Dynamical Corrections to the DFT-LDA Electron Conductance in Nanoscale Systems

Abstract

Sai et al. Reply: In Ref.[1], we demonstrated that there exists a nontrivial correction, arising from the viscosity of the electron liquid [2], to the conductance of nanoscale junctions calculated within the adiabatic local-density approximation to time-dependent density-functional theory (DFT). This dynamical correction cannot be captured by any static DFT functional, even the exact one. To provide an estimate of these effects, we derived Eq. (14) for the dynamical viscous resistance and evaluated that ex

Electron transport through single conjugated organic molecules: Basis set effects in ab initio calculations

“…Regarding the comparison between theory and experiment, however, large discrepancies have been found for some conjugated molecules between theoretical predictions 13,14,15,16,17 and experimental conductances. 1,2,3,4,5,6 Many efforts have been made on the theoretical side 18,19,20,21,22,23 to understand this discrepancy and several possible reasons have been discussed. For example, the self-interaction error (SIE) and the underestimation of the transport gap in density functional theory (DFT) based ab initio calculations, 21,24 and the neglect of the dynamical correlation effect in the non-interacting Landauer formalism.…”

“…For example, the self-interaction error (SIE) and the underestimation of the transport gap in density functional theory (DFT) based ab initio calculations, 21,24 and the neglect of the dynamical correlation effect in the non-interacting Landauer formalism. 20,22 In this paper, we investigate a fundamental technical issue in all ab initio transport calculations: the effect of the basis set. We find that this effect can also cause a significant change, up to orders of magnitude, in the resulting conductance, depending on the molecular feature and the molecule-lead coupling strength.…”

Time-dependent density functional theory for quantum transport

“…Despite the success, the ALDA approximation is expected to become inadequate for addressing transient dynamics of open electronic systems in circumstances where electron correlations dominate. Moreover, the frequency dispersion part of the XC potential is completely missing from an adiabatic XC functional such as ALDA, 18 which may lead to loss of crucial transient features in the electronic dynamics. The AWBL approximation for dissipation functional may lead to nontrivial errors for electrodes of finite band widths and strongly inhomogeneous energy bands, or when non-Markovian memory effects play a significant role (such as in multi-channel conductors).…”

Current-Induced Effects in Nanoscale Conductors

“…We are implicitely assuming that static DFT gives a reasonable account of the scattering properties of a nanoscale system, at least in linear response. While this may be true for metallic junctions it is not obvious for molecular junctions [18]. However, in linear response and far from the resonant regime, we expect basic physical trends for these systems to be reproduced well by static DFT [18].…”

“…While this may be true for metallic junctions it is not obvious for molecular junctions [18]. However, in linear response and far from the resonant regime, we expect basic physical trends for these systems to be reproduced well by static DFT [18]. The full Hamiltonian of the system is H = H 0 + V , where H 0 is the Hamiltonian due to the bare biased electrodes, and V is the scattering potential.…”