Comment on: Accuracy of a b i n i t i o C–H bond length differences and their correlation with isolated C–H stretching frequencies

Publisher Website
Google Scholar
Abstract
Relative C–H bond distances in n‐alkanes, obtained from a b i n i t i o geometry optimizations with the 4‐31G, 4‐21G, and 5‐31G** basis sets, are compared with experimental isolated C–H stretching frequency data. It is found that results from different calculations all give essentially linear frequency‐bond length correlations, but with different slopes. For each of the sets of calculated C–H bonds investigated, the slope of the frequency‐bond correlation is too small compared to the equation given by McKean [J. Mol. Struct. 1 1 6, 313 (1984)] for correlations between isolated C–H stretching frequencies and experimental r 0 bonds. Thus it is concluded that, in contrast to work previously published by Aljibury e t a l. [J. Chem. Phys. 8 4, 6872 (1986)] and Snyder e t a l. [J. Chem. Phys. 8 1, 5352 (1984)], precise correlation between calculated bond lengths and isolated C–H stretching frequencies does not determine the relative differences between the various C–H bonds in n‐alkanes to ±0.0001 Å.