Group VB to Transition Metal Bonds

Abstract

The structure of this introductory essay requires some explanation, because for much of the discussion the behaviour of platinum complexes of phosphorus-containing ligands is taken as the starting point. Considerable emphasis on phosphorus-containing ligands is unavoidable: parameters such as bond lengths, stretching frequencies, n.m.r. shifts and coupling constants and many more that bear reasonably directly on the question of the nature of the metal-ligand bonds are very much more plentiful for phosphorus ligands than for the ligands containing the heavier Group VB elements. Also, the study of platinum compounds has been very closely associated with the development of theories of metal-Group VB element bonding. During the 1940s some similarity had been noted between carbon monoxide, phosphorus donors and arsenic donors in their ability to form complexes with transition elements. In the late 1940s J. Chatt had the idea, based on Pauling’s theory of metal-carbonyl double bonding, that ligands such as phosphines had vacant d-orbitals which could accept electrons from filled metal d-orbitals, thus forming a ‘dative π-bond’ in addition to the ligand-metal σ-bond. This transfer of electron density from the metal to ligand would be promoted by electronegative groups on phosphorus, so Chatt tested the theory by studying the interaction of phosphorus trifluoride with platinum(II) chloride and boron trifluoride. Platinum, which has suitable d-electrons, formed stable complexes, whereas boron, which has no electrons available for dative π-bond formation, did not have any significant interaction with phosphorus trifluoride.