Comment on ``Spin and cyclotron energies of electrons in GaAs/GaAlAs quantum wells''

Abstract

In a recent publication, Pfeffer and Zawadzki [cond-mat/0607150; Phys. Rev. B 74, 115309 (2006)] attempted a calculation of electron g factor in III-V heterostructures. The authors emphasize that their outcome is in strong discrepancy with our original result [Ivchenko and Kiselev, Sov. Phys. Semicond. 26, 827 (1992)] and readily conclude that ``the previous theory of the g factor in heterostructures is inadequate''. We show here that the entire discrepancy can be tracked down to an additional contribution missing in the incomplete elimination procedure of Pfeffer and Zawadzki. This mistake equally affects their ``exact'' and approximate results. When the overlooked terms stemming from the nondiagonal Zeeman interaction between light hole and spin-orbit-split valence states are taken into account in the effective electron dispersion, the results of the both approaches applied to the three-level kp model become identical.