Comment on “Critique of the foundations of time-dependent density-functional theory”

Abstract

Two particular subjects tackled in the recent paper by Schirmer and Dreuw (SD) [J. Schirmer and A. Dreuw, Phys. Rev. A 75, 022513 (2007)] are commented on by us. By providing a convincing example, we answer positively the question posed by SD whether a local operator can be reconstructed from the knowledge of its particle-hole matrix elements (when the number of particle states exceeds one). In connection with the problem of the utmost reduction of the density-functional theory (DFT) to a solution of one equation for the density amplitude—the radical Kohn-Sham (rKS) approach—used by SD as a valuable tool to investigate the time-independent and time-dependent DFT, we demonstrate that the calculational scheme implementing the rKS formulation is rather uncompetitive with the traditional KS scheme.