A question of bonding

J. Am. Chem. Soc. 133, 14168–14171 (2011)

Clusters or cages of main-group elements that incorporate transition metal atoms have recently interested chemists because of their unusual shape and bonding, as well as their similarity to Zintl-type phases. These generally fall into two groups: cages with an encapsulated transition metal dimer or alloy-like clusters.

Credit: © 2011 ACS

Now, Stefanie Dehnen from the Philipps University of Marburg and colleagues have made an endohedral cluster anion where three atoms of palladium are held by a 14-atom tin–bismuth cage: [Pd₃Sn₈Bi₆]⁴⁻. The triangularly arranged palladium atoms are close enough to each other to suggest bonding interactions. The tin–bismuth cage is similar in structure to some previous examples, but these enclosed either no atoms or just one. To account for the −4 charge of the cluster, Dehnen and colleagues considered two bonding models. In one, the Pd₃ centre is neutral and the tin–bismuth cage has two extra valence electrons: [Pd₃]⁰[Sn₈Bi₆]⁴⁻. In the other model, the Pd₃ centre is a dianion: [Pd₃]²⁻[Sn₈Bi₆]²⁻.

To test the two models, Dehnen and colleagues calculated the difference between the electron densities of the anion as a whole and the sum of the constituent parts of each of the two models (pictured; blue shows the model gives a relative excess of electron density and red shows a relative lack). For the model with the neutral palladium centre (left), there were places with too much or too little electron density on the surface of the cages, whereas the dianion model (right) shows only a lack. As might be expected, modelling the Pd₃ centre as a dianion seems to give too much electron density to this area. Although neither model is perfect, Dehnen and colleagues suggest that the [Pd₃]⁰ gives the better fit.