July 5 , 1964
COMMUNICATIONS TO THE EDITOR
verted to a stable transannular p e r o ~ i d e ~which ~s~~ was isolated, characterized, and identified as the 5,8-peroxide of A5~7~24-cholestatrien-3/3-ol, m.p. 138139') [ a ] -42' ~ (CHCl,), [ 4 ] -182j; n.m.r., r 9.10, 9.17 (C-18, C-19, and C-21 methyls); 8.28, 8.36 ((2-23 vinyl methyls); 3.51, 3.80 (C-6and C-7 vinyl protons); 4.96 (C-24 vinyl proton). AnaZ. Found: C , 77.93; H , 10.17, in agreement with Cz?H4203. Structure was adduced by hydrogenation with PtOz as catalyst to the cholestane-3/3,50i,8ar-triol(111, R = H),26 m.p. 185-187', [ a ] D -21.4' (pyridine), [+] -88. Anal. Found: C, 77.20; H,'11.43, in agreement with C Z ~ H ~ ~ OThe : . triol (111) proved to be identical with an authentic sample prepared from I1 by photosensitized transannular addition of oxygenz3 followed by hydrogenation. Our findings reflect the role of 7-dehydrocholesterol as intermediate on the major pathway of hepatic cholesterol synthesis6 and indicate t h a t A5,7, *4-cholestatrien-3/3-01 (I) is a precursor of 7-dehydrocholesterol and of 24-dehydrocholesterol and is not a metabolite of the However, the quantitative importance of I as an obligatory intermediate on the normal pathway of cholesterol biosynthesis depends on the degree of side-chain reduction occurring a t any one point in the normal p a t h ~ a y . ~ Acknowledgment.--We thank Dr. E. Greselin for the animal work and Drs. G. Schilling and G. PapineauCouture for the gas-liquid chromatographic data. (22) Attempts to isolate pure I from livers of 30 rats treated with triparanol and AY-9944 were not successful. However, gas-liquid chromatography of the hydrolyzed neutral fraction indicated the presence of a peak with retention time corresponding t o t h a t calculated for I by the method of R . B. Clayton, Biochemistry, 1, 357 (1962). This finding lends support to the recent report on the probable presence of I in rat liver homogenates [M. E. Uempsey, J. D. Seaton, I f . G. Sanford, and R . W. Trockman, Federation Proc., 23, 425 (1961)l. (23) F. Schenck, K . Buchholz, and 0. Wiese, B e r . , 69, 2696 (1936). The peroxide (0.8 9.) was separated from cholesterol by chromatography on florisil and purified by "thick" layer chromatography. (24) I t is interesting t o note t h a t the same procedure, applied t o the brain of the same pig, has led t o t h e isolation of the peroxide of 11, thus suggesting inability of triparanol t o cross the blood-brain barrier, (25) Molecular rotation [cf. P . I f . Jones and W. Klyne, J . Chem. Soc., 871 (1960);. (26) T h e corresponding acetate (111, R = C H C O ) exhibited m.p. 160162O, [a]D -41.2' (CHClz), [ $ ] -190. n.m.r., 7 4.96 (carbinolic proton). Anal Found: C , 75.09; H . 10.81, in agreement with CzpHaoOl.
DEPARTMENT OF BIOCHEMISTRY AYERSTRESEARCH LABORATORIES MONTREAL, CAXADA RECEIVED APRIL22, 1964
D. DVORNIK M. KRAML J. F. BAGLI
The Vertical Ionization Potentials of Phenyl and Phenoxy Radicals
Sir: Recent measurements' of the appearance potentials of C6H5+ ions from phenyl halides lead to AHf(C6HS+) = 288 kcal./mole. This value is significantly lower than earlier values of -300 kcal./mole given by the appearance potentials of C6HSt ions from benzene and and phenyl halides. From A H f (C6H5) = 70 kcal./mole5s6 and A H ~ ( C ~ H S +=) 288 kcal./mole Majer and Patrick obtained 9.4 v. for the ionization (1) J . R. hlajer and C. R . Patrick, "Advances in Mass Spectrometry," Val. 2 , R M Elliott, E d . , Pergamon Press, Oxford, 1963, p. 555. (2) R . J. Kandel, J . Chem. P h y s , 22, 1496 (1954). (3) F. H. Field and J. I*. Franklin, ibid.,22, 1895 (1954) (4) J . Momigny, Bzill. S O L .Roy. Sci. L i e g e , 38, 251 (1959). (5) J. S. Roberts and H . A. Skinner, T r a n s . Faraday Soc., 46, 339 (1949). (6) hl. Szwarc, Chem. R e v . , 47, 75 (1950).
274 1
potential of the phenyl radical. This result is much lower than earlier values of I(C6H5) = 9.9 v.*-* based on Afff(C6H5+)= 300 kcal./mole, and is close to the ionization potential of benzene itself, 9.25 v.7,8 The vertical ionization potential of the phenyl radical has been recently measured in this laboratory by electron impact ionization of phenyl radicals generated by the thermal decomposition of azobenzene a t 800' in a reactor coupled to a mass spectrometer. I n addition to the phenyl radical, benzene and biphenyl were produced. The apparatus and the method of evaluating the vertical ionization potential by comparison with a rare gas standard have been described.$ The average of six determinations gives Ivert(C&) = 9.20 v., in reasonable agreement with Majer and Patrick's value.' Ivert(C6H6) measured in the same apparatus was 9.50 v., i.e., 0.25 v. higher than the adiabatic value.? Since i t may reasonably be expected that the adiabatic ionization potential of phenyl radical will be lower than the vertical value of 9.2 v. by a roughly similar amount, it appears that I(C&) is less than I(C&) by about 0.3 v. An upper limit for AHr(C6H5+jcan be obtained from the present result, using the relationship
dlthough Afff(C&) is gqnerally taken a t 70 kcal./ mole5z6 the bond stretching frequency in benzene suggests a slightly higher value, 73 kcal./mole.'O Taking this value and IVert(C6H5)= 9.20 v. (212 kcal./mole), Afff(C6Hjf) 283 kcal./mole, in close agreement with Majer and Patrick's value of 288 kcal./ mole (which is also an upper limit), This result clearly shows that the appearance potentials for C6HS+ions from aromatic hydrocarbons include considerable excitational energy. This is consistent with labeling experiments" which show that the formation of ions from toluene and other aromatic hydrocarbons is accompanied by extensive reshuffling of H atoms, and perhaps even the loss of the phenyl structure. On the other hand, the agreement of the upper limit for AHf(CsH5+) derived from the present result with that obtained from the dissociative ionization of phenyl halides' indicates that C6HSTions from phenyl halides have probably retained the phenyl configuration. The measurement of the vertical ionization potential of the phenoxy radical C6H60was carried out in the same apparatus. The radicals were produced by the thermal decomposition of allyl phenyl ether. The phenoxy radical has previously been detected by flash photolysis in absorption'? and in the thermal decomposition of anisole by mass ~ p e c t r o m e t r y . ' ~In the thermal decomposition of allyl phenyl ether, which decomposes a t a much lower temperature than anisole, a good yield of phenoxy radicals was obtained. Comparison of the ionization efficiency curve for phenoxy radical (mass 93) with that of a Kr standard gave ( 7 ) P . G Wilkinson, J . C h r m Phys , 24, 917 (1956) (8) K. Watanabe, ibid., 26, 542 (1957). (9) R . Taubert and F. P Lossing, J . A m . ChPm. S O L ,84, 1523 i l B 6 2 ) . (10) H. J Bernstein, Speclvochrm. A c t a , 18, 161 (1962). (11) P. N. Rplander and S. Ifeyerson, J . Chew. P h y s , 27, 1116 11957). (12) I . Piorman and G Porter, Puoc. Roy. Soc. (I,ondon), 8 2 3 0 , 3BY (1955). (13) A . G. Harrison, I-. R. Honnen, H . J . Dauben, Jr ., and F. P. Lossing, J . A m . Chem. S o c . , 82, 5593 (1960); R . F . Pattie and F P. Lossing, ibid., 86, 269 (1963).
2742
COMMUNICATIONS TO THE EDITOR
8.84 v. for Ivert(C6HsO). As would be expected, this is considerably higher than the ionization potential of the isoelectronic benzyl radical (7.76 v.).l 4 The appearance potential of the C6H50+fragment ion from anisole was found to be 11.92 f 0.1 v., in good agreement with the value obtained by Harrison, et ~ 1 . l Assuming this process to be a simple bond rupture in
Vol. 86
carbon atom as one of the vicinally coupled protons.' In vinyl compounds (11) it has been demonstrated that the cis-vicinal coupling constant (JH") is more sensitive than the trans-vicinal coupling constant (J"") t o the electronegativity of R," in qualitative ~agreement with the theoretical requirements.
-
-
(14) J . B. Farmer, I . H. S. Henderson, C. A. McDowell, and F. P . Lossing, J . Chem. Phys., aa, 1948 (1954). (15) J . h l . S. T a i t , T . W . Shannon, and A . G. Harrison, J . A m . Chem. S O L ,84, 4 (1902). (16) H. A. Skinner, "Modern Aspects of Thermochemistry," Lectures, hlonographs and Reports S o . 3 , T h e Royal Institute of Chemistry. London,
7
/-
\H
c=c
K
which the structural identity of the C6H50fragment is preserved, the relationship D(R-Y) .4(R+) - I(R) gives D(C&,O-CH3) L 71 kcal./mole. Evidence for the retention of the phenoxy structure in C6H50+ions from anisole a t the dissociation threshold is provided by the results of Harrison, et al., who found meta-para orientation to be preserved in R C s H 4 0 +ions produced in the dissociative ionization of RC6H40CH3 isomers. One might reasonably expect the 0-CH3 bond in anisole to be appreciably lower than this limit, however. The phenoxy radical is isoelectronic with benzyl radical, and should therefore exhibit considerable resonance stabilization. To a very rough approximation, then, D(CH30-CH3) - D(C6HsO-CH3) 2 D(CHx-CH3) D(C6H5CH2-CH3). Taking AHf(CH3) = 32.5 kcal./ mole16 and AHf(CsH5CHz) 43 kcal./mole,'' and the standard heats of formation of ethane and ethylbenzene,18 D(CH3-CH3) - D(C6HsCHz-CH3) 17 kcal. /mole. If D(CH30-CH3) = 77 k ~ a l . / m o l e , then ' ~ D(C6H50CH3) is: 60 kcal./mole. On this basis, the appearance potential of C6H50+ from anisole would include 11 kcal./mole of excitational energy. The validity of such a comparison is rather doubtful, however. Other phenoxy derivatives (phenol, phenyl ethyl ether, diphenyl ether) have such small intensities for C6H50+ ion that estimates of AHf(C6H50f)cannot be made from them. The value given above for A(C6H60+), together with AHr(aniso1e) = - 18.5 kcal./mole, leads to AHf(C6H50f) 224 kcal./mole and AH
