The geometric and electronic structures of the title compounds are calculated with scalar relativistic, gradient-corrected density functional theory. The most stable geometry of ThCp(4) (Cp = eta(5)-C(5)H(5)) and UCp(4) is found to be pseudo-tetrahedral (S(4)), in agreement with experiment, and all the other AnCp(4) compounds have been studied in this point group. The metal-Cp centroid distances shorten by 0.06 A from ThCp(4) to NpCp(4), in accord with the actinide contraction, but lengthen again from PuCp(4) to CmCp(4). Examination of the valence molecular orbital structures reveals that the highest-lying Cp pi(2,3)-based orbitals split into three groups of pseudo-e, t(2) and t(1) symmetry. Above these levels come the predominantly metal-based 5f orbitals, which stabilise across the actinide series, such that in CmCp(4), the 5f manifold is at more negative energy than the Cp pi(2,3)-based levels. The stability of the Cm 5f orbitals leads to an intramolecular ligand-->metal charge transfer, generating a Cm(III) f(7) centre and increased Cm-Cp centroid distance. Mulliken population analysis shows metal d orbital participation in the e and t(2) Cp pi(2,3)-based orbitals, which gradually decreases across the actinide series. By contrast, metal 5f character is found in the t(1) levels, and this contribution increases four-fold from ThCp(4) to AmCp(4). Examination of the t(1) orbitals suggests that this f orbital involvement arises from a coincidental energy match of metal and ligand orbitals, and does not reflect genuinely increased covalency (in the sense of appreciable overlap between metal and ligand levels). Atoms-in-molecules analysis of the electron densities of the title compounds (together with a series of reference compounds: C(2)H(6), C(2)H(4), Cp(-), M(CO)(6) (M = Cr, Mo, W), AnF(3)CO (An = U, Am), FeCp(2), LaCp(3), LaCl(3) and AnCl(4) (An = Th, Cm)) indicates that the An-Cp bonding is very ionic, increasingly so as the actinide becomes heavier. Caution is urged when using early actinide/lanthanide comparisons as models for minor actinides/middle lanthanides.

The geometric and electronic structures of the title complexes have been studied using scalar relativistic, gradient-corrected density functional theory. Extension of our previous work on six-coordinate M[N(EPH 2) 2] 3 (M = La, Ce, U, Pu; E = O, S, Se, Te), models for the experimentally characterized M[N(EP (i)Pr 2) 2] 3, yields converged geometries for all of the other 4f and 5f metals studied and for all four group 16 elements. By contrast, converged geometries for nine-coordinate M[N(EPPh 2) 2] 3 are obtained only for E = S and Se. Comparison of the electronic structures of six- and nine-coordinate M[N(EPH 2) 2] 3 suggests that coordination of the N atoms produces only minor changes in the metal-chalcogen interactions. Six-coordinate Eu[N(EPH 2) 2] 3 and Am[N(EPH 2) 2] 3 with the heavier group 16 donors display geometric and electronic properties rather different from those of the other members of the 4f and 5f series, in particular, longer than expected Eu-E and Am-E bond lengths, smaller reductions in charge difference between M and E down group 16, and larger f populations. The latter are interpreted not as evidence of f-based metal-ligand covalency but rather as being indicative of ionic metal centers closer to M (II) than M (III). The Cm complexes are found to be very ionic, with very metal-localized f orbitals and Cm (III) centers. The implications of the results for the separation of the minor actinides from nuclear wastes are discussed, as is the validity of using La (III)/U (III) comparisons as models for minor actinide/Eu systems.