Transuranic Computational Chemistry

A covalent chemical bond carries the connotation of overlap of atomic orbitals between bonded atoms, leading to a buildup of the electron density in the internuclear region. Stabilization of the valence 5f orbitals as the actinide series is crossed leads, in compounds of the minor actinides americium and curium, to their becoming approximately degenerate with the highest occupied ligand levels and hence to the unusual situation in which the resultant valence molecular orbitals have significant contributions from both actinide and the ligand yet in which there is little atomic orbital overlap. In such cases, the traditional quantum-chemical tools for assessing the covalency, e.g., population analysis and spin densities, predict significant metal-ligand covalency, although whether this orbital mixing is really covalency in the generally accepted chemical view is an interesting question. This review discusses our recent analyses of the bonding in AnCp3 and AnCp4 (An = Th-Cm; Cp = η(5)-C5H5) using both the traditional tools and also topological analysis of the electron density via the quantum theory of atoms-in-molecules. I will show that the two approaches yield rather different conclusions and suggest that care must be taken when using quantum chemistry to assess metal-ligand covalency in this part of the periodic table. The implications of this work for minor actinide partitioning from nuclear wastes are discussed; minor actinide extractant ligands based on nitrogen donors have received much attention in recent years, as have comparisons of the extent of covalency in actinide-nitrogen bonding with that in analogous lanthanide systems via quantum-chemical studies employing the traditional tools for assessing the covalency.

Chlorine K-edge X-ray absorption spectroscopy (XAS) and ground-state and time-dependent hybrid density functional theory (DFT) were used to probe the electronic structures of O(h)-MCl(6)(2-) (M = Ti, Zr, Hf, U) and C(4v)-UOCl(5)(-), and to determine the relative contributions of valence 3d, 4d, 5d, 6d, and 5f orbitals in M-Cl bonding. Spectral interpretations were guided by time-dependent DFT calculated transition energies and oscillator strengths, which agree well with the experimental XAS spectra. The data provide new spectroscopic evidence for the involvement of both 5f and 6d orbitals in actinide-ligand bonding in UCl(6)(2-). For the MCl(6)(2-), where transitions into d orbitals of t(2g) symmetry are spectroscopically resolved for all four complexes, the experimentally determined Cl 3p character per M-Cl bond increases from 8.3(4)% (TiCl(6)(2-)) to 10.3(5)% (ZrCl(6)(2-)), 12(1)% (HfCl(6)(2-)), and 18(1)% (UCl(6)(2-)). Chlorine K-edge XAS spectra of UOCl(5)(-) provide additional insights into the transition assignments by lowering the symmetry to C(4v), where five pre-edge transitions into both 5f and 6d orbitals are observed. For UCl(6)(2-), the XAS data suggest that orbital mixing associated with the U 5f orbitals is considerably lower than that of the U 6d orbitals. For both UCl(6)(2-) and UOCl(5)(-), the ground-state DFT calculations predict a larger 5f contribution to bonding than is determined experimentally. These findings are discussed in the context of conventional theories of covalent bonding for d- and f-block metal complexes.