A Comparative Relativistic DFT and Ab Initio Study on the Structure and Thermodynamics of the Oxofluorides of Uranium(IV), (V) and (VI)

Abstract

All the possible uranium(VI, V, IV) oxides, fluorides and oxofluorides were studied theoretically by using density functional theory (DFT) in the generalised gradient approximation (GGA), and three different relativistic methods (all-electron scalar four component Dyall RESC method (AE), relativistic small-core ECPs, and zeroth order regular approximation ZORA). In order to test different correlation methods, for the two former relativistic methods hybrid DFT, and, for the AE method, MP2 molecular orbital calculations were performed as well. Single-point AE-CCSD(T) energies were calculated on MP2 geometries as well. Energies of the uranium(VI) and (V) oxofluorides dissociation, uranium(VI) fluoride hydrolysis and oxofluoride disproportionation were calculated and compared against the available experimental thermochemical data. AE-CCSD(T) energies were the closest to the experiment. For GGA DFT methods, all the relativistic methods used yield similar results. For thermochemistry, the best quantitative agreement with the experimental and CCSD(T) values for both U=O and U-F bond strengths was obtained with hybrid DFT methods, provided that a reliable basis set was used. Both the GGA DFT and MP2 MO methods show overbinding of these bonds; moreover, this overbinding was found to be not uniform but strongly dependent on the coordination environment of the uranium atom in each case. U=O vibrational frequencies given by hybrid DFT, however, are systematically overestimated, and are better reproduced by GGA DFT; MP2 values usually fall in-between. Reaction enthalpies, U=O frequencies and complex geometries given by the PBE, MPBE, BPBE, BLYP and OLYP GGA functionals are quite similar, with OLYP performing slightly better than the others but still not as good as hybrid DFT. The geometries of the molecules are found to be influenced by the following factors: the inverse transinfluence (ITI) of the oxygen ligand and, for U(V), and U(IV), the Jahn-Teller distortion.

Binuclear Uranium(VI) Complexes with a “Pacman” Expanded Porphyrin: Computational Evidence for Highly Unusual Bis‐Actinyl Structures

“…[65,[69][70][71][72][73][74] Careful comparison with other relativistic methods as implemented in codes such as Gaussian 03 and ADF confirmed that Priroda is entirely reliable for actinide complexes. [65,[69][70][71] Very recently, investigations of binuclear UO 2 2 + ···M 2 + (An = U, Np and Pu; M = Mn, Fe, Co and Zn) complexes have given reasonable agreement with experimental results. [74] Scheme 1.…”

“…here to slightly underestimate bond strengths and corresponding stretching frequencies [69][70][71]86] ) and similar to those at 895-907 cm À1 for [{(thf)UO 2 ···MA C H T U N G T R E N N U N G (thf)}(L)] (M = Mn, Fe and Co), [25] 830 and 933 cm À1 for [UO 2 …”

Oxygen Exchange in Uranyl Hydroxide via Two “Nonclassical” Ions

“…12 As for issues (i) and (ii), hybrid DFT and small-core relativistic effective core potentials (ECPs) or all-electron relativistic treatments such as the zeroth-order regular approximation (ZORA) have proven fairly robust in describing the thermochemistry of actinides. 13 Because 1 is a closed-shell f 0 system, spin-orbit coupling can safely be neglected. Based on these considerations, there is little reason to doubt the accuracy, at least in a qualitative sense, of Shamov and Schreckenbach's results for 1.…”

“…For the thermochemistry of reactions involving the uranyl unit, hybrid functionals such as B3LYP tend to perform better than pure GGAs (such as BLYP). 13 Unfortunately, CPMD is impracticable with hybrids. Since our CPMD/BLYP-derived overall barrier is related to the relative stability of cis-and trans-uranyl hydroxide, however, to a first approximation we can use the static energetic data from Table 1 as incremental corrections to the barrier.…”

Structure and stability of aquotetrafluorouranyl(VI) in the solid state — Density functional study of [UO2F4(H2O)][NMe4]2·2H2O,

“…pure, gradient corrected functionals, which are sometimes outperformed by hybrid variants when it comes to reaction energies involving actinides. 26 As mentioned above, in the CPMD/PTI studies of aqueous uranyl complexes BLYP has performed quite well so far, but if it turns out to fail for the aquo tetrafluoro complex, the predictive power of this approach may be at stake. An utter failure would for instance be the incorrect prediction that 1a should also be unstable in the environment of a known, stable crystal.…”