Predictive DFT-Based Approaches to Charge and Spin Transport in Single-Molecule Junctions and Two-Dimensional Materials: Successes and Challenges

Abstract

CONSPECTUS: The emerging field of flexible electronics based on organics and two-dimensional (2D) materials relies on a fundamental understanding of charge and spin transport at the molecular and nanoscale. It is desirable to make predictions and shine light on unexplained experimental phenomena independently of experimentally derived parameters. Indeed, density functional theory (DFT), the workhorse of first-principles approaches, has been used extensively to model charge/spin transport at the nanoscale. However, DFT is essentially a ground state theory that simply guarantees correct total energies given the correct charge density, while charge/spin transport is a nonequilibrium phenomenon involving the scattering of quasiparticles. In this Account, we critically assess the validity and applicability of DFT to predict charge/spin transport at the nanoscale. We also describe a DFT-based approach, DFT+Σ, which incorporates corrections to Kohn-Sham energy levels based on many-electron calculations. We focus on single-molecule junctions and then discuss how the important considerations for DFT descriptions of transport can differ in 2D materials. We conclude that when used appropriately, DFT and DFT-based approaches can play an important role in making predictions and gaining insight into transport in these materials. Specifically, we shall focus on the low-bias quasi-equilibrium regime, which is also experimentally most relevant for single-molecule junctions. The next question is how well can the scattering of DFT Kohn-Sham particles approximate the scattering of true quasiparticles in the junction? Quasiparticles are electrons (holes) that are surrounded by a constantly changing cloud of holes (electrons), but Kohn-Sham particles have no physical significance. However, Kohn-Sham particles can often be used as a qualitative approximation to quasiparticles. The errors in standard DFT descriptions of transport arise primarily from errors in the Kohn-Sham energy levels (self-energy errors). These errors are small in the strong-coupling regime where the molecular levels are significantly broadened at the Fermi level but are large in the coherent off-resonant tunneling regime where DFT overestimates conductance by orders of magnitude. The DFT+Σ approach uses a physically motivated, parameter free estimate of the self-energy corrections to correct the energy levels in DFT, giving conductance in quantitative agreement with experiment for a large but nonexhaustive class of single-molecule junctions. In 2D materials, the self-energy error is relatively small, and critical issues stem instead from the large length scales in experiments, which make it necessary to consider band-bending within the 2D material, as well as scattering due to electron-phonon interactions, spin-flip interactions, defects, etc.

Quantum Interference Enhanced Chemical Responsivity in Single‐Molecule Dithienoborepin Junctions

“…We predict a conductance that is a bit larger for DTB‐A, a behavior that disagrees with the experimental trend. However, these two compounds have very similar conductances and it is very likely that the usual errors of the DFT approximations prevent us from predicting the exact ordering . In general, we have a qualitative agreement as we predict the same order of magnitude for the conductance of both isomers, prior to fluoride ion capture.…”

Effect of H2O Adsorption on Negative Differential Conductance Behavior of Single Junction

“…Since the HOMO is much closer to the Fermi level than the LUMO, the electron-transport properties are mainly governed by the HOMO. According to the DFT + Σ method in the literatures 61–63 and the experimental results 60 , we can roughly estimate that, for the TADHA molecular junction without H 2 O adsorbate, the HOMO is overestimated by not more than 0.3 eV relative to the Fermi level, which is much smaller than those of weak coupling systems 61–63 .…”

“…According to the studies of Quek et al . 61–63 , the HOMO-LUMO (LUMO: lowest unoccupied molecular orbital) gap is usually underestimated in standard density functional theory (DFT), which results in overestimate of conductance of single molecular junction. Thus an alternative DFT-based approach 61–65 (e.g.…”

The low-bias conducting mechanism of single-molecule junctions constructed with methylsulfide linker groups and gold electrodes

“…In contrast, computer simulations can take into account the atomic details of molecular junctions and improve our understanding of their electronic transport properties. Although many theoretical investigations have been devoted to exploring the conducting mechanism of molecular junctions constructed using the thiol, pyridine, and amine groups, [34][35][36][37][38][39][40] only a few theoretical studies focus on the intrinsic properties of the molecule-electrode interfaces formed by the Au-SMe bonds. 31,41 Therefore, it is highly desirable to carry out first-principles calculations on representative molecular junctions constructed with the Au-SMe bonds.…”