Erratum: Nonvanishing spin Hall currents in disordered spin-orbit coupling systems [Phys. Rev. B<b>71</b>, 041304(R) (2005)]

K. Nomura,Jairo Sinova,T. Jungwirth,Q. Niu,A. H. MacDonald
DOI: https://doi.org/10.1103/physrevb.73.199901
IF: 3.7
2006-01-01
Physical Review B
Abstract:In Ref. 1 we reported on a numerical study of the influence of disorder on the spin Hall conductivity of the k-linear Rashba model, concluding incorrectly that it remains finite in the thermodynamic limit. Our original conclusion is at odds with new numerical findings2 in which the extrapolation to infinite system size is carried out systematically. The new numerical results are consistent with the conclusion from perturbation theory calculations that the dc spin Hall conductivity of this model is zero.3,4 These numerical studies are complicated by strong frequency dependence and large fluctuations in the spin Hall conductivity values obtained in finite-size calculations. The earlier calculations erred by using the frequency dependence of the longitudinal conductivity, which has corrections that vary as 2 compared to the 1 dependence of the spin Hall conductivity discussed below, to judge whether or not the dc limit was reached. The new findings in Ref. 2 supersede the conclusions reached in Ref. 1 with regards to the thermodynamic dc limit of the k-linear Rashba model. The findings in Ref. 2 resolve the controversy that had been associated with the linear Rashba model. The expression used for numerical evaluation of the spin Hall conductivity is
What problem does this paper attempt to address?