Metal–CO photodissociation in transition metal complexes:: The role of ligand-field and charge-transfer excited states in the photochemical dissociation of metal–ligand bonds

Abstract

The photodissociation of a CO ligand from a series of d 6 metal–carbonyl complexes, with various other substituents (Cl, α-diimine, H, Mn(CO) 5) is discussed. The considerations are based on calculations of potential energy curves, using a density functional method. Particular attention is given to the long-standing question of the possible role of charge-transfer excited states in the photodissociation. Such CT states exist at relatively high energy (ca. 4 eV) in an unsubstituted complex like Cr(CO) 6, but at much lower energy if α-diimine ligands are introduced. For Cr(CO) 6, it is argued that the classical interpretation of the Cr–CO photodissociation upon excitation into the weak intensity low energy band in the spectrum at ca. 4 eV, due to the occupation of a ligand-field excited state having to be revised: the lowest excited state, from which the photodissociation occurs, has a charge-transfer character at R e. The LF excited states are much higher in energy. They decrease in energy rapidly, however, upon Cr–CO bond lengthening, so the dissociation is due to an avoided crossing of the CT state with a LF state. This photodissociation mechanism does not depend on the nature of the lowest excited state and is shown to be operative in a variety of d 6 systems Mn(CO) 5L (L=H, Cl or Mn(CO) 5). This immediately explains the relatively high yield of CO dissociation upon low energy excitation in Mn 2(CO) 10 and MnCl(CO) 5, even though the excitation is to an Mn–L σ* LUMO orbital. Mn–L bond breaking has very different quantum yields in these compounds, which is also explained from the electronic structure. Finally, equatorial substitution with an α-diimine ligand with low-lying π* orbitals generates MnCl(CO) 3( α-diimine). The CT states in this case are so low that excitation does not provide enough energy for CO dissociation if the atomic configuration is kept fixed. Nevertheless, dissociation can occur by an altogether different mechanism that involves relaxation of the Cl from axial to equatorial position upon equatorial CO departure.