Ligand Field Theory for Linear ML2Complexes

Abstract

AbstractPoint‐charge crystal field theory (CFT) for linear ML2systems guarantees dz2>dxz/dyz>dxy/dx2‐y2 (i. e. dσ>dπ>dδ). This is not what is found for CuCl2and other linear, divalent MX2complexes where dπ>dσ. This failure of CFT has also been attributed to its successor, ligand field theory (LFT). However, taking LFT to be parameterised CFT, any d orbital sequence is possible. The real test is whether the LFT parameters make chemical sense. A new cellular ligand field (CLF) analysis of CuCl2, including the equatorial toroidal void, which affects all the d orbitals, confirms eπ(Cl)>eσ(Cl). To account for this, the ‘distributed donation’ model is proposed linking σ and π bonding such that enhanced π donation is at the expense of reduced σ bonding and vice versa. This is the π‐donor analogue of the Dewar‐Chatt‐Duncanson model for π‐acceptor ligands. Ab initio LFT (AI LFT) calculations support a progressive π‐donation increase/σ donation decrease from Cu(II) to Sc(II) with eσfinally becoming negative. The model further predicts that adding equatorial ligands will force the π density back into the σ bond which is also confirmed by AI LFT calculations fortrans‐[CuCl2(NH3)4] where eσ(Cl)>eπ(Cl) even though the CuCl2unit is unchanged. The distributed donation model for linear σ‐only [MII(Lσ)2]nsystems (Lσ=NH3, H), where increasing π donation is not possible, implies that the M−L bonding in linear systems should be inherently weak.