5940
MARTIND. BARNETT,ALLANCRATVSIIAW AND GEORGE B. BUTLER
beam instrument using an identical solvent-filled cell in the reference beam. The percentage of I and I1 in each mixture was then calculated froin extinction coefficients previously determined at each wavc leiigtii. 'l'hese data are summarizcd in Table I . Isomerization Study.-The isumerization of I to I1 was carried out at 61.2, 80.1 and 100.0' by thermostating solutions of known concentration uf I in n-decane. Samples were removed for analysis (done as described previously) a t times which gave a good conversioii-time curve a t each temperature. A plot of In X , / ( X , - X ) us. time for each temperature is shown in Fig. 1. Thermodgnamic values and rate d a t a are summarized in Tables TI and 111,respectively. Lithium Aluminum Hydride Reduction.-Bromine was added to butadiene in chloroform a t -15" in the manner previously described. After addition of the bromine was complete the chloroform was flashed off under reduced pressure a t the reaction temperature and tetrahydrofuran was added to replace the chloroform. Lithium aluminum hydride in tetrahydrofuran was added a t - 18 to - 15' and a temperature of -15' was maintained for two hours. T h e reaction mixture was then heated to reflux temperature (67") and held at that temperature for three hours. The reaction products were distilled into a cold trap ( D r y Ice-acetone),
[CONTRIBUTION FROM
THE
Yol. 81
weighed and analyzed by gas cliromntograpliy (Fig. 2 ) . Only one product was formed, luans-2-butene, and the X I X terial balance on the butadiene was 99%. Figure 2 also contains data obtained from the lithium aluminum hydride reduction of a mixture of 3,4-dibromo-1butene and 1,I-dibromo-2-butene. Similar treatment ( ~ f 92.5% 3,I-dibromo-1-butene and pure 1,4-dibromo-%-butene gave trans-2-butene as the only product with a yield of 95c/u or higher. Analysis of a mixture of the three olefins possible (1-butene, cis- and tram-2-butene) and of a 9770 trans- 3% cis-2-butene mixture are given in Fig. 3. The gas chromatographic equipment used for analysis consisted of a 10 f t . X 3 / 4 in. copper column with dinonyl phthalate on fire brick as the packing. The detector unit was a Gow-Mac thermal conductivity cell and helium w a s the carrier gas.
Acknowledgment.-The authors are indebted to The Robert A. Welch Foundation for the financial support of this research and to W. 13. R. Shaw for valuahle discussions concerning the kinetic studies. AUSTIX12, TEX.
DEPARTMENT OF CHEMISTRY, UNVERSITY O F
FLORIDA]
The Formation of Linear Polymers from Diene Monomers by a Cyclic Polymerization Mechanism. 111. A Study of the Structure of Cyclic Polymers of Certain Unsymmetrical 1,6- and 1,7-Dienes1>* B Y hfARTIN D. BARNETT,** ALLANCRAWSHAW3b AND GEORGEB. BUTLER RECEIVEDAPRIL25, 1959 The methyl allyl, methyl 2-butenyl and methyl 3-butenyl mixed esters of maleic and fumaric acid have been synthesized and polymerized in bulk using benzoyl peroxide to afford a series of low molecular weight linear polymers containing 37-77% residual unsaturation. The properties of the polymers indicated t h a t polymerization had occurred by a n intramolecularor (in the case of intermolecular mechanism leading, in part, to a poly- [3(5)-methylene-4-carbomethoxy-6-valerolactone] the 3-butenyl esters) to a poly- [8(5)-methylene-l-carbomethoxy-~-caprolactone] structure. In addition, poly-(methyl allyl maleate) and poly-(methyl 2-butenyl maleate) showed absorption at 1789 em. - 1 characteristic of a y-butyrolactone ring system. This band was lacking in the corresponding fumarates. Possible explanation; of the various modes and relative degrees of cyclization in the polymers are presented.
Recent work4-l0 has verified the original proposal" that certain monomers containing a 1,Gdiene system can, on polymerization, undergo an alternating intramolecular-intermolecular chain propagation or cyclization, leading to the formation of saturated linear chains containing alternating six-membered rings and methylene groups. I n additioii, extensions of this new chain propagation mechanism to diene systems capable of forming rings containing more than six carbon atomsI2 (1) T h i s research was supported b y t h e United S t a t e s Air Force through t h e Air Force Office of Scientific Research of t h e Air Research and Development C o m m a n d , under Contract Kumber AFlS(603)-116. Reproduction in x hole or in p a r t is permitted for any purpose of t h e United S t a t e s government. (2) T h i s paper w a s presented before t h e Division of Polymer Chemistry, American Chemical Society, Boston, Mass., April, 1069. (3) (a) Post-doctornl fellovrr, 1958-1059; (b) Post-dortnral fellow, 1956-1968. (4) G. B Butler, A. Crawshaw a n d W. L. Miller, THISJ O U R N A L , 80, 3615 (1958). ( 5 ) A. Crawshaw and G. B. Butler, i b i d . , 80, 5464 (195s). (R) C. S. hIarvel and R. D. Vest, i b i d . , 79, 5771 (1957). (7) C. S. Rlarvel a n d R. I). Vest, ibid., 81, 984 (1959). ( 3 ) C. S. Marvel and J . K . Stille, ibid., 80, 1740 (195b) (9) J . F. Jones, J . Polymer Sci., 33, 7 (1968). (10) J . F. Jones, i b i d . , 33, 15 (1958). (11) G. 13. Butler and R. J . Angelo, T H I S J O U R N A L , 79, 3128 (1957). ( 1 2 ) (a) C. S. Marvel a n d W. E. Garrison, Jr., ibid., 81, 1737 (1959); (b) T. H a l t a n d W. Simpson, Proc. R o y . Soc. (Londoiz), 8238, 154 (195G); (c) 51. Oima a n d Y. Ogata, J . Chem. SOC.J a p o n , 79, 1506 (1958).
as well as into the copolymer field13have now been reported. Work in this area thus far has been largely limited t o the study of symmetrical 1,B-dienes. Unsymmetrical 1,6-dienes, however, also should be capable of homopolymerization via cyclization, especially those monomers containing double bonds of comparable reactivity toward copolymerization. Although the homopolymerizations of such unsymmetrical 1,e-dienes as allyl acrylate,l 4 allyl ~ n e t h a c r y l a t eand ~ ~ , substituted ~~ allyl metha c r y l a t e ~ ~have ' been reported, there appears to have been E O previous study directed specifically toward verification of the presence of cyclic units in the polynier derived from such dienes. Unsaturated esters of maleic and fun-aric acid seemed particularly suited for such a study because systems of this type should permit a qualitative coniparison of the relative reactivities of various unsaturated alcohols as well as a comparison of the degree of cyclization in the cis-trans isomers. (13) W. H. Schuller, J. A . Price, S. T. Moore and W. 51. Thomas.
J. Chem. E n g . Data, 1, i n Dress (1959). (14) L. Gindin, S. h1edvedev a n d E. Flesher. J . Gen. C h e n . (U..v.S . R . ) , 19, a127 (1949). (15) E. R . Blount and B. E. Ostberg, J . Polymer Sci.. 1, 230 (1946). (10) S. G. Cohen, B. E. Ostberg, D. B. Sparrow a n d E. R . Rluunt, ibid., 3. 264 (1948). (17) G. M. Bristow, Trans. Faraday Sac., 64, 1004 (19%).
5947
CYCLIC POLYMERS O F vNSYMUETRICAL 1,G- AND 1,7-DIENES
c PHYSICAL PROPERTIES OB R~ONOMERS,
11
c
/ \
Ri Yield, Compd. H Ia H Ib
%
Rz
Rl
IC
H
IIa IIb IIC
-COOCHzCH=CHa -COOCHzCH=CHCHa -COOCHzCHzCH=CHx
-COOCHzCH=CHz --COOCHCH=CHCHa -COOCHzCH2CH--CIL H H
€1
39 21 48 G2 01 72
Boiling point "C. Mm. 80-81 1.0 0,6 73.5-75 0.0 71.5-72 65.5-70.5 1.1 71-71.5 0.7 63.5-64.5 0.6
Rt
$ 2 2 ~
d22
1.4580 1.4580 1.4561 1.4590 1.4622
1,1031 1,0762
1.0730 1,0990 1,0683 1.0051
1.4580
Carbon, % Calcd. F o u n d 56.35 56.46 58.68 58.60 58.79 48.68 56.48 46.46 58.80 58.08 58.57 88.68
Hydrogen. % Calcd. Found 5.94 5.90 6.57 6.68 8.57 0.59 5.94 6.15 6.72 6.57 G.73 6.57
TABLE I1 PHYSICAL PROPERTIES OF POLYMERS Pvlymer
Poly-Ia Poly-Ib Poly-IC Poly-IIa Pol>r-IIb Poly-IIC
--Carbon, Calcd.
56.46 58.65 58.68 56.46 58.68 58.68
%-
Found
Hydrogen, 70 Calcd. Found
56.51 58.34 58.56 56.59 58.70 58.44
5.94 6.57 6.57 5.94 6.57 6.57
Experimentall* Fumaryl Chloride Monomethyl Ester (III).-A procedure which obviates the necessity of isolating methyl hydrogen fumarateig or separating I11 from fumaryl chloride by fractional distillation20 is: A mixture of 392.0 g. (4.0 moles) of maleic anhydride and 800 ml. of dry methanol was allowed to stand a t room temperature overnight and the excess methanol removed under reduced pressure. T o the residue was added cautiously with sxirling a mixture of 375 ml. of thionyl chloride and 400 ml. of dry benzene and the slurry heated t o reflux. The mixture became homogeneous within 1.0 hr. and refluxing was continued for an additional 18.0 hr. After removal of solvent and excess thionll chloride under reduced pressure there was obtained 394.3 g. of fsintly \-ellow liquid, b.p. 93-95' (29-30 mm.). Redistillation \.ielded 332.0 g. (56%) of colorless acid chloride (HI), b.p. 83-84' (24 mm.), reportedzGb.p. 70-il" (14 mm.). Preparation of Monomers. ( a ) Methyl Alkyl Maleates (La, b , c).-The synthesis of methj-l 3-butenyl maleate illustrates the procedure followed. A mixture of 29.4 g. (0.30 mole) of maleic anhydride and 200 ml. of dry methanol was allowed to stand at room temperature overnight and the excess methanol removed under reduced pressure. T o the residual methyl hydrogen maleate was added 21.6 g. (0.30 mole) of 3-buten-1-01,~~ 150 ml. of dry benzene, 1.0 g. of p toluenesulfonic acid and 0.5 g. of shiny copper wire and the mixture refluxed under a Dean-Stark trap until the theoretical amount of water had been collected (ca. 24 hr.). After extracting the mixture with three SO-ml. portions of 10% sodium bicarbonate solution, water, and drying over anhydrous sodium sulfate there was obtained upon distillation 38.0 g. of colorless liquid, b.p. 80-85" (1.0 mm.). Redistillation afforded 26.6 g. (48%) of methyl 3-butenyl maleate ( I a ) , b.p. 71.5-72" (0.6 mm.). The physical properties of the methyl alkyl maleates (Ia,h,c) are recorded in Table I. (b) Methyl Alkyl Fumarates (IIa,b,c).-The preparation of methyl 3-butenyl fumarate illustrates the general method. To a stirred solution of 22.0 g. (0.30 mole) of 3-buten-l-o1,21 73.7 g. (0.30 mole) of dry pyridine and 200 ml. of dry ether contained in a 1-l., three-necked flask fitted with a dropping funnel, stirrer and condenser (calcium chloride tube) was (18) All melting a n d boiling points a r e uncorrected. Infrared spectra were recorded using a Perkin-Elmer model 21 double beam spectrophotometer. Analyses were carried o u t b y Galbrtrith Laboratories, Knoxville, Tenn., or b y Weiler and Strauss, Oxford. England. (19) EL Erlcnmeyer and W. Schoenauer, H e h . Chim. Acta, 20, 1008 (1937). (20) R. E . L u t z , THISJ O U R N A L , 62, 3423 (1980). (21) Prepared b y t h e lithium aluminum hydride reduction of vinylacetic acid according t o t h e method of R. T. Arnold, R. W.Amidon a n d R. M . Dodson, THISJOURNAL, '72, 2871 (1550).
5.84 6.58 6.58 5.91 6.60 6.39
Approximate softening "C. point,
95 110 105 100 75 100
lnherent viscosity , concn., g./lOO ml.
0.061 (0.306) ,036 ( .297) Not detd. Not detd. ,044 ( .243) ,072 ( ,262)
added dropwise 44.5 g. (0.30 mole) of redistilled acid chloride I11 over a 1.5-hr. period. T h e reaction mixture rapidly darkened and deposited pyridine hydrochloride. it;hen addition was complete the mixture was refluxed for 1.0 hr., 100 ml. of water added, and the ether layer washed with water, loyosodium bicarbonate solution, water, and dried over anhydrous sodium sulfate. Distillation afforded 42.5 g. of colorless product, b.p. 69-67' (0.6 mm.), which was redistilled to give 39.8 g. (72y0) of methyl 3-butenyl fumarate ( I I c ) , b.p. 63-64" (0.6 mm.). The physical properties of the methyl alkyl fumarates (IIa,b,c) are recorded in Table I. A strong absorptiou band a t 965-970 mi.-' characteristic of a trans structure \vas found in the infrared spectra of the fumarates (trans-acid bond) and methyl 2-butenyl maleate (trans-alcohol bond). This band was absent in the other maleates. Mass Polymerizations.-The general method was t o heat approximately 10.0 g. of monomer containing the appropriate amount of benzoyl peroxide (Table 111)in a screw-capped bottle under dry nitrogen at 02-63' for a period just short of the gel time (Table 111). The soluble polymers were obtained by pouring the polymerized mixture into pentane, decanting the clear supernatant liquid, dissolving the semisolid residue in acetone and reprecipitating the polymer from pentane. This was repeated twice and the residual solid dried at room temperature for 12-24 hr. at 0.5 mm. After being finely ground, the polymer was dried at 0.5 mm. for an additional 12-24 hr . The polymers mere soluble in chloroform, carbon tetrachloride, acetone, acetic acid and acetonitrile and insoluble in ether, methanol, ethanol, cyclohexane, benzene and pentane. T h e analyses and approximate softening points of the polymers are presented in Table 11. The infrared data are presented in the Discussion. Determination of Residual Unsaturation.-The method of SiggiaZz was modified as follows. An iodine flask was charged with 50 ml. of glacial acetic acid, a 20-257, excess of 0.1 N potassium bromate-potassium bromide solution (as determined by previous titration for residual alcohol unsaturation), 5 ml. of 6 N sulfuric acid and the solution stirred magnetically for 2-3 min. A 10-ml. aliquot of polymer solution (prepared by dissolving ca. 1 g. of polymer in glacial acetic acid and diluting to 100 ml.) was added rapidly and the mixture stirred in the dark for 10 min. Fifteen milliliters of 2OY0 potassium iodide solution was added, stirring continued for another minute, and the solution rapidly titrated with 0.05 N sodiuni thiosulfate (starch end-point) to give residual alcohol unsaturation. T o determine residual acid urisaturstion an iodine flask was charged with 50 ml. of glacial acetic acid, a n amount of (22) S. Siggia, "Quantitative Organic Analysis uia Functional Groups," John Wiley a n d Sons,I n c . , New York, N. Y . , 1954,p. 69.
5948
LIARTIN
D. BARNETT,ALLAN CRAWSHALT
0.1 ,V bromate-bromide solution exactly equivalent t o the residual alcohol unsaturation, 5 rnl. of 6 N sulfuric acid and the solution stirred magnetically for 2-3 min. rl 10-ml. aliquot of pol)-mer solution was introduced :md stirring continued in the dark for 10 min. 'Then, wliile keeping the flask as shielded from light as possible, an excess of 0.1 "V bromatebromide solution was added, followed by a mixture of 25 mi. mercuric sulfate of glacial acetic acid aiid 25 nil. of 0.2 solution. The mixture was stirred in the dark for 1.0 hr., after which 15 nil. uf 2 ,Vsodium chloride solution and 15 ml. of 20yOpotassiuiii iodide solution were added, stirring continued for 2-3 min., ruid the solution titrated with 0.05 iV sodium thiosulfate to give residual acid unsaturation. ~1 blank vas run using 50 i d . of glacial acetic acid and about one-half the volume of bromate-bromide solution necr y for the titration of the total residual unsaturation. T h e above procedure was chccked on each of the six monomers. Gel Times and Percentage Conversion at Gelation.-The gel time of each monomcr XIS determined by heating a 1-g. sample containing the requisite ninount of initiator under drl- nitrogen a t 62". T!ie percentage conversion was determined under the same coiitlitions using 4-5 one-gram samples of monomer in sinal1 scrcw-c:ippeti vials. Samples were removed periodically, cartfull!: poured into pentane, arid the precipitated pol>-rricr iiltered Ionto previously weighed sintered glass funnels. The pol)-nic.r wa$ 11-ashed thoroughly with pentane aiid dried at room temperature and 0.5 rnm. overnight. h plot of time iss. weight of polymer gave 311 cssentially straight liue which, upon esrrapo1:rtioii to the gcl titnc, aflorded perceiitage conver'iion. Tlic results are recorded in Table 111. GEL
TABLE I11 TIMES,P E R C E S T A C E CONVERSIOT AT GELATIOXA S D CATALYST-TO-MOSOMER RATIO
I'd> mer
Gel time lit-
Benzoyl Cmversion peroxide as 5% a t gelation, % o f monnrner
Poly-Ia 9 1s 2 Poly-ib >2S0 2 Poly-ic 8.25 12 > Pol>--IIa 11.5 (1~1'' 9 0.1 Poly-iIb 56 17 2 Pol>--IIc 12 j l f 1.1 0.1 (6 1-0 gelation after 243 hr. The ""i;:conversion a t gelation" represents the yield of soluble palymer after this period of lieating a t 62". Using 2:lO benzoyl peroxide. 3
Viscosity Measurements.-The
inherent viscosities of the
poi!-iner-; !yere determined in glacial acetic acid at 30.0' using ;I:I Ubbclohde viscometer. T h e values are recorded in
'rnbie 11.
Results and Discussion 'The syntheses of the mixed maleate and fumarate esters are summarized in Fig. 1. Fumaryl chloride niononiethyl ester (111) could not be obtained in good yield by published procedures.19.20 However, refluxing a benzene solution of methyl hydrogen maleate niid thionyl chloride readily :dir,rded 111.2 3
The polymerizatioris were initiated by benzoyl peroxide a t 62' and carried out in bulk to yield relatively low melting, non-hygroscopic white powders readily soluble in acetone. chloroform and a number of other polar solvents. Inherent viscosity measurements (Table lI), together with the known tendency for allyl groups to terminate chains through degradative chain transfer, indicxt.ed that the polymers were of low molecular v-eight. ( 2 3 ) During t h e course of this work Spatz a n d Stone [ J . 01.2.C h e m . , 23, l:>9 (1058) ] described t h e thiourea-catalyzed isomerization of methyi hydrogen maleate t o the fumarate. Treatment u f t h e fumarate n i t h thionyl chloride yields 111 1 b '21: 1) T. T.niirlc. ClL / , I R r - c 6 8 , i o i 1 1 1 1 7 4 3
AND
GEORGE B. BUTLER
1-01. 81
CHCRO !i >O CHC NO
Fig. 1 -SJ
nthesis nf rnniirmc-r-
Residual alcohol and acid uiisaturatiori was determined by bromine titration essentially according to Siggia.22 The slightly modified procedure (see Experimental) was first checked by titration of the monomers. The analytical data obtained by titration of the polymers are summarized in Table IV. The approximate percentage cycliTABLE IV L"~XALYTICALDATA DERIVEDFROM
Polymer
Poly-Ia Poly-I b Poly-IC Poly-IIa Poly-11 b Poly-IIC
Residual alcohol bond,
7c
35 8 i 0 . 9 2 1 . 1 rt . 6 32.8 i 29.3 69.7 i 4 3 . 4 rt
*
.1 .4 .6 .2
BROMINETIIRATIONS
Residual acid bond, %
Approximate %. cyrlization
15.3 i 2 . 7 18.6 i. 2 . 1 2 9 . 2 =k 0 . 2 7.76 + .23 6 . 5 2 k .20 1 4 . 4 zk . 3
49 60 38 63 23 43
zation was assumed to be the difference between total residual unsaturation and 10()yc,for, if linear polymerization but no cyclization had occurred, the total residual unsaturation would be 100%. Any further loss of uiisaturatioii may be accounted for by three processes : (1) cross-lixiking, ( 2 ) branching and (3) cyclization. Since, from the physical properties of the polymers, cross-linking has not taken place aiid branching has occurred to, at most. a few per cent., i t follows that cyclization must account for essentially all of the "lost" unsaturation. The infrared spectra of the six soluble polymers (potassium bromide pellet) showed considerable absorption in the C-C stretchiiig regioii (16%-1655 cm-I) as well as a strong absorption band at 1720-1755 ciii,-'. Saturated esters (17&-1750 cm.-1j,23 ct,d-unsaturated esters (1717-1'73l cm.-1),25 6-valerolactone (174) cxri.-1)!6 x : ( I E caprolactone (1727 c m - ' all absorb i n this latter region. I n addition, poly-(methyl allyl iiialeatej (poly-Ia) and poly-(methyl 2-butenyl maleate) (poly-Ib) exhibited a band a t 17SO c m - I iiidicatixig the presence of a ybutyro1acto:ie ring systeiii. 2i llolecular models iiidicxte soinew1i'Lt less steric hindrance to attack of the allyl rndical :it C 2 of the maleate ester (X)than a t C,. ( 2 5 ) L. J. Bellamy. " T h e Infrared Spectra of Cumpiex \ I John Wiley and Sons, Inc., S e w York, K . Y . , 1954, p. 1.i.i ( 2 6 ) 11 K . I3all :ind R . Zbinden, THISJ O U R N A L . 8 0 , tA2h 2 i . Rrfcrenc es 1 i i . j c n - 1 for -,-l>iit>r
