Metallaboranes of the earlier transition metals: Relevance to the cluster electron counting rules

Structures of the known hypoelectronic metallaboranes of tungsten and rhenium are used to explore the applicability of the electron counting rules to metallaborane clusters containing earlier transition metal fragments. It is demonstrated that for metallaboranes with formal skeletal electron pairs (sep) less than the p + 1 closo count, p = the number of vertices, the cluster shapes adopted are intermediate between the most spherical deltahedral shapes found in the closed borane anions (canonical structures) and the multicapped skeletons found in metal systems. For a given cluster size the total connectivity, or number of bonded edges, is a constant and independent of the skeletal electron count. What varies is the distribution of vertex connectivities with the hypoelectronic skeletons exhibiting a greater number of high (occupied by the metal atoms) and low (occupied by the boron atoms) connectivity vertices. Consequently, as the number of sep falls below the p + 1 count, the average coordination number of the metals increases and the average coordination number of the boron atoms decreases. For each missing sep, the observed shape is related to the canonical form by one, in some cases more than one, diamond-square-diamond rearrangement or a cross cage metal-metal bond. In the specific case of the rhena- and tungstaboranes, it is shown that two of the missing sep are delocalized over the cluster bonding network and one is localized on the metal centers in the form of a Re-Re bond.