TODW. CAMPBELL AND WILLIAMG. YOUNG
296
acid, m. p. 255-$?57" (after sublimation in vacuo) (reported m. p. 255-256 , subl.);ls the yield was 10.7 g. (71%). The acetyl derivative melted at 197.5-198.5" (reported 198.5) The ether filtrate was evaporated yielding 12 g. (66%) of benzophenone. The semicarbazone melted at 103-1M0 (reported m. p. 164').21
Summary 1. Benzohydrylamine and certain of its deriva(19) Zelinsky and StadnikofT, Ber., 18, 1726 (1906). (20) Heilbron, "Dictionary of Organic Compounds," Vol. I, Oxford University Press, New York, N.Y., 1043. p. 112 (21) See ref. 6,p. 2 f X
f CONTRIBUTION FROM
THE
Vol. T1
tives react with potassium amide in liquid ammonia to form red colored solutions indicating the formation of the corresponding carbanions. 2. Carbonation of the potassium salt of benzohydrylamine has given the corresponding aamino acid. 3. Carbonation of the carbanion of benzalbenzohydryldmine gives the N-benzohydrylidene Q phenylglycine instead of N-benzal a, a-diphenylglvcine. UURHAM,NORTH CAROLINA RECEIVEDAUGUST2, 1948
DEPARTMENT OF CHEMISTRY,
UNIVERSITY OF CALIPORNIA, LOS
ANCELES]
Allylic Rearrangements. XXV. Reaction of the Sodium Derivative of Allylbenzene with a-Haloacetic Acids BY T O D
w.CAMPBELLAXD ~VILLIAM G. YOUNG
The sodium derivative of allylbenzene, which has been the subject of recent investigations' undergoes reactions which can best be expressedlc as nucleophilic attacks by the resonating anion
as benzylamine saltse2 The composition of the original mixtures was estimated to about f 2% from the characteristic absorption spectrums for the conjugated system. From the results listed in Table IV, it will be seen that the polarizing effect of the carboxylate ion adjacent to the carbon atom being attacked is 1 ZI probably more important than simple steric Since most reactions studied in which steric effects hindrance, since the unsubstituted a-haloacetate are not serious lead to mixtures of products ions are certainly less hindered than a compound derived from both of the principal resonance such as a-niethallyl chloride.'* It will be noted forms of the allylbenzene anion, i t seemed odd that the addition of methyl groups on the athat the reaction with simple halides1* gave, at haloacetate ions leads to an increasing amount of least principally, products derived from form I1 the product arising from the primary form of the carbanion, as might be predicted from steric efabove. In an attempt to clarify the issue, the reaction fects,lc It is also interesting to note that in this of sodium allylbenzene with a series of a-halo- series the nature of the halogen being displaced acetate ions was examined. This particular definitely affects the composition of the mixture series was chosen since it should be possible to obtained from the resonating carbanion, the effect determine the effect of (a) steric hindrance, (b) becoming more apparent as the number of methyl polarization and (c) the nature of the ion being groups is increased. This effect of the halogen displaced, on the manner in which the resonating being displaced on the products has not been allylbenzene carbanion reacts in displacement noted before in any of the reactions of sodium reactions. When the chosen reactants were allylbenzene. mixed, a rapid reaction ensued. From the reacExperimental P.vt action product was isolated a mixture of carboxylic All melting points are uncorrected. acids, whose formation can be represented by the Reaction of the Sodium Derivatives of Allylbenzene equation with a-Haloacetic Acids-General Method.-The sodium
C&LCH-CH=CHy
I
+ CsHi-CH~CH--CH&!R&O~-
CReCOe-
All possible combinations of X and R were examined. The individual acids produced were separated and carefully characterized, principally (1) (a) Levy and Cope. THxa JOURNAL, 66, 1684 (1944); (b) Campbell and Young, ibid.. BO. 1388 (1947); (c) ibid., 69, 8066 (19471 ~
derivative was prepared in 300 ml. of liquid ammonia in the usual manner1 from 0.1 M of allylbenzene. The solution was forced over as rapidly as possible (by the vapor pressure of the liquid ammonia solvent) into a twoliter three-necked flask containing 0.05 mole of the appropriate a-haloacetic acid dissolved in 20 ml. of ether (anhydrous). The flask was equipped with a stirring device and a wide outlet tube to provide for escape of the ammonia vapors. The characteristic red color of the sodium derivative was discharged rapidly. After evaporation of the solvent, water was added, and the carboxylic acids were separated from the non-acidic material. Table I is o rGsumC of results obtained by this method. Data are Civerages of several experiments. (2) Buehler, Carson and Wda, ibid.. 57, 2181 (1936). ( 8 ) Csmpbell Liaden. Godnhalk and Young.,*hl.. 69, 880 (1847)
Jan., 1949
SODIUM DERIVATIVES OF ALLYLBENZENE WITH a-HALOACETIC ACIDS
297
drogen at 750 mm. and 23" (theory 2750) over platinum TABLE I oxide in ethyl acetate. The acid melted at 58" after rePRODUCTION OF ACIDSBY COUPLING OF AN WHALOACBTICcrystallization from warm (35") methanol, to which water ACIDWITH SODIUM ALLYLBENZENE had been added t o the cloud point. Its benzylamine salt melted a t 96.5". Neither the melting point of the acid, nor the melting point of the benzylamine salt, was depressed by addition of the corresponding product obtained by hydrogenation of 6-vinylhydrocinnamic acid described above. CICH&OnH 176 177 43 647 440 Separation of the Reaction Products from a-BromoCICH(CH8)COzH 190 189 54 784 545 propionic Acid and Sodium Ally1benzene.-The methods CIC(CH,)&02H 204 203 40 1100 742 described above could not be used to separate this mixture BrCHpC02H 176 176 74 650 450 of acids. Benzylamine was added t o the mixture, and the BrCH(CH8)COZH 190 188 67 927 663 mixture of solid benzylamine salts was fractionally crystallized. It was possible to isolate the spectroscopically BrC(CH3)ZCOzH 204 204 72 1270 919 Dure benzvlamine salt of the secondarv acid. a-methvla Yield based on a-haloacetic acid. *Maximum deviation in values of maximum averaged was 5%, mean &vinylhydrocinnamic acid, m. p. 1 4 6 5O , by fractional crystallization from ethyl acetate. Anal. Calcd. for deviation about 2%. ClpHnaN02:C, 76.73; H. 7.80. Found: C, 76.35; H , Identification of Acidic Components of the Reaction 7.87. The benzylamine salt of the primary acid could Mixture from Bromoacetic Acid and Sodium Allylben- not be obtained optically pure, by direct recrystallization. The residual impure benzylamine salt, which was rich zene.-The mixture of acids, obtained in 7470 yield, was partially separated by filtration of the chilled product. in the primary acid, was dissolved in glacial acetic acid. The solid separated in this manner from the residual oil An equivalent amount of bromine in acetic acid was added, was recrystallized several times from ligroin. It formed followed by dilution with water, to give an oily dibromoshining white scales, m. p. go", which was identified as 4- acid and a solid dibromo-acid. The solid was isolated phenyl-3-pentenoic acid by its melting point and the melt- and purified by recrystallization from aqueous ethanol ; ing point (162') of its d i b r ~ m i d e . ~Its equivalent weight long, lustrous needles, m. p. 133'. This solid dibromowas 176 (calculated for C11HleOl, 176), and it yielded ben- acid was refluxed with zinc dust in isopropyl alcohol. The zoic and succinic acids on ozonization. The benzylamine resulting oily acid was isolated and converted to a benzylamine salt. which was recrvstallized from Skellvsolve C. salt*of 4-phenyl-3-pentenoic acid melted a t 108". Anal. Calcd. for ClsHtlNOe: C, 76.29; H , 7.47. giving long needles, m. p. i l 9 " . Anal. Calcd: for CI& HzrNOz: C, 76.73; H, 7.80. Found: C, 76.58; H, 7.92. Found: C, 76.40; H, 7.61. The absorDtion s ~ e c t r u mof~the salt indicated that it was The oily mixture of acids remaining after filtration was derived from the primary acid. dissolved in four times its volume of benzene, and an equivSeparation of Acids Derived from the Sodium Derivaalent amount of potassium hydroxide was added as a tive of Allylbenzene and a-Bromoisobutyric Acid.-The 25y0 solution in methanol. The solid which precipitated primary acid was obtained by filtration of the partially was filtered off and found to be nearly pure potassium 4- solidified mixture. It was obtained optically pure on rephenyl-3-butenoate. This separation was purely fortui- crystallizing from methanol-water. I t formed massive tous and was not expected. The residual solution was prisms, m. p. 104". Its benzylamine salt melted at 145.5". treated with dilute sulfuric acid, dried, and the solvent Anal. Calcd. for CZ0Hz6N02:C, 77.13; H, 8.09. .Found: removed. The oily acid remaining gave a benzylamine C, 77.23; H. 8.28. The secondary form of the acid could salt melting a t 97-98'. not be obtained spectroscopically pure. Anal. Calcd. for CI~HZINOZ; C, 76.29; H, 7.47. Absorption Spectra of the Pure Acids Derived from the Found: C, 76.26; H, 7.55. Primary Form of the Allylbenzene Carbanion.-The abThe oily acid was converted by hydrogenation t o a solid sorption spectra of these acids and their benzylamine salts acid, m. p. 58". The benzylamine salt of the latter melted were taken on the Beckmann ultraviolet spectrophotometer and are tabulated in Table 11. a t 96-97". TABLE I1 293 Max.
Acid
Absorption spectra of acid 284 Max.
79 1
CsHsCH=CHCH&HL!OiH
CBH~CH=CH--CH~CH(CH~)CO~H(850)' a
CsHsCH=CHCH2C\CH&COzH 978 Estimated values from benzylamine salt.
1120 (1250)" 1380
Anal. Calcd. for ClsH23NOa: C, 75.75; H, 8.12. Found: C, 75.94; H, 8.21. The hydrogenated acid w2s shown to be p-phenylvaleric acid (see below), thus the liquid acid WRS 8-vinylhydrocinnamic acid. It could not be 2-phenyl-2-pentenoic acid, since the absorption spectrum of the liquid acid indicated that the double bond was out of conjugation with the benzene ring.a Synthesis of @-Phenylvaleric Acid.-p-Ethylcinnamic acid, m. p. 94", was synthesized through the Reformatsky reaction from propiophenone and ethgl bromoacetate.' Its benzylamine salt melted a t 139-140 Anal. Calcd. for CIIHPINO?: C, 76.29; H, 7.47. Found: C, 76.34; H, 7.43. This acid (19.70 8.) absorbed 2740 ml. of hy-
.
(4) Fichter and Bauer, Be?., 81, 2002 (1898). (5) Shriner, "The Heformatsky Reaction" in "Organic Reactions." Vol I John Wilcp & Sons,Iac.. New York. N Y.. 1942. p 1
250 Max.
293 Max.
18,470 (18,800)" 19,030
820 850 971
Ahsorption spectra of benzylamine salt 284 250 Max. Max
1260 1385 1498
17,020 18,500 19,980
Composition of Original Reaction Mixture by Spectral Analysis.-From the extinction coefficients quoted above, the percentage composition was calculated from the formula
s mixture
(% of Compound I)
=
%"la1
X 100
TABLE I11 a-Haloacetic acid
XCHiCOiH XCH(CHdCozH XC(CH&CO?H
Composition of Composition of mixture mixture x = CI,per cent x = Br. per cent from anion from anion primary secondary primary secondary
57.2 63. 77.2
42.8 36. 22.8
57.0 76. 93.3
42.4 24. 0.7
ALBERTL. HENNEAND WILLIAM G. FINNEGAN
298
where (1) is derived from the Primary form of the a W b a zene carbanion. It was assumed that the acid derived from the secondary form of the allylbenzene carbanion would have extinction coeffieients of approximately zero at 294 mB, 20 at 283 m p and &
