[COXTRIBUTION FROM THE K O Y E S CHEMICAL
LABORATORY, UNIVERSITY
O F ILLINOIS]
T H E STABILIZATION OF POLYSULFONES TOWVARD HEAT' C. S. MARVEL
AND
W. 13. SHARKEY
Received November 19, 1943
Wien a polysulfone derived from an olefin and sulfur dioxide is heated to its melting or softening point, it tends to decompose with the liberation of sulfur dioxide (1). In a patent (2) it has been claimed that heating a polysulfone with an olefinic compound containing a negative group linked to one of the unsaturated carbon atoms, at a temperature not to exceed loo", increases the stability of the polysulfone toward further heating. The compounds specifically mentioned as having this effect are vinyl acetate, chlorobutadiene, and acrylic esters. Wilder (3) has improved the heat stability of butenepolysulfones by treating them at 60-100" with such olefinic compounds as partially polymerized vinyl chloride, esters of acrylic and methacrylic acid, coumarin, furfural, or chloroprene. It has been reported (4) that the removal of volatile impurities, by heating a polysulfone in a current of inert gas to a temperature just below the temperature a t which the polymer itself begins to decompose, will improve the molding properties of the polymer. It has been found ( 5 ) that polysulfones are more stable to heat after they have been treated with acetyl chloride or acetic anhydride. Harmon (6) has observed that the molding properties of a polysulfone are improved by washing the polymer with a mixture of solvents, one of which exhibits a swelling action on the polymer. It is of interest to note that any of these procedures might be expected to destroy residual peroxides in the polymer. This may be the reason that these treatments have practical merit. In connection with a study of thermal stability of certain samples of l-pentenepolysulfone it was found that the results were not concordant when different samples were used. Further investigation showed that samples of this polymer prepared from 1-pentene, obtained by the action of ethylmagnesium bromide on allyl bromide, were exceptionally stable to heat. This property was found to be due to traces of allyl bromide left in the 1-pentene. Addition of allyl bromide in ext,remely small amounts (1 t o 2 drops t o 10 cc. of olefin) t o an olefin which was then converted t o a polysulfone, uniformly resulted in a polysulfone of greater thermal stability than was characteristic of that polysulfone prepared in the absence of allyl bromide. This effect seemed t o be rather specific for allyl bromide as the following compounds tested similarly were ineffective: allyl chloride, ethyl bromide, allyl alcohol, crotyl bromide, camphene, 1-bromoheptene, undecylenyl bromide, P-bromostyrene, ethyl acrylate, chloromethyl ether, chloroform, heptylmercaptan, carbon tetrachloride, p-bromobenzyl chloride, benzyl bromide, benzyl alcohol, furfuryl alcohol, furfurylacrylic acid, and &chloroisodurene. Some improvement in the thermal stability of a preformed polysulfone could be effected by dissolving the polymer in a suitable solvent, adding some allyl bromide, and heating. However, this effect was not marked. 1 This is the fourteenth communication on polysulfones. For the thirteenth communication see J . Am. Chem. Soc., 64,1229 (1942). 113
114
C. S. MARVEL AND W. H. SHARKEY
It was also found that purification of a polysulfone by solution and reprecipitation improves its thermal stability to some extent. Likewise, a thermal treatment somewhat like that suggested by Frey and Bury (4)was effective in giving a product which had improved thermal stability. Incorporation of small TABLE I EFFECT O F A412LYI, BROMIDE ON THE
-___
WEIGHT LOSS AT
THERMAL STABILITY
140' POP TWO HOURS
O F POLYSULFONE
160' FOR TWO BOUPS
WEIGHT LOSS AT
POLYSULFONE FPOY SAMPLE
G.
70
.33
0.4051 .2574
0.0081 .ow7
2.00 1.83
3.99 3.54
.3892 ,3511
.0474
.0049
.1798 ,1733
,0016 ,0019
0.812 1.10
.l523 .lo13
.0758 .IO40
.ooO7 .0017
0.92 1.63
1-Butene (trace allyl bromide)
.23GO .2026
.0033 .001s
1.40 0.89
1-Butene (control)
.2029 .2311
.OS23
1-Hesene (trace allyl bromide)
.a12 .2482
.0011 .O008
0.38 .32
1-Hesene (cont~rol)
.2G22 .2679
.0284 ,0296
10.80 11.05
1-Heptene (trace allyl bromide)
.3214 .2598
.0017 .0018
0.53 .69
1-Heptene (control)
.2673 .2597
.0250 .0305
9.36 11.75
Cyclohexene (trace allyl bromide)
,7536 .4523
.0461 .0284
6.14 6.28
Cyclohexene (control)
.2082 .2025
.0375 .033 1
18.00 16.35
SAMPLE
G.
0.3898 .2758
0.0018 .oO09
0.4G
,1556 .1386
.0062
2-Butene (trace allyl tlromide) 2-Butene (control)
1-Pcntcne (trace allyl bromide) 1 -Pentene (control)
.05GO
26.2 24.2
.0442
12.2 12.6
0042 .0025
2.75 2.46
.1308 .112GG
'0045
.mi3
3.44 4.98
.2300 .2774
.0113 .0160
6.22 5.78
.2375 .2602
.0392 .0388
*
16.5 14.9
amounts of benzoyl peroxide into a polysulfone gives a product of greatly diminished stability toward heat. EXPERIMENTAL
The heat stability tests were made by placing weighed samples of the polymers in small test tubes, which were then placed inside larger glass tubes, and these in turn were sus-
115
STABILIZATION OF POLYSULFONES
pended in a n oil-bath held a t the desired temperature for the specified time. The tubes containing the samples were then removed and weighed. In some cases considerable foaming occurred. Traces of ally1 bromide in the olefin. In Table I are summarized a number of experiments which show the effect of traces of allyl bromide in t h e olefin used for making a polysulfone on t h e thermal stability of t h e polymer. The polymers were prepared by a standard method described earlier (7), using 10 cc. of olefin, 10 cc. of liquid sulfur dioxide, 10 cc. of ethyl alcohol, and 0.1 t o 0.2 g. of ascaridole in the reaction mixture. For t h e stability
p!FFECT OF
POLYSULFONE
1
TABLE I1 HEATT R E A T M E S T O S STABILITY
1 WEIGHT SAMPLE
I
O F POLYSULFONES
LOSS I N WEIGHT 1Sf TWO HOUBS AT 140
LOSS IN WEIGHT ZND T w o HOURS AT 140'
%
G.
2.28
0.0032 2-Hutencb
.1383 .0409
.0108 ,0061
7.8 14.9
2-i'entene
.I718
.0268
15.6
1-Pcntenc
.2762 .2294
.0546 .0456
'
1.9.9 19.8
0
o.oO0o .oO06
0.44 0
.m
I
.0023
TABLE I11 EFFECTOF HEATTREATMENT FOLLOWED B Y REPRECIPITATION ON STABILITY OF POLYSULFONES POLY SULFONE
SAMPLE G .
LOSS I N WEIGHT AT 140' FOB 2 HBS. G.
%
FWTHEB HEAT TPEATMENT AT 140", HES.
WEIGHT OF BEPBECIPITATED SAMPLE, G.
I
LOSS I N WEIGHT AT 140' FOR G.
70
0.98
1-Ileptene
0.2751 .2797
0.0167 .0102
6.06 3.65
0.2867
0.0028
1-Pen tene
.3501 .3246
,0097 .0082
2.76 2.52
.4324
.0008
1-Pentene
.7767 .3398
.0185 .0082
2.38 2.42
.4589 .4163
.oO03 .0003
--
2 HES.
9
19
.06
.07
tests tlie polymers were isolated by pouring the reaction mixture into water, allowing t h e excess sulfur dioxide t o escape, heating the water t o boiling t o wash the polysulfone, cooling, separating the polymer cake, powdering the cake, washing i t thoroughly with ethyl alcohol and ether, and drying. E f e c t of allyl bromide on a preformed polysulfone. A preformed 1-pentenepolysulfone which lost about 1.5-2.0% its weight when heated t o 140" for two hours was rendered more stable by boiling a 3-g. sample in 30 cc. of dioxane with 10 g. of allyl bromide for five hours. The treated sample after isolation, lost less than 0.5% its weight under the same heat treatment a t 140'. Crotyl bromide and 1-bromoheptene also showed a stabilizing effect of t h e same order, but ethyl acrylate, camphene and cyclohexene did not.
116
C. S. MARVEL AND
W. H. SHARXET
Eflect of heat treatment. Heat treatment alone was found t o improve the heat stability of polysulfones. This was shown by measuring the weight loss in two consecutive twohour heating periods. and also by heating one two-hour period, dissolving the polymer in acetone, reprecipitating, and then heating the purified polymer. I n Tables I1 and I11 some of these data are collected. Effect on heat stnbilitg of repeated purijicntions. A sample of 1-pentenepolysulfone was purified by taking the polymer up in acetone, then adding two t o three volumes of ethyl alcohol, evaporating t o one-half volume, adding alcohol t r ) the original volume and again evaporating t o hnlf-volume t o precipitate the polymer, filtering. drying, powdering in a mortar and drying under 5 mm. a t 56" for four hours. When this material was heated t o 140" for two hours it lost 12-140, of its weight. A sample of 1-pentenepolysulfone made by the usual procedure n i t h ascaridole as a catalyst was taken up in chloroform, the solution boiled t o remove sulfur dioxids and then ether was added t o throw out the polymer. After filtering and drying nt 50" a t 3 mm. for four hours, the material lost 19-20% of its weight on heating t n o hours a t 140". -4 portion of this recovered polympr was boiled with water for two and a half hours, then powvdered, washed with alcohol and ether, and dried a t 78" and 5 nini. for five hours. It then lost 8-12% of its veight a t 140" in two hours. This material was then dissolved in wetone and precipitated TI ith alcohol as described for the first sample mentioned. This material then lost only 2-3y0 of its weight when heated t o 140' for two hours. Another trentment of the same sort yielded a polymer which lost only 1-1.37$ of its weight in the same heAt treatment. Further purification by this process did not improve heat stabilitl.. A sample of 1-hevmepolysulfone which was purified by solution in acetone and reprecipitated with alcohol as described above lost only 0.9% its weight a t 140' for two hours. Repwification by the same process gave a product which lost only 0.07-0.270 its weight on the heat treatment. A sample of 2-butenepolysulfone after repeated purification showed a loss of less than ly0its weight a t 140" for two hours. Eflect of added henzoyl peroxide. -4ddition of 2.74% benzoyl peroxide t o a sample of 1-pentenepolysulfone which had been found t o lose less than 0.5% its weight a t 140" for two hoiirs, gave a product nrhich lost 12% its weight in two hours a t 140". SUhl&LkRY
Polysulfones made from olefins containing a trace of allyl bromide show a remarkably improved stability toward heat over that of polysulfones made from pure olefins. The preformed polysulfone can also be treated with allyl bromide to bring about some stabilizing action but the effect is less marked. Heat treatment of polysulfones appears to remove some of the readily decomdecomposable material so that the remaining sample is more stable toward heat. Presence of peroxides in polysulfones increases the amount of decomposition which occurs when they are heated. URBANA, ILL. REFERENCES (1) (2) (3) (4) (5) (6) (7)
STAUDINGER AND RITZENTHALER, Bey., 88, 455 (1935). E. I. DUPONT DE NEMOURS AND Co., INC., French Patent 816,024 (1937). WILDER,U. S. Patent 2,146,276 (1939). FREY AND BURY,U. S. Patent, 2,154,444 (1939). MARVEL AND FREDERICK, U. S. Patent 2,169,364 (1939). HARMON, U. 8. Patent, 2,190,836 (1940). RYDEN,GLAVIS,. ~ N DMARVEL, J. Am. Chern. soc., 69, 1014 (1937).