3586
COMMUNICATIONS TO THE EDITOR
for the relative rates of abstraction of halogen atoms from alkyl halides by methyl radicals. These reactivities are closely parallel to each other, and different from those for S N reactions. ~ One example is provided by the fact that benzyl bromide is more reactive than t-butyl bromide in the reduction, whereas t-butyl chloride solvolyses faster than benzyl chloride by a factor which depends strongly on the s ~ l v e n t . The ~ high reactivity of carbon tetrachloride (72.4), comparable to benzyl bromide (33.5), constitutes another. The reactivity sequence bromide < iodide is shown by the fact that n-heptyl iodide is more reactive than n-butyl bromide by a factor of 13.3. The data presented here are consistent with a mechanism in which the rate of formation of an alkyl free radical is the prinie factor in determining reactivity of a halide. This might involve a chain mechanism in which an organotin radical abstracts a halogen atom from the halide in one step, and the resulting alkyl radical abstracts a hydrogen atom from the organotin hydride in the other step.1° Catalysis of the reaction by azobisisobutyronitrile confirms the existence of a chain mechanism: tri-n-butyltin hydride in toluene a t 80 =t 2' reduced benzyl chloride, chlorocyclohexane, and bromobenzene to the extents of 100, 70, and 41y0 in 35 min. in the presence of 1.5 mole % of the catalyst, and to the extents of 26, 1, and 5% in its absence.
Vol. 84
methane. However, all attempts to trap the carbene intermediates JAth cyclohexene were unsuccessful. In fact, none of their products'{ require tlic postulation of carbenes as reaction intermediates. They can be accounted for by displacement and/or elimination reactions. In our original work' on the chemistry of dimethylsulfoniumfluorenylide we obtained several products which were best accounted for by the ylid first decomposing to a carbene. We have now succeeded in demonstrating conclusively that sulfur ylids can be decomposed to carbenes by trapping the carbene formed from the ylid (IIb). Franzen and Wittig4 previously reported the decomposition of trirnethylamnioniummethylide to inethylene which was trapped with cyclohexene to form norcarane. R-cH~-&(c~H~)~
+
R-cII-&c~FI~),-+ IT
Ia, R C3H Ib, R = CsIls
R--CH : I11
+ (CsHs)rS
Addition of 4 mmoles of n-butyllithium solution to a slurry of 4.0 mmoles each of benzyldiphenylsulfonium tetrafluoroborate (Ib) and acenaphthylene in anhydrous tetrahydrofuran a t -40' afforded 4.0 mmoles of lithium tetrafluoroborate (100%) ; 3.3 mmoles of phenyl sulfide (83%), determined and identified by oxidation to phenylsulfone, m.p. 122.5-124.5 (lit.,5 m.p. 124'); 1.1 mrnoles of un(9) A. Streitwieser, Jr., Chem. Reus., 66, 616 (1956). (10) This last statement cannot, of course, apply to the dehalogenareacted acenaphthylene; and 1.2 mmoles of 7tion of vicinal dihalides, which proceeds according t o the equation phenyl - 7H - Gb,7a - dihydrocycloprop [alacenaphthylene (IV) (43% based on unreacted acenaphthylI 1 I / 2RsSnH + -C-C+-C=C2RsSnBr HP ene, 31% based on Ib), m.p. 170' dec. I I The adduct (IV) was identified by its inertriess to Br Br cold permanganate or bromine; by its infrared (11) National Science Foundation Senior Postdoctoral Fellow spectrum (no olefinic absorption, Amax 13.70 and California Institute of Technology, 1959-1960. 14.83 p, characteristic of monosubstituted aryl) ; DEPARTMENT OF CHEMISTRY HENRYG. K U I V I L A ~by ~ its ultraviolet spectrum (Ama, 232 mu (log E 4.3), OF NEWHAMPSHIRE UNIVERSITY 296 (3.6), 310 (3.5) and 323 (3.1)); and by oxidaLAWRENCE W. MENAPACE CONTRIBUTION No. 2855 GATESAND CRELLINLABORATORIES CHARLES R. WARNER tion with chromic acid to benzoic acid (m.p. 120121') and naphthalene-] ,8-dicarboxylic acid anINSTITUTE OF TECHNOLOGY CALIFORNIA PASADENA, CALIFORNIA hydride, m.p. 272.5-274' (lit.,6m.p. 274), identified RECEIVED AUGUST 2, 1962 by admixture melting point and infrared spectra comparison with an authentic sample.
+
+
T H E DECOMPOSITION OF SULFUR YLIDS TO CARBENES'
Sir : Franzen and co-workers2 recently studied the reaction of alkyldiphenyl-sulfonium salts with strong bases. On the basis of isolating triphenylmethane (42y0), 1,1,l-triphenylpentane(15y0)and 1-butene (7y0) from the reaction of tritylsodium with n-butyldiphenylsulfonium tetrafluoroborate (Ia) they postulated that this salt was first converted to the corresponding ylid (IIa) (by proton abstraction) which then decomposed to phenyl sulfide and the carbene (IIIa). Similar results were reported for the benzyldiphenylsulfonium salt (Ib), affording 1,1,1,2-tetraphenylethane by an insertion reaction of the carbene (IIIb) with triphenyl(1) This is paper VI1 of our series "The Chemistry of Ylids"; for paper VI see Johnson and LaCount, J . A m . Chem. SOC.,83, 417 (1961). (2) V. Franzen, H. J. Schmidt and C. Mertz, Chem. Bev., 94, 2942 (1961).
IV
We have found that other alkyldiphenylsulfonium ylids can be decomposed to carbenes which can be trapped with acenaphthylene and other olefins. We have also noted thlrt this deconlposition can be avoided by working a t even lower temperatures. This work is continuing and will be published in detail a t a later date. (3) An unspecified amount of a (5-phenylisopropyl)-ethyl ether resulting from an insertion reaction of (possibly) the benzyl carbene on ethyl ether was reported and identified by vapor phase chror1.ntography analysis and infrared spectrum. (4) V. Franzen and G. Wittig, AngPw. Chem., 7 2 , 117 (IWiO). ( 5 ) 0. Hinsberg, Chem. Ber.. 43, 290 (1910). (6) C. Graebe and E. Gfeller, ibid., 26, 652 (1925).
COMMUNICATIONS TO THE EDITOR
Sept. 20, 1962
3587
Acknowledgment.-We are indebted to the National Science Foundation for financial support (Grant No. G-17345).
Chlorophyll a, isolated from spinach and carefully purified, exhibited three sharp, low field resonance bands ( T = 0.62, 0.85, and 1.81, measured in CDC13, tetramethylsilane as internal indicator). DEPARTMENT OF CHEMISTRY UNIVERSITY OF NORTH DAKOTA VICTOR J. HRUBY These bands correspond to the three bands found GRANDFORKS,NORTHDAKOTA A. WILLIAM JOHNSON in chlorin e6 trimethyl ester and assigned to the a-, RECEIVED JUNE 22, 1962 p- and &bridge hydrogen atoms by Woodward and Skaric.5 All other hydrogen resonances in chlorophyll a, including the C-10 hydrogen, lie a t higher SITE OF EXCHANGEABLE HYDROGEN IN fields. When chlorophyll a was treated with excess CHLOROPHYLL D FROM PROTON MAGNETIC RESONANCE MEASUREMENTS ON DEUTERIOCH30D in CClk for 48 hours,2 the band a t r = 1.81 CHLOROPHYLL 01 disappeared, in a manner entirely analogous to Sir : that reported for chlorin by Woodward and S k a r i ~ . ~ When the exchanged chlorophyll a was treated with As shown by an indirect infrared method, chlorophyll a possesses one slowly exchangeable hydrogen CHaOH, the band reappeared. These facts indicate atom when the chlorophyll is treated with methanol that the &bridge hydrogen atom undergoes the exin carbon tetrachloride solution a t room tempera- change. To minimize uncertainties and ambiguities in the ture in the dark.2 Although the position of the exchangeable hydrogen could not be established by interpretation of the very complex chlorophyll the exchange method, considerations had indicated n.m.r. spectrum, we followed the hydrogen exthat the active hydrogen probably was located a t change in deuterio-chlorophyll a.4 This substance showed no proton resonances. On treatment with the C-10 position (I).3 We have now carried out proton magnetic reso- CHaOH in CC1, for 48 hours, one, and only one, = 1.81. Again, nance measurements (Varian A-60 n.m.r. spec- proton resonance appeared a t treatment with CH30D caused t h e resonance peak trometer equipped with an audio oscillator and counter) that confirm the presence of but one to disappear. These results with deuterio-chloroexchangeable hydrogen atom in chlorophyll a phyll a demonstrate unequivocally that only one (under our exchange conditions). The exchange- hydrogen atom undergoes exchange under the reacable hydrogen is not located a t the C-10 position, tion conditions used here, and that the labile hydrobut is in fact on the &bridge carbon atom of the gen is located on the porphyrin ring a t the 6 posiporphyrin ring (I). The n.m.r. observations were tion. The remarkable observations of Woodward and made with two materials; not only was theexchange followed in ordinary chlorophyll a, but also in the Skaric5 on the chemical and exchange behavior of fully deuteriated chlorophyll a that we have avail- chlorins thus also appear to be valid for chlorophyll able.4 The use of the deuterio-chlorophyll made it a itself. That the hydrogen a t the 6 position is possible to detect hydrogen resonances that might more labile than the one a t C-10 could only be otherwise have been obscured in ordinary chloro- established by measurements on chlorophyll itself rather than on derivatives lacking the central phyll. magnesium atom and the cyclopentanone ring. Although the conditions used for the hydrogen exchange are certainly not physiological, nevertheless the lability of the &hydrogen atom in chlorophyll a as established here may also be pertinent to the chemical behavior of chlorophyll a in photosynthesis, We are deeply indebted to Dr. Gerhard L. Closs of the University of Chicago for guidance and instruction in the interpretation of the n.m.r. spectra. ( 5 ) R. B. Wnndwarrl and V. Skaric. J . A m . Chem. Sor., 8 3 , 4076 (1961).
CH2
I
I
"-;;IO COOCH3 C00G20H39 CHZ
I
I
JOSEPH J. KATZ ARGONNEXATIONAL LABORATORY MARYR. THOMAS ARGONNE, ILLINOIS HAROLD II. STRAIN RECEIVED JULY 25, 10632
c. 0 9
(1) Based on work performed under the auspices of the U . S.
Atomic Energy Commission. (2) J. J. Katz, M. R. Thomas, H. L. Crespi and H. H. Strain, J . A m . Chem. Soc., 83, 4180 (1961). (3) H. Fischer and S. Goebel, A n n . , 633, 168 (1936); S. Aronoff, Encyclopedia of Plant Physiology, 6, 234 (1960); W. Vishniac and I. A. Rose, Naluie, 133, 1089 (1958). (4) H. H. Strain, M. R. Thomas, H. L. Crespi, M. I. Blake and J. J. Katz, A n n . New York Acad. Sci., 84, 617 (1960).
THE EFFECT OF H?O ON THE y-RADIOLYSIS OF AERATED CHaOH'
Sir : We wish to report a previously overlooked2 effect of Ha0 on the y-radiolysis of aerated methanol which gives promise of serving as a useful tool in the study of radiolytic mechanism. (1) Research performed under the auspices of the United States Atomic Energy Commission. (2) E. Hayon and J. J. Weiss, J. Chem. Soc., 3970 (1961).
