Theoretical studies of heavy-atom magnetic shielding in some small polyatomic molecules

Abstract

Perturbed Hartree-Fock calculations are performed on the molecules CH4, C2H2, C2H4, C2H6, CH3F, CH3OH, H2O, NH3, and C6H6 to obtain the heavy-atom NMR shielding constants and magnetic susceptibilities. Slater-type atomic orbitals (STO's) are used as the basis set. Various aspects of this model, including the dependence on basis-set size and on the origin of the vector potential, are examined. The results are compared with the available experimental data and with other ab initio and semiempirical treatments.

Coupled-Hartree-Fock calculations of susceptibilities and magnetic shielding constants

“…[13][14][15]. We were mainly interested in the question concerning the number of polarization functions.…”

“…Ab initio calculations have been performed by Ditchfield et al [t6] and by Appleman et al [15] using conventional atomic basis sets, and by Ditchfield [12] using GIAOs. Only the latter computations give good results, especially for the shifts relative to CH4.…”

“…In recent years several theoretical investigation on magnetic properties of molecules have been published [1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17]. From these calculations it is clear that the situation is much more complicated than in the case of the computation of molecular energies in the absence of external fields.…”

Calculations of the magnetic shielding constants of heavy nuclei in polyatomic molecules

“…As was previously found by other authors [11,12], double-zeta sets are still insufficient to achieve accurate shielding values. We have not attempted to improve Set I11 with polarization functions since this might introduce some unbalancing of the entire basis and other disagreeable features.…”

“…It is interesting to note that, in the case of ammonia, all the components of the experimental paramagnetic shielding at 14N, which have been recently determined by spin-rotation techniques [33], are reproduced with great accuracy: assuming the gauge on 14N and applying the formulas [lld], we got c+:,(N)= a;,(N) = -76.57 ppm; uiz(N) = -115.37 ppm, the corresponding experimental values being -76 and -117 ppm [33]. For methane we computed an average total shielding of 195.96 ppm, compared to the "best experimental value" 196zt10ppm inferred by Appleman et al [12]. Our theoretical value for 1 7 0 shielding in the water molecule is about 17 ppm lower than the experimental one, 334 ppm, according to the same authors [33] …”