VOl. ti7
PAULD. BARTLETTAND ROLFALTSCHUL
812
TABLE
Acetyl 2-Pyridyl 2-Thiazolyl
I
(Concluded) ‘
R
Alcohol Acetone Acet. alc.
R‘ = OCOCIII R’ = NHCOCHI
Prisms 9.88 9.95 100” 85* 198-199 12.17 12.00 Prisms 100 203.5 Cluster of h’eavy 100 207 12.01 12 14 plates 2-Pyrimidyl Acetone Yellow-orange Fine needles 100 195-196 15.18 15.46 Acetylguanyl Acetone Orange Needles 67 205-206 14.99 15.14 Acetyl Acet. alc. Orange Prisms 64‘ 90d 213.5-214.5 13.21 13.34 2-Pyridyl Acetone Orange Cluster of needles 91 149-150 15.25 15.28 2-Thiazolyl Acet. alc. Orange-red Clusters of needles 100 160-161“ 15.05 15.12 2:Pyrimidyl Acet. alc. Yellow-orange Cluster of prisms 99 208-209’ 18.26 18.36 Acetylguanyl Alc. petr. ether Orange Fine needles 87 246 18.02 18.14 a Acetylation of 4-(3’-methyl-4’-hydroxynaphthylaz~)-benzenesulfonamide. Acetylation of 4-(3’-methyL4’-hydroxynaphthy1azo)-N-acet ylbenzenesuffonamide. Acetylation of 4-(3’-methyl-4’-aminonaphthylazo)-benzenesulfonamide. Acetylation of 4-(3‘-methyl-4’-aminonaphthylazo)-N-ac~tylbenzenesulfonamide. a Two polymorphic forms m. p. 280’ and m. p. 218” yielded the same acetyl derivative. f Two polymorphic forms m. p. 283” and m. p. 248-249’ yielded the same acetyl derivative.
+
Orange Yellow-orange Red-orange
+ + + +
-
procedure of Lesse9 with slight modification. The yield Methyl - 4’ hydroxynaphthylazo) - heiizenesulfonamide was about W 9 0 % . and 4-(3’-methyl-4‘-hydroxynaphthylazo)-sulfa1iilylguani2-Methyl-I-naphthylamine hydrochloride ahd 2-methyl- dine yielded diacetyl derivatives, respectively. The prop1-naphthol were, respectively, coupled with sulfatiilamide. erties of these acetyl derivatives are listed in Tahle I The hydrochloride of 4-(3’-methyl-4’-aminanaphthylazo)benzenesulfonamide thus obtained was treated with a 10% Summary sodium hydroxide solution to liberate ,the free amine. 4(3’-Methyl-4’-aminonaphthylazo)-benzeneThe analogs from albucid, sulfapyridine, sulfathiazole, sulfadiazine and sulfaguanidine were similarly prepared. sulfonamide, 4-(3’-methyl-4’-hydroxynaphthylThe yields and properties of these compounds are listed in azo)-benzenesulfonamide and their analogs derived Table I. from albucid, sulfapyridine, sulfathiazole, sulfadiAcetyl Derivatives of 4-(3’-Methyl4’-aminonaphthylazine and sulfaguanidine have been respectively azo) -benzenesulfonamide, 4-( 3 ’-Methyl4 ’-hydroxynaphthy1azo)-benzenesulfonamide and Analogs.-A sample of synthesized. Also corresponding acetyl deriva0.001 g. mole of 4-(3‘-methyl-i‘-aminonaphthylazo)- tives have been prepared. benzenesulfonamide was refluxed with 1 cc. of acetic anA preliminary test has shown that 4-(3‘hydride and about 3 drops of pyridine for an hour. It hydroxynaphthylazo) benzenewas poured into cold water and then filtered. The pre- methyl - 4’ cipitate was recrystallized from a suitable solvent. This sulfonamide possesses an antrhemorrhagic activwas identified as the diacetyl derivative by analysis and ity almost equal to that of 2-methyl-l,4-naphthomix-melting point with the acetyl derivative obtained by quinone. An extensive study of antihemorrhagic acetylation of 4-(3’-niethyl-4’-aniirioriaphthylazo)-K-acetylbenzenesulfonaniidc.. 4-(3’-Jiethyl-l’-aminonaphthyl-activities of these compounds is in progress. 2-Methyl-1-naphthylamine has been prepared azo)-sulfa~iilylguanidinealso forms diacctyl derivative which occludes acetic acid tenaciously and has to be stirred by refluxing 2-methyl-1-nitronaphthalene with with a lOy0SaOH befsre recrystallization. Acetyl derivaRaiiey nickel only, in a yield of 96%. tivcs of 4-(3’-methyl-i‘-liydroxynaphthylazo)-benzeneRECEIVED FEBRUARY 10, 1945 sulfonamide and analogs were similarly prepared. 4-(3’- AUSTIK:TEXAS
-
[CONTRIBUTION FROM
THE
-
CON.VVERSE MEMORIAL LABORATORY OF HARVARD UNIVERSITY]
The Polymerization of Allyl Compounds. I. Factors Governing the Acyl PeroxideInduced Polymerization of Allyl Acetate, and the Fate of the Peroxide BY PAULD. BARTLETT AND ROLFALTSCHUL’ If pure allyl acetate is heated a t 80’ with 6% of its weight of benzoyl peroxide, the slow decomposition of the peroxide induces polymexkation of the ester. The peroxide is 90% decomposed after thirteen hours; only an undetectable amount remains after forty-eight hours, at which time about 50y0 of the Driginal allyl acetate has polymerized. The soft, transparent polymer is obtained by distillation. of the volatile monomer. It is soluble in a number of organic solvents (1) Pittsburgh Plate Glass Fellow, 1941-1944; present address. Department of Chemistry, Bryn Mewr College, Bryn Mawr, Pa.
and has an average molecular weight of about 1300, hence an average degree of polymerization of 13. This low degree of polymerization and high requirement of initiating peroxide makes allyl acetate a favorable material for studying the mechanism of .polymerization, especially with regard to the fate of the peroxide. Different samples of allyl acetate, prepared with care to remove free alcohol and acid and distilled in the absence of air, yielded reproducible results. The polymerizations described in this paper were followed by bromometric titration of
ACYLPEROXIDE-INDUCED POLYMERIZATION OF ALLYLACETATE
May, 1945
weighed samples of the monomer-polymer mixture. At times we also made use of two other methods, the dilatometric and the gravimetric, for following the polymerizations. All three methods checked closely when applied40 the same reaction. Table I shows the reproducibility of the polymerization using three samples of allyl acetate prepared a t different times, and also adding to one of these samples substances which might occur as impurities from the pteparation and treatment of the ester. Oxygen is capable of retarding the polymerization of allyl acetate, but the effect of simple exposure to air is not great. In order to demonstrate it, a sample undergoing polymerization must be agitated vigoiously with air or oxygen. A solution containing 6.10% of benzoyl peroxide 'In allyl acetate, sealed under nitrogen, was 20.5% p l y merized after 2.5 hours at 80'; a similar solution, sealed in air and spun mechanically during polymerization, contained 8.3% polymer after the same length of time. Comparable tests showed that a series of evacuations and nitrogen flushings using a water-pump a t ' 0 was as effective in eliminating air inhibition as the most careful series of freezings with liquid nitrogen and exhaustions with a diffusion pump; the effect of any of these precautions was about a 10% increase in rate over that observed in air without agitation (Table 11). TABLE I
SIB
I11 shows the chlorine balance for a polymerization of allyl acetate containing 5.90% of p-chlorobenzoyl peroxide, which was carried out to total decomposition of the peroxide, and followed by recovery of the free and extractable chlorobenzoic acid, determination of the bound and saponifiable chlorobenzoate groups, and finally an estimation of the bound and unsaponifiable chlorophenyl groups. TABLE I11 DISTRIBUTION OF CHLORINE AFTER POLYMERIZATION OF ALLYL ACETATEWXTH 5.9070 BY WEIGHT OF P-CHLOROBENZOYL PEROXIDE AT 80" %
Mg.
Polymer sample, 0.765 g. Chforine present in original peroxide 21.73 100 Chlorine in free chlorobenzoic acid isolated 3.60 16.8 Chlorine bound to polymer, by analysis 16.76 72.5 Chlorine unbound and unaccounted for 2.31 10.7 Chlorine in pure chlorobenzoic acid hydrolyzed from polymer 11.32 52.1
During a run similar to this the carbon dioxide evolved was swept through with nitrogen and cbllected quantitatively in an ascarite tube. It amounted to 0.24 mole per mole of initial peroxide, or 12% of the maximum carbon dioxide possible if each p-chlorobenzoate radical formed had undergone decarboxylation. The Evolution of Carbon Dioxide from Benzoyl Peroxide.-The form in which peroxide fragREPRODUCIBILITY OF POLYMBRIZATION OF ALLYLACETATE ments are attached to the polymer-whether as AT 80.0 * 0.2' IN THE PREWNCE OF ADDEDSUBSTANCES benzoate or as phenyl groups-is related to the AND 4.58% BENZOYL PEROXIDE evolution of carbon dioxide during the polymeriBatch of Substance added % Polymerization zation, since for each phenyl group attached to and moles % after 6.00 hours monomer None 31.0 1 the polymer one carbon dioxide molecule must 31.1 2 None have appeared. The reactions have been writtens 3 3 3 3
None HtO (15) HCI (2) HIO (5) Pyridine (tr.)
+
30.4
,31.1 31.1 30.3
-
c$Iscoo--ooccdIs PcsHsCOOCaHrCOO- +cot + casC&COO- + CHFCHR ---+ QH&OOCHICHRC a s - + CH+HR +G"CH&HR-
We have determined the amount of carbon dioxide TABLE I1 evolved during the polymerization of allyl acetate EFFECTOF DEGASSING PROCEDURE ON RATE OF POLYI+~EWATIONOF A u n ACETATE AT 80.0 * 0.2'; BENZOYL with conceutrations of benzoyl peroxide varying -OXIDE,
1.12%
For a description of the procedures, see the Experimental Part. Procedure
Air, no agitation A B C D
3t2s3:2 11.2 12.5 12.2 12.4 12.5
Determination of Pate of Peroxide.-It has been shown in a number of cases that when polymerization is induced. by the decomposition of acyl peroxide, fragments of the peroxide are permanently attached to the polymer.* Table (2) Price, KelI and %elm, T E I ~ JOUBNAL, 64, 1103 (lerZ); Rice and Durham, iMd., p. 2608: Ian sad plmmerer, J . @aY. C h . , 161, 181 (lMZ); Price and Taw, T8xn JOUXNAL, 66, 617 (1943); B M e t t s a d Cohan, ibid., p. 64%.
from 2.68 to 10% by weight, with the results shown in Table IY. The total amount of carbon dioxide produced by a given amount of peroxide is seen to be independent of the initial peroxide contentration in the absence of oxygen, but to TABLE IV CARBON DIOXIDE EVOCUTION DURING BENZOYL PBROXZDECATALYZED POLYXBBJZATION OF ALLYL ACETATEAT
80.5 Puoxide wt. % molw/l. 0.389 10.0 6.13 .233 2.68 ,103
10.0 10.0 2.88
* 0.5'
%
Atmosphere
Pole
m&t n attained
MoleaCOI/ mole peroxide
NI NI
71
0.64,0.60,0.85
4s
0.64
NI
26
.70 .88 1.68
,389 Os, slow stram .389 Os, lost stream .lo3 a,fMt Strum
44:s 26.0 19
(3) Price and Itell, ibid., 6% 2800 (1821).
1.76
314
PAULD. BARTLETTAND ROLFALTSCIIUL.
L'ol. (;i
be much increased by the passage of a rapid was so great, due to solubility of t h e pu1y~Ayl stream of oxygen through the polymerizing solu- alcohol in water, that the result is of little signifition. This effect of oxygen is seen again when cance. Were we to assume that the small anioun t benzoyl peroxide is decomposed in solution in of reacetylated polymer isolated was typical of toluene a t 100' or in nitrobenzene at 80' (Table the whole, then the p-chlorophenyl groups in the V). In these solvents a mole of peroxide yields polymer would account for 1.7 X 0.758/0.16 = almost twice as much carbon dioxide as in allyl 8% of the original chlorine (correction being ncetate under comparable conditions. made for that part of the 0.765-g. sample removed as benzoic acid by hydrolysis). This is a miniT A B L E \' mal figure for p-chlorophenyl groups, since there CARBON DIOXIDE EVOI.UTION DURING DECOMPOSITION OF would be more loss of the lowq-molecular-weight J ~ E K Z O P L PEROXIDE I N TOLUENE AT 100 * 0.5" AND components of the polyallyl alcohol than of the NITRORENZENE AT xO.5 * 0.5" higher. However, there cannot have been more Moles chlorophenyl groups attached to the polymer I'eroxide, .ttmos- Temp., k COi/niole Solvent rnolesil. plirre 'C. (hr. peroxide than there was carbon dioxide eliminated, and 0 , 208 Toluene N? 100 1.38 1.282 this sets an upper limit of 12%. AH the measureCI., 100 ... 1.45 208 Toluene ments are therefore consistent with the estimate ,126 Nitrobenzene N2 80 0.175 1.013 that of the chlorine in the original peroxide, .l26 Nitrobenzene (A 80 ,130 1.272 (j0.50/, is attached to the polymer as p-chlorobenDiscussion.-Of the lfi.Syo of fiee p-chloro- zoate groups, 12% as p-chlorophenyl groups, benzoic acid isolated from the polymerization, 16.8% appears as free p-chlorobenzoic acid, and not more than one-fourteenth, or 1.2y0,can pos- 10.7% is converted into this or other unattached sibly have been present as an impurity in the origi- products not isolated. A comparison of the amount of carbon dioxide nal peroxide, for this peroxide showed 98.8% purity by iodoirietric titration. This indicates evolved from benzoyl peroxide with that from p that the most important path by which peroxide chlorobenzoyl peroxide shows that the figures fragments are lost in allyl acetate, without giving arrived a t above are no guide to the nature of the rise to pulyiner, is tlie taking of hydrogen from end-groups in polyallyl acetate prepared with somewhere by benzoate radicals. The source of unsubstituted benzoyl peroxide. The unchlorithis hydrogen is coilsidered in a later discussio~i.~nated peroxide evolves nearly three times as much The molecular weight of the polymer produced carbon dioxide (30-35% instead of 12y0)as does by the chlorobenzoyl peroxide was 1330, and p-chlorobenzoyl peroxide under the same conditherefore the 0.765-g. sample of this polymer de- tions. The assumption that carbon dioxide is evolved scribed in Table I11 contained 0.575 millimole. The amount of chlorine found in,it was 15.761 by unimolecular decomposition of a benzoate 35.46 = 0.444 milligram-atom. Hence, on an radical leads to the prediction that less carbon average, only 77% of the polymer molecules con- dioxide should be evolved, fbe greater the concentain fragments from the peroxide. The growth of tration of monomu molecules a t hand to react at least 23% of the polymer molecules must have with the benzoate radicals and shorten their free been initiated otherwise than by the attachment lifetime. 'The figures contradict this prediction, of a chlorine-containing radical. If any of the since in a run in which 71% polymerization occhains were terminated by union with chloro- curred the evolution of gas was not proportionbenzoate or chlorophenyl radicals, thus having ately greater than in a run reaching only two peroxide fragments per molecule, it would polymerization. This seems to indicate that the mean a corresponding increase in the number of carbon dioxide does not issue from a unimolecular polymer molecules having no peroxide fragment decomposition of benzoate radicals. The facts attached. Presumably, then, a t least 23y0 of all would be in accord with either of two interpretathe growing chains i n this experiment were ini- tions: either &at all the carbon dioxide evolved tiated (while others were terminated) by chain is produced spontaneously. by decomposition of transfer.b Since 52.1% of all the initial chlorine benzoyl peroxide at once into phenyl radical, was isolated as purified p-chlorobenzoic acid from carbon dioxide, and possibly benzoate radical6 the hydrolysis of the polymer, a t least 52.1/72.5 or 72%, of the end-groups were p-chlorobenzoate groups. This is a minimal figure, since there was or alternatively that carbon dioxide is eliminated some loss in recrystallizing the acid. from benzoate radicals only at the moment of An attempt was made to determine the p- reaction with another molecule, as for example chlorophenyl groups by reacetylating the polyC+"COO CHFCHR & COz 4- GF€rCHdXR allyl alcohol resultmg from the previous hydrolysis and determining the chlorine content of this or ca,coo + Ot +cot 4- CiHrOO reconstituted polymer. The loss in this process (the last radical above being possibly involved' in (4) Part I1 of this series: TEISJOURNAL, 67,816 (1945). -1)
+
(51 Flory,
