Effect of Substituents in the Styrene Ring on the Dynamic Mechanical Relaxations of a Styrene-

Crosslinked Unsaturated Polyester Resin

WILLIAM DOUGLAS and GEOFFREY PRITCHARD, School of Chemical and Physical Sciences, Kingston Polytechnic, Penrhyn Road, Kingston-

upon-Thames, Surrey, KT1 2EE, United Kingdom

Synopsis

The dynamic mechanical properties of a series of polyester resins made from a maleic/phthalic anhydride-based unsaturated polyester crosslinked with each of styrene, 4-methyl styrene, 4-ethyl styrene, 4-n-butyl styrene, 4-isopropyl styrene, tertiary butyl styrene, 4-chlorostyrene, and 3,4- dichlorostyrene were studied. The order of the a transition temperatures was as expected from that for the homopolymers, except in the case of the chlorostyrenes, for which dipolar interactions with the polyester chain may be important. The styrene bridges appeared to be involved in a steric interaction (and in the case of the chlorostyrenes, a dipolar interaction) with the p relaxing ester species. It is suggested that both the y and y’ relaxations involve similar interactions between the matrix and the relaxing moieties. For the 4-n-butyl styrene resin, an additional relaxation below -170°C was observed, and is ascribed to relaxation of the n-butyl group.

INTRODUCTION

Four dynamic mechanical relaxations are observed for styrene-crosslinked unsaturated polyester resins up to and including the glass transition temperature (a relaxation).’ They are labeled a, p, y, and y’ in descending order of transition temperature. The a relaxation involves movement of substantial parts of the whole network.

A model2 involving water bridges between ester groups on neighboring chains has been proposed to explain the water-induced y relaxation which is also ob- served in the linear polyester. The y’ relaxation is thought to originate in the diol segments of the crosslinked polye~ter.~ Both the y and y’ relaxations are influenced to some extent by the styrene crosslinks.

There is doubt about the role of the styrene crosslinks in the p relaxation. Some workers have ascribed the p relaxation to motion of the styrene-based bridging Others deduce that the /3 relaxation arises from motion of the saturated dioic acid-diol grouping in the polyester segments, with the bulky styrene units restricting motion of some potential relaxation sites6 The unre- solved problem as to the extent of involvement of the styrene crosslinks in the

relaxation has previously been studied by variation in the number and length of styrene crosslinks. It occurred to us that an investigation into the effects of substituents in the styrene ring should show whether or not the styrene-based bridging units are in fact the main p relaxation species. It should also help to establish the extent of involvement of the styrene crosslinks in the y and y’ re- laxations. In order to do this, development of an unsaturated polyester com- patible with a wide range of ring-substituted styrenes was necessary.

Journal of Polymer Science: Polymer Physics Edition, Vol. 20, 1223-1232 (1982) 0 1982 John Wiley & Sons, Inc. CCC 0098-1273/82/041223-10$02.00

1224 DOUGLAS AND PRITCHARD

EXPERIMENTAL

The Unsaturated Polyester

The unsaturated polyester was synthesized, under nitrogen, by standard techniques from maleic anhydride, phthalic anhydride, adipic acid, and 2,2- dimethyl-1,3-propanediol in the molar ratio 1.00:0.700.30:2.08. The acid value was 26, and the lH NMR spectrum indicated that 74% of the maleic groups had isomerized to the trans form. It showed compatibility with a wide range of substituted styrenes.

The Crosslinking Monomers

The crosslinking monomers and resins studied are listed in Table I, together with their codes. The styrene, 4-methyl styrene, tertiary butyl styrene, 4- chlorostyrene, and 3,4-dichlorostyrene were commercial samples and used as such. The remaining monomers were synthesized by the following modification of the described procedure7: 6 M e C C ; - G NaBH, -6 KH,SO,, 6

COMe CHOHMe CH=CH, Scheme I

The IR spectra indicated the substituents to be in the para position. The ‘H NMR and mass spectra showed that the monomers were of high purity.

The Resins

The polyester was dissolved in each of the crosslinking monomers to give SO-

lutions of molar ratio of monomer to fumarate plus maleate of 1.75/1. In the

(“4 +10 +50 +90 +130

Fig. 1. Dynamic mechanical spectra of CS resin in region of a and @ relaxations [(o) loss modulus G ” (MN m-2), (A) shear modulus G’ (GN m-2), (0 ) logarithmic decrement A].

DYNAMIC MECHANICAL RELAXATIONS 1225

r- e m 0 a w a m . ( D o o r - w m 0 0

‘ 0 0 0 3 0 0 V

‘ 9 9 9 - i 4 9 zz

1226 DOUGLAS AND PRITCHARD

TABLE I1 Temperatures of a and p Transitions ("C, sh Denotes Shoulder)

T , of substituted Resin T,(A) TJG") To(& To(G") styrene homopolymerg

LS 59 58(sh) 58(sh) S 88 45 MS 86 80(sh) 53(sh) 45 93

ES 80 34

33 }loo ... ... 27 ... ...

nBS 50 ... ... 10 ~1 48 ... ... iPS 78 34

tBS 106 90(sh) 58(sh) 45 130 (para isomer)

110 cs 83 80(sh) 60(sh) 42

diCS 63 30 128 ... ...

case of styrene an additional solution was made with a molar ratio of 0.9911 (resin code LS). These resins each contained 100 ppm tertiary butyl catechol as in- hibitor.

Except in the case of the ES resin, the resins were cured by using 2% wlw of a 50% solution of methyl ethyl ketone peroxide in a mixture of phthalate plas- ticizers (active oxygen ca. 9%)) and 0.054% wlw of 9% cobalt alsynate solution in white spirit, to give cast sheets. The gel time was 30-60 min at RT. The initially unreactive ES resin was cured with an additional 50% of peroxide and 100% of accelerator. Immediately after gelation, the cast sheets were cured for 21 h at 40°C followed by 3 h at 80OC.

The residual monomer concentrations (Table I) in each of the resins at various stages of cure were determined by dichloromethane extraction followed by GLC.

Dynamic Mechanical Properties

The variations of dynamic mechanical properties of strips of each resin (ca. 9 X 3 X 120 mm) with temperature were measured with a torsional pendulum at two or three resonant frequencies obtained by altering the moment of inertia. Two temperature ranges were examined, from RT up to +120"C and from RT

TABLE 111 Temperatures of y and y' Transitions (Dry Samples) ("C, sh Denotes Shoulder)

Resin T , ( 4 TJG") T,4& T,,(C")

LS -83(sh) -82(sh) -133 -137 S -90 -94 -135 -137 MS -80(~h) -81(sh) -134 -137 ES -92 -96 -137 -143 nBS -90(sh) -90(sh) -144 -154 iPS - 100 - 100 -138 -143 tBS -100 -100 -140 -140 cs - 90 - 90 -135 -137 diCS -76(sh) -76(sh) -131 -137

DYNAMIC MECHANICAL RELAXATIONS 1227

-160 -120 -80 -40 0 ("4 Fig. 2. Dynamic mechanical spectra of dry nBS resin in region of y and y' relaxations. Symbols

as in Fig. 1.

down to -170°C. In the high-temperature range, the sample was heated in air in an oven, and the properties were calculated at the reference frequency of 0.4 Hz by linear extrapolation of a semilogarithmic plot of resonant frequency against the relevant property. The results are to be found in Table 11, and a typical spectrum (CS resin) is given in Figure 1. For the low-temperature range, the sample was cooled in dry nitrogen gas and the properties were calculated at the reference frequency of 0.7 Hz (Table I11 and Fig. 2). The samples had previously been dried to constant weight over silica gel. Additionally, in the case of the low-temperature range, the dynamic mechanical measurements were repeated on samples which had attained constant weight in air saturated with water vapor (Table IV and Figs. 3,4) . The weights of water absorbed are given in Table IV.

RESULTS AND DISCUSSION All the resins (see Table I for codes) were made from the same base polyester

dissolved in equimolar quantities (except for the LS resin) of crosslinking monomer. The IR spectra of the cured resins are very similar. In particular,

TABLE IV Temperatures of y and y' Transitions (Wet Samples) ("C) and Water Absorption

T,4& T,*(G") Water Resin T J A ) T,(G") (shoulders) (shoulders) (% w/w)

LS -88 -92 -130 -134 0.636 S -102 -105 -148 - 148 0.958 MS -94 -97 -136 -139 0.505 ES -107 -108 -148 -151 0.910 nBS -95 -99 -138 -141 0.781 iPS -105 -107 -150 -152 0.899 tBS -101 -104 ... -140 0.721 cs -92 -93 -143 -144 0.702 diCS -89 -92 -132 -134 0.320

1228

50

40

30

20

-

-

-

-

DOUGLAS AND PRITCHARD

t I - - G’ A

-

-

-

I 2.3 .08

1.9 .06

1.5 .04

-1.1 .02

-

-160 -120 -80 -40 0 Pc) Fig. 3. Dynamic mechanical spectra of wet tBS resin in region of y and y’ relaxations. Symbols

as in Fig. 1.

1.9 .06

-1.5 .04

-1.1 .02

there is no appreciable difference in the relative intensity of the residual fumarate absorption at 771 cm-l. Therefore a comparable proportion of fumarate groups has reacted in each resin, and thus the average bridge lengths must be similar for all the resins except LS.

The concentrations of monomer extracted from each resin at various stages of cure are given in Table I. The cured resins contained some unreacted monomer, which was still present after prolonged heating. Since the concen- trations are low, i t seems unlikely that the dynamic relaxations will have been appreciably affected. However, it has been suggested8 that residual styrene may cause the y relaxation near 120 K which is sometimes observed for polystyrene. Residual monomer may be partly responsible for the additional low-temperature relaxation observed for the nBS resin (see below).

50 t 42.3 .08

-160 -120 -80 -40 0 (“4 Fig. 4. Dynamic mechanical spectra of wet nBS resin in region of y and y’ relaxations. Symbols

as in Fig. 1.

DYNAMIC MECHANICAL RELAXATIONS 1229

The a Relaxation

The variations with temperature of the loss modulus G ”, the shear modulus G‘, and the logarithmic decrement A in the region of the a and 0 relaxations for the CS resin are shown in Figure 1. All are at the reference frequency of 0.4 Hz. Similar figures, not shown here, were obtained for the other resins. The a and 0 relaxation temperatures obtained from these figures are given in Table 11. It will be seen that the a relaxation is shown clearly by the maximum in A, whereas the less intense 0 relaxation is shown by the maximum in G”. In some cases shoulders are observed which give TJG”) or Tp(A).

The a relaxation temperatures of the resins from the alkyl-substituted styrenes decrease in the order

tBS > S - MS > ES N iPS > nBS

This is the same order as that observed for the alkyl-substituted styrene ho- mopolymers (Table II).9 A rigid substituent raises the T , and a flexible one lowers it. The bulky tertiary butyl group restricts molecular rotation by inter- action with other chain segments and the T, is thus raised above that of poly- styrene. However, linear substituents reduce interaction and lower the T,, as is seen in the cases of poly(4-n-butyl styrene) and poly(4-ethyl styrene). In general, the longer the chain, the greater the reduction in T,. The effect of methyl substitution is slight. Similar effects for polyolefinslO and poly(n -alkyl methacry1ates)ll have been observed.

In the case of the substituted-styrene-crosslinked resins, the differences in T , are smaller than those for the homopolymers. This reflects the fact that only 30-40% of the resin is substituted styrene. The a relaxation temperature of the iPS resin indicates that the isopropyl substituent, unlike the tertiary butyl group, is not bulky enough to restrict molecular movement. Rather, it lowers interac- tions and behaves much as an ethyl substituent.

Reduction in styrene content (cf. S and LS resins) results in a lowering of the a relaxation temperature. This effect is generally observed for unsaturated polyester resins.12

In the case of the resins made from the chloro-substituted styrenes, the a re- laxation temperatures decrease in the order

S > CS > diCS

This is the reverse order from that shown by the homopolymers (Table 11), and is thus a t first sight surprising. However, since chlorine and methyl sub- stituents are of similar size, dipolar interactions have been assumed13 to explain the increase in T, of the chlorostyrene homopolymers. In the case of the CS and diCS resins, these dipolar interactions presumably do not occur, perhaps as a result of the shortness of the chlorostyrene bridges. Indeed, there may be some dipolar interactions with neighboring ester groups, facilitating relaxation and thus giving rise to a lower T, than that for the S resin. The dielectric relaxational spectra may help to clarify this.

1230 DOUGLAS AND PRITCHARD

The fl relaxation

The data are to be found in the same table and figure as for the a relaxation. The p relaxation temperatures decrease in the order:

S - MS - tBS > CS > ES - iPS - LS > diCS > nBS.

This order lends support to the suggestion6 that the bulky benzene rings at- tached to the styrene bridges are involved in a steric interaction with the 0 re- laxing species. It is not consistent with the alternative h y p o t h e ~ i s ~ . ~ that the p relaxation is the result of motion of the styrene-based bridging units. It has been suggested6 that when the average length of the styrene bridges exceeds a critical value, any additional styrene units in the bridge are too remote to interfere with the relaxing polyester segments, so that further increase in the styrene concentration has little effect on Tp Thus in this case the S resin bridge length exceeds the critical value and Tp is at a maximum [Tp(G”) = 45”CI. The bulky tertiary butyl group does not increase the steric interaction with the /3 relaxing ester species since this interaction is already at a maximum. Thus the same maximum Tp(G”) of 45OC is observed for the tBS resin. However, linear sub- stituents (ethyl, n-butyl) and substituents of low bulk (isopropyl) lead to a lowering of the steric interaction with the p relaxing species and thus a lowering of Tp. The reduction in To observed for the chlorostyrene resins is perhaps a result of dipolar interactions with the p relaxing species. The dielectric relax- ation spectra may help to clarify this. As already discussed, the chloro sub- stituent is similar in size to a methyl group. Reduction in styrene content (cf. S and LS resins) leads to lowering of Tp. This is the result of the styrene bridge length falling below the critical value.

The y Relaxation

The variations with temperature of the loss modulus G”, the shear modulus G’, and the logarithmic decrement A in the region of the y and y’ relaxations for the dry nBS resin are shown in Figure 2; values of T , and T,) for all the dry resins are given in Table 111. The analogous data for the wet samples are to be found in Table IV, and spectra for the wet tBS and nBS resins in Figures 3 and 4, respectively. The percentage weight of water absorbed by each resin is given in Table IV.

For the dry samples, the temperature T , of the y relaxation decreases in the order

diCS > MS > LS > CS - nBS > S > ES > tBS N iPS

The corresponding order for the wet samples is

LS > diCS > CS > MS > nBS > tBS > S > iPS > ES

The water absorption decreases in the order

S > ES > iPS > nBS > tBS > CS > LS > MS > diCS

An increase in water content causes a reduction in T,, and an increase in in- tensity of the r e l a ~ a t i o n . ~ , ~ ~ The variation in water absorption may therefore be partly or wholly responsible for the fact that the order of the T , peak is dif- ferent for the dry and the wet samples.

DYNAMIC MECHANICAL RELAXATIONS 1231

An increase in styrene content causes a decrease in T,.2J4 This effect is ob- served for the S and LS resins.

Apart from the above observations it is difficult to rationalize the observed order of T,. It seems that there is some slight tendency towards a reduction in T , when the styrene ring bears an alkyl substituent, but this is by no means clear.

A water-bridging model has been proposed to explain the water-induced re- laxation y, which is also observed in the linear polyester.2 The relaxing species are of the type

‘C’ I1 0

0 /“’ ‘H.

0 Scheme I1

The following explanations for the observed dependence of the y relaxation on styrene concentration have been suggested2: (i) The presence of crosslinks in the polymer modifies the relaxation behavior of those species near a crosslink. (ii) Crosslinks induce an entirely different relaxation motion which is superim- posed on the relaxation modes of linear polyesters. (iii) Crosslinks affect the distance between adjacent polyester chains and thus the extent of water bridging.

The results of this work are consistent with the third explanation, since only in that case would change of the ring substituent of styrene lead to no systematic change in T,. The observed differences in T , for the various resins would therefore be the result of changes in the resin matrix.

The y‘ Relaxation

The data are to be found in the same figures and table as those for the y re- laxation. For the dry samples, the temperature T,) of the y’ relaxation, decreases in the orders

for T,t(A): diCS > LS > MS > CS - S > ES > IPS > tBS > nBS

for T,J(G’’): diCS - CS - MS - S - LS > tBS > IPS - ES > nBS

In the case of the wet samples, the corresponding orders are

for T,!(A): LS > diCS > MS > nBS > CS > S - ES > iPS

for T,r(G”): LS - diCS > MS - tBS - nBS > CS > S > ES > iPS

It will be seen that these orders are very similar to those for T , except in the case of the dry nBS resin, for which an abnormally low T,) is observed. Also, for the wet nBS resin (Fig. 4), the start of an additional relaxation around -170OC

1232 DOUGLAS AND PRITCHARD

is observed, which presumably involves the n-butyl group (see below). It may be that in the dry nBS resin (Fig. 2), this additional low-temperature relaxation is superimposed on the y‘ relaxation, thus leading to an abnormally low T,.. Such a superimposition of several relaxations to give a broad y’ peak has been postulated for some polyester resin^.^ The otherwise similar orders of T, and T,! are consistent with the hypothesis that the y and y’ relaxations are both influenced in similar ways by the resin matrix. The nature of the styrene sub- stituent has little effect on the intensity of the y’ relaxation.

It has been suggested that the y’ relaxation originates in movement of the diol units of the polye~ter .~ Absorption of water was found to cause a reduction in intensity of the y’ relaxation, as is observed in this work. The dependence on styrene concentration of both T,, and the intensity of the y’ relaxation was found to be complicated, and the mechanism of the interaction between the relaxing species and the surrounding matrix was not clear. It seems from the results of this work that for both the y’ and y relaxations similar interactions between the matrix and the relaxing species are involved.

nBS Resin Relaxation below -170°C An incipient relaxation in the region of -170°C is clearly shown (Fig. 4) in the

case of the wet nBS resin. This is not observed for the other resins. It seems likely that this additional relaxation is the result of movement of the n-butyl chain. For both poly(n-propyl methacrylate) and poly(n-butyl methacrylate), a relaxation at -190°C (the y’ loss peak) occurs, whereas the corresponding re- laxation for poly(ethy1 methacrylate) is at -232°C and that for poly(methy1 methacrylate) probably below -269”C.15 For these poly(n-alkyl methacrylates), the y’ relaxation has been ascribed to movement of the n-alkyl chain.

In this case, the relaxation observed may be partly the result of the relatively high concentration of n-butyl styrene monomer remaining in the resin (Table I).

We thank the Dow Chemical Company for providing a sample of tertiary butyl styrene.

References 1. W. E. Douglas and G. Pritchard, Deuelopments in Reinforced Plastics-1, G. Pritchard, Ed.,

2. W. D. Cook and 0. Delatycki, J. Polym. Sci. Polym. Phys. Ed., 15,1967 (1977). 3. W. D. Cook and 0. Delatycki, J. Polym. Sci. Polym. Phys. Ed., 15,1953 (1977). 4. R. S. Lenk and J. C. Padget, Eur. Polym. J., 11,327 (1975). 5. J. R. Dombroski, 31st Ann. Tech. Conf. S.P.I. Reinf. Plas. Comp. Inst., Washington D.C., 1976,

6. W. D. Cook and 0. Delatycki, Eur. Polym. J., 14,369 (1978). 7. C. G. Overberger, C. Frazier, J. Mandelman, and H. F. Smith, J . Am. Chem. SOC., 75,3326

8. R. J. Morgan and L. E. Nielsen, J. Polym. Sci. A-2 , 10, 1575 (1972). 9. J. Brandrup and E. H. Immergut, Polymer Handbook, Interscience, New York, 1966.

Applied Science, London, 1980, Chap. 8.

Paper 14-B.

(1953).

10. M. L. Dawnis, J. Appl . Polym. Sci., 1, 121 (1959). 11. S. S. Rogers and L. Mandelkern, J. Phys. Chem., 61,985 (1957). 12. W. D. Cook and 0. Delatycki, J . Polym. Sci. Polym. Phys. Ed., 12,2111 (1974). 13. W. G. Barb, J . Polym. Sci., 37,515 (1959). 14. W. D. Cook and 0. Delatycki, J . Polym. Sci. Polym. Phys. Ed., 13,1049 (1975). 15. Introduction to Polymer Science and Technology, H. S. Kaufmann and J. J. Falcetta, Eds.,

Wiley, New York, 1977, p. 279.

Received March 24,1981 Accepted February 16,1982