2480
Journal of the American Chemical Society
of the Hebrew University for his critical reading of the manuscript.
References and Notes (1) Part 1: Hoz, S;Aurbach, D. Tetrahedron,in press. (2) For a review of cyclobutane stereochemistry, see: Moriarty, R. M. Top. Stereochem. 1974, 8, 271-423. (3) Meiboom, S.;Snyder, L. C. J. Chem. Phys. 1970, 52, 3857-3863. (4) Cox, K. W.; Harmony, M. D.: Nelson, G.; Wiberg, K. B. J. Chem. Phys. 1969, 50, 1976-1980. ( 5 ) Newton, M. D.; Schulman, J. M. J. Am. Chem. SOC. 1972, 94, 767-773. (6) Wiberg, K. B.; Peters, S. K.; Ellison, G. 5.; Alberti, F. J. Am. Chem. SOC. 1977, 99,3946-3951. (7) Bordwell, F. G.; Jarvis, 8 . B. J. Am. Chem. SOC. 1973, 95, 3585-3594. (8) The nonlinear least-squares program was written by R. Williamson and M. J. Goldstein of Cornell University and modified by M. Ben-Zion of this department. (9) Bartsch. R. A. Acc. Chem. Res. 1975, 8,239-245. (IO) Svoboda, M.; Hapala, J.; Zavada, J. Tetrahedron Len. 1972, 265-268. (11) Nickon, A.; Werstiuk, N. H. J. Am. Chem. SOC. 1967, 89, 3914-3915. (12) (a) Walborsky, H. M.; Motes, J. M. J. Am. Chem. SOC. 1970, 92, 2445-2450; (b) Cram, D.J.; Kingsbury, C. A.; Rickborn, B.; Haberfield, P. ibid. 1961, 83,3678-3687. (13) Bunnett, J. F. Tech. Chem. (N.Y.)1974, 6, 472-475.
1 101.9 1 April 25, 1979
(14) Zavada, J.; Svoboda, M. Tetrahedron Lett. 1972, 23-26. (15) Bartsch, R. A,: Mintr, E. A,; Parlman, R. M. J. Am. Chem. SOC. 1974, 96, 4249-4252. (16) Wiberg, K. 8.; Lampman, G. M. J. Am. Chem. SOC. 1966, 88, 44294433. (17) Bordweil, F. G.; William, J. B. J. Am. Chem. SOC. 1972, 94, 3907-3911; 1975, 97,3447-3452. (18) Zimmerman, H. E. J. Org. Chem. 1955, 20, 549-557. Zimmerman. H. E.; Nevins, T. E. J. Am. Chem. SOC. 1957, 79,6559-6561. Zimmerman, H . E.; Mariano, P. S. ibid. 1968, 90, 6091-6096. (19) Malhotra, S. K.; Johnson, F. J. Am. Chem. SOC.1965, 87, 5492-5493. Bordwell, F. G.; Yee, K. C. ibid. 1970, 92, 5939-5944. (20) Cremer, D. J. Am. Chem. SOC. 1977, 99, 1307-1309. (21) Baciocchi, E.; Corsano, S.:Ruzziconi. R. J. Chem. SOC.Perkin Trans. 2 1977,436-439. (22) von Schleyer, P. R.; Williams, J. E.; Blanchard, K. R. J. Am. Chem. SOC. 1970, 92, 2377-2386. (23) Klasinc, L.; Maksic, 2.; Randic, M. J. Chem. SOC.A 1966, 755-757. (24) Cram. D.J. "Fundamentals of Carbanion Chemistry", Academic Press: New York, 1965; pp 88-92. (25) Hall, K. H.; Blanchard, E. P.; Cherkofsky, S. C.: Sieja, J. B.;Shcppard. W. A. J. Am. Chem. SOC.1971, 93, 110-120. (26) Nevill, W . A ; Frank, D. S.: Trepka, R . D. J. Org. Chem. 1962, 27, 422428. (27) Gokel, G. W.; Cram, J.: Liotta, C. L.; Harris, H. P.; Cook, F. L. J. Org. Chem. 1974, 39, 2445-2446.
Communications to the Editor cis-Methyldiazene Sir: cis- Diazene (diimide) has been implicated as an intermediate in stereospecific reductions by in situ generated diazene' and in nitrogenase action on nitrogen.* This molecule and its rather well-studied trans isomer3y4 have been the objects of a number of theoretical studies5 Yet, only fragmentary and inconclusive evidence has been reported for the direct observation of the cis i ~ o m e r . W ~ .e~now report the isolation of the closest homologue of cis-diazene, cis-methyldiazene, and its infrared spectrum. Based on the assignment of this spectrum, we conclude that cis-diazene is yet to be observed. Hutton and Steel obtained cis-dimethyldiazene by photolysis of solid trans-dimethyldiazene with near-UV light at liquid nitrogen temperature.8 A similar experiment with transCH3N=ND9 has given a new substance that has properties appropriate to cis-CH3N=ND. Approximately 0.5 mmol of the trans isomer were distilled onto a CsI crystal which was held at -196 OC in a conventional low-temperature infrared cell. Prior to photolysis the glassy deposit was annealed into polycrystalline material. Photolysis was carried out for 6 h with illumination from a 45-W, low-pressure mercury lamp. New infrared absorption bands appeared a t 2180, 1560, and 1060 cm-I. The bands due to trans-CH3N=ND had decreased in intensity and had reverted toward the structureless appearance of glassy material. After the deposit was annealed again, the new features were seen more clearly and a small amount of noncondensible gas, presumably nitrogen, was pumped from the cell. Cautious cycling of the temperature between -196 OC and values progressing through the range of - 125 to - 1 13 0002-78631791 I501-2480$01 .OO/O
OC, as well as repeated evacuation of the cell (at - 196 "C), caused the relatively large amount of unconverted transCH3N=ND to sublime from the CsI plate to the surfaces of the surrounding copper support and liquid nitrogen well. The large difference i n volatility between the two isomers of methyldiazene is consistent with the boiling point difference of nearly 100 OC found for cis- and trans-dimethyldiazenes.Io Figure 1 gives the infrared spectrum of the residual deposit. Table I gives the vibrational assignment for cis-CH3N=ND in comparison with that of the trans i ~ o m e rThat . ~ many of the frequencies of the fundamentals of this molecule are nearly coincident with those of the trans isomer is reasonable and is supported by zero-order normal coordinate calculations.' Bands are present for methyl group stretching and bending, for N=N stretching, and for N D stretching and bending. The band due to N=N stretching is a much more intense feature of this spectrum than in that of the trans isomer. The largest frequency shift, a decrease of -120 cm-I, occurs in the N D stretching mode. Two satellite bands in the spectrum are attributed to hydrogen bonding. Several features are due to cis-CH3N=NH which was formed from the isotopic impurity present in the trans isomer. Several other possible products of the photolysis were considered but discarded. 'These included CH,jD, CH3CH3, trans-DN=ND, and trans-HN=ND, for which published spectra were available. Furthermore, these substances should have vaporized during the temperature cycling. The hydrazone of methyldiazene-dl, C H l = N N H D , which is a possible rearrangement product, does not fit the observed spectrum. This molecule has no methyl group and would have a C=N
0 I979 American Chemical Society
Communications to the Editor
248 1 Table I. Infrared SDectra of cis- and trans-CH1N=ND
cis isomer trans isomer" freq, i n t b freq, intb crn-l cm-I
?t
4000
3000
2000 FREOUENCY 1500(Chf')
1000
Figure 1. Infrared spectrum of polycrystalline cis-methyldiazene at -196 OC.
stretching frequency near 1600 cm-I, NH stretching above 3 100 cm-I, and C H stretching above 3000 cm-'.'O As a further test of the identification of the new material as cis-CH3N=ND, it was photolyzed again without changing the temperature from -196 "C. After photolysis for 1 h, bands due to the trans isomer were evident, especially in the N D stretching region between 2320 and 2250 cm-I. Photolysis for a second hour caused an increase in the intensity of the bands attributable to the trans isomer. After an annealing cycle, bands due to the trans isomer were even more apparent. The same sequence of experiments, including the rephotolysis step, was performed on trans-CD3N=ND with comparable results. However, when trans-CH3W=NH was tried, we were unsuccessful in isolating the cis isomer even though bands attributable to this species were observed after photolyzing and annealing the solid. As the temperature was cycled upward above - 120 "C, the intensity of the cis band a t 1570 cm-I decreased along with the intensity of the neighboring trans band. All bands were gone after the - 113 "C cycle. Thus, it appears that a narrow temperature range exists in which the kinetic isotope effect gives the ND-containing cis molecule sufficient additional stability to remain intact, while the much more volatile trans isomer sublimes. In the experiments with cis-CH3N=ND, it was finally lost from the CsI crystal a t -100 "C. The marked decrease in the N D stretching frequency in going from trans-methyldiazene to cis-methyldiazene is consistent with a correlation to which McKeanI2 and Bellamy and MayoI3 have drawn attention. These investigators have shown that a CH, N H , or O H bond that is oriented trans to an unshared electron pair on a neighboring atom has a significantly decreased vibrational frequency. This effect correlates with a weakened X H bond owing to interaction between the orbital of the unshared electron pair and the antibonding orbital of the X H bond. cis- Methyldiazene and cis-diazene have the requisite geometry for this effect: the trans isomers do not. Normal coordinate calculations for cis-diazene-do and -d2 gave 2994 and 2956 cm-' and 2195 and 2150 cm-l for N H and N D stretching frequencies, respectively.I4 Reasonable changes in force constants other than the one for N H ( D ) stretching, which had been fit to cis-methyldiazene-dl and -d4, did not cause significant changes in the frequencies of interest here. In view of these predicted frequencies, it is doubtful that cis-diazene has been observed to date. Rosengren and Pimentel assigned an infrared band at 3074 cm-l to cis-diazene as one of several products of the photolysis of matrix-isolated hydrazoic acid.6 However, the highest fundamental frequency for cis-diazene should be
