首页   按字顺浏览 期刊浏览 卷期浏览 Thermoreversible gelation in polymer systems. Part 2. Gel/sol transition in vinylidene ...
Thermoreversible gelation in polymer systems. Part 2. Gel/sol transition in vinylidene chloride/methyl acrylate copolymer gels

 

作者: M. A. Harrison,  

 

期刊: Faraday Discussions of the Chemical Society  (RSC Available online 1974)
卷期: Volume 57, issue 1  

页码: 38-46

 

ISSN:0301-7249

 

年代: 1974

 

DOI:10.1039/DC9745700038

 

出版商: RSC

 

数据来源: RSC

 

摘要:

Themoreversible Gelation in Polymer Systems Part 2. Gel/Sol Transition in Vinylidene Chloride/Methyl Acrylate Copolymer Gels BY M. A. HARRISON, P. H. MORGAN AND G. S . PARK* Chemistry Department, UWIST, King Edward VII Avenue, Cardiff, CFl 3NU Received 1 st November 1973 Those vinylidene chloride/methyl acrylate copolymers that show crystallinity in the solid state form reversible gels when dissolved in benzene, chlorobenzene, rn-dichlorobemene or o-dichloro- benzene. The temperature Tm of the gel/sol transition increases with the vinylidene chloride content of the copolymer and with decreasing dipole moment of the solvent. Polymer concentration and molecular weight have only a small effect on Tm and so the Ferry and Eldridge relationship, a In C/aT;l = AHJR, gives very large values (80-240 kJ mol-') for the junction point energy AH,.This increases with the vinylidene chloride content of the copolymer and with the dipole moment of the solvent. Equating the gel/sol transition with the crystalline melting point enables Flory's relationships between copolymer composition, solvent power and the melting point to be applied to Tm. This gives x values for poly(viny1idene chloride) increasing from 0.67 to 0.80 as the dipole moment of the solvent decreases and a heat of fusion of 2.64 kJ mol-' per vinylidene chloride unit. It is suggested that the network junction points of the copolymer are between 30 and 100 units long and this ac- counts for the absence of gelation in copolymers having less than 80 mol % of vinylidene chloride units.Many solutions of poly(viny1 chloride) form weak gels on standing or cooling and these revert to the sol state on heating. This reversible gelation has been attri- buted to network formation in which the network linking points consist of minute crystallites formed from the occasional syndiotactic runs that occur in the predom- inantly atactic polymer. Since there is some uncertainty about the syndiotactic content of these polymers it is of interest to investigate reversible gel formation in solutions of polymeric substances that contain known amounts of crystallizable and non-cry s t allizable units. The methylacrylate/vinylidene chloride copolymers are such substances. The reactivity ratios in the free radical polymerization are unity for both monomers and so random copolymers of constant composition can easily be made.This enables the stability of the resulting gels to be related to polymer composition as well as to other factors such as molecular weight and concentration. The gel melting point (the temperature at which the gel/sol transition occurs) gives a ready means of investigating gel stability and can be related to the energy changes involved in network formation. EXPERIMENTAL COPOLYMERS Vinylidene chloride/rnethylacrylate copolymers were prepared by free radical polymer- ization of well out-gassed solutions in cyclohexanone using azoisobutyronitrile as initiator at 40°C. After about a 10 % conversion the polymers were separated by precipitation in methanol and were purified by repeated solutions in tetrahydrofuran followed by precipit- 38M .A . HARRISON, P . H . MORGAN A N D G . S . PARK 39 ation in further quantities of methanol. The polymers were dried at 6040°C under vacuum and the copolymer compositions were found from C and H determinations. Molecular weights were measured in tetrahydrofuran solution using a Hewlett-Packard 502 automatic membrane osmometer. The characteristics of the copolymers are given in table 1. TABLE 1 .-VINYLIDENE CHLORIDE : METHYL ACRYLATE COPOLYMER CHARACTERISTICS mol % vinylidene chloride designation 13 18 17 19 11 22 20 21 24 23 26 25 27 28 16 29 10 14 nominal 83.3 83.3 83.3 83.3 91.2 91.2 91.2 91.2 93.8 93.8 93.8 93.8 95.2 95.2 95.2 95.2 95.5 87.5 actual b 81.1 83.1 85.2 83 .O 91.3 91.5 89.8 92.4 93.2 92.8 93.9 94.0 95.6 95.6 97.6 95.0 84.0 96.6 Id/dl g-l 0.423 0.561 0.575 0.835 0.543 0.602 0.71 3 0.877 0.489 0.584 0.663 0.675 0.540 0.603 0.683 0.748 - - G n x c 32 030 0.378 77 900 0.390 84 500 0.399 20 900 0.405 70 600 0.333 74 500 0.381 0.402 88 900 23 O00 0.377 58 700 0.369 79 400 0.349 79 600 0.408 84 700 0.383 70 100 0.3 16 71 700 0.41 5 74 700 0.408 82 300 0.404 36 560 - - - 15 88.9 88.3 - 49 280 0.354 12 90.0 88.3 - - - a Calculated from composition of monomer feed ; b calculated from C and H analysis data ; x values in tetrahydrofuran from the dependence of osmotic pressure on concentration. GEL MELTING POINTS Uniform gels of the copolymers were prepared in sealed tubes as described previously for poly(viny1 chloride) gels.' The temperature at which a pressure head caused flow through a fine capillary was used as a measure of the gel melting point, Tm.This was determined using a technique described elsewhere.2 In many of the gels, syneresis occurred on standing for a day or more and so the gel melting points were determined soon after the gels had been formed. In some experiments at high solvent concentrations, syneresis was so rapid that gel melting points could not be determined. No significant change in the observed value of Tm was found on changing the pressure head or altering the diameter of the capillary. RESULTS Benzene, chlorobenzene, m-dichlorobenzene and o-dichlorobenzene could all be used as solvents for the copolymer at high temperatures. Solutions of polymers containing less than 80 mol % of vinylidene chloride units remained mobile even on prolonged standing. Polymers containing about 83 % of vinylidene chloride units gave solutions that set to weak gels on prolonged standing for a few days.Polymers containing 90 % or more of vinylidene chloride gave gels on cooling to room temper- ature and standing for only a few minutes. Most of the gels were cloudy.40 VINYLIDENE CHLORIDE COPOLYMER GELS Fig. la gives the gel melting points T, for gels of various concentrations of co- polymer 11 having a nominal vinylidene chloride content of 91 rnol % and having a molecular weight of 70 600 in all four solvents, while fig. lb, l c and Id give data for solutions in o-dichlorobenzene for copolyniers having various compositions and molecular weights. (4 FIG. l.-Dependence of gel melting points Tm on concentration C.(a) Polymer 11 (nominal vinyl- idene chloride = 91.2 mol %, Gn = 70 600) in various solvents. 0, CsHs ; A, rn-C6H4Cl2 ; 0, C6H5Cl ; 0, o-C6H4CI2. (b) Polymers 11, 22, 20, 21 (nominal vinylidene chloride = 91.2 rnol %) in o-dichlorobenzene. Mn : 0, 70 600 ; 0, 74 500 A, 88 900 ; 0,123 OOO. - (c) Polymers 24, 23, 26, 25 (nominal vinylidene chloride = 93.8 rnol %) in o-dichlorobenzene. Mn : 0, 58 700 ; D, 79 400 ; A, 79 600 ; 0, 84 700. (d) Polymers 27, 28, 16, 29 (nominal vinylidene chloride = - 94.2 rnol %) in o-dichlorobenzene. Mn : 0, 70 100 ; 0, 71 700 ; A, 74 700 ; 0, 82 300. DISCUSSION THE FERRY AND ELDRIDGE RELATIONSHIP Ferry and Eldridge treated reversible gelation as an equilibrium between actual network junctions and potential network junction sites.At the gel melting point theM. A . HARRISON, P. H. MORGAN AND G . S. PARK 41 number of actual network junctions is the minimum required for a continuous net- work. This enabled the equilibrium constant for network formation to be related to both concentration and molecular weight. By application of the vant’ Hoff isochore to this equilibrium Ferry and Eldridge obtained the relationship Here, Tm is the gel melting point, M , is the molecular weight of the polymer, C is the concentration and, in the original theory, AH, and AH, were both equal to the heat of formation of a mole of network junctions from potential network junction points. An extention of this treatment to junctions involving n polymer chains retains the significance of the AHm but gives AH, = AH,,&- l), (3) FIG.2.-Ferry and Eldridge plots for the concentration dependence of Tm (a), (b), (c), (d) and the symbols have the same meaning as in fig. 1.42 VINYLIDENE CHLORIDE COPOLYMER GELS so that AH, is more closely related to the energy change on uniting a mole of potential network junction points to partially formed network junctions. It is interesting to see if eqn (1) and (2) can be used to give the junction point energies in vinylidene chloride/methyl acrylate copolymer gels. Since data for copolymers covering a range of molecular weights for copolymers with a sufficiently constant composition are not available (fig. l), it was not possible to obtain values of TABLE 2.-JUNCTION FORMATION ENERGIES, &.HX FOR COPOLYMER NO.11 (91 % NOMINAL VINYLIDENE CHLORIDE CONTENT M , = 70 600) IN VARIOUS SOLVENTS dipole moment/ TnZ/"C for 6 % AH=/ crystallite length* solvent debye solution k J mol- 1 C6H6 0 90.5 - 80 31 C~HSCI 1.70 74.8 134 52 m-C 6H4C12 1.72 74.8 163 63 O-C6H4C12 2.48 70.8 180 69 * no. of vinylidene chloride units AH, from eqn (1). The relative insensitivity of T, to changes in concentration in fig. 1 indicates that AH, in eqn (2) must be large. The logarithmic plots in fig. 2a enable the AH,,, values given in table 2 for various solvents to be obtained. As the dipole moment of the solvent is increased on passing from benzene to o-dichlorobenzene, the gel melting point decreases as might be expected for increasing solvent power but, surprisingly, the values of AH, increase.TABLE 3 .-JUNCTION FORMATION ENERGIES AHx FOR VARIOUS COPOLYMERS IN O-DICHLORO- BENZENE nominal vinylidene TmlOC polymer chloride content/ - for 5 % A&/ AH2/kJ mol-1 crystallite no. (mol O A Mn soln . kJ mol-1 mean value length* - 50 113 11 91.2 70 600 22 91.2 74 500 76.5 20 91.2 88 900 - 21 91.2 123 OOO - 109 130 24 93.8 58 700 - 163 23 93.8 79 400 - 172 66 26 93.8 79 600 87.2 25 93.8 84 700 - 243 27 95.2 70 100 - 28 95.2 71 700 - 243 93 16 95.2 74 700 101.3 29 95.2 82 300 - 117 ii4 * no. of vinylidene chloride units The dependence of AH, on polymer composition is given by the slopes of the plots for gels in o-dichlorobenzene in fig. lb, lc, and Id. Table 3 indicates a large amount of scatter in the data but average values for each composition show an increase in AH, with increasing vinylidene chloride content and this is accompanied by an increase in gel melting point.It is interesting to compare these results with those for poly(viny1 chloride) ge1s.l In poly(viny1 chloride) gels high T, values were obtained when more regular polymer containing larger crystallizable syndiotacticM. A . HARRISON, P . H . MORGAN A N D G . S. PARK 43 runs were used. In the present work this is paralleled by the increase in T, for larger runs of vinylidene chloride units in the copolymers but in these gels and in contrast with the poly(viny1 chloride) gels, higher AHx values accompany the higher T, values as would be expected if network junctions are small crystallites which are larger when the runs of vinylidene chloride units are larger.CRYSTALLINITY A N D GEL FORMATION The hypothesis that the copolymer gels are held together by minute crystallites is in accord with the increased difficulty of gel formation with decreasing vinylidene chloride content. X-ray examination showed no evidence of crystallinity in solid copolymers containing 80 % or less of vinylidene chloride, while crystalline reflec- tions were obtained with polymers containing 83 % vinylidene chloride units or more. This composition was also the critical one below which gel formation did not occur. Direct evidence of crystallite formation in actual gels is not as easy to obtain from X-ray examinations and so indications of crystallinity were sought from differential thermal analysis of a 10 % gel in o-dichlorobenzene of a copolymer containing 95 % of vinylidene chloride units.Fig. 3 shows the trace obtained using a DuPont model 900 instrument. The rapid endotherm at 102°C coincides with the melting temper- ature of this gel and gives good evidence that the disappearance of the gel structure on heating coincides with the end of a period of rapid crystallite dissolution. 2.c 1.5 Y iz' d 1.0 0.5 ----- b I I I I I I 8 0 120 160 temperature/'C FIG. 3.-D.T.A. plot for a 10 % gel of polymer 16 (95.2 nominal mol % vinylidene chloride Mn = 74 700 in o-dichlorobenzene: a, heating ; 6, cooling ; heating rate 20°C min-'. The zero point on the temperature difference AT scale is arbitrary. EFFECT OF METHYLACRYLATE UNITS O N THE GEL MELTING POINT If gel formation is due to a crystallite-linked network the gel melting point should be related to the crystalline melting point of the vinylidene chloride crystallites in the gel.In copolymer/solvent mixtures both copolymer composition and solvent concentration affect the melting point. These factors have been treated separately by Flory to give the relationship44 VINYLIDENE CHLORIDE COPOLYMER GELS Here TE is the melting point of pure vinylidene chloride, and T, is the melting point observed for a copolymer containing mol fraction x of non-crystallizable comonmer with volume fraction 4 of solvent. AH, is the enthalpy change on converting pure perfectly crystalline polyvinylidene chloride to unit volume of melt ; V, and V are the molar volumes of a crystallizable unit in the amorphous polymer and of the pure solvent, while x is the Flory-Huggins interaction parameter for the poly(viny1idene chloride) +solvent system.0.8 n I R' / 0 . 6 - e b4 2 W 0.4'- _1 - c - I I I 1 I j I I rnol fraction methyl acrylate, x FIG. 4.-Gel melting point Tm variation with methyl acrylate content of copolymer. Polymer concentrations : 0, 3 % ; 0, 6 % ; 0, 9 %. T," = 463 K. Actual x values for C and H deter- minations on copolymers 22, 26, 28, 16. Identifying T, with T,, eqn (4) predicts a linear relationship between l/Tm and x. The data for four of the copolymers at three concentrations in o-dichlorobenzene are plotted in fig. 4. There is some indication of curvature but taking the best straight line through the points the values for the energy of melting per mole of crystallizable units, Vv(AH,), shown in table 4 are obtained.The mean figure of about 2.64 kJ TABLE 4.-HEAT OF MELTING V"(A&) FROM THE EFFECT OF COPOLYMER COMPOSITION ON THE GEL MELTING POINTS concentration/(g l.-l) 30 60 90 mean V,(AH,)/(kJ mol-I) 2.52 2.57 2.83 2.64 mo1-1 for the heat of fusion is small. Values of about 10 W mol-l are typical for most crystalline polymers. Low values are, however, commonly obtained from the application of this sort of theory to the melting point of copolymers and a similar treatment for poly(viny1 chloride) gels led to heats of fusion that were only a quarter of the established value for large poly(viny1 chloride) crystallites. This is, at least partly, due to the large surface energy that results from the small crystallite size in the gel but errors may also arise from the persistence of crystallinity at temperatures above T,.This has been well established in the poly(viny1 chloride) system and may also occur with the present copolymers.M. A . HARRISON, P . H. MORGAN AND G. S . PARK 45 EFFECT OF SOLVENT PROPERTIES ON THE GEL MELTING POINT Eqn (4) enables the gel melting point to be correlated with solvent properties. At considerable dilution, as the volume fraction 4 approaches unity, eqn (4) approxi- mates to Values of (1 -x)/V can be obtained from this relationship by combining the AH, values with the data of fig. l(a) after extrapolating to infinite dilution. The results together with the derived x values, are shown in table 5. As expected, x increases on TABLE 5.-x VALUES FOR CO~OLYMER 11 (91 % NOMINAL VINYLIDENE CHLORIDE CONTENT M , = 70 600) IN VARIOUS SOLVENTS molar volume/ V Tm 1°C at (i-X)/vx 1031 solvent cms zero conc.cm-3 X C6H6 89 77 2.30 0.80 C6HjCl 102 67 2.76 0.72 m-C 6H4C12 115 69 2.66 0.69 0-C6H4C12 113 64 2.90 0.67 Molar volume of vinylidene chloride units V, = 58.4 cm3 mol-' ; melting point of pure poly- (vinylidene chloride), T," = 463 K. going from nz-dichlorobenzene to benzene. All of the values are greater than 0.5 which indicates that these liquids are non solvents. If these x values are correct then presumably the solution that occurs is due to the low molecular weight of the poly- (vinylidene chloride) sequences. A similar treatment for poly(viny1 chloride) gels in which l/Tm is plotted against (1 -x)/V led to T," values that were too small.This may have arisen in part because the copolymeric character of the poly(viny1 chloride) was ignored, i.e., a zero value of x was assumed; but it is also likely that the minute size of the fringed micellar crystallites linking the network structure would also lead to a low T," value. This same effect may be occurring in the vinylidene chloride copolymer gels and so the true values for poly(viny1idene chloride) could be less than those in table 5. SIZE OF NETWORK JUNCTION POINTS The AH, values from the Ferry and Eldridge theory give a measure of the size of the network junction crystallites. The number of continuous vinylidene chloride units along a polymer chain that are needed to form a network junction crystallite can be calculated from the ratio of AHx to the heat of fusion per mole of vinylidene chloride units. If the figure of 2.64 kJ mol-1 for V,(AH,) is taken from the treatment of the effect of copolymer composition on T,, crystallite lengths of between 30 and 100 vinylidene chloride units are obtained. The actual values are included in tables 2 and 3. As would be expected, the size increases with the vinylidene chloride content of the copolymer. The large number of units needed to form a crystallite presumably accounts for the absence of crystallinity or gel formation when the vinylidene chloride content is 80 mol % or less. This contrasts with the findings for poly(viny1 chloride) gels in which runs of only about 10 crystallizable units are needed to form junction points.46 VINYLIDENE CHLORIDE COPOLYMER GELS This paper reports work carried out with the support of the Procurement Execu- tive, Ministry of Defence. M. A. Harrison, P. H. Morgan and G. S. Park, European Polymer J., 1972,8, 1361. M. A. Harrison, P. H. Morgan and G. S. Park, Brit. PoZymer J., 1971, 3, 154. J. E. Eldridge and J. D. Ferry, J. phys. Chem., 1954,58,992. P. J. Flory, J. Chern. Phys., 1949, 17,223.

 

点击下载:  PDF (558KB)



返 回