4566
11. 17.CURTIS,I € . 11.. Jom-sos, J R . ,
trans-K-Benzylamide, m.p. 141-142". Anal. Calcd. for CI~HZIOZN:C, 72.84; H , 8 . 5 6 ; N, 5.66. Found: C, 73.20; H , 8.69; N,5.84. cis-N-Benzylamide, m.p. 97-980, C , 72.3; H , 8.80; N, 5.75.
Acknowledgments.-Microanalyses
CONTRIBUTIOS FROM
were
h-OYES
per-
E. G. STEINER
WD
VOl. 77
formed by Dr. Ching S. Yeh and Mrs. Jean Fortney. The infrared spectra were obtained with the generous assistance of Dr. C. A. Ward, Mrs. LaVerne Walsh and Mr. P. A. Kinsey. LAFAYETTE, IXDIANA
CHEMICAL LABORATORY, UNIVERSITY OF ILLINOIS]
Stereoisomeric Vinyllithium Compounds. 111. Reactions with Aldehydes, Ketones and Methyl Iodide1 W. JOHNSON, B\ DAVIDY. CURTIN,HARRY
J R . , ~AND
EDWING. STEINER
RECEIVED MARCH19, 1955 The vinyllithium reagents prepared from cis- and trans-l-bromo-2-(p-chlorophenyl)-l,2-diphenylethylene have been shown to react with formaldehyde, benzophenone and methyl iodide to yield the corresponding primary alcohol, tertiary alcohol and methyl-substituted olefin, respectively, in good yield and with a high degree of over-all stereospecificity. Configurations assigned by relating the products of the formaldehyde and methyl iodide reactions to each other and to the known cis- and trans-3-(p-chlorophenyl)-2,3-diphenylacrylic acids show that configuration is retained in each case. Attempts to prepare tetraarylethylenes by treating triphenylvinyllithium with phenyldiazonium ion or diphenyliodonium ion were unsuccessful.
It was shown by Curtin and Harris3 that cisand trans- 1-bromo-2-(p-chlorophenyl) - 1,2-diphenylethylene (cis- and trans-I) could be converted to the corresponding cis- and frans-vinyllithium compounds with a high degree of stereospecificity. These vinyllithium derivatives were shown to undergo stereospecific reactions with methanol to give the corresponding cis- and trans-olefins and with carbon dioxide to give the corresponding cis- and trans-acids. It was the purpose of the present research to extend the study of the reactions of these isomeric lithium compounds. It was of interest, first, to make a brief survey of the reactions of vinyllithiuni compounds with a number of substances. For this purpose triphenylvinyllithium was employed. This reagent was treated with formaldehyde, benzophenone, benzaldehyde, benzoyl chloride, ethyl benzoate and methyl iodide to give the prodsicts listed in Table I in the yields shown. Attempts to carry out replacement of the lithium atom by an aryl group using either phenyldiazonium ion4 or diphenyliodonium ion5 were unsuccessful. illthough the configuration of the vinyl bromide cis-I had been established by the dipole moment study of Bergmann,6 he had concluded from the dipole moments that the trans isomer was not pure but contained as much as 50% of the cis isomer.6 Since this latter conclusion was not in agreement with the work of Curtin and Harris, samples of the vinyl bromides (cis- and trans-I) were re-examined by Dr. Max Rogers who found the dipole moments (1) T a k e n in p a r t f r o m t h e P h . D . Thesis submitted t o t h e University of Illinois b y Edwin C. Steiner. (2) National Science Foundation Fellow, 1953-1954. T h e authors are indebted t o E. I. d u P o n t de Nemours a n d Co., Inc., for a G r a n t in-aid which supported part of this work. (3) D . Y . Curtin a n d E. E. Harris, THISJOURNAL, 73, 2716, 4519 (1951). (4) Compare R . Oda and K. N a k a n o , Rcpts. Inst. Chcm. Research Kyoto Univcr., 19, 91 (1949) [C. A , , 4 5 , 7543 ( 1 9 5 l ) l . ( 5 ) F. M . Beringer, M. Drexler, E. M. Gindler and C. C. Lumpkin, THISJOURNAL, 75, 2708 (1953). (6) E. Bergmann. J. Chem. Soc., 402 (1936).
TABLE I CsHs c""5)c=C(~,,5 PRODUCTS
FROM
REACTIONO F TRIPHENYLVINYLLITHIUM Product
Reagent
R=
L i t . ref. to product
7
Yield,
%
-
77 GI
8 10 10 11
65l
I
..
52 71 40
0 0
to be those expected of the pure cis and pure trans isomers, respectively. l 2 When the cis-vinyl bromide (cis-I) was converted by means of butyllithium to the corresponding lithium derivative and the latter treated with formaldehyde, a single alcohol, here assigned the structure cis-3- (p-chlorophenyl)- 2,3-diphenylallyl alcohol (cis-II), was obtained. Similar treatment of trans-I yielded the secondary isomeric alcohol, trans-11. The configurational assignments of the alcohols, cis- and trans-11, were based on the fact that the alcohol obtained from the methyl ester (cis-111) of 3-(~-chlorophenyl)-2,3-diphenylacrylicacid3 by treatment with lithium aluminum hydride was identical with cis-11, while the alcohol from the ester (trans-111) was identical with trans-11. When the lithium compound prepared from the bromide (cis-I) was treated with methyl iodide a single olefin, cis-l-p-chlorophenyl-l,2-diphenylpro( 7 ) C. F. Koelsch, THISJOURNAL,6 4 , 2046 (1932). (8) C. F. Koelsch, i b i d . , 54, 3384 (1932). (9) Isolated a s t h e ring-closed product, 1,2,3-triphenylindene. (10) E. P. Kohler a n d E. M. Nygaard, THIS J O U R N A L , 6.8, 4128 (1930). (11) W. Schlenk a n d E. Bergmann, Ann., 4 6 3 , 45 (1928). (12) We are indebted to D r . Rogers for communicating these results t o US.
STEREOISOMERIC VINYLLITHIUM COMPOUNDS
Sept. 5, 1955
pene (cis-IV) was obtained. The steric configuration of this substance was established by its preparation from the alcohol (cis-11) by successive treatments with 9-toluenesulfonyl chloride, thiophenolate ion and, finally, Raney nickel. The transhydrocarbon (trans-IV) was prepared both from the vinyl bromide (trans-I) and from the alcohol (trans-11) by sequences of reactions similar to those employed with the cis isomers described above. These reactions are summarized in Chart I below. The product obtained from each of the isomeric vinyllithium compounds and formaldehyde showed no evidence of contamination by the alcohol obtained from the second stereoisomer. It appears then that these reactions of vinyllithium compounds with carbonyl compounds and also with methyl iodide are highly stereospecific as are the reactions with carbon dioxide and methanol previously examined by Curtin and Harris3
4567
exchange reaction is appreciably more stereospecific than is the reaction of vinyl halides with metallic lithium. I n order to determine more precisely the effect of temperature upon the stability of the vinyllithium compounds the two bromides, cis- and trans-I, were treated with butyllithium in ether a t temperatures of 25, 0 or -35' and, after 30 minutes, with methanol. Examination of the infrared spectra of the product olefins indicated that neither olefin obtained from the reactions a t -35' was detectably contaminated with its geometric isomer. On the other hand, the products of the reactions carried out a t 0 and a t 25' showed successively more contamination by the stereoisomer formed in minor amount. The infrared spectra also indicated that a t the two higher temperatures there was contamination by a substance containing aliphatic hydrogen, presumably 1-(p-chlorophenyl)-1,2-diphenyl1-hexene. I t seems likely, then, that CHARTI a t temperatures of -35' and lower, COOCHI P-X-C&a cis isomers X = C1 the exchange reactions proceed with Y=H an extremely high degree of stereop-Y-Ce.H, specificity and the isomeric vinyltrans isomers X = H Y = c1 'I1 LiAlHn lithium compounds are configurationC-x-c~Ha c H 2 0 H ally stable. However, a t 0' and above there is apparently some interCHzO p-Y-CeH4 )c'cc'C-ield of red platelets of 2,3-diphenylindenone, m.p. 149 130". Similar results were obtained with thionyl chlrxide in 2,g-lutidine and with oxalyl chloride. IJRBANA,ILLISOIS
