POLYMERIZATION AND POLYCONDENSATION PROCESSES

3 Polymer Synthesis at High Pressures N. L. ZUTTY and R. D. BURKHART

Downloaded by UNIV OF CALIFORNIA SANTA BARBARA on July 1, 2016 | http://pubs.acs.org Publication Date: January 1, 1962 | doi: 10.1021/ba-1962-0034.ch003

Research and Development Department, Union Carbide Chemicals Co., South Charleston, W. Va.

Experiments at high pressures are useful for gaining insight into polymerization mechanisms, and for preparing polymers that cannot be easily made in other manners.

High pressure equipment can

be obtained commercially for carrying out almost any type of addition polymerization and kinetic experiments under pressure can be designed to determine the pressure effect on the various steps of a free radical polymerization.

The effect of

pressure on comonomer reactivity ratios has revealed that radical selectivity is decreased with increasing pressure, resulting in more ideal copolymerizations at higher pressures.

Finally, the

high pressure copolymerization of ethylene with various comonomers allows the use of ethylene rather than styrene as a more logical base point for the Q-e scheme.

p o l y m e r syntheses at high pressures present some unique problems not usually met under more normal polymerization conditions. These involve both the use of specialized h i g h pressure equipment and pressure-induced changes i n the physical and chemical properties of the substances being studied. H i g h pressure equipment is generally cumbersome to handle, a n d only i n rare instances is it possible to make meaningful visual observations on a reacting system. It is almost always better to rely on indirect measurements. It is also unfortunate that P - V - T relations are known for only the most common monomers at high pressures, and i n most cases it is necessary to estimate densities. D u e largely to the work of Bridgman (4) some excellent F - V - T data are available for most common solvents. In carrying out high pressure polymerizations one also should be constantly on the guard for both pressure-induced phase and solubility changes. A n ex treme example of the latter effect is found in the report that gases, widely different in polarity, may become partially immiscible at high pressures, even though they are above their critical temperature (14). This phenomenon is of course con trary to what is taught i n the textbooks, where it is said that true gases are always completely miscible under any conditions. Therefore, i n dealing w i t h h i g h pres-

52 PLATZER; POLYMERIZATION AND POLYCONDENSATION PROCESSES Advances in Chemistry; American Chemical Society: Washington, DC, 1962.


ZUTTY AND

BURKHART

Polymer Synthesis at High Pressure

53

sures it may be necessary to reorient thinking as to gas behavior. I n the authors' experience, it is best to consider gases at high pressure as more nearly l i q u i d i n character. F o r example, ethylene above its critical temperature at 100° C . and 23,000 p.s.i. has a density of 0.526 gram per c c , while its critical density is only 0.22 gram per cc. Is it proper to consider such a substance a l i q u i d or a gas? T h e pressure-induced freezing of solvents may also lead to anomalous results. In this connection it is noteworthy that the freezing point of benzene is raised to about room temperature at 1000 atm. pressure (4). Although the difficulties involved i n carrying out high pressure polymeriza tions seem great, the possible yield i n terms of new knowledge of polymerization mechanisms, a n d as an alternative route to new polymers, is commensurate. W a l l i n g (15) has pointed out the advantages of using high pressure as a means of studying free radical polymerizations and some further studies along these lines are discussed below. Parallel to these mechanism studies the opportunity arises for using high pressures to polymerize monomers w h i c h are unreactive to free radi cals at atmospheric pressure. The use of h i g h pressure also expands the number of monomer pairs w h i c h can be copolymerized. F i n a l l y , high pressures can also affect degrees of polymerization, chain transfer, and chain branching. Thus, i n many cases, the polymerization pressure may have a profound effect on polymer structure and, hence, polymer properties. High Pressure Equipment for Polymerization Work T h e choice of a reactor for a high pressure polymerization is governed by the pressure required and the quantity and type of polymer desired. F o r reactions where efficient mixing is desired, both stirred autoclaves a n d rocking bombs are available commercially w h i c h are capable of maintaining at least 2000- to 3000atm. pressure. Less efficient stirring is obtained i n continuous tubular reactors, but the pressure limitations here are equal only to the pressure limitation of the tubing used and the pumps available. It is not unusual to find systems such as these capable of operating at 5000 atm. Static reactors providing no mixing may operate at m u c h higher pressures (10,000 to 50,000 atm.), but the higher pres sures inevitably result i n a large decrease i n reactor volume. Pumps, gages, fittings, and valves for operation at these pressures are avail able from a number of manufacturers. Diaphragm pumps capable of pres surizing a fluid without contamination by p u m p o i l have simplified the job of working w i t h chemically pure systems. Pressure Effects in Free Radical Polymerization Most quantitative work on high pressure polymerizations has centered on reactions proceeding by a free radical mechanism a n d , among free radical poly merizations, ethylene has received the greatest attention, because high molecular weight polyethylene can be produced only by high pressure techniques when nor mal free radical catalysis is used. F o r a free radical reaction, where électrostriction forces may be neglected ( 9 ) , the effect of pressure on the magnitude of the specific rate constant i n any step of the reaction is given b y Equation 1 (7) ÔJogJ: -AF* =

àP

RT

K

}

where k is the specific rate constant; A V * is the difference i n volume between activated species and reactants i n cubic centimeters per mole; Ρ is the pressure; PLATZER; POLYMERIZATION AND POLYCONDENSATION PROCESSES Advances in Chemistry; American Chemical Society: Washington, DC, 1962.

54

ADVANCES IN CHEMISTRY SERIES

and R a n d Γ have the usual meaning. M o s t experimental work has shown that E q u a t i o n 1 is reliable only i f the compressibilities of the activated species a n d the reactants are the same. I n those cases where compressibilities are unequal, it is customary to take the value of A V * extrapolated to atmospheric pressure. A s w o u l d be predicted from E q u a t i o n 1, the rate of dissociation of free radical initiators is decreased b y the application of pressure. Thus azobisisobutyronitrile dissociates w i t h a rate constant equal to 4.47 X 10~ sec." at 1500 atm. but at 1 atm. the dissociation rate constant is 5.5 X 10~ seer (8). Studies con cerning the effect of pressure o n the decomposition of benzoyl peroxide reveal that the rate of this reaction also decreases w i t h increasing pressure ( I I , 18). T h e extent to w h i c h the radical-induced decomposition of this peroxide at h i g h pressures affects the rate is not clear, but it appears that some complications arise from this cause. F o r practical purposes it is usually necessary to use a free radical initiator about 10° higher for each 1000-atm. pressure, i n order to achieve the catalytic activity normally expected. O n the basis of experience w i t h a wide variety of initiators, the temperatures at w h i c h some of the more common initiators m a y be used at 1000 a n d 3000 atm. have been summarized i n Table I. 5

1


5

1

Table I.

Most Efficient Use Temperature Range for Some Common Initiators Temperature, °C. Initiator lOOO atm. 3000 atm. Diisopropyl peroxydicarbonate 30-50 50-80 Diacetyl peroxide 40-60 60-70 Azobisisobutyronitrile 65-75 75-90 Dibenzoyl peroxide 70-90 110-130 Potassium persulfate 70-90 120-130 fcr/-Butyl hydroperoxide 120-140 160-180 T h e effect of pressure on the rate of free radical propagation reactions has been studied i n homopolymerizations only for styrene. Since A V * is negative for this reaction, the application of pressure increases the propagation rate. Nicholson and Norrish (12) list the propagation rate constant for styrene as 72.5 liters-mole —sec." at atmospheric pressure a n d 3 0 ° C . This increases to 206 liters-mole —sec. at 2000 atm. a n d 4 0 0 liters-mole- —sec." at 3000 atm. F r o m these data it is possible to calculate the value of A V * for the propagation step to be —13.3 cc. per mole. W a l l i n g and Pellon (16) report a value of —11.5 cc. per mole for the same reaction measured b y a different technique. B y studying the effect of pressure o n copolymerizations it is possible to ob tain and activation volume difference for various homo- a n d cross-propagation reactions. F o r instance, i n t w o of the propagation reactions occurring i n the styrene-acrylonitrile polymerization : -1

1

-1

-1

1

1

~vCH —GH- -f C H = C H

~vCH —CH-

φ φ ~vCH —CH- + C H = C H

φ —CH —CH—CH —CH-

2

2

I

I

2

2

I

I

(2)

2

I

2

(3)

2

I

I

φ CN φ CN T h e ratio k /k is found to vary w i t h pressure, a n d A V * — A V * has a value of —9.8 cc. per mole ( 5 ) . U s i n g the value of A V * given b y Norrish, A V * is found to be —3.5 cc. p e r mole. It w o u l d be interesting to compare A V * values for other cross-propagation reactions as a possible route to a n understanding of the nature of the activated complex. lx

l2

U

n

1 2

PLATZER; POLYMERIZATION AND POLYCONDENSATION PROCESSES Advances in Chemistry; American Chemical Society: Washington, DC, 1962.

1 2


ZU11Y AND BURKHART

55


A l t h o u g h E q u a t i o n 1 predicts that the rate of a radical termination reaction increases as pressure is applied, i t appears that a point is reached where these reactions become diffusion-controlled, resulting i n a positive rather than a negative A V * . Recent work suggests that even at atmospheric pressure a n d at l o w con versions the termination step of many polymerizations m a y be diffusion-controlled (2, 3, 13). It is not surprising, therefore, that increasing pressure decreases the rate of radical termination reactions (12) rather than increasing the rate as might be expected. T h e over-all rates for free radical polymerizations increase w i t h increasing pressure. T h i s means simply that the pressure-induced retardation of the initia tor decomposition rate is more than offset b y the increase i n the rate of chain propagation a n d the decrease i n the rate of chain termination. T h i s is formally stated i n terms of activation volumes i n E q u a t i o n 4 ( 1 5 ) AV* = AV* + i/2AV* - \/2AV*t (4) where a biradical reaction is assumed as the termination step. Subscripts d, p, and t refer to initiator decomposition, chain propagation, and chain termination, respectively. e

P

d

It is also to be expected that pressure w i l l affect the rate of c h a i n transfer reactions to monomer, polymer, a n d solvent. I n the polymerization of allyl ace tate, where degradative chain transfer to monomer occurs, the rates of the propaga tion a n d transfer reactions increase b y about the same amount for a given increase i n pressure (17). T h e transfer reaction becomes less degradative—i.e., the allyl acetate radicals become more reactive—as pressure is increased. Another aspect of free radical polymerizations under pressure w h i c h has been recently studied is the effect of pressure on comonomer reactivity ratios (5). I n two copolymerization systems—styrene-acrylonitrile a n d methyl methacrylateacrylonitrile—it was found that the product of the reactivity ratios, r r , approaches unity as the pressure is increased. T h e monomer-polymer composition curves for these two copolymerizations at 1 a n d 1000 atm. are illustrated i n Figures 1 a n d 2. T h e effect of pressure on the individual reactivity ratios and on the r r p r o d uct is given i n Table I I . 1

2

x

WT.% STYRENE IN INITIAL MONOMERS

Figure 1. Monomer-polymer composition curve for the styrene acrylonitrûe polymerization at 1 and 1000 atm. pressure


2


56


2

WT.%

METHYL METHACRYLATE IN INITIAL MONOMERS

Figure 2. Monomer-polymer composi tion curve for the methyl methacrylateacrylonitrile copolymenzation at 1 and 1000 atm. pressure Table II. Pressure, Aim. 1 100 1000 1 100 1000

Effect of Pressure on Monomer Reactivity Ratios

Monomer 7 Styrene Styrene Styrene MMA MMA MMA

Monomer 2 Acrylonitrile Acrylonitrile Acrylonitrile Acrylonitrile Acrylonitrile Acrylonitrile

ri

Γ2

nr

0.37 0.43 0.55 1.34 1.46 2.01

0.07 0.13 0.14 0.12 0.37 0.45

0.03 0.06 0.08 0.16 0.54 0.91

2

The methyl methacrylate-acrylonitrile copolymerization is nearly ideal at 1000 atm. (r r approaches very close to unity)—i.e., the free radicals have almost completely lost their individual selectivity. In the case of styrene-acrylonitrile the free radicals are tending toward a loss in selectivity w i t h increasing pressure, a l though the copolymerization is still far from ideal even at 1000-atm. pressure. These results suggest that of the two copolymerization parameters, Ç and e, only the e values are affected appreciably by an increase in pressure. Calculations have been made obtaining relative Q and e values w h i c h bear out this thesis. The decreasing radical selectivity found in these experiments implies an increased radical reactivity. This offers further support to Walling's suggestion that the allyl acetate radical increases i n reactivity w i t h increasing pressure. These results, of course, mean that the use of high pressures allows one to carry out copolymerizations w h i c h do not occur readily at atmospheric pressure because of widely different monomer-radical reactivities. F o r example, at several thousand atmospheres styrene and vinyl acetate, w h i c h w i l l not appreciably copolymerize at 1 atm. (10), may be made to give copolymers at a reasonable rate. As well as increasing monomer-radical reactivity, increasing pressure w i l l also increase monomer density, and if one or both of the monomers is normally gaseous, this w i l l allow the preparation of polymers w h i c h cannot normally be made at atmospheric pressure. Chief among such monomers is ethylene w h i c h , mainly because of its low concentration, w i l l not give high polymers nor copoly mers rich in ethylene at low pressures with free radical initiators. x

2


ZUTTY AND BU UKHAUT

57



T h e fact that ethylene w i l l copolymerize at h i g h pressures is rather fortunate from not only a commercial but also a theoretical view, since it has long been be lieved that monomer reactivity ratios might be better correlated according to a scheme based on ethylene rather than the currently used Q-e correlation i n w h i c h styrene is taken as the reference standard. However, until recently no quantitative data on ethylene copolymerizations have been available upon w h i c h to base such a scheme. T h e advantages to be gained b y utilizing an ethylene-based Q-e scheme are readily illustrated b y reference to Equations 5 and 6. η = Qiexp -(erf

(5)

r. - 1

(6)

These are simply the equations of Alfrey and Price (1 ), w h i c h relate monomer reactivity ratios to Q and e values, and i n w h i c h the reasonable values of e = 0 and Q = 1 are substituted, w i t h the convention that the reference standard, ethylene, is monomer 2. In Equation 6 it is seen that the Q value is simply a ratio of propagation rate constants unmodified by the presence of differences i n e values, as is the case i n the styrene-based scheme. This w o u l d seem to be a more desirable type of parameter to deal w i t h , simply because its meaning is perfectly straightforward. T h e reactivity ratio product is given b y 2

2

x

nr = exp — (e f

(7)

x

2

Thus, the copolymerization ideality i n ethylene copolymerizations—i.e., the proxim ity of r ^ to unity—is strictly dependent on the e value of the comonomer. Hence, we see that since ethylene, because of its lack of substituent groups, resides at the center of the e scale, relatively large positive or negative e values m a y be toler ated without seriously affecting the ideality of the copolymerization. Table III.

Reactivity Ratios and Q and e Values for Ethylene-Vinyl Chloride and Ethylene-Vinyl Acetate Copolymerizations Based on Based on C2H4 Styrene

Copolymerization Vinyl acetate-ethylene Vinyl chloride-ethylene 0

r\ 1.08 ± 0.19 3.60 ± 0.3

r* 1.07 ± 0.06 0.24 =fc 0.07 2

Qi 0.93 4.2

*i 0 +0.37

Qi 0.03 0.04

*i -0.3 -0.1

Ethylene is monomer 2.

Some data recently obtained on high pressure ethylene copolymerizations illustrate the quantitative aspects of an ethylene-based Q-e scheme (6). I n F i g ures 3 and 4 copolymer composition curves for the ethylene-vinyl chloride and the ethylene-vinyl acetate copolymerizations are given. T h e monomer reactivity ratios for these two systems are tabulated i n T a b l e III along w i t h Q values a n d e values for v i n y l chloride and vinyl acetate calculated using ethylene as the stand ard (Q = 1.0 and e = 0 ) . These Q and e values may be compared w i t h those obtained using styrene as the standard. These ethylene-based Q and e values may be used to calculate the reactivity ratios for the copolymerization of vinyl acetate w i t h vinyl chloride. Agreement is good when these values are compared w i t h experimental values. I n Table I V reactivity ratios calculated from ethylene- and styrene-based Q and e values are shown. PLATZER; POLYMERIZATION AND POLYCONDENSATION PROCESSES Advances in Chemistry; American Chemical Society: Washington, DC, 1962.



58

CHLORIDE IN INITIAL MONOMERS

Figure 3. Monomer-polymer composition curve for ethylene-vinyl chloride copolymerization made at 90° C. and 15,000 p.s.i.

WT.% VINYL ACETATE IN INITIAL MONOMERS

Figure 4. Monomer-polymer composition curve for ethylene-vinyl acetate copolymenzation made at 90° C . and 15,000 p.s.i.

Table IV.

Reactivity Ratios for Vinyl ChloridefM^Vinyl Acetate(M >) Copolymerization

Ethylene Q-e calculated Styrene Q-e calculated Experimental

r\ 3.7 1.36 2.1

f2

0.22 0.67 0.3

It appears that these ethylene-based Q and e values are capable of forming an internally consistent correlation scheme. It w i l l be interesting to see whether this scheme is capable of yielding good results over the w i d e variety of monomers for w h i c h the styrene-based scheme has been so successful. PLATZER; POLYMERIZATION AND POLYCONDENSATION PROCESSES Advances in Chemistry; American Chemical Society: Washington, DC, 1962.

ZUTTY AND BURKHART


59


Literature Cited (1) Alfrey, T., Price, C. C., J. Polymer Sci. 2, 101 ( 1947 ). (2) Allen, P. Ε. M., Patrick, C. R., Makromol. Chem. 47, 154 ( 1961 ). (3) Ibid., 48, 89 ( 1961 ). (4) Bridgman, P. W., "The Physics of High Pressure," pp. 128-30, 198, G. Bell and Sons, London, 1958. (5) Burkhart, R. D., Zutty, N. L., IUPAC Symposium on Macromolecular Chemistry, Montreal, July 1961; J. Polymer Sci., in press. (6) Ibid., in press. (7) Evans, M. G., Polanyi, M., Trans. Faraday Soc. 31, 875 ( 1935 ). (8) Ewald, A. M., Discussions Faraday Soc. 22, 138 ( 1956 ). (9) Hamann, S. D., "Physico-Chemical Effects of Pressure," pp. 163-6, Academic Press, New York, 1957. (10) Mayo, F. R., Lewis, F. M., Walling, C., Discussions Faraday Soc. 2, 285 ( 1947 ). (11) Nicholson, A. E., Norrish, R. G. W . , Ibid., 22, 97 ( 1956 ). (12) Ibid., p. 104. (13) North, A. M., Reed, G. Α., Trans. Faraday Soc. 57, 859 ( 1961 ). (14) Tsiklis, D. S., Dokkdy Akad. Nauk, S.S.S.R. 86, 1159 ( 1952 ). (15) Walling, C., J. Polymer Sci. 48, 335 ( 1960 ). (16) Walling, C., Pellon, J., J. Am. Chem. Soc. 79, 4776 ( 1957 ). (17) Ibid., p. 4782. (18) Ibid., p. 4786. RECEIVED January 4, 1962.


POLYMERIZATION AND POLYCONDENSATION PROCESSES

Recommend Documents