GIAO 13C NMR Calculation with Sorted Training Sets Improves Accuracy and Reliability for Structural Assignation

Abstract

GIAO 13C NMR calculation plays important roles in structural assignment for small organic molecules. Especially, higher accuracy and confidence are required for the structural assignation of complex organic molecules. In this GIAO 13C NMR calculation strategy, carbons were sorted according to their type of hybridization, radii of solvation cavity, or solvent interactions. The calculated shielding tensors of carbons in each sorted training set were linear-regressed with experimental data separately, and the obtained linear parameters were used to convert calculated shielding tensors into calculated chemical shifts. This approach shows significantly improved accuracy, especially for sp2 carbons, comparing to conventional GIAO 13C NMR calculation protocols. A statistic-based probability algorithm was proposed to quantify the reliability of structural assignation. With empirical linear parameters for three commonly used NMR solvents as well as an easy-to-use script and a spreadsheet, this 13C NMR calculation protocol provides a useful tool for structural validation or assignation of ambiguous organic structures.