3224
COMMUNICATIONS TO THE EDITOR
VOl. 78
these four products are formed at -75’ in the ration 10 :25 : 25 : 40 and at 15’ in the ratio 11: 26 : 26: 37. Here again one product, l-niethylcyclohexene, appears to be formed in amount differing slightly from the statistical, but the estimated error of 1070is too large to justify further comment a t this time. Finally diazomethane reacts with cyclopentane to give methylcyclopentane unaccompanied (less than 0. 1y0)by cyclohexane. Within much increased limits, one concludes that methylene does not react (9) W. D . Paist, E . R. Blout, F. C . Uhle and R . C . Elderfield, J . with the carbon-carbon bond. O v g . Chem., 6, 273 (1911). Within a few per cent. it is clear that methylene MEDICITALCHEMICAL RESEARCH SECTIOK fails to distinguish between different types of RESEARCH DIVISIOS AMERICANCYASAMID COMPASV ~ ‘ I L L I A M S. ALLEX carbon-hydrogen bonds, reacting a t random with LEDERLE LABORATORIES SEYMOUR BERNSTEIN l o ,2’ and 3’ aliphatic C-H bonds as well as with PEARL RIVER,NEWPORK “allylic” and “vinyl” types. Even more surprisRECEIVED . ~ P R I L 6, 1956 ing is the fact that the reaction cross-section of the double bond is only about three times that of a INDISCRIMINATE REACTION OF METHYLENE hydrogen atom. One might reasonably have exWITH THE CARBON-HYDROGEN BOND pected an overwhelming preference similar to that Sir: shown by dichlorocarbene. Methylene must be PIIethyle ne shows essentially no discriminatory classed as the most indiscriminate reagent known power in its reaction with the carbon-hydrogen in organic chemistry. bonds’ in the saturated hydrocarbons, n-pentane The experiments were carried out in the usual and 2,3-dimethylbutane and in the olefin, cyclo- way by irradiating very dilute solutions of diazohexene. When diazomethane is irradiated with methane in the hydrocarbon with sun-lamps, rethese hydrocarbons either a t -75’ or at 15”, the moving most of the solvent by fractional distilladistribution of products resulting from insertion tion in a 50-plate column and analyzing the product of the carbene into a carbon-hydrogen bond is by gas-liquid partition chromatography using a very close to statistical. For example, diazo- Perkin-Elmer Model 154 Vapor Fractometer and methane may react with n-pentane a t carbon atoms by infrared spectroscopy. In all cases but one, 1, 2 or 3 to give n-hexane, 2-methylpentane and 3- samples of pure product were isolated from the methylpentane in the statistical ratio 6 : 4 : 2 or Fractometer and identified by comparing their 50: 33.3 : 16.7, respectively; in fact, these three infrared spectra with those of authentic samples products are formed a t -75’ in the ratio 48:35: 17; (neat, 0.1-mm. cell). In the cyclohexene reaction and a t 15’ in the ratio 49:34:17. where the Fractometer mould not resolve 3- and 42,3-Dimethylbutane which contains primary and methylcyclohexene but easily separated this mixtertiary hydrogen in the ratio of 12 : 2 or e5.7 : 14.3 ture from the other components, analysis was made reacts with diazomethane to give 2J-dimethyl- infrared spectroscopically comparing with known pentane and 2,2,3-trimethylbutane in the ratio mixtures of the authentic olefins. 53:17. It is possible that the small difference (2) W. v a n E . Doering and A . K . H o f f m a n n , THISJ O U R N A L , 1 6 , between the statistical and actual values lies out- 6162 (1954). (3) Sterling Chemistry Laboratory, Yale University, Kew Haven, side the experimental error but higher resolution in the form of considerably refined techniques Connecticut. W , \‘ON E. DOERINC3 would be required to decide. R. G. BUTTERY HICKRILL CHEMICAL In cyclohexene there are three types of carbon- RESEARCH R. G. LAUCHLIN FOUNDATIOS K. CHAUDHURI hydrogen bonds of widely differing chemical char- KATOSAH,NEWYORK RECEIVED J U N E 4, 1956 acteristics : the “vinyl” hydrogen atoms a t carbon atoms 1 and 2 , the “allylic” hydrogens a t Cs and Ce and the “aliphatic” hydrogens a t C4 and Cg. SYNTHESIS OF APOGOSSYPOL HEXAMETHYL ETHER Reaction with methylene can occur in principle a t each type of carbon-hydrogen bond to give 1- Sir: methylcyclohexene, 3-niethylcyclohexene and 4hpogossypol hexamethyl ether, the primary degmethylcyclohexene, respectively, in the statistical radation product of gossypol (I),has been formulated ratio of 2:4:4. Reaction can also occur by addi- as 2,2’-bi-5-isopropyl-1,6,7-trimethoxy - 3 - methtion to the double bond to give norcarane. In fact, ylnaphthyl (11) by Adamsl as a result of his extensive and thorough studies. The Adams structures (1) T h e photochemical decomposition of diazomethane was first in\.estigated b y H. Meerwein, H. R a t h j e n a n d H . Werner, B e y . , 75, however, have been criticized recently by three I O I O (1942), in ether where ethyl n-propyl a n d ethyl isopropyl ether separate groups of i n ~ e s t i g a t o r s . ~The ~ ~ , present ~ were formed. Although t h e l a t t e r product appears t o be formed b y H , 7.96), positive legal t e ~ t . ~Treatment ,~ of I with acetic anhydride-pyridine a t room temperature yielded only starting material. The compound formed a mono-2,4-dinitrophenylhydrazone at C-3, m,p, 2560 d,, chloroform-sba.&IC. 257 ( E 17,300) and 357-388 mp (E 29,000) (Anal. Calcd. for C27H32N408:C, 59.99; H , 5.97; N, 10.37. Found: C, 59.71; H, 6.15; N, 10.25). il detailed report of this work will be forthcoming in this JOURKAL.
direct reaction of a carbon-hydrogen bond, t h e presence of t h e oxygen a t o m and t h e possibility i t affords for initial “ylid” formation [see R . Huisgen, Angew. Chenz., 6 1 , 439 (195513 makes t h e mechanism ambiguous. T h e direct reaction of carbethoxy carbene with s a t u rated hydrocarbons h a s been reported b y W. v a n E . Doering a n d L. H. K n o x , 119th Meeting of t h e American Chemical Society, Boston, >lass., April 2, 1951 “Abstracts of Papers,’’ p . 21.1, a n d THISJ O U R N A L . in press.
(1) R. Adams, R. C . Morris, T. A . Geissman, D . J . B u t t e r b a u g h a n d E. C. Kirkpatrick, THISJOURNAL, 60, 2193 (1938). (2) D . A. Shirley a n d W. L. D e a n , ibid., 11, 6077 (1955). (3) R . T. O’Connor, P H a a r , E. F. D u P r e , L. E. Brown a n d C . H. Pominski, ibid.,1 6 , 2368 (1954). (4) C. H. Boatner, “Pigments of Cottonseed a n d Cottonseed Products,” Edited b y A. E . Bailey, Interscience Publishers, I n c . , Xew York, N. Y., 1948, Chap. V I , p p , 215-262.
July 5, 1956
COXMUNICATIONS TO THE EDITOR
3225
communication reports an unambiguous synthesis of ring size on the nuclear magnetic resonance of apogossypol hexamethyl ether, identical in all signals. Accordingly, the proton signals from a respects with that obtained from I (natural prod- number of polycyclic compounds have been exuct m.p. 277-278"; synthetic product m p . 277- amined (Table I). The shifts are measured rela279" ; mixed m.p. 277-278' ; infrared spectra in- tive to the main peak due to the rest of the arodistinguishable),j which establishes conclusively the matic hydrogens. Usually the distance of the correctness of the original Adams formulation. two peaks from a standard marker (cyclohexane or Compound I1 was synthesized by the oxidative decalin when the latter was used as solvent) were measured. The strong interaction found in benzocoupling of S-isopropyl-G,7-dimethoxy-3-methyl-1naphthol (111) m.p. 129-13O0, found; C, 73.76; phenanthrene is considerably weakened by the H, 7.78)6 with. ferric chloride in dioxane as the change in angle introduced by the 5-membered solvent and subsequent methylation of the resulting ring in dibenzofluorene, and the initially weaker 2,2' - bi - 1 - (5 - isopropyl - 6,7 - dimethoxy - 3- phenanthrene interaction is similarly much remethyl)-naphthol (m.p. 271-275", found: C, duced in fluorene and completely removed in 74.21; H, 7.39). Consideration of models demon- fluoranthene, where the hydrogen atoms are even strates that no coupling in the 4position of I11 farther apart. can occur due to steric hindrance. TABLE I H-H distance A Shift, cycles Compound I11 was prepared in the following manner : formylation of 3-isopropyl- 1,2-dimethoxyben- 3,4Benzophenanthrene 0.53 68 1.71 45 zene7,* gave 2-isopropyl-3,4-dimethoxybenzalde-1,2,5,6-Dibenzanthracene hyde (IV), b.p. 98-102" (0.13 mm.) (2,4-dinitro- 1,2,5,6-Dibenzfluorene 1,73 40 1.76 42 phenylhydrazone, m.p. 190-192', found: C, 55.79; Phenanthrene 2.72 12 uncertain H, 5.58). Oxidation of IV gave 2-isopropyl-3,4- Fluorene 0 uncertain dimethoxybenzoic acid as shown by comparison Fluoranthene 2.95 with an authentic sample.g The Stobbe reaction A fairly simple relation exists between the of IV and diethyl succinate with subsequent ring nuclear magnetic resonance shift and the hydrogenclosure (sodium acetate and acetic anhydride) and saponification in the conventional manner hydrogen distance. However, since the H-H distances are calculated on the basis of a planar molegave l-hydroxy-5-isopropyl-G,7-diniethoxy-3-naphthoic acid (m.p. 226-227", found: C, 66.32; H, cule and the H-H repulsion must distort the mole6.34). Reduction with lithium aluminum hydride cule somewhat out of plane, these values are not the true equilibrium H-H distances. Allowance for gave . 3-hydroxymethyl-5-isopropyl-6,7-dimethoxythis fact would increase the small H-H distances 1-naphthol (m.p. 207-209", found: C, 69.71; H, 7.20). Hydrogenolysis with palladium on char- proportionally more than the large ones. X consideration of other molecules with interactcoal in methanol with a trace of hydrochloric acid ing protons shows some interesting effects. Perylgave 111. The interest and encouragement of Professor ene with two sterically interfering pairs each tending to twist the molecule out of plane about the R. B. Turner is gratefully acknowledged. central pair of bonds shows a shift of only 25 cycles, ( 5 ) Perkin-Elmer Model 21 Spectrophotometer, 0.11-mm. cells, solabout half that observed for phenanthrene. The v e n t chloroform, concn. 7Y0 doubled torque has clearly effected an increased (6) All analyses b y Huffman Xicroanalytical Laboratories, Wheatridge, Colorado. twist, so that the protons in each pair are con(7) R Adams, M .H u n t a n d R. C. Morris, THIS J O U R N A L60, , 2972 siderably farther apart than in phenanthrene. (1938). Chrysene also has two interacting pairs, but now ( 8 ) J. D. Edwards, Jr., a n d J. L . Cashaw, J . Ovs. Chem., 20, 817 the torque is across a 3-bond rather than a two(1955). bond system. Correspondingly we find a shift (9) R. Adams a n d B. R. Baker, THISJ O U R N A L , 61, 1138 (1939). intermediate between that of phenanthrene and RESEARCH LABORATORIES J . D. EDWARDS, JR. VETERANS-%DMISISTRATIOX HOSPITAL J. L. CASHAW perylene (35 cycles). A more detailed examination HOCSTOX, TEXAS of this kind of structural effect will be published RECEIVED MAY21, 1956 shortly. One is easily convinced from the molecular models that the 4-5 proton repulsion in phenanNUCLEAR MAGNETIC RESONANCE STERIC SHIFTS threne puts the maximum strain on the C9-Cl0 IN -CH GROUPS bond. The anomalously high "K region" reactivSir : ity which is greater than predicted by molecular Sterically interfering hydrogen atoms in the 4-5 orbital theory based on a planar molecule is no position of phenanthrene and many similar mole- doubt due to this fact. A study of the shifts in cules shows up strongly in high resolution nuclear carcinogenic hydrocarbons is also in progress. magnetic resonance experiments. Since the work It has been pointed out to the author2 that the of Arnold, et uZ.,I on the steric effects of methylene closeness of the proton to two aromatic rings groups in hydrindene and tetralin with a-sub- rather than one may result in the kind of shift obstituents in the aromatic ring shows clearly that served. An investigation of the relative importhe interaction is much greater when both rings tance of the two factors is under way. are six-membered than when one of them is fiveThe work reported here was done using a Varian membered, it is of interest to investigate the effect Associates Model V4300 spectrometer working a t ( I ) R. T. Arnold, V. J. Webers a n d R . M. Dodson. THISJOURNAL, 7 4 , 868 (1952).
(2) W . G. Schneider, H. J. Bernstein a n d J. A. Pople, private communication
