Jan., 1951
'THERMAL
[CONTRIBUTIOS FROM
DECOMPOSITIONS O F 2-hfETAYL-1-INDANOL ISOMERS THE
59
XOYESCHEMICAL LABORATORY, UNIVERSITY OF ILLINOIS ]
Studies on the Mechanism of Chugaev and Acetate Thermal Decompositions. 111. cis and trans-2-Methyl-1-indanol BY ELLIOTTR. ALEXANDER AND ANTONMUDRAK'
In the second paper of this series? it was shown that, a t 98-loo', inethyl trans-2-methyl-1-tetralylxanthate (I) was readily decomposed to 2-methyl3,4-dihydronaphthalene (11) while the corresponding cis-isomer (111) was stable. Similarly, cis-2-methyl-1-tetralylacetate was found to be OCSICH~
H I
OH 1V
indanol meet these requirements. This paper describes their preparation and a study of their thermal decomposition.
Experimenta17 a-Bromoisobutyryl Bromide.-The preparation of 01I 11 bromoisobutyryl bromide has been rather poorly described. COS CHaSH (1) The following procedure, adapted from that of Berger,*was OCSCHr found to be satisfactory. A mixture of 244 g. (2.77 moles) of isobutyric acid and 25 g. (0.81 mole) of red phosphorus8 was placed in a 2-liter, 3-necked flask fitted with a reflux condenser, a dropping funnel and an efficient mechanical stirrer. Vigorous stirring was begun and 265 ml. (5.1 moles) of bromine was added a t such a rate that the vapors above the solution were I11 never red in color. When about half of the bromine had inore stable to pyrolysis a t 350-550' than trans- been added, the mixture was warmed to about 40" and the addition of bromine u'as continued. A total of about five 2-methyl-1-tetralyl acetate. Decomposition of the hours was required for the addition. When all of the cis-isomer, however, was found to lead to 2-methyl- bromine had been added, the mixture was heated on a naphthalene rather than 2-methyl-3,4-dihydro- steam-bath until the evolution of hydrogen bromide ceased (20 to 30 hours). The product was then decanted from the naphthalene. . viscous residue into a 500-ml. Claisen flask and distilled Although these experiments support the sugges- under reduced pressure. The fraction boiling at 60-75" tion that Chugaev3v4or acetate5 thermal decom- (23 mm.) was collected (484 g., 76sc) and used without positions proceed through the formation of a cyclic further purification in the next step. 2-Methyl-1 -indanone.-The procedure for the prepatransition state complex (for example, equation 3 ) , of 2-methyl-1-indanone was essentially that used by they are not ideal for two reasons. First, in order ration Kishner'o but modified by omitting the alcoholic potassium 1 I U hydroxide washing operation.11 Thus from 470 g. (3.56 l U moles) of anhydrous aluminum chloride, 950 ml. (10.7 moles) of thiophene-free benzene, and 375 g. (1.63 moles) of a-bromoisobutyryl bromide was obtained 95.1 g. (40y0) of 1.5555.12 2-methyl-l-indanone, b. p. 80-81' (2 mm.) ; n 2 0 ~ The melting points of the sernicarbazone,lo the p-nitrophenylhydrazone,l3 the brorno-de~ivative~~ and the oxirnelo all agreed with the values reported in the literature. trans-2-Methyl-l-indanol.-A mixture of 43.8 g. (0.3 to purify methyl cis-2-methyl- 1-tetralylxanthate, a mole) of 2-methyl-l-indanone, 15 g. of copper-chromium catalyst and 55 ml. of anhydrous ether was hydrovacuum distillation was necessary. This operation oxide genated at 130" and a pressure of 1800-2600 p.s.i. The has been reported to convert some xanthates into a theoretical amount of hydrogen was absorbed in 35-40 form having increased thermal stability.6 Secondly, minutes. The catalyst was removed by filtration and the in the pyrolysis of cis-2-methyl-l-tetralylacetate, solvent was evaporated in vacuo. The residue crystallized petroleum ether (42-50') to give a product melting at the reaction did not simply fail as it had with the from 56-88'. The yield in several runs varied from 44 to 63";. corresponding xanthate, but rather the reaction An analytical sample was obtained by further rrcrystallizaapparently took another course involving simul- tion from petroleum ether, m . p . 88-89'. . t m / . Calcd. for CloHI?O: C, 81.04; H , 8.16, I'ountl: taneous elimination and dehydrogenation. For an ideal experiment, therefore, the xanthates should C , 80.97; 1.1, X.32. The 3,s-dinitrobenzoate i i i c l t d at 12!1 l:3!l.53 (90 -130" be readily isolated froin the reaction mixture by ethcr) crystallization and they should be of such structure petroleum _ _ __
+
r-,
.
that eliniination could lead only to one product. I t has been found that the acetates and inethyl xanthates of cis(1V)- and tmns(T'~-~-nietllyl1(1) Colgate, Palmolive-Peet C o m p n y , Jersey City, N.J . (2) Alexander and Mudrak, THISJOURNAL, 71, 3194 (1950). (3) Hiickel. Tappe and Legutke, A m . , 648, 191 (1940). (4) Stevens and Richmgnd, THISJOURNAL, 68, 3132 (1941). ( 5 ) Hurd and Blunck, ibid., 60, 2421 (1938). (6) McAlpine, J . Chcm. SOL.,1114 (1931); W G 11932): Bulrner and bfann, ibid., 666 (1045).
(7) h l l nic1tin.g points and boiling pnints ai-7erd. Geol , ala, No. 8 ( 1 9 4 5 ) 1.5) P. J. Flory, A-. Rabjohn a n d hl. Shaffer, .I. P o l y t n e ~Sei, 4, 225. 43.5 (1949). (6) P. J. Flory, T m s J O ~ R K A L ,63, 3086,3091. 3096 (1911); 69, 30 (1947). (7) W. H. Stockmayer, in "High Polymers," edited hy S. H. Twiss. Reinhold Publishing Co , S e w I'ork. S Y , 19-13, p 61.
JOHN
r). FERRY
found by difference. The quantity of cross-linking agent employed is sufficiently small so the gel point is not reached. After preliminary trials with various systems, we chose the reaction between a 1 : l copolymer oi styrene and maleic anhydride as the initial linear molecule and a glycol or diamine as the crosslinking agent. This reaction is simple and would be expected to be quantitative. The only solvent found to be suitable was dioxane, which is inert with respect to both polymer and cross-linking agents, and has a dielectric constant sufficiently low so that any occasional hydrolyzed anhydride groups on the polymer remain un-ionized and the complications exhibited by polyelectrolytes with respect to osmotic pressure and viscosity in dilute solution are avoided. Experiments with crosslinking by hexamethylene glycol and benzidine are presented here. Other experiments with hexaniethylenediamine gave qualitatively similar results, but are not reported because in this case the reaction was so rapid that i t is doubtful whether a homogeneous mixture was achieved before the cross-linking process commenced. Materials and Methods The copolymer of styrene and maleic anhydride (SYHbl, Blend 1) was furnished by Union Carbide and Carbon Corporation.* It was dissolved in methyl ethyl ketone and fractionated roughly b y precipitation with benzene. The first two fractions, A and B, comprising 31 and 297,, rcspectively, of the original material, were used for crosslinking by benzidine and hexamethylene glycol, respectively. These two cuts did not differ greatly in numberaverage molecular weight, suggesting that the original material was relatively homogeneous. The fractions wcre dried by aspiration followed by heating iit vacuo at 65" for ti0 to 70 hours. Although the portions used for cross-linking experiments were not heated above 65' for fear of possiblc degradation, aliquots heated to 100" in w c u o showed a small additional weight loss, and conceutrations of solutioiii macle up by weight were expressed in terms of thi< Itilly dried material. Hexamethylene glycol (111. p . 38-41 -) was obtaiiicd from 15. I . du Pont de Semours and Co., and bcuzidiue (n1.p. 127~-128')from H. Rubin.8 The benzidine w a s kept under nitrogen. Solutions of the glycol and benzidine were prepared immediately before use. Dioxane was purified by refluxing with hydrochloric acid, neutralization, refluxing _- .. (8) We art. indebted t o Mr.
