Relativistic effects for NMR shielding constants in transition metal oxides using the zeroth-order regular approximation

Citation
R. Bouten et al., Relativistic effects for NMR shielding constants in transition metal oxides using the zeroth-order regular approximation, J PHYS CH A, 104(23), 2000, pp. 5600-5611
Citations number
46
Categorie Soggetti
Physical Chemistry/Chemical Physics
Journal title
JOURNAL OF PHYSICAL CHEMISTRY A
ISSN journal
10895639 → ACNP
Volume
104
Issue
23
Year of publication
2000
Pages
5600 - 5611
Database
ISI
SICI code
1089-5639(20000615)104:23<5600:REFNSC>2.0.ZU;2-Z
Abstract
Relativistic effects for NMR shielding constants have been calculated using the zero order regular approximation (ZORA) for relativistic effects. Isot ropic NMR shielding constants were obtained using density functional theory with gauge including atomic orbitals (GIAO) in a spin-fi ec formalism for the metal nuclei in transition metal oxides MO4n- (M = Cr, Mn, Fe, Mo, Tc, Ru, W, Re, Os) and carbonyl complexes M(CO)(6) (M = Cr, Mo, W). The ZORA is otropic shieldings are compared with results from an extended version of th e relativistic method employing the Pauli Hamiltonian developed earlier by Schreckenbach and Ziegler. Comparison between ZORA and Pauli shieldings, em ploying the restrictions necessary for the Pauli approach-frozen cores, res tricted basis sets in the core region-show the ZORA shieldings to be signif icantly different from Pauli ones, but the chemical shifts of the metal oxi des with respect to the carbonyl complexes do not differ much. However, ext ending the ZORA calculations (no frozen core, extended basis sets) gives si gnificant changes, proving the limitation to frozen cores and restricted ba sis sets of the Pauli method to be unwarranted. Comparison to experiment sh ows that the errors of the most precise ZORA chemical shifts are ca. 10% (i n the order of a few 100 ppm) for both the light and heavy transition metal s. Error sources may be, apart from deficiencies of the density functional, the neglect of spin-orbit coupling and the neglect of solvent effects.