“…Generally, the agreement with experiment increases from WF1 to WF2 to WF3, and all parameters, apart from the harmonic bending frequency, 2 , are within 0.5% of the experiment at the final WF3 level (with WF1 and WF2 better values of 2 were obtained than with WF3). The accuracy of WF3 results other than 2 is comparable with CCSD(T)/cc-pVQZ numbers (41,42). We have performed the calculation of the shielding surfaces for three functions, WF1, WF2, and WF3; thus we may estimate the relationship between the differences in the potential and the differences in the rovibrational effects on the shielding.…”