Beruflich Dokumente
Kultur Dokumente
Low-Temperature Behaviour
of Elastomers
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.
Foreword 1
Introduction 3
1 Glass transition of elastomers 11
1.1 Methods of glass-transition temperature determination 11
1.1.1 Structural and other physical methods 11
1.1.2 Mechanical methods 15
1.1.3 Comparison of the data obtained by different methods 17
1.2 Modern views of glass transition 18
1.2.1 Kinetic views of glass transition 18
1.2.2 Thermodynamic views of glass transition 25
1.2.3 Morphological views of glass transition 26
1.3 Effect of the molecular parameters of rubbers and the composition
of rubber compounds on glass-transition temperature 28
1.3.1 Effect of molecular mass 28
1.3.2 Effect of the structure of macromolecules 30
1.3.3 Effect of intermolecular interaction 32
1.3.4 Glass transition of copolymers 33
1.3.5 Effect of the composition of rubbers 34
1.4 Effect of a mechanical action on glass-transition temperature 41
2 Mechanical properties of elastomers near the glass transition temperature 43
2.1 Mechanical properties in the region of transition from the rubberlike to
the glassy state 43
2.1.1 Static mechanical properties 43
2.1.2 Dynamic mechanical properties 48
2.1.3 Prediction of the mechanical properties of elastomers in the
transition region from the glassy to the rubberlike state 52
2.2 Mechanical properties at temperatures below glass-transition temperature 55
3 Crystallization of elastomers at low temperatures 59
3.1 Methods of crystallization studies 60
3.1.1 Structural and other physical methods 60
3.1.2 Mechanical methods 63
3.2 Modern views of crystallization 63
3.2.1 Morphology of crystals in elastomers 64
3.2.2 Kinetics of isothermal crystallization 65
vi CONTENTS
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 elas-
tomers and elastomeric goods is low-temperature resistance. This factor preserved its sig-
nificance 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 combi-
nations make it possible to significantly broaden the temperature ranges of elastomer
applications.
There are two most studied physical processes determining the low-temperature resist-
ance 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 rap-
idly 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 In-
stitute of Elastomeric Materials and Articles (Moscow) and Lebedev Synthetic Rubber Re-
search 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 re-
ported 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 re-
commended 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. Novi-
kova 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 elasto-
mers, developing and using elastomeric materials and articles, and also for university stu-
dents, 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 be-
tween shear modulus and bulk modulus, and significant losses of mechanical energy in dy-
namic 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 com-
ponents:
• Purely elastic deformation similar to elastic deformation of common solids, with elas-
ticity modulus of |103 MPa. It is reversible and is related to the change of inter-
molecular and interatomic distances, as well as angles formed by valent bonds, which
connect atoms in macromolecules; the development of elastic deformation is accom-
panied 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 nat-
ural 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 rub-
bers and raw rubber compounds. In deformation of vulcanized rubbers, i.e., elas-
tomers 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.
ε
I II III IV
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.
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 calcu-
lated from the assumption of incompressibility:
INTRODUCTION 5
f ; σ, MPa
2 2'
10
8
2''
1
6
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 rubber-
like 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
2
NkT O12 O2 2 O32 . (4)
6 INTRODUCTION
Here N is the number of molecular segments forming the network, k is Boltzmann’s con-
stant and T is absolute temperature, and
NkT G, (4c
For uniaxial tensile/compressive deformation, from the relation (4) for normal f and true V
stresses it follows:
f G O 1/ O , (6)
V
G O 2 1/ O . (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 represen-
tations of the nature of a deformation. Thus, based on the assumption of the linear depen-
dence between stress and deformation in simple shear for an isotropic incompressible
material, an expression, which is usually called the Mooney or Mooney–Rivlin strain en-
ergy 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 , (9)
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 co-
ordinates )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. In-
troduction of additional restrictions for the motion of molecular chains between chemical
junctions made it possible to describe experimental data better. However, more new addi-
tional 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 experi-
mental data from classical theory, we can assume that its significance is the greater, the larg-
er 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 de-
formation 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 vol-
ume 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 en-
ergetic) component of the deformation is great; this results also in some peculiarity of low-
temperature 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 de-
formation 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:
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 dy-
namic 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:
where
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 low-
temperature 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 sub-
stances. 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 sub-
stances 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 poly-
ethylene, 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 crys-
tallize at all or crystallize slowly, for most of them glass transition is the major process,
which determines their low-temperature resistance.
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, cor-
responding 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 charac-
terize 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 refer-
ence 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 ob-
tained 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 con-
ductivity 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'
Figure 1.2 Dependence of the temperature change in cells, 'T c, on scanning temperature T c
(a thermogram, DTA method).
tan δe.102
2 4
3
2
1 1
0
-60 -40 -20 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 deter-
mined 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 temperature-
change 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
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
1
10 2
0
-150 -50 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 tri-
fluorochlorethylene 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 tempera-
ture 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 mo-
bility 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
-20 0 T, °C
TgNMR
Figure 1.6 A temperature dependence of the transverse relaxation time T2 (arb. units) for a poly-
chloroprene vulcanizate.
εa.102
2
1 4 5
2
3
1
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).
103 1
102 10
2
10 1.0
3
1 0.1
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 mod-
ulus 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 fre-
quency 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 glass-
transition 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/ 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)
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 pro-
posed [198, 293].
The kinetic nature of glass transition is revealed especially demonstrably in the mea-
surements 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 prop-
erties 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 increas-
ing 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 ac-
tion 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 process-
es, 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 tem-
perature 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
-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 tem-
perature 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, be-
sides 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:
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 , (1.7)
where c1g and c2g are constants; for most polymers, c1g = 17.44 and c2g = 51.6 K.
GLASS TRANSITION OF ELASTOMERS 21
The temperature–time superposition principle is used for predicting the low tempera-
ture 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 cg
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 quanti-
tative estimate of the change of Tg with the cooling rate:
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)
The geometrical free volume is the difference between the macroscopic volume of a
22 CHAPTER 1
VfG Vf
Ph
Vf
VTPh V0 VT
Tg T
Figure 1.10 A schematic illustrating the views of the geometric (VfG) and physical (VfPh) free vol-
umes, 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):
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 mol-
ecules. Considering the thermal expansion above and below the glass-transition tempera-
ture 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 – geo-
metric (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)
The free-volume concept is used the most extensively to describe the relaxation pro-
cesses and glass transition. It relates unequivocally the mobility of segments to the remain-
ing 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 tem-
peratures 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 glass-
transition 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:
K pac V0 VT (1.17)
This observation is the basis for creating the Tg calculation scheme based on the chemical-
structure 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
1 1
f gPh 0.434 c1g ; E f 0.434 c1g c2g . (1.19)
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 specific-
heat 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]:
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]:
where 'c1 is the conformational contribution to specific heat; 'c2 and 'c3 are the contribu-
tions associated with a freezeout of the free volume and a change of frequency and ampli-
tude of structural units’ vibrations.
In spite of the indisputable fruitfulness of the isofree-volume concept, it is not rigor-
ous. Thus, the critical free volume corresponding to glass transition may not be a “univer-
sal” 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]. Treat-
ment 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. Here-
with, 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, poly-
isobutylene) 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 as-
sumption 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 pos-
sible 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.
* Elucidation of the causes and conditions of such anomalies to appear in elastomers requires separate
studies.
26 CHAPTER 1
Here
'E E1 E 2 , (1.24)
'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 tran-
sition 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 vol-
ume; herewith, the change of Tg with p changing is the larger, the lower the glass-transition
temperature is. However, usually in the experiment
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 indic-
ative 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 poly-
mer, 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 oc-
currence 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 respec-
tive 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-tran-
sition) 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 seg-
ments) 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 as-
pects 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 tran-
sitions, successfully used for polymers, approaches to the description of the kinetics of
second-order phase transitions have been developed insufficiently. Development of glass-
transition theory is of great practical significance both for better understanding plastics op-
erated in the glassy state and for refining the methods of predicting the behaviour and im-
proving elastomers’ low-temperature resistance.
1
Tg Tgf K c M n , (1.28)
where Kc is the constant for a given homologous series. Equation (1.28) describes satisfac-
torily 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 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 con-
stant for polymers of a given chemical composition, eq. (1.30) can be presented as
1
Tg Ac B M n , (1.31)
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 dif-
ferences 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:
(xc is the number of monomer units), which describes the experimental data for polyiso-
prene at Tgf = 203 K; then Mnf = 9500, i.e., also approximately 104.
30 CHAPTER 1
Tg, °C
-70
-80
-90
-100
-110
-120 _
103 104 105 106 Mn
R im , % tan δ
a b
80 1
1.2 6
5
60 4
0.8 3
40 2
0.4
20 2
5 4 3 1
6 0
-80 -40 0 40 80 T, °C -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]
1
Tg Tgf AN 2 M n , (1.35)
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 polybutadi-
ene (2) on the content of units w 3,4 (1) and 1,2 (2).
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 polyiso-
prene, 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]:
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 three-
dimensional 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 halo-
gens 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.
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 – inter-
molecular 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, poly-
vinyl acetate and polyvinyl chloride have close solubility parameters (9.4 and 9.5), howev-
er, 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 33
c, % Tg, °C
Polybutadiene – from –105 up to –112
Carboxylate SKD-1 1.5 –81
10 –55
20 +1
Carboxylate SKS-30-1 0.3 –56
5 –50
15 –30
Styrene-butadiene SKS-30 – –56
where Tg1 and Tg2 are the glass-transition temperatures of homopolymers; w1 and w2 are
their mass fractions in the copolymer:
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:
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] (rub-
ber 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 con-
sidered 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 comono-
mer 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
a -50 b
-20
1
3
-40
2 2
-60 1' -100
-80
1
3
-100
-150
0 20 40 60 80 w, % 0 50 w, %
Figure 1.14 A dependence of the glass-transition temperature Tg for statistic copolymers of buta-
diene (a) on the content of comonomers: isoprene (1, 1c), styrene (2), acrylonitrile (3) and silicon rub-
bers (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 1,2-
and 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 copol-
ymers 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% con-
tent 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).
Δl K
1 1.0
0.5
0
-150 -100 T* -50 0 T, °C
Figure 1.15 A temperature dependence of the changes of length 'l (1) and recovery K (2) for a rub-
ber from a blend of butadiene (SKD) and isoprene (SKI-3) rubbers at a ratio of 1:1. T*, conventional
glass-transition temperature.
Two glass-transition regions corresponding to the glass transition of the initial compo-
nents 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 (SKMS-
10) 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 low-
resistant polymers, information of the low-temperature behaviour of the blends of elas-
tomers 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 com-
pounds based on rubber blends, which are vulcanized systems containing, besides, plasti-
cizers, 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 plasti-
cizers into rubber compounds based on polar rubbers is the most efficient. Another impor-
tant 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)
and
• Kargin–Malinsky equation:
* 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]:
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 an-
other polymer);
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, calcula-
tions 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; how-
ever, 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 plasti-
cizers (hydrocarbons C8 –C12) on the glass-transition temperature and viscosity of blends
of plasticizers with nonpolar rubbers have shown that the efficiency of plasticizers dimin-
ishes 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 impor-
tance 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 plasti-
cizer 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 re-
placement 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 poly-
propylene, 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 rub-
ber 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 increas-
ing the intermolecular interaction of elastomers. As the result of crosslinking, the free vol-
ume 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 be-
comes 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-di-
mensional network, as, in particular, ebonites or epoxy resins. Considerably less deter-
mined 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 relax-
ation time by the NMR method is the same for non-vulcanized polychloroprene, its ther-
movulcanizate 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 max-
imum 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 dif-
ference 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 coeffi-
cients above and below the glass-transition temperature are small, which reduces the accu-
racy 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 ac-
celerators, 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 vulca-
nization 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 di-
sulphide, 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 as-
sociated 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
250
150
0 500 P, MPa
Figure 1.16 A dependence of the glass-transition temperature Tg on pressure p for natural (1), sty-
rene-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 ex-
pansion 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 re-
laxation 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 meth-
ods (see Chapter 2). For large concentrations of active carbon black, as well as other, espe-
cially 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 tem-
perature 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 increas-
ingly 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], introduc-
tion of small doses of fullerens does not practically change Tg determined by the tempera-
ture 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 low-
temperature resistance of rubbers is exhibited, first and foremost, in the transition region.
* 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
The results of the comparison of Tg values obtained by different techniques show a signi-
ficant 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 mechan-
ical measurements and by the loading. Besides, depending on the service conditions of elas-
tomeric 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. Here-
with,
is usually positive, i.e., Tlow is higher than Tg; for particular applications, however, the con-
dition 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 mechan-
ical properties near this temperature. Therefore, when considering the low-temperature be-
haviour 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.
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 vul-
canizing 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 in-
cludes the transition regions of all rubbers in the blend. It begins with the onset of the tran-
sition 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 degen-
erates, 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 co-
efficients 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; here-
with, it contributes to the improvement of low-temperature resistance of rubbers in the tran-
sition region [563, 728].
Plasticizers. Introduction of plasticizers leads to a decrease of Tg and can cause a re-
duction 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; here-
with, 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 swell-
ing of a vulcanizate (as compared with its introduction by way of mixing [368, 641]) is as-
sociated 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 vulcaniza-
tion 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 pro-
duction of rubbers with low Tlow and high values of low-temperature resistance coeffi-
cients. 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 rub-
bers.
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 butadi-
ene 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
0.8 2
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 sig-
nificantly decreased (Fig. 2.1). Possibly, in the latter case the decrease of T is also contrib-
uted 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 low-
temperature 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 rub-
bers 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 tran-
sition 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 low-
temperature 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], intro-
duction 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 de-
formed under conditions of the complete development of rubberlike deformation at a tem-
perature 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:
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
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 param-
eter 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 tech-
nique has not been carried out due to the difficulties in applying any considerable deforma-
tions 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 behav-
iour of elastomers.
Measurements of KH and K by the former method make it possible to trace the equiv-
alency 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
where S10 and S20 are the areas of free and loaded surfaces of the non-deformed specimen.
48 CHAPTER 2
) d 0 2h 0 1 2G , (2.4c)
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 ob-
tain 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 spec-
imen 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 contri-
bution 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 espe-
cially clearly seen at temperatures close to Tg. The same pattern is also observed for recov-
ery 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 low-
temperature resistance of rubber goods (see Chapter 6).
* 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 prac-
tically change during the deformation.
MECHANICAL PROPERTIES OF ELASTOMERS NEAR GLASS TRANSITION 49
σk', MPa
6
5
4.0
4
3.0 3
2
2.0
1
1.0
0 0.25 0.50 ε
Figure 2.2 Dependence of the “true” mean stress Vkc on the compression deformation H for cylindri-
cal 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 maxi-
mum of tan G towards higher temperatures, as a rule, is unequivocally matched with an in-
crease 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 con-
ditions 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 character-
istics, 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 mea-
sured in the corresponding frequency – temperature range. Thus, the value of tan G deter-
mined 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 re-
spectively. 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 re-
gion 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 broad-
er. 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 dy-
namic 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 den-
sity 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 radiation-
vulcanized 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
-5 -4 -3 -2 -1 0 log (t/aT)
-0.5
4
2 -1.0
1 log K
log aT
5.0 b
1
4.5 2
4.0
3.5 3
3.0 +
4
2.5
2.0
+
1.5
+
1.0
+
0.5 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 NR-
based rubber (1) and filled butadiene acrylonitrile-based rub-
bers (2–4); the reference temperatures T0 = –25°C (1) and 0°C
-60 -40 -20 0 T, °C (2–4).
The applicability of the TTS to the dependences of recovery on recovery time t2, ob-
tained 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 dy-
namic characteristics (see Fig. 1.9 and Fig. 2.3, curve 1), i.e., is described by the WLF equa-
tion (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 tem-
perature 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 ob-
tained 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 relax-
ation processes occurs in the temperature and time (or frequency) range considered, in par-
ticular, 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 ef-
fect of deformation and temperature: the low-temperature resistance increases with the de-
formation 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 non-
linear behaviour of polymers and, in particular, elastomers, have been developed to date in-
sufficiently. Therefore, the TDS principle can be considered as an attempt to describe the
nonlinear processes using techniques well developed for linear physical relaxation process-
es, 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 low-
temperature 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 55
log aTc
c
c1g T Tg c
gc
2
T Tg , (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 extend-
ing the analogy with the TTS, the same relation was suggested to exist for c1gc and c2gc . Cor-
respondingly, 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, con-
sidering 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.
σ, MPa
10
80 9 8
7
60
6
40 5
4
2 3
20
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).
100 1
80 2
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. Stress-
induced deformation is reversible by its nature, but the specimen recovered only when heat-
ed 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 corre-
sponding to the development of stress-induced deformation is observed only at tempera-
tures 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 tempera-
0
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 elon-
gations are. The dependence of Vbr (below Tbr0) and VL on the testing temperature for rub-
bers 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 amor-
phous 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 temper-
ature 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 mea-
sured 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 elas-
tomers 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 re-
gion and the relation of the width of this region to the structure of an elastomer and its prop-
erties 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 de-
velopment 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 tem-
perature 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 elasto-
mers 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
cooperative motion of segments. There is also a view [79] that the value of Tbr 0 – the lower
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 phase-
transition 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 pos-
sibilities 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 second-
order 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 tran-
sition as well as the identity of three temperatures: the temperatures T2, the E-transition and
the lower boundary of the stress-induced elasticity region Tbr 0 (see also Chapter 1).
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 prob-
lems 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 par-
ticular, 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 co-
incides 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 prop-
erties 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 flex-
ibility 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 resis-
tance. 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.
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 fac-
tors makes elastomers distinctly different from other crystallizable polymers: high flexibil-
ity 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
* 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. Small-
angle 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 stress-
induced 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 crys-
tallinity. 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 stress-
induced crystallization. Unfortunately, the difficulties associated with low-temperature
crystallization studies restrict its wide use for characterizing structural changes in crystal-
lization 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 infor-
mation, 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 poly-
urethane 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 tech-
niques are imperfect. Therefore, the most interesting results have been obtained for poly-
chloroprene, 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 birefrin-
gence 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 re-
quired, 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 crystalliza-
tion and melting kinetics (which, in fact, could be less labour-intensive using more tradi-
tional 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 char-
acteristics of a polymer in crystallization makes it possible to use this method for crystalli-
zation 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 glass-
transition studies, has not found wide application in crystallization studies due to the un-
equivocal 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 mea-
surements 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 dilatom-
etry is considered in detail in [95]. Unfortunately, no automated volume-measurement sys-
tems 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, re-
spectively, 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 crystal-
lization studies. In this case one should take into account the increase of thermal conduc-
tivity 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 car-
ried out, and no respective methods of crystallization studies have been worked out. Meth-
ods 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
* 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 temper-
atures lower than the equilibrium melting temperature Tm0 . However, for polymers the
supercooling interval
where much broader crystallization can be observed than for low-molecular-mass substanc-
es. 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.
* 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 ob-
served in thin-film crystallization from solution.
Due to a high viscosity of the melt, no folding of chains seems to occur during the in-
crease 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 devel-
oped 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 forma-
tions. By analogy with low-molecular-mass substances, such formations can be called poly-
crystals. 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 elas-
tomers; 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].
a 1 - C/C ∞
a
I II III
1.0
0.5
τ0 τ1/2
0 100 300 t, min
b 200
C/C∞ 213 208 218 204 223 225 228
1.00
200
0.80 I
0.60
0.40
0.20
0.00
0 5 10 15 20 25 30 35 40 45
C/C∞ 218 232 240 t, min
244 248 250
1
0.8 II
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 crys-
tallization 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 poly-
mers, including elastomers, is to use a modified Kolmogorov–Avrami equation*, which re-
lates 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 67
Ct / Cf
1 exp k ct n , (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; usual-
ly, 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]:
n
Ct / Cf 1 exp ª« t / t0 º» , (3.3c)
¬ ¼
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 pro-
cesses 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 nucle-
ation process is called homogeneous or thermal. If foreign impurities serve as nuclei, in-
cluding 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 crys-
tallization 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 condi-
tions). In most cases, the Kolmogorov–Avrami equation gives a reasonable idea of the crys-
tallization 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 pro-
cess by the same set of constants.
The treatment of the kinetic curves for elastomers containing an active filler or sub-
jected 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 isother-
mal 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
ª D Dº
v v0 exp « U / kT \ Tm0 / T 'T » ; (3.4)
¬ ¼
ª D Dº
N N 0 exp « U / kT \ c Tm0 / T 'T » ; (3.4c)
¬ ¼
ª D Dº
G G0 exp « U / kT \ cc Tm0 / T 'T » ; (3.4cc
¬ ¼
D 1
E V
\ D
. (3.5)
k 'H U kD
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 depend-
ing 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 pos-
sible only in exceptional cases [19, 96, 174, 569], so the nucleation rate is usually charac-
terized 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) nucle-
ation.
The problem of singling out the nucleation and crystal-growth processes and their con-
tributions 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:
D
ln W1/ 2
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 low-
molecular-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) occur-
ring 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 en-
ergy efficiency of the crystallization process); the second is the decrease of the crystal for-
mation 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
4 1
2
2 4
5 3
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-poly-
butadiene 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 tem-
perature decrease, the crystallization rate is negligibly small already near the glass-
transition 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 de-
termined by different contributions of the interface-formation energies in the case of pri-
mary nucleation and subsequent crystal growth. The result of these differences is that the
nucleation maximum-rate temperature both for low-molecular-mass substances and poly-
mers 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 val-
ues 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 trans-
ferring 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 crys-
tal-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 suffi-
ciently 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
\ . (3.5c)
k 'H U k
Ve
k\ / 0.4b02 . (3.7c)
A
¬
ln W1/ 2 c1 / ª k c2 T Tg º ;
¼
(3.8)
Thorough analysis of experimental data for elastomers has shown that at constant val-
ues of the parameters in eq. (3.4), an inflection is observed in the region of small supercool-
ings 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 frag-
ments (i.e., at small and large supercoolings 'T, respectively) is close to 0.5.
In accordance with the Lauritzen–Hoffman crystallization theory [395], the occur-
rence of inflections in the –A –F1 curve is due to the presence of three different crystalli-
zation 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
\ . (3.5cc)
k 'H U k
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 depen-
dence 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 de-
pendence 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 elas-
tomeric 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 measure-
ments of the temperature dependence of the crystallization rate, for specimens of 1,4-cis-
polybutadiene this boundary does not coincide with T1, but is shifted to higher temperatures
[774].
* 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
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 speci-
men 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 en-
tropy; this change is a sum of three components:
where 'S1, 'S2 and 'S3 are contributions stipulated by position, orientation and conforma-
tion 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 tem-
peratures of the beginning (Tmb) and end (Tmf) of melting. Herewith, the width
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 mate-
rial 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).
ΔT' dQ/dt
a b
4 2
2 4
6
5
3
2
1
4 1.0
5 1
0.5
-90 -50 -10 30 T', ºC
238 268 T', K 260 270 T', K
Fig. 3.4 Fig. 3.5
Figure 3.4 Melting thermograms (dependences of the temperature difference in cells 'Tc on scan-
ning 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.
T m1 ,T m2 ,T mf , K
3,3'
250 2
2' 1,1' 4
230
210
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 occur-
rence of various crystalline modifications. For elastomers (excluding gutta percha), this ef-
fect 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 pos-
sible 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 dif-
ferent 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 tempera-
ture, 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-temper-
ature (Tm1) and, second, high-temperature (Tm2) melting peaks on crystallization tempera-
ture, 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 “decrystalliza-
tion” methods for rubbers and rubber articles after their crystallization in storage. In decrys-
tallization, 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 poly-
mers. 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:
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]:
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 some-
times 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 melt-
ing temperatures for crystals of known size to Tm0 of an infinitely large crystal by the
Thomson–Gibbs equation [751]
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 ex-
trapolation of the melting data using the Hoffman–Weeks method. An exception, as we not-
ed, 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 par-
ticular, 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 ini-
tial 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 equi-
librium values [95, 437a].
* 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.
80 CHAPTER 3
σ/σ0 K
a 1.0 b
3 2 1
1.0
0.5
5
0.5 4
4 1
3 2
0
102 103 t, s 1 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).
where logW1/2
0 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 elonga-
tion [95, 453]. The characteristic straight lines are parallel at various temperatures. How-
ever, 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 predomi-
nantly 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]. Espe-
cially 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 re-
covery 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 com-
pletion 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 crys-
tallization, 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 crystalli-
zation, 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 poly-
chloroprene (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 polycrys-
tals (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 crys-
tals – the most equilibrium type of crystals. The formation rate of these crystals is extremely
high, so it is extremely difficult to follow the crystallization kinetics. The temperature de-
pendence 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 pres-
sure (see Chapter 1). As for polyethylene, for which studies of crystallization at high pres-
sures 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.
('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 equilibri-
um 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. Ow-
ing 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 poly-
ethylene [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 in-
terface formation energy V e [158].
Effect of shape factor. As in the case of glass transition, changes of mechanical prop-
erties due to crystallization are the less, the larger the shape factor of a specimen is [113,
343, 643, 688, 689].
* 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 crystal-
lization 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 occur-
ring 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 poly-
mer’s microstructure, the most important from the point of view of crystallization is the reg-
ularity 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 sys-
tems [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 crys-
tallization 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 con-
taining 1,2- and 3,4-units does not crystallize, either; the capability of crystallization is de-
creased 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-poly-
isoprene 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 sen-
sitive 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) dif-
fers 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, rang-
ing from those intended for fabricating adhesives with the maximum rate and ultimate de-
gree 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]. Crystalli-
zation 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 sta-
tistical distribution, a decelerating effect is maximal, at a deviation from statisticity the co-
polymers 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 con-
tent 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 char-
acterized by the formation of microcrystals.
Crystallization of polymer blends. During the crystallization of blends of a crystalliz-
ing 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'
Figure 3.8 Melting thermograms obtained by the DTA method for vulcanized specimens of the rub-
ber 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 incom-
patible 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 prac-
tice 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 melt-
ing 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, po-
lybutadiene [420]), crystallization is performed mainly at the expense of the rapidly crys-
tallizing 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 vulcani-
zation 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 den-
sity 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-in-
duced crystallization (see Chapter 4). An increase of the network density leads to a suppres-
sion 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 param-
eter 1/Mc for rubbers based on NR with poly- and disulphide (1) and monosulphide and C–C (2)
bonds.
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 three-
dimensional network density increasing can be observed using a light microscope [95, 96].
The same conclusions are suggested by the consideration of vulcanizates as copoly-
mers 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 melt-
ing 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 crystal-
lization 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 crys-
tallization process. Therefore, different results can be obtained, even studying the same sys-
tems 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 de-
crease 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]. Nonisothermic-
ity 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, chloro-
prene, 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 direc-
tions: (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 acceler-
ation 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 insig-
nificantly 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'
1
4.0 2
0
1 2 log τ1/2, min
Figure 3.10 Dependence between stress in a specimen before crystallization V and log W1/2 for rub-
bers 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).
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 polysilox-
anes with various fillers have shown that both the melting character and value Tm0 are usu-
ally 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 broad-
ening, 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 in-
creased 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 sur-
face of filler particles was shown during the measurement of a thermal expansion of elas-
tomers 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 crys-
tallization 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 ma-
trix. This conclusion is suggested by the analysis of the effect of elastomer crystallites them-
selves 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 deter-
mined by their nature. The traditional approach to plasticizers as products decreasing the
crystallization rate is wrong.
90 CHAPTER 3
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 de-
pendence 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
6
5 7
4 I
2
3
1 2 II
4
1
endo
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) with-
out 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 tem-
perature 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 Chap-
ter 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 prop-
erties of elastomers, i.e., in this region of “average” compositions the capability of crystal-
lization of both polyethylene and polypropylene is significantly disturbed. Earlier, these
polymers were considered amorphous; however, studies of their low-temperature proper-
ties 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 pro-
cess [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
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 poly-
ethylene 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 copoly-
mer; studying it enables a more exact characterization of the microstructure of such elas-
tomers. 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 93
K
a
1.0
3
2
1
0.5
3' 4'
4''
3''
0.5
4'''
1' 3'''
1''
1'''
0 20 0 80 0 20 40 t, h
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 dy-
namic 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 recov-
ery 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 ab-
sence of an induction period [132, 133].
The dependence of melting temperature of microcrystals of ethylene-propylene rub-
bers 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 mi-
crocrystals 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 co-
polymers.
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 represent-
ing 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 micro-
crystallization. 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 pe-
culiarities of the crystallization kinetics (see above). Thus, transition from the microcrys-
tallization 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 resis-
tance of rubber and rubber articles deteriorates (see Chapter 6). At the same time, the co-
hesion strength of rubbers is improved, which facilitates the use of elastomers capable of
microcrystallization for fabrication of complex components.
* 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.
Prediction by time. Knowing the value of W1/2 0 , one can findW required to achieve any
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 activa-
tion energy U does not depend on crystallization temperature and D = 2; the values of
Tm0 are related one-to-one with W1/20 at the maximum crystallization rate temperature T [95,
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
prediction accuracy, but replaces the bulky experiment for determining W1/2 0 at T by a more
1
compact determination of Tm [380, 450]. At a change of composition, Tm0cc is assumed to
0
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 com-
pletely abandon experiments with concrete rubbers, which are a must when using the earlier
developed approaches.
96 CHAPTER 3
4 Stress-induced
Crystallization
of Elastomers
a c
1.5
0 1
1.0 4'
3
1 2
0.5 3'
4
2 1
0
102 103 104 log t, min 1 0 1 log t, min
C, %
36 b
28 4
20
12 1
4
4 8 12 16 20 24 t, h
n C,% C,%
a b c
3
15 2
4 4'
30 6
4
10
3 1 5 20
2 5 7
1 2-5 10
1 2 3 4 5 6 λ 3 4 5 6 λ 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 deforma-
tion. 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 polysilox-
anes [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 un-
equivocal. 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 con-
ditions 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 impossi-
bility of registering the initial stages of the process at large H.
Gent and Zhang [239] compared the kinetics and degree of crystallization of three elas-
tomers – NR, butadiene and butyl rubbers – using weakly crosslinked and then cooled spec-
imens. 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 meth-
od* [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). Here-
with, an increase of H or O indicates not only an increase of the crystallization devel-
opment 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 con-
ditions 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 super-
cooling 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 asso-
ciated with the need to rearrange part of the crystals formed during the stretching [75]. How-
ever, 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 com-
paring 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]:
* 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 ǻ
¬ 1/ 2 º
6 / S O Ns ª 2
º
»¼ «¬ O / 2 1/ O / Ns »¼ ;
(4.1c)
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 crys-
tallization. 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 non-
deformed 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 crystal-
lization, 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 crys-
tallization 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 poly-
chloroprene [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 ob-
tained 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 regis-
tered 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 forma-
tion 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 fore-
most, its abnormally high rate, is the formation of the preceding liquid-crystalline phase
[240, 253].
with osmium tetroxide. Some information could also be obtained by the method of light mi-
croscopy [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, espe-
cially 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 spec-
imen is deformed up to H < Hk at T | Tm0, and then cooled (two-stage stress-induced crys-
tallization). 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 defor-
mations, 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 com-
pleted 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 orien-
tation 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 pos-
sibility 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 (“unfold-
edness”) 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 un-
folding parameter E is presented in Fig. 4.3a. In the absence of an external orientational ef-
fect, 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 unfold-
ed 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 poly-
mers 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 ap-
plication of an intensive mechanical field to a polymer solution, in particular, in stirring [553]. As ap-
plied 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 b
3 2
0.6
1 1
2 0.4
3 βcr
0.2
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 temper-
ature T prior to the onset of crystallization [644] (Mc was taken to be 116 [141]).
Ms 3RT U / Ef *, (4.2cc)
* 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 de-
scribing the difference of crystals formed in three kinetic regions of crystallization (see Chapter 3).
STRESS-INDUCED CRYSTALLIZATION OF ELASTOMERS 103
0
1/ Tm 1/ Tm, O R / 'H ª¬« 6 / S m 1/ 2 O O 2 / 2m 1/ O m º¼» , (4.3)
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 un-
suitability 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
2.50 1
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).
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 constan-
cy of the right-hand side part of the equation yield:
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 manu-
facturing conditions); for rubbers from polychloroprene it is within the range of
120–150°C, which explains their comparatively high thermal resistance. For common rub-
bers 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
a I
II
-20 0 20 40 60 T', °C
-dε/dt -dε/dt
25 b c
20 2
1
15 2
10 1
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 val-
id 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 lim-
itations for crystallization, and its occurs very rapidly close to the equilibrium melting tem-
perature 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 in-
termediate 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 thermo-
grams 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-polyb-
utadiene 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 de-
termined by the birefringence method at different temperatures. The assignment of the high-
er 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 exper-
imental 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 temper-
atures 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 crys-
tallization can proceed, and crystals do not melt after the stress is released).
Contraction of oriented elastomers during the melting, used in measurements of resid-
ual elongations, makes it possible to determine the melting temperatures of oriented spec-
imens 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. Nat-
urally, 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 macro-
molecules, 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-equilib-
rium 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 experimen-
tal 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 CHAPTER 4
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 crystal-
lization 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 macromol-
ecules 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 char-
acteristic 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-polyiso-
prene. 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 extend-
ed-chain crystals for polymers with a broad MMD. In the first case, macromolecules of dif-
ferent 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 di-
rectly 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 exper-
imental points is much greater. This is due mainly to the strong influence of the other mo-
lecular 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-cis-
polybutadiene 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 Tm2,ε, °C
-5 a 4 b
-10 2
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).
* 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 sup-
pressing isotropic crystallization better (sulphur with sulphenamides, mercaptobenzo-
tiazole 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 den-
sities. At the same time, there are data on the independence of the deformation correspond-
ing to the onset of stress-induced crystallization (Hk) for NR and synthetic 1,4-cis-
polyisoprene 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 net-
works. 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 ther-
modynamic 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 stress-
induced and isotropic) the vulcanization network junctions do not enter the monocrystal
[333]. Therefore, the distribution by size of extended-chain crystals occurring in stress-
induced 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 in-
tensity 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 stress-
induced 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 small-
er than Mc.
** This distinction restricts the possibilities of using parameter B in eq. (3.17) to predict stress-
induced crystallization.
112 CHAPTER 4
-dε/dT
6
4
2 6
6
4
5
2
5
3 4
1
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.
Tm, °C Tm, °C
20 a b
4 16
16
3 12 4
12 3
2 8
8 2
1 4
4 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.
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 rein-
forcing chalk. Introduction of carbon black PM-100 also leads to the appearance of two ad-
ditional 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 tempera-
tures [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 mechan-
ical properties of vulcanized and raw rubber compounds.
Introduction of a filler into the SKD rubber also enhances its capability of stress-in-
duced 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 con-
traction 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 crystalliza-
tion 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 defor-
mation 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 crystalliza-
tion. 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 stress-
induced 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.
σr
I' I''
II
III 2
IV 1
Tbr0 Tg Tmax T
Figure 5.1 Dependence of the breaking stress Vr on the temperature of amorphous (1) and crystal-
lizable (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 temper-
atures, the role of relaxation processes increases, and they become predominant at temper-
atures higher than the glass-transition temperature Tg.
There is no unequivocal views of the nature of the fracture in the stress-induced rub-
berlike 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 stress-
induced (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 deter-
mined 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 temper-
ature 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 transi-
tion region take on the value characteristic of the rupture strength of elastomers in the rub-
berlike 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
6.0 b
5.0
4.0
3.0
log σr(TS/T), MPa
2.0
a
1.8
1.0
1.4
0 .
-8 -6 -4 -2 0 2 4 6 -log aT ε
1.0
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 temper-
atures from –67.5 up to +93.3°C.
1.0
0.5
-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 enve-
lope”) 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 enve-
lopes 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 crys-
tallization.
The applicability of the Bartenev equation
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 pro-
cesses (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 condi-
tions, 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 pro-
nounced when the stress in a specimen is determined to a great extent by the elastic com-
ponent, 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].
σr, MPa
εr 4
28
12 3
24
8
1
20 2
4
b
16
0
-40 0 40 80 120
12 T', °C
4
8
3 a
4 1
2
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).
* 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
of uniaxial tension, but also in abrasion. Apparently, this is true for the tearing strength un-
der other conditions, too, when an elastomer capable of stress-induced crystallization is un-
der 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 pre-
ceded 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 de-
gree of strengthening due to stress-induced crystallization depends in these cases on the life-
time 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 noncrys-
tallizable polymers never makes it possible to achieve the values of tensile strength provid-
ed 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 re-
inforcement of elastomers as compared even with active fillers is the small size of unit crys-
tallites 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 macromole-
cules, bordering on crystallites, can be crystallites’ components, too. Both these factors en-
able 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.
degree of crystallinity. This was shown for 1,4-cis-polybutadiene (SKD) [513] and poly-
siloxane [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 meth-
ods. 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
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, extra-
polation for w = 0.
ε K
0.10 a b
1 1.0 1
2 4
0.05 2 4
3 0.5
3
0 0
-40 -30 -20 -10 0 10 20 T, °C -40 -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]. How-
ever, 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 crystal-
lization. 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 crys-
talline 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, al-
though Tg measured by both this and other methods does not change during the crystalliza-
tion 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 crys-
tallization 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 crystal-
lization develops in the temperature region, which corresponds to its maximum, the monot-
ony 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 crys-
tals 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.
Development of the crystallization process in a glassy elastomer is made difficult. The iso-
thermal 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 nu-
cleus. If the isothermal holding (annealing) is performed in the transition region in imme-
diate 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.
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 crystal-
lized 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 crystalliz-
able 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 proper-
ties 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 crystalliza-
tion. 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]).
* 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 crystalli-
zation, too.
130 CHAPTER 6
including thin-layer rubber-metal elements and coatings at temperatures lower than it fol-
lows from the tests of common specimens, is well known to practical engineers. An illus-
trative 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 applica-
ble 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 com-
ponents 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 im-
provement 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 ther-
moelastic 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 lim-
iting 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 mate-
rial or articles do not limit the temperature range of its performance, is the use of local heat-
ing, 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 op-
erational 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 Shut-
tle 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 per-
form at temperatures higher than the ambient temperature. For shock absorbers, this tem-
perature 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 frac-
ture 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 cor-
rectly 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 me-
dia 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 elas-
tomeric materials with low Tlow remains very topical.
* 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 (gas-
kets) 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 elas-
tomer, 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 tran-
sition 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 crystalliza-
tion 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 crystal-
lization). This was first shown [38] for crystallizable rubbers, the loss of hermeticity of im-
movable 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 contrib-
uting 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 com-
pared 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 rub-
ber 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 133
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 vol-
ume 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 rub-
bers 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 corre-
sponding 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].
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 accumu-
lation 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. Here-
with, 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 prop-
erties 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 break-
down 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 ob-
served 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 spec-
imens 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 in-
crease 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 de-
crease 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 mois-
ture). 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 long-
term storage of articles under winter conditions in Canada was shown by the example of
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 de-
scribing 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 dif-
ficult to obtain quantitative data, which could be considered as physical constants charac-
terizing 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 low-
temperature 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 mea-
surements, the literature data are used [221, 444, 476, 670, 744, 746].
30% of styrene units are isolated, about 40% are arranged as diads [221]. An increased con-
tent of 1,2-units determines a higher value of Tg obtained by extrapolation of the depen-
dence 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 attri-
buted 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 acrylo-
nitrile 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).
* 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 ex-
cellent model for experimental and theoretical crystallization studies of elastomers in gen-
eral, 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 lamel-
lar 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 pro-
cessing 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 “steam-
ing” 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 crystalliza-
tion, 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 distribu-
tion), 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 de-
gree of crystallization. Reprecipitation of extracted rubber with chloroform leads to the
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 tem-
peratures of –8°C and +14°C. Introduction of acetone extract products to synthetic poly-
isoprene 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 con-
sidered 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 elas-
tomers, 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 dif-
ferent types of rubbers was considered in detail in a number of reviews [69, 424, 640, 739].
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 crystal-
lization on the processing of rubber compounds the most completely [43, 275, 699, 700].
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 crys-
tallization 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 rubber-
based 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.
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 de-
crease 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 for-
mation 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 resis-
tance is completely determined by glass transition (see Table 7.1). Homopolymer SKT-
FT-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 copoly-
mer of 8% phenyl units (SKTFV-803), with the blocks containing up to 30% of homoge-
neous 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 meth-
od of their introduction, as well as for studies of interchain exchange of polyorganosilox-
anes [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 py-
rolytic 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 crystal-
lization [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 substitu-
ent, crystallization of copolymer can be suppressed, so in this case the low-temperature re-
sistance 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-dimen-
sional network rises up to a certain limit, which depends on the type and content of filler
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 pos-
sibilities 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-temper-
ature 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 litera-
ture. In the mean time, such data are of interest not only from the point of view of low-tem-
perature 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 ac-
celeration 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.
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 low-
temperature resistance. However, by decreasing the regularity of polyesters forming a flex-
ible 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].
7.10 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, re-
spectively [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 re-
sistance 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 re-
sistance 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 resis-
tance is accompanied by a deterioration of the properties of oxygen-containing fluoroelas-
tomers during their prolonged holding at low temperatures in the process of micro-
crystallization, 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 pos-
sibility of developing various copolymers with rather low Tg [22, 184, 488]. Low-
temperature resistant fluoroelastomers have been developed and are commercially manu-
factured [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 mi-
crocrystallization even at small deviations of the distribution of such units from the statis-
tical 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 devel-
opment 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].
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 pro-
cess 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 en-
tering 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 tem-
perature 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 temper-
atures; herewith, their melting temperature Tm | 50–60°C (at the crystallization tempera-
ture 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 microcrys-
tallization; 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 ethyl-
ene-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 vulcaniza-
tion 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 com-
position 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.
7.14 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-configu-
ration; 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 rub-
bers 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 de-
gree 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 crystalliza-
tion, of considerable interest are blends of this rubber with other polymers, e.g., with poly-
isoprene [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 age-
ing. However, the tear resistance and low-temperature resistance (determined by low-tem-
perature crystallization) are slightly decreased [517, 520].
1. Abramova, N.L., Zlobina, M.B. and Akopyan, L.A. (2006) Kauchuk i Rezina, No 6: 5
(in Russian).
2. 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.
3. 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).
4. Ajroldi, G., Bonardelli, P., Fumagalli, M. and Moggi, G. (1986) Proc. Int. Rubber Conf.
Göteborg, vol. 2, p. 467.
5. Akkerman, S. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere,
Finland, C6.
6. Alegra, G. (1986) In Lae, L. and Mark, Y.E. (eds.) Advances in Elastomers and Rubber
Elasticity. London–New York: Plenum Press, p. 303.
7. Aleksandrov, A.P. (1945) Proc. 1st and 2nd Conf. on High-molecular-weight Compounds.
Moscow–Leningrad: USSR Acad. Sci. Publishers, p. 49 (in Russian).
8. Aleksandrov, A.P. and Lazurkin, Yu.S. (1944) Dokl. Akad. Nauk SSSR, 45: 308 (in Russian).
9. Aliguliyev, R.M., Khiteyeva, D.M., Mamedov, A.A. and Oganyan, V.A. (1983) Dokl. Akad.
Nauk Az. SSR, 39(2): 87 (in Russian).
10. Aliguliyev, R.M., Ovanesova, G.S., Khiteyeva, D.M. and Oganesyan, V.A. (1985)
Vysokomol. Soed., ser. A, 28(7): 1452 (in Russian).
11. Aloev, V.Z., Kozlov, G.V., and Zaikov, G.E. (2004) Kauchuk i Rezina, No 3: 38 (in Russian).
12. Andrews, E.H. (1962) Proc. Roy Soc., A270(1341): 232; (1964) A277(1371): 562.
13. Andrews, E.H. (1979) In Andrews, E.H. (ed.) Development in Polymer Fracture. London:
Applied Science Publications, vol. 1, p. 1.
14. Andrews, E.H. and Gent, A.N. (1963) In Bateman, L. (ed.) The Chemistry and Physics of
Rubber-like Substances. London: Maclaren, p. 225.
15. 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.
16. 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.
17. Andrews, E.N. and Reeve, B. (1971) J. Mater. Sci., 6: 547.
18. Andrews, E., Owen, P.J. and Reed, P.E. (1971) Proc. Int. Rubber Conf. 1969. Moscow:
Khimiya Publishers, p. 95 (in Russian).
19. Andrews, E.H., Owen, P.J. and Singh, A. (1971) Proc. Roy. Soc., A324(1556): 79.
20. Ansarifar, A., Ibrahim, N., Bennett, M. (2005) Rubb. Chem. Techn., 78(5): 93.
21. 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).
22. Arcella, V. (1989) Prepr. Scandinavian Rubber Conference “Arctic Rubber”. Tampere,
Finland, F4.
23. Arctic Rubber (1989) Prepr. Scandinavian Rubber Conference. Tampere, Finland.
24. Arenz, R.J. and Landel, R.F. J. (1982) J. Polym. Sci. Polym. Phys. Ed., 20(3): 363.
25. Arlman, J.J. (1950) Rubb. Chem. Technol., 23(2): 306.
158 REFERENCES
26. Arlman, J.J. and Goppel, J.M. (1950) Rubb. Chem. Techn., 23(2): 319.
27. Askadsky, A.A. (1973) Deformation of Polymers. Moscow: Khimiya Publishers,
448 pp. (in Russian).
28. Askadsky, A.A. (1995) Vysokomol. Soed., ser. A, 37(2): 332 (in Russian).
29. Askadsky, A.A. and Matveyev, Yu.I. (1983) Chemical Structure and Physical Properties of
Polymers, 248 pp. (in Russian).
30. Avitable, G., Napolitano, R. and Riva, F. (1980) J. Polym. Sci. Polym. Phys. Ed., 7: 1663.
31. Avrami, M.A. (1939) J. Chem. Phys., 7: 1103; (1940) 8: 212; (1941) 9: 177.
32. 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).
33. Babitsky, E.L. (1984) Prepr. Int. Rubber Conf. Moscow, B 56.
34. Bach, D., Van Gool, G., Steffens, J. and Kataev, A.D. (2004) Abstr. Int. Rubber Conf.
Moscow, p. 274.
35. Bagdasaryan, E.I., Arutyunyan, L.S., Zakharov, N.D. and Kabalyan, Yu.K. (1978) Kauchuk i
Rezina, No 4: 9 (in Russian).
36. Bahani, M., Laupretre, F. and Monnerie, L. (1995) J. Polym. Sci. Polym. Phys. Ed., 33(2):
167.
37. Baldwin, P.P. and Ver Strate, G. (1972) Rubb. Chem. Technol., 45(3): 709.
38. Baranov, N.S. and Elkin, A.I. (1974) Kauchuk i Rezina, No 5: 33; (1978) Kauchuk i Rezina,
No 6: 37 (in Russian).
39. Baranov, N.S., Sokolov, V.E. and Elkin, A.I. (1982) Kauchuk i Rezina, No 8: 25 (in Russian).
40. Baranov, V.G. (1977) Khim. Volokna, No 3: 11 (in Russian).
41. Baranov, V.G., Khrenov, V.V. and Frenkel’, S.Ya. (1975) Fiz. Tverd. Tela, 17: 1550 (in
Russian).
42. Baranov, V.G., Ovanesov, G.T., Gasparyan, K.A. et al. (1974) Dokl. Akad. Nauk SSSR,
217(1): 119 (in Russian).
43. Barruel, P. (1984) Prepr. Int. Rubber Conf. Moscow, C41.
44. Bartenev, G.M. (1979) Structure and Relaxation Properties of Elastomers. Moscow: Khimiya
Publishers, 288 pp. (in Russian).
45. Bartenev, G.M. and Barteneva, A.G. (1992) Relaxation Properties of Polymers. Moscow:
Khimiya Publishers, 382 pp. (in Russian).
46. Bartenev, G.M. and Barteneva, A.G. (2005) Kauchuk i Rezina, No 1: 8 (in Russian).
47. Bartenev, G.M. and Gorbatkina, Yu.A. (1959) Vysokomol. Soed., 1(5): 769 (in Russian).
48. Bartenev, G.M. and Kilimnik, O.V. (1983) Koll. Zh., 45(2): 299 (in Russian).
49. Bartenev, G.M. and Kuchersky, A.M. (1970) Mekhan. Polim., No 3: 544 (in Russian).
50. Bartenev, G.M. and Kuchersky, A.M. (1970) Vysokomol. Soed., ser. A, 12(4): 79 (in Russian).
51. Bartenev, G.M. and Lazorenko, M.V. (1984) Vysokomol. Soed., ser. B, 26(5): 330 (in
Russian).
52. Bartenev, G.M. and Lukyanov, I.A. (1955) Zh. Fiz. Khim., 29: 1486 (in Russian).
53. Bartenev, G.M. and Lukyanov, I.A. (1963) Plast. Massy, No 12: 46 (in Russian).
54. Bartenev, G.M. and Vishnitskaya, L.A. (1952) Zh. Techn. Phys. 22: 426 (in Russian).
55. Bartenev, G.M. and Vishnitskaya, L.A. (1962) Vysokomol. Soed., 4: 1324 (in Russian).
56. Bartenev, G.M. and Voevodskaya, M.V. (1964) Kauchuk i Rezina, No 3: 21(in Russian).
57. Bartenev, G.M. and Voevodskaya, M.V. (1964) Kauchuk i Rezina, No 12: 14 (in Russian).
58. Bartenev, G.M. and Voevodskaya, M.V. (1966) Kauchuk i Rezina, No 3: 25 (in Russian).
59. Bartenev, G.M., Kuchersky, A.M. and Radaeva, G.I. (1977) Vysokomol. Soed., ser. B, 19(8):
564 (in Russian).
60. Bartenev, G.M., Kuchersky, A.M. and Radaeva, G.I. (1981) Vysokomol. Soed., ser. A, 23(2):
283 (in Russian).
61. Bartenev, G.M., Lyalina, N.M. and Revyakin, B.I. (1982) Vysokomol. Soed., ser. B, 24(8):
567 (in Russian).
62. Bartenev, G.M., Kuchersky, A.M., Fedyukina, L.P. and Radaeva, G.I. (1979) Kauchuk i
Rezina, No 12: 35 (in Russian).
63. Beardsley, K.P. and Wortman, W.A. (2000) Abstr. Int. Rubber Conf. Helsinki, p. 102.
64. Beatty, C.L. and Karasz, F.E. (1975) J. Polym. Sci. Polym. Phys. Ed., 13(5): 971.
65. Begishev, V.P., Bolgov, S.A., Malkin, A.Ya. et al. (1980) Vysokomol. Soed., ser. B, 22(2): 124
REFERENCES 159
(in Russian).
66. Begishev, V.P., Kipin, I.A. and Malkin, A.Ya. (1982) Vysokomol. Soed., ser. B, 24(9): 656
(in Russian).
67. 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).
68. Bekkedahl, N.J. (1934) J. Res. Nat. Bur. Stand., 13(2): 411; (1935) Rubb. Chem. Technol.,
8(1): 5.
69. Bekkedahl, N.J. (1967) Rubb. Chem. Technol., 40(3): 655.
69a. 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.
70. Bell, C.L.M., Stinson, D. and Thomas, A.G. (1982) Rubb. Chem. Technol., 55(1): 66.
71. Belousov, V.N., Kotsev, G.Kh. and Mikitayev, S.A. (1985) Dokl. Akad. Nauk SSSR, 280(5):
1140 (in Russian).
72. Belyaev, O.F., Kunizhev, B.I. and Zelenev, Yu.V. (1979) Vysokomol. Soed., ser. B, 21(4): 271
(in Russian).
73. 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).
74. 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).
75. Berezkina, A.P., Nelson, K.V. and Kurlyand, S.K. (1986) Kauchuk i Rezina, No 12: 4
(in Russian).
76. Berghmans, H. (1991) Abstr. Int. Conference “Polymer Networks”. Moscow, L-20.
77. 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.
78. Bershtein, V.A. and Egorov, V.M. (1985) Vysokomol. Soed., ser. A, 27(11): 2440 (in Russian).
79. Bershtein, V.A. and Egorov, V.M. (1990) Differential Scanning Calorimetry in Polymer
Physical Chemistry. Leningrad: Khimiya Publishers, 255 pp. (in Russian).
80. Bershtein, V.A., Egorov, V.M. and Emelyanov, Yu.A. (1985) Vysokomol. Soed., ser. A,
27(11): 2451 (in Russian).
81. Bershtein, V.A., Egorov, V.M. and Ryzhov, V.A. (1986) Vysokomol. Soed., ser. B, 28(4): 268
(in Russian).
82. Bershtein, V.A., Egorov, V.M., Marikhin, V.A. and Myasnikova, L.P. (1985) Vysokomol.
Soed., ser. A, 27(4): 771 (in Russian).
83. Bianchi, U. (1965) J. Phys. Chem., 69(5): 1497.
84. 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).
85. Boochathum, P. and Prajudlake, W. (2001) Eur. Polym. J., 37(3): 411.
86. Boyer, R.F. (1963) Rubb. Chem. Technol., 36(5): 1301.
87. Boyer, R.F. (1974) Macromolecules, 7(1): 142.
88. Boyer, R.F. and Keineth, S.E. (ed.) (1980) Molecular Motion in Polymers by ESR. New
York: Harwood Acad. Publishers, 328 pp.
89. Boyer, R.F. and Simha, R. (1973) J. Polym. Sci. Polym. Lett. Ed., 11(1): 33.
90. Brazier, D.W. and Nickel, G.H. (1979) Rubb. Chem. Technol., 52(4): 735.
91. Brown, R.P. (1987) Physical Testing of Rubber, 2nd edition. London: Elsevier, 348 pp.
92. Bruzzone, M. and Sorta, E. (1978) Polymer, 19(4): 467; (1979) Rubb. Chem. Technol.,
52(1): 207.
93. Bruzzone, M., Cesca, S. and Carbonaro, A. (1978) Rubb. Chem. Technol., 51(5): 907.
94. Bukhina, M.F. (1963) Vysokomol. Soed., 5(11): 1725 (in Russian); (1964) Rubb.
Chem. Techn., 37(2): 404.
95. Bukhina, M.F. (1973) Crystallization of Rubbers. Moscow: Khimiya Publishers, 239 pp.
(in Russian).
96. 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).
97. Bukhina, M.F. (1984) Technical Physics of Elastomers. Moscow: Khimiya Publishers,
224 pp. (in Russian).
160 REFERENCES
98. Bukhina, M.F. (2000) Abstr. Int. Rubber Conf. Helsinki, p. 120.
99. 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: TsNIITENefte-
khim, 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 161
123. Bukhina, M.F., Voevodskaya, M.V. and Karandashov, B.P. (1971) Vysokomol. Soed., ser. A,
13(4): 775 (in Russian).
124. 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).
125. 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).
126. 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.
127. 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.
128. 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).
129. 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.
130. 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).
131. 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.
132. Bukhina, M.F., Zorina, N.M., Voloshin, V.N. and Rudenko, G.A. (1989) Prepr. Scandinavian
Rubber Conference “Arctic Rubber”. Tampere, Finland, C10c.
133. 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).
134. Bulavin, L.A., Aktan, E.Yu. and Zabashta, Yu.F. (2003) Vysokomol. Soed., ser. A, 45(10):
1684.
135. 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.
136. Burfield, L. (1984) Polymer, 25(12): 1823.
137. 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.
138. Cesari, M., Galderazzo, F. and Cucinella, S. (1976) J. Polym. Sci. Polym. Lett. Ed., 14: 107.
139. Cesari, M., Perego, G., Zauetta, A. and Gargani, L. (2003) Macromol. Chem., 181(5): 1143
140. Chalykh, A.E. (1987) Diffusion in Polymer Systems. Moscow: Khimiya Publishers, 510 pp.
(in Russian).
141. 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).
142. 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).
143. Chernov, V.M. and Fedotov, V.D. (1981) Vysokomol. Soed., ser. A, 23(4): 939.
144. 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).
145. Chi, M. (1981) J. Polym. Sci. Polym. Chem. Ed., 19(7): 1767.
146. Chinnosamy, T.V. and Harimur, H. (1983) Angew. Makromol. Chem., 112: 45.
147. Clark, E.I. and Hoffman, I.D. (1984) Macromolecules, 17(4): 878.
148. Clarson, S.I. (1985) Polymer, 26(6): 930.
149. Coe, D.G. and Howgate, P.G. (1986) Proc. Int. Rubb. Conf., Göteborg, vol. 2, p. 418.
150. Cohen, M.H. and Grest, G.S. (1979) Phys. Rev., B20(3): 1077.
162 REFERENCES
151. Cole, Y.H. and Pierre, L.E. (1978) J. Polym. Sci. Polym. Symp., 63: 205.
152. Collins, E.A. and Chandler, L.A. (1966) Rubb. Chem. Technol., 39(2): 193.
153. Conchman, P.R. (1981) Polymer Eng. Sci., 21(7): 377.
154. Cook, S. (2000) Abstr. Int. Rubber Conf. Helsinki, p.97.
155. Copley, B.C. (1982) Rubb. Chem. Technol., 55(2): 416.
156. 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.
REFERENCES 163
186. Edwards, B.C. and Phillips, P.J. (1975) J. Pol. Sci. Polym. Phys. Ed., 13(11): 2117.
187. Edwards, B.S., Phillips, P.J. and Sorenson, D. (1980) J. Polym. Sci. Phys. Ed., 18: 1737.
188. 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.
189. Eisenberg, A. and Shen, M.C. (1970) Rubb. Chem. Technol., 43: 156.
190. Eldred, R.J. (1984) Rubb. Chem. Technol., 57(2): 320.
191. 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).
192. Enle, Z., Guiping, J. and Yanshou, L. (1982) Polym. Commun., No 3: 178; (1986) No 4: 251.
193. Fedors, R.F. (1979) Polymer, 20(9): 1055; 20(10): 1255.
194. Fedotov, V.D., Chernov, V.M. and Khazanovich, T.N. (1978) Vysokomol. Soed., ser. A, 20:
919 (in Russian).
195. Fedyukin, D.L. and Zakharenko, N.V. (1960) Kauchuk i Rezina, No 9: 43 (in Russian).
196. Fedyukin, N.D., Severina, N.L., Ionov, Yu.A. and Bukhina, M.F. (1984) Kauchuk i Rezina,
No 7: 14 (in Russian).
197. 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).
198. 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.
199. Fisher, E.W. (1990) Proc. 2nd Int. Workshop Non-Crystal. Solids: “Basic Features of Glassy
States”. Singapore, p. 172.
200. Flory, P.J. (1947) J. Chem. Phys., 15(6): 397; 15(9): 684.
201. Flory, P.J. (1949) J. Chem. Phys., 17(3): 223.
202. Fomina, L.G. and Yurovsky, V.S. (2000) Kauchuk i Rezina, No 2: 39 (in Russian).
203. Fomina, L.G., Sedov, V.V. and Kaporovsky, B.M. (2000) Kauchuk i Rezina, No 4: 26
(in Russian).
204. Fox, T.G. and Flory, P.J.J. (1950) J. Appl. Phys., 21: 581.
205. Freakley, P.K. and Payne, A.R. (1978) Theory and Practice of Engineering with Rubber,
London: Applied Science Publications, 666 pp.
206. Frenkel, S.Ya. (1973) In Physics Today and Tomorrow. Leningrad: Nauka Publishers, p. 176
(in Russian); (1977) Khim. Volokna, No 3: 11 (in Russian).
207. Frenkel, S.Ya. (1992) Abstr. Int. Conference “Orientational Phenomena in Polymers”.
St. Petersburg, p. 3.
208. Frenkel, S.Ya. and Elyashevich, G.K. (1971). Vysokomol. Soed., ser. A, 13(2): 493
(in Russian).
209. Frisch, H.L. and Mark, J.E. (196) Chem. Mater., 8(8): 1735.
210. Fuller, K.N.G. (2000) Abstr. Int. Rubber Conf. Helsinki, p. 86.
211. Furukawa, I. (1973) Polym. J., 4(4): 358.
212. Galil-Ogly, F.A., Novikov, A.S. and Nudelman, Z.N. (1966) Fluoroelastomers. Moscow:
Khimiya Publishers, 235 pp. (in Russian).
213. Gallymov, I.I., Bukhina, M.F., Maklakov, A.I. (1978). Vysokomol. Soed., ser. B, 20: 336
(in Russian).
214. Galperina (Zorina), N.M. and Severina, N.L. (1982) In Application of Mathematical Methods
in Rubber Research. Moscow: TsNIITENeftekhim, p. 25 (in Russian).
215. 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).
216. Gamlitsky, Yu.A. and Bass, Yu.P. (2003) Inzh.-Fiz. Zh. (J. Eng. Phys. Thermophys.), 76(5):
101 (in Russian).
217. 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.
218. Gargani, L., De Ponti, P. and Bruzzone, M. (1986) Proc.Int. Rubb. Conf. Göteborg, vol. 1,
p. 171.
219. Gargani, L, Giuliani, G., Perego, G. and Zarretta, A. (1977) Proc. 6th Europhysics Conf. on
Macromolec. Physics. Varna, Bulgaria, p. 10.
220. Garmonov, I.V. (ed.) (1982) Phase-Aggregate State and Properties of Substances. Moscow:
164 REFERENCES
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
293. Hoffman, I.D. and Weeks, Y.Y. (1962) J. Chem. Phys., 37(8): 1722.
294. Hordgies, K.H. (1986) Kautschuk und Gummi Kunststoffe, 39(7): 599.
295. Hotge, J.M. and Huvard, G.S. (1983) Macrom., 16(3): 371.
296. 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.
REFERENCES 167
382. 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.
170 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 Phase-
Aggregate 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 175
565. 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
(Russian translation).
670. 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
REFERENCES 179
(in Russian).
704. 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).
180 REFERENCES
733. Vishnyakov, I.I. and Tulenev, A.I. (1984) Prepr. Int. Rubber Conf. Moscow, A39.
734. 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 181
762. Zakharov, N.D. (1978) Chloroprene Rubbers. Moscow: Khimiya Publishers, 271 pp.
(in Russian).
763. Zgaevsky, V. E. and Yanovsky, Yu.G. (2003) Dokl. Akad. Nauk SSSR, 392(5): 627
(in Russian).
764. Zhang, Z.-M. and Mark, J.E. (1982) J. Polym. Sci. Phys. Ed., 20(3) 473.
765. 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).
766. Zhurkov, S.N., Sanfirova, T.P. and Tomashevsky, E.E. (1962) Vysokomol. Soed., 4:196
(in Russian).
767. Ziabicki, A. (1967) Polymery zworygwa wielkocrasecz kowe, 12(9): 405.
768. Ziabicki, A. (1979) Abstr. 9th Europhys. Conf. Macromolec. Phys. Jablonna, p. 199.
769. Ziabicki, A. and Jareckii, L. (1976) Abstr. 5th Europhys. Conf. Macromolecules, Budapest,
p. 166.
770. Ziegel, K.D. and Romanov, A. (1973) J. Appl. Polym. Sci., 17: 1119.
771. Zlatkevich, L.Yu. and Nikolsky, V.G. (1971) Vysokomol. Soed., ser. B, 13(11): 810
(in Russian).
772. Zorina, N.M. (2006) Abstr. 23rd Symp. Rheology, Valdai, Russia, p. 56 (in Russian).
773. Zorina, N.M. and Bukhina, M.F. (1984) Vysokomol. Soed., ser. A, 26(9): 1855 (in Russian).
774. Zorina, N.M. and Bukhina, M.F. (2002) Abstr. 1st All-Russian Rubber Conference. Moscow,
p. 67.
775. Zorina, N.M. and Bukhina, M.F. (2002) Kauchuk i Rezina, No 6: 3 (in Russian).
776. Zorina, N.M., Bukhina, M.F. and Morozov, Yu.L. (2004) Abstr. Int. Rubber Conf. Moscow,
p. 108.
777. Zorina, N.M., Bukhina, M.F. and Morozov, Yu.L. (2006) Abstr. 23rd Symp. Rheology, Valdai,
Russia, p. 57 (in Russian).
778. Zorina, N.M., Bukhina, M.F., Morozov, Yu.L. and Kuznetsova, E.I. (2004) Kauchuk i Rezina,
No 6: 37 (in Russian).
779. Zorina, N.M., Rudenko, G.A., Voloshin, V.N. and Bukanova, N.N. (1988) Kauchuk i Rezina,
No 3: 10 (in Russian).
780. Zuev, Yu.S. (1980) Fracture of Elastomers under Conditions Typical for Service. Moscow:
Khimiya Publishers, 288 pp. (in Russian).
781. Zuev, Yu.S. and Degteva, T.G. (1986) Resistance of Elastomers under Service Conditions.
Moscow: Khimiya Publishers, 268 pp. (in Russian).
782. Zuev, Yu.S. and Stern, T.M. (1982) Kauchuk i Rezina, No 7: 34. (in Russian).
783. Zuev, Yu.S., Bukhina, M.F. and Astaf’ev, A.N. (1983) Mechan. Komposit. Mater., 2: 359
(in Russian).
784. Zuev, Yu.S., Bukhina, M.F. and Borschevskaya, A.Z. (1979) In Structure and Properties of
Polymer Materials. Riga: Zinatne, p. 111 (in Russian).
785. Zuev, Yu.S., Revazova, M.A. and Bukhina, M.F. (1969) Vysokomol. Soed., ser. B, 11(12): 898
(in Russian).
786. Zuev, Yu.S., Revazova, M.A. and Bukhina, M.F. (1971) Proc. Int. Rubb. Conf. 1969.
Moscow: Khimiya Publishers, p. 146 (in Russian).
787. 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 Characteristic time of crystallization, 67
Activity of filler, 45, 51, 93, 121 Chemical ageing, 3, 4, 64
Adhesive, 84, 101, 103, 115, 121, 145 Chemical crosslinks/junctions, 7, 50, 51
Aerosil, 89, 149 Cis-trans isomerization, 144
Alternating rubbers, 155, 156 Clapeyron–Clausius equation, 58, 82, 105,
Amorphous halo, 26 Clay, 45, 121
Amorphous phase, 61–64, 68, 74, 75, 93, Coefficient of thermal expansion, 12, 18,
100, 127, 145 21, 24, 25, 33, 37, 39, 40, 44, 54, 127,
Anisotropy, 62 130, 134
Annealing, 25, 126 Coefficient of volume thermal expansion,
Articles, see Components, see Elastomeric 24
articles, see Rubber goods, 1, 2, 32, 46, Compatibility, 34–37
52, 60, 72, 74, 75, 78, 81, 94, 103, 115, Complex dynamic modulus, 8, 93, 107
122, 129–135, 140 Components, see Articles, see Elastomeric
Axial fibrillae, 65, 80, 101–103, 106–108 articles, see Rubber goods, 1, 3, 8, 16,
22, 26, 32, 34–37, 44, 74, 80, 85, 89,
Belts, 132, 134, 135 94, 109, 122, 130, 131, 132, 134, 135,
Bend, 15 149, 150
Birefringence, 61, 81, 99, 107, 119 Composition of rubber compounds, 28, 83,
Block copolymers, 65, 72, 94 108
Breakdown, 47, 51, 57, 91, 109, 116–118, Compression, 6, 8, 15, 16, 26, 47–49, 55,
121, 122, 130, 134 57, 63, 80, 81, 93, 94, 120, 130, 132,
Bridge bearings, 130 146
Brittleness, 57, 115, 116, 123, 127, 151, Conformation of macromolecular chains,
152 3, 82, 84
Brittleness temperature, 57, 115, 116, 123, Conveyor belts, 132, 134, 135
127, 151, 152 Cooling rate, 11, 12, 18, 21, 25, 91, 134
Butadiene acrylonitrile rubber (BNR), 24, Copolymer, 1, 14, 30, 33, 34, 49, 65, 72, 73,
33–35, 139, 140, 145, 156 79, 81, 84, 87, 91–94, 97, 111, 126, 137,
Butyl rubber (BR), 45, 59, 97, 99, 100, 104, 139, 143, 144, 146–155
108, 119–121, 129, 138, 146, 147, 150 Covulcanization, 140
Crosslinks, 39, 44, 45, 50, 51, 85–88, 94,
Calorimetric method, 12, 74, 76 104, 111, 141
Catalyst, 140, 142, 143, 153–155 Crystal growth rate, 77, 83, 155
Chalk, 45, 46, 56, 114 Crystallizable plasticizers, 38, 91
Characteristic straight lines, 81, 88 Crystallization, 1, 7, 11, 26, 31, 34, 38, 40,
184 SUBJECT INDEX
41, 44, 47, 51, 52, 53, 58–156 Dynamic modulus, 8, 17, 39, 43, 51, 93,
Crystallization half-time, 66, 81, 84, 86, 98, 107
148, 150
Crystallization kinetics, 63, 74, 75, 81, 82, Ehrenfest equation, 26
88, 94, 97, 98, 141, 155 Elastic deformation, 3, 4, 8, 43, 48, 116
Crystallization prediction, 94 Elasticity modulus, 3, 55, 125
Crystallization rate, 11, 38, 59, 64, 66–79, Elastomeric articles, see Articles, see
83–87, 89, 91, 94, 95, 109, 119, 125, Components, see Rubber goods, 1, 2,
129, 141–156 32, 43, 75
Crystallization rate constant, 67 Electrical conductivity, 62, 63
Crystallization temperature, 67, 75–77, 85, Electrical properties, 12
88, 89, 93–95, 110, 143, 145, 147, 153 Electron microscopy, 61, 69, 70, 83, 87,
Crystals with extended chains, 64, 65, 81, 88, 101, 141
82, 101, 106–108, 117–121 Electron paramagnetic resonance (EPR),
Crystals with folded chains, 65, 82, 102, 14, 62
108, 119, 121 Elementary cell, 64, 100, 141, 155
Crystals with “unfolded” chains, 108, Elongation, 4, 16, 47, 55, 57, 80, 81, 98,
115–117, 121 102, 104, 107, 117, 119, 120, 122, 150
Cyclopentene, 154 Energy of interface formation, 69
Entanglements, 50
Deceleration of crystallization, 87, 90 Enthalpy, 63, 64
Decrystallization, see Steaming, 75, 78, Entropy, 3, 63, 64, 74, 105
133 141, 145 Equilibrium melting temperature, 58, 63,
Deformation components, 3, 16 64, 70, 72, 77–79, 82, 90, 92, 94, 95,
Deformation properties, 7, 46, 115, 116, 100, 106, 108, 133, 141, 143, 145, 147,
118 149, 153, 154
Deformation rate, 23, 116, 120 Ethylene-propylene rubber, 35, 51, 91–94,
Degree of crystallinity, see Degree of 97, 120, 152, 153
crystallization, 60, 61, 66, 83, 92, 123,
148 Filler, 25, 36, 38–40, 44–46, 50, 51, 67,
Degree of crystallization, see Degree of 73, 85, 87–90, 93, 95, 110, 112–114,
crystallinity, 11, 31, 60, 62, 63, 67, 84, 118, 119, 121, 122, 125, 134, 139, 140,
98, 99, 103, 104, 110, 113, 122, 145–147, 149, 150, 153, 155
125–127, 133, 141, 144, 145, 147, 150, First-order phase transitions, 28, 58
154, 155 Flory equation, 78, 87, 99, 104, 107, 143
Density of amorphous polymer, 102 Flory–Higgins constant, 90
Dicumyl peroxide, 7 Fluctuation, 7, 27, 50, 51, 68, 89, 90, 100,
Dielectric loss, 13, 27, 39, 57, 62 115, 128
Differential scanning calorimetry, 12, 62, Fluoroelastomer, 24, 94, 151
92 Folds, 65, 72, 80
Differential thermal analysis, 12, 62 Forced elasticity, 116, 127, 128, 130, 151
Dilatometry method, 18, 31, 35, 39, 41, 75, Form factor, 47, 49, 55, 57, 82, 118,
77 129–131, 151
Dimethyl siloxane rubber, 30 Fox–Flory equation, 28
Disulphide bonds, 86 Free volume, 21–26, 33, 36–40, 54, 55
Double peak of melting, 76, 113, 114
Dynamic mechanical characteristics, 50, Gaskets, 130, 132
51 Glass transition, 1, 11–44, 51, 52, 55–63,
SUBJECT INDEX 185
68, 70, 82, 91, 94, 115, 116, 125–134, 28, 59, 64, 65, 70, 72, 74, 106
137, 139, 140, 142, 144–156 Low-temperature resistance, see Freeze
Glass transition, conventional, 35 resistance, 1–3, 11, 15, 16, 26, 28, 34,
Glass transition, “chemical”, 32 35, 38, 40, 41, 44–48, 53–55, 59, 90,
Glass transition temperature, 11, 13–16, 94, 125, 128–156
18–22, 26–44, 51, 55–68, 82, 91, 94,
115, 116, 126, 128–133, 144, 147–156 Master curve, 52–54, 117, 118
Goods, see Articles, see Components, see Mastication, 121
Elastomeric articles, see Rubber goods, Maximum crystallization rate, 59, 70, 72,
1, 43, 48, 61, 135 84, 85, 91, 95, 119, 141, 143, 146, 147,
Grains, 65, 150 149, 150, 153, 155, 156
Gutta percha, see polyisoprene-1,4-trans Mechanical losses, 49–51
64, 77, 81, 83, 103, 140 Mechanical methods, 15, 27, 40, 43, 57,
60, 63, 89, 129
Hardness, 16, 52, 63, 93, 125 Mechanical properties, 3, 4, 9, 15, 17, 18,
Heat of fusion, 69, 78, 79, 92, 141, 142 28, 41, 43–60, 82, 89, 92, 104, 106,
Heating rate, 12, 14, 17, 18, 21, 39, 76, 77, 107, 114, 122, 125, 127, 129
79, 141 Melting temperature, 38, 58, 59, 63, 64, 68,
Heterogeneous nucleation, 69 70, 72, 74, 75–95
Homopolymer, 29, 30, 33, 34, 78, 79, 84, Melting thermograms, 76, 85, 114
94, 140, 142, 144, 147, 148, 151, 152, Mesophase/liquid-crystalline structure, 82,
154, 155 103, 148
Hoses, 122, 132, 134 Micellar model, 65
Hysteresis, 8, 18, 50, 75, 119, 134 Microfractionation of polymer in
crystallization, 109, 111
Identity period, 156 Microstructure of polymer, 30, 33, 50, 83,
Impact resilience, 17, 30, 50, 63, 127 91, 92, 108, 110, 125, 140, 142, 143
Incompatibility, 35, 44, 50, 84, 109 Monosulphide bonds, 39, 45, 86, 120, 143,
Induction period, 65, 66, 69, 80, 87, 93, 97, 144
112, 146, 147 Mooney–Rivlin strain energy function, 6
Internal energy, 3 Morphology, 15, 26, 27, 61, 64, 65, 80, 81,
IR spectroscopic method, 61, 111 88, 97, 100, 101, 104, 141, 145, 150
Isothermal crystallization, 62, 64–67, 73, Mullins–Patrikeyev effect, 47
75, 76, 86, 97, 100, 110, 112
Nairit, 70, 88, 90
Kargin–Malinsky equation, 36 Nanofillers, 40, 46, 89, 121, 122
Kinetic curves of crystallization, 66 Natural rubber (NR), 20, 24, 35, 44,
Kolmogorov–Avrami equation, 66, 67, 74, 51–53, 61, 64, 65, 70, 75, 80, 83, 84,
94 86, 88, 90, 97–102, 107–111,
Krigbaum–Roe equation, 104 119–122, 129, 134, 140, 142
Neck, 55, 56
Lamellae, 65, 70, 82, 83 Neoprene, 84
Light microscopy, 61, 101, 145, 150 Network, 3, 5–7, 32, 38, 39, 45, 46, 50, 51,
Light scattering, 61, 145 76, 84–87, 90, 94, 95, 101, 105,
Long-term low temperature 109–111, 113, 118, 120, 141–143,
behaviour/resistance, 38, 90, 128, 130 146, 149, 150, 155
Loss modulus, 17, 93 Network density, 38, 39, 45, 46, 50, 51, 76,
Low-molecular-mass substances, 11, 22, 85, 111, 149
186 SUBJECT INDEX
NMR method, 14, 38, 40, 60, 88, 93, 127, Polysulphide rubbers, see Thiocols, 39, 87,
145 121, 150, 151
Nonisothermal crystallization, 67, 73–75, Polyurethane, see Urethane rubber, 59, 61,
87, 89, 112 70, 90, 103, 121, 129, 138, 149, 150
Nucleation, 67–72, 74, 75, 77, 81, 83, 84, Precipitated silica, 40, 46, 149
87, 89, 90, 112, 145, 149 Prephase fluctuation, 68, 89, 94, 100
Pressure effect on glass transition, 25, 40
O-rings, 130 Pressure effect on melting, 58, 79
Orientation of crystals, 88, 113 Primary nucleation, 67, 68, 70, 77
Orientation of rubber, 97, 100, 112, 148 Principle of temperature–time
Overall crystallization rate, 67–70, 84, 88, superposition, 19, 23, 50, 52–55, 116,
90 132
Processing, 2, 4, 41, 60, 74, 80, 94, 115,
Payne–Warnaka effect, 51 122, 133, 134, 141, 142, 146, 153
Phase transition, 25, 26, 28, 58, 63, 92
Photoelasticity, 61 Quenching, 35, 101
Physical junctions, 7, 47, 50, 53, 68, 89, 91,
94, 100, 109, 111, 128, 133 Radiothermoluminescence (RTL), 13, 14,
Physical methods, 43, 60, 62, 68 35, 57, 62
Plastic deformation, 3, 4, 109 Raw rubber compounds, 3, 4, 44, 114, 149,
Plasticizers, 36–39, 45, 50, 54, 89–91, 155
134, 139, 140, 145, 149 Recovery, 16, 35, 41, 43–48, 51–54, 63,
Plasticizing, 38 79–81, 93, 94, 104, 127, 131, 132, 154
Polybutadiene, 1,4-cis, 31, 32, 33, 35, 40, Reference temperature, 20, 53, 54, 117
50, 51, 59, 66, 70, 76–78, 80, 84, 85, Relaxation properties, 16, 40, 52, 68, 91,
90, 103–105, 107, 108, 110, 111, 119, 93, 125
121, 123, 129, 141–144 Rigidity, 3, 24, 43, 46–48, 81, 94, 122,
Polybutadiene, 1,4-trans, 31 125, 130, 131, 149
Polychloroprene, 15, 38, 39, 59, 61, 62, Rubber blends, 34– 38, 44, 45, 50, 79, 81,
64, 65, 70, 77, 81, 83, 88 –90, 100, 84, 85, 121, 144, 145, 155
101, 103, 105, 113, 120, 121, 122, Rubber goods, see Articles, see
127, 129, 130, 134, 145, 146, 155 Components, see Elastomeric articles,
Polycrystals, 61, 65, 69, 80, 81, 87, 90, 122 48, 61, 135
Polyethylene, 11, 30, 38, 64, 78, 82, 89, 91, Rubber/fabric goods, 132, 134
92, 101, 103, 106, 109, 125, 126, 149, Rubber/metal goods, 130, 132, 134
153 Rubberlike deformation, 3–5, 8, 43, 46,
Polyisobutylene, 25, 35, 146 47, 55, 58, 116
Polyisoprene, 1,4-cis, 28, 41, 59, 65, 69, Rubberlike plateau, 47, 51, 63, 68, 116,
72, 78, 79, 81–85, 88, 101, 102, 122, 133
108–111, 113, 119–121, 129, Rupture envelope, 117
140–142, 144, 145
Polyisoprene-1,4-trans, see gutta percha, Seals, 81, 130–132, 135
83, 140 Secondary crystallization, 66, 154
Polypropylene, 11, 38, 91, 125, 126 Secondary nucleation, 68, 70, 71, 77
Polysiloxane, see Silicon rubber, 38, 40, Second-order phase transition, 25, 28, 58
61, 78, 83, 84, 85, 89, 99, 100, Shear, 3, 5, 6, 8, 15, 17, 20, 80, 130
104–106, 123, 127, 129, 148, 149 Shift factor, 19, 52, 117
Polysulphide bonds, 86, 87, 144 Shish-kebab structure, 80, 101, 102
SUBJECT INDEX 187