COMMUNICATIONS TO THE EDITOR
870
Vol. 84
CH-CsHn ‘I;=”
The reaction of l a with acetic anhydride in the presence of a catalytic amount of p-toluenesulfonic acid gave l-acetyl-3,5-diphenyl-2-pyrazoline (III), m.p. 125-125.5’, identical (infrared spectra and mixed melting point) with an authentic sample prepared by the reaction of acetic anhydride with II.5a.b Compound I a is the first stable 1pyrazoline6 having an e-hydrogen capable of
pyrazoline (11) a t 200’ gave both cis- and trans1,2-diphenylcyclopropanes(about 50 : 50 mixture) as was shown by Curtin, et a1.,8 and checked by us. Acknowledgment.--n‘e thank the Susan Greenwall Foundation of the National Starch Company and Grant NSF-G-17448 from the National Science Foundation for their generous support of this work. DEPARTMENT OF CHEMISTRY CHARLES G. OVERBERGER POLYTECHNIC INSTITUTE UF BROOKLYS 1, XETTYORK JEAN-PIERRE ANSELME BROOKLYS RECEIVED JASUARY 24. 1961
Ia
INHERENTLY DISSYMETRIC CHROMOPHORES. OPTICAL ROTATORY DISPERSION OF or,@-UNSATURATED KETONES AND CONFORMATIONAL ANALYSIS OF CYCLOHEXENONES‘J
COCH3 111
Sir: I
H
I1
isomerizing to a conjugated system. This type of 1-pyrazoline usually rearranges spontaneously to a 2-pyrazoline under conditions of their synthesis The isomerization of Ia to I1 in dilute ethanolic hydrochloric acid was followed spectrophotometrically by the appearance of the hydrazone peak a t 291 mp. Thermal decomposition of Ia, presumably a mixture of the D,L-trans-isomers a t 7 0 ° , resulted in only one product, trans 1,2-diphenylcyclopropane7 (D,L pair), n * 5 ~1.5951 (%*OD 15995, prepared by base catalyzed decomposition of the corresponding 2-pyrazoline8) in contrast t o the nonselectivity exhibited by the seven- and eightmembered ring compounds (I, n = 3, 4). Photochemical decomposition in hexane solutioii a t 25’ also gave this same selectivity One explanation may be that there is much less time for the radicals to become relatively free in the formation of the cyclopropane than in the case of formation of the cyclopentane or the cyclohexane This might be clue to the closer proximity of the two radicals in the five-membered ring case l y e can safely discount the possibility that the five-membered ring azo compound is the czs isomer and that the formation of the trans cyclopropane represents the formation of the most stable isomer since the base catalyzed decomposition of 3,5-diphenyl-2( 5 ) (a) S G Beech, J H Turnbull and W Wilson, J Chmn Soc , 4684 (1952) (b) M Hamada, Bolyu Kagaku, 21, 22 (1956) (6) S m i t h and coworkers J Ain Chenz S o c , 66, 159 (19431, cldirned t o habe isolated some 1 pyrazolines but these claims have not been substantiated by any d a t a ( 7 ) T h e n m r spectrum clearly shows t h a t t h e lraizs i m m e r is t h e u n l y product formed and was idenhcal with t h a t reported in reference 8 (8) D Y Curtin H Gruen, Y G Hendrickson and H E Knipmeyer, J A m Lhem Soc , 83, 4838 (1961)
Derivation of the octant rule3 for the 290 mp n-a* transition of saturated ketones rests, inter alia, on the local symmetry of the carbonyl chromophore which exhibits two orthogonal reflection planes. In a,P-unsaturated ketones, one or both of these planes is lost and hence the octant rule is in general no longer applicable. This reduction in local symmetry can be turned to advantage if one shifts attention from the n - x * to the shorter wave length H - T * transition of the C=C--C=O grouping, which, if nonplanar, may be regarded as an inherently dissymmetric chromop h ~ r e . ~Because ,~ of the similarity in the asystems of C-C-C=O and C-C-C-C, the helicity rule for cisoid dienes6 is expected to apply also to a$-unsaturated ketones. Calculations utilizing simple Hiickel type wave functions bear out this ,point: cisoid conformation A and transoid conformation B are to be associated with a positive Cotton effect centered a t the wave length corresponding to the K-band, i . c . , the lowest x - x * transition near 240-860 nip, while a negative Cotton effect would be exhibited by their mirrorimage representations. For sufficiently large angles of skew, the rotational strengths of the a - x * transitions of the above discussed inherently dissymmetric chromophore will be a11 order of magni(1) “Optical Rotatory Ilispersilln Studies.” LXXII. For paper LXXI. see C. S. Barnes and C. lljerassi. J . A m . Chein. Soc., 84, in press (1962). (2) Financial support by t h e Alfred P. Sloan Foundation (K. M . ) and by the National Science Foundation (grants ?io. Gl5746 a n d G-19905) is gratefully acknowledged. (3) W. Moffitt, R . B. Woodward, A . Moscowitz, W. Klyne and C . Djerassi, J . A m Chem. Soc., 83,4013 (1961). (4) A . Moscowitz, T e f v a h e d v o z , 13, 48 (1961). ( 5 ) A. hfoscowitz, K . Uislow, 11, A. W. Glass and C. Djerassi, J . d i n . Chem Soc., in press; K. hfislow and J. G. Berger, ibid., in press; K. hlislow, M . A. W.Glass, R . E. O’Brien, P. Rutkin, D. H . Steinberg, J. Weiss and C . Djerassi, ; b i d , in press. (6) A. hloscowitz, E. Charney, t-.Weiss and H . Ziffer, J . A m . Chenz. Soc., 83, 4661 (1961).
March 5 , 1962
COMMUNICATIONS TO THE EDITOR
tude greater4s5,6than the rotational strengths of the n-ir* transitions, which, in previous considerations involving saturated ketones, were most conveniently associated with asymmetrically perturbed symmetric chromophores.
Y 0
A
871
possible t o assign a chirality to the chromophore and hence a conformation to the cyclohexenone ring by taking into account the sign of the strong background dispersion which is in virtuaIly all cases superimposed on the 330 mp Cotton effect measured earlier.' In such instances, the sign and magnitude of the background dispersion may be correlated with geometries A, B or the respective mirror images. The conformational conelusions which may be reached by this approach are somewhat complicated by the fact that a t times the less populous conformer may conceivably
B
Utilizing a spectropolarimeter built by E. B. which employs a Faraday effect-type analyzer, we now have been able to measure (cell path lengths 1-10 mm., concentrations, c, 0.03-0.002 in methrn I anol) the Cotton effects in the region of the K-band (240-260 mp, E 8000-20,000) of a number of a,@unsaturated ketones, the long-wave length n-p* Cotton effects of which have been recorded prev i o ~ s l y . ~The cisoid a,p-unsaturated ketones jervine (I) and 4a-methyl-A8(14)-cholesten-3~-ol-7one acetate (11) show strong negative Cotton effects (Fig. 1) centered near 250 mp, while that of A4-cholesten-4-one (111) occurred a t slightly lower wave length. Similarly, the transoid ketones 4methyl - A4 - cholesten - 3 - one (IV), 4 - bromo(VI) testosterone (V) and B-nor-A4-cholesten-3-one exhibit high amplitude Cotton effects (Fig. I) in the 240-260 mp region. In the above examples, the sign of the Cotton effect reflects the absolute conformation of the C=C -C=O grouping which makes the dominant contribution to the observed optical activity; a positive sign indicates conformation A (cisoid) or B (transoid), and a negative sign the appropriate enantiomers. Since Dreiding models8 of many a,@-unsaturatedsteroids or terpenoids permit construction of two (or more) conformations differing in the chirality of the C=C-C=O grouping, optical rotatory dispersion-through the sign of the Cotton efect of the K-band-may now often permit for thefifirst time a choice between such conformational Fig. 1.-Optical rotatory dispersion curves (methanol) of representations. When Dreiding models8 point definitely to a given chirality-viz., the negative steroidal a,p-unsaturated ketones I-VI11 in the region 200helicity of 11-the sign of the relevant Cotton effect 300 mp, (Fig. 1) proves to be in agreement with prediction. The C- 1 isomeric 1-methyl-19-norprogesterones possess a much greater amplitude and thus con(VII, VIII) are particularly interesting since their trol the sign of the observed Cotton effect. NeverC=C-(2-0 groupings have opposite chiralities, as theless, the present rotatory dispersion results can demonstrated by the positive (VIII) and negative be taken to indicate a significant population of (VII) K-.band Cotton effects, respectively. This boat-like form for ring A in P-cholesten-+one reversal is the result7j9of the interaction between (111), and for ring B in A4-cholesten-6-one. The the equatorial I@-methylgroup (VII) and the l l a - present method also indicates which conformation hydrogen atom, which forces ring A into an alter- is primarily responsible for the rotatory dispersion nate conformation, thus reversing the sense of twist curves of B-nor-A4-cholesten-3-one (VI) (Fig. 1) and 8,19-epoxy-3-oxo-A4-etienic acid,1° which alof the C=C-C=O functionality. ready exhibit antipodal curve^^^^^ in the 300-400 When the K-band Cotton effect is not experimentally realized in its entirety, i t may still be mp region as compared to those7 of the usual A43-ketones, this inversion in sign again being due (7) C. Djerassi, R. Riniker and B. Riniker, J. A m . Chem. Soc., 78, to the oppositive signs of the K-band Cotton effect 6362, 6377 (1956). which makes itself felt in the background rotation (8) A. Dreidin'g, Helel. Chim.Acta, 4S, 1339 (lW9). (9) C. Djerassi, E. Lund and A. A . Akhrem, J . Am. Chem. Sot., 84, April (1962).
(10) K. (1961).
Otto and
M
Ehrenstein, J
Ovg
C h e m , 26, 2871
872
COMMUNICATIONS TO THE EDITOR
Vol. 84
above 300 mp superimposed on the n-n* Cotton effect.
50 (8%) cation exchange column and elution with 1 12.I perchloric acid. I t consisted mainly of CrC12+ and a smaller amount of Cr(Hz0)63+. (11) O n leave from the Department of Chemistry, University of The same reaction was carried out in the absence of Minnesota, Minneapolis 14, Minnesota. chloride ion (which was replaced by the pure solDEPARTMENT OF CHEMISTRY CARLDJERASSI STANFORD UNIVERSITY RUTHRECORDS vent) and was found to proceed quite slowly a t STANFORD, CALIFORNIA E. BUNNENBERG- 50°, thus demonstrating the marked catalytic activity of C1-. The only chromic species proDEPARTMENT OF CHEMISTRY XEWYORKUNIVERSITY KURTMISLOW duced in this experiment was the hexaaquochromic NEWYORK53, NEW YORK ion, as expected. OF PHYSICAL CHEMISTRY DEPARTMENT The assumption that FeCIZ+is not formed prior UNIVERSITY OF COPENHAGEN ALBERTMOSCOIVITZ~~ to the oxidation reaction was confirmed by another COPENHAGEN, DENMARK blank experiment in which Cr2+ was replaced RECEIVED JANUARY 9, 1962 by the pure solvent into which the ferric and chloride solutions were injected a t - 50' as above. After 60 seconds the yellow color of F e c i 2 + was OXIDATION OF CHROMIUM(I1) BY IRON(II1) IN THE not yet observed. A very faint color was first PRESENCE OF CHLORIDE ION noted after three minutes, and even after one Sir: hour equilibrium concentration was not attained. The oxidation of Cr2+ by Fe3+ in aqueous per- These results are in accord with the kinetic data chloric acid produces Fez+ and Cr(H20)63+,but given by Connick and Coppel5 for this reaction a t in the presence of chloride ion the complex Cr- room temperature. Using their values of the (Hz0)6C12+is produced along with Cr(H20)e3+inpro- rate constants and activation enthalpies, we found portions depending on the concentrations of C1- and that a t the initial ferric and chloride concentrations H+.' This reaction may proceed via the chloro used by us (0.075 molar) less than 0.47'( of the complexofFe(II1) FeC12++ C r 2 + + F e 2 + + CrClZ+ total Fe3f was converted to FeC12+ a t -50' in an inner sphere activated complex mechanism, after GO seconds. The results of these experiments show that similar to that observed in the oxidation of Cr2+by Co +(NH3)6X. ,2 Alternatively, Fe3+ may enter the the chloride-catalyzed oxidation of Cr2+ by Fe3+ activated complex without prior substitutionby C1-, does not proceed via the FeClz+complex a t -X', in a mechanism similar to that taking place in the but they do not rule out this path a t room temperaoxidation of Cr2+ by C O ( N H ~ )in ~ ~the + presence ture. of Cl-.3 It is difficult to distinguish between the (5) R. E. Connick and C. P. Coppel, J. A m . Chern. SOC.,81, G389 two mechanisms a t room temperature because the (1959). equilibrium Fe3+ C1FeC12+ is attained DEPARTMENT OF ISORGANIC A N D ANALYTICAL CHEMISTRY OF JERUSALEM MICHAEL ARDON too rapidly. At -50' the formation of the Fe- THEHEBREWUXIVERSITY JACOB LEVITAX ISRAEL ClZ+ complex proceeds quite slowly, therefore JERUSALEM, HERBERT JOSESLABORATORY this temperature was chosen to distinguish between GEORGE DEPARTMENT OF CHEMISTRY the two alternative paths. THEUNIVERSITY OF CHICAGO HENRYT A ~ B E ILL.,U. S. A . The solvent used in this investigation was a CHICAGO, eutectic aqueous solution of perchloric acid (5.27 RECEIVED DECEMBER 8, 1961 M ) having a melting point of -59.7°.4 A Pyrex reaction vessel consisting of three compartments ABSORPTION INTENSITIES AND ELECTRONIC was used : two side compartments were connected STRUCTURES OF TETRAHEDRAL COBALT(I1) through a common T-shaped mixing chamber to COMPLEXES the main compartment. An inert atmosphere Szr: was maintained by a slow stream of CO2 and the We wish to report soirie observations which see111 whole vessel was immersed in an ethanol bath, A chromous solution was to us to be both novel and important iii respect to maintained a t -50'. placed in the main compartment, a ferric per- the problem of the intensities of electronic absorpchlorate solution in one of the side compartments tion bands in transition nietal complexes. Using the ligaiid I, the enolate anion uf diand a hydrochloric acid solution in the other. By applying a pressure of COz the ferric and the pivaloylmethane, hereafter abbreviated 131'11, chloride solutions were driven through the common we have prepared the complexes Co(DP11)2 mixing chamber into the chromous solution, which 0 0they reached within 0.1 sec. after mixing. The 11 H I concentrations of the reactants after mixing were: ( CHI)JC-C-C=C-C( CHJ)J I [Fe3+]0.075 M , [Cl-] 0.075 Ad, [Cr2+]0.031 A I . The oxidation reaction was "instantaneous" and Zn(DPM)2. It has been shown previously and produced a green solution. After warming that in Ni(DPM)2 the size of the t-butyl groups to room temperature, this solution was separated prevents the trimerization which occurs in Ni(I1) chromatographically by absorption on a Dowex complexes with less hindered P-diketone enolates so that Ni(DPM)2 is a planar, diamagnetic mono(1) H. Taube and H. Myers, J. A m . Chem. Soc., 78,2103 (1954). mer.' C O ( D P M ) ~has three unpaired electrons
+
(2) H. Taube, { b i d . ,77, 4481 (1955). (3) H. Taube, Chem. SOC.Spec. Publ., 13, 57 (1959). (4) L. H. Brickwedde, J . Research Nnll. Bur. Standards, 42, 309 (1949).
(1) F A Cotton and J. P. Fackler, J r , J . A m . Chem. SOL.,83, 2818 (1961). J. P. Fackler, Jr., and F. A. Cotton, rbid., 88, 3775 (1961).
