New Concepts in Polymer Science Low-Temperature Behaviour of Elastomers
New Concepts in Polymer Science Previous titl...
263 downloads
1582 Views
3MB 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
New Concepts in Polymer Science Low-Temperature Behaviour of Elastomers
New Concepts in Polymer Science Previous titles in this book series: Active Sites of Polymerization. Multiplicity: Stereospecific and Kinetic Heterogeneity Y.B. Monakov, N.N. Sigaeva and N. Urazbaev Stationary and Non-Stationary Kinetics of the Photoinitiated Polymerization Yu.G. Medvedevskikh, A.R. Kytsya, L.I. Bazylyak, A.A. Turovsky and G.E. Zaikov Thermal Stability of Engineering Heterochain Thermoresistant Polymers E.V. Kalugina, K.Z. Gumargalieva and G.E. Zaikov The Concept of Micellar-Sponge Nanophases in Chemical Physics of Polymers Yu.A. Mikheev, G.E. Zaikov Structure of the Polymer Amorphous State G.V. Kozlov and G.E. Zaikov The Statistical Nature of Strength and Lifetime in Polymer Films and Fibers B. Tsoi, E.M. Kartashov and V.V. Shevelev Cyclolinear Organosilicon Copolymers: Synthesis, Properties, Application O.V. Mukbaniani and G.E. Zaikov Modern Polymer Flame Retardancy S.M. Lomakin and G.E. Zaikov Polymeric Biomaterials. Part I. Polymer Implants M.I. Shtilman Biodegradation and Durability of Materials under the Effect of Microorganisms S.A. Semenov, K.Z Gumargalieva and G.E. Zaikov Fire Resistant and Thermally Stable Materials Derived from Chlorinated Polyethylene A.A. Donskoi, M.A. Shashkina and G.E. Zaikov Polymers and Polymeric Materials for Fiber and Gradient Optics N. Lekishvili, L. Nadareishvili, G. Zaikov and L. Khananashvili Polymers Derived from Isobutylene. Synthesis, Properties, Application Yu.A. Sangalov, K.S. Minsker and G.E. Zaikov
Ecological Aspects of Polymer Flame Retardancy S.M. Lomakin and G.E. Zaikov Molecular Dynamics of Additives in Polymers A.L. Kovarski Structure and Properties of Conducting Polymer Composites V.E. Gul' Interaction of Polymers with Bioactive and Corrosive Media A.L. Iordanskii, T.E. Rudakova and G.E. Zaikov Immobilization on Polymers M.I. Shtilman Radiation Chemistry of Polymers V.S. Ivanov Polymeric Composites R.B. Seymour Reactive Oligomers S.G. Entelis, V.V. Evreinov and A.I. Kuzaev Diffusion of Electrolytes in Polymers G.E. Zaikov, A.L. Iordanskii and V.S. Markin Chemical Physics of Polymer Degradation and Stabilization N.M Emanuel and A.L. Buchachenko
New Concepts in Polymer Science
Low-Temperature Behaviour of Elastomers
M.F. Bukhina and S.K. Kurlyand
LEIDEN • BOSTON - 2007
A C.I.P. record for this book is available from the Library of Congress ISBN: 978 90 04 15720 0 © Copyright 2007 by Koninklijke Brill NV, Leiden, The Netherlands. Koninklijke Brill NV incorporates the imprints Brill Academic Publishers, Martinus Nijhoff Publishers and VSP All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Brill provided that the appropriate fees are paid directly to the Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, MA 01923, USA. Fees are subject to change.
Printed and bound in The Netherlands.
Contents
Foreword
1
Introduction
3
1 Glass transition of elastomers 1.1 Methods of glass-transition temperature determination 1.1.1 Structural and other physical methods 1.1.2 Mechanical methods 1.1.3 Comparison of the data obtained by different methods 1.2 Modern views of glass transition 1.2.1 Kinetic views of glass transition 1.2.2 Thermodynamic views of glass transition 1.2.3 Morphological views of glass transition 1.3 Effect of the molecular parameters of rubbers and the composition of rubber compounds on glass-transition temperature 1.3.1 Effect of molecular mass 1.3.2 Effect of the structure of macromolecules 1.3.3 Effect of intermolecular interaction 1.3.4 Glass transition of copolymers 1.3.5 Effect of the composition of rubbers 1.4 Effect of a mechanical action on glass-transition temperature
11 11 11 15 17 18 18 25 26 28 28 30 32 33 34 41
2 Mechanical properties of elastomers near the glass transition temperature 2.1 Mechanical properties in the region of transition from the rubberlike to the glassy state 2.1.1 Static mechanical properties 2.1.2 Dynamic mechanical properties 2.1.3 Prediction of the mechanical properties of elastomers in the transition region from the glassy to the rubberlike state 2.2 Mechanical properties at temperatures below glass-transition temperature
43
52 55
3 Crystallization of elastomers at low temperatures 3.1 Methods of crystallization studies 3.1.1 Structural and other physical methods 3.1.2 Mechanical methods 3.2 Modern views of crystallization 3.2.1 Morphology of crystals in elastomers 3.2.2 Kinetics of isothermal crystallization
59 60 60 63 63 64 65
43 43 48
vi
CONTENTS
3.3 3.4 3.5 3.6
3.2.3 Temperature range of crystallization 3.2.4 Kinetics of nonisothermal crystallization 3.2.5 Melting Effect of mechanical action and other factors on crystallization Effect of the molecular parameters of rubbers and the composition of rubber compounds on crystallization Microcrystallization Prediction of the crystallization rate
68 73 74 79 82 91 94
4 Stress-induced crystallization of elastomers 4.1 Kinetics of stress-induced crystallization 4.2 Morphology in stress-induced crystallization 4.3 Melting of oriented elastomers 4.4 Effect of the molecular parameters of rubbers and the composition of rubber compounds on stress-induced crystallization 4.4.1 Effect of chain chemical structure, molecular mass and microstructure 4.4.2 Effect of vulcanization and filling
97 97 100 104
5 Strength properties of elastomers at low temperatures 5.1 Strength of amorphous elastomers at low temperatures 5.2 Strength of crystallizable elastomers 5.2.1 Strength determined by stress-induced crystallization 5.2.2 Effect of precrystallization on the strength at low temperatures
115 115 118 118 122
6 Contribution of crystallization and glass transition to low-temperature resistance of elastomers 6.1 Mutual interaction of crystallization and glass transition 6.1.1 Effect of crystallization on glass transition 6.1.2 Crystallization from the glassy state 6.2 The ratio of the contributions by crystallization and glass transition to the low-temperature resistance of elastomers 6.3 Low-temperature resistance of rubber articles 6.3.1 Ways to improve the low-temperature resistance of rubber articles 6.3.2 Assessment of low-temperature resistance of rubber articles from the low-temperature behaviour data for elastomers 6.4 Effect of repeated exposure to low temperatures on the performance of rubber articles 6.4.1 Reversibility of changes due to glass transition and crystallization 6.4.2 Effect of temperature cycling on the performance of rubber articles 7 Low-temperature resistance specifics of particular elastomers 7.1 Styrene-butadiene rubbers 7.2 Butadiene acrylonitrile rubbers 7.3 Isoprene rubbers 7.4 Butadiene rubbers 7.5 Chloroprene rubbers 7.6 Butyl rubber 7.7 Silicon rubbers
108 108 110
125 125 125 128 128 129 129 131 133 133 133 137 137 139 140 142 144 146 147
CONTENTS
7.8 7.9 7.10 7.11 7.12 7.13 7.14 7.15
Urethane rubbers Thiocol rubbers Fluoroelastomers Acrylate rubbers Ethylene-propylene rubbers Epoxide rubbers Polypentenomers Alternating rubbers
vii
149 150 151 151 152 153 154 155
References
157
Subject index
183
viii
CONTENTS
Foreword Intensive development of many engineering fields, aerospace including, raised the problem of a significant broadening of the operational temperature range of elastomeric materials. In a number of cases, one of the essential factors determining the performance of elastomers and elastomeric goods is low-temperature resistance. This factor preserved its significance for space-system and aviation components, in developing new aircraft and helicopters. In the ostensibly advancing era of global warming, unusual local temperature drops in the normally warm areas make low-temperature resistance of elastomers even more an issue. The currently existing ways of increasing low-temperature resistance by developing novel elastomeric materials and composites with new efficient ingredients and their combinations make it possible to significantly broaden the temperature ranges of elastomer applications. There are two most studied physical processes determining the low-temperature resistance of elastomers: glass transition and crystallization. Extensive information has been accumulated on the relation of these processes to the structural and molecular parameters of rubbers, their compositions and deformation conditions. These processes have been found to contribute to the low-temperature resistance of elastomeric materials and articles. Ever-widening applications and development of elastomers of regular structure make one to consider the low-temperature resistance of elastomeric materials in a new light. Attention of investigators is currently attracted by microcrystallization, a specific manifestation of the crystallization of elastomers. There are grounds to believe that this process determines the properties of a number of relatively new elastomers, which are copolymers including rapidly crystallizing sequences. Due to the inclement climate of Russia, low-temperature studies of elastomers have been vitally important for the successful development of industries and transport in Extreme North regions and Siberia. That is why these studies have been the most intensive namely in Russia. The book written by specialists of two Russian research institutes (Research Institute of Elastomeric Materials and Articles (Moscow) and Lebedev Synthetic Rubber Research Institute (St.-Petersburg)), where such studies have been carried out for many years, summarizes the results of these works and other numerous investigations by Russian and Western scientists. The book was published in Russian in 1989. That was the year when a major cycle of works on the low-temperature behaviour of elastomers had been completed. Those works, however, are little known to specialists from other countries. The book also includes the results obtained by Russian and Western specialists in the recent years. Major attention is given to Russian publications, not always known to investigators and engineers in other countries. The results published in special Russian editions [220, 377, 439] and reported at special Russian and international conferences [23, 99] are also considered.
2
FOREWORD
The book also contains reference material, which could help specialists in choosing the compositions of rubbers and the designs of rubber articles intended for low-temperature applications. The authors also strived to analyze the not-yet solved scientific and technical problems, essential for increasing the quality of these items. Considerable attention is given in the book to the effect the compositions of rubbers and the conditions of their processing have on low-temperature resistance. Herewith, the issues related to chemical changes that occur in rubbers are not discussed. The interested reader can find respective data in many other books. These are, for instance [178, 180, 346, 372, 373, 390, 391]. For convenience, the authors thought it expedient to give a list of references and recommended literature in the end of the book in an alphabetical order (by the first author’s names). Chapters 1–3, 6 and 7 were written by the authors with participation of N.M. Zorina and N.L. Severina (Moscow). In preparation of the text of the present edition, the authors used the data by N.N. Petrova (Yakutsk). The authors thank G.D. Garmashova, G.V. Novikova and G.P. Petrova (St.-Petersburg) for their assistance in preparing the material and for valuable comments; and B.M. Polyakova, N.N. Kutsankina (Moscow), G.P. Maksimchuk and G.S. Solodovnikova (St.-Petersburg) for their help in preparing the manuscript of the Russian edition. The authors use the occasion to sincerely thank Prof. V.A. Shershnev, who took upon himself an uneasy task of reviewing the manuscript of the Russian edition. The great majority of his numerious comments were taken into account. The book will be of use for researchers and engineers working in the field of elastomers, developing and using elastomeric materials and articles, and also for university students, postgraduates and teachers studying the physics of elastomers and majoring in this field.
Introduction The purpose of Introduction is to remind the reader of basic information on the mechanical properties of elastomers, required to understand the physical processes that occur in these materials at low temperatures. The major differences of elastomers from all other materials are their capability of large reversible deformation, low rate of deformation, low rigidity, large difference between shear modulus and bulk modulus, and significant losses of mechanical energy in dynamic deformation. These unique properties are realized in the temperature range close to room temperature. The low-temperature resistance of elastomers is their ability to preserve this set of unique mechanical properties at a temperature decrease. When considering deformation of polymers, we distinguish between three of its components: • Purely elastic deformation similar to elastic deformation of common solids, with elasticity modulus of |103 MPa. It is reversible and is related to the change of intermolecular and interatomic distances, as well as angles formed by valent bonds, which connect atoms in macromolecules; the development of elastic deformation is accompanied with a change of internal energy of the system. • Rubberlike deformation characteristic of only polymers in the rubberlike state; it is also reversible but is related to the change of macromolecular conformation, i.e., to the change of system’s entropy. The entire set of unique mechanical properties of natural and synthetic rubbers is determined by this type of deformation. • Plastic deformation related to the irreversible motion of macromolecules relative to one another. This type of deformation plays an especially important role in raw rubbers and raw rubber compounds. In deformation of vulcanized rubbers, i.e., elastomers that possess a three-dimensional network of chemical bonds, this third component of deformation develops only when chemical bonds are broken, i.e., as the result of chemical ageing. Depending on the deformation conditions of an elastomer, the ratio of its deformation components changes, which is determined by the difference in the rates of each type of deformation: the rate of elastic deformation is very high, and it develops in a negligibly small time. The rate of the rubberlike deformation component is hundreds of times lower even at high temperatures, and it sharply drops down at a temperature decrease. Thus, by changing the temperature of an elastomer, the predominant development of each of these types of deformation can be realized.
4
INTRODUCTION
ε
I
III
II
IV
T
Figure 1 A temperature dependence of deformation H under a given load. I, glassy-state region; II, transition region; III, region of the complete development of rubberlike deformation (rubberlike-state plateau); IV, viscous-flow region.
Figure 1 is a schematic representation of a temperature dependence for an elastomer. In region I, where the elastomer is in the glassy state, mainly elastic deformation develops. Transition region II is characterized by the coexistence of elastic and rubberlike deformations. On the plateau of the rubberlike state III, rubberlike deformation determines the unique mechanical properties of elastomers. Mainly plastic deformation develops in the viscous-flow region IV; it determines the processing. properties of elastomers. The low temperature performance of an elastomer is mainly determined by its behaviour in region II – in transition from the rubberlike to the glassy state (or the other way round, from the glassy to the rubberlike state). The difference of the temperature dependence and the absolute values of the rates of three deformation types peculiar to polymers leads to the change of their properties in time, i.e., to the occurrence of a set of their relaxation properties. The main type of physical relaxation processes observed in polymers is related to the transition from “instantaneous” elastic to “slow” rubberlike deformation. These relaxation processes are called the major physical relaxation processes or D-processes of relaxation. Besides physical relaxation processes, polymers are characterized by chemical processes related to the change of their chemical structure: rupture or formation of new chemical bonds, i.e., to chemical ageing. In the deformation of rubbers at a very low rate, obviously lower than the rate of rubberlike deformation, the latter develops completely. However, for raw rubbers and rubber compounds the development of rubberlike deformation is obligatorily accompanied by plastic deformation, so that no equilibrium deformation can be achieved at a given load. For a three-dimensional-network elastomer, which vulcanized rubber is, a certain equilibrium value of deformation corresponds to each value of stress, and, vice versa, a very certain value of equilibrium stress different from zero corresponds to each value of deformation. A typical dependence between a nominal equilibrium stress f and a deformation ratio O at an uniaxial tension is presented in Fig. 2. Here
f
P / S0 ,
(1)
where P is the load on the specimen, measured after an equilibrium has been established; S0 is the cross-section area of the specimen before the deformation. Note that the length of the specimen in these measurements should be considerably larger than the dimensions of its cross section. The dependences between the true equilibrium stressV and the elongation ratio Oare represented in Fig. 2 by curves 2. Here, the true cross-section area S was calculated from the assumption of incompressibility:
INTRODUCTION
5
f ; σ, MPa 2
10
2'
8
2'' 1
6
4
1'
2
0
1
3
5
7
9
λ
Figure 2 Dependence of the equilibrium stress, normal f (1, 1c ) and true V (2, 2c, 2cc ), on the deformation ratio O at a uniaxial tension of unfilled vulcanized natural rubber: 1, 2, experiment; 1c, 2c, 2cc, calculation: 1c, by (6); 2c, by (7); 2cc, by (3).
Sl
l0 S0 ,
S
S0 / O .
(2)
where l0 is the length of the undeformed specimen; l, the length of the deformed specimen. For the values ofO close to unity the dependenceV – O (or V – H, where the deformation H = O – 1) can be considered to be linear [55]:
V
EH .
(3)
Here E is the Young’s modulus. For a deformation not accompanied by a volume change, E = 3G, where G is the shear modulus, and the Poisson ratio P = 0.5. Complex (three-axial) deformation is described by the dependence of strain energy function W, i.e., the specific work of deformation, on the extension ratio O1, O2, O3 along the three principal axes of deformation. Considering an individual molecule as a statistical set of segments, and a specimen as a whole as a network formed by the statistical set of molecules, and accepting that rubberlike deformation of elastomers is of entropic nature, an expression for the specific work of deformation or for strain energy function was obtained in the form:
W
1 NkT O12 O2 2 O32 . 2
(4)
6
INTRODUCTION
Here N is the number of molecular segments forming the network, k is Boltzmann’s constant and T is absolute temperature, and NkT
(4c
G,
where G is the shear modulus. The dependences between the principal stresses and deformations are determined by differentiation wW / wOi
(5)
fi , i 1, 2,3.
For uniaxial tensile/compressive deformation, from the relation (4) for normal f and true V stresses it follows:
f
G O 1/ O ,
V
G O 2 1/ O .
(6)
(7)
Figure 2 presents the dependences of f – O from the relation (6), curve 1c, and V – O from the relation (7), curve 2c. The difference of the values of V obtained by (3) and (7) does not exceed 5% in the region* of –0.5 < H < 2.5 at G < 5 MPa; as G is increased, the deformation limits within which (3) and (7) virtually coincide become narrower. The experimental results obtained at an equilibrium uniaxial tension of unfilled vulcanizates based on most natural rubbers is described well by eq. (3) [44, 55]. The applicability limits of (3) and (7) are the broader the smaller G (or E) is. The expression for strain energy function can be also obtained without using representations of the nature of a deformation. Thus, based on the assumption of the linear dependence between stress and deformation in simple shear for an isotropic incompressible material, an expression, which is usually called the Mooney or Mooney–Rivlin strain energy function, was obtained [715]:
W
C1 O12 O22 O32 3 C2 1/ O12 1/ O2 2 1/ O32 3 .
(8)
At C2 = 0, it changes to (4). The relation between stress and deformation at uniaxial deformation, which follows from (8):
f
2 O 1/ O 2 C1 C2 / O ,
* The negative values of Hcorrespond to compression.
(9)
INTRODUCTION
7
ΦM 0.4 5 4
0.3
3
0.2
2
1 0.1
0
0.2
0.4
0.6
0.8
1/λ
Figure 3 Dependence of )M = f/(O – 1)O2) on 1/O for natural rubbers vulcanized with various amounts of dicumyl peroxide (numbers at the curves, %).
describes the experimental data in the region of large deformation better than (6). At C2 = 0 and 2C1 = NkT, eq. (9) changes to (6). In the case of the validity of relation (9), experimental data form straight lines in coordinates )M = f/(O– 1)O2) and 1/O(Fig. 3). The slope of these lines gives the values 2C2; and the segment cut on the ordinate axis, 2C1. As a rule, at large O (small 1/O) one observes deviations from this dependence, which are due to a steeper character of the curve f – O than it follows from eq. (9). For natural and polychroloprene rubbers these deviations are related, first of all, to the development of stress-induced crystallization (see Chapter 4). The fact that the classical theory of rubberlike elasticity fails to describe deformation curves at large tensions implies that its assumptions are not fulfilled for real elastomers. Introduction of additional restrictions for the motion of molecular chains between chemical junctions made it possible to describe experimental data better. However, more new additional assumptions need to be introduced for the data to coincide with the experiment. Considering the constant C2 in eqs. (8) and (9) as a measure of deviation of experimental data from classical theory, we can assume that its significance is the greater, the larger the number of physical junctions in the network is and the more intensively they are formed in the process of deformation. There is no common point of view on the nature of the physical junctions. Presumably, the greatest role in the deformation of elastomers is played by the physical junctions, which are pre-phase fluctuations and are structurally close to crystals. They become stable crystals only when their size exceeds the critical size of the nucleus* (see Chapter 3). Therefore, understanding of the mechanism of crystallization, in * Formation of junctions of this nature is also possible in non-crystallizing elastomers, but they fail to become stable under any conditions there.
8
INTRODUCTION
particular, that of stress-induced crystallization, can also be of help in describing the deformation properties of elastomers. When considering the entropic component of deformation, we do not have to take into account the change of volume during the deformation. However, we can not neglect the volume change both in the case of large tensile deformations and at large compressions, as well as in compression of specimens with a large shape factor )= S2o/S1o, for which the area of the loaded surface S2o is considerably greater than that of the free surface S1o, i.e., when the process is close to volume compression. In this case, the contribution of elastic (or energetic) component of the deformation is great; this results also in some peculiarity of lowtemperature resistance (see Chapter 2). The low rate of rubberlike deformation as compared with elastic deformation, and its dependence on temperature, i.e., the relaxation behaviour, are manifested the most vividly in the dynamic loading of elastomers, when both the stress and deformation change in time. A typical case of dynamic loading is a periodic application of a load. The higher the loading frequency Q, the smaller (at a constant amplitude of the stress) the time of the deformation development is and, respectively, the smaller the amplitude of the deformation for one cycle of loading. Due to the low rate of rubberlike deformation, it is delayed in phase from the load, so part of the mechanical energy is not returned completely but passes into heat, i.e., a hysteresis is observed. The simplest way of periodic loading is the deformation at a constant frequency, so that deformation H and stress f of the specimen change by a harmonic law: f
f0 cos Zt ;
H
H 0 cos Zt G ,
(10)
where t is time, Z is the angular frequency (Z = 2SQ ); G is the phase shift between the stress and deformation, which usually (tan G) can serve as a measure of energy losses. The dynamic modulus of tension/compression E = f0/H0. The dynamic moduli of tension/compression E and shear G in harmonic loading are characterized by complex values: E
E c iE cc;
G
G c iG cc.
(11)
where Ec
E cos G ; E cc
E sin G ; tan G
E cc / E c.
(11c
The relations for the shear modulus G are written down in a similar way. The values Ec, Gc and Ecc, Gcc are called, respectively, the storage and loss moduli. The temperature dependence of the deformation amplitude at a given frequency Q has the same shape as the temperature dependence of deformation (see Fig. 1) at static loading. The dependence of the real and imaginary components of the dynamic modulus on frequency is similar to that on temperature: as frequency is increased, Gcand Ecgo up as they do when temperature is decreased, and tan G, Gccand Ecc pass through a maximum. The temperature and frequency dependences of the mechanical properties determine the lowtemperature resistance of elastomers operated under dynamic conditions.
INTRODUCTION
9
The interested reader can find out more about the peculiar features of the properties of elastomers in the books [27, 29, 44, 45, 97, 161, 178, 180, 198, 205, 340, 372, 373, 390, 391, 403, 462, 462a, 464, 550, 600, 624, 715].
10
INTRODUCTION
1
Glass Transition of Elastomers
Glass transition – the transition of a substance from the liquid state into the solid (but not crystalline) state – is characteristic of both polymers and many low-molecular-mass substances. A substance proves glassy when the energy of molecular motion, decreasing in cooling, approaches the energy of molecular interaction in the order of magnitude. Glass transition occurs when the cooling rate of a substance exceeds that of crystallization, i.e., it can be observed for substances with a low crystallization rate or those that do not crystallize at all. The higher the viscosity of a liquid, the more probable it is that it will pass into the glassy state when cooled. Due to the high viscosity of the melt, glass transition is realized, e.g., for low-molecular-mass silicate glasses. Vice versa, rapidly crystallizable substances are very difficult – and sometimes impossible – to obtain in the glassy state. This explains the fact that crystals are much more frequent than glasses among low-molecular-mass substances in the solid state. Similar patterns are observed for polymers. The impossibility of crystallization (due to the non-regularity of the chain for a number of polymers) is also responsible for glass transition. For many polymers, the glass transition range is at room temperature and higher, which determines their use as structural hard materials. These are polystyrene, polymethyl methacrylate, polycarbonate and others. Rapidly crystallizable polymers, such as polyethylene, polypropylene, polyethylene terephthalate etc., occurring in the crystallized state at room temperature, can not in practice be produced in the glassy state due to the high rate and degree of crystallization. Only their amorphous part can undergo glass transition, and the properties of the amorphous part can be significantly affected by the crystallizable part (if the degree of crystallization is large enough, see Chapter 6). In elastomers, where the degree of crystallization does not exceed 30–60%, cooling leads to glass transition. In high-crystallization rate elastomers only the amorphous part can be glassy, and low-crystallization rate elastomers can be produced completely glassy, the same way as non-crystallizable elastomers (Fig. 1.1). Since many elastomers do not crystallize at all or crystallize slowly, for most of them glass transition is the major process, which determines their low-temperature resistance.
1.1
Methods of glass-transition temperature determination
1.1.1 Structural and other physical methods Dilatometry. The measurement of the volume or linear size is a traditional and reliable means of detecting glass transition of elastomers, proposed by Wood and Bekkedahl as
12
CHAPTER 1
Δl, mm 0.1 1 0.2 2 0.3 0.4 -150
-100
-50
0
T, °C
Tg
Figure 1.1 Temperature dependences of the change of linear size 'l for vulcanized rubbers from 1.4-cis-polybutadiene (SKD rubber) with the cis-unit content of 92% (1), 96% (2); the cooling rate v = 5°C/min.
applied to natural rubbers [68, 604, 745]. The Tg value is determined as a temperature, corresponding to the point of intersection of linear segments of the curve, which describes a change of volume or linear size of an elastomer specimen during the temperature change (see Fig. 1.1). The value of Tg determined in this way is a temperature at which the bulk (E ) or linear (D) thermal expansion coefficient
D
E / 3,
(1.1)
undergoes a jumplike change. The value of Tg determined in this manner depends on the rate of temperature change. The values of Dand E at temperatures higher than Tg characterize the properties of elastomers in the transition region. Calorimetry. The slope of the heat content–temperature curve [745] also changes at Tg, so a jump of specific heat ('cp) is observed in passing from the glassy to the rubberlike state. A jump of specific heat in glass transition is registered on thermograms recorded by differential scanning calorimetry (DSC), which represent a dependence of the heat flow dQ/dt on scanning temperature Tc, or on differential thermal analysis (DTA) thermograms, which represent the dependence of the temperature difference in the measuring and reference cells 'Tc on scanning temperature Tc[79, 246, 247, 249, 752] (Fig.1.2). Often, Tg is determined as the point of inflection on thermograms or by a maximum on the curve obtained by differentiation of the experimental thermogram. However, to obtain the Tg values coinciding with the data obtained by mechanical static methods, it is more convenient to use the temperature of the onset of the transition from the glassy to the rubberlike state* (see Fig. 1.2). The value of Tg measured by the calorimetric method also depends on the cooling and heating rates. The shape and width of the specific heat jump at Tg characterize the properties of elastomers in the transition region, too. Measurements of the electrical properties. The change of shape of the specific conductivity vs. temperature dependence can also serve as a method of determining Tg [162, 637]. However, specific conductivity measurements in the region of low temperatures to * This is especially important in the case when the termination of the transition is characterized by an abnormal rise of specific heat (see further).
GLASS TRANSITION OF ELASTOMERS
13
ΔT'
Tg1 Tg2Tg3 Tg4
T'
Figure 1.2 Dependence of the temperature change in cells, 'T c, on scanning temperature T c (a thermogram, DTA method). tan δe.102 4 3 4
2 1
1 0 -60
-40
-20
3
2
0
T, °C
Figure 1.3 Temperature dependences of the loss tangent tan Ge at different frequences of the electric field O, kHz: 0.1 (1), 1 (2), 10 (3), 100 (4).
determine Tg and the structure of elastomers in the glassy state failed to become common use due to the absence of reliable and well-reproducible methods. For (especially polar) elastomers, the glass-transition temperature Tg is often determined by the dielectric-loss maximum (Fig. 1.3) [35, 159, 162, 331, 358, 404, 407, 465, 568, 571, 594, 594a, 637]. However, when using this method, one should be certain that the maximum observed does refer to glass transition and consider not only the temperaturechange rate but also the frequency of electric-field changes. Radiothermoluminescence (RTL). An ingenious way of registering the increase of molecular mobility during the transition from the glassy to the rubberlike state is to measure the luminescence intensity during the heating of a polymer irradiated in the glassy state, usually at a temperature of liquid nitrogen (–190°C). The RTL method is widely used in studies of the structure and imperfection of low-molecular-mass crystals. It is also used to determine the glass-transition temperature of elastomers [468, 771] (Fig. 1.4). As in the dielectric loss method, it is important to make certain that the rise of luminescence intensity is determined namely by the glass-transition process. In contrast to that technique, the RTL
14
CHAPTER 1
I
-150
-100
-50 T, °C
Figure 1.4 A temperature dependence of the luminescence intensity I (arb. units), obtained by the RTL method for unfilled vulcanized rubber based on styrene-butadiene natural rubber SKS-30. 2
ΔH2 20
10
0 -150
2
-50
1
50 T, °C
Figure 1.5 A temperature dependence of the second moment 'H22 obtained by the broad-line NMR method for rubber-like copolymers of vinylidene fluoride with trifluorochlorethylene (1) and with trifluorochlorethylene and perfluoromethoxy perfluoropropyl acrylate (2).
method is the most efficient for less polar elastomers, because the luminescence intensity sharply goes down owing to its absorption by the polar (halogen-containing, nitrile) groups in the thin surface layer of polymer. Therefore, the glass-transition studies by the RTL method require in this case a significant instrument sensitivity increase. Nuclear magnetic resonance (NMR). This method also makes it possible to register segmental mobility changes in the glass-transition region. The values of Tg determined by broad-line NMR do not differ in practice from those determined under static conditions [51, 52, 637] (Fig. 1.5) with the cooling and heating rates equal. A “dynamic” modification of the method is the pulsed NMR method. In this case, a change of mobility of kinetic units in the glass-transition region leads to a change in the shape of the free induction decay curve. In the glassy state, the free induction decay curves have a Gaussian shape, whereas in the rubberlike state it is Lorentzian. The temperature at which the shape of the lines and the character of the temperature dependence of the spin–lattice relaxation T2 change is considered as the dynamic glass-transition temperature TgNMR (Fig. 1.6) [143, 194, 213, 431]. Due to the high frequency of the effect of the magnetic field, TgNMR is higher than Tg determined by static methods. Electron paramagnetic resonance (EPR). The determination of Tg by this method is based on the change of mobility of free-radical labels in transition of polymer from the glassy to the rubberlike state [88, 737]. It should be taken into account that changes in mobility may fail to register if non-standard techniques, which expand the range of measured times, are used [611]. This method of Tg determination has not become widespread.
GLASS TRANSITION OF ELASTOMERS
15
T2
TgNMR
-20
0
T, °C
Figure 1.6 A temperature dependence of the transverse relaxation time T2 (arb. units) for a polychloroprene vulcanizate.
Chromatography. The use of chromatography for glass-transition studies is based on the changes of the diffusion characteristics of polymer during its transition from the glassy to the rubberlike state. It enables not only registering the value of Tg, but also making conclusions on the morphology of polymer in the glassy state [71]. 1.1.2 Mechanical methods
The most simple and readily available Tg determination techniques are the methods based on the measurements of polymers’ mechanical properties, which change radically in passing from the rubberlike to the glassy state. The mechanical methods are widely used in glass-transition studies of elastomers as they enable the measurements both in the glassy and rubberlike states and, which is especially important, in the transition region. For other polymers, these methods are less suitable. The parameters determined at temperatures higher than Tg are called low-temperature resistance coefficients and are widely used to characterize the properties of elastomers in the transition region. The temperature dependence of deformation. The method is based on the measurement of the temperature dependence of deformation H under a given load P = const (see Fig. 1). The temperature at which the deformation begins to sharply increase is determined as the glass-transition temperature Tg [90, 313, 444, 600, 704, 705]. Under small loads, the deformation of a specimen in the glassy state does not depend on temperature; therefore, the glass-transition temperature can be taken to be the highest temperature at which there is no deformation of the specimen. A specimen may be loaded in the glassy state and then heated or loaded at each temperature studied. The accuracy is greater if Tg is determined as a temperature at which the extension of the linear segment of the dependence of H on T in the transition region intersects with the temperature axis. Sometimes, Tg is determined by the inflection point of the curve in the transition region. The method is successfully used for various kinds of deformations – tension, compression, shear, torsion (the Gehman method), bending [195] (the Clash–Berg method) [91, 600, 624].
16
CHAPTER 1
εa.102 4 3 2 1
2
1
4 5
3
0 -80 -60 -40
-20
0
20 T, °C
Figure 1.7 A temperature dependence of the amplitude of deformation Ha for unfilled vulcanized natural rubber at different frequences O(min–1): 0.1 (1). 1.0 (2), 10 (3), 100 (4), 1000 (5).
Recovery. The glass-transition temperature and the properties of elastomers in the transition region are determined by the method, which is based on the registration of the change of the ratio between the reversible and irreversible deformation components. The method is widespread; it depends on the capability of a specimen to restore its size after deformation and characterizes the relaxation properties of elastomers. If a specimen is loaded at a temperature TD on the plateau of the rubberlike state (for instance, at room temperature), then cooled down to a temperature T and held at this temperature for some time t1, the size of the specimen at T will be recovered when the load is removed. As the temperature T is decreased, the recovery values measured after a given time t2 and the recovery rate will go down to their minimum, which corresponds to Tg. The method can be used both in tension and compression. The reproducibility test of a version of this method presented in [677] gave very poor results, which, in our opinion, is due to an incorrect choice of the measuring instrument. With the proper choice of the method and the make of the instrument, the data spread does not exceed 2%. The recovery values in compression (the parameter of the low-temperature resistance during compression) is one of the most widely used in Russia to characterize the properties of elastomers in the transition region. A version of the method is also used, when TD = T, i.e., the specimen is loaded at a testing temperature [441]. The recovery method is described in detail in [95, 97]. Another variety of the recovery method is the temperature retraction (TR) method [91, 624], which consists in the measurement of specimen’s recovery in the scanning mode, when the specimen is stretched at room temperature, then cooled down to a temperature lower than Tg, after which its temperature is increased at a given rate. In this case, Tg depends on the rate of temperature change. In the TR method, one usually makes note of the temperatures corresponding to 2, 10, 50 or 90% recovery [91, 600, 624]. The reproducibility of this method is assessed in [677]. In principle, Tg can be determined using the temperature dependence of any mechanical parameter, including hardness, strength and elongation at break. Dynamical methods. The temperature dependence of the deformation amplitude Ha at given load and frequency has the shape similar to the low-temperature segment of the temperature dependence of deformation under static loading, and the glass-transition temperature can be determined in the same way (Fig. 1.7) [8]. There are several types of
GLASS TRANSITION OF ELASTOMERS
17
tan δ
G'; G'', MPa 1
103
10
102 2 10
3
1.0 0.1
1 10-1
-20
0
20 T, °C
Figure 1.8 A temperature dependence of the shear storage Gc (1) and loss Gcc (2) modulus and the tangent of mechanical loss tan G (3) for a specimen of vulcanized styrene-butadiene rubber containing 50% of styrene units: O= 100 min–1.
instruments, which enable the thermomechanical analysis (TMA) of elastomers, i.e., during the scanning they measure the dependence of the deformation amplitude or dynamic modulus on temperature. The potentialities of TMA and respective instruments, in particular, those manufactured by Monsanto, for studies of the mechanical properties of elastomers were considered in [622]. It should, however, be taken into account that at a frequency O> 5 Hz the curve of Ha vs. T obtained under dynamical conditions is shifted towards higher temperatures. An increase of the rate or frequency of the loading by an order of magnitude leads to a Tg rise by 3–5°C. The curves of the temperature dependence of storage modulus (Fig. 1.8) are shifted with respect to the static curves in the same way. As the temperature dependences of the mechanical loss, e.g., tan G or the elasticity Ecc or shear Gcc loss modulus are of an extreme character and a maximum is in the region of transition from the glassy to the rubberlike state, the magnitude of Tg is determined as the temperature corresponding to a maximum of tan G (Ecc or Gcc). However, the maximum Ecc or Gcc is usually shifted relative to the maximum of tan G towards a lower temperature [198, 483] (see Fig. 1.8), and both of these maxima are always shifted to the region of a higher temperature as compared with Tg, which is determined by the cross-section of the continuation of the curve for the dependence of Ha vs. T or Gc vs. T with the axis of T. Thus, the values of Tg determined by different methods differ even at the same frequency Q. The magnitude of Tg determined by the loss maximum usually corresponds to the point of inflection on the curve of Ha vs. T. Using the dynamical methods, the glasstransition temperature can be determined not only in the harmonic loading mode, but also in a single impact, the way it is done in the impact-resilience method [443, 645, 646, 651]. The temperature dependence of the resilience, the value of which is inversely proportional to the mechanical loss, has a minimum in the region of temperatures of the transition from the glassy to the rubberlike state. 1.1.3 Comparison of the data obtained by different methods
Considering each Tg determination method, special attention should be paid to the influence of the experimental conditions: the cooling and heating rates, the rate or frequency of loading [44, 53]. Ways to determine Tg by the experimental curves of the temperature
18
CHAPTER 1
dependence of a respective property could be different: at the beginning, end or middle of the transition region or by the position of the maximum. In principle, if the measurements are made at the same cooling or heating rates and frequency, and a respective Tg determination method is used, the results obtained by different methods should coincide. However, this is observed only for elastomers with the very narrow and well-pronounced transition region. In volume or linear-size measurements, it is 2–3°C; in specific heat changes, 5–7°C; in changes of the mechanical properties, 10–15°C. For composite materials, which rubbers are, these differences can be much greater. In practice, all these conditions often prove impossible to meet. Thus, Tg determined as the inflection point by the DSC thermogram is always higher than that found by the dilatometry method. This is because the DSC curves are usually registered in heating, which is preceded by cooling at a very high rate of >300°C/min, but the dilatometry curves are often obtained in cooling at a rate of 2–5°C/min; Tg determined by the maximum of the mechanical-loss temperature dependence is always higher than that found dilatometrically not only owing to the effect of the frequency but also due to the differences in the transition-region width characteristic of measurements by different methods. Therefore, the results obtained by different methods should be compared with great caution.
1.2
Modern views of glass transition
To date, the issue of the nature of glass transition can not be considered to be solved, and the existing theories of this process be comprehensive and generally recognized. The views of glass transition as a second-order thermodynamic transition, which arose in 1930s, proved inconsistent. In 1940s, the classical works by the Russian scientists Kobeko, Aleksandrov, Lazurkin and Gurevich [7, 8, 281, 340] formulated the kinetic theory of glass transition. However, further studies in this field led to the necessity of considering both these aspects. The theoretical views of the glass transition of polymers, in particular, the cluster and fluctuation theory of transition, are described in detail in [296, 357, 611, 615]. 1.2.1 Kinetic views of glass transition
The kinetic nature of glass transition is supported by a number of experimental facts. First and foremost, this is the dependence of glass-transition temperature Tg on the ratio of the cooling (vc) and heating (v h) rates, which is observed for both polymers and low-molecular-mass liquids. For elastomers, the dependence of Tg on vc has been studied by the method of linear dilatometry [47, 261]. An increase of vc by an order of magnitude leads to a rise of Tg by 3–4°C. The slope of the curves, i.e., the value of the linear thermal expansion coefficient D, does not change. The dependence of Tg on vc is described satisfactorily by the Bartenev equation [44, 47]: 1/ Tg
D1 D 2 log vc ;
(1.2)
D2/D1 = 0.031, as for other high- and low-molecular-mass glasses. If the specimen heating rate differs from the cooling rate, the phenomenon of hysteresis is observed. Interesting data on the effect of the cooling and heating rates were obtained for polyvinyl chloride [145,
GLASS TRANSITION OF ELASTOMERS
19
146]. The isothermal changes of volume of a specimen near the glass-transition temperature also depend on its thermal prehistory. As the transition temperature is approached from the high-temperature side, the time required to reach an equilibrium volume increases. In the glassy state, the volume relaxation rate is also the larger, the higher T is. The viscosity increase of the medium, K, which accompanies glass transition, can be approximately described by the equation:
K K0 exp U / kT ,
(1.3)
where K0 is a constant and U is the activation energy. However, at U = const, eq. (1.3) describes the experimental data satisfactorily only at T > Tg. To use eq. (1.3) to describe the changes of the polymer properties near Tg, Bartenev and coauthors proposed an equation for calculating the changes of U with temperature [44]: U
U fT / T T0 ,
(1.4)
where Uf and T0 are constants; Uf has the meaning of activation energy at T of, T0 is a characteristic temperature. A number of other equations to calculate the changes of U with temperature were proposed [198, 293]. The kinetic nature of glass transition is revealed especially demonstrably in the measurements of polymers’ dynamic, mechanical or dielectric properties, carried out at a given frequency O (see Fig. 1.7). It is seen that the values of Tg go up as O rises. Thus, under a dynamic loading the position of Tg on the temperature scale depends not only on the properties of a substance and conditions of temperature determination but also on the rate of a mechanical action. So, the polymer can be transferred from the rubberlike to the glassy state by either increasing the relaxation time W, i.e., by decreasing the temperature, or by increasing the frequency of action of a force. For the latter case, Bartenev proposed the term “mechanical glass-transition temperature” (Tgm). Bartenev has shown that the dynamic action is equivalent to the static action at O < 0.5 Hz [44]. The above-considered experimental data testify unequivocally to the kinetic and relaxation nature of glass transition and enable considering it as the result of a slowdown of physical relaxation processes at a temperature decrease. The physical relaxation processes, the change of the rate of which is determined by glass transition, were called “D-relaxation” [44]. The equivalency of the action of temperature and time (or frequency) shown in the works by Aleksandrov and Lazurkin for natural rubber [8] and in the subsequent works by Bartenev [44] and Gul’ [279] was, in fact, the basis of the temperature–time superposition principle. It can be formulated as follows: a temperature decrease in the region of transition from the glassy to the rubberlike state is equivalent to a decrease of observation time; thus, the dependence of any viscoelastic function on frequency O (or time t) obtained at one temperature can, by a shift along the axis of log O (or log t), be referred to any other temperature in the region where the physical relaxation processes of the same nature develop [198]. The value of the shift along the axis of time, DT (the horizontal-shift factor), is constant for each temperature and does not depend on the type of characteristic the shift is made for. The temperature dependence of DT for natural rubber [198, 549] is presented in Fig. 1.9. The magnitude of DT preserves the same values for vulcanizates. During the
20
CHAPTER 1
log aT 10 8 6 4 2 0 -2 -4 -40 -20
0
20
40
T - TS, K
Figure 1.9 A dependence of the logarithm of the horizontal-shift factor aT on T – Ts (T is the temperature of the experiment; Ts is the reference temperature) for NR, plotted by the data of the temperature–frequency dependence of the shear storage and loss moduli Gc and Gcc.
transition from one temperature (T1) to another (T2), one should also take into account, besides the shift along the time or frequency axis, the change of density U with temperature (from U1 to U2), which is given by the coefficient U1T1/U2T2 [198, 223]. Usually, the shift to one standard temperature (Ts) is made. The temperature dependence of DT is described by the Williams–Landel–Ferry (WLF) equation:
log aT
c10 T TS c20 T TS ,
(1.5)
where Ts is the reference temperature; c10 and c20 are empirical coefficients calculated by treating the temperature dependence of DT, when the value (T – Ts)/log DT is put off on the abscissa axis, and the value (T – Ts) on the ordinate axis; the values c10 and c20 depend on the choice of Ts. Consideration of the data for many polymers makes it possible to write down the WLF equation (1.5) in the form:
log aT
c1 T TS c2 T TS .
(1.6)
Here the coefficients c1 and c2 are constant for all polymers; only the reference temperature Ts changes [198]. Virtually for all polymers Ts = Tg + 50 (where Tg is the glass-transition temperature); c1 = 8.86; c2 = 101.6 K. Thus, the WLF equation can be also written down directly relative to Tg:
log aT
c1g T Tg
c
g 2
T Tg ,
where c1g and c2g are constants; for most polymers, c1g = 17.44 and c2g = 51.6 K.
(1.7)
GLASS TRANSITION OF ELASTOMERS
21
The temperature–time superposition principle is used for predicting the low temperature resistance of elastomers, see Chapter 2. The WLF equation in the form of eq. (1.7) was the basis for one of the ways to describe the temperature dependence of the activation energy U in eq. (1.3) [293]: U
c1gT
c
g 2
T Tg .
(1.8)
Equation (1.8) coincides in its form with eq. (1.4); herewith, if in eq. (1.4) T0 | 50 K, the calculations by eqs. (1.4) and (1.8) lead to the same results [44]. Their coincidence at T > Tg was also noted in [198]. Differentiation of the WLF equation gives a rough quantitative estimate of the change of Tg with the cooling rate: d 'Tg d log aT | 3 K.
(1.9)
This indicates that Tg shall change by about 3 K at a change of the cooling rate by an order of magnitude even if the heating rate remains constant; this is observed experimentally both for elastomers [47] and other polymers [145]. An immediate development of the relaxation concept of glass transition is to use the concept of free volume to describe the process. This concept, as the relaxation theory itself, is successfully used to describe glass transition (and other physical processes) both in low-molecular materials and polymers. A change of solid bodies’ size at a temperature change is determined by the change of free volume, i.e., the volume not occupied by molecules of a substance. Owing to intensive thermal motion, polymers in the rubberlike state are characterized by the presence of a large free volume, which is one order of magnitude greater than that for glasses and several orders larger than that for crystalline solid bodies. Therefore, the thermal expansion coefficients of polymers in the rubberlike state are at least one order of magnitude larger than those for low-molecular-mass materials in the solid state. In transition to the glassy state, the values of D and E for elastomers usually decrease 5–6-fold [from (2–3)×10 –4 down to (5–8)×10 –5 1/K] and approach the values of Dand E for some low-molecular-mass solids. The free volume Vf is the difference between the volume occupied by a body at some temperature T (VT), and the van-der-Waals volume of the molecules at 0 K (V0) (Fig. 1.10):
Vf
VT V0 .
(1.10)
In Fig. 1.10, it is shown for the glass-transition temperature Tg. According to [198], one should distinguish between the fluctuation free volume and empty volume, which are usually called the physical (VfPh ) and geometric (VfG ) free volumes, respectively. The physical free volume is the difference between the volume (VTPh ) occupied by thermally vibrated molecules of a body at a temperature T and the van-derWaals volume of the molecules at 0 K (V0) (Fig. 1.10)
VfPh
VTPh V0 .
(1.11)
The geometrical free volume is the difference between the macroscopic volume of a
22
CHAPTER 1 V
VfG Vf Ph
VTPh
Vf
V0 VT
Tg
T
Figure 1.10 A schematic illustrating the views of the geometric (VfG) and physical (VfPh) free volumes, which are shown for T = Tg.
body at a temperature T and the volume occupied by thermally vibrated molecules at the same temperature T (VTPh):
VfG
VT VTPh .
(1.12)
Or:
VfG
VT VfPh V0 .
(1.12c)
Equation (1.12c) relates the physical free volume reflecting the intensity of molecules’ thermal vibrations and the geometrical free volume reflecting the loose packing of the molecules. Considering the thermal expansion above and below the glass-transition temperature Tg, one can determine these free-volume components and the value of V0 (see Fig. 1.10). The physical properties of materials are described usually using the specific – geometric (f G ) and physical (f Ph ) – free-volume values:
fG
VfG V0 ,
(1.13)
f Ph
VfPh V0 .
(1.13c)
It seems that the concept of free volume was first used to describe the viscosity of simple liquids. The Doolittle’s empirical equation relates viscosity (K) and specific physical free volume as ln K
a B f Ph ,
(1.14)
where a and B are constants, with B |1. The free-volume concept, which is essentially based on relation (1.14), made it possible to develop the diffusion theory of low-molecular-mass substances in polymers [140]; it is also used to describe the thermal conductivity [588, 655] and behaviour of solutions.
GLASS TRANSITION OF ELASTOMERS
23
The free-volume concept is used the most extensively to describe the relaxation processes and glass transition. It relates unequivocally the mobility of segments to the remaining free volume at any temperature, so that not temperature but free volume should be used as an independent variable to describe the deformation rate; in this case, Tg is a temperature at which the free volume reaches some constant magnitude. This means that for the specific geometric free volume at Tg f gG
Tg 'E ,
(1.15)
where 'E = E1 – E2 is the difference between the volumetric expansion coefficients at temperatures above (E1) and below (E2) Tg. This empirical equation was proposed by Simha and Boier [88, 89, 654]. They suggested that the specific geometric free volume at the glasstransition temperature was constant: f gG
const.
(1.15c)
This relation reflects the main content of the concept of isofree volume (the constancy of free volume in glass transition). This concept was developed further and used with various modifications as the basis for a number of glass-transition theories [79, 246, 247, 249, 611, 635, 655]. The relation between the specific physical and geometric volumes in glass transition was established based on the consideration of the thermodynamics of mixing the molecules and free-volume elements: f gG
fgPh ln f gPh .
(1.16)
The concept of free-volume constancy finds confirmation in calculations of the packing coefficient Kpac [5]. It has been shown that K pac
V0 VT
(1.17)
is the same for all polymers at Tg: Kpac, g
0.667.
(1.18)
This observation is the basis for creating the Tg calculation scheme based on the chemicalstructure data [29]. There is a one-to-one correspondence between the parameters of free-volume theory and empirical constants, c1g and c2g, which are present in eq. (1.7) for the calculation of aT and characterize the temperature–time superposition. The correspondence is revealed if one takes into account that the value of aT is used to assess the changes of viscosity with temperature. Then we have an equation for aT identical to eq. (1.7) in its form, which makes it possible to express the parameters of free-volume theory through the constants c1g and c2g:
24
CHAPTER 1
f gPh
1 0.434 c1g ; E f
1 0.434 c1g c2g .
(1.19)
Here Ef is the free-volume thermal expansion coefficient. g g Using the known values of c1 and c2 , we obtain:
f gPh
0.025; E f
4.8 u 10 4 1 K .
(1.19c)
Then, using eq. (1.17), we have:
f gG
0.113.
(1.19cc)
Averaging of the fgG value calculated for a large number of polymers from the empirical equation (1.15) results in the same magnitude [635]. Based on the isofree-volume concept, we can also calculate the value of the specificheat jump 'cp in glass transition (at a constant pressure). Wunderlich [751, 753] proposed the “rule of 'cp constancy”, according to which the fraction of 'cp per chain element (bead) whose vibrational motion can be considered to be independent, is constant. In terms of 1 mol of these beads, this fraction is equal to [247, 249, 753]:
'cpc
11.3 r 2 J mol K .
(1.20)
For molecules of complex composition, it could be difficult to determine what a bead is; however, for a number of elastomers this rule is met satisfactorily [196, 247, 248, 249, 753]. The jump of specific heat, 'cp, in glass transition adds up from three constituents [248, 249]:
'cp
'c1 'c2 'c3 ,
(1.21)
where 'c1 is the conformational contribution to specific heat; 'c2 and 'c3 are the contributions associated with a freezeout of the free volume and a change of frequency and amplitude of structural units’ vibrations. In spite of the indisputable fruitfulness of the isofree-volume concept, it is not rigorous. Thus, the critical free volume corresponding to glass transition may not be a “universal” value. This is clearly seen for elastomers: relation (1.20) is observed for NR, styrene-butadiene and butadiene acrylonitrile rubbers, but the value of fgG calculated by eq. (1.15) for one of the most rigid-chain elastomers – fluoroelastomer – is fgG = 0.13, while for an elastomer with the minimal chain rigidity – silicon rubber – fgG = 0.08 [544]. Treatment of all available literature data for elastomers gives an average value fgG = 0.106±0.013, i.e., the isofree-volume concept is fulfilled in glass transition only in the first approximation and worse than for other polymers, and the mean value of fgG is less. Herewith, a tendency for a rise of fgG with Tg can be traced [544]. Mandelkern proposed to consider constant not the value fgG but 'E; then from eq. (1.15) it follows that fgG increases linearly with Tg. For elastomers, this is satisfied only in
GLASS TRANSITION OF ELASTOMERS
25
the region of mean values of Tg [97, 544]. For a number of polymers (polystyrene, polyisobutylene) it has been shown that at Tg the free volume weakly increases with molecular mass. In some cases, large deviations of the measured values of fgG can be due to the use of filled or plasticized polymers [408, 409]. However, the treatment of available literature data for filled elastomers, taking account of the volume fraction of a filler and in the assumption that the linear expansion coefficient changes additively with filler’s content, gives the values of fgG close to those for unfilled elastomers [544]. Modifications of free-volume theory are the hole and fluctuation glass-transition theories. The relaxation views of the nature of glass transition in combination with the concept of free volume describe well a major amount of experimental factors observed, make it possible to develop methods for predicting the course of glass transition, and can be used to obtain information on the structure of polymers, elastomers in particular. 1.2.2 Thermodynamic views of glass transition
The success of kinetic glass-transition theories temporarily pressed back the views of glass transition as a second-order phase transition. It became clear that the experimentally measured Tg is not an equilibrium temperature of this transition. However, similar features between glass transition and second-order phase transition make investigators return again to the thermodynamic aspects of the glass-transition process. Indeed, a characteristic feature of second-order phase transitions is a jumplike change of the second derivatives of thermodynamic potential – the coefficients of thermal expansion E and specific heat cp – which takes place in glass transition. At the same time, a long-time holding (annealing) of polymers near Tg may lead to maxima on the curves of the temperature dependence of E and cp [295]. This character of change of the curves is also observed during the increase of the heating* or cooling rate [222]. Besides, the experimentally measured Tg, as we pointed out above, appreciably depends on the rate of change of the temperature or frequency, definitely indicating the relaxation character of glass transition. All this gave grounds to several authors [198, 304] to suggest the existence of an ultimate transition of the second order, which is characterized by some temperature T2. The experimentally measured Tg is higher than the true equilibrium transition (T2) by about 50 K. Interestingly, the presence of a temperature about 50–55°C lower than Tg measured under static conditions is also indicated by eqs. (1.4) and (1.8), which are based on purely relaxation approaches. The existence of an extra mobility in polymer glasses in the temperature region immediately adjacent to Tg is also indicated by the stress-induced rubberlike state observed in the same temperature range (see Chapter 2). At the same time, the identification of even the ultimate transition during the glass transition process with the secondorder phase transition draws a number of strong objections [44]. However, the data on the effect of pressure p on Tg are in favour of the adequacy of glass transition as a second-order phase transition. Experimentally, they have been studied best of all for polystyrene [224, 296]; however, there are also data for other polymers and, in particular, for elastomers by Bridgman and some other authors [3, 159, 305, 546, 617]. Within the pressure changes from 1 up to 500 MPa, Tg changes linearly with pressure so that * Elucidation of the causes and conditions of such anomalies to appear in elastomers requires separate studies.
26
CHAPTER 1
dTg dp
A,
const
(1.22)
with A|0.1–0.3 K/MPa. To describe the glass-transition temperature changes with pressure, the Ehrenfest equation is used, which describes the change of second-order transition temperature T* with pressure: dT dp
T 'E 'cp
'æ 'E .
(1.23)
Here 'E
E1 E 2 ,
(1.24)
'cp
cp1 cp2 ,
(1.25)
'æ = æ1 - æ 2 ,
(1.26)
where E1, cp1, æ1 and E2, cp2, æ2 are the coefficients of volume expansion, specific heat at constant pressure and compressibility in the liquid (rubberlike for polymers, subscript 1) and the glassy (subscript 2) state. The constancy of the values dTg/dp and dT*/dp implies the linear dependence of transition temperature on pressure, which for Tg is observed experimentally (see further, Fig. 1.16). The linear dependence of Tg on p at small p also follows from the theory of free volume; herewith, the change of Tg with p changing is the larger, the lower the glass-transition temperature is. However, usually in the experiment dTg dp 'æ 'E .
(1.27)
Interesting conclusions on the effect of pressure on Tg were also made based on Gibbs–Di Marzio theory [242]. It could be suggested that the relation (1.27) is determined by kinetic effects [83]. Similar to a pressure increase eliminating the kinetic complications of the first-order phase transition – crystallization, a pressure increase also makes it possible to eliminate kinetic complications in the case of glass transition. Therefore, a detailed study of the dependence of Tg on p would not only provide data important in practice (in particular, on the low-temperature resistance of rubber components operated under pressure and uniform compression conditions) but would also greatly contribute to understanding the nature of the glass-transition process. 1.2.3 Morphological views of glass transition
Direct studies of the morphology of polymer (and low-molecular) glasses are rather numerous. However, information obtained as the result of these studies is equivocal. The presence of separate formations in glassy polymers has been registered by the electron-microscopy
GLASS TRANSITION OF ELASTOMERS
27
method [140, 758]; X-ray data, in particular, amorphous halo analysis [422], indicates the occurrence of fluctuation formations. The results of spectroscopy studies [199] lead to the same conclusions. Unfortunately, it is rather difficult to separate instrument noise and the signal from the specimen. A number of features on the thermograms recorded by the DSC method are also indicative of the structural inhomogeneity of polymers in the glassy state and in the transition region. The occurrence of a maximum in the region of the end of transition from the glassy to the rubberlike state is registered especially clearly. It was noted above that, as a rule, the maximum emerges as the result of holding a specimen at a temperature lower than Tg [295] or if the scanning rate is increased [222]. As is known, the mechanical methods are rather sensitive to the morphology of polymer, in particular, elastomer, materials. A large body of information can be obtained by studying the frequency or temperature dependence of mechanical or dielectric loss or by plotting relaxation times spectrum based on static experiments, e.g., by the relaxation of stress. As we noted above, in the region of transition from the glassy to the rubberlike state these curves have a maximum, which corresponds to D-relaxation processes. However, it is not the only maximum in these curves. From the data on the multiplicity of these curves, we can make some conclusions on the nature of glass transition. Thus, the data on the occurrence of these structures registered by the appearance of several low-intensity transitions (O-transitions) in the transition region are in favour of density fluctuations in the glass. Comparison of the parameters of these transitions with the data of the spectral methods and the character of the temperature and frequency dependences of the positions of the respective maxima support the views of their fluctuation nature [46]. The most low-temperature E-transition is usually associated with the freezeout of the mobility of side and end groups; however, transitions below Tg were found for more than 80% of all polymer systems studied, as well as for some low-molecular-mass systems. Bersthein [78–80] believes that the E-transition is indeed the true glass transition related to the change of segmental mobility, and the temperature usually determined as Tg (or D-transition) is the result of a freezeout of the cooperative motion of segments. The distinction in the nature of D- and E-transitions is indicated by the different frequency dependences for the temperatures of these transitions leading to their temperatures approaching each other as the frequency is increased. However, the different size (a segment or a group of segments) may also lead to a distinction in the frequency dependence of the temperatures of the transitions. Interestingly, at high frequencies the dielectric loss maxima, corresponding to D- and E-processes, merge [637]; the mechanical loss maxima also get nearer one to another [611]. Theoretical consideration based on the assumption of the same mechanisms of the D- and E-processes leads to the same results [270, 271]. A direct indication of the multiplicity of the glass-transition process are the data obtained using the method of inverse gas chromatography for polymethyl methacrylate, polycarbonate and polysulphone [71]. They agree with the views of the cluster structure of material in the glassy state and in the region of transition to it. However, the issue of the nature of clusters remains unclear. Some investigators [611] believe that clusters represent groups of atoms or molecules packed such that they can not crystallize in principle (“ideal” glass* cells). Other authors think that clusters can represent fluctuation formations oooooooo * The views of ideal glass are developed in the works by Bernall as applied to low-molecular glasses.
28
CHAPTER 1
structurally close to crystalline formations but smaller than the critical size of the nucleus* formed in cooling but, owing to the freezing of molecular motion, fail to develop to a critical size (or can not do so due to the insufficient length of regular sequences of molecules). Views of two types of glasses even appeared – with clusters not crystallizable at all and those that can, in principle, get crystallized. The cluster views are the basis of the Cohen–Grest [274, 150] and Rao–Rao [596] theories. At present, there is no comprehensive theory, which would enable combining all aspects of glass transition – thermodynamic, kinetic and morphological. The development of this theory is apparently possible based on the consideration of the kinetics of second-order phase transitions. However, unlike the description of the kinetics of first-order phase transitions, successfully used for polymers, approaches to the description of the kinetics of second-order phase transitions have been developed insufficiently. Development of glasstransition theory is of great practical significance both for better understanding plastics operated in the glassy state and for refining the methods of predicting the behaviour and improving elastomers’ low-temperature resistance.
1.3
Effect of the molecular parameters of rubbers and the composition of rubber compounds on glass-transition temperature
The glass-transition temperature of polymers is determined by their chemical structure. The use of Tg assessment methods based on the account of the volume and mobility of macromolecule-forming structural units [28, 29, 86, 753] gives satisfactory results. 1.3.1 Effect of molecular mass
The effect of the molecular mass Mn on Tg of polymers was first studied in the works by Kargin and Sogolova [313, 668], who showed that in one homological series in transition from low-molecular-mass substances to oligomers and then polymers there is a critical molecular mass Mnf, above which Tg (Tgf) does not depend on Mn. The values of Mnf exceed the size of a statistical segment. Properly speaking, starting from Mnf, a substance can be considered as a true polymer. As low-molecular-mass rubbers and oligomers find ever wider application, determination of the dependence of Tg on Mn is of undoubted interest [87, 157, 193, 355, 479, 610]. Several attempts are known of a quantitative description of the Tg dependence on the degree of polymerization. Thus, Fox and Flory [204] obtained a linear dependence of Tg on the inverse degree of polymerization: Tg
1
Tgf K c M n ,
(1.28)
where Kc is the constant for a given homologous series. Equation (1.28) describes satisfactorily the data for 1,4-cis-polyisoprene [741]. The following equation [153] was proposed for calculating the Tg values of blends of polymer and oligomer: * This type of elements of physical structure plays the role of physical junctions and appears [44, 97] to determine some features of the mechanical properties of elastomers in the rubberlike state.
GLASS TRANSITION OF ELASTOMERS
1 Tg 1 Tgf
m / n 1 Tgm 1 Tgf ;
29
(1.29)
here m and n are the degrees of polymerization of oligomer and polymer; Tgm and Tgf, their glass-transition temperatures. Based on the cluster theory of glass transition [596], the dependence of Tg on M has been obtained: Tg
Tgf M m M n Tgf Tgm ,
(1.30)
where Mn is the molecular mass of a monomer; Tgm is its Tg. As Tgf, Tgm and Mn are constant for polymers of a given chemical composition, eq. (1.30) can be presented as Tg
1 Ac B M n ,
(1.31)
which is analogous to the Fox–Flory equation (1.28). Glass-transition temperature can be also calculated using the molecular-mass distribution (MMD) [153]. In this case, the real MMD of a homopolymer is identified with the distribution of a blend of oligomers and a polymer. This assumption substitutes the problem of the position of Tg of a blend depending on its composition for the problem of the dependence of Tg on the degree of polymerization. Considering that the changes of specific heat in glass transition of a polymer 'cpf and of an oligomer 'cpm are close: 'cpf | 'cpm ,
(1.32)
their difference affects Tg insignificantly, so that approximately we can write: ln Tg
ln Tgf m n ln Tgm Tgf .
(1.33)
The dependences of Tg on M calculated by eqs. (1.29) and (1.33) are close despite the differences in the form of the equations. The experimental study of the dependence of Tg on M was carried out for polyisoprene (Fig. 1.11) [355]. The data obtained are satisfactorily described by the Fox–Flory equation (1.28) at Kc = 17.6 K; Tgf = 204.4 K. The values of Mnf are close to 104, and for Mn |300, Tg = 163 K, whereas for isoprene Tg = 88 K. An empirical equation was also proposed: Tgf
372.6 log xc 595,
(1.34)
(xc is the number of monomer units), which describes the experimental data for polyisoprene at Tgf = 203 K; then Mnf = 9500, i.e., also approximately 104.
30
CHAPTER 1 Tg, °C -70 -80 -90 -100 -110 -120
103
104
_ 106 Mn
105
Figure 1.11 A dependence of Tg on the molecular mass Mn for polyisoprene. R im , %
tan δ b
a 1
80
1.2
6
60 2
40
5
4
0.8
3
0.4
20 -80
6 -40
5 0
4 40
3 80
0 T, °C
2 1 -80
-40
0
40
80 T, °C
Figure 1.12 A temperature dependence of the impact resilience Rim and the loss tangent tan G for butadiene rubbers of lithium polymerization with narrow MMD and different M: 106 (1), 3.05u105 (2), 9.5u104 (3), 6.1u104 (4), 3.8u104 (5), 1.8u104 (6).
The effect of branchings on Tg can be predicted using the empirical equation [355] Tg
1
Tgf AN 2 M n ,
(1.35)
where N is the number of chain ends and A is a constant. Concluding the analysis of the results on the effect of M on Tg, let us recall that usually rubbers have molecular masses of hundreds of thousands, and the number of fractions with M < Mf is vanishing. Therefore, Tg of such polymers virtually does not depend on Mn [741], as is seen, in particular, from the data of Fig. 1.12 [443]. 1.3.2 Effect of the structure of macromolecules
The value of Tg is related to the flexibility of macromolecules, which is determined by the energy barrier of the rotation around the main valent bonds ('U). Therefore, the value of Tg is often used as a measure of flexibility of macromolecules. The lowest values of 'U are characteristic of elastomers, hence their low Tg. One of the most flexible-chain polymers – polydimethyl syloxane – has a rather low Tg = –125°C; polyethylene is also characterized by the low values of 'U, so Tg = –90°C for it.
GLASS TRANSITION OF ELASTOMERS
31
Tg, °C -20 1 -40 -60
2
-80 -100 -120
20
40
60
80 w, %
Figure 1.13 A dependence of the glass-transition temperatures of polyisoprene (1) and polybutadiene (2) on the content of units w 3,4 (1) and 1,2 (2).
Changes of Tg at a change of the microstructure and intermolecular interaction, composition and position of side groups, as well as in transition from homopolymers to copolymers can usually be reduced to changes of flexibility of macromolecules [608]. Comparison of the values of Tg of cis- and trans-polymers shows their small difference. The differences observed in some cases experimentally are determined, as a rule, by the different content of 1,2- and 3,4-units in the polymers compared. This difference is contributed to even more by crystallization. Indeed, trans-isomers are characterized by a higher rate and degree of crystallization (see Chapter 3), and the presence of the crystalline phase (at its high content) can lead to the rise of Tg of the amorphous component. As the result, the measured differences of Tg may reach 25°C, which is seen from the below data on the glass transition of isomers*: 1,4-cis Polybutadiene (PB) Polyisoprene (PI) Polypentenamer (PP) Polyoctenamer (PO)
–105 –73 –114 –108
1,4-trans –80 –60 –97 –96
1,2- or 3,4-units –5 +12
The occurrence of 1,2- and 3,4-units has a much greater effect on Tg. These changes are determined by a decrease of flexibility of the chains. The linear dependences of Tg on the content of 1,2- and 3,4-units were obtained for polyisoprene [741] and polybutadiene [438, 460] (Fig. 1.13). Thus, 1,2-polybutadiene has Tg = –5°C [221, 284a]. For polyisoprene, which contains 20–25% of 1,2-units and 70–75% of 3,4-units, Tg = 12°C [355]. The differences of the Tg values for syndio- and isotactic polymers are determined by
* If not specified otherwise, here and further we adduce data for Tg obtained by the dilatometry method at vc = 2–5°C/min.
32
CHAPTER 1
those of the intermolecular interaction of the side chains in respective positions. Thus, for any pair of stereoisomers the following dependence can be proposed [438]: Tg1 Tg 2
0.59 'E k ,
(1.36)
where 'E is the difference of the energy of isomers in the syndio- and isotactic positions; Tg1 and Tg2 are the values of Tg for the syndio- and isotactic polymers; k is the Boltzmann’s constant. The change of Tg at a change of the length of alkyl radical was considered in [444]. Cyclization, by determining a decrease of macromolecular flexibility, also leads to a significant rise of Tg. Thus, for 1,4-cis-polybutadiene the value of Tg can increase by 100–150°C as the result of the cyclization under the action of high pressure and temperature [389, 390, 440, 731]. Halogenation (chlorination, bromination, fluorination) of elastomers also leads to a rise of Tg up to temperatures exceeding room temperature. An increase of Tg in developing ebonites is, evidently, due not only to the rise of density of the threedimensional network but also to a modification of the rubber chain. To denote the increase of Tg as the result of the chemical changes of the macromolecular chains, Kuzminsky and Sedov proposed the term “chemical glass transition” [389, 390]. Formation of a surface layer, which underwent glass transition as the result of the chemical modification of a polymer can lead to a decrease of the friction coefficient and diffusion rate of gases and liquid. However, such a glassy polymer cracks under the action of stresses, and the formation of surface cracks may deteriorate the operational properties of elastomer articles [390]. This effect should be taken into consideration when using halogens for surface modification of components from elastomers. Note that it is usually very difficult to differentiate between the cases of Tg rising as the result of decreased molecular flexibility or increased intermolecular interaction. 1.3.3 Effect of intermolecular interaction
The energetics of intermolecular interaction in a liquid is described using the density of cohesion energy (Vc):
Vc
Eev V ,
(1.37)
(where Eev is the molar energy of evaporation and V is the molar volume). The value of Vc is related to the solubility parameter Ps, which can be determined experimentally [399]. As a rule, Tg increases with Ps; a linear dependence [189, 304, 639] between Tg and Ps was established for a large number of polymers. Herewith, Tg of rigid-chain polymers increases faster than that of flexible-chain polymers. A linear empirical equation relating Tg and Vc was proposed. However, in real cases it is difficult to single out the effect of only one factor – intermolecular interaction – because other molecular characteristics, which determine Tg, then also change. Vice versa, at a close intermolecular interaction the values of Tg can change significantly, in particular, due to the differences in volume of the side groups. Thus, polyvinyl acetate and polyvinyl chloride have close solubility parameters (9.4 and 9.5), however, their Tg differ by 58°C (polyvinyl chloride, Tg | 87°C; polyvinyl acetate, Tg = 29°C). As an example illustrating the effect of enhancing the intermolecular interaction, below we give Tg of rubbers with various contents c of methacrylic acid [438, 444]:
GLASS TRANSITION OF ELASTOMERS
Polybutadiene Carboxylate SKD-1
Carboxylate SKS-30-1
Styrene-butadiene SKS-30
c, %
Tg, °C
– 1.5 10 20 0.3 5 15 –
from –105 up to –112 –81 –55 +1 –56 –50 –30 –56
33
1.3.4 Glass transition of copolymers
From the concept of free volume, Tg of a copolymer
Kw1Tg1 w2Tg2
Tg
Kw1 w2 ,
(1.38)
where Tg1 and Tg2 are the glass-transition temperatures of homopolymers; w1 and w2 are their mass fractions in the copolymer:
K
D1,2 D2,2 D1,1 D2,1 ,
(1.39)
Here D1,2, D2,2, D1,1, D2,1 are the thermal expansion coefficients of the homopolymers above (D1,1, D1,2,) and below (D2,1, D2,2) Tg. Usually, Tg of statistical copolymers is described satisfactorily by the linear equation:
Tg
w1Tg1 w2Tg2 ,
(1.40)
into which eq. (1.38) goes over at close values of the thermal expansion coefficients of homopolymers [745]. An approximate validity of eq. (1.40) can be illustrated by the example of the copolymers of butadiene with isoprene by the data of [297] and [525] (rubber SKDI), curve 1, 1c, Fig. 1.14a). The linearity of the dependence of Tg on the content of copolymer is also fulfilled for copolymers of butadiene with styrene (styrene-butadiene rubbers) and acrylonitrile (butadiene acrylonitrile rubbers), curves 2 and 3, Fig. 1.14a*. However, extrapolation for the 100% content of styrene or acrylonitrile gives values of Tg equal to, respectively, 24°C and 72°C, i.e., significantly lower than those for polystyrene (100°C) or polyacrylonitrile (140°C). Extrapolation for the 100% content of polybutadiene gives in the three cases considered different values of Tg, which is due, as we mentioned above, to the differences in the microstructure of the butadiene segments of the chain in these copolymers. From these * Some deviations from linearity in the Tg dependence of butadiene rubbers on the content of comonomer can be due both to the difference in the content of 1,2- and 3,4-units in the specimens studied (which is determined by the differences in the conditions of their synthesis) and to the differences in the real content of comonomer.
34
CHAPTER 1 Tg, °C
Tg, °C
-50
a
b
-20 1 3
-40
2
-60
1'
-80
2
-100
1
3
-100 0
20
40
60
80 w, %
-150
0
50
w, %
Figure 1.14 A dependence of the glass-transition temperature Tg for statistic copolymers of butadiene (a) on the content of comonomers: isoprene (1, 1c), styrene (2), acrylonitrile (3) and silicon rubbers (b) on the content of modifying units: phenyl (1), fluorosilicon (2) and ethyl (3).
extrapolated values of Tg, one can, using the data of Fig. 1.13, determine the content of 1,2and 3,4-units. The data of Fig. 1.14a show that it is minimal in the case of copolymers with isoprene of the type of rubber SKDI. A slight increase of Tg is also caused by an increase of the content of piperylene units in 1,4-cis-polybutadiene. Thus, butadiene rubber containing a minor amount (| 10%) of piperylene units (grade SKDP) has Tg = –98°C, i.e., several degrees higher than Tg of the homopolymer (grade SKD) [108, 314]. The dependence of Tg on the content for the copolymers of ethylene and propylene is peculiar (see Chapter 6). For silicon rubbers, Tg increases with the rise of the fluorosilicon content (in the series SKTFT-50 and SKTFT-100 with the 50 and 100% content of these units) and phenylsilicon units (for instance, rubbers of grades SKTFV-803 and SKTFV-2101 with 8 and 20% content of these units) [256, 319, 421, 498, 499] and decreases with the rise of the content of ethyl units, i.e., those that have side groups with a smaller density of cohesion energy (see Fig.1.14b). For a polymer containing 100% of such units, the glass-transition temperature is observed to be rather low (Tg = –141°C) [64, 421]. A similar effect is well known for polyesters. Thus, when developing a formulation of low-temperature resistant rubber compounds, one should first of all take a rubber with the sufficiently low glass-transition temperature. However, in the case of rubbers of regular structure the low-temperature resistance can be limited by another process – crystallization (see Chapter 3). 1.3.5 Effect of the composition of rubbers Rubber blends. A promising way of developing low-resistant rubbers is to use blends of rubbers, one of which has low Tg and the other enables formulating rubber compounds with a required set of physical and chemical properties [563, 565, 728]. Often, this also contributes to the suppression of crystallization (see further). If two polymers are compatible, their blend has one Tg lying between the Tg of the components [371, 472, 577, 578]. For elastomers, such systems are blends of butadiene
GLASS TRANSITION OF ELASTOMERS
35
K
Δl 1
1.0 2 0.5
-150 -100 T* -50
0 T, °C
0
Figure 1.15 A temperature dependence of the changes of length 'l (1) and recovery K (2) for a rubber from a blend of butadiene (SKD) and isoprene (SKI-3) rubbers at a ratio of 1:1. T*, conventional glass-transition temperature.
rubbers of various types, blends of styrene-butadiene and butadiene acrylonitrile rubbers with various contents of styrene and acrylonitrile groups (e.g., between SKS-30 and SKS-50, or between SKN-18, SKN-26 and SKN-40). Butadiene and styrene-butadiene rubbers are of limited compatibility. For compatible polymers, as for copolymers, the additivity condition is satisfied (see eq. (1.40)): Tg
w1Tg1 w2Tg2 ,
(1.41)
where w1 and w2 are the contents of the components. One glass-transition temperature and its additivity with the change of the content is also observed for the blend of SKN-40 with polyvinyl chloride [371]. However, most elastomers are incompatible. The first sign of this is the occurrence of two glass-transition temperatures determined by any of the methods considered* [36, 294, 371, 412, 468, 577, 578, 609, 650, 707, 771]. In Fig. 1.15 (curve 1), the occurrence of two glass-transition temperatures for blends of 1,4-cis-polybutadiene (SKD, Tg1) and 1,4-cispolyisoprene (SKI-3, Tg2) is illustrated by the results obtained by the dilatometry method [707]. However, the low-temperature resistance of rubbers from a blend of such incompatible elastomers can be conditionally characterized by some intermediate temperature T*. This temperature can be obtained by extending to their intersection the segments of the lines corresponding to the change of length of a specimen, when both elastomers are in the rubberlike and the glassy state (curve 1, Fig. 1.15). For rubbers from a blend of SKI-3 and SKD, the zero-recovery value corresponds to just this temperature (curve 2, Fig. 1.15). Two luminescence-intensity maxima corresponding to Tg of the components were found for a blend of SKI-3 and SKD by the RTL method [468]. Their positions do not depend on the composition of the blend and the preparation technique. A change of the blend composition leads to a change in the ratio between the intensities of the respective maxima.
* In fact, the occurrence of two glass-transition temperatures is only indicative of no compatibility in the temperature range close to Tg (if the measurements were not preceded by quenching, i.e., a sharp cooling down to a temperature much lower than Tg, which prevents phase separation).
36
CHAPTER 1
Two glass-transition regions corresponding to the glass transition of the initial components were also found in studies of blends of NR and butadiene acrylonitrile rubber SKN-40, NR and fluororubber SKF-32, PVC and butadiene acrylonitrile rubber SKN-18 by scanning calorimetry [577]. For blends of polymers with close Tg (NR–polyisobutylene, NR–ethylene-propylene rubber SKEP), attempts to single out particular glass-transition regions failed; however, the heterogeneity of the blends in these cases is manifested in the change of shape of the specific heat jump during the glass transition and an extension of its temperature interval, which embraces the transitory regions of both components* (see, e.g., [155, 752, 754]). A blend of SKD with styrene-butadiene rubber with a small styrene content (SKMS10) is characterized by two glass-transition temperatures; however, a slight increase of Tg for the SKD phase is observed, which is indicative of a partial compatibility of this rubber with SKMS-10. Interestingly, vulcanization, which fixes the compatibility of these rubbers at high temperatures, leads to the occurrence of a stretched transition region and one Tg for rubbers based on their mixture [707, 708]. When blending three rubbers, e.g., SKD, SKI-3 and SKMS-10, attempts to find three glass-transition temperatures prove unsuccessful due to the close Tg for SKI-3 and SKMS-10. However, the shape of the transition in the region between Tg of these rubbers is complicated. Considering the important role of blending two and more rubbers to produce lowresistant polymers, information of the low-temperature behaviour of the blends of elastomers can not be considered to be exhaustive. Of especial significance are the studies of glass transition and changes of the properties in the transition region for real rubber compounds based on rubber blends, which are vulcanized systems containing, besides, plasticizers, fillers and other ingredients. Plasticizers. After the type of rubber (or a blend) is chosen, introduction of plasticizers is the most efficient way of decreasing the glass-transition temperature of elastomers and, thus, improving their low-temperature resistance related to glass transition [181, 417, 740]. The mechanism of decreasing Tg during the introduction of plasticizers consists, first and foremost, in reducing the intermolecular interaction. Therefore, introduction of plasticizers into rubber compounds based on polar rubbers is the most efficient. Another important cause of decreasing Tg in the introduction of plasticizers is to increase free volume, so that the critical free volume fgG (or the value of 'E ) corresponding to the glass transition of the system are reached at lower temperatures. The decrease of Tg during the introduction of plasticizers (which is a measure of the glass-transition efficiency) is described by two most widespread equations: • Zhurkov equation: 'Tg
Kn ;
(1.42)
K cV1 ,
(1.43)
and • Kargin–Malinsky equation: 'Tg
* Therefore, one glass-transition temperature for a blend of rubbers with close Tg is not the univocal feature of their compatibility.
GLASS TRANSITION OF ELASTOMERS
37
where 'Tg = Tge – Tg (Tge and Tg are the glass-transition temperatures of an elastomer with and without a plasticizer); n is the number of moles of a plasticizer; V1 is its volume fraction; K and Kc are constants. Equation (1.42) was obtained for polar polymers and plasticizers and performs well for polar elastomers. Equation (1.43) is usually used for nonpolar elastomers. However, the experimental data for the decrease of Tg of both polar and nonpolar elastomers during the introduction of a plasticizer are usually well described by eq. (1.43) [174]. A correction for this equation, which takes into account the nature of a plasticizer, was proposed [673, 674]: 'Tg
K1V1 'Tgc ,
(1.44)
where 'T gc = Tg – Tgp; Tgp is the glass-transition temperature of a plasticizer; K1 is a constant. Based on the free-volume concept, an equation similar to eq. (1.38) was obtained for describing the effect of plasticizers on Tg:
Tg
KwpTgp weTge Kwp we ,
(1.45)
where Tgp and Tge are the glass-transition temperatures of a plasticizer and an elastomer (or another polymer); wp and we are the mass fractions of a plasticizer and an elastomer (or another polymer);
K
a1p a2e a1p a1p ,
(1.45c)
where D1p, D2p, D1e, D2e are the coefficients of thermal expansion of a plasticizer (D1p, D2p) and a polymer (D1e, D2e) above (D1e, D1p) and below (D2e, D2p) Tg. This equation can not be considered to be universal, as it assumes an equal decrease of Tg of the blend during the introduction of the same amount of the same plasticizer into the polymers with the same Tg, which is not the case in reality. On the other hand, calculations by eq. (1.45) show that Tg of the blend can not be lower than that in the component that undergoes glass transition at a lower temperature. A detailed analysis of the applicability of the equation was carried out [182] for polar rubber–polar plasticizer and nonpolar rubber– nonpolar plasticizer systems. It was found that when Tg of a plasticizer was lower than that of a polymer, eq. (1.45) described experimental data satisfactorily; however, the convergence of the Tg values of the components led to an increase of the difference between the calculated and experimental values of Tg. Other equations were also proposed for describing Tg changes [145, 182]. The most important factor determining the efficiency of a plasticizer, i.e., the value 'Tg, is its compatibility with the polymer, determined by the value of equilibrium swelling Qf of the polymer in the plasticizer. The value of the Flory–Huggins constant F correlates with the value of Qf. However, the values of Qf and F do not always univocally determine the 'Tg value [417]. Studies of the effect of Fand the chemical structure of model plasticizers (hydrocarbons C8 –C12) on the glass-transition temperature and viscosity of blends of plasticizers with nonpolar rubbers have shown that the efficiency of plasticizers diminishes in the series: linear, branched, cyclic. Phase diagrams have been obtained, which show
38
CHAPTER 1
the demixing region of the system and the critical concentration restricting the efficient use of a plasticizer; this concentration diminishes in the same series [229]. Analysis of the phase diagrams of the polymer–plasticizer systems is of great importance for understanding the efficiency of plasticizer action [140, 696, 697]; however, their plotting in the region of low temperatures and with account for the equilibrium kinetics is associated with significant experimental difficulties. NMR data are also useful for understanding the changes of the low-temperature resistance of elastomers in the presence of a plasticizer [431, 511, 512]. The specific action of crystallizable plasticizers has been studied insufficiently [330, 509, 592]. It is known that during the introduction of crystallizable plasticizers Tg of rubbers decreases below the melting temperature of the plasticizer itself. This is due to the change of the crystallization conditions of the plasticizer in the rubber, where the size of its crystals is limited by the three-dimensional network and the presence of a filler (similar to how the size of the crystals in crystallizable rubbers is limited, see Chapter 3). The crystallizable plasticizers include one of the most efficient plasticizers, dibutyl cebacinate. For the rubbers containing it, Tg depends not only on the concentration of the plasticizer introduced and type of polymer, but also on the type of the vulcanizing system and filler. The efficiency of a plasticizer depends on the method of its introduction into the rubber blend. Thus, it has been shown that the swelling in a plasticizer enables producing rubbers with the glass-transition temperature 5–10°C lower than in rubbers into which a plasticizer was introduced by a traditional method [509, 510]. The ultimate concentration of a plasticizer leading to a change of Tg also depends on the type of the vulcanizing system and the filler [417]. From the practical point of view, it is important to develop permanent plasticizers. Their action (and presence in rubber) is preserved for the entire term of service of a rubber component, in the oil medium including [190, 563, 564]. Diffusion on a surface in long-term storage at low temperatures or elution by aggressive liquids (as a rule, with replacement of part of the plasticizer with these liquids) results in an inevitable worsening of low-temperature resistance of elastomeric materials. These processes make necessary the assessment of long-term low-temperature resistance of not only crystallizable, but also non-crystallizable rubbers. The possibility of using low-molecular-mass polymers [577] was investigated with the aim to develop permanent plasticizers, e.g., low-molecular-mass modified atactic polypropylene, grafted polysiloxane or low-molecular-mass polyethylene were added to rubber compounds. However, the use of these polymers failed to solve the problem of developing low-temperature-resistant rubbers. Grafting of plasticizing fragments to the main chain of an elastomer did not receive due attention, either. The role of permanent plasticizers can be partially performed by rubbers with low Tg and low crystallization rate, i.e., the problem is reduced to the development of low-temperature-resistant rubber compounds based on rubber blends [563, 728]. Three-dimensional network. The effect of the three-dimensional network on Tg and the properties of rubbers near Tg are much less pronounced and, therefore, probably, have been less studied. Crosslinking of elastomers can be considered as a special case of increasing the intermolecular interaction of elastomers. As the result of crosslinking, the free volume decreases, and Tg should go up. However, at usual densities of crosslinking accepted for elastomers, Tg changes little with the density of the network (until the value of Mc becomes comparable with the magnitude of the kinetic segment). Experimental data on the effect of vulcanization, leading to the formation of a three-
GLASS TRANSITION OF ELASTOMERS
39
dimensional network, on glass transition can be divided into two groups. A larger part of them are the data on the increase of Tg during the formation of a sufficiently dense three-dimensional network, as, in particular, ebonites or epoxy resins. Considerably less determined are the data related to networks of low density formed in vulcanization of ordinary rubbers. Interestingly, the molecular mobility assessed by the change of spin-lattice relaxation time by the NMR method is the same for non-vulcanized polychloroprene, its thermovulcanizate and the vulcanizate obtained in the presence of metal oxides [100, 213]. On the other hand, the change of the character of molecular mobility in polybutadiene networks depending on their density was determined by the same method [555]. For the vulcanizate of polychloroprene, a shift by about 5°C of the dielectric loss maximum at a frequency of 1 kHz towards higher temperatures and its narrowing as compared with non-vulcanized rubber, as well as smaller changes of the dynamic modulus in glass transition were observed [35, 506]. At lower frequences, an abnormally large shift of the mechanical loss maximum was observed for the same specimens. The causes for this difference are not clear. An absolutely clear-cut increase of the glass-transition temperature (at a frequency of 1 Hz) with the increase of the sulphur content was shown by the DMA method for vulcanizates of styrene-butadiene rubber [622]. The data on the rise of Tg by 7–10°C at an increase of the three-dimensional network density 5- to 10-fold were obtained for atactic polyvinyl acetate and polyisobutyl acrylate vulcanized by benzoyl peroxide for different times [463, 634]. However, these polymers have a lower flexibility than ordinary elastomers, i.e., their values of Tg are within the region of 30–50°C. Besides, for these polymers the differences of the volume expansion coefficients above and below the glass-transition temperature are small, which reduces the accuracy of Tg determination. In practice, it is well known that when passing from vulcanizing groups, which create C–C and monosulphide bonds, to groups containing sulphur (| 2 weight fractions) with accelerators, the glass-transition temperature increases. The efficiency of plasticizers also changes [418, 419]. There is no satisfactory explanation of these effects at present. Thus, in [181, 418, 419] they are related to the different character of the physical structure of the vulcanizate formed by various-type bonds. An undoubted contribution to this difference is introduced by an elastomer chain modification, which usually leads to the rise of the glass-transition temperature. This is the result of the flexibility change of the polymer chains owing to the presence of branchings and side groups, formed in vulcanization, as well as the junctions of the vulcanization network itself; an important role is also played by the isomerization and cyclization processes in vulcanization. However, the effect of vulcanization on the glass-transition temperature has not been studied in detail, though such data could be not only of practical significance, but would also provide valuable information on the structure of the network. Nevertheless, vulcanization by peroxides can be recommended for production of more-temperature-resistant rubber compounds from non-crystallizable rubbers. Some effect could also be given by the use of sulphur-containing vulcanizing agents, which form crosslinks of low sulphidity (dithiomorpholin, tetramethyl thiuram disulphide, etc.). Fillers. Information on the effect of filling on the glass transition of polymers is rather contradictory. For rigid-chain polymers, Tg is often observed to increase [410], which is associated with a decrease of the molecular mobility of a polymer at the boundary with a filler. For elastomers, as works by Bartenev et al. [56, 58] show, the glass-transition temperature measured, in particular, by the dilatometry method, does not change during the introduction of such fillers as carbon black in the amount of up to 30 vol. %. The absence of Tg change
40
CHAPTER 1 Tg, K 350 2
1
3 250
150
0
500
P, MPa
Figure 1.16 A dependence of the glass-transition temperature Tg on pressure p for natural (1), styrene-butadiene (2) and butadiene (3) rubbers.
during the introduction of up to 100 parts by weight of carbon black into some elastomers, e.g., thiocols, is also registered by the DTA method [196]. For these systems, the linear expansion coefficient Dchanges additively during the change of the filler concentration, and as calculated per polymer matrix the values of D, free volume [544] and specific heat 'cp at Tg do not depend on the content of carbon black [196, 544]. Coincidence of the Tg values of filled and unfilled polymer is usually registered at average cooling and heating rates [567]. However, an additional maximum emerges in the relaxation spectrum to fix the relaxation processes in the filler-associated part of the polymer, which is accompanied with a decrease of the intensity of the basic D-maximum [44, 61]. The use of the NMR method in a pulsed modification also makes it possible to reveal an additional relaxation process in carbon black-filled elastomer system [100]. A change of properties in the elastomer layer associated with the filler [179, 216, 743 754, 755, 757, 757a, 757b, 763] (formation of the so-called “pseudoglassy” layer [784]) and the emergence of an additional transition caused by this change lead (at an invariable Tg) to an apparent broadening of the transition region, which is registered by mechanical methods (see Chapter 2). For large concentrations of active carbon black, as well as other, especially fibrous, types of fillers, formation of a “pseudoglassy” boundary layer and changes of the relaxation properties are accompanied with noticeable changes of free volume and the emergence of non-additivity of the concentration dependence of the linear-expansion coefficient [784]. In this case, the low-temperature properties can also change significantly. Thus, for rubbers filled with short fibres, the dependence of the linear expansion D on temperature can be abnormal. The values of D higher than Tg (D1) prove lower than those at temperatures lower than Tg (D2). This is due, on the one hand, to the considerable decrease of D1 during the introduction of short fibres, and on the other to the separation of fibres from the rubber matrix during the passage through Tg in the case of their insufficiently strong bond. Then the value of D2 is determined only by the matrix. However, in this case, too, the value of Tg remains invariable. In the recent years, precipitated silica began to be increasingly used as an active filler of rubbers, especially in fabrication of tyres [285, 484, 573, 574, 580, 691]. Unfortunately, no direct data are available on the effect of this filler on the glass-transition temperature. The effect of nanofillers on Tg has not been systematically
GLASS TRANSITION OF ELASTOMERS
41
studied, either [209, 285, 306, 580, 589, 576, 721, 722]. Thus, according to [306], introduction of small doses of fullerens does not practically change Tg determined by the temperature dependences of the storage modulus and loss tangent. Low-temperature resistance of rubbers has been shown to improve at the introduction of ceolites [563]. On the whole, it can be considered that the glass-transition temperature of elastomers practically does not depend on the type and dosage of a filler, and its effect on the lowtemperature resistance of rubbers is exhibited, first and foremost, in the transition region.
1.4
Effect of a mechanical action on glass-transition temperature
Pressure. The effect of the uniform pressure on Tg was mentioned in Section 1.2.2. An increase of Tg under the action of pressure, described by eq. (1.22), can be hundreds of degrees. Thus, according to the data by Bridgman, the glass-transition temperature of a number of rubbers reaches 20±3°C and higher at pressures p | 500–800 MPa [97]. Figure 1.16 presents dependences of Tg on p for some elastomers* plotted also from the data of [617]. The data for polysiloxanes are given in [3]. Thus, information on the effect of pressure on the glass-transition temperature of elastomers is not numerous [3, 159, 305, 546, 617]. Still, they are needed both for understanding the nature of glass transition and for practical purposes. They make it possible, first and foremost, to estimate the low-temperature resistance of elastomers operated in the region of high pressures, as well as to choose correctly the processing parameters, in particular, the regimes of degradation or vulcanization under high pressure. Stress. Immediate data on the effect of stress on Tg are rather contradictory [226, 463, 579, 633, 634]. Mention of both the rise and decrease of Tg at an increase of stress can be found. Some contradictions are, apparently, due to the fact that the measured value is not always the true glass-transition temperature. For such elastomers as 1,4-cis-polyisoprene, deformation can be accompanied by crystallization, which significantly complicates the data interpretation [97]. The change of Tg during the orientation of filled elastomers [579] is, evidently, also due to the peculiarities of the method of experiment. The rise of the glass-transition temperature by 5–7°C with the increase of deformation for polymers with Tg | 30°C can be both the consequence of a relatively low Tg determination accuracy in stretched specimens by the dilatometry method or the contraction of a specimen, and the low flexibility of the macromolecules of the specimens studied. Both the deformation conditions and Tg determination methods differ. Thus, the measurement of the effect of tensile deformation on the change of specific heat during the glass transition has shown that within the limits of deformation H = 1 this dependence is not observed if Tg is measured by the point of inflection of the cp –T dependence [579]. If Tg is fixed as an initial point of the transition region, the glass-transition temperature tends to decrease. A considerably less pronounced tendency for the decrease of Tg was observed in measurements using the dilatometry method or by the temperature dependence of recovery. These changes for specimens deformed in the rubberlike state exceed the experimental error only insignificantly. The greater the flexibility of the macromolecules of a polymer and the lower Tg, the weaker this tendency is [545]. * Extrapolation of the data adduced in [617] for 1,4-cis-polybutadiene for P = 0 (curve 3 in Fig. 1.16) gives the values of Tg larger than that for the initial polymer. This is due, evidently, to the cyclization of polybutadiene under the action of high pressure and temperature [731]; it is also possible that these results refer to crystallization not glass transition [97].
42
CHAPTER 1
Thus, if an elastomer is deformed in a rubberlike state and is not crystallized herewith, the effect of an applied stress on aTg is not large. The effect of a stress on the mechanical properties of elastomers in the transition region and in the glassy state will be considered in the next chapter, as well as the phenomena observed during the dynamic impact.
2
Mechanical Properties of Elastomers near the Glass Transition Temperature
The results of the comparison of Tg values obtained by different techniques show a significant difference between the values determined by physical and mechanical methods. As we have pointed out, an important role can be played both by the frequency in the mechanical measurements and by the loading. Besides, depending on the service conditions of elastomeric materials, they can be characterized by a low temperature boundary Tlow; this boundary is determined by the properties of the material itself and the level of the properties required to provide the operation of elastomeric goods. This issue will be considered in detail in Chapter 6; here we would only note that Tlow and Tg can differ significantly. Herewith,
'Tlow
Tlow Tg
(2.1)
is usually positive, i.e., Tlow is higher than Tg; for particular applications, however, the condition can be achieved when Tlow < Tg (see below and also Chapter 6). The temperature Tlow is determined not only by Tg but also by the level of the mechanical properties near this temperature. Therefore, when considering the low-temperature behaviour of rubbers, it is important not only to know Tg and understand how it changes under the action of various factors, but also to consider all the mechanical properties of elastomers near (both higher and lower than) Tg and the character of their change with the temperature and under the action of various factors.
2.1
Mechanical properties in the region of transition from the rubberlike to the glassy state
When considering the effect of one factor or another, one should bear in mind that all glass transition-related changes in the mechanical properties of elastomers at a temperature decrease are determined by a deceleration of the relaxation processes and, as a consequence, an increase of the ratio of elastic deformation and rubberlike deformation. It is this that determines increased rigidity, decreased deformability and recovery, greater dynamic modulus and mechanical loss changes.
44
CHAPTER 2
2.1.1 Static mechanical properties The change of the static mechanical characteristics of elastomers in the transition region is determined not only by Tg but also by the character of changes of the properties in transition to the glassy state, as well as by their level below Tg. Conventionally, the changes of a property (A) in the transition region can be characterized by a slope (T ) of a segment of the curve, which describes the dependence of this property on temperature, to the temperature axis. Important characteristics are also the value of change of this property ('A) in transition from the rubberlike to the glassy state, and the width of the temperature interval 'T, within which it occurs. Conventionally, by linearizing the curve, which describes the transition, we can write:
T
'A 'T .
(2.2)
If we believe that in transition from the rubberlike to the glassy state a relative change of A is from 1 to 0, which takes place for all low-temperature resistance coefficients, i.e., 'A = 1, then
T
1 'T .
(2.2c)
The values of T are different for different rubbers. Thus, for rubbers from NR (in the absence of crystallization) T is usually larger than for styrene-butadiene rubbers. However, such comparisons are difficult to make, as changes of composition, in particular, of the vulcanizing group, and the density of crosslinks, of a filler, of its amount, as well as the type and amount of plasticizer significantly change the value of T. Still, knowledge of this effect is necessary, because both Tlowl and 'Tlow in eq. (2.1), as the values of the low-temperature resistance coefficients in the transition region (see Chapter 1), are directly related toT. The effect of the characteristics of rubber on the mechanical properties of elastomeric materials in the transition region is mainly reduced to their influence on Tlow. A specific effect on the dynamic properties of raw rubbers and other non-crosslinked elastomer systems is exerted by the molecular mass and molecular-mass distribution. This effect is rather difficult to trace as applied to their static mechanical properties in the transition region. However, for vulcanizates this effect is levelled down. Therefore, below we will consider the effect of the composition of rubbers on the static mechanical properties for vulcanized systems only. Rubber blends. Vulcanized rubbers from rubber blends are characterized by a broad transition region and low values of T. If rubbers are incompatible, the transition region includes the transition regions of all rubbers in the blend. It begins with the onset of the transition region of a rubber with the lowest Tg (Tgl) and completes with the end of the transition region of a rubber with the highest Tg (Tgh). However, usually at temperatures close to Tgl rubbers behave as glassy polymers up to a temperature T* (Tgl < T* < Tgh), which can be called the “conditional glass-transition temperature” and can be determined by eq. (1.41). Exactly at T* for rubbers from rubber blends the capability of recovering their size degenerates, i.e., the recovery K = 0 (in Fig. 1.15, this is illustrated by the data for a rubber based on a blend of isoprene (SKI-3) and butadiene (SKD) rubbers). The use of a graphic method for T* determination shows that this value depends on the ratio of the linear expansion coefficients of the components in the rubberlike and the glassy state; it does depend, too, on the type of the vulcanizing group and, on the whole, on the composition of rubbers.
MECHANICAL PROPERTIES OF ELASTOMERS NEAR GLASS TRANSITION
45
As we have already shown in Chapter 1, the use of rubber blends opens a possibility of developing rubbers with satisfactory low-temperature resistance and a required set of properties. In this blend, a rubber with Tgl plays the role of a permanent plasticizer; herewith, it contributes to the improvement of low-temperature resistance of rubbers in the transition region [563, 728]. Plasticizers. Introduction of plasticizers leads to a decrease of Tg and can cause a reduction of T. As we have shown in Chapter 1, Tg decreases monotonely (virtually linearly) during an increase of the content of plasticizer. However, the level of elastic properties in the transition region, in particular, the recovery K, increases only up to certain limits; herewith, these ultimate values of K depend on the type of three-dimensional network. These specific features of the relaxation behaviour of plasticized systems are associated with the presence of more or less ordered regions in the elastomer and with the effect on K mainly of that part of the plasticizer, which is localized in unordered regions [169, 171, 368, 419]. An improved low-temperature resistance upon introduction of a plasticizer by way of swelling of a vulcanizate (as compared with its introduction by way of mixing [368, 641]) is associated with this effect. The presence of plasticizers also changes the character of the effect of a stress on the deformability and recovery of rubbers above Tg [545], which is explainable by a change of the ability of macromolecules for orientation in the presence of a plasticizer. Unquestionable resources for increasing the level of preservation of the properties above Tg for rubbers operated in liquid media are opened by using permanent plasticizers. Experiments also show that a major role in preserving the properties in this case is played by the nature of the liquid medium used. However, no systematic works of this kind appear to have been carried out. Three-dimensional network. An increase of the density of crosslinks always leads to a rise of T and a decrease of 'T and, therefore,'Tlow. Herewith, the use of vulcanizing groups forming C–C and monosulphide bonds is more efficient, i.e., leads to larger Tand smaller 'Tlow [370, 728]. The low values of Tg for some butyl rubber-based vulcanizates are determined, besides Tg of the rubber itself, by its low unsaturation, i.e., by the low density of crosslinks in the rubbers produced after vulcanization. The same results are stipulated by the low vulcanization capability of some types of silicon rubbers. Therefore, although the increase of the density of crosslinks can lead to an insignificant rise of Tg (see Chapter 1), networks of sufficiently high density should be chosen for production of rubbers with low Tlow and high values of low-temperature resistance coefficients. The networks should preferably contain C–C and monosulphide bonds, which lead to a slower increase of Tg [362, 363, 369]. These vulcanizing groups are recommended to be used for developing low-temperature-resistant rubbers based on non-crystallizable rubbers. Fillers. As we have shown in Chapter 1, addition of a filler does not change Tg but an increase of its activity invariably leads to a decrease of T [370, 552, 728] and, therefore, to a reduction of the low-temperature resistance coefficients. The effect is more pronounced for rubbers from nonpolar rubbers [362, 367]. Thus, introduction of carbon black causes a decrease of the low-temperature resistance coefficients of rubbers based on styrene butadiene rubber (grade SKMS-30ARKM-15) to a greater extent as compared with rubbers based on butadiene acrylonitrile SKN-26. The slope of the curve for the dependence of recovery K or deformation H on temperature in the transition region for chalk-, clay- or inactive carbon black-filled rubbers is practically the same as that for unfilled rubbers. At the
46
CHAPTER 2 ε/ε0 1.0 2
0.8
1
0.6 0.4 0.2 0 -60
-40
-20
0
T, °C
Figure 2.1 Temperature dependences of the relative change of deformation H /H0 during the stretching for rubbers from styrene-butadiene SKS-30 ARKM-15 filled with active carbon black (1) and chalk (2).
introduction of the same amount of active carbon black the slope, i.e., the value of T, is significantly decreased (Fig. 2.1). Possibly, in the latter case the decrease of T is also contributed to by the smaller efficient density of the network, which is due to the sorption of part of the vulcanizing agents on the filler. At a carbon-black content of 80 weight fractions, the differences in the lowtemperature resistance of rubbers with carbon black of various activities are levelled off [632, 728]. Detailed information on the effect of fillers on the deformation properties of rubbers in the transition region at small deformations is given in a series of works [361– 367]. There are data on the increase of rubber low-temperature resistance coefficients in the transition region at the introduction of ceolites [563]. In the recent years, precipitated silica began to be increasingly used as an active filler, especially in rubbers for manufacturing of tyres (see Chapter 1). Unfortunately, no systematic information on its effect on the lowtemperature resistance parameters is available. Practically unavailable are also any data on the effect of nanofillers on low-temperature resistance. It appears to be not large, as, first of all, nanofillers are used only in small doses. Thus, according to the data of [306], introduction of small doses of fullerens practically does not change the low-temperature part of the temperature dependences of the storage modulus and loss tangent. As the data on the effect of pressure on the behaviour of rubbers in the transition region are practically absent, consider the influence of deformation and shape of a rubber article. Deformation. The effect of deformation or stress on the behaviour of elastomers in the transition region manifests itself differently depending on whether the specimen was deformed under conditions of the complete development of rubberlike deformation at a temperature T0 [101, 544, 545] or in the transition region, at a temperature of the experiment T < T0 [62, 361, 732]. This refers both to the rigidity of the specimen and to such a relaxation characteristic as recovery. In the former case, in the deformation measurement, a specimen predeformed at T0 to a deformation H0 is loaded by an additional load 'P at a temperature T, and its additional deformation 'H is measured. As the low-temperature behaviour characteristic, the value of the low-temperature resistance coefficient with respect to deformation is used: KH
'H 'H 0 ,
(2.3)
where'H0 is 'Hat T = T0. The value of KH measured in this way and the respective value
MECHANICAL PROPERTIES OF ELASTOMERS NEAR GLASS TRANSITION
47
of T increase with the rise of predeformation up to H0 | 0.5 both in elongation and compression. Similar to KH, the value of recovery K, which is measured in the same way and is called the low-temperature resistance coefficient with respect to elastic recovery (and the respective value ofT) increases with H0 going up both in tension and compression [544, 545]. At large elongations (starting from H0 = 1), the values of K and KH do not change, which can be explained by the intensive formation of the physical junctions in orientation and cooling. Indeed, the ultimate development of this process – crystallization – leads to a decrease of K (see Chapter 3). Otherwise, the effect of deformation (or stress) can manifest itself in the latter case – i.e., if the deformation is performed at the testing temperature – not on the rubberlike state plateau, but in the region of transition from the rubberlike to the glassy state or even near Tg. Then the parameter characterizing the low-temperature behaviour of rubbers, KHc
HT HT0 ,
(2.3c)
whereHT and HT0 are the deformations of the specimen under a given loading at the testing temperature T and under conditions of rubberlike deformation (i.e., at T = T0). The parameter KHc thus measured is the lower the higher H is, i.e., the larger the load applied to the specimen is [62, 361, 732]. This effect can be due first and foremost to the dependence of the modulus on the deformation in the region of small deformations. This dependence, namely a decrease of the modulus with the rise of H in the region of small H, was observed when the deformation curve was obtained in the temperature range of the rubberlike-state plateau [49, 50, 59, 60, 364]. The same dependence takes place in the high-temperature part of the transition region [44, 62, 361, 732] and can be due to the breakdown of some physical junctions present in the specimen before the deformation. For filled rubbers, this effect is well known in the region of large deformations (Mullins – Patrikeyev effect [495]; Payne’s effect under dynamic deformation is of the same nature [549, 550]). A detailed experiment to reveal the effect ofH on the recovery in this deformation technique has not been carried out due to the difficulties in applying any considerable deformations at sufficiently low temperatures and low accuracy of measurements at small H. This is the major limitation of this deformation method for assessing the low-temperature behaviour of elastomers. Measurements of KH and K by the former method make it possible to trace the equivalency of the effect of temperature and deformation on the parameters, which characterize the decrease of the rate of the relaxation processes in elastomers at a temperature decrease. Indeed, both at a deformation increase and a temperature rise [177] the parameters, which characterize the low-temperature behaviour, go up. This effect can be used to predict the low-temperature behaviour of elastomeric materials (see further). Shape of specimen. It is known that the change of shape of a specimen leads to a change of shape of the deformation characteristics under compression [97, 161, 550]. “Low” specimens have a greater rigidity under compression than “high” specimens. The shape of the specimen is usually characterized by the shape factor )
S20 S10 ,
(2.4)
where S10 and S20 are the areas of free and loaded surfaces of the non-deformed specimen.
48
CHAPTER 2
Thus, for cylindrical specimens,
)
d 0 2h 0
1 2G ,
(2.4c)
i.e., it is determined by the ratio of diameter d0 to height h0. Figure 2.2 presents a dependence of the mean stress Vkc in the specimen on the deformationH under compression at room temperature; the dependence was plotted by the data of Payne [161]. Here
V kc
P S2 ,
(2.5)
i.e., it is a stress referred to the “true” cross-section area of the compressed specimen, S2, calculated from the condition of the constant volume of the specimen under compression*:
S2
S 20
1 H .
(2.5c)
From the data of Fig. 2.2, it is seen that if the diameter of a specimen is considerably larger than its height, i.e., G > 1, then not only a significantly larger force is required to obtain the same deformation, but also the mean stress in the specimen, Vkc , should be much greater**. An increase of rigidity with G (or )) occurs because in the mid-part of the specimen there are no possibilities for the displacement of the material without the volume change; the material is under conditions of uniform compression, its volume changes and the contribution of the elastic component to its deformation sharply increases. This contribution is the larger, the relatively “lower” the specimen is, i.e., the smaller its free surface is. This increase in the share of elastic deformation in the specimen under compression leads to a change of its low-temperature resistance [113, 129]. Thus, if the low-temperature resistance coefficient KH in compression (determined from eq. (2.3)) is considered as a low-temperature behaviour parameter, then, irrespective of the initial deformation H0, the relative decrease of KH at a temperature decrease is the lower, the higher ) is. This is especially clearly seen at temperatures close to Tg. The same pattern is also observed for recovery K in compression. An improvement of low-temperature resistance, determined under compression, with )rising is due to the fact that the elastic component of the deformation, which goes up with )rising, does not decrease at a temperature decrease, and, therefore, the low-temperature resistance is partially preserved. An increase of ) is a reserve for improving the lowtemperature resistance of rubber goods (see Chapter 6). 2.1.2 Dynamic mechanical properties
Figures 1.7 and 1.8 present the changes of the dynamic characteristics with temperature in the transition region. The effect of frequency changes on Tg, as well as on the position of * Strictly speaking, for specimens with large ) this condition is not fulfilled. ** For specimens with large ) within the limits of attainable values of H the value of S2 does not practically change during the deformation.
MECHANICAL PROPERTIES OF ELASTOMERS NEAR GLASS TRANSITION
σk', MPa
4.0
6
5 4
3
3.0
2
2.0
1
1.0
0
49
0.25
0.50 ε
Figure 2.2 Dependence of the “true” mean stress Vkc on the compression deformation H for cylindrical specimens of vulcanized NR with different shape factors G= 2): 1 (1), 2.2 (2), 4.25 (3), 6.15 (4), 8 (5), 12 (6).
the transition region, was discussed in detail in Chapter 1. As for the static characteristics (see Section 2.1.1), all changes in the polymer structure leading to an increase or decrease of Tg also result in a change in the position of transition region. Thus, a shift of the maximum of tan G towards higher temperatures, as a rule, is unequivocally matched with an increase of Tg, e.g., if copolymers are used [218, 623]. A major part of the formulation recommendations, considered above and leading to changes of the mechanical characteristics obtained in the transition region under static conditions are also valid for the storage moduli (Ecor Gc). Generally speaking, these values are less sensitive to the structure of a rubber or the composition of an elastomeric material than the mechanical losses characterized by the loss moduli (Eccor Gcc) and the loss tangent tan G,
50
CHAPTER 2
as well as the hysteresis during the stretching at a constant rate. Therefore, when discussing below the effect of the formulation and other factors on the dynamic mechanical characteristics, we shall mainly consider the mechanical losses. Research in the recent years has shown that the use of the principle of temperature– time superposition opens the possibilities to characterize adequately the service properties of rubbers intended for operation under some conditions or other. Practically for the first time, this was shown for tribomechanical properties, see [276, 486]. Correlation of the data on tyre wear and the results of rubber tests are achieved only when using the correct temperature– time (or frequency) range in the tests [276]. Other characteristics of tyres also correlate with the dynamic properties of rubber, in particular, with the value of tan G measured in the corresponding frequency – temperature range. Thus, the value of tan G determined at –20 to –10°C and a low (e.g., 5 Hz) frequency, characterizes the road traction of a tyre on ice- and snow-covered roads. The value of tan G determined at 0°C characterizes the wet road traction [574]. This enables studies of some formulation factors or other on the service properties by measuring the value of tan G under respective conditions [574]. Molecular mass. Of all parameters that characterize the microstructure of rubber, the specific effect on the dynamic characteristics in the transition region can be traced only for molecular mass M. Data of this kind were obtained by Vinogradov with coworkers [730] and also by Marei and Sidorovich [443, 644]. The change of mechanical losses at T < Tg with M rising is illustrated by the data of Fig. 1.12b, obtained by recalculations of impact resilience for the value of tan G. It is seen that the position of the maximum does not change at M changing; however, the level of loss in the transition region is the lower the higher M is. The broadening of molecular-mass distribution leads to that of the high-temperature arm of tan G and a shift of the maximum itself towards lower temperatures or higher frequencies. Rubber blends. The transition region for elastomers based on blends of incompatible rubbers has two loss maxima, so their value is smaller and the region where the loss is high is broader. For technical-grade rubbers, the regions of mechanical loss maxima usually overlap, as the presence of a plasticizer and filler leads to a broadening of each maximum, so technical rubbers from rubber blends are characterized by a broad region with ill-defined maxima. The character of the frequency and temperature dependences of tan G changes respectively. However, the ways of directed changes of the level of mechanical loss in the prearranged frequency– temperature range for technical rubbers based on rubber blends have been developed insufficiently. Plasticizers. At the introduction of oils [497] and other plasticizers, the transition region is observed to broaden due to the decrease of Tg; as the result, the position of the loss maximum is shifted to the region of lower temperatures and the maximum becomes broader. However, detailed data on the effect of plasticizers on the character of the frequency and temperature dependence of mechanical losses in the transition region are insufficient in the literature (in contrast with the data on the effect of plasticizers on Tg). Three-dimensional network. The effect of the three-dimensional network on the dynamic properties have been studied in a greater detail [24, 223]. Thus, the studies of model 1,4-cis-polybutadiene vulcanized rubbers have shown that an increase of the network density leads to a decrease of Gcc maximum or tan G [223]. At the same density of crosslinks, this effect is more pronounced for sulphur-containing vulcanizates than for radiationvulcanized rubbers; the maximum is observed to degenerate totally at high degrees of crosslinking. These changes of mechanical losses are related to the degradation of the network of fluctuation entanglements or physical junctions as the result of the formation of chemical crosslinks. A decrease in the number of physical junctions as the result of
MECHANICAL PROPERTIES OF ELASTOMERS NEAR GLASS TRANSITION
51
vulcanization undoubtedly contributes to the decrease of the mechanical loss maximum in the transition region. However, the fact that the maximum in the frequency range studied vanishes is indicative, in our view, of the shift of Tg towards higher temperatures (or lower frequencies), which is supported by the data considered below [287] and also by [551]. The role of the fluctuation network of the physical junctions is seen especially clearly at low densities of the network of the chemical crosslinks [287]. Thus, an additional maximum in the spectrum of delay times L, obtained by the treatment of the data of creep and free torsional vibrations and reflecting the character of mechanical losses, which is close in its shape to the temperature dependence of tan G, was found for NR and ethylene-propylene rubber specimens [287]. An extra maximum in weakly crosslinked elastomers was also found in the transition region in [223]. This maximum can be determined by the relaxation processes related to the rearrangement of the physical junctions, the number of which is the larger, the smaller the density of the chemical-crosslink network is. These physical junctions, occurring in elastomers in the transition region, apparently serve as centres of subsequent crystallization or microcrystallization. It is not to be ruled out that the degeneration of the mechanical loss maximum for 1,4-cis-polybutadiene observed in [223] is related not to the D-maximum, but exactly to this additional maximum determined by the existence of a network of the physical junctions of a fluctuation nature. Fillers. Introduction of active fillers leads to a significant change of mechanical losses [583, 584]. The mechanical properties of a rubber blended with carbon black differ from the properties of a rubber in the matrix. This rubber, “rigid” by the definition of Mullins [495], is characterized, as shown by Bartenev, by the presence of an additional mechanical loss maximum, which lies at higher temperatures than the principal D-maximum [44, 48]. Herewith, the height of the principal maximum is decreased, it is blurred, so that the level of mechanical losses in the transition region slightly decreases [223]. Variations in compliance during the changes of activity and content of a filler are shown in [551]. Changes of the dynamic properties of elastomers in the presence of fillers are the larger the greater the amplitude of deformation is, which is due to the nonlinearity of the relaxation properties of filled elastomers [241]. The effect of a dynamic modulus decrease for filled elastomers with the deformation amplitude rising (the Payne–Warnaka effect) is well known in the range of temperatures and frequencies corresponding to the rubberlike plateau [198, 205, 549, 736]. The data on the broadening of the mechanical loss maximum with the content and activity of a filler rising are well consistent with the data on recovery (the decrease of T with the rise of filler’s activity) and can be considered to be the consequence of the transition of the polymer layers occurring on the surface of the filler particles to the “pseudoglassy” state or (which can be equivalent) the result of an intensive formation of physical junctions of a fluctuation nature on filler’s particles. Other factors. The effect of pressure on the dynamic mechanical characteristics in the transition region appears not to have been studied. As for the effect of deformation or, to be more exact, the amplitude of deformation, this effect, as we mentioned above, is well known for filled rubbers (the Payne– Warnaka effect) in the range of room temperatures and average frequencies; in the transition region the storage modulus is the smaller and the mechanical loss is the larger, the greater the deformation amplitude is, i.e., the nonlinearity of the dynamic properties of filled rubbers is preserved. Near the glass-transition temperature in the transition region the Payne– Warnaka effect was also observed for unfilled rubbers, which can be due to the breakdown of the fluctuation junctions [543]. The effect of the shape of a specimen on the dynamic characteristics in the transition region was not studied systematically. When measuring the dynamic mechanical
52
CHAPTER 2
characteristics of elastomeric materials in the transition region, it is necessary either to carry out the measurements at the values of ) characteristic of rubber articles, or to learn how to predict the effect of). 2.1.3 Prediction of the mechanical properties of elastomers in the transition region from the glassy to the rubberlike state Use of the principle of temperature– time superposition. In Chapter 1, when considering the relaxation character of glass transition, we formulated the principle of temperature– time superposition (TTS) and presented the temperature dependence of the horizontal-shift factor aT for NR (see Fig. 1.9), described by the Williams–Landel–Ferry (WLF) equation (eq. (1.7)). As applied to elastomers, the TTS principle was used successfully to describe the dynamic characteristics of NR [161] and a number of technical-grade rubbers based on NR and other rubbers [24, 223, 497]. Applicability of the TTS and the WLF equation, and the possibility of plotting a master curve for Gc and Gcc (or tan G ) for unfilled rubbers from NR in the absence of crystallization was clearly shown by Payne [548]. Applicability of the TTS for technical rubbers of not too high hardness (less than 50) is shown in [223]. Techniques of accounting for rubber density changes with temperature [223, 610], as well as the means of passing from the creep data (at room and higher temperatures) to dynamic characteristics were also developed. However, the nonlinearity of the dynamic characteristics of filled rubbers, i.e., their dependence on deformation, renders difficult the use of the TTS technique, which was developed for linear viscoelastic processes. These limitations pertain to the possibility of passing from one viscoelastic function to another, and from one deformation to another. Studies of filled elastomers have shown that superposition of the temperature dependences of Gc and tan G even at relatively small deformations requires not only horizontal but also vertical shift to be available (the value of the vertical shift depending on the temperature and type of carbon black and not related to the density changes). Applicability of the TTS principle was also shown for oil-containing elastomers [497]. It proved possible to perform superposition with respect to oil concentrations, too. However, as for filled elastomers, attempts to obtain superposed curves failed during the changes of the deformation amplitude. The nonlinearity is not an obstacle for using the TTS and WLF equation for passing from the frequency dependences obtained for Gc and tan G at one temperature, to the data for the same characteristics at another temperature and the same deformation using the master curves of the shape presented in Fig. 1.8. This also follows from the data of [223]: in the absence of a plasticizer, eq. (1.7) with universal constants is valid for filled rubbers from NR, and the values of Gc can be assessed from the knowledge of hardness H of the rubbers. The TTS principle can also be used to predict the relaxation properties in the transition region under static conditions [117, 125]. Thus, the direct measurement of such a relaxation parameter as recovery within a broad temperature range is not associated with any problems. In the absence of crystallization, the change of the time of isothermal holding of rubbers in the transition region does not change the results obtained. However, an experimental change of the recovery time t2 is made difficult, and for very small times is not possible at all. At the same time, when using the recovery to characterize the low-temperature resistance of rubber articles, the value of recovery should be assessed at various, often very small t2 (see Chapter 6). In this case, prediction using the TTS principle seems to be the only possibility to obtain reliable results.
MECHANICAL PROPERTIES OF ELASTOMERS NEAR GLASS TRANSITION -4
-5
-3
-2
-1
53
0 log (t/aT)
a
-0.5 3 4
2
-1.0
1
log K
log aT b
5.0 1 2
4.5
4.0 3
3.5
3.0
+ 4
2.5
2.0
+
1.5
+
1.0
+
0.5
-60
-40
-20
0
Figure 2.3 Master curves of log K vs. log (t/aT) (a) and the dependence of log aT on temperature T (b) for unfilled NRbased rubber (1) and filled butadiene acrylonitrile-based rubbers (2–4); the reference temperatures T0 = –25°C (1) and 0°C T, °C (2–4).
The applicability of the TTS to the dependences of recovery on recovery time t2, obtained at various temperatures, was shown for unfilled rubber from NR under conditions, which exclude crystallization [114]. Their superposition for the reference temperature T0 = – 25°C (Fig. 2.3a, curve 1) gives a master curve, using which we can calculate the recovery values at any T and t2. The dependence of the horizontal-shift factor aT used to plot this
54
CHAPTER 2
master curve on temperature coincides with the dependence obtained by Payne for the dynamic characteristics (see Fig. 1.9 and Fig. 2.3, curve 1), i.e., is described by the WLF equation (1.7) with standard constants. The same values of the constants are obtained for the same rubber from the data on the free volume fgG at Tg calculated from eq. (1.15). The temperature dependences of log aT for butadiene acrylonitrile-based rubbers of various compositions are presented in Fig. 2.3b (curves 2 – 4). The dependences of log aT vs. T and log K vs. log (t/aT) are of the same character* as for NR-based rubbers (T0 = 0°C was taken as the reference temperature). However, the change of composition of the rubbers from nitrile rubbers, in particular, the change of type and content of plasticizers, leads to a shift of the dependence of log aT on T (see Fig. 2.3b, curves 2 – 4). The master curves for butadiene acrylonitrile-based rubbers given in Fig. 2.3a can be used to predict the low-temperature behaviour of these rubbers in various regimes. Note that the values of the constants c1g and c2g in eq. (1.7) and also fgG, all calculated from the temperature dependence of aT, differ from the universal values but coincide with those obtained from independent experiments for measuring the linear-expansion coefficients D above and below Tg. The possibility of using the TTS to describe the recovery of styrene-butadiene thermoelastoplasts in the transition region [122] was also shown. Thus, the TTS principle can be used in predicting the relaxation characteristics obtained both under dynamic and static conditions, even at a nonlinearity of the viscoelastic properties but provided that the prediction is made within one type of function and for one type and degree of deformation. The passage from one type of function to another is possible only if one type of relaxation processes occurs in the temperature and time (or frequency) range considered, in particular, the D-relaxation process. Application of temperature– deformation superposition principle. When considering the effect of deformation on low-temperature resistance, we note the equivalency of the effect of deformation and temperature: the low-temperature resistance increases with the deformation going up, the same way as with the temperature increasing. This equivalency was shown in [7, 8, 44, 101, 281]. Thus, similar to the TTS principle, the principle of temperature– deformation superposition (TDS) can be proposed [101]. In essence, the effect of deformation on low-temperature behaviour is a manifestation of the nonlinearity of the relaxation processes in polymers; however, the theories of the nonlinear behaviour of polymers and, in particular, elastomers, have been developed to date insufficiently. Therefore, the TDS principle can be considered as an attempt to describe the nonlinear processes using techniques well developed for linear physical relaxation processes, i.e., to use the techniques available for the TTS to describe the nonlinear behaviour of elastomers at low temperatures. Similar to how it is made using the TTS, the deformation dependence of the lowtemperature resistance parameters K and KH or of another relaxation function obtained at one temperature can be referred to any other temperature in the temperature range, where the same physical relaxation processes develop, by shifting along the log H axis (whereHis deformation). The temperature dependence of the log aT value to which this shift should be made can be described by the equation similar to the WLF equation (1.7) used to describe the temperature dependence of log aT in the case of the TTS: * In passing from one temperature to another, besides the shift along the time axis, we should take into account the change of density with temperature, which change is given by the coefficient U1T1/U 2T2. In the work considered, it was taken to equal unity.
MECHANICAL PROPERTIES OF ELASTOMERS NEAR GLASS TRANSITION
log aTc
c
c1g T Tg
c
gc 2
T Tg ,
55
(2.6)
where c1gc and c2gc are empirical constants that differ little from the values c1g = 17.44 and c2g = 51.6 K used as constants for most polymers [198]. Considering that the values c1g and c2g are related to the specific free volume in glass transition by relations (1.19 ), and extending the analogy with the TTS, the same relation was suggested to exist for c1gc and c2gc . Correspondingly, the values fgPh and fgG calculated from the TDS for a number of elastomers lie within the limits of 0.022– 0.029 [101], i.e., the deviations from the universal values do not exceed ±13%, as for the values calculated for the same rubbers directly from the volume changes at Tg [544]. A satisfactory coincidence of the free-volume values, obtained from the temperature– deformation dependences of K and KH, with the universal values of these magnitudes, is one of the criteria of the applicability of the superposition method for describing the temperature– deformation dependences of the low-temperature resistance parameters. Using the temperature– deformation superposition principle similar to the principle of temperature– time superposition can help avoid the problems that occur when describing the nonlinear behaviour of deformed elastomers at low temperatures and offers a method for predicting their low-temperature resistance depending on deformation. However, considering the effects of nonlinearity, which are to be avoided by using the TDS principle, the possibilities of its use for passaging from one type of viscoelastic functions to another seem to be rather limited.
2.2
Mechanical properties at temperatures below glass-transition temperature
The glass-transition temperature Tg, which separates the glassy state of a polymer and the region of transition from the glassy to the rubberlike state, is usually the low temperature boundary of the development of rubberlike deformation. However, under conditions of a stress, large deformations in polymers can also develop at below Tg. Herewith, the kinetic hindrances of the deformation are eliminated and, in contrast with the case of the deformation of specimens with a large shape factor, the deformation of an entropic nature develops. This deformation was called “stress-induced deformation” and the phenomenon, “stressinduced elasticity” (the term was proposed by Aleksandrov [7]). This phenomenon, first studied in detail by Lazurkin and Fogelson [396], consists in the development of large deformations in a polymer specimen at an application of sufficiently large stresses (elongation or compression) at temperatures lower than Tg. The major patterns of the behaviour of elastomers at temperatures below Tg [57, 397] are the same as for other polymers. In the recent years, the discussion of the nature of stress-induced or “forced” elasticity of polymers has been re-opened [356, 401, 756]. Figure 2.4 presents the deformation curves for unfilled rubber from styrene-butadiene rubber SKS-30 at temperatures T < Tg [57]. In the initial section, the relation between stress V and deformation H is linear at all T < Tg. In this temperature region, the elasticity modulus E | 103 MPa, i.e., is of the order characteristic of glassy polymers. Upon reaching a value V = VL (the stress-induced elasticity limit), which is the smaller, the closer the experimental temperature to Tg is, the values of V pass through a maximum, and then a region is observed, where the deformation increases almost without any change of stress. This section of the curve corresponds to the formation of a “neck”, when the central part of the specimen is
56
CHAPTER 2
σ, MPa 10
80
9
8 7
60
6 40
5
4 2
20
3 1
0
4
8
12
16 ε . 102
Figure 2.4 Dependence of stress V in a specimen on tensile deformation H for vulcanizates based on styrene-butadiene rubber SKS-30 at temperatures T, °C: –53.0 (1), –55.7 (2), –72.0 (3), –88.5 (4), –96.0 (5), –116.5 (6), –138.5 (7), –141.0 (8), –175.6 (9), –196.0 (10). σbr; σL, MPa
100
1 2
80 60
3
40 20 0
-180
-140
-100
-60 T, °C
Figure 2.5 Dependence of stress in brittle fracture Vbr (z, , S) and limit of stress-induced (or forced) elasticityVL ({,
, U) on temperature T for vulcanizates based on styrene-butadiene rubber SKS-30, filled with carbon black (1), unfilled (2) and filled with chalk (3).
deformed; by the end of the section, the “neck” spreads on to the entire specimen. Stressinduced deformation is reversible by its nature, but the specimen recovered only when heated up to above Tg. This effect is used for developing thermosetting materials. As it follows from the data of Fig. 2.4, the section of the deformation curve corresponding to the development of stress-induced deformation is observed only at temperatures not too far away from Tg. The temperature range 'TL, where stress-induced elasticity develops, is limited from the above by the glass-transition temperature Tg, and from below by a temperature, which is interpreted as the “ultimate” brittle temperature Tbr. At T < Tbr, the curves of V vs. H have only a linear section and the brittle fracture begins at V = Vbr before the stress-induced elasticity develops. Thus, the static or “ultimate” brittle tempera0 ture Tbr can be taken to be a temperature, at which for a specimen deformed at a small rate the stress is observed to diminish with the deformation rising, i.e., the limit of stress-induced
MECHANICAL PROPERTIES OF ELASTOMERS NEAR GLASS TRANSITION
57
elasticity VL emerges. The closer to Tg, the smaller VL is, and the larger the rupture elongations are. The dependence of Vbr (below Tbr0) and VL on the testing temperature for rubbers from SKS-30 is presented in Fig. 2.5. The horizontal section of the curve corresponds to brittle fracture and shows an independence of Vbr on temperature. Thus, the region of the glassy state consists of two parts: the region where the brittle fracture of a polymer occurs, and the region of stress-induced elasticity 'TL, where amorphous glassy polymers are used the most frequently. It is reputed that the major requirement to structural plastics from the class of amorphous polymers is that their operational temperature range, which includes room temperature, be the region of stress-induced elasticity. The less flexible the polymer molecules, i.e., the higher Tg, the broader the region of stress-induced elasticity 'TL is. An increase of the rate of applying a load leads to a rise of the temperature, at which a brittle fracture of the polymer occurs. This temperature (Tbr c ) can go up to such an extent so as to approach the glass-transition temperature Tg measured under static conditions [780, 781], and sometimes even to exceed it, however, still remaining always lower than Tg measured at the same loading rate. The brittle fracture of polymers is accompanied by the breakdown of chemical bonds [598]. This is also observed for elastomers [320, 388]. This means that the breakdown mechanism of elastomers in the rubberlike state, accompanied with their deformation and orientation, differs from the mechanism of their breakdown in the brittle state. Zuev et al. [781, 782] have shown that in the case of a limited (hindered) deformation of elastomers this mechanism approaches the brittle fracture mechanism. The brittleness temperature measured under standard loading conditions is widely used to characterize the low-temperature behaviour of elastomers. However, we should bear in mind that this is a conditional value, which depends on the velocity of applying the load and the strength properties of an elastomeric material in the glassy state. Therefore, the ratio between Tg and Tbr c can be most diverse depending on both the strength properties and on how much the load-application rate shifts Tg and on the width of the stress-induced (forced) elasticity range 'TL. The available scarce data give grounds to believe that the possibilities of using elastomers at T < Tg are related to the value of 'TL, which, as for other polymers, is the greater, the higher Tg is. However, the behaviour of elastomers in the stress-induced elasticity region and the relation of the width of this region to the structure of an elastomer and its properties in the transition region have not been studied at all. In the meantime, prediction of the properties of elastomers at below Tg by their properties in the transition region becomes all the more topical problem. There is a point of view, according to which the effect of the shape of a specimen on its low-temperature behaviour under the compression deformation is also related to the development of stress-induced elasticity under the action of large stresses characteristic of the deformation of specimens with the large shape factor ). However, the relative role of this effect as compared with the preservation of the elastic component of deformation at a temperature decrease has not been studied. Another interesting problem is to elucidate the nature of the relaxation processes in the region of stress-induced elasticity. The occurrence of such transitions in a number of elastomers is registered by the mechanical methods [44, 489] and by the specific heat changes [78]. The most reliable registration is by the RTL (for nonpolar elastomers) and dielectric loss (for polar elastomers) methods. Bershtein et al. [78, 79, 81] believe that E-transition is the transition related to the emergence or loss of segmental mobility, i.e., true glass
58
CHAPTER 2
transition. And the transition at Tg (D-transition) is due to the emergence or freezing of the 0 – the lower cooperative motion of segments. There is also a view [79] that the value of Tbr temperature range of the stress-induced elasticity region – coincides with the temperature of the E-transition, and this temperature, in turn, coincides with the second-order phasetransition temperature T2. Then the stress-induced elasticity region is the region in which the transition kinetics is predominant, and only the application of a stress reveals the possibilities for the development of a rubberlike deformation. Here, as in the consideration of the effect of pressure (see Chapter 1), the use of the formalism of the description of secondorder phase transitions proves fruitful*. The complexity of the problem is, however, that the experimentally measured Tg increases at the application of pressure and decreases at the application of stress. Based on similar considerations, some authors refute the idea of the existence, below the measured Tg, of the kinetically complicated second-order phase transition as well as the identity of three temperatures: the temperatures T2, the E-transition and 0 (see also Chapter 1). the lower boundary of the stress-induced elasticity region Tbr All these considerations, both theoretical and practical, make one think that studies of the behaviour of elastomeric materials near and below Tg is one of the most important problems of the physics of elastomers.
* An analogy with crystallization can be traced here. Thus, the experimentally measured effect of a mechanical stress and uniform pressure on the equilibrium melting temperature of polymers, in particular, elastomers, is described by the Clapeyron–Clausius equation obtained for first-order phase transitions. This is related to the fact that the melting temperature measured under the action of a mechanical field is closer to the equilibrium temperature and at high pressures and stresses even coincides with it, as the mechanical field decreases the kinetic hindrance of the transition (see Chapter 3).
3
Crystallization of Elastomers at Low Temperatures
For stereoregular elastomers, changes of properties at low temperatures can be due not only to glass transition, but to yet another physical process – crystallization. The rubbers capable of crystallization include natural and synthetic 1,4-cis-polyisoprene, 1,4-cis-polybutadiene, polychloroprene, butyl rubber, most siloxanes, some types of polyurethanes (Table 3.1). Technical rubbers from these polymers crystallize, too. In crystallization, their elastic properties deteriorate at low temperatures, but it is stress-induced crystallization that determines the strength properties of elastomeric materials based on crystallizable rubbers. This is one of the reasons for a widespread use of stereoregular elastomers in rubber industry. Another reason is low glass-transition temperatures for most of them, which is due to the high flexibility of macromolecular chains. This makes possible the development of rubbers with good short-term low-temperature resistance. A natural consequence of decreasing Tg is an increased capability of crystallization, which, in turn, deteriorates low-temperature resistance. The contribution of crystallization and glass transition to low-temperature resistance will be considered in Chapter 6. Table 3.1 Crystallization parameters of some elastomers: maximum crystallization rate temperature T1, crystallographic system and parameters of the unit cell, density Uof completely crystallized Uc and amorphous Ua material, melting heat 'H. Elastomer
1,4-cis-Polyisoprene 1,4-cis-Polybutadiene Polymethyl vinyl siloxane Polychloroprene Butyl rubber
T1, °C
Crystal system
–25 –55 –80 –10 –40
monoclinic monoclinic monoclinic orthorhombic orthorhombic
Unit cell parameters a, Å
b, Å
12.46 8.89 4.60 9.50 13.0 8.3 9.0 8.23 6.24 11.96
c, Å
J, °
8.10 92 8.60 100 7.75 60 4.79 – 18.63 –
U, kg/m3 Uc
'H, J/g3
Ua
1000 906 56.7 1010 |00 85 1070 980 63 1657 1229 104 937 913 208
The crystallization patterns are the same for all polymers and have much in common with the crystallization features of low-molecular substances. A combination of three factors makes elastomers distinctly different from other crystallizable polymers: high flexibility of macromolecular chains (low Tg), comparatively low melting temperatures Tm and high viscosity in the amorphous rubberlike state. These features taken together (and
60
CHAPTER 3
characteristic of different crystallizable elastomers to different degrees) is what determines the crystallization specifics of elastomers. Its manifests itself in the lower temperature region, lower rate and degree of crystallization as compared with other polymers* [95]. Therefore, a change of the mechanical properties of elastomers in crystallization has its specific features, too. Crystallization of elastomers develops in storage or in operation, as, unlike other polymers, under usual conditions elastomers are melts in the rubberlike state. Crystallization of highly crystalline polymers, which have high Tg, occurs in their processing from the melt, which is not in the rubberlike state but in a viscoelastic state; and it does not deteriorate the entire body of their operational properties but forms them. Therefore, the process techniques of fabricating rubber materials and articles therefrom aim at developing the optimal crystallization conditions. However, the views on the nature of crystallization, developed for highly crystalline polymers [135, 435, 436a, 751, 751a,b,c] can be successfully used to describe the crystallization of elastomers.
3.1
Methods of Crystallization Studies
As in the determination of glass-transition temperature, the basis of the methods of crystallization studies are changes of elastomers’ physical and mechanical properties in their transition from the amorphous to the partially crystalline state. The major distinctions between crystallization and glass transition are that, on the one hand, regions of far order are formed in crystallization, and, on the other hand, all changes related to crystallization develop in time. These distinctions determine the specific features of the physical and mechanical methods used for crystallization studies and the importance of the structural methods of studies in this case. Methods of crystallization studies of elastomers have been considered in detail in [95]. Here we would only briefly dwell on the difference between the use of the structural/physical and mechanical methods in crystallization and glass-transition studies. 3.1.1 Structural and other physical methods X-ray diffraction. Diffraction of X-rays is the most widespread method of studying the crystallization of both low-molecular substances and polymers. However, the small size of crystals, low degree of crystallization and its slow development make the use of X-ray diffraction for low-temperature crystallization studies rather labour-intensive. Special care should be taken when interpreting the obtained data; methods of interpreting differ for different elastomeric materials [26, 95, 413, 467, 492]. Recall that the minimum degree of crystallinity registered by the X-ray diffraction method is 3–5%, the minimum size of crystals is 10–100 Å, and the minimum exposure required to obtain an X-ray pattern is 3–5 min. The main problem being solved by the X-ray diffraction method as applied to elastomers – the determination of lattice parameters (by the X-ray patterns of oriented specimens) – has already been solved for most elastomers (see Table 3.1) [30, 135, 476, 492]); the data on the structure of oriented elastomers obtained by this method will be considered in Chapter 4. New results on the stress-induced crystallization of elastomers were obtained using synchrotron X-ray radiation [711], see Chapter 4.
* Such polymers as polyvinyl chloride and polysuccinates are close to elastomers by the degree of crystallization and its specific features.
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
61
The main results on elastomers were obtained by large-angle X-ray scattering. Smallangle scattering (SAXS) measurements, successfully used in crystallization studies of other polymers [240, 597, 605, 720] have shown that for such elastomers as polyisoprene and polybutadiene the scattering intensity changes observed in orientation are due not to stressinduced crystallization but to the change of the character of distribution of stearic acid or other substances present in the vulcanizing group [180, 724]. The absence of SAXS changes for elastomers is due not only to the low degree of crystallinity. Indeed, the degree of crystallinity of polysiloxanes is sufficiently high (up to 60%), but no SAXS changes for them have been observed yet.* A significant role is, evidently, also played by the difference in the densities of the amorphous and crystalline phases. Of all elastomers, it is maximum for polychloroprene (see Table 3.1), and this is the elastomer, for which SAXS changes in crystallization have been observed [173, 602, 603]. Infrared (IR) spectroscopy. Changes of IR spectra in crystallization and melting are well known [338, 502, 503, 505, 625]. This method is the most fruitful in studies of stressinduced crystallization. Unfortunately, the difficulties associated with low-temperature crystallization studies restrict its wide use for characterizing structural changes in crystallization of elastomers. Electron and light microscopy, electron diffraction, light scattering, birefringence. No new data on crystals in elastomers obtained by the method of electron microscopy have appeared in the recent years. It appears that the most interesting results obtained by Andrews using OsO4 pertain to 1960–1970s [12, 18]; attempts to get more detailed information, in particular, on the structure of crystals in crosslinked and filled elastomers, have failed due to the not-yet-overcome specimen-preparation problems at low temperatures [12, 18, 95, 96]. All above-said is totally applicable to the data obtained by electron diffraction methods. Using light microscopy, the morphology of polycrystals in elastomers was studied, especially in those that crystallize at room temperature: polychloroprene and polyurethane as well as polysiloxane [95, 96, 102, 628, 785, 786, 787]. The use of lasers as light sources made it possible to extend information provided by the light scattering method [353, 354]; however, the low-temperature measurement techniques are imperfect. Therefore, the most interesting results have been obtained for polychloroprene, which crystallizes at room temperature [713, 714]. Measurements of the refractive index n and birefringence index 'n for nonoriented polymers have not been refined after the first classical works by Smith and Saylor for NR [663, 664] and Trapeznikova and Novikova for polychloroprene [713, 714]; the birefringence indexes for other elastomers in the crystalline state have never been measured. After the first works by Treloar [715], studies of birefringence changes in orientation have been refined [163, 164, 582, 585, 695, 764]; a laser was used as a light source. Recall that the account of the changes in the refractive index n in crystallization is required, when crystallizable elastomers are used as a material in studies of a stressed state of rubber goods by the photoelasticity method [197]. Nuclear magnetic resonance. This method enables studies of not only the crystallization and melting kinetics (which, in fact, could be less labour-intensive using more traditional thermal methods). It also makes it possible to single out the changes due to crystallization proper and to its effect on that part of material, which is in the rubberlike oooooooooooooooooooo * Probably, this is due to the difficulties of studies in the temperature range from –50 to –80°C, where these elastomers crystallize the most intensively even in the oriented state.
62
CHAPTER 3
state. Thus, the data obtained by the pulsed NMR method for polychloroprene [100, 213] show that the molecular mobility of the rubberlike component of the elastomer is decreased in the process of crystallization, which manifests itself in a change of the shape of free induction decay (FID). Chromatography. As in the case of glass transition, the change of the diffusion characteristics of a polymer in crystallization makes it possible to use this method for crystallization studies. However, as applied to elastomers at low temperature, it is almost not used. Electron paramagnetic resonance can also be used for crystallization and melting studies. However, this kind of research is not yet widespread. Radiothermoluminescence (RTL). The RTL method, used very efficiently for glasstransition studies, has not found wide application in crystallization studies due to the unequivocal data interpretation. However, the use of this method for polychloroprene enabled tracing the changes of the properties of the amorphous component in crystallization [72, 376], similar to the way this is done by the pulsed NMR method; still, the accuracy of measurements is much lower in this case. Dilatometry. Studies of crystallization-accompanying volume decrease, which is up to 2–2.5% for elastomers, is the classical method [68, 69, 452]. The use of volume dilatometry is considered in detail in [95]. Unfortunately, no automated volume-measurement systems of satisfactory quality, operated at low temperatures, have been developed in the recent years, so this method is not widespread. The use of linear dilatometry does not always give correct results due to the anisotropy of volume changes. Calorimetry. This method or its variety, differential scanning calorimetry (DSC), gains ground. The method makes it possible to register with sufficient accuracy both the evolution of heat in crystallization, and its absorption in melting by the emergence of, respectively, exothermal and endothermal peaks on the thermograms. Experimental data for heat flow changes, dQ/dt, versus scanning temperature, Tc, can be recalculated to obtain absolute values of specific heat [79, 246, 247, 249, 462, 749, 752–754]. This method is especially widespread in melting studies. It is also efficient in the case of rapid isothermal crystallization. Close information is also provided for by the method of highly sensitive differential thermal analysis (DTA) [246, 247, 249]. The use of calorimetry for stress-induced crystallization studies is also rather promising [246, 247, 248, 249, 253, 264, 265–268, 269, 332, 334, 618]. Other physical methods. Thermal conductivity changes can also be used for crystallization studies. In this case one should take into account the increase of thermal conductivity along the direction of stretching during the orientation, i.e., its anisotropy. Electrical conductivity determined by ionic conduction decreases with the degree of crystallization going up. However, no systematic measurements of this kind for elastomers have been carried out, and no respective methods of crystallization studies have been worked out. Methods based on the changes of dielectric loss tan Ge and dielectric permeability H care used for crystallization studies of polar polychloroprene [35, 539], as well as polysiloxane [404] and polyisoprene [158]. The method of determining the degree of crystallization based on sound-speed measurements [554] can also be attributed to physical methods. The speed of ultrasound waves cu in an amorphous cua and a crystalline cuc polymer is different, as their densities differ. For elastomers, this method has found a restricted application yet [84], owing both to the small difference of the densities of the crystalline and amorphous phases, and to the relatively low degree of crystallization of elastomers. However, as it is highly sensitive, it can be used in the cases when other methods prove inapplicable. But, as in the case of
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
63
thermal and electrical conductivity measurements, correct determination of cu requires a good contact of a specimen with the emitter and receiver. 3.1.2 Mechanical methods The complete body of mechanical methods considered in Chapter 1 with respect to glasstransition studies are used for crystallization studies. Static methods, described in detail in [95, 97], have been developed the most. First and foremost, this can be said of the measurements of the compression/tension modulus at small and large deformations [384, 400, 447, 686], as well as hardness measurements. The hardness-measurement method for assessing elastomer’s capability of crystallization is the basis of the ISO standard 3387 [678]. The crystallization process is rather successfully characterized by the use of such relaxation characteristics as stress relaxation [230–235, 239, 661, 747] and recovery, i.e., the ability of rubbers deformed in the region of the rubberlike plateau to recover their size at low temperatures after the load is removed [95, 97, 441, 593] (see Chapter 1). The method of recovery in compression, widely used for crystallization kinetics measurements, is the basis of the ISO standard 6471 and the Russian state standard GOST 13270-85 [508, 679]. Works of the recent years confirmed the high sensitivity of these methods (along with thermal methods) to crystallization of elastomers.* The method of impact resilience [386, 450] also found its use in crystallization studies. The method of determining the degree of crystallization based on sound-speed measurements [84] (see above) is also attributed by the authors to the relaxation methods. Extensive use of mechanical methods for crystallization kinetics studies is characteristic of namely crystallization studies of elastomers. All these characteristics can be readily measured in their rubberlike melts. In conclusion, we note that in a crystallized specimen all effects due to the transition from the glassy to the rubberlike state (specific heat jump, dielectric and mechanical loss maxima, etc.) are attenuated. Given the absence of the effect of crystallization on glass transition, this can also serve a source of information on the crystallization of elastomers [754, 756].
3.2
Modern views of crystallization
Crystallization of elastomers is a first-order phase transition and has much in common with the formation and melting of common molecular crystals. Transition from the crystalline to the amorphous state is characterized by the equilibrium melting temperature Tm0 . A thermodynamic condition of the transition is the equality of the Gibbs energies of the crystalline and amorphous phases achieved under isobaric conditions at the melting temperature. During the melting, there occurs a jumplike change of the first derivatives of the isobaric potential with respect to temperature and pressure, i.e., enthalpy H, entropy S and volume V, and the second derivatives (e.g., the temperature coefficients of expansion Eand D and 0 meets the condition thermal capacity cp) turn to infinity. The value Tm
Tm0
'H / 'S ,
(3.1)
* Criticisms of the recovery method in [91] and [676] appear to be inconsistent and are, in our mind, caused by the incorrect choice of measuring instruments and methods (see also Chapter 1).
64
CHAPTER 3
where 'H and 'S are enthalpy and entropy changes during the melting.* The elementary cells of all crystallized elastomers belong to known crystallographic types (Table 3.1) [95, 228, 476, 751]. The crystallographic axis c in the elementary cell of a polymer usually coincides with the direction along the molecular chain. At the same time, crystallization of polymers has its specific features. The major peculiarity of the crystalline state of polymers is a large content of amorphous phase; in the case of elastomers it can be 70–80%. In the case of microcrystallization (see further), the content of crystalline phase can be a mere 3–5%. Crystallization of all substances occurs under supercooling conditions, i.e., at temperatures lower than the equilibrium melting temperature Tm0 . However, for polymers the supercooling interval 'T
Tm0 T ,
(3.2)
where much broader crystallization can be observed than for low-molecular-mass substances. It is especially broad for elastomers; 'T is tens and hundreds of degrees. This is the result of the low crystallization rate of elastomers; this makes its kinetics studies especially important. 3.2.1 Morphology of crystals in elastomers
The morphology of crystals in polymers depends on the supercooling of the crystallizing melt. At low rates of supercooling, crystallization of polymers occurs under conditions maximally close to the equilibrium conditions. Herewith, the most perfect crystals are formed. For polymers, these are extended-chain crystals characterized by a minimum free energy. The emergence of these crystals, evidently, explains the rapid increase of the large period registered in polyethylene by the small-angle X-ray scattering at temperatures close to Tm0 [323]. During the melting of these crystals, the most sharp increase of the volume and a narrow melting interval are observed. There were reports on the production of single crystals with extended chains at a long (several weeks) holding of polyethylene at a temperature lower than Tm0 by 1°C; under isothermal crystallization near Tm0 such crystals were observed in polyethylene oxide. Extended-chain crystals can also form in crystallization in the process of polymerization [323, 751]. Formation of such crystals at a low-temperature (–180°C) polymerization was observed for polychloroprene. Under close-to-equilibrium conditions, single crystals with extended chains should form in elastomers, too, but the crystallization rate near Tm0 is rather small for the predominant majority of elastomers. However, crystallization of NR for two years at room temperature yielded crystals whose melting temperature practically coincided with Tm0. For trans-polymers with higher Tm0 (polychloroprene, trans-polybutadiene, gutta percha etc.) the crystallization rate is comparable with the rate of chemical changes (chemical ageing). The most widespread way of producing extended-chain crystals is crystallization of a polymer under pressure. This technique has been studied in detail for 1,4-cis-polyisoprene [158, 185–187, 569, 570, 571, 750]. In stretching, which sharply accelerates crystallization near Tm0 similar to pressure, a certain proportion of extended-chain crystals are formed; they are located in axial fibrillae (see Chapter 4). * Usually for elastomers 'H is smaller and 'S larger than for other polymers, so Tm0 is lower (see Table 3.1).
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
65
Crystals with extended chains are the only type of crystal in polymers, where defects characteristic of all other polymer crystals can practically completely be absent in principle, and the amorphous component is completely absent [228, 323, 435, 569, 636, 751]. As the supercooling 'T is increased, the size of crystals along the axis c decreases and can be 1–2 orders of magnitude lower than the length of the molecular chain (which is of the order of 104 Å). Polymers crystallized under supercooling conditions are characterized not only by the small size of crystalline regions, but also their imperfection, i.e., deviations from the correct three-dimensional order inside the crystals. It should be noted that clearcut techniques for separating the effects associated with the small size of crystals and their imperfection do not exist. Polymer single crystals produced from solution have a lamellar structure and consist of separate laminate formations, lamellae [228, 323, 435, 569, 636, 751]. Single crystals of n-paraffins obtained under similar conditions have the same shape [135]. The thickness of polymer lamellae (l |102 Å) corresponds to the thickness of n-paraffin crystals; molecular chains in such crystals fold to form a defective end surface. Crystals with folded chains also form during the crystallization from melt [82]. For elastomers, their presence is reliably observed in thin-film crystallization from solution. Due to a high viscosity of the melt, no folding of chains seems to occur during the increase of supercooling. In this case, ordered fragments of macromolecular chains, different parts of which can be in various crystalline regions, gather in crystalline regions. This model of crystallized polymer with a large number of tie chains corresponds to the model developed well back in 1930s and called the “micellar” model. For elastomers, which have a high viscosity of the melt, this is, probably, the most widespread type of crystal. This model is also, apparently, the most adequate to the structure of crystals formed in block copolymers. Individual single crystals are usually constituents of more complex crystalline formations. By analogy with low-molecular-mass substances, such formations can be called polycrystals. A distinctive feature of polycrystals in polymers is the large portion of amorphous material in them, which can be considered as their defect part. It is especially large for elastomers; it is here that the major part of amorphous material is localized. Typical polymer polycrystals are spherulites. At large supercoolings, the major type of polycrystals in elastomers are grains, which represent centres of undeveloped spherulites. Polycrystals of elastomers can be observed using a light microscope [95, 96, 424]. Classical data of the morphology of elastomers (NR and polychloroprene) were obtained by Andrews using an electron microscope (during the treatment of specimens with OsO4) [12]. 3.2.2 Kinetics of isothermal crystallization
Figure 3.1 presents the kinetic curves for isothermal crystallization of some elastomers; the curves were obtained by different methods. The initial fragment of the kinetic curve, where no crystallization-induced changes are observed, characterizes the induction period W0. The value of W0 is related not only to the character of nucleus formation in the system studied, but also depends on the sensitivity of the method. The second fragment reflects an intensive increase of the amount of crystalline material, accompanied by a drastic change of all physical properties of the elastomer. The final fragment of the curve characterizes the processes of secondary crystallization. The crystallization rate of polymers is determined either by the tangent of slope of the curve at the second fragment or by the crystallization half-time W1/2. The higher the
CHAPTER 3
66
a
1 - C/C ∞ a I
III
II
1.0 0.5
b C/C∞ 1.00
τ1/2
τ0
0
100
213
200
300 t, min
204
208 218
223
225
228 200
0.80
I
0.60 0.40 0.20 0.00
5 0 C/C∞ 218 232 240 1
10 244
15
20
25 248
30
35 250
40
45 t, min II
0.8
253
0.6 0.4 0.2 0
0
5
10
15
20
25
30
35
40
45 t, min
Figure 3.1 Kinetic curves of crystallization of elastomers (dependence of relative changes of the degree of crystallization C/Cf on time t) for (a) polymethyl vinyl siloxane (primary crystallization at –50°C; W0, induction period; W1/2, crystallization half-time) and (b) 1,4-cis-polybutadiene with cis-unit content w, mol. %: 90.2 (I) and 97 (II), numbers at the curves, temperature, K. The curves were obtained by the dilatometry (a) and DSC (b) methods.
crystallization rate, the smaller the value of W1/2 is. Any time Wx, corresponding to any other fraction x of crystallization-related changes can be chosen to characterize the rate of crystallization instead of the W1/2 value. Often, the values of W0.1 and W0.2 are used. The crystallization completion time is usually denoted as Wf. The change of the properties for this time corresponds to the ultimate degree of crystallinity of material, Cf, under given conditions. The most widespread way to describe the kinetics of isothermal crystallization of polymers, including elastomers, is to use a modified Kolmogorov–Avrami equation*, which relates the degree of crystallization Ct (referred to its ultimate degree Cf) to crystallization time t [31, 342]:
* This equation was first obtained by Kolmogorov [342] for Cf = 1 for metals at n = 4, and then independently by Avrami [31] at n from 1 up to 4 for ceramics.
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
Ct / Cf
1 exp k ct n ,
67
(3.3)
where kc has the meaning of the constant of the total crystallization rate; n is the Avrami parameter, having the meaning of dimensionality of crystalline formations’ growth; usually, n has integer values of 1 up to 4 [95, 435, 636]. A convenient form of writing this equation was proposed by Reznikovsky, see [95, 96]:
Ct / Cf
n 1 exp ª« t / t0 º» , ¬ ¼
(3.3c)
where again Ct is the degree of crystallization at time t, and Cf is the ultimate degree of crystallization (for elastomers, Cf < 1); t0 is the characteristic time of crystallization. The crystallization process of elastomers, as of all other substances, consists of the processes of primary nucleation and crystal growth; the latter is determined by the process of secondary nucleation, i.e., formation of nuclei on the growing surface of a crystal. Primary nucleation can be of homogeneous and heterogeneous character. If the formation of primary nuclei from crystallizing material proceeds simultaneously with their growth, the nucleation process is called homogeneous or thermal. If foreign impurities serve as nuclei, including the nuclei formed under other conditions, e.g., in cooling to the crystallization temperature, the nucleation process is taken to be heterogeneous or athermal. Note that the real shape of the curve obtained in studies of the overall crystallization rate depends on the ratio of the rate of a process and the speed of its registration (the crystallization rate of elastomers may change from fractions of a minute to hours and tens of days, and depends significantly on the structure of the polymer and crystallization conditions). In most cases, the Kolmogorov–Avrami equation gives a reasonable idea of the crystallization mechanism consistent with the results of direct structural methods.* However, the use of different experimental methods, whose time parameters differ from the kinetic parameters of an investigated process, may lead to different values of the constants of these equations [246]. In some cases, this equation fails to describe the entire crystallization process by the same set of constants. The treatment of the kinetic curves for elastomers containing an active filler or subjected to a stress sometimes yields fractional values of n. Changes of the parameters in eq. (3.3) in crystallization and fractional values of n are due to, first and foremost, the fact that assumptions taken when deriving it (in particular, on the constancy of shape, growth rate and nucleation rate) are not always observed for polymers, even in the case of isothermal crystallization. Description of the nonisothermal crystallization process requires eq. (3.3) to be modified.
* Equation (3.3), in fact, coincides with the Erofeyev equation obtained to describe chemical reactions [717, 718]. The kinetics of isothermal crystallization of rapidly crystallizable polymers is also described using other equations, which describe the kinetics of chemical reactions, in particular, of the autocatalytic type [66].
68
CHAPTER 3
3.2.3 Temperature range of crystallization
As we have already noted, crystallization of polymers develops within the temperature range below the melting temperature and above the glass-transition temperature. The first stage of the crystallization process – primary nucleation – consists in the formation of crystalline-phase nuclei in the matrix of amorphous material. The process is of a fluctuation character. A forming nucleus becomes stable when its size reaches a critical value, i.e., if the decrease of free energy in its formation is larger than the energy rise stipulated by the formation of a new interface between the crystalline and amorphous phases. At T > Tm0, these fluctuations are always of a smaller size than the critical size of a crystalline-phase nucleus and are, thus, unstable. However, for elastomers of high viscosity the time of their existence is sufficiently large, and the size can reach tens of angströms. Apparently, these fluctuations are the major type of physical junctions, the presence of which changes the character of elastomers’ deformation and relaxation properties in the region of the rubberlike plateau [44, 46, 97]. The high viscosity of elastomers stipulates the occurrence of these fluctuations at T < Tm0, too; in that case their lifetime is even larger and the size is limited by the critical size of the nucleus at a given supercooling. Noncrystallizable polymers are those for which, up to the glass-transition temperature, no formation of nuclei of a size equal to the critical value is possible. This is due to the disturbances of the regularity of macromolecular chains of such elastomers. A “borderline” case, when the size of part of the forming crystals reaches the value of a critical nucleus due to the occurrence of sufficiently long sequences of the crystallizing comonomer, leads to the phenomenon of microcrystallization (see Section 3.5). The same phenomenon can be observed in the case of close values of the temperature of maximum nucleation rate and glass transition. The views of the physical junctions as prephase fluctuations of a size smaller than that of the nucleus are fruitful, as they make it possible to describe anomalies of elastomers’ rubberlike properties using the nucleation data for crystallizing elastomers. At the second stage of crystallization – in crystal growth – amorphous material is transferred to the boundary of a nucleus, and “secondary” nuclei are formed on the surface of a growing crystal. The temperature dependences of the overall crystallization rate v, as well as the rates of thermal nucleation N and growth G are described by equations similar in their form: D ª Dº v0 exp « U / kT \ Tm0 / T 'T » ; ¬ ¼
(3.4)
N
D ª Dº N 0 exp « U / kT \ c Tm0 / T 'T » ; ¬ ¼
(3.4c)
G
D ª Dº G0 exp « U / kT \ cc Tm0 / T 'T » ; ¬ ¼
(3.4cc
v
where the values \, \ c and \ cc can be presented as
\
D 1
E V
D
k 'H U kD
.
(3.5)
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
69
Here v0, N0, G0 are constants; U is the activation energy for the transfer of molecules of a substance; k is the Boltzmann’s constant; U k is the density of totally crystallized material; 'H, heat of fusion; V, mean energy of interface formation; E and D are coefficients depending on the crystal formation and growth mechanism, crystal growth coefficient Dcan take the values of 1 or 2; 'T, supercooling (see eq. (3.2)). The value of N is determined experimentally by measuring the change in the number of growing crystalline sites by the microscopy method. However, for elastomers this is possible only in exceptional cases [19, 96, 174, 569], so the nucleation rate is usually characterized by the induction period W0, which is the larger, the smaller* N is. The linear growth rate is usually also measured by the microscopy method by the changes of crystals’ linear size, which is also made difficult for elastomers.** An exception are electron-microscopy measurements of the growth rate of 1,4-cis-polyisoprene single crystals performed by Andrews et al. [19] and Phillips and Edwards [570]. The use of the slope of the crystallization kinetic curve on its second fragment (see Fig. 3.1a) as a growth-rate characteristic is possible only in the case of heterogeneous (athermal) nucleation. The problem of singling out the nucleation and crystal-growth processes and their contributions to the crystallization rate of elastomers crystallizable at low temperatures could not be solved based on direct measurements. Its solution required the use of the data on the reorganization of crystals during the melting [774, 776, 777]. To characterize the overall growth rate, it is more convenient to use the inverse value W1/2 (or Wx, if all the kinetic curve could not be obtained). In the logarithmic form, we can write: ln W1/ 2
D
ln C U / kT \ Tm0
D
/ T 'T ;
(3.4ccc)
where C is constant. The curves of the temperature dependence for the parameters characterizing both the overall crystallization rate and the rates of nucleation and growth have an extremum and are called the Tamman curves. Their shape is the same both for polymers and lowmolecular-mass substances, but for polymers the temperature range for the existence of crystals is usually much broader. The extremum of the Tamman curve is due to the fact that the second and third terms in the right-hand side of eq. (3.4ccc), as two terms in square brackets (two exponents) occurring in eqs. (3.4), (3.4c) and (3.4cc), describe two tendencies whose competition determines the position of a maximum on the temperature dependences of the crystallization rate V, N and G (or a minimum on the curves for W1/2, Fig. 3.2). The first tendency is the increase of crystal formation rate with the supercooling 'T going up (owing to the increase of the energy efficiency of the crystallization process); the second is the decrease of the crystal formation rate with temperature going down owing to the increase of viscosity of the system. * Taking into account the dependence ofW0 on the sensitivity of the method used, the nucleation rate should be characterized via W0 with great care and only for comparison of close systems investigated by the same method. ** It appears that the growth rate measurements of polycrystals could not be used to obtain correct values of the constants in eq. (3.4c).
70
CHAPTER 3 log τ1/2, min 4
1 2 4
2
3
5 0 -100
-50
0
T, °C
Figure 3.2 Dependence of log W1/2 on temperature T for unfilled rubbers from polyurethane SKU-8 (1), NR (2), polychloroprene (nairit NP) (3), polymethyl vinyl siloxane SKTV-1 (4) and 1,4-cis-polybutadiene SKD (5).
In accordance with eq. (3.4), the crystallization rate tends to zero at the equilibrium melting temperature Tm0 and at T o 0, and then passes through the maximum at the maximum crystallization-rate temperature T1. However, owing to a sharp increase of U with the temperature decrease, the crystallization rate is negligibly small already near the glasstransition temperature. The curve describing the temperature dependence of the overall crystallization rate is a superposition of two curves describing the processes of primary nucleation (N–T curve) and crystal growth (G–T curve). The temperature shift of these dependences is mainly determined by different contributions of the interface-formation energies in the case of primary nucleation and subsequent crystal growth. The result of these differences is that the nucleation maximum-rate temperature both for low-molecular-mass substances and polymers is slightly lower than the maximum growth rate temperature. At the early stages of analysis of the crystallization-rate temperature dependence, the value of U was taken to be constant [95, 233, 380, 450]. However, as we have shown in Chapter 1, description of the changes in elastomers’ properties near Tg using constant values of U is impossible. Therefore, to describe the changes of viscosity accompanying glass transition and leading to a decrease of the crystallization rate owing to difficulties of transferring material to the interface, the temperature dependence of U is used in the form of eq. (1.8) [92, 93, 96, 126, 246, 437, 534, 535, 569, 586, 751]. Then U / kT
c1 / ª k c2 T Tg º . ¬ ¼
(3.6)
Based on the single-crystal three-dimensional growth model, the value of the crystal-growth coefficient is taken to be D = 2 and for two-dimensional growth D = 1. Treatment of experimental data both for elastomers and for other polymers does not make it possible to choose between these two values of D: the same experimental data are described sufficiently well by eq. (3.4) both at D = 1 and at D = 2 [95, 246, 435, 437]. Electron-microscopy studies of the growth of polymer single crystals (lamellae) showed the process to form two-dimensional secondary nuclei, i.e., to be two-dimensional. Therefore, the value D = 1 has been used more often recently in the treatment of the
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
71
temperature dependence of the crystallization rate [92, 93, 96, 246, 437, 533–535, 586, 751]. The choice of the value of E also depends on the chosen crystal-growth model. In the three-dimensional growth for cubic crystals, E = 32; for a cylinder-shaped nucleus,E = 4S. Considering that the two-dimensional growth of nuclei is more probable, the value E = 4b0 is used, where b0 is the thickness of the nucleus (parameter b in the crystal unit cell, see Table 3.1). Taking the two-dimensional growth model of a lamellar crystal, one should consider not the averaged energy of growing-crystal interface formation V, as it is done in eq. (3.5), but divide the formation energies of the side (Vu) and end (Ve) faces. Then at D= 1, eq. (3.5) takes the form:
\
4b0V eV u . k 'H U k
(3.5c)
The value of Vu is usually calculated using an empirical relation:
Vu
0.1b0 'H U k .
(3.7)
The value of Ve is then found from the values of \ :
Ve
k\ / 0.4b02 .
(3.7c)
The value of \ is determined from the slope of an experimentally obtained temperature dependence forW1/2 (or Wx) processed in the coordinates –A – F1, where A
ln W1/ 2 c1 / ª k c2 T Tg º ; ¬ ¼
(3.8)
Tm0 / T 'T .
(3.8c)
F1
Thorough analysis of experimental data for elastomers has shown that at constant values of the parameters in eq. (3.4), an inflection is observed in the region of small supercoolings on the curves in –A –F1 coordinates. Its emergence was suggested to be due to a change of the character of the forming crystals and, therefore, with the change of Ve [96]. The ratio of the values of Ve (obtained at D = 1 and E = 4b0) in the high- and low-temperature fragments (i.e., at small and large supercoolings 'T, respectively) is close to 0.5. In accordance with the Lauritzen–Hoffman crystallization theory [395], the occurrence of inflections in the –A –F1 curve is due to the presence of three different crystallization regimes (see [147, 569, 572, 599]). Regime I is observed at small 'T, when the rate of transfer of material to a face of a growing crystal is high, and all the face of the growing crystal is filled in each act of secondary nucleation. At intermediate supercoolings, when the rates of material transfer and secondary nucleation are the same, regime II is on. And, finally, at large supercoolings, when the transfer of material to a face of a growing crystal is made difficult, regime III takes place. At this regime, the major role is played by
72
CHAPTER 3
nucleation. According to this theory, not the value of Ve changes in passing from one regime to another, but coefficient i introduced into the numerator of eq. (3.5), written down as*
\
4ib 0V uV eTm0 . k 'H U k
(3.5cc)
For regimes I and III, i = 1; for regime II, i = 0.5. For elastomers, crystallization at very small supercoolings'T is not observed, and regime I is, apparently, not realized under usual conditions. Transition from regime II to regime I was observed only for narrow fractions of 1,4-cis-polyisoprene with a low molecular mass (MM = 3.13 × 105) at a temperature T = –13.8°C [572]. An inflection in a –A –F1 curve and directly in the temperature dependence of W1/2 [774] indicates** that regimes II and III are realized; however, the major role for elastomers seems to be played by regime III. According to [147, 292, 395], this regime implies a decrease of the number of molecules “folding” into themselves and an increase of the number of tie chains, i.e., crystallization at large 'T is accompanied by the formation of small crystals representing the integration of the most regular sections of adjacent macromolecules, i.e., “micelles” to use the terminology of the old theory. Apparently, regime III is mainly performed in the crystallization of copolymers, too, especially block copolymers. One of the most important parameters to characterize the crystallization process is the maximum crystallization-rate temperature T1. For rapidly crystallizable polymers and low-molecular-mass substances, crystallization is usually completed in the process of cooling already at T > T1. For elastomers, most of the process proceeds near T1, and exactly in this temperature range it is the most harmful for rubber articles. In most cases, this temperature is determined experimentally (see Fig. 3.2). From eq. (3.4), it is readily assessed as an extremum point, i.e., a temperature at which dV / dT
0.
(3.9)
Thus, Gent, using eq. (3.4), solved an inverse problem: he found the value of U (taking it to be constant) knowing the value of T1 [233]. The other way round, taking the dependence of U on T in the form of eq. (1.8) and knowing the equilibrium melting temperature Tm0, one can calculate T1 or, knowing T1, to assess Tm0. For an estimate of T1 we can also use empirical equations, e.g., T1
D Tm0 Tg / 2,
(3.10)
where D d1. The value of the coefficient reflects the asymmetry of the Tamman curve described by eq. (3.4). The smaller D, the stronger the maximum is shifted towards larger supercoolings 'T. * Analysis of data available in the literature shows that the parameters of eq. (3.4) can always be chosen in such a way that there will be no inflection in the –A –F1 curve. ** A change of slope of –A – F1 curves can be interpreted either as the result of the twofold change of the value of Ve or the result of the change of parameter i from 1 to 0.5.
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
73
Generalization of a large body of experimental data for many polymers has shown that there is an empirical relation between Tm0 and Tg: T1
D1Tm0 ,
(3.10c)
in which D1 = 0.82–0.83 and is constant for a large group of polymers. For elastomers, this relation is also fulfilled satisfactorily, D1 varies within slightly larger limits (0.75–0.85). The relation (3.10c) can be rather useful for finding the region of possible values of T1 (if Tm0 is known) or Tm0 (if T1 is known). Often, an empirical equation Tg
D2Tm0
(3.11)
with D2 ~ 0.67 for most polymers.* For elastomers, D2 varies from 0.60 up to 0.74. Tg and T1 are related as Tg
D3T1 ,
(3.11c)
where D3 changes from 0.76 up to 0.89, so that we can approximately think that Tg / T1 | T1 / Tm0 .
(3.12)
The validity of eq. (3.12) is indicative of a kind of a symmetry of the temperature dependence of the crystallization rate, and makes it possible to use it for a rough estimate of T1 or Tm0. As for most slowly crystallizing elastomers it is easy to determine only Tg, the system of equations (3.10)–(3.12) enables assessing T1 and Tm0 to know the temperature region to carry out experiments for detecting and studying the crystallization of new elastomeric materials. Equation (3.10c) is the most widespread of the above considered relations. It is used to assess Tm0 for copolymers. Close results were obtained based on the theoretical relations between Tm0, Tg and T1 [533, 534]. Other methods of assessing Tm0 will be considered below. It is believed that the boundary between the temperature regions corresponding to regimes II and III considered above coincides with T1. However, as shown by the measurements of the temperature dependence of the crystallization rate, for specimens of 1,4-cispolybutadiene this boundary does not coincide with T1, but is shifted to higher temperatures [774]. 3.2.4 Kinetics of nonisothermal crystallization
For rapidly crystallizing elastomers, as well as during the crystallization in the presence of fillers or under the action of a stress, part of or all the crystallization process can occur under * Limited applicability of eq. (3.11) is seen for cis- and trans-polymers; their Tg are practically equal, and Tm0 for trans-polymers is usually tens of degrees higher than for cis-polymers.
74
CHAPTER 3
nonisothermal conditions during the cooling. A heating of an elastomer rapidly cooled to below Tg (quenched) can also be accompanied by nonisothermal crystallization [111, 702, 702a,b, 703, 754, 754a]. Nonisothermal crystallization during the heating is also observed in the melting process [666, 667, 719, 773, 774, 776]. Major nonisothermal-crystallization studies have been performed on highly crystalline polymers. For them the problems associated with nonisothermal crystallization are of ever growing importance in relation with the development of technologies for production of large-sized articles, extrusion of fibres, filaments, films, pressure moulding. This process should be also taken into account during the synthesis and processing of thermoelastoplasts. Most approaches used for description of nonisothermal crystallization kinetics are based on a modification of Kolmogorov–Avrami equation (3.3) [541, 613]. Herewith, it is rather fruitful to take into account the temperature dependence of the crystallization rate in the form of eq. (3.4) and eq. (3.6). Godovsky et al. [613], besides, added to the consideration the sensitivity of the measuring instrument, and also separated the contributions of nucleation and polycrystal growth to obtain a system of equations, which make it possible to predict the temperature of the onset of crystallization in the nonisothermal regime. The equations obtained make it possible to calculate the kinetics of nonisothermal crystallization from the results of isothermal-crystallization experiments, and also to determine the kinetics of nonisothermal crystallization at an arbitrary temperature change from the results of nonisothermal-crystallization experiments at another regime of temperature change. Another approach to the description of nonisothermal-crystallization kinetics is to use equations describing autocatalytic processes [65–67, 374]. This approach is fruitfully used to describe crystallization in the process of polymerization. The most successful studies of nonisothermal-crystallization kinetics have been carried out using the DSC method [65, 302, 613]. 3.2.5
Melting
Differences in the behaviour of low-molecular-mass substances and polymers are the most pronounced in melting, i.e., in transition from the crystalline substance to the melt, or, in other words, from the crystalline to the amorphous state. Usually, the melting registered by the absorption of heat using calorimetric methods or by the increase of volume of a specimen in dilatometric experiments, is studied in the process of heating a crystallized specimen at a constant rate. From eq. (3.1) it follows that melting is accompanied by a change of entropy; this change is a sum of three components: 'S
'S1 'S2 'S3 ,
(3.13)
where 'S1, 'S2 and 'S3 are contributions stipulated by position, orientation and conformation changes, respectively. Thus, elastomers with an increased intermolecular interaction or slight conformations disordering after the melting shall have a high melting temperature; Tm0 of elastomers are within the range of –50 up to +100°C. Most elastomers increase in volume during the melting within the limits of 1 up to 6%, which coincides with the change of volume in metals. For other low molecular liquids, this value can reach 20%. Melting, as crystallization, is characterized by a certain kinetics, which depends on the rate of the process at the interface. The melting rate is tens and hundreds of times higher
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
75
Tmb;Tmf, °C 40 20 2
0
1
-20
3
-40 -40
-20
0
20 T, °C
Figure 3.3 Dependence of the temperatures of the beginning Tmb (1) and end Tmf (2) of the melting interval measured by the dilatometry method on the crystallization temperature T for NR (according to the data from [69a]); T = Tm (3).
than the crystallization rate, though the forms of the kinetic curves are similar. To describe the melting kinetics, the same techniques can be used as those used for describing the nonisothermal-crystallization kinetics.* The difference in the crystallization and melting rates is due to the fact that the crystallization-limiting stage is nucleation, which occurs only at considerable supercoolings, whereas liquid-phase nuclei exist practically always on the edges of the crystal faces. Therefore, the melting rate is determined only by how fast the surface of the crystal–melt interface advances into the crystal. Under usual conditions, when no special measures are taken, transition of a polymer from the partially crystalline to the amorphous state occurs not at any one temperature Tm as for low molecular substances, but within the whole range of melting, limited by the temperatures of the beginning (Tmb) and end (Tmf) of melting. Herewith, the width 'Tm
Tmf Tmb
(3.14)
and position of the melting interval depend on the crystallization temperature T. The first data on the melting temperature range of polymers were obtained [748] for NR (Fig. 3.3) by the method of volume dilatometry. Using this method, the heating of crystallized material is accompanied by its increase in volume; at temperatures close to T it corresponds to thermal expansion of partially crystalline material. Then at Tm1 the melting proper begins, which is accompanied by a larger volume change. Finally, at a temperature Tm2 all material passes into an amorphous state, and further heating is accompanied by a volume increase corresponding to the thermal expansion of amorphous material. From Fig. 3.3, it is seen that the entire melting interval is above the curve Tm = T. The presence of a melting interval for elastomers is the cause of a kind of a “hysteresis” observed at a periodic change of operation or storage temperature [121]. This should be taken into account when developing the melting regimes for crystals (decrystallization, “steaming”) of elastomeric materials and articles (see Chapter 6).
* For elastomers this was done in [773].
76
CHAPTER 3 ΔT'
dQ/dt b
a
1
8 t, min 223
4
243 T', K
1 Δcp, kJ/g
Δcp, kJ/g d
endo
c 4
2
6
5
3 2
5 -90
-50
-10
Fig. 3.4
2 4
1 1.0
4
0.5 30 T', ºC
238
268 T', K
1
260 270 T', K
Fig. 3.5
Figure 3.4 Melting thermograms (dependences of the temperature difference in cells 'Tc on scanning temperature Tc) obtained by the DTA method for 1,4-cis-polybutadiene specimens (crystallized at –40°C for 3 h) with different network densities (1/Mc ·104): 0 (1), 1.06 (2), 1.58 (3), 1.79 (4), 2.64 (5). Figure 3.5 Crystallization (a) and melting (b–d) thermograms obtained by the DSC method, initial (a, b) and recalculated for specific heat changes 'cp (c, d), for 1,4-cis-polybutadiene specimens with cis-unit content of 92% (a–c) and 97% (d); crystallization temperatures 218 K (a, b), 238 K (c) and 260 K (d); crystallization time 11 min (b), 120 min (c) and 160 min (d); heating rate vh (K/min) 5 (1), 20 (2), 40 (3), 80 (4), 160 (5); 6, a calculation curve corresponding to the melting of a specimen with the initial distribution of crystals.
Dilatometry is an integral method of studying crystallization and melting. When using calorimetric methods, a maximum heat absorption (the endothermal effect) on the thermograms corresponds to melting. This is the heat flow maximum dQ/dt (corresponding to the maximum of specific heat changes 'cp) in the case of DSC or the temperature-difference maximum between the cells 'T c in the case of DTA. Figure 3.4 (curve 1) presents a DTA thermogram for 1,4-cis-polybutadiene [121]. The melting thermograms obtained by the DSC method (Fig. 3.5, b – d) [773] are of the same character. As seen in Figs. 3.4 and 3.5, not one but two (or more) melting peaks can be registered [336, 773]. Studies performed for many polymers [246, 247, 249 751, 752, 754a,b], including for elastomers [336, 666, 667, 719, 754, 754a,b, 760, 773, 774, 776] have shown that a melting interval is due not only to the existence of a set of different-size crystals in a crystallized specimen, but also to their reorganization, which is a consequence of the metastability of finite-size crystals. Before they melt finally, these crystals can melt and crystallize on nuclei or pass from one metastable state into another, more perfect, state, so the experimentally
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
77
T m1 ,T m2 ,T mf , K 3,3' 250
2 4
1,1'
2' 230
210 200
220
240
T, K
Figure 3.6 Dependence of the temperatures of the first Tm1 (1, 1c) and second Tm2 (2, 2c) peaks and of the end Tmf (3, 3c) of melting on crystallization temperature T for 1,4-cis-polybutadiene at heating rates vh (K/min): 5 (1, 2, 3), 80 (1c, 2c, 3c); T = Tm (4).
observed Tm exceeds that of the initial crystals. As the result, multiple melting curves are observed. Multiplicity of the melting curves can also be due to polymorphism, i.e., to the occurrence of various crystalline modifications. For elastomers (excluding gutta percha), this effect has not been observed until the present time.* The multiplicity can be also due to the character of the distribution of the initial crystals by size [336, 437, 751, 752]. The cause of the emergence of multiple melting curves for elastomers is the reorganization of crystals in the process of melting. Detailed melting studies at different heating rates vh made it possible to establish this unequivocally for polychloroprene [666, 667, 719]. Similar results were obtained for 1,4-cis-polybutadiene [751, 754, 773, 774, 776] (in contrast with earlier assumptions). Thus, the melting studies of 1,4-cis-polybutadiene specimens precrystallized for different times and heated at different rates vh established that in this case too the multiplicity of the melting curves is determined by the reorganization process (see Fig. 3.5c), and the initial distribution of crystals by size is unimodal (curve 6, Fig. 3.5c). The low-temperature peak on the melting curves, the beginning of which is close to the crystallization temperature, reflects the melting of part of the initial crystals, which did not reorganize in time. These studies made it possible to separate the nucleation and growth processes. They also showed that the process limiting the elastomer crystallization rate is primary nucleation. The rate of secondary nucleation determining the crystal growth rate in the melt is several orders of magnitude higher than the primary nucleation rate [103, 774, 776]. The dependence of temperatures corresponding to the maximum of, first, low-temperature (Tm1) and, second, high-temperature (Tm2) melting peaks on crystallization temperature, presented in Fig. 3.6 is similar to the temperature dependence of the melting interval determined by the dilatometry method (see Fig. 3.3). Treatment of such dependences ooooooooooooo
* Multiplicity of the melting curves for oriented, crystallized and then unloaded specimens is associated with a kind of a topomorphism due to the presence of crystals of different morphologies (see Chapter 4).
78
CHAPTER 3
enables assessing the equilibrium melting temperature* Tm0, which was done for 1,4-cis-polyisoprene [160], 1,4-cis-polybutadiene of different 1,4-cis-unit content [380, 450, 773, 774, 776] and polysiloxanes of different structures [104]. Determination of the equilibrium melting temperature Tm0 for elastomers is of special importance for two reasons. First of all, this is the necessity of developing “decrystallization” methods for rubbers and rubber articles after their crystallization in storage. In decrystallization, it is necessary to avoid the repeated, faster crystallization (see Chapter 6). This could be achieved only performing decrystallization at T > Tm0. The second reason, which makes the determination of Tm0 so important, is the need for developing crystallization-predicting methods, which is especially topical for elastomers with their low crystallization rates. There are several ways of determining the values of Tm0. The most widespread method (especially as regards elastomers) was proposed by Hoffman and Weeks. The method makes use of linear extrapolation of the Tm2 dependence (or of the melting end temperature Tmf) to infinitely small supercooling, i.e., up to the intersection with the curve Tm – T [95, 96, 160, 247, 437a, 751, 774]. Herewith, only the high-temperature part of the dependence should be used; usually for elastomers, in contrast with highly crystalline polymers, it is easier to obtain its low-temperature part, where Tmf is practically independent of T (see Figs. 3.3 and 3.6).** Another method of assessing Tm0, which was in practice used only for polyethylene, is extrapolation of the Tm0 dependence on molecular mass M (or degree of polymerization x1) using empirical dependences of Tm and 'H on x1. In this case, extrapolation is justified if the low-molecular homologues crystallize to form the same crystalline structure as the polymers. The difficulty of using this method is the nontriviality of the procedure of producing low-molecular homologues, completely identical by their chemical structure (absence of functional end groups) and stereo structure (the same content of isomeric forms). One more way of extrapolation to yield Tm0 is based on the dependence of this value for copolymer (to be more exact, the maximum melting temperature of copolymer, Tmn) on the number n of foreign units at their statistical distribution:
1/ Tmn 1/ Tm0
R / 'H u ln n ,
(3.15)
where Tm0 is the equilibrium melting temperature of homopolymers; 'Hu, the heat of fusion per repeating unit; R, gas constant. This equation differs little from the Flory equation [201]:
1/ Tmn 1/ Tm0
2 R / 'H u ª« ln n F 1 vc º» , ¬ ¼
(3.15c)
where F is the interaction parameter, which reflects the component mixing heat; vc is the volume fraction of crystallizing units. * At Tm close to Tm0 no reorganization is observed, and the melting curves are unimodal (see Fig. 3.5d). ** The use of this method draws objections from some investigators, but it is this method that provides reliable data for Tm0 and Tmn (see below) for elastomers.
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
79
The complexity of using both these equations is that it is none the easier – and sometimes much more difficult – to obtain the maximum melting temperature of copolymer Tmn than Tm0 of homopolymer.* The same principle can be used if, instead of copolymer, blends of copolymers with one another or with a solvent, are used. Thus, in a number of cases the determination of Tmn of an elastomer in a blend with a plasticizer proves an easier task [96, 174]; however, in principle, this method is as difficult to use as that based on eq. (3.15). Yet another way of determining Tm0 is extrapolation of experimentally obtained melting temperatures for crystals of known size to Tm0 of an infinitely large crystal by the Thomson–Gibbs equation [751] Tm
Tm0 ª¬1 2V e / l 'H º¼ ,
(3.16)
where Ve is the specific surface energy of end faces; l is the size of a crystal along axis c; 'H is the specific heat of fusion of a crystal.
As we have already noted, it is usually impossible to determine l for elastomers by an independent method, so eq. (3.16) is usually used to solve an inverse problem – to assess l from the melting data. The value of Ve used in this equation can be obtained by treatment of the temperature dependence of the crystallization rate (see Section 3.2.3), and Tm0 by extrapolation of the melting data using the Hoffman–Weeks method. An exception, as we noted, is 1,4-cis-polyisoprene. Using eq. (3.16), the values obtained are Tm0 = 35.3±0.3°C [570], which is slightly lower than the value Tm0 = 40°C obtained by other methods, in particular, by extrapolation to the zero width of the melting interval. It should be noted that the use of eq. (3.16) is the only reliable way of assessing the size of elastomers’ single crystals; however, it is of approximate character, too, due to the inaccuracies of determining Ve and the necessity of obtaining the melting curves for the initial crystals not those reorganized during the melting** [751]. Considering the experimental problems in determining Tm0, it is convenient to assess it using the data on the effect of stress and pressure on melting temperature, as namely under the action of these external factors the crystallization and melting processes approach equilibrium values [95, 437a].
3.3
Effect of mechanical action and other factors on crystallization
Effect of stress. The first data on the acceleration of crystallization in deformed specimens were obtained by Gehman et al. [227], who studied the decrease of the recovery of com* Due to this, assessment is often done using not Tmn but Tm obtained at some crystallization temperature T < Tm0, and a random heating rate, i.e., under conditions far from the equilibrium, which leads to underestimated values of Tm0. ** Frenkel and Baranov developed the concepts [40, 41, 191, 206], according to which Tm0 for crystals with extended and folded chains are different. As the result, the crystal–melt transition on the diagram for the temperature dependences of Gibbs energies should be shown not by a dot, but by a four-sided melting region restricted by the intersection lines of respective phase “corridors”. Experimentally, this should result in the appearance of a number of discrete melting temperatures, as in the case of reorganization during the melting. It appears, however, that the very term “equilibrium melting temperature” is inapplicable for folded-chain crystals.
CHAPTER 3
80 σ/σ0
K 1.0
a 3
1.0
0.5
5
2
b
1 0.5 4 4
1 3
0
102
103
t, s
1
2
2 3
4
5
6
t, h
Figure 3.7 Dependence of the relative change of stress V /V0 and recovery K on time t for rubbers from polychlroprene at –10°C (a) and 1,4-cis-polybutadiene at –40°C (b) at stretching deformations: 0.16 (1), 0.95 (2), 1.61 (3), 2.19 (4), 2.85 (5) (a) and compression deformations: 0.15 (1), 0.30 (2), 0.50 (3), 0.70 (4) (b).
pressed specimens in crystallization; accelerated crystallization with deformation rising was observed under the action of compression, uniaxial and biaxial stretching, as well as shear [95, 96, 100a, 183, 230, 234, 249a, 264, 335, 450, 455, 458, 685, 686] and under dynamic deformation [566]. Studies of elastomers’ crystallization acceleration under the action of deformation (or stress) are very important from the practical point of view, as most rubber components are operated in a stressed state. Stretching and shear deformations occurring in the processing of elastomers also lead to acceleration of crystallization [275, 478, 699]. The accelerating action of a stress on crystallization enables significantly speeding up its studies; this approach is the basis of the standardized methods (ISO 6471 [679] and Russian standard GOST 13270-85). Systematic studies of the effect of elongation on the crystallization on NR, performed by Gent [230, 234] showed that the kinetic curves of the crystallization of stretched rubbers based on NR at not too large deformations H have the same shape as at H = 0. The same applied to compression deformation [95, 96, 686]. However, in the region of large stretching and compression deformations the character of the kinetic curves changes. Already at H > 1–3 in elongation and H > 0.5–0.7 in compression the induction period in the curves disappears (Fig. 3.7), but the kinetic curves are satisfactorily described by eq. (3.3). Herewith, parameter n even for unfilled rubbers (see Section 3.4 below) decreases with the stress rising [95, 96, 230, 234, 686]. As the deformation increases, the morphology of the crystals forming under the action of stress changes, too [12, 18, 40, 41, 206, 322, 325, 569]. If a polymer is deformed before crystallization, then, as shown by Andrews et al. [12, 17, 18], polycrystals called “shish kebab” are formed. They consist of axial fibrillae oriented along the direction of the stretching and folded (or micellar) crystals formed on these fibrillae (see Chapter 4). Gent also showed that there was a linear dependence in the region of H < 2 between logW1/2 and stretching deformation H; but that at H |2–3 this dependence deviated from the linear shape. Quantitatively, accelerated crystallization under the action of stress can be described by an empirical equation for a characteristic straight line [94–96]: log W1/ 2
0 log W1/ 2 BV ,
(3.17)
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
81
0 where logW1/2
is the crystallization half-time in the absence of a stress; V is the stress in the melt prior to crystallization, referred to the true cross-section area of a specimen; B is the slope of the characteristic straight line to axis V ; the B value is the measure of the effect of the stress on crystallization. This equation is valid at not too large deformations of both compression and elongation [95, 453]. The characteristic straight lines are parallel at various temperatures. However, in the region of large elongations the experimentally observed values of logW1/2 are not described by eq. (3.17). This is due to the change of morphology of the crystals. Extrapolation of logW1/2 values for V = 0 using eq. (3.17) yields the values close to W1/2 for a nondeformed specimen. This implies that within the range of deformations, where eq. (3.17) is valid, the crystallization kinetics of a deformed specimen is determined predominantly by the same mechanism as for nondeformed specimen. The effect of deformation in this case is related only to the increase of nucleation rate. The same is indicated by the data on the morphology of deformed specimens. Studies of crystallization under the action of a stress make use of modified methods of dilatometry [446], DTA and DSC [246, 264] and modulus measurement [384, 447, 686]. The use of the birefringence measurement method is efficient [582, 585, 715, 764]. Especially widespread are the relaxation methods – recovery at compression [95–97] and at elongation [384], as well as stress relaxation [230, 234, 335, 661]. Comparison of the recovery and stress changes with the volume decreasing at crystallization, which is a classical method of its studies, shows that three cases are possible: (1) Recovery and stress decrease to zero (K = 0, V = 0), and crystallization-induced volume changes continue. This can take place at high ultimate crystallizations (usually Cf > 0.3) or large deformations (H > 0.8). Herewith, extrapolation by eq. (3.17) for V = 0 can give values not coinciding with those measured in the absence of a stress. This case is the most critical in operation of elastomers, especially as seals. (2) By the completion of crystallization, K > 0 and V > 0. This effect is characteristic of blends of polymers and copolymers, namely in this case the development of crystallization may not lead to the loss of performance of a rubber article. (3) By the completion of crystallization, K = 0 and V = 0; this case is observed the most often.* A fourth case can be observed in stress measurements, too – at the final stages of crystallization, V increases. This effect is due to the fact that the measured stress
V1
V V c,
(3.18)
where V is the true stress in a specimen;V c is the stress emerging at the deformation of the specimen during the measurement process. Even if V decreases in the process of crystallization, the increase of V c as a consequence of the rise of rigidity of the specimen can lead to the overall rise of V 1, which was observed in crystallization of gutta percha and polychloroprene (see Fig. 3.7a) [235]. Effect of pressure. Crystallization studies of 1,4-cis-polyisoprene [160, 185–187, 571] at not too large pressures p |0.1–02 MPa have shown that its morphology has much in common with the morphology of crystals in the absence of pressure. The same polycrystals (of the “shish kebab” type) as at intermediate stretching stresses are formed in this case, too. However, as p rises, an increasing part of material crystallizes as extended-chain crystals – the most equilibrium type of crystals. The formation rate of these crystals is extremely * Probably, due to the insufficient accuracy of measuring small values of K and V.
82
CHAPTER 3
high, so it is extremely difficult to follow the crystallization kinetics. The temperature dependence of the crystallization in this case is described by eq. (3.4), as at an atmospheric pressure. However, at p |0.05 MPa an apparent rise of interface formation energy V e was observed. It can not be due to the passage from regime III to II, as in this case V e decreases. Therefore, the authors associate this effect with the change of conformation of molecular chains in the folding, which can occur during the formation of single crystals with folded chains. It seems, however, that this apparent effect is due to the calculation not considering the rise – taken into account in eq. (3.4)) – of the glass-transition temperature Tg with pressure (see Chapter 1). As for polyethylene, for which studies of crystallization at high pressures are the most numerous [351, 352, 751], in this case, too, the formation of an intermediate liquid-crystalline phase was suggested. As for low-molecular substances, pressure p increases melting temperature, so the Clapeyron–Clausius equation is applicable to describe this effect. d ln Tm dp
'V / ' H ,
(3.19)
('V is the change of volume in melting). The validity of this relation at its right-hand side constant was shown for natural rubber [95, 158]. Similar dependences are also fulfilled for polyethylene [95, 352, 751]. It is the conditions of the action of pressure in polymers that are favourable for extended-chain single crystals – thermodynamically the most equilibrium kind of polymer crystals – to form. The rate of their formation under pressure is very high, so pressure can be thought to eliminate the kinetic hindrances of crystallization. Owing to this, crystallization proceeds near the equilibrium melting temperature characteristic of a given value of p. It can be considered to be established that in this case, as in the case of orientation [240, 277, 680, 681], crystallization proper is preceded by the formation of a mesophase (see Chapter 4). Extrapolation of the linear high-baric part of the dependences of Tm on p for p = 0 obtained for 1,4-cis-polyisoprene [95], and polyethylene [95, 751], gives values close to Tm0 for the respective polymer at normal pressure* and can be one of the ways to determine this value. Note that the melting temperature of crystals produced at high pressure and measured at normal pressure is equal to Tm0, which was shown for polyethylene [751]. Herewith, from the data of the dependence between the melting temperature and thickness of lamellae at different pressure, one can, using eq. (3.16), calculate the interface formation energy V e [158]. Effect of shape factor. As in the case of glass transition, changes of mechanical properties due to crystallization are the less, the larger the shape factor of a specimen is [113, 343, 643, 688, 689].
3.4
Effect of the molecular parameters of rubbers and the composition of rubber compounds on crystallization
Effect of molecular mass. Studies of the effect of molecular mass M and molecular-mass distribution MMD on crystallization of elastomers are not of systematic character. Considering the rise of melting temperature with M, it should be expected that at the same temperature of crystallization its rate shall be the greater, the higher M is (as a consequence of an * At the extrapolation from the values of p = 10–102 MPa the values of p = 0 and p = 0.1 MPa are indistinguishable.
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
83
increased supercooling 'T). This effect was indeed found for polysiloxanes [428, 429, 658]; the authors relate it, first and foremost, to the improvement of the nucleation conditions. According to the data of [572], the crystal-growth rate G is also observed to increase with M within the investigated range of M for 1,4-cis-polyisoprene. It is possible, however, that changes of M affect the nucleation and growth rates in opposite directions. A decrease of the molecular mass in milling contributes to the decrease of the crystallization rate [275, 699]; however, in this case, an even stronger impact than the decrease of M proper is rendered by the addition of oxygen in the points of chain rupture. This effect is sometimes overlapped by acceleration of crystallization under the action of a stress occurring in the specimen as the result of milling [275, 699]; still, after the rubber compound is heated up, the effect of crystallization deceleration is clearly pronounced. A decrease of the crystallization rate with M rising [35], observed for polychloroprene, is apparently due not directly to the rise of M, but to an increase of the macromolecular-chain branching, whose effect proves predominant. Thus, the fact that it is difficult to get unequivocal conclusions on the effect of M on crystallization is due to the difficulties of obtaining narrow fractions of polymer without changing the structure of the macromolecules, without specific end groups, and the same MMD [192, 436]. Effect of the microstructure of a chain. Of the parameters characterizing the polymer’s microstructure, the most important from the point of view of crystallization is the regularity of chains. The higher it is, the higher the rate and ultimate degree of crystallinity Cf are. For a stereospecific polymer, the capability of crystallization is determined, first and foremost, by the configuration of chains. It is known that trans-polymers crystallize much faster than cis-polymers; the values of Tm and Cf for trans-polymers are higher than for cis-polymers. Classical examples are natural rubber and gutta percha, i.e., 1,4-cis- and 1,4-trans-polyisoprenes. Polychloroprene having the highest melting temperature among elastomers is a trans-polymer. However, the presence of a small amount of trans-units in cis-polymer hinders crystallization [95, 96, 338, 398, 452, 501, 516, 625, 773, 774]; what is more, the increase of trans-unit content during the modification of polyisoprene (NR) leads to a linear increase of log W1/2 [398]. The difference in the crystallization of NR and synthetic 1,4-cis-polyisoprene SKI-3 is partially due to the different content of 1,4-cis-units. Improved conditions of synthesis and especially the use of new catalytic systems [2, 420], in particular, neodymium systems, made it possible to produce synthetic 1,4-cis-polyisoprene with an extremely high 1,4-cis-unit content, which determined its high crystallization rate. The same catalytic systems are used to produce polybutadiene with an extremely high 1,4-cis-unit content; as a consequence, it also has an extremely high crystallization rate (see Fig. 3.1b) [282, 359, 360, 659, 660, 761]. Polybutadiene containing the same number of cis- and trans-units does not practically crystallize; the rubber SKB containing 1,2- and 3,4-units does not crystallize, either; the capability of crystallization is decreased in the butadiene rubber of lithium polymerization (SKDL). Thus, the rate of crystallization can serve as a measure of stereoregularity of rubber’s microstructure [516]. Crystallization of different types of polyisoprene and polybutadiene is discussed in more detail in Chapter 7. Electron-microscopy studies of the dependence of the growth rate G of 1,4-cis-polyisoprene lamellae on the content of 1,4-trans-units in specimens have shown [19, 371] that there exists a linear dependence between log G and the content of these units; an increase of the content from 0 up to 0.1 (i.e., up to 10%) makes G decrease thousands of times.
84
CHAPTER 3
Similar data [599] were obtained on the effect of other types of regularity disturbances and impurities on G; all those data form a unified dependence. Comparing these data with those on the effect of trans-unit content on the overall crystallization rate v, which adds up from the nucleation and growth rates, one can conclude that the growth rate G is much more sensitive to the change of conformation of the molecular chains, as to the occurrence of other types of irregularity, than the nucleation rate. The same conclusions were arrived at in the treatment of data on the reorganization in melting [772–774]. Crystallization of copolymers. Crystallization of rubbers, which are copolymers or contain different-structure side units (which could also be considered as copolymers) differs from crystallization of homopolymer the stronger, the larger the number of statistically distributed units of homopolymer is; also, copolymers usually crystallize at a lower rate than homopolymers. An example can be chloroprene rubbers of different structures, ranging from those intended for fabricating adhesives with the maximum rate and ultimate degree of crystallization to copolymers specially designed to reduce the crystallization rate (of the type of neoprene WRT). A decrease of the crystallization rate is also achieved in copolymerization of 1,4-cis-polybutadiene with 1,4-cis-polyisoprene (rubber SKDI) [167, 282, 525]. Crystallization deceleration is the more efficient, the closer the distribution of isoprene units to the statistical distribution is. The occurrence of side groups usually also slows down crystallization, especially in the case of their statistical arrangement along the chain. However, for polysiloxanes it has been shown that in the region of small content of side groups the values of the crystallization rates and Cf pass through a maximum (and, respectively, the crystallization half-time W1/2 through a minimum) [123, 494, 519]. A significant role is also played by the distribution of units along the chain. At a statistical distribution, a decelerating effect is maximal, at a deviation from statisticity the copolymers are produced at a higher crystallization rate, which was clearly shown for the silicon rubber SKTFV-803 [457]. An interesting example of a change of the capability of crystallization with the change of the content of side groups is crystallization of ethyl siloxane rubbers. Its rate decreases as the content of ethyl groups goes up. But then for the polymer with the predominant content of these groups it rises again, in the region of higher temperatures [64] (this, in a sense, is similar to the results of changing the trans-unit content in the chain of diene polymers). Formation of the liquid-crystalline phase was found for ethyl siloxane homopolymer [251, 253–255]. Below we consider a specific case of crystallization of copolymers, which include short segments of the chain of rapidly crystallizing homopolymers; this case is characterized by the formation of microcrystals. Crystallization of polymer blends. During the crystallization of blends of a crystallizing rubber and a noncrystallizing rubber, the system is diluted, and crystallization slows down, even in the case of incompatible rubbers, the more so, the higher the content of the additive in the blend is; herewith, the temperature of the maximum crystallization rate T1 remains practically constant, and the character of melting does not change (Fig. 3.8, curves 1 and 2) [96, 708, 710]. A decrease of the crystallization rate in blends is, evidently, related to the crucial role of homogeneous nucleation in small particles of the crystallizing rubber; this was shown for model blends of NR and styrene-butadiene rubber [85]. However, if crystallizing and noncrystallizing rubbers are compatible, e.g., at high temperatures, and the formation of the vulcanizing network prevents their separation in cooling, as it takes place for blends of 1,4-cis-polybutadiene and styrene-butadiene rubber
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
85
ΔT'
1
2 3
-100
-60
-20 T', °C
Figure 3.8 Melting thermograms obtained by the DTA method for vulcanized specimens of the rubber SKD (1), its blend at a ratio of 1:1 with SKI-3 (2) and with styrene-butadiene rubber SKMS-10 (3); crystallization temperature T = –55°C, crystallization time t = 48 h.
with small styrene content, the melting character changes (Fig. 3.8, curves 1 and 3) [96, 708]. When a noncrystallizing and incompatible rubber in the blend proves to be in the glassy state, this rubber can behave as a low-active filler [96, 107]. If two components in the blend are crystallizable [466], they can both suppress and enhance each other’s crystallization. An example of the system where rubbers are incompatible and mutually slow down crystallization are blends of 1,4-cis-polyisoprene and 1,4-cis-polybutadiene (SKI-3 and SKD) [96, 457, 522, 707, 708, 710], widely used in practice for manufacturing low-temperature-resistant rubbers. Within a broad range of ratios of the components, this system has two crystallization-rate maxima coinciding (to an accuracy of up to 1°C) with maximum crystallization rate temperatures of the rubbers in the blend. Independent crystallization of these two rubbers in the blend is also supported by the melting character (see Fig. 3.8). The presence of a noncrystallizable third polymer in the blend suppresses the crystallization of the specimens even more [709]. In the case of a blend of a slowly crystallizing rubber and its rapidly crystallizing chemical analogue, e.g., upon addition of trans-polymer to cis-polymer (polyisoprene, polybutadiene [420]), crystallization is performed mainly at the expense of the rapidly crystallizing trans-isomer. The nucleus-forming action of the rapidly additive is observed only if its crystal lattice is the same as in the slowly crystallizing elastomer, which was shown for blends of chloroprene and also isoprene rubbers of different regularity. Effect of the three-dimensional network. Formation of crosslinks during the vulcanization disturbs the regularity of polymer’s chain, so an increase of the network density leads to the same changes of crystallization parameters as other disturbances of regularity. In the general case, as for other disturbances of regularity [527], W1/2 changes with a network density increase along the curve with a minimum, which is especially pronounced for rubbers from polysiloxanes [123, 493, 518, 560]. For rubbers from other crystallizable polymers, this minimum is shifted to the region of really used networks only in the case of stress-induced crystallization (see Chapter 4). An increase of the network density leads to a suppression of reorganization in melting (see Fig. 3.4) [103, 774]. The effect of crosslinks on
86
CHAPTER 3 τ1/2, h
1
600
400
2
200
0
1
2
3 1/Mc.104
Figure 3.9 Dependence of crystallization half-time W1/2 on network density determined by parameter 1/Mc for rubbers based on NR with poly- and disulphide (1) and monosulphide and C–C (2) bonds.
crystallization can be considered as the result of the emergence of a prohibited volume around the links, crystallization in whose volume is impossible [230, 398]. Not all types of vulcanizing groups slow down crystallization similarly efficiently (Fig. 3.9). Well back the first works by Bekkedahl and Wood [69a] have shown that an increase of the sulphur content sharply slows down isothermal crystallization of unfilled rubbers based on NR. Treatment [95, 96] of the data obtained by Russell [612] for a large series of NR-based rubbers with different network type and density, as well as the subsequent experiments, suggested [95, 96] that for rubbers from NR, SKI-3 and SKD the most efficient increase of W1/2 and decrease of Cf are achieved using vulcanizing groups leading to the formation of polysulphide bonds. Partially, this effect occurs as the result of modification of the polymer chain in the presence of accelerators. A rather efficient way of slowing down crystallization in NR, SKI-3 and SKD is to use vulcanizing groups containing sulphur with sulphamide accelerators. In these cases, W1/2 can increase tens and hundreds of times, and Cf decreases 1.5–2 times [85, 95, 96, 110, 111, 263]. Crystallization is slowed down much less by the networks formed in vulcanization by tetramethyl thiuram disulphide (thiuram) without sulphur or with its minor content (0–0.5 phr), as well as by peroxide, i.e., those formed predominantly by monosulphide or C–C bonds [85, 95, 96, 110, 111, 215, 263]. According to the data of [398], log W1/2 rises with the number of C–C bonds increasing. An increase of the number of units modified the same way as in vulcanization also results in a linear rise of log W1/2. However, the efficiency of this modification is always lower than in the case of an increase of density of C–C bonds, i.e., the very low-efficient crosslinks for slowing down crystallization [398, 527]. This implies, apparently, that the difference in the effect of various types of networks on crystallization is not reduced only to the changes related to disturbances occurring when using various vulcanizing groups [103]. The difference in distribution of disturbances along the chain is also a possible source of this effect. An attempt to take into account the modification of the chain when assessing the effect of vulcanization on crystallization was made in [385]. Differences in the effect of various vulcanizing groups on crystallization also manifest themselves in the change of the melting character [103, 109, 110] at the same density of the three-dimensional network. However, Tm0 depends little on the density and type of
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
87
crosslinks [103, 110, 111]. At the statistical distribution of the disturbances along the chain this weak dependence obeys the Flory equation (3.15c). Electron-microscopy photographs obtained by Andrews show that the junctions of the vulcanizing network are not in single crystals, but can be in polycrystals [14, 18]. A decrease of grain size with the threedimensional network density increasing can be observed using a light microscope [95, 96]. The same conclusions are suggested by the consideration of vulcanizates as copolymers of the regular part of the chain and the part with disturbed regularity [334]. The change of length of the crystallizing sequences in a vulcanizate as a consequence of the presence of the junctions themselves and other regularity disturbances results in a change of the melting character, in particular, to the suppression of reorganization during the melting [103]. Effect of fillers. Literature data on the effect of filling on crystallization of elastomers are contradictory. One can find information on both acceleration or deceleration of crystallization in the presence of a filler, and on the independence of the crystallization rate on its content. Crystallization and melting studies of elastomer systems with different character of polymer–filler interaction have shown [626, 627] that this is due to the different effect of a filler on the nucleation and growth rate and, therefore, on the different stages of the crystallization process. Therefore, different results can be obtained, even studying the same systems by methods with different characteristic time, e.g., dilatometry and DSC [629]. The kinetic curves for filled elastomers are described by eq. (3.3); the values of n decrease with filler content rising and at fillings of about 30–40 phr usually n = 1 [95, 96, 151, 231, 232, 448, 527, 575, 626a, 627]. Herewith, the induction period can be absent on the kinetic curves, as its time becomes comparable with the time constants of the methods used. At large fillings, when, as the result of a nucleating action of a filler, crystallization partially proceeds during the cooling, and only its final part occurs under isothermal conditions, the curves are not described by eq. (3.3) with one value of n; often n < 1 [626]. Nonisothermicity of the crystallization of filled systems should be taken into account in the analysis of the data on the kinetics of their crystallization, obtained by various methods. Generalization of the studies of filled systems based on butadiene, isoprene, chloroprene, silicon rubbers, as well as trans-polypentenamer [626a, 627–629] has shown that the effect of a filler on crystallization of elastomers manifests itself in three major directions: (1) increase of the nucleation rate in the presence of filler particles; (2) orientation of polymer molecules at the interface with the filler; (3) decrease of molecular mobility in the polymer layer adjacent to the filler’s surface. The stronger the interaction of the filler and polymer, the stronger its effect on crystallization is; the effect leading both to the acceleration of nucleation and to slowing down crystal growth. This conclusion is also valid for other polymers [409, 410, 434]. A change of content and type of filler can change the ratio between the nucleating and crystal growth-inhibiting effect of the filler; as the result, the overall effect can slightly change. Thus, as a small content of a filler (2–10 phr) and at a very large content (over 60 phr) crystallization is predominantly slowed down; at an intermediate content, it is insignificantly accelerated (W1/2 decreases), the acceleration being the larger, the more active the filler is [380a]. However, for nonstressed rubbers (and at a deformation H < 2) at an average content of a filler its effect on W1/2 is not large* and is more pronounced for rubbers with polysulphide bonds [626, 628]. * Internal stresses due to the introduction of large amounts of filler can lead in some systems to a change of the character of observed phenomena.
88
CHAPTER 3 σ, MPa 2'
1 2
4.0
2.0
0
1'
3'
3
1
3''
2
2''
log τ1/2, min
Figure 3.10 Dependence between stress in a specimen before crystallization V and log W1/2 for rubbers from polybutadiene (SKD) with different contents of carbon black (phr): 0 (1, 1c), 30 (2, 2c, 2cc), 60 (3, 3c, 3cc); crystallization temperature –50°C (1, 2, 3); –40°C (1c, 2c, 3c) and –30°C (2cc, 3cc).
Suppression of crystallization at superlarge doses of an active filler is of the same nature as amorphization observed in thin films [410, 434], and is stipulated by the comparable thicknesses of the rubber layers between filler particles with the critical size of the nucleus [411, 586, 587], suppression of the molecular mobility in the thin layer of polymer at the interface with filler (which is registered by the NMR method [213, 415, 416]), as well as the concentration of the part of polymer, capable of crystallization directly on its surface [626a]. The change of structure of elastomer on the surface of a filler is also indicated by the data obtained for 1,4-cis-polyisoprene filled with active carbon black. Using X-ray diffraction analysis, it was shown that regions characterized by different densities of polymer-chain packing can be singled out in a specimen at room temperature: a region of the nondisturbed polymer matrix with density of 903 kg/m3, a region of increased-density rubber 990 kg/m3, and that of a smaller density 890 kg/m3 [379]. Herewith, the more active carbon black, the closer the density of densely packed regions to that of the crystals is. These data are consistent with the results obtained by the method of electron microscopy on the formation from solution of NR crystals oriented perpendicular to the surface of the active filler particle [738]. Based on the example of rapidly crystallizing polychloroprene (nairit NP), the occurrence of orientation of crystalline formations at the interface with active filler was shown [95, 96, 102, 106]. Comparison of the morphology of crystalline formations of oriented and filled elastomers, of the character of the crystallization kinetics change, as well as of parameter n in eq. (3.3) made it possible to draw an analogy between the action of filling and stress on crystallization and to conclude that the effect of filling is reduced to the establishment of an additional stress, in the field of which a stress-induced crystallization occurs [95, 96, 106]. Indeed, if stressed rubbers are crystallized, a filling always results in an increase of the overall rate of crystallization (decrease of W1/2), so that the characteristic straight lines of the rubbers containing various amounts of filler form a fan (Fig. 3.10). This effect is the stronger, the higher the sulphidity of elastomer’s crosslinks is. The more active the filler, the stronger parameter B in eq. (3.17) changes, i.e., the faster the rubber will crystallize at the same deformation. However, the body of the data on the effect of the type and content of fillers on crystallization indicate that its mechanism could not be reduced to the action of an additional stress. This, first and foremost, follows from the melting data of filled elastomers [95, 96, 106, 109, 111].
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
89
In accordance with the views on the effect of stress on melting temperature, one should expect its increase for filled elastomers. At the same time, thorough studies of the melting of filled composites based on polyisoprene, polybutadiene, polychloroprene and polysiloxanes with various fillers have shown that both the melting character and value Tm0 are usually preserved the same as in unfilled specimens. Some changes in the character of the melting peaks are observed for vulcanizates at the introduction of active types of carbon black and Aerosil: the peaks themselves slightly broaden, and the height of their maxima decreases, so that the area of a peak practically does not change. However, this effect can be due, primarily, to the interaction of the vulcanizing system components with a filler [109]. A noticeable change of the melting peaks, which is expressed mainly in their broadening, was observed at the introduction of asbestos [629], but systems with asbestos are characterized by thermoelastic stresses in the elastomer at the interface with a filler [784]. The literature data on the melting temperature change are in reality related to the fact that crystallization of filled elastomers is nonisothermal. Acceleration of nucleation as the result of introducing a filler leads to a situation when a filled polymer begins to crystallize during the cooling at higher temperatures than an unfilled polymer. As the result, since the melting temperature Tm2 depends on the crystallization temperature T, a filled polymer may have a higher Tm than an unfilled one; and the other way round, if crystallization is observed during the heating, Tm of a filled elastomer can be lower than that of an unfilled one [626a, 627, 629]. This, apparently, is the major reason for the ambiguity of the data on the effect of filling on the melting of elastomers. Thus, the specific features of the effect of fillers on crystallization are determined by the formation of a layer with changed properties at the boundary with the filler. Presumably, the layer concentrates crystallization nuclei in it, and in noncrystallizable polymers an increased concentration of physical junctions is observed on the surface of filler particles. These junctions represent the prephase fluctuations of high stability but of a size smaller than the critical size of the crystallization nuclei. Apparently, it is these physical junctions that contain the “rigid phase” (according to Mullins [495]) of the rubber on the surface of a filler. The presence of this phase is well registered by mechanical methods [329, 770]. The possibility of forming an increased-density layer (called “pseudoglassy”) on the surface of filler particles was shown during the measurement of a thermal expansion of elastomers based on noncrystallized rubbers filled with asbestos fibres [784]. Studies of the recent years, both experimental [216] and theoretical [763], confirm the presence of a layer with changed properties on the surface of filler particles. Unfortunately, the effect of nanofillers on the crystallization of elastomers has not been studied. Meanwhile, data for rapidly crystallizing elastomers are indicative of both a strong nucleating action of nanofillers [500] and of the possibility of slowing down crystallization in the presence of nanofillers [299]. A significant role, apparently, should be played not only by the size of particles but also by their interaction with the elastomer matrix. This conclusion is suggested by the analysis of the effect of elastomer crystallites themselves on their mechanical properties [124]. The effect of silica on crystallization has not been studied, it could only be assumed to be similar to the effect of active carbon black. It is not clear if the peculiar features of crystallization of the boundary layers in filling with silica, studied for polyethylene glycols [134], are characteristic of crystallization of elastomers. Effect of plasticizers. The effect of plasticizers on crystallization is primarily determined by their nature. The traditional approach to plasticizers as products decreasing the crystallization rate is wrong.
90
CHAPTER 3
Crystallization studies of four elastomers (NR, 1,4-cis-polybutadiene, polychloro0 changes linearly with the prene and polyurethane) made it possible to establish that W1/2 plasticizer content rising. Herewith, the stronger a plasticizer decreases Tg, the less it slows down crystallization (or the more accelerates it) [96, 174, 175]. In contrast to filling, introduction of a plasticizer leads to the deterioration of the nucleation conditions and acceleration of the growth of polycrystals, so that the overall effect of introducing a plasticizer is determined by which of these effects prevails.* Therefore, an increase of the overall rate of crystallization is observed for those plasticizers, introduction of which leads to the strongest change of macromolecular mobility, i.e, to the strongest decrease of Tg (and the other way round) [174, 330]. Comparison of the effect of plasticizers on crystallization of various-composition blends (filled and unfilled, vulcanized and nonvulcanized) shows that a change of composition of a blend does not lead to a significant change of the character of the effect of a plasticizer on crystallization [174–176]. Attempting to improve the low-temperature resistance by introducing a plasticizer, one should take into account possible changes related to acceleration of crystallization. Thus, for polychloroprene, with a plasticizer introduced with the aim to increase low-temperature resistance (nairit M), Tg was observed to decrease and, therefore, short-term low-temperature resistance to improve. But in this case crystallization is accelerated, and as the result the long-term low-temperature resistance decreases. A deceleration of crystallization in the presence of some other plasticizers is used for preventing crystallization of NR in storage. The data on the effect of various plasticizers on crystallization of polychloroprenes of different types have been studied in detail by example of chloroprene rubbers manufactured by Bayer [473, 742]. In the presence of plasticizers, as of other solvents, the melting temperature decreases too. The performed measurements [96, 174] do not make it possible to determine the true equilibrium melting temperature Tm0, but the change of ultimate temperature (Tm0 c ) obtained by an approximate method of extrapolation [174], is satisfactorily described by the Flory–Higgins equation:
1 1/ Tm0 0
1/ Tm0
ª¬ RV / 'HV1 º¼ Q1c FQ1c2 , ,
(3.20)
where R is the gas constant; V, V1 are the molar volumes of the monomer unit of polymer and plasticizer; vc1 is the volume fraction of a plasticizer in the system; F is the Flory–Higgins constant; (Tm0 c)0 is the equilibrium melting temperature of polymer without plasticizer; 'H is the melting heat. Substitution into eq. (3.20) of values of F for the polymer–plasticizer system, obtained from the data on swelling, and Tm0 c instead of Tm0, gives a good match of the calculated dependence of Tm0 c on vc1 with the experimental value [96, 174, 175]. Using eq. (3.20), one can also determine the melting heat 'H by the change of Tm0 c, which leads to the results matching with those obtained by other methods. * It is well known that for swollen networks the stress–strain curves are well described by equations following from statistical theory (eqs. (6) and (7)), i.e., in eq. (9) C2 = 0 [97, 715] and the deviations, which could be related to the formation of physical junctions as prephase fluctuations, are not observed. This is well consistent with the deterioration of nucleation conditions in these systems. In a number of cases, for such networks this effect proves prevailing, and a decrease of their capability of crystallization is observed as compared with nonswollen networks.
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
91
dQ/dt 5
6
7
4
2
3 1
2
4
I II
endo
1
-100
-50
0
50
100 T', °C
Figure 3.11 DSC thermograms (dependence of heat flow dQ/dt on scanning temperature Tc) for specimens of ethylene-propylene rubbers with propylene content w, mol. %: 43 (I) and 48.6 (II) without isothermal holding (1), after a holding for 24 h at 293 K (2) and 40 min at: 223 K (3), 233 K (4), 250 K (5), 268 K (6) and 273 K (7); heating and cooling rates, 80°C/min.
According to eq. (3.20), a decrease of the melting temperature depends on the volume content of the solvent in the compound and on the value of the thermodynamic parameters of the polymer–solvent system. A larger decrease of temperature should be expected in the case of a good solvent (smaller values of F) rather than a bad solvent. As the glass-transition temperature is observed, as a rule, to shift during the introduction of a plasticizer, the temperature position of the whole range of crystallization and maximum crystallization rate temperature T1 should change. Specific problems are associated with studies of the effect of crystallizable plasticizers on the crystallization of elastomers and on their glass transition [330, 380b, 592] (see Chapter 1). As in the case of glass transition, such studies are far from being completed.
3.5
Microcrystallization
Peculiarities of rubber microstructure can lead to a restricted growth of crystals owing to the small size of chain segments capable of crystallization. This takes place in the case of copolymers with relatively short crystallizing sequences of rapidly crystallizable polymers. Typical representatives of such polymers are copolymers of ethylene and propylene. The microcrystallization processes have been studied in detail, using them as an example. At a propylene content w from 25 up to 65 mol. %, these copolymers have the properties of elastomers, i.e., in this region of “average” compositions the capability of crystallization of both polyethylene and polypropylene is significantly disturbed. Earlier, these polymers were considered amorphous; however, studies of their low-temperature properties showed the presence of anomalies [95, 105, 132, 133, 156, 475, 532, 619, 620, 682]. Static relaxation properties of these copolymers change depending on preliminary holding time, and additional maxima are observed on the relaxation spectra [9, 10, 61]. These effects were associated with either a peculiar manifestation of the crystallization process [95] or with the breakdown and formation of physical junctions [9, 10, 44, 61] from sequences of one type, which by their structure are close to crystallization nuclei [10].
92
CHAPTER 3
However, no X-ray data on the crystallization of copolymers of ethylene and propylene of an average composition have been obtained. The use of differential scanning calorimetry (DSC) along with a set of mechanical properties made it possible to arrive at a conclusion that these anomalies are determined by a first-order phase transition. However, due to the small size of crystals and a low crystallinity their presence is not registered by an X-ray method. This process of forming a minor amount of small-size crystallites not registered by an X-ray method was called microcrystallization. Recall that a higher (than in the X-ray method) sensitivity of changes in specific heat and mechanical properties to the crystallization of elastomers is well known [95, 97]. Figure 3.11 [132, 133] presents DSC thermograms for two copolymers. Isothermal holding of the specimens at temperatures from –40°C to +20°C prior to scanning results in the appearance of pronounced endothermal peaks* on the thermograms, characteristic of the melting of crystallized elastomers. Herewith, the endothermal effect** is the larger, the smaller w is, i.e., the larger the content of methylene units is. This implies that microcrystallization of investigated copolymers is the result of forming small crystals from sequences of such units. Proceeding from this, by the ratio of heats of the endothermal transition ('Hc to the heat of fusion of ultimately crystallized polyethylene ('H), one can calculate the degree of crystallinity Cc. As w decreases within the investigated limits, the values of Cc rise from 1 up to 4%. In the further decrease of w, for instance, for copolymer with w = 17 mol. %, crystallization is registered by the X-ray method at room temperature; the degree of crystallinity being C = 12. For the same specimen, the value of Cc= 11.4%, i.e., the values of the degree of crystallinity calculated by the thermograms and X-ray patterns practically coincide [470]. What is more, the data obtained at different laboratories coincide too [132, 133, 470]. The Thomson equation (3.16) yields for a crystallite size l [132, 133]: l
2V eTm0 / Tm0 Tm 'H ,
(3.21)
(where Tm0, 'H and Ve are equilibrium melting temperature, heat of fusion and energy of interface formation). This made it possible to assess the crystal size l along the crystal axis c. If one uses the values of Tm0, 'H and Ve for polyethylene, then at T = 233 K l = 10–12 monomer units. This is indicative of the presence of sufficiently long segments of polyethylene chains in copolymers. The data of Fig. 3.11 show that an increase of temperature of the isothermal holding results in an increase of respective melting temperature, and, therefore, of l. Apparently, the distribution of the l values, corresponding to the distribution of Tm of microcrystals reflects the distribution of the lengths of crystallizing sequences in copolymer; studying it enables a more exact characterization of the microstructure of such elastomers. Thus, for ethylene-propylene rubber the microcrystallization data indicate the presence of units containing 9–10 ethylene sequences. At the same time, the most perfect
* Anomalies of specific heat in the region of transition from the glassy to the rubberlike state were observed for specimens with smaller w even in the absence of isothermal holding (see Fig. 3.11), which is characteristic of crystallizable polymers. ** As anomalies of specific heat in the absence of holding.
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES K
93
a
1.0
3 2 1 0.5
-60
-40
-20
0
T, °C
K b
1.0
4'
3'
4''' 1' 1'''
0
4''
3''
0.5 3''' 1'' 20
0
80
0
20
40 t, h
Figure 3.12 Dependences of recovery K at compression deformation H = 0.3 for ethylene-propylene vulcanized rubbers on temperature T at holding time t = 5 min (a) and on holding time t at temperatures T, °C: –18 (1c, 3c, 4c), –25 (1cc, 3cc, 4cc), –40 (1ccc, 3ccc, 4ccc) (b); propylene content w, mol. %: 43 (1, 1c, 1cc, 1ccc), 47.5 (2), 48.6 (3, 3c, 3cc, 3ccc) and 49.8 (4c, 4cc, 4ccc).
NMR methods yield only the number of units containing more than 5–7 of such sequences, to say nothing of the IRS method, which makes it possible to take account of even shorter sequences. Therefore, the method of measuring the melting temperature of microcrystals can be used to obtain information of the true structure of copolymer. Microcrystallization, as common crystallization, leads to an increase of the modulus and hardness of elastomers. Thus, the presence of less than 3% of microcrystals (impossible to determine by an X-ray method) leads to a rise of hardness of the ethylene-propylene-based rubber from 50 up to 80 Shore hardness units. Relaxation properties change significantly, too, such as the dynamic modulus and loss modulus [620]. Changes of the relaxation properties in the process of microcrystallization can be also illustrated by recovery changes. Figure 3.12a shows the temperature dependences of recovery K measured at a holding for 5 min. By their shape, these curves are similar to typical curves for elastomers in amorphous state (see Chapter 1). An increase of the holding time leads to a decrease of K, and, therefore, of T, similar to how it takes place at an increase of the content and activity of a filler or crystallization of elastomers. A distinctive feature of the kinetic curves of the recovery change for these copolymers (see Fig. 3.12b) is the absence of an induction period [132, 133]. The dependence of melting temperature of microcrystals of ethylene-propylene rubbers on crystallization temperature is linear; the lines assigned to the different contents of crystallizing units in the copolymer are parallel [132, 133]. The melting temperatures of microcrystals formed at room temperature are | +60°C. This leads to the necessity of taking
94
CHAPTER 3
account of this process when considering the processing and some other properties of copolymers. Studies of the low-temperature properties of oxygen-containing fluoroelastomers also showed a possibility of forming microcrystals in these elastomers at low temperature [129, 131]. It was shown that microcrystallization in these elastomers develops mainly in the gel fraction. For these copolymers, too, the dependence of melting temperature of microcrystals on crystallization temperature is linear. Preliminary data show that the microcrystallization process is also characteristic of some other elastomers, representing copolymers, which include chain segments of rapidly crystallizable polymers. Special significance of the microcrystallization process is that it is a direct transition from common crystallization of polymers to the formation of physical junctions representing prephase fluctuations and, by their structure, similar to crystallization nuclei, but of a size smaller than the critical size of the crystallization nucleus. Crystallization of rigid blocks in block copolymers also has features similar to microcrystallization. An undoubtful analogy exists between microcrystallization and formation of the network of physical crosslinks in “physical” gels [76, 324]. In a sense, crystallization of elastomers near the glass-transition temperature Tg can be also called microcrystallization. Limitation of crystal growth in this case is due to the peculiarities of the crystallization kinetics (see above). Thus, transition from the microcrystallization of ethylene-propylene copolymers to common crystallization is reflected on the dependence of the glass-transition temperature of copolymers Tg on composition (see Chapter 6). The microcrystallization pattern and its distinction from common crystallization are related to the impossibility of the growth and reorganization of crystals during the melting, which is due to both steric hindrances and large supercooling 'T = Tm0 – T, where Tm0 is the equilibrium melting temperature of rapidly crystallizable homopolymer, whose chain segments are a part of the copolymer. As common crystallization, microcrystallization leads to an increase of the rigidity of elastomers, i.e., to the deterioration of their processing properties. Low-temperature resistance of rubber and rubber articles deteriorates (see Chapter 6). At the same time, the cohesion strength of rubbers is improved, which facilitates the use of elastomers capable of microcrystallization for fabrication of complex components.
3.6
Prediction of the crystallization rate
A comparatively low rate of crystallization of most elastomers, and the extreme character of its temperature dependence prevent direct experimental studies of the process. The basis of the methods of its prediction is the possibility of accelerating it under the action of a stress, and also the use of Kolmogorov–Avrami equations (3.3) and the temperature dependence of the crystallization time in the form of eq. (3.4ccc) [116, 126, 214]. Prediction by stress. An accelerating effect of a stress on crystallization is described by eq. (3.17). Having obtained the kinetic curves at two or three values of stress*V (e.g., by the recovery method at compression deformations H equal to 0.3, 0.5 and 0.7), one * V is the static stress prior to the onset of crystallization (at room temperature) referred to the true cross-section area of a specimen.
CRYSTALLIZATION OF ELASTOMERS AT LOW TEMPERATURES
95
determines the values of W1/2 and plots the data in coordinates V – log W1/2 (a characteristic straight line). Extrapolating it for required values ofV or forV = 0, one obtains the required values of W1/2. It is important to remain within the values of V, for which eq. (3.17) is valid. The results can be obtained tens and hundreds of times faster. 0 , one can findW required to achieve any Prediction by time. Knowing the value of W1/2 x other power x of the crystallization process, using eq. (3.3). For common technical rubbers with large filling, one can take n = 1; for unfilled rubbers in the absence of deformation, n = 3. Dependences of n on filling can be used [626a]. As a consequence of an indefiniteness of the value of n, this stage of prediction is the least accurate (| Prediction by temperature. Equation (3.4ccc) is used for prediction by temperature. The first prediction system was based on the assumption of the constancy of all constants in the equation except the equilibrium melting temperature Tm0. It was also assumed that activation energy U does not depend on crystallization temperature and D = 2; the values of 0 at the maximum crystallization rate temperature T [95, Tm0 are related one-to-one with W1/2 1 116, 233]. Considering the independence of parameter B in eq. (3.17) on temperature, this approach makes it possible to determine W1/2 at required V and T, knowing the values of W1/2 at T1 and two or three more values of V, i.e., under extreme conditions [95, 96]. The accuracy of prediction improves if, instead of the assumption that U = const, we use the temperature dependence of U in the form of eq. (1.8) and the value of D = 1 [92, 93, 96, 126, 246, 437, 533–535, 586, 751]. Yet another system [380, 450] is based on the same equations but assumes the direct experimental determination of Tm0. As not real but some approximate values of Tm0cc were determined, this does not yield a large improvement in 0 at T by a more prediction accuracy, but replaces the bulky experiment for determining W1/2 1 0 compact determination of Tm [380, 450]. At a change of composition, Tm0cc is assumed to change [92, 93]. The above experimental data on the study of the melting temperature range for crosslinked and filled elastomers, which showed an independence of the true equilibrium melting temperature Tm0 on the type and density of the three-dimensional network, as well as the type and content of filler, make the core of the new system of prediction. It is based on eq. (3.4ccc), where U is described by eq. (1.8) andD= 1. It is assumed that the effect of composition can be mainly assigned to constant C. This system can make it possible to completely abandon experiments with concrete rubbers, which are a must when using the earlier developed approaches.
96
CHAPTER 3
4
Stress-induced Crystallization of Elastomers
As we noted in Chapter 3, transition to large deformations significantly changes both the crystallization kinetics and morphology of crystals formed. It is stress-induced crystallization that determines the character of the strength properties of crystallizing elastomers, in particular, the high tensile strength of elastomeric materials from isoprene, chloroprene, urethane and butyl rubbers, as well as ethylene-propylene copolymers. Attempts to produce high-modulus and high-strength materials, from elastomers including, are related exactly to this phenomenon. To achieve this aim, it is necessary to develop elastomeric materials with an optimal structural organization.
4.1
Kinetics of stress-induced crystallization
As shown in Chapter 3, crystallization is observed to proceed faster under the action of a stress; herewith, the kind of deformation is of no crucial importance. However, starting with some tensile deformation Hk, the pattern of the crystallization kinetics is changing. This value can be conditionally defined as a deformation at which in the isothermal crystallization kinetic curves for unfilled rubbers the induction period vanishes, parameter n in eq. (3.3) approaches* unity and eq. (3.17) ceases to be true. Starting with Hk, crystallization can be considered to be predominantly stress-induced. Major part of this crystallization occurs either directly during stretching or in cooling, i.e., does not meet the isothermality conditions**. Therefore, the kinetic curves of stress-induced crystallization at H > Hk (or O = Ok) not only fail to have an induction period, but the initial part of the curve often can not be obtained at all. For not too large H, such kinetic curves are similar to those obtained for filled rubbers (see Chapter 3). Kinetic curves of stress-induced crystallization for NR, when orientation is accompanied with subsequent cooling, are given in Fig. 4.1. As concerns the character of change of parameter n with H rising, the data by various authors obtained under different conditions do not coincide. Thus, according to [335, 685, 686], the change from n = 3 or n = 4 atH = 0 to n = 1 at H > Hk is gradual (curve 5, Fig. 4.2a), while by the data of [267, 268] n = 1 within the entire region of stress-induced crystallization (curves 1–4, * Equation (3.3) can be changed as applied to stress-induced crystallization [685, 686]. ** In the former case, due to the continuous change of Tm0, and, therefore, due to an increase of supercooling 'T.
CHAPTER 4
98
log |E*|, MPa
ΔV/Va, % a
c
1.5 1
0 3
1
1.0
4'
2 3'
0.5
4
1
2 0 102
104 log t, min
103
1
0
log t, min
1
C, % b
36 28
4
20 12
1
4 4
12
8
20
16
24 t, h
Figure 4.1 Kinetic curves of a two-stage stress-induced crystallization of vulcanized NR at –25°C (a, b) and –15°C (c), obtained by the relative changes of volume, 'V/Va (Va is the volume of an amorphous specimen) [335] (a), specific heat recalculated for the degree of crystallization C [267] (b) and dynamic modulus |E*| [217] (c) at elongations H : 0 (1), 0.5 (2), 1 (3), 1.31 (3c), 2 (4), 2.64 (4c). n
C,% b
a
2
15 4 3 2 1
2
4
3
4
5
6 λ
4'
30
5
5
2-5 1
3
10
1
C,% c
20
7
10 3
4
5
6
λ
6
1
5
10
15 λ
Figure 4.2 Dependence of parameter n in eq. (3.3) by the data of [267, 335] (a) and degree of crystallization C by the data of [733] (b) and [387] (c) on the ratio O for vulcanized NR at temperatures T, °C: –25 (1), 5 (2), 15 (3), 25 (4) and 44 (5) (a, b); vulcanized rubbers: NR (4cc), SKI-3 (6 ) and SKIL (7) at T = 25°C (c).
Fig. 4.2a). This difference can be due to various factors. Besides the different inertiality of the methods used, differences in the chemical characteristics of the systems studied can contribute, too. However, the most significant factor is the temperature regime of deformation. Thus, if a specimen is stretched near Tm0 and then cooled to a temperature T < Tm0, at which the crystallization kinetics is determined (two-stage stress-induced crystallization
STRESS-INDUCED CRYSTALLIZATION OF ELASTOMERS
99
[265, 267, 268]), n gradually decreases with H rising. A gradual decrease of n under these conditions was observed for NR [95, 96, 230, 335, 685, 733], polybutadiene and polysiloxanes [95, 96]. If crystallization is registered at the same temperature near Tm0, at which the stretching was performed [267–269] (stress-induced crystallization proper), the fragment of the kinetic curves at H = const is always described by eq. (3.3) with n = 1. Experimental data on the effect of stress on the degree of crystallization are not unequivocal. As a rule, they indicate that the ultimate degree of crystallization C of oriented and then cooled specimens (two-stage stress-induced crystallization) is practically the same as C at H = 0 (see Fig. 4.1) [95, 96, 217, 230, 239, 335, 392, 585, 733]. The data that the ultimate degree of crystallization C achieved under isothermal conditions decreases withH rising [259, 447, 685, 686] are related, first and foremost, to the specific features of the kinetics of stress-induced crystallization, namely, to the impossibility of registering the initial stages of the process at large H. Gent and Zhang [239] compared the kinetics and degree of crystallization of three elastomers – NR, butadiene and butyl rubbers – using weakly crosslinked and then cooled specimens. The values of C obtained by various methods (by stress relaxation, dilatometry, DSC and X-ray method) support these conclusions. A different situation arises when the degree of crystallization C is measured directly in stretching at a temperature close to Tm0 or in subsequent holding at the same temperature (stress-induced crystallization proper). These measurements carried out by the X-ray method* [25, 139, 239, 260, 315–318, 422, 711, 712], by heat release [163, 164, 165a, 167, 239, 246., 247, 267–269], changes of birefringence in the infrared [75] or visible [164, 582, 585, 695, 764] regions show that C is the greater, the larger H orO is (see Fig. 4.2b,c). Herewith, an increase of H or O indicates not only an increase of the crystallization development time but also an increase of supercooling 'T, as Tm0 rises with the deformation increasing. However, the ultimate values of C in isothermal holding under these conditions proves impossible to achieve as the development of crystallization at H = const is accompanied with stress relaxation [230, 239, 758, 758a] (see Chapter 3), the supercooling decreases and crystallization slows down.** The differences between the values of C measured during the stretching and Cv are also due to the morphological factors. The shape of the C– O curves obtained by different authors under these conditions slightly differs, too. In all cases, starting with some valueOk (or Hk), C slowly increases, and then is observed to rise sharply. Some works also single out a third fragment of the curves, where C rises not so rapidly [32, 74, 75]. The occurrence of this fragment is associated with the need to rearrange part of the crystals formed during the stretching [75]. However, the works where C is registered some time after a given value of O is reached usually fail to find a third segment in the C– O curves. Here we come across the problems of comparing different experimental data on stress-induced crystallization. These problems are due, first and foremost, to the conditions of stretching and registering the value of C. To describe the dependence of C on O, the Flory equation is often used [200]:
1 C
1/ 2
ª¬3 / 2 M O º¼
>3 / 2 T @1/ 2 ,
(4.1)
* The absolute values of C obtained at the early stages of NR stress-induced crystallization studies by the X-ray method [315–318, 422] are overestimated because of the particularities of the method of X-ray pattern treatment used. ** This is true at T |Tm0, i.e., at the right-hand side branch of the Tamman curve (see Fig. 3.3).
100
CHAPTER 4
where
M O «ª ¬
1/ 2 º
6 / S O Ns
T
ª 2 º »¼ «¬ O / 2 1/ O / Ns »¼ ;
'H / R 1/ Tm0 1/ T ,
(4.1c)
(4.1cc)
where Ns is the number of statistical segments; R is the gas constant; 'H and Tm0 are the heat and the equilibrium melting temperature; T is the temperature of the experiment. Equation (4.1) gives a practically linear relation between C andO. Göritz and Müller [267, 268] think that, in principle, it pertains only to the values obtained in isothermal crystallization. By singling out this part, they did obtain a linear C– O dependence. However, the coincidence with the experiment is achieved only using unreal values of Tm0 of a nondeformed elastomer. While for NR the values determined in stretching are C < Cv, for butyl rubber where the rate of low-temperature isothermal crystallization is small the pattern is reverse: stress-induced crystallization proper can reach up to 50%, and the low-temperature crystallization, no more than 15% [95, 239, 442, 711]. That is, the differences in the kinetics and morphology are opposite. It is well known that for crystallizable elastomers the stress–strain curve is determined by stress-induced crystallization. The determination of the degree of stress-induced crystallization by the shape of the deformation curve is successfully used for rubbers based on NR [715], polybutadiene [463a, 694], polysiloxanes [764], butyl rubber [239] and polychloroprene [474, 762]. The most convenient and exact way of registering the onset of stress-induced crystallization is by the deviation from linearity of the deformation curve obtained under equilibrium conditions and treated in Mooney–Rivlin coordinates (see Fig. 3). This technique can be also successfully used for treating the deformation curves obtained at a constant stretching rate. The onset of stress-induced crystallization can be also registered by the change of sign of the changes of volume in stretching (from the increase to the decrease) [332, 715]. Attempts to predict the shape of the stress–strain curve based on the crystallization data are scarce [332, 671]. It seems that it can be done, if one takes into account the formation of stress-induced physical junctions representing prephase fluctuations, which are formed at the initial stages of deformation and reach a critical size as it increases. Synchrotron X-ray radiation studies have shown the high orientation of the amorphous phase of both NR and synthetic polyisoprene before the onset of crystallization [711, 712]. A possible factor determining the features of orientation crystallization and, first and foremost, its abnormally high rate, is the formation of the preceding liquid-crystalline phase [240, 253].
4.2
Morphology in stress-induced crystallization
First of all, we shall note that the structure of an elementary cell is the same for an oriented and nonoriented elastomer, which was shown for NR [240]. The elastomer elementary-cell parameters are determined most often under stretching conditions. Most data on the morphology of crystals formed during the orientation of elastomers have been obtained [17, 18] by the method of electron microscopy using shadow contrasting
STRESS-INDUCED CRYSTALLIZATION OF ELASTOMERS
101
with osmium tetroxide. Some information could also be obtained by the method of light microscopy [95, 786]. Depending on deformation conditions, one should distinguish between two cases of formation of oriented crystals: (1) deformation of an amorphous specimen followed by its crystallization (deformation + crystallization); (2) crystallization followed by deformation (crystallization + deformation). The former case is of the greatest practical importance for elastomers operated at low temperatures in a stressed state. As applied to elastomers, especially to 1,4-cis-polyisoprene, this case has been studied the most completely [163, 164, 165a, 264–269, 335, 387, 478, 537, 585, 640, 653, 662, 685, 686, 695, 733, 739]. Data for polychloroprene are also available [17, 18]. But in its pure form it is realized only if a specimen is deformed up to H < Hk at T | Tm0, and then cooled (two-stage stress-induced crystallization). As we pointed out in Chapter 3, the morphology under these conditions differs little from the morphology of a nonoriented specimen. In the region of intermediate deformations, Hk < H< Hkc, “shish kebab” crystals are formed. Axial fibrillae consisting of crystals with extended chains are clearly registered here. And, finally, at some deformation H= Hkc the axial fibrillae become the major type of crystals and their formation is practically completed during the stretching of the rubberlike melt (stress-induced crystallization proper). The same morphology is also characteristic of other polymer crystals produced at the orientation of an amorphous melt [322, 323, 325, 758, 758a]. However, in most cases, due to the low viscosity of the melt, viscous flow predominates, and it is impossible to achieve the orientation required for axial fibrillae to form. Special techniques, which enable orientation of the amorphous melt of polyethylene (development of a weak three-dimensional network by irradiation or rapid cooling, i.e., pre-orientation quenching) provided for a possibility to obtain an oriented structure by the orientation + crystallization protocol [191].* To characterize the capability of stress-induced crystallization of a polymer, S.Ya. Frenkel and coworkers [191, 206–208] proposed to use the unfolding parameter (“unfoldedness”) of macromolecular chains:
E
h / L,
(4.2)
i.e., the ratio of the distance between the ends of a chain, h, to the contour length of the chain, L. Schematically, the macromolecule distribution function by the value of the unfolding parameter E is presented in Fig. 4.3a. In the absence of an external orientational effect, the unfolding values of flexible-chain polymer macromolecules are within the limits of 0 < E < 0.2. As molecular orientation increases, the distribution function is shifted towards larger values of E. For crystallization to begin according to the chain-unfolding mechanism, the value ofE should not be greater than unity [40, 41, 141, 191, 206–208]. The complete unfolding of the chains in the melt is impossible, and the length of an unfolded fragment could be considerably less than the length of a macromolecule. Intensive formation of axial fibrillae begins from some value of E =Ek, for all polymers Ek being approximately equal to 0.3, which is easy to achieve under orientation conditions. The deformation values Hk, at which E =Ek, are different for different polymers.
* Exactly this orientation technique is used in the orientation scheme proposed by Pennings, at an application of an intensive mechanical field to a polymer solution, in particular, in stirring [553]. As applied to elastomers, this route (orientation in solution) has not been in fact studied, though it takes place in fabrication and, in some cases, application of adhesives [155, 284, 286].
102
CHAPTER 4
w(β)
β
a
0.8 1
b
2
2
3
0.6
1
0.4 βcr 0.2
3
0 0.1 0.2 0.3 0.4
β
0
2
4
ε
Figure 4.3 (a) Distribution of molecules, w (E ), with respect to the unfolding parameter E at different molecular orientations: no orientation (1), average orientation (2), maximum orientation (3); (b) dependences ofE on the degree of stretching H for NR: unfilled milled NR (1) and NR filled with 30 (2) and 50 (3) (phr) of carbon black.
At E < Ek, only folded-chain crystals form, and at E >Ek, both folded-chain and fibrillar crystals. It is parameter Ethat determines the difference of natural and synthetic polyisoprene (NR and SKI-3) in stress-induced crystallization [141]. Thus, for NR the critical unfolding value E = 0.3 is achieved at a stretching Hk = 2, while for SKI-3, at Hk = 4–5. According to [40, 41], the unfolding parameter E can be determined from the formula
E
O/
2 / 3 Ns ,
(4.2c)
where O is the elongation ratio; Ns is the number of statistical segments between physical junctions; Ns = Mc /Ms, Mc is the molecular mass of a chain between physical junctions and Ms is the molecular mass of a statistical segment. Using 1,4-cis-polyisoprene as an example, it was proposed to determine Ms by the magnitude of the arbitrarily equilibrium modulus Ef* measured at the experiment temperature T prior to the onset of crystallization [644] (Mc was taken to be 116 [141]). Ms
3RT U / Ef *,
(4.2cc)
where U is the density of an amorphous specimen and R is a gas constant. Despite the predominant formation of axial fibrillae, at E >Ek (and at H >Hk) crystals with folded chains can exist, too. Besides direct morphological data (“shish kebab” crystals), the coexistence of two types of crystals is also indicated by the two-stage character of the crystallization kinetic curves observed in some cases [217, 323]. The occurrence of two melting temperatures for such systems is sometimes related to this, too. The coexistence of two types of crystals, which are characterized by the same lattice parameters but different packing structures, and, due to this, different values of Gibbs energy, was attributed by S.Ya. Frenkel [206] to a special type of crystalline topomorphism, which differs by the impossibility of direct transition from one structure to another. There are only regions of their coexistence.* * Introduction of the concept of topomorphism does not appear to be obligatory in this case. If it is to be used in consideration of stress-induced crystallization, then it should be also introduced for describing the difference of crystals formed in three kinetic regions of crystallization (see Chapter 3).
STRESS-INDUCED CRYSTALLIZATION OF ELASTOMERS
103
If orientation is not accompanied by cooling, the coexistence of two types of crystals can be due to the above-considered mechanism of stress relaxation in crystallization [757, 758]. A significant role is played by the molecular-mass distribution in the initial polymer, which, in turn, leads to different conditions of macromolecules’ orientation and different values of E(larger or smaller Ek). One of the causes of the coexistence of crystals with extended and folded chains is considered to be the stress-distribution inhomogeneity in the oriented specimen [266]. The existence of overstressed regions in a stretched specimen is indicated by considerations voiced by Patrikeyev [547]. If orientation is accompanied by cooling, the major mass of crystals are formed at low temperatures with folded chains, and extended-chain crystals serve as nuclei for them. The presence of the distribution of E and the nonuniformity of stress distribution in a specimen result in formation of axial fibrillae only in a part of material capable of crystallization during the stretching. This determines the difference of the degree of crystallization of elastomers in stress-induced crystallization proper (i.e., T | Tm0), and at the two-stage crystallization, when the orientation at T | Tm0 is accompanied by the subsequent cooling. Crystals formed by the former route of deformation, i.e., from the oriented melt, are characterized by high orientation, which does not practically change as H changes from 2 up to 4 [19, 387]. Herewith, the axis c of the crystals is oriented along the stretching direction [26, 315, 332]. The issue of what the axial fibrillae are shall be discussed below, in the analysis of the melting data of oriented elastomers. Keller [325], using rapidly crystallizable polymers, in particular, polyethylene, as an example, has shown that there is an interval of extrusion pressures and rates, inside which the formation of oriented axial fibrillae is facilitated. Taken together, the data on the abnormally high rates of formation of axial fibrillae and on the role of the mesophase in the formation of extended-chain crystals, suggests that this effect, as in the action of pressure, is due to the emergence of the mesophase. The latter route of stress-induced crystallization – crystallization + deformation – i.e., orientation of partially crystalline polymers, is of smaller importance for elastomers than the former. Until very recently, it was the major way to produce fibres from crystallizing polymers. It is its drawbacks that determined the need for the search of the possibilities of moving to the former route, which gives a big advantage in strength properties [191]. In elastomer systems, this route is realized for adhesive bonds obtained using cold-hardening (nonvulcanized) adhesives from polychloroprenes, as well as crystallizable coatings, e.g., polyurethane coatings. One should take into account that oriented crystals can form via this route under deformation of rubber articles crystallized in storage, as well as under deformation of nonvulcanized trans-analogues of elastomers that possess an extremely high rate of isotropic crystallization at room temperature: gutta percha, trans-polybutadiene, etc. As a rule, their deformation begins from a precrystallized state. Data on the deformation by this route as compared with the former route have been obtained for polychloroprene and polyurethane [18, 95, 96, 785, 786]. Special experiments were also conducted for specimens of polydimethylsiloxane [459] and 1,4-cis-polybutadiene crystallized at low temperatures. In this case, in the region of small deformations a specimen is deformed at the expense of the amorphous material at the boundaries between spherulites. Spherulites are extended into ellipses with the large axis coinciding with the direction of stretching [40, 41, 95, 96, 785, 786]. After the load is removed, the specimen contracts, and they acquire the previous shape. During the subsequent deformation, the spherulite structure passes into the fibrillar structure.
104
CHAPTER 4
The further deformation increase is due to amorphous intercrystalline regions, which are defects of the structure of such partially crystalline elastomers. Orientation of parts of macromolecules in crystals is close to unity; but the crystals themselves are not highly oriented, as well as macromolecules in amorphous regions; during the deformation, they can be oriented further. Depending on the way of orientation of an elastomer (before or after crystallization), the level of the mechanical properties at small deformations will differ. But for elastomers with a low ultimate degree of crystallization Cv both the morphology under large elongation and tensile strength can be the same [95, 96, 785, 786, 787]. The difference between the two routes of developing an oriented structure for highly crystalline polymers is also preserved at high degrees of orientation. It also manifests itself in the measurements of the thermal effects in stretching [167, 246, 247, 250, 266, 331, 334, 665]. Various aspects of the theory of stress-induced crystallization of polymers have been considered by many investigators [6, 206, 207, 250, 303, 333, 767–769].
4.3
Melting of oriented elastomers
The experiment shows that the larger the stress applied to a specimen before crystallization, the higher its melting temperature is (Fig. 4.4). This increase can be completely registered, if the melting is also performed in a stressed specimen. The first experiments on the effect of V on the melting temperature Tm were carried out on NR-based rubbers [54]; Tm was taken to be the temperature, at which the V –T dependence for a prestretched and then cooled specimen deviated from the linear shape (see curve 1 in Fig. 4.4). The dependence of Tm on V was also obtained in [538], where Tm was defined as the temperature, at which a preoriented and then radiation-vulcanized and crystallized NR specimen was contracted (curve 2 in Fig. 4.4). A dependence of Tm on deformation was also obtained for butadiene rubber SKD (curve 3) [95, 96] and polysiloxane (curve 4) [100a]. An increase of Tm with the rising deformation for stretched and then cooled weakly crosslinked vulcanizates of NR, butadiene and butyl rubber was also determined by the recovery of the crystallization-relaxed stress by heating [764]; and for 1,4-cis-polybutadiene, by the X-ray data [139]. Theoretical equations describing the dependence of Tm on deformation were obtained by Flory [201], Krigbaum and Roe [606], as well as in [6] and [769]. The Flory equation 0 1/ Tm 1/ Tm, O
R / 'H ª¬« 6 / S m 1/ 2 O O 2 / 2m 1/ O m º¼» ,
(4.3)
(where O = H + 1; m, the number of statistical segments in the chain between crosslinks) describes an experiment only at intermediate deformations. The Krigbaum and Roe equation 0 1/ Tm 1/ Tm, O
Rm / 2'H O 2 2O 3 ,
(4.3c)
gives underestimated values of Tm within the entire region of deformations. This, as the unsuitability of eq. (4.1) for describing the dependence of C on O, is directly determined by the character of the models used to plot them [95]. Nevertheless, they continue to be used for describing the stress-induced crystallization of elastomers [11].
STRESS-INDUCED CRYSTALLIZATION OF ELASTOMERS
105
log Tm 3 2
1
2.50
2.45 4 2.40 2.35 0
2.5
5.0
7.5
10.0 σ, MPa
Figure 4.4 Dependence between the melting temperature of a stretched specimen, Tm, and stress V, calculated for the true cross section, for vulcanizates of NR (1, 2), 1,4-cis-polybutadiene (3) and polysiloxane (4).
Another approach is based on the use of an equation similar to the Clapeyron–Clausius equation (3.19) [95, 96, 374, 752] in the form d ln Tm / df
'l / 'S ,
(4.4)
where pressure is replaced by the stretching force f, and the change of volume, by the change of length (contraction) during the melting, 'l ('S is the change of entropy). Passage to stress V, referred to the true cross section of a stretched specimen, and the assumption of constancy of the right-hand side part of the equation yield: Tm,V
Tm0 eaV ,
(4.4c)
Tm0 1 aV .
(4.4cc)
where a is a constant; at small a and V Tm,V
Experimental data for elastomers are described well by eqs. (4.4c) and (4.4cc) in the range of H > 0.3–0.5, which enables assessing the changes of Tm with the stress rise. Thus, for stresses close to tensile values (Vr), the value of Tm,V for NR-based rubbers is within the range of 100–120°C (depending on the type and density of the network and rubber manufacturing conditions); for rubbers from polychloroprene it is within the range of 120–150°C, which explains their comparatively high thermal resistance. For common rubbers from SKD, Tm,V atV = Vr fails to reach 20°C, which determines their low rupture strength at room temperature. The use of 1,4-cis-polybutadiene with a high cis-unit content (see Chapter 7) enables production of rubbers for which Tm,V atV = Vr can exceed 20°C [163, 165]. For rubbers based on polysiloxanes, Tm,V atV = Vr fails to reach 20°C [100a]. Extrapolation of the data, obtained at large Vand H, to V = 0 makes it possible to obtain an equilibrium melting temperature of the nondeformed specimen, Tm0. The values obtained
106
CHAPTER 4 dQ/dt I
a
II -20
0
20
25
40
60
T', °C
-dε/dt
-dε/dt
c
b
20
2
15
1
2
10
1
5 -10
-5
0
5
10
15 T', °C
-10
-5
0
5
10 T', °C
Figure 4.5 Dependence of heat flow dQ/dt (a) and deformation changes dH/dT during the contraction in heating (b, c) on scanning temperature T c for specimens from NR (a, b) and SKI-3 (c) stretched at 20°C (a) and –26°C (b, c) and crystallized at –26°C: a, melting under stress; I, II, melting regions; H0 = 2; b, c, melting after the release of stress for the initial (1) and milled (2) specimens; H0 = 3 (b) and H0 = 5 (c).
are close to those obtained by other methods. As all extrapolations of this kind, they are valid only at the identity of crystalline structures occurring in various deformation regions. This fact, as well as the formation of axial fibrillae in elastomers immediately in the process of stretching suggest that axial fibrillae are monocrystals with extended chains. Owing to the equilibrium character of such crystals, they melt under close-to-equilibrium conditions, too. The data of [79, 81] that crystals formed in the ultimate orientation of polyethylene by the first route (orientation + crystallization) melt, after a stress is removed, at Tm = 141.5°C, i.e., at Tm = Tm0 for polyethylene, are also in favour of these views. Herewith, the melting occurs within an extremely narrow temperature range, close by its width to the melting range of low-molecular-mass substances. This is indicative of the equilibrium character of the crystals formed. Note that the mechanical properties of such crystals (modulus and strength) are close to the mechanical properties of an ideal crystal. Apparently, similar to pressure, application of stress leads to the release of kinetic limitations for crystallization, and its occurs very rapidly close to the equilibrium melting temperature Tm0, characteristic of the given stress. This, in turn, makes it possible to produce close-to-equilibrium crystals. The causes of this accelerated crystallization close to Tm0 for oriented polymers are, probably, due, as in the case of pressure, to the formation of an intermediate liquid-crystalline phase. The melting curves obtained by the calorimetry method for oriented and then cooled specimens, i.e., at two-stage stress-induced crystallization, have two melting temperatures:
STRESS-INDUCED CRYSTALLIZATION OF ELASTOMERS
107
a higher one, which pertains to the melting of extended-chain crystals, and a lower one, which characterizes the melting of folded-chain crystals [322, 323, 553]. These data were also obtained for elastomers.* Thus, two melting regions have been found on the thermograms of NR oriented at room temperature and then crystallized at –26°C (Fig. 4.5a). Two melting regions at O > 2.5 were also observed under similar conditions for 1,4-cis-polybutadiene with high cis-unit content both by the DSC method [265] and by the change of the dynamic modulus [217]. Two melting temperatures at O > 4 for NR specimens oriented and crystallized at one temperature were obtained [585, 733] by extrapolating for C = 0 the C– O dependences determined by the birefringence method at different temperatures. The assignment of the higher temperature to the melting of a less defective part of the crystals (extended-chain crystals) is supported by the X-ray method. The Flory equation (4.3) is inapplicable for describing the dependence of Tm,O on O as this equation describes only the low-temperature melting region [265], whereas the experimental data usually pertain to the end of the high-temperature region. Meanwhile, eq. (4.3) should in principle be applicable only to the equilibrium values of T 0m,O, and the end melting temperatures of the crystals formed at T = T1 are in no way equilibrium. Only the temperatures corresponding to the end of the high-temperature melting regions, i.e., to the melting of extended-chain crystals, can be considered to be equilibrium. Crystals formed in elastomers during the orientation melt at lower temperatures after the stress is released. However, these temperatures are still higher than Tm of nonoriented elastomers (if a temperature can be chosen at which both isotropic and stress-induced crystallization can proceed, and crystals do not melt after the stress is released). Contraction of oriented elastomers during the melting, used in measurements of residual elongations, makes it possible to determine the melting temperatures of oriented specimens under such conditions by the differential melting curves (see Fig. 4.5) [141, 142, 381, 383, 454, 562]. Under these conditions, as in the case of the melting atO = const (or H = const), one can single out two melting temperatures (Tm,H,1 and Tm,H,2). Their presence gives grounds for views of the topomorphism of the crystals formed during the orientation. Naturally, the values of Tm,Hobtained under these conditions for axial fibrillae (Tm,H,2) at even larger initial deformations (O0 or H0) are much lower** than T 0m,O. This gives grounds to believe that the axial fibrillae that melt under these conditions are not crystals with extended chains, but those with “unfolded” chains. These crystals formed in an oriented specimen differ from extended-chain crystals by that their length is smaller than that of the macromolecules, and the crystals are linked by a large number of tie chains [206]. It is these tie chains that mainly determine the mechanical properties of such elastomeric materials. Thus, the melting temperatures Tm,H,1 and Tm,H,2 measured by this method can be considered to reflect the melting of the crystals with folded and “unfolded” chains, respectively. The above data also suggest that when an elastomer is oriented under close-to-equilibrium conditions, the axial fibrillae consist mainly of extended-chain crystals. Unfortunately, significant effects of the crystallization and melting conditions on Tm and on the character of the melting curves makes difficult the comparison of the data by different authors. Also, it should be taken into account that the major part of the experimental data were obtained under far from equilibrium conditions. * The contribution of reorganization to the melting of oriented elastomers has not been studied. ** Tm,H,2 at H0 > Hk does not practically depend on H0.
108
4.4
CHAPTER 4
Effect of the molecular parameters of rubbers and the composition of rubber compounds on stress-induced crystallization
Owing to the different morphologies of crystals at low-temperature and stress-induced crystallization, the same techniques may lead to different changes in them. Moreover, the high rate of formation of crystals with folded chains (or micellar crystals) can presumably be an obstacle for the formation of axial fibrillae, i.e, crystals with extended chains or crystals with “unfolded” chains. Thus, for elastomers with a high rate of formation of folded-chain crystals crystallization with chain unfolding can occur only in those temperature ranges, where folded-chain crystals are formed comparatively slowly, i.e., the critical value of the unfolding parameter Ek should be reached faster than the crystallization by the mechanism of folded-chain crystals begins. Besides the choice of respective temperature conditions for crystallization, this requirement can be fulfilled by suppressing crystallization by changing the chain structure or composition of rubbers. 4.4.1
Effect of chain chemical structure, molecular mass and microstructure
Available literature data give grounds to assume that, by changing the chemical structure of the polymer chain, conditions for the predominant development of stress-induced crystallization can be created. What is more, crystallization to form folded-chain crystals may not take place at all at a large content of foreign units, and only micellar-type crystals would form. Proceeding from the views of the competition of the formation rates for extended- and folded-chain crystals, all crystallizable rubbers can be divided into three groups. The first group includes the rubbers of the great rate of isotropic crystallization. Its time is small up to temperatures close to the equilibrium melting temperature Tm0. For these rubbers, formation of axial fibrillae at large supercoolings, i.e., far from Tm0, is the least probable. This group comprises siloxane rubbers and 1,4-cis-polybutadiene with high cis-unit content [558, 634a]. Herewith, in the range of temperatures close to Tm0 stress-induced crystallization of these elastomers proceeds to form extended-chain crystals. For this, the time of reaching Ek during the stretching should be less than that required to form folded-chain or micellar crystals in the close vicinity of Tm0. The second group are elastomers with a moderate rate of low-temperature crystallization. The time of their crystallization at a temperature of its maximum rate is tens of minutes and more. During the stretching of a specimen, the time of reaching Ek is significantly (10 times) less than that of the onset of the formation of folded-chain or micellar crystals. Under these conditions, crystals with extended or “unfolded” chains are formed; in turn, they serve as nuclei of the crystals with folded chains or micellar crystals. This group includes both NR and synthetic 1,4-cis-polyisoprenes, as well as chloroprene and urethane elastomers. The third group of elastomers are materials characterized by an extremely low rate of folded-chain crystal formation. A classical example of rubbers in this group is butyl rubber, the low-temperature crystallization time of which is rather high. However, the low rate of isotropic crystallization is a necessary but insufficient condition for stress-induced crystallization to develop. Another condition is the possibility of achieving Ek during the deformation. This requires, first of all, the absence of breakdown before Ek is reached. In the case of nonvulcanized rubber, plastic deformation should not begin, at which macromolecules shift one relative another. An obstacle for the development
STRESS-INDUCED CRYSTALLIZATION OF ELASTOMERS
109
of plastic deformation can be the formation of the network of physical junctions, which takes place in the case of NR, owing to the presence of biological components. This is what determines the difference in the stress corresponding to the onset of stress-induced crystallization in NR and in synthetic polyisoprene even with equal cis-unit content [139, 712]. The most stable structure of the crystal is achieved when the end groups of the macromolecules are localized exclusively on the surface of the crystal. The specific features of the effect of M and MMD on stress-induced crystallization of nonvulcanized elastomers are, apparently, related to the role of the network of physical junctions in the achievement of given values of unfolding of the molecular chain. As the molecular mass increases, the number of physical junctions rises, and the critical value of the macromolecule unfolding parameter Ek is achieved at lower stretchings Hk [345, 557]. The melting temperature Tm,H,2 of “unfolded”-chain crystals* rises with M going up, the rise being faster for fractionated specimens than for nonfractionated ones with the same characteristic viscosity. Broadening of the MMD of the fractions brings the melting temperature values to those that are observed for nonfractionated rubber [557]. For natural rubber, the dependence of the melting temperatures for crystals with folded and “unfolded” chains on viscosity is less pronounced than for synthetic 1,4-cis-polyisoprene. As the other differences in stress-induced crystallization, it is due to the presence of protein fragments [139, 557] forming the network of weak physical junctions. A decrease of the molecular mass of rubbers in milling also results in a change of the stress-induced crystallization rate and a redistribution of the intensity of the peaks corresponding to the melting of crystals with folded and “unfolded” chains (Fig. 4.5b). Wunderlich [751b] considered two possibilities in principle of constructing extended-chain crystals for polymers with a broad MMD. In the first case, macromolecules of different lengths are not separated one from another; as a result, crystals of this type are a solid solution of components. These crystals should grow by the mechanism providing for the butt jointing of macromolecules. In the second case, there is microfractionation, i.e., the separation of macromolecules by length during the crystallization. This corresponds to the ultimate case of the eutectic phase separation in crystallization, when the components are compatible in the melt, but incompatible in the crystalline state. In this macromolecule packing method there are not limitations and difficulties associated with butt packing. This microfractionation was directly observed for polyethylene crystallized at high pressures [751b]. The dependence of the melting temperature of “unfolded”-chain crystals, Tm,H,2, for oriented SKD** on the cis-unit content w is linear (Fig. 4.6a) [557]. For polyisoprene, this dependence can also be considered to be linear (Fig. 4.6b), though the scatter of the experimental points is much greater. This is due mainly to the strong influence of the other molecular parameters of polyisoprene on stress-induced crystallization. Extrapolation of the melting temperature of “unfolded”-chain crystals, Tm,H,2, to w = 100% gives the value Tm,H,2 = 6°C, coinciding satisfactorily with the value for denatured NR under the same conditions. Comparison of the effect of the microstructures of 1,4-cis-polyisoprene and 1,4-cispolybutadiene on the kinetics of isotropic and stress-induced crystallization has shown that a microstructure disturbance suppresses more the formation of folded- rather than * Measured by the contraction of a specimen oriented at –25°C up to a given value of H0 and then released. ** As earlier, Tm,H,2 was determined in contraction of oriented specimens.
110
CHAPTER 4 Tm2,ε, °C -5
Tm2,ε, °C
a
b
4 2
-10
0
-15
-2
-20
-4 -25 -6 85 90
95
100 w, %
97
98
99 w, %
Figure 4.6 Dependence of the melting temperature of “unfolded-chain crystals, Tm,H,2, measured by the contraction of an oriented specimen, on the cis-unit content w for 1,4-cis-polybutadiene (a) and 1,4-cis-polyisoprene (b); initial deformation H 0 = 2 (a) and H0 = 5 (b), crystallization temperature –40°C (a) and–26°C (b).
“unfolded”-chain crystals. Therefore, in some cases the disturbance of microstructure can (within small limits) be used as a technique, which makes it possible to preserve the enhancing action of stress-induced crystallization while almost completely suppressing the crystallization in the process of prolonged holding, e.g., during the storage in the absence of stresses or at small stresses.* 4.4.2 Effect of vulcanization and filling Vulcanization. The presence of a three-dimensional network suppresses elastomers’ capability of viscous flow and, thus, improves the conditions of orientation of polymer chains. However, in the process of vulcanization the regularity of the chains is disturbed, which makes crystallization difficult. A pronounced maximum on the dependence of the degree of crystallization in orientation on 1/Mc found for NR-based rubbers [258, 315, 338, 625] and other elastomers is related to these effects. At the same time, as for isotropic specimens crystallized at T < Tm0, an increased density of the network leads to a systematic decrease of the melting temperature. The decrease measured by the contraction of stretched specimens takes place both for Tm,H,2 and Tm,H,1, and is the most pronounced at a density of the network junctions 1/Mc > 1.5 mol/cm3. This is clearly observed when using nonaccelerated sulphur vulcanization for 1,4-cis-polyisoprene (SKI-3) and 1,4-cis-polybutadiene (SKD) [142]. Unfortunately, due to the difficulties associated with Tm measurements under stress conditions, no direct data on the changes of the melting temperatures for extended-chain crystals depending on the density of the network and deformation. However, the data * Such microstructure disturbances should not be in the region, where an acceleration of isothermal crystallization can be observed (see Chapter 3).
STRESS-INDUCED CRYSTALLIZATION OF ELASTOMERS
111
obtained by the X-ray [315] and IR spectrometry [338, 625] methods support the conclusion on the decrease of Tm,Hwith 1/Mc rising. The effect of the type of vulcanizing group on stress-induced crystallization has been studied insufficiently. However, available data [85, 239, 315, 338, 625, 711], as well as the results related mainly to the tensile strength measurements indicate that the networks suppressing isotropic crystallization better (sulphur with sulphenamides, mercaptobenzotiazole and other accelerators) do not inhibit stress-induced crystallization. The maximum on the curves of the dependence of the degree and rate of stress-induced crystallization on the density of the network is shifted for them to the region of large values of network densities. At the same time, there are data on the independence of the deformation corresponding to the onset of stress-induced crystallization (Hk) for NR and synthetic 1,4-cispolyisoprene on the density of the vulcanization network [711, 712]. This result can be due both to the features of the method used (synchrotron X-ray radiation) and to the Mc interval studied. According to the data of [164], stress-induced crystallization of the networks from NR produced in a swollen state (the so-called ideal networks) is slower than for common networks. However, the differences become less as H increases. This is indicative of the role of the initial physical junctions as nuclei of stress-induced crystallization and its decrease with H going up. Recall that vulcanizates can be considered as copolymers of the regular part of the chains and the part with disturbed regularity. The results obtained by combining the thermodynamic analysis of the behaviour of copolymers as a eutectic and the theory of van-der-Waals networks [334] suggest that in the process of crystallization (both stressinduced and isotropic) the vulcanization network junctions do not enter the monocrystal [333]. Therefore, the distribution by size of extended-chain crystals occurring in stressinduced crystallization corresponds to the distribution by size of chain fragments between crosslinks (Mc).* This implies that microfractionation (also noted in consideration of the effect of the molecular mass) is under way in the process of stress-induced crystallization and extended-chain crystal formation. In turn, the distribution by size of extended-chain crystals is also reflected on the character of melting of oriented vulcanizate. This approach made it possible to describe well the deformation curves of peroxide vulcanizates of NR and a change of heat release during their stretching at temperatures of 22–90°C. The different character of the effect of the type and density of the three-dimensional network on stress-induced and low-temperature (isotropic) crystallization, i.e., on the intensity of forming different types of crystals** enables using this factor, as well as some disturbance of molecular-chain regularity, to obtain systems with a high capability of stressinduced crystallization and a reduced capability of low-temperature crystallization. Filling. As the application of an external stress to the system considerably increases the rate of inherent nucleation, the effect of a filler as a nucleating agent should decrease in the case of stress-induced crystallization. Indeed, both for filled and oriented specimens the induction period in the kinetic curves is absent; herewith, the effect (described in Chapter * This is also true for networks with other disturbances of microstructure. In this case, the distribution by size of extended-chain crystals reflects the distribution of not Mc but of chain fragments between all regularity disturbances, i.e., the distribution of crystallizing sequences, which, naturally, are smaller than Mc. ** This distinction restricts the possibilities of using parameter B in eq. (3.17) to predict stressinduced crystallization.
112
CHAPTER 4 -dε/dT 6 4 2
6
6 4 2
5
5 3 1
4
15 10 5
3
20 15 10 5
2
20 15 10 5
1 -10
0
10 T', °C
Figure 4.7 Dependence of the deformation change dH/dt on temperature T c for SKI-3 containing 0 (1), 10 (2), 20 (3), 30 (4), 50 (5) and 75 (6) phr of active carbon black; H = 3.
3) due to nonisothermal crystallization of filled rubbers is enhanced in the case of a stress. Therefore, the main effect of a filler on stress-induced crystallization should be reduced to the change of orientation conditions in the presence of a filler. On the whole, crystallization under orientation conditions aggravates the features of the process for filled rubbers considered in Chapter 3. The X-ray patterns at H = 3 preserve the distribution of intensity characteristic of c-grain patterns within the entire range of filling. As, at the same deformation, the stress in a specimen increases with the filling increased, the unfolding parameter E determined by eq. (4.2) also increases (see Fig.4.3b) and the value of Ek is achieved at smaller H0 [141, 383a] (transition to crystallization by the chain-unfolding mechanism occurs at Ek = 0.3). In principle, such a filling can be achieved, when the formation of oriented crystals would develop without any application of external stress, which was indeed observed on particles of active filler for NR [379, 738] and polychloroprene [95, 96, 102, 106, 618].
STRESS-INDUCED CRYSTALLIZATION OF ELASTOMERS Tm, °C a
20 16
12 8 4
4
Tm, °C b 16
3
12
2
8
1
4
113
4 3 2 1
0
0
4 2
4
2
4
ε
Figure 4.8 Dependence of the melting temperature of folded-chain Tm,H,1 (1) and “unfolded”-chain Tm,H,2 (2) crystals and additional peaks (Tcm,H,2) (3) and (Tccm,H,2) (4) on the initial stretching deformation H0 for specimens of NR (a) and SKI-3 (b) filled with 50 phr of active carbon black.
X-ray measurements of the dependence of the degree of crystallization on deformation for filled vulcanizates of NR [379] have shown that the deformation corresponding to the onset of crystallization, Hk (at the same stretching rate) is the smaller, the higher the content of carbon black is. However, the calculation of the true deformation of the rubber matrix H r suggests that crystallization begins at the same value H k, which, apparently, corresponds to Ek = 0.3. Consideration of filled rubbers as eutectic systems, as in the case of the occurrence of the vulcanization network and other regularity disturbances (see above) makes it possible to describe the produced deformation curves and heat release at deformation [332, 333]. A decrease of Hk was observed upon introduction of small amounts of oligoester epoxides into rubbers from SKI-3 [74], which is due, apparently, to the specific role of oligoester compounds in the formation of the three-dimensional network in the rubber matrix. For vulcanized rubbers, the effect of filling on stress-induced crystallization is the same. The rate of stress-induced crystallization for rubbers (in particular, from SKD [350]) filled with more active fillers can go down during the crystallization in the process of stretching. This is due to the reorientation of crystals produced at the initial stages of stretching; their number is larger is the filler is more active. As unfilled elastomers, filled rubbers in stress-induced crystallization in the range H > Hk are characterized by two melting regions. They are reliably registered by the method based on the contraction of a stretched specimen. Using 1,4-cis-polyisoprene with active carbon black as an example, it has been shown that, as the content of the filler is increased from 0 up to 75 phr, there occurs a redistribution of the intensities of these melting peaks. The melting temperature of folded-chain crystals, Tm,H,1, decreases, and the melting temperature of “unfolded”-chain crystals, Tm,H,2, increases (Fig. 4.7). Besides, additional melting peaks at temperatures Tcm,H,2 and Tccm,H,2, which are higher than Tm,H,2, occur on the melting thermograms of filled specimens released from the load. As the amount of filler is increased, the number of additional melting peaks rises (see Fig. 4.7) [141]. Figure 4.8 presents the dependences of four main melting temperatures on the initial degree of stretching, H0, of milled NR and SKI-3 filled with 50 phr of carbon black PM-100. The position of the additional peaks does not depend on H0. The temperature of the main peak at Tm,H,2 depends little on H0, too. Additional melting peaks in a filled system are, apparently, due to the existence of regions of greater “unfoldedness” of the macromolecules. A greater stability of these regions
114
CHAPTER 4
in the process of unloading and melting is determined by the character of interaction with the filler.* Thus, introduction of KCl, which is not a reinforcing filler, leads to the formation of only one additional melting peak. One more peak emerges in the case of the weakly reinforcing chalk. Introduction of carbon black PM-100 also leads to the appearance of two additional peaks shifted to the region of higher temperatures, and the filling with modified carbon black makes the high-temperature peak shift to the region of still higher temperatures [141]. Thus, the position of the peaks in the temperature scale reflects the efficiency of interaction of filler and rubber, which is of crucial importance for predicting the mechanical properties of vulcanized and raw rubber compounds. Introduction of a filler into the SKD rubber also enhances its capability of stress-induced crystallization, but yields no additional melting peaks, which is indicative of the weak interaction of the investigated rubber with filler [627, 629]. Note that the use of the method of determining the melting temperatures by the contraction of released oriented specimens, together with the views of chain “unfoldedness” E and its critical value Ek = 0.3 makes it possible to reveal many features of stress-induced crystallization and to determine its contribution to the strength properties of rubbers (see Chapter 5). At the same time, it is the melting temperature obtained under conditions, when orientation is preserved up to the melting of the elastomer, that determines the temperature dependence of the strength of elastomers in stretching. The data on the increase of macromolecule “unfoldedness” in the presence of fillers and the change of the crystallization mechanism (transition from folded-chain crystals to “unfolded”-chain or extended-chain crystals may also shed light to some extent on the mechanism of reinforcement of filled composites as a whole.
* A certain contribution to this effect is, apparently, made by the nonisothermicity of the crystallization process, similar to the case of the melting of nonstressed filled elastomers [96, 626a].
5
Strength Properties of Elastomers at Low Temperatures
The possibility of using elastomers in various fields is determined by their specific deformation properties. Namely these properties are the most sensitive to glass-transition and crystallization processes. Therefore, it is to their changes at low temperatures that the low-temperature behaviour of elastomers is related. These changes have been considered in Chapters 1–3. The level of the strength properties also determines the applicability boundaries of elastomeric materials. This pertains, first and foremost, to the strength properties of elastomers at below the glass-transition temperature; herewith, as a rule, the limiting factor is not the stress but the breaking deformation. This chapter will also consider the strengthening effect of stress-induced crystallization. It is of the greatest significance for elastomers crystallizing during the stretching at room temperatures. The high rupture strength, including the cohesion strength, of rubbers and nonvulcanized rubber compounds and, therefore, the possibility of their processing, as well as the development of nonvulcanized adhesives, are associated with exactly stressinduced crystallization. The contribution of stress-induced crystallization to the strength properties under small deformations manifests itself in a rather specific way [8, 97, 237, 616, 785, 786]. It has been studied insufficiently, whereas namely under these conditions most rubber articles are operated.
5.1
Strength of amorphous elastomers at low temperatures
Chapter 2 partially considered the strength properties of elastomers in the glassy state. Thus, the left-hand part of the curves in Fig. 2.5 represents a temperature dependence of the brittle breaking stress Vbr under static loading conditions. The region, where the brittle fracture of a polymer takes place, is characterized by the independence of the breaking stress on temperature. It corresponds to region I (Ic and Icc) in Fig. 5.1, curve 1, which represents schematically a temperature dependence of the breaking stress of an amorphous elastomer. This region, in turn, can be divided into two parts. The athermal breaking mechanism is characteristic of the low-temperature part (Ic). At higher temperatures, the athermal mechanism is replaced by the brittle thermofluctuation mechanism (region Icc). Its action is not complicated by the relaxation processes up to the brittleness temperature Tbr0 (see Chapter 2). The fundamental views of the kinetic thermofluctuation theory of strength, formulated by Zhurkov [598], are applicable to this temperature range. The upper temperature boundary
116
CHAPTER 5
σr I'
I'' II
III
2 IV
Tbr0
Tg
1 Tmax
T
Figure 5.1 Dependence of the breaking stress Vr on the temperature of amorphous (1) and crystallizable (2) elastomers; in regions Ic and Icc, Vr = Vbr.
of this range is the “ultimate” brittleness temperature Tbr0 (see Chapter 2). At temperatures above Tbr0, the thermofluctuation mechanism is complicated by local relaxation processes due to the development of stress-induced elastic deformation. As we pass to higher temperatures, the role of relaxation processes increases, and they become predominant at temperatures higher than the glass-transition temperature Tg. There is no unequivocal views of the nature of the fracture in the stress-induced rubberlike region II (Fig. 5.1) [278, 280, 320, 598, 611,785, 786]. It is clear, though, that in the region between Tg and Tbr0 the polymer has some molecular mobility stored, but for stressinduced (forced) elasticity to be realized, large tensile stresses and small deformation rates are required. It is not clear whether the stress-induced rubberlike elasticity can develop at other kinds of stress. There are some grounds to believe that the boundary between regions I and II (Tbr0) coincides with the region of the development ofE-relaxation processes (see Chapters 1, 2). However, this point of view is not generally accepted. We should note that the stress corresponding to the forced elasticity limit VL (see Fig. 2.5) is always higher than the breaking stress in this temperature range. Since the development of forced elasticity, as the realization of the thermofluctuation breaking mechanism, requires considerable time, at sufficiently large stress–strain rates the brittle fracture can take place up to Tg. Herewith, it should be taken into account that Tg itself rises with the deformation rate going up. Therefore, no correlation between Tg determined under static conditions and Tbr determined using the standardized methods by high-speed impact, could be observed. To determine the true temperature boundary of the breakdown of an elastomeric material at low temperature, one should measure the temperature dependence of the strength at equal stress–strain rates (which is not done in practice) or at least measure the dependence of Tbr on the loading rate [786]. At temperatures higher than Tg the breaking stresses are sharply decreased (region III) and in the end of the transition region take on the value characteristic of the rupture strength of elastomers in the rubberlike state. In the transition region from the glassy to the rubberlike state, the strength properties are determined by the same physical relaxation processes stipulated by the change of the rate of development of the rubberlike deformation determining the deformation properties. This is unequivocally testified by the applicability of the principle of the temperature–time superposition to the strength properties of elastomers; herewith, the deformation rate H is used in this case as the time characteristic [77, 96, 188, 198, 236, 662]. Figure 5.2 [188] presents the results of the superposition of the data for NR-based rubbers obtained within
STRENGTH PROPERTIES OF ELASTOMERS AT LOW TEMPERATURES
117
εr b
6.0 5.0 4.0 3.0 log σr(TS/T), MPa 1.8
2.0
a
1.0
1.4
0
1.0
-8
-4
-6
-2
0
2
4
6
. -log aT ε
0.6 0.2 -0.2 -0.6 -8
-6
-4
-2
0
2
4
6
. 8 -log aT ε
Figure 5.2 A generalized dependence of the reduced strength at the uniaxial tension logVr(Ts/T) (a) and elongation at break Hr (b) on the logarithm of the reduced stress–strain rate log aTH for unfilled NR-based rubber; the reference temperature Ts = –10°C; different points refer to the testing temperatures from –67.5 up to +93.3°C. log σr(TS/T), MPa 1.5 1.0 0.5 0 -0.5
0.1
0.3
0.5
0.7 log εr
Figure 5.3 Dependence of the logarithm of reduced strength logVr(Ts/T) on rupture strain Hr at the uniaxial tension (rupture envelope) for unfilled styrene-butadiene-based rubber; different points refer to the testing temperatures from –67.5 up to +93.3°C; the reference temperature Ts = –10°C.
a broad temperature range. The temperature dependence of the shift factor aT obeys the Williams–Landel–Ferry equation (1.7) already considered in Chapter 1. It has been shown [188, 662] that there is a universal dependence (a “rupture envelope”) between the breaking stressVr and the elongationHr; this dependence points to pairs of values Vr and Hr at which the breakdown can occur (Fig. 5.3) [188]. The rupture envelopes have been plotted for many amorphous elastomers with different chemical structures [13–16, 221]. At temperatures significantly higher than Tg, the rupture envelopes obtained for 15 different elastomers form approximately a unified curve, if the breaking stress is
118
CHAPTER 5
referred to the unit concentration of the efficient chains of the network per unit volume [221]. In the first approximation, this reflects the fact that the deformation properties are not dependent on the chemical structure. This is one more confirmation that deformation and breakdown processes have much in common. Deviations of the experimental data from the master curves are related to the particular features of the changes of the flexibility of the molecular chains near Tg for different elastomers, as well as of the stress-induced crystallization. The applicability of the Bartenev equation
Vr
A0H1/ m exp >U / mkT @ ,
(5.1)
in which H is the tensile strain rate; k is the Boltzmann constant; m = 1 + B, where B is a constant in the equation for the long-term strength
W
C V B exp >U / kT @ ,
(5.1c)
is limited by the range of temperatures with a constant value of activation energy U, i.e., by the rubberlike region. The values of activation energy U determined in accordance with eq. (5.1) are a convincing proof that the kinetics of the breaking process of elastomers at above Tg is determined mainly not by the rupture of the chemical bonds but by the physical processes (recall that in the temperature range close to room temperatures eqs. (1.8) and (1.4) give practically equivalent results) [44]. The case of the breakdown at T < Tbr0 is different. Here, as it has been shown for glassy polymers, the major role is played by the breakdown of the chemical bonds [388, 598]. Experimental data obtained in studies of the breakdown products [320] also support this conclusion for elastomers. It has been shown [780, 781] that some features of low-temperature brittle breakdown are also characteristic of the breakdown of elastomers under close-to-operational conditions, when large deformations do not develop. In particular, under these conditions and in brittle breakdown the breakdown energy values are close [782]. This is especially pronounced when the stress in a specimen is determined to a great extent by the elastic component, i.e., for specimens with large shape factors. This analogy is also preserved in the effect of various factors on the breaking stress. Thus, the brittle strength (as the strength parameters determined at small deformations) does not practically depend on the presence and content of active fillers [57, 780, 781] (see, e.g., Fig. 2.5). Inactive fillers can even reduce Vbr, as the number of defects in a specimen, at which the breakdown can begin, increases. Orientation and crystallization have the same effects in this case [95, 97, 504, 598, 785, 786].
5.2
Strength of crystallizable elastomers
5.2.1 Strength determined by stress-induced crystallization The relation of stress-induced crystallization and strength of elastomers is well known and does not admit of doubt [8, 13–16, 225, 239, 463a, 715, 744]. The crucial significance of
STRENGTH PROPERTIES OF ELASTOMERS AT LOW TEMPERATURES
119
stress-induced crystallization in forming the strength properties of elastomers under conditions of uniaxial tension is indicated by the fact that at room temperature the strength of unfilled elastomers crystallizable under tension in these conditions (isoprene and chloroprene rubbers, butyl rubber) is by an order or two higher than for non-crystallizable polymers or those that crystallize only at low temperatures. In the opinion of Aleksandrov and Lazurkin [8], as well as Wood [744], Gee [225] and Treloar [715], the role of crystals during the deformation of crystallized rubbers is similar to that of active filler particles. The value of Vr increases linearly with the rise of crystallinity C, measured during the stress-induced crystallization by the X-ray method [25, 387, 316-318, 259, 260], by the release of heat [163, 243] and birefringence [243, 715, 733, 764]. Differences in the strength of NR and synthetic 1,4-cis-polyisoprene [139, 239, 712] are also related to the distinctions in stress-induced crystallization (see Chapter 4). The temperature of the decrease of the strength of crystallizable rubbers is the melting temperature of the crystals formed under deformations close to rupture deformations, i.e., the melting temperature of extended-chain crystals (measured under conditions of a stress). The character of the effect of the tensile strain rate H onVr for elastomers is also related to their capability of stress-induced crystallization under experimental conditions [278–280, 735, 766]. For rubbers crystallized during the stretching at the testing temperature, an increase of H can result in a decrease of the degree of stress-induced crystallization and, therefore, also to a decrease of Vr, whereas for rubbers from noncrystallizable polymers Vr increases with H rising [278, 735]. The development of stress-induced crystallization, as the melting of crystals under orientation conditions, i.e., under the action of a stress, determines the extreme character of the curves for the tensile strength Vr and the elongation at break Hr (at uniaxial tension) on temperature [163, 387, 513, 558]. As a rule, the temperature dependence of tensile strength Vr (referred to the true section of a specimen) for crystallizable elastomers (see curve 2 in Fig. 5.1) at T > Tg can have two maxima. One, low-temperature, maximum is due mainly to the formation of crystals with folded chains in the temperature range of maximum crystallization rate. The other, high-temperature, maximum is related to the formation of crystals with extended or “unfolded” chains (which is facilitated owing to the rise of Hr with temperature and to the possibility of achieving Ek). The upper temperature boundary of this maximum corresponds to TmV , i.e., to Tm measured at the action of the stress. The character of the Vr –T dependence differs for rubbers with different chain structures [387, 556a]: the more regular a rubber and higher Tm, the more pronounced the maximum of the strength properties (related to the development of stress-induced crystallization) at elevated temperatures is. All techniques contributing to its development lead to an increase of the tensile strength. For an elastomer to have a high strength at a temperature close to room temperature and higher, crystals with extended or “unfolded” chains should be formed in exactly this temperature range. Thus, for 1,4-cis-polyisoprene, 1,4-trans-polychloroprene, ethylene-propylene, urethane and butyl rubbers the room temperature is within the range of the development of stress-induced crystallization. For 1,4-cis-polybutadiene, the temperature range of stress-induced crystallization reaches room temperature only when the content of cis-1,4-units w |98% [634a]. Thus, a rubber with w < 98% is characterized by a decrease of Vr with temperature going up, and for an elastomer with w > 98% the temperature dependence of Vr is observed to have a characteristic maximum with a peak at |40°C [163]. Such a rubber is well processed in mixing rolls [478]; vulcanized rubbers have good strength and hysteresis properties (see Chapter 7).
120
CHAPTER 5
σr, MPa
εr
28
4
12
24
3
8
20
1
16
b
0
0
-40
12
40
80
120 T', °C
4
8 4
2
4
1 -40
a
3 2 0
40
80
120
T', °C
Figure 5.4 Dependence of strength Vr (a) and deformation at break Hr (b) for rubbers based on 1,4-cis-polyisoprene, grade SKI-3, vulcanized with sulphur for various times tv, min: 0 (1), 100 (2), 250 (3) and 350 (4).
The effect of the three-dimensional network on the stress-induced crystallization parameters also determines the character of the strength properties of elastomers crystallizable under respective conditions. By example of rubber SKI-3 vulcanized by sulphur for various times (Fig. 5.4) [556], it has been shown that the formation of the three-dimensional network manifests itself first and foremost in the change of the temperature range of the strength preservation. Thus, for a nonvulcanized rubber the segment due to the stressinduced crystallization is weakly pronounced (see Chapter 4). For a weakly crosslinked vulcanizate, the temperature range of strength preservation TV |140°C, and the further increase of the network density changes little the width of this region. However, the value of Vr rises, as does the rate of stress-induced crystallization. Only at a sufficiently high density of the network, together with the rise of Vr, the decrease of TV is observed. This correlates with the decrease of the melting temperature of the crystals of extended or “unfolded” chains with the density of the vulcanization network rising* (see Chapter 4). The temperature dependence of elongation also reflects the peculiarities of orientation strengthening of polyisoprene. If the density of the vulcanization network changes within a broad range, then, similar to the degree of stress-induced crystallization, the value of Vr passes through a maximum. The position of this maximum depends on the type of rubber and the vulcanizing group, and the value of Vr proper on the deformation rate. This was observed for NR and polyisoprene [387], polychloroprene [238, 762], butyl rubber [442], ethylene-propylene rubber [449, 765]. The effect of the type of the vulcanizing group on tensile strength is also determined by its influence on stress-induced crystallization. Thus, rubbers containing monosulphide * Recall that the melting temperature under conditions of a stress is always higher than that measured during the compression of a specimen released from a load.
STRENGTH PROPERTIES OF ELASTOMERS AT LOW TEMPERATURES
121
and C–C bonds are characterized by a smaller development of stress-induced crystallization (in contrast with low-temperature crystallization, see Chapters 3 and 4) than rubbers vulcanized with sulphur and an accelerator. Respectively, the strength of rubbers of the former group is lower [95, 315, 474, 706]. Similar data were also obtained for polychloroprene [474, 762]. As the temperature resistance of rubbers capable of stress-induced crystallization, i.e., the temperature of the loss of strength, is determined by the melting temperature of crystals with extended or “unfolded” chains under stress conditions, the correct choice of the vulcanizing group is required not only to ensure the required strength properties of the rubbers, but also to obtain stable data on temperature resistance. Therefore, the compositions of rubbers from NR and SKI-3 and their testing temperatures should be chosen such that the melting temperature of the crystals formed in stress-induced crystallization be 10–15°C higher than the testing temperature. Otherwise, even minor deviations from the test conditions (defects in specimens during their preparation, instability of compositions, inaccuracy in temperature regimes) shall lead to changes of strength from the values characteristic of rubbers capable of stress-induced crystallization up to the values characteristic of amorphous elastomers [70, 474], i.e., by an order of magnitude. The effect of mastication on the tensile strength of 1,4-cis-polyisoprene, grade SKI-3, and a modified rubber SKI-3-01, is also due to the change of stress-induced crystallization conditions. A decrease of M in mastication results in a suppression of stress-induced crystallization of SKI-3, as the result of which the values of Vr in the region of the second (and first) maximum are decreased [556]. Introduction of an active filler leads in all cases to a decrease of the critical deformation (H k) at which stress-induced crystallization begins; therefore, Vr also increases, almost within the entire temperature range studied, i.e., the maxima in theVr –T curves degenerate [556]. A modification of carbon black [141] leading to an increase of its activity contributes to the rise of Vr of rubbers based on synthetic 1,4-cis-polyisoprene (SKI-3) and to an increase of the temperature of the loss of strength. This effect is especially strongly pronounced for 1,4-cis-polybutadiene (SKD). Thus, for rubbers filled with carbon black this temperature rises up to 100°C and higher, while for unfilled vulcanizates with w < 98% it only approaches the room temperature. However, attempts have failed to separate in this case the effect of filling on the stress-induced crystallization and directly on the strength of an elastomer not crystallizable under these conditions. The temperature interval of strength preservation was observed to broaden and the absolute values of Vr to increase at the filling of NR [307, 706], butyl rubber [442, 662], trans-polypentenamers [517, 520], thiocols [456], and polychloroprenes [762]. The strengthening efficiency of nanofillers on the strength due to stress-induced crystallization has been shown by example of clay nanoparticles in 1,4-cis-polyisoprene [20]. Various techniques of raising the cohesion strength of blends based on SKI-3 are associated in one way or another with enhancing the ability of this rubber for stress-induced crystallization and for the shift of the stress-induced strengthening region above the room temperature. Development of good nonvulcanizing adhesives is also associated with production of materials capable of stress-induced crystallization [257, 284, 286 581]; however, experimental data on the character of this association are scarce. Also scarce are the data on the effect of stress-induced crystallization on the strength of elastomers at other types of loading, other than uniaxial tension. Gent [236] believes that stress-induced crystallization contributes to elastomer breakdown not only under conditions
122
CHAPTER 5
of uniaxial tension, but also in abrasion. Apparently, this is true for the tearing strength under other conditions, too, when an elastomer capable of stress-induced crystallization is under breakdown in a rubberlike state but large deformations, as in the case of fatigue endurance, do not develop [616]. Evidently, the breakdown under these conditions is preceded by the formation of highly-oriented microregions (microstrands). This was shown for the cases of tearing [706], ozone attack [96, 96, 785–787] and cutting [780, 781]. The degree of strengthening due to stress-induced crystallization depends in these cases on the lifetime of the microstrands. Comparison of the effects of crystallization and filling on the strength of elastomers shows that crystallization is more efficient. The use of fillers in rubber based on noncrystallizable polymers never makes it possible to achieve the values of tensile strength provided by the development of stress-induced crystallization in rubber compounds based on crystallizable polymers. The same refers to fatigue strength and abrasive wear (though here the difference is not so great). The advent of fillers with nanosize particles and their high efficiency suggest that one of the causes of the larger efficiency of crystallization in the reinforcement of elastomers as compared even with active fillers is the small size of unit crystallites in elastomers [124]. Thus, crystallites are natural nanofillers of elastomers. However, their high efficiency is due to not only excess surface energy determined by their small size. A rather significant factor is that they are directly linked to the elastomer matrix. Indeed, parts of macromolecules, bordering on crystallites, can be crystallites’ components, too. Both these factors enable crystallites move under the action of stresses (including thermoelastic stresses) together with the rubberlike matrix. Analysis of the effect of crystallization on the mechanical properties of elastomers enables formulating the requirements to the properties of nanofillers. It shows that a way to increase the efficiency of nanofillers in elastomers is both to decrease the size of particles and to improve the conditions for their linking to the surface of elastomer’s matrix. Crystallization and microcrystallization, which develop in storage, have a significant effect on the entire set of processing properties, first and foremost, by increasing the rigidity of both rubbers and rubber compounds and by making processing difficult. A positive effect of these processes is in providing for, due to stress-induced crystallization, a high cohesion strength, which is required in manufacturing complex-shape articles, such as tyres, hoses, etc. 5.2.2 Effect of precrystallization on the strength at low temperatures
Crystallization prior to deformation also usually results in an increase of elastomer strength; this rise of strength can be accompanied with a decrease of elongations at rupture. An increase of strength at low temperatures (below Tbr0) for preoriented and crystallized specimens has been shown by example of an NR-based rubber [13, 15, 16]; similar data were also obtained for polychloroprene and polyurethane [95, 96, 786]. Herewith, Vr of nonoriented specimens at temperatures T < Tbr0 are the larger, the higher the degree of crystallization and the smaller the size of polycrystals (spherulites) is. The value of Vr does not depend on these parameters, if crystallization and deformation of these polymers crystallizable at room temperature are performed in the region of the rubberlike state plateau, when polycrystals are completely reorganized in the process of deformation [15, 16, 95, 96, 785, 786]. Due to the limited deformability of the specimens crystallized in the transition region from the glassy to the rubberlike state, their strength, as at T < Tbr0, depends on the initial
STRENGTH PROPERTIES OF ELASTOMERS AT LOW TEMPERATURES
123
degree of crystallinity. This was shown for 1,4-cis-polybutadiene (SKD) [513] and polysiloxane [459]. An increase of strength as the result of isotropic crystallization at above Tg for rubber SKD and a dependence of the strength of polysiloxanes on the size of spherulites were also noted in [459]. A reinforcing effect of crystallization also apparently manifests itself in the value of brittleness temperature Tbr of crystallized rubbers, measured by the standard methods. In a number of cases, this temperature is lower than Tbr of amorphous specimens [781]. However, a direct effect of the presence of different-type crystals on the strength of elastomers within a broad temperature range has been studied insufficiently.
124
CHAPTER 5
6
Contribution of Crystallization and Glass Transition to Low-temperature Resistance of Elastomers
Low-temperature resistance of elastomers is determined by the combined action of two above considered processes – glass transition and crystallization. Both processes lead to an increase of hardness, elasticity modulus; a change of the dynamic and other relaxation properties. In contrast to glass transition, which involves all the material, crystallization develops only in small volumes. However, formation of even a minor amount of crystallites results in significant changes of the mechanical properties. The thing is that elastomer monocrystals of the size within the limits of 5–100 nm [124] are typical nanostructures. Comparison of the effect of crystallization and filling on the mechanical properties of elastomers shows that the reinforcing action of crystals exceeds the action of filler particles by tens of times. The efficiency of the action of the microcrystallization process on the mechanical properties is especially pronounced (see Chapter 3). The smaller the maximally possible size of crystallites, the relatively more efficiently their presence affects the rigidity and the relaxation properties. The strength properties are affected by crystallization more efficiently than by fillers (see Chapter 5). Depending on the position of the temperature regions for the manifestation of crystallization and glass transition and the crystallization rate, one process or the other proves to be dominant. Their mutual effect, one on the other, is also possible [98, 117, 125, 751c].
6.1
Mutual interaction of crystallization and glass transition
6.1.1 Effect of crystallization on glass transition For highly-crystalline polymers, crystallization is well known to affect the glass-transition process. Thus, for polyethylene Tg was found to increase with the degree of crystallization rising. The result of the effect of crystallization is that the true value of Tg is unequivocal and depends on the degree of crystallization. The same is true for polypropylene and other polymers of high degree of crystallization.
126
CHAPTER 6
Tg, K
220
1
2 210
3
200 3' 190 180
20
30
40
50 w, %
Figure 6.1 Dependence of the glass-transition temperature Tg on the content of propylene, w, for copolymers of ethylene and propylene according to the data by different authors (1–3); 3c, extrapolation for w = 0.
Such an effect of crystallization on the glass transition of the amorphous component can be due to two causes. The first is the immediate change of mobility of the amorphous component in the presence of crystallites. At their high content, the major part of the amorphous component represents tie chains of chain segments directly adjacent to or occurring in the crystallites’ surface layer. A reduction of their mobility determines the increase of Tg as compared with the initial amorphous polymer. The second cause is the enrichment of the amorphous part by macromolecules or their segments incapable of crystallization and, as a rule, less regular. The effect of crystallization on glass transition can be illustrated by the example of copolymers of ethylene and propylene. Figure 6.1 presents a dependence of Tg of such copolymers on the propylene content, w, according to the data by different authors [37, 132, 471, 477] (a slight discrepancy between these data can be due to the different methods of Tg measurements). The cause of the extreme character of this dependence is the effect of the crystallization of one-type sequences on glass transition. The effect is especially pronounced at a small content of propylene, when the microcrystallization process changes over to the crystallization (see Chapter 3). Apparently, both above considered factors play a role here – the direct effect of the crystallization of long sequences of methylene units on the glass transition of noncrystallized segments of polypropylene and the depletion of these segments in the amorphous component and, the other way round, its relative enrichment by polypropylene segments. If we assume that for polypropylene (for which the effect of crystallization on Tg is slightly lower than for polyethylene) Tg = –20°C, then the extrapolation of the right-hand side branch of the curve for w = 0 gives for polyethylene the value of Tg = –90°C, which is well consistent with the other data. Available literature data on the decrease of Tg with the degree of crystallization rising are, apparently, due to the fact that different degrees of crystallization were obtained not owing to different crystallizations of one polymer, but using polymers of different microstructures. To obtain a polymer with a larger degree of crystallization, specimens with more regular structures were chosen, which is what led to decreased values of Tg. A major part of experimental data for elastomers indicate that Tg is independent of the
CONTRIBUTIONS TO LOW-TEMPERATURE RESISTANCE OF ELASTOMERS
ε 0.10
K b
a 1
0.05
127
1.0 4
2
2 3
0 -40 -30 -20
1
0.5
-10
0
10
20 T, °C
0 -40
4
3
-20
0
20 T, °C
Figure 6.2 Dependence of deformation under a given loadH (a) and recovery K (b) on temperature T for NR-based vulcanized rubber under an isothermal holding for times, h: 2 (1), 5 (2), 16 (3), 48 (4) and 120 (5).
degree of crystallization. These data were obtained for NR [68, 663], polychloroprene [308], polybutadiene [95, 96] (see Fig. 1.1), some types of polysiloxanes [123, 326]. However, for polysiloxanes Tg was noted to be dependent on the degree of crystallization. The difference is outside the limits of Tg determination accuracy [465]. Evidently, the major role in whether or not the effect of crystallization is revealed is played by the degree of crystallization. The higher it is, the larger the probability of revealing a Tg increase with the degree of crystallization is. However, even if the Tg value itself does not change, the properties of elastomers in the transition region do change. First and foremost, in the presence of the crystalline phase, the density goes up and the thermal expansion coefficient goes down [68, 69, 308]. Studies of polychloroprene by the pulsed NMR method [100] have shown that, although Tg measured by both this and other methods does not change during the crystallization process, the mobility of the amorphous phase undergoes significant changes. Measurements of the mechanical properties in the transition region under conditions of partial crystallization have shown their significant changes [95]. Thus, in the initial crystallization times only a slight decrease of the level of the mechanical characteristics can be observed (of deformation H or recovery K ) in the transition region; they look like a decrease of the slope of both the curves H – T or K – T (Fig. 6.2a, curve 2). However, as the crystallization develops in the temperature region, which corresponds to its maximum, the monotony of the curves is violated (Fig. 6.2a, curves 2 and 3; Fig. 6.2b, curves 2 and 3). Further development of the crystallization results in the degeneration of the transition region and in its apparent shift to the region of higher temperatures (Fig. 6.2a and b, curves 4). This effect is also observed in the microcrystallization process (see Fig 3.12a). Owing to the development of the crystallization, one can also observe the degeneration of the loss maximum, corresponding to the D-relaxation process; however, data of this kind for elastomers are extremely scarce, unlike the data on the effect of crystallization on the impact-resilience minimum [386, 450]. Of interest is also to reveal the changes introduced by the crystalline component into the forced elasticity phenomenon. However, data of this kind are absent. Not numerous are also the data on the effect of the crystalline phase on the brittleness temperature Tbr [781]. It is not clear if the changes of Tbr are related in this case directly to the occurrence of crystals or to the change of mobility of the amorphous component and, therefore, to the change
128
CHAPTER 6
of the forced elasticity region. One may think that a respective contribution is made by both processes.
6.1.2 Crystallization from the glassy state Development of the crystallization process in a glassy elastomer is made difficult. The isothermal holding of elastomers at below Tg (annealing) usually fails to lead to the emergence of the crystalline component. However, the appearance of specific heat anomalies at the end of the transition region, described in Chapter 1, can be considered as the result of a structural rearrangement in the glassy state, which rearrangement manifests itself if the heating time is significantly less than the rearrangement time. This structural rearrangement occurs, as we think, by way of forming the physical junctions of a fluctuation nature of the size less than the critical size of the crystalline nucleus. If the isothermal holding (annealing) is performed in the transition region in immediate vicinity to Tg, crystals may form but their size insignificantly exceeds the critical size of the nucleus. In essence, this corresponds to the third region of crystallization by the Hoffman–Lauritzen model or even to microcrystallization (see Chapter 3). Thus, when crystallization can develop in immediate vicinity to the glass-transition temperature, it proceeds to form the smallest crystals or even crystals of the size close to the critical size, as in the case of microcrystallization.
6.2
The ratio of the contributions by crystallization and glass transition to the low-temperature resistance of elastomers
Three ultimate cases of the ratio of glass transition and crystallization in the low-temperature behaviour of elastomers can be singled out [97]: • for rubbers from noncrystallizable polymers, the low-temperature resistance is determined only by the slowdown of the relaxation processes, which lead to glass transition; • for rubbers from slowly crystallized polymers, the low-temperature resistance at a short action of low temperatures – “short-term low-temperature resistance” – is determined by glass transition, and the service life of these rubbers at temperatures higher than Tg – “long-term low-temperature resistance” – is determined by crystallization; • for rubbers from rapidly crystallizable polymers, the low-temperature resistance is determined only by crystallization. Each of these three cases of the contribution of crystallization and glass transition to low-temperature resistance needs to be approached in its own way both with respect to the choice of parameters to assess low-temperature resistance and to the optimal formulation. In the absence of crystallization, low-temperature resistance can be characterized by one temperature Tlow, up to which the required level of the properties is preserved. As we pointed out in Chapter 2, this temperature depends both on Tg and on the character of changes of the properties in the transition region, and is determined by the application of an elastomer. However, even in this case the low-temperature resistance of an elastomer may deteriorate with time, if there is a possibility for its plasticizer, providing for its good low-temperature resistance, to migrate (see Chapter 1). For crystallizable elastomers, the main characteristic of low-temperature resistance is the time of preserving a given level of properties at a certain temperature. This time (tcr) is
CONTRIBUTIONS TO LOW-TEMPERATURE RESISTANCE OF ELASTOMERS
129
minimal at the temperature of the maximal crystallization rate T1; therefore, crystallization poses the greatest danger at temperatures close to T1. Thus, the problem of determining the low-temperature resistance of rubbers based on crystallizable polymers in the general case consists of two parts: first the determination of the lowest temperature Tlow, at which the required level of elastic properties is preserved, and then the determination of the time of keeping the given level of properties as the result of crystallization at the temperature of the maximal crystallization rate T1 [126]. This general scheme is applicable for rubbers based on comparatively slowly crystallized polymers*: NR, 1,4-cis-polyisoprene, butyl rubber, some types of polysiloxanes, polychloroprene and polyurethane, rubbers based on blends of 1,4-cis-polybutadiene with noncrystallizable polymers and 1,4-cis-polyisoprene. For rubbers based on rapidly crystallizable polymers, the low-temperature resistance is totally determined by the crystallization process. This refers to rubbers from crystallizable silicons, as well as to rubbers containing a large amount of 1,4-cis-polybutadiene. For these rubbers, crystallization develops either directly during the cooling, usually even at temperatures higher than T1, or 5–10 min after T1 is reached. For such rubbers, the properties at low temperatures are completely determined by crystallization. Respectively, the usual mechanical methods of low-temperature resistance determination, which envisage measurements at small times, register for such rubbers only the crystallization processes, not the glass-transition processes as for slowly crystallizable or noncrystallizable rubbers [97, 501]. When choosing the composition of rubbers to be operated at low temperatures, it should also be made clear which of the three above considered cases of the crystallization and glass transition ratio takes place. In the first case, all effort should be made to reduce the glass-transition temperature; in the second, it is necessary to provide for a reduction of Tg and an increase of the crystallization time; in the third case, the problem of developing low-temperature resistant rubbers is reduced only to sharply slowing down the crystallization. The values of Tlow for noncrystallizable rubbers, and Tlow, T1 and tcr for crystallizable rubbers should be the basis for the classification of rubbers with respect to low-temperature resistance (see, e.g., [781]).
6.3
Low-temperature resistance of rubber articles
6.3.1 Ways to improve the low-temperature resistance of rubber articles Low-temperature resistance of rubber articles is determined, first and foremost, by the low-temperature behaviour of rubbers from which they are manufactured. Still, a significant role is also played by the design and operational conditions. As it follows from the data presented in Chapters 2 and 3, at an increase of the shape factor ) of an elastomer specimen the changes of the mechanical properties due to the glass-transition and crystallization processes are decreased, i.e., Tlow goes down and tcr increases. Therefore, to improve the low-temperature resistance of rubber articles, it is recommended to increase their shape factor ) [33, 38, 39, 129, 131, 643]. The performance of rubber articles with a larger ), * For rubbers from polymers capable of microcrystallization the same scheme is valid, but one should take into consideration that short-term low-temperature resistance is largely determined by crystallization, too.
130
CHAPTER 6
including thin-layer rubber-metal elements and coatings at temperatures lower than it follows from the tests of common specimens, is well known to practical engineers. An illustrative example of such articles are rubber-metal bridge bearings [149, 154, 480, 601, 607, 685, 688, 689, 692]. They are believed to preserve their performance in the crystallization of polychloroprene or NR, from which their rubber elements are usually fabricated. The cause of this are small changes of rigidity, characteristic of the rubber elements with a large shape factor. One should, however, take into account that these considerations are applicable to rigidity in compression; changes of rigidity in shear have not been assessed for such specimens in the amount required. These data are especially necessary for assessing the low-temperature resistance of such components, as helicopter ball bearings and other components operated under complex loading conditions. The tightness of long-size seals at low temperatures can also be achieved at the expense of a high value of their shape factor [642]. Studies of the low-temperature resistance of sealing gaskets for rotating shafts [33, 144, 203, 312, 403, 432–433a, 607, 760, 760a] have shown that in this case, too, an improvement can be achieved by choosing a proper design. Besides choosing low-temperature resistant rubbers and “low-temperature resistant” designs, other ways to improve the low-temperature resistance of rubber articles can be used. They are associated with creating conditions under which an elastomeric material would not essentially need to preserve its elastic properties at low temperatures. Thus, the difference in the thermal expansion coefficients of elastomers and metals, in contact with which they should be operated, can have a negative effect on the low-temperature resistance of elastomeric components. This difference can lead to seal failure and emergence of thermoelastic stresses, which can be the cause of breakdown of the elastomeric material. But the same effect, at a respective choice of material and design, can be used to develop an additional deformation in the glassy material, which makes it possible to operate elastomers at temperatures lower than Tg [262, 310–312]. In this case, the brittle fracture is also a limiting factor. Examples of advanced designs of seals to be operated at low temperatures can be found in the reports at the conference “Arctic Rubber” [23], a special seminar [99] and a conference on seals [202], as well as in [488, 684]. Another example, when the low-temperature resistance proper of an elastomer material or articles do not limit the temperature range of its performance, is the use of local heating, as is recommended, e.g., for shock absorbers of diesel locomotives, which are heated by low-voltage electric current [327]. However, it is evidently difficult to maintain the operational temperature at a much higher level than the ambient temperature in this way. The heating of the sealing unit was also used in US space shuttles after the Space Shuttle Challenger disaster in 1986, when an O-ring seal in the rocket booster of the shuttle failed. The failure was caused by the poor long-term service ability of the fluorine rubber seal [487] at temperatures higher than its Tg, apparently due to its microcrystallization (see Chapter 3). In essence, most rubber articles operated under the dynamic loading conditions perform at temperatures higher than the ambient temperature. For shock absorbers, this temperature increase is small, but even this increase can make it possible to operate articles at slightly lower ambient temperatures than Tlow. It is important, however, that no brittle fracture occur at the initial load application time, i.e., the forced elasticity region be broad enough, and the brittle strength be high. Operation of tyres at temperatures lower than the glass-transition temperature for rubber is based on the same effect. The initial motion at a very small velocity provides for the low values of Tbr, and owing to the self-heating, which takes place even under these
CONTRIBUTIONS TO LOW-TEMPERATURE RESISTANCE OF ELASTOMERS
131
conditions, a temperature higher than Tg settles in the tyre if the rubber composition is correctly chosen. One should, however, bear in mind that a multiple load at temperatures lower than Tg can lead to formation of microcracks, the accumulation of which affects the general performance of the rubber article (see further). The same effect takes place in the operation of seals of rotating shafts. However, in this case, the sealed medium may leak under static conditions. For non-freeze-resistant media such leakages are not observed, but the startup of the machine is made difficult. Thus, while providing for the performance of rubber articles at low temperatures, the above considered techniques have limitations. Therefore, the problem of developing elastomeric materials with low Tlow remains very topical. 6.3.2 Assessment of low-temperature resistance of rubber articles from the low-temperature behaviour data for elastomers As it follows from the above, for correct determination of the low-temperature resistance of rubber components they should be tested directly in the operated units and under conditions close to operation conditions. Herewith, one should take into consideration both the specific features of the design and the features of low-temperature resistance and thermal conductivity of elastomeric materials. This requires a correct, scientifically substantiated choice of the testing regime in each particular case. Of special importance is the correct choice of thermostatting time. This is important not only for crystallizable elastomers, but also for bulky components and units. In this case, it takes time for the ambient temperature to settle.* One should also take into account the possible prolonged processes of plasticizer migration [563] (see Chapter 1). Reliable data on the low-temperature resistance of rubber articles, besides those considered above, are comparatively few [5, 393, 488, 590, 672, 693]. As the low-temperature resistance tests of rubber components are usually associated with great difficulties, an important task is to assess the low-temperature resistance of rubber articles by the low-temperature resistance of rubbers. This requires choosing a correct testing method, which correlates with the low-temperature resistance of a given type of rubber articles (a critical method) and finding a critical value of its result. The only type of rubber articles for which this has been done completely are immovable sealing components. For them, it is accepted that the critical method is recovery. Long-term experience accumulated in comparing the low-temperature resistance of rubbers with that of ring seals has shown that for round cross-section rings with the ratio d/D < 5 of cross-section diameter d to ring diameter D, the critical values of recovery Kcr = 0.2. This is supported both by glass-transition [344] and crystallization [490, 491] tests. As the shape factor )in the case of rectangular cross-section rings is high, Kcr can go down to 0.05. Comparison of the low-temperature resistance of various elastomeric materials, in particular, their glass-transition temperature Tg with the low-temperature resistance of movable seals, assessed by the average number of cycles nr before fracture, showed that there was no equivocal correlation [496]. This is due to the effect of the strength of the material under the testing conditions on the value of nr. This correlation is, naturally, also absent if a seal is made from an elastomer crystallizable under the testing conditions. * An increase of the dynamic rigidity of large-size shock absorbers at a temperature decrease is the larger, the greater the deformation amplitude is, which is explainable by an increased nonuniformity of the temperature pattern [729].
132
CHAPTER 6
It has also been shown that the low-temperature resistance of rubber/ metal seals (gaskets) for rotating shafts, determined as the sealing-preservation temperature, is stipulated by the recovery rate of the gasket edge and is related to the rate of recovery of the elastomer after compression* [125, 607, 656, 760a]. Good results are obtained by comparison with the recovery determined by loading the specimen at the testing temperature [432, 433] (a variety of the recovery method, see Chapter 1). The critical value of recovery of an elastomer, Kcr, depends in this case on the design and quality of the sealing unit [120, 125, 607]. The loss of performance of immovable ring seals owing to crystallization or glass transition is reversible. In the heating of a sealing component at above Tlow for the case of the glass transition or higher than the melting temperature Tm in the case of the crystallization, the hermeticity is recovered (if there was no brittle fracture of a sealing component). Thus, the hermeticity of the units with ring seals from silicon rubber SKTV-1 lost in crystallization at –55°C is restored by heating the unit** at above Tm = –35°C. However, for immovable sealing articles, contacting the sealed surface for a long time, the seal failure temperature can be significantly lower than Tlow, and time tcr is higher (tcr is the time corresponding to the values of Kcr = 0.2 for the rubber in the process of crystallization). This was first shown [38] for crystallizable rubbers, the loss of hermeticity of immovable components began at time t > tcr. The loss of hermiticity in this case is due to the formation of an adhesion bond between the surface of the seal and the unit to be sealed. The development of an adhesion of the rubber surface to the metal surface to be sealed can be the way to improve the low-temperature resistance of immovable seals [168, 170, 172, 328]. An improved adhesion and, therefore, low-temperature resistance are promoted in this case by a preliminary heat ageing, which leads to the release of substances contributing to the improvement of adhesion, on the surface of the rubber. As the result, at optimal ageing regimes of seal failure temperature could be decreased by 100°C and more as compared with Tlow. However, the reliable operation of such joints at low temperatures is rather limited. The mechanical break of contact, e.g., in vibrations, can disturb the adhesion bonds, which can result in seal failure. Performance of conveyor belts, hoses and other rubber-reinforced articles is usually provided for up to a temperature, where the recovery of rubber Kcr < 0.2*** (e.g., Kcr | 0.05). Special techniques have been proposed to assess the low-temperature resistance of rubber/fabric materials [641]. Special care should be taken when determining the low-temperature resistance of rubber articles from crystallizable or microcrystallizable elastomers. Correct temperature, time and deformation conditions of tests should be chosen. As these processes develop in time, at temperatures above Tg, and are the faster, the larger a stress is, one should always bear in mind these underlying potential problems.
* The possibility was shown of using the recovery kinetics data obtained by the usual method (see Chapter 1) in combination with the principle of temperature–time superposition for assessing the low-temperature resistance of rubber components operated under dynamic conditions [122, 656]. ** Data by V.V. Sedov. *** At Kcr < 0.2, the accuracy of recovery determination is small.
CONTRIBUTIONS TO LOW-TEMPERATURE RESISTANCE OF ELASTOMERS
6.4
133
Effect of repeated exposure to low temperatures on the performance of rubber articles
6.4.1 Reversibility of changes due to glass transition and crystallization A specific feature of the changes occurring in elastomeric materials under the action of glass transition and crystallization, i.e., at low temperatures, is their reversibility. Heating up to a higher temperature than the glass-transition temperature Tg (in the absence of crystallization) or equilibrium melting temperature Tm0 (if the changes are due to crystallization) returns the elastomer or an elastomeric article into the initial state. The temperature in the elastomer should be checked to be certain that it did settle at a respective level. The time required for this is the greater, the larger the volume of the elastomer-containing material, component or unit to be heated is. Crystallizable elastomers require not only higher temperature (as Tm0 < Tg) but also the supply of additional heat corresponding to the melting heat of the crystallized material:
Q
'HCm ,
(6.1)
where 'H is specific melting heat in terms of a totally crystallized elastomer; C is the degree of crystallization of a rubber; m is the mass of the crystallized rubber in the considered volume of the elastomeric material. It is known that changes of the mechanical properties related to crystallization prevent the normal processing of rubbers and rubber compounds crystallized in storage (i.e., at room and close temperatures). First and foremost, this applies to isoprene and chloroprene rubbers and rubber compounds based on them. Decrystallization (“steaming”), i.e., a heating prior to processing, is widely used in rubber industry. For noncrystallizable rubbers, i.e., to rule out the effect of glass-transition-related changes in the mechanical properties of a rubber, it is necessary that a temperature corresponding to the onset of the rubberlike plateau settle in the entire bulk of the rubber. For most elastomers, as we pointed out above, this temperature is close to room temperature. It is also necessary to stabilize if not eliminate the physical junctions in the noncrystallized rubbers. However, their role in elastomer processing has not in fact been studied. When working out the heating regimes, it is necessary to take account of both the weight of the material heated and its thermal conductivity. It is to be recalled that the thermal conductivity of carbon black-filled rubbers is 2–3 times higher than that of unfilled rubber [97]. 6.4.2 Effect of temperature cycling on the performance of rubber articles Reversibility of changes due to the impact of low temperatures is one of the reasons for insufficient attention to the low-temperature resistance of rubber articles. However, it is not always that complete reversibility takes place. Thus, if a crystallized elastomer is heated to the melting temperature Tm but not the equilibrium melting temperature Tm0, the properties of the elastomer would not recover completely. This is a manifestation of the “memory” of crystallizable elastomeric materials. Indeed, even if the major part of the crystals melted during the heating, but some nuclei remained, the repeated crystallization of such a system would proceed much faster, and the disturbances in the performance of rubber articles
134
CHAPTER 6
would start earlier than in the first cycle of crystallization. Repetition of the crystallization–melting cycles in the case if the latter occurs at Tm < Tm0 leads to accumulation of the nuclei of crystallization and its increasing acceleration [631]. The role of this effect is rather significant in long-term storage of rubbers from crystallizable polymers, e.g., NR, synthetic polyisoprene and polychloroprene, under Extreme North conditions. Herewith, a peculiar effect of a “hysteresis” is observed [115, 121]. It should be also taken into account that all changes in the mechanical and other properties of elastomers, due to glass transition or crystallization, are reversible only in the case if these processes are not accompanied with brittle fracture of the material in the glassy or the crystallized state. It may start if an elastomeric material in such a state was subject to impact loads. Another source of fracture can be thermoelastic stresses, which occur in an elastomer at the interface with other materials of lower thermal expansion coefficient values (metal, polymer fibres etc.). And, finally, a decrease of the volume of an elastomer during the crystallization, both isotropic and anisotropic (in a stressed state) can be a source of breaking stresses. In the two latter cases, these breaking stresses usually lead not to a macroscopic breakdown of an elastomeric material, but to the emergence of microbreaks. These microbreaks occurring at the interface of the elastomer with the reinforcing material can accumulate in the multiple repetitions of the glass transition or crystallization–melting cycles. They can cause poor performance of rubber/metal and rubber/fabric components. This effect was observed in rubber/fabric belts from rubbers based on polychloroprene [244, 245]. A triple repetition of the crystallization–melting cycles led to a threefold decrease of the life service of these belts at room temperature. The same effects can be a consequence of the changes of volume in glass transition. Thus, after thermal cycling from –70 up to +90°C the room-temperature strength of specimens modelling the cable-reinforced conveyor belts and fabricated from rubber based on styrene-butadiene polymer SKS-30 decreased twofold. Herewith, the scatter of the results increased (no changes of the strength properties were observed after a control holding of the specimens at +90°C) [118, 119]. Microseparation of the elastomer from short filling fibres leads to an abnormal increase of the linear expansion coefficient in transition of the rubber/fibre material from the rubberlike to the glassy state [783]. There are also reasons to believe that a significant role in the microbreaking processes is played by moisture. Getting into the cavities formed by microbreaks and crystallizing*, water contributes to their increase. A multiple cyclic temperature change may lead to a decrease of service life of such rubber components as belts, conveyor belts, hoses, cables and tyres. A certain effect on the formation of microbreaks is also rendered by the cooling rate. The higher it is, the more probable the emergence of such breaks is (in the absence of moisture). This should be taken into account when using the cryogenic method of processing with the aim to trim flashes in rubber/metal articles. A specific role in deteriorating the low-temperature resistance of rubber articles in cyclic temperature changes plays the behaviour of plasticizers. The processes of their sweating or washing out in an oil medium at a long-term storage even at room temperature can result in a sharp deterioration of the low-temperature resistance of the components. The role of plasticizer’s sweating in deteriorating the low-temperature resistance in longterm storage of articles under winter conditions in Canada was shown by the example of * It is to be recalled that water increases in volume during crystallization.
CONTRIBUTIONS TO LOW-TEMPERATURE RESISTANCE OF ELASTOMERS
135
conveyor belts [590]. A deterioration of the low-temperature resistance was confirmed by round-the-year prolonged exposure of elastomer seals and their specimen models on a testing ground in Yakutsk, Russia [121]. Exposure under the action of aggressive liquids, in particular, petroleum, results in a significant deterioration of the low-temperature resistance and other properties of the components [563–565]. A long-term storage of specimens of both crystallizable and noncrystallizable elastomers under Extreme North conditions (at a testing ground in Yakutsk) and in the absence of aggressive media led to a decrease of their breaking strength, which, apparently, is also due to the cyclic action of low temperatures [121]. All these effects, in our view, represent the main source of a decrease of the performance times for components operated under North conditions. Unfortunately, they have been studied absolutely insufficiently. At the same time, the development of measures preventing the emergence of microbreaks, is an undoubted reserve of increasing the reliability of rubber goods intended for operation under conditions of low temperatures [115, 130].
136
CHAPTER 6
7
Low-temperature Resistance Specifics of Particular Elastomers
This chapter considers the specific features of the low-temperature resistance of individual rubbers, which are stipulated by the distinction of their structure. The complexity of describing the major physical properties, including the glass transition and crystallization of elastomers, is in the absence of elastomer specimens, which could be considered to be really standard [670]. Differences in the catalytic systems and polymerization conditions lead to production of rubbers, which differ by microstructures, character of alternation of monomer units, molecular mass and molecular-mass distribution. Sometimes not even registered by the traditional structural and chemical methods, these differences still have an impact on the physical properties. A significant role can be also played by the differences in the testing conditions. Therefore, within even one chemical structure of an elastomer, it could be difficult to obtain quantitative data, which could be considered as physical constants characterizing its properties. In full measure, it also pertains to the phenomenon of crystallization whose kinetic, thermodynamic and morphological features for one type of rubber can change within broad limits. Glass transition is slightly less sensitive to small changes of chemical and physical structure, but for this process, too, chemical-structure changes within the range of one rubber grade can lead to changes of the characteristics of this process, in particular, Tg and T. Besides, contributions of crystallization and glass transition to lowtemperature resistance are different for different types of elastomers, which also requires special consideration. Table 7.1 presents averaged data on the glass-transition and crystallization parameters practically for all currently used elastomers. Besides the results of the authors’ own measurements, the literature data are used [221, 444, 476, 670, 744, 746].
7.1
Styrene-butadiene rubbers
Statistical copolymers of butadiene and styrene are widely used as general-purpose rubbers. The styrene content w can be from 10 up to 85%. The larger it is, the higher the glasstransition temperature Tg is (see Fig. 1.14a). Styrene-butadiene rubbers are produced by methods of emulsion polymerization and polymerization in solution [221]. Depending on the production method and temperature, rubbers can differ by the structure of the butadiene part. Thus, for an emulsion-polymerization rubber (polymerization temperature, 5°C) the butadiene part contains 7% of 1,4-cis-units, 72% of 1,4-trans- and 21% of 1,2-units. About
*Microcrystallization; ** no reliable data.
Polysulphide trans-Polypentenamer Propylene oxide
Urethanes
Silicons
Chloroprene Ethylene-propylene Butyl rubber Fluoroelastomers
Butadiene acrylonitriles
Styrene-butadienes
Natural isoprene Synthetic isoprene Butadienes
Rubbers
–(68–73) –(68–71) –(102–112) –(95–100) –(48–52) –78 –75 –(52–54) –30 –(47–50) –(40–42) –(25–30) –(40–42) –(55–65) –69 –(18–20) –(40–50) –(125–130) –(115–110) –110 –105 –78 –(130–132) –127 –(35–30) –(35–40) –(50–55) –(100–105) –75
Tg, °C
Table 7.1 Low-temperature parameters of elastomers.
–25 –25 –55 –55 – – – – – – – – –10 * – – * –78 –78 –78 – * –78 –78 +5 +5 –30 –30 **
T1, °C +40 +40 +10 0 – – – – – – – – +80 +140 ** – ** –25 –40 ** – ** ** –30 +80 ** ** ** **
Tm0, °C 0–(– 40) –(5–40) –(25– 80) –(40–70) – – – – – – – – +10–(– 30) –50– (+70) –(20–40) – ** –(35–100) –(55–90) ** – ** ** – (50–90) +10–(–30) ** –(15–30) –(20–80) **
–(63– 68) –(63–65) –(30–50) –(90–95) –(40–45) –(70–75) –(65–70) –(45–48) –(25–30) –(40–52) –(35–45) –(20–30) –(35–40) –(50–55) –(55–60) –(12–15) –(30–45) –(50–60) –(65–80) –(95–100) –(90–95) –(45–65) –(125–128) –(50–65) –(25–30) –(30–35) –(45–50) –(40–50) –(60–65)
102 –105 102 –105 1–103 102 –104 – – – – – – – – 102 –104 102 –105 104 –106 – 102 –106 0.01–102 0.01–103 ** – ** ** 0.01–102 10–102 ** 102 –104 0.01–102 **
Tlow, °C
tcr, min
For vulcanized rubber compounds 'Tk, °C
138 CHAPTER 7
LOW-TEMPERATURE RESISTANCE SPECIFICS OF PARTICULAR ELASTOMERS
139
30% of styrene units are isolated, about 40% are arranged as diads [221]. An increased content of 1,2-units determines a higher value of Tg obtained by extrapolation of the dependence of Tg on the styrene content w for w = 0 than for rubbers with the high content of 1,4-cis-units (see Fig. 1.14a). The butadiene part of solution-polymerization rubbers contains a much larger amount of 1,4-cis-units (from 40 up to 50%), 47% of 3,2-trans- and from 3 to 13% of 1,2-units. The traditional styrene-butadiene rubbers of emulsion polymerization and elastomeric materials based on them are typical noncrystallizable elastomers. Their low-temperature resistance is totally determined by the glass-transition process. Copolymers with w < 30% can be attributed to low-temperature resistant elastomers. Rather low-temperature-resistant is the solution-polymerization rubber DSSK-18, for which Tg = –(75–79)°C [221], i.e., much lower than it would have been for an emulsion-polymerization rubber at the same w, as its butadiene part contains a smaller number of 1,2-units. Addition of styrene-butadienes, especially with small w, to the SKD rubber efficiently suppresses the crystallization of the latter. Considering the effect of the type of the vulcanizing group on Tg (see Chapter 1), the most low-temperature-resistant rubbers based on styrene-butadienes are produced using peroxide vulcanization. Introduction of plasticizers results in a decrease of Tg and an improvement of the low-temperature resistance of styrene-butadiene rubbers [417, 419]. They themselves (at a styrene content w < 20%) can also serve as permanent plasticizers for butadiene acrylonitrile rubbers [563, 564, 728]. The pattern of the impact of filling on the low-temperature resistance of these rubbers are the same as for other elastomers (see Chapters 1 and 2).
7.2
Butadiene acrylonitrile rubbers
Statistical copolymers of butadiene and nitrile of acrylic acid are the most widely used rubbers, which are resistant to oils and solvents. The content of acrylonitrile w can change from 10 up to 40%. The larger w, the higher Tg is (see Fig. 1.14a), i.e., the worse the lowtemperature resistance of these polymers and rubbers based on them is, but the better their resistance to aggressive media is. These rubbers are produced by emulsion polymerization of butadiene and acrylonitrile. These polymers and rubbers based on them, are, as styrene-butadienes, typical noncrystallizable elastomers, and their low-temperature resistance is totally determined by glass transition. The content of 1,2-units in the butadiene part is practically the same as in styrene-butadiene rubbers of emulsion polymerization (see Fig. 1.14a). As it follows from the data of Fig. 1.14a, only the rubbers with w < 20%, e.g., SKN-18, can be considered to be low-temperature-resistant. Deviations from the statistical character of distribution of acrylonitrile lead to the emergence of several Tg. As for the structure of the butadiene part, it is the same as in styrene-butadiene rubbers of emulsion polymerization. A change of polymerization conditions, in particular, of an emulsifier, can deteriorate the low-temperature resistance of vulcanized rubber compounds without changing Tg of the rubbers, which is due to changes of the vulcanizate structure in the presence of the emulsifier [127, 426, 469]. Change-over to alternating copolymers of the same chemical composition, but with a high content of alternating 1,4-trans-units (see further), leads to a decrease of the glasstransition temperature; however, a clearly pronounced tendency to the crystallization of the repeating units, which includes both comonomers, appears.
140
CHAPTER 7
To reduce Tg and increase the low-temperature resistance of these rubbers, plasticizers are widely used [417, 418]. However, it is rather difficult to produce rubbers, which are low-temperature-resistant down to –60°C, especially after thermostatting in oils. To achieve a better low-temperature resistance (in combination with good oil resistance), styrene-butadiene additives to blends based on butadiene acrylonitriles are used [563, 564, 728]. The problem of plasticizer sweating and, especially, washing-out, considered in Chapter 6, pertains primarily to rubbers based on butadiene acrylonitriles. As the result of these processes, the long-term low-temperature resistance of rubbers based on these noncrystallizable polymers deteriorates and should be checked. Vulcanization with a small amount of sulphur [690], tetramethyl thiuram disulphide and peroxides, as well as introduction of low-active fillers (no more than 60 weight fractions) results in the improvement of low-temperature resistance [728]. Introduction of ceolites as fillers leads to some improvement of low-temperature resistance [564]. A set of techniques, which enable the best low-temperature resistance of rubbers and articles therefrom, was developed by N.N. Petrova in Yakutsk [563]. In the recent years, hydrogenated butadiene acrylonitrile rubbers find ever increasing applications [1, 39, 288, 290, 301, 488]. Owing to low unsaturation, their resistance to various kinds of ageing noticeably exceeds that of common butadiene acrylonitrile rubbers. However, no systematic data on the low-temperature resistance of these elastomers are available in the literature. Special attention should be paid to their long-term low-temperature resistance, as the possibility of microcrystallization is not to be ruled out in some cases.
7.3
Isoprene rubbers
Both natural and synthetic 1,4-cis-polyisoprenes are stereoregular homopolymers containing more than 90% of 1,4-cis-units*. Depending on the type of catalytic system and degree of conversion, rubbers of different stereoregularity, molecular mass and molecular-mass distribution are fabricated [221, 445]. Polyisoprenes produced using lithium catalysts contain up to 93% of 1,4-cis-units, up to 2% of 1,4-trans- and up to 5% and more 3,4-units; of 1,4-units, 1–2% are attached head to head and 2%, tail to tail. During the polymerization on Ziegler–Natta catalysts, the content of 1,4-cis-units is up to 98%; 1,4-trans-, up to 2%; 3,4-units, up to 1%; herewith, 1% of 1,4-units are attached head to head and approximately the same amount, tail to tail. The use of cobalt and neodymium catalysts [726] enables production of a rubber of greater stereoregularity. As for natural rubber, it is thought that all 100% of 1,4-cis-units are attached exclusively head to tail [283]. However, a number of signals in the NMR spectra can be interpreted as proof of three units of a trans-structure at the beginning of the molecular chains from cis-sequences [701]. The low-temperature resistance of polyisoprenes is determined by their microstructure. Thus, introduction of 1,4-trans- and 3,4-units in the amount of up to 20% at their statistical distribution leads to breaking of a structure and suppression of crystallization. As the result, low-temperature resistance is determined exclusively by glass-transition processes; * Gutta percha – 1,4-trans-polyisoprene – also pertains to isoprene elastomers. However, this material failed to find wide application in production of rubber articles, so its properties are not considered here. Blends of NR and gutta percha do not covulcanize, so no reinforcement as compared with NR is observed [85].
LOW-TEMPERATURE RESISTANCE SPECIFICS OF PARTICULAR ELASTOMERS
141
what is more, with the content of 3,4-units going up, Tg rises linearly (see Fig. 1.13). For NR, Tg = –72°C; for commercial synthetic isoprene rubbers, it is slightly higher (Tg = –(68–72)°C). Long-term low-temperature resistance of stereoregular polyisoprenes is determined by the crystallization processes. It should be noted that NR have been an excellent model for experimental and theoretical crystallization studies of elastomers in general, so the crystallization of NR has been studied the most completely. Most data obtained carry over to synthetic polyisoprenes. The temperature of the NR maximum crystallization rate T1 = –(25–26)°C [69, 95, 96, 233, 702b], the equilibrium melting temperature Tm0 = 39.1°C [221] or 40°C [95, 96]. The crystallization parameters of NR at T1 essentially depend on its grade. Within one grade, the crystallization parameters depend on protein content [559, 723, 734], products of acetone extract [231] etc. Gel and sol fractions of NR have different crystallization rates [454]. NR melting data obtained by the dilatometric method [69, 748] are given in Fig. 3.3. At usual heating rates, DSC thermograms for NR, as for SKI-3, register two melting peaks [186, 336], which are sometimes ascribed to crystals of different morphology. Edwards and Phillips [186], based on the data of thin-film electron microscopy, found two types of lamellar crystals to form. The crystals have different thicknesses l, grow at different rates and have different free energiesVe: 24 and 50 MJ/m2. However, these crystals do not belong to different crystallographic modifications, as the elementary cell parameters are the same for all crystals of 1,4-cis-polyisoprene (see Table 3.1). NR melting studies performed by the DSC method at different heating rates* have shown that the melting process proceeds in a more complex way than that described in Chapter 3 for 1,4-cis-polybutadiene. The data obtained give no unequivocal answer on the distribution of the initial crystals by size; it is not clear if the high-temperature peak at Tm2 is due to the melting of crystals formed during the rearrangement (as it takes place for 1,4-cis-polybutadiene) or the initial crystal size distribution is also bimodal [336, 751]. The relatively high value of Tm0, which determines the crystallization of NR in storage at room and lower temperatures, leads to the necessity to melt its crystals (decrystallization) prior to the treatment – “steaming”. An increase of the content of 1,4-cis-units in synthetic isoprene rubber makes it necessary to decrystallize this rubber, too. The presence of the crystalline phase increases the rate of carbon black–rubber gel formation and makes processing difficult [321]. In Malaysia, NR is filled with special crystallization-inhibiting oils with the aim to prevent crystallization in storage and transportation and rule out the “steaming” routine. The development of crystallization in NR is greatly affected by the content and type of protein impurities, which play the role of low-temperature crystallization nuclei [139, 723, 734]. These impurities are especially important during the stress-induced crystallization, as they form a network of weak crosslinks, which contributes to chain unfolding, i.e., to a decrease of the critical deformation value Hk and to the transition to crystallization with unfolded chains. This, first and foremost (along with a different molecular-mass distribution), is what explains the difference in the cohesion strength of unfilled rubber compounds based on NR and synthetic polyisoprene [239, 711, 712], in particular, SKI-3. The effect of acetone extract substances on the NR crystallization kinetics was studied in detail in [136]. Purification of NR leads to a strong decrease of the measured heat of fusion, i.e., of the degree of crystallization. Reprecipitation of extracted rubber with chloroform leads to the * Data by N.M. Zorina.
142
CHAPTER 7
further decrease of the heat of fusion. The melting curves of the acetone extract products indicate the presence of at least two crystallized substances with the melting peaks at temperatures of –8°C and +14°C. Introduction of acetone extract products to synthetic polyisoprene leads to some acceleration of crystallization; however, the crystallization rate is lower than that in NR. The use of various organic additives to synthetic polyisoprene makes it possible to increase slightly the cohesion strength of rubbers based on it [581]; however, attempts to reach the level of NR have been unsuccessful until now. A combined use of neodymium catalysts with a selection of additives actively promoting the stress-induced crystallization would enable, in our mind, production of synthetic polyisoprene similar in its properties to NR. However, its use instead of the existing grades would bring forth the problem of obligatory “steaming” and crystallization-related low-temperature resistance. Stress-induced crystallization of NR and synthetic polyisoprene has been studied in detail in [141, 142, 309, 382, 383a, 387, 556]. Interest in its pattern and its relation to the strength properties has renewed in the recent years [139, 239, 711, 712], see Chapters 4 and 5. Regularities of the effect of density and type of three-dimensional network were considered in Chapter 3 using 1,4-cis-polyisoprene as an example, so we would not discuss them here. Recall that the effect of the type and density of three-dimensional network on stress-induced crystallization of polyisoprene differ from their influence on the isotropic low-temperature crystallization [139, 239, 711, 712], see Chapter 4. As for other elastomers, addition of noncrystallizable styrene-butadiene rubber slows down crystallization of NR [210]. The effect of crystallization on the processing of rubber compounds, and then on the properties of technical rubbers was shown in [321] by the example of 1,4-cis-polyisoprene. Crystallization of natural and synthetic 1,4-cis-polyisoprene and its parameters for different types of rubbers was considered in detail in a number of reviews [69, 424, 640, 739].
7.4
Butadiene rubbers
Polybutadiene with the predominant 1,4-cis-unit content (w |90% and greater) is, as polyisoprene, a homopolymer of regular structure, so its low-temperature resistance is determined by both glass transition and crystallization. The glass-transition and crystallization parameters of 1,4-cis-polybutadiene essentially depend on its microstructure, which, in turn, depends on the type of catalytic system. Catalysts used commercially are based on titanium, cobalt and nickel [221, 669, 670]. Titanium polybutadiene (rubber SKD) is characterized by a narrow molecular-mass distribution and a practically absent branching. The content of 1,4-cis-units w is from 87 up to 93%; 1,4-trans-units, from 3 up to 8% and 1,2-units, from 3 up to 5%; Tg = –105°C. Depending on the MMD regulation conditions in polymerization, Mw/Mn may vary from 1.5 up to 5 [503]. Rubber (SKD-2) obtained using cobalt catalysts is characterized by a considerable branching, which increases with the rising conversion; w varies from 93 up to 98%; the content of 1,4-trans-units, from 2 up to 4%; and 1,2-units, from 2 up to 4%. MMD of these rubbers is sufficiently broad: Mw/Mn = 3–5; Tg = –107°C. Regulation of the properties of cobalt catalysts makes it possible both to regulate the rate of synthesis and produce rubbers with improved properties [359, 360, 659, 660]. Catalysts based on S-allyl nickel complexes enable production of polymers (e.g., SKD-3) containing up to 98% of 1,4-cis-units, from 1 up to 3% of 1,4-trans- and from 1 up to 3% of 1,2-units. The rubber is characterized by a broad MMD (Mw/Mn = 5–8) and a
LOW-TEMPERATURE RESISTANCE SPECIFICS OF PARTICULAR ELASTOMERS
143
considerable branching; Tg = –110°C. In recent years, catalytic systems based on rare-earth metals have been widely used. They proved unique in their stereoselectivity: the content of 1,4-cis-units in polymers produced exceeds 98%; Tg = –110°C. The dependence of Tg of polybutadiene on the content of 1,2-units is presented in Fig. 1.13, curve 2. The use of neodymium catalysts enabled production of polymers, providing improved strength and dynamic properties of technical rubber obtained [282, 574, 761]. However, low-temperature resistance of rubbers based on these polymers is determined completely by the crystallization process [128, 501, 778]. Crystallization of 1,4-cis-polybutadiene and rubbers based on it has been well studied (see, e.g., [30, 92, 93, 95, 96, 109–112, 138, 163, 192, 217–219, 266, 314, 338, 380, 441, 452, 516, 522, 524, 625, 626a, 694, 702a, 754, 773, 774–776, 778]). The maximum crystallization rate temperature T1 = –(55–56)°C and does not practically depend on the microstructure [775]. At this temperature, polybutadienes with w > 93% crystallize very rapidly. The cis-unit content affects the crystallization rate very much [128, 441, 452, 516, 522, 524, 774, 775, 778], see Fig. 3.1b. An increase of the crystallization rate in passing from w = 93% to w = 96% is apparent already during the cooling at a given rate (see Fig. 1.1). If the cis-unit content is higher, then even short-term low-temperature resistance is totally determined by crystallization even in the case of rubbers from blends with SKI-3 [128, 501, 778]. At w < 75%, the crystallization of polybutadiene does not develop for a sufficiently large observation time [516, 522]; at the same content of cis- and trans-units, polybutadiene does not crystallize at all. In principle, isotropic crystallization of 1,4-cis-polybutadiene can be observed within the temperature range from –90 up to –10°C. The melting studies of polybutadienes of various structures revealed the melting properties of elastomers: occurrence of two melting peaks [152] (or three peaks at the crystallization temperature much lower than the melting temperature) [110, 773, 774]. The role of reorganization in the melting of elastomers with different w and crystallized in different conditions was shown (see Chapter 3 and Fig. 3.5). The dependences of the temperatures corresponding to the melting peaks and of the melting-termination temperature on the crystallization temperature for polybutadiene are presented in Fig. 3.6. Their extrapolation to T = Tm can give the value of the equilibrium melting temperature of 1,4-cis-polybutadiene Tm0. It has been shown to weakly depend on w, the dependence obeying the Flory equation (3.15c)*. Therefore, some discrepancy of the literature data takes place; thus, Wunderlich [751] gives the value Tm0 = +6.3°C obtained experimentally, and Tm0 = +11.7°C calculated from the decrease of the melting temperature of copolymers. There are also other data in the literature [139]; e.g., Tm0 = 12.5°C [669]. The equilibrium melting temperature Tm0 little depends on w, however, Tm2 at a given crystallization temperature T may substantially depend on w; this dependence is linear, and its slope is the larger, the lower T is [110, 380, 773]. In a number of cases, the dependence was found to be more complex, which is due to the disturbance of the statistical distribution of the “disturbing” units. The dependence of Tm2 on the density of the network, as determined by the value 1/Mc, is also linear for the network produced by monosulphide bonds, and its slope is the smaller, the higher the crystallization temperature is [110, 773]. To inhibit the ability of 1,4-cis-polybutadiene to crystallize, its copolymerization with isoprene is used [282, 360, 525, 561, 761] (for the Tg data of copolymers, see Fig. 1.13, curve 1). However, an increase of Tg is not always compensated for by the required level of the crystallization rate decrease to ensure long-term low-temperature * Data by N.M. Zorina.
144
CHAPTER 7
resistance. Copolymerization of butadiene with piperilene was also tested [108, 727], as well as its combination with trans-polybutadiene [93]. Copolymers with alkyl butadienes and pentadiene [221, 669] were obtained. Some decrease of the crystallization rate is also rendered by cyclization; the suppression of crystallization is significantly affected by cis-trans-isomerization. An efficient way of inhibiting the crystallization of 1,4-cis-polybutadiene is to combine it with other rubbers (Chapters 1 and 3) [108, 707–710]. The most efficient crystallization-suppressing method is to use blends with styrene-butadiene rubbers, especially those that contain a minor amount of styrene [708]. Rather efficient is to combine 1,4-cis-polybutadiene with noncrystallizable polybutadiene of lithium polymerization; herewith, the ultimate degree of crystallization in terms of crystallizable rubber remains constant, as in the case of blends with 1,4-cis-polyisoprene [108, 707–710]. The specific effect of stress on crystallization has been studied insufficiently. It has been shown that the equation of the characteristic curve (3.17) is observed for technical rubbers based on this polymer within the same limits as for 1,4-cis-polyisoprene. Only specimens of 1,4-cis-polybutadiene with w > 98.5% are capable of stress-induced crystallization at room temperature [163, 139, 478, 711]. Crystallization at stress begins at the value of Hk | 4.5. Due to the large crystallization rate of 1,4-cis-polybutadiene, there is the specificity of the effect of the vulcanizing group on its crystallization: for specimens with w > 92%, the low-temperature resistance of rubbers on its basis with C–C and monosulphide bonds is totally determined by crystallization. So the advantages due to its low Tg are not realized; even short-term tests show a deterioration of the elastic properties at T = T1. The use of vulcanizing groups forming di- and polysulphide bonds changes the situation. Short-term low-temperature resistance is determined by glass transition, i.e., rubbers are very low-temperature-resistant, but a small increase of the holding time (up to 1–2 hours) leads to the loss of the elastic properties. At w > 98%, crystallization determines, as we noted above, the low-temperature resistance of rubbers with any vulcanizing group. Blends of 1,4-cis-polybutadiene with 1,4-cis-polyisoprene and noncrystallizable rubbers belong to the class of elastomers, whose short-term low-temperature resistance is determined by glass transition. Usually, the lower boundary of the performance, Tlow, is determined by the conditional glass-transition temperature Tg1 (see Chapter 2), which is 30–40°C higher than Tg of 1,4-cis-polybutadiene. Long-term low-temperature resistance is determined by crystallization in this case, too. When using blends of PB with styrene-butadiene rubbers at a small styrene content, rubbers can be produced with a small ultimate degree of crystallization and the preservation of a high level of elastic properties.
7.5
Chloroprene rubbers
1,4-cis-Polychloroprene is also a homopolymer of regular structure capable of crystallization; its Tg = –40°C. Diversity of the types of chloroprene rubbers is due to the requirement of elastomers with different crystallization abilities, achieved both by developing different structures of homopolymer as the result of its regulation during the synthesis and by producing copolymers [95, 106a, 137, 291, 473, 742, 762]. The lower the polymerization temperature, the more regular the rubber structure is. The largest regularity was observed in a polymer obtained by solid-phase polymerization at –180°C from crystalline monomer, i.e., crystallized during polymerization. This is a fibre-like material with the melting temperature Tm = +100°C. Rubbers of the most regular structure with the greatest rate and ultimate
LOW-TEMPERATURE RESISTANCE SPECIFICS OF PARTICULAR ELASTOMERS
145
degree of crystallization (Cf |24%) are intended for manufacturing adhesives. The equilibrium melting temperature of common types of polymers is Tm0 = +80°C [95, 481] and their crystallization temperature range is from –30 up to +50°C. The maximal crystallization rate temperature T1 = –10°C, however, its rate changes little in the region from –10°C up to 0°C [95, 275, 702, 716]; sometimes, it is taken that T1 = 0°C [42, 43]. As for 1,4-cis-polyisoprene, decrystallization (“steaming”) of the rubber stored at room and lower temperatures is required for polychloroprene. Short-term low-temperature resistance of rubbers from polychloroprene is determined by glass transition even in the case of the most regular-structure (adhesive) rubbers. However, development of rubbers with long-term low-temperature resistance requires a combination of techniques, both decreasing Tg and slowing down crystallization. Therefore, blends of chloroprene rubber with butadiene-acrylonitrile rubbers of low acrylonitrile content (of the type of SKN-18) are widely used to produce oil- and low temperature-resistant rubbers. Addition of plasticizers and use of a correctly chosen vulcanizing group and a filler makes it possible in this case to produce rubbers resistant down to –(55–58)°C [348]. The use of plasticizers decreasing Tg in the case of polychloroprene-based rubbers results in acceleration of crystallization [43, 96, 174, 473, 716, 742]. Introduction of a plasticizer at the stage of synthesis was used to produce a rubber with low values of Tg (–55°C); however, in this case crystallization is accelerated so much that both long-term and short-term low-temperature resistance of the rubbers is deteriorated [631]. Detailed studies of the effect of plasticizers on the crystallization of different-type chloroprene rubbers have been conducted by Bayer AG [473, 742]. The possibilities of slowing down the crystallization of polychloroprene by choosing the vulcanizing group are not large [43, 95, 96, 347, 349, 762]; herewith, particles of MgO and ZnO in the vulcanizing groups traditional for polychloroprene can serve as nucleation centres, and crystallization is accelerated [347]. The effect of filling on the crystallization of chloroprene rubbers is of the same character as for other elastomers [106a, 110, 349, 540, 638]. The melting of polychloroprene reveals the most vividly the reorganization phenomena, studied in detail by N.V. Sochava et al. [666, 667, 719]. Therefore, it is rather difficult to trace the effect of vulcanization on the melting behaviour. During the introduction of a filler, the melting character, as for other rubbers, practically does not change [110]. Owing to the high values of Tm0, chloroprene rubbers are characterized by the capability of stress-induced crystallization at room and elevated temperatures, which is what determines their high cohesion strength and the tensile strength of rubbers on their basis [238]. Stress-induced crystallization of these elastomers has been studied sufficiently completely; the differences in the structure of oriented specimens due to different orientations have been shown [18, 95, 96, 785–787]. With reference to polychloroprene, the effect of crystallization on the properties of the amorphous component was also shown (by the method of pulsed NMR) [100, 213]. This is the only elastomer, for which large periods were found by the method of small-angle X-ray scattering [602, 603]. Apparently, this is due to a larger difference in the densities of the amorphous and crystalline phases as compared with other elastomers (see Table 3.1), related, in turn, to the trans-configuration of the polymer chain [481]. Owing to the capability of crystallization at room temperature, namely chloroprene elastomers served as objects in studies of the morphology of crystalline formations by the methods of light scattering [713, 714], light microscopy [95, 96, 785–787] and electron microscopy [17, 18] both in oriented state and in the absence of orientation, during the
146
CHAPTER 7
crystallization from solution and as a block, in particular, also in the presence of a filler. Exactly with reference to polychloroprene, several works have shown the effect of crystallization on the processing of rubber compounds the most completely [43, 275, 699, 700].
7.6
Butyl rubber
Butyl rubber (BR) is a copolymer of isobutylene with a minor amount (1–4.5 mol.%) isoprene. It has the linear structure of macromolecular chains. At a comparatively low Tg of –(68–61)°C, this elastomer, as polyisobutylene, is characterized by a broad temperature range of transition from the rubberlike to the glassy state, 'T and the low values of T, as well as the broad region of mechanical loss, which creates prerequisites for using butyl rubber in formulations of rubbers with good dissipative properties. The steric features of the structure of the polyisobutylene chain determine the very small rates of isotropic crystallization of butyl rubber. This rubber was considered for a long time to be incapable of crystallization at low temperatures. It was established by the dilatometric method [442] that at –33°C polyisobutylene crystallizes in 20 days, and butyl rubber with 1 mol.% of isoprene units, in 1.5–2 years. According to other data [95, 96], 30 days is required to complete the crystallization of technical rubber based on butyl rubber (apparently, of a lower unsaturation) in the temperature range from –25 up to –40°C, and the induction period W0 = 10 days. As isotropic crystallization of these polymers takes a long time, its basic regularities have been studied insufficiently. According to the data by Wunderlich [751], for pulyisobutylene Tm0 = +44°C. The maximum crystallization rate temperature T1 = –35°C [95, 96, 442]. Under compression conditions (H | 0.5), crystallization of compounds based on butyl rubbers of small unsaturation, i.e., low-density network, at T = –40°C completes in approximately 20 days. The effect of the density of the three-dimensional network on crystallization is determined by its unsaturation. The larger it is, the higher the density of the network is, and the smaller the capability of crystallization is. As for polyisoprene and polybutadiene, the peroxide vulcanization of butyl rubber slows down its crystallization to a smaller degree than the sulphur vulcanization. The effect of filling on crystallization of rubbers based on butyl rubber have been studied insufficiently. Thus, it has been shown that, in contrast with other elastomers, the value of W1/2 determined by extrapolation of the characteristic curves from eq. (3.17) decreases with the content of active carbon black, apparently, due to the reduction of the induction period W0, which for butyl rubber makes the major part of the value of W1/2 [95, 96]. This feature is, apparently, related to the particularities of the effect of stress on the crystallization of polyisobutylene and butyl rubber. Introduction of a plasticizer (transformer fluid) leads to the temperature shift of the maximum crystallization rate up to –50°C and a decrease of the crystallization rate, especially at –40°C and higher. At –50°C, the differences in the crystallization rates of filled rubbers with or without a plasticizer level off [442]. All these data show that the lowtemperature resistance of rubbers from BR is determined by glass transition, and only at very large times the effect of crystallization shows. As we have already noted in Chapter 4, a distinctive feature of polyisobutylene and butyl rubber is a combination of a small rate of isotropic low-temperature crystallization with a high tendency to stress-induced crystallization. The equation for the characteristic curve (eq. 3.17), which describes the effect of stress on crystallization is valid both in compression [95, 96] and in stretching (up to H = 4) [442]. However, the stretching accelerates crystallization of butyl rubber-based elastomers much stronger [764].
LOW-TEMPERATURE RESISTANCE SPECIFICS OF PARTICULAR ELASTOMERS
147
Thus, stretching of the sulphur vulcanizate up toH = 6 decreases W1/2 by about 105 times; at H = 3 and T = –40°C, W1/2 |70–80 min, and the induction period W0 < 5 min. During the stretching, not only the rate increases, but also the maximum achievable degree of crystallization C. Thus, at H = 2, C |50%. Introduction of fillers (up to 50 weight fractions of carbon black) reduces the stress-induced crystallization time by 1.5–2 orders of magnitude [442]. The change of the melting temperature TmH, measured by the contraction of a stretched, crystallized and then released specimen revealed a greater dependence of this value on H than for other elastomers. At T = –40°C and H = 3, TmH = –16°C, and at H = 5, TmH = +35°C [442]. Stress-induced crystallization determines good strength properties of butyl rubberbased elastomers; however, its features and, first and foremost, the effect of unsaturation as well as its temperature range, have been studied insufficiently. Not abundantly clear, either, are the causes of an anomalously high effect of stress on crystallization. Low-temperature resistance of chloro- and bromobutyl rubbers has been practically not investigated.
7.7
Silicon rubbers
The polymer chains of polyorganosiloxanes are shaped as helices with outer orientations of organic groups. This shape is due to a large freedom of rotation around bonds Si–O and Si–C and the lability of the Si–O–Si angles, which endows the molecules with a high segmental mobility. A consequence of this are rather low glass-transition temperatures. Thus, for polydimethylsiloxane (PDMS), Tg = –126°C. A high segmental mobility, which impedes the dense packing, leads to a weakening of intermolecular interaction. Therefore, in PDMS the cohesion-energy density (226 MJ/m3) and the viscous-flow activation energy (15–16 kJ/mol) are much lower than in other elastomers [221]. Introduction of substituents into the principal chain leads to a change of Tg (see Fig. 1.14b). Thus, the presence of phenyl substituents results in an increase of Tg, the effect from each diphenyl unit being the same as from two phenyl units. Introduction of fluorine-containing substituents also results in an increase of Tg (see Fig. 1.14b). Copolymers containing ethyl units have a lower Tg, and in the homopolymer containing 100% ethyl units Tg = –141°C [95, 96, 100a, 123, 148, 256, 319, 421, 465, 498, 499, 519, 521, 523, 557 594a]. That is, Tg of silicon rubbers is determined by the cohesion energies of the substituents (see Chapter 1). A larger chain flexibility and structure regularity of silicon rubbers provide for their high capability of crystallization [95, 96, 100a, 123, 148, 404–407, 428, 429, 536]. The maximum crystallization rate temperature of PDMS is T1 = –80°C, but a high rate of crystallization in this temperature range hampers the exact determination of T1, so the available literature data are not unequivocal. For practical purposes, it is recommended to take T1 = –(80r5)°C [95]. Within these limits, the value of T1 varies depending on the type and amount of substituents [104]. In contrast with earlier data [95, 404–407], recent research has shown that the equilibrium melting temperature of PDMS Tm0 > –35°C, and at crystallization temperatures T > –35°C the melting temperature Tm starts to rise as for other elastomers [104]. PDMS has the highest ultimate degree of crystallization among rubbers, C = 60% [246, 404, 406, 492–494]. As we have already mentioned in Chapter 3, with the content of substituents w rising
148
CHAPTER 7
the values of W1/2 for polysiloxanes pass through a minimum, the values of w being the same (4%) for different substituents [428, 429, 492–494, 519]. A further increase of w slows down the crystallization, the slowdown being different for different substituents. When choosing a copolymer additive, it is important not only to suppress the capability of crystallization, but also to preserve the low value of glass-transition temperature. This requires introduction of units, which have a low cohesion energy and, therefore, do not decrease the flexibility of the principal chain. These requirements are satisfied the most by diethylsiloxane units. However, the homopolymer containing 100% ethyl units crystallizes, as we mentioned in Chapter 3; herewith, the crystallization and melting temperature range is much greater than that for the copolymer with low ethyl-unit content [64]. Besides usual crystallization, ethylsiloxane and propylsiloxane homopolymers reveal the capability of forming a liquid-crystalline structure [251, 252, 254, 255]. If the crystallization proceeds from the liquid-crystalline phase, its rate is much higher than that from the rubberlike melt, and the ultimate degree of crystallinity C = 90%. In Chapter 4, mention was made of the liquid-crystalline phase as a possible precursor of stress-induced crystallization, the formation of which during the orientation of elastomers provides for an anomalously high rate of stress-induced crystallization and crystallization under high pressure [253, 680, 698]. Fluorosilicon copolymer SKTFT-50 does not crystallize, its low-temperature resistance is completely determined by glass transition (see Table 7.1). Homopolymer SKTFT-100, whose Tg is close to T1 for PDMS, is characterized by anomalous low-temperature properties, close by its nature to microcrystallization (see Chapter 3). For all crystallizable copolymers based on PDMS, with the small content of comonomer, a characteristic feature is the formation of the crystal lattice of the same type as for PDMS homopolymer. Efficient inhibition of crystallization requires the statistical distribution of modifying units along the chain. A disturbance of the statistical distribution of units leads to a sharp acceleration of the crystallization process. Thus, a block specimen of phenyl vinyl copolymer of 8% phenyl units (SKTFV-803), with the blocks containing up to 30% of homogeneous units, crystallizes at –78°C for 30–40 min, whereas crystallization of an elastomer of the statistical structure of the same composition starts in 1000–1200 min [521, 523]. A high sensitivity of the crystallization parameters to the character of unit distribution in polysiloxanes was used to assess the distribution of vinyl groups depending on the method of their introduction, as well as for studies of interchain exchange of polyorganosiloxanes [521]. Comparison of the data of crystallization half-time* W1/2 with the results of the chromatographic measurements made it possible to develop a method for assessing the block structure of rubbers of the type of SKFV-803. The method is based on the use of pyrolytic gas chromatography [394]. As a results of the high crystallization rate of most polysiloxanes is that even their short-term low-temperature resistance is determined not by glass transition but by crystallization [95, 96, 104, 405, 494, 523]. This should be taken into account when assessing the low-temperature resistance of these rubbers by standard methods, which do not consider the effect of the testing time and stress. However, at a certain content of each type of substituent, crystallization of copolymer can be suppressed, so in this case the low-temperature resistance is determined by the glass-transition process, too. A distinctive feature of these rubbers is that their capability of crystallization increases as the density of the three-dimensional network rises up to a certain limit, which depends on the type and content of filler
* Data obtained by N.M. Zorina.
LOW-TEMPERATURE RESISTANCE SPECIFICS OF PARTICULAR ELASTOMERS
149
and anti-texturing agent [95, 96, 123, 519, 626]. Standard vulcanizates of polysiloxanes also crystallize faster than raw rubbers. The further increase of the density of network (as well as the number of substituents) leads to a slowdown of crystallization. But the limited possibilities of choosing the vulcanizing group stipulate the low efficiency of vulcanization for slowing down the crystallization. Even at maximum admissible contents of the vulcanizing agent, crystallization of rubbers is slowed down not as much as to make their low-temperature resistance be determined by glass transition. Introduction of fillers (e.g., various types of Aerosil) accelerates crystallization at the initial stages; the efficiency of their action is higher for vulcanizates than for raw blends and changes in the presence of modifiers [414, 415, 523, 560, 626, 759]. Unfortunately, no data on the effect of such a promising filler as precipitated silica are available in the literature. In the mean time, such data are of interest not only from the point of view of low-temperature resistance, but can help assess the quality of a filler. Introduction of zinc oxide, redoxide (Fe2O3) and other additives into rubber compounds leads to an insignificant acceleration of crystallization of raw compounds as the result of the nucleation rate increase and causes practically no change of the process rate in vulcanizates. Attempts to choose plasticizers, which would have significantly affected glass transition or crystallization of siloxane elastomers gave no positive results.
7.8
Urethane rubbers
Variation of the properties of urethane elastomers is achieved by changing the ratio of three components: high-molecular-mass polyols, diisocyanates and chain extenders [21, 248, 289, 514, 652, 630, 705]. Two kinds of segments – flexible and rigid – can be said to alternate in the molecular chain of polyurethanes. Changes in content and nature of flexible and rigid segments within broad limits result in the polyurethane glass-transition temperature changing from –60°C up to –20°C. Depending on the chemical structure, molecular mass M, degree of microphase separation and other factors, crystallization can be observed both in flexible and rigid blocks. Crystallization of rigid segments based on hexamethylene diisocyanate or diphenylmethane diisocyanate promote microphase separation. Crystallization of flexible segments deteriorates elasticity, increases the modulus of polyurethanes and their rigidity. As crystallization occurs at low temperatures, it determines the low-temperature resistance of polyurethanes. Depending on the type, content and molecular mass of a flexible segment, the maximum crystallization rate temperature T1 can change from –10°C up to +5°C. Data on the equilibrium melting temperature Tm0 are not available. Flexible segments based on adipic-acid polyesters and polymethylene series glycols crystallize as they have a linear structure. An increase of M of a polymethylene fragment enhances the tendency to crystallization [669]. The tendency decreases as the result of both the interaction of polyethylene adipinate with diisocyanate, and the decrease of the polyester’s molecular mass; the decrease of the crystallization ability can also be achieved by the correct choice of diisocyanate. Crystallization can be suppressed by modifying polydiol or disturbing the regularity of its structure [669]. Polyurethanes containing polytetrahydrofurane-based flexible segments have a good short-term low-temperature resistance and hydrolytic stability. It has been shown [21] that crystallization of polyurethanes based on copolymers of tetrahydrofurane and ethylene oxide (at –30°C) occurs only at oligomer’s Mn greater than 1700. With a greater number of rigid segments, i.e., with the rise of the number of efficient junctions of the physical and chemical network, the ability of flexible segments to crystal-
150
CHAPTER 7
lize decreases and can be totally suppressed at a definite ratio of the components. Usually the changes in the composition of polyurethanes leading to a decrease of Tg contribute to the manifestation and acceleration of crystallization, deteriorating long-term lowtemperature resistance. However, by decreasing the regularity of polyesters forming a flexible block, but not changing its molecular mass, the tendency of polyurethanes to crystallize can be significantly decreased at practically the same Tg [21, 248, 515, 630, 647]. Thus, long-term low-temperature resistance of polyurethanes depends first and foremost on the tendency of the polyester component to crystallize. Crystallization of polyurethanes at room and not too low temperatures makes possible polycrystal morphology studies by light and electron microscopes [95, 96, 785, 786].
7.9
Thiocol rubbers
Liquid thiocols based on di(E-chloroethyl)formal [675], with Tg within the limits of –40°C up to –60°C [95], have become the most widespread at present. Introduction of up to 100 weight fractions of fillers of various activities does not lead to the change of Tg and to a jump of specific heat in glass transition, 'cp; crosslinking does not practically change Tg, either, but 'cp decreases [196]. Elastomeric materials based on thiocols are used mainly as sealing agents (hermetics). The regularity of structure of this polymer gives grounds to consider it to be capable of crystallization. However, the fact of crystallization of thiocol could not be observed for a long time due to the extremely slow development of the process [95, 675]. At temperatures from –15°C up to –25°C, the dilatometric method failed to establish the development of crystallization even within 1.5 months. The absence of crystallization of thiocols studied at –25°C for two months was also noted in [196]. As in the case of butyl rubber, crystallization of thiocols can be studied the most conveniently during the deformation of vulcanizates. The maximum crystallization rate temperature T1 = –25°C [95, 96]. Films obtained from a solution of the same thiocol in benzene were found by the method of light microscopy to have grains 5–6 Pm in size. The grains were formed after three days of holding at room temperature; the use of copolymers made it possible to reduce the rate and degree of crystallization. For thiocol of a different composition, it was found by the method of residual elongations [456] that at H = 2 and holding time of 30 min vulcanizates filled with 30 weight fractions of carbon black crystallize within the temperature range of –45°C to –10°C. The maximum crystallization rate temperature T1 = –28°C. At this temperature, the major part of the process is complete within the first 30 min; however, the equilibrium degree of crystallinity settles in only after three or four days. For an unfilled vulcanizate at –25°C and H = 2, the crystallization half-time W1/2 |40 days [456]. To suppress the crystallization of thiocol vulcanizates, monomers of another chemical nature are used, which can result both in the decrease of Tg (down to –60°C and lower) and in its increase up to –40°C. However, this substitution fails to inhibit completely the ability of the polymer to crystallize under deformation conditions. The most reliable method is copolymerization of different monomers. For vulcanizates of copolymers, the calculated value of W1/2 at –25°C in a nonoriented state increases from 2 up to 5–6 years [456]. As for other copolymers, the efficiency of crystallization suppression depends on the distribution of monomers along the chain. Due to the low crystallization rate, the melting parameters of thiocols have not been studied; stress-induced crystallization proper has been little investigated, too. The lowtemperature resistance of thiocols is mainly determined by the glass-transition process.
LOW-TEMPERATURE RESISTANCE SPECIFICS OF PARTICULAR ELASTOMERS
151
7.10 Fluoroelastomers Fluoroelastomers, which are copolymers of vinylydene fluoride with hexafluoropropylene at a ratio of 7:3 (e.g., grade SKF-26) or with trifluorochloroethylene at a ratio of 1:1 up to 3:7 (e.g., SKF-32), have the glass-transition temperatures Tg = –20° and Tg = –18°C, respectively [430], i.e., are not low-temperature resistant [212, 648, 528–531a]. These are some of the most rigid-chain rubbers. They are amorphous, and their low-temperature resistance is determined by the glass-transition temperature. However, a specific feature of fluoroelastomers is the greatest difference between Tg and the brittleness temperature Tbr [341, 378a], which is due to the presence of a broad region of forced elasticity*. Thus, for SKF-26 Tbr = –60°C, and for SKF-32 Tbr = –62°C. These rubbers are also characterized by the largest value of specific free glass-transition volume f Gg = 0.13 [101, 544], as well as the lowest jump of specific heat cp = 0.14–0.19 kJ/g and the strongest effect of the shape factor ) of the specimen on the parameters characterizing the low-temperature resistance of elastomers in the transition region [129, 131] (see Chapter 2). The capabilities of the formulation effect on the decrease of Tg and improvement of the low-temperature resistance of fluoroelastomers (on condition that the high thermal resistance is preserved) are rather limited [485, 530, 531, 531a, 563]. Therefore, the major way to improve the low-temperature resistance is to develop copolymers with lower Tg, which is provided for, first and foremost, by introduction of oxygen-containing groups into the polymer chain [4, 341, 482, 529, 531a, 556a]. The presence of a joint oxygen atom in the side group significantly facilitates rotation of perfluoroalkyl radical, which results in a decrease of Tg. Thus, for copolymers of vinylydene fluoride and perfluoromethyl vinyl ester (of the type of SKF-260) Tg = –40°C and Tbr = –62°C. However, the improvement of the parameters of short-term low-temperature resistance is accompanied by a deterioration of the properties of oxygen-containing fluoroelastomers during their prolonged holding at low temperatures in the process of microcrystallization, which mainly develops in the rubber’s gel fraction [129, 131], see Chapter 3. Studies of the Tg dependence of fluoroelastomers on their composition showed a possibility of developing various copolymers with rather low Tg [22, 184, 488]. Lowtemperature resistant fluoroelastomers have been developed and are commercially manufactured [22, 34, 184, 423, 531a, 590, 594, 594a, 621, 725]. However, the long-term low-temperature resistance of these polymers has not been studied. At the same time, they are all polymers, containing units of at least one rapidly crystallized homopolymer, e.g., such as vinylydene fluoride. The low Tg values of these copolymers are provided for by the high flexibility of their molecular chains, which in principle creates the possibilities for microcrystallization even at small deviations of the distribution of such units from the statistical value. The high strength of some of them during the stretching, probably related to their stress-induced crystallization, is in favour of this suggestion [298]. Apparently, the development of microcrystallization of a heat-resistance fluoropolymer was the cause of the loss of tightness of the seal, which had led to the Space Shuttle disaster [487].
7.11 Acrylate rubbers Acrylate rubbers are copolymers of acrylic-acid esters and various polar vinyl monomers containing the functional group capable of further vulcanization. As a rule, these are linear * The data by G.P. Petrova.
152
CHAPTER 7
saturated amorphous polymers, the physical properties and low-temperature resistance of which are mainly determined by the structure of the acrylate block [221]. With the length of the side alkyl chain extended their low-temperature resistance is improved; however, the oil resistance drops down in this case. For commercially available acrylates and rubbers based on them the glass-transition temperature Tg = –(31–35)°C; the brittleness temperature Tbr, can, in contrast with fluoroelastomers, be much higher than Tg. Thus, for acrylate rubbers with Tg = –35°C, Tbr = –29°C; at Tg = –31°C, Tbr = –22°C. Acrylate rubbers with Tbr = –(11–14)°C are produced, too [669]. Systematic studies of polyacrylates’ low-temperature resistance depending on the structure of the polymer chain have shown*, according to the data of dynamic mechanical spectroscopy, that an increased length of polyacrylates’ alkyl groups leads to a systematic decrease of theD-transition temperature** TD. Thus, for polyethyl acrylate, TD = +3°C; for polybutyl acrylate, TD = –25°C; and for polyoctyl acrylate, TD= –33°C. The width of the D-transition temperature range increases in this progression, too. A sharp decrease of the glass-transition temperature of polyacrylates is achieved by addition of units of crystalline homopolymer, e.g., hexadecyl acrylate at polymerization [305a]. In practice, acrylate rubbers are often produced using ethoxyethyl acrylate as a comonomer in the synthesis of oil- and petrol-resistant polymers with increased lowtemperature resistance. Namely these rubbers give an optimal combination of lowtemperature and oil resistance. The minimum temperature for the operation of commercial acrylate rubbers is of the order of –40°C; the maximum temperature, from +150°C (long-time operation) up to 200°C (short-time operation) [221]. Low-temperature-resistant types of acrylate rubbers with Tg from –50 to –55°C have also been developed. A decrease of Tg from –35 to –60°C can be achieved by increasing the number of carbon atoms in alkyl acrylate from 2 up to 7. However, an improvement of low-temperature resistance is accompanied with a considerable deterioration of heat resistance; the strength properties at room temperature also become worse. An improvement of low-temperature resistance is also achieved using ethylene acrylate rubbers [184, 221, 273]. On the whole, acrylate rubbers are a promising class of heat- and oil-resistant elastomers. By their heat resistance and resistance to aggressive media, they occupy an intermediate position between butadiene acrylonitrile and fluoroelastomers. However, the low-temperature behaviour of this class of rubbers have not been practically studied. At the same time, as fluoroelastomers, acrylate rubbers should be expected to have a peculiar behaviour at low temperatures.
7.12 Ethylene-propylene rubbers Ethylene-propylene rubbers and elastomeric materials based on them find ever increasing applications owing to their low unsaturation and, as a consequence, resistance to many kinds of ageing. Commercially produced types are double copolymers of ethylene and propylene (EPM) and triple copolymers, with diene-hydrocarbon additives (usually ethylidene norbornene or dicyclopentadiene) to provide for sulphur vulcanization (EPDM). Elastomers are copolymers of average compositions, i.e., with the ratio of monomers close to * Data by E.A. Sidorovich. ** Remember that TD corresponds to the midpoint of the transition region, and its value is always higher than the glass-transition temperature measured under static conditions (see Chapter 1).
LOW-TEMPERATURE RESISTANCE SPECIFICS OF PARTICULAR ELASTOMERS
153
1:1 [272, 337, 477]. Recall that the values of Tg characteristic of them are the lowest (see Fig. 6.1). The main features of low-temperature behaviour of ethylene-propylene rubbers were discussed in Chapter 3, as these copolymers were used as an example to consider the process of microcrystallization. We would note here that the X-ray and DSC methods have shown that the crystalline structure formed in copolymers with the low propylene content w is very defective, probably, due to the possibility of methyl side groups of propylene entering the polyethylene lattice [532, 595]. The maximum crystallization rate temperature depends on composition: it is the lower, the larger w is. As methylene units are involved in the microcrystallization of EPM and EPDM (as well as rubbers based on them), its degree is the greater, the higher their content is, i.e., the smaller w is. But the melting temperature, i.e., size of microcrystals (their length l) does not depend on it. Calculations based on the modified Thomson equation (eq. 3.21) gives the value of l equal to 10–20 methylene units depending on the crystallization temperature and the values of polyethylene equilibrium melting temperature Tm0 used for the calculations [132, 133]. It is essential that microcrystals continue to be formed up to room and higher temperatures; herewith, their melting temperature Tm | 50–60°C (at the crystallization temperature T = 20°C) and even higher at higher crystallization temperatures. Therefore, the role of microcrystals in the processing of ethylene-propylene elastomers is very great [63], though insufficiently studied. The presence of a third, diene, monomer does not change the character of microcrystallization; however, due to different conditions of synthesis, copolymers with different numbers of long methyl sequences can be obtained. For the same conditions of synthesis, it appears that the number of long sequences of the methyl units is the larger, the greater their content is [532, 595]. The main method of providing for the required low-temperature resistance of ethylene-propylene rubbers is to choose a monomer ratio close to 1:1, so that Tg be minimal, and the microcrystallization process be suppressed. The possibilities of the effect of vulcanization and filling on the microcrystallization process and, therefore, on the low-temperature resistance of rubbers based on ethylene-propylenes are rather limited [105, 132, 133, 300, 427, 779]. Stress-induced crystallization of ethylene-propylene rubbers depending on the composition of copolymer, as well as on the presence of a plasticizer are described in [156, 166, 449, 475, 619, 620, 765]. Stress-induced crystallization and, therefore, cohesion strength of rubber compounds and the strength of vulcanizates are the higher, the larger the content of methyl units is. That is, the choice of the optimal composition of a polymer depends on the operation conditions.
7.13 Epoxide rubbers Epoxide rubbers are produced by scission of stressed oxygen-containing cycles under the action of catalysts. Epoxide rubbers include propylene oxide elastomer – a copolymer of propylene oxide and allyl glycidyl ether – and epichlorohydrine elastomers – a homopolymer of epichlorohydrine and a copolymer of epichlorohydrine with ethylene oxide or propylene oxide [331, 331a, 507, 649, 669]. A specific feature of the chemical structure of epoxide rubbers is the presence of ether groups in the polymer chain. The low-temperature resistance of different types of epoxide rubbers is different. Thus,
154
CHAPTER 7
for propylene rubber Tg = –(72–74)°C. For epichlorohydrine homopolymer, it is –28°C; for copolymers, –45°C. Propylene oxide homopolymer has Tg = –55°C. It crystallizes at room temperature; the degree of crystallization measured by the X-ray method is 70%; the melting temperature Tm = 80°C. Propylene oxide rubber is one of the most low-temperature-resistant rubbers. Depending on the polymerization conditions, both stereoregular propylene oxide polymer capable of crystallization can be obtained, and a polymer of nonregular structure whose crystallization is not observed. By changing the composition of the catalytic complex, in particular, by increasing the acetyl acetone content in it, and also by introducing an ether into the catalyst, the regularity of the polymer can be changed and, thus, its capability of crystallization be affected. The presence of even 2 mol.% allyl glycidyl ether units leads to a decrease of the degree of crystallization. An increase of the polymerization temperature and of the concentration of monomer enhances the secondary processes and also contributes to the chainregularity disturbance. The small number of double bonds in the principal chain of the polymer determines the high ozone and heat resistance (up to +150°C) of rubbers based on propylene oxide polymer; its swelling in petrol does not exceed 30%. Propylene oxide rubber is one of the most promising low-temperature-resistant rubbers [563]. Elastomeric materials based on it are characterized by the high values of T ; the elastic recovery coefficient at –50°C, K = 0.63 [563]. It possesses a high capability of stress-induced crystallization at room and higher temperatures; therefore, unfilled rubbers based on it are characterized by a high tensile strength and flexing life, low heat formation, high cut-growth resistance [273a]. Depending on the composition and synthesis conditions, however, this rubber, as ethylene-propylene, could be expected to be subject to microcrystallization, i.e., long-term low-temperature resistance should be controlled. Epichlorohydrine rubbers, with their poor low-temperature resistance, possess good oil resistance. Therefore, to develop rubbers efficient in liquid media they are recommended to be used in blends with other rubbers that possess good low-temperature resistance [331].
7.14 Polypentenamers Polypentenamers are produced by ring-scission polymerization of cyclopentene [517, 520, 669]. The double bonds in the polypentenamer molecules can be in cis- or trans-configuration; respectively, the glass transition temperatures are Tg = –114°C and Tg = –97°C. Owing to the regularity of structure, they are capable of crystallization*. The equilibrium melting temperatures obtained by extrapolation are, respectively, Tm0 = +38°C and Tm0 = +34°C. The crystallization of trans-polypentenamer, which is an amorphous rubber at room temperature but rapidly crystallizes upon cooling, has been studied to date in detail [392, 517, 520]. This polymer is characterized by a high crystallization rate, as well as a strong dependence of the crystallization rate on temperature. According to X-ray data, the ultimate degree of crystallization Cf |35% and does not depend on T [392]. Using an empirical equation (3.10), which relates Tg, Tm0 and the maximum crystallization rate temperature T1, the value T1 = –50°C was obtained. An increase of the content of trans-units from 80 up to 84% leads to a acceleration of crystallization by an order of magnitude. At the same content * A higher Tg of the trans-polymer can be due to a higher rate and degree of crystallization, and to its effect on glass transition (see Chapter 6).
LOW-TEMPERATURE RESISTANCE SPECIFICS OF PARTICULAR ELASTOMERS
155
of trans-units, a decrease of the characteristic viscosity [K] from 3.4 down to 1.26 leads to an increase of the crystallization rate of the trans-polypentenamer, in contrast to other rubbers of regular structure. This, apparently, implies that an increase of [K], i.e., a rise of M, is accompanied by an increase of branching, as it takes place for polychloroprene. Formation of the three-dimensional vulcanization network decreases the ultimate degree of crystallization, probably, owing to the suppression of the crystal growth rate. The effect of filling on the crystallization of trans-polypentenamer is on the whole the same as for other elastomers.* Owing to a high tendency of the trans-polypentenamer to stress-induced crystallization, of considerable interest are blends of this rubber with other polymers, e.g., with polyisoprene [517, 520]. Its addition to SKI-3 increases the cohesion strength of raw compounds, enhances the elasticity of rubbers, their resistance to abrasion and thermal ageing. However, the tear resistance and low-temperature resistance (determined by low-temperature crystallization) are slightly decreased [517, 520].
7.15 Alternating rubbers Alternating rubbers, which represent alternating copolymers of dienes and olefins, are a new type of general-purpose rubbers [73, 669]. Copolymers of propylene with trans-butadiene and butadiene with nitrile of acrylic acid have been studied the most. Depending on the type of catalytic system [669] and temperature of the process, block structures of butadiene (up to 3%) can be formed along with alternating copolymers. A feature of the structure of alternating copolymers of dienes and olefins is the presence of six carbon atoms in the elementary unit. In copolymers of propylene and trans-butadiene, the butadiene part usually contains 80–90% of 1,4-trans-units. If, in this case, the content of 1,2-units is within 2–5%, then Tg = –(75–80)°C, i.e., the copolymer belongs to elastomers with good short-term low-temperature resistance. Despite the alternation of units, the capability of crystallization is preserved in the copolymer (though it is much smaller than for respective homopolymers). In this case, the elementary cell comprises, apparently, the whole elementary unit of the copolymer. The maximum crystallization rate temperature T1 |–(50–55)°C, and the crystallization kinetics is determined by the type of catalytic system. In the case of titanium catalysts, at T1 = –50°C W1/2 = 1500–3000 min; if the catalysts are vanadium ones, providing for a greater stereoregularity, W1/2 = 200–300 min. Within the limits of one catalytic system, the capability of crystallization of the copolymer depends on the content of trans-units: when their number decreases from 89 to 69%, W1/2 increases two times, and the degree of crystallization as measured by the dilatometric method decreases 1.5-fold [461]. The degree of crystallization measured by the X-ray method decreases in this case from 20 to 5%. As the copolymer’s unit cell parameters are determined insufficiently reliably, it is difficult to assess the absolute values of the degree of crystallization. Melting of copolymers is observed starting from –35°C. For specimens crystallized in the oriented state, Tm measured by the contraction method little depends on the degree of stretching of the specimens; however, the crystallization rate sharply increases. Herewith, the maximum crystallization rate at H = 5 and T = –58°C is 20–22% and is achieved in 20 min. * Data by N.L. Severina.
156
CHAPTER 7
On the whole, the short-term low-temperature resistance of rubber and elastomeric materials on its basis is determined by glass transition; the long-term low-temperature resistance, by crystallization. The tendency to stress-induced crystallization provides for a high level of their strength properties. Alternating butadiene-acrylonitrile rubbers, in contrast with butadiene-acrylonitrile rubbers of emulsion polymerization are characterized by a more uniform structure, the absence of microblocks from nitriles of acrylic acid. At the same content of nitriles of acrylic acid, their glass-transition temperature is lower. The regularity of their structure stipulates the development of stress-induced crystallization at room temperature at H k | 7; the identity period d = 7.145 [211, 451]. As for polypentenamers, the short-term low-temperature resistance of alternating butadiene-propylene rubber and elastomers based on it is determined by glass transition; the long-term parameter, by crystallization [461]. Methods of regulating the low-temperature resistance (both glass transition and crystallization) of elastomeric materials from these rubbers by changing the formulation have not been practically studied.
References 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25.
Abramova, N.L., Zlobina, M.B. and Akopyan, L.A. (2006) Kauchuk i Rezina, No 6: 5 (in Russian). Aggarwal, S.L., Hardis, J.G., Livigni, R.A., Fabris, H.G. and Marker, J.E. (1986) In Lae, L. and Mark, J.E. (eds.) Advances in Elastomers and Rubber Elasticity, London–New York: Plenum Press, p. 17. Ainbinder, S.B., Alksne, K.I., Tyunina, E.L. and Laka, M.G. (1973) Properties of Polymers at High Pressures. Moscow: Khimiya Publishers, 214 pp. (in Russian). Ajroldi, G., Bonardelli, P., Fumagalli, M. and Moggi, G. (1986) Proc. Int. Rubber Conf. Göteborg, vol. 2, p. 467. Akkerman, S. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, C6. Alegra, G. (1986) In Lae, L. and Mark, Y.E. (eds.) Advances in Elastomers and Rubber Elasticity. London–New York: Plenum Press, p. 303. Aleksandrov, A.P. (1945) Proc. 1st and 2nd Conf. on High-molecular-weight Compounds. Moscow–Leningrad: USSR Acad. Sci. Publishers, p. 49 (in Russian). Aleksandrov, A.P. and Lazurkin, Yu.S. (1944) Dokl. Akad. Nauk SSSR, 45: 308 (in Russian). Aliguliyev, R.M., Khiteyeva, D.M., Mamedov, A.A. and Oganyan, V.A. (1983) Dokl. Akad. Nauk Az. SSR, 39(2): 87 (in Russian). Aliguliyev, R.M., Ovanesova, G.S., Khiteyeva, D.M. and Oganesyan, V.A. (1985) Vysokomol. Soed., ser. A, 28(7): 1452 (in Russian). Aloev, V.Z., Kozlov, G.V., and Zaikov, G.E. (2004) Kauchuk i Rezina, No 3: 38 (in Russian). Andrews, E.H. (1962) Proc. Roy Soc., A270(1341): 232; (1964) A277(1371): 562. Andrews, E.H. (1979) In Andrews, E.H. (ed.) Development in Polymer Fracture. London: Applied Science Publications, vol. 1, p. 1. Andrews, E.H. and Gent, A.N. (1963) In Bateman, L. (ed.) The Chemistry and Physics of Rubber-like Substances. London: Maclaren, p. 225. Andrews, E.H. and Reed, P.E. (1973) In Kaush, H.H., Hassel, J.A. and Jaffer, B.Y. (eds.) Deformation and Fracture of High Polymers. New York: Plenum Press, p. 17. Andrews, E.H. and Reed, P.E. (1979) In Andrews, E.H. (ed.) Development in Polymer Fracture. London: Applied Science Publications, vol. 1, p. 17. Andrews, E.N. and Reeve, B. (1971) J. Mater. Sci., 6: 547. Andrews, E., Owen, P.J. and Reed, P.E. (1971) Proc. Int. Rubber Conf. 1969. Moscow: Khimiya Publishers, p. 95 (in Russian). Andrews, E.H., Owen, P.J. and Singh, A. (1971) Proc. Roy. Soc., A324(1556): 79. Ansarifar, A., Ibrahim, N., Bennett, M. (2005) Rubb. Chem. Techn., 78(5): 93. Apukhtina, N.P., Marei, A.I., Novikova, G.E. and Miller, B.E. (1965) Vysokomol. Soed., 7(6): 1117; (1974) ser. B, 16(1): 3 (in Russian). Arcella, V. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, F4. Arctic Rubber (1989) Prepr. Scandinavian Rubber Conference. Tampere, Finland. Arenz, R.J. and Landel, R.F. J. (1982) J. Polym. Sci. Polym. Phys. Ed., 20(3): 363. Arlman, J.J. (1950) Rubb. Chem. Technol., 23(2): 306.
158
26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65.
REFERENCES
Arlman, J.J. and Goppel, J.M. (1950) Rubb. Chem. Techn., 23(2): 319. Askadsky, A.A. (1973) Deformation of Polymers. Moscow: Khimiya Publishers, 448 pp. (in Russian). Askadsky, A.A. (1995) Vysokomol. Soed., ser. A, 37(2): 332 (in Russian). Askadsky, A.A. and Matveyev, Yu.I. (1983) Chemical Structure and Physical Properties of Polymers, 248 pp. (in Russian). Avitable, G., Napolitano, R. and Riva, F. (1980) J. Polym. Sci. Polym. Phys. Ed., 7: 1663. Avrami, M.A. (1939) J. Chem. Phys., 7: 1103; (1940) 8: 212; (1941) 9: 177. Avstriyskaya, E.E., Berezkina, A.P. and Nelson, K.V. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftehkim, p. 83 (in Russian). Babitsky, E.L. (1984) Prepr. Int. Rubber Conf. Moscow, B 56. Bach, D., Van Gool, G., Steffens, J. and Kataev, A.D. (2004) Abstr. Int. Rubber Conf. Moscow, p. 274. Bagdasaryan, E.I., Arutyunyan, L.S., Zakharov, N.D. and Kabalyan, Yu.K. (1978) Kauchuk i Rezina, No 4: 9 (in Russian). Bahani, M., Laupretre, F. and Monnerie, L. (1995) J. Polym. Sci. Polym. Phys. Ed., 33(2): 167. Baldwin, P.P. and Ver Strate, G. (1972) Rubb. Chem. Technol., 45(3): 709. Baranov, N.S. and Elkin, A.I. (1974) Kauchuk i Rezina, No 5: 33; (1978) Kauchuk i Rezina, No 6: 37 (in Russian). Baranov, N.S., Sokolov, V.E. and Elkin, A.I. (1982) Kauchuk i Rezina, No 8: 25 (in Russian). Baranov, V.G. (1977) Khim. Volokna, No 3: 11 (in Russian). Baranov, V.G., Khrenov, V.V. and Frenkel’, S.Ya. (1975) Fiz. Tverd. Tela, 17: 1550 (in Russian). Baranov, V.G., Ovanesov, G.T., Gasparyan, K.A. et al. (1974) Dokl. Akad. Nauk SSSR, 217(1): 119 (in Russian). Barruel, P. (1984) Prepr. Int. Rubber Conf. Moscow, C41. Bartenev, G.M. (1979) Structure and Relaxation Properties of Elastomers. Moscow: Khimiya Publishers, 288 pp. (in Russian). Bartenev, G.M. and Barteneva, A.G. (1992) Relaxation Properties of Polymers. Moscow: Khimiya Publishers, 382 pp. (in Russian). Bartenev, G.M. and Barteneva, A.G. (2005) Kauchuk i Rezina, No 1: 8 (in Russian). Bartenev, G.M. and Gorbatkina, Yu.A. (1959) Vysokomol. Soed., 1(5): 769 (in Russian). Bartenev, G.M. and Kilimnik, O.V. (1983) Koll. Zh., 45(2): 299 (in Russian). Bartenev, G.M. and Kuchersky, A.M. (1970) Mekhan. Polim., No 3: 544 (in Russian). Bartenev, G.M. and Kuchersky, A.M. (1970) Vysokomol. Soed., ser. A, 12(4): 79 (in Russian). Bartenev, G.M. and Lazorenko, M.V. (1984) Vysokomol. Soed., ser. B, 26(5): 330 (in Russian). Bartenev, G.M. and Lukyanov, I.A. (1955) Zh. Fiz. Khim., 29: 1486 (in Russian). Bartenev, G.M. and Lukyanov, I.A. (1963) Plast. Massy, No 12: 46 (in Russian). Bartenev, G.M. and Vishnitskaya, L.A. (1952) Zh. Techn. Phys. 22: 426 (in Russian). Bartenev, G.M. and Vishnitskaya, L.A. (1962) Vysokomol. Soed., 4: 1324 (in Russian). Bartenev, G.M. and Voevodskaya, M.V. (1964) Kauchuk i Rezina, No 3: 21(in Russian). Bartenev, G.M. and Voevodskaya, M.V. (1964) Kauchuk i Rezina, No 12: 14 (in Russian). Bartenev, G.M. and Voevodskaya, M.V. (1966) Kauchuk i Rezina, No 3: 25 (in Russian). Bartenev, G.M., Kuchersky, A.M. and Radaeva, G.I. (1977) Vysokomol. Soed., ser. B, 19(8): 564 (in Russian). Bartenev, G.M., Kuchersky, A.M. and Radaeva, G.I. (1981) Vysokomol. Soed., ser. A, 23(2): 283 (in Russian). Bartenev, G.M., Lyalina, N.M. and Revyakin, B.I. (1982) Vysokomol. Soed., ser. B, 24(8): 567 (in Russian). Bartenev, G.M., Kuchersky, A.M., Fedyukina, L.P. and Radaeva, G.I. (1979) Kauchuk i Rezina, No 12: 35 (in Russian). Beardsley, K.P. and Wortman, W.A. (2000) Abstr. Int. Rubber Conf. Helsinki, p. 102. Beatty, C.L. and Karasz, F.E. (1975) J. Polym. Sci. Polym. Phys. Ed., 13(5): 971. Begishev, V.P., Bolgov, S.A., Malkin, A.Ya. et al. (1980) Vysokomol. Soed., ser. B, 22(2): 124
REFERENCES
66. 67. 68. 69. 69a. 70. 71. 72. 73. 74. 75. 76. 77. 78. 79. 80. 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. 91. 92. 93. 94. 95. 96. 97.
159
(in Russian). Begishev, V.P., Kipin, I.A. and Malkin, A.Ya. (1982) Vysokomol. Soed., ser. B, 24(9): 656 (in Russian). Begishev, V.P., Kipin, I.A., Andrianova, Z.S. and Malkin, A.Ya. (1983) Vysokomol. Soed., ser. B, 25(5): 343; ser. A, 25(11): 2441 (in Russian). Bekkedahl, N.J. (1934) J. Res. Nat. Bur. Stand., 13(2): 411; (1935) Rubb. Chem. Technol., 8(1): 5. Bekkedahl, N.J. (1967) Rubb. Chem. Technol., 40(3): 655. Bekkedahl, N.J. and Wood, L.A. (1941) J. Chem. Phys., 9: 193; (1941) Ind. Eng. Chem., 33: 381; (1941) Rubb. Chem. Technol., 14: 514. Bell, C.L.M., Stinson, D. and Thomas, A.G. (1982) Rubb. Chem. Technol., 55(1): 66. Belousov, V.N., Kotsev, G.Kh. and Mikitayev, S.A. (1985) Dokl. Akad. Nauk SSSR, 280(5): 1140 (in Russian). Belyaev, O.F., Kunizhev, B.I. and Zelenev, Yu.V. (1979) Vysokomol. Soed., ser. B, 21(4): 271 (in Russian). Berezkina, A.P. and Novikova, N.N. (1982) In Garmonov, I.V. (ed.) Phase-Aggregate State and Properties of Substances. Moscow: TsNIITENeftehkim, p. 126 (in Russian). Berezkina, A.P., Avstriyskaya, E.E. and Nelson, K.V (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekkim, p. 117 (in Russian). Berezkina, A.P., Nelson, K.V. and Kurlyand, S.K. (1986) Kauchuk i Rezina, No 12: 4 (in Russian). Berghmans, H. (1991) Abstr. Int. Conference “Polymer Networks”. Moscow, L-20. Berry, J.R. (1972) In Leibowitz, H. (ed.) Fracture. An Advanced Treatise. V. 7. Fracture of Nonmetals and Composites. New York–London: Academic Press; Russian translation (1976) Moscow: Mir Publishers, p. 8. Bershtein, V.A. and Egorov, V.M. (1985) Vysokomol. Soed., ser. A, 27(11): 2440 (in Russian). Bershtein, V.A. and Egorov, V.M. (1990) Differential Scanning Calorimetry in Polymer Physical Chemistry. Leningrad: Khimiya Publishers, 255 pp. (in Russian). Bershtein, V.A., Egorov, V.M. and Emelyanov, Yu.A. (1985) Vysokomol. Soed., ser. A, 27(11): 2451 (in Russian). Bershtein, V.A., Egorov, V.M. and Ryzhov, V.A. (1986) Vysokomol. Soed., ser. B, 28(4): 268 (in Russian). Bershtein, V.A., Egorov, V.M., Marikhin, V.A. and Myasnikova, L.P. (1985) Vysokomol. Soed., ser. A, 27(4): 771 (in Russian). Bianchi, U. (1965) J. Phys. Chem., 69(5): 1497. Bityukov, V.K., Khvostov, A.A., Titov, S.A., Sotnikov, P.A. and Zaichikov, M.A. (2006) Kauchuk i Rezina, No 5: 26 (in Russian). Boochathum, P. and Prajudlake, W. (2001) Eur. Polym. J., 37(3): 411. Boyer, R.F. (1963) Rubb. Chem. Technol., 36(5): 1301. Boyer, R.F. (1974) Macromolecules, 7(1): 142. Boyer, R.F. and Keineth, S.E. (ed.) (1980) Molecular Motion in Polymers by ESR. New York: Harwood Acad. Publishers, 328 pp. Boyer, R.F. and Simha, R. (1973) J. Polym. Sci. Polym. Lett. Ed., 11(1): 33. Brazier, D.W. and Nickel, G.H. (1979) Rubb. Chem. Technol., 52(4): 735. Brown, R.P. (1987) Physical Testing of Rubber, 2nd edition. London: Elsevier, 348 pp. Bruzzone, M. and Sorta, E. (1978) Polymer, 19(4): 467; (1979) Rubb. Chem. Technol., 52(1): 207. Bruzzone, M., Cesca, S. and Carbonaro, A. (1978) Rubb. Chem. Technol., 51(5): 907. Bukhina, M.F. (1963) Vysokomol. Soed., 5(11): 1725 (in Russian); (1964) Rubb. Chem. Techn., 37(2): 404. Bukhina, M.F. (1973) Crystallization of Rubbers. Moscow: Khimiya Publishers, 239 pp. (in Russian). Bukhina, M.F. (1974) Crystallization of Rubbers and its Effect on Mechanical Properties. DSc (Techn.) Thesis. Moscow: Karpov Physical Chemistry Research Institute, 258 pp. (in Russian). Bukhina, M.F. (1984) Technical Physics of Elastomers. Moscow: Khimiya Publishers, 224 pp. (in Russian).
160
98. 99.
REFERENCES
Bukhina, M.F. (2000) Abstr. Int. Rubber Conf. Helsinki, p. 120. Bukhina, M.F. (ed.) (1981) Low Temperature Resistance of Elastomers and Elastomer-based Goods. Abstr. Papers Presented at the Seminar at the All-Union Exhibition of the Achievements of National Economy. Moscow: TsNIITENeftekhim, 168 pp. (in Russian). 100. Bukhina, M.F. and Gallymov, I.I. (1979) Vysokomol. Soed., ser. B, 21(3): 198 (in Russian). 100a. Bukhina, M.F. and Karandashov, B.P. (1972) Vysokomol. Soed., ser. B, 14(3): 174 (in Russian) 101. Bukhina, M.F. and Parizenberg, M.D. (1981) Vysokomol. Soed., ser. B, 23(4): 453 (in Russian). 102. Bukhina, M.F. and Severina, N.L. (1982) Abstr. 3rd Conf. “Surface Phenomena in Polymers”. Kiev: p. 18 (in Russian). 103. Bukhina, M.F. and Zorina N.M. (1991) Abstr. Int. Conference “Polymer Networks”. Moscow, P-22. 104. Bukhina, M.F. and Zorina, N.M. (1992) Abstr. Advanced Research Workshop “Crystallization of Polymers”. Mons (Belgium), P4. 105. Bukhina, M.F., Basova, V.P., Voloshin, V.N. and Mikhlin, V.E.(1984) Elektrotekhn. Prom., ser. kabeln. tekhn., No 8: 9 (in Russian). 106. Bukhina, M.F., Derbeneva, A.P. and Vodzinskaya (Severina), N.L. (1971) In Lipatov, Yu.S. (ed.) Macromolecules at the Interface. Kiev: Naukova Dumka, p. 233 (in Russian). 106a. Bukhina, M.F., Derbeneva, A.P. and Zhitlovskaya, A.I. (1971) Kauchuk i Rezina, No 1: 7 (in Russian). 107. Bukhina, M.F., Devirts, E.Ya., Lavrova, A.A. and Severina, N.L. (1975) Vysokomol. Soed., ser. B, 17(8): 632 (in Russian). 108. Bukhina, M.F., Devirts, E.Ya., Lavrova A.A. and Titarenko, S.A. (1979) Kauchuk i Rezina, No 3: 22 (in Russian). 109. Bukhina, M.F., Galperina (Zorina), N.M. and Severina, N.L. (1978) Prepr. Int. Rubber Conf. Kiev, A 11. 110. Bukhina, M.F., Galperina (Zorina), N.M. and Severina, N.L. (1979) Vysokomol. Soed., ser. B, 21(8): 591 (in Russian). 111. Bukhina, M.F., Galperina (Zorina), N.M. and Severina, N.L. (1984) Prepr. Int. Rubber Conf. Moscow, A 37. 112. Bukhina, M.F., Severina, N.L. and Galperina (Zorina), N.M. (1975) Vysokomol. Soed., ser. B, 17(2): 127 (in Russian). 113. Bukhina, M.F., Severina, N.L. and Galperina (Zorina), N.M., (1986). Abstr. 4th Conf. on Methods of Calculation of Elastomer Goods. Riga, p. 103 (in Russian). 114. Bukhina, M.F., Severina, N.L. and Zorina, N.M. (1988) Kauchuk i Rezina, No 2: 15 (in Russian). 115. Bukhina, M.F., Severina, N.L. and Zorina, N.M. (1993) Kauchuk i Rezina, No 3: 26 (in Russian). 116. Bukhina, M.F., Severina, N.L., Galperina (Zorina), N.M. and Donner, A.D. (1978) In Prediction of the Properties of Rubbers and Rubber Articles. Moscow: TsNIITENeftekhim, p. 11 (in Russian). 117. Bukhina, M.F., Severina, N.L., Galperina (Zorina), N.M., Titarenko, S.A. and Parizenberg, M.D. (1977) Abstr. All-Russian Conference “Properties and Application of Materials under Low Temperatures”, Yakutsk: p. 55 (in Russian). 118. Bukhina, M.F., Severina, N.L., Leonov, I.I. and Artem’ev, V.N. (1993) Prepr. 5th Symp. “Problems of Tyres and Rubber/Cord Compositions”. Moscow, p. 21 (in Russian). 119. Bukhina, M.F., Severina, N.L., Novikova, T.N., Artem’ev, V.N. and Leonov, I.I. (1989) Abstr. Conf. “Cryophobia and Cryophobic Ice Resistant Materials”. Yakutsk, p. 25 (in Russian). 120. Bukhina, M.F., Severina, N.L., Yurovsky, V.S. et al. (1986) Abstr. Conf. on Problems of Increasing the Quality of Rubbers and Rubber Articles. Moscow: TsNIITENeftekhim, p. 7 (in Russian). 121. Bukhina, M.F., Severina, N.L., Zorina, N.M., Golshtrach, I.S. and Morova, L.G. (1994) Proc. Int. Rubber Conf. Moscow, vol. 4, p. 368. 122. Bukhina, M.F., Severina, N.L., Zorina, N.M., Meladze, R.A. and Petropavlovskaja, M.V. (1994) Proc. Int. Rubber Conf. Moscow, vol. 4, p. 375.
REFERENCES
123. 124.
125.
126. 127. 128. 129. 130. 131.
132. 133. 134. 135. 136. 137. 138. 139. 140. 141. 142. 143. 144. 145. 146. 147. 148. 149. 150.
161
Bukhina, M.F., Voevodskaya, M.V. and Karandashov, B.P. (1971) Vysokomol. Soed., ser. A, 13(4): 775 (in Russian). Bukhina, M.F., Zorina, N.M. and Morozov, Yu.L. (2004) Abstr. Int. Rubber Conf. Moscow, p. 53; Prepr. 15th Symp. “Problems of Tyres and Rubber/Cord Compositions”. Moscow, vol. 1, p. 74 (in Russian) (2005); Inzh.-Fiz. Zh. (J. Eng. Phys. Thermophys.), 78(5): 19 (in Russian). Bukhina, M.F., Zorina, N.M. and Severina, N.L., (1981) In Bukhina, M.F. (ed.) Low Temperature Resistance of Elastomers and Elastomer-based Goods. Abstr. Papers Presented at the Seminar at the All-Union Exhibition of the Achievements of National Economy. Moscow: TsNIITENeftekhim, p. 5 (in Russian). Bukhina, M.F., Zorina, N.M. and Severina, N.L. (1995) Prepr. 6th Symp. “Problems of Tyres and Rubber/Cord Compositions”. Moscow, vol. 1, p. 33 (in Russian); (1996) Kauchuk i Rezina, No 3: 4 (in Russian); (1996) Int. Polym. Sci. Techn., 23(11): T/36. Bukhina, M.F., Zorina, N.M., Lysova, G.A. and Morozov, Yu.L. (2000) Kauchuk i Rezina, No 4: 28 (in Russian); (2001) Int. Polym. Sci. Techn., 28(2): T/34. Bukhina, M.F., Zorina, N.M., Morozov, Yu.L. and Kuznetsova, E.I. (2003) Prepr. 14th Symp. “Problems of Tyres and Rubber/Cord Compositions”. Moscow, vol. 1, p. 92 (in Russian). Bukhina, M.F., Zorina, N.M., Severina, N.L. and Morozov, Yu.L. (1998) Prepr. 2nd Int. Conf. “Oilfield Engineering with Polymers”. London, p. 236. Bukhina, M.F., Zorina, N.M., Severina, N.L. and Morozov, Yu.L. (2002) Prepr. 13th Symp. “Problems of Tyres and Rubber/Cord Compositions”. Moscow, vol. 1, p. 74 (in Russian). Bukhina, M.F., Zorina, N.M., Severina, N.L., Morozov, Yu.L. and Sankina, G.A. (1998) Prepr. 9th Symp. “Problems of Tyres and Rubber/Cord Compositions”. Moscow, p. 37 (in Russian); (1999) Kauchuk i Rezina, No 1: 6 (in Russian); (1999) Int. Polym. Sci. Techn., 26(4): T/28. Bukhina, M.F., Zorina, N.M., Voloshin, V.N. and Rudenko, G.A. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, C10c. Bukhina, M.F., Zorina, N.M., Voloshin, V.N., Rudenko, G.A. and Kotova, I.P. (1989) Vysokomol. Soed., ser. A, 31(5): 1106 (in Russian). Bulavin, L.A., Aktan, E.Yu. and Zabashta, Yu.F. (2003) Vysokomol. Soed., ser. A, 45(10): 1684. Bunn, C.W. (1964) Crystals: Their Role in Nature and in Science. New York: Academic Press, 286 pp; Russian translation (1970) Moscow: Mir Publishers, 312 pp. Burfield, L. (1984) Polymer, 25(12): 1823. Catton, N.L. (1953) The Neoprenes. Principles of Compounding and Processing. Wilmington, Delaware: Rubber Chemicals Division E.I. Du Pont Nemours & Co., 245 pp.; Russian translation (1958) Leningrad: Goskhimizdat, 207 pp. Cesari, M., Galderazzo, F. and Cucinella, S. (1976) J. Polym. Sci. Polym. Lett. Ed., 14: 107. Cesari, M., Perego, G., Zauetta, A. and Gargani, L. (2003) Macromol. Chem., 181(5): 1143 Chalykh, A.E. (1987) Diffusion in Polymer Systems. Moscow: Khimiya Publishers, 510 pp. (in Russian). Cherbunina, G.D., Kurlyand, S.K. and Garmashov, E.P. (1982) In Garmonov, I.V. (ed.) Phase-Aggregate State and Properties of Substances. Moscow: TsNIITENeftekhim, p. 69 (in Russian). Cherbunina, G.D., Vasil’eva, M.A., Pankratova, E.D. and Mikheev, A.I. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 39 (in Russian). Chernov, V.M. and Fedotov, V.D. (1981) Vysokomol. Soed., ser. A, 23(4): 939. Chersky, I.N., Popov, S.N. and Golshtrakh, I.Z. (1992) Design and Calculation of Cold-Resistant Flexible Seals. Novosibirsk: Nauka Publishers Siberian Division of RAS, 123 pp. (in Russian). Chi, M. (1981) J. Polym. Sci. Polym. Chem. Ed., 19(7): 1767. Chinnosamy, T.V. and Harimur, H. (1983) Angew. Makromol. Chem., 112: 45. Clark, E.I. and Hoffman, I.D. (1984) Macromolecules, 17(4): 878. Clarson, S.I. (1985) Polymer, 26(6): 930. Coe, D.G. and Howgate, P.G. (1986) Proc. Int. Rubb. Conf., Göteborg, vol. 2, p. 418. Cohen, M.H. and Grest, G.S. (1979) Phys. Rev., B20(3): 1077.
162
REFERENCES
Cole, Y.H. and Pierre, L.E. (1978) J. Polym. Sci. Polym. Symp., 63: 205. Collins, E.A. and Chandler, L.A. (1966) Rubb. Chem. Technol., 39(2): 193. Conchman, P.R. (1981) Polymer Eng. Sci., 21(7): 377. Cook, S. (2000) Abstr. Int. Rubber Conf. Helsinki, p.97. Copley, B.C. (1982) Rubb. Chem. Technol., 55(2): 416. Corbelli, L., Martini, E. and Milani, F. (1987) Abstr. 9th Int. Rubb. Conf. Gotvaldov, Czech Republic, A 5. 157. Cowie, J.M. (1975) Eur. Polym. J., 11: 297. 158. Dalal, E.N. and Phillips, P.J. (1983) Macromol., 16(11): 1754; (1984) 17(2): 248; (1984) J. Polym. Sci. Polym. Lett. Ed., 22(1): 7. 159. Dalal, E.N. and Phillips, P.J. (1983) Macromol., 16(6): 890; 160. Dalal, E.N., Taylor, K.D. and Phillips, P.J. (1983) Polymer, 24(12): 1623. 161. Davey, A.B. and Payne, A.R. (1964) Rubber in Engineering Practice. London: Maclaren – New York: Palmerton, 501 pp. 162. Daw, N.P. and Bhuniya, R.C. (1983) Bull. Mater. Sci., 5(5): 425. 163. De Candia, F., Romano, G. and Russo, R. (1984) Prepr. Int. Rubber Conf. Moscow. A 110; (1985) Macromol. Chem. Rapid Comm., 6(1): 15. 164. De Candia, F., Romano, G., Russo, R. and Vittoria, V. (1982). J. Polym. Sci. Polym. Phys. Ed., 20(9): 1525. 165. De Candia, F., Romano, G. and Vittoria, V. (1973) J. Polym. Sci. Polym.Phys. Ed., 11: 2291. 165a. De Candia, F., Romano, G. and Vittoria, V. Macromol. Chem., (1974) B175(11): 2983. 166. De Candia, F., Russo, R. and Vittoria, V. (1980) J. Macromol. Sci., part B, 18(2): 257. 167. De Cherico, L., De Candia, F., Romano, G. and Russo, R. (1975) Polymer, 16(11): 845. 168. Degteva, T.G., Dontsov, A.A. and Gudkova, V.M. (1984) Proizvodstvo Shin, RTI i ATI, No 5: 27 (in Russian). 169. Degteva, T.G., Dontsov, A.A., Gudkova, V.M. and Kondrat’eva, V.F. (1983) Kauchuk i Rezina, No 7: 27 (in Russian). 170. Degteva, T.G., Dontsov, A.A., Khodosh, T.S. and Kondrat’eva, V.F. (1981) Kauchuk i Rezina, No 12: 37 (in Russian). 171. Degteva, T.G., Granovskaya, I.Ya., Gudkova, V.M. and Dontsov, A.A. (1979) Kauchuk i Rezina, No 4: 26 (in Russian). 172. Degteva, T.G., Gudkova, V.M. and Smyslova, RA. (1984) Promyshlennost’ SK, Shin i RTI, No 4: 31 (in Russian). 173. Della Monica, C., Forte, A., Maglio, J. and Riva, F. (1984) Prepr. Int. Rubber Conf. Moscow, A 36. 174. Dobrokhotov, N.E., Bukhina, M.F., Litvinova, T.V., Popova, P.I. and Lukashevich, N.P. (1973) Proizvodstvo Shin, RTI i ATI, No 5: 33 (in Russian). 175. Dobrokhotov, N.E., Litvinova, T.V. and Bukhina, M.F. (1975) Vysokomol. Soed., ser. A, 17(4):705 (in Russian). 176. Dobrokhotov, N.E., Litvinova, T.V., Bukhina, M.F. and Zubov, P.I. (1972) Kauchuk i Rezina, No 8: 7; (1974) Proizvodstvo Shin, RTI i ATI, No 9: 3 (in Russian). 177. Dogadkin, B.A., Bartenev, G.M. and Reznikovsky, M.M. (1949) In Medvedev, S.S. (ed.) Studies in the Field of High-molecular Compounds. Moscow: USSR Acad. Sci. Publishers, p. 297 (in Russian). 178. Dogadkin, B.A., Dontsov, A.A. and Shershnev, V.A. (1981) Elastomer Chemistry, 2nd edition. Moscow: Khimiya Publishers, 376 pp. (in Russian). 179. Donnet, J.B. (1994) Proc. Int. Rubber Conf. Moscow, vol. 1, p. 129. 180. Dontsov, A.A. (1978) Processes of Structure-Formation of Elastomers. Moscow: Khimiya Publishers. 287 pp. (in Russian). 181. Dontsov, A.A., Kanauzova, A.A. and Litvinova, T.V. (1986) Rubber/Oligomer Composites in Rubber Goods Production. Moscow: Khimiya Publishers, 216 pp. (in Russian). 182. Drapkin, A.E., Marei, A.I. and Dmitriev, V.S. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 33 (in Russian). 183. Dunning, D.J. and Pennells, P.J. (1967) Rubb. Chem. Techn., 40(5): 1381. 184. Dutt, B. (1984) In Stevenson, A. (ed.) Rubber in Offshore Engineering. Bristol: Adam Hilger Ltd., p. 208. 185. Edwards, B.C. and Phillips, P.J. (1974) Polymer, 15(8): 491; (1975) J. Mat. Sci., 10(7): 1233. 151. 152. 153. 154. 155. 156.
REFERENCES
186. 187. 188. 189. 190. 191. 192. 193. 194. 195. 196. 197. 198. 199. 200. 201. 202. 203. 204. 205. 206. 207. 208. 209. 210. 211. 212. 213. 214. 215. 216. 217. 218. 219. 220.
163
Edwards, B.C. and Phillips, P.J. (1975) J. Pol. Sci. Polym. Phys. Ed., 13(11): 2117. Edwards, B.S., Phillips, P.J. and Sorenson, D. (1980) J. Polym. Sci. Phys. Ed., 18: 1737. Eirikh, F.R. and Smith, T.L. (1972) In Leibowitz, H. (ed.) Fracture. An Advanced Treatise. V. 7. Fracture of Nonmetals and Composites. New York–London: Academic Press; Russian translation (1976) Moscow: Mir Publishers, p.100. Eisenberg, A. and Shen, M.C. (1970) Rubb. Chem. Technol., 43: 156. Eldred, R.J. (1984) Rubb. Chem. Technol., 57(2): 320. Elyashevich, G.K. and Frenkel, S.Ya. (1980) In Malkin, A.Ya. and Papkov, S.P. (eds.) Orientational Phenomena in Polymer Solutions and Melts. Moscow: Khimiya Publishers, p. 9 (in Russian). Enle, Z., Guiping, J. and Yanshou, L. (1982) Polym. Commun., No 3: 178; (1986) No 4: 251. Fedors, R.F. (1979) Polymer, 20(9): 1055; 20(10): 1255. Fedotov, V.D., Chernov, V.M. and Khazanovich, T.N. (1978) Vysokomol. Soed., ser. A, 20: 919 (in Russian). Fedyukin, D.L. and Zakharenko, N.V. (1960) Kauchuk i Rezina, No 9: 43 (in Russian). Fedyukin, N.D., Severina, N.L., Ionov, Yu.A. and Bukhina, M.F. (1984) Kauchuk i Rezina, No 7: 14 (in Russian). Fetdman, G.I., Maiskaya, M.A. and Gorelik, B.M. (1976) Determination of Stresses in Rubber Goods by Photoelasticity. Moscow: Khimiya Publishers, 109 pp. (in Russian). Ferry, J.D. (1961) Viscoelastic Properties of Polymers. 1st edition; Russian translation (1963) Moscow: Inostr. Lit. Publishers, 535 pp.; (1980) 3rd edition, John Wiley & Sons, 672 pp. Fisher, E.W. (1990) Proc. 2nd Int. Workshop Non-Crystal. Solids: “Basic Features of Glassy States”. Singapore, p. 172. Flory, P.J. (1947) J. Chem. Phys., 15(6): 397; 15(9): 684. Flory, P.J. (1949) J. Chem. Phys., 17(3): 223. Fomina, L.G. and Yurovsky, V.S. (2000) Kauchuk i Rezina, No 2: 39 (in Russian). Fomina, L.G., Sedov, V.V. and Kaporovsky, B.M. (2000) Kauchuk i Rezina, No 4: 26 (in Russian). Fox, T.G. and Flory, P.J.J. (1950) J. Appl. Phys., 21: 581. Freakley, P.K. and Payne, A.R. (1978) Theory and Practice of Engineering with Rubber, London: Applied Science Publications, 666 pp. Frenkel, S.Ya. (1973) In Physics Today and Tomorrow. Leningrad: Nauka Publishers, p. 176 (in Russian); (1977) Khim. Volokna, No 3: 11 (in Russian). Frenkel, S.Ya. (1992) Abstr. Int. Conference “Orientational Phenomena in Polymers”. St. Petersburg, p. 3. Frenkel, S.Ya. and Elyashevich, G.K. (1971). Vysokomol. Soed., ser. A, 13(2): 493 (in Russian). Frisch, H.L. and Mark, J.E. (196) Chem. Mater., 8(8): 1735. Fuller, K.N.G. (2000) Abstr. Int. Rubber Conf. Helsinki, p. 86. Furukawa, I. (1973) Polym. J., 4(4): 358. Galil-Ogly, F.A., Novikov, A.S. and Nudelman, Z.N. (1966) Fluoroelastomers. Moscow: Khimiya Publishers, 235 pp. (in Russian). Gallymov, I.I., Bukhina, M.F., Maklakov, A.I. (1978). Vysokomol. Soed., ser. B, 20: 336 (in Russian). Galperina (Zorina), N.M. and Severina, N.L. (1982) In Application of Mathematical Methods in Rubber Research. Moscow: TsNIITENeftekhim, p. 25 (in Russian). Galperina (Zorina), N.M., Bukhina, M.F., Chertkova, V.F. and Dudinova, N.P. (1982) Proizvodstvo Shin, RTI i ATI, No 1: 3 (in Russian). Gamlitsky, Yu.A. and Bass, Yu.P. (2003) Inzh.-Fiz. Zh. (J. Eng. Phys. Thermophys.), 76(5): 101 (in Russian). Gargani, L. and Bruzzone, M. (1986) In Lae, L. and Mark, Y.E. (eds.) Advances in Elastomers and Rubber Elasticity, London–New York: Plenum Press, p. 233. Gargani, L., De Ponti, P. and Bruzzone, M. (1986) Proc.Int. Rubb. Conf. Göteborg, vol. 1, p. 171. Gargani, L, Giuliani, G., Perego, G. and Zarretta, A. (1977) Proc. 6th Europhysics Conf. on Macromolec. Physics. Varna, Bulgaria, p. 10. Garmonov, I.V. (ed.) (1982) Phase-Aggregate State and Properties of Substances. Moscow:
164
REFERENCES
TsNIITENeftekhim, 173 pp. (in Russian). Garmonov, I.V. (ed.) (1983) Synthetic Rubbers, 2nd edition. Moscow: Khimiya Publishers, 559 pp. (in Russian). 222. Garn, P.D. and Menis, O. (1977) J. Macrom. Sci., B13(4): 611. 223. Gartzman, V.I. and Gorelik, B.M. (1978) Kauchuk i Rezina, No 4: 24; (1972) Mechan. Polym., No 6: 1136 (in Russian). 224. Gee, G. (1966) Polymer, 7(4): 177. 225. Gee, G. (1970) Contemp. Phys., 11: 31. 226. Gee, G., Hartly, J., Herbert, J.B.M. and Laneeley, H.D. (1960) Polymer, 1(1): 365. 227. Gehman, S.D., Jones, P.Y., Wilkinson, C.S. and Woodford, D.E. (1950) Ind. Eng. Chem., 42(3): 475. 228. Geil, P. H. (1963) Polymer Single Crystals. New York: Interscience; Russian translation (1965) 552 pp. 229. Geisler, H., Schuster, R.H. and Schölz, G. (2003) Proc. Int. Rubb. Conf. Prague, p. 47. 230. Gent, A.N. (1954) Trans. Farad. Soc., 50: 521; (1955) Rubb. Chem. Technol., 28(1): 36. 231. Gent, A.N. (1954) Trans. Inst. Rubb. Ind.,: 30:139; (1955) Rubb. Chem. Technol., 28(2): 457. 232. Gent, A.N. (1955) Trans. Inst. Rubb. Ind.,: 30:144; (1955) Rubb. Chem. Technol., 28(2): 460. 233. Gent, A.N. (1955) J. Polym. Sci., 18: 321. 234. Gent, A.N. (1965) J. Polym. Sci., part A, 3(11): 3787. 235. Gent, A.N. (1966) J. Polym. Sci., part A-2, 4(4): 447. 236. Gent A.N. (1972) In Leibowitz, H. (ed.) Fracture. An Advanced Treatise. V. 7. Fracture of Nonmetals and Composites. New York–London: Academic Press; (1976) Russian translation (1976) Moscow: Mir Publishers, p. 66. 237. Gent, A.N. (1986) In Lae, L. and Mark, Y.E. (eds.) Advances in Elastomers and Rubber Elasticity, London–New York: Plenum Press, p. 253. 238. Gent, A.N. and Bhowmick, A.K. (1983) Rubb. Chem. Technol., 56(4): 845. 239. Gent, A.N. and Zhang, L.-Q. (2002) Rubb. Chem. Technol., 75(5): 923. 240. Gerasimov, V.I. and Tsvankin, D.Ya. (1969) Vysokomol. Soed., ser. A, 11(12): 2652 (in Russian). 241. Gerspacher, M., Jang, H.H. and Storita, J.M. (1991) L’actualite Chimique, No 2: 106. 242. Gibbs, I.H. and Di Marzio, E.A. (1958) J. Chem. Phys., 39: 373. 243. Giuliani, G. (1976) La Chimica e l’Industria, 58(8): 553. 244. Glushko, V.V., Bukhina, M.F., Lagover, I.E. and Grigor’eva, N.I. (1982) Proizvodstvo Shin, RTI i ATI, No 2: p. 5 (in Russian). 245. Glushko, V.V., Lagover, I.E., Bukhina, M.F. and Tamulevich, G.D. (1978) Kauchuk i Rezina, No 3: 43 (in Russian). 246. Godovsky, Yu.K. (1976) Thermophysical Methods of Polymer Investigation. Moscow: Khimiya Publishers, 213 pp. (in Russian). 247. Godovsky, Yu.K. (1982) Thermophysics of Polymers. Moscow: Khimiya Publishers, 280 pp. (in Russian). 248. Godovsky, Yu. K. (1982) Coll. Polym. Sci., 260: 461. 249. Godovsky, Yu.K. (1992) Thermophysical Properties of Polymers. New York–Berlin–Heidelberg: Springer-Verlag, 310 pp. 249a. Godovsky, Yu.K. (1993) Abstr. 28 Europhysics Conference on Macromolecular Physics “Transitions in Oligomer and Polymer Systems”. Ulm, Germany, P-6. 250. Godovsky, Yu.K. and Bessonova, N.N. (1977) Vysokomol. Soed., ser. A, 19(12):2731 (in Russian). 251. Godovsky, Yu.K. and Makarova, N.N. (1994) Phil. Trans.: Phys. Sci. Eng., 348(1686): 45. 252. Godowsky, Ju.K. and Papkov, V.S. (1991) Abstr. Int. Conference on Polymer Networks. Moscow, L-23. 253. Godovsky, Yu. K. and Valetskaya, L A. (1991) Polym. Bull., 27(2): 221. 254. Godovsky, Yu.K., Makarova, N.N., Papkov, V.A. and Kuz’min, N.N. (1985) Vysokomol. Soed., ser. B, 25(3): 359 (in Russian); (1985) Macromolecular Chemistry, Rapid Commun., 6(6): 443. 255. Godovsky, Yu.K., Volegova, I.A., Valetskaya, L.A., Rebrov, A.V., Novitskaya, A.A. and Rotenburg S.I. (1988) Vysokomol. Soed., ser. A, 30(2): 359. 256. Goldovsky, E.A., Korotkikh, L.A., Bukhina, M.F., Kuz’minsky, A.S. et al. (1975) In Studies 221.
REFERENCES
165
in the Field of Physics and Chemistry of Rubber. Leningrad: Khimiya Publishers, p. 140 (in Russian). 257. Goncharova, Yu.E., Sakharova, E.V. and Potapov, E.E. (1995) Kauchuk i Rezina, No 6: 17 (in Russian). 258. Goppel, G.M. (1948) Proc. 2nd Rubb. Techn. Conf. London, p. 164. 259. Goppel, G.M. (1948) Rubb. Chem. Techn., 21(4): 773; 783. 260. Goppel, J.M. and Arlman, J.J. (1950) Rubb. Chem. Technol., 23(2): 310. 261. Gorbatkina, Yu.A. (1960) Vysokomol. Soed., 2(10): 1456 (in Russian). 262. Gorelik, B.M. and Kaporovsky, B.M. (1969) Khim. Neftekhim. Mashinostroen., No 4: 33 (in Russian); (1969) Kauchuk i Rezina, No 1: 38 (in Russian). 263. Gorelik, B.M., Bukhina, M.F. and Kapshtyk, V.I. (1965) Kauchuk i Rezina, No 2: 20 (in Russian). 264. Göritz, D. (1992) Angew. Makromol. Chem., 202–203: 309. 265. Göritz, D. and Gassler, R. (1987) Rubb. Chem. Technol., 60(2): 217. 266. Göritz, D. and Kiss, M. (1986) Rubb. Chem. Technol., 59(1): 40. 267. Göritz, D. and Müller, F.H. (1970) Koll. Z. u. Z. Polymere, 241: 1075. 268. Göritz, D. and Müller, F.H. (1986) Rubb. Chem. Technol., 59(2): 210. 269. Göritz, D., Müller, F.H. and Sietz, L.E. (1977) Progress in Colloid and Polym. Sci., 62(2): 114. 270. Gotlib, Yu.Ya. (1993) Abstr. 28th Europhysics Conference on Macromolecular Physics “Transitions in Oligomer and Polymer Systems”. Ulm, Germany, G-5. 271. Gotlib, Yu.Ya., Darinsky, A.A. and Svetlov, Yu.E. (1986) Physical Kinetics of Macromolecules. Leningrad: Khimiya Publishers, 271 pp. (in Russian). 272. Govorova, O.A., Frolov, A.E. and Sorokin, G.L. (1986) Properties of Vulcanized Rubbers Based on Ethylene-Propylene Rubbers. Moscow: TsNIITENeftekhim, 62 pp. (in Russian). 273. Govorova, O.A., Galil-Ogly, F.A. and Zakirova, M.A (1972) Acrylic Rubbers. Moscow: TsNIITENeftekhim, 38 pp. (in Russian). 273a. Govorova, O.A., Vishnitsky, A.S., Chubarova, G.V. and Morozov, Yu.L. (1999) Kauchuk i Rezina, No 2: 18 (in Russian). 274. Grest, G.S. and Cohen, M.H. (1980) Phys. Rev., B21(9): 4113. 275. Grmela, V. and Megarska, L. (1968) Kauchuk i Rezina, No 9: 31; Vysokomol. Soed., ser. A, 10(5): 1032 (in Russian). 276. Grosch, K.A. (1984) Prepr. Int. Rubber Conf. Moscow, B1. 277. Gubina, T.I., Talroze, R.V., Dakin, V.I., Shmakova, N.A., Shibaev, V.P., Sukhov, F.F. and Plate, N.A. (1988) Vysokomol. Soed., ser. B, 30: 920 (in Russian). 278. Gul’, V.E. (1978) Structure and Strength of Polymers, 3rd edition. Moscow: Khimiya Publishers, 328 pp. (in Russian). 279. Gul’, V.E. (1984) Prepr Int. Rubber Conf. Moscow, 57. 280. Gul’, V.E. and Kuleznev, V.N. (1979) Structure and Mechanical Properties of Polymers, 3rd edition. Moscow: Vysshaya Shkola Publishers, 349 pp. (in Russian). 281. Gurevich, G.I. (1947) Zh. Tekhn. Fiz., 17(12): 1491 (in Russian). 282. Gusev, A.V., Konovalenko, N.A., Kharitonov, A.G. and Rachinsky, A.V. (2004) Kauchuk i Rezina, No 3: 2, 4 (in Russian). 283. Hager, T., McArthur, A., McIntyre, D. and Seeger, R. (1979) Rubb. Chem. Technol., 52(4): 693. 284. Hamed, G.R. (1981) Rubb. Chem. Technol., 54(3): 576. 284a. Hamed, G.R. (1981) Rubb. Chem. Technol., 55(1): 151. 285. Hamed, G.R. (2000) Rubb. Chem. Techn., 73(3): 524. 286. Hamed, G.R. and Shieh, C.H. (1985) Rubb. Chem. Technol., 57(5): 1469. 287. Havranek, A. (1982) Polymer Bull., 8: 133. 288. Hayashi, S., Sakakida, H., Oyama, M. and Nakagawa, T. (1991) Rubb. Chem. Techn., 64(4): 534. 289. Hepburn, C. (1982) Polyurethane Elastomers. London–New York: Applied Science Publications, 462 pp. 290. Hertz, D. L., Bussem, H. and Ray, T.W. (1995) Rubb. Chem. Techn., 68(3): 540. 291. Ho, R.Y., Ho, C.S., Hwan, H.G. and Ha, H.G. (1980) Bull. Acad. Sci. D.P.R. Korea, 1: 30. 292. Hoffman, I.D. (1983) Polymer, 24(1): 3.
166
REFERENCES
Hoffman, I.D. and Weeks, Y.Y. (1962) J. Chem. Phys., 37(8): 1722. Hordgies, K.H. (1986) Kautschuk und Gummi Kunststoffe, 39(7): 599. Hotge, J.M. and Huvard, G.S. (1983) Macrom., 16(3): 371. Howard, R.N. (ed.) (1973) The Physics of Glassy Polymers. London: Applied Science Publications, 605 pp. 297. Hsieh, H.L. and Yeh, M.C. (1985) Rubb. Chem. Technol., 58(1): 117. 298. Hsu, J.H. and Mark, J.E. (1987) Pap. 194th Nat. Meet. Am. Chem. Soc., p. 1116. 299. Huang, J.W., Kang, C.C., Lin, C.-J. and Lee, C.-T. (2005) Polym. J., 37(8): 418. 300. Huhn, G. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, F5. 301. Ito, S., Oyama, M., Kawanaka, T. and Nomoto, H. (2005) Book of Papers Int. Rubber Conf. Maastricht, p. 363. 302. James, H.H. and Hay, H. (1980) Anal. Proc., 17(6): 240. 303. Jaracki, L. and Ziabicski, A. (1992) Abstr. Int. Conference “Orientational Phenomena in Polymers”, St. Petersburg, p. 72. 304. Johnston, W.V. and Shen, M. (1969) J. Polym. Sci., A-2, 7(1): 1985; (1970) Amer. Chem. Soc. Polym. Prepr., 11: 154. 305. Jones Parry, E. and Tabor, D. (1974) J. Mater. Sci., 9: 289. 305a. Jordan, E. (1971) J. Polym. Sci., A1(9): 3367. 306. Jurkowski, B. and Jurkowska, B. (2004) Abstr. Int. Rubber Conf. Moscow, p. 285; (2005) Kauchuk i Rezina, No 2: 2. 307. Kadir, A. and Thomas, A.G. (1984) J. Polym. Sci. Phys. Ed., 22(9): 1623. 308. Kaimin’, I.F., Zuev, Yu.S., Bukhina, M.F. and Kozlov, P.V. (1971) Kauchuk i Rezina, No 5: 6 (in Russian). 309. Kamaeva, O.A. and Sire, E.M. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 32 (in Russian). 310. Kaporovsky, B.M. (1969) In Advances of Science and Technology in the Field of Rubber. Moscow: Khimiya Publishers, p. 273 (in Russian). 311. Kaporovsky, B.M. and Yurtsev, N.N. (1986) Modern Designs of Rubber Sealing Valves and Ways of Improving their Service Properties. Moscow: TsNIITENeftekhim, 43 pp. (in Russian). 312. Kaporovsky, B.M. and Yurtsev, N.N. (1990) Kauchuk i Rezina, No 12: 24 (in Russian). 313. Kargin, V.A. and Slonimsky, G.L. (1967) An Outline of the Physical Chemistry of Polymers. 2nd edition, 232 pp. 314. Karlina, I.A., Diner, E.Z. and Boikova, I.N. (1982) In Garmonov, I.V. (ed.) Phase-Aggregate State and Properties of Substances. Moscow: TsNIITENeftekhim, p. 132 (in Russian). 315. Kasatochkin, V.I. and Lukin, B.V. (1949) In Medvedev, S.S, (ed.) Study in the Field of High-molecular-weight Compounds. Moscow: USSR Acad. Sci. Publishers, p. 289 (in Russian). 316. Kasatochkin, V.I. and Lukin, B.V. (1949) Zh. Techn. Phys., 19: 76 (in Russian). 317. Kasatochkin, V.I. and Lukin, B.V. (1950) Zh. Techn. Phys., 20: 670 (in Russian). 318. Kasatochkin, V.I. and Lukin, B.V. (1951) Izv. Akad. Nauk SSSR, ser. fiz., 15: 209 (in Russian). 319. Katayama, Y., Kato, T., Ohyanagi, M., Ikeda, E. and Sekine, Y. (1986) Macromol. Chem. Rapid Commun., 7(7): 465. 320. Kausch, H.H. (1978) Polymer Fracture. 1st edition. Berlin: Springer-Verlag, 456 pp. (1987) 2nd edition. Berlin: Springer-Verlag; Russian translation (1981) Moscow: Mir Publishers, 440 pp. 321. Kazansky, V.V., Tamarkin, V.F., Bukhina, M.F. and Devirts, E.Ya. (1980) Proizvodstvo Shin, RTI i ATI, No 10: 29 (in Russian). 322. Keller, A. (1976) Proc. 5th Eur. Conf. on Macromolec. Phys. Budapest, p. 130; 323. Keller, A. (1985) In Polym. Liquid Crystal and Low Dimensional Solids, pp. 71, 117. 324. Keller, A. (1991) Abstr. Int. Conference on Polymer Networks. Moscow, L-19. 325. Keller, A. (1992) Abstr. Int. Conference “Orientational Phenomena in Polymers”. St. Petersburg, p. 1; (1993) Abstr. 28th Europhysics Conference on Macromolecular Physics “Transitions in Oligomer and Polymer Systems”. Ulm, Germany, P-3. 326. Kennedy, M.A., Turturro, G., Brown, G.R. and Pierre, L.E. (1983) J. Polym. Sci. Polym. Phys. 293. 294. 295. 296.
REFERENCES
167
Ed., 21(8): 1403. Kennon, D.B. (1984) Prepr. Int. Rubber Conf. Moscow, B62. Khodosh, T.S., Degteva, T.G. and Dontsov, A.A. (1981) Proizvodstvo Shin, RTI i ATI, No 10: 3 (in Russian). 329. Khotimsky, M.N., Krishtal, B.F. and Dontsov, A.A. (1982) Proc. All-Russian Conference on New Phenomena in Polymers. Kiev, p. 120 (in Russian). 330. Khrustalev, Yu.M., Nezabudkina, O.A., Vaschev, B.V. and Berkovich, M.A. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 56 (in Russian). 331. Khvostik, G.M., Ramsh, A.S., Kurlyand, S.K., Vasilev, V.A. and Timchenko, B.N. (2002) Abstr. 1st All-Russian Rubber Conference. Moscow, p. 79 (in Russian). 331a. Khvostik, G.M., Vaschev, B.V., Ventseslavskaya, K.K. et al. (2002) Abstr. 1st All-Russian Rubber Conference. Moscow, p. 135 (in Russian). 332. Killian, H.G. (1984) Prepr. Int. Rubber Conf. Moscow, A71; (1991) Abstr. Int. Conference on Polymer Networks. Moscow, L-22. 333. Killian, H.G. (1992) Abstr. Int. Conference on Orientational Phenomena in Polymers. St. Petersburg, p. 5. 334. Killian, H.G., Höhne, G.W., Trogele, P. and Ambacher, H. (1984) J. Pol. Sci. Polym. Symp., No 71: 222. 335. Kim, H. and Mandelkern, L. (1968) J. Polymer Sci., A2(6): 181. 336. Kim, H. and Mandelkern, L. (1972) J. Polym. Sci. Polym. Phys. Ed., 10: 1125. 337. Kisin, K.V., Lovchikov, V.A., Mironyuk, V.L. and Romanichin, V.B. (1981) Kauchuk i Rezina, No 6: 5 (in Russian). 338. Klauzen, N.A., Semenova, L.P. and Zaides, A.L. (1967) Vysokomol. Soed., ser. B, 9: 263 (in Russian). 339. Kobatake, T., Kodama, K., Hayashi, S. and Yoshioka, A. (1997) Rubb. Chem. Techn., 70(5): 839. 340. Kobeko, P.P. (1952) Amorphous Substances. Moscow–Leningrad: USSR Acad. Sci. Publishers, 432 pp. (in Russian). 341. Kojima, G. and Wachi, H. (1985) Proc. Int. Rubb. Conf. Kyoto, Japan, p. 242, 820. 342. Kolmogorov, A.N. (1937) Izv. Akad. Nauk SSSR, ser. math., No 3: 355. 343. Kolosova, V.I., Galperina (Zorina), N.M., Bukhina, M.F. and Artem’ev, V.N. (1987) Kauchuk i Rezina, No 2: 22 (in Russian). 344. Kolyadina, N.G. and Bartenev, G.M. (1961) Kauchuk i Rezina, No 9: 27 (in Russian). 345. Kormer, V.A., Kurlyand, S.K., Kovalev, N.F. and Tverdov, A.I. (1987) Kauchuk i Rezina, No 2: 20 (in Russian). 346. Koshelev, F.F., Kornev, A.E. and Bukanov, A.M. (1978) General Technology of Rubber. Moscow: Khimiya Publishers, 566 pp. (in Russian). 347. Kostrykina, G.I., Epshtein, V.G., Zakharov, I.D. and Bukhina, M.F. (1971) Vysokomol. Soed., ser. B, 13(4): 277 (in Russian). 348. Kostrykina, G.I., Tamarkin, V.F., Krishtal, I.V. and Voronov, V.A. (1986) Kauchuk i Rezina, No 5: 11 (in Russian). 349. Kostrykina, G.I., Zakharov, N.D. and Epshtein, V.G. (1970) Zavodsk. Lab., 36: 179 (in Russian). 350. Kostrykina, G.I., Zakharov, N.D., Kolpakov, N.I. and Babyuk, V.N. (1975) In Chemical Technology, ser. Rubber. Yaroslavl, p. 25 (in Russian). 351. Kotov, N.M., Bakeev, N.F., Belov, G.P. et al. (1975) Vysokomol. Soed., ser. A, 17(5): 1141 (in Russian). 352. Kotov, N.M., Zubov, Yu.A. and Bakeev, N.F. (1976) Dokl. Akad. Nauk SSSR, 229: 1375 (in Russian). 353. Kotov, Yu.I., Gerasimov, V.I. and Tsvankin, D.Ya. (1970) Zh. Prikl. Spektroscopii, 13: 1097 (in Russian). 354. Kotov, Yu.I., Serdan, A.A. and Pavlova, N.M. (1968) Vysokomol. Soed., ser. B, 10(2): 225 (in Russian). 355. Kow, C., Morton, M., Fetters, L.Y. and Hadjechristidis, N. (1982) Rubb. Chem. Technol., 55(1): 245. 356. Kozlov, G.V., Beloshenko, V.A., Gazaev, M.A. and Novikov, V.U. (1996) Mekhanika 327. 328.
168
REFERENCES
Kompozitnykh Materialov, 32(2): 270 (in Russian); (1996) Mechanics of Composite Materials 32(2): 186. 357. Kozlov, G.V. and Lipatov, Yu.S. (2003) Vysokomol. Soed., ser. A, 45(4): 660. 358. Kremer, F. (1993) Abstr. 28th Europhysics Conference on Macromolecular Physics “Transitions in Oligomer and Polymer Systems”. Ulm, Germany, G-12. 359. Kropacheva, E.N., Smirnova, L.V., Sakhabutdinov, A.G. and Burganov, T.G. (2004) Abstr. Int. Rubber Conf. Moscow, p. 131. 360. Kropacheva, E.N., Smirnova, L.V., Sakhabutdinov, A.G. and Burganov, T.G. (2005) Kauchuk i Rezina, No 1: 2 (in Russian). 361. Kuchersky, A.M. (1973) Proc. All-Union Scientific-Technical Conference “Working Capacity of Mashines and Constructions under Conditions of Low Temperatures. Cold Resistance of Materials”. Yakutsk, p. 111 (in Russian). 362. Kuchersky, A.M. (1991) Kauchuk i Rezina, No 7: 3 (in Russian). 363. Kuchersky, A.M. (2002) Abstr. 1st All-Russian Rubber Conference. Moscow, p. 69 (in Russian). 364. Kuchersky, A.M., Fedyukina, L.P. and Bartenev, G.M. (1978) Prepr. Int. Rubber Conf. Kiev, B2: 34. 365. Kuchersky, A.M., Fedyukina, L.P. and Glezer, L.G. (1975) Kauchuk i Rezina, No 10: 46 (in Russian). 366. Kuchersky, A.M., Fedyukina, L.P. et al. (1974) Kauchuk i Rezina, No 4: 46 (in Russian). 367. Kuchersky, A.M., Glezer, L.G. and Kornienko, I.V. (1990) Kauchuk i Rezina, No 6: 12 (in Russian). 368. Kuchersky, A.M., Khodosh, T.S. and Vasileva, T.N. (1983) Kauchuk i Rezina, No 9: 11 (in Russian). 369. Kuchersky, A.M. Varaksin, M.E. and Glezer, L.G. (1987) Kauchuk i Rezina, No 11: 18 (in Russian). 370. Kuchersky, A.M., Varaksin, M.E., Radaeva, G.I. and Glezer, L.G. (1983) Kauchuk i Rezina, No 2: 12 (in Russian). 371. Kuleznev, V.N. (1980) Polymer Blends. Moscow: Khimiya Publishers, 303 pp. (in Russian). 372. Kuleznev, V.N. and Shershnev, V.A. (1988) The Chemistry and Physics of Polymers. Moscow: Vysshaya Shkola Publishers, 311 pp. (in Russian); (2007) 2nd edition. Moscow: KolosS Publishers, 367 pp. (in Russian). 373. Kuleznev, V.N. and Shershnev, V.A. (1990) The Chemistry and Physics of Polymers. Moscow: Mir Publishers, 335 pp. 374. Kulichikhin, V.G. and Malkin, A.Ya. (1980) In Malkin, A.Ya. and Papkov, S.P. (eds.) Orientational Phenomena in Polymer Solutions and Melts. Moscow: Khimiya Publishers, p. 91 (in Russian). 375. Kulichikhin, V.G., Dzyura, E.A., Malkin, A.Ya. and Vinogradov, G.V. (1970) Vysokomol. Soed., ser. A, 12(3): 568 (in Russian). 376. Kunizhev, B.I., Belyaev, O.F., Kuyumdzhi, E.S. and Zelenev, Yu.V. (1981) Vysokomol. Soed., ser. B, 23(3): 186 (in Russian). 377. Kurlyand, S.K. (ed.) (1986) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, 184 pp. (in Russian). 378. Kurlyand, S.K. and Marei, A.I. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: Khimiya Publishers, p. 48 (in Russian). 378a. Kurlyand, S.K., Gubanov, V.A., Volkova, M.A. et al. (2002) Abstr. 1st All-Russian Rubber Conference. Moscow, p. 124 (in Russian). 379. Kurlyand, S.K., Illarionova, G.I. and Marei, A.I. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 31 (in Russian). 380. Kurlyand, S.K., Marei, A.I. and Novikova, G.E. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: Khimiya Publishers, p. 51 (in Russian). 380a. Kurlyand, S.K., Marei, A.I., Novikova, G.E. and Chetvolinskaya, E.N. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: Khimiya Publishers, p. 54 (in Russian). 380b. Kurlyand, S.K., Marei, A.I. and Novikova, G.E. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: Khimiya Publishers, p. 57 (in Russian). 381. Kurlyand, S.K., Petrova, G.P. and Cherbunina, G.D. (1981) Kauchuk i Rezina, No 7: 57 (in Russian).
REFERENCES
169
Kurlyand, S.K., Petrova, G.P., Cherbunina, G.D. and Baranets, P,V. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 19 (in Russian). 383. Kurlyand, S.K., Petrova, G.P., Cherbunina, G.D. and Garmashov, E.P. (1982) Vysokomol. Soed., ser. B, 24(10): 774 (in Russian). 383a. Kurlyand, S.K., Petrova, G.P., Cherbunina, G.D. and Zhuravleva, I.P. (1986) Kauchuk i Rezina, No 5: 8 (in Russian). 384. Kurlyand, S.K., Petrova, G.P., Marei, A.I. and Novikova, G.E. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: Khimiya Publishers, p. 6 (in Russian). 385. Kurlyand, S.K., Shagov, V.S., Novikova, G.E., Marei, A.I. and Yakubchik, A.N. (1970). Vysokomol. Soed., ser. A, 12(11): 2472 (in Russian). 386. Kurlyand, S.K., Sidorovich, E.A. and Marei, A.I. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 40 (in Russian). 387. Kurlyand, S.K., Zhuravleva, I.P. and Marei, A.I. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 3 (in Russian). 388. Kusano, T., Kobayashi, K. and Murakami, K. (1979) Rubb. Chem. Technol., 52(4): 773. 389. Kuzminsky, A.S. and Sedov, V.V. (1979) Kauchuk i Rezina, No 12: 33 (in Russian). 390. Kuzminsky, A.S. and Sedov, V.V. (1984) Chemical Transformation of Elastomers. Moscow: Khimiya Publishers, 292 pp (in Russian). 391. Kuzminsky, A.S., Kavun, S.M. and Kirpichev, V.P. (1976) Physico-Chemical Basics of Production, Processing and Application of Elastomers. Moscow: Khimiya Publishers, 366 pp. (in Russian). 392. Kuznetsova, S.M., Priss, L.S. and Vorob’eva, V.A. (1982) Kauchuk i Rezina, No 2: 9 (in Russian). 393. Lakio, M. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, C7. 394. Lapshova, A.A., Ivanova, M.P. and Chechetkina, L.N. (1984) Prepr. Int. Rubber Conf. Moscow, p. 70. 395. Lauritzen, J.L. and Hoffman, J.D. (1973) J. Appl. Phys., 44(10): 4340. 396. Lazurkin, Yu.S. and Fogelson, R.L. (1951) Zh. Techn. Phys., 21(3): 267 (in Russian). 397. Lazurkin, Yu.S., Ushakov, V.S., Bartenev, G.M. and Voevodskaya, M.V. (1963) Zh. Techn. Phys., 33(6): 288 (in Russian). 398. Lee, T.K. and Porter, M. (1979) Proc. Int. Rubb. Conf., Venice, Italy. p. 991. 399. Lee, W.A. and Sewell, I.H. (1968) J. Appl. Polym. Sci., 12(6): 1397. 400. Leitner, M. (1955) Trans. Farad. Soc., 51: 1015. 401. Leonov, A. (2002) Int. J. Solids and Structures, 39: 5913. 402. Lepetov, V.A. and Yurtsev, L.N. (1987) Calculation and Design of Rubber Articles. Leningrad: Khimiya Publishers, 406 pp. (in Russian). 402a. Lepetov, V.A. and Yurtsev, L.N. (2005) Calculation and Design of Rubber Articles and Processing Equipment. Moscow: Istek Publishers, 420 pp. (in Russian). 403. Lepetov, V.A., Yurovsky, V.S., Komarnitsky-Kuznetsov, V.K. and Astaf’eva, V.D. (1972) Kauchuk i Rezina, No 4: 36 (in Russian). 404. Levin, V.Yu. and Zhdanov, A.A. (1984) Kompoz. Mater., No 22: 35 (in Russian). 405. Levin, V.Yu., Godovsky, Yu.K., Moskalenko, V.A., Andrianov, K.A., Slonimsky, G.L. and Zhdanov, A.A. (1972) J. Polym. Sci., A1, 10(1): 1. 406. Levin, V.Yu., Slonimsky, G.L., Andrianov, K.A., Zhdanov, A.A., Godovsky, Yu.K., Papkov, V.S. and Lubavskaya, E.A. (1973) Vysokomol. Soed., ser. A, 15: 224 (in Russian). 407. Levin, V.Yu., Zhdanov, A.A., Slonimsky, G.L. et al. (1982) Vysokomol. Soed., ser. A, 27(10): 2115 (in Russian). 408. Lipatov, Yu.S. (1973) J. Polym. Sci. Polym. Symp., 42: 855. 409. Lipatov, Yu.S. (1977) Physical Chemistry of Filled Polymers. Moscow: Khimiya Publishers, 303 pp. (in Russian); (1979) Shrewsbury: Rubber and Plastics Research Association of Great Britain (RAPRA), 227 pp. 410. Lipatov, Yu.S. (1978) Usp. Khimii, 67(2): 332 (in Russian). 411. Lipatov, Yu.S. and Privalko, V.P. (1972). Vysokomol. Soed., ser. A, 14(7): 1643; (1973) ser. B, 15(10): 749 (in Russian). 412. Lipatov, Yu.S., Kosyanchuk, L.F., Yarovaya, N.V. and Nesterov, A.V. (2003) Vysokomol. 382.
170
REFERENCES
Soed., ser. A, 45(3): 401. Lipatov, Yu.S., Shilov, V.V., Gomza, Yu.P. and Kruglyak, N.E. (1982) X-Ray Study of Polymer Systems. Kiev: Naukova Dumka, 296 pp. (in Russian). 414. Lipatov, Yu.S., Shilov, V.V., Gomza, Yu.P. et al. (1986) Vysokomol. Soed., ser. B, 28(6): 428 (in Russian). 415. Litvinov, V.M. (1986) Abstr. 6th All-Russian Conference on the Chemistry and Application of Organosilicon Substances. Riga, p. 309. (in Russian). 416. Litvinov, V.M. and Zhdanov, A.A. (1987). Vysokomol. Soed., ser. A, 29(5): 1021 (in Russian). 417. Litvinova, T.V. (1981) Plasticizers for Rubber Production. Moscow: TsNIITENeftekhim, 88 pp. (in Russian). 418. Litvinova, T.V., Galil-Ogly, F.A., Shevchenko, A.R. and Kur’yanova, I.S. (1979) Kauchuk i Rezina, No 1: 21 (in Russian). 419. Litvinova, T.V., Khodosh, T.S., Shevchenko, A.R. and Dontsov, A.A. (1981) Kauchuk i Rezina, No 8: 32 (in Russian). 420. Livigni, R.A. and Aggarwal, S.L. (1978) Prepr. Int. Rubber Conf. Kiev. A 1. 421. Lobkov, V.D., Novikova, G.E., Kolokoltseva, I.G. and Novinskaya, L.A. (1980) Kauchuk i Rezina, No 12: 5 (in Russian). 422. Lukin, B.V and Kasatochkin, V.I. (1950) In Evstratov, V.F. (ed.) Studies of the Physics and Chemistry of Rubber. Moscow: Goskhimizdat, p. 177 (in Russian). 423. Lynn, M.M. and Shool, R. (1987) Rubb. Plast. News, 16(27): 18. 424. Lysenko, E.E. (1940) Zh. Techn. Phys., 10: 1151; 1651 (in Russian). 425. Lysova, G.A. and Dontsov, A.A. (1991) Hydrogenated Butadiene Acrylonitrile Rubber: Properties, Application. Moscow: TsNIITENeftekhim, 58 pp. (in Russian). 426. Lysova, G.A., Ovsyannikova, M.A., Morozov, Yu.L. and Sigov, O.V. (2000) Kauchuk i Rezina, No 4: 16 (in Russian). 427. Ma, Y. Wu, Y.-P. Wang, Y.-Q. and Zhang, L.-Q. (2006) J. Appl. Polym. Sci., 99(3): 914. 428. Magill, J.H. and Joseph, H. (1986) Macromol. Chem., 187(2): 455. 429. Magill, J.H. and Li, H.M. (1978) Polymer, 19(11): 416. 430. Makhlis, F.A. and Fedyukin, D.L. (1989) Terminological Handbook of Rubber. Moscow: Khimiya Publishers, 400 pp. (in Russian). 431. Maklakov, A.I., Skirda, V.D. and Fatkullin, N.F. (1987) Self-diffusion in Polymer Solutions and Melts. Kazan: Izd. KazGU, 150 pp. (in Russian). 432. Malanichev, V.I., Chersky, I.V. et al. (1984) Kauchuk i Rezina, No 5: 24 (in Russian). 433. Malanichev, V.I. and Morova, L.Ya. (1985) Kauchuk i Rezina, No 3: 27 (in Russian). 433a. Malanichev, V.I. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, B10a. 434. Malinsky, Yu.M. (1970) Usp. Khimii, 34(8): 1511 (in Russian). 435. Mandelkern, L. (1964) Crystallization of Polymers. New York–San Francisco–Toronto– London: McGraw-Hill Book Company; Russian translation (1966) Moscow: Khimiya Publishers, 332 pp. 436. Mandelkern, L. (1967) J. Pol. Sci., C(18): 51. 436a. Mandelkern, L. (1985) Polym. J., 17(1): 337. 437. Mandelkern, L., Jain, N.L. and Kim, H. (1968) J. Pol. Sci., A2(6): 165. 437a. Mandelkern, L., Stack, G.M. and Mathieu, P.J.M. (1983) In Jonson, P.J. and Gill, P.S. (eds.) Analytical Calorimetry. New York: Plenum Press, vol. 5, p. 223. 438. Marei, A.I. (1960) Kauchuk i Rezina, No 2: 1 (in Russian). 439. Marei, A.I. (ed.) (1975) Physical Properties of Elastomers. Moscow: Khimiya Publishers, 136 pp. (in Russian); (1978) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 88 (in Russian). 440. Marei, A.I. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 31 (in Russian). 441. Marei, A.I. and Petrova, G.P. (1969) Kauchuk i Rezina, No 8: 4; (1970) No 5: 19 (in Russian). 442. Marei, A.I. and Petrova, G.P. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 70 (in Russian). 443. Marei, A.I. and Sidorovich, E.A. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 102 (in Russian). 444. Marei, A.I., Altshuler, M.Z. and Pankratova, A.M., (1975) In Marei, A.I. (ed.) Physical 413.
REFERENCES
171
Properties of Elastomers. Leningrad: Khimiya Publishers, p. 28 (in Russian). Marei, A.I., Garmonov, I.V., Novikova, G.E., Drozdov, V.A. and Nilva, S.Ya. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 12 (in Russian). 446. Marei, A.I., Grigoryants, I.K., Novikova, G.E., Petrova, G.P., Kurlyand, S.K., Rumyantsev, V.N. and Galkin, M.F. (1970) Mekhan. Polym., No 5: 780 (in Russian). 447. Marei, A.I., Kurlyand, S.K. and Novikova, G.E. (1969) Kauchuk i Rezina, No 5: 37 (in Russian). 448. Marei, A.I., Kurlyand, S.K. and Novikova, G.E. (1971) Vysokomol. Soed., ser. A, 13(3): 696 (in Russian). 449. Marei, A.I., Kurlyand, S.K., Baburin, B. G., Zhuravleva, I.P. and Nezabudkina, O.A. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 129 (in Russian). 450. Marei, A.I., Kurlyand, S.K., Sidorovich, E.A. and Novikova, G.E. (1981) Zhurn. VKhO im. Mendeleyeva, 26(3): 281 (in Russian). 451. Marei, A.I., Kurlyand, S.K., Zhuravleva, I.P., Nezabudkina, O.A. and Aref’eva, N.G. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 108 (in Russian). 452. Marei, A.I., Novikova, G.E. and Kuznetsov, N.P. (1964) Kauchuk i Rezina, No 10: 7; (1965) No 3: 41 (in Russian). 453. Marei, A.I., Petrova, G.P., Kurlyand, S.K. et al. (1970) Mekhan. Polym., No 5: 929 (in Russian). 454. Marei, A.I., Petrova, G.P. and Kurlyand, S.K. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 9 (in Russian). 455. Marei, A.I., Petrova, G.P. and Novikova, G.E. (1978) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 45 (in Russian). 456. Marei, A.I., Petrova, G.P. and Novikova, G.E. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 93 (in Russian). 457. Marei, A.I., Petrova, G.P., Novikova, G.E. and Kurlyand, S.K. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 77 (in Russian). 458. Marei, A.I., Petrova, G.P., Novikova, G.E. and Kurlyand, S.K. (1971) Proc. Int. Rubber Conf. 1969. Moscow: Khimiya Publishers, p. 153 (in Russian). 459. Marei, A.I., Tkachenko, T.G., Novikova, G.E. and Berestneva, Z.Ya. (1968) Dokl. Akad. Nauk SSSR, 179(5): 1114 (in Russian); (1969) Vysokomol. Soed., ser. A, 11(9): 1900 (in Russian). 460. Marei, A.I., Rokityansky, I.V. and Samoletova, V.V. (1959) Kauchuk i Rezina, No 2: 9 (in Russian). 461. Marei, A.I., Vasil’eva, M.A., Novikova, G.E., Petrova, G.P., Kurlyand, S.K. and Fedorov, Yu.N. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 100 (in Russian). 462. Marikhin, V.A., Bershtein, V.A., Egorov, V.M. and Myasnikova, L.P. (1986) Vysokomol. Soed., ser. A, 28(9): 1983 (in Russian). 463. Mark, J.E. (1994) Science and Technology of Rubber. London: Elsevier. 753 pp. 463a. Mark, J.E. (1979) Pol. Eng. Sci., 19(6): 403. 464. Mark, J.E., Ngai, K., Graessley, W., Mandelkern, L., Samulski, E., Koenig, J. and Wignall, G. (2004) Physical Properties of Polymers, 3rd edition. Cambridge University Press, 536 pp. 465. Martirosov, V.A., Levin, V.Yu., Zhdanov, A.A. and Slonimsky, G.L. (1981) Vysokomol. Soed., ser. A, 23(4): 896 (in Russian). 466. Martuscelli, E. (1984) Polym. Eng. Science, 24(1): 563. 467. Martynov, M.A. and Vylegzhanina, K.R. (1972) X-ray Imaging of Polymers. Leningrad: Khimiya Publishers, 93 pp. (in Russian). 468. Masagutova, L.V., Zlatkevich, L.Yu., Guseva, V.I. and Nikolsky, V.G. (1971) Vysokomol. Soed., ser. B, 13(2): 881 (in Russian). 469. Maskalyukaite, O.E., Morozov, Yu.L., Sukhinin, N.S. et al. (2006) Kauchuk i Rezina, No 3: 14; No 4: 9 (in Russian). 470. Mathot, V. B. F. (ed.) (1994) Calorimetry and Thermal Analysis of Polymers, Munich–Vienna–New York: Hanser Publishers, 368 pp. 445.
172
471. 472. 473. 474. 475. 476. 477. 478. 479. 480. 481. 482. 483. 484. 485. 486. 487. 488. 489. 490. 491. 492. 493. 494. 495. 496. 497. 498. 499. 500. 501. 502. 503. 504. 505. 506.
REFERENCES
Maurer, I.I. (1965) Rubb. Chem. Technol., 38(4): 979. Menson, J.A. and Sperling, L.H. (1976) Polymer Blends and Composites, 1st edition. 513 pp. Mezger, D.M. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, E5. Migulina, N.P., Epshtein, V.G. and Zakharov, N.D. (1969) Kauchuk i Rezina, No 12: 13 (in Russian). Milani, J. and Martini, E. (1986) Proc. Int. Rubb. Conf. Göteborg, Suppl. Prepr. Miller, R. and Nelson, L. (1960) J. Polym. Sci., 44: 391; (1961) 55: 1381. Mironyuk, V.L., Sidorovich, E.A., Afanasyev, I.D. and Broi-Karre, G.V. (1981) Kauchuk i Rezina, No 3: 8 (in Russian). Mitchell, J.S. and Windle, A.H. (1967) Polymer, 8(7): 1562. Mitlin, V.S. and Manevich, L.I. (1989) Vysokomol. Soed., ser. A, 31(5): 1120 (in Russian). Miyauchi, Y, Nishi, T., Yazaki, F. and Ohnaka, T. (2004) Proc. Symp. Int. Rubb. Conf. Beijing, China, v. C, p. 234. Mochel, W.E. and Maynard, J.T. (1954) J. Polym. Sci., 13: 235. Modji, J., Bondaretti, P., Turturro, A. and Kapazza, P. (1984) Prepr. Int. Rubb. Conf. Moscow. A35. Mohsin, M.A., Berry, J.P. and Treloar, L.R.G. (1985) Polymer, 26: 1463. Mokhnatkina, E.G., Volfson, S.I., Portnoi, Ts.B. and Ilyasov, R.S. (2004) Kauchuk i Rezina, No 2: 16, 19 (in Russian). Monroe, C. (1983) Automot. Eng., 8(1):40. Moore, D.F. (1975) The Friction of Pneumatic Tyres. Amsterdam–Oxford–New York: Elsevier Sci. Publishers, 220 pp. Moore, M. (1989) Eur. Rubb. J., No 1: 17. Morgan G.J. (1998) Prepr. 2nd Int. Conf. “Oilfield Engineering with Polymers”. London, p. 175; (2000) Kauchuk i Rezina, No 3: 35 (in Russian). Morgan, R.J., Nielsen, L.E. and Buchdahl, R. (1971) J. Appl. Phys., 42(11–12): 4653; (1972) Rubb. Chem. Technol., 45(4): 1033. Morris, R.E., Hollister, Y.W. and Barret, A.Y. (1950) Ind. Eng. Chem., 42: 1581. Morris, R.E., Hollister, Y.W. and Mallard, P.P. (1946) Rubb. Chem. Technol., 19: 151. Moskalenko, V.A. and Tsvankin, D.Ya. (1969) Vysokomol. Soed., ser. A, 1(12): 383 (in Russian). Moskalenko, V.A., Levin, V.Yu., Tsvankin, D.Ya. et al. (1971) Vysokomol. Soed., ser. A, 13(12): 2685 (in Russian). Moskalenko, V.A., Tsvankin, D.Ya. and Galil-Ogly, F.A. (1970) Vysokomol. Soed., ser. A, 12(3): 548 (in Russian). Mullins, L. (1950) J. Phys. Coll. Chem., 54: 239; (1959) J. Appl. Pol. Sci., 2: 1. Nagi, K. (1988) Kautschuk und Gummi Kunststoffe, 41(7): 717. Nakajima, N. and Harrell, E.R. (1983) Rubb. Chem. Technol., 56(4). 784. Nakamura, K., Katayama, Y. and Amano, T. (1973) J. Appl. Polym. Sci., 17(4): 1031. Nakamura, K., Watanabe, T., Katayama, Y. and Amano, T. (1972) J. Appl. Polym. Sci., 16(5): 1077. Nam, J., Shem, K.B., Kim, S.H. and Kim, B.C. (2004) Polym. J., 36(7): 450. Nefedova, E.V., Zorina, N.M. and Kuznetsova, E.I. (2002) Kauchuk i Rezina, No 6: 42 (in Russian). Nelson, K.V. and Kulkova, A.B. (1965) In Nelson, K.V. (ed.) Vibration Spectra and Molecular Processes in Rubbers. Moscow–Leningrad: Khimiya Publishers, p. 43 (in Russian). Nelson, K.V. and Novikova, N.N. (1965) In Nelson, K.V. (ed.) Vibration Spectra and Molecular Processes in Rubbers. Moscow–Leningrad: Khimiya Publishers, p. 5 (in Russian). Nelson, K.V., Khabibulaev, Kh. and Saidov, D. (1982) In Garmonov, I.V. (ed.) Phase-Aggregate State and Properties of Substances. Moscow: TsNIITENeftekhim, p. 94 (in Russian). Nelson, K.V., Kurlyand, S.K. and Berezkina, A.P. (1981) In Garmonov, I.V. (ed.) Advances in the Synthesis and Studies of Elastomer Structure. Moscow: TsNIITENeftekhim, p. 79 (in Russian). Nersesyan, D.A., Bagdasaryan, E.I. and Kabalyan, Yu.K. (1978) Kauchuk i Rezina,
REFERENCES
507. 508. 509. 510. 511. 512. 513. 514. 515. 516. 517. 518. 519. 520. 521. 522. 523. 524. 525. 526. 527. 528. 529. 530. 531.
173
No 11: 11 (in Russian). Nezabudkina, O.A.,Sokolova, S.G., Ventseslavskaya, K.K. and Gorin, Yu.A. (1982) In Garmonov, I.V. (ed.) Phase-Aggregate State and Properties of Substances. Moscow: TsNIITENeftekhim, p. 47 (in Russian). Nikitin, O.V., Kuchersky A.M. and Bukhina, M.F. (1979) Proc. Int. Rubber Conf. Venice, p. 471. Nikolaeva, Yu.K., Kuchersky, A.M., Zakharenko, N.V. and Dontsov, A.A. (1983) Proizvodstvo Shin, RTI i ATI, No 11: 3 (in Russian); (1984) Kauchuk i Rezina, No 10: 33 (in Russian). Nikolaeva,Yu.K., Zakharenko, N.K. and Kuchersky, A.M. (1987) Kauchuk i Rezina, No 8: 17 (in Russian). Novikov, N.A. and Shashkov, A.S. (1973) Vysokomol. Soed., ser. A, 15(5): 856; ser. B, 15(11): 8 (in Russian). Novikov, N.A., Shashkov, A.S. and Galil-Ogly, F.A. (1974) Kauchuk i Rezina, No 5: 19 (in Russian). Novikova, G.E. (1982) In Garmonov, I.V. (ed.) Phase-Aggregate State and Properties of Substances. Moscow: TsNIITENeftekhim, p. 111 (in Russian). Novikova, G.E., Korennaya, A.B., Baranets, I.V. and Gordeeva, S.B. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 122 (in Russian). Novikova, G.E., Lukina, N.E., Myuller, B.E. et al. (1976) In Apukhtina, N.P. (ed.) Synthesis and Properties of Urethane Elastomers. Leningrad: Khimiya Publishers, p. 64 (in Russian). Novikova, G.E., Marei, A.I. and Altshuler, M.Z. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: Khimiya Publishers, p. 15 (in Russian). Novikova, G.E., Marei, A.I. and Babitsky, B.D. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 52 (in Russian). Novikova, G.E., Marei, A.I. and Berkovich, M.A. (1969) Vysokomol. Soed., ser. B, 11: 502 (in Russian). Novikova, G.E., Marei, A.I. and Chetvolinskaya, E.N. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: Khimiya Publishers, p. 39 (in Russian). Novikova, G.E., Marei, A.I., Kholodnitskaya, G.V. and Berkovich, M.A. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 59 (in Russian). Novikova, G.E., Marei, A.I., Pankratova, E.D., Karlina I.A., Lobkov, V.D. and Kolokoltseva, I.G. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 134 (in Russian). Novikova, G.E., Marei, A.I., Pankratova, E.D., Kurlyand, S.K. and Petrova, G.P. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: Khimiya Publishers, p. 64 (in Russian). Novikova, G.E., Marei, A.I., Petrova, G.P. and Pankratova, E.D. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Moscow: Khimiya Publishers, p. 11 (in Russian). Novikova, G.E., Pankratova, E.D., Karlina, I.A. and Mongait, E.Z. (1984) Kauchuk i Rezina, No 2: 11 (in Russian). Novikova, G.E., Shesternina, N.V., Gorbenko, G.M. and Churlyaeva, L.A. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 62 (in Russian). Novikova, G.E., Smirnov, V.P., Marei, A.I. and Kovalev, N.F. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 18 (in Russian). Novikova, G.E., Smirnov, V.P., Marei, A.I., Kovalev, N.F. and Chetvolinskaya, E.N. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 36 (in Russian). Novitskaya, S.P., Nudelman, Z.N. and Dontsov, A.A. (1990) Fluoroelastomers. Moscow: Khimiya Publishers, 240 pp. (in Russian). Nudelman, Z.N. (2003) Kauchuk i Rezina, No 1: 8 (in Russian). Nudelman, Z.N. (2005) Kauchuk i Rezina, No 3: 33 (in Russian). Nudelman, Z.N. (2006) Kauchuk i Rezina, No 4: 27 (in Russian).
174
REFERENCES
531a. Nudelman, Z.N. (2007) Fluoroelastomers. Moscow: RIAS Publishers, 380 pp. (in Russian). 532. Odenhamn, R. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, E4. 533. Okui, N. (1990) Polymer, 31(1): 92. 534. Okui, N. (1990) J. Mater. Sci., 25: 1623. 535. Okui, N. (1992) Proc. NATO Advanced Research Workshop “Crystallization of Polymers”, University of Mons, Mons, Belgium, p. 22. 536. Okui, N. and Magill, J.H. (1977) Polymer, 18(4): 1152. 537. Oono, R., Miyasaka, K. and Ishikawa, K. (1973) J. Polym. Sci. Polym. Phys. Ed., 11: 1477. 538. Oth, J.FM. and Flory, P.J.J. (1958) J. Am. Chem. Soc., 80: 1297; (1958) Rubb. Chem. Technol., 31: 485. 539. Ovanesov, G.T., Baranov, V.G., Kabalyan, Yu.K. and Gasparyan, K.A. (1975) Izv. Akad. Nauk Arm. SSR, ser. fiz., 10(2): 128 (in Russian). 540. Ovanesov, G.T. et al. (1983) Vysokomol. Soed., ser. A, 25(8): 1708; (1986) 28(5): 1052 (in Russian). 541. Ozawa, T. (1971) Polymer, 12(3): 150. 542. Pai, P.C-H. and Meier, D.J. (1992) Rubb. Chem. Technol., 65(2): 396. 543. Pan, X.-D. and Kelley, E. D. (2003) Polym. Eng. Sci., 43(8): 1512. 544. Parizenberg, M.D. and Bukhina, M.F. (1979) Kauchuk i Rezina, No 10: 43 (in Russian). 545. Parizenberg, M.D., Bukhina, M.F. and Kuchersky, A.M. (1980) Proizvodstvo Shin, RTI i ATI, No 10: 29 (in Russian). 546. Paterson, M.S. (1964) J. Appl. Phys., 35(1): 176. 547. Patrikeyev, G.A. (1981) Dokl. Akad. Nauk SSSR, 256 (1): 129 (in Russian). 548. Payne, A.R. (1958) Plastics, 23(248): 182. 549. Payne, A.R. (1960) J. Appl. Polym. Sci., 4: 127. 550. Payne, A.R. and Scott, J.R. (1960) Engineering Design with Rubber. London: Maclaren, 256 pp. 551. Pechhold, W. (1991) Abstr. Int. Conference “Polymer Networks”. Moscow, L-16. 552. Pechkovskaya, K.A. (1968) Carbon Black as Rubber Reinforcement. Moscow: Khimiya Publishers, 214 pp. (in Russian). 553. Pennings, A.Y. and Zwijnenburg, A. (1979) J. Polym. Sci., A-2 17(6): 1011. 554. Perepechko, I.I. (1973) Acoustic Methods in Polymer Studies. Moscow: Khimiya Publishers, 296 pp. (in Russian). 555. Pestryakov, E.M., Skirda, V.D. and Storozhuk, I.P. (1981) Vysokomol. Soed., ser. A, 23(4): 874 (in Russian). 556. Petrova, G.P. and Kurlyand, S.K. (2002) Abstr. 1st All Russia Rubber Conference. Moscow, p. 66. 556a. Petrova, G.P. and Kurlyand, S.K. (2002) Abstr. 1st All-Russian Rubber Conference. Moscow, p. 75. 557. Petrova, G.P. and Marei, A.I. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 85 (in Russian). 558. Petrova, G.P., Berkovich, M.A., Nikolaeva, N.I. and Mongait, E.Z. (1982) In Garmonov, I.V. (ed.) Phase-Aggregate State and Properties of Substances. Moscow: TsNIITENeftekhim, p. 139 (in Russian). 559. Petrova, G.P., Nikandrova, N.E., Karlina, I.A. and Nikolaeva, N.I. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 13 (in Russian). 560. Petrova, G.P., Novikova, G.E. and Marei, A.I. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 90 (in Russian). 561. Petrova, G.P., Vasil’eva, M.A., Zhuravleva, I.P. and Shibaeva, A.N. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 68 (in Russian). 562. Petrova, M.A., Kogan, L.M., Monastyrskaya, N.B. and Smirnov, V.P. (1982) In PhaseAggregate State and Properties of Substances. Moscow: TsNIITENeftekhim, p. 88 (in Russian). 563. Petrova, N.N. (2005) Kauchuk i Rezina, No 6: 2 (in Russian). 564. Petrova, N.N. and Sokolova, M.D. (2003) Kauchuk i Rezina, No 5: 11 (in Russian).
REFERENCES
565.
175
Petrova, N.N., Portnyagina, V.V. and Fedorova, A.F. (2004) Abstr. Int. Rubber Conf. Moscow, p. 190. 566. Pettifor, J. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, B2. 567. Peyser, P. and Bascom, W.D. (1977) J. Macromol. Sci. Phys., 14(4): 597. 568. Phillips, P.J. (1979) J. Polym. Sci. Pol. Phys. Ed., 17(3): 409. 569. Phillips, P.J. (1990) Rep. Progr. Phys., 53: 549. 570. Phillips, P.J. and Edwards, B.C. (1975) J. Polym. Sci. Polym. Phys. Ed., 13(9): 1819. 571. Phillips, P.J. and Edwards, B.C. (1976) J. Polym. Sci. Polym. Phys. Ed., 14(2): 377. 572. Phillips, P.J. and Vatansever, N. (1987) Macromolecules, 20(9): 2138. 573. Pichugin, A.M. (2005) Kauchuk i Rezina, No 5: 32 (in Russian). 574. Pichugin, A.M., Stepanova, L.I. and Scherbakov, Yu.V. (2006) Kauchuk i Rezina, No 2: 13 (in Russian). 575. Pierre, L.E.S., Yim, A. and Chahal, R.S. (1971) Abstr. 8th JUPAC Microsymposium on Macromolecules “Morphology of Polymers”. Prague, A. 4. 576. Pokid’ko, B.P. and Tutorsky, I.A. (2004) Abstr. Int. Rubber Conf. Moscow, p. 195. 577. Pol, D. and Newman, S. (eds.) (1981) Polymer Blends, vol. 1. Moscow: Mir Publishers, 550 pp. (Russian translation). 578. Pol, D. and Newman, S. (eds.) (1981) Polymer Blends, vol. 2. Moscow: Mir Publishers, 543 pp. (Russian translation). 579. Polmanteer, K.E. and Thorne, J.A. (1966) Rubb. Chem. Technol., 39(5): 1403. 580. Ponomarenko, S.M., Privalko, E.G., Privalko, V. P., Schön, F. and Gronski, W. (2004) J. Macromol. Sci. Part B: Physics, 43(6): 1231. 581. Potapov, E.E. and Imadze, E.G. (1991) Kauchuk i Rezina, No 3: 27 (in Russian). 582. Priss, L.S. (1991) Abstr. Int. Conference “Polymer Networks”. Moscow, P-31. 583. Priss, L.S. (1993) Int. Polym. Sci. Techn., 20(3/4): 201. 584. Priss, L.S. (1996) Int. Polym. Sci. Techn. 23(7): T/53. 585. Priss, L.S., Tulenev, A.I. and Vishnyakov, I.I. (1982) Vysokomol. Soed., ser. A, 24(1): 51 (in Russian). 586. Privalko, V.P. (1986) Molecular Structure and Properties of Polymers. Leningrad: Khimiya Publishers, 238 pp. (in Russian). 587. Privalko, V.P. and Lipatov, Yu.S. (1974). Vysokomol. Soed., ser. A, 16(7): 1562 (in Russian). 588. Privalko, V.P., Novikov, V.P. and Yanovsky, V.V. (1991) The Basics of Thermo- and Rheophysics of Polymer Materials. Kiev: Naukova Dumka, 231 pp. (in Russian). 589. Privalko, V.P., Ponomarenko, S.M. and Privalko, E.G. (2004) Abstr. Int. Rubber Conf. Moscow, p. 199. 590. Pugh, T.L. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, B9. 591. Questad, D.L., Pae, K.D., Newman, B.A. and Scheinbeim, J.I. (1980) J. Appl. Phys., 51(7): 5100. 592. Rabinovich, Yu.V., Martynenko, L.Ya. and Ovchinnikov, Yu.V. (1970) Vysokomol. Soed., ser. B, 12(9): 696 (in Russian). 593. Ramella, D., Grosso, M., Kucherskii, A.M., Bukhina, M.F. and Varaksin, M.E. (1984) Prepr. Int. Rubber Conf. Moscow, 23. 594. Ramsh, A.S., Kurlyand, S.K., Volkova, M.A. and Gubanov, V.A. (2002) Abstr. 1st All-Russian Rubber Conference. Moscow, p. 73. 594a. Ramsh, A.S., Nikolaev, G.A. and Lobkov, V.D. (2002) Abstr. 1st All-Russian Rubber Conference. Moscow, p. 74. 595. Ranschaert, J. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, E1. 596. Rao, K.Y. and Rao, C.R. (1982) Mater. Res. Bull., 13(5): 1337. 597. Rault, J. and Robelin-Sourffache, E. (1989) J. Polym. Sci. B. Polym. Phys., 27: 1349. 598. Regel, V.R., Slutsker, A.I. and Tomashevsy, E.E. (1974) The Kinetic Nature of the Strength of Solids. Moscow: Nauka Publishers, pp. 430 (in Russian). 599. Rensh, G.J., Phillips, P.J. and Vatansever, N. (1986) J. Polym. Sci. B. Polym. Phys., 24: 1943. 600. Reznikovsky, M.M. and Lukomskaya, A.I. (1968) Mechanical Tests of Rubbers. 2nd edition. Moscow: Khimiya Publishers, 525 pp (in Russian).
176
REFERENCES
Richie, D.F. (1989) Rubb. World, 200(2): 27. Riva, F. and Forte, A. (1982) Rubbercon-82, Paris, p. 1. Riva, F., Forte, A. and Della Monica, C. (1980) Abs. 5th International Conference on Small Angle Scattering (SMAS). West Berlin, p. 113. 604. Robbins, R.F., Ohori, V. and Weitzel, D.H. (1963) Adv. Cryog. Eng., 8: 287. 605. Robelin-Sourffache, E. and Rault, J. (1989) Macromolecules, 22: 3581. 606. Roe, R.J. and Krigbaum, W.R. (1962) J. Polym. Sci., 61: 167; (1963) A1: 2049; (1965) Polymer, 6: 231. 607. Roeder, C.W., Stanton, J.F. and Feller, T. (1990) J. Cold Regions Engineering, 4(3): 113. 608. Roland, C.M. and Ngai, K.L. (1991) Macromolecules, 24: 5315. 609. Roland, C. M. Santangelo, P. G. Baram, Z. and Runt, J. (1994) Macromolecules, 27(19): 5382. 610. Roovers, J. and Toporowsky, P.M. (1990) Rubb. Chem. Techn., 63(5): 734. 611. Rostiashvili, V.G., Irzhak, V.I. and Rozenberg, B.A. (1987) Glass Transition of Polymers. Leningrad: Khimiya Publishers, 190 pp. (in Russian). 612. Russell, E.W. (1952) Rubb. Chem. Techn., 25(3): 397. 613. Sabsai, O.Yu., Barshtein, G.P., Fridman, M.L. and Godovsky, Yu.K. (1985). Vysokomol. Soed., ser. A, 27(8): 1697 (in Russian). 614. Sandberg, O. and Bäckström, G. (1979) J. Appl. Phys., 50(7): 4720. 615. Sanditov, D.S., Zydynov, Sh.B. and Bapnova, L.B. (2003) Vysokomol. Soed., ser. B, 45(7): 401. 616. Santangelo, P.G. and Roland, C.M. (2003) Rubb. Chem. Techn., 76(4) 892. 617. Sasuga, T. and Takehisa, M. (1977) J. Macromol. Sci. Phys. B, 13: 215. 618. Schenk, K.H. and Killian, H.G. (1991) Kautschuk und Gummi Kunststoffe, 44(7): 631. 619. Scholtens, B.J.R. (1984) Rubb. Chem. Techn., 57(3): 703; (1985) Brit. Polymer J., 17(2): 134. 620. Scholtens, B.J.R., Riande, E. and Mark, J.E. (1984) J. Polym. Sci., 22(7): 1223. 621. Schroeder, H. (1986) In Seymour, R.B. and Kirshenbaum, G.S. (eds.) Proc. Symp. “High Performance Polymers: Their Origin and Development”. New York: Elsevier Science Publishers, p. 389. 622. Schubnell, M. and Schawe, J.E.K. (2003) Proc. Int. Rubb. Conf. Prague, p. 471. 623. Schwarzl, F.R. (1980) Rheology, 1: 243. 624. Scott, J.R. (1968) Physical Testing of Rubbers. Moscow: Khimiya Publishers, 316 pp. (Russian translation). 625. Semenova, L.P., Klauzen, N.A. and Zaides, A.L. (1970) Kauchuk i Rezina, No 7: 263 (in Russian). 626. Severina, N.L. and Bukhina, M.F. (1983) Vysokomol. Soed., ser. A, 25(3): 518 (in Russian). 626a. Severina, N.L. and Bukhina, M.F. (1983) Vysokomol. Soed., ser. A, 25(11): 2397 (in Russian). 627. Severina, N.L. Bukhina, M.F. and Galperina (Zorina), N.M. (1983) Kauchuk i Rezina, No 10: 7 (in Russian). 628. Severina, N.L., Bukhina, M.F. and Galperina (Zorina), N.M.(1984). Kauchuk i Rezina, No 5: 8 (in Russian). 629. Severina, N.L. Bukhina, M.F. and Yurtseva, E.S. (1983). Vysokomol. Soed., ser. B, 25(8): 557 (in Russian). 630. Severina, N.L., Lyusov, Yu.N., Bukhina, M.F. and Dontsov, A.A. (1982) Proizvodstvo Shin, RTI i ATI, No 3: 3 (in Russian). 631. Severina, N.L., Tolmacheva, L.A., Bukhina, M.F. and Lysenkova, L.N. (1979) Proizvodstvo Shin, RTI i ATI, No 12: 3 (in Russian). 632. Severina, N.L., Veselov, N.M. and Bukhina, M.F. (1986) In Chersky, I.N. (ed.) Methods of Evaluation of Environment Resistance of Polymer Materials. Yakutsk: Yakutsky Branch of the Siberian Division of the USSR Acad. Sci. Publishers, p. 79 (in Russian). 633. Shanks, R.A. (1986) British Polymer J., 18(2): 72. 634. Sharaf, M.A. and Mark, J.E. (1980) Rubb. Chem. Technol., 53(4): 982. 634a. Sharaf, M. A., Mark, J.E. and Cesc, J.E.S. (2004) Polym. Eng. Sci., 26(4): 304. 635. Sharma, S.C., Mandelkern, L. and Stehling, F.C. (1972) J. Pol. Sci. Polym. Lett., B10(5): 345. 636. Sharples, A. (1966) Introduction to Polymer Crystallization. London: Edward Arnold Publishers; Russian translation (1968) Moscow: Mir Publishers, 200 pp.
601. 602. 603.
REFERENCES
637. 638. 639. 640. 641. 642. 643. 644. 645. 646. 647. 648. 649. 650. 651. 652. 653. 654. 655. 656. 657. 658. 659. 660. 661. 662. 663. 664. 665. 666. 667. 668. 669.
rd
177
Shazhin, B.I. et al. (1986) Electrical Properties of Polymers, 3 edition. Leningrad: Khimiya Publishers, 224 pp. (in Russian). Shelomentseva, I.V., Kostrykina, G.I., Listikova, I.V. and Korsakov, V.G. (1981). Kauchuk i Rezina, No 8: 14 (in Russian). Shen, M. C. and Eisenberg, A. (1970) Rubb. Chem. Technol., 43: 95-155; (1966) Progr. Solid State Chem., 3: 407. Shimomura, Y., Spruiell, I.E. and White, J.L. (1982) J. Appl. Polym. Sci., 27(9): 3553. Shpakov, V.P., Kuchersky, A.M., Titarenko, A.G. and Sazhina, O.S. (1981) Proizvodstvo Shin, RTI i ATI, No 10: 39 (in Russian). Shpindler, V.M., Bondareva, V.S., Sveshnikov, N.A. and Chulukina, A.V. (1991) Kauchuk i Rezina, No 10: 32 (in Russian). Shpindler, V.M., Severina, N.L., Sveshnikova, A.A. and Chulyukina, A.V. (1987) Promyshlennost’ SK, Shin i RTI, No 2: 25 (in Russian). Sidorovich, E.A. (1982) In Garmonov, I.V. (ed.) Phase-Aggregate State and Properties of Substances. Moscow: TsNIITENeftekhim, p. 102 (in Russian). Sidorovich, E.A. and Marei, A.V. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 97 (in Russian). Sidorovich, E.A. and Marei, A.V. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 114 (in Russian). Sidorovich, E.A. and Marei, A.V. (1975) In Marei, A.I. (ed.) Physical Properties of Elastomers. Leningrad: Khimiya Publishers, p. 126 (in Russian). Sidorovich, E.A., Fomicheva, M.M., Kotleva, N.S., Marei, A.V., Grinblat, M.P. and Dolgopolsky, I.M. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 149 (in Russian) Sidorovich, E.A., Kurlyand, S.K., Marei, A.V., Sokolova, S.G. and Gorin, Yu.A. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 117 (in Russian). Sidorovich, E.A., Marei, A.V., Bogdanov, V.V. and Metelkin, V.I. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 47 (in Russian). Sidorovich, E.A., Pavlov, G.N. and Marei, A.V. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 166 (in Russian). Sidorovich, E.A., Surkoa, N.S., Marei, A.V., Rappoport, L.Ya. and Trostyanskaya, I.I. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 158 (in Russian). Siesler, H.W. (1984) Colloid Polym. Sci., 262(3): 223. Simha, R. and Boyer, R.F. (1962) J. Chem. Phys., 37(5): 1003. Simha, R. and Carri, G. (1994) J. Polym. Sci. B: Polym. Phys., 32(16): 2645. Sleptsova, M.I. (2003) Kauchuk i Rezina, No 6: 18 (in Russian). Slonim, I.Ya. and Lyubimov, A.N. (1966) Nuclear Magnetic Resonance in Polymers. Moscow: Khimiya Publishers, 340 pp. (in Russian). Slonimsky, G.L., Levin, V.Yu. et al. (1966) Vysokomol. Soed., 8(11): 1936 (in Russian). Smirnova, L.V., Kropacheva, E.N., Sakhabutdinov, A.G. and Khamitov, I.I. (2004) Abstr. Int. Rubber Conf. Moscow, p. 222; (2005) Kauchuk i Rezina, No 1: 5 (in Russian). Smirnova, L.V., Tikhomirova, I.N., Kropacheva. E.N., and Zolotarev, V.L. (1996) Vysokomol. Soed., ser. A, 38(3): 458 (in Russian). Smith, K.Y. (1981) Polym. Prepr., 22(2): 154. Smith, T.L. (1978) Rubb. Chem. Technol., 51(2): 225. Smith, W.H. and Saylor, C.P. (1939) Rubb. Chem. Technol., 12(1): 18; (1934) J. Res. Nat. Bur. Stand., 13: 453; (1937) Science, 85: 204. Smith, W.H., Saylor, C.P. and Wing, H.Y. (1933) J. Res. Nat. Bur. Stand., 10: 479. Sobottka, J. and Donth, E. (1982) Plaste und Kautschuk, 29(11): 660. Sochava, I.V. and Tsereteli, G.I. (1971) Vysokomol. Soed., ser. A, 13(7): 1501 (in Russian). Sochava, I.V., Tsereteli, G.I. and Smirnova, O.I. (1972) Fiz. Tverd. Tela, 14: 553 (in Russian). Sogolova, T.M. (1965) Mekhan. Polim., No 1: 5 (in Russian). Soltmen, V. (ed.) (1981) Stereoregular Rubbers, vol. 1. Moscow: Mir Publishers, 320 pp.
178
REFERENCES
(Russian translation). Soltmen, V. (ed.) (1981) Stereoregular Rubbers, vol. 2. Moscow: Mir Publishers, 508 pp. (Russian translation). 671. Song, N.S. and Zhu, Q.B. (1992) Abstr. Int. Conference “Orientational Phenomena in Polymers’. St. Petersburg, p. 47. 672. Soos, M. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, C9. 673. Sorokin, G.A. (1971) Vysokomol. Soed., ser. A, 13(11): 2577 (in Russian). 674. Sorokin, G.A., Varfolomeeva, E.A., Kutsenko, A,I., and Merzlikina, V.P. (1979) Kauchuk i Rezina, No 7: 17 (in Russian). 675. Sotnikova, E.N. and Apukhtina, N.P. (1981) Zhurn. VKhO im. Mendeleyeva, 24(3): 310 (in Russian). 676. Spetz, G. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, B1. 677. Spetz, G. (2005) Book of Papers Int. Rubber Conf. Maastricht, p. 479. 678. Standard ISO 3387: Rubber. Determination of Crystallization Effects by Hardness Measurements. 679. Standard ISO 6471: Rubber Vulcanized – Determination of Crystallization under Compression. 680. Stannarius, R., Aksenov, V., Bläsing, J., Krost, A., Rössle, M. and Zentel, R. (2006) Phys. Chem. Chem. Phys., 8: 2293. 681. Stannarius, R., Khler, R., Rössle, M. and Zentel, R. (2004) Liquid Cryst., 31: 895. 682. Starkweather, H.W. (1983) Amer. Chem. Soc. Polym. Prepr., 24(1): 116; (1982) Macromolecules, 15(6): 1600. 683. Starkweather, H.W. and Howard, W. (1980) Macromolecules, 13(4): 892-897. 684. Stevenson A. (ed.) (1984) Rubber in Offshore Engineering. Bristol: Adam Hilger Ltd., 346 pp. 685. Stevenson, A. (1984) Kautschuk und Gummi Kunststoffe, 37(2): 105. 686. Stevenson, A. (1986) Polymer, 27(8): 1211–1218; (1983) J. Pol. Sci. Polym. Phys. Ed., 21: 553. 687. Stevenson, A. (2000) Abstr. Int. Rubb. Conf. Helsinki, p. 112. 688. Stevenson, A. and Maxwell, B.L. (1986) Proc. Int. Rubber Conf. Göteborg, No 2(5). 689. Stevenson, A. and Price, A.R. (1986) Proc. 2nd World Congress on Joints and Bearings. Antonio, Texas, SP 94-6, p. 113. 690. Stickney, P.B. and Müller, W.I. (1956) Rubber World, 134(2): 234. 691. Stober, R. (2004) Abstr. Int. Rubber Conf. Moscow, p. 303; (2005) Kauchuk i Rezina, No 3: 2. 692. Stranaghan, D. (1984) Prepr. Int. Rubb. Conf. Moscow, B58. 693. Stromrall, H. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere, Finland, B3. 694. Su, T.-K. and Mark, J.E. (1978) Rubb. Chem. Technol., 51(2): 285. 695. Suzuki, A., Oikawa, H. and Murakami, K. (1985) Polymer, 26: 247-252. 696. Tager, A.A., Adamova, L.V., Blinov, V.S. et al. (1987). Vysokomol. Soed., ser. A, 29(11): 2327 (in Russian). 697. Tager, A.A., Suvorova, A.I., Ikanina, Y.V. et al. (1984) Vysokomol. Soed., ser. B, 26(5): 323 (in Russian). 698. Talroze, R.V., Gubina, T.I., Shibaev, V.P., Plate, N.A., Dakin, V.I., Shmakova, N.A. and Sukhov, F.F. (1990) Macromol. Chem. Rapid Commun., 11(2): 67. 699. Tamarkin, V.F., Bukhina, M.F., Derbeneva, A.P., Solovova, T.A. and Kuzminsky, A.S. (1970) Kauchuk i Rezina, No 8: 10 (in Russian). 700. Tamarkin, V.F., Chubarova, G.V., Bukhina, M.F. and Kuzminsky, A.S. (1971) Dokl. Akad. Nauk SSSR, 200(1): 156 (in Russian). 701. Tanaka, J. (1985) Proc. Int. Rubb. Conf. Kyoto, Japan, p. 141. 702. Teitelbaum, B.Ya. and Anoshina, N.G. (1965) Vysokomol. Soed., 7(6): 978 (in Russian). 702a. Teitelbaum, B.Ya. and Anoshina, N.G. (1965) Vysokomol. Soed., 7(7): 1188 (in Russian). 702b. Teitelbaum, B.Ya. and Anoshina, N.G. (1965) Vysokomol. Soed., 7(12): 2112 (in Russian). 703. Teitelbaum, B.Ya. and Anoshina, N.G. (1968) Vysokomol. Soed., ser. B, 10(1): 26 670.
REFERENCES
179
(in Russian). Teitelbaum, B.Ya., Dianov, M.P., Beregovskaya, M.G. and Yagfarova, T.L. (1962) Kauchuk i Rezina, No 8: 3 (in Russian). 704a. Teitelbaum, B.Ya., Gubanov, E.F. and Naumov, V.A. (1962) Dokl. Akad. Nauk SSSR, 145(5): 1077 (in Russian). 705. Terent’eva, N.N., Vasileva, S.Yu., Kolyamshin, O.A. and Koltsov, N.I. (2006) Kauchuk i Rezina, No 2: 16 (in Russian). 706. Thomas, A.G. and Whittle, J.M. (1970) Rubb. Chem. Technol., 43(2): 222. 707. Titarenko, S.A., Bukhina, M.F. and Klyuchnikova, L.F. (1981) Proizvodstvo Shin, RTI i ATI, No 3: 2 (in Russian). 708. Titarenko, S.A., Bukhina, M.F., Klyuchnikova, L.F. and Kozlova, T.L. (1980) Kauchuk i Rezina, No 5: 18 (in Russian). 709. Titarenko, S.A., Klyuchnikova, L.F., and Bukhina, M.F. (1975) Proizvodstvo Shin, RTI i ATI, No 5: 9 (in Russian). 710. Titarenko, S.A., Klyuchnikova, L.F. and Bukhina, M.F. (1986) Abstr 1st All-Russian Conf. Polymer Blends. Ivanovo: p. 42 (in Russian). 711. Toki, S., Sics, I., Ran, S., Liu, L., Hsiao, B. and Murakami, S. (2004) Rubb. Chem. Technol., 77(2): 317. 712. Tosaka, M., Kohjiya, S., Murakami, S., Ikeda, Y., Toki, S., Sics, I. and Hsiao, B. (2004) Rubb. Chem. Technol., 77(4): 317. 713. Trapeznikova, O.N., Novikova, G.E. and Minakova, S.V. (1961) Optika i Spektroskopiya, 11: 353. 714. Trapeznikova, O.N. and Smirnova (Novikova), G.E. (1957) Optika i Spektroskopiya, 3: 631 (in Russian). 715. Treloar, L.R.G. (1949) The Physics of Rubber Elasticity, 1st edition; Russian translation (1953) Moscow: Inostr. Lit. Publishers, 240 pp.; (1975) 3rd edition. Oxford: Clarendon Press, 310 pp.; (2005) Oxford: Oxford University Press, 310 pp. 716. Trick, G.S. (1960) J. Appl. Polym. Sci., 3(8): 253. 717. Tsaikova, S. (1984) Prep. Int. Rubber Conf. Moscow, A11. 718. Tsaikova, S. and Mladenov, I. (1978) Prepr. Int. Rubber Conf. Kiev, A12. 719. Tsereteli, G.I. and Sochava, I.V. (1971) Vysokomol. Soed., ser. A, 13(8): 2612 (in Russian). 720. Tsvankin, D.Ya. (1969) Kristallografiya, 14:431 (in Russian). 721. Tutorsky, I.A. and Pokid’ko, B.P. (2004) Abstr. Int. Rubber Conf. Moscow, p. 234. 722. Tutorsky, I.A. and Pokid’ko, B.P. (2004) Kauchuk i Rezina, No 5: 23; No 6: 33 (in Russian). 723. Tutorsky, I.A., Dontsov, A.A., Filipovich, B.V. and Orlov, V.A. (1979) Kauchuk i Rezina, No 5: 26 (in Russian). 724. Udagava, Y. (1984) Prepr. Int. Rubb. Conf. Moscow, A 95. 725. Van Cleef, A. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”.Tampere, Finland, F3. 726. Vasilev, V.A., Kormer, V.A., Bubnova, S.V. and Dolinskaya, E.V. (1986) Kauchuk i Rezina, No 1: 5 (in Russian). 727. Vasileva, M.A., Fedorov, Yu.N. and Nikandrova, N.E. (1982) In Garmonov, I.V. (ed.) Phase-Aggregate State and Properties of Substances. Moscow: TsNIITENeftekhim, p. 148 (in Russian). 727. Vasileva, M.A., Pankratova, A.M. and Mikheev, A.I. (1986) In Kurlyand, S.K. (ed.) Structure and Physical Properties of Elastomers. Moscow: TsNIITENeftekhim, p. 39 (in Russian). 728. Veselov, V.M. and Bukhina, M.F. (1986) Abstr 1st All-Russian Conf. Polymer Blends. Ivanovo: p. 171 (in Russian). 729. Vinnitsky, L.E., Evseeva, L.G. and Charkin, L.G. (1980) Kauchuk i Rezina, No 10: 43 (in Russian). 730. Vinogradov, G.V. and Malkin, A.Ya. (1977) Rheology of Polymers. Moscow: Khimiya Publishers, 437 pp. (in Russian); (1980) Berlin: Springer-Verlag; (1980): Moscow, Mir Publishers, 467 pp. 731. Vinogradov, G.V., Govorova, O.A., Frenkin, E.I. and Podol’sky, Yu.Ya. (1983) Kauchuk i Rezina, No 9: 8 (in Russian). 732. Vishnitskaya, L.A., Kuchersky, A.M., Sokolskaya, V.D. and Perelygina, V.P. (1970) Kauchuk i Rezina, No 6: 31 (in Russian). 704.
180
733. 734.
REFERENCES
Vishnyakov, I.I. and Tulenev, A.I. (1984) Prepr. Int. Rubber Conf. Moscow, A39. Volgin, V.A., Tutorsky, I.A., Lazorev, Yu.A. and Polivanov, O.F. (1983) Kauchuk i Rezina, No 6: 3 (in Russian). 735. Walters, M.H. (1963) J. Polymer Sci., A1, 1: 3091. 736. Warnaka, G.E. (1963) Rubb. Chem. Technol., 36: 407. 737. Wasserman, A.M. and Kovarsky, A.L. (1986) Spin Labels and Probes in Polymer Physico-chemistry. Moscow: Nauka Publishers, 241 pp. (in Russian). 738. Westlinning, M., Butenuth, G. and Leinweber, S. (1961) Makromol. Chem., 50: 253. 739. White, J.L. (1985) J. Polym. Eng., 5(3): 235. 740. Whittington, W.H. (1981) Rubber World, 183(6): 33. 741. Widmaier, J.M. and Meyer, G.C. (1980) Macromolecules, 14: 450; (1981) Rubb. Chem. Technol., 54(5): 940. 742. Wolf, W. (1965) Paper at International Exhibition “Chemistry”, Moscow. 743. Wong, W.K. (2004) Abstr. Int. Rubber Conf. Moscow, p. 306. 744. Wood, L.A. (1939) Rubb. Chem, Techn., 12(2): 130. 745. Wood, L.A. (1958) J. Polym. Sci., 28(117): 319; (1964) J. Polym. Sci. B, 1(7): 703. 746. Wood, L A. (1976) Rubb. Chem, Techn., 49(2): 189. 747. Wood, L.A. (1981) Rubb. Chem, Techn., 54(2): 331. 748. Wood L.A. and Bekkedahl, N.J. (1945) Rubb. Chem. Techn., 19: 1145. 749. Wood, L.A. and Bekkedahl, N.J. (1968) Rubb. Chem. Technol., 41(3): 564. 750. Wood, L.A., Bekkedahl, N.J. and Gibson, R.E.J. (1945) J. Chem. Phys., 13: 475; (1945) J. Res. Nat. Bur. Stand., 35: 375. 751. Wunderlich, B. (1973) Macromolecular Physics, vol. 1: Crystal Structure, Morphology and Defects, New York: Academic Press.568 pp.; Russian translation (1976) Moscow: Mir Publishers, 623 pp. 751a. Wunderlich, B. (1976) Macromolecular Physics, vol. 2: Nucleation, Crystallization, Annealing New York: Academic Press, 476 pp.; Russian translation (1979) Moscow: Mir Publishers, 573 pp. 751b. Wunderlich, B. (1980) Macromolecular Physics, vol. 3: Crystal Melting, New York: Academic Press, 363 pp.; Russian translation (1984) Moscow: Mir Publishers, 484 pp. 751c. Wunderlich, B. (1993) Abstr. 28th Europhysics Conference on Macromolecular Physics “Transitions in Oligomer and Polymer Systems”. Ulm, Germany, G-8. 752. Wunderlich, B. (2005) Thermal Analysis of Polymeric Materials. Berlin: Springer-Verlag, 894 pp. 753. Wunderlich, B. and Baur, H. (1970) Heat Capacities of Linear High Polymers. Berlin–Heidelberg–New York: Springer-Verlag, 320 pp. 754. Yagfarov, M.Sh. (1968) Vysokomol. Soed., ser. A, 10(6): 1264; (1979) Vysokomol. Soed., ser. A, 21(4): 888; (10): 2379; (1980) Vysokomol. Soed., ser. A, 22(11): 2609 (in Russian). 754a. Yagfarov, M.Sh. (1982) Vysokomol. Soed., ser. A, 24(12): 2537; (1984) Vysokomol. Soed., ser. A, 26(3): 551; (1988) Vysokomol. Soed., ser. A, 30(1): 79 (in Russian). 754b. Yagfarov, M.Sh. and Mitrofanova, E.P. (1986) Vysokomol. Soed., ser. A, 28(5): 908 (in Russian). 755. Yamada, K. and Fuvayama, Y. (1990) Rubb. Chem. Techn., 63(5): 669. 756. Yang, X.-Z. and Sun, J. (2002) J Polym Sci. B: Polym Phys., 40(23): 2646. 757. Yanovsky, Yu.G., Gamlitsky, Yu.A. and Filipenkov, P.A. (2004) Abstr. Int. Rubber Conf. Moscow, p. 269. 757a. Yanovsky, Yu.G., Gamlitsky, Yu.A., Molchanov, S.P. and Teplukhin, A. (2004) Abstr. Int. Rubber Conf. Moscow, p. 270. 757b. Yanovsky, Yu.G., Gamlitsky, Yu.A., Zgaevsky, V. E. and Bass, Yu.P. (2002) Kauchuk i Rezina, No 5: 20 (in Russian). 758. Yeh, G.C.J. (1976) Proc. 5th Europh. Conf. Macrom. Budapest, p. 86. 758a. Yeh, G.C.J. and Nong, K.L. (1979) Polym. Eng. Sci., 19(6): 401. 759. Yim, A. and Pierre, L.E. (1970) J. Polym. Sci. B, 8: 241. 760. Yurovsky, V.S., Sirota, R.M. and Igonin, A.S. (1973) Kauchuk i Rezina, No 5: 37 (in Russian). 760a. Yurtsev, N.N. and Kuchersky, A.M. (1973) Kauchuk i Rezina, No 10: 36 (in Russian). 761. Zaboristov, V.N. and Ryakhovsky, V.S. (2004) Abstr. Int. Rubber Conf. Moscow, p. 101.
REFERENCES
762. 763. 764. 765. 766. 767. 768. 769. 770. 771. 772. 773. 774. 775. 776. 777. 778. 779. 780. 781. 782. 783. 784. 785. 786. 787.
181
Zakharov, N.D. (1978) Chloroprene Rubbers. Moscow: Khimiya Publishers, 271 pp. (in Russian). Zgaevsky, V. E. and Yanovsky, Yu.G. (2003) Dokl. Akad. Nauk SSSR, 392(5): 627 (in Russian). Zhang, Z.-M. and Mark, J.E. (1982) J. Polym. Sci. Phys. Ed., 20(3) 473. Zhuravleva, I.P., Kurlyand, S.K. and Marei, A.I. (1978) In Marei, A.I. (ed.) Physical and Mechanical Properties of New Elastomers. Moscow: TsNIITENeftekhim, p. 124 (in Russian). Zhurkov, S.N., Sanfirova, T.P. and Tomashevsky, E.E. (1962) Vysokomol. Soed., 4:196 (in Russian). Ziabicki, A. (1967) Polymery zworygwa wielkocrasecz kowe, 12(9): 405. Ziabicki, A. (1979) Abstr. 9th Europhys. Conf. Macromolec. Phys. Jablonna, p. 199. Ziabicki, A. and Jareckii, L. (1976) Abstr. 5th Europhys. Conf. Macromolecules, Budapest, p. 166. Ziegel, K.D. and Romanov, A. (1973) J. Appl. Polym. Sci., 17: 1119. Zlatkevich, L.Yu. and Nikolsky, V.G. (1971) Vysokomol. Soed., ser. B, 13(11): 810 (in Russian). Zorina, N.M. (2006) Abstr. 23rd Symp. Rheology, Valdai, Russia, p. 56 (in Russian). Zorina, N.M. and Bukhina, M.F. (1984) Vysokomol. Soed., ser. A, 26(9): 1855 (in Russian). Zorina, N.M. and Bukhina, M.F. (2002) Abstr. 1st All-Russian Rubber Conference. Moscow, p. 67. Zorina, N.M. and Bukhina, M.F. (2002) Kauchuk i Rezina, No 6: 3 (in Russian). Zorina, N.M., Bukhina, M.F. and Morozov, Yu.L. (2004) Abstr. Int. Rubber Conf. Moscow, p. 108. Zorina, N.M., Bukhina, M.F. and Morozov, Yu.L. (2006) Abstr. 23rd Symp. Rheology, Valdai, Russia, p. 57 (in Russian). Zorina, N.M., Bukhina, M.F., Morozov, Yu.L. and Kuznetsova, E.I. (2004) Kauchuk i Rezina, No 6: 37 (in Russian). Zorina, N.M., Rudenko, G.A., Voloshin, V.N. and Bukanova, N.N. (1988) Kauchuk i Rezina, No 3: 10 (in Russian). Zuev, Yu.S. (1980) Fracture of Elastomers under Conditions Typical for Service. Moscow: Khimiya Publishers, 288 pp. (in Russian). Zuev, Yu.S. and Degteva, T.G. (1986) Resistance of Elastomers under Service Conditions. Moscow: Khimiya Publishers, 268 pp. (in Russian). Zuev, Yu.S. and Stern, T.M. (1982) Kauchuk i Rezina, No 7: 34. (in Russian). Zuev, Yu.S., Bukhina, M.F. and Astaf’ev, A.N. (1983) Mechan. Komposit. Mater., 2: 359 (in Russian). Zuev, Yu.S., Bukhina, M.F. and Borschevskaya, A.Z. (1979) In Structure and Properties of Polymer Materials. Riga: Zinatne, p. 111 (in Russian). Zuev, Yu.S., Revazova, M.A. and Bukhina, M.F. (1969) Vysokomol. Soed., ser. B, 11(12): 898 (in Russian). Zuev, Yu.S., Revazova, M.A. and Bukhina, M.F. (1971) Proc. Int. Rubb. Conf. 1969. Moscow: Khimiya Publishers, p. 146 (in Russian). Zuev, Yu.S., Revazova, M.A. and Bukhina, M.F. (1972) Mechan. Polym., 2: 222 (in Russian).
182
REFERENCES
Subject Index Activation energy, 19, 21, 69, 95, 118, 147 Activity of filler, 45, 51, 93, 121 Adhesive, 84, 101, 103, 115, 121, 145 Aerosil, 89, 149 Alternating rubbers, 155, 156 Amorphous halo, 26 Amorphous phase, 61–64, 68, 74, 75, 93, 100, 127, 145 Anisotropy, 62 Annealing, 25, 126 Articles, see Components, see Elastomeric articles, see Rubber goods, 1, 2, 32, 46, 52, 60, 72, 74, 75, 78, 81, 94, 103, 115, 122, 129–135, 140 Axial fibrillae, 65, 80, 101–103, 106–108 Belts, 132, 134, 135 Bend, 15 Birefringence, 61, 81, 99, 107, 119 Block copolymers, 65, 72, 94 Breakdown, 47, 51, 57, 91, 109, 116–118, 121, 122, 130, 134 Bridge bearings, 130 Brittleness, 57, 115, 116, 123, 127, 151, 152 Brittleness temperature, 57, 115, 116, 123, 127, 151, 152 Butadiene acrylonitrile rubber (BNR), 24, 33–35, 139, 140, 145, 156 Butyl rubber (BR), 45, 59, 97, 99, 100, 104, 108, 119–121, 129, 138, 146, 147, 150 Calorimetric method, 12, 74, 76 Catalyst, 140, 142, 143, 153–155 Chalk, 45, 46, 56, 114 Characteristic straight lines, 81, 88
Characteristic time of crystallization, 67 Chemical ageing, 3, 4, 64 Chemical crosslinks/junctions, 7, 50, 51 Cis-trans isomerization, 144 Clapeyron–Clausius equation, 58, 82, 105, Clay, 45, 121 Coefficient of thermal expansion, 12, 18, 21, 24, 25, 33, 37, 39, 40, 44, 54, 127, 130, 134 Coefficient of volume thermal expansion, 24 Compatibility, 34–37 Complex dynamic modulus, 8, 93, 107 Components, see Articles, see Elastomeric articles, see Rubber goods, 1, 3, 8, 16, 22, 26, 32, 34–37, 44, 74, 80, 85, 89, 94, 109, 122, 130, 131, 132, 134, 135, 149, 150 Composition of rubber compounds, 28, 83, 108 Compression, 6, 8, 15, 16, 26, 47–49, 55, 57, 63, 80, 81, 93, 94, 120, 130, 132, 146 Conformation of macromolecular chains, 3, 82, 84 Conveyor belts, 132, 134, 135 Cooling rate, 11, 12, 18, 21, 25, 91, 134 Copolymer, 1, 14, 30, 33, 34, 49, 65, 72, 73, 79, 81, 84, 87, 91–94, 97, 111, 126, 137, 139, 143, 144, 146–155 Covulcanization, 140 Crosslinks, 39, 44, 45, 50, 51, 85–88, 94, 104, 111, 141 Crystal growth rate, 77, 83, 155 Crystallizable plasticizers, 38, 91 Crystallization, 1, 7, 11, 26, 31, 34, 38, 40,
184
SUBJECT INDEX
41, 44, 47, 51, 52, 53, 58–156 Crystallization half-time, 66, 81, 84, 86, 148, 150 Crystallization kinetics, 63, 74, 75, 81, 82, 88, 94, 97, 98, 141, 155 Crystallization prediction, 94 Crystallization rate, 11, 38, 59, 64, 66–79, 83–87, 89, 91, 94, 95, 109, 119, 125, 129, 141–156 Crystallization rate constant, 67 Crystallization temperature, 67, 75–77, 85, 88, 89, 93–95, 110, 143, 145, 147, 153 Crystals with extended chains, 64, 65, 81, 82, 101, 106–108, 117–121 Crystals with folded chains, 65, 82, 102, 108, 119, 121 Crystals with “unfolded” chains, 108, 115–117, 121 Cyclopentene, 154 Deceleration of crystallization, 87, 90 Decrystallization, see Steaming, 75, 78, 133 141, 145 Deformation components, 3, 16 Deformation properties, 7, 46, 115, 116, 118 Deformation rate, 23, 116, 120 Degree of crystallinity, see Degree of crystallization, 60, 61, 66, 83, 92, 123, 148 Degree of crystallization, see Degree of crystallinity, 11, 31, 60, 62, 63, 67, 84, 98, 99, 103, 104, 110, 113, 122, 125–127, 133, 141, 144, 145, 147, 150, 154, 155 Density of amorphous polymer, 102 Dicumyl peroxide, 7 Dielectric loss, 13, 27, 39, 57, 62 Differential scanning calorimetry, 12, 62, 92 Differential thermal analysis, 12, 62 Dilatometry method, 18, 31, 35, 39, 41, 75, 77 Dimethyl siloxane rubber, 30 Disulphide bonds, 86 Double peak of melting, 76, 113, 114 Dynamic mechanical characteristics, 50, 51
Dynamic modulus, 8, 17, 39, 43, 51, 93, 98, 107 Ehrenfest equation, 26 Elastic deformation, 3, 4, 8, 43, 48, 116 Elasticity modulus, 3, 55, 125 Elastomeric articles, see Articles, see Components, see Rubber goods, 1, 2, 32, 43, 75 Electrical conductivity, 62, 63 Electrical properties, 12 Electron microscopy, 61, 69, 70, 83, 87, 88, 101, 141 Electron paramagnetic resonance (EPR), 14, 62 Elementary cell, 64, 100, 141, 155 Elongation, 4, 16, 47, 55, 57, 80, 81, 98, 102, 104, 107, 117, 119, 120, 122, 150 Energy of interface formation, 69 Entanglements, 50 Enthalpy, 63, 64 Entropy, 3, 63, 64, 74, 105 Equilibrium melting temperature, 58, 63, 64, 70, 72, 77–79, 82, 90, 92, 94, 95, 100, 106, 108, 133, 141, 143, 145, 147, 149, 153, 154 Ethylene-propylene rubber, 35, 51, 91–94, 97, 120, 152, 153 Filler, 25, 36, 38–40, 44–46, 50, 51, 67, 73, 85, 87–90, 93, 95, 110, 112–114, 118, 119, 121, 122, 125, 134, 139, 140, 145–147, 149, 150, 153, 155 First-order phase transitions, 28, 58 Flory equation, 78, 87, 99, 104, 107, 143 Flory–Higgins constant, 90 Fluctuation, 7, 27, 50, 51, 68, 89, 90, 100, 115, 128 Fluoroelastomer, 24, 94, 151 Folds, 65, 72, 80 Forced elasticity, 116, 127, 128, 130, 151 Form factor, 47, 49, 55, 57, 82, 118, 129–131, 151 Fox–Flory equation, 28 Free volume, 21–26, 33, 36–40, 54, 55 Gaskets, 130, 132 Glass transition, 1, 11–44, 51, 52, 55–63,
SUBJECT INDEX
68, 70, 82, 91, 94, 115, 116, 125–134, 137, 139, 140, 142, 144–156 Glass transition, conventional, 35 Glass transition, “chemical”, 32 Glass transition temperature, 11, 13–16, 18–22, 26–44, 51, 55–68, 82, 91, 94, 115, 116, 126, 128–133, 144, 147–156 Goods, see Articles, see Components, see Elastomeric articles, see Rubber goods, 1, 43, 48, 61, 135 Grains, 65, 150 Gutta percha, see polyisoprene-1,4-trans 64, 77, 81, 83, 103, 140 Hardness, 16, 52, 63, 93, 125 Heat of fusion, 69, 78, 79, 92, 141, 142 Heating rate, 12, 14, 17, 18, 21, 39, 76, 77, 79, 141 Heterogeneous nucleation, 69 Homopolymer, 29, 30, 33, 34, 78, 79, 84, 94, 140, 142, 144, 147, 148, 151, 152, 154, 155 Hoses, 122, 132, 134 Hysteresis, 8, 18, 50, 75, 119, 134 Identity period, 156 Impact resilience, 17, 30, 50, 63, 127 Incompatibility, 35, 44, 50, 84, 109 Induction period, 65, 66, 69, 80, 87, 93, 97, 112, 146, 147 Internal energy, 3 IR spectroscopic method, 61, 111 Isothermal crystallization, 62, 64–67, 73, 75, 76, 86, 97, 100, 110, 112 Kargin–Malinsky equation, 36 Kinetic curves of crystallization, 66 Kolmogorov–Avrami equation, 66, 67, 74, 94 Krigbaum–Roe equation, 104 Lamellae, 65, 70, 82, 83 Light microscopy, 61, 101, 145, 150 Light scattering, 61, 145 Long-term low temperature behaviour/resistance, 38, 90, 128, 130 Loss modulus, 17, 93 Low-molecular-mass substances, 11, 22,
185
28, 59, 64, 65, 70, 72, 74, 106 Low-temperature resistance, see Freeze resistance, 1–3, 11, 15, 16, 26, 28, 34, 35, 38, 40, 41, 44–48, 53–55, 59, 90, 94, 125, 128–156 Master curve, 52–54, 117, 118 Mastication, 121 Maximum crystallization rate, 59, 70, 72, 84, 85, 91, 95, 119, 141, 143, 146, 147, 149, 150, 153, 155, 156 Mechanical losses, 49–51 Mechanical methods, 15, 27, 40, 43, 57, 60, 63, 89, 129 Mechanical properties, 3, 4, 9, 15, 17, 18, 28, 41, 43–60, 82, 89, 92, 104, 106, 107, 114, 122, 125, 127, 129 Melting temperature, 38, 58, 59, 63, 64, 68, 70, 72, 74, 75–95 Melting thermograms, 76, 85, 114 Mesophase/liquid-crystalline structure, 82, 103, 148 Micellar model, 65 Microfractionation of polymer in crystallization, 109, 111 Microstructure of polymer, 30, 33, 50, 83, 91, 92, 108, 110, 125, 140, 142, 143 Monosulphide bonds, 39, 45, 86, 120, 143, 144 Mooney–Rivlin strain energy function, 6 Morphology, 15, 26, 27, 61, 64, 65, 80, 81, 88, 97, 100, 101, 104, 141, 145, 150 Mullins–Patrikeyev effect, 47 Nairit, 70, 88, 90 Nanofillers, 40, 46, 89, 121, 122 Natural rubber (NR), 20, 24, 35, 44, 51–53, 61, 64, 65, 70, 75, 80, 83, 84, 86, 88, 90, 97–102, 107–111, 119–122, 129, 134, 140, 142 Neck, 55, 56 Neoprene, 84 Network, 3, 5–7, 32, 38, 39, 45, 46, 50, 51, 76, 84–87, 90, 94, 95, 101, 105, 109–111, 113, 118, 120, 141–143, 146, 149, 150, 155 Network density, 38, 39, 45, 46, 50, 51, 76, 85, 111, 149
186
SUBJECT INDEX
NMR method, 14, 38, 40, 60, 88, 93, 127, 145 Nonisothermal crystallization, 67, 73–75, 87, 89, 112 Nucleation, 67–72, 74, 75, 77, 81, 83, 84, 87, 89, 90, 112, 145, 149 O-rings, 130 Orientation of crystals, 88, 113 Orientation of rubber, 97, 100, 112, 148 Overall crystallization rate, 67–70, 84, 88, 90 Payne–Warnaka effect, 51 Phase transition, 25, 26, 28, 58, 63, 92 Photoelasticity, 61 Physical junctions, 7, 47, 50, 53, 68, 89, 91, 94, 100, 109, 111, 128, 133 Physical methods, 43, 60, 62, 68 Plastic deformation, 3, 4, 109 Plasticizers, 36–39, 45, 50, 54, 89–91, 134, 139, 140, 145, 149 Plasticizing, 38 Polybutadiene, 1,4-cis, 31, 32, 33, 35, 40, 50, 51, 59, 66, 70, 76–78, 80, 84, 85, 90, 103–105, 107, 108, 110, 111, 119, 121, 123, 129, 141–144 Polybutadiene, 1,4-trans, 31 Polychloroprene, 15, 38, 39, 59, 61, 62, 64, 65, 70, 77, 81, 83, 88 –90, 100, 101, 103, 105, 113, 120, 121, 122, 127, 129, 130, 134, 145, 146, 155 Polycrystals, 61, 65, 69, 80, 81, 87, 90, 122 Polyethylene, 11, 30, 38, 64, 78, 82, 89, 91, 92, 101, 103, 106, 109, 125, 126, 149, 153 Polyisobutylene, 25, 35, 146 Polyisoprene, 1,4-cis, 28, 41, 59, 65, 69, 72, 78, 79, 81–85, 88, 101, 102, 108–111, 113, 119–121, 129, 140–142, 144, 145 Polyisoprene-1,4-trans, see gutta percha, 83, 140 Polypropylene, 11, 38, 91, 125, 126 Polysiloxane, see Silicon rubber, 38, 40, 61, 78, 83, 84, 85, 89, 99, 100, 104–106, 123, 127, 129, 148, 149 Polysulphide bonds, 86, 87, 144
Polysulphide rubbers, see Thiocols, 39, 87, 121, 150, 151 Polyurethane, see Urethane rubber, 59, 61, 70, 90, 103, 121, 129, 138, 149, 150 Precipitated silica, 40, 46, 149 Prephase fluctuation, 68, 89, 94, 100 Pressure effect on glass transition, 25, 40 Pressure effect on melting, 58, 79 Primary nucleation, 67, 68, 70, 77 Principle of temperature–time superposition, 19, 23, 50, 52–55, 116, 132 Processing, 2, 4, 41, 60, 74, 80, 94, 115, 122, 133, 134, 141, 142, 146, 153 Quenching, 35, 101 Radiothermoluminescence (RTL), 13, 14, 35, 57, 62 Raw rubber compounds, 3, 4, 44, 114, 149, 155 Recovery, 16, 35, 41, 43–48, 51–54, 63, 79–81, 93, 94, 104, 127, 131, 132, 154 Reference temperature, 20, 53, 54, 117 Relaxation properties, 16, 40, 52, 68, 91, 93, 125 Rigidity, 3, 24, 43, 46–48, 81, 94, 122, 125, 130, 131, 149 Rubber blends, 34– 38, 44, 45, 50, 79, 81, 84, 85, 121, 144, 145, 155 Rubber goods, see Articles, see Components, see Elastomeric articles, 48, 61, 135 Rubber/fabric goods, 132, 134 Rubber/metal goods, 130, 132, 134 Rubberlike deformation, 3–5, 8, 43, 46, 47, 55, 58, 116 Rubberlike plateau, 47, 51, 63, 68, 116, 122, 133 Rupture envelope, 117 Seals, 81, 130–132, 135 Secondary crystallization, 66, 154 Secondary nucleation, 68, 70, 71, 77 Second-order phase transition, 25, 28, 58 Shear, 3, 5, 6, 8, 15, 17, 20, 80, 130 Shift factor, 19, 52, 117 Shish-kebab structure, 80, 101, 102
SUBJECT INDEX
Shock absorbers, 130, 131 Short-term low temperature behaviour, 59, 90, 128 Silica, 3, 40, 46, 89, 149 Silicon rubber, see Polysiloxane, 24, 34, 45, 84, 87, 132, 147 Single crystals, 64, 65, 69, 79, 82, 87 Small-angle X-ray method, 61, 64, 145 Specific heat, 12, 18, 24, 25, 26, 29, 35, 39, 41, 57, 62, 63, 76, 92, 98, 150, 151 Spherulites, 65, 103, 104, 122, 123 Steaming, see decrystallization, 75, 134, 141, 142, 145 Storage modulus, 8, 17, 40, 46, 49, 51 Strain energy function, 5, 6 Strength, 16, 57, 59, 94, 97, 103–106, 111, 114–123, 125, 130, 131, 134, 135, 141–143, 145, 147, 151–156 Stress, 4–8, 27, 41, 45–49, 55–58, 63, 67, 73, 79–81, 83, 88, 89, 94, 97, 99, 102–113, 115–121, 132, 144, 147, 148 Stress-induced crystallization, 7, 59–62, 85, 88, 97, 99–101, 104, 107, 109, 111, 114, 119–121, 141, 142, 145, 147, 153, 155, 156 Stress-strain curve, 100 Stretching, 46, 50, 62, 65, 80, 81, 97, 99–106, 108, 109, 111, 112–115, 119, 146, 147, 151, 156 Structural methods, 11, 60, 67, 137 Styrene-butadiene rubber (SBR), 14, 17, 24, 33, 36, 41, 44, 46, 55, 56, 84, 85, 117, 134, 137, 138, 139, 144 Supercooling, 64, 65, 68, 69, 71, 72, 75, 78, 83, 94, 97, 99, 108 Swelling, 37, 38, 45, 90, 154 Tamman curve, 69, 72, 99 Tear resistance, 155 Temperature dependence of crystallization rate, 69, 71 Temperature range of crystallization, 68–74 Tetramethyl thiuram disulfide, 39, 86, 140 Thermal capacity, 63 Thermal conductivity, 22, 62, 63, 131, 133 Thermal cycling, 133–135
187
Thermal resistance, 105, 151 Thermodynamic parameters of crystallization, 91 Thermodynamic potential, 25 Thermoelastic stress, 89, 122, 130, 134 Thiocols, see Polysulphide rubbers, 39, 87, 121, 138, 150, 151 Thomson–Gibbs equation, 79, 92, 153 Three-dimensional network, 3, 4, 32, 38, 39, 45, 50, 85, 86, 95, 101, 111, 112, 120, 142, 146, 149, 155 Tie chains, 65, 72, 107, 126 Torsion, 15 Transition of amorphous part of elastomer to glassy state, 11 Transition region, 4, 12, 15, 27, 36, 40, 44–46, 50, 52, 57, 116, 127, 152 Trans-polypentenamer, 31, 87, 121, 138, 154, 155 TR method, 16 True stress, 81 Tyres, 40, 46, 50, 122, 130, 134 Ultimate degree of crystallization, 66, 67, 81, 84, 99, 104, 144, 145, 147, 155, Uniaxial tension/stretching deformation, 4, 5, 80, 117, 119, 121, 122 Urethane rubber, see Polyurethane, 97, 108, 119, 149, 150 Viscosity, 11, 19, 22, 23, 37, 59, 65, 68–70, 101, 109, 155 Viscous flow, 4, 101, 110, 147 Volume decrease, 52 Vulcanization, 36, 38, 39, 41, 45, 51, 85, 86, 87, 110, 111, 113, 120, 139, 140, 145, 146, 149, 152, 153–155 Williams–Landel–Ferry (WLF) equation, 20, 52, 117 X-ray diffraction, 60, 88, X-ray method, 60, 92, 93, 99, 100, 107, 111, 119, 153–155 Zhurkov equation, 36 Ziegler–Natta catalysts, 140
188
SUBJECT INDEX