May20, 1961 [CONTRIBUTION FROM THE
STEREOCHEMISTRY OF ALKYL-SUBSTITUTED ACETONITRILE ANIONS DEPARTMEKT OF CHEXISTRY OF THE UNIVERSITY OF 24, CALIF.]
2363
CALIFORSIA AT L O S ANGELES, L O S .&NGELES
Electrophilic Substitution at Saturated Carbon. XII. Stereochemical Capabilities of Alkyl-substituted Acetonitrile Anions1 BY DONALD J. CRAM'4ND
PAUL
HABERFIELD
RECEIVED OCTOBER4. 1960 The steric course of the base-catalyzed decarboxylation of ( +)-2-cyano-2-methyl-3-phenylpropanoicacid ( I ) , and the steric course and kinetics of the cleavage of (+)-3-cyano-2,3-dimethyl-4-phenpl-2-butanol (111) to ( j- or ( )-2-cyano-1phenylpropane (11) has been studied. Both rates and steric course were found to vary with solvent, and in some solvents with the nature of the cation of the base. Reaction conditions were found which gave product with greater than 8% net retention, and greater than 16% net inversion. In ethylene glycol, the steric course was invariably net inversion, and the stereospecificity was effected little if a t all by the nature of the cation of the base, but was dependent on the leaving group (carbon dioxide in system I and acetone in system 111). In cleavage of I11 in n-butyl alcohol, the steric direction of t h e reaction depended on the nature of the cation of the base. The rates of cleavage of 111 correlate with the hydrogen-bonding properties of the solvent, and x i t h changes in the character of the cation of the base.
+
-
The steric course and mechanism of electrophilic ester to optically pure (+)-3-cyano-2,3-dimethyl-4substitution a t saturated carbon has previously phenyl-2-butanol. Since determination of the stereospecificity of the reactions under study debeen studied in systems in which carbon-carbon bonds cleave heterolytically, and the carbanions pends on a knomiedge of the rotation of both optiproduced are captured by proton donors in the cally pure reactants and products, (+)-2-cyano-Imedium. The carbanion intermediates were stabi- phenylpropane was prepared with opiically pure lized by only a phenyl group in most of the systems 2-methyl-3-phenylpropancic acid6 as starring studied,? and by both cyano and phenyl groups in material. The substance was converted to its one of the systems e ~ a m i n e d . ~ The less stabilized amide which was dehydrated with phosphorus carbanion appeared to be consumed by fast pro- pentoxide to the desired nitrile. Three separate ton transfer reactions,2 but the more stabilized and conversions of the same acid to nitrile gave rotalonger lived carbanion had a somewhat different tions of cy% +34.3', +35.6" and $37.4' ( I 1 dm., stereochemical fate. neat). Use of phosphorus oxychloride as dehydrat~ (1 1 dm., neat), a value The present study deals with two systems which ing agent gave a Z 5+33.0' cleave to give an anion stabilized only by a cyano identical with that obtained by Kipping and group. The base-catalyzed decarboxylation4 of Hunter.6 These authors prepared nitrile from 2-cyano-2-methyl-3-phenylpropanoic acid (I) and optically pure acid by dehydration of amide with cleavage of 3-cyano-2,3-dimethyl-4-phenyl-2-butaphosphorus oxychloride. WThen hydrolyzed, the no1 (111) both lead to 2-cyano-1-phenylpropane nitrile gave acid which had a ZY0 higher rotation (11). The steric courses of these electrophilic than the starting acid. However, the authors consubstitution reactions have been investigated, clusion that no racemization occurred in any of and the results compose this paper. the stages of this cyclic series of conversions was negated by their crystallization of the amide, which might have led to removal of racemic Contaminant. Their results taken together with ours suggest that the value of QZ5D 137.4' ( I 1 dm., neat) is a t least close to maximum rotation. Unfortunately, no completely conclusive means f R-H were found by which the configurations of the two starting materials could be related to that of the product. However, the pattern of results of the decarboxylation and cleavage reactions of I and I11 (see Tables I and 11) in ethylene glycol is CN CH, similar to that observed for other systems whose .L I1 internal configurational relationships have been unequivocally established. I n this solvent or in Results the closely similar diethylene glycol, cleavage of Optically pure (+)-%-cyano-2-methyl-3-phenyl- two ketones7 two sac-alcohols,8seven t - a l c ~ h o l s ~ ~ ~ ~ ~ i)ropnoic acid ( ( + ) - I ) 5 was converted through its and one decarboxylation of an acid3 gave net in( I ) This work was supported b y a g r a n t from t h e Petroleum Reversion, irrespective of the nature of the metal or warch F u n d admiuistered b y t h e American Chemical Society. Gratequaternary nitrogen cation employed. In runs f u l acknowledgment is herehy made to donors of this f u n d . TTT
( 2 ) D. J . C r a m , A. Langemann, J . Allinger a n d K. R. Kopecky, J . . 4 1 n C h z m . Suc., 81, 5 i 1 0 11959); (b) D. J. C r a m , F. Hauck, K . R. Koprcky a n d W. D Nielson, ibid., 81, 5767 (1959); ( c ) D. J. C r a m ,
J . L Mateos, F. Hauck, A. L a n g e m a n n , K. R. Kopecky, W. D. Nielsen a n d J. Allinger, ibid., 81,5774 (1959); (d) D. J. C r a m a n d B. iiickborn, i b i d . , 83,2178 (1961); ( e ) D. J . C r a m , L. K. Gaston a n d H.Jhger. ibid., 83,2183 (1961). i.31 D. J. C r a m a n d P.Haberfield, i b i d . , 83, 2354 ( l g t i l ) , I 1) See ref. 3 f o r a s u m m a r y of previous literature. (31 J . Renyon a n d W. A . Ross, J . Chern. .Sot, 3407 (1951).
( 6 ) F. S. Kipping, a n d A . E. H u n t e r , i b i d . , 1005 (1903). ( 7 ) D. J. C r a m , K. R. Kopecky, F. Hauck a n d A. Langemann A m . Chepn. Soc., 81, 5751 (1959). ( 8 ) D. J. C r a m , A . Langemann a n d F. H a u c k , ibid., 81, 575U
(1959). (9) Of t h e large number of systems examined, only 2-methyl-3phenyl-2,3-butanediol cleaved in ethylene glycol t o give product with net relention. T h i s anomalous result is undoubtedly d u e t o t h e presence of an internal proton source in t h e carbanion intermediate (ref. 2e).
DONALD J. CRAMAND PAUL HABERFIELD
2364
KESULl S
OF
DECARBOXYLATION REACTIOXS OF
THE
TABLE I ANIONO F ( f )-2-CYAXO-2-XETHYL-3-PHENYLPROPANOIC T,
Bas--
r
Vol. 83
Time, hr.
Product b
%
ACID
( I )"
S e t steric course
Solvent Nature Concn., M "C. yld. pll--tlDc l iOCI12CIi20H SHI 0.025 135 110 72 +1.69" 5%i inv. 3 IX0CHzCH:OH KzCOz ,0125 125 47 69 $1.60 fiyGin\.. 3 CBFISOH XH3 + .10 135 288 53 +O. 53 1.55&i n v . 1 CoH50H (CH3)dNOH .10 125 114 33 - .l7 0.5y0 ret. 3 (CILhCO€I SH3 .10 125 47 38 - .20 0.5%,ret. Solutions 0.10 i l l in optically pure acid were used. b Optically pure 2-cyano-1-phenylpropane has a presumed rotation of 2%+37.4" ( I 1 dm., neat). 1 = 1 dm., neat. I:un
1
-
rCESL'LTS O F
TABLE I1 CLEAVAC:E I k , but k,' k , and h d > unchanged. After 47 hours a t 125", the rotation had 5 ~ ( 1 1 dm.). The solution was poured kd'. Since only low stereospecificities were ob- diminished to ( ~ 2 0.00" 100 mI. of water and extracted with two 200-ml. portions served in the cleavages of system 111, detailed into of pentane. The pentane phases were washed with three interpretation of the steric results is impossible. 100-ml. portions of water and dried. The pentane was evaporated and the residue distilled a t 2 mm. (pot temperature 120') to give 0.11 g. of 2-cyano-l-phenylpropane, Experimental nZ5D1.5093, aZ7~-O.2O0 (1 1 dm., neat). ( +)-2-Cyano-2-methyl-3-phenylpropanoicacid (I) was Runs 3 and 4 were carried out and the product isolated prepared and resolved as reported p r e ~ i o u s l y ;m.p. ~ 88-89', by a procedure similar t o that employed in decarboxylation [aI29~ +25.7" (c 2.4, CHCla); literature5 m.p. 87.5of 2-cyano-2-phenylbutanoic acid.3 Run 1 was similar in 88.5", [CY]"D +25.1° ( ~ 2 ~CHCI,). 4, procedure to that employed in runs 2 and 5 . The infrared ( )-3-Cyano-2,3-dimethyl-4-phenyl-2-butanol (111).spectrum of the 2-cyano-1-phenylpropane produced in runs T o 10.0 g. of optically pure (+)-2-cyano-2-methyl-31, 4 and 5 was compared with that of authentic material, phenylpropanoic acid dissolved in 20 ml. of ether was added and all samples were found to be identical in detail with one a solution of excess diazomethane in 250 ml of ether. The another. solution was allowed t o stand for 1 5 minutes at 25', the Representative Cleavage Reactions of ( +)-3-Cyano-2,3excess diazomethane was decomposed with formic acid, dimethyl-4-phenyl-2-butanol(111): Run 6.-Optically pure and the solution was washed with saturated sodium caralcohol, 0.203 g., was dissolved in 10 ml. of a solution of bonate solution, with 22V hydrochloric acid and finally with 0.10 M potassium hydroxide i n pure ethylene glycol. After water. The solution was dried, evaporated and the residual standing a t 83" for 39 hours, t h e solution was diluted with oil was distilled a t 2 mm. (pot temperature 160') to yield 100 ml. of water, and extracted with two 30-ml. portions 10.0 g. of (+)-methyl 2-cyano-2-methpl-3-phenylpropan- of pure pentane. The organic layers were washed with 100 oate, n25D1.5055, d '+26.72' ~ ( I 1 dm., neat). ml. of water and dried. The pentane layer was evaporated To 200 ml. of a n ethereal solution of a Grignard reagent and the residue absorbed on a 2 by 30 cm. column of neutral prepared from 5.0 g. of magnesium and 30 g. of iorlomethane activated alumina. The product was eluted with 150 nil. was added 10 g. of t h e above ester. The reaction mixture of 5 % ether-pentane, and the solvent was evaporated !vas stirred at 0 ' for 3U minutes, a saturated ammonium The residue was distilled a t 2 mm. a t a pot temperature of chloride solution was thetl added and the ether Dhase seaa- _-___
V-P
1
i
+
'
+
+.
>>
+
>>
>>
STABILITY IN PHENYL ~ A N I S YSERIES L
May 20, 1961
2367
120" to give 0.050 g. of 2-cyano-l-phenylpropane, NZ5D m.p. 108-109". A solution of 0.203 g. of optically pure 111 in 10 ml. of ethylene glycol showed no rotational change 1.5087, @D $5.32" ( 1 1 dm., neat). Further elution of when held a t 100" for 44 hours. The initial and final rothe column with pure ether yielded no additional material. Extraction of the aqueous layer (from the pentane extrac- tations were a Z 5 D+0.44' ( 1 1dm.). Typical Rate Measurements for Cleavage of( W C y a n o tions) with 200 ml. of ether and evaporation of the ether gave 0.050 g. of starting alcohol, m.p. 103-106", [ c Y ] * ~ D 2,3-dimethyl-4-phenyl-2-butanol (111): Run 14.-A solution of 0.203 g. of optically pure I11 in 10 ml. of a 0.10 M solu17.6' ( C 4.9, CHCI,). tion of lithium n-butoxide in n-butyl alcohol was placed in Run 10.-Optically pure alcohol 111, 0.203 g., was refluxed in a solution of 10 ml. of 0.10 -If lithium rnethoxide a 65' constant temperature bath; aliquots were taken a t in methanol for 159 hours. The cooled solution was di- intervals, and their rotations measured in a 1-din. tube. First-order rate constants were calculated and averaged. luted with 50 ml. of water and extracted with two 20-ml. portions of pentane. The pentane layer was washed with 'Time, hr. [O]''D ki, set.-' X 10j two 50-ml. portions of water, dried, evaporated, and the residue was distilled a t 2 mm. t o give 0.11 g. of 2-cyano-l0.0 +0.29' .. phenylpropane, ? z Z f i ~1.5089, [ C Y ] ~ ~-2.96" D ( I 1 dm., neat). 2.0 4- .26 1.7 Run 15.-Optically pure alcohol 111, 0.203 g., was heated 4.0 .l9 2.8 in 10 ml. of 0.10 114 tetramethylammonium hydroxide a t 65" for 105 minutes. The solution was acidified with glacial i.5 .13 2 8 acetic acid, and most of the solvent evaporated at 100" and 10.5 .08 20 mm. The residual oil was chromatographed on a 2 by AV, 2.4 20 cm. column of neutral activated alumina with 5% etherpentane as developer. The first 200 ml. of eluant was The rate constants, which are recorded in Table 11, are evaporated and the residue distilled a t 2 mm. (pot temperature 120') t o give 0.11 g. of 2-cyano-l-phenylpropane,ltZ5D clearly very approximate. Approximate Rates of Racemization of ( -)-2-Cyano-11.5089, a Z 5 D -2.30'. Other Runs.-Runs iand 8 were carried out by the pro- phenylpropane Under Conditions of Its Formation.--XI)cedure for run 6; runs 9, 11, 12, 16 and 17 by the procedure proximate rates of racemization were measured for ( -)-?for run 10; and runs 13 and 14 by the procedure for run 15. cyano-1-phenylpropane (11), under the exact conditions of The indices of refraction of the product were in all cases its formation in runs 6, 7, 9, 10, 12, 14, 15, 16 and 17, except ~ to n Z 51.5093. ~ Infrared spectra of the that an 0.10 M solution of I1 was substituted for the 0.10 between n Z 51.508i product of runs 8, 10, 16 and 17 were taken and found t o be M solution of I11 used in the cleavage reaction. -1liquot techniques were employed, five to eight points were taken, identical in detail with that of authentic material. Stability of ( +)-3-Cyano-2,3-dimethyl-4-phenyl-2-butanol and the reaction was carried through several half-lives. (111) in Solvent in the Absence of Base.-Optically pure Probable errors for all values for the first-order rate cou111 (0.101 g.) was dissolved in 5 ml. of methanol, CPDstants (Table 11) are less than 50$0.12 f0.33" / Z 1 dm.). The solution was heated a t 65" for 23 (12) 4 n extensive kinetic investigation of racemization r a t e s of I1 hours, after which the solution had C Y ~ +0.34' ~ D ( I 1 dm.). with a v a r i e t y of bases a n d solvents has been carried o u t , t h e results The solvent was evaporated and the residue recrystallized of which will be reported in a later study. from ether-pentane to give 0.097 g. of starting material,
+
+
+ + +
[CONTRIBUTION FROM
THE
DEPARTMENT OF CHEMISTRY OF COLUMBIA USIVERSITY,SEW YORK27, S . Y.]
Triarylcyclopropenium Ions.
Synthesis and Stability in the Phenyl p-Anisyl Series1
BY RONALD BRESLOW AND HAIWONCHANG RECEIVED NOVEMBER 2, 1960
A4novel synthesis is reported for triarylcyclopropenium ions, involving the reaction of an arylchlorocarbene with a diarylacetylene. Using this reaction, the triphenyl-, diphenyl-p-anisyl-, di-p-anisylphenyl- and tri-P-anisylcyclopropenium ions have been prepared, and their spectra and @KR+'s determined. The pK's are correlated with the predictions of molecular orbital theory.
Although the synthesis of the triphenylcyclopropenium ion was described some time ago,a'B the approach was not suited to the preparation of substantial amounts of materials, and more seriously the scheme, involving reaction of phenyldiazoacetonitrile with diphenylacetylene, and subsequent conversion of the resulting triphenylcyclopropenyl cyanide to the cation with boron trifiuoride, has proved unsuitable for extension to the misyl series, and thus does not promise to be a general method. We accordingly have investigated an alternative approach, and have found that phenylchlorocarbene, generated from benzal chloride and potassium t-butoxide, reacts with diphenylacetylene to form triphenylcyclopropenyl chloride. (1) T h i s work was supported b y a g r a n t from t h e National Science Foundation. (2) R. Breslow, J . A m . Chem. Soc., 19, 5318 (1857). (3) R. Breslow a n d C. Y u a n , ibid., 80, 6991 (1968). Through a typographical error, AD.E.ion for trlphenylcyclopropenyl cation WDS reported t o be 4.38 inatead of %a#, tha correct vslua.
Under the conditions of the experiment this is of course converted to triphenylcyclopropenyl tbutyl ether, and interestingly this is hydrolyzed to bis-triphenylcyclopropenyl ether during aqueous washing of the reaction product. Neither of these compounds need be isolated, however, for they are converted quantitatively to triphenylcyclopropenyl bromide (I)3by treatment with hydrogen bromide, and since this salt is insoluble in non-polar solvents i t can be isolated easily from the crude extracts. Accordingly, with the proper choice of reaction conditions largea mounts of diphenyacetylene can be converted quantitatively to triphenylcyclopropenyl bromide in a few hours. The reaction works also on p-anisylphenylacetylene and on di-p-anisylacetylene, or with p-anisal chloride instead of benzal chloride, and using the appropriate combinations it has been possible to (4) This method was first reported a t t h e Bostnb, M r a a
vf the Amuicsn Chamleal Bociety, April, 1868,
I
Meeting
