Вы находитесь на странице: 1из 22

Molecular Theory of Solutions and Blends of

Heteropolymers. I. Thermodynamics of Amorphous


Multicomponent Polymer Systems
S. I. KUCHANOV, 1 S. V. PANYUKOV 2
1

Keldysh Institute of Applied Mathematics of Russian Academy of Science, Moscow 125047, Russia

Theoretical Department, P. N. Lebedev Physics Institute, Russian Academy of Sciences, Moscow 117924, Russia

Received 1 October 1996; revised 3 November 1997; accepted 5 November 1997

ABSTRACT: On the basis of a variational principle a quantitative theory is developed

enabling a thermodynamic description in terms of mean field approximation of heteropolymer mixtures of macromolecules with an arbitrary distribution for both degree of
polymerization and composition. Rather simple general equations are derived to calculate compositions and volume fractions of spatially homogeneous macroscopic phases
as well as to find the cloudpoint curve, spinodal, and critical points. Potentialities
of general theory are illustrated for copolymers synthesized by traditional methods.
q 1998 John Wiley & Sons, Inc. J Polym Sci B: Polym Phys 36: 937958, 1998

Keywords: thermodynamics; heteropolymers; phase diagram

INTRODUCTION
When elaborating the molecular theory of solutions and blends of actual polymers some essential peculiarities, qualitatively distinguishing
them from traditional systems containing lowmolecular weight compounds, have to be taken
into account.
The first such peculiarity is that each macromolecule possesses its specific chemical structure
G. The latter is unequivocally characterized by its
structural formula reflecting the fashion in which
monomer units are arranged in a given macromolecule. Even for homopolymers there is a possibility of a variety of different structural isomers (regio- or topological ones, for instance), so that the
degree of polymerization does not, generally
speaking, completely characterize the chemical
structure of a homopolymer macromolecule. It is
obvious that the latter could be described as a
copolymer molecule consisting of several types of
monomer units.
Correspondence to: S. I. Kuchanov
Journal of Polymer Science: Part B: Polymer Physics, Vol. 36, 937958 (1998)
q 1998 John Wiley & Sons, Inc.
CCC 0887-6266/98/060937-22

The second important feature peculiar to actual


polymers is that particular specimens consist, as
a rule, of an enormous (practically infinite) set
of different chemical compounds whose molecules
differ in the degree of polymerization, composition, and chemical structure. So, each polymer
specimen is characterized by its Molecular Structure Distribution (MSD) n(G), i.e., by the number
of macromolecules with given the chemical structure G. The specific form of the MSD of a polymer
specimen depends, obviously, on the conditions of
its synthesis. Problems of the calculation of this
distribution and its statistical characteristics may
be solved by different methods elaborated for
main processes of polymer synthesis (free radical
polymerization, polycondensation, macromolecular reactions).1 All such problems of the statistical
chemistry of polymers will not be discussed in the
present and further articles of this series, and are
taken to be already solved.
The main purpose of the authors is to develop a
general approach to calculate thermodynamic and
correlation characteristics of polymer solutions
allowing for their MSD. These approaches, relevant
to the statistical physics of polymers, usually an937

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

polpa

W: Poly Physics
9610001

938

KUCHANOV AND PANYUKOV

swer important questions concerning the role of the


chemical structure of macromolecules for the dissolution or mixing of polymers. Despite the high degree of universality of the theory, its final results
are presented below as simple expressions suitable
for treating experimental data.
In this article we focus only on the consideration of statistical thermodynamics of polymer
systems with spatially homogeneous macroscopic
phases, having in mind to discuss the formation
of mesophases in subsequent publications. This
article is organized as follows. A short review devoted to the evolution of theoretical approaches
to the thermodynamics of amorphous polymers is
followed by the section where a rather general
variational principle is proposed for a thermodynamic description of such polymers with allowance for their chemical inhomogeneity. Proceeding from this principle within the framework of a
mean field approximation, we derive the expression for the free energy of a polymer system comprising macromolecules distributed arbitrarily for
numbers of different type monomer units. For this
system the conditions are presented of phase equilibrium and the equations are derived to determine the cloudpoint curve, spinodal, and critical
points. The last section of this article contains the
results of the application of general theory to the
calculation of thermodynamical characteristics of
the type of copolymers of utmost importance
among those synthesized by traditional methods.

APPROACHES DEVELOPED IN
THERMODYNAMICS OF POLYMER
LIQUIDS
The theory of polymer solutions goes back to the
fundamental studies carried out independently by
Flory 2 and Huggins.3 They obtained a simple formula for the entropy of a thermal mixing of a monodisperse polymer with low molecular liquid. Having
added to this contribution of combinatorial entropy
the heat of mixing, which in terms of the lattice
model takes into account the energy of pairwise contacts between monomer units and solvent molecules, they obtained a well-known expression for the
free energy of a polymer solution. Further progress 46 in the FloryHuggins theory was associated
with an account for polydispersity of macromolecules and the selection of more complicated dependencies of the enthalpy of mixing on concentration
and temperature. This theory was extended later
to the solutions of homopolymer blends, 5 where the

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

expression for the Gibbs free energy of mixing per


one lattice site DG has the form

DGv
Fa(la)

ln Fa(la) / xabFaFb ,
TV
la
a la
ab
Fa Fa(la), Fa 1
la

(1)

Hereinafter Fa(la) and Fa are volume fractions of


the a-th component containing la units and that
of a-th type units Sa , respectively, xab represents
a binary interaction parameter of the pairwise interaction between such units of the a-th and b-th
type, T is the absolute temperature multiplied by
Boltzmanns constant, while v and V stand for the
volume of one lattice cell and the whole system.
The index a in expressions (1) can refer either
to the polymer or to the solvent for which Fa(la)
Fad (la , 1), where d (la , 1) is the Kronecker delta.
Due to the reasonable character of the assumptions underlaying the FloryHuggins theory, and
mainly due to the remarkable simplicity of its final formulas, this theory up to now remains
widely popular among investigators dealing with
experimental data treatment. However, the above
theory, in principle, fails to provide an adequate
explanation for a number of experimentally discovered important regularities, such as the presence in some polymer solutions and blends of the
lower critical temperature of mixing, the possibility of the heat evolvement in the course of the
latter as well as the dependence of the parameters
xab on the concentration. The most serious shortcoming of the FloryHuggins theory is, apparently, a disregard of the volume change occurring
at dissolution.
This effect is taken into account in the framework of molecular theories of polymer solutions proceeding from their equation of state. Such approaches peculiar to the second stage of the development of statistical thermodynamics of polymer
systems seem to originate from fundamental studies by Prigogine and co-workers, 79 who, for the
first time, demonstrated the possibility of applying
the principle of corresponding states to polymer liquids. Further refinement 1013 of the Prigogines theory enabled to predict the existence of upper and
lower critical temperature, to explain the phenomenon of volume change of polymers on their mixing,
as well as to elucidate some other experimentally
observed regularities. The main difficulty hampering the application of these approaches for the
quantitative description of particular systems con-

polpa

W: Poly Physics
9610001

MOLECULAR THEORY OF HETEROPOLYMERS. I.

sists in the necessity to previously find for them


the values of reduction parameters occurring in the
theory of corresponding states.
This difficulty seems to a great extent to be
obviated in two subsequent versions of the theory
of polymer solutions, based on the equation of
state, i.e., the new Florys theory 14 17 as well as
the lattice liquid theory.18 22 Free energy of the
system, in accordance with both mentioned theories, can be divided into two groups of items, the
first of which comprises all items taking into consideration combinatorics of transposition of the
molecules in the solution. The items of the second
group allow for various pairwise interactions between monomer units and molecules of the solvent containing energetic parameters, which
characterize both attraction and repulsion between these constituents. The distinction in equations of state, derived in the framework of the
two mentioned theories, resembles that between
equations of state of the Van der Vaals and the
lattice gas. The reason for such a distinction is
the difference in approximations used for the account of the repulsion at short distances, i.e., of
the account of excluded volume. Written down in
a compact matrix form formulas for both theories
of polymer solutions based on the equation of state
are presented in ref. 23.
Later, Florys new theory underwent a number
of modifications mainly connected with due regard for inhomogeneity of mixing 24 26 and some
other factors.27,28 Detailed analysis of peculiarities
of different models based on the equation of state
can be found within the reviews listed as refs. 29
33. Up to now studies, devoted to further modifications of this approach happen to appear in literature (see, for instance, refs. 34, 35).
The construction of phase diagrams for equilibrium systems proceeding from their known thermodynamic potential is regarded to be one of the
crucial thermodynamic problems. First results in
this direction have been reported for multicomponent systems with participation of polymers about
40 years ago in pioneer articles by Scott, 36,37
Tompa, 38,39 and Stockmayer.40 In these studies,
where the expression for free energy obtained in
terms of the FloryHuggins formalism is applied,
authors presented for some of the simplest systems
conditions of phase coexistence as well as expressions for the spinodal, critical points, and osmotic
pressure. Besides, the possibility of the appearance
in the phase diagram of coexistence area of three
liquid phases has been theoretically predicted.36,38
For the first time this possibility was experimen-

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

939

tally verified by Koningsveld.41 The series of works


with his participation, devoted to theoretical and
experimental investigation of thermodynamics of
multicomponent polymer solutions, 4247 is undoubtedly of first-rate importance for physical chemistry
of polymers. Remarkable contribution to this
branch of polymer science has also been made in
collaboration with Koningsveld 4850 by Gordon and
co-workers, 4853 who, particularly, considered the
influence of polydispersity of polymers upon thermodynamic characteristics of their solutions. To
take this factor into consideration the method called
continuous thermodynamics has been laid down
(see articles 54,55 and references therein). Within the
context of this treatment the molecular weight of
macromolecules in the solution or polymer blend
is considered as a continuous variable, so that the
operations of summing in expressions for thermodynamic characteristics is replaced by integrating
over this variable.
A similar approach was employed earlier by
Solc 56,57 when theoretically considering the cloud
point curves for solutions of polydisperse homopolymers. Rigorous thermodynamic consideration
of multiphase equilibrium in such systems 58 61 allowed him to make a number of very interesting
conclusions concerning possible mechanisms of
phase separation under the variation of external
parameters. In particular, the correctness of the
three-phase separation mechanism proposed by
Tompa 38 for a three-component system as well as
the groundlessness of the criticism62 of this mechanism have been convincingly proven.58
Of great importance for the thermodynamics of
polymer solutions is the consideration of the topology of their phase diagrams, which implies the determination of the type of each critical point as well
as the localization of specific lines and surfaces
within the phase space of external parameters. As
the most significant in this field, articles should be
mentioned by Solc, 5161,6366 who thoroughly studied binary systems along with ternary ones, 6366
containing the low molecular solvent and two chemically different polymers. No such comprehensive
theoretical analysis, to the author knowledge, has
been performed so far for analogous polydisperse
systems, although the expressions for the spinodal
and critical points of such quasiternary systems
have been repeatedly reported.49,67,68
While there is an appreciable progress in theoretical description of solutions and blends of
homopolymers, the quantitative theory of heteropolymer systems is still to be developed. The main
obstacle that an explorer encounters elaborating

polpa

W: Poly Physics
9610001

940

KUCHANOV AND PANYUKOV

such a theory is that macromolecule of a heteropolymer, unlike that of a homopolymer, cannot


completely be characterized by only one number,
i.e., its molecular weight. To describe chemically
a heteropolymer molecule in an exhaustive manner it is indispensable to indicate numbers of all
its units of each type (composition) as well as the
pattern of their connection (structure). Samples
of actual polymers contain assemblies of macromolecules differing in composition and structure.
A natural question arises in this respectwhich
particular statistic characteristics of this assemble prescribe thermodynamic behavior of solutions
and blends of heteropolymers. Up to now, as it follows from the data reported in reviews, 6972,47
there is no rigorous molecular theory that answered this question on the quantitative level,
even if the phenomenon of microphase separation,
typical for block copolymers 73 75 is left out of consideration. The above phenomenon, calling for
some special theoretical treatment, is supposed to
be taken in our next article.
Scott 76 seems to be the first who examined the
problem of thermodynamic compatibility of copolymers. He noticed that for such a compatibility,
sufficiently narrow composition distribution of
macromolecules in the sample is indispensable.
Almost 30 years later, approaching the same problem, Leibler 77 managed in the framework of the
FloryHuggins formalism to obtain the expression for the spinodal and critical point for the
blend of three fractions of the copolymer with different composition.
The commencement of a new stage of quantitative description of the copolymer thermodynamics
was marked in articles, 78 80 examining compatibility conditions of binary blends of a homopolymer with a two-component copolymer 78,80 or mixture of two fractions of: (a) one such copolymer, 78,80 (b) two copolymers containing one
identical type of monomer units, 80 or (c) two copolymers with different pairs of monomer units.79
In subsequent works 81 85 the analogous consideration was carried out for ternary blends of: (a)
binary copolymers, 81 (b) two homopolymers and
a copolymer composed of their monomer units 82
as well as a binary blend of terpolymers with different compositions, obtained from the same triple of monomers.84 Koningsveld and Kleintjens 81
put forth a rather general, in terms of the lattice
model, expression for free energy of mixing of low
molecular solvent with an arbitrary number of
types of binary copolymer macromolecules differing in the degree of polymerization and the

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

composition. This intriguing approach enabled


them to take into account such important factors,
influencing conditions of phase separation, as copolymer composition inhomogeneity and the distinction in dimensions of monomer units and molecules of the solvent.
The case of arbitrary blend of copolymers was
examined by Paul and co-authors, 8385 who made
use of the algorithm, permitting them to write down
the simple expression for the heat of mixing DHm .
The essence of this algorithm, applied for the calculation of DHm within the mean field theory formalism, consists of the necessity of subtracting the certain term from traditional one [i.e., the second item
in the right-hand part of formula (1)]. This term
equals the heat of the formation of copolymer molecules from monomer units 83 and allows for intramolecular contacts between different type units. The
same expression for DHm can be put down as a
quadratic form with respect to volume fractions of
macromolecules of different types. Coefficients of
this form, having the sense of the binary interaction
parameters for macromolecule interactions, are
readily expressed through analogous parameters of
monomer units.85
Another approach to the description of the liquidliquid equilibrium in solutions and blends of
polydisperse random copolymers has been proposed
by German scientists 8690 in terms of the continuous thermodynamics concept. They have analyzed
the effect of polydispersity of macromolecules on
the cloudpoint curve, shadow curve, the spinodal,
and critical points as well as having discussed the
feasibility of the existence of three-phase equilibrium regions. In spite of an impressive volume of
thermodynamic information available in articles, 8690 its practical value remains vague because
of an arbitrary character of the expression for free
energy of mixing of polydisperse copolymers, which
authors applied in both the primary 86,87 and
later 8890 version of their theory.
Further development of thermodynamic theory
of solutions, melts, and blends containing copolymers is consistent with the allowance for freevolume effects. For this purpose they normally
use (see, e.g., articles 91 95 ) different extensions
of the equation-of-state approaches, considering
these effects in terms of either lattice models or
off-lattice (continuous space) models. In the first
case, the models slightly differing in details are
based on some version of compressible lattice (cell
model, hole model, cellhole model). Along with a
number of indisputable advantages, these models
exhibit one very serious shortcoming because they

polpa

W: Poly Physics
9610001

MOLECULAR THEORY OF HETEROPOLYMERS. I.

leave out the consideration of the distinction in


size and shape of different type monomer units.
The above difficulties are possible to overcome using the Lattice Cluster Theory (LCT) introduced
by Freed and Dudowicz.96 99 Their generalization
of the traditional lattice model permits a single
monomer unit to occupy several neighboring lattice sites. Within the framework of such an extension of the hole model it becomes possible to elaborate a theory taking into account both compressibility of the polymer system and differences in
shapes of the monomer units.
The same goal is attainable also by means of
approaches using current concepts of the theory
of low molecular liquids. A key role is played here
by an off-lattice microscopic statistical mechanical approach developed by Schweizer and Curro
(see ref. 100 and references therein) for calculating thermodynamic characteristics of polymer
liquids. The general methodology is known as
the Polymer Reference Interaction Site Model
(PRISM) theory, and is based on RISM approach
for low molecular liquids.101 Under such an approach some equations have been derived for
which, upon suitable closure approximation, either analytic 102,103 or numerical 104 solutions have
been found for model binary polymer blends. The
analytic PRISM theory has been recently extended by Schweizer 105 to the description of thermodynamics of random copolymer alloys.
A very promising approach to the description
of equilibrium polymer system, introduced by
Prausnitz and co-workers (see ref. 106 and references therein), suggests considering as a reference system some simple model, permitting one
to take into account as accurately as possible the
excluded volume of monomer units, and then to
add a mean field perturbation term allowing for
the attractive interactions. The Prausnitz central
idea is to use as a reference system an a thermal
ensemble of hard-sphere chains, each possible to
picture as a series of freely joint tangent hard
spheres. This model has been much studied, 107
and despite its simplicity, takes into account some
significant features of real polymers such as excluded volume of units and its connectivity in
chain. A Perturbed Hard-Sphere Chain (PHSC)
theory appears rather prospective when performing thermodynamic calculations of actual
polymer systems because it allows for in a natural
way distinctions in size of different type monomer
units 108 as well as because this theory input parameters (the effective hard-sphere diameters
and attractive energy of pair interactions) are

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

941

theoretically based functions of temperature. An


extension of the PHSC equation of state was presented 109 for two types of copolymer mixtures for
which theoretical cloudpoint curves and miscibility maps were computed. The efficiency of the
employment of the PHSC theory in thermodynamics of blends with participation of copolymers
has been supported in recent articles by Prausnitz
with co-workers.110
One more interesting formalism111116 merits special attention, enabling the explanation within the
context of the mean field approximation of the effect
of the sequence distribution character in copolymers
on the compatibility of their blends. This approach
implies the necessity, when calculating a number of
binary contacts between monomer units, to distinguish their pair interaction parameters, depending
on types of units neighboring with a given one. To
make final thermodynamic formulas suitable for
their practical application, authors of refs. 111116
assumed the majority of these parameters to be
equal to each other and practically confined the consideration to copolymers with the Markov statistics
of sequence distribution characterized by the sole
parameter u. For many actual systems this approximation seems to be good enough because it permits
one varying the value of u to embrace alternating,
random, and block copolymers. As in the above-cited
articles, 7885 the enthalpy of mixing is defined by
effective values of molecular parameters of pair interactions that comprise, however, two groups of
items. Those constituting the first group take into
account composition distribution of copolymer macromolecules in a blend and look just like the parameters appearing in the previous approach.7885 The
addition of the second group of items, taking into
consideration the character of the sequence distribution in monodisperse copolymers, enables the interpretation of known experimental data.
Proceeding from brief analysis of voluminous
literature cited above and recent reviews and
monographs 47,69 72,117 126 it is possible to come to
the inference that sufficiently rigorous general
molecular theory of solutions and blends of heteropolymers, taking a comprehensive account of
their molecular structure characteristics, is not
available so far. The present article is the first
in the series of the intended articles where the
authors attempt, to a certain extent, to make up
this deficiency. The theory we propose, being of
rather universal character, by no means refutes
and even, on the contrary, does accumulate main
ideas of the aforementioned generally accepted
approaches. Elaborating on this theory, the au-

polpa

W: Poly Physics
9610001

942

KUCHANOV AND PANYUKOV

thors tried to adjust it for the description of as


great variety of different systems as possible. On
the other hand, when selecting molecular models
they always preferred those that could lead to final formulas simple enough to be applied for the
treatment of experimental data. Such an approach allows one to follow all approximations
used for obtaining final results as well as to estimate possible areas of their applicability for the
description of particular systems.

VARIATIONAL PRINCIPLE
In accordance with general ideas of statistical
physics volume interactions between monomer
units in the framework of the mean field approximation are normally taken into account via introduction of a certain external field h(r). Each of
its component ha(r) ( a 1, . . . , m) acts upon
units Sa in the space point r forming certain density distribution ra(r) of such units. Further, by
means of the self-consistence condition another
relation between h(r) and r(r) is established,
which leads to a closed set of equations, enabling
the calculation of these distributions.
The present article deals with the situation
when scales of spatial change of both density and
field are large enough compared to the geometrical dimensions of the macromolecules. In this
case, it is possible to consider them as 0-dimensional points by assuming all Sa units of any macromolecule to be acted upon by the same fields
ha(r). Such a consideration implies that external
field contribution to the energy of the macromolecule located in the point r
m

h(l; r) ha(r)la h(r)l

(2)

ing n(l) by overall number P of molecules we obtain the distribution of copolymer macromolecules
for size and composition (SCD) generally accepted
in polymer chemistry and calculated for many
types of copolymers.1 In the present article conditions of macrophase separation occurring in solutions and blends are found for polymer systems
with arbitrary SCD.
Because macromolecules are thought of here as
points, they are possible to consider in terms of
classical thermodynamics as independent components, each being characterized by a certain value
of vector l. The joint distribution of concentrations
C(l; r) of these components within the Euclidean
space describes, generally speaking, the nonequilibrium macroscopic state of the system under examination. To this state a point corresponds in
the function space where one, following the Leontovich principle, 127 can introduce functional F [C]
of the Helmholtz free energy. This functional under the fixed temperature T, system volume V,
and given SCD reaches constrained local minimum at equilibrium distribution C(l; r) c(l; r).
The problem of finding the latter is equivalent to
that of finding free minimum of other functional

L [C] F [C] / PV 0 m(l)n(l)

where the pressure P and chemical potentials m(l)


conjugate, respectively, to V and n(l) are introduced as the Lagrange multipliers. Employing
definitions

F [C]

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

* f (r; [C])d r,
3

n(l)

* C(l; r)d r
3

(4)

one can write down the functional (3) as follows

a1

is characterized only by numbers la in this macromolecule of monomer units Sa and does not depend on the pattern of their connection. In this
case, a detailed chemical structure of a polymer
molecule proves to be unessential, so that it is
enough to differentiate them by values of vector
l, defining the chemical size (the degree of polymerization l l1 / rrr / lm ) and composition ( j1
l1 /l, . . . , jm lm /l) of such a molecule. Thus,
in the framework of the approach under consideration a polymer sample appears to be exhaustively characterized by the set of number n(l) of
molecules with given value of the vector l. Divid-

(3)

L [C]

* [ f (r; [C]) / P 0 m(l)C(l; r)]d r


3

(5)

Equilibrium distribution of concentrations c(l;


r) should be searched among the functional (5)
extremals, which are, by definition, solutions of
the equation

dL [C]
dF [C]

0 m(l) 0
dC
dC(l; r)

(6)

Solving this equation with respect to C(l; r) it is


possible using normalization conditions (4) to find

polpa

W: Poly Physics
9610001

MOLECULAR THEORY OF HETEROPOLYMERS. I.

the relation between m(l) and n(l), in that way


excluding chemical potentials of molecules.
Among extremals found, the functional (5)
reaches local minimum only at those that provide
the positive definiteness of the quadratic form

l=

l0

** K (l*, l9; r*, r9)


1 DC(l*; r*) DC(l9; r9)d 3r*d 3r9 (7)

at any deviations DC(l*; r*), DC(l9; r9) of the


distribution of concentrations C(l; r) from the extremal c(l; r). The function K (l*, l9; r*, r9) represents the second-order variational derivative
d 2L [C]/ dC(l*; r*) dC(l9; r9) of the functional (5)
calculated at the extremal. Necessary and sufficient condition of positive definiteness of the quadratic form is the absence of negative eigenvalues
in the spectrum of integral operator with the kernel K (l*, l9; r*, r9). The calculation of functional
F [C] values at functions C(l; r) c(l; r) where
it reaches local minimum

F [c] F m(l)n(l) 0 PV

(8)

permits one to find values of free energy F for all


metastable states of the system and thermodynamically stable state, for which the minimum of
the functional F is the deepest.
The density of the Leontovich unequilibrium
free energy (4) for the approximation in hand can
be presented (see Appendix A) as the sum of two
items
f (r; [C]) fCB (C(l; r)) / f *( r(r))

(9)

describing the systems of Chemical Bonds and


Separate Units, respectively.128,129 The first of
them is the ideal gas of macromolecules, defined
by the well-known expression

943

The second item in the right-hand part of expression (9) is equal to the density of free energy
f * fSU 0 fIG of the Separate Units system fSU
minus analogous magnitude fIG for the ideal gas
with the same values of densities r . The form of
the function f *( r) depends on the choice of the
model of liquid of Separate Units and is not of
critical importance for the theory introduced.
Here, free energy density f * as well as chemical
potentials ma* and pressure P*

m*a

f *
r *a

, P* m*a ra 0 f *

(11)

are functions of temperature T and densities r(r)


of monomer units

ra(r) C(l; r)la

(12)

considered to be known from the theory of low


molecular liquid. When calculating thermodynamic parameters of particular systems one can
choose as functions f *, m *a, and P* those presented
in Appendix B.

HOMOPHASE SYSTEMS
Spatially homogeneous equilibrium distribution
of concentrations C(l; r) c(l) is always an extremal of the functional (3). For this distribution the
expression for chemical potential obtained from
eq. (6) has a simple form

m(l) 0T

f * ra
s
/
c(l)
ra c(l)
a

T ln n(l) / m*a ( r )la / min (l) (13)


a

The equation of state of the system under consideration

fCB (C(l; r)) fin 0 Tsc ,


sc 0 C(l; r)ln
l

P m(l)c(l) 0 f T c(l) / P*( r) (14)

C(l; r)
(10)
e

and the expression for its entropy density


which, along with the contribution fin (A.7) to the
free energy of the internal degrees of freedom of
macromolecules (responsible for their conformation isomerization) also takes into account combinatoric (i.e., translational) entropy, whose density is sc .

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

s0

f
sC / s* / sin ,
T
s* 0

polpa

f *
fin
, sin 0
(15)
T
T

W: Poly Physics
9610001

944

KUCHANOV AND PANYUKOV

are also easy to derive. The expressions for chemical potentials of molecules, pressure, and entropy
density (1315) are connected by well-known in
thermodynamics the GibbsDuhem equation 130
sdT 0 dP / c(l)dm(l) 0

(16)

Substituting into this formula the expression for


the heat function (enthalpy) H and the entropy
S
H
1

TM F

gef 2
F
T

0ln(1 0 F ) 0 F 0 2

Applying formulas of lattice fluid model of lowmolecular liquid to define P* we obtain from eq.
(14) the equation of state of polymer liquid

/Y/
n(l)
1
n(l)
S
0
ln
0 [(1
M
M
M
F
l

Pv
g ef 2
0ln(1 0 F ) 0 F 0
F / Y F (17)
T
T
containing as parameters, along with v, the effective energy of pair contacts of units g ef as well as
the number P of all molecules of the system reduced to overall number of units M

g ef gab Xa Xb Y
a

P
n(l)

(18)
M
M
l

Remarkably, expression (17) appears to be quite


similar to that derived by Sanchez and Lacombe
in the framework of the Lattice Fluid theory, 19 22
provided parameters n *ab and e a*b of their equation
are substituted for v and gab , respectively. Unlike
the simplified version of the Lattice Fluid model,
whose formulas are presented in Appendix B, the
SanchezLacombe treatment enables the correct
account of the distinction in values of excluded
volume of different type units. Under such a treatment the v is usually written as quadratic form
with phenomenological parameters as coefficients. No objective difficulties arise if for thermodynamic calculations instead of relations presented in Appendix B for thermodynamic potentials of the Separate Units system one will use
corresponding expressions of the SanchezLacombe theory, which takes more an accurate account of the excluded volume effect. However, for
the simplicity sake we will confine our examination in this article to the simplest version of a
lattice model containing as few adjusting parameters as possible.
When dealing with thermodynamics of mixing
of polymers it is convenient to employ the Gibbs
free energy G(P, T )

0 F )ln(1 0 F ) / F] 0 Y ln

/ 8Q4F$$0001
8Q4F

n(l)
G
n(l)
g ef

ln
02
F
TM
M
M
T
l
0 ln(1 0 F ) / Y ln F /

Gin
(22)
TM

which, along with the equation of state (17), defines in an implicit way the function G(P, T ). It
is easy to prove that expression (22) coincides in
the particular case of a blend of homopolymers
up to unessential linear in n(l) items with the
expression derived earlier by Sanchez and Lacombe 19 22 in a distinct way.
Because thermodynamic potentials are normally referred to a standard state we will also
write down expressions for the enthalpy DH and
the entropy DS of mixing

DH PDV 0 n(l)l gab[ FXaXb


l

ab

0 F(l) zazb] (23)

DV v n(l)l 1 0
l

S
FS

DS 0 n(l)ln
l

/ n(l)l
l

polpa

1
F(l)

D
D
S D

(24)

n(l) F
M F(l)

1
0 1 ln(1 0 F(l))
F(l)

(19)

02-14-98 16:01:42

F Sin
/
(21)
e
M

found in terms of the simplest model of low-molecular liquid (see Appendix B) we will obtain the
relationship

0
G H 0 TS

Hin
(20)
TM

1
0 1 ln(1 0 F )
F

W: Poly Physics
9610001

(25)

MOLECULAR THEORY OF HETEROPOLYMERS. I.

The reduced density F (B.3) of the blend with


given values of number of molecules Y (18) and
their average composition X (B.2) can be found
from the equation of state (17). For the melt of
the pure component characterized by the vector l
lz the value F(l) in expressions (23) (25) can
be calculated using the same eq. (17), where Y
and X should be replaced, respectively, by the reciprocal degree of polymerization 1/ l and by vector z of composition of macromolecules of this particular component.

F
F

n(l)exp 0
c i (l)

1
( m*a ( r i )la
T a

G
G

1
( V exp 0 ( m*a ( r j )la
T a
j 1
j

945

(30)

Substitution of concentrations c i (l) (30) in expression (29) and stoichiometry conditions

r ai c i (l)la (i 1, . . . , r; a 1, . . . , m) (31)
l

MULTIPHASE SYSTEMS
Because of the local character of free energy density (9) only stepwise distribution C(l; r) can be
an extremal of functional (3). Each such function,
C(l; r), is exhaustively characterized by the number r of steps (phases), their volumes V i and amplitudes C i (l) (i 1, . . . , r). Thus, functional is
known to reduce to the function of these independent arguments C i (l) and V i
r

L ({C i (l), {V i }}) V i[ f (C i (l))


i1

/ P 0 m(l)C i (l)] (26)


l

Conditions for calculating extremals of this function


L
L
0,
0 (i 1, . . . , r) (27)
C i (l)
V i
result in expressions

m(l) T ln c i (l) / m*a ( r i )la

yields a closed set of (m / 1)r equations with


respect to the same number of variables r ai and
V i . Between these internal parameters there are
m stoichiometry conditions
r

r ai V i Ma n(l)la ( a 1, . . . , m) (32)
l

i1

that ensue from formula (31) and might, if necessary, verify the correctness of the calculation accomplished. Values r ai and V i calculated in such
a way permit one to find via expression (30) all
extremals of function L (26).
The next problem to be solved consists in the
necessity to separate out those extremales in
which this function shows a local minimum. Because, according to eq. (26), the function in hand
represents the sum of contributions of the r
phases, positive definiteness of quadratic form (7)
for each phase with distribution of concentrations
c(l), known from the solution of eqs. (30) and
(31), is necessary and a sufficient condition of
local minimum of the multiphase system. Therefore, this problem obviously consists in analyzing
the spectrum of eigenvalues of matrix with elements

(i 1, . . . , r) (28)
P T c (l) / P*( r )
i

K (l*, l9)

1
2 f
T C(l*)C(l9)

testifying the equality both of chemical potentials


of molecules (13) and pressure (14) in all r coexisting phases.
By means of relationships (4) and (28) it is
possible to exclude the m(l) and to express concentrations of molecules in each phase through their
volumes and densities of monomer units they contain

/ 8Q4F$$0001

Cc

(29)

8Q4F

02-14-98 16:01:42

d (l* 0 l9)
0 Cabl*a l9b
c(l*)
ab

(33)

where d (l* 0 l9) is the Kronecker delta symbol,


whereas elements Cab of the matrix C of direct
correlation functions 131 are defined by expression

Cab 0

polpa

1 m*a ( r)
1 2 f *( r)
0
T rb
T rarb

W: Poly Physics
9610001

(34)

946

KUCHANOV AND PANYUKOV

Because the matrix (33) is symmetrical K (l*, l9)


K (l9, l*), all its eigenvalues lk are real and can
be enumerated in order of their increasing l1
l2 rrr . Because we are interested to know
only the sign of principal eigenvalue, l1 , it is sufficient instead of the matrix (33) to consider a far
more simple one
M k 01 0 C

1
n(l)lalb
V l

(36)

Despite the fact that the spectra of matrices (33)


and (35) are quite different, the number of their
negative eigenvalues, as can be readily proven, is
the same. This enables one to reduce the analysis
of positive definiteness of quadratic form to specifying the sign of the matrix (35) principal eigenvalue. If the latter is negative for at least one
phase, we will have the case of local instability
of the whole multiphase system. Upon excluding
such unstable states it is expedient to separate
out an absolutely stable state of the system from
metastable ones.
In view of this purpose the free energy value F
should be calculated at all local minima of the
function (26) and the deepest of them should be
selected. To calculate F by formula (8) it is necessary to find values of chemical potentials of molecules (28) and pressure (29) for each (meta)stable state. It seems pertinent here to express m(l)
and P using formula (30) through volume fractions of phases y i and densities r i of units in them

m(l) 0T ln V y i
i1

1 exp 0

PT
l

GY J

(37)

Vi
V

(38)

1
m*a ( r i )la
T a

n(l)

n(l)
/ y i P*( r i ),
V
i1
where y i

In terms of these variables the expression for

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

F 0T n(l)ln y i
l

1 exp 0

(35)

where the elements of matrix k are proportional


to the second-order statistical moments of SCD of
macromolecules in the phase of interest

kab c(l)lalb

the free energy of the system in any (meta)stable


state looks as follows

i1

GJ

1
m*a ( r i )la
T a

0 V y i P*( r i )
i1

/ T n(l)ln
l

n(l)
/ Fin (39)
eV

Noteworthy, expression (38) can be obtained


from eq. (29) via summing them after preliminary
multiplication by V i . That is why considering systems where not pressure but volume V is given it
is convenient to exclude the one of eqs. (29) in
which external pressure P occurs. The solution of
the remainder r 0 1 eqs. (29) jointly with rm eqs.
(31) enables us to find all (meta)stable states of
the system with given distribution n(l), volume
V, and temperature T. Substituting the dependencies on these external parameters of volume
fractions y i of phases and densities r i into expression (29) leads to the equation of state, P(V, T ),
of the system under examination.
This equation, just like expressions (37) and
(39) for chemical potential m(l) and free energy
density f F/V, comprises parameters n(l) and
V as their ratio n(l)/V only. The inference made
remains, of course, valid for any intensive thermodynamic variable whose expression will contain
just the distribution for molar concentrations c(l)
c(l)

n(l) n(l) F

V
M v

(40)

which up to the factor ( F /v)Y coincides with the


number SCD of a polymer sample. This distribution for a particular polymer sample is known to
be entirely prescribed by its synthesis conditions
and can be calculated theoretically for many important actual polymers.1 The value of the system
volume reduced to M, designated as v/ F, is
readily obtained from the equation of state, at a
given value of external pressure.
According to formula (39) free energy F of multiphase system depends on the n(l), V and T not
only in explicit but also in implicit manner
through the dependence of concentrations r i of
units in different phases and their volume fractions y i on aforementioned external variables.
However, when calculating free energy first deriv-

polpa

W: Poly Physics
9610001

MOLECULAR THEORY OF HETEROPOLYMERS. I.

atives proceeding from the formula (39), there is


no necessity of taking into account this implicit
dependence because the function F reaches an extremum
F
F
0,
0
y i
r ai
(i 1, . . . , r 0 1; a 1, . . . , m) (41)
with respect to internal variables y i and r ai at any
fixed values of external variables.
The fact that the polymer system comprises,
generally speaking, an infinite number of chemically different type individual components is responsible for some qualitative peculiarities of
thermodynamics of such a system compared to
the traditional thermodynamics of low molecular
compounds. The first such peculiarity consists,
naturally, in the absence of fundamental Gibbs
phase rule, 130 because the theory imposes no restrictions on the number r of coexisting phases in
solutions of actual, i.e., polydisperse, polymers.
The general algorithm we have introduced permits one at given values of temperature T, overall
volume V (or external pressure P) and distribution n(l) of the polymer system to find all its rphase states. Among them, metastable ones may
occur, which are known to play an especially important role in the thermodynamics of polymers,
because here, unlike for most low molecular systems, such states exist for a rather long time by
virtue of kinetic reasons (extremely high viscosity, for instance).132
Another specific feature of thermodynamics of
polymer mixtures is that the problem of calculation of internal parameters y i , r i of their multiphase states reduces to the solution of a set of
relatively small number of algebraic equations.
This latter is defined along with r only by number
m of monomer unit types irrespective of how many
chemically individual components the polymer solution is composed of. That is why such a mixture
is called quasi-m-component. Monomer units
Sa , unlike solvent molecules, are none other that
quasi-components for which even the very concept
of chemical potential becomes physically meaningless. One more substantial peculiarity inherent to such thermodynamic consideration in
terms of quasi-component is the fact that the free
energy (39) of the system is not divided into several items, each being dependent only on the densities r i of monomer units in i-th phase and its
volume V i .

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

947

When constructing phase diagrams of the solution of the heteropolymer with fixed SCD, fractions of molecules of low-molecular solvents and
a total fraction of monomer units in the polymer
along with temperature and pressure (or volume)
should be considered as independent external
variables. For mixtures of several polymers (with
given SCD each of them) the system is defined by
fractions of their units.

CLOUDPOINT CURVE, SPINODAL,


AND CRITICAL POINTS
In the framework of the approach proposed the
character of variation of the equilibrium state of
the polymer system with an arbitrary SCD, resulting from the change of any of its n parameters
(thermodynamic or structural), can be predicted
without much effort. In the course of this change,
the point characterizing the equilibrium state of
a system will drift within n-dimensional space of
mentioned external parameters. For such a drift
this point can intersect some specific ( n 0 1)dimensional hypersurfaces.
The most important among them is a boundary
hypersurface upon intersecting of which the second liquid phase appears for the first time in a
homophase system. Equations defining such a
hypersurface of the cloud points ensue directly
from expressions (29) (31) and can be written
in a following simple form

F *a FGa(s) ( a 1, . . . , m)
sa exp{[ m*a ( F) 0 m*a ( F*)]/T}

(42)
(43)

TFG(s) / vP*( F*) TFG(1)


/ vP*( F) vP (44)
where components Ga(s) of the vector G(s) represent derivatives

Ga(s) sa

G(s)
G(s)

sa
ln sa

(45)

of generating function
G(s)
l

polpa

n(l)
M

s ala

a1

W: Poly Physics
9610001

(46)

948

KUCHANOV AND PANYUKOV

of the SCD of all constituents including low molecular solvents. Let us consider the set of eqs. (42)
and (43) with respect to vector F* vr* with
components equal dimensionless concentrations
of different type units. These equations by virtue
of stoichiometry relations (32) and (B.2) always
have the trivial solution

F *a U atr ( F) Fa XaF, sa 1
( a 1, . . . , m) (47)
where the dependence of total density of units
F in a single-phase system on pressure and temperature at a given SCD can be found from the
equation of state (17). Within certain region of
values of the mentioned external parameters P,
T, and function SCD eqs. (42) and (43) in parallel
with trivial solution can possess a nontrivial one

F *a Ua( F)

(48)

which determines the dependence of densities F*


of units in the first drops of precipitated phase
on their values F in principal phase. Analogous
dependencies of F* on F in the case of traditional
thermodynamics of low molecular compounds are
normally obtained 130 from the equality of chemical potentials ma( F*) ma( F) of constituents in
both phases. These equations, as easy to prove,
can be derived from general eqs. (42) and (43),
provided they are employed for low molecular solvents. Substituting every F*a for functions Ua( F)
in the expression (44), which has a meaning of
the equality of pressure of coexisting phases, one
will arrive at the equation for the hypersurface of
cloud points. To each point F located on it there
corresponds value F* of the densities of units in
a precipitating phase, so that the manifold of all
such points F* will constitute the shadow hypersurface.
Another important specific hypersurface within
the parameter space is the spinodal where the
loss of the local stability of spatially homogeneous
single-phase state takes place. This corresponds
to the vanishing of the principal eigenvalue in the
spectrum of the matrices (33) and (35), a necessary condition of which is turning into zero of their
determinants. When carrying out practical calculations of a spinodal it is more convenient to proceed from the condition M 0, resorting to find
matrix M (35) to a simple expression

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

kab

F
Gab(1),
v
where Gab(s)

2G(s)
(49)
ln sa ln sb

for matrix k elements (36), relating them with


second-order derivatives of generating function
(46) at point s 1.
Proceeding from the expression (B.5) for chemical potential we can reduce the spinodal condition, M 0, to the more simple form by dividing
the matrix (49) into two terms
M MR / MS,
where M Rab ( k 01 )ab 0 2vgab/T,
M Sab v/(1 0 F ) (50)
exhibiting at F r 1 regular and singular behavior,
respectively. The determinant of the matrix M is
expressible through the sum D R of all elements
of matrix adjoint 133 for M R and its determinant
M R, so that the equation for spinodal looks as
follows
M M R / vD R (1 0 F ) 01 0

(51)

The (n 0 2)-dimensional manifold (hyperline) of


critical points on the cloudpoint hypersurface
can be found provided the densities F *a of units of
all types within the incipient phase coincide with
their values Fa within the initial system. Such a
coincidence means that the set of eqs. (44) along
the whole hyperline mentioned has a multiple
root F* F , pertaining jointly to two different
branches, (47) and (48), of the solution of these
equations. Mathematical condition of multiplicity
of trivial root F* F is vanishing of the determinant of matrix E 0 kC, which agrees with spinodal condition M 0. Thus, the manifold of common points of both the cloudpoint hypersurface
and a spinodal one will contain the hyperline of
critical points.
The equation defining such a hyperline can be
derived in different manners, some of which have
been already employed earlier for several simple
polymer systems.36,40,50,60,64,68 Below, dealing with
the solution of a general problem, we resort to the
approach whose key idea, going back, for polymer
systems, to the pioneer study by Stockmayer, 40 is
based on the Landau expansion of nonequilibrium
potential (26) into the Tailor series of powers of

polpa

W: Poly Physics
9610001

MOLECULAR THEORY OF HETEROPOLYMERS. I.

deviation DC(l; r) C(l; r) 0 c(l) of molecule


concentration C(l; r) from their equilibrium values c(l) in a single-phase state. As shown in Appendix C, applying such an approach one readily
arrives at the following simple condition

(3)
abg

cacbcg 0

949

with the number of rows equal to that of the types


of monomer units. Into eq. (57) enters submatrices of the complete matrix (34)
C

(52)

C PP C PS
C SP C SS

abg

Kij

3)
3)
(3)
G (abg
k (ijk
( k 01 )ia( k 01 ) j b( k 01 )kg / f *abg
/T (53)

dij
1 2 fSU ( r)
0C ijSS /
T ri rj
ri

(58)

ijk

which, together with the spinodal condition M


0, yields the equation of the hyperline of critical
points of polymer system with arbitrary SCD.
Along with components ca of the eigenvector c
of matrix M (35), corresponding to its vanishing
principle eigenvalue, and elements (36), (49) of
matrix k , formula (53), contains also components
3)
k (abg
c(l)lalblg
l

F (3)
G abg (1),
v

3)
G (abg
(s)

( 3 ) G(s)
(54)
ln sa ln sb ln sg

which take into account volume interactions of


different types: (polymer / polymer), (polymer
/ solvent) and (solvent / solvent). In the reduced
system the interactions of such a kind proceed, as
it follows from expression (57), either directly
(the first item) or through solvent molecules (the
second item).
When calculating critical points in a reduced
system it can be also resorted to general formulas
(52) and (53), substituting in them the components of tensor f * ( 3 ) for their modified values
(3)
(3)
3)
f *abg
/ (L (aij
pibpj g
fO *abg
ij

(3)

(3)
f *abg

f *( r)
rarbrg

/L

(55)

(3)
ibj

3)
piapj g / L (ijg
piapj b)

3)
3)
3)
pig / J (aig
pib / J (ibg
pia)
/ (J (abi
i

of the two third-order symmetric tensors k ( 3 )


and f * ( 3 ) . Noteworthy, that because of obvious
identity

( k 01 )abcb Cabcb
b

(56)

CO C PP / C PSK 01C SP

/ 8Q4F$$0001

ijk

Here, summing is over all types of solvents and


the following designations are used

the matrix k 01 elements in formula (53) can be


substituted for those of matrix C (34).
When constructing a phase diagram of solutions of multicomponent heteropolymers it appears more convenient to reformulate the problem
in terms of macromolecules only, having formally
reduced the number of quasi-components due to
an exclusion from their low molecular solvents.
Under such an approach the expression for SCD,
entering into all thermodynamic formulas, describes polymer exclusively. However, all characteristics of the system of Separate Units appearing in these formulas are to be replaced by
their modified values. Thus, for instance, to calculate the spinodal using equation M 0 one
should substitute in it the matrix of direct correlation functions (34) for square matrix

8Q4F

3)
/ K (ijk
piapj bpkg (59)

(57)

02-14-98 16:01:42

3)
pia (K 01 )ijC jSP
L (aij

a ,
j
3)

J (abi

1 3 fSU ( r)
,
T rari rj

1 3 fSU ( r)
1 3 fSU ( r)
3)
, K (ijk

T rarbri
T ri rj rk

(60)

In the above formulae (58) (60) Greek and


Latin indexes are reserved for polymer and solvent components of the density, respectively.
Equations (57) and (59) can be derived by minimizing the free energy over solvent densities and
calculating corresponding derivatives of the resulting expression with respect to the components
{ ra} of polymer density.

ERGODIC HETEROPOLYMERS
To demonstrate some potentialities of the general
thermodynamic approach put forward above, er-

polpa

W: Poly Physics
9610001

950

KUCHANOV AND PANYUKOV

godic heteropolymers have been chosen. Among


these latter are high-molecular compounds synthesized by copolycondensation and azeotropic
free radical copolymerization as well as the products of polymer analogous transformations.1 Once
thermodynamics of oligomers is left apart, the
SCD of high-molecular heteropolymers is conveniently considered in terms of continuous variables l and z and can be presented as the product
of the distribution f (l) of chains for their length
and conditional distribution W (lz) of macromolecules with a fixed length l for composition z . The
latter of these distributions in the case of ergodic
heteropolymers is known to be a Gaussian one, 139
exhaustively characterized by its first- and second-order statistical moments, i.e., by the vector
of average composition X and covariance composition matrix l
Xa za lab ( za 0 Xa)( zb 0 Xb)

Dab
l

( a, b 1, . . . , m) (61)
In view of an apparent condition z1 / rrr
/ zm 1 number of independent quantities Xa and
Dab will be (m 0 1) and (m 0 1)m/2, respectively.
These quantities are not controlled by l, and can
be expressed through stoichiometric and kinetic
parameters of the process of copolymer synthesis
within the framework of known kinetic models.1,139 It is readily shown that the generating
function (46) for ergodic heteropolymers looks as
G(s)

* f (l)exp{lR(h)} dl

(62)

where following designations are used


f (l)

second- and third-order statistical moments of the


SCD are found as derivatives (49) and (54) of the
generating function (62) at point s 1
Xab Gab(1) PW Xa Xb / Dab
Xabg G

(3)
abg

(65)

(1) PZ PW Xa Xb Xg

/ PW ( Xa Dbg / Xb Dag / Xg Dab) (66)


Weight-average, PW , and z-average, PZ , degree of
polymerization of macromolecules in a simple way
are expressible through statistical moments of the
distribution f (l) (63).
The matrix M (50), whose determinant turns
into zero at the spinodal (51), looks like

Mab v

1
Pab
1
/
/
0 2gab
PW F
DF
10F

(67)

Here D D ab represents a cofactor of any element Dab of the matrix D, which up to the factor
(01) a/b is the determinant of the matrix obtained
from D by means of elimination from the latter of
the elements of a-th row and b-th column. Similar
procedure can be extended to finding the cofactor
D ab,gd of an arbitrary pair of elements Dab and Dgd ,
which do not appear at the same row ( a x g ) and/
or at the same column ( b x d ). The quantity D ab,gd
is equal 140 to the product of the signed factor (01) w
and the determinant of the matrix that is obtained
from D upon the elimination of the elements pertaining to the rows a and g as well as to the columns b and d. A signed factor depends on whether
the positive integer w a / b / g / d / t is even
or odd. Index t, which characterizes the ordering
of indices of chosen pair of elements Dab and Dgd of
the matrix D, will be

n(l)
,
M

0, if ( a g, b d )

R(h) Xaha /
a

1
Dabhahb (63)
2 ab

or ( a g, b d )
1, if ( a g, b d )

(68)

or ( a g, b d )
while the components of vectors h and s are related in a simple fashion ha ln sa . Functions
Ga(s) (45) entering into equation (42) for finding
cloud points can be obtained via the expression
Ga(s) (Xa / Dabhb)
b

Once all cofactors D ab,gd of the matrix D have been


found, the elements of matrix P involved in expression (67) can be determined using relationship

* lf (l)exp{lR(h)} dl (64)

Pab D ab,gdXgXd

meeting an obvious condition Ga(1) Xa . The

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

( a, b 1, . . . , m) (69)

gd

In the case of a homopolymer (m 1) the sec-

polpa

W: Poly Physics
9610001

MOLECULAR THEORY OF HETEROPOLYMERS. I.

ond item in the right-hand part of expression (67)


is missing, whereas matrices g and M degenerate
into scalars. Here, the equation for the spinodal
(67) is trivial.
Less trivial this equation proves to be for ergodic copolymers even for the simplest among
them, i.e., binary copolymer ( m 2). Matrices D
and P , for it will read
DD

1
01

01
1

D S
P

X 22

0X1 X2

0X1 X2

X 21

(70)

while the equation for the spinodal looks as follows


b
a
/ 0d0
10F F

(71)

Coefficients of this equation


a102

Dg x
T

b102

g ef
T

04DG
(72)
T2

are expressed through homogeneity parameter


D, 139 effective energy of pair contacts g ef (18), as
well as through the Flory x-parameter and the
determinant G of the matrix g

x x12 gx/T ( g11 / g22 0 2g12 )/T


G det g g11g22 0 g 212 (73)
The circumstance that average length of macromolecules does not enter into coefficients (72)
is due to the smallness of the first item compared
to the second one in the right-hand part of expression (67), which makes it possible to ignore this
small item for calculations. This property holds
at any number m of types of monomer units in
high-molecular ergodic heteropolymer, whose
combinatoric entropy of mixing essentially exceeds that of homopolymer and practically does
not depend on average length of polymer chains.
Construction of the spinodal (i.e., a curve
within the coordinate plane of thermodynamic
variables ( F , T ) where monophase state becomes
absolutely unstable) turns out to be complicated
by the fact that the elements of the matrix g can,
generally speaking, be temperature dependent.
For particular systems where such a dependence
is known the calculation of the spinodal proceeding from eq. (71) presents no problem. However,
its analysis in terms of coefficients a, b, d provides

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

951

a possibility to come to some general conclusions


about the form of the spinodal curve making no
recourse to explicit dependence of parameters gab
on T.
Equation (71) at ab 0 has the only solution
within the interval 0 F 1. This means that
when coefficients a and b differ by sign, the dependence of F on T along the spinodal curve will be
monotonic. Alternatively, if the sign of coefficients
a and b are the same, eq. (71) can have two roots
F1 and F2 possessing physical meaning. To ensure
such a situation the coefficient d is supposed, first,
to have the same sign as a and b and, second, his
modulus
should
be qlarge enough for the inequalq
q
ity d a / b to be fulfilled. If at least
one of these two conditions does not hold, eq. (71)
at ab 0 has no roots within the interval 0 F
1. This means that in three-dimensional Euclidean space of the coefficients ( a, b, d) within the
regions (a q 0, b q0) andq(a 0, b 0), there
is a surface d a / b where the confluence of two roots having physical meaning into
one takes place. At one qside of this surface such
multiple root F1 F2 b/d is split into two simple roots, whereas at the other
sideqit vanishes.
q
q
Consequently, the surface d a / b
within the space of coefficient (72) is bifurcational
one, where either the appearance of a pair of roots
within the interval 0 F 1 or their annihilation
is the case. If for temperature evolution, the point
with coordinates (a, b, d) crosses at some value
T Textr biffurcational surface situated in the
above-mentioned space, the spinodal curve does
exhibit extremum. To avoid misunderstanding, it
should be stressed that this extremum coincides,
generally speaking, for heteropolymers under consideration with no critical point located on the
spinodal.
Within the framework of the solubility parameter approach introduced by Hildebrand, 141 the elements of matrix g factorize gab gagb , which
results in x ( g1 0 g2 ) 2 /T, g ef (X1g1 / X2g2 ) 2 ,
d 0. Here eq. (71) has either one root within
the interval 0 F 1 (if ab 0), or does not
have such a root (if ab 0). The last case is
always realized under rather low ( T r 0) and
rather high (T r `) temperature. As for the extreme points on spinodal curve they are impossible under the approach discussed.
In supposition of the independency of parameters gab on temperature general analysis of spinodal curve can be essentially advanced. Moreover,
here an opportunity opens to find its position on

polpa

W: Poly Physics
9610001

952

KUCHANOV AND PANYUKOV

the plane of thermodynamic variables ( F, T ).


Spinodal eq. (71) under the above assumption can
be put as
T 2 / 2( g ef 0 gx D)T F 0 4GDF 2
0 2g ef T / 4GDF 0 (74)
The solution of this equation, depending on the
values of the parameters of the system entering
into its coefficients, can be either hyperbola or
ellipse.142 The first of them appears when the
main invariant I 04GD 0 ( g ef 0 gx D) 2 of eq.
(74) is negative. The fulfillment of the inverse
inequality I 0 is, however, only necessary condition for the hyperbola to appear. Besides, negativeness of the second invariant (referred to as
discriminant) J 4GD 2 ( gxg ef 0 G ) of eq. (74) is
also required. If both invariants I and J are positive, eq. (74) has not real roots at all, which corresponds to the absence of the spinodal at the phase
diagram. Noteworthy, that in terms of the solubility parameter approach, where G 0, the ellipse
is absolutely impossible since the main invariant
I is always negative. Obviously, only those parts
of hyperbola or ellipse, obtainable from the solution of eq. (74), represent the spinodal that fall
within the area 0 F 1, 0 T of the thermodynamic variables.
When considering thermodynamics of ergodic
heteropolymers the tensor (53) as it might be
shown will read
2

3)
G (abg
0

PZ Pab / Pag / Pbg


v
f *abg
(75)
/
/
2
2
D
T
F PW

and the condition (52) can be reduced to the following form


v 2c

1
c2
0 2
2
(1 0 F)
F

DG

3 m
PZ 2
c / Pabcacb
2
PW
D ab
0

/ 8Q4F$$0001

c2

02-14-98 16:01:42

1
D

1
01/a
F
2

Pabcacb
ab

1
0 1 / b (77)
10F

using which, eq. (76) can be written down as follows:


au 2 0 2u 0 3b 0, where u

F
(78)
10F

The condition for finding the critical point is the


existence of positive root, which is common for eq.
(78) and that of the spinodal (71), the latter being
rewritten in terms of the variable u
au 2 / (a / b 0 d)u / b 0

(79)

The above condition is fulfilled at some surfaces


in the three-dimensional space of coefficients
(72), the equations of which are obtainable by
equating to zero the resultant of two square trinomials appearing in the left-hand part of eqs. (78)
and (79). The set of such surfaces comprises two
planes, a 0 and b 0, as well as two surfaces
s/ and s0 , described by equations
q

d a / b / 23 (1 { 2 1 / 3ab)

(76)

where c c1 / rrr / cm denotes the sum of all


components of the eigenvector c corresponding to
the zeroth main eigenvalue of the matrix M (35).
In the case of a homopolymer we will have Pab
0, c x 0, so that the value F F cr at critical
point is possible to find from simple relationship
F 2 /(1 0 F) 2 Pz /P 2w and it will be always positioned within the region F ! 1. Quite different is
the situation with ergodic heteropolymers where

8Q4F

the volume fraction F cr is practically independent


from average length of macromolecules and can
assume any values. The absence of the dependence mentioned is due to the smallness of the
first item in comparison with the second one in
the parentheses of eq. (76), which permits one to
omit this small item from further consideration.
It can be readily shown that the retained item is
necessarily positive in view of positive definiteness of matrix P .
For binary copolymers one can easily arrive at
the following relationships

(80)

The first of them, s/ , is located above the right


half-plane, whereas the second one, s0 , is positioned above lower half-plane of the Cartesian
plane of coefficients (a, b). When the dependencies of coefficients a, b, and d on T are known, eq.
(80) enables determination of the value of temperature T cr at every critical point located on each
surface s/ and s0 . Critical values of polymer volume fraction F F cr at these two surfaces can

polpa

W: Poly Physics
9610001

MOLECULAR THEORY OF HETEROPOLYMERS. I.

953

Figure 1. The surface of critical points of the melt of binary ergodic copolymer. The
parameters a and b are defined by eq. (72).

be calculated, respectively, proceeding from the


expressions
u/

q
q
1
1
(1 / 1 / 3ab) u0 (1 0 1 / 3ab)
a
a

(81)
where coefficients a and b are taken at values T
T cr . Highly helpful in the understanding of the
expected results of such a calculation can be
three-dimensional graph depicted on Figure 1.
The number of critical points appearing on the
phase diagram of the melt of binary ergodic copolymer coincides, obviously, with the number of the
roots T cr of eq. (80) having physical meaning at
both surfaces s/ and s0 .

CONCLUSIONS
An extension of the Leontovich treatment to polymer systems with an infinite set of types of molecules enables us to formulate a simple variational

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

principle of the construction of thermodynamics


of such systems. According to this principle, the
nonequilibrium thermodynamic potential L (3) is
introduced, having an absolute minimum at thermodynamically equilibrium state of a system and
local minimum at each its metastable state.
Neglecting fluctuations of the density of units
on scales comparable to the dimensions of the
macromolecule, expression (9) has been derived
in terms of the mean field theory for thermodynamic potential L of the system, containing an
arbitrary mixture of polymers and solvents.
Within the context of such an approximation
this mixture is defined by the distribution (SCD)
n(l)/ P of molecules for number l of different type
monomer units. Besides, the potential L is known
to depend on parameters describing physical interaction between units in the framework of some
model of low molecular liquid. Such parameters
of the model considered in the present study are
an excluded volume of a unit as well as parameters of pairwise interaction of such units. These
thermodynamic parameters, along with the SCD,
exhaustively describe polymer system in terms of

polpa

W: Poly Physics
9610001

954

KUCHANOV AND PANYUKOV

approximation chosen and define all its thermodynamic characteristics such as, in particular,
chemical potentials (13) of molecules, enthalpy
(20), entropy (21), the Gibbs free energy (22), as
well as the equation of state (14) and (17) of the
homophase system.
When polymer fluid consists of several spatially
homogeneous phases, a closed set of eqs. (29)
(31) is derived to determine densities of various
units in each phase and their volumes. Calculating free energy (39) values for obtained solutions
of eqs. (29) (31), one can separate the absolutely
stable state of the system from the metastable
ones.
Simple equations for finding cloud points (42)
(44), spinodal (51), and critical points (52) (53)
are derived for systems with an arbitrary SCD. It
is essential that these equations are of trivial form
in terms of the generating function (46) of the
aforementioned distribution. The latter can be often found analytically by virtue of modern theoretical approaches of statistical chemistry 1,139 for
great variety of products of the synthesis and
chemical modification of polymers, even if the expression for the very SCD is impossible to obtain.
To illustrate the efficiency of the general thermodynamic approach, ergodic copolymers, which
are universally widespread in polymer manufacturing, have been chosen. For these copolymers
the problems have been considered connected
with the construction of the spinodal and critical
points on phase diagram of their melts.
All thermodynamic relationships derived within
the present study take into account in the simplest way polymer system compressibility.
One of the authors (S.I.K.) is very much indebted to
J. Prausnitz, K. Schweizer, D. Paul, and J. Cowie for
sending the reprints of their publications on Polymer
Thermodynamics.

APPENDIX A
To derive expressions (9) and (10) for the free
energy functional (4), let us proceed from its
definition, proposed for the first time by Leontovich 127

F [C] F[H e ] 0
l

* H (l; r)C(l; r) d r
e

/ 8Q4F$$0001

C(l; r) dF[H e ]/ dH e (l; r)

(A.2)

Considering the system with volume interactions in terms of the mean field approximation,
we should replace it by equivalent system of ideal
(i.e., noninteracting between each other) molecules, located in the self-consistent field h(l; r)
defined by formula (2) at values ha(r) m*a ( r(r)).
The concentration distribution of such molecules
is characterized by variational (i.e., functional)
derivative
C(l; r) dFCB[H]/ dH(l; r)

(A.3)

of the free energy functional FCB[H] of so-called


system of Chemical Bonds 128,129 being in the
field
H(l; r) H e (l; r) / h(l; r)

(A.4)

Free energy density of this system, representing


an ideal gas of molecules, is known 130 to equal
1
c(l)
FCB[H] T c(l)ln
/ fin (A.5)
V
eQ(l; [H])
l
where following designations are used
1
V

Q(l; [H])

* exp[0H(l; r)/T] d r
3

fin c(l) min (l); c(l)


l

n(l)
V

(A.6)
(A.7)

To find the Gibbs free energy min (l) of sole molecule with the given value of vector l in the absence
of both external fields and volume interactions,
it is straightforward to have a recourse to wellelaborate methods of the solutions of such problems.134,135
Relationships (A.2) (A.4) permit one to reconstruct the equilibrium functional
F[H e ] FCB[H e / h] 0

* P*( r(r)) d r
3

(A.8)

(A.1)

Here, F[H e ] is free energy of the equilibrium sys-

8Q4F

tem in an external field H e (l; r) forming the distribution of concentrations of molecules

02-14-98 16:01:42

Setting expressions (A.5) and (A.6) into formulas (A.8) and performing the Legendre transformation (A.1), (A.2) leads to expressions (9) and (10).

polpa

W: Poly Physics
9610001

MOLECULAR THEORY OF HETEROPOLYMERS. I.

APPENDIX B
In the framework of the hole version 9,136138 of the
lattice model of low molecular liquid each of M its
molecules occupies one cell of volume v, whereas
some of them can be empty, containing no particles.
The number M of molecules in a system with volume V is, obviously, less than total number V/v of
cells of corresponding lattice while the ratio

Vmin vM
r

vr
V
V
rmax

(B.1)

has a meaning of fraction of the V falling on the


minimum volume Vmin vM, which the liquid
would occupy if its molecules were packed firmly
in contact with each other. The free volume of
a liquid is considered to represent the difference
between its actual volume V and minimum one
Vmin . The F, by definition (B.1), is none other than
dimensionless density of particles reduced to its
maximum value rmax 1/v corresponding to the
closest packing of molecules.
The free-volume concept can be also used for the
description of systems comprising several types of
particles Sa ( a 1, . . . , m). Thus, each of m components parallel with its molar fraction Xa
m

Xa

Ma
, M Ma ,
M
a1

Xa 1 (B.2)
a1

will be defined by the value of reduced density Fa .


Here, we restrict ourselves to the simplest variant
of the cellular hole model with the parameter of
reducing rmax 1/v where v is a volume of one cell
that has physical meaning of the average closepacked volume of a particle in the fluid. Under
this approximation the empirical parameter v depends only on geometrical size of molecules irrespective of their mixing conditions. Stoichiometry
relationships for Fa analogous to (B.2) look as
follows

ra
Fa
vra ,
rmax

Fa F vr (B.3)

/ 8Q4F$$0001
8Q4F

Sa and Sb are defined by parameters gab , the relationships for thermodynamic potentials of a low molecular liquid will be of rather simple form.9,136138
Employing these relationships for the description of low molecular system of Separate Units,
occurring in our treatment, it is easy to write
down equations for the Helmholtz free-energy
density, chemical potentials of components and
pressure. These equations lead to the following
expressions for magnitudes (9) and (11) characterizing the polymer system
f *v
(1 0 F )ln(1 0 F )
T
/F0

02-14-98 16:01:42

1
gabFaFb (B.4)
T ab

m*a
2
0ln(1 0 F ) 0 gabFb
T
T b
P*v
1
0ln(1 0 F ) 0 F 0 gabFaFb
T
T ab

(B.5)
(B.6)

APPENDIX C
Let us consider the expansion of functional L (3)
into the Tailor series of powers of deviations DC
C 0 c of l-mers concentrations C from their
spatially homogeneous distribution c(l). The
zeroth and the first terms of this expansion are
identically equal to zero. To calculate the rest of
terms it is necessary to find expansion coefficients, which for the case under consideration, are
proportional to derivatives of the free energy with
respect to l-mer concentrations. For the secondand the third-order terms these expansion coefficients are equal to VT K ( 2 ) (l*, l 9)/2! and
VT K ( 3 ) (l*, l9, l -)/3!, where elements of tensors k
k(2)and k(3) are defined, respectively, by eq. (33)
and the following expression

K ( 3 ) (l*, l9, l -)

1
3 f
T C(l*)C(l9)C(l -)

a1

In parallel with the excluded volume v, characterizing mutual repulsion of particles, low molecular
liquid models usually contain parameters that take
into account the attraction caused by the van der
Vaals interaction. If these latter are accounted in
terms of the well-known van LaarHildebrand approximation, where pairwise interactions of particles

955

d (l* 0 l9) d (l9 0 l -)


C 2 (l*)

1
3 f *
lalblg (C.1)

T abg rarbrg

Any function DC(l) can be expanded by the eigenvectors of the matrix (33). In this expansion solely

polpa

W: Poly Physics
9610001

956

KUCHANOV AND PANYUKOV

the first item, corresponding to eigenvalue l1 0,


can be retained in the vicinity of the spinodal.
Such a reduction is due to the fact that at the
spinodal the loss of local stability of the spatially
homogeneous state occurs only along one of main
directions of free energy n-dimensional hypersurface, where the alteration of the sign of its curvature takes place. Changes DC(l) along all other
main directions of this hypersurface at the extremal C c(l) can be neglected. The closer a system
is to spinodal, the more accurate is such an approximation. Quadratic with respect to the DC
term of the expansion of the functional L (3)
equals 0 on (n 0 1)-dimensional spinodal hypersurface, while along (n 0 2)-dimensional hyperline of critical points located on this hypersurface,
cubic with respect to DC term of this expansion
also vanishes
VT
K ( 3 ) (l*, l9, l -) DC(l*) DC(l9) DC(l -)
3! l = l 0 l 0

(C.2)

It can be readily shown that this condition is


equivalent to the expression (52).

REFERENCES AND NOTES


1. S. I. Kuchanov, Methods of Kinetic Calculations
in Polymer Chemistry (in Russian), Chemistry,
Moscow, 1978.
2. P. J. Flory, J. Chem. Phys., 10, 51 (1942).
3. M. L. Huggins, Ann. NY Acad. Sci., 43, 1 (1942).
4. P. J. Flory, Principles of Polymer Chemistry, Cornell Univ. Press, Ithaca, NY, 1953.
5. H. Tompa, Polymer Solutions, Butterworth, London, 1956.
6. M. L. Huggins, Physical Chemistry of High Polymers, Wiley, New York, 1958.
7. I. Prigogine, N. Trappeniers, and V. Mathod, Discuss. Faraday Soc., 15, 93 (1953).
8. I. Prigogine and A. Bellemans, J. Polym. Sci., 18,
147 (1955).
9. I. Prigogine, The Molecular Theory of Solutions,
North Holland Publ. Comp., Amsterdam, 1957.
10. J. Hijmans, Physica, 27, 433 (1961).
11. S. N. Bhattacharaya, D. Patterson, and T. Sominsky, Physica, 30, 1276 (1964).
12. D. Patterson, Macromolecules, 2, 673 (1969).
13. D. Patterson and G. Delmas, Discuss. Faraday
Soc., 98 (1970).
14. P. J. Flory, R. A. Orwoll, and A. Vrij, J. Am. Chem.
Soc., 86, 3507, 3515 (1964).
15. P. J. Flory, J. Am. Chem. Soc., 87, 1833 (1965).

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

16. B. E. Eichinger and P. J. Flory, Trans. Faraday


Soc., 64, 2035 (1968).
17. P. J. Flory, Discuss. Faraday Soc., 7 (1970).
18. I. C. Sanchez and R. H. Lacombe, J. Phys. Chem.,
80, 2352 (1976).
19. R. H. Lacombe and I. C. Sanchez, J. Phys. Chem.,
80, 2568 (1976).
20. I. C. Sanchez and R. H. Lacombe, Macromolecules,
11, 1145 (1978).
21. I. C. Sanchez, J. Macromol. Sci., B17, 565 (1980).
22. I. C. Sanchez, in Polymer Compatibility and Incompatibility, K. Solk, Ed., Harwood Acad., New
York, 1982, p. 59.
23. C. J. Glover and W. A. Ruff, Macromolecules, 21,
3051 (1988).
24. J. A. R. Renuncio and J. M. Prausnitz, Macromolecules, 9, 898 (1976).
25. V. Brandani, Macromolecules, 11, 1293 (1978).
26. V. Brandani, Macromolecules, 12, 883 (1979).
27. F. Hamada, T. Shiomi, K. Fujisawa, and A. Nakajima, Macromolecules, 13, 729 (1980).
28. K. Fujisawa, T. Shiomi, F. Hamada, and A. Nakajima, Polym. Bull., 3, 261 (1980).
29. J. G. Curro, J. Macromol. Sci., Rev. Macromol.
Chem., C11, 321 (1974).
30. E. F. Casassa, J. Polym. Sci., Part C, N54, 53
(1976).
31. R. A. Orwoll, Rubber Chem. Technol., 50, 451
(1977).
32. I. C. Sanchez, in Polymer Blends, D. R. Paul and
S. Newman, Eds., Academic Press, New York,
1982.
33. E. J. Beckman, R. S. Porter, R. Koningsveld, and
L. A. Kleintjens, in Integration of Fundamental
Polymer Science and Technology, Vol. 2, P. J.
Lemstra and L. A. Kleintjens, Eds., Elsevier Applied Science, London, 1988, p. 197.
34. R. K. Jain and R. Simha, Macromolecules, 22, 464
(1989).
35. M. Wolf and J. H. Wendorff, Polym. Commun., 31,
226 (1990).
36. R. L. Scott, J. Chem. Phys., 17, 268 (1949).
37. R. L. Scott, J. Chem. Phys., 17, 279 (1949).
38. H. Tompa, Trans. Faraday Soc., 45, 1142 (1949).
39. H. Tompa, Trans. Faraday Soc., 46, 970 (1950).
40. W. H. Stockmayer, J. Chem. Phys., 17, 588
(1949).
41. R. Koningsveld, Dr. Sci. Dissertation, Heerlen
University, Leiden, Holland, 1967.
42. R. Koningsveld, in Polymer Science (A Materials
Science Hand Book), Vol. 2, A. D. Jenkins, Ed.,
North Holland Publishing, Amsterdam, 1972, p.
1047.
43. R. Koningsveld, Br. Polym. J., 7, 435 (1975).
44. R. Koningsveld and L. A. Kleintjens, J. Polym.
Sci., Part C, N61, 221 (1977).
45. R. Koningsveld, L. A. Kleintjens, and M. H. Onclin, J. Polym. Sci., Part B, 18, 363 (1980).

polpa

W: Poly Physics
9610001

MOLECULAR THEORY OF HETEROPOLYMERS. I.

46. R. Koningsveld, M. H. Onclin, and L. A. Kleintjens, in Polymer Compatibility and Incompatibility, K. Solk, Ed., Harwood Acad. Publ., New York,
1982, p. 25.
47. R. Koningsveld, Pure Appl. Chem., 61, 1051
(1989).
48. M. Gordon, H. A. G. Chermin, and R. Koningsveld, Macromolecules, 2, 207 (1969).
49. R. Koningsveld, H. A. G. Chermin, and M. Gordon, Proc. R. Soc. Lond., Ser. A, 319, 331 (1970).
50. J. W. Kennedy, M. Gordon, and R. Koninsveld, J.
Polym. Sci., Part C, N39, 43 (1972).
51. M. Gordon, P. Irvine, and J. W. Kennedy, J.
Polym. Sci., Part C, N61, 199 (1977).
52. P. Irvine and M. Gordon, Proc. R. Soc. Lond., Ser.
A, 375, 397 (1981).
53. P. Irvine and J. W. Kennedy, Macromolecules, 15,
473 (1982).
54. Y. Hu, X. Yang, D. T. Wu, and J. M. Prausnitz,
Macromolecules, 26, N25, 6817 (1993).
55. M. T. Ratzsch and H. Kehlen, Progr. Polym. Sci.,
14, 1 (1989).
56. K. Solc, Macromolecules, 3, 665 (1970).
57. K. Solc, J. Polym. Sci., Part B, 12, 555 (1974).
58. K. Solc, J. Polym. Sci., Part B, 20, 1947 (1982).
59. K. Solc, Macromolecules, 16, 236 (1983).
60. K. Solc, L. A. Kleintjens, and R. Koningsveld,
Macromolecules, 17, 573 (1984).
61. K. Solc and K. Battjes, Macromolecules, 18, 220
(1985).
62. H. A. G. Chermin, Br. Polym. J., 9, 195 (1977).
63. K. Solc, Macromolecules, 19, 1166 (1986).
64. K. Solc, Macromolecules, 20, 2506 (1987).
65. K. Solc and Y. C. Yang, Macromolecules, 21, 829
(1988).
66. K. Solc, Y.-H. Huang, and Y. C. Yang, Macromolecules, 22, 4334 (1989).
67. R. Koningsveld, Discuss. Faraday Soc., 49, 144
(1970).
68. K. Kamide, S. Matsuda, and H. Shirataki, Polym.
J., 20, 949 (1988).
69. F. E. Karasz and W. J. MacKnight, in Multicomponent Polymer Materials, S. R. Paul and L. H.
Sperling, Eds., Advances in Chemistry Series A,
American Chemical Society, Washington, DC,
1986, p. 67.
70. R. J. Roe and D. Rigby, Adv. Polym. Sci., 82, 103
(1987).
71. W. J. MacKnight and F. E. Karasz, in Comprehensive Polymer Science, Vol. 7, G. Allen, Ed., Pergamon Press, New York, 1989, p. 111.
72. M. T. Ratzsch and H. Kehlen, in Frontiers of Macromolecular Science (Proc. of the IUPAC 32 nd Intern. Symp on Macromolecules, Japan, 1988) T.
Saegusa, T. Higashimura, and A. Abe, Eds.,
Blackwell Sci. Publ., New York, 1989.
73. A. Noshey and J. E. McGrath, Block Copolymers,
Academic, New York, 1977.

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

957

74. M. Folkes, Ed., Processing, Structure and Properties of Block Copolymers, Elsevier, London, 1985.
75. R. P. Quirk, D. J. Kinning, and L. J. Fetters, in
Comprehensive Polymer Science, Vol. 7, G. Allen,
Ed., Pergamon press, New York, 1989, p. 1.
76. R. L. Scott, J. Polym. Sci., 9, 423 (1952).
77. L. Leibler, Makromol. Chem., Rapid Commun., 2,
393 (1981).
78. R. P. Kambour, J. T. Bendler, and R. C. Bopp,
Macromolecules, 16, 763 (1983).
79. G. ten Brinke, F. E. Karasz, and W. J. MacKnight,
Macromolecules, 16, 1827 (1983).
80. D. R. Paul and J. W. Barlow, Polymer, 25, 487
(1984).
81. R. Koningsveld and L. A. Kleintjens, Macromolecules, 18, 243 (1985).
82. D. Rigby, J. L. Lin, and R.-J. Roe, Macromolecules,
18, 2269 (1985).
83. J. W. Barlow and D. R. Paul, Polym. Eng. Sci., 27,
1482 (1987).
84. M. Nishimoto, H. Kekkula, and D. R. Paul, Polymer, 30, 1279 (1989).
85. G. R. Brannock and D. R. Paul, Macromolecules,
23, 5240 (1990).
86. M. T. Ratzsch, H. Kehlen, and D. Browarzik, J.
Macromol. Sci., Ser. A, 22, 1679 (1985).
87. M. T. Ratzsch, H. Kehlen, and D. Browarzik, J.
Macromol. Sci., Ser. A, 23, 1349 (1986).
88. M. T. Ratzsch, D. Browarzik, and H. Kehlen, J.
Macromol. Sci., Ser. A, 26, 903 (1989).
89. M. T. Ratzsch, C. Wohlfarth, D. Browarzik, and
H. Kehlen, J. Macromol. Sci., Ser. A, 26, 1497
(1989).
90. M. T. Ratzsch, D. Browarzik, and H. Kehlen, J.
Macromol. Sci., Ser. A, 27, 809 (1990).
91. C. Panayuotou, Makromol. Chem., 188, 2733
(1987).
92. P. Dafniotis, E. Karayianni, and C. Panayioton,
Makromol. Chem., 190, 1103 (1989).
93. T. Shiomi, H. Ishimatsu, T. Eguchi, and K. Imai,
Macromolecules, 23, 4970 (1990).
94. T. Shiomi, T. Eguchi, H. Ishimatsu, and K. Imai,
Macromolecules, 23, 4978 (1990).
95. T. Shiomi and K. Imai, Polymer, 32, 73 (1991).
96. J. Dudowicz and K. F. Freed, Macromolecules, 24,
5076, 5112 (1991).
97. K. F. Freed and J. Dudowicz, Theor. Chim. Acta,
82, 357 (1992).
98. J. Dudowicz and K. F. Freed, J. Chem. Phys., 96,
1644, 9147 (1992).
99. K. F. Freed and J. Dudowicz, J. Chem. Phys., 97,
2105 (1992).
100. K. S. Schweizer, Macromolecules, 26, 6033
(1993).
101. D. Chandler and H. C. Andersen, J. Chem. Phys.,
57, 1930 (1972).
102. K. S. Schweizer and J. G. Curro, Chem. Phys.,
149, 105 (1990).

polpa

W: Poly Physics
9610001

958

KUCHANOV AND PANYUKOV

103. K. S. Schweizer and J. G. Curro, J. Chem. Phys.,


94, 3986 (1991).
104. K. S. Schweizer and J. G. Curro, Phys. Rev. Lett.,
60, 809 (1988).
105. K. S. Schweizer, Macromolecules, 26, 6050
(1993).
106. Y. Song, S. M. Lambert, and J. M. Prausnitz,
Chem. Eng. Sci., 49, 2765 (1994).
107. Y. C. Chiew, Mol. Phys., 70, 129 (1990).
108. Y. Song, S. M. Lambert, and J. M. Prausnitz, Macromolecules, 27, 441 (1994).
109. Hino, T. Y. Song, and J. M. Prausnitz, Macromolecules, 27, 5681 (1994).
110. T. Hino, Y. Song, and J. M. Prausnitz, Macromolecules, 28, 5717, 5725 (1995).
111. A. V. Balazs, I. C. Sanchez, I. R. Epstein, F. E.
Karasz, and W. J. MacKnight, Macromolecules,
18, 2188 (1985).
112. P. C. Balazs, F. E. Karasz, W. J. MacKnight, H.
Veda, and I. C. Sanches, Macromolecules, 18,
2784 (1985).
113. J. van Hunsel, A. C. Balazs, W. J. MacKnight, and
R. Koningsveld, Macromolecules, 21, 1528 (1988).
114. A. C. Balazs and M. T. De Meuse, Macromolecules, 22, 4260 (1989).
115. H.-J. Cantow and O. Schulz, Polym. Bull., 15, 449
(1986).
116. H.-J. Cantow and O. Schulz, Polym. Bull., 15, 539
(1986).
117. D. R. Paul, in Multicomponent Polymer Materials,
S. R. Paul and L. H. Sperling, Eds., Advances in
Chemistry Series, American Chemistry Society,
Washington, DC, 1986, p. 3.
118. D. J. Walsh, in Comprehensive Polymer Science,
Vol. 2, G. Allen, Ed., Pergamon Press, New York,
1989, p. 135.
119. P. Munk, P. Hattam, and Q. Du, J. Appl. Polym.
Sci., Appl. Polym. Symp., 49, 373 (1989).
120. L. A. Utracki, Polymer Alloys and Blends: Thermodynamics and Rheology, C. Hanser Verlag,
Munchen, 1989.
121. R. Koningsveld, W. H. Stockmayer, and E. Nies,
Makromol. Chem., Macromol. Symp., 39, 1 (1990).
122. K. Kamide, Thermodynamics of Polymer Solu-

/ 8Q4F$$0001
8Q4F

02-14-98 16:01:42

123.
124.
125.
126.
127.
128.
129.
130.
131.

132.
133.
134.

135.
136.
137.
138.
139.

140.
141.
142.

polpa

tions. Phase Equilibria and Critical Phenomena,


Elsevier, Amsterdam, 1990.
C. Wohlfarth, Makromol. Chem., Theory Simul.,
2, 605 (1993).
H. W. Kammer, J. Kressler, and C. Kummerloewe, Adv. Polym. Sci., 106, 31 (1993).
K. Binder, Adv. Polym. Sci., 112, 183 (1994).
J. M. G. Cowie, Macromol. Symp., 78, 15 (1994).
M. A. Leontovich, Introduction to Thermodynamics (in Russian), 3rd ed., Nauka, Moscow, 1983.
I. M. Lifshitz, A. Yu. Grosberg, and A. R. Khokhlov, Rev. Mod. Phys., 50, 683 (1978).
S. I. Kuchanov, S. V. Korolev, and S. V. Panyukov,
Adv. Chem. Phys., 72, 115 (1988).
L. D. Landau and E. M. Lifshitz, Statistical Physics, Addison Wesley, Reading, MA, 1969.
C. A. Croxton, Liquid State PhysicsA Statistical
Mechanical Introduction, Cambridge Univ. Press,
Cambridge, 1974.
A. A. Tager, Vysokomol. Soedin., Ser. A, 30, 1347
(1988).
R. A. Horn and C. R. Johnson, Matrix Analysis,
Cambridge Univ. Press, Cambridge, MA, 1986.
M. V. Volkenstein, Configuration Statistics of
Polymer Chains, Interscience Publishers, New
York, 1963.
P. J. Flory, Statistical Mechanics of Chain Molecules, Interscience Publishers, New York, 1969.
E. A. Guggenheim, Mixtures, Oxford Univ. Press,
London, 1952.
E. A. Guggenheim, Applications of Statistical Mechanics, Oxford Univ. Press, London, 1966.
N. A. Smirnova, Molecular Theories of Solutions
(in Russian), Khimia, Leningrad, 1987.
S. I. Kuchanov, in Mathematical Methods in Contemporary Chemistry, S. Kuchanov, Ed., Gordon & Breach Publ., Amsterdam, 1996, p. 267.
F. R. Gantmakher, Theory of Matrix, Nauka Publ.,
Moscow, 1966.
J. H. Hildebrand and R. L. Scott, The Solubility
of Non-Electrolytes, Dover Publ., New York, 1964.
G. A. Korn and T. M. Korn, Mathematical Handbook for Scientists and Engineers, McGraw-Hill
Book Company, New York, 1961.

W: Poly Physics
9610001

Вам также может понравиться