Experimental and theoretical investigations of CB8-: towards rational design of hypercoordinated planar chemical species.

We demonstrated in our joint photoelectron spectroscopic and ab initio study that wheel-type structures with a boron ring are not appropriate for designing planar molecules with a hypercoordinate central carbon based on the example of CB(8), and CB(8)(-) clusters. We presented a chemical bonding model, derived from the adaptive natural density partitioning analysis, capable of rationalizing and predicting planar structures either with a boron ring or with a carbon atom occupying the central hypercoordinate position. According to our chemical bonding model, in the wheel-type structures the central atom is involved in delocalized bonding, while peripheral atoms are involved in both delocalized bonding and two-center two-electron (2c-2e) sigma-bonding. Since carbon is more electronegative than boron it favors peripheral positions where it can participate in 2c-2e sigma-bonding. To design a chemical species with a central hypercoordinate carbon atom, one should consider electropositive ligands, which would have lone pairs instead of 2c-2e peripheral bonds. Using our extensive chemical bonding model that considers both sigma- and pi-bonding we also discuss why the AlB(9) and FeB(9)(-) species with octacoordinate Al and Fe are the global minima or low-lying isomers, as well as why carbon atom fits well into the central cavity of CAl(4)(2-) and CAl(5)(+). This represents the first step toward rational design of nano- and subnano-structures with tailored properties.