Bisphenol-A Polycarbonate-Poly( butylene Terephthalate) Transesterification. 11. Structure Analysis of the Reaction Products by IR and lH and 13C NMR J. DEVAUX, P. GODARD,* and J. P. MERCIER, Laboratoire des Hauts PolymGres, and R. TOUILLAUX and J. M. DEREPPE, Laboratoire de Chimie Physique et de Cristallographie, Uniuersit6 Catholique de Louuain, R- 1.348-Louvain-la-Neuve, Belgium Synopsis The structure of the four-component copolyester resulting from the exchange reaction between molten bisphenol-A polycarbonate and poly(buty1ene terephthalate) is analyzed as a function of the reaction time by infrared and nuclear magnetic resonance spectroscopy. By applying a statistical method developed earlier, the mean chain length of the various sequences as well as the degree of randomness is computed. The exchange reaction leads initially to the formation of a block copo- lyester with reduced solubility. As the reaction proceeds, a solutde random copolycondensate is progressively formed. INTRODUCTION As previously mentioned,' an exchange reaction takes place in molten bis- phenol-A polycarbonate (PC)-polybutylene terephthalate (PBTP) mixtures. In the present work, the structure of the reaction products was determined by infrared and NMR analysis, and it is shown here that a four-component copo- lyester is formed. By application of the statistical methods described in the first article of this series, the structure of this four-component copolycondensate is quantitatively analyzed and the mean chain length of the various sequences as well as their degrees of randomness is computed. EXPERIMENTAL Materials The PBTP used in this work is a commercial product manufactured by Eastman Kodak under the trade name Tenite 6 PRO. Its structure is A number-average molecular weight of 36,500 was determined from intrinsic * Research Associate of the National Fund for Scientific Research (Belgium). ,Journal of Polymer Science: Polymer Physics Edition, Vol. 20, 1881-1894 (1982) (C' 1982 John Wiley & Sons, Inc. CCC 0098- 1273/82/101881- 14$02.,40
Transcript
Bisphenol-A Polycarbonate-Poly( butylene Terephthalate) Transesterification. 11. Structure Analysis of the Reaction Products by IR and lH and
13C NMR
J. DEVAUX, P. GODARD,* and J. P. MERCIER, Laboratoire des Hauts PolymGres, and R. TOUILLAUX and J. M. DEREPPE, Laboratoire de
Chimie Physique et de Cristallographie, Uniuersit6 Catholique de Louuain, R- 1.348-Louvain-la- Neuve, Belgium
Synopsis
The structure of the four-component copolyester resulting from the exchange reaction between molten bisphenol-A polycarbonate and poly(buty1ene terephthalate) is analyzed as a function of the reaction time by infrared and nuclear magnetic resonance spectroscopy. By applying a statistical method developed earlier, the mean chain length of the various sequences as well as the degree of randomness is computed. The exchange reaction leads initially to the formation of a block copo- lyester with reduced solubility. As the reaction proceeds, a solutde random copolycondensate is progressively formed.
INTRODUCTION
As previously mentioned,' an exchange reaction takes place in molten bis- phenol-A polycarbonate (PC)-polybutylene terephthalate (PBTP) mixtures. In the present work, the structure of the reaction products was determined by infrared and NMR analysis, and it is shown here that a four-component copo- lyester is formed. By application of the statistical methods described in the first article of this series, the structure of this four-component copolycondensate is quantitatively analyzed and the mean chain length of the various sequences as well as their degrees of randomness is computed.
EXPERIMENTAL
Materials
The PBTP used in this work is a commercial product manufactured by Eastman Kodak under the trade name Tenite 6 PRO. Its structure is
A number-average molecular weight of 36,500 was determined from intrinsic
* Research Associate of the National Fund for Scientific Research (Belgium).
,Journal of Polymer Science: Polymer Physics Edition, Vol. 20, 1881-1894 (1982) (C' 1982 John Wiley & Sons, Inc. CCC 0098- 1273/82/101881- 14$02.,40
1882 DEVAUX ET AL.
viscosity measurements at 25°C in a phenol-tetrachloroethane mixture (60-40 by weight) on applying the equation2
[q] = 2.15 x 1 0 - 4 M ; ~ ~ ( d ~ g-1)
The PC sample is also a commercial product sold by General Electric under the trade name Lexan 135. Its structure is
Its weight-average molecular weight 46,500 was computed from viscosimetry at 30°C in dioxane by using the relation3
[q] = 4.76 X 10-4Mi68 (dL g-1)
From GPC measurements, a number-average molecular weight of 15,000 was also determined.
The bisphenol-A polyterephthalate used here for identification of NMR peaks was prepared by interfacial polyconden~ation.~ Its structure can be represented as follows:
L CH3 _I,
The reagent concentrations were adjusted to obtain a low-molecular-weight polymer soluble in deuterated chloroform.
Solvents such as methylene chloride, tetrachloroethane, and methanol were bidistilled technical products; all other products used were p.a. grade.
Study of the Exchange Reaction
PC-PBTP mixtures of various compositions were prepared by blending the polymers on a two-roll mill for 5 min at 240°C. Before mixing, the polymers were dried for a t least 24 h at 60°C under reduced pressure of lo-' torr. The polymer blends were then compression molded for 5 min at 240°C and granulated. They were again dried under vacuum for 24 h at 60°C before proceeding further. The exchange reaction was carried out in the barrel of an Instron 3211 capillary rheometer a t temperatures ranging from 224 to 255°C. The temperature was determined with an accuracy of f0.25"C. During increasing elapsed times, samples of approximately 1 g were extruded for solubility, IR, and NMR mea- surements.
Solubility
The extruded samples were dissolved in boiling tetrachloroethane and pre- cipitated in methanol. The dried samples were then extracted for 7 h in a Ku-
PC-PBTP TRANSESTERIFICATION. I1 1883
magawa apparatus by a selective solvent for PC (CH2C12). The dissolved fraction w1 (precipitated in methanol) and the undissolved polymer w2 were dried and weighed. The degree of solubility was expressed by the ratio w1Iw1 + w2.
Infrared Spectroscopy
IR spectra of methylene chloride or tetrachloroethane solution-cast films were obtained in a Perkin-Elmer 157 G infrared spectrophotometer.
Proton and 13Carbon NMR
Spectra were obtained by using a XL 100 Varian instrument working under usual conditions with tetramethylsilane as the internal standard. For proton NMR, the solvent used was deuterated chloroform and the spectrum width was 250 or 1000 Hz whereas for l3C NMR, the solvent was trifluoroacetic acid and the spectrum width was 4000 Hz.
Quantitative analysis of 13C NMR results necessitates special precautions as in such cases the peak intensities depend not only on the concentration of the species analyzed but also on their relaxation time. Therefore, we shall only compare peaks of the same 13C chemical species5 and assume, as a first approx- imation, that substitutions modify very little the motions of the nuclei near the carbon studied. Therefore these changes have been ignored.
RESULTS AND INTERPRETATION
Solubility Tests and Infrared Spectroscopy
The occurrence of an exchange reaction between molten PC and PBTP can be clearly established from solubility tests coupled with infrared analysis.
In a first series of experiments, the solubility of mixtures of PC and PBTP of various compositions was analyzed as a function of the reaction time at 253°C.
The reaction products were extracted with methylene chloride. In this solvent, PC is completely soluble, whereas PBTP remains practically undissolved (weight loss, under standard conditions, less than 1%). Figure 1 summarizes the ex- perimental results. A t short reaction times, a sharp decrease in solubility in CHzCl2 is observed; the solubility then passes through a minimum at a time which varies with the blend compositions. At a later stage, the solubility again increases and, finally, a completely soluble product is obtained.
Similar results were observed by Yakubovitch et al. on the PC-bisphenol-A polyterephthalate system.6
Important structural changes corresponding to the solubility variations were detected by infrared spectroscopy. As is shown in Figure 2, where the evolution of the infrared spectra from 1700 to 1800 cm-l is reported for a 50150 PC-PBTP mixture, PC sequences characterized by their C=O stretching absorption at 1780 cm-' progressively appear in the insoluble fraction, while PBTP blocks with their C=O band a t 1720 cm-l are identified in the soluble part.
At reaction times greater than 30 min, new infrared absorptions were detected at 1740 and 1070 cm-l, whereas the C=O band of the carbonate group is slightly shifted from 1780 to 1770 cm-l.
1884 DEVAUX ET AL.
%
75
50
25
20 40 60 80 t (mint
Fig. 1. Solubility of PC-PBTP blends as a function of reaction time a t 253°C. Solvent methylene chloride. Weight composition of the blends: (A) P B T P 20/PC 80, (A) PBTP 4 0 E C 60, (0) PBTP 50/PC 50, (@) P B T P 60/PC 40, ( w ) P B T P SO/PC 20.
The comparison with various ester spectra shows that the peaks a t 1070 and 1740 cm-' are characteristic of an aromatic ester structure of the following type:
1800 1700 cm-' 18a) 1700 cm-'
Fig. 2. Evolution of the C=O stretching bands in a PC/PBTP blend (50/50 by weight) as a function of reaction time. Temperature 243.5OC. (1) 0, (2) 5, (3) 10, (4) 15 min; (-) soluble fraction; ( - - -) insoluble fraction.
PC-PBTP TRANSESTERIFICATION. I1 1885
The attribution of the 1740-cm-' band to the C=O stretching is quite evident, whereas we believe that the 1070-cm-I band results from a complex vibration of the para-disubstituted phenyl to the right of the ester structure influenced by the neighboring -COO- group. The absorption a t 1770 cm-' results from the C=O stretching of a mixed aliphatic aromatic carbonate as follows:
n
The slight variation of the carbonyl stretching of the carbonate unit from 1780 to 1770 cm-' can be attributed to the progressive formation of this structure in the PC-PBTP mixtures.
I t may be concluded from these studies on solubility and infrared absorption that an exchange reaction (transesterification) takes place between PC and PBTP which, a t short reaction times, leads to the production of block copo- lyesters with reduced solubility. As the reaction times increase, the blocks be- come shorter and a level of complete solubility is reached.
Table I gives the main IR absorptions of the PC-PBTP copolyesters as well as their attributions.
Proton Nuclear Magnetic Resonance
Important information concerning the structure of the PC-PBTP copolyesters can be obtained from 100-Hz 'H NMR as already described by Gouinlock et aL7 in a study on bisphenol-A-neopentyl glycol-terephthalic acid copolyesters. The
TABLE 1 Characteristic Infrared Ahsorptions of the Ester and Carbonate Groups Detected During the
PC-PBTP Reaction
IR absorption Groups (cm-')
Aliphatic ester:
Aromatic ester:
Aromatic carbonate: n
Mixed aliphatic-aromatic carbonate: n - - @O-C-0-CH, II -
172W
17408 1070b
17808
17708
a C=O (stretching). Para-disuhstituted phenyl coupled to the oxygen atom.
1886 DEVAUX ET AL.
region between 8 and 8.4 ppm is of special interest. The terephthalic unit being represented by
there occurs a chedcal shift of the proton peaks, depending on the nature of R1 and R2. The shifts7 are noted in Table 11. When R1 and Rz, respectively, are aromatic and aliphatic, four peaks appear in this portion of the spectrum. The two main peaks are found around 8.16 and 8.20 ppm, while, in 100-Hz NMR, the two smaller absorptions at 8.06 and 8.28 ppm are superimposed on the two sin- glets corresponding to symmetrical substitutions of the terephthalic unit.
The resolution of this spectrum can be performed by the method of Gouinlock et al. using the following relation between the peak intensities S1, S2, S3, and S4 and the resonance frequencies v1,v2, v3, and v4 of the four peaks:
- s1 - _- s4
s2 s3 w + b 2 - v3)
v2 - v3 - -
Peaks are noted in order of decreasing chemical shifts. J is a coupling constant equal to 8.03 f 0.4 Hz.
The NMR spectra between 8 and 8.4 ppm of PC-PBTP (50/50 by weight) mixtures, observed after increasing reaction times at 253OC, are shown in Figure 3 together with the NMR spectrum of a low-molecular-weight bisphenol-A polyterephthalate. The PBTP spectrum cannot be obtained under the same conditions because it is insoluble in CDCl3.
The experimental results clearly disclose the progressive appearance in the PC-PBTP mixtures of terephthalic ester units substituted by one and two aro- matic groups. The bisphenol-A polyterephthalate spectrum confirms the po- sition of the terephthalic protons at 8.28 ppm for symmetrically substituted iiromatic ester units.
TABLE I1 Chemical Shifts of Praton Peaks as a Function of R1 and I22 far a Terephthalate Unita
Nature of units Chemical ehift RI Rz protone 6 (ppm)
L- Fig. 3. NMR peaks of protons of terephthalate units in the PC-PBTP system (50/50 by weight)
after different reaction times at 253OC. Solvent CDC13. (1) 30, (2) 60, (3) 100, (4) 200 min; (5) bis- phenol-A polyterephthalate.
This NMR analysis, which gives results in complete agreement with the work of Gouinlock, confirms the observations made by the infrared method.
We also notice significant modifications at the level of the bisphenol-A group aromatic protons (7.0-7.5 ppm) simultaneous with a progressive change of the spectrum around 4.2 ppm which can be attributed to a modification in the sur- roundings of the methylene group adjacent to an oxygen atom (CHZ-0). However, lack of resolution in these two regions renders a complete analysis of these data difficult.
In order to obtain further confirmation of the modifications resulting from the exchange reaction, a 13C NMR study was undertaken. Tables I11 and IV summarize the results obtained in a 50/50 mixture for increasing reaction times at 253°C.
To facilitate the interpretation, the carbons belonging to the PBTP units were labeled C and those of the PC group indexed C’. The various indices have the following significance. For PBTP:
For PC:
3 4 I 9 1 0
TABL
E 11
1 Ev
olut
ion
of P
BT
P 'w
Peaks a
nd R
elat
ive I
nten
sitie
s Dur
ing
the
Reac
tion'
(C
hem
ical
Shi
fts in
mrn
Rel
ativ
e to T
MS)
Attr
ibut
ion
C&Q C2
.s C3
.8 C,
.? C6
.6
Che
mic
al
Rea
ctio
n Sh
ifts
23.8
65.7
68.1
68
.7 16
8.3
167.5
13
2.3
132.9
13
3.4
128.
9 12
9.6
time
(mli:
ii (P
P4
Rel
ativ
e 43
30
NM
b N
M
31
NM
N
M
30
NM
10
0 N
M
inte
nsiti
es
41
26
NM
5
29
3 3
27
3 86
15
31
units
) 35
19
4
8 21
14
5
22
5 60
41
100
33
18
7 7
17
16
7 21
7
56
45
200
(arb
itrar
y 39
22
3
9 20
11
6
21
6 68
35
75
3 + r 0
0
a Id
entif
icat
ion o
f ind
ices
6
6
b N
ot m
easu
rabl
e.
TA
BL
E I
V
cd
Evo
lutio
n of
'W P
C P
eaks
and
Rel
ativ
e Int
ensi
ties D
urin
g th
e R
eact
ionP
(Che
mic
al Sh
ifta
in p
prn
Rel
ativ
e to T
MS)
?
Shift
a 15
4.8
155.
8 15
6.3
147.
7 14
7.3
119.
1 11
9.6
127.
2 14
8.9
41.6
28.7
time
(PP
d
(min
)
15
31
Rel
ativ
e 9
NM
N
M
23
NM
75
N
M
91
30
18
30
inte
nsiti
es
6 4
NM
19
6
59
19
85
31
20
31
(arb
itrar
y 2
6 N
M
14
11
46
41
75
27
21
33
75
unita
) N
.M.
9 N
M
15
18
36
47
74
30
20
34
100
2 N
.M.
6 4
12
21
34
46
67
28
17
30
a Id
entif
icat
ion o
f ind
ices
:
Not
mea
sura
ble.
Y
U
1890 DEVAUX ET AL.
As is clearly shown in Table 111, the relative intensities of the C signals corre- sponding to PBTP carbons decrease progressively. The new C3,8 band appearing in the spectra is due to the substitution of a butylene group by a bisphenol A. This confirms the IR and proton NMR observations.
The splitting of the C4,7 peak at 132.9 ppm into two symmetrical absorptions at f0.5 ppm of the initial band is also the result of the substitution. Therefore it is evident that two identical aromatic substitutions can cancel the splitting and replace the NMR absorption a t 132.9 ppm.
As a result of a single aromatic substitution, the observed chemical shift variations of the ortho- and meta-carbons (C, and C,) are identical. Only one peak at 129.6 ppm appears for unsymmetrically substituted terephthalic groups. We also note that the absorption at 128.9 ppm characterizes the C5 and CS car- bons of the terephthalic unit substituted either by two butylene groups or by two bisphenol-A groups.
The C2,g peak corresponding to the carbon of the CH2-0 groups of PBTP is located at 65.7 ppm. The two new bands at 68.1 and 68.7 ppm probably result from the appearance, in the copolyester, of aliphatic-aromatic carbonate and aliphatic-aliphatic carbonate units:
0
This splitting would appear to be logical since 13C NMR permits differentiation of substitution effects up to a distance of a t least four atoms.8
The results given in Table IV confirm the qualitative interpretation deduced from the results of Table 111. We, indeed, observe the presence of three types of carbonate ester units (Ci,12) bearing respectively two butylene groups (156.3 ppm), one butylene and one bisphenol-A (155.8 ppm), and two bisphenol-A groups (154.8 ppm).
signals of the bisphenol-A groups are split into two bands due to the simultaneous presence of bisphenol-A carbonate and bispheno1-A- terephthalate units in the copolyester.
The behavior of the Ck,g peak is similar to that of C ~ , J O of the PBTP group: a weakening and broadening of the peaks is observed at long reaction times.
The Ch8, Cg, and C; bands remain practically unchanged during the reaction and, therefore, the C; band may be chosen as a measure of the solution concen- tration for the quantitative analysis described in the next section.
The Ci,ll and
Statistical Structure Analysis
As is shown in the preceding paper,l exchange reactions in molten PC-PBTP mixtures lead to the formation of a four-component copolycondensate which can be represented by the general formula
[(AlB1),-(A2Bl),I,-[(A1B2)z-(A2B2)wln
PC-PBTP TRANSESTERIFICATION. I1 1891
TABLE V Fractions of Triads Centered on Bl in Molten PC-PBTP Blend Determined by Proton NMR and
Dcarees of Randomness"
Reaction time ( m i d / A ~ R , A ~ / A ~ B ~ A ~ / A ~ B ~ A ~ hI
Theoretical values for a 0.29 0.50 0.22 1 statistical coDolvester
a Temperature of reaction 253°C; 50/50 by weight = 46.4h3.6 mole ratio.
Al: butylene, -(CH2)4-; A2: bisphenol-A, *-F@ CH,
-0- By: carbonate, -9-C-0- II 0
B1: terephthalate, 4 - C
0 0 II
The meanings of the various indices are given in the preceding paper. In the case of PC-PBTP copolyesters, Al, B1, A2, and B2 are identified in the key to Tables V and VI. As already shown in ref. 1, it is possible to determine from the dyad mole fractions obtained from NMR a degree of randomness B:
Similarly, and taking into account that fA,B,Ak = fAkR,A,, the triad analysis gives the degree of randomness BB,:
BB, = fA,B,Ak c - (i # k) (i:l ;A,)
TABLE VI Fractions of Dyads in Molten PC-PBTP Blend Determined by I3C NMR and Degrees of
Randomnessa
Reaction t i m e b i n ) F A ~ I ~ ~ F A ~ B ~ F A ~ H ~ F A ~ B ~ H
Theoretical values for blend 0.536 0 0 0.464 0 Experimental values for 31 0.49 0.09 0.05 0.35 0.36
Theoretical values for 0.29 0.25 0.25 0.22 1 statistical copolyester
F A ~ B ~ : C3.8, FA~B?: C&O. a PCPBI'I' ratio, A], A2. €31, B2 as in Table V. I3C species studied for FA~B~: C3.8, F A ~ R ; C2,9.
1892 DEVAUX ET AL.
where fAiB,Ak stands for the fraction of AiB;Ak triads having a central B; unit.
The triad fractions around the terephthalate unit (B1) determined from proton NMR and calculated by the method of Gouinlock et al.7 are reported in Table V, together with calculated values of the degree of randomness B B ~ for 50/50 PC-PBTP mixtures (50/50 parts by weight = 46.4b3.6 mole ratio). The theo- retical values of BB1 are also given for the unreacted PC-PBTP mixture and for a statistical copolyester of the same composition.
From the results of the 13C NMR (Table 111) four dyad fractions ( F A ~ B ~ , F A ~ B ~ , F A ~ B ~ , and F A ~ B ~ ) were calculated for the 50/50 PC-PBTP copolyester. These figures are reported in Table VI, where the calculated values are compked with the theoretical values for the unreacted polycondensate mixtures and for the statistical copolyester. In Table VI, the I3C species, upon which the calculations are based, is also given, as well as the calculated and theoretical values for the degree of randomness B. The data obtained after 15 min of reaction were not taken into consideration in these calculations because it is very difficult to measure the induced modifications with sufficient accuracy at this reaction time.
We note from Tables V and VI that, as the reaction proceeds, the degrees of randomness B and B B ~ tend toward unity, which is the anticipated theoretical value for a statistical copolyester.
Average Sequence Lengths
As developed in the theoretical study, the average sequence length can be computed from either dyad or triad determinations provided that a substitution on one side of the central unit of a triad is not affected by a substitution on the other side.
From proton NMR determination of triads based on the terephthalate unit B1, we calculated the evolution of the mean lengths of the butylene terephthalate ( x ) and bisphenol-A terephthalate (y ) sequences (Table VII).
l3C NMR analysis allowed us to calculate the evolution of the mean sequence lengths x , y, z , and w of the copolyester. The results of these calculations are summarized in Table VIII, together with the theoretical values for the unreacted
TABLE VII Evolution of the Average Length of Butylene (x) and Bisphenol-A (y) Terephthalate Sequences
as a Function of Reaction Time a t 253'Ca
Reaction time (min) X Y
~ ~~~ ~~ ~
... Theoretical value for blend 166 Experimental values 30 6.46 1.15
a Calculation based on dyads.a Mixture of homopolyesters as in Table VII. x is the average length of butylene terephthalate sequences, y the average length of bisphenol-A terephthalate sequences, z the average length of butylene carbonate sequences, w the average length of bisphenol-A carbonate sequences, m the average length of terephthalate sequences, and n the average length of carbonate sequences.
blend and for the statistical copolyester. We note excellent agreement between the x and y values computed from dyad and triad analysis, respectively. As the reaction proceeds, the copolyester sequence distribution becomes statistical.
DISCUSSION
The infrared and NMR analyses give a very coherent picture of the PC-PBTP copolyester structure which results from an exchange of aliphatic ester (A1B1)
and aromatic carbonate (AzB2) sequences simultaneous with the bisphenol-A terephthalate (A2B1) and butylene carbonate (AlB2) formation in equimolecular amounts.
The statistical analysis (degree of randomness as well as mean sequence lengths) clearly shows that the exchange reaction gradually leads to the formation of a copolyester with a random distribution (statistical copolyester). As shall be discussed in a future paper, this salient point will greatly facilitate the cal- culation of the kinetic equations.
The appearance, at the beginning of the exchange reaction, of an insoluble fraction results from the initial formation of a copolyester with a long insoluble PBTP sequence. As the reaction proceeds, the PBTP sequences become shorter and a fully soluble product is progressively formed. From the statistical analysis, we may conclude that a sequence length shorter than approximately 7 units is necessary to obtain a complete solubility of the copolyester.
We note finally that the m and n values in Table VIII are, in fact, upper limits since thermal degradation2 was not taken into account.
CONCLUSION
We have shown that an exchange reaction takes place in mixtures of molten PC and PBTP. At short reaction times, block copolyesters with reduced solu- bility are formed. As the reaction proceeds, the sequence distribution in the copolyester becomes statistical and a completely soluble product is obtained.
In subsequent papers of this series, the mechanism of the exchange reaction
1894 DEVAUX ET AL.
will be deduced from the study of model reactions and the overall transesterifi- cation kinetics determined on blends of various compositions.
References
1. J. Devaux, P. Godard, and J. P. Mercier, J. Polym. Sci. Polym. Phys. Ed., preceding paper,
2. J. Devaux, P. Godard, and J. P. Mercier, Makromol. Chem., 179,2201 (1978). 3. F. Gallez, R. Legras, and J. P. Mercier, J. Polym. Sci. Polym. Phys. Ed., 14,1367 (1976). 4. N. R. Sorensen and T. W. Campbell, Preparative Methods ojPolymer Chemistry, Interscience,
5. Y. G. Urman, S. G. Alekseyeva, and I. Y. Slonim, Vysokomol. Soedin. Ser. A , A19, 299
6. A. Y. A. Yakubovitch, C. Y. A. Gordon, L. I. Masslennikova,E. M. Grobman,K. I.Tret’Yakova,
7. E. V. Gouinlock, R. A. Wolfe, and J. C. Rosenfeld, J. Appl. Polym. Sci., 20,949 (1976). 8. K. J. Ivin, J. Polym. Sci. Polym. Symp., 62,89 (1978).
20,1875 (1982).
New York, 1961.
(1977).
and N. I. Kokoreva, J. Polym. Sci., 55,251 (1961).
