The effects of including (a) implicit solvent in geometry optimizations, (b) conformationally flexible molecules in test sets, and (c) empirical dispersion D3(BJ) on scaling factors for predicting 1 H and 13 C NMR chemical shifts were explored. Scaling factors with optimizations performed in the gas phase and with a Polarizable Continuum Model (PCM) solvent model were obtained for 12 organic solvents, including 2,2,2-trifluroethanol and chlorobenzene, for which scaling factors have been developed for the first time. Scaling factors for aromatic solvents were split into primary and secondary scaling factors to account for CH-π effects. Including empirical dispersion D3(BJ) did not lead to significant improvement.
Keywords: 13C; 1H; NMR; computational NMR; density functional theory.
© 2019 John Wiley & Sons, Ltd.