Comment on “The hyperpolarizability of <i>trans</i>-butadiene: A critical test case for quantum chemical models” [J. Chem. Phys. <b>106</b>, 1827 (1997)]

Abstract

It has been suggested that the random phase approximation is superior to many-body/coupled cluster methods for calculating the second hyperpolarizability of trans-butadiene and, by extension, other π-conjugated species. We show that this conclusion is premature because of basis set inadequacies; approximations in the treatment of frequency dispersion; and omission of vibrational effects. We also hypothesize that trans-butadiene may not provide the most critical test of various computational approaches.