Comment on "Stability of conductance oscillations in monatomic sodium wires"

The discussion in a recent paper by Khomyakov and Brocks [Phys. Rev. B 74, 165416 (2006)] suggests that it is the use of a uniform-background (jellium) model for the electrodes, which are linked by a short chain of Na atoms, in Lang [Phys. Rev. Lett. 79, 1357 (1997)] that leads to a calculated conductance for the Na wire appreciably smaller than the quantum unit (particularly for the one-atom wire), in disagreement with experiment. These calculations, however, used electrodes with a fairly high electron density, roughly that of Al, and so they should not be expected to reproduce the results of experiments that are more appropriately modeled using electrodes whose density corresponds to that of Na (which is much lower than that of Al). © 2007 The American Physical Society.