p n orbital (Figure 1). (iii) Hydrogen bonding with the solvent, stronger when R = H than Me, which may invert the usual trend of electron-donating ability H < Me. Other examples of enhanced electrondonating properties of acidic hydrogens are known. (iv) Stabilization
191
VINYL UATIONS
FIU.1
of the transition state when R = H by non-hindered solvation. The arguments under (ii) and (iii) are supported by the fact that the kinetic deuterium isotope effect, kHCrcPh/kDC~CPh = 1.1 (Noyce et al., 1965; Noyce and Schiavelli 1968b) and k H h C s E t / k D b c s E t = 1.03, (Hogeveen and Drenth, 19638) are substantially higher than calculated (0.77 and 0.83 respectively). The values of activation entropy are negative and small ( - 1 to - 6 cal mol-lK-l) in the case of compounds (3). Higher negative values ( - 18 to - 28 cal mol-lK-l) have been estimated for the more slowly reacting phenylacetylene derivatives. In spite of the large spectrum of AS' values observed, it has been argued that they are all consistent with an A-SE2 mechanism (Matesich, 1967). It is remarkable that the mechanistic parameters, rates included, for the acid-catalyzed hydration of compounds (2) and (3), are in general, similar to those observed in the reaction of the corresponding alkenes. This has been illustrated in detail by Richey and Richey (1970) and will be discussed in section IIIA2a. The hydration of triple bonds may be faster than that of double bonds, as shown by the main product obtained from the alkynyl alkenyl ether under 9 of Table 1. The reaction products, shown in Table 1, are mainly those expected to be formed by interaction of vinyl cations with water, reasonably via enol derivatives as indicated in equations (2) and (3). Acetic acid and olefins are also formed from compounds (3) (Y = 0; R' =i-Pr, t-Bu) besides the expected carboxylic esters. In these cases, splitting of the vinyl cation formed initially to keten and a carbonium ion following equation (4) (R'=t-Bu) takes place to approximately 12% when R' =i-Pr and 60% when R' = t-Bu (Stamhuis and Drenth, 1963b). H2C&-0
+
H&=C-O--C-Me
Ha0
Me:CO.OH
I I Me
(4)
Me
I
+&Me
I
Me
a
Me'
192
G I O R U I O MODENA A N D UMBERTO TONELLATO
The possibility that the vinyl cation may lose a proton other than collapse with the solvent has been investigated by Bott et al. (1965)and Noyce and Schiavelli (1968a). Whereas the former authors reported no detectable loss of tritium from PhCECT under the conditions of 1 of Table 1, the latter ones were able to show that 0.6% exchange occurs in the unreacted material when phenylacetylene is allowed to react up to 50% completion in 40% sulphuric acid-t. Since base- and water-catalyzed exchange under the conditions used can be excluded, the outcome of this experiment would kdicate that the protonation of a triple bond may, at lea& in some conditions, be reversible. (b) Addition of Carboxylic Acids. The addition of trifluoroacetic acid to alkynes (hexynes and pentynes) has been investigated by Peterson and Duddey (1963, 1966) and Peterson and Bopp (1967) and the intermediacy of vinyl cations generated by protonation of the triple bond has been convincingly suggested. The arguments offered are mainly based on the close similarity of the kinetic behaviour with that of the addition of trifluoroacetic acid to the corresponding alkenes for which the reaction mechanism was shown to involve carbonium ions intermediates (Peterson and Allen, 1963a, b ; Peterson et al. 1965). The rate of addition at 60" to (5-substituted)-1-pentyne derivatives is retarded by electron withdrawing substituents such as halogens and methoxy, acetoxy, and cyano groups. Anchimeric assistance effects by some of these groups will be discussed below. A scrutiny of kinetic data shows that 1-hexyne reacts only 1.1 times more slowly than 2-hexyne and faster = 2.2. I n this than expected on the basis of the ratio k3-hexyne/k2-haxyne case also, the intriguing terminal hydrogen us. methyl effect discussed in the preceding section is apparently involved, perhaps more attenuated than in the hydration cases previously examined. The product distribution from simple alkynes (Scheme 1)is consistent with the hypothesis that a discrete vinyl cation is involved. I n particular, the formation of both cis and trans adducts, under kinetic control, is significant since it indicates that a free linear vinyl cation is the actual intermediate. The formation of hexaethylbenzene is another point of interest. Its yield increases with increasing concentrations of 3-hexyne, thus supporting the hypothesis that the trimerization process involves in its early stages interaction between the vinyl cation and the alkyne. Other examples of polymeric products formed via vinyl cations, probably through a concerted electrocyclic process, have been reported (see infra) although generally dimers are the ultimate products. Addition of acetic acid to the 2-chlorobutyn-3-oic ester (5) has been suggested to occur via a vinyl action intermediate (Verny and Vessiere, 1969) to give the primary 1,2-adduct ( 6 ) together with cis ester (7).
193
V I N Y L OATIONS SOHEME 1
-J
/--=
CFaCOtH
H 1
1 98-34%
2-16%
The latter product is suggested to arise from the primary acetate ester via an internal proton transfer as indicated in equation ( 5 ) .
I
An example of intramolecular addition of a carboxylic function to the triple bond of 2,2'-tolandicarboxylic acid (8) has been reported by Letsinger et al. (1965). A mechanism is proposed in which an ortho carboxylic group acts as a proton donor to the triple bond thus favouring the attack of the second carboxylate group to give the product of formal 1,2 addition (9). Although the intermediacy of a vinyl c&tioncannot be excluded, the process quite probably involves a concerted mechanism through a transition state of type (10).
TAB= 2 Addition of Hydrogen Halides (HX)to Acetylene Derivative8
Substrate
Is H
M Me
lb
Hx Solvent
Me(X)CLC€&
MeCX#e
MeCH=cHx
HCI
-
-70
66%
44%
-
HBr
-
-70 -70
17-24% 36%
24-34%
30%
10
HI
2 NC.CrC.NEta
HCI
CHzCIa
-5
0
References
Products
t°C
66%
-
ma Me
z s z
Me Me
X
0 0
3 HC=C.C(Me)3
HCl
HCI
ratio0 1:l 5: 1
HOAc
.
.
1O:l
4 Me.C=-C,C&Hs
IvIe2CCl.CC1 Mea (rearranged) ClCHz C(Me).CCl Me2 31% 52% 64%
HzC=CCI. CMe3 (normal) Me. CClz CMe3 69% 48% 36%
25
5%
70%
~n
H 5 EtkCEt
HCI
HOAc
HCI HzO HBr HzO HI HzO HCI HzO
25
25 26 25
H
60% X
8%
'c4'
HA H ' 83-75% 82-98% 98%
ratio HCI :t-butylacetylene.
2#
+4 ?l
4040%
R
COaH
&C' X H ' 17-25% 2-28%
Fahey and Lee, 1967
z
0
H
u)
Bowden and Price, 1969
< 2%
80-100
0
Fahey and Lee, 1966
cl
< 1% COzH
ariesbaum and Rehman, 1970
Borkent and Drenth, 1970
196
O I O R Q I O MODENA A N D UMBERTO TONELLATO
(c) Additibn of Hydrogen Hulides. This common type of reaction has received very little attention with regard to the possible intermediacy of vinyl cations. Addition of various hydrogen halides to propyne has been investigated by Griesbaum et al. (1965) (la+ of Table 2). The product distribution parallels that obtained from allene under identical conditions and the most likely mechanism involves formation of a propenyl cation which may react with halide ions to give the expected addition products or cyclodimerize to 1,3-dihalocyclobutane (see discussion in section 11, Blb). The great-reactivity of ynamines in electrophilic additions is well documented (Viehe, 1969) and ketenimmonium ions, electromers of -I-
+
vinyl cations (R-CH : C : NR,++RCH : C-NR,), have been suggested as intermediates in several cases. Addition of hydrogen chloride to diethylaminocyanoacetylene results in a surprisingly high yield of cyclodimeric product (2 of Table 2). Vinyl cation 11 is reasonably indicated as the reaction intermediate and a stepwise mechanism, involving addition of 11 to the cyanoacetylene followed by chloride ion attack and ring closure, has been proposed. However, the facile cyclodimerization process (equation 6), may well occur via a concerted [2 21 addition. It has been argued that a vinyl cation may be an ideal component for a thermal [,,2, + ,,2J addition which is allowed by the orbital symmetry selection rules (Woodward and Hoffmann, 1969) (equation 6).
+
Addition of hydrogen chloride to t-butylacetylene has been studied in neat mixtures in the liquid phase at room temperature (3 of Table 2). Both “normal” and “rearranged” products have been isolated and the ratio of the two types of products was found to diminish by increasing the ratio HCl/t-butylacetylene. The detection of rearranged products has been explained by assuming that the initially formed vinyl cation
197
VINYL CATIONS
12 may undergo a 1-2 methyl shift to generate the ally1 cation 13 which leads to rearranged products. Following the authors’ suggestion, mixtures of high HCljt-butylacetylene ratio provide a medium of higher polarity which may prolong the lifetime of the cation and hence favour rearrangements. Me
+ I Haw-C-Me I
__f
Me
[
Me
Me
+‘
HzC--C-c,
H&=L/Me ‘Me
Me
The kinetics and stereochemistry of addition of HC1 to l-phenylpropyne and to 3-hexyne in acetic acid have been found to be different in the two cases (4 and 6 of Table 2) (Fahey andLee, 1966,1967,1968). I n the case of 1-phenylpropyne a vinyl cation is suggested as intermediate. The kinetic law is :rate = k, [--C=C-] [HCl]2, The second order dependence on HC1 is explained by assuming that proton transfer to the triple bond results in the anion hydrogen dichloride, HCI;. The product distribution and the stereochemistry, under kinetic control, have been explained by assuming that a cis-oriented intimate ion-pair, 14 is initially formed which, following Scheme 2, may either collapse to cis chloride, undergo anion displacement by acetic acid to form trans acetate or a randomly oriented (solvent separated) ion-pair 15 which gives “racemic ” material. SCaErdE
2
Me-CdSPh
l(w
HOAo EtCOPh
cis
T
+tram acetate
C-
[
MeC€I=6-Ph]
HClr HOAc (15)
___+
+
cis tram chloride
198
UIORQIO MODENA A N D UMBERTO TONELLATO
A similar mechanism ww recently suggested for the hydrochlorination of styrene and t-butylethylene by Fahey and McPherson (1969). The proposed mechanistic scheme appears reasonable in a weakly dissociating solvent, such as acetic acid, although it seems somewhat surprising that the product distribution and the stereochemistry are insensitive to addition of diluenta such a8 formic acid up to 1 . 1 ~ . The addition of HCI to 3-hexyne shows an unexpectedly different picture. The rate law is rather complicated and the reaction is accelerated by addition of chloride ion (as tetramethylammonium chloride). The product distribution ranges from that shown under 6 of Table 2 to the almost exclusive formation of trans-chloride in the presence of ,added ammonium chloride. The postulated mechanism excludes the formation of vinyl cations intermediates and indicates a concerted transition state for the reaction of type 16.
The difference in mechanisms of addition of trifluoroacetic acid (see above) and HCl (in acetic acid) to 3-hexyne has been tentatively explained as due to the different nucleophilic and ionizing power of the two solvents. The addition of hydrogen halides to propiolic acid (6 of Table 2) in water occurs by a mechanism entirely different from that of the acidcatalysed hydration of phenylpropiolic acid (see above). Here again vinyl cations are probably not involved as intermediates and the major mechanistic pathway is the one involving formation of 17 which stereospecifically adds a proton to form trans adducts following equation (7).
1 A similar mechanism is apparently involved in the stereospecific (trans) addition of hydrogen chloride to 1-dkynylphosphines (7 of
V I N Y L UATIONB
199
Table 2). The kinetic data and, in particular, the effect of substituents (both R and R') on the rate of addition indicate that the rate-determining step involves attack of the chloride ion to the triple bond of a protonated l-alkynylphosphine molecule (equation 8).
(d) Neighbouring-Croup Participation in Proton Addition Reactions. Examples of proton addition to the triple bond assisted by intramolecular nucleophilic intervention have been reported. The acid catalyzed hydration of compounds 18 to give both the expected /?-hydroxy ester 19 and the unsaturated ester 20 proceeds at a rate which is larger by a factor of 50-100 than calculated on the basis of the Taft p*o* relationship (see 7b and 8 of Table 1) (Hekkert and Drenth, 1961). The rate enhancement has been explained in terms of anchimeric assistance efiects due to the hydroxyl group, and a four membered ring structure of type 21 has been proposed for the cationic intermediate.
. (22)
(23)
The study of the addition of trifluoroacetic acid to several 5-substituted 1-pentyne derivatives 22 led Peterson and Duddey (1963, 1966) and Peterson and Bopp (1967) to the conclusion that for some of the compounds investigated the intermediate resulting from protonation of the triple bond may be a five-membered ring cation of type 23, rather than a linear vinyl cation. This is indicated by the observation of rearranged products deriving from 1-5 shift when Y = C1, F, OMe, and OAc (see scheme 3). Kinetic data indicate that Y-participation is important in the transition state of the reaction. Anchimeric assistance effects, expressed in terms of the ratio of assisted to unassisted mechanisms, kA/ks (Heck and Winstein, 1957), have been estimated to be 3.4 for Y =C1 and 6-6for Y = OMe. The entire kinetic picture almost matches
QIORQIO M O D E N A A N D UMBERTO TONELLATO
200
SCHEME 3
+unshited produata 15%.
86%
QaCCFs (22) (Y=F)
I
Me*C*(CHa)aaOgCCFs +unehidpmducte
I
F (22) (Y=OMe)
16%
80%
+ Me-CO*(CHa)o*OaCCF3+(MeO-CO*CFa)
+
(22) (Y=OAc) + Me *CO.(CHa)a OaCCFa Me *CO (CHa)aOAa
that observed for the corresponding 6-substituted 1-pentenes (Peterson et al., 1966). However, there is a significant difference in the product distribution : more shifted products are obtained from alkynes (86% in the case of 6-chloro I-pentyne; see Scheme 3) than from alkenes (0.6% from 6-chloro I-pentene in identical conditions). Since cyclic intermediates are likely to be formed in each case, the authors argued that the breaking of the C-C1 bond in the vinyl-onium ion 23a must occur exclusively at the C5-Cl position and preferentially at the C,-Cl bond of the saturated cation 24.
R’
424,
MeQ H
++
The authors point out that, unlike vinyl halides, the vinyl chloronium ion 23a could hardly show double bond character in the C-Cl bond because of unfavourable charge repulsion in 23b. The unreactivity of vinyl halides may therefore primarily be a consequence of the strength of the (I bond to the spz carbon. However if opening of 23a is assumed to be due to nucleophilic attack of the solvent (or any other nucleophile) to either C5 or C2,the greater amount of shifted products obtained from akynes than from alkenes is a direct consequence of the relative facility of nucleophilic attack at the sp3 carbon and sp2 carbon (Rappoport, 1969; Modena, 1971). 2. Addition of Carbonium Ions
The addition of carbonium ions to acetylene derivatives has often been exploited for synthetic purposes but the meohanism of such
V I N Y L CATIONS
201
additions was not usually considered (Winterfeldt, 1969; see, in paxticular, Yen, 1962). (a) Addition of adawntyl cation to acetylene. Attention has recently been devoted to the particular caae of addition of the adamantyl cation to acetylene and its mechanistic implications. Sasaki et al. (1968), observed formation (at least 40%) of adamantyl methyl ketone (25) and no detectable amounts of adamantyl metaldehyde (26) in the reaction of 1-adamantyl bromide in hexane-90% H2S04at 5'.
Br+€IWH
IIaSOI.
Ad*CO.Me (25)
More recently Kell and McQuillin (1970),for the same reaction in 98% H2S04, detected, besides 25 (76%), small amounts (8%) of adamantyl aceta-ldehyde (26) and homoadamantan-4-one (27). Bott (1969 a, b), by allowing adamantan-1-01 and acetylene to react in 96% H,SO4 at &To, observed the formation of aldehyde 26 as the main product, together with ketones 25 and 27. The yields of ketone 25
relative to 27 was found to increase when three methyl substituents were introduced into the bridgehead positions of the adamantyl moiety. From adamantan-1-01 and acetylene in 90% H2S04 Kell and McQuillin (1969) obtained exclusively the aldehyde 26 and ketone 27. Addition to substituted adamantyl acetylenes, AdCZCR (R=Ph, n-Pryn-Hexyl) resulted in the formation of ketones AdCH2COR. There is no doubt among the above authors that a vinyl cation 28 is + AdCH=CH (28)
202
Q I O R Q I O M O D E N A A N D UMBERTO TONELLATO
primarily formed. Its fate is the object of different interpretations to accommodate the confusing results. According to Kell and McQuillin, cation 28 may in the presence of Br- undergo deprotonation to adamantylacetylene AdC-CH which then, by hydration, would give predominantly ketone 25 (see, however, 6 of Table l), whereas, in the absence of Br-, 28 may go to AdCH=CHOSOsH and AdCH2*CH2 (O.SO,H), which would yield homoadamantan-4-one (27) in a synchronous process. Bott instead considers that cation 28 may undergo hybride shift to form cation 29, react with water to give ketone 25 or rearrange to cation 30, from which the ketone 27 is formed by double hydration and pinacolic rearrangement.
(b) Triple-Bond Participation. Triple bond participation may be considered a special case of addition of a carbonium ion (or a carbon atom having partial carbonium ion character) to an acetylenic b-ond although, in some caaes, it may be envisaged as a direct nucleophilic displacement by the triple bond in competition with the solvent. The relevant studies on the subject are summarized in Table 3. I n all cases shown, conditions were such as to exclude solvent addition to the triple bond. Triple-bond participation has been mainly studied (by Hanack and his school) in reactions of “homopropargyl ” derivatives 31 under solvolytic conditions. I n all cases of Table 3, except 1 and 2, compounds 31 yield, besides (and very often instead of) the expected solvolysis products, cyclobutanone derivatives (32) and alkyl cyclopropyl ketones (33).
TABU 3 Triple Bond Participation in Solvolytic Reactiona ~-
-
8ubstraW
Reaction Medium (SOH) 8°C
Products, Yields
R
1 MeCkC.(CH2)2Br 2e Me, W , (CH2)zBr 2b
70
20
50
3a 3b 3c 4&
4b 8a
5b $a
6b
HzO (AgzO) MeOH HCOaH CF3C02H M e w . ( a 2 ) 8 .m-NBS AcOH CF3COzH CFaCOaH (Hg++) E t W . ( C H & 60n-NBs CFsCOzH CFsCOzH (Hg++) i - P r M .(CH2)z. CFaCOzH m-NBS m3COzH (Irg++) H W . ( C H ~ ).(3,5B CF3COzH DNBS) AcOH
70 50 60 50 50 60
60
exclusively exclusively 83%
-
L
-
-
16% 85% 98% 98% i)% 95% 10% 98%
25% 20%
12%
~
__
~
References
-
-
1% 9%
-
1% 97% 1% 90% 1%
76%
Hanack et al., 1965 Hanack et al., 1965 Hanack et al., 1965 Hanack et al., 1965 Hanack et al., 1967 Hanack et al., 1967 Hanack et al., 1967 Hanack et al., 1967 Hanack et al., 1967 Hanack et al., 1967 Hanack et al., 1967 Hanack et al., 1968 Hanack et al., 1968
tw 0 W
(EtWO-)20
.
9 EtO C k C . (CHS)~.OTe
8a XdkG.GHs. CH{I&) (m-NBS)
-
MencoJrpO
68
CFaCO2H
32%
60%
71%
15%
21%
Hanack et d.,1968
25% R&nacketd.,1968
46%
ab
M08CO-XaO 1:2v/v
80
115
MezCO-BaO OTs
1 :4 v]v
70
90%
10%
Q
(tram:60%)
Bnmk el d.,1968
0
x0
d cl
KM
w
La
Y 0 U
Wilson, 1969
Hansck and Heumsnn, 1969
10b
CH3COzH
70
(trans >60%)
-
"0
llb 1 Ic
2%
CF3COeH
.
HCOzH CH3COzH
25 26
40% 30%
38%
50%
1 : 4 HCOpH
c&co,&
15% 20%
Peterson and Kamat. 1969
22%
Peterson End Kamat, 1969 Peterson and Kamat, 1969
19%
9 : l 49%
51%
CFSCOaH
7%
1
I
12a Me .W.(CHZ)~. CHMe .om 12b 120
elimination products
90%
J S 11s H W . ( C H & . CHMe OTs
Hanack and Heumann, 1969
30%
60%
6Oy0
Peterson and %at,
6% 20%
1969
Peterson and %at, 1969 Peterson and Kamat, 1969
\
60%
OAc 36%
Clown and Roman, 1966
~0 m-NBS=m-nitrobenzenesulphonete; 3,6-DNBS=3,1-dinitrobenzenesulphona~ ; OTs =p-toluenesulphonate; OBa=pbromobenmneulphomte.
m
0 01
206
(XIORBIO MODENA A N D UMBERTO TONELLATO
The relative yields of cyclized products increase greatly with the decreasing nucleophilicity (2a-c, 3 and 4 of Table 3) and with the increasing ionizing power of the solvent. Cyclobutanone derivatives (32)are usually obtained in much greater amounts than cyclopropyl ketones (33)but, when mercuric ions are added, formation of 33 is overwhelming. The effect of R in 31 on the product distribution is illustrated in cases 3b, 4a, 5a, 6a, and 7. All the above facts are of importance when homopropargylic participation is exploited as a valuable synthetic route (Hanack and Herterich, 1966; Hanack et al., 1970) but also offer information concerning the probable mechanism. This is shown in Scheme 4. The bridged ion 34
L
(34)
(37)
is likely to be the actual intermediate when triple-bond participation is involved. The intermediacy of vinyl cations 35 and 37 has been tentatively suggested to explain the formation of ketones 32 and 33,respectively although, at least 35,appears unlikely in view of its very strained arrangement. Moreover, if 35 exists, it is not a primary intermediate since analysis of the cyclobutanone derivative 32 obtained by trifluoroacetolysis of 31 (R =Me, X =m-NBS) deuteriated a t C1 showed that scrambling of positions 1 and 2 of the original material had occurred during the reaction (Hanack and Vott, 1968). Derivative 36 has been shown to be a labile precursor of 32 (Hanack and Vott, 1968). The effect of substitution of a methyl group for a hydrogen at C1of 31 on the product distribution is illustrated in cases 8a and 8b of Table 3. Substitution of two methyl groups for the hydrogens of Cz is a special case investigated by Wilson (1969). Under the conditions shown in Table 3 (9), compound 38 reacts some 3000 times faster than calculated on the basis of the rate of the saturated analog and of the inductive effects due to the presence of the triple bond. It is interesting to note that anchimeric assistance effects on the rate are larger than those for
V I N Y L OATIONS
207
homoallylic participation (factors of 440-2000) and smaller than those calculated for homoallenyl participation (see infra) (factor of 830058000) on the basis of data on neopentyl-like substrates (Bly et al., 1967). The only product isolated at the end of the formolysis reaction (90-95%) was ketone 39 but lH n.m.r. analysis of the reaction allowed Me
1
Me*CH2*CO*W(Me)s
MeCksC-C-CH2-0Te
I
the author to observe the successive formation and disappearance of intermediates 40 and 41 (S=CHO). Cyclized products of type 32 and 33 and unrearranged solvolysis products were not observed. Here too a vinyl cation is indicated as possible intermediate, as shown in the sequence of Scheme 5 . Following the Scheme the intermediate goes over into the primary solvolysis product 40 which, by addition and elimination of a solvent molecule, yields 41 as precursor of the final reaction product 39. Somm 5
Triple-bond participation is involved also in 6-heptyn-2-yl and 6-octyn-2-yl derivatives, in spite of the fact that the triple bond is four carbon atoms removed from the reaction centre. Values of kA/k,(Heck and Winstein, 1957) have been evaluated to be 6.5, 0.7 and 0.1 in cases lla-c of Table 3, respectively, and 84, 5.6, and 2.3 in the sequence 12a-c. These values indicate that participation occurs in most cases also in the transition state of the reaction. The observation of products with sixand five-membered rings was first interpreted by assuming that a “bent” secondary vinyl cation 42 as well as a linear vinyl cation 43
208
GIOROIO M O D E N A A N D UMBERTO TONELLATO
are involved as intermediates in the solvolysis. However, the estimated energy difference between the cations 42 and 43, in the light of the observed product distribution, led Peterson and Kamat (1969) to conclude that the transition states for solvolysis of 6-alkyn-2-yl tosylates “bear almost no resemblance to linear or bent vinyl cations ”. A bridged ion of structure 44 is therefore suggested as the most likely intermediate.
I
R
R
The large amounts of five- and six-membered cyclic tosylates, products of internal return (49% yield in one case : 12a of Table 3) points also to a highly selective intermediate, although the reason for such a remarkable selectivity towards external nucleophiles is not clear. On the other hand, acetolysis of 6-phenyl-5-hexynyl brosylate, which is apparently = 1.6) yields only anchimerically assisted by the triple bond (kobs./kcalc. the five-membered ring product in addition to open-chain solvolysis products (13of Table 3). The effect of the phenyl group in orienting the cyclization reaction would indicate that intermediate species like 43 may become important when R’ =Ph. Finally, when the triple bond is three carbon atoms removed from the reaction centre, participation effects are apparently much less important, as is indicated by the fact that 4-pentyn-l-yl tosylate does not cyclize on acetolysis (Kwiatkowski, as reported by Closson and Roman, 1966). 3. Addition of Halogens
This type of reaction again has received little mechanistic attention. The intermediacy of vinyl cations in the addition of halogens to some acetylenic derivatives has only recently been convincingly demonstrated (Pincock and Yates, 1968, 1970; Wilson and Berliner, 1971). Prior to this report, the addition of bromine to alkynes in acetic acid was suggested to occur via an electrophilic process (although not explicitly involving vinyl cations as intermediates) by Robertson et d. (1960). The reaction rate was found to obey mixed second and thirdorder kinetics and to be enhanced by electron-supplying substituents. It wm also noted that strong electron-withdrawingresidues bonded to the triple bond may switch the reaction pattern towards a nucleophilic
VINYL CATIONS
209
mechanism. This picture closely parallels that observed in the caae of the corresponding alkenes except for the large difference in rates: the bromination of oleic acid occurs 50,000 times faster than that of stearolic acid and that of styrene 3,000 times faster than that of phenylacetylene in identical conditions. The substrates, reaction conditions and products investigated by Pincock and Yaks are shown in Table 4. The rate of bromine disappearance follows in each case the equation : rate = {kz[Bra]+k3 [Bra]a+k~~-[Bra] [Br-I} [-G=C-]
Activation entropiesof - 29 and - 51 cal. mo1-lK-l have been caIculated for k2 and k3 respectively in the case of tolylacetylene and the difference is that expected between a bimolecular and a termolecular event. A value of AS* of - 41 cal. mol-l K-l for k2in the case of 3-hexyneindicates a more ordered transition state than in the case of phenylacetylene derivatives. The values of k2 for ring-substituted phenylacetylenes are well correlated by the Hammett relationship, using u+ constants, and p is very large, - 5.2. The p value compares well with those obtained in the acid-catalysed hydration of arylacetylene derivatives (see Table 1 ) and is entirely consistent with the formation of a vinyl cation 45 in which the positive charge is at the carbon next to the phenyl ring and is largely shared by it. Interestingly, the p value (by use of a+)based on k2 rate coefficients for the bromination of styrene in acetic acid is - 4.5.
The product distribution from phenylacetylene in the absence of added salts shows that the bromination is not stereospecific (under kinetic control) although trans &bromide is obtained in higher yields than the cis isomer. The preferential formation of trans adduct as well as the formation of bromide-acetate (solvent-incorporated)adducts may be accounted for by the formation of intimate and solvent-separatedion pairs, approximately as outlined in Scheme 2. From tolyacetylene, solvent-incorporated products were not obtained (see Table 4). The reasons for such an abrupt change are not clear, although it has been suggestedthat in this case a more stable and discriminating ion is formed. ~ the bimolecular kinetic term is only a In the presence of 0 . 1 LiBr minor contributor and a bromide-ion catalysed process largely prevails
TABLE 4
E:
:
Addition of Bromine (Pincock and Yates, 1970) Substrate
Reaction Medium
2 0
Products
a0 Ar
1 la
HkCPh
2 2a
HWTol
Br, /C=C\Br
3 3a
.
CH3.CHa C=C. CHz .CHI
BrkCAr ‘OAc
42% 99% 56% 81%
HOAc HOAcSLiBr (0.1~) HOAc HOAc + LiBr ( 0 . l M )
HOAc HOAc+LiBr ( 0 . 1 ~ )
H
19%
-
44%
19%
f-
72% 99%
‘Br
u
Ar
Br\ ,c=c / H \Br
4-
unidentified material
M
P
VINYL UATIONS
211
through a transition state like 46. This change of mechanism is consistent with the fact that, when LiBr is added, the only product is the trans dibromide (see Table 4).
From alkylacetylenes trans dibromide is formed in large or quantitative yields and no detectable amounts of cis or solvent-incorporated diadducts were observed. The indications are that in this case the electrophilic addition occurs via a cyclic “bromonium” ion (47) and that this is probably more stable as an intermediate than the linear alkyl vinyl cation. From 47 only trans addition is expected to occur. ALK-C-C
\+Br/
ALK
Also from this study the difference in the rate of bromination of alkenes and alkynes is quite evident. Thus the bimolecular coefficient Ik2) for styrene is 2 x lo3 times that for phenylacetylene and k2 for 3hexene is 1.4 x lo6 times the value for 3-hexyne. Since the first two compounds are believed to react via carbonium ions and the latter two via “bromonium ions” there seems to be an extra factor of lo2 in the stability of bromonium ions from alkenes relative to those from alkynes. The addition of iodine to sodium phenylpropiolate in aqueous solutions has been thoroughly investigated by Wilson and Berliner (1971). I n - 7 x 1 0 - 4 ~ the ) rate the presence of small amounts of iodide ion ( law contains three terms :
where K 1 is the dissociation constant for the triiodide ion. The term k, has been interpreted as the rate constant for a termolecular reaction which becomes dominant at high iodide ion concentrations (see below) and k’ as the rate constant for a bimolecular process involving molecular iodine. On the other hand, k” would be the rate coefficient for a reaction pathway involving a rate-determining addition of the (hydrated) iodine cation (Scheme 6 ) . The detection of a t least three iodinated ketoacids among the reaction products is offered as a supporting argument for the
212
GIORGIO M O D E N A A N D UMBERTO TONELLATO
Somm 6 P h a * C O z - + I+(HaO)
___+
several (iodo)ketoacids
k
I
P h . C s C I A PhCI=CIa
formation of a cyclic iodonium ion 48 rather than an open vinyl cation. ) reaction At moderately high iodide ion concentrations ( 0 - 0 2 - 0 . 1 0 ~ the rate depends on iodine, iodide, and phenylpropiolate concentrations and trans- cr,/3-diiodocinnamic acid is obtained in high yield. The most reasonable mechanism is a termolecular concerted iodine-iodideaddition through a transition state of type 49.
Addition of pseudohalogens, such as BrN, and INs, to unsaturated compounds has been shown to occur either by a free radical (via N;) or by an electrophilic (viaX+) stepwise mechanism, depending on the reaction conditions (Hassner and Boerwinkle, 1969). Electrophilic addition of BrN, to 1-phenylpropyne is non-regiospecific(Hassner, 1969) and both products 50 and 51 are formed. The addition of IN, is instead Ph-CEC-Me
XNa _____+
(acetonitrlle)
Ph(N&=CXMe (50)
PhXC=C(Ns)Me (51)
regiospecific, the only product being adduct 51 (Hassner et al., 1969). The orientation, which occurs in the opposite sense to hydration (see under 2 of Table l ) ,would suggest, following the authors, the formation of an unsaturated three-membered cyclic iodonium ion (52), in which the
VINYL CATIONS
213
double bond is resonance-stabilized by the phenyl group and in which the positive charge, in a resonance form 52a, is effectively stabilized by the purely inductive effect of the methyl group. However, the reaction with INs is largely non-stereospecific and the geometric isomers of 51 are obtained in the ratio 2:l under kinetic control. The formation of large amounts of the cis isomer of 51 has been tentatively explained in terms of cis oriented ion-pairs (see Scheme 2) but it may be a serious argument against the hypothesis of cyclic iodonium ions (Storr, 1970). The addition mechanism is probably more complicated than suggested and the idea of the formation of 52, although reasonable, requires further support. It is evident from the above arguments that the addition of halogens t o alkynes may occur by a multiplicity of mechanisms. One more argument is offered by a kinetic study of the addition of bromine to diphenylacetylene in bromobenzene reported by Sinn et al. (1965). I n this case kinetic data are better interpreted on the basis of a nucleophilic mechanism involving the intermediacy of a vinyl anion, rather than an electrophilic reaction. 4. Addition of Sulphenyl Halides
Addition reactions of sulphenyl halides (in particular chlorides) to acetylene derivatives have been extensively explored and recently reviewed (Modena and Scorrano, 1968). Although free radical processes may be involved under specific conditions, the addition of both areneand alkanesulphenyl halides normally occurs by an ionic mechanism, the sulphenyl halide sulphur being the electrophilic centre. The idea of an electrophilic process is supported by the fact that the rate of addition is strongly affected by the solvent (Orrand Kharasch, 1953, 1956; Hogg and Kharasch, 1956; Dondoni et al. 1964) and by the presence of alkyl substituents at the acetylenic carbons as indicated by the fact that p-toluenesulphenyl chloride reacts with 3-hexyne 2 x lo4 faster than with acetylene (Dondoni et al., 1964). On the other hand, the rates of addition of p-toluenesulphenyl chloride to substituted diphenyl acetylenes are correlated by the Hammett relationship using o+ values but the p values, although negative, are rather small ( - 1* 3 in chloroform and - 1-8in ethyl acetate) (Di Nunno et al., 1965)when compared with those observed in addition reactions proceeding through linear vinyl cations (see previous sections). Moreover, the effect of substituents on the rate of addition of various benzenesulphenyl chlorides to 1butyne and to diphenylacetylenes results in a flat bell-shaped pa plot (Dondoni et al., 1964; Di Nunno et al., 1965). The addition of sulphenyl halides is in each case trans stereospecific
214
GIOROIO M O D E N A A N D UMBERTO TONELLATO
(Truce and Boudakian, 1966; Schmid and Heinola, 1968; Calo’ and Scorrano, 1968;Calo’ et al., 1968)but not regiospecific. The orientation depends both on the structure of the acetylene and the solvent and the results may be summarized as follows : (i)anti-Markownikoff adducts are generally formed from alkylacetylenes (equation 9) (Calo’ and Scorrano, 1968); (ii)the orientation of the addition to phenylacetylene is solvent dependent (equation 10) (Calo’ et al., 1968); (iii) Markownikoff-type
ethyl acetate
100%
--
chioroform
65%
35%
acetic acid
29%
71%
adducts are obtained as main products in the case of monosubstituted diphenylacetylenes (Di Nunno et al., 1966); (iv) the orientation is not sensibly affected by the nature of the sulphenyl halide with the notable exception of the 2-nitro- and 2,4-dinitrobenzenesulphenylhalides which show a greater tendency to give Markownikoff oriented products (Dondoni et al. 1964;Kharasch and Yannios, 1964). Both kinetic data and the nature of the products concur to exclude that linear vinyl cations are formed in any case in the addition reaction. There is little doubt that a bridged species is involved as intermediate and its structure may well be that of a thiirenium ion (53) more or less
R (53)
tightly paired with the counter-ion. Opening of the three-membered ring of 53 in a SN2-typeprocess may lead to the anti-Markownikoff orientation observed in the case of alkylacetylenes. I n the case of alkynes which possess a stabilizing group, such as an aryl group, or when a polar solvent is used, the ring opening process may be S,1 in character to give the Markownikoff-type products.
VINYL CATIONS
216
It is noteworthy that recent observation of bridged compounds with a tetraco-ordinate sulphur atom (Owsley et at., 1969) may support the hypothesis that a species of type 54 is initially formed. From 54 the h a 1 products may be formed through different paths depending on the nature of the acetylene and on the reaction medium.
Electrophilic addition of sulphenyl halides t o alkenes occurs, by all the evidence, via cyclic thiiranium ions (Mueller, 1969)and a comparison of the rates of addition to the double and triple bond would be quite interesting. Unfortunately, direct kinetic data for strictly comparable and typical cases are not available. Phenylacetylene has been reported (Kharasch and Yannios, 1964) to react lo2 times slower than styrene (in acetic acid at 25') with 2,4-dinitrobenzenesulphenylchloride. On the other hand, Thaler (1969),by means of competitive experiments carried out in dilute paraffin solutions at - 70°,estimated that methanesulphenyl chloride adds to mono- (and di-)alkylacetylenes 60-100 times more slowly than to the corresponding alkenes (cis) (but only ca. twice slower than to trans dialkylethylenes). The paucity of information does not allow generalizations and further work in this area seems desirable also with respect to the much larger rate differences observed in those bromine additions to triple and double bonds which also occur via bridged species. 5. Other additions
Many other addition reactions are likely to involve an electrophilic mechanism via vinyl cations or bridged cationic species. A particularly interesting field which has been exploited for synthetic purposes but little understood from a detailed mechanistic viewpoint is that of the additions catalyzed by salts of mercury, silver and other metals (Winterfeld, 1969;Miocque et at. 1963).
B. Electrophilic Additions to Allene Derivatives Electrophilic additions to allene derivatives were shown to occur, in number of cases via attack of a positively charged electrophile to a 8
216
G I O R Q I O MODENA A N D UMBERTO TONELLATO
terminal carbon atom of the cumulate double bond to generate a vinyl cation intermediate (equation 11). I n the simple allene such direction
of attack is favoured by the polarization of the bonds in the molecule : 6-
a+
6-
CHz=C=CH2
However, in many cases, in particular for allene derivatives bearing substituent which can stabilize a positive charge at a terminal carbon atom, the electrophilic attack is favourable at the central carbon atom to give an allylic-type ion (equation 12). However, it is to be emphmized
that the intermediate cation of equation (12) is not resonance-stabilized unless it can survive a 90" rotation around the 0 bond joining the terminal carbonium atom to the central carbon atom (Scheme 6) (Jacobs and Johnson, 1960).
In poorly nucleophilic media such as FS03H-SbF,, allylic carbonium ions can easily be generated by protonation of the central carbon of 1,3-dirnethylallene and of tetramethylallene. These ions have been directly observed (Pittman, 1969) and identified from the 'H n.m.r. and U.V. band positions and splittings and kept unchanged for 1 week at
- 70". The alternative modes of addition under equations (11) and (12) can be easily distinguished on the basis of the primary product distribution. However, bridged species other than open ions are frequently involved as intermediates and ring opening may occur by direct nucleophilic displacement processes (Poutsma and Ibarbia, 1971).
217
VINYL CATIONS
1. Proton Addition
(a) Addition, of water (acid-mtalysed) and carboxylic acids. Recent mechanistic studies on the acid-catalysed hydration of allenes are not available. There is however accord among the most recent reviewers (Taylor, 1967; Griesbaum, 1965; Richey and Richey, 1970; Mavrov and Kucherov, 1967) on the idea that the hydration in sulphuric acid of allene, 1-alkylallenes,and 1’3-dialkylallenesproceeds via protonation of a terminal carbon as in equation (1 1) and hydration of the resulting vinyl cation to give acetone or acetone derivatives (eq. 13) aa final RCH=CECHR (R=H, Alk)
R+,Hi0
+ RCHz * CO .CH&
(13)
products (Petrov and Fedorova, 1964 and references cited therein). The case of 1,tdialkylallenes is apparently the turning point towards central protonation which is expected for more alkylated allene derivatives (see infra). As a matter of fact, Pittman’s report (1969) of the direct observation of the allylic carbonium ion 55 from 1,3-&methylalleneand, in particular of the ion containing deuterium bound at the central carbon when FS03D-SbF6was used, is in apparent contrast with the nature of the products obtained in sulphuric acid.,
Protonation of the central carbon atom (equation 14) probably occurs in the case of alkoxyallene derivatives (56) which, by reaction in dilute sulphuric acid at 60”, yield the corresponding cc,p-unsaturated aldehyde 57 (van Boom et al., 1965; Richey and Richey 1970). H+
+
R’CH=C=CH*OEt +R’CH=CH * C H OEt
__C
(R‘=t-Bu) (56)
R’CH==CH*CH(OH)(OEt) +R’CH=CH*CHO (57)
(14)
218
GIORGIO MODENA AND U M B E R T O T O N E L L A T O
The addition of acetic acid at reflux temperature to the aIlenic ester 58 gives adduct 6, ester 7, and the product of cyclodimerization 59 (Verny and Vessibre, 1969).
The product distribution, except for the formation of 59, resembles that obtained from the corresponding p-acetylenic ester (equation 6 ) , and the formation of a vinyl cation of type 60 is suggested for both cases.
+
CHa=C-CHCl-C02Et (601
Very recently, Bergmann (personal communication) prepared allene 61 and found that it reacts in acetic acid at 100' to give two major products, 62 and 63. These compounds are consistent with a mechanism
involving protonation predominantly or exclusively at the methylene carbon of 61 to generate the vinyl cation 64 which formally undergoes ring opening and ring enlargement at a rate greater than that of solvent trapping. These results are reminiscent of those obtained in the case of homopropargyl participation except for the absence of cyclopropyl methyl ketones (see Scheme 4). Crandall et al. (1968) reported the reaction of allene 65 in acetic acidsulphuric acid. The reaction products were identi6ed as a mixture of 66 and 67 (S = H, MeCO) suggesting terminal protonation of 65, an orientation different from that obtained with tetramethylallene and hydrogen halides (see infra).
219
VINYL CATIONS
Me
I
M Me e > & C q i
H
Me
I /Ha I eM' Me C I c---;c--c--C
Me
I
I
Me
Me4
Me
Me Me
I I I I
H-CL-CECC448
Me Me
Further investigations on electrophilic additions to (65) (Poutsma and Ibarbia, 1971) found that the products of the reaction with trifluoroacetic acid in benzene at - 15" to - 25" are mainly 66 and 67 (S=CF,CO), but minor amounts of 68 and 69 were also observed. A similar product distribution was obtained from 65 with acetic acid, with water in tetrahydrofuran-sulphuric acid, and with methanol in the presence of acids. The detection of compound 68 indicates that terminal protonation is not exclusive and some central attack also occurs.
However, the vinyl cation 70 is likely to be the major primary intermediate which may either be trapped by the solvent to yield compound 69 or undergo fast ring opening to the stable tertiary carbonium ion 71 aa precursor of products 66 and 67. The open-chain materials resemble Me
1
H--CCZ&--CC
I
Me
Me
Me
I
Me
Me
I I
Me
,Me + e M '
that obtained in the solvolysis of the neopentyl-type homopropargyl derivative investigated by Wilson (sect. 11, A, 2, b) and both reactions probably proceed largely through a closely similar geometry of type 72 (see Scheme 5).
220
OIOROIO MODENA A N D UMBERTO TONELLATO
(b) Addition of Hydrogen Halides. Recent studies are available for this type of addition and the proposed mechanisms have been inferred on the basis of the product distribution. Relevant cases are reported in Table 5. Allene reacts with HC1 and HBr (1 and 2 of Table 5 ) to give 2-substituted propenes, 2,2-disubstituted propanes and considerable amounts, depending on the conditions, of cis and trans 1,3-dihalo-l,3-dimethylcyclobutanes which were shown not to arise from head-to-tail dimerization of 2-halopropenes. At least in the case of HBr, the same product distribution was obtained from propyne (1 of Table 2). Cyclodimeric products were not obtained from allene or propyne with HI. According t o Griesbaum et al. (1965), (see also Griesbaum, 1966, 1969)the addition reaction for allene and propyne may proceed as illustrated in Scheme 7. Vinyl cation 73 is formed and may be trapped by halide ions X- to give 2-substituted propenes (the exclusive pathway in the case of I-, a relatively strong nucleophile) or add to allene (or propyne) through the cyclodimerization routes which are represented as distinct multi-step processes. As noted above, thermal [,2, + .2,] concerted cycloadditions of a vinyl cation to double and triple bonds are probably involved: these are allowed by the orbital symmetry selection rules according to Woodward and Hoffmann (1969). It has recently been shown that, at least in the presence of HC1 and BiCl,, there is no acid-catalysed isomerization of allene to propyne and that the vinyl cation 73 is formed independently from both unsaturated compounds. I n fact, allene and DCl, in the presence of BiCl,, give, at very low conversion, 2-chloropropene containing deuterium exclusively in the methyl group, and the recovered allene does not contain deuterium (Charleston et al., 1969). 1,2-Butadiene and HCl(4 of Table 5) yield a mixture of trans and cis chloro-2-butene together with considerable amounts of 2-butyne which, in turn, adds HCl much more slowly than the parent isomer. The orientation is that predicted following equation (11) and indicates the intermediacy of a vinyl cation. When two methyl groups are attached to one of the terminal carbon atoms of the allene, the orientation is changed. The addition of HC1 to allene 74 (6 of Table 5 ) , which is accompanied by rearrangement to isopropene, results in the formation of allylic chlorides 75 and 76 in the ratio 2 :1. On the other hand, addition of hydrogen chloride to isoprene (77) via ally1cation 78 gives the same products under the same conditions, but the ratio is 6 :1. This shows that not all of the addition to allene 74 occurs through isoprene and that different cationa result from protonation of the allene and of isoprene. Jacobs and Johnson (1960)estimated that, since rearrangement of allene 74 to isopropene is 1.5-3 times faster
TABLE5 Addition of Hydrogen Halides t o Allene Derivatives Substrate
HX
2OC
ax’a
Products
I\leCX=CHz
MeCXzMe
References
(%conversion)
x
1 H&=C=CHz 2 H2C=C=CHz 3 H2C=C=CHa
4 lleCH=C=CH2
HCI ( 1 : l ) HBr ( 1 : l ) HI ( 1 : l )
HCI
- 70 - 70
- 70 - 78
8 Yo 69% 95% C1 Me
5% 30% 6”’
53.5% 53.1%
40% 40%
Me I CI-C-CH=CHa
I
6 (Me)2C==C==CH2 6 R-CH=C==CHMe
HC1 HBr
7 (Me)&=C=C(Me)z
HBr
- 78
-40
+ 20
Me 64%
.
RCH=CBr CHzMe
77% 1 x 44%
C1
10% 1% 8%
1 58 42
Griesbaum et al., 1965 Griesbaum et al., 1965 Griesbaum et al., 1965
20 59
Jacobs and Johnson, 1960
H
6.5% 8.1%
Me \C==C-CH&l / Me 36% RCH=CH ,CHBr .Me
(Me)ZC=CH-CBr
(Me)2
Jacobs ar.d Johnson, 1960 Fedorova, 1963 Bianchini and Guillemonat, 1968 Bianchini and Guillemonat, 1968
222 QIORGIO MODENA A N D UMBERTO TONELLATO
h
E
VINYL CATIONS
223
the addition of acid, direct addition of HCl to 74 gives at least two-thirds and probably exclusively the chloride of structure 76. Making allowance for the fact that central protonation of 74 does not generate aresonancestabilized ally1 cation, one might expect formation of a tertiary carbonium ion 79 (precursor of 75) and not a primary carbonium ion 80 (precursor of 76). This is a quite puzzling argument to be accommodated in the hypothesis of a carbonium ion mechanism.
Me ‘C==C-CHO* Me’
Mixed orientation has been observed in the case of 1,3-dialkylallene derivatives (6 of Table 5 ) and also in the case of allene 65 which gives a mixture of ring-opened isomers 66 (major) and 68 (minor) by reaction with HCl (Poutsma and Ibarbia, 1970). It is interesting to note that HCl addition to allenes under the conditions used by Jacobs and Johnson ( 5 of Table 5 ) is faster than addition to the corresponding alkynes but slower than addition to the oorresponding 1,3-dienes. The heat of hydrogenation of allene is 71-3 kcal mol-’ more than twice that of propylene (30.1 kcal mol-’). This indicates that the double bond in allenes is somewhat destabilized by the cumulative system. As pointed out by de la Mare and Bolton (1966), the destabilization energy would provide the driving force for a faster protonation of cumulate than of conjugated diene. The lower reactivity indicates that the cations generated by protonation of allenes are less stable than those (allylic) formed from conjugated dienes. Finally, tetramethylallene (7 of Table 6) as well at3 other tetraalkylallenes react with hydrogen halides by an electrophilic mechanism involving exclusively protonation of the central carbon (Bianchini and Guillemonat, 1968; Poutsma and Ibarbia, 1971).
TABLE 6 Homoallenic Participation Substrate
X5 Solvent (HOS) t°C
--0s -la =
v
X Br
HzO (As+)
References
Products
Me-CO
t.a.
44%
32%
lb
ONs MeOH
60
92%
-
lc
ONs HCOzH
60
5%
80%
Id le
ONs MeCOzH OTs MeCOzH
60 100
61% 76%
20% 11%
0
Hanack and Haffner, 1964, 1966 Hanack and Haffner. 1964, 0 1966 Hanack and Haffner. 1964.- w Hanack 1966 and Haffner, 1966 M U
2
-x
3%
6%
Jacobs and Macomber, 1969
P b
!4
U. 2a = = \ t X
OTs
H2O
28%
70%
2b
OTs OTs
MeCOzH MeCOzH
35% 65%
43% 22%
2C
85
Bertrand and Santelli, 1968
11%
Santelli and Bertrand, 1969 Jacobs and Macomber, 1969
4Et.mq 4 y't--Jos
L=vos -= 3a \= V X 3b
ON6 HCOzH OTs MeCOzH
13%
100
14%
31
51% 5%
8%.
7%
a td
a 0 H 0
w
M
Hanack and Haffner, 1966 36% JacobsandMacomber, 1969 P I5 U.
0
4L,
3,5-DNB Dioxan-He0
55%
16%
18%
Jacobs and Macomber, 1969
Ba 5b
(3
7
-x --
X OBs MeCOzH OBs EtOH
-Y 55 55
=v =;=/7tx
77% 95%
11%
11%
-
OTs MeCOzH
37%
34%
OTs MeCOzH
40%
20%
-
19%
Bly el d.,1967
22%
Santelli and Bertrand, 1969
14%
SantelliandBertrand, 1969c
=aR .“i; Q;
*l?
fi
d P 0
R
9
OTs
MeCOzH
(70)
OTs
MeCOzH
(70)
OTs
MeCOzH
=%
4.5
70%
17%
Santelliand Bertrand, 1969c
18%
45%
29%
SantelliandBertrand, 1969~
50 YO
30%
20%
Jacobs and Macomber, 1969
ONs = 1-naphthdenesulphonate;3,5-DNB = 3,5-dinitrobenzoat.e; OTs = p-toluenesulphonate;OBs = p-bromobenmnesulphonate
2u
u
N &3
h
226
GIORGIO M O D E N A A N D UMBERTO TONELLATO
2. Cumulate Double Bond (Homoallenic)Participation An example of addition of vinyl carbonium ion to allene has been discussed in the caae of cyclodimeric products formed from allene and hydrogen halides. Allenic participation may be considered a case of intermolecular addition of an electrophilic carbon to an allenic bond. The participation has been extensively investigated in the caae of homoallenic compounds of type 81 and “homoallenic participation” is the current term for the phenomenon. Earlier reports by Hanack and
Haffner (1964, 1966) were concerned with the products of solvolysis of unsubstituted homoallenic derivatives. The observed formation of cyclopropyl methyl ketone ( 1 4 of Table 6) and the higher yield of this product with solvents of higher dielectric constants and lower nucleophilicity were found consistent with the hypothesis of homoallenic
participation and formation of a cyclopropyl vinyl cation 82, in competition with direct solvolytic substitution. Four-membered cyclic compounds are also obtained from simple homoallenic derivatives (leof Table 6). The effect of substitution of a methyl group for hydrogen at C1of 81 on the product distribution is illustrated by cases 2a-c of Table 6: the yield of cyclized products generally increases. The stereochemistry of the reactions shown under 2a and 2b of Table 6 waa also investigated. Optically active tosylate 83 waa shown to yield the cyclopropyl derivative 84 (two geometric isomers) with virtually complete inversion of configuration at the reaction centre (Bertrand and Santelli, 1968).
V I N Y L CATIONS
227
The “normal ” solvolysis products were found to be racemic or partially retained material depending on the solvent (2a and 2b of Table 5). Therefore, the “normal ” solvolysis may occur either via “free ” carbonium ion or via back-shielded carbonium ions. gem-Dimethylsubstitution at C1of 81prevents the formationof cyclized products and only “normal ” substitution or elimination derivatives are obtained (4 of Table 6). gem-Dimethyl substitution at C2 also prevents the formation of cyclic material but all of the linear compounds obtained are rearranged products in which the gem-dimethyl groups are at C1. The effect of methyl substitution at Cs and C6 on the product distribution is shown under 3 , 6 and 7 of Table 6 and by other cases investigated by Bertrand and Santelli (1968),Santelli and Bertrand (1969 a, b, c), Bly et al. (1967) and Jacobs and Macomber (1969). Extensive kinetic measurements also showed that homoallenic participation is relevant in the transition state of the reaction except for the 1,l-dimethyl substituted compounds where participation is excluded on the basis of both rate data and product distribution. The lower limit of kA/k8in the case of simple homoallenic tosylate in acetic s acid at 65” is ca unity. The effect of methyl substitution on ~ A ’ relative to unsubstituted 81 in acetic acid at 65” is illustrated below (Jacobs and >C=c=C-C-~--OTe 1 3.2 3.8 -
(one Me)
i I
2 h (two Me)
I lh \ (one Me)
Macomber, 1969). The larger anchimeric assistance effect on rate is observed in the case of gem-dimethyl substitution at Cz. Bly and Koock (1969) estimated rate enhancements due to homoallenic participation of 8-3-58 x los for the 2,2-dimethyl brosylate (5a of Table 6) and as large as 1.7 x lo6 in the case of 2,2,5,5-tetramethyl substituted brosylate in acetic acid at 76”. It is noteworthy that these rate enhancements are larger by factor of at least 10 than in the cme of homoallylic participation in related compounds. From the large amount of .rr-electron participation it is evident that considerable charge delocalieation is developed in the transition state of the assisted solvolysis. This may be represented as 89. The multiplicity of products ranging from linear to three- and four-membered cyclic derivatives may be accommodated by the idea of a non-identical “ bicyclobutonium ” cationic intermediate 90 (Jacobs and Macomber, 1969 ; Santelli and Bertrand, 1 9 6 9 ~ ) .Interconversion of various bicyclobutonium ions and possibly conversion of these to various other cations,
228
QIORQIO M O D E N A A N D U M B E R T O T O N E L L A T O
including vinyl cations (depending on the type of methyl substitution) is illustrated in Scheme 8.
Solvent attack formally occurs exclusively at C2 when two methyl groups are attached to it, reasonably via formation of a tertiav carbonium ion. Attack at C8 leads to four-membered cyclic products and attack at C4to three-membered cyclic products possibly via 82. Elimination of a proton may also occur to give a cyclopropylacetylene derivative, as observed in case 3b and from related compounds (Jacobs and Macomber, 1969). With homoallenic derivatives 91 (under 8 and 9 of Table 6), cyclobutane derivatives are formed but anchimeric assistance effects were not observed. Obviously, homoallenic participation may occur only
from axial conformation 91b which is less stable than 91a in particular when R =Me. Therefore, the rate of solvolysis is controlled by the rate of interconversion of the two conformers.
229
VINYL CATIONS
3. Addition of Halogens
The addition of Br2,Clzand BrCl to the simple allene has been investigated by Peer (1962) and discussed by recent reviewers (Griesbaum, 1966; Taylor, 1967). The primary products of addition of bromine, irrespective of the solvent used is the 1,2-dibromidewhich adds a second bromine molecule at a much slower rate than allene. The nature of the products resulting from addition of chlorine depends on the solvent : in inert media the 1,2-dichlorideis formed, together with some propargyl chloride 92, whereas in acetic acid the solvent-incorporated adduct 93
is obtained. The same dependence of product distribution on solvent is observed in the case of addition of BrC1, as indicated below :
+ - , CHz=C=CHz
+ BrCl
\
CHz=CBr-CH&l
HOAC
~
CHz-CBr
- CHzOAc + (93)
The results would indicate initial formation of a bridged cation 94 which may open in an SN1process to generate an ally1 cation and react with all available anions according to equation (16) (Y =Br, C1, OAc).
(94)
On the other hand, the formation of propargyl chloride has been explained by assuming the formation of species 94a which goes to product via a “four-centre elimination reaction’’ (equation 16). As an
230
GIOROIO MODENA AND UMBERTO TONELLATO
alternative hypothesis, ion 94 may undergo an SN2-typering-opening process or a "fragmentation "-type elimination reaction whose transition state may be of type 95.
Similar arguments can be used to explain the products observed in the addition of halogens to symmetrical and unsymmetrical alkylallenes (Fedorova, 1963; Fedorova and Petrov, 1961). More recently the formation of 96 (80-90%), 97 (10-15%), and 98 (143%) was observed by Poutsma (1968) for the chlorination of 3-methyl-l,2-butadiene under a variety of conditions in the presence of oxygen to inhibit free-radical side processes. C1\ /Me Me,
Me\CC/CHB CH2/ \C1
Me
,c=c\
/CHzC1 c1
Me'
C
>C=CH2
c1
Poutsma and Ibarbia's (1971) work on 2-methyl-l-(tetramethylcyc1opropylidene)propene (65) showed that Chlorination (in CC14) yields derivative 99 exclusively. Bromination of 65 (methanol, 25') gives a more complex mixture of products ranging from the bromine analogue of 99 to compounds of type 100. Terminal electrophilic attack or, a t least if bridged species are involved, SN1-typering opening is apparently involved in these cases. Me
I
I Cl-c-kCCC
I
Me
I
Me Me
Me
Me /CH2
\Me
Me (99)
I
Br-C-C5C-C-C-Y
I
I ,I I I
Me Me
Me (100)
(Y=Br,OMe)
4 . Addition of Sulphenyl Chlorides
Addition of sulphenyl chlorides to allene gives anti-Markownikoff adducts, as shown in equation (17) (Jacobs and Johnson, 1960; Mueller
231
V I N Y L (IATIONS
and Butler, 1968). Addition of 2,4-dinitrobenzenesulphenylchloride to the optically active allene 101 gives the optically active product 102 (Jacobs et d.,1967) which indicates that a bridged cation of type 103
is very likely to be an intermediate. The anti-Markownikoff orientation suggests that ring opening occurs by nucleophilic attack at the sp3 carbon, although unimolecular opening to generate an ally1 cation cannot be excluded in some cases. Bridged ions have been also suggested as intermediates in the reactions of allenes with peracids and mercuric salts (for a recent discussion, see Poutsma and Ibarbia, 197 1). C. Heterolytic Fission of Bonds Attached to a Vinyl Carbon Atom Vinyl cations can be generated by heterolytic fission of the C-X bond of an appropriate vinyl compound (equation 18). There is a growing number of recent reports that this type of reaction may occur also in
'
c,=c
.
\
A -
)C=d
-+
X-
__f
products
(18)
mild conditions, depending on the nature of X and on the structure of the vinyl moiety. Exceptionally good leaving groups, such as trifluoromethane sulphonate (triflate), fluoromethanesulphonate, and 2,4,6trinitrobenzenesulphonate, have proved to be of great help in the study of the unimolecular reactivity of vinyl derivatives. Loss of nitrogen from vinyldiazonium ion precursors is a special case of generation of vinyl cations according to equation (18) and will be considered separately. 1. Nitrogen Loss from Vinyldiazonium Ions Curtin et al. (1966) tried to prove the existence of vinyl diazonium
ions, which were thought to be of intermediate stability between that of aliphatic and aromatic diazonium ions from obvious considerations. However, the authors' attempts to intercept these intermediates in the deamination of vinylamine failed.*
* Stable vinyldiazonium salts have reoently been isolated (Bott, 1970).
232 OIOROIO MODENA A N D UMBERTO TONELLATO
+
I I t t
T
V I N Y L CATIONS
233
The idea of vinyldiazonium ion precursors of vinyl cations, seems to apply to the products obtained from vinylamine (104) and nitrosyl chloride in dichloromethane (Scheme 9). Vinyl cation 105 may undergo rearrangement to the more stable cation 106 before being intercepted by nucleophilic species. The non-stereospecific addition of C1- to 106 under kinetic control accords with the idea of a free linear cation intermediate. However, a t least in the case of the reaction of the vinylamine 107 with isoamyl nitrite in cyclohexene the main reaction product is that expected from a carbene, 108, rather than from a vinyl cation intermediate. The formation of vinyl oations by deamination of vinyl amines in neutral media, is therefore not established and the possibility that a carbene intermediate is actually involved must be considered. Thus, the products of the alkaline decomposition of N-nitroso-oxazolidones 109 were earlier explained (Newman and Kutner, 1961; Newman and Weinberg, 1966) in terms of a mechanistic scheme involving vinyl cation intermediates (equation 19). Such a scheme has recently been
re-proposed by Newman and Beard (1970). However, at least in some conditions, Hine’s (1964) original hypothesis of the intermediacy of species like 110 and 111 (when R’ = H or alkyl, respectively) was proven to be fully consistent by carbene-trapping experiments (Newman and Okorodudu, 1968, 1969). R >C=C:
I
B-C-C\
I
R
.. R’
(111) (B=bm)
Convincing evidence that vinyl cations, possibly resulting from diazonium ion precursors, are the actual intermediates of the acidcatalysed decomposition of 1-aryl-3-vinyl triazene derivatives 112 (equation 20) has been discussed by Jones and Miller (1967).
234
Q I O R O I O M O D E N A A N D UMBERTO TONELLATO
L
Rearranged products of type R,YC=CR,R, are obtained from the vinyl triazene 112 when R2is an aryl group and R, is hydrogen or methyl and also when R2 is tolyl and R1 is phenyl. The acid-catalysed decomposition of vinyl triazene which occurs in very mild conditions has been exploited as a useful method for the synthesis of vinyl sulphonates (Jones and Maness, 1969, 1970; Modena and Tonellato 1971). The photolysis of pyrazolenine derivatives 113 yields mainly allene 114 and 1,3-diene derivatives 115. According to Day and Whiting
(113)
(114)
(115)
(1967),the reaction does not involve decomposition of a possible diazoalkene intermediate but rather the formation of vinyl cation 116 which may be stabilized by allylic resonance.
(116)
The products obtained in the nitrous acid deamination of amine 117 in acetic acid are shown in Scheme 10 (Nishimura et al., 1967, 1970). Nitrogen loss occurs from a primary saturated carbon atom and this case is therefore not strictly pertinent to this section. To account for the formation of ketone 118 among the reaction products the authors
VINYL UATIONS
236
SOHEME 10
indicate the cyclopropylvinyl cation as the possible intermediate precursor and emphasize the fact that it must be ultimately generated by cleavage of carbon-carbon bond. The product distribution is reminiscent of that observed in the solvolysis of homoallenic (see section 11,B, 2) and a-cyclopropylvinyl derivatives (see below) and the nature of the intermediate(s) is likely to be similar. 2. solvolysis
Solvolysis may proceed via vinyl cations (equation 21) to give substitution (SN1-type)products and/or, when a ,&hydrogen is available, elimination (El-type) products.
Reactions of vinyl derivatives under solvolytic conditions may occur d a mechanisms other than a SN1-(El)-typeprocess and yet lead to the
same type of products. The following mechanistic possibilities are to be considered: (i) in acidic solvents, a proton-initiated addition-elimination process which may lead to products of formal substitution (equation 22) ;
(ii)in basic media, the nucleophilic processes (Rappoport, 1969;Modena, 1971)of displacement (equation 23), of nucleophilic addition-elimination
TABLE7
AS*
AHS Ri
Ra
R3
H Me
Ph
Ph H Me Me
1 2 3a 3b 4a 4b 4c 4d 6a 6b 6 7
Me Me cM03
8
C H a . CHa '3%
9 10 11 12 13
16 17 18
Me
Me Me Me Me Me
Me
.
.
H H H Me Me Me Me Me Me Ph H
CHa (C3a)a. m a C H a . (CHr)a.CHa CHa.(CHa)s.CHs CH;(CHs)iCHa CH=CHa H
CH=CMea H CH==CMea Me CH=CMea H
X U
Solvent*
t"C k / m x 104 Kcal.mo1-1
E-W 4: 1 E-W 4:l E-W4:1 M-W 1:l E-W 4: 1 E-W1:1 M-W 1:l M-W 1:l E-W1:l N-M9:1 E-W 4: 1 E-W1:1 E-W 1:1 E-W 1:l E-W 1:l E-W 1:l E-W 1:l E-W 4:l
126 76 76 130 76 100 130 130 100 100
Me H H H
OTf OTf OTf OTs OTf OTf OTS OBS OTf TNBS OTf OTf OTf OTf OTf OTf OTf Br
160
5.77 61.1 0.016
H
Br
E-W 4: 1
100
0.016
26.9
H
Br
E-W 4: 1
100
0.031
24.9
H
Br Br Br
E-W 4: 1 E-W 4: 1 E-W 4: 1
100 100 100
H H H H Me Me Ph H H H
H Me
76 100 100 100 100 100
0-49 6.29 0.71 0.226 17.7 0.07 0.27 29.0 0.66
24.7 23.8
9.14 2 x 10-4 0-0066
23-7 33 31.2 28.1 26.3 26.7
0.064
8.29 11.1 4.61
25.3
26.2
23.3 22.6 24.2
cal. mol-1
deg.-1
-7.7' -6.36
References
Imhoff el al., 1970 Stmg and Summerville,1969 Stang mnd Summerville, 1969
Peterson and Indelicato, 1969 Stang and Summerville, 1969 Heifer et al., 1971 Peterson and Indelicato, 1969 Peterson and Indelicato, 1969 Pfeifer et al., 1971 - 12 Burighel et al., 1971 Imhoff et al., 1970 4.8f Martinez et al., 1970 (- 6.9)g Pfeifer et al., 1970 Pfeifer et al., 1970 -4.1 7.7 Pfeifer et al., 1970 - 3.9h Pfeifer el al., 1970 - 2.4' Pfeifer et al., 1970 Grob and Spaar, 1970
- 7.7'
-
- 16.2 - 17.2 - 10.9 - 12.1 - 9.6
Grob and Spaar, 1970 Grob and Spaar, 1970 Grob and Spaar, 1969,1970 Grob and Spaar, 1969,1970 Grob and Spcutr, 1969,1970
19 20
C H 4 M e a Me H c-CsH5
Me H
Br I
Me I 21 c - C ~ H ~ H Me H I 22 c-CsH5 H H Br 23 (X) .CsH4. Br H H CeH5 24 OTs H H C&5 26 OTf C&5 H H 26a 26b Ca5 Me H Br 27 28 Ca5 H Me Br H Br 29 Me0 . C a 4 Me0 .C6H4 H Me0 .CeH4 Br 30 Me0 C a 4 CsHs Cs&.(Y) I 31 (X).CeH4 C&5 Br ceH5 C6H5 32 CeH5 OTS C&5 C&5 33 C&5 OsSF Ca5 34 C&5 C&€5 CeH5 CeH5 OTf 35 36 C&5 Ca5 C a 5 TNBS Ca5 C&5 CaH5 TNBS 37 38MeO.CeH5 Me0 C a 6 M e 0 .C6H5 Br 39MeO.CZH5MeO . C a 5 Me0 C&5 C1 40&feO.C&5 MeO.CeH5 Me0 .CsH5 osSc6H4 .X Me MeS Me TNBS 41 .CeH4S (Z) .C&4 TNBS 42 (X). C a 4 (Y) 43 ca5 ca5s csH5 TNBS
.
.
.
E-W 4:l M-W
100 150
11.8
HOAc (Ag+) 25 (1)C HOAc (Ag+) 25 (9.5)C E-W 4: 1 loo 4.2 x 10-56 0.305 189.5 DMF-W 7:3 130 1.74 M-W 1:l 71.8 5-6 HOAc E-W 4: 1 25 1.12 i-P(AgOsCCF3) 25 i-P (Ago2 CCF3) 25 HOAc 120-2 0.825 HOAc 120.2 0-042 DMF-W 7:3 189.5 0.084d 189-5 0.213 DMI?-W7:3 25 1.4~10-6 HOAC 0-058 HOAC 25 HOAC 25 0.112 HOAC 25 0-014 N-M 19:l 25 0.183 E-W4:1 120 4-08 E-W 4: 1 120 0.071 A-W7:3 N-M9:1 N N-M 19: 1
75 25 25 25
1.41d 4-1 3.44s 3-76
23.4
-9.8
33.7
- 7.8
23-7 26.4 22.4
-0 4 - 1.5
23.5
- 16.3
23-8 22.6 23.6 24.7
- 20.5 - 6.5, -2.0
21.2 26.8 (25-27) 21.6 22.8
-4.7 - 20 - 14 (
Grob and S p m , 1969,1970 Sherrod and Bergman, 1969 Hanmk and Biissler, 1969 Kelsey and Bergman, 1970 Kelsey and Bergman, 1970 Grob and Cseh, 1964 Miller and Kaufman, 1968 Peterson and Indelicato, 1969 Jones and Maness, 1970 Hargrove et al., 1970 Kernaghan and Hof€mann, 1970 K e r n a g h and HoEmann, 1970 Rappoport and Atidia, 1970 Rappoport and Atidia, 1970 Miller and Kaufman, 1968 Miller and Kaufmm, 1968 Jones and Maness, 1970 Jones and Maness, 1970 Jones and Maness, 1970 Modena and Tonellato, 1971a Modena and Tonellato, 1969 Rappoport and Gal, 1969 Rappoport and Gal, 1969
-5 - 1) Rappoport and Kaspi, 1970 -
1 +2
Burighel et uZ., 1971 Modena and Tonellato, 1971a Modena and Tonellato, 19718
5 OTf =trifluommethanesulphonate; OTs=p-tolueneeulphonah; OBs =p-bromobenzenesdphonate; TNBS =2 , 4 , 6 - t r i n i t r o b sulphonate. A =aaetone;DMF =dimethylformamide ;E =ethanol; M =methanol; N =nitromethane ;p =penhne c relative mtea d rste constant for the unsubstituted (X=Y==ZEH) term rs w 6at25O; fat60-4'; QinE-W3:2; bat75.2"; Cat49.4O; 5at61.5O; kat67O 4
*
238
G I O R G I O MODENA A N D UMBERTO TONELLATO
(equation 24), and, when a /3-hydrogen is available, of base-catalysed
\
; C - G p/x
\
+HX
>Ch=C
H
elimination (equation 25). Mechanisms in equations (231, (24), and (26) ‘C==C<x
H’
-%
4 k C L +SOH
may be distinguished from the mechanism via vinyl cation, on the basis of the effect on rate and/or product distribution due to changes in lyate ion or added base concentration. Reaction according to equation (22) is also distinguishable from that involving vinyl cations but it must be carefully taken into account before reaching a definite conclusion on the solvolysis mechanism (Hanack, 1970). I n answer to a paper by Schubert and Barfnecht (1970) which is very critical of the current mechanistic interpretation of “some ” solvolytic reactions of vinyl compounds, Rappoport et al., (1970) gave 20 distinctive criteria for mechanisms of equations (21) and (22), in particular the strong rate dependence upon proton concentration in the reaction media. The products of solvolysis in aqueous solvents are generally ketones (119) presumably deriving from the corresponding enols (120). Vinyl ethers (121) and esters (122) are obtained in non-aqueous solvents or in ,OH
R R>H-CO-R
”>&C R
‘R
R
,OR’
R\
,C=C
,OZCR’
‘R
mixtures of low water content, usually together with relatively large amounts of ketones (130), often deriving from solvolysis of labile ethers or esters. Alkynes are also formed when vinyl hydrogens are available. Table 7 shows the relevant examples of substrates investigated, listed according to structural criteria, the solvolytic conditions used and, when available, the first-order rate coefficients and activation parameters. The reactivity of vinyl derivatives with equal leaving groups strongly depends on the nature of the a-residue. Early reports emphasized the importance of an ol-aryl group for an SN1-typereaction to occur.
239
VINYL CATIONS
In fact, in the absence of a-substituents that would be able to delocalize the positive charge, vinyl compounds are extremely unreactive. This is illustrated by the observation (Kaufman and Miller, 1969) that 8-bromostyrene (119), in the presence of silver nitrate, does not solvolyze
but undergoes an unexpected substitution reaction (equation 26) via a free-radical process. A similar reaction was found to occur with other a-unsubstituted vinyl bromides. However, by the use of the exceedingly reactive trifluoromethanesulphonate as leaving group, Imhoff et al. (1970)were able to solvolyse the 2,2-diphenylvinyl derivative, (1 of Table 7) and observed the formation of aldehyde 120 (96%) and of diphenylacetylene (4%0). Ph ph>CH *COH
+
Ph-C=C
/H ‘Ph
Although the latter product may suggest the formation of the rearranged vinyl cation 106, the main reaction mechanism may well be one of addition-elimination, rather than ionization. a-Methylvinyl compounds, (2, 3 and 4 of Table 7) react by what has been d e h e d as a “delicate balance between concerted elimination and unimolecular ionization ” (Stang and Summerville, 1969). The presence of a 8-hydrogen trans to the leaving group favours the concerted elimination route. This is shown for the trans 2-buten-2-yl triflate (3a of Table 7) by the quantitative yield of 2-butyne, by the primary !-deuterium kinetic effect (kH/kD=2.09)and by the rate ratio of 40 relative to the cis isomer (4a of Table 7). From this derivative, the products observed [2-butyne (68%), 2-butanone (33%) and methylallene (9%)] and the 8-deuterium kinetic effect (kH/k, = 1.20) are those expected for an E1SNI process. A similar picture and virtually identical conclusion follow from the study of related substrates (3b, 4c and 4d of Table 7). (Peterson and Indelicato, 1969). l-Methy1-2,2-diphenylvinyltriflate (6 of Table 7) reacts to give exclusively the rearranged ketone 121. The formation of 121 has been explained by assuming that a 8-phenyl group migrates across the
240
QIORQIO MODENA A N D UMBERTO TONELLATO
originally formed vinyl cation 122 to generate the more stable cr-phenylvinyl cation 123. No rearrangement WM found to occur in the solvolysis of the trimethylvinyl triflate (5a of Table 7)as indicated by the products ,Me \Ph
Ph. CO *CH
Ph c ‘C=CMe ph’
+
Ph-C=C
,Me ‘Ph
of solvolysis of the deuteriated compound shown below. This is expected in view of the low migrating ability of the methyl group and of the fact that migration would lead to an energetically identical cation. 4:1 p
MeaCkC(0Tf)CDs
\
CDs’COsrD
t
M~~CD.CO.CD~+M~~C=CDZ
Me&=C(O&.CD3)CD3
The intermediacy of a vinyl cation aa intermediate in the solvolysis of 1-t-butylvinyl trifluoromethanesulphonate (124) (7 of Table 7) was mainly inferred on the basis of the large spectrum of rearranged products (Scheme 11; cf. 3 of Table 2). SCHEME 11
L
1.6%
1.6%
9%
t-Butylvinyl cation 12 (see sect. 11, A, 1) is likely to be involved: by methyl migration this may readily go over into the more stable allylic carbonium ion precursor of the rearranged material. 1-Adamantyl triflate (125) was investigated by Imhoff et al. in order
VINYL CATIONS
241
to compare the products arising from the 1-adamantylvinyl cation from solvolysis and from electrophilic additions (seesection IIA2). I n ethanolwater and buffered acetic acid, adamantylacetylene (126) is the main product and an E2 elimination mechanism has been suggested for its formation. I n less basic solvents (CFs* CH,OH), rearranged materia1 (127 and 27) was obtained in high yields together with minor amounts of unrearranged products (126 and 25).
Cycloalkenyl derivatives show a great difference in reactivity depending on the ring size. These compounds (8-12 of Table 7) are believed to react by an ionization mechanism to give exclusively cycloalkanones; the addition-elimination route was excluded, at least for the acetolysis of 1-cyclohexenyl trifluoromethanesulphonate, by the observation that little or no deuterium uptake in the products occurs in deuteriated acetic acid. Since in cyclic systems, vinyl cations are more or less rigidly held in “bent ” arrangements, the decrease in reactivity with decreasing ring size is considered to be evidence of the instability of non-linear vinyl cations (Pfeifer et al., 1971). The presence of an a-cyclopropyl residue is rather effective and apparently more effective than that of an a-phenyl group in promoting the ionization of vinyl compounds (Sherrod and Bergman, 1969; Hanack and Biissler, 1969). The intermediacy of a cyclopropylvinyl cation which may be trapped by nucleophiles, undergo proton elimination or rearrangement via ring enlargement was proven by the product distribution (Scheme 12) obtained from the two isomeric l-cyclopropyl-1-iodopropenes (21, 22 of Table 7). The product distribution is virtually the same from both isomers and it is particularly significant that the ratio of cis and trans vinyl acetates is ca. 1 :1 under kinetic control. The product distribution is close to that obtained in the solvolysis of homoallenic derivatives (see section 11,B, 2) and in the acid addition to cyclopropylallene derivatives (see section 11, B, 1 ; see also Sherrod and Bergman, 197121, b). The fact that the cis isomer reacts in the presence of Ag+ at least 9.5 times faster than the trans isomer excludes the possibility of the incursion of a concerted elimination and is explained by the authors in terms of relief of steric strain as the cation from the iodide approaches linearity.
242
GIORGIO M O D E N A A N D UMBERTO TONELLATO r
M e T H G
+ Mc--CH&CH(CH~)~OAo
OAc OAc
10-13%
0.6%
2-4%
Conjugated 2-bromodiene derivatives 128 display a remarkable reactivity under the conditions used by Grob and Spaar (1969, 1970) (13-19 of Table 7). The rate is insensitive to added triethylamine, which excludes the occurrence of acid-catalysed hydration or base-induced
elimination processes. The hypothesis of an ionization mechanism is supported by the high sensitivity of the reaction rate to the ionizing power of the solvent, the m value (Winstein et al. 1961) being ca. unity (16 of Table 7). The rate of solvolysis strongly depends on the methyl substitution a t the terminal C4 atom. One methyl (R'l or R', of 128; 14 and 16 of Table 7) increases the rate by a factor of lo2 and two gemmethyl groups (16-19 of Table 7) by a factor of lo4. The high reactivity of these compounds is explained in terms of stabilization of the primarily formed vinyl cation 129a by limiting structures like 129b.
H (129a)
(129b)
VINYL CATIONS
243
Overlap of the neighbuuring double bond with the developing p n orbital occurs only in a non-planar configuration whose achievement is possible by a 90" rotation around the C,-C, bond of the linear cation 129. When such a geometry cannot be attained, as in the case of 2-
bromocyclohexa-1,3-diene derivatives 130, the reaction of solvolysis does not occur even in the presence of silver ions or a t temperatures as high as 180' (Grob and Pfaendler, 1970). On the other hand, mesomeric stabilization through rotation is favoured by methyl substitutions (in
R Q R (130)
particular a t C,) which, by steric interactions with the 2-bromine or with substituents at C1,cause considerable departure from planarity in the ground states of the reagents 128. The intermediacy of a cation of structure 129 is confirmed by the observation of products 131 (main in all cwes, exclusive in the case 19 of Table 7), 132 and 133.
or-Arylvinyl derivatives have been more extensively studied. Grob and Cseh (1964) in their pioneering work investigated the solvolysis in ethanol-water of a series of substituted (fromp-NOzto p-NH,) a-bromostyrenes (23 of Table 7). With the exception of the p-nitro derivative which reacts only in the presence of added base by a bimolecular elimination reaction, all the compounds investigated undergo solvolysis by what has been indicated as a clear unimolecular mechanism yielding hydrolysis products (acetophenones) and elimination products (acety-
244
U I O R G I O MODENA A N D UMBERTO TONELLATO
lenes). The hypothesis was firmly based on the insensitivity of rates to added base, on the rate increase with increasing polarity of the solvent and on the very large effect of substituents in the phenyl ring (p = - 6.6 by use of o+;Grob and Pfaendler, 1971). A different mechanism involving an acid-catalysed addition-elimination process in the particular case of the p-amino derivative was claimed by Schubert and Barfnecht (1970). However, Grob (1971)and Grob and Pfaendler (1971) were able to dismiss their criticism by a careful study of the solvolysis in ethanol-water and dioxan-water at various pH and buffer concentrations. In fact, the rates are unaffected by change of acid and buffer concentration in the apparent pH range from 3 to 12. Moreover the ratio of acetylene to acetophenone derivatives increases significantly with pH and buffer concentration, thus indicating a different sensitivity of the intermediate vinyl cation to undergo attack at carbon or at ,&hydrogen with different nucleophiles. The products from cis and trans 1-phenylpropenyl bromides (27 and 28 of Table 7) formed in a suspension of silver trifluoroacetate in isopentane were found to be 1-phenylpropyne and cis and trans l-phenylpropenyl trifluoroacetates. The yield of the elimination products is virtually the same (53-57%) from both isomers, a fact which would rule out the concurrence of an E2 mechanism. The substitution reaction involves overall ( 13%) retention of configuration under kinetic control and the stereochemical course, at variance with that usually observed, i.e. “racemization” (Rappoport and Apeloig, 1969; Kelsey and Bergman 1970), is tentatively explained in terms of surface interaction between the vinyl cation and the trifluoroacetate ion in the crystal lattice. The higher retention (ca. 30%) of configuration observed in the substitution products from the cis bromide and silver trifluoroacetate in ether (where the reaction is homogeneous) has been suggested to be due to nucleophilic intervention of the basic solvent and double inversion at the reaction centre. Such an argument should however be substantiated by the observation of the stereochemistry of the trans isomer. Moreover, although kinetic data were not reported, it is apparent that the cis isomer is substantially more reactive than the trans derivative, as found in analogous cases by Kelsey and Bergman, (1970) and by Rappoport and Atidia (1970). The latter authors found that the solvolysis of 1,2-dianisylvinyl bromides occurs by way of a mixed E2 and E l mechanism in ethanolwater mixtures in the presence of NaOH and by an E l mechanism in acetic acid in the presence of NaOAc (cases 29 and 30 of Table 7). Under these conditions the cis isomer reacts ca. 20 times faster than the trans isomer and the difference is explained in terms of the relief of steric
V I N Y L OATIONS
245
strain between the two anisyl groups as the cis derivative approaches linearity in the transition state. The alternative hypothesis of /3-anisyl participation in the transition state seems to be excluded by other results.
Perhaps little attention has been devoted to a disturbing report by Frydman and Mazur (1970) concerning, in general, the reactivity of vinyl tosylates which possess at least one /I-hydrogen and, in particular, that of the isomeric 1,%diphenylvinyltosylates (134). These compounds, when heated at temperatures of at least 100’ or irradiated with a tungsten filament lamp, are converted into the j?-ketosulphone 135 and the mechanism is a free-radical one as indicated in Scheme 13. SCHEME 13
This type of reaction is accelerated by small amounts of peroxide and may occur both in inert and solvolytic solvents, such as methanol and 2-propanol. Triarylvinyl compounds provide a safer framework for the study of SN1processes and have been the subject of a number of full reports. Triphenylvinyl and substituted triphenylvinyl iodides have been investigated by Miller and Kaufman (1968) (31 of Table 7). The products obtained are the expected ketones in essentially quantitative yields, the reaction rate is not enhanced by added nucleophiles, is depressed by the common-ion effect, and the p value relative to substituents on the or-phenyl ring (p-MeO, H, p-C1) is -3.6 by the use of u+. The effect of substituents in the g-ghenyl ring tmns to the leaving group (p-MeO, H) is very small corresponding to a p value of -0.6 by the use of u constants. Several triphenylvinyl sulphonates were also investigated (cases 33-37 of Table 7) and the reaction mechanism shown to be an Sxl-type
246
UIORUIO MODENA AND U M B E R T O TONELLATO
processes in a number of solvolytic conditions (Jones and Maness, 1969, 1970;Modena and Tonellato, 1971). Here again, the rate is not changed by changing the concentration of added base, kHos/kDosis unity in acetic acid (for the fluorosulphonate and tosylate) and virtually quantitative yields of substitution products were obtained in acetic acid and ethanol from the more reactive sulphonates. The large leaving group effect observed in the acetolysis is one more argument in favour of the proposed mechanism. The relative rates calculated at 25" of p-toluenesulphonate, 2,4,6-trinitrobenzenesulphonate, fluorosulphonate and trifluoromethanesulphonate are 1 :25 000 : 42 000 :81 000. The reactivity of trianisyl derivatives under solvolytic conditions has been thoroughly investigated. The bromide and chloride (38and 39 of Table 7) were shown to react in buffered solutions at a rate which is independent of the concentration of added base or of nucleophiles, such as thiolate ions. The ratio kB,/ka = 68, the Grunwald and Winstein m values of 0.34-0-63in ethanol-water mixtures, the pronounced commonion rate depression observed also during a single kinetic run in acetic acid (Rappoport and Gal, 1970) were presented as convincing evidence that the solvolysis of these compounds proceeds via rather stable and selective vinyl cations. Similar conclusions were reached for the solvolysis of substituted arenesulphonates in buffered aqueous acetone (see 40 of Table 7), mainly on the basis of the leaving-group effects. The p value for changes in the leaving group is + 1-67and kOTs/kBris 32 in acetone-water 7 :3 at 76". The stereochemical course of the solvolysis of isomers 136 and 137 was determined by Rappoport and Apeloig ( 1969). The reaction of both isomers in ethanol-water 4 :1 in the presence of thiolate ions, in acetic acid in the presence of acetate and chloride ions and in dimethylformamide in the presence of chloride ion affords cis and trans substitution products in the ratio 1 :1. The observed "racemization " indicates that the intermediate precursor of the products is a linear dissociated vinyl cation. The remarkable reactivity of trianisyl derivatives is certainly due to the u-anisyl group and not to the /l-groups. The rate of compound 138 relative to 139 in buffered acetone-water 7 : 3 at 120" is 830 whereas
(136)
(137)
the relative rate of 140, 141, and 142 in ethanol-water 4 :1 at 120' are 1 :0*7:1-7.
247
VINYL CATIONS
The very small effect due to changes at the p-carbon is a significant point since it virtually excludes the possibility of S-aryl anchimeric assistance effects. PhzC :C(0l's)Ph
PhzC :C(0Ts)An HaC :C(Br)An PhzC :C(Br)An
(138)
(139)
(140)
(141)
An& :C(Br)An (142)
The absence of such an effect has been definitely excluded by Rappoport and Apeloig (1970) by the observation that the two geometric isomers 143 and 144 react at virtually the same rate in ethanol-water 4 : 1 and acetic acid at 120" as well as in formic acid-acetic acid 1 : 1 at 99.7". Moreover, the rate of 143 and 144 is intermediate between that of 141 and 142, as expected for the additivity of purely inductive effects. An,
,OTs
Ph
'An
/c=c
Ph>C=C An
/OTY
'An
As a general rule, the effect on rate due to /? substitution with alkyl or aryl groups in the vinyl system is very sinall even when the presence of a bulky a-group would lead to an expected rate increase due to steric relief upon ionization. This effect is likely to be counterbalanced by (i) a hyperconjugative effect when hydrogen(s) is bound at the /?-carbon; (ii) steric hindrance to solvation. Argument (i)is supported by the fact that in the study of the solvolysis of a-stpyl trifluoromethanesulphonate (26b of Table 7) which occurs by all the evidence via a SN1-Elmechanism, Hargrove et al. (1970) measured the S-deuterium isotope effect, kIIIC=c(oTf)ph/k&c=C(OTf)ph= 1.42 at 25". This value is more than twice that observed in the generation of trisubstituted carbonium ions (Shiner et al., 1968) and it has been explained as due to the fact that in the transition state leading to the a-styryl cation the C,-H bond is closer to the vacant p orbital than in the saturated analog and ideally situated for hyperconjugative overlap (see Fig. 1). The argument (ii) has been advocated by Miller and Kaufmann (1968) ; Rappoport and Gal (1970) and Jones and Maness (1970). The relative rates of acetolysis at 120" and the Grunwald-Winstein m values reported by Rappoport and Gai (1970) for the compounds shown in Scheme 14 are illustrative. The rather large rate difference between the cis and trans dianisylvinyl derivatives (affected by a comparable degree of steric hindrance to solvation as indicated by the similar m values) is likely to be due to 9
248
G I O R G I O MODENA A N D UMBERTO TONELLATO
relief of steric strain. Such an effect should be magnified on going from the monoanisyl to the trianisylvinyl derivative also in view of the favourable inductive effects but it is almost cancelled out by the increasing steric hindrance to solvation as revealed by the large decrease in the rn values. SCHEME14
X-rel.
1
0.25
4.9
4.0
m
0-71
0.67
0.63
0.4
The effect of /I-methyl substitution in the case of 1-cyclohexenyl derivatives (Pfeifer et al., 1971) is an interesting case which is not entirely understood. The trifluoromethanesulphonate 145 in ethanolwater 3 :2 at 128" gives mainly rearranged cyclopentyl derivatives. The ten-fold rate increase due to /?-methyl substitution (see 9 and I 0 of llie I
Me I
(60%)
(40%)
(10%)
Table 7) is substantially larger than that observed in acyclic derivatives (4b and 5a of Table 7) and could indicate anchimeric assistance effects. The authors, however, tend to exclude this possibility on the basis of the large rate enhancement ( lo3 relative to cyclohexenyl triflate) observed in the case of compound 146 which gives mainly rearranged products formed by methyl shift and none by ring contraction.
Me
Vinyl derivatives 147 which possess a /?-arylthio (or alkylthio) group trans to the leaving group are a special case which has been extensively
investigated.
The rate of 1,2-diaryl-2-arylthiovinyl2,4,6-trinitro-
249
V I N Y L CATIONS
benzenesulphonate was found to be affected by positive solvent and salt-effects, by the “special” salt effect (Winstein et al. 1954) in acetic acid in the presence of lithium perchlorate, and common-ion depression in the presence of lithium trinitrobenzenesulphonate. Where 35S
lithium trinitrobenzenesulphoiiate was used, the unreacted vinyl sulphonate was found to incorporate radioactivity. In nitromethane and nitromethane-methanol 19 : 1 at 25” the rate is virtually the same but the product distribution is dramatically different: in the former solvent benzo[b]thiophen, 148, or benzil, 149, derivatives are obtained, in the latter vinyl methyl ethers 150 (Modena et al., 1968; Capozzi et al., l971a, b, c; Modena and Tonellato 1971a, b).
A11 this leaves little doubt about the S,l-type reactivity of compounds 147. However, the actual nature of the intermediate does not correspond to that of a digonal vinyl cation. The following facts are indicative. The solvolysis in acetone-methanol mixtures (equation 27) and in acetic acid is entirely stereospecific and retention of the t r a n s configuration is iiivolved. Migration of the arylthio residue was found to occur in the Tol, PhS’
,03STNB
c=c ‘To1
acetonc-nietlmnol4:1
25”
Tol,
,OMe
PhS
To1
/c=c,
(TNB =2,4,6- t,rinitrophenyl)
inethsnolysis reaction of both structural isomers (151 and 152) shown in Scheme 15 (Modena and Tonellato 1971b). Virtually complete scrambling of the label between the two ethylenic carbons was observed in the products of reaction of 1- 14C-1,Z-diphenyl-2-phenylthiovinyl brosylate (Capozzi et al., 1969).
250
GIORGIO MODENA A N D UMBERTO TONELLATO
It has been concluded that the cationic intermediate involved in the reaction of esters 147 has a bridged structure indicated as that of a thiirenium ion (53). SCHEME15
n (53)
Bridging by nucleophilic intervention of the sulphur atom occurs in most cases in the transition state of the reaction. The kinetic results for the most relevant compounds are briefly summarized in Scheme 16, which gives the p values relative to various substituents on the phenyl rings and the relative rates in nitromethane-methanol 19 : 1 at 25" (Modena and Tonellato, 1971a, c; Capozzi et al. 1971; Burighel et al., 1971). For the a-phenyl derivatives (153-155) the degree of anchimeric assistance is very sensitive to changes in the 15 position (R,of 147) and increases on going from 15-hydrogen to 8-phenyl and 15-methyl compounds. This is shown by the large increase in reactivity and by the decrease in the difference between the p value for substituents on the a-phenyl ring and that for substituents on the P-phenylthio ring. The trend of p values is consistent with a gradual change of the transition state geometry from 157 and 158 and may indicate a comparable stability for the open vinyl cation and for the thiirenium ion structures. The rate enhancement due to sulphur anchimeric assistance is virtually nil for ester 153 and rather small (factors of 20-35) in the case of esters of type 154 as estimated by comparison with rate data measured on the triphenylvinyl analogues (see 37 and 42 of Table 7) (Modena and Tonellato, 1971~).For these compounds the a-phenyl group shares a large fraction of the positive charge developing at the reaction centre, as indicated by the relatively high p values and the use of a+ constants. In the absence of an a-phenyl group the demand for sulphur assistance
9*
0 u3
0
VINYL CATIONS
n
U
-
21 II
a
n
II 0 U
261
2 52
QIORGIO M O D E N A A N D U M B E R T O T O N E L L A T O
66, S+.-03STNB , ,,03STNB ,c=c; C=C'
Ar-S
Ar
\s,$>Ar I
(157)
increases and the rate enhancements may become spectacular indeed. Compound 156 (41 of Table 7) reacts at 25" ca 4 x lo4 times faster than calculated for the trimethylvinyl analogue (5b of Table 7 ) . Clearly in the a-alkyl case the stability of the thiirenium ion intermediate structure is much greater than that of the open vinyl cation. It is apparent that, whereas for an open vinyl cation the effect of a substituents is very large in the order : H < Alk < Ph and the effect of ,!I substituents is very small, in the case of thiirenium ions, the effect of substituents bound to the carbon atoms of the three-membered ring of the thiirenium ion 53 follows the order H < < Ph < Alk which is, by the way, that observed for cyclopropenium ions (Breslow et al., 1962). Compounds 159 which possess a ,!I-morpholino group have been suggested to undergo carbon-halogen fission to generate the corresponding cation 160 (Huang and Lessard, 1968). The mass spectra fragmentation patterns of 159 show that an ion whose m/e value correspondsto that of 160 is the most important fragment. Compounds 159 have been found to react in water in the presence of acids or silver ions to give benzoin, b e n d and morpholine salts as the major products, and a possible SN1 solvolysis has been suggested to be involved. However, under acidic conditions, an acid catalyzed hydration followed by morpholine elimination, (equation 28) may be involved in the earlier steps of the reaction. Such a mechanism has been proposed by Sollenberger and Martin (1970) for the hydrolysis of morpholine enamine derivatives.
(159)
(160)
Finally, the values of activation entropies of s N 1 vinylic solvolyses reported in Table 7 range from small positive or negative values to rather large negative values. Assisted solvolyses are associated with the smallest AS+ values, as expected (Winstein et al., 1956; Lancelot and Schleyer, 1969). On the other hand, in a series of structurally similar
VINYL CATIONS
263
vinyl derivatives (e.g. triphenyl vinyl, 33-36 of Table 7, and a-styryl, 23 and 26b of Table 7) the AS' values (and not the AH' values) increase
as the leaving group becomes worse (e.g. from trifluoromethanesulphonate to p-toluenesulphonate). This apparent trend which is at variance with that observed for the solvolysis of saturated systems (Streitwieser et al., 1967; Su et al., 1969) has received little attention and probably requires further experimental support.
D. Electron, Removal from Neutral Species Generation of cations by electron removal from neutral species is certainly more of theoretical than of practical interest. Loss of one electron from a vinyl radical to give a vinyl cation can be achieved in the mas9 spectrometer (eq. 29). The minimum energy
required by an electron to promote the ionization (ionization potential) of the vinyl radical C2H, generated by pyrolysis of methyl vinyl mercury is 9.45 eV (Harrison and Lossing, 1960)correspondingto 218 kcal mol-l. This value is 11.5 kcal mol-1 lower than that of the methyl radical and 5.8 and 15.5 kcal mol-l higher than those of phenyl and ethyl radicals (Lossing, 1963; Fisher et al., 1964). Electron removal from vinyl compounds, accompanied by rupture of the C-X bond, may also occur in the mass spectrometer and a vinyl cation may be generated (equation 30). \
/c=c
/
\
X -
e-
\
+
,C=C-+X+e-
254
GIORGIO MODENA AND UMBERTO TONELLATO
The appearance potential of the vinyl cation from vinyl chloride is 12-81eV. Such a value is ca. 14 kcal mol-l higher than that of the ethyl cation from ethyl chloride (Maccoll, 1962). Vinyl cations and radical cations can be generated from the corresponding alkenes (equation 30). The appearance potentials and fates of the species thus generated are more of interest in mass spectrometry than in carbonium ion chemistry, and have recently been reviewed by Loudon and Maccoll(l970). Ion cyclotron resonance (i.c.r.) spectroscopy has allowed Buttrill (1970) to observe the ion-molecule reaction between hydrogen sulphide and acetylene (acetylene-d,). The products are HCS+(DCS+) and C2H,S+ (C2D,HS+). For the latter ion, the structure consistent with the labelling experiments is that of the thiirenium ion :
111. GENER~L PROPERTIES OF VINYLCATIONS A. Geometry and Stability 1. Theoretical Calculations Semi-empirical calculations for the simple vinyl cation C,H,+ have been reported by Hoffmann (1964) and by Yonezawa et al., (1968). More rigorous calculations by Sustmann et al. (1969)are based on a semiempirical method based on the neglect of diatomic differential overlap (NDDO) calibrated to results of ab initio Hartree-Fock-Roothaan SCF calculations. Recent work by Hopkinson et al. (1971) is entirely based on a non-empirical LCAO-MO-SCF method. The calculations on the vinyl cation have been mainly concerned with the problem of linear (161) versus bridged structures (162).
There is general agreement on the prediction that cation 161 is more stable than 162 and ab initio calculations indicate a difference of 18-5-25 kcal mol-l which probably corresponds to the real difference in the gas phase (Sustmann et al., 1969).
VINYL C A T I O N S
266
Vinyl cation 161 in its most stable arrangement is a planar structure with the G-H bond collinear with the C-C bond. When the cc-hydrogen is removed out of the line of the C-C bond but kept in the plane of the molecule, the plot of calculated total energy versus the Cj-C,-H angle, @, shows one minimum at @ = O ” and no partial minimum a t @=60° corresponding to a sp2 hybridization arrangement of C., As a matter of fact, the 60” “bent” geometrical situation is less stable than the linear one* by ca. 47-65 kcal. mol-l. When the H is moved vertically out of the plane of the molecule, a similar plot is obtained with one minimum, again at @ = 0”. The optimized bond lengths and angles for the most stable geometries 161 and 162 are shown in Figure 2 (Hopkinson et al., 1971) together with the NDDO charge distribution (Sustmann et al., 1969).
H0.245 118 5
JC
0.139 0.209
2e50.3101 065 0.215
C-CLH
-C-H =.\1,7;//1.20S
~/-o.01$180~~
H (161)
(162)
FIU.2
Calculations of methyl substituted cations show that again a linear geometry is always favoured over any bridged geometry with either hydrogen or methyl as bridging groups. Whereas /3-methyl substitution in 161, to give ion 163, provides no extra stabilization, a-methyl substitution, to give ion 164, involves ca. 23-39 kcal mol-l in extra stabilization
(163)
(164)
energy (Sustmann et al., 1969). This additional gain is greater than that calculated (18.5 kcal mol-I) for the difference between a primary and a secondary propyl cation and at least as great as the experimental difference in energy between primary and secondary saturated carbonium ions. On the other hand, methyl substitution in the bridged structure 162 does not provide additional stabilization energy. Dissection of charge distribution in (T and 7~ orbitals for the vinyl cation 161 indicates that the ?T electrons are polarized toward the
* It has been shown that in the case of vinyl radicals the “bent” configuration is more stable than a linear one and the energy barrier for interconveraion of the cis and trans geometries is ca. 2 kcal mol (Kasai and Whipple, 1967).
266
QIORQIO M O D E N A A N D U M B E R T O T O N E L L A T O
a-carbon which holds 1.23 r electrons, as shown in Fig. 3. Also u electrons are attracted toward the electron deficient a-carbon and the “empty ” p-orbital is filled with 0.23 o-type electrons. I n the case of the a-methyl
vinyl cation, 164, the “empty” p-orbital contains 0.35 a-type electrons thus indicating the electron donating effect of the substituent. I n the cwe of the /3-methyl vinyl cation 163 the calculated u and r electron distribution is similar to that of the unsubstituted species. Recent calculations based on the extended Huckel (EHT) molecular orbital method for the 1-cyclopropylvinyl cation indicate that the ion is more stable in the linear “bisected’’ conformation (Kelsey and Bergman 1971). A somewhat special case is represented by the cation involved in the homopropargyl participation. Calculations based on the modified 0 0 method for the system C4HS+,reported by Fischer et al., (1969)
indicate that the most stable geometry is that of the bridged cyclobutenyl cation shown in a non classical formulation 165. The cyclopropylidenemethyl cation structure 166 is less stable by 6.4 kcal mol-’. The greater stability of the digonal vinyl cation 161 than of the bridged cation 162 would imply that, at least in the gas phase, “racemization ” is involved as stereochemical course of any reaction with nucleophiles since both lobes of the crempty” p-orbital of 161 are available for a nucleophilic attack. Efforts to rationalize (Burnelle, 1964) early reports of stereospecific trans addition to acetylene derivatives have recently been complemented by a theoretical calculation of the energy profile for addition of hydrogen fluoride to acetylene in the gas phase (Hopkinson et al., 1971). The results of these calculations suggest that the bridged cation is possibly a transition state and not an intermediate.
257 calculations on the C2H8S+
VINYL CATIONS
Against this, non-empirical SCP-MO ion show (Denes et al., 1971) that the bridged structure (type 162) is about 66 kcal mole-l more stable than the linear structure. These results, although preliminary, accord very well with the chemical evidence (see infra) and point to a great influence of the nature of the electrophile on the relative stability of open and bridged structures. 2. Experimental Evidence
(a) Geometry. The duality of geometric structures 161a and 162a discussed in the preceding section for the vinyl cations in the gas phase is also reflected by the stereochemical and kinetic results of reactions involving vinyl cations in solution. Apparently, the geometry of the vinyl cation depends on the nature both of the (potential) bridging entity, Y, and of the residues bound to the unsaturated carbons.
(1610)
(162a)
Addition reactions in which a proton is added to acetylenic derivatives are non-stereospecific and the cation formed is accordingly linear. What is apparently (see infra) the most straightforward evidence is the observation by Peterson and Duddey (1966) that the addition of trifluoroacetic acid to 3-hexyne gives equal amounts of cis and trans adducts under kinetic control. Addition of hydrogen halides to acetylenic derivatives giving unequal mixtures of geometric isomers have been better explained in terms of oriented ion-pairs or by the operation of a concerted mechanism than by the formation of bridged ions (see section II,A,2). On the other hand, the reactivity of 1-cycloalkenylt d a t e s in ethanol water 1 :1 has been shown to increase with ring size. A factor of a t least lo5, has been observed between five- and eight-membered derivatives (see section II,C,2 and Table 7). These results indicate (i)that the vinyl cation is not a bridged species with H at the bridgehead position, since its stability would have not so dramatically affected by the size of the cycloalkenyl moiety; and (ii)that the stability of the vinyl cation increases with the increasing possibility of attaining a linear arrangement, according with theoretical predictions (see preceding section). However, when a linear geometry cannot be achieved, as in the case of the cyclohexenyl cation 167, a bridged geometry of type 168 (in a nonclassical formulation) may become more stable, a t least when R = Alkyl. This type of bridged ion is likely to be generated in the triple
268
OIOROIO MODENA AND UMBERTO TONELLATO
bond msisted solvolysis of 6-hepten-2-yl tosylate studied by Peterson and Kamat (1969) (section II,A,2b). Such a cation is apparently remarkably stable and selective as shown by the relatively large amounts
(167)
(168)
of five- and six-membered cyclic products of internal return (see Table 3). Bridged cations of type 169 and 170 are expected to be formed for the same reasons when smaller ring are involved. Whereas structure 170 has been indicated as the most stable geometry of the cation involved in homopropargylic participation, there is no experimental indication that cation 169 is formed.
The absence of cyclized products in the acetolysis of 4-pentyn-l-yl tosylate (see section II,A,b2) may indicate that the stability of the bridged cation 169 is lower than that of the homologues 168 and 170. Bridged cations have been indicated as intermediates of the addition of bromine to dialkylacetylenes on the basis of the exclusive trans mode of the reactions. Bridged cations are in each case likely to be formed in the addition of sulphenyl derivatives to alkyl- as well as to arylacetylenes. The “retention ” of trans configuration and the anchimeric assistance effects observed in SN1-type reactions of jl-arylthiovinyl sulphonates indicate that the bridged geometry is always favoured over a linear one. On the other hand, the influence of the residue bound to the unsaturated carbon (which may carry the positive charge in the ion) on the geometry of the intermediate is quite relevant. There is, for instance, a greater tendency towards linear structures when an aryl, instead of an alkyl group, is bound to the a-carbon. This is illustrated by the fact that, whereas bridged intermediates are formed in the addition of bromine to dialkylacetylenes, a linear cation is generated in the addition to arylacetylene derivatives, (Pincok and Pates, 1971). The high -p and -p* values in the case of rate-limiting protonation of arylacetylene and alkoxy- and thioalkoxy acetylenes and in the solvolysis of a-arylvinyl
V I N Y L CATIONS
269
derivatives, the “racemization ’’ observed in solvolytic reactions of a-aryl- and a-cyclopropylvinyl halides as well as the absence of sizable neighbouring-assistance effects in the solvolysis of a-arylvinyl derivatives are clear-cut arguments,discussed in the pertinent sections,in favour of vinyl cations of linear geometry when they may be resonance stabilized by an appropriate a-group (aryl, alkoxy, thioalkoxy, cyclopropyl, alkenyl, etc.). Only in the case of addition of sulphenyl derivatives and, perhaps, of IN3are bridged a-arylvinyl cations involved. However, the study of the unimolecular reactivity of a-arylthiovinyl sulphonates indicates that the stability of bridged a-arylvinyl cations is comparable to that of the corresponding linear ions whereas, in the case of the a-alkylvinyl ions the bridged structures are much more stable than the linear ones. To sum up, the experimental evidence indicates that a linear geometry of type 161a is favoured over the bridged one 162a by the nature of Y in the order H > C< > Br > I > S-, following an approximate scale of “hardness” and also by the increasing efficiency of cc groups capable of delocalizing the positive charge. (b)Xtability. (i) Vinyl us. Saturated Cations. Theoretical calculations do not provide a reliable ground for comparison of the stabilities of unsaturated and saturated cations such as the simple vinyl and ethyl ions. This is a point of concern for organic chemists when looking for reasons for the lethargic rate of solvolysis of vinyl derivatives. Mass spectral data indicate that the energy required to generate the vinyl cation lies between that required to form methyl and ethyl cations from the corresponding radicals. I n particular, the difference in ionization potentials for vinyl and ethyl radicals is 15.5 kcal mol-l. The stability of the above radicals is obviously not the same. However, it has been argued (Richey and Richey, 1970) that the differences in stabilization of the radicals should not be as serious as the differences in the stabilization of the resulting cations. Therefore it seems reasonable to assume that in the gas phase, the vinyl cation is less stable than the ethyl cation by some 10-20 kcal mol-l. Recently, Jones and Maness (1970) estimated from the “best ” values available for heats of formation, that the gas-phase ionization of vinyl chloride requires ca. 17 kcal mol-l more than that of ethyl chloride. The relative stability of vinyl and saturated cations in solution can in principle be evaluated by following three approaches: (a) from the competitive formation of vinyl and saturated cations in electrophilic addition to allenes ; (b)from the relative rates of electrophilic addition to alkynes and alkenes ; (c)from the relative rates of solvolysis of vinyl and saturated derivatives.
260
GIORGIO M O D E N A A N D U M B E R T O T O N E L L A T O
(a) Electrophilic additions to allenes, when bridged cations are not involved, as in the case of proton additions, should provide a straightforward test for the relative stabilities of vinyl and tri-coordinated cations. As reported in detail in section II,B,l, terminal protonation is favoured in the case of the simple allene and monoalkylalleneand in some instances in the case of 1,3-&alkylalleneswhereas central protonation is observed in the case of more highly allcylated allenes. This suggests the following sequences for the apparent stability of vinyl cations and trisubstituted carbonium ions, obviously assuming that the last ones are not stabilized by allylic resonance. Whereas sequences (A) and (C) are those expected, at least in the gas phase, from theoretical calculations
Me
I
+
2 RCH=C<
R--CH=C-CHyMe
+C-H
H
(B)
Me
I C+ c RCH=C( ‘Me Me H
Me
RCH=&CH<
(C)
and mass spectral data, sequence (B)would surprisingly indicate that a secondary vinyl cation is more stable than a secondary tri-coordinate carbonium ion. The latter one is, however, certainly destabilized by the inductive electron-withdrawing effect of the adjacent vinyl residue. Moreover, whereas terminal protonation of the allene derivative does not affect the linear arrangement of the allenic carbon atoms, central protonation requires a simultaneous bending of the C-C-C bond, a factor which, on the basis of the principle of least motion, makes the formation of vinyl cations from allenes more favoured than expected from the actual stability of the intermediate species which might be formed. (b)The available kinetic data on electrophilic additions to acetylenes and structurally related ethylenes in similar conditions are summarized in Table 8. The rate of protonation of simple alkynes is only slightly slower than that of the corresponding alkenes, whereas the protonation of a-aryl or a-alkoxyacetylenes is faster than addition to the double bond. The relative rates of protonation are in each case within a factor
TABLE8 Relative rates of electrophilic additions to acetylene and ethylene derivatives
Substrates
1-pentyne/l-pentene 3-hexyne/3-hexene
Electrophile
Reaction Conditions
CFsCOOH st 60"
H+
l
rp1are1
Peterson and Duddey (1963,1966); Peterson et al. (1965) 2
7x104
-
2.3
1.16
10-2 5x
1.15
Br* phenylpropiolic acidleis-cinnamic acid
H+
HzS04 at 2.5'
20
1.12
ethoxyacetylene/ethoxyethylene
H+
HzO (buffers) at 25"
102
-
a -p
acetylenes/-p ethylenes; by the use of U +
References
2
0.23
Brz in acetic acid a t 25' HzSO4-ethanol95:5 at 25" 2,4-dinitrobenzenesulphenyl chloride in acetic acid at 25" Bra in acetic acid at 25'
phenylacetylenelstyrene
Br+ H+ ArS+
h
Pincock and Yates (1970)
4 F
Noyce et al., (1955)
d
Kharasch and Yannios, 1964 Pincock and Yates (1970)
0
P
Noyce et al., (1967); Mortimer, (1962) Stamhuis and Drenth, Skrabal (1939)
1963b);
H H
262
QIORQIO M O D E N A A N D U M B E R T O T O N E L L A T O
of lo2. It has been argued (Richey and Richey, 1970) that, since the protonation of a triple bond to give a vinyl cation involves the breaking of a weaker bond (by ca. 18 kcal mol-l) than in the case of a double bond to give saturated a carbonium ion, this energy factor must be tightly counterbalanced by the formation of a less stable cation. The stability of a simple linear vinyl cation is, therefore, probably lower than that of the corresponding saturated ion by 10-20 kcal mol-1 in good agreement with the estimated value for gaseous ions. The fact thab the rate enhancements observed for the homopropargylic participation are of the same order of magnitude of those for homoallylic participation may indicate that also in the case of the addition of carbonium ions, the rates of addition to the triple and double bond are similar. On this basis, the very large rate difference for the bromine addition to phenylacetylene and styrene, a fact which would indicate that the linear vinyl cation thus generated is exceedingly less stable than the saturated analog, is certainly surprising and, at the present, difficult to explain (Perrin, 1968, Pincock and Yates, 1971). I n the case of formation of bridged ions (addition of sulphenyl derivatives and bromination of dialkylacetylenes) the rate is slower for alkynes than for alkenes (see Table 8). The difference in the rate of addition of sulphenyl derivatives is not very large (although, as argued in section II,A,4, the values do not refer to strictly typical conditions) but in the case of cyclic “bromonium” type intermediates the rate factor is spectacular (lo-’). This may indicate that an extra factor of instability is involved in the case of unsaturated bridged ions. Several arguments may be invoked to accommodate the facts: the three-membered unsaturated ring may suffer more severe strain than the corresponding saturated ring (this, however, does not prevent the existence of the relatively stable cyclopropenium ion). Another factor, probably more important, is the repulsion between the T electrons of the double bond and the non-bonded electrons of the sulfur and bromine atoms. However, too few systems have so far been investigated, and not yet in full detail, to attempt anything more than a vague hypothesis. (c) Less straightforward arguments about the relative stability of vinyl cations are obtained from a scrutiny of kinetic data on unimolecular reactions of vinyl and alkyl derivatives. Rate differences of lo0-lo8 have been estimated for solvolytic reactions between vinyl derivatives and their saturated counterparts (Rappoport and Gal, 1969; Miller and Kaufmann, 1968; Peterson and Indelicato, 1969; Jones and Maness, 1969, 1970). It is difficult to evaluate how much these rate differences reflect the different stabilities of saturated and unsaturated cations,
VINYL UATIONS
263
although it could be argued that they areapproximately of the magnitude expected from the mass-spectral data and also in line with the arguments advanced to explain the relative rates of addition to double and triple bonds (see above). Several authors prefer to interpret the difference in solvolytic reactivity in terms of reactant stabilization. Such stabilization is mainly due to the stronger (r bond between the sp3 vinyl carbon and the leaving group than that linking the sp2 carbon to the leaving group, because of the different hybridization state. (Millerand Kaufman, 1968). Earlier suggestions of resonance stabilization of a non-bonded electron pair of the leaving group, illustrated by limiting structures such as 171, have been indicated as unimportant (Peterson and Duddey, 1966; Jones and Maness, 1970).
(171)
Probably both reactant stabilization and the already evaluated relative instability of the cationic transition state contribute to the slowness of the solvolysis of vinyl components, but other factors are certainly involved. The most obvious experimental problem is whether the compounds compared react by a unimolecular mechanism or nucleophilic attack by the solvent is involved to a certain extent. I n the case of vinylic systems, for instance, nucleophilic solvation from the rear is in general much more hindered than in the case of saturated compounds and the transition state is likely to be stabilized only by electrophilic solvation of the leaving group (Rappoport and Atidia, 1970). The low rn values observed in the case of vinyl halides or sulphonates may be taken as a strong indication of poor solvation of the transition state in solvolytic reactions of vinyl derivatives. These and other complications, such as differences in hyperconjugation, differences in and >C- bonds (Jones and Maness, electronegativity of the -C= 1970), make the comparison of the unimolecdar reactivity of vinyl and saturated compounds of little help even for a semi-quantitative evaluation of the relative stability of vinyl and alkyl cations. (ii) Internal Factors Influencing the Stability of Vinyl Cations. Semi-quantitative indications of the effect of internal factors on the stability of the vinyl cations can be obtained from a scrutiny of available kinetic data. The relative rates of proton additions to structurally different acetylene and ethylene derivatives shown in Table 8 indicate that the effect of groups, such as phenyl and alkoxy, which can conjugate with the positive charge, is greater for alkenes. Moreover, it is evident from Table
264
G I O R G I O M O D E N A A N D U M B E R T O TONELLATO
8 that the negative p values for substituents on the a-aryl group is 12%-16% higher in the alkyne series. The above effect is also evident
from the relative rates for solvolysis via carbonium ions of a-alkylvinyl, a-arylvinyl, conjugated 2-dienyl derivatives (roughly estimated for aqueous ethanol solutions at 100' from data of Table 7) and the saturated analogue in ethanol-water 4 :1 at 76' (Streitwieser, 1966). The relative rates within each series should reflect the relative stability of the unrearranged carbonium ions. H\,C=C-Me + Me 1
+
Me-CH-Me 1
H >C=CP-h H
+
104-105
+
Me-CH-Ph 104
H&=&-CH=CHz 104-105
+
Me-CH-CH=CH2 102
The greater conjugative effect of substituents in linear vinyl than in alkyl cations may well be the effect of shorter bonds to sp carbon than to sp2 carbon and hence of a greater overlap between the empty p orbital of the vinyl a-carbon and the available p orbital of the adjacent atom as suggested by Matesich (1966) (Richey and Richey, 1970). A second factor which may be related to the inherently greater instability of the vinyl cation compared to the ethyl cation-and which may be in itself an explanation of the stronger demand of electron relay in the former than in the latter case-is the different state of hybridization of one of the carbon atoms next to the cationic centre. I n the vinyl cation the /3carbon is sp2 hybridized and hence more electronegative than the sp3 carbon of the ethyl cation. Consequently a greater fraction of the positive charge is concentrated on the vinyl cation centre than on the ethyl-cation carbon. On the other hand, from the available data it is difficult to evaluate the effect of /3-substituents on the stability of vinyl cations. The effect of changes at the /3 carbon on the rate of solvolysis of vinyl derivatives is generally very small, as discussed above. Although this certainly + suggests that a structure of the type R,' C-&R is a relatively unimportant resonance contributor (Miller and Kaufmann 1968), the small kinetic effect observed may be due to other factors than the influence on the stability of the reaction intermediates, such as steric hindrance to solvation, steric or anchimeric assistance effects. The effect of internal factors in the stability of bridged cations has been discussed in the case of thiirenium ions generated by addition of
266
VINYL CATIONS
sulphenyl derivatives and by SN1-type reactions of /3-thiovinyl derivatives. It has been argued that the effect of substituents on the carbon atoms of the cyclic ion follows the series H < Ph < Alk, which is not the order of substituents at the p-carbon of the linear vinyl cation. There are no indications whether this is a general trend in the case of cyclic vinyl cations or is an effect peculiar to the thiirenium ion, which may have at least partial aromatic character (Pilgrim, 1968; Volpin et al., 1962).
B , Reactions of Vinyl Cations The vinyl cations so far investigated are energetic intermediates and their reactions are too rapid for a direct study. The results discussed in detail in section I1 show that a variety of alternative paths are available to them. As electrophilic species, vinyl cations react by interaction with electron pairs which may be, following Bethell and Gold’s (1967) classification : (i) non-bonded electrons ; (ii)r-bonded electrons; (iii) ubonded electrons. (i) Interactions with non-bonded electrons of anions or electrically neutral species is the most frequently observed type of reaction (nucleophilic substitution, solvolysis, etc.). (ii) Interactions of vinyl cations with r electrons of triple bonds or cumulate double bonds are involved in the cyclodimerization reactions of propyne and allene in the presence of hydrogen halides (section II,B,l) and in the formation of hexaethylbenzene from 3-hexyne in trifluoroacetic acid (section II,A, lb). Interactions with aromatic r electrons has been suggested to occur in the formation of benzo[b]thiophen derivatives from 1,2-diary1-2-arylthiovinylcompounds (equation 31) (Capozzi et al., 1970b). Ar
A=\-
‘c-c/
Ar
+
(iii) El-type reactions have been suggested to occur in a variety of conditions when a /3-hydrogen is available. The small but real tritium exchange found by Noyce and Schiavelli (1968) in the hydration of phenylacetylene in a medium of high acidity, such as (tritiated) 40% H2S04solution, indicates that a hydrogen may be easily lost as B proton from a vinyl cation.
266
G I O R O I O M O D E N A A N D UMBERTO TONELLATO
Intramolecular rearrangements involving interaction of the positive vinyl carbon with u electrons of C-H and C-C bonds have been observed. Migration across the double bond (equation 32) is one type of
such interaction. Hydride transfer probably occurs in the vinyl cation generated by addition of adamantyl cation to acetylene prior to formation of adamantyl methyl ketone. Migration of aryl and alkyl groups has been observed in several cases and in particular in the case of vinyl cations generated from diazonium precursors. Such migration can occur in a process concerted with the heterolytic cleavage or subsequent to it. In the case of assisted SN1reactions of 2-methyl-1-cyclohexylderivatives and of some p-thiovinyl derivatives, migration occurs as expected. Migration apparently occurs also in anchimerically unassisted solvolytic reactions such as the silver-catalysed acetolysis of 2,2-dianisyl-l-phenyl bromide (Rappoport; Hanack, 1970).
Migration to the double bond (equation 33) is a second type of interaction and is well illustrated by the nature of rearranged products obtained in the hydrogen chloride addition to t-butylacetylene (section II,A, lc) and in the solvolysis of t-butylvinyl derivatives and a-cyclopropylvinyl derivatives (section II,C,2). R
+ I ,c=c-c\
R _3.
I
,c-c=c1 \+
/-
(33)
\
Reactions of bridged unsaturated cations have not been investigated in detail and are currently interpreted in terms of ring opening to generate a back-shielded vinyl cation which may react with avadable electron pairs or as a direct nucleophilic attack at one of the carbon atoms of the cyclic cation to give the h a 1 products. The stereochemistry of the reactions of vinyl cations with nucleophiles is predictably different depending on their geometry. (a) High stereospecihity is expected from bridged ions (transaddition to acetyIenesand “retention ” of configuration in substitution reactions of vinyl derivatives) and is experimentally observed in the case of thiirenium ions. (b) From “free” linear cations with two /3 substituents of equal size, complete “racemization” is expected and is fully verified in the substibromide (sectionII,C,2). tution products from 1,2-dianisyl-2-phenylvinyl
VINYL CATIONS
267
(c) Prom “free” linear cations which possess two 8 , residues of different size or different polarity (e.g. H and Me or Ar), preferential attack of the nucleophile a t the Iess hindered lobe of the vacant pn orbital (and hence a certain degree of stereoselectivity) may be predicted. The experimental data are contradictory. On one hand, the prediction is verified in the addition of bromine to arylacetylenes in acetic acid (section II,A,3), in the addition of hydrogen chloride to phenylpropyne in acetic acid ( 4 of Table 2) and in the addition of hydrogen bromide to dianisylacetylene in acetic acid and carbon tetrachloride (Rappoport and Atidia, 1970),to cite a few examples where the thermodynamically less stable cis isomer is preferentially formed. On the other hand, complete racemization has been observed in the substitution products from 1iodo-l-cyclopropylpropenesin acetic acid (section II,C,2) and in the sdducts of trifluoroacetic acid to 3-hexyne (section II,A,lb). Very often, as discussed in the pertinent sections, oriented intimate ion-pairs have been invoked to explain the observed stereoselectivity. Specific interactions (which may well be steric in origin) of the anion with the cis or the trans substituents may give rise to two different ion-pairs and the stereochemical course determined by their relative stabilities and interconversion rates compared to the rate of covalent return.
IV. RELATED SPECIES Some of the species most closely related to vinyl cations will be briefly considered. These are the propargyl cations and some isoelectronic analogues of vinyl cations, such as acyl cations, nitrilium cations and the imminium ions. Aryl cations are related to vinyl cations insofar as they are disubstituted carbonium ions, although their chemistry is more related to the aromatic properties of the nucleus. This subject has recently been reviewed (Richey and Richey, 1970). Other disubstituted carbonium ions such as R,C+ and R,C2+ as well as monosubstituted cations, R2C=C+and RC+ have not been suggested ; ~ 8 relevant reaction intermediates and only for some of them mass spectral evidence has been reported. Although acetylium ions R C A Y may be reasonably thought to be formed in the decomposition of diazonium precursors (Robson and Tedder, 1963)no clear evidence of their existence is as yet available. A. Propargyl Cations Yropargyl cations (alkynylcarbonium ions) (172) may be viewed as the “unsaturated” analogues of ally1 cations (173). The close relationship with vinyl cations is shown by the vinyl cation-type structure 172a
268
G I O R G I O MODENA A N D UMBERTO TONELLATO
which is an important resonance contributor. In contrast to allyl cations, which must be in a favourable “planar” stereoelectronic configuration, propargyl cations may achieve conjugative stabilization without rotational requirements, because of their cylindrical symmetry. In spite of the favourable steric arrangement, propargyl cations, although more stable than vinyl and ethyl cations, are less stable than allyl cations. This is indicated by the ionization potentials of the simple parent radicals, 8.25 and 8.16 eV (Lossing, 1963)respectively, and by the lower (by a factor of ca. lo4) rate of unimolecular solvolysis in ethanolwater 4 :1 of propargyl (174) than of allyl halides (175) (Burawoy and Me
I I
H--CC--CC1
Me (174)
Me
I I
H&=C--C--CI lnrc
(175)
Spinner, 1954 ; Vernon, 1964). The opposing destabilization is probably due to the greater (electron-withdrawing) inductive effect of the triple bond than that of the double bond in allyl cations. As a matter of fact, it has been evaluated from the solvolysis rate of derivatives 174 and 175 that the resonance contributions of the limiting forms (172a) and ( 173a) to their respective hybrids are similar (Richey and Richey, 1970). Spectral studies of propargyl cations have established their existence in solution. The IH n.m.r. absorptions of the ions are in each case downfield from those of the neutral precursors and the extent of the shift is a valuable tool for mapping the distribution of the positive charge on the carbon atoms of the ions (Richey et al., 1965a, b ; Pittman and Olah, 1965). In solution, propargyl cations may in principle be generated by electrophilic addition to vinylacetylenes (equation 34). However, by all the evidence, the electrophilic attack by protons occurs more readily at the terminal acetylenic carbon than at the terminal vinyl carbon, and cation 176 is generated instead (de la Mare and Bolton, 1966).
VINYL CATIONS
269
(176)
This is illustrated by the products of addition of hydrogen bromide to vinyl acetylene (177) (Traynard, 1962) (equation 35). Only in the case of halogen addition have products formally arising from electrophilic attack to the double bond been isolated in substantial amounts (Petrov et al., 1960).
Propargyl cations may be generated in solution by heterolytic fission of the C-X bond of derivatives 177. Direct spectroscopic evidence of their formation from several a,lcohols 177 in “magic” acid mixtures has
been obtained. Examples of solvolysis of halides 177 (in particular of chlorides 174 and related structures) have been reported (Burawoy and Spinner, 1954; Shiner and Wilson, 1962; Shiner et al., 1962; le Noble, 1965) and the evidence for an SN1 mechanism discussed in a recent review (Richey and Richey, 1970). Propargyl cations may also be generated by heterolytic fission of the C-X bond of allenyl derivatives (178). Early attempts to solvolyse chloride 178 (R = H ; R’ =Me) in alcoholic solution, even in the presence of silver nitrate, did not furnish reliable conclusions concerning the possibility of unimolecular solvolysis through the propargyl cation (Pudovik, 1951 ; Hennion and Maloney, 1951). The apparently slow
270
B1:ORBIO M O D E N A A N D U M B E R T O T O N E L L A T O
x\ ,c=c=c It
/ \
R‘
R’
(178)
rate of solvolysis of the model considered is not surprising in view of the fact that it is a primary allenyl derivative. Very recently, Schiavelli et al. (1970) reported a preliminary study of the SolvoIysis of triphenylallenyl chloride (179), in acetone-water mixtures. The reaction occurs readily at mild temperatures and the major product is the alcohol 180 (909’0). Changing the a-phenyl into an Ph
cl\ Ph/c=c=c,
/
I
Ph
Ph-C4--C---OH
Ph
I
Ph (180)
a-anisyl group results in a modest rate enhancement corresponding to a value of - 2. Both the effect of substituents on the a-ring and the nature of the products are consistent with the hypothesis of an intermediate propargyl cation in which the positive charge is substantially delocalized at the C, position (resonance structure 172). The m value of 0.7 is larger than that observed (0-4) in the solvolysis of trianisylvinyl halides and arenesulphonates (Rappoport and Gal, 1969 ; Rappoport and Kaspi, 1970). Such a difference may indicate a 1a.rgernucleophilic assistance by the solvent in the case of triarylallenyl derivatives consistent with lower steric hindrance to “back-side” solvation in the case of allenyl derivatives.
p(u+)
B. Nitrilium Ions Nitrilium ions (azacarbonium, azomethinium ions) (182) are stabilized by resonance structure (182a). -hT=c-
+
(182)
-
+
-N=C-(182~)
They have been suggested as intermediates in acid-catalysed additions
t o nitriles. The hydration of nitriles probably involves initial protonation of the N end of the -C=N bond, as indicated by the nature of the product of the acid-catalysed reaction between acetonitrile and ethanol (Hill and Rabinowitz, 1926) (equation 36).
V I N Y L CJATIONS
MebN
EtOH (HCl) ____f
271
MeC(OEt)=NH
Nitrilium ions are generally indicated aa discrete intermediates in the Beckmann rearrangement of oximes (equation 37, Y = OH2) and in the Schmidt rearrangement (equation 37, Y = N,) of ketones and aldehydes with hydrogen azide. The first stage of these reactions has been defined as an “ionization by rearrangement ’’ process through a transition state of type 183 (Smith, 1963).
(183)
The hydrolysis in acetone-water of the azomethine derivatives 184 to the corresponding amides has been investigated by Ugi et al., (1962). An SN1mechanism, probably involving nitrilium chloride ion-pairs, has
(1W
been suggested, based on the kinetic solvent and substituents (GI and G,) effects. The p(G1)value - 1.2 (Aylward and Scott 1969), although of the appropriate sign for a carbonium ion mechanism, is rather small in magnitude and indicates that a large fraction of the positive charge developed in the transition state is held by the nitrogen atom in a resonance structure of type 182a. The reactivity of 184 in benzene is being reinvestigated by Rappoport and Ta-Shma (1971). The solvolysis in dioxan-water 4 :1 of p-nitrophenylbenzoylhydrazonyl bromides (185) has been investigated by Aylward and Scott (1969). From the general kinetic behaviour (mass-law effects, strong primary salt effects) and the nature of the products (186) the solvolysis was suggested to occur via a rather stable cation. The calculated p(G1) value of - 0.93 indicates that also in this case much of the positive charge developed in the transition state is largely removed from the or-carbon. Structure 187 is indicated as an important resonance contributor of the intermediate cation and the possibility of further charge delocalization via tautomerization into the arylhydrazine ring (188) is also suggested. A review on this subject is due to appear (Scott’andButler, 1972). 10
272
G I O R G I O MODENA A N D U M B E R T O T O N E L L A T O
C. Imminium Ions Imminium ions (190) are relatively rare species. It had been suggested (Stieglitz and Leech, 1913; Waters, 1936; Hammett 1940) that the Beckmann, Schmidt and related rearrangement processes might involve imminium ions as transient intermediates but this hypothesis has later been rejected on the basis of the observed retention of the configuration of the migrating group (Kenyon and Young, 1941; Campbell and Kenyon, 1946). It is however not entirely excluded that imminium ions are actually formed before undergoing very fast rearrangement to the more stable azomethinium ions. When such a rearrangement cannot be easily achieved, as in the case of the decomposition of oxime 191, the corresponding imminium ion is probably involved (Lansbury et al., 1964; Lansbury and Mancuso, 1964) (equation 38).
273
V I N Y L OATIONS
D. Acyl Cations The acyl cation (acylium ion, oxocarbonium ion) is a resonance hybrid of the two main contributing structures 192 and 192a. The importance of structure 192a is indicated, for instance, by the high frequency carbonyl absorption (2200-2300 cm-l) observed in acyl cations generated from acyl halides and Lewis acids (Olah et ab., 1962, 1963; Bethell and Gold, 1967). +
0CC-R
CJ
+
OEC-R (192a)
(192)
A detailed survey of this important type of reactive intermediates is outside the scope of this chapter. Reviews on the subject are available [Praill, 1963; Olah (Ed.), 19651 or due to appear (Olah and White, 1972). Early evidence of the existence of acyl cations in solution from conductivity (Seel, 1943 ; Mackenzie and Winter, 1948) and cryoscopic (Treffers and Hammett, 1937; Gillespie and Robinson, 1968) measurements have been complementedby more recent spectroscopic studies and X-ray structure analysis of stable salts (Boer, 1966). I n particular, IH and l9Fn.m.r., spectroscopy has been widely exploited (Deno et al., 1964; Olah et al., 1962, 1963, 1964, 1969; Olah and Comisarow, 1966). Acyl cations are postulated as intermediates of the hydrolysis of some substituted benzoic acid esters (Newman, 1941 ; Treffers and Hammett, 1937) and ,!?-lactones(Olson and Hyde, 1941) in strong acid solutions via the AAcl mechanism (Ingold, 1953, 1969). Acylium ion pairs, as well as the related oxonium complexes with Lewis acids, are recognized as the effective intermediates of aromatic (Friedel-Crafts) acylations. Kinetic studies apparently exclude that the electrophilic attack at the aromatic nucleus is by “free” acyl cations (Brown and Jensen, 1958). Acyl cations are formed in the reversible addition of carbon monoxide to carbonium ions (Balaban and Nenitzescu, 1959; Koch and Haaf, 1958a, b ; 1961). The formation of the acetyl cation 193 from carbon monoxide and methane in SbF, has been directly observed by lH n.m.r. CHs++CO
CHa6O
%
CHsCOzH+H+
(39)
(193)
spectroscopy and acetic acid is eventually obtained (equation 39) by quenching with water (Hogeveen et al., 1969). Investigation on similar carbonylation equilibria has proved to be a valu&bletool for the study of the stability of carbonium ions (Brouwer and Hogeveen, 197 1).
274
G I O R G I O MODENA A N D UMBERTO TONELLATO
ACXNOWLED(XMENT
The authors wish to acknowledge the receipt of information and of manuscripts prior to publication from Drs. R. G. Bergman, M. Hanack, Z. Rappoport, and K. Yates.
REFERENCES
’
Baeyer, A. (1905). Ber. 38, 569. Balaban, A. T., and Nenitzescu, C. D. (1965). Annalen 625, 66. Banthorpe, D. V. (1970). Chem. Rev. 70, 295. Bertrand, M., and Santelli, M. (1968). Chem. Comm. 718. Bethell, D., and Gold, V. (1967). “Carbonium Iona”, Academic Press, London. Bianchini, J. P., and Guillemonat, A. (1968). Bull. SOC. chim. France 2121. Bly, R. S., Ballentine,A. R., and Koock, S. U. (1967).J . Amer. Chem.Soc. 89,6993. Bly, R. S., and Koock, S. U. (1969). J . Amer. Chem. SOC. 91, 3292. Boer, F. P. (1966).J . Amer. Chem.Soc. 88, 1672. Borkent, G., and Drenth, W. (1970). Rec. Traw. chim. 89, 1057. Bott, K. (1969a). Tetrahedron Letters 1747. Bott, K. (1969b). Chem. Conam. 1349. Bott, K. (1970). Angew. Chem. Internat. Edn. 9,964. Bott, R. W., Eaborn, C., and Walton, D. R. M. (1964). Organometullic Chem. 1,420. Bott, R. W., Eaborn, C., and Walton, D. R. M. (1965). J . Chem. Roc. 384. Breslow, R., Hover, H., and Chang, H. W. (1962).J . Amer. Chem.SOC.84, 3168. Brouwer, D. M., and Hogeveen, H. (1971). Prog. Phys. Org. Chem., to bepublished. Brown, H. C., and Jensen, F. R. (1958). J . Amer. Chem. SOC.80, 2291. Burawoy. A., and Spinner, E. (1951). J . Chem. Soc. 3752. Burighel, A., Modena, G., and Tonellato, U. (1971). To be published. Buttrill, S. E. (1970). J . Amer. Chem. SOC.92, 3560. Calo’, V., Modena, G., and Scorrano, G. (1968).J . Chem. SOC.C, 1344. Calo’, V., and Scorrano, G. (1968). Gazzettu 98, 646. Campbell, A., and Kenyon, J. J. (1946).J . Chem. SOC. 26. Capozzi, G., Melloni, G., andModena, G. (1970a).J . Chem.Soc. C , 2617. Capozzi, G., Melloni, C., and Modena, G. (1970b).J . Chem.SOC. C, 2621. Capozzi, G., Melloni, G., and Modena, G. (19700).J . Chem. SOC.C, 2626. Capozzi, G., Melloni, G., Modena, G., and Tonellato, U. (1969). Chem.Comm. 1620. Capozzi, G., Modena, G., andTonellato, U. (1971). J . Chem.SOC.B , 1700. Charleston, B. S., Dalton, C. K., Waahburne, S. S., and Dalton, C. K. (1969). Tetrahedron Letters 69. Closson, W. D., and Roman, 8. A. (1966). Tetrahedron Letters 6016. CrandalI, J. K., Paulson, D. R., and Bunnell, C. A. (1968). Tetrahedron Letter8 6063.
Curtin, D. Y . ,Kampmeier, J. A., and Farmer, M. L. (196513). J . Amer. Chem.SOC. 87, 874. Curtin, D. Y . ,Kampmeier, J. A., and O’Connor, B. R. (1966a). J . Arner. Chem. SOC.87, 863.
Day, A. C., and Whiting, M. C. (1967). J . Chem. Roc. B , 991. de la Mare, P. B. D., and Bolton, R. (1966). “Electrophilic Additions to Unsaturated Systems”, Elsevier, London.
V I N Y L UATIONS
275
Denes, A. S., Csizmadia, I. G., and Modena a.(1971).Chem. Comm. (inpress). Deno, N. C., Pittman, C. U., and Wistosky, M. J. (1964). J. Amer. Chem.SOC. 86, 4370. Dewar, M. J. S. (1949). “Electronic Theory of Organic Chemistry”, Clarendon Press, Oxford. Di Nunno, L., Melloni, G., Modena, G., and Scorrano, G. (1966). Tetrahedron Letter8 4406. Dondoni, A., Modena, G., and Scorrano, G. (1964). BoEZ. mi. Fm. C h h . id. Bologna 22, 26. Drenth, W., and Hogeveen, H. (1960). Rec. Truv. chim. 79, 1002. Fahey, R. C., and Lee, D. J. (1966). J. Amer. Chem.SOC.88, 6656. Fahey, R. C., and Lee, D. J. (1967) J. Amer. Chem.SOC. 89, 2780. Fahey, R. C., and Lee, D. J. (1968). J. Amer. Chem. SOC.90,2124. Fahey, R. C., and McPherson, A. (1969). J. Amer. Chem. SOC.91, 3866. Fedorova, A. V. (1963)J. Qen. Chena. (U.S.S.R.)33, 3608. Fedorova,A. V.,andPetrov, A.A. (1961). J. Gen.Chem. (U.S.S.R.)31,3273. Fischer, H., Hummel, K., and Hanack, M. (1969). Tetrahedron Letters 2169. Fisher, I. P., Palmer, T. F., and Lossing, F. P. (1964). J . Amer. Chem.SOC. 86,2741. Frydman, N., and Mazur, Y. (1970). J. A w r . Chem. SOC. 92,3203. Gillespie, R. J., and Robinson, E. A. (1968). I n “Carbonium Ions” (G. A. Olah and P. v. R. Schleyer Eds.). Vol. 1, pp. 111-134. Interscience, New York. Griesbaum, K. (1969). Angew. Chem. Internat. Edn. 8, 966. Griesbaum, K. (1966). Angew. Chem. Internat. Edn. 5,933. Griesbaum, K., Naegele, W., and Wanless, G. G. (1965). J. Amer. Chem. SOC.87, 3162. Griesbaum, K., and Rehman, Z. (1970). J. Amer. Chem. Soc.92,1416. Grob, C. A. (1971).Chimia, 25, 87. Grob, C. A., and Cseh, G. (1964). Helv. China. Acta 47, 194. Grob, C. A., andpfaendler, H. R. (1970). Helv. Chim. Acta 53,2130. Grob, C . A., and Pfaendler, H. R. (1971). Helv. Chim. A& (in press). Grob, C. A., and S p a r , R. (1969). Tetrahedron Letter8 1439. Grob, C. A., and Spaar, R. (1970). Helv. Chim. Acta 53,2119. Hanack, M. (1970). Account Chem. Rea. 3, 209. Hanack, M., and Bassler, T. (1969). J. Amer. Chem.SOC. 91,2117. Hanack, M., Bocher, S., Herterich, I., Hummel, K., and Vbtt, V. (1970). Annulen, 733, 5. Hankk, M., Bocher, S., Hummel, K., and Vbtt, V. (1968). Tetrahedron Letlers 4613. Hanack, M., and HiifYner, J. (1964). Tetrahedron Letters 2191. Hanack, M., and Hiifher, J. (1966). Chem. Ber. 99,1077. Hanack, M., Hliffner, J., and Herterich, I. (1965). Tetrahedron Letter8 876. Hanack, M., and Herterich, I. (1966). Tetrahedron Letter8 3847. Hanack, M., Herterich, I., and Vott, V. (1967). Tetrahedron Letter8 3871. Hanack, M., and Heumann, A. (1969). Tetrahedron Lettera 5117. Hanack, M., and Vott, V. (1968). Tetrahedron Lettera 4617. Hantzsch, A. (1921). Ber. 54, 2673. Hammett, L. P. (1940). “Physical Organic Chemistry”, McGraw-Hill,New York. Hargrove, R. J.,Dueber, T. E., and Stang,P. J. (1970). Chem.Comm. 1614. Harrison, A. G., and Lossing, F. P. (1960). J. Amer. Chem. SOC. 82.619. Hassner, A., Ibister, R. J., and Friederang, A. (1969). Tetrahedron Leitws 2939.
276
G I O R G I O MODENA A N D UMBERTO TONELLATO
Hasaner, A., and Boerwinkle, F. (1969). Tebrahedron Letters 3309. Haasner, A. (1968). J . Org. Chem. 33, 2684. Heck, R., and Winstein, S. (1957). J . Amer. Chem. SOC.79, 3105. Hennion, G. F., and Maloney, D. E. (1951). J . Amer. Chem.SOC.73,4735. Hekkert, G. L., and Drenth, W. (1961). Rec. Trav. chim. 80, 1285. Hekkert, G. L., and Drenth, W. (1963). Rec. Trav. chirn. 82, 405. Hill, A. J., and Rabinowitz, I. (1926). J . Amer. Chem.SOC.48, 732. Hine, J. (1964). “Divalent Carbon”, pp. 88-90. The Ronald Press Co., New York. Hogeveen, H., and Drenth, W. (19634. Rec. Trav. chim. 82,375. Hogeveen, H., and Drenth, W. (1963b). Rec. Trav. chim. 82, 410. Hogeveen, H., Lucaa, J., and Roebeck, C. F. (1969). Chem. Comm. 920. Hogg, D. R., and Kharasch, N. (1956). J . Amer. Chem. SOC. 78,2728. Hoffiann, R. (1964). J . Chem. Phys. 40,2480. Hopkinson, A. C., Yates, K., and Csizmadia, I. G. (1971). J . Chem. Phys. 55 (inpress). Huang, S. J., and Lessard, M. V. (1968). J . Amer. Chem.SOC.90, 2432. Imhoff, M. A., Summerville, R. H., Schleyer, P. v. R., Martinez, A. G., Hanack, M., Dueber, T. E., and Stang, P. J, (1970). J . Amer. Chem. SOC.92, 3802. Ingold, C. K. (1953). “Structure and Mechanism in Organic Chemistry”, Cornell Univ. Press, Ithaca, New York, and Bell, London. Jacobs, T. L., and Johnson, R. N. (1960). J . Amer. Chem. SOC.82, 6397. Jacobs, T. L., and Macomber, R. S. (1969). J . Amer. Chem. SOC.91,4824. Jacobs, T. L., Macomber, R. S., and Zunker, D. (1967). J . Amer. Chem. SOC.89, 7001.
Jacobs, T. L., and Searles, S. (1944). J . Amer. Chem. SOC.66, 686. Jones, W. M., and Maness, D. D. (1968). J . Amer. Chem.SOC.91,4314. Jones, W. M., and Maness, D. D. (1970). J . Amer. Chem.SOC.92,5457. Jones, W. M., andMiller, F. W. (1967). J . Amer. Chem.Soc. 89, 1960. Kasai, P. H., and Whipple, E. 3. (1967). J . Amer. Chem. SOC.89, 1033. Kaufman, D. A. and Miller, L. L. (1969). J . Org. Chem. 34, 1496. Kell, D. R., and McQuillin, F. J. (1970). Chem. Comm. 599. Kelsey, D. R., and Bergman, R. G. (1970). J . Amer. Chem. SOC.92, 228. Kelsey, D. R., and Bergman, R. G. (1971). J . A w r . Chem.SOC.93, 1963. Kenyon, J., and Young, D. P. (1941). J . Chem. SOC.263. Kernaghan, G. F. P., and Hoffiann, H. M. R. (1970). J . Amer. Chem.Soc. 92, 6988. Kharasch, N., and Yannios, C. N. (1964). J . Org. Chem. 29,1190. Koch, H., and Haaf, W. (1958a). Angew Chem. 70,311. Koch, H., and Haaf, W. (1968b). Annalen 618, 251. Koch, H., and Haaf, W. (1961). Chem. Ber. 63, 2823. Lancelot, C. J., and Schleyer, P. v. R. (1969). J . Amer. Chem.SOC.91,4291. Lansbury, P. T., Colson, J. G., and Mancuso, N. R. (1964). J . Amer. Chem. SOC. 86, 6225.
Lansbury, P. T., andMancuso, N. R. (1966). J . Amer. Chem.SOC.88,1205. le Noble, W. J. (1965). J . Amer. Chem. SOC.87, 2434. Letsinger, R. L., Oftendahl, E. N., and Nazy, J. R. (1965). J . Amw. Chem. SOC. 87, 742. Lossing, F. P. (1963). I n “Mass Spectrometry” (C. A. McDowell, Ed.), McGrawHill, New York. Loudon, A. G., and Maccoll, A. (1970). In “The Chemistry of Alkenes” (J. Zabicky, and S. Patai, Eds.). Vol. 2, pp. 327-358. Interscience, London.
VINYL UATIONS
277
Maccoll, A. (1962). “The Transition State”, Spec. Publ. No. 16. The Chem. Soc., London, p. 169. Mackenzie, H. A. E., and Winter, E. R. S. (1948). Tram. Faraday Soc. 44,169. Martinez, A. G.,Hanack, M., Summerville, R. H., Schleyer, P. v. R., and Stang, P.J. (1970). Angew. Chem. Internat. Edn. 9, 302. Matesich, M. A. (1966).Ph.D. Thesis. University of California, Berkeley. Matesich. M. A. (1967).J . Org. Chem. 32, 1268. Mavrov, M. V., and Kucherov, V. F. (1967). Ruaa. Chem. Rev. 36,233. Miller, L. L., and Kaufman, D. A. (1968). J . Amer. Chem. SOC.90,7282. Miocque, M., Hung, N. M., and Yen, Y. Q.(1963).Ann. Chim. (France)[3], 8,167. Modena, G.(1971). Accounts Chem. Res. 4,73. Modena, G.,and Scorrano, G. (1969). I n “Mechanism of Reactions of Sulfur Compoundi3” (N. Kharasch, B. 5. Thyagarajan, and A. I. Khodair, Edrs.). Vol. 3,p. 116. Intra-Science Research Foundation, Santa Monica. Modena, G., and Tonellato, U. (1968). Chem. Cmnm. 1676. Modena, G.,and Tonellato,U. (19718).J . Chem.SOC. B , 374. Modena, G.,and Tonellato, U. (197lb). J . Chem. SOC.B , 381. Modena, G.,and Tonellato, U. (19710).J . Chem. SOC.B, 1669. Modena, G.,Tonellato, U., and Naso, F. (1968). Chem. Cmm. 1363. Mortimer, C. T. (1962). “Reaction Heats and Bond Strengths”, Chap. 3. Pergamon Press, New York. Mueller, W. H. (1969).Angew Chem. Internat. Edn. 8, 482. Mueller, W. H., and Butler, P. E. (1968). J . Org. Chem. 33, 1633. Newman, M. S. (1941).J . Amer. Chem. SOC.63,2431. Newman, M.S.,and Beard, C. D. (1970).J. Amer. Chem.SOC.92,7664. Newman, M. S.,and Kutner, A. (1961).J . Amer. Chem. SOC.73,4199. Newman,M. S . , andokorududu, A. 0.M. (1968).J.Amer. Chem.Soc. 90,4189. Newman, M. S.,and Okorududu, A. 0. M. (1969). J. Org. Chem. 34, 1220. Newman, M. S.,and Weinberg, A. E. (1966).J . Amer. Chem. SOC.78,4664. Nishimura, A.,Kato, H., and Ohta, M. (1967). J . Amer. Chem. SOC.89,6083. Nishimura, A.,Ohta, M., and Kato, H. (1970). BulC. Chem.SOC. Japan 43,1630. Noyce, D. S., and De Bruin, K. E. (1968). J . Amer. Chem. Soo. 90,372. Noyce, D. S.,Matesich, M. A., and Peterson, P. E. (1967). J . Amer. Chem.SOC.89, 6226. Noyce, D. S., Matesich, M. A., Schiavelli, M. D., and Peterson, P. E. (1966). J . Amer. Chem. SOC.87,2295. Noyce, D. S., and Schiavelli, M. D. (1968s~).J . Amer. Chem. SOC.90, 1020. Noyce, D.S.,and Schiavelli, M. D. (196810). J . Amer. Chem. SOC.90,1023. Olah, G.A. (Ed.) (1966). “The Friedel-Craftsand Related Reactions”, Vol. 111. Wiley, New York. Olah, G. A., andComieaxow,M. B. (1966).J . Amer. Chem.Soc. 88,3313,4442. Olah, G . A.,Lukas, J., and Lukrss, E. (1969). J . Amer. Chem.SOC.91,5319. Olah, G.A.,Kuhn, 8.J., Tolgyesi, W. S., and Baker, E. B. (1962).J . Amer. Chem. Soc. 84,2733. Olah, G. A., Moffatt, M. E., Kuhn, S. J., and Hardie, B. A. (1964). J . Amer. Chem. SOC.86, 2198. Olah, G. A., Tolgyesi, W. S., Kuhn, S. J.,Moffatt, M. E., Bastien, I. J.,and Baker, E.B. (1963). J . Amer. Chem. SOC.85, 1328. Ohh, G. A., and White, A. E. (1972). I n “Carbonium Ions” (G. A. Olah, and P. v. R.Schleyer, Eds.). Vol. IV. Wiley-Interscience, New York.
278
QIORQIO MODENA A N D UMBERTO TONELLATO
Oleon, A. R., and Hyde, J. (1941). J. Amer. O h m . SOC. 63,2469. Om, W. L., and IULeresch, N. (1963). J. Amer. Chem. Soo. 75,0030. Orr, W. L., and Kharasch, N. (1966). J. Amer. Chem. SOC.78,1201. Owsley, D. C., Helmkamp, G. K., &d Rettig, M. F. (196s). J. Amer. C h .SOC. 91, 6239. Peer; H. G. (1962). Rec. Trav. c h h . 81,113. Perrin, C. L. (1908). Quoted by R. C. Fahey in “Topics in Stereochemistry” (E. L. Eliel, and N. L. Allinger, Ede.). Vol. 111, p. 328. Interscience, New York. Peterson, P. E., and Allen, G. (1962a). J. Org. Chem. 27, 1606. Peterson, P. E., and Allen, G. (1962b). J. Org. Chm. 27,2290. Peterson, P. E.,and Bopp, R. J. (1967). J. Amer. Chem. SOC.89,1283. Peterson, P. E., Casey, C., Tao, E. V. P., Agtarap, A., and Thompson, G. (1966). J . Amer. C h .Soc. 87,6103. Peterson, P. E., and Duddey, J. E. (1903). J. Amer. Chem.Soo. 85,2866. Peterson, P. E., and Duddey, J. E. (1966). J. Amer. Chem.SOC. 88,4990. Peterson, P. E., and Kamat, R. J. (lQ00).J. Amer. Chern. SOC.88,3162. Peterson, P. E., and Kamat, R. J. (190B). J. Amw. C h m . SOC.91,4621. Peterson, P. E.,and Indelicato, J. M. (1909). J. Amer. Chem.SOC.91,6194. Petrov, A. A., and Fedorova, A. V. (1904). Rues. Chem. Rev. 33, 1. Petrov, A. A., Porfir’eva, Yu. I., and Yakovleva, T. V. (1900). Zhur. obahohei Khim. 30, 1441. Petrov, A. A., Semenov, G. I.,and Sopov, N. P. (1967). Zhur. obehchi K h h . 27, 928. Pfeifer, W. D., Bahn, C. A., Schleyer,P. v. R., Bocher, S., Harding, C. E., Hummel, K., Hanack, M., and Stang, P. J. (1971). J. Amer. Chem.Roc. 93,1613. Pilgrim, W. R. (1969). Ph.D. Thesis. Queen’s Univ., Ontario, Canada. Pincock, J. A., and Yates, K. (1968). J. Amer. Chem. SOC. 90, 6043. Pincock, J. A., and Yates, K. (1970). C a d . J. Chem. 48, 3332. Pittman, C. U. (1909). Chem. Comm. 122. Pittman, C. U., and Olah, GI. A. (1966). J. Amer. Chem. SOC. 87, 6632. Poutema, M. L. (1968). J. Org. Chem. 33,4080. Poutsma, M. L., and Ibarbia, P. A. (1970). Tetrahedron Lettera 4907. Poutsma, M. L., and Ibarbia, P. A. (1971). J. Amer. Chem. SOC. 93, 440. Praill, P. F. C. (1903). “Acylation Reautions”, Pergamon Press, Oxford. Pudovik, A. N. (1961). Z h w . obahchei K h h . 21, 1811. Rappoport, Z. (1969). I n “Advances in Physical Organic Chemistry” (V. Gold, Ed.). Vol7. Academic Press, London. Rappoport, Z., and Apeloig, Y. (1969). J. Arner. Chem. SOC.91,0734. Rappoport, Z., and Apeloig, Y. (1970). Tetrahedron Lettera 1817. Rappoport, Z . , and Atidia, M. (1970). Tetrahedron Letters 4086. Rappoport, Z . , and Gal, A. (1969). J. Amer. Chem. SOC.91, 6246. Rappoport, Z., and Gal, A. (1970). Tetrahedron LMera 3233. Rappoport, Z . , BiirJsler, T., and Hanaak, M. (1970). J. Amer. Chem. SOC.92,4986. Rappoport, Z., and Kaspi, J. (1970). J. Amer. Chem. Soo. 92, 3220. Rappoport, Z., and Patai, 5. (1964). In “The Chemistry of Alkenes” (S. Patai Ed.). Vol. 1, p. 646. Intersoience, London. Rappoport, Z. and Ta-Shma, R. (1971). Tetrahedron Letter8 (in press). Richey, H. Q. (1970). In “The Chemistry of Alkenes” (J.Zabicky, and S. Patai, Eds.).Vol. 2, pp. 39-49. Interscience, London. ,
V I N Y L UATIONS
279
Richey, H. G., Philips, J. C., and Rennick, L. E. (1966a). J . Amer. Chem.SOC. 87, 1381. Richey, H. G., Rennick, L. E., Kushner, A. S., Richey, 3. M., and Philips, J. C. (1965b). J . Amer. Chem. SOC.87, 4017. Richey, H. G., and Richey, J. M. (1970). I n “Carboniwn Ions” (G. A. Olah, and P. v. R. Schleyer, Eds.). Vol. 2, pp. 899-967. Wiley-Interscience,New York. Robertson, P. W., Daeent, W. E., Milburn, R. M., and Oliver, W. H. (1960). J . Chem. SOC.1628. Robson, E., and Tedder, J. M. (1963). Proo. Chem. Soo. 344. Santelli, M., and Bertrand, M. (196%). Tetrahedron Letters 2611. Santelli, M., and Bertrand, M. (1969b). Tetrahedron Letters 2616. Santelli, M., and Bertrand, M. (196%). Tetrahedron Letters 3699. Sasaki, T., Eguchi, S., and Toru, T. (1968). Chem. Comm. 780. Sasaki, T., and Kojima, A. (1969). Tetrahedron Letters 3639. Sohiavelli, M. D., Hixon, S. C., and Moran, H. W. (1970). J. Amer. Chem. SOC. 92, 1082. Schmid, G. H., and Heinola, H. (1968). J . Amer. Chem. Soo. 90, 3466. Schubert, W. M., and Barfnecht, 0. W. (1970). J . Anaer. Chem.Soo. 92,207. Scott, F. L., and Butler, R. N. (1972). I n “Carbonium Ions” (G. A. Olah, and P. v. R. Schleyer, Eds.). Vol. 4. Wiley-Interscience, New York. Seel, F. (1943). 2. A w g . Chem. 252,24. Sherrod, 8.A., and Bergman, R. G. (1969). J . Amer. Chem. SOC.91, 2116. Sherrod, S. A., and Bergman, R. G. (1971a, b). J. Amer. Chem.S00.,93,1926,1941. Shiner,V. J., Buddenbaum, W. E., Mum, B. L., and Lamaty, Q. (1968). J . Amer. O h m . Soo. 90, 418. Shiner, V. J., Wilson, J. W., Heinemann, G., and Solliday, N. (1962). J . Amer. Chem. SOC. 84,2408. Shiner,V . J., and Wilson, J. W. (1962). J . Amer. Chem. SOC.84, 2402. Shriner, R. L., Fuson, R. C., and Curtin, D. Y. (1966). “The SystematicIdentification of Organic Compounds”, IV ed. Wiley, New York. Skrabal, R. (1939). 8.phys. Chem. A185,81. Sinn, H., Hopperdietzel, S., and Sauermann, D. (1965). Monateh. 96, 1036. Smith, P. A. S. (1963). In “Molecular Rearrangements” (P. de Mayo, Ed.). P. 467. Interscience, New York. Sollenberger,P. Y. and Martin, R. B. (1970). J . Amer. Chem. Soc. 92,4261. Stamhuis, E. J., and Drenth, W. (1961). Rec. Trav. chim. 80, 797. Stamhuk, E. J., and Drenth, W. (1963a). Reo. Trav. d i m . 82, 386. Stamhuis, E. J., and Drenth, W. (1963b). Rec. Trav. chim. 82, 394. Stang, P. J., and Summerville,R. H. (1969). J. Amer. Chem.SOC. 91,4600. Stieglitz, J., and Leech, P. N. (1913). Ber. 46,2147. Storr, R. C. (1969). I n “Organic Reaction Mechanisms” (E. Capon, and C. W. Rees, Eds.). P. 179. Interscience, London. Streitwieser, A. (1966). Chem. Rev. 56,671. Streitwieser,A., Wilkins, C. L., and Kiehlmann (1967). J . Amer. Chem. SOC.90, 1698. Su, T. M., Sliwinski, W., and Schleyer, P. v. R. (1969). J . Amer. Ohm. SOC.91, 6386. Sustmann, R., Williams, J. E., Dewar, M. J. S., Allen, L. C., and Schleyer, P. v. R. (1969). J . Amer. C h .SOC. 91, 6360. Taylor, D. R. (1967). Chem. Rev. 67,317.
280
GIORGIO MODENA A N D UMBERTO TONELLATO
Thaler, W. A. (1969). J . Org. Chem. 34, 871. Traynard, J. C. (1962). Bull. SOC.chim. France 19. Treffers, H. P., and Hammett, L. P. (1937). J . Amer. Chem. SOC.59, 1708. Truce, W. E., and Boudakian, M. M. (1956). J . Amer. Chem. SOC.78, 2748. Ugi, I., Beck, F., andFetzer, U. (1962). Chem. Ber. 95, 126. van Boom, J. H., Montijin, P. P., Brandsma, L., and Arens, J. F. (1965). Rec. Trav. chim. 84, 31. Vernon, C. A. (1954). J . Chem. SOC.423. Verny, M.,and Vessiere, R. (1969). Bull SOC.chim. France 1729. Viehe, H. G. (1969). In “Chemistry of Acetylenes” (H. G. Viehe, Ed.). Pp. 861-912. Dekker, New York. Volpin, M. E., Koreshkov, Yu. D., Dulove, V. G., and Kursanov, D. N. (1962). Tetrahedron 18, 107. Waters, W. A. (1936). “Physical Aspects of Organic Chemistry”, p. 346. Van Nostrand, New York. Wilson, J. W. (1969). J . Amer. Chem. SOC.91, 3238. Wilson, M. H., and Berliner, E. (1971). J . Amer. Chem.SOC. 93, 208. Winstein, S., Clippinger, E., Fainberg, A., and Robinson, G. C. (1954). J . Amer. Chem. SOC.76,2597. Winstein, S., and Heck, R. (1956). J . Amer. Chem. SOC.78, 4801. Winterfeldt, E. (1969). I n “Chemistry of Acetylenes” (H. G. Viehe, Ed.). Pp. 267-334. Dekker, New York. Woodward, R. B., andHofhann, R. (1969). Angew.Chem. Internat. Edn. 8,781. Yonezawa, T., Hakatsuji, H., and Kato, H. (1968). J . Amer. Chem.Soc.90,1239. Yen, V.-Q. (1962). Anmlen 785. Yukawa,Y., andTsuno,Y. (1959a, b). Bull.Chem.Soc. Japan 32,965; 32,971.
AUTHOR INDEX Numbers in italics refer to the pages 012 which references are listed at the end of each artkle. Figures in parentheses are reference numbers.
A
Balasubramanian, A., 163, 174 Baliga, B. T., 174 Ballentine, A. R., 207, 226, 274 Abramovici, M., 142,174 Bamford, C. H., 130, 134,174 Abramovitch, R . A., 142, 174 Banes, E. M., 27(366), 122 Adamic, K., 148, I74 B d e l d , F. H., 164,174 Adams, G. E., 142,171,174 Banica, R., I24 Agtarap, A., 192,200,278 Banthorpe, D. V., 186,274 Aksnes. G., 27(40), 29(40), 123 Barb, W. G., 134,174 Albery, W. J., 174 Barfnecht, G. W., 238,244,279 Aleksandrov, A. L., 163,176 Barr, J., 6, 7, 9, 12, 13, 14, 16, 23, 24 Allen, G., 192, 264, 278 Bartell, L. S., 27(26a,b), 28(26a,b). 42 Allen, L. C., 279 (26a,b), 63(26a,b), 82(26a,b), 122 Andersen, K. K., 182 Bartlett, P. D., 129, 131, 174, 178 Anderson, W. R., Jr., 161, 166,174 Bascombe, K. N., 16, 23 Andrews, L. J., 146,177 Basolo, F., 119(133), 126 Andronov, C. M., 168, 166, 168,174 Bass, S. J., 6, 23 Anschel, N., 101(120a,b), 125 Biissler, T., 237, 238, 241, 244, 275, 278 Apeloig, Y., 244, 246, 278 Baatien, I. J., 273, 277 hens, J. F., 217,280 Basu, S., 163,175 Armstrong, N. R., 163,182 Bauer, J., 27(11e), 78(11e), 121 Arrhenius, S., 167, 174 Baver, R. F., 163, I74 Arthur, N. L., 133,166,174,177 Bazilevskii, M. V., 132,175 Atidia, M., 237, 244, 263, 267, 278 Beard, C. D., 233, 277 Atkins, P. W., 166,174 Becconsall, J. K., 161,175 Ayscough, P. B., 138, 140,168,174 Beck, D., 27(30), I22 Azori, M., 130, 182 Beck, F.. 271,280 Becker, E. D., 160, 161, 163,174, 177 Behrman, E. J., 131,175 Bell, R. P., 16, 142, 143, 169, 23, 173, 175 Bellamy, L. J., 163, 163, I75 Bennett, J. E., 131, I75 Baciocchi, E., 148, 161,174 Benson, S. W., 129, 141, 142, 167, 175, Back, M. H., 173,174 177 Baeyer, A., 186,274 Bagdasaryan, Kh. S., 128, 129, 130, 134, Bent, H. A., 28 (61), 68(61), 123 Berezhnykh-Foldes,T., 128, 166, 167,175, 142,174,182 182 Bahn, C. A., 236,241,248,278 Berezin, I. V., 133, 167, 169, 175 Baker, A. W., 160,174 Bergesen, K. 27(40), 29 (40), 123 Baker, E. B., 273,277 Bergman, R. G., 237, 241, 244, 266, 276, Baker, K. M., 131,162,174 279 Baker, R. T. K., 174 Balaban, A. T., 60(88a,b,c,d), 214, 273, Berliner, E., 208, 211, 280 Bernard, D., 80(110), 125, 126 274 Bernstein, A. J., 68(98), 125 Balendin, A. A., 174 281
B
282
AUTHOR INDEX
Bern~tein.H. J., 160,181
Brownlie, I. T., 138, 148, I75 Berry,R.S.,27(63),28(63),40(63),42(63), Brubeker. C. H. Jr., 124 77(63), 82(63), 119(63), 123 Bertrand, M., 224, 226,226, 227, 274, 279 Bethell, D., 186,266,273,274 Beveridge, D. L., 60(83), 68(83), 124 Bevington, J. C., 134,175 Bh&r, K. R., 140,167,182 Bhatia, S. B., 27(11c), 78(11c), 80(111b), 98(111b), 100(111b), 121,125 Bhowmik, B. B., 163,164, I75 Bbis, J., 168,175,177.181 Bianchini, J. P., 221,223,274 Bickel, A. F., 130, 137, 175,179 Biczb, G., 160, 175 Bigeleisen, J., 7, 24, 142, 175 Bigler, A. J., 27(11a), 29(67a,b), 78(11a), 124(67a.b), 121, I24 Birchall, T., 18, 23 Bird, R. A., 137, 139, 140, 142, 146, 149, 176 Binun, G. H., 27(12), 100(12), 121 Blade-Font, A., 27(29c), 122 Blake, J. A., 147, I75 Bly, R. S., 207,226, 227,274 Bocher, S., 203,204,236,241,248,275,278 Bodenstein, M., 137,175 Boer, F. P., 273, 274 Boerwinkle, F., 212,275, 276 Bolland, J. L., 136, I75 Bollinger, M.J., 21,24 Bolton, P. D., 148, I75 Bolton, R.,223,268,274 Boozer, C. E., 136, 137, 138, 146, 148, 169, 175,178 Bopp, R. J., 192, 199,278 Borkent, G., 196,274 Borowitz, I. J., 101(120a,b), 125 Bott, K., 274 Bott, R. W., 188,189, 192,274 Boudakian. M., M., 214,280 Boyd, D. B., 69(93), 68(93), 125 Brand, J. C. D., 7, 23 Bmndsma, L., 217,280 Brahk, S., 167,175 Breitenbeoh, J. W., 142, 175 Breslow, R., 262,274 Bridger, R. F., 130, I75 Briers, F., 132, 176 Brigge, E. R.,130, 142,183 Brodskii, A. I., 142,164,175,179 Brook Hart, M., 18,23 Brouwer, D. M., 20,23,273,274 Brown, D. H.. 27(27), 122 Brown, H. C., 146, 160, 175, 180, 273, 274
Buchachenko,A.L., 141,164,168,176,179, 182 Buchanan, I. C., 166,174 Buddenbaum, W. E., 279 Bullock, G., 166, 166,176 Bunn, D., 166,176 Burawoy, A., 268,269,274 Burchill, a. E., 166,176 Burgada, R., 80(llO), 111(122a), 112 (l22a), 116(122a), 116(128), 125, 126 Burighel, A., 236, 237, 260, 274 Burnett, G. M., 134, 168, I76 Butler, P. E., 230, 231, 277 Butler, R. N., 271, 279 Buttrill, S. E., 264, 274 Bunnell, C. A., 218, 274
C Cahn,R.S.,36(76),86(76),91(76),92(76), 124 Calo V., 214,274 Caldin. E. F., 173,176 Caldwell, R. Q., 137, 138, 146, 176 Calin, M., 24 Campbell, A., 272, 274 Cannon, C. G., 169, I76 Capozzi, G., 249,260,266,274 Cmlton, T. S., 166,176 Carter, R. P. Jr., 27(24a,b). 82(24a,b), 122 Caaey, C., 192,200,278,277 Caughlan, C. N.. 27(41), 29(41), 32(68), 33(69), 91(69), 123,124 Chakrovorty, K., 131, I76 Chang, B. C., 27(14c), 110(14c), 122 Chmg, H. W., 262,274 Chapman, D. L., 132,176 Charleston, B. S., 220, 274 Cheburkov, Yu. A., 76(108a,b), 101 (108a,b), 125 Chekunov, A. V.. 163,187 Chen, H. L., 143, I79 Cher, M., 166, 176 Chorvat, R. J., 27(61), 29(61), 123 Chou, C. C., 131, I76 Christensen, R. L., 166,176 Christian, S. D., 168, 164, I76 Christiansen, J. A., 167,176 Clark, R. J., 126 Clementi, E., 69(89), 68(89), 124 Clippinger, E., 249,280 Closson, W. D., 206,208,274
283
AUTHOR I N D E X
Colson, J. G., 272, 276 Cornismow, M. B., 1,24,273,277 Cook,R. L., 162,163,176,178 Cooper, R., 166, 166,176 Coppinger, G. M., 138,176 Corbridge, D. E. C., 36(72), 124 Coffield, J. R., 27(60b), 29(60b), 123 Coskran, K. J., 97 (109a), 125 Cotton, F. A,, 126 Cottrell, T. L., 136, 176 Covitz, F., 27(46), 27(47d), 29(46), 123 Cowley, A. H., 118(132a), 118(132a), 126 Cox, J. R. Jr., 27(47e,b), 123 Craig, D. P., 28(64), 68(64), 68(97), 124, 125 , Crandall, J. K., 218,274 & m o , J. C., 137, 181 Crawford, B. Jr., 162,176 Creak, G. A., 129,176 Cremer, S. E., 27(61), 29(61), 123 Crooks, J. E., 160,176 Cruickshmk, D. W. J., 68(99), 125 Crutchfield, M. M., 68 (103), 125 Ceeh, G., 186,237,243,275 Csizmadia, I. G., 264,255,275,276 Cupas, C. A., 1,24 Curci, R., 158, 176 Curtin, D. Y., 186,231,274,279
D Dainton, F. S., 129, 176 Dalchau, S., 144, 178 Dalton, C. K., 220,874 Daly, J. J., 36(71), 124 Danti, A., 126 DaRooge, M. A., 139, 141, 146, 146, 163, 164,176,180 Das, R. C., 166,176 Daaent, W. E., 208,279 Daves, G. D. Jr., 175,175 Davies, D. S., 137, 146, 152, 176 Davis, G. T., 166,182 Day, A. C., 234,274 De, 5. K., 166,180 Dean, P. A., 17,23 Dearden, J. C., 168,176 DeBruin, K.E., 27(48b,c), 29 (48b,c), 123, 188, 189,277 De Clercq, M., 60(87a), 124,126 Deiters, R. M., 27(25a,b,c), 122 De Jeu, W. H., 161.176 de la Mare, P. B. D., 223,268,274 De Mad,G. R., 132,176
Denes, A, S., 267,275 Deniaov, E. T., 163,176 Denney, D. B., 27(14a,b.c). 110(14a,b,c), 131,122,176 Denney,D. Z.,27(14c), 110(14c), 131,122, 177 Dennis, E. A,, 27(39), 27(45), 29(39), 29(46), 123 Deno, N. C., 21, 23, 273,276 D e Palme, D., 164,181 Dershowitz, S., 27(6a), 27(6b), 27(6c), 100(6a,b), 121 Desai, N. B., 27(6a,c,36a,b,37,38), 29 (37,38), 100(6a,~), 101(68,C), 109 (129), 119(129), 121,125,126 Detoni, S., 162,176 Dever, J. L., 27(12), 100(12), 121 Dewar, M. J. S., 186,264,275,278 Deyrup, A. J., 2, 4,24 Di Nunno, L., 213,214,275 Dierdorf, D. S., 118(132a), 118(137a), 126 Dimroth, K., 73(106), 125 Diprete, R. A., 168, 176 Doak, G. O., 126 Dobis, O., 133,167, 169,175, 176 Dolgoplosk, B. A., 142,176, 177 Dondoni, A., 213,214,275 Dos Santos, J., 168, 162,176,177 Downs, A. J., 69(90), 68(90), 124 Drago,R. S., 69(92), 68(92), 161,125,183 Dreeskamp, H., 126(132b), 119(132b), 126 Drenth, W., 188, 189, 190, 191, 196, 199, 261,274,276,279 Dubrin, J., 131,177 Duddey, J. E., 192, 199, 267, 261, 263, 278 Deuber, T.E., 236,237,247,275,276 Dulova, V. G., 266,280 Dungan, C. H., 68(103), 125 Dunitz, J. D., 60(86),124 Dunn, M., 148,174
E Eaborn, C., 188, 189,192,274 Eastman, J. W., 168,177 Echols, J. T., 130, 181 Edwards, J. O., 131, 168,175,176 Eguchi, S., 201, 279 Ehrenson, S., 60(86), 124 E l i m n , M. A., 172,177 Eliel, E., 122 Ellis, J., 148, 175
284
AUTHOR INDEX
Ellison, H. R., 163, I77 Emanuel, N. M., 138, 168, 164, 168, 174, I83 Engberts, J. B. F. N., 163,177 Englund, B., 111(123a), 123 Esson, W., 173,178 Evans, M. G., 177 Evans, M. J. B., 136, 147,175
F Fahey, R. C., 196, 197, 198,275 Fainberg, A., 249, 280 Farcario D., 60(88a), 124 Farcaaio, D., 60 (88a), I24 Farkaa, A., 129,177 Farkas, L., 129,177 Farmer, M. L., 231,274 Fedorova, A. V.. 217, 221, 230, 275, 278 Felder, W., 131, 177 Fendley, T. A., I75 Ferguson, K. C., 166,177 Ferris, F. C., 141, 163,180 Feesenden, R. W., 126 Fetzer, U., 271, 280 Firstenberg, S., 101 (120b), 125 Fischer, H., 266,275 Fisher, I. P., 263, 275 Fischer, J. P., 168,177 Fletcher, A. N., 168, 161, 162, 164, 177 Flowers, R. H., 6,9, 23 Foldes-Berezhnykh, T., 130, 146, 177 Foner, S. N., 136, 177 Fox, I. R., 166, 182 Frank, D. S., 27(43), 29(43), 123 Fraser, G. W., 27 (27), 122 Friederang, A., 212, 275 Friedmann, J., 132,178 Friedrich, S. S., 146,177 Frydman, N., 246,275 Fuller, D. L., 129, 130, 181 Funderburk, L. H. 143,166,179 Fung, M. K., 73(106c), 77(106c), 125 Furtsch, T. A., 118(132a), 119(132a), 126 Fuson, R. C., 186, 279
G Gagnaire, D., lll(122b), 112(122b), 116 (122b), 126
Gal, A.. 237,246,247,262,270,278 Gallagher, M. J., 27(34), 122 Gambaryan, N. P., 76 (108a), 76 (108b), 101(108a,b), 125 Gardner, D. V., 138, 139, 148,177 Geib, K. H., 129,177 Geiseler, G., 166,180 Gergely, J., 160,175 Gerhold, G. A., 164,179 Gielen, M., 60(87a,b), 124, 126 Gilbert, B. C., 166,180 Giles, R. D., 132, I77 Gillespie, P., 27(66-69), 28(66-69,63), 36, 38, 40(66), 42(68), 60, 68, 63, 66(81), 73(66-68), 76(69), 76, 80, 86, 86, 96 (66-69), 117, 118(66), 123,124 Gillespie, R. J., 1,4,6,6,7,9,11, 12,13,14, 16,16,17,18,21,22,23,28(66),69(81), 68(63, 66), 82(68). 84(68), 86(63), 124 Gillespie, R. J., 273, 275 Gilliom, R. D., 146, 177 Ginns, I. S., 166, 176 Glaser, S. L., 27(11a), 78(11a), 121 Glaastone, S., 172, I77 Glockling, F., 166, I77 Gnanadesian, S. G., 162,181 Godsay, M. P., 140, 142, 146, I77 Gokel, G., 27(66), 28(66), 36(66), 38(66), 40(66), 42(66), 44(66), 73(66), 77(66), 80(66), 96(66), 123 Gold, V.. 160,177, 186, 266,273, 274 Golden, D. M., 142,177 Goldschmidt, S., 136, 177 Goldsmith, H. L., 137, 146, 162,176 Goldstein, M., 161, 177 Goldwhite, H., 111(124), 126 Golen, J. A., 27(26c), 43(79), 122, 124 Golomb, S. W., 36(73b), 124 Goodall, D. M., 142, 143,175 Goodrich, R. A., 63(96), I25 Gorenstein, D. G., 36(77), 37(77), 38(77), 96(118a,b), 96(77), 104(77), 107 (118,b), 124,125 Gough, S.T.,27(14b), 110(14b), I22 Gragerov,I. P., 131, 163, 164, 177 Gramstad, T., 161, 162,177 Gray, P., 133, 136, 166, 166,174,177 Greco, C. V., 27(37b), 29(37b), 122 Green, M., 27(31), 122 Gregg, R. A., 134,169,178 Greiner, N. R., 171, I77 Griesbaum,K., 194,196,196,217,220,221, 229,275 Griffith, E. J., 122 Griffiths, J. E., 27(24a), 82(24a), 122 Grob, C. A., 186,236,237,242,243,275
AUTHOR I N D E X
Gruege, F., 162,177 Grunwald, E., 169, 171,180 Guggenheim, E. A., 136, 173,177,178 Guillemonat, A., 221,223,274 Gulati, A. S., 29(66b), 33(66b), 76(107), 80(111b), 89(66b), 98(111b), 100 ( l l l b ) , 101(107), 116(127c,d), 118 (66b), 124,125,126 Gunning, H. E., 131,178 Gupta,A. K., 137, 146,162,176 Gurd, R. C., 166,174 Guyer, J. W., 125
H Haaf, W., 273,276 Haake, P. C., 27 (36), 28 (36), 29 (36), 122 HadEi, D., 177 Hiifher, J., 203, 224,226, 275 Hakatsuji, H., 264, 280 Hall, F. M., 148, 175 Hamilton, C. E., 136, 137, 138, 146, 148, 169,175,176 Hamilton, W. C., 28 (6011, b, c), 29 (60a, b,o), 30(60a,b,o), 123 Hammett, L. P., 2,4,24,272,273,275,280 Hammond, G. S., 131, 133, 136, 137, 138, 146,148,169,175,178,183 Hanack, M., 186, 203, 204, 206, 206, 224, 226, 236, 237, 238, 241, 244, 248, 266, 266,275,277,278 Hanna, S. B., 143,178 Hansen, B., 27(36a), 122 Hansen, K. W., 27(26a,b), 28(26a,b), 42 (26a,b), 63(26a,b), 82(26a,b), 122 Hantzsch, A., 4, 20, 24, 186,275 Hardie, B. A., 273, 277 Herding, C. E., 236, 241,248, 278 Hargrove, R. J.. 237,247,275 Haque, M. U., 33(69), 91(69), 124 Harary, F., 60(88a), 124 Harcourt, A. V., 173,178 Harle, 0. L., 164, 178 Harpell, G. A., 137, 139, 140, 142, 175, 177 Harris, G. S., 27(16), 122 Harrison, A. G., 166,181,263,275 Hassner, A., 212, 275, 276 Harteck, P., 129,177 Hartland, E. J., 161,178 Hartter, D. R., 2, 24 Hasselbarth, W., 36(74), 42(74), 124 Hawser, K. H., 148,178 Hauteoloque, S., 170,178 Hawes,W.,27(6Oe), 29(60), 123
285
Hawranek, J., 162,177 Heck, R., 199,207,262,276,280 Heckner, K. H., 144,178 Heidt, J., 129, 164, 181 Heilman, W. J., 133, 142, 178 Heinemann, G., 269, 279 Heinola, H., 214, 279 Heitsch, C . W., 73(106b), 77(106b), 125 Hekkert, G. L.,189,199,276 Heller, C. A., 162, 177 Heller, S. R., 27(38a,b), 29(38a,b), 80 ( l l l a ) , 98(111a), 100(111a),100(114), 123,125 Hellwinkel, D., 27(2), 27(4a,d), 29(2), 114 (2,4a,b,c,d), 118(2,4a), 121 Hehnkamp, G. K., 216,278 Hennion, G. F., 269,276 Herk, L., 133, 182 Herkes, F. E . . 132, 178 Herod, A. A., 166, 166,177 Herterich, I., 203, 206, 275 Heumann, A., 204,206,275 Higley, S. W., 17, 24 Hildebrand, K., 118(132b), 119(132b), 126 Hill, A. J., 270,276 Hine, J., 276 Hirata, T., 131, 178 Hirschfelder, J. O., 172,177 Hixon, S. C., 270,279 Hobbs, M. E., 162,163,183 Hodge, J. D., 23 Hogg, J. S.,141,146,148,149,162,178 Hoffmann, H. M. R., 237,276 Hoffmann, P., 27, 28, 73(66, 68, 69), 76(69), 77(66, 68, 59), 82, 84(68), 86(66), 86(69),96(66,68,59), 117(66), 118(66), 123 Hoffmann, R., 196,220,264,276,280 Hogeveen,H., 188,189,190,191,273,274, 275,276 Hogg, D. R., 213,276 Holly, S., 140, 142, 143, 146, 168, 165, 166, 181 Holmes, R. R., 27(24a,b), 27(26a,b,c), 43(79), 82(24a,b), 122, 124 Homer, J., 161, 178 Hong, H. K., 163,182 Hopkinson, A. C., 264,266,266, 276 Hopperdietzel, S., 213, 279 Horeisohy, K., 142,175 Horner, L., 27(30), 138,122,178 Horning, W. C., 7, 23 Horsfield, A., 166,180 Houalla, D., 111(122b), 112(122b), 116 (122b), 126
AUTHOR INDEX
286
House, J. E., Jr., 162, 163,176,178 Houeer, J. J., 23 Hover, R.. 262,274 Howard, J.A., 134,137,138,139,141,146, 146, 147, 148, 161, 162, 164, 168, 169, 166,166,167,168,177,178 Howles, J. R., 146, 177 H u n g , S. J.. 262,276 Hudson, R. F., 27(31), 68(101), 122, 125 Hudson, R. L., 136,177 Huggins, M.L.,168,178 Hulett, J. R.,172,175,178 Hulme, R.,16, 17, 23 Hummel, K., 203, 204, 236, 241, 266, 276, 278 Humphreys, D. A., 16,17,23 Hung, N.M., 216,277 Huong, P.V., 162,178 Hutternan, J. J., 73(1060),77(1060),126 Huybrechts, G.,132,176 Hyde, J., 273,278 Hyman, H. H.,16.24
James, D. G. L.,178 James, R.E.,128,166,178 Jenkins, A. D.,129, 130, 134, 174, 178, 179 Jenkins, I. D., 27(34),122 Jensen, F. R.,273,274 Jentschura, U.,160,179 Jermini, C., 143,178 Johnson, C. D.,2, 24 Johnson, R. N., 216, 220, 221, 230, 276 Johnston, K. M., 129,179 Johnston, R.,130,174 Jones, J.. R.,142, 143, 169,179 Jones, W.M.,233, 234, 237,246,247, 269, 262,263,276 Jorgenson, M.J., 2,24 Jousaot-Dubien,J., 164,179 Jumonville, S., 129, 130, 181
K
Kaganskaya, K. Ye., 168,175 Kagarise, R. E., 161, 162, 163, 179, 183 I Kagiya, T.,136,179 Ibarbia, P. A., 216, 219, 223, 230, 231. Kaiser, E.T.,27(470), I23 K~leahnikova,L.A., 141,179 278 Ibieter, J., 212,275 Kalatzis, E., 130,179 Ihrig, J. L., 130, 137, 138, 146, 176, Kernat, R.J., 206,208,268,278 Kampmeier, J. A., 231,274 183 Illumimti, G., 148, 161,174 Ktteai, P. H., 266,276 Imhoff, M.A., 236,239,276 Kaspi, J., 237, 270, 278 Imre, L., 162,180 Kato, H., 234,264,277, 280 Indelicato, J. M., 236, 237, 239, 262, Katritzky, A.R.,2,24 Katz, J. J., 16,24 278 Ingold, C. K., 36(76), 86(76), 91(76), 92 Ksufman, I). A., 237, 238, 244, 246, 247, (76),124,178, 186,273, 276 262,263,264,276,277 Ingold, K. U., 133, 134, 137,138, 139, 141, Kauzmm, W., 173,179 146, 146, 147. 148, 149, 161, 162, 163, Keefer, R.M.,146,177 164, 168, 169, 166, 167, 168,174,175, Kell, D.R.,201,276 177,178 Kelsey, D.R.,237,244,266,276 Inoue, T., 136,179 Kemppainen, A. E.,131,183 Ishii, T.,142,180 Ken, V. Q.,280 I&, K. J., 129,176 Kenachuk, K.. 142,174 Kende, I., 128, 130, 166,179,182 Kennedy B.A., 124 Kennedy, B. R.,141,146,179 Kenyon, J., 164,174,272,276 Kenyon, J. J., 272, 274 Jackson, C. J., 161,178 Kerepes, R.,167,177 Jacobs, T. L.,189,216,220,221,224,226, Kernaghan, a. F.P., 237,276 227,228,230,231,266 Kerr, J. A., 166, I79 J d 6 , H.H.,60(84), 124 Kerst, F,,27(42), 29(42), I23
a.
J
AUTHOR I N D E X
Kharesch, N., 213, 214, 216, 261, 276, 278 Kice, J. L., 134,179 Kiehlmann, 263,279 Kilpatrick J. E., 27(64), 123 Kilpatrick, M., 16, 17, 24 King, R. W., 73(106a,b,c), 77(106a,b,c), 125 Koch, H., 273,276 Kiousky, T. E., 24 Kirby, J. A., 146,182 Kluger, R., 27(42), 27(46), 29(42), 29(46), 123 Klusecek, H., 27(66, 67, as), 28(66, 67, 68), 36(66), 38(66), 40(66), 42(66), 44(66), 73(66, 67, 68). 77(66, 67, 68), 80(66, 67, 68). 82(68), 84(68), 96(66, 67, 68), 123 Knunymts, I. L., 76(108a), 76(108b), 101 (108a,b), 125 Kojima, A., 194, 279 Kolbe, A., 162, 166, 179 Kondo, Y.,166,179 Kondratiev, V. N., 131, 170,179 Koock, S. U.,207,226,227,274 Kooyman, E. C., 130, 137, 175, 179, 182 Koreahkov,Yu.D., 266,280 Korotkina, D. Sh., 142,176,177 Kotowycz, G., 161,181 Kouba, J., 143,180 K o v b s L., 160,181 Kreilick, R. W., 129,168,179 Kresge, A. J., 143,179 Kucherov, V. F., 217,277 Kugler, A. J., 27(44), 29(44), 116(126a,b, 127b), 123,126 Kuhn, 8.J., 273,277 Kukhtin, V. A., 27(13). 100(13), 122 Kumamoto, J., 27(47a), 123 Kumli, K. F., 27(29a), 122 K m n o v , D. N., 266,280 Kushner, A. S., 268, 279 Kutner, A., 233,277 Kuts, V. S., 164,179 Kwart, H., 129,174
L Lmksonen, E., lll(123b). 126 Ladel, J., 28(60c), 29(6Oc), 29(60c), 123 Ladik, J., 160,176 Laflair, R. T., 163.174
287
Laidler, K. J., 172, 173, 174, 177 Lsmsty, G., 279 Lamotte, M., 164,179 Lancelot, C. J., 262,276 Land, E. J., 142,171,174 Lmge, C. E., 138,179 Landsberg, R., 144,178 Lansbury, P. T., 272,276 La Placa, S. J., 28 (608,b), 29(60a, b), 30 (6h,b), 123 J. C., 162,178 L-guea, Laurence, C., 146,179 Lauterbur, P. C., 36(76), 37(76), 38(76), 44(76), 47(76), 60(76), 124 Leery, U., 161,179 Lee, D., 27(42), 29(42), 123 Lee, D. F., 144,182 Lee, D. J., 196, 197,275 Leech, P. N., 272,279 Leffler, J. E., 169, 170, 171,179 Lemanoeau, B., 168,175,181 le Noble, W.J., 269, 276 Le Roy, D. J., 129,179 Le Roy, L. R., 173,179 L e w d , M. V., 262,276 Lester, G. R., 137,146,176 Letcher, J. H., 68(103), 125 Letsinger, R. L., 193, 276 Levy, G. C., 18.23 Lewis, E. S., 143,179 Lewis, G. N., 7,24 Lewis, I. C., 166,166,182 Leyland, L. M., 169,179 Lezine, V. P., 166,183 Lien, A. P., 17, 24 Limbach, H. H., 161,179 Lippert, E., 160,179 Lipscomb, W. N., 69(93), 68(93), 125 Liu, J. S., 23 Lloyd, A. C., 133,177 Lloyd, W. G., 138,179 Loewengmt, G. V., 27(11d), 78(11a), 121 Lohmann, D. R., 141, 146, 148, 149, 162, 177,178 Long, F. A., 3, 7, 24, 143, 171, 181 Longridge, J. L., 143,180 Loeeing, F. P., 263,268,275,276 Loudon, A. U., 264,276 Lowe, J. P., 161,180 Luboraky, F., 17,24 Lucas, J., 273,276 Lukas, E., 263,277 Lukee, J., 263,277 Lum. K. K., 146,180,182
AUTHOR INDEX
288
M McCauley, D. A., 17,24 McClellan, A. L., 167, 160, 180 McCoubrey, J. C., 16,23 McDaniel, D. H., 146, 180 McDonel, E. T., 137,181 McEwen, W. E., 27 (29a, b,c), 27 (33), 28(33), 122 McGowan, J. C., 138, 142, 146, 148, 149, 162, I80 McLauchlan, K. A., 166,174 McPherson, A., 198,275 McQuarrie, D. A., 124 McQuillin, F. J., 201,276 Maccoll, A.,28(64), 68 (64),124,264,276,277 Mackenzie, H. A. E., 273,277 Mackor, E. L., 20,21,23 Macomber, J. D., 120(137), 126 Macomber, R. S., 224,226,227,228,231,276 MacLean, C., 20,23 Madan, 0. P., 27(36b), 27(38c), 27(41), 29(38c), 29(4l), 92(116), 96(116), 100 (114), 122, 123, 125 Maarcker, A., 27(2), 29(2), 114(2), 118(2), I21 Magnuson, E. A., 68(97), 125 Mahler, W., 27(18e,b), 28(18a,b), 82 (18a,b). 86(18a,b), 122 Mahoney, L. R., 139, 141, 146, 146, 163, 164,176,180 Maizus, Z. K., 138,168,164,166,169,174, 183 Majer, J. R., 131, 169, 179,180 Malhotra, K. C., 6, 7,23 Malinowski, E. R., 161,180 Maloney, D. E., 269, 276 Mancuso, N. R., 272,276 Manesa, D. D., 234,237,246,247,269,262, 263,276 Margraf, J. H., 181 Mark, V., 68(103), 125 Marks, R. E., 142, 143, I79 Marquarding,D., 27(66,67,68,69), 28(66, 66,67,68, 69,), 36(66), 38(66), 40(66), 42(66), 73(66, 66, 67, 68, 69), 76(69), 77 (66,66,67,68,69), 80(66,66,67,68, 69), 82(68), 84(68), 86(66), 86(69), 96 (66, 66, 67, 68, 69), 117(66). l l S ( M ) , 123 Marsi, K. L., 27(14c), 27(62), 29(62), 110(14c), 122,123 Martin, R. B., 262,279 Martinez, A. G., 236,277 Matesich, M. A., 188, 189, 190, 191, 261, 264,276,277
Mathis, F., 111(122a), 112(122a), 116 (1224,126 Matsui, T., 166,179 Mavrov, M. V., 217,277 Mayence, G., I26 Mayer, I., 172, 180 Mayo, F. R., 130, 134, 166, 168, 169, 177, 180,183 Mazur, Y., 246,275 Melander, L., 142, 180 Melloni, G., 213,214,266,274,275 Melville, H. W., 132, 180 Mentrup, A., 27(30), 122 Manzinger, M., 172,180 Merivuari, K., l l l ( l 2 3 b ) , 126 Merrill, J. R., 162,180 Mesrobian, R. B., 131,182 Meyer, B. W., 163,177 Meyerson, S., 27(28), 27(36b), 122 Michael, B. D., 142, 171, 174 Migita, T., 131, 180 Milburn, R. M., 208,279 Mile, B., 131, 175 Miller, F. W., 233, 239, 247, 264, 276 Miller, L. L., 237, 244, 246, 262, 263, 276, 277 Milne, J. B., 9, 11, 12, 13, 14, 23, 24 Minoura, Y., 142,180 Miocque, M., 216,277 Mislow, K., 27(48, 49), 28(49), 29(48, 49), 36(49), 37(49), 38(49), 43(49), 123 Mitchell, H. G., 43(80), 44(80), 82(80), 119(80), 124 Modem, G., 168, 176, 200, 213, 214, 234, 236, 236, 237, 246, 249, 260, 266, 274, 275,277 Moffatt, M. E., 273, 277 Montijin, P. P., 217,280 Moran, H. W., 270,279 More O’Ferrall, R. A., 143,180 Morris, E. R., 133,166,180 Mortimer, C. T., 261, 277 Morton, M. J., 21,23 Moss, K. C., 17, 23 Mueller, W. H., 216, 230, 231, 277 Muetterties, E. L., 27 (18a,b), 27 (22, 23), 28(18a,b), 40(78), 42(78), 44(78), 82 (18a,b), 86(18a,b), 122,124 Mukherjee, S., 166,180 Miiller, G., 166, 180 Mullins, C. B., 161,177 M m , B. L., 279 Murrell, J. N., 167,180 Musher, J. I., 69(91), 68(91), 125
AUTHOR INDEX
N
289
Palit, S. R., 166, 180 Palmer, T. F., 253,275 Naegele, W., 194, 196, 220, 221, 275 Panar, M., 27(47c), 123 Nagabhuehanam, M., 109(121), 114(121), Parfenova, G. A., 142, I77 Parker, A. J., 168,169,180 125 Nonnan, S-A. M. A., I80 Passmore, J., 22, 23 Patai, S., 186, 278 Namanworth, 18,20,24 Patwardhan, A. V., 27(37b,38a,b, 44, Namba, H., 138,182 Nandi, U. S., 137, I78 107). 29(37b,38a,b, 44). 80, 98, 100 Nasillski, J., 60(87a), 124,126 ( l l l a ) . 101(107),116(126a,b, 127a.b). Nwo, F., 249,277 122,125,126 Naumann, K., 27 (48a,b,c), 29 (48a, b , ~ ) , Paul, M. A., 3, 7, 24 Pauling, L., 69(94), 125 123 Peulson, D. R., 218, 274 Nave, P. M., 144,180 Nazy, J. R., 193, 276 Payne, D. S., 27(16, 17), 122 Neiman, M. B., 141,168,176,179 Peake, 8.C., 118(131c), 126 Nelson, J., 164,180 Pearson, J. M., 131, I76 Nemes, I., 167,177 Pemson, J. T., 166,177 Nenitzeffiu. C. D., 273, 274 Peerson, R. G., 119(133), 126 Peer, H. G., 229,278 NeszmBlyi, A., 160, 162, I80 Peddle, G. J. D., 126 Newman, M. S., 233,273,277 Peel, T. E., 6, 9, 16, 23 Nkhimura, A., 234,277 Perkms, M. J., 166,180 Norman, R. 0. C., 166, I80 Noyce, D. S., 188, 189, 190, 191, 192, 261, Perrin, C. L., 262,278 266,277 Peterson, C., 169.175 Peterson, G. E., 27(24b), 82(24b), 122 Nurrenboch, A., 73(105), 125 Nyholm, R. S., 28(64), 28(66), 68(64,66), Peterson,P.E., 188,189,190,191,192,199, 200, 206, 208, 236, 237, 239, 267, 258, 124 261,262,263,277,278 Petersen, R. C., 181 Petrhnek, J., 141, 161, 163,180 0 Petrov, A. A., 217,230,269,275.278 Pez, G. P., 12, 13, 16, 23 Oae, S., 138,182 O’Brien, D. H., 24 Pfmndler, H. R., 243,275 Pfeifer, W. D., 236, 241,248, 278 O’Connor,B. R., 231, 274 Oftendahl, E. N., 193,276 Philips, J. C., 268, 279 Ohta, M., 234, 277 Pfohl, S., 27, 28, 73(66, 68, 69), 76(69), Okada, Y., 131,180 77, 80(66, 68, 69), 82, 84(68), 86(66), Okamoto, Y., 146, 160, I75 86(69), 96(66, 68, 69), 117, 118(66), Okorududu, A. 0. M., 233,277 123 O l d , G. A., 1, 18, 19, 20, 21, 24, 268, 273, Phillippe, R. J., 162, 163,183 Pickai, R., 162, 181 277 Pierce, S. B., 63(96), 125 Oliver, W. H., 208, 279 Olson, A. R., 273,278 Pilaf, J., 141, 180 Onyon, P. F., 134,174 Pilgrim,W. R., 266,278 Orgel, L. E., 28(64), 68(64), 124 Pilot, J. F., 27(11a), 27(66), 27(69), ZS(66, 69), 29(66a,b), 32(68), 33(66a,b), 33 Orr, W. L., 213,278 Oubridge, J. V., 9, 23 (69), 73(66. 69), 76(69), 76(66a,b). 77(66, 69), 78(11a), 86(66), 86(66), Ouchi, K., 12, 13, 16, 23 86(69), 89(66a,b), 91(69). 92(116, Owsley, D. C., 216, 278 116), 96(116, 116), 96(66, 69), 98 ( l l l b ) , 100(111b), 110(67a,b), 117 (66), 118(66a,b), 118(66), 121,123 P Pimentel, G. C., 167, 160, 180 Packer, K. J., 27(18b), 28(18b), 82(18b), Pincock, J. A., 208, 210, 268, 261, 262, 86(18b), 122 278
AUTHOR INDEX
290
96(116, 116, 117), 96(10, 66-69), 98 (10, lll), lOO(6-lo, 111, 114), 101 (10, 107, 119), 104(117), 106(117), 107(114-117), 109(67, 121, 129), 110(66), 114(121), IlS(ll0, 126, 127), 117(66), 118(10,66,66,129), 121,122, 123,124,125,126 Rao, C. N. R., 140,167,163,164,166,174, 181,182 Rapp, A., 27(30), 122 Rappoport, Z., 186,200,236,237,238,244, 246, 247, 262, 263, 266, 267, 270, 271,278 Rathke, J. W., 125 Raw, R., 138, 146, 148, 149, 162, 180 Reddy, 0. S., 118(131b), 126 Redl, G., 126 h e , I. H., 163, 181 Rehfeld, S. J., 168,177 Rehman, Z., 196, 275 Reioh, L., 134, 181 Relles, H. M., 27(14a), 110(14a), 122 Rembaum, A,, 133,142,278 Renn, K., 136,177 Rennick, L. E., 268,279 Rettig, M. F., 216, 278 Richey, H. G., 23, 186, 191, 217, 269, 262, 264,267,268,269,278, 279 Richey, J. M., 186,191,217,269,262,264, 267,268,269,279 Richter, B., 36(74), 42(74), 124 Ridley, B. A., 129,180 Rieber, M., 27(1), 121 Ritchie, C. D., 146,169, 181 Quee, M. Y., 166,181 Quartermain,L. A., 16,24 Robb, J. C., 168, 169, 179,180,181 Quickert. K. A., 129,180 Robert, J. B., 111(122b), 112(122b), 116 (122b), 126 Robertson, P. W., 208,279 Robinson, B. H., 160,174,176 Robinson, E. A., 4, 6, 6, 9, 11, 16, 23, 273, 275 Rabinowits, I., 270, 276 Robinson, G. C., 249,280 Ragimova, A.M., 167,176 Robimon. J. K., 179 Raimondi, D. L., 69(89), 68(89),124 Robson, E., 267,279 Rammathan, N., 27 (6b,c 37), 29 (37a,b, c), Rochester, C. H., 163,181 100(6b,c), 101(6b,c), 121,122 Roebeck, C. F., 273,276 Rameswamy, K., 162,181 Rokhlin, E. M., 76(108b), lOl(lO8b), ram ire^, F., 27(6-11, 28, 36-38, 41, 44, 125 66-69), 28(66-60), 29(4, 8, 9, 10, 37, Roman, S. A., 206,208,274 38, 44, 60, 66), 30(60), 32(68), 33(66, Rosantsev, E. G., 179 69), 36(66), 36(76), 38, 40, 42(66), Rosenberg, H. M., 131,181 44(56, 76)s 47(76), 60(76), 73(66-69), ROBB, S. D., 181 76(9,10,69), 77(56-69, 66), 78(7-ll), Rossclll, B., 163,181 80(66-69, lll),82(68), 84(68), 86(66, Rothenbury, R. A., 12,13, 14,24 66), 86(69), 89(66. 113), 91(69), 92, Roualt, M., 27(16), 122
Pineau, P., 168,162,177 Pines, H., 142,174 Pittmm, C. U., 1,21,23,24, 216,217,268, 273,275,278 Pitzer, K. S., 27(64), 123 Pogorelyi, V. K., 163, 164,177 Pokhodenko,V. D., 141,164,179,180 Polenyi, M., 136,177 Polya, U., 36(73a), 124 Pople, J. A., 60(82a,b,c,e), 60(83), 68 (82a,b,c,e). 60(83), 160, 124, 181 PorfVeva, Yu. I., 269,278 Porter, G., 127, 180 Pospigil, J., 140, 142, 143, 146, 146, 147, 149, 160, 161, 162, 163, 168, 166, 181, 182 Poutsma, M. L., 216,219,223,230,231,278 Powell, T.,138, 142, 146, 148, 149, 162, 180 Prctoejua, H., 162,179 Praill. P. F. (3.. 273.278 Prelog, V., 36(76), 60(86),86(76), 91(76), 92176). 124 Price, E., 166,182 Proll, P. J., 141,180 Pryor, W. A., 129, 130, 168, 180, 181 Pudovik, A. N., 269,278
0
R
AUTHOR INDEX Rowland, F. S., 131,176 Ruch, E., 36(74), 38(74), 42(74), I24 Rundle, R. E., 62(a,b,c), 68(62a,b,c), 86 (62a,b,c), 123 Russell, G. A., 130,168,175, I81 Russell,K. E., 137,138,139,140,141,142, 146, 147, 148, 149, 162, 163, 168, 174, 175,177, I78 Ryabokobylko, Yu. S., 163, 181 Ryba, O., 161,163,181
S Sagatys, D. S., 143,179 Sager, W. F., 146, 169,181 Salmi, E. J., 111(123b), 126 Sanchez, M., 111(122a), 112(122a), 116 (122a), 126 Santelli, M., 224, 226, 226, 227, 274, 279 Santry, D. P., 60(82a,d), 68(82a,d), I24 Sasaki, T., 194,201,279 Sauermann, D., 213,279 Sbar, N., 131,177 Schaefer, T., 168, 161,181 Schaleger, L. L., 171,181 Schiavelli, M. D., 188, 189, 190, 191, 192, 261,266,270,277,279 Schleyer, P. v. R., 238, 241,248,262,263, 264,277,278,279 Schmid, G. H., 214,279 Schmidt, W., 136,171 Schmutzler, R., 27(18a.b), 27(19, 20, 21a,b), 28(18a,b), 69(90), 63(19, 20), 68(90), 82(18-21), 84(19, 20), 84 (21a,b), 86(18a,b), 13l(a,b,o), 120 (19,20, 21a,b), 124,126 Schneider, W. G., 160, 181 Schott, H. N., 131,183 Schubert, W. M., 238,244,279 Schulz, G. V., 168, I77 Schutt, J. R.. 27(60b), 29(60b), 123 Schwalm, W. J., 139,164, 166, 178 Schwenk, E.. 168,178 Schwenk, R., 138,183 Schunn, R. A., 27(22), 122 Scorrano, G., 213, 214, 274, 275, 277 Scott, F. R., 271,279 Scott, R., 164,181 Searlee, S., 189,276 qegd,G.A.,60(82a,b,o,d),68(82a,b,c,d), 124
29 1
Seltzer, S., 86(60), 124 Semenov, G. I., 278 Semenov, N. N., 132, I81 Sen, J. N., 136, 137,138,148, I78 Servanton, M., 168,181 SeN6, P.,131,181 Shahin, M., 168,181 Shannon, T. W., 166,181 Shapiro, S. A., 2, 24 Shctraevskii, A.P., 132,182 Sharp, D. W. A., 27(27), 122 Shaw, R., 166.177,181 Shchedrin, V. P., 163, I76 Shelton, J. R., 137, 138, 181 Shenod, S. A., 237,241,279 Shiner, V. J., 269, 279 Shlyapintokh, V. Ya., 168,182 Shorter, J., 146,181 Shriner, R. L., 186,279 Shtraichman, G. A.,130,183 Shulgin, A. T., 160,174 Sicilio, F., 166,178 Silbert, M., 174 S i m h d i , T. L., I82 Simonyi, M., 129, 130, 140, 142, 143, 146, 146, 147, 148, 149, 160, 161, 162, 164, 168, 160, 166, 166, 167, 168, 169, 172, 180,181,182 Sknpson, A. F., 166,174 Singh, S., 138, 140, 167, 164, 166, 181 182 Sinitsyna, 2. A., 129, 130, 142, 174, 182 S b , H., 213,279 Skrabal, R., 261,279 Sleiter, Q., 148, 174 Sliwinaki, W., 263,279 Smerkolj, R., 162, 177
Smith,C.P.,27(11~,28,38~,41,44,66,69),
28(66, 69, Bob), 29(60b, 66, 67, 38c. 41, 44), 30(60b), 33(66, 69), 73(66, 59), 76(69, 66, 107), 77(66, 69), 78(11c), 86(66, 66), 86(69), 89(66), 91(69), 92, 96(116, 116), 96(66, 68), 98, 100(111b), 101(11a, 107), 109 (67, 121, 129), 110(67), 114(121), 116(26, 127), 117(66), 118(66, 66), 119(129), 121, 122, 123, 124, 125, 126 Smith, P. A. S., 271, 279 Smith, G. G., 146,180,182 Smith, K., 130,181 Sobczyk, L., 162,177 Socrates, G., 163,182 Sollenberger, P. Y., 262,279 Solliday, N., 269,279
AUTEOR INDEX
2!3!
Solodovnikov, S., 128, 166, 182 Solomons, C., 6, 23 Sommer, J., 18, 20, 24 Sommer, L. H., 126 Solov, N. P., 278 Spall, W. D., 163,182 Spaar, R., 236,237,242,275 Spinner, E., 268,269,274 Spitzer, R., 27(64), 123 Spratley, R. D., 28(6Oc), 29(6Oc), 30(60c), I23 Springer, A., 142, I75 Stamhuis, E. J., 188, 189. 191, 261, 279 Stang, P. J., 236, 237, 239, 241, 247, 248, 275, 277,278,279 Steeper, J. R., 166,176 Stefani, A. P., 133, 142, 182,183 Steinacker, K., 73(104), 125 Stephens, A., 166,179 Stepukhovich, A. D., 132,182 Stetter, H., 73(104), 125 Stewart, R., 144,182 Stiegerwald, C., 136,177 Stieglitz, J., 272, 279 Stivala, S. S., 134, 187 Storr, R. C., 213,279 Streitwieser, A., 263, 264, 279 Strohbusch, F., 161,179 Su, T. M., 263,279 Su, Y. S., 163,182 Subbe Rao, 5. C., 142, 143, I79 h c h a , L., 148, 182 Suchhnek, M., 148,182 Sujishi, S., 131,178 Sukhanova, 0. P., 182 Siimegi, L., 166, 179 Sumida, V.. 179 Summerville, R. H., 236, 239, 277, 279 Sustmann, R., 264,266,279 Sutcliffe, L. H., 141,180 Sutton, L. E., 28(64), 68(64), 124 Swank, D., 27(41), 29(41), 123 Swank, D. D., 32(68), 124 Smbb, Z. G., 136,182 Sz&cs, S., 130, 177 Szwarc, M., 131, 133, 142, 176, 178, 182, 183
T Taft, R. W., 166, 170, I82 Tao, E. V.P., 192,200,278,277
Tasaka, K., 109(129), 118(129), 126 Ta-Shma, R., 271,278 Taylor, D. R., 161, 163, 217, 229, 279 Taylor, M. J., 68(100), 125 Tedder, J. M., 128, 166, 182, 267, 279 Telefus, C. D., 27(11b,c,e), 78(11b,c,e), 121 ten Have, P., 136, I75 Thaler, W. A., 216, 280 Thiagarajan, V., 144,182 Thomas, A., 131,175 Thomas, J. R., 148, 164,178,182 Thomas, R. K., 162,183 Thompson, G., 192,200,278,278 Thompson, H., 162,182 Thompson, R. C., 9, 11, 12, 13, 16, 23, 24 Thorn, R. J., 172,182 Thornley, J. D., 7,23 Thyme. J. C. J., 133, 166, 177, 180, 181 Toboleky, A. V., 131, 182 Tokura, N., 166,179 Tolgyesi, W. S., 273, 277 Tolman, C. A,, 148,164,183 Tolman, R. C., 172,182 Tomlinson, C., 160,177 Tonellato, U., 129, 130,181,234,236,237, 246,249,260,274,277 Topart, J., 126 Toru, T., 201, 279 Trahanovsky, W. S., 144,180 Traynard, J. C., 269,280 Treffera, H. P., 4,24,273,280 Treichel, P. M., 63 (96), 125 Trippett, S., 27(32), 27(60a,b), 29(60a,b), 122,123 Trivedi, B. C., 27(61), 29(61), 123 Trosman, E. A., 132,175 Troughton, G. E., 128, I78 Truce, W. E., 214,280 Tsepalov, V. F., 168, 182 Tsolis, E. A., 27 (66), 27(68, 69), 28 (66, 68, 69), 73(66, 68, 69), 76(69), 77(66, 68, 69), 82(68), 84(68), 85(66), 86(59), 89(113), 96(66, 68, 69), 117(66), 118(66), 123, 125 Tsuno, Y., 188,280 Tucker, E. E., 168, 164, I76 Tudos, F., 128, 129, 130, 134, 136, 140, 142, 143, 146, 146, 147, 148, 149, 160, 161, 162, 164, 166, 168. 169, 160, 166, 167, 168, 169, 177, 179, 180, 181, 182
AUTHOR I N D E X
295
Ward, B., 131,175 Ward, B. F.Jr., 146,177 Ward, P.,156,180 Udovich, C. A., 126 Ugi, I., 27,28 (66-69),36,38,40,42,44(66), Warren, J. D.,126 60, 68, 69, 63, 66(8l), 73(66-69), 76 Washburne, S. S.,220,274 (69), 80(66-69), 82, 84(68), 86(66), Wasylishen, R.,168,183 86(69), 96(66-69), 117, 118(66),123, Waters, W. A.,272, 280 Waugh, J. S., 126 271, 280 Weinberg, A. E.,233, 277 Ulitskii, V. A., 132, 182 Weiner, P.H., 161,180 Umanskii, V. M., 132, 182 Weiner, S. A., 132, 133,183 Uneyama, K.,138,182 Weiss, J., 136, 178 Urner, Z.,148,182 Weissman, S. I., 129, 168, 279 Usher, D.,27(43),29(43),123 Werner, R.L., 163,181 Westheimer, F. H., 27 (36,39, 42, 46, 46, 47a,b, c,d), 28 (36,46), 29 (36,39, 42, V 46, 46), 36(77), 96(118a,b), 96(77), 104(77,llEa,b), 122,142,183 Van Binst, G., 162,177 Whalley, E., 174 van Boom, J. H., 217,280 Wheatley, P.J., 36(70),124 Vanderborgh, N. E.,163,182 Whetse1,K. B.,161,162,163,279,183 Van der Voorn, P. C., 69(92), 68 (92), Whipple, E.B., 266, 276 125 White, A.E.,273,277 Vander Werf, C. A., 27 (29a,b,c),122 White, A.M., 1, 19,21, 24 van Helden, R.,130, 179,182 Whitesides, G.M.,43(80), 44(80), SZ(SO), Van Lautem, N.,60(87), 124 119(80),124 Van Wazer, J. R., 68(102),68(103),125 Whiting, M. C., 234, 274 Vanwuytswinkel, H., 126 Whittemore, I. M., 133, 142, 183 Varshnyi, G., 162,182 Whittle, E.,132, 177 Venkatasubramanian, N.,144, 182 Wiberg, K.B., 142,183 Verkade, J. G.,73 (106a,b,c), 77(106a,b,c), Widom, J. M., 162, 163,183 125 Wilfinger, H. J., 27(4d), 114(4d), 118(4d), Vernon, C. A., 268,280 121 Verny, M.,192,218,280 Wilkins, C. L., 263,279 Vessiere, R.,192,218,280 Wilkinson, F.,129,176 Vetchinkina, V. N.,138,183 Williams, A., 136, 177 Viehe, H. G.,196,280 Williams, G.H., 129, 130, 179 Vincent, D.N.,138,181 Williams, H. J., 129, 179 Vinogradov, S., 164,181 Williams, J. E.,264,279 Voevodskii, U. V., 128, 166,182 Williams, L.D.,27(46),29(46),123 Vogel, G.C.,161,183 Williams, R.L.,163,163,175 Volpin, M. E.,266,280 Willis, H.A., 161, 177 Vott, V.,203, 204,206, 275 Wilmshurst, J. K.,68(98),125 Wilson, J. W., 204, 206, 269, 279, 280 Wilson, M. H., 208,211,280 W Wilson, R. Q.,131, 162.174 Winkler, H., 27(30), 122 Wagner, P. J., 131,166,183 Winstein, S., 18, 23, 199, 207, 242, 249, Wall, R.E.,27(47b),123 262,276,280 Walling, C.,130, 132, 142, 166,183 Winter, E. R. S., 273, 276 Walsh, A. D.,68(96),125 Winterfeldt, E.,201, 216, 280 Waltem, E.,132,176 Wisotsky, M. J., 21,23,273, 275 Walton, D.R. M., 188, 189, 192,274 Wittig, G.,27(1), 27(2),27(3), 29(2), 114 Walton, J. C.,166,182 (2,3), 118(2,3), 121 Wanless, a. G., 194, 196, 220, 221, Wojtkowiak, B., 146, 179 275
U
AUTHOR INDEX
294
Wolf, R., 111(122a), 111(122b), 112(122e, b), 116(122a,b),126 Wolfgang, R.,I74 Wolfgang, R. L., 172, 180 Wolfstirn, K.B.,130,183 Woodward. L.A., 68(100),125 Woodward, R.B.,196,220,280
Yen, Y.Q., 201,216,277 Yonezawa, T.,254,280 Young, D.P.,272,276 Young, J. C., 166,179 Yukewa, Y., 188,280
Z G. E.,168,164, 166, 168,174, I83 Zaitsev, B.A., 130, 183 Yakovleve, T. V., 278,278 Zanger, M.,27(29b), 122 Yannios, C.N.,214,216,261,276 Zeifman, Yu. V., 95(108b), 101(108b),125 Yasumoto, N.,142,180 Z h 6 m l 3 M , H., 160,179.183 Y8teS,K., 208,210,264,266,268,261,262, Zohger, H., 143,178 Zuidema, G.,163,177 276,278 Yetee, W. R., 130,138,183 Zunker, D., 231,276
Y
Zaikov,
CUMULATIVE INDEX OF AUTHORS Anbar, M., 7,116 Bell, R. P., 4, 1 Bennett, J. E., 8, 1 Bentley, T. W., 8, 161 Bethell, D., 7, 163 Brand, J. C. D., 1,366 Brown, H. C., 1, 36 Cecace, F., 8, 79 Collins,C. J., 2, 1 Crampton, M. R., 7,211 Fender, E. J., 8, 271 Fender, J. H., 8,271 Ferguson, G., 1,203 Fields, E. K., 6, 1 Frey, H. M., 4, 147 Gilbert, B. C., 6 , 6 3 Gillespie, R. J., 9, 1 Gold, V.,7,269 Greenwood, H. H., 4, 73 Johnson, 5. L., 6,237 Johnstone, R. A. W., 8,161 Kohnstam, G., 6,121 Kreevoy, M. M., 6 , 6 3 Long, F. A., 1 , l Meccoll, A., 8,91 McWeeny, R., 4,73 Mile, B., 8, 1 Miller, 5. I., 6, 186 Modena, G., 9,186 More O’Ferrall, R. A., 6,331 Norman, R. 0. C., 6 , 6 3 Olah, G. A., 4,306 Parker, A. J., 6, 173 Peel, T. E., 9, 1 Perkampus, H.-H., 4, 196 Pittmann, C. U., Jr., 4, 306 Ramirez, F., 9 , 2 6 Rappoport, Z., 7 , l Reeves, L. W., 8, 187 Robertson, J. M., 1,203 Samuel, D., 8, 123 Sohaleger, L. L., 1, 1 Soheraga, H. A., 6,103 Shatenshtein, A. I., 1, 166 Silver, B. L., 8, 123 Simonyi, M., 9,127 Stock, L. M., 1,36 29K
296 Symom, M. C. R., 1,284 Thorn-, A., 8 , 1 Tonellato, U., 9, 186 Tudbs, F.. 9, 127 Turner, D. W., 4, 31 Ugi, I., 9, 26 W a d , B., 8 , 1 Whalley, E., 2,93 William, J. M., Jr., 6, 63 Williamson, D. cf., 1,366 Wolf, A. P., 2,201 Zollinger, H., 2, 163 Zuman, P., 6, 1
CUMULATIVE I N D E X
CUMULATIVE INDEX OF TITLES Abstraction, hydrogen atom, from 0-H bonds, 9, 127 Acid solutions, strong, spectroscopic observation of alkylcarbonium ions in, 4, 306 Acids, reactions of aliphatic diazo compounds with, 5, 331 Activation, entropies of, and mechanisms of reactions in solution, 1, 1 Activation, heat capacities of, and their uses in mechanistic studies, 5, 121 Activation, volumes of, UBB for determining reaction mechanisms, 2, 93 Aliphatic diazo compounds, reactions with acids, 5, 331 Alkylcarbonium ions, spectroscopic observation in strong acid solutions, 4, 306 Ammonia, liquid, isotope exchange reactions of organic compounds in, I, 166 Aromatic substitution, a quantitative treatment of directive effects in, I, 36 Aromatic substitution reactions, hydrogen isotope effects in, 2, 163 Aromatic systems, planar and non-planar, 1,203 Arynes, mechanisms of formation and reactions a t high temperatures, 6 , l A-S& reactions, developments in the study of, 6 , 6 3
Base catalysis, general, of ester hydrolysis and related reactions, 5, 237 Basicity of unsaturated compounds, 4, 196 Bimolecular substitution reactions in protic and dipolar aprotic solvents, 5, 173 Carbene chemistry, structure and mechanism in, 7,163 Carbon atoms, energetic, reactions with organic Compounds, 3,201 Carbonium ions (alkyl), spectroscopic observation in strong acid solutions, 4, 306 Carbonium ions, gaseous, from the decay of tritiated molecules, 8, 79 Carbonyl compounds, reversible hydration of, 4, 1 Catalysis, general base and nucleophilic, of ester hydrolysis and related reactions, 5,237 Catalysis, micellar, in organic reactions: kinetic and mechanistic implications, 8,271 Cations, vinyl, 9, 186 Conformationsof polypeptides, callcultations of,6, 103 Conjugated molecules, reactivity indices in, 4, 73 Diazo compounds, aliphatic reactions with acids, 5, 331 Dipolar aprotic and protic solvents, rates of bimolecular substitution reactions in, 5, 173 Directive effects in aromatic substitution, a quantitative treatment of, 1 , 3 6 Electron spin resonance, identi6cation of organic free radicals by, I, 284 Electron spin resonance studies of short-lived organic radicals, 5 , 6 3 Electronically excited molecules, structure of, 1, 366 Energetic tritium and carbon atoms, reactions of, with organic compounds, 2,201 Entropies of activation and mechanisms of reactions in solution, 1, 1 Equilibrium constants, N.M.R. measurements of, as a function of temperature, 3, 187 Ester hydrolysis, general base and nucleophilic catalysis, 5,237 Exchange reactions, hydrogen isotope, of organic compounds in liquid ammonia, 1. 166 Exchange reactions, oxygen isotope, of organic compounds, 3, 123 Excited molecules, structure of electronically, 1, 366
Free radicals, identification by electron spin resonance, 1, 284
Free radicals and their reactions a t low temperature using a rotating cryostat, study of, 8. 1 297
298
UUMOLA!FIVE INDEX
Gaseous carbonium ions from the decay of tritiated molecules, 8,79 Gas-phaseheterolysis, 3, 91 Gas-phase pyrolysis of small-ring hydrocarbons, 4,147 General base and nucleophilic catalysis of ester hydrolysis and related reactions, 5, 237 HaO-DsO Mixtures, protolytic processes in, 7, 259 Heat capacities of activation and their uses in mechanistic studies, 5, 121 Heterolysis, gas-phase, 3,91 Hydrated electrons reactions of, with organic compounds, 7, 115 Hydration, reversible, of carbonyl compounds, 4,1 Hydrocarbons, small-ring, gas-phase pyrolysis of, 4,147 Hydrogen atom abstraction from 0-H bonds, 9,127 Hydrogen isotope effects in aromatic substitution reactions, 2, 163 Hydrogen isotope exchange reactions of organic compounds in liquid ammonia, 1, 166 Hydrolysis, ester, and related reactions, general base and nucleophilic catalysis of, 5,237 Ionization potentials, 4,31 Isomerisation, permutational, of pentavalent phosphorus compounds, 9, 26 Isotope effects, hydrogen, in aromatic substitution reactions, 2, 163 Isotope exchange reactions, hydrogen, of organic compounds in liquid ammonia, 1. 150 Isotope exchange reactions, oxygen, of organic compounds, 3,123 Isotope and organic reaction mechanisms, 2, 1 Kinetics, reaction, polarography and, 5, 1 Mass spectrometry, mechanism and structure in: a comparison with other chemical processes, 8,162 Mechanism and structure in carbene chemistry, 7,163 Mechanism and structure in mass spectrometry: A comparison with other chemical processes, 8,162 Mechanisms, organic reaction, isotopes and, 2, 1 Mechanisms, reaction, w e of volumes of activation for determining, 2, 93 Mechanisms of formation and reactions of arynes at high temperatures, 6,1 Mechanisms of reactions in solution, entropies of activation and, 1, 1 Mechanistic studies, heat capacities of activation and their uses in, 5, 121 Meisenheimer complexes, 7,211 Micellar catalysis in organic reactions: kinetic and mechanistic implications, 8,271 N.M.R. measurements of reaction velocities and equilibrium constants as a function of temperature, 3,187 Non-planar and planar aromatic systems, 1,203 Nuclear magnetic resonance, 8ee N.M.R. Nucleophilic catalysis of ester hydrolysis and related reactions, 4,237 Nucleophilic vinylic substitution, 7, 1 0-H bonds, hydrogen atom abstraction from, 9,127 Oxygen isotope exchange reactions of organic compounds, 3, 123 Permutational isomerization of pentavalent phosphom compounds, 9,26 Phosphorous compounds, pentavalent, turnstile rearrangement and pseudorotation in permutational isomerization, 9,26 Planar and non-planar aromatic systems, 1,203 Polarizability, molecular refractivity and, 3, 1 Polarography and reaction kinetics, 5, 1 Polypeptides, calculations of conformations of, 6,103 Protic and dipolar rtprotic solvents, rates of bimolecular substitution reactions in, 5, 173
CIJNULATIVE INDEX
299
Protolytic processes in HaO-DaO mixtures, 7, 269 Pseudorotation in isomerization of pentavalent phosphorous compounds, 9, 26 Pyrolysis, gas-phaee, of small-ring hydrocarbons, 4,147 Radicals, organic free, identification by electron spin resonance, 1, 284 Radicals, short-lived organic, electron spin resonance studies of, 5, 63 Reaction kinetics, polarography and, 5, 1 Reaction mechanisms, use of volumes of activation for determining, 2 , 9 3 Reaction mechanisms in solution, entropies of activation and, 1, 1 Reaction velocities imd equilibrium constants, N.M.R. measurements of, as a funotion of temperature, 3, 187 Reactions of hydrated electrons with organic compounds, 7, 116 Reactivity indices in conjugated molecules, 4, 73 Refractivity, molecular, and polarizability, 3, 1 Resonance, electron-spin, identification of organic free radicals by, 1,284 Resonance, electron-spin, studies of short-lived organic radicals, 5, 63 Short-lived organic radicals, electron-spin resonance studies of, 5,63 Small-ringhydrocarbons, gas-phase pyrolysis of, 4, 147 Solution, reactions in, entropies of activation and mechanisms, 1, 1 Solvents, protic and dipolar aprotic, rates of bimolecular substitution reactions in, 5, 173 Spectroscopic observation of alkylcarbonium ions in strong acid solutions, 4, 306 Stereoselection in elementary steps of organic reactions, 6, 186 Structure and mechanism in carbene chemistry, 7,163 Structure of electronically excited molecules, 1, 366 Study of free radicals and their reactions at low temperature using a rotating cryostat, 8, 1 Substitution, aromatic, a quantitative treatment of directive effects in, 1, 36 Substitution reactions, bimolecular, in protic and dipolar aprotic solvents, 5, 173 Substitution reactions, aromatic, hydrogen isotope effects in, 2, 163 Superacid systems, 9 , l Temperature, N.M.R. measurements of reaction velocities and equilibrium constants 88 a function of, 3, 187 Tritiated molecules, gaseous carbonium ions from the decay of, 8, 79 Tritium atoms, energetic, reactions with organic compounds, 2, 201 Turnstile rearrangement in isomerization of pentavalent phosphorus compounds, 9, 26 Unsaturated compounds, basicity of, 4, 196 Vinyl cations, 9, 186 Volumes of activation, use of, for determining reaction mechanisms, 2, 93
This Page Intentionally Left Blank