COMMUNICATIONS TO THE EDITOR
3910
TABLEI NI4 COUPLING CONSTANTS Negative ion radical“
)(IN),g>lUVS
Nitroethane 25.2 1-Nitropropme 24.8 2-Nitropropane 2.5.4 1-Nitrobutane 24.3 2-Nitrobu tane 24.5 e All generated in background electrolyte of 0.4 with 10% 1-propanol for solubility.
M KCI
diphenylamine and several derivatives of chloropromazine has been carried out in aqueous buffers. Acknowledgment.-We wish to express our appreciation to Robert Lorimer for aid in the design and construction of the electrochemical cells. This work was sponsored in part by the Atomic Energy Comniission through contract AT(11-1)686 and by the Air Force through AFOSR. DEPARTMENT OF CHEMISTRY OF KANSAS UNIVERSITY LAWRENCE, KANSAS RECEIVED JULY 10, 1901
Apparently7 the fundarnen tal difference between hyperconjugation and a-hydrogen bonding is that whereas hyperconjugation depends upon a non-vanishing magnitude for the exchange integral between a carbon pn-orbital and the tetrahedral hybrid on an a-carbon, a-hydrogen bonding requires a non-vanishing magnitude for the exchange integral between the pn-orbital and an a-hydrogen 1s-orbital. Until calculations have demonstrated that acceptable estimates of the latter integral can account for the magnitudes and signs of observed long range couplings, it seems more reasonable to interpret these couplings in terms of hyperconjugation. More elaborate arguments and evidence in support of this view may be found in some recent works.* (7) M M . Kreevoy, Tetrahedron, 6 , 233 (1959). (8) (a) R. A. Hoffman and S. Gronowitz, Arkir Kcmi, 16,471 (1060); (b) R. A. Hoffman, ibid., 17, 1 (1960).
L. H PIETTE INSTITUTE OP PHYSICS RAGNAR A. HOFFMAN P. LUDWIG INSTITUTE OF CHEMISTRY SALOGRONOWITZ R. N. ADAM UNIVERSITY OF UPPSALA UPPSALA, SWEDEN RECEIVED JULY 24, 1961
ON THE MECHANISM FOR LONG RANGE PROTON SPIN COUPLINGS
Sir: Jn a recent paper, Kreevoy, ct aLI1question the evidence for hyperconjugation as a mechanism for long range coupling in unsaturated compounds2 such as propargyl derivatives and suggest that ahydrogen bondinga should be involved instead. They refer to recent theoretical work by Karplus4 which, they claim, utilized essentially the a-hydrogen bonding model. This latter assertion is not valid. Though Karplus states explicitly4 that “the orbitals on the protons H and H’ whose coupling is being determined are taken to be part of the siginaelectron system, this does not imply that hyperconjugation is not involved in his theory, since the proton spin coupling, as calculated by Karplus, is expressed in terms of hyperfine coupling constants and triplet state energies. Now the hyperfine coupling constants are obtained from e.s.r. data on related free radicals and insofar as the fragments \
VOl. 83
*
H-C-C are concerned all theoretical interpreta/ tions6J involve hyperci *ijugation either in its 1110lecular orbital5 or valencz bond6 formulation. Thus the successful theoretical treatment of long range coupling4 depends ultimately on hyperconjugation, rather than on the use of the a-hydrogen bonding model. (1) M. M. Kreevoy, H. B. Charman and D. R . Vinard. J . Am. Chem. Soc., 83, 1978 (1951). (2) (a) R . A. Hoffman, M o l . Phys.. 1, 328 (1858); (b) E. B. Whlpple, J. H . Goldstein and L. Mandell, J. Chcm. Phys., S O , 1108 (1959): (c) E. E. Whipple. J. H. Goldstein and W. E. Stewart, J . Am. Chcm. Soc., 81, 4761 (1959); (d) A. D. Coben and N . Sheppard. Proc. Roy. Soc. (London). A262, 488 (1959); (e) P. L. Corio and I. Weinberg. J. Chcm. Phys., 81, 5G9 (1959; (f) R . A. Hoffman and S. Gronowltz, Acta Chem. Scand., 13, 1477 (1958). (3) M. M. Kreevoy and H. Eyring, J . Am. Chcm. Soc., 79, 6121 (1857). (4) hl. Karplus, J . Chem. Phys.. 33, 1842 (1960). (5) (a) R. Bersohn, ibid., 24, 1066 (1955); (b) D. B. Chesout. ibid., 19, 43 (1958). (6) (a) A. D. McLachlan, Mol. Phys., 1, 233 (1858); (b) P. G . Lgkor, J . Chcm. Phys., 32, 625 (1860).
O N d HYBRIDIZATION IN CHLORINE
Sir : The use of hybridization in valence bond theory in order to explain observed molecular geometries is well known. The classic example is of spa hybridization in the group I V elements. In this case the bonding configuration (valence state) is considerably above the ground state in energy. The necessary promotion energy is regained upon bond formation yielding, of course, a stable molecule. It is known that compounds of fluorine and chlorine exhibit many differences in structural properties. These differences frequently have been attributed to d hybridization in chlorine. Since chlorine has valence electrons in the 3s and 3p shells, exritation of these electrons to the 3d orbital has been assumed to be relatively easy in contradistinction to fluorine where the necessary excitation is 2s and 2p to 3d. The argument of easy orbital excitation when the principal quantum nuniber remains unchanged is based in essence upon the assumption that the field in which the valence electrons move is reasonably close to coulombic. I n fluorine the energy of the center of gravity of states arising from 2s22p4(3P)3d is 15.9 e.v.’ Until recently the energy location of states arising from the configuration 3s23p43din chlorine was unknown. Humphreys and Paul2 and Minnhagena have analyzed the chlorine spectrum. From their assignments the center of gravity of states arising from 3 ~ ~ 3 p ~ ( ~isP 11.2 ) 3 de.v. In both fluorine and chlorine the excitation energy is large, in each instance being slightly less than 2 e.v. below the ionization limit. It would thus appear arbitrary to select the configuration of chlorine involving 3d electrons from the wealth of configurations near the ionization limit. I n view of (1) Charlotte E. Moore, “Atomic Energy Levels,” Vol. I, Unlted States Department of Commerce, 1949, Washiugton. D. C . (2) C. J. Humphreys and E. Paul, Jr.. J . Opl. SOC.Am., 49, 1186 (1958). (3) L. Mionhagen, ibid.. 61, 298 (1951).
Sept. 20, 1961
COMMUNICATIONS TO THE EDITOR
3011
ture, [ala10 +534', from which the trans-ketone (IV, R = CBH17, It1 = 0, G a H; n1.p. 136-137'; [a]D 10'; [a1310 1017', negative Cotton effect) was isolated.' This therinodynamic equilibrium corres onds to 77% cis i3 23% trans (as calculated The two ketones also were formed by from passing 111 through alumina (Merck, acid washed). The structure and stereochemistry of the cis-ketone was proved through its conversion (65%) by WolffKishner reduction t o the hydrocarbon (IV, K = CsHi7, R1 = Hz, 6 @ H; m.p. 71-72': [.ID +29.5') which was identical with the one obtained by conversion of B-norcoprostan-3-one6with diazomethane in situ to A-homo-B-nor-6@H-cholestan-4-one (m.p. 78-80'; [ a ] $21') ~ and subsequent reducDEPARTMENT OF CHEMISTRY HARVARD UNIVERSITY WILLIAM KLEMPERER tion. CAMBRIDGE 38, MASS. Similarly, androstane-5a,6a,17@-triol 6-tosylate RECEIVED JULY5, 1901 [ a ] 4-34') ~ was 17-benzoate (m.p. 129-130'; converted to (IV, R = OBz, R1 = 0, 6PH; m.p. ~ The corresponding alco120-122'; [ a ] 4-51'). SYNTHESIS OF PERHYDROAZULENES hol (IV, R = OH, R1 = 0, G@H; m.p. 152-153'; Sir: [ a ]+5'; ~ CY]^^^ +275', positive Cotton effect) We wish to report a facile stereospecific conver- was stable to acid and base. sion of bicyclo [4,4,0]decanes to bicyclo[5,3,0]deCholestane-4a,5a-diol4-tosylate (map.135-136'; canes. This conversion involves a pinacolic type [ a ]$So), ~ stable on alumina (Merck acid washed), rearrangement of vicinal cis-glycol monotosylates. was rearranged to the ketone (VI R = CBH17, 10-Methyl-A1~g-octalin2 in pyridine was treated 4@H; n1.p. 87-88'; [ a ] D $125'; [a1310 +1945', with osmium tetroxide to give 10-methyl-decal- positive Cotton effect) with 1 mole equiv. of potasl,g-diol, m.p. 96-98', which with tosyl chloride in sium t-butoxide or by boiling with dimethylformpyridine a t room temperature yielded 10-methyl- amide-calcium carbonate for 24 hr. This ketone decal-1,g-diol 1-tosylate (I). Treatment of I with was converted to the equilibrium mixture, [ a1310 potassium t-butoxide in t-butyl alcohol or with $1520' (8570 cis F! 15% trans) from which (V, alumina (Alcoa F-20) gave 10-methylbicyclo- R = C8H17, 4aH; m.p. 136-137': [a]D -27', [5,3,0]decan-l-one (TI), b.p. 110' (25 mm.); [a1710-717', negative Cotton effect) was isolated. dinitrophenylhydrazone, m.p. 118', Amax 366 mp In the same way androstane-4a,5a,17j3-triol-4( E 25,900); semicarbazone, m.p. 169-171', Xmax 226 tosylste 17-benzoate (m.p. 138-139'; [ a ]$53') ~ inp ( E ll,800).a The stereochemistry arid exact was rearranged to (V, R = OBz, 4@H; map. 166conditions of this rearrangement were studied on a 16s'; [ a ]$33'). ~ The equilibrium mixture of more convenient system, on the steroidal nucleus. the two corresponding ketoalcohols (V, R = OH, @ H ; m.p. 126-128'; [ a ] D +20'; [ a 1 3 2 0 $looo, It R positive Cotton effect), and (V, R = OH, 4aH; ~ [a]320 +2375', posim.p. 162-163'; [ a ]$155'; tive Cotton effect) corresponds to 15% cis Fs: 85% trans. The trarisition state for this rearrangement demands coplanarity of the four centers involved, 0 which together with the concerted electron shift, V causes retention of configuration a t the bridgehead Cholestaiie-Sa,Ga-diol 6-tosylate (111, 1n.p. 122- of the product.s 123'; [ a ] D $18') on treatment a t 25O with 1 mole It is of particular interest that there are marked equiv. of potassium t-butoxide in t-butyl alcohol or variations in the relative stabilities of the cis F? a t 100' with dimethylformamide and calcium car- trans ketones in the systems studied. Differences bonate (for 8 hr.), gave a quantitative yield of cis- in the substitution in these systems may change ketone (TV, R = C8H17, R1 = 0, 6 @ H: m.p. the conformation of the cycloheptane ring as in125-126'; [ a ] +19'; ~ [a1310 +1105', positive ferred from the directions of the respective CotCotton effect). Heating with aqueous sulfuric ton effects. This is in accord with the flexibilities acid-dioxane for 2 hr. yielded an equilibrium mix(4) This is compatible with the Cotton effect of similar systems: (1) Cf., inter at.. P. D. Bartlett and R . H. Hosenwald, J . Am. Chcn.. D. N. Kirk and V. Petrow, J . Chem. Soc., 4057 (I9GO); G. Bilchi, S. W. the similarity in magnitude of 3d excitation in chlorine and fluorine it seems unreasonable to base gross differences in structural properties on these relatively small differences in energy. In view of the fact that most modern calculations require an extensive linear combination of atomic orbitals for a basis set, the selection of a single excited configuration as predominant for qualitative discussion would appear useful only if this configuration lies uniquely lower in energy than the remaining member of the set. It appears that the differences in chemical bonding of chlorine and fluorine cannot be iiaturally explained simply in terms of 3d hybridization.
Soc., 66, 1090 (1934): R. B . Bates, G.Bdchi, T. Matsuura and R . R. Shaffer, ibid., 83, 2327 (1000); E. J. Corey, hf. Ohno, P. A. Vatakencherry and R. B . Mitra, ibid., 83, 1251 (1901) N. L Wendler, Tetrahedron, 11, 213 (1900). (2) P. Sondheimer and D. Rosenthal, J . Am. Chem. S O C ,80, 3995 (1 958).
(3) All the compounds described gave analytical results. infrared and ultraviolet spectra compatible with the assigned structures. Rotations were taken In chloroform and rotatory dispersion measurements in dioxane.
-
-
L]wI).
Chaw, T. Matsiiura, T.L. Popper, H . 13. Rennbnrd and M. Schach von Wittenau, Tctrahcdran Lcffeur,6, 14 (1959). (5) J. Joska. J. Pajkos and F. Sorm, Coll. C E C CChetn. ~. Comm., 26, 234.1 (1900). ( 6 ) cf.,inter al., R. J. W. Cremlyn, D. L. Garmaise and C. W. Shoppee, J . Chcm. Soc., 1847 (1953); J. Elks, G. H. Phillipps, D. A. H . Taylor and L. J. Wyman, ibid., 1738 (1954); R . Hirschmann, C. S. Snoddy, C. F. Hiskey and N. L. Wendler, 1.Am. Chem. SOC.,76,4013 (1954): R . Anliker, 0. R o k and H. Heusser, Hela. Chim. Acta, 38, 1171 (1955).
