Resumen de Fun with Computational Chemistry: Solving Spectral Problems Using Computed 13C NMR Chemical Shifts. A Comparison of Empirical and Quantum Mechanical Methods
D. A. Forsyth, Leon J. Tilley, Shawn J. Prevoir
The viability of using 13C NMR shift prediction as an aid to solving undergraduate spectral problems was tested. The chemical shifts of structures corresponding to four typical problems were calculated on a PC using two approaches: (i) a quantum mechanical method for calculating carbon shieldings and (ii) empirically based shift predictions. The quantum mechanical approach used GIAO B3LYP/3-21G (X, 6-31+G*) calculations of isotropic shieldings based on molecular mechanics geometries (MM3). Four commercial programs were used for the empirical predictions. The average shift differences between the predicted and known values were compiled and used to determine the correct structure. Results indicated that the quantum mechanical method could be used to identify the correct structure in all cases. The empirically based shift prediction software was much less effective. In particular, the important issue of stereochemistry is ignored in most of the empirically based predictions.
The process of computing chemical shifts is fun and relatively easy to follow. While primarily serving as an aid in solving problems, it can also reinforce visualization and molecular modeling skills and pique student interest in computational and other branches of chemistry.
