Compensation laws and phase segregation in polymer blends

Solid State Communications, Vol. 69, No. 6, pp. 707-711, 1989. Printed in Great Britain.

0038-t098/89 $3.00 + .0( Pergamon Press ph

COMPENSATION LAWS AND PHASE SEGREGATION IN POLYMER BLENDS J.J. del Val and J. Colmenero Departamento de Fisica de Materiales, Facultad de Quimica, Universidad det Pais Vasco, Apartado Postal 1072, 20080 San Sebastifin, Spain and C. Lacabanne Laboratoire de Physique des Solides, Universit6 Paui Sabatier, 31062 Toulouse Cedex, France

(Received 2 February 1988 by R. Fieschi) Thermally Stimulated Depolarization Current (TSDC) and Thermally Stimulated Creep (TSC) as well as compensation law analysis have been used in order to study miscibility/compatibility problems in polymer blends not only in the glass-transition range, but also in the fl-relaxation range. The samples used were blends of Poly(Vinyl Chloride) (PVC) and Chlorinated Poly(Ethylene) (CPE) with different chlorine distribution, i.e., statistically random distribution in one case and Blocky distribution in the other. The results obtained show that the blends studied are immiscible as far as cooperative motions liberated at the glass-transition are concerned. However, some short range miscibility can be deduced from/%relaxation behaviour in the case of PVC/CPE blocky samples. adequate in the study of both, glass-transition proces~ (a-relaxation) and sub-Tg secondary processes [6-13] POLYMER-POLYMER miscibility or compatibility This kind of technique has been proved to show more are badly defined concepts which are frequently used accuracy than the classical mechanical or dielectric to describe that a polymer-polymer alloy behaves as techniques in the detection and study of the differenl only a one phase material. As a consequence, the most molecular motions involved in a given relaxatio~ extended criterium for determining the miscibility process. This is a consequence of the resolution capabetween two amorphous polymers is the presence of a bility of neighboring processes owing to the lowunique glass-transition process at a temperature inter- equivalent frequencies (,-~ 10-3 Hz) of these techmediate between those of pure unblended polymers niques and to those of the fractional polarizationj [1-5]. This behaviour is in general shown by means of stress procedure. Differential Scanning Calorimetry (DSC) technique. This experimental procedure allows to separate However, the structural homogeneity in a multicom- the different contributions involved in a given process ponent system can be considered as a scale problem. and to characterize them by kinetic parameters capable The shorter the scale the higher the level of homo- of interpretation in terms of compensation laws [6-9. geneity. In this sense, the glass-transition process corre- 12-15]. These compensation laws [16, 17], which allox~ sponds to molecular motions involving several struc- to distinguish the different relaxation processes al tural units. Therefore, the study of these kind of molecular level, should be used as powerful tools in processes gives information about the homogeneity at order to study miscibility or comptability in polymer that scale level. The study of secondary relaxation blends. processes like/~-relaxation, which are related to local On the other hand, in recent years large attention molecular motions, could give information about has been paid to the influence of the preparation ot miscibility or compatibility at shorter scale levels. polymer blends on their morphology and properties However, DSC is not an adequate technique in order [2-5]. Blends of Poly(Vinyl Chloride) (PVC) and to study these kind of secondary processes. In this Chlorinated Poly(Ethylene) (CPE) have been found to sense, thermally stimulated techniques like Thermally present good impact strength so that several studies Stimulated Depolarization Current (TSDC) or Ther- have been carried out on the effect of the preparation mally Stimulated Creep (TSC) have been shown very 1. INTRODUCTION

707

C O M P E N S A T I O N LAWS A N D P H A S E S E G R E G A T I O N

708

procedure on their morphology [l 8, 19]. These characteristics have a main role in the mechanical properties and therefore relaxation experiments performed on PVC/CPE blends of different microstructure and morphology could be very interesting. The fine structure of the different relaxation modes will permit the correlation of the chlorine distribution in polyethylene and the miscibility of the blends. To do this we have used in this work both, thermally stimulated techniques and compensation law analysis. 2. E X P E R I M E N T A L

2.1. Thermally stimulated creep and depolarization currents Both, the basis of TSC and TSDC techniques and the general operating principle have been widely described in other papers [7-9, 12-14]. The measurement procedure used in this work to study relaxation processes which occur between 100 and 390 K can be summarized as follows. The sample is cooled from high temperature to 100K at a cooling rate q,, of 20 K/min. A mechanical stress a (electric polarizing field Ep) is set-on at a temperature T~,o,, (Tp,on) during cooling and set-off when the sample reaches a temperature To,on < T~,on (Tp,on < Tp,olr) where the molecular motions that one wishes to consider are completely hindered. This procedure allows that the structural units (dipoles in TSDC case) which have a relaxation time at any temperature between To.on (Tp.o,) and T~.olr(Tp,on-) short enough to orientate and contribute to the strain ? (polarization P) of the sample. The sample is subsequently heated at a constant rate, qh, of 10 K/min so that the structural units (dipoles in the TSDC case) can return to random orientation. The corresponding time derivative of the strain 7 (depolarization current J in the TSDC case) is recorded versus sample temperature. If the interval To,on Ta,ofr(Tp.on - Tp,orr)covers a wide temperature range, TSC and TSDC responses are respectively known as TSC and TSDC whole spectra. In polymer, TSC and T S D C whole spectra are in general too broad to be associated to only one Debyelike relaxation process, that is to only one relaxation time [10-13, 20]. However it is well known that these techniques can separate different elementary processes that contribute to such a complex relaxation by means of the fractional stress/polarization procedure. In this procedure the mechanical stress (polarization field) are only applied to the sample in a narrow temperature window, T,,o, - To.oft = 5 K (Tp,on -- Tp.o~ = 5K), in our case, during cooling. By varying the interval To,o. - T,,o~ (Tp,o, - Tp,oU) in the temperature region where global TSC or TSDC spectra respectively -

Vol. 69, No. 6

appear, these ones can be resolved in quasielementary peaks. So obtained peaks can be considered in a first approximation, as activated Debye-like peaks [10, 11]. The retardation time r (T) at a temperature T be obtained from the relationship [10-13]. r(T)

=

~'(T)/I';'(T)I

(1)

for the TSC case, and

r(T)

= P(T)/IJ(T)I

(2)

for the TSDC case. Usually the temperature dependence of the relaxation times can be well approximated by the Arrhenius equation r(T)

=

~0 exp (E/KBT)

(3)

where r0 is a preexponential factor, KB the Boltzmann's constant and E the activation energy for the process. If the logarithm of the preexponential factor z0 varies linearly with the activation energy E, as is often experimentally found in polymeric solids, one observes a compensation law [15-17]. ro =

r, exp ( - E / K B T , ) .

(4)

Then, expression (3) becomes

,ex I (, ')J T

(5)

where T, and 7], are phenomenological parameters, respectively known as compensation time and compensation temperature. The physical significance of these parameters is not yet clear; but the most general interpretation consists of considering them as characteristic of processes arising from the same molecular mechanism [15-16]. 2.2. Samples Two blends of commercial PVC produced by B.F. Goodrich Chemical Company (EP-103-F-76) with 15 wt % CPE were prepared by precipitation of dilute solutions in dichlorobenzene. CPE had approximately the same chlorine content (30-32 wt %) but different chlorine distribution. In the former case, statistically random distribution was obtained by solution chlorination; in the latter one, preparation by suspension chlorination gave rise to Blocky distribution of chlorine. The samples are designated S and B for PVC - Statistically CPE and PVC - Blocky CPE respectively. Rectangular plates with dimensions 65 x 7 x 0 . 2 m m 3 and 8 x 5.5 x 0 . 2 m m 3 were prepared for TSC and T S D C experiments respectively. 3. R E S U L T S A N D D I S C U S S I O N Global TSC and T S D C spectra are schematized in

Vol. 69, No. 6

C O M P E N S A T I O N LAWS A N D PHASE S E G R E G A T I O N

709

\

6 -50

) ........

0 4

'~

B

TSDC

.

.

.

.

.

'~

.

.

.

.

\\A

.

-/

-100

< ~ 2 -150

0 6

\. I

~ 4

1

I

I

3

I

I

5

\1 E(eV)

% 2 Fig. 2. Compensation plot for the kinetic parameters obtained by TSC (full symbols) and TSDC (empty symbols) in ct relaxations of PVC and CPE for pure PVC (O, O) and PVC/Statistically CPE (A, A) and PVC/Blocky CPE ( L El) blends.

0

0

100

150

200

250 T(K)

300

350

Fig. i. Global TSC and TSDC spectra obtained around ~ relaxations of PVC and CPE and fl relaxation of PVC for PVC/Statistically CPE (a) and PVC/Blocky CPE (b) blends. Fig. 1 for samples S and B. Maxima are found in three temperature regions assigned to the fl relaxation of PVC (185-205K) [6, 7, 20], the ~ relaxation (glass transition) of CPE (230-255K) [21, 22] and the relaxation (glass transition) of PVC (350-365 K [6, 7, 20]. At temperatures higher than the glass transition of PVC, sample B shows another maximum in the TSDC case. This maximum has been attributed to the motion of free charges in PVC and is known as ~ peak [7]. The two glass transitions of the pure components are observed in the two blends. So, according to the rule of a unique glass transition above commented, the two homopolymers are as far as cooperative motions liberated at the glass transition are concerned. In order to obtain the distribution of the kinetic parameters associated to each whole relaxation process, fractional stress/polarization technique was used to decompose them into elementary spectra and the kinetic parameters E and z0 were calculated from equation (!) and (2). First of all it is important to point out that both, the high values of E and the exceptional low values of r0 obtained in the glass-transition range (for example E ~ 5eV and r0 ~ 10-8°s, see below) cannot be associated to thermally activated molecular site changes between two equilibrium positions separated by a true molecular energy barrier. These values of E

and r0 could be only explained in terms of cooperative motions corresponding to the long-range conformational changes characteristic of the a-relaxation. In this sense, although a definitive theoretical model has not yet been built, there are different approaches to interpreting the unrealistic values of both E and %. These range from cooperative or correlated two site models [23, 24] to free volume models like what has been previously developed by some of the authors [25]. In the case of cooperative site change models, an actual activated process associated to a true energy barrier, is assumed. Then the experimental parameters E and r0 are related to the corresponding ones of the true molecular activated process through a correlation index. From a completely different point of view, free-volume models assume that molecular transport around the glass-transition is driven by the fraction of free-volume. In this framework E and r0 are directly related to free-volume-characteristic parameters. In our opinion, the free-volume models are the most adequated approaches to the molecular transport in the glass-transition range while cooperative models would be appropriate explanations to the anomalous values of E and % frequently obtained at the fl-range (see below). In any case E and r0 should be considered only as apparent kinetic parameter without the physical meaning habitually used in the framework of the activated rate processes. However, we can use these phenomenological apparent kinetic parameters as useful tools in order to compare the TSDC behaviour of different systems. This is the point of view followed in this work. Figure 2 shows a compensation plot for the kinetic parameters obtained by TSC and TSDC in both relaxation regions of PVC and CPE. As can be seen

C O M P E N S A T I O N LAWS A N D PHASE S E G R E G A T I O N

710

peaks isolated in the glass transition region of PVC show wide dispersion in their characteristic kinetic parameters (2-6eV for E and 10 25-10 8o s for %). In the case of the two blends here studied, the values of E and z0 corresponding to the or-region of PVC follow the same compensation law than the previously obtained for the pure PVC [7] with compensation parameters T, =

356.5K

z,

=

21.9s.

z, =

!1.4s.

(7)

These facts reinforce the above mentioned conclusion of incompatibility of PVC and statistically and blocky CPE. A complete segregation of two PVC and CPE phases might be concluded in a first step for the two blends. After the study of the motions involved in the relaxations of the two components of the blends, we will now pay attention to local molecular motions taking place in the fl relaxation of PVC. In a recent work [7], we found that fl relaxation in PVC is composed by two relaxation modes characterized by two different compensation laws defined by the following parameters: T,.I =

200K

z,.i

=

T,2 =

245K

z,2 =

man°~Oan"o

I

I

o.a

I

I

0.6

(6)

So, blending with CPE does not alter the molecular mechanism of the ct relaxation of PVC. On the other hand, peaks isolated in the glass transition region of CPE have shown kinetic parameters less dispersed than in the above commented PVC case (0.8 to 1.7eV for E and 10 -~n to 10 -33 S for %). The compensation law obtained is the same for the two blends S and B, with parameters: T, = 260.8K

-2C -

Vol. 69, No. 6

0.13s

(8)

105.

(9)

This behaviour has been observed in different kinds of PVC samples: syndiotactic, atactic and isotactic PVC samples [6, 26] and commercial PVC used to be blended with CPE in this work [7, 26]. The molecular origin of these two relaxation modes (two compensation laws) is not clear. It will be the subject of future works. Nevertheless, another conclusion carried out from Refs. [7] and [26] is that, in this region, mechanical and dipolar motions are equivalent: the shape of TSC and TSDC curves is not different and obtained E and % parameters are not distinguishable. We can see in Fig. 1 that both TSC and TSDC whole spectra in this region are very similar. In this sense, the decomposition of the whole spectra into elementary peaks was done only in the TSC case for the two blends. The kinetic parameters obtained from the analysis of the decomposition of the global TSC

I

E(eV)

Fig. 3. Compensation plot for the kinetic parameters

obtained by TSC in the fl relaxation of PVC for pure PVC (O) and PVC/Statistically CPE (A) PVC/Blocky CPE (m) blends. spectra show wide dispersion in both cases (0.2-0.8 eV for E and 10 7-10-18 S for %). A compensation plot of them is shown in Fig. 3. It is clear that sample B shows without modification the two compensation processes characteristic of the fl relaxation of PVC. This suggests that segregation between the two phases is complete in the PVC-Blocky CPE blends, even in the range of localized motions typical of this secondary relaxation. Contrarily, PVC-Statistically CPE blend presents only the first of the two compensation processes. In this case, the relaxation behaviour of PVC has been altered by the presence of Statistically CPE chains. In our point of view, some short range compatibility might exist, explaining this result. One of the two processes may have been hindered by the presence of CPE units which do not allow PVC short segments to move.

4. CONCLUSIONS PVC-Statistically CPE and PVC-Blocky CPE blends show two ~t relaxation processes corresponding respectively to the glass transitions of CPE and PVC in the order of increasing temperatures; so, the two blends are immiscible in the level of long range motions taking part at the glass transitions. The fl relaxation of PVC is obtained without modification in the PVC-Blocky CPE blend confirming immiscibility even at short range. However, in the other blend, the compensation law of the fl relaxation of PVC is partially modified, allowing us to conclude the existence of some short range miscibility. As general conclusion, we think that short range molecular motions (i.e., fl relaxation) may give complementary information about the miscibility/ immiscibility level of two homopolymers found to be not compatible by the unique glass transition rule. To do this, thermally stimulated techniques, TSC and TSDC, and the compensation law analysis have been shown as very useful tools.

Vol. 69, No. 6

COMPENSATION LAWS AND PHASE SEGREGATION

Acknowledgements - J.J. del Val is indebted to the Patronato de la Universidad del Pais Vasco for a fellowship which initiated this work. J. Colmenero and J.J. del Val thank Gipuzkoako Foru Aldundia for partial financial support.

11. 12. 13. 14.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

S. Krause, J. Macromol. Sci.-Revs. Macromol. Chem. 7, 251 (1972). D.R. Paul & S. Newman, Polymer Blends. Academic Press, New York (1978). O. Olabisi, L.T. Robeson & M.T. Shaw, PolymerPolymer Miscibility. Academic Press, New York (1979). L.E. Nielsen, Mechanical Properties of Polymers and Composites. Marcel Dekker, New York (1974). J.A. Manson & L.H. Sperling, Polymer Blends and Composites, Marcel Dekker, New York (1976). J.M. Barandiarfi.n, J.J. del Val, C. Lacabanne, D. Chatain, J. Millfin & G. Martinez, J. Macromol. Sci.-Phys. B22, 645 (1983). J.J. del Val, A. Alegria, J. Colmenero & C. Lacabanne, J. Appl. Phys. 59, 3829 (1986). J.J. del Val, A. Alegria, J. Colmenero & J.M. Barandiarfin, Polymer 27, ! 771 (1986). J. Colmenero, A. Alegria, J.M. Alberdi, J.J. del Val & G. Ucar, Phys. Rev. B35, 3995 (1987). P. Braunlich (Ed.), Thermally Stimulated Relaxation in Solids, p. 135. Springer Verlag, New York (1979).

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26.

711

G.M. Sessler (Ed.), Electrets, p. 81, Springer Verlag, New York (1979). C. Lacabanne, D. Chatain & J.C. Monpagens, J. Macromol. Sci.-Phys. B13, 537 (1977). J.C. Monpagens, D. Chatain & C. Lacabanne, Electrostatics 3, 87 (1977). P. Demont, D. Chatain, C. Lacabanne, D. Ronarch & J.L. Moura, Polym. Eng. Sci. 24, 127 (1984). C. Lacabanne, D. Chatain, J.C. Monpagens, A. Hiltner & E. Baer, Solid State Commun. 27, 1055 (1978). J.P. Crine, J. Macromol. Sci.-Phys. B23, 201 (1984). J.D. Hoffmann, G. Williams & E. Passaglia, J. Polym. Sci. C14, 173 (1966). J.L. Work, Polym. Eng. Sci. 13, 46 (1973). G. Menges, N. Bendtsen & N. Opferman, J. Plast. Rubber Proc. 4, 156 (1979). N.G. McCrum, B.E. Read & G. Williams, Anelastic and Dielectric Effects in Polymeric Solids. Wiley, New York (1976). F. Lindhardt, Kunststoffe 59, 787 (1969). G. Humbert, B.M. Quenum, P. Berticat & G. Valet, Makromol. Chem. 175, 1611 (1974). A.J. Owen & R. Bonart, Polymer 26, 1034 (1985). T.V. Ramakrishaw & M. Raj Lakshmi (Eds.), Non-Debye Relaxation in Condensed Matter. World Scientific, Singapore (1987). A. Alegria, J.M. Barandiarfin & J. Colmenero, Phys. Status Solidi (b) 120, 349 (1983). J.J. del Val & J. Colmenero, Polym. Bull. 17, 489 (1987).