Studies of Polylactone/Poly(vinyl Chloride), Polylac tone/Poly (vinyl Fluoride), and

Polylac t one/Pol y (vin ylidene Fluoride) Blends

MADELEINE AUBIN and ROBERT E. PRUD’HOMME,* Groupe de recherche sur les macromolBcules, Chemistry Department, Lava1

University, Quebec, Canada G l K 7P4

Synopsis

It is shown that polyvalerolactone/poly(vinyl chloride) (PVL/PVC) blends are miscible over all compositions since a single glass transition temperature Tg is observed, intermediate between those of pure PVL and pure PVC. Melting points, enthalpies of fusion and morphologies of PVLPVC blends are also reported. It is also shown that polyvalerolactone, poly(a-methyl-a-n-propyl- @-propiolactone), poly(a-methyl-a-ethyl-@-propiolactone), and poly(capro1actone) are immiscible with poly(viny1 fluoride) and poly(viny1idene fluoride), despite the fact that all these polylactones are miscible with PVC. Differences in electronegativity, in atomic radius, and in molar attraction between the fluoride and the chlorine atoms are probably responsible for this difference in be- havior.

INTRODUCTION

It now well established that polycaprolactone (PCL) is miscible with a large number of polymers, and particularly with poly(viny1 chloride) (PVC).I The miscibility of PCL and PVC has been attributed to a specific interaction between the carbonyl groups of PCL and the a hydrogens of PVC.2 Indeed, Fourier- transform infrared spectroscopy3 clearly indicates that the PCL-PVC interaction involves the carbonyl groups of PCL and inverse gas chromatography indicates that it involves the chlorine atoms of PVC.*

On the basis of the presence of such an interaction, we previously postulated that a large number of polylactones should be miscible with PVC. We were then able to show that the poly(a,a-substituted-P-propiolactones) are miscible with PVC since for such blends, a single glass transition temperature Tg, a depression in melting point T,, and an increase in density were ~ b s e r v e d . ~

More recently, in a preliminary account of measurements reported fully in this paper, we indicated that polyvalerolactone (PVL) is also miscible with PVC? according to the accepted Tg criteri0n.l

However, if PVC is miscible with such a large number of polylactones, it is expected that other poly(viny1 halides) will present the same phenomenon. It is the purpose of the present paper to verify the miscibility or the immiscibility of polylactone/poly(vinyl fluoride) (PVF), and of polylactone/poly(vinylidene fluoride) (PVF2) blends by measurements of Tg. More specifically, it will be shown that PVL/PVF, PVL/PVFz, PMPPLPVF, PMPPLPVFz, PMEPL/ PVF2, PCL/PVF, and PCL/PVFZ blends are immiscible, where PMPPL is poly(a-methyl-a-n-propyl-@-propiolactone) and PMEPL is poly(a-methyl- a-ethyl-P-propiolactone).

* To whom correspondence should be addressed.

Journal of Polymer Science: Polymer Physics Edition, Vol. 19, 1245-1253 (1981) 0 1981 John Wiley & Sons, Inc. CCC 0098-1273/81/081245-09$01.00

1246 AUBIN AND PRUD’HOMME

EXPERIMENTAL

In Table I, a list of the different polymers used in the present study is given, along with Tg, T,, and molecular weight. Several of these samples were obtained from the Aldrich Chemical Corporation, including PVF (No. 18,266-4), PVF2 (No. 18,270-2), and PCL (No. 18,160-9). PVC was supplied by the Shawinigan Chemical Corporation. PVL, PMPPL, and PMEPL were prepared in our lab- oratories by ring-opening polymerization.- These polymers have the following values of intrinsic viscosity: PVF 1.7 dL/g (in N-methyl-2-pyrrolidone at 298 K), PVF2 0.96 dL/g (in dimethyl formamide at 298 K), PCL 0.26 dL/g (in tet- rahydrofuran at 298 K), PVL 0.53 dL/g (in tetrahydrofuran at 298 K), PMPPL 0.48 dL/g (in tetrahydrofuran at 298 K), and PMEPL 0.55 dL/g (in CH2C12 at 303.5 K). The Tg and T, values reported in Table I were determined in the present study. Tg was determined as the temperature corresponding to the half-height of the ACp increase in the glass transition zone. T, was taken at the end of the melting curve. All samples are of relatively high molecular weight.

Polymer films were generally cast from 1% solutions. Various solvents were used: dimethylformamide (DMF) for PVF, PVF2, and blends involving these polymers; tetrahydrofuran (THF) or CH2Cl2 for PVL, PVC, and their mixtures; N-methyl-2-pyrrolidone for PMPPL, PVF, and their blends. Solvent evapo- ration was normally conducted at room temperature. The resulting films were dried under vacuum until they reached constant weight. Film thicknesses were about 40 pm.

Differential scanning calorimetry (DSC) measurements were conducted on the solvent-cast films and on films subjected to the following thermal regimes: for blends involving PVL, PVC, and PVF, the solvent cast films were melted at 360 K for 5 min, and quenched at 140 K in the DSC; for samples involving PVF2, the solvent-cast films were melted at 460 K for 5 min before quenching at 140 K, finally PMPPLPVF solvent-cast blends were melted at 380 K for 5 min before quenching at 140 K. Generally, the DSC results are the same whatever the method of preparation of the films. Unless otherwise specified, the results re- ported herein were obtained on the quenched films.

DSC measurements were conducted on a Perkin-Elmer DSC-i ,$paratus

TABLE I Characterization of Polymers

Molecular Polymer Acronym Repeat unit T,/K T,/K weighta

Poly(viny1 chloride) PVC -CH&HCl- 338 -a* 80,OOO (aw) Poly(viny1 fluoride) PVF -CH&HF- 336 --- ... Poly(viny1idene fluoride) PVFz -CHzCF2- 233 444 ... Polycaprolactone PCL -(CHZ)~-COO- 215 330 20,OOO(7i;i,) Polyvalerolactone PVL -(CH2)4-C00- 206 332 33,000 (aw) Poly(a-methyl-a-n-propyl- PMPPL -CH&(CH3)COO- 275 353 131,000 (aw)

C3H7 80,000 (an) P-propiolactone) I

P-propiolactone) I Poly(a-methyl-a-ethyl- PMEPL -CHzC(CH&OO- 264 406 80,000 (aw)

CzHs

a a,, is the number-average molecular weight and aw the weight-average molecular weight.

POLYLACTONE/PVC BLENDS 1247

which was calibrated with n-octane (216 K) and indium (429 K). A heating rate of 20 K/min was used in all cases.

In this paper, the following terminology is adopted: a sample is declared miscible when it gives for all compositions a single Tg intermediate between those of its pure components, even if crystals of components 1 and/or 2 form. A sample presenting two Tg’s at a given composition is declared immiscible. According to this terminology, it suffices to observe two Tg’s at one composition in order to conclude that the sample is immiscible, even if the possibility of observing a single Tg at other compositions is not excluded.

RESULTS AND DISCUSSION

PVL/PVC Blends

DSC scans in the Tg and T,,, zones are presented in Figure 1 for PVL, PVC, and their mixtures. All blends give a single Tg, which increases regularly with composition. PVL has a Tg of 206 K, which is very close to that of PCL.l PVC has a Tg of 338 K, which is slightly lower than the published

The regular increase in Tg of the blends is shown in Figure 2 and in Table 11. These results are satisfactorily represented (Fig. 2) by the Fox equationlo

where Tg, Tgl, and Tg2 are the glass transition temperatures of the blend, of PVL, and of PVC, respectively, and w1 and w2 are the corresponding weight frsctions.

w

. . . .

R = 4

. . R = 4

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . R =4

: .

........ .......... R = 4 + ....................... R=10 . . . .

......................

90% P V C

75% PVC

50% PVC

25% PVC

PVL

190 210 230 250 270 290 310 330 350 370 390

TEMPERATURE (K)

Fig. 1. DSC curves for PVLPVC blends. The lower curve is for PVL and the higher one for PVC. The intermediate curves, from bottom to top, correspond to blends with 25,50,75, and 90% by weight of PVC. The Tg portions of the curves were recorded at higher apparatus sensitivity ( R = 4) than those in the T, zone (R = 10 or R = 20).

1248 AUBIN AND PRUD'HOMME

10 20 30 40 50 60 70 80 90 100

% PVC

Fig. 2. Glass transition temperature Tg of PVLPVC blends as a function of weight percent PVC. The solid line is drawn according to eq. (1).

The experimental data can alternatively be represented by the Gordon-Taylor equationll or the Bueche-Kelley equation.12 This behavior indicates extensive mixing between PVL and PVC in the blends.

Figure 1 also indicates that the width of the Tg zone changes with composition. For PVL and PVC, the width of the transition zone is about 10 K. But for most blends, the width of the transition is of the order of 20 K. A similar broadening was observed for PMPPLRVC blend^.^ It indicates partial miscibility at the molecular leve1,13-15 and has been interpreted as due to microheterogeneities where local composition fluctuations are in excess of normal density and tem- perature fluctuations.l3 For PVLPVC blends, the presence of a single Tg for each composition certainly indicates an extensive degree of miscibility, but the broadening of the transition zone suggests that the miscibility does not extend down to the molecular level.

Blends containing 50% or less PVC present a PVL melting peak while other do not. In fact, the 50-5070 blend corresponds to a critical composition. When

TABLE I1 Glass Transition Temperature, Melting Point, and Enthalpy of Fusion of PVC, PVL, and Their

Mixturesa

0 206 332 65, 81b 25 229 332 51 50 261 332b 33b 75 302 ... ... 90 318 ... ...

... ... 100 338

a Uncertainties are estimated to be f 2 K in Tg and T , values and 1 3 J g-' in AH. Value obtained for a film prepared from THF solution.

POLYLACTONE/PVC BLENDS 1249

prepared from THF or CH2C12 solutions, this mixture presents a well-defined melting peak, as indicated in Table 11. When the sample is melted and quenched, the melting peak disappears, as is shown in Figure 1. In the former case, the crystallization is conducted in a medium (the solution) having a low viscosity, where it is favored. In the latter case, the crystallization cannot occur owing to the high viscosity of the polymer liquid medium and to the efficiency of the quenching process. This interpretation is in agreement with earlier observations made on high-molecular-weight polyethylene, where drastic changes in mor- phology were seen between samples prepared from the melt and from solution.16 Only the latter presented a well-defined spherulitic morphology.

The 75-259’0 blend can also be partially quenched. But additional crystalli- zation occurs during the heating cycle in the DSC, as shown by the exothermic peak in Figure 1. When the 75-25% blend is melted and cooled to 290 K for 10 min, and then heated, a reduced melting curve is observed, indicating unam- biguously that the exothermic peak is a crystallization peak. A similar exo- thermic peak does not occur in the 50-5096 blend because it should appear in the Tg transition zone. The viscosity of this blend is then too high to allow crys- tallization.

The absence of crystallization for blends with high PVC content is due to ki- netics and not to thermodynamic requirements. It is due to the increase in Tg leading to an increase in the viscosity of the melt and to a slowing down of the crystallization rate. Proper annealing conditions could eventually lead to the crystallization of PVL, whatever the blend composition.

Table I1 indicates that the melting temperatures of the blends are the same as that of pure PVL. The absence of melting point depression for blends which are clearly miscible according to the Tg data can indicate a thermodynamic in- teraction parameter xl2 close to zero. If xl2 is zero, the enthalpy of mixing is zero, but the free energy (free enthalpy) of mixing is still negative owing to the entropic contribution. Miscibility is then produced thermodynamically under such conditions.

Another explanation of the absence of melting point depression would be possible if one admits the presence of a lower critical solution temperature (LCST) for PVLPVC at a temperature below the melting point of PVL. The sample would then exhibit two phases at 332 K, but a single amorphous phase at lower temperatures in the Tg zone. However, it is not known whether LCST exists in the PVL/PVC sample. It is surprising that the depression in melting point already seen in several other similar b l e n d ~ ~ , ~ J ~ J ~ is not observed here.

Table I1 also gives the enthalpy of fusion AH per gram of PVL in the blend for PVLPVC blends. AH decreases regularly as the PVC content increases, indicating that PVC is a macromolecular diluent that reduces the crystallization rate of PVL. Similar examples can be found in the l i t e r a t ~ r e . ~ J ~ ~ ~ ~

PVLIPVC blends having a melting point give rise to a well-defined morphol- ogy, as shown in Figure 3, where H, small-angle light-scattering (SALS) patterns are presented. The H, polarization mode indicates that the polarizer of the SALS apparatus is oriented vertically and the analyzer horizontally. Figures 3(A) and 3(B), corresponding, respectively, to pure PVL and to the 7525% blend, have the cloverleaf appearance characteristic of a spherulitic structure.21 From the intensity maximum of the patterns, spherulitic sizes of 8 and 14 pm were calculated. Such an increase in spherulitic radius with blend composition was

1250 AUBIN AND PRUD’HOMME

Fig. 3. H , small-angle light-scattering patterns for PVLPVC blends: (A) pure PVL, (B) 75-25%, (C) 50-50%.

also observed in PCLPVC blends by Khambatta et a1.,20 who attributed it to a decreasing nucleation rate due to the observed increase in Tg and to the thermal history of the blend. Figure 3(C) is of a different nature. It shows an intensity maximum rather than a minimum at the center of the pattern, as in Figures 3(A) and 3(B). This pattern is associated with a more disordered morphology, usually found for samples of low crystallinity. A rodlike model leads to this type of patternz2

Blends involving PVF or PVFz

PVL/PVF and PVLPVF2 blends can be prepared and their properties com- pared with those of the PVLPVC sample presented in the previous section of this paper. All measurements made on these two samples have been summarized in Table 111. For PVFz, a Tg of 232-235 K and a T, of 444 K were found, in agreement with the literature?J7 For PVF, a Tg of 336 K, a value slightly larger than that in the literature? was found.

For PVLPVF blends, we observed a Tg value of 207-209 K equal to that of pure PVL, within experimental error. The Tg of PVF could not be observed, probably because it was too close to the melting point of PVL. The AH values of PVL do not show a tendency to decrease with PW content. The unperturbed Tg of PVL in PVLPVF blends as compared to the Tg of pure PVL indicates that the PVLPVF sample is immiscible.

TABLE111 Glass Transition Temperature, Melting Point, and Enthalpy of Fusion of PVL, PVF, PVF2, and

of PVLPVF and PVLPVFz Mixturesa

PVF or PVF2 weight content (%) Tg (W T m (K) AH (J g-9

0 206 332 65 25% PVF 209 326 61 50% PVF 211 326 49 51% PVFz 234b 334,456 50,34 75% PVF2 232 325,450 50,36

100% PVF 336 ... ... 100% PVF2 232-235 444 29

a In this and the following tables, uncertainties in Tg and T, are estimated as f 3 K, and in AH as f 3 J g-l).

Value obtained for a film prepared from a DMF solution.

POLYLACTONEDVC BLENDS 1251

For PVL/PVFz blends, a single Tg was also observed at 234-238 K, in agree- ment with the Tg of pure PVF2 at 232-235 K. Two Tm’s were found for each blend. The first one at 334-343 K corresponds to the melting of PVL crystals and the second at 455-456 K to the melting of PVF2 crystals. Two AH’S are also reported in Table 111. The Tg of PVL could not be detected for any of these blends. But the unperturbed Tg of PVF2 indicates clearly that the PVL/PVF2 sample is immiscible.

PMPPL/PVF and PMPPL/PVF2 blends have also been prepared and their properties compared with those of the PMPPL/PVC blends that have been previously shown to be mis~ible.~ All measurements made on these two samples are summarized in Table IV.

The 30-70% PMPPL/PVF blend presents a single Tg at 278 K equal, within experimental error, to that of PMPPL at 275 K, indicating the immiscibility of the blend. In general, PMPPL/PVF2 blends give two Tgk, one at 238-239 K close to that of PVF2 at 232-235 K, and one at 275-277 K corresponding to the Tg of PMPPL. The melting points and the enthalpies of fusion of PVF2 in these blends are almost unperturbed as compared with the values observed for pure PVF2. The PMPPL/PVF2 sample is then clearly immiscible.

Similarly, in Table V, the 33.1-66.9% PMEPL/PVF2 blend shows two Tg’s, one at 236 K close to that of PVF2, and one at 266 K close to that of pure PMEPL. The 66.6-33.4% PMEPL/PVF2 blend has only one Tg at 263 K, the Tg of PVF2 being unobserved owing to the small amount of PVF2 in this blend. Two Tm’s and two AH’S are observed in both cases. Clearly the PMEPL/PVF2 sample is immiscible, unlike the PMEPLEVC sample, which has been previously shown to be mi~cible .~

TABLE IV Glass Transition Temperature, Melting Point, and Enthalpy of Fusion of PMPPL, PVF, PVF2,

and of PMPPLPVF and PMPPLPVF2 Mixtures

PVF or PVF2 weight content (%) Tg (W T m (K)

0 275 353 70% PVF 278 ... 30% PVF2 277 444 50% PVF2 238,275 447 70% PVF2 239,277 444

100% PVF 336 ... 100% PVF? 232-235 444

...

... 21 38 28

29 ...

TABLE V Glass Transition Temperature, Melting Point, and Enthalpy of Fusion of PMEPL, PVF2, and

Their Mixtures

0 264 406 29 33.4 265 399,451 28,29

(66.9)a 236,266 399,450 14,37 100 232,235 444 29

a All values reported for this blend were obtained for a film cast from a DMF solution.

1252 AUBIN AND PRUD’HOMME

Finally, PCLPVF and PCL/PVF2 blends have been prepared and their be- havior has been compared to that of the PCLPVC sample, which is known to be miscib1e.l All measurements are summarized in Table VI.

For a 50-50% PCLPVF blend, a single Tg is observed at 213-216 K, in agreement with that of PCL at 215 K. The Tg of PVF is not seen in this blend, probably because it is too close to the melting point of PCL at 330 K. Immisci- bility is again found. Similarly, PCLPVF2 blends exhibit a Tg at 232-236 K close to that of pure PVF2. Two T,’s and two AH’S are, however, observed for these blends. These results again lead to the conclusion that the PCLPVF2 sample is immiscible.

CONCLUSIONS

We have been able to show that PVL is fully miscible with PVC. This con- clusion is clearly indicated by the presence, a t each composition, of a single Tg intermediate between the values for pure PVL and pure PVC.

Since several polylactones are miscible with PVC, it can be expected that they will also be miscible with other poly(viny1 halides). Blends of PVL, PMPPL, PMEPL, and PCL have been prepared with PVF and PVF2, and in all cases it has been shown that these blends are immiscible. This result is surprising in view of the similarity between PVC and PVF. In fact the differences between the chlorine and the fluorine atoms are small. However, the fluorine atom is smaller than the chlorine atom. It has an ionic radius of 0.136 nm as compared to 0.181 nm for the chlorine atom. It is also more electronegative, having an electronegativity of 4.0 as compared to 3.0 for the chlorine atom on Pauling’s scale. These differences are not large enough to give significantly different Tg’s and Tm’s for PVC and PVF, but they are still large enough to lead to significantly different behavior in polylactone/poly (vinyl halide) blends.

It may be noted in passing that this difference in behavior seems to be taken into account in the tables of molar attraction constants of Hoy and Molar attraction values given for the fluorine and the chlorine atoms are very different, suggesting different miscibility behavior of compounds involving these two atoms. However, it must be kept in mind that the solubility parameter criterion derived from such tables does not explain the miscibility of polylac- tone/PVC blends, as was shown bef01-e.~

TABLE VI Glass Transition Temperature, Melting Point, and Enthalpy of Fusion of PCL, PVF, PVF2, and

of PCLPVF and PCLPVF2 Mixtures

PVF or PVFz weight content (%) T g W) Tm (K) (J g-’)

0 215 330 71 50% PVF 213-216a ... ...

216 322 63 (33.7% PVF2)’ 232 327,449 91,30

100% PVF 336 ... ... 100% PVFz 232-235 444 29

a All values reported for this blend were obtained for a film cast from DMF or N-methyl-2-pyr-

(65.1% PVFz)’ 236 334,452 7a,36

rolidone solutions.

POLYLACTONE/PVC BLENDS 1253

Smaller differences in electronegativity, in ionic radius, and in molar attraction between the bromine and the chlorine atoms than between the fluorine and the chlorine atoms, suggest the possibility of better miscibility between polylactone and poly(viny1 bromide), than between polylactone and PVF. This possibility has not yet been tested.

The authors thank Francois R. Prud‘homme for experimental assistance in the small-angle light-scattering work. They also thank the National Sciences and Engineering Research Council of Canada and the Ministry of Education of the Province of Quebec (FCAC program) for the research grants that supported this work.

References

1. J. V. Koleske, in Polymer Blends, D. R. Paul and S. Newman, Eds., Academic, New York 1978,

2. 0. Olabisi, L. M. Robeson, and M. T. Shaw, Polymer-Polymer Miscibility, Academic, New

3. M. M. Coleman and J. Zarian, J. Polym. Sci. Polym. Phys. Ed., 17,837 (1979). 4. 0. Olabisi, Macromolecules, 8,316 (1975). 5. M. Aubin and R. E. Prud’homme, Macromolecules, 13,365 (1980). 6. M. Aubin, and R. E. Prud’homme, Polymer, to appear. 7. N. Spassky, A. Leborgne, M. Reix, R. E. Prud’homme, E. Bigdeli, and R. W. Lenz, Macro-

8. D. Duchesne and R. E. Prud’homme, Polymer, 20,1199 (1979). 9. J. Brandrup and E. I. Immergut, Eds., Polymer Handbook, 2nd ed., Wiley, New York

Chap. 22.

York 1979.

molecules, 11,716 (1978).

1975. 10. T. G. Fox, Bull. Am. Phys. SOC., 1,123 (1956). 11. M. Gordon and J. S. Taylor, J. Appl. Chem., 2,495 (1952). 12. F. N. Kelley and F. Bueche, J. Polym. Sci., 50,549 (1961). 13. W. J. Macknight, F. E. Karaz, and J. R. Fried, in Polymer Blends, D. R. Paul and S. Newman,

14. M. Bank, J. Leffingwell, and C. Thies, J. Polym. Sci. Polym. Phys. Ed., 10,1097 (1972). 15. C. Elmqvist and S. Swanson, Eur. Polym. J., 12,559 (1976). 16. L. Mandelkern, S. Go, D. Peiffer, and R. S. Stein, J. Polym. Sci. Polym. Phys. Ed., 15,1189

17. T. Nishi and T. T. Wang, Macromolecules, 8,909 (1975). 18. T. K. Kwei, T. Nishi, and R. F. Roberts, Macromolecules, 7,667 (1974). 19. C. H. Ong and F. P. Price, J. Polym. Sci. Polym. Symp., 63,45 (1978). 20. F. B. Khambatta, F. Warner, T. Russell, and R. S. Stein, J. Polym. Sci. Polym. Phys. Ed., 14,

21. R. S. Stein and M. B. Rhodes, J. Appl. Phys., 31,1873 (1960). 22. M. B. Rhodes and R. S. Stein, J. Polym. Sci. Part A-2,7,1539 (1969). 23. S. Krause, in Polymer Blends, D. R. Paul and S. Newman, Eds., Academic, New York 1978,

Eds., Academic, New York 1978, Chap. 5.

(1977).

1391 (1976).

Chap. 2.

Received August 5,1980 Accepted March 2,1981