Infrared intensities of : a true challenge for DFT methods

Absolute infrared intensities of $$ were evaluated with a great variety of DFT and ab initio methods and basis sets. It is shown that the intensities calculated by different levels of theory may not agree with each other even in the qualitative (weak/strong) sense. Geometrical parameters, as well as net atomic charges evaluated from multiple partition schemes, did not vary as much as the intensities and thus cannot explain the tremendous differences found for the latter. As there are no experimental estimates for the intensities to guide the theoretical evaluation, deciding on the best level of theory is reduced to comparisons between the different DFT methods and QCISD or CCSD, believed to be the best theoretical estimates in the set. The differences found among the various DFT methods suggest the development of new methods, instead of converging to a focal point, is rather diverging.