Computational studies of the dimerization of the diboradiazacyclobutadienes (HBNH) 2 and (MeBNMe) 2 to form the tetraboratetraazacyclooctatetraenes (RBNR) 4 suggest that the preferred pathway involves "face-to-face" dimerization of the four-membered ring, followed by asynchronous, one-step scission of the two transannular BN bonds to form the eightmembered monocyclic product. Detection of the proposed intermediates in either system is unlikely, as they lie in shallow potential energy wells. That in certain situations the fourand eight-membered rings establish equilibria is consistent with the comparatively small energy difference between (MeBNMe) 2 and (MeBNMe) 4 .Iminoboranes RBtNR′ prefer to oligomerize rather than react with other molecules. 1-3 Paetzold 4 and coworkers, among others, 5 have examined in detail the relationship between the size of R/R′ and the degree of oligomerization, finding in broad terms an inverse correlation, as one would expect. However, inexplicable subtleties exist, and these are readily seen in the behavior of the iminoborane tetramers, 6 here termed tetraboratetraazacyclooctatetraenes. 7 For example, NMR studies show that equilibria between the iminoborane dimers (diboradiazacyclobutadienes) and the tetramers can occur at or somewhat above room temperature for R ) Me, R′ ) tBu 8 and for R ) R′ ) iPr 9 but not for R ) nPr, R′ ) iPr, R ) R′ ) iBu, or several other combinations. 10 The mechanism by which the equilibria are established remains lightly studied. Paetzold proposed the mechanism shown in Scheme 1. 4 The diboradiazacyclobutadiene 1 dimerizes to the tricyclic "Dewar cyclooctatetraene" analogue 2. This undergoes sequential breakage of the internal transannular BN bonds to form first fused bicyclic 3 and then the product tetraboratetraazacyclooctatetraene 4. All steps proceed reversibly, so that equilibria between the four species can occur. Experimentally, however, only the end points 1 and 4 were observed, suggesting that 2 and 3, if they are intermediates, lie in very shallow potential energy wells.Close inspection of the mechanism suggests several issues meriting computational examination. First, does the diboradiazacyclobutadiene dimerization occur "edgeto-edge", to give a staircase-like tricycle, or "face-to-face", to give a tub-like tricycle? Can these tricycles equilibrate, possibly through decomposition back to the fourmembered rings and then recombination? Second, need the internal bond breaking occur stepwise, or can it occur synchronously (or nearly so)? Third, given that rather strong bonds (ca. 88-90 kcal mol -1 ) 11,12 must break to convert the tricyclic intermediate to the bicyclic intermediate and monocyclic product, why are the proposed intermediates 2 and 3 so difficult to observe? Are the wells they lie in really so shallow?We report here an examination of these questions through calculations of the structures and energetics of the parent perhydrogen and permethyl homologues represented in Scheme 1 and the transition states connecting them. We find that t...