Four reaction pathways from diborane and ammonia to borazine, (HBNH)3, have been studied computationally at the density functional level (B3LYP/6-311+G(2d,p)//B3LYP/6-31G(d)). The cycloaddition of H2BNH2 to 1,3-diaza-2,4-diborabuta-1,3-diene and subsequent elimination of two molecules of H2 was found to be the lowest-energy pathway to (HBNH)3. In the other pathways, the formation and conversion of the intermediates 1,3,5-triaza-2,4,6-triborahexatriene, cyclotriborazane, and 1,3,5-triaza-2,4,6-triborahexa-1,5-diene into (HBNH)3 were investigated. The formation of 1,3-diaza-2,4-diborabuta-1,3-diene and, subsequently, the formation and electrocyclization of 1,3,5-triaza-2,4,6-triborahexatriene and the cycloaddition of H2BNH2 to 1,3-diaza-2,4-diborabuta-1,3-diene are predicted to be the kinetically favored pathways to (HBNH)3 in the gas phase. At low concentrations of 1,3-diaza-2,4-diborabutene, high concentrations of H2BNH2, and a temperature of 298.15 K, the formation of the polyolefins H3BNH2(H2BNH2)nNHBH2 (n=1,2) is predicted to be competitive with the formation of 1,3-diaza-2,4-diborabuta-1,3-diene.

2379Z > 2.58~(1) were used in the least-squares refinement.The structure was solved by using direct methods (MULTAN) and weighted difference Fourier methods. The positions of 20 non-hydrogen atoms were deduced from an E map. Subsequent difference Fourier calculations revealed the positions of the remaining non-hydrogen atoms. The quantity minimized by the leasbsquares program was Cw(volwhere w = 2.73/(u(F,J2 + (pFo)2). The analytical approximations to the scattering factors were used, and all structure factors were corrected for both the real and imaginary components of anomalous dispersion. Methyl carbon atoms C22424 were disordered over four sites, A-D, with occupancy factors of 0.38 (l), 0.20 (l), 0.12 (l), and 0.31 (l), respectively. In the final cycle of least-squares, all non-hydrogen atoms were independently refined with anisotropic thermal coefficients; for C22-C24, the 12 disordered positions were refined with separate isotropic thermal parameters. A group isotropic thermal parameter was varied for the methyl hydrogen atoms, which were fiied in "idealized" positions with C-H = 0.95 A, while the remaining hydrogen atoms were refined with independent isotropic thermal coefficients (hydrogen atoms attached to C22-C24 were excluded from the calculations; hydrogen atoms on the methyl groups of the mesityl ring were placed in arbitrary rotational conformations about the C 4 H B bonds). An empirical extinction coefficient was also varied to give a value of 1.7 (2) X lo-'. Successful convergence was indicated by the maximum shiftlerror of 0.23 in the last cycle. Final refinement parameters are given in Table 11. The final difference Fourier map had no significant features, and there were no apparent systematic errors among the final observed and calculated structure factors. (1988)(1989)(1990). We particularly thank Dr. Scott Wilson and Charlotte Stern of the University of Illinois X-Ray Crystallographic Laboratory for performing the X-ray crystal structure determinations.Registry No.

No abstract

No abstract