4486
If spin density in the T framework were the only mechanism causing the shifts, the following relationships are expected to apply, which relate the carbon-13 and proton isotropic shifts to the appropriate uc and uH value for the ring positions of the molecular fragments 2 and 3, where pciT is the spin density in the pr orbital centered at Ci.5
pmr studies. It has been notedsarc that, if the p-tolyl group is replaced by ethyl or n-propyl, large negative contact shifts are observed for the CH2 and CH, groups of the chain. The only reasonable mechanism of placing spin density at these positions is uia the u framework, and thus it is not unreasonable to expect nuclei at ring positions to also exhibit isotropic shifts resulting from such a mechanism. On this basis, the uconCi/uconHi = (YHIYC)(35'5PCir observed small isotropic shifts of C-1 and C-2 seem 14'0[pch" Pc,"l)/(-22*5pci") (3) quite reasonable. The u and T effects, though large, should be opposite in sign and could be close to equivE -6.2 2.5(~c: PC,")/PC," (4) alence at these positions. The present results indicate CHI CH3 = that spin densities calculated from proton isotropic ucon /ucon (YH/YC)(- 14.0PCk"/27.5PC~") (5) shifts, assuming purely T delocalization, are likely to = -2.1 (6) be unreliable. This study is the first reported of the Because only the resonances of the a-, 6-, ortho-, nmr spectrum of a paramagnetic transition metal complex capmeta-, and methyl carbons were assigned with real able of showing multiple resonances, and although not certainty, the following discussion will be centered all of the carbon resonances were observed, it seems around these positions. It is readily seen that the predicclear that 13C contact shifts will be extremely useful and observed (uobsdCi/bobsdHi) values ted (uconc~/uconH~) in probing the nature of metal-ligand interactions. are in major disagreement for the p and meta positions whereas good agreement is obtained for uconCHa/ Acknowledgments. We are indebted to Dr. R. E. uConCHa. However, in all cases, the signs of the ratio, Benson of the Central Research Department of E. I. pC,")/pc: predicted to be negative because (pch" du Pont de Nemours and Company for samples of the is negative, of the carbon and corresponding proton ligand and complex and the reviewers for their helpful isotropic shifts are reproduced experimentally. No comments. experimental determinations for the ratio are possible David Doddrell, John D. Roberts Cor the a and ortho positions because P C - ~ " and p ~ - ~ " Contribution No. 4026 are unknown; however, if the calculated valuesza of The Gates and Crellin Laboratories of Chemistry these parameters are used it is readily seen that the California Institute of Technology Pasadena, Califorilia 91109 predicted ratios are too low. The discrepancies beReceiaed March 30, 1970 tween the predicted and observed ratio of contact shifts for carbon and hydrogen are sizable but might be regarded as tolerable in view of the approximate nature of the Karplus-Fraenkel treatmentsb and the The Structure of a Novel Lipid from the serious reservations that have been suggested for its Antibiotic Diumycin a p p l i ~ a t i o n . ~However, ~ in our view, the discrepancies Sir: fall into a pattern which indicates that there is an additional important mechanism contributing to the The isolation and characterization of a new group of isotropic shifts and involve transmission of spin density phosphorus-containing antibiotics, the diumycins, have by the u framework of the ligand. The argument been reported.]r2 The antibiotics are highly active in is based on the expectation that positive spin density uitro against gram-positive bacteria and exhibit a rearising from u delocalization at the B and metu position markable duration of action in ciuo. A single dose of would reduce the predicted upfield carbon shifts and 6.7 mg/kg of diumycin, administered subcutaneously to increase the downfield proton isotropic shifts.6 The mice 14 days prior to challenge with a lethal dose of overall result would be to reduce uconCi/uconHi for these Streptococcus pyogenes Czo3, provides protection to two positions, in agreement with experiment. The 50% of the mice injected.' reverse situation is expected for the a and ortho posiRecently, structures have been assigned to the options, again in accord with experiment. The methyl tically inactive lipids obtained by hydrolysis of the regroup should exhibit a ratio in better agreement with lated antibiotics prasinomycin3 and m ~ e n o m y c i n . ~ theory, because effect of u delocalization should be We now wish to present evidence for the structures of the markedly reduced at this position as the result of attenuation through the intervening bonds. Because a (1) E. Meyers, D. S . Slusarchyk, J. L. Bouchard, and F. L. Weisenborn, J . Antibior., 22,490 (1969). tetrahedral Ni(I1) complex has only unpaired electrons (2) The diumycins are members of a family of antibiotics that inof T symmetry available for interaction with the ligand, cludes: prasinomycin,za moenomycin,Zb 11,837RP,*c 8036RP,Id a likely pathway of introducing spin density into the u 19,402RP,le and macarbomycin.zf (a) F. L. Weisenborn, J. L. Bouchard, D. Smith, F. Pansy, G. Maestrone, G. Miraglia, and E. Meyers, framework would be an indirect T-u polarization. Nature (London), 213, 1092 (1967); (b) G. Huber, U. Schacht, H. L. Further evidence in support of positive spin density Wiedenmuller, J. Schmidt-Thome, I. Duphorn, and R . Tschesche, in the u framework of the ligand comes from previous Antimicrob. Ag. Chemother. 1965, 737 (1966); (c) Rhone-Poulenc, Bel-
+
+
+
+
(5) (a) H. M. McConnell, J . Chem. Phys., 24, 632, 764 (1956); Proc. Nar. Acad. Sci., 43, 741 (1957); (b) M. Karplus and G. K. Fraenkel, J . Chem. Phys., 35, 1312 (1961); G. K. Fraenkel, Pure Appl. Chem., 4, 143 (1962). (6) The predictions about the direction of the shifts are based on consideration of X-x configurational interactionsZawhich suggest that negative spin density at the carbon will produce positive spin density at the proton and a downfield shift.
Journal of the American Chemical Society 1 92:14
gian Patent 653,168 (1965); (d) Rhone-Poulenc, South African Patent 6516204 (1966); (e) Rhone-Poulenc, Netherlands Patent 68,02093 (1968); (f) S. Takahashi, A. Okanishi, R. Utahara, K. Nitta, I
