Calculation of 33S isotropic nuclear shielding

After some sparse papers published by different authors in the 1970s (see e.g. refs. 111-113), in 1987 Tossel et al.114 performed ab initio CHFPT (coupled Hartree-Fock perturbation theory) calculations of Gd and ap/ the diamagnetic and 

As a general rule, ab initio methods reproduce the 33S experimental chemical shift trend well, independently of the level of approximation used, provided that sufficiently expanded basis sets are used (according to Schindler, basis sets should include at least two sets of polarization functions for sulphur115). However, in many cases the introduction of electron correlation is mandatory. When this is the case, DFT methods117 are usually preferred to Moeller-Plesset (MP) and coupled-cluster calculations because they are less time-consuming and can also be easily applied to complex molecular systems. 

To avoid this problem, Chesnut119 has proposed the use of a scaling factor derived by a least squares fit of DFT-calculated crp against the difference between the observed isotropic shielding and the calculated xd. The relationship derived for values calculated using the GIAO approach at B3LYP/6-311+G(2d,p) level is 

Jackowski et al.120 have examined the temperature dependence of the rovibrational corrections to the 33S nuclear shielding constant in OCS by the GIAO method used at the MCSCF level with different basis sets. In the temperature range 170—420 K, the rovibrational correction has a deshielding effect that depends slightly on the basis set (—15 to —18.5 ppm ca. and —12.5 to —14.5 ppm ca. with two different basis sets). 

Some interesting applications of theoretical calculations of 33S shielding constants have been reported in the literature. 

---