“…Further, an energy-dependent correlation decreases the Hartree-Fock band gap by raising the valence-band energy and lowering the conduction-band energy. There is some empirical evidence that supports the idea that even in the first iteration (that is, using just the noninteracting Green's function G 0 ) one obtains quite accurate results for oneelectron properties such as the excitation energy Louie, 1985, 1986;Godby et al, 1986Godby et al, , 1987Godby et al, , 1988Aryasetiawan and Gunnarsson, 1998) and the quasiparticle lifetime (Campillo et al, 1999;Schö ne et al, 1999;Echenique et al, 2000;Campillo, Silkin, et al, 2000;Keyling et al, 2000;Silkin et al, 2001;Spataru et al, 2001). This is important for practical applications of the GW approach since, despite its formal simplicity, the practical solution of the self-consistent GW equations is a formidable task, which has been carried out only recently: self-consistent calculations were performed for the homogeneous electron gas (Holm and von Barth, 1998;Holm and Aryasetiawan, 2000;García-Gonzá lez and Godby, 2001), simple semiconductors, and metals (Shirley, 1996; Schö ne and Eguiluz, 1998).…”