Bĝldea and Köppel have generalized our correlated transport formalism to constrain the occupancies of single-electron orbitals, and then claimed that zero conductivity G=0 in linear response follows and that our method is invalid. We show here that it was their incorrect choice to constrain the occupancies of real orbitals that resulted in G=0. In a scattering state incident to a barrier with values at ± differing by eV, only the incoming plane wave component has the bias-independent amplitude of 1, while the outgoing has r which depends on the voltage V and the barrier geometry. Thus, if occupancies are to be constrained to bias-independent values in transport, those of complex orbitals such as an incoming plane wave on each side are suitable, while real orbitals such as sines are not. This is true whether constant current constraints are additionally imposed or not.
|Physical Review B - Condensed Matter and Materials Physics
|Published - 11 Jul 2011
ASJC Scopus subject areas
- Electronic, Optical and Magnetic Materials
- Condensed Matter Physics