Electronic Transport as a Driver for Self-Interaction-Corrected Methods

Abstract

© 2015 Elsevier Inc. While spintronics often investigates striking collective spin effects in large systems, a very important research direction deals with spin-dependent phenomena in nanostructures, reaching the extreme of a single spin confined in a quantum dot, in a molecule, or localized on an impurity or dopant. The issue considered in this chapter involves taking this extreme to the nanoscale and the quest to use first-principles methods to predict and control the behavior of a few "spins" (down to 1 spin) when they are placed in an interesting environment. Particular interest is on environments for which addressing these systems with external fields and/or electric or spin currents is possible. The realization of such systems, including those that consist of a core of a few transition-metal (TM) atoms carrying a spin, connected and exchanged-coupled through bridging oxo-ligands has been due to work by many experimental researchers at the interface of atomic, molecular and condensed matter physics. This chapter addresses computational problems associated with understanding the behaviors of nano- and molecular-scale spin systems and reports on how the computational complexity increases when such systems are used for elements of electron transport devices. Especially for cases where these elements are attached to substrates with electronegativities that are very different than the molecule, or for coulomb blockade systems, or for cases where the spin-ordering within the molecules is weakly antiferromagnetic, the delocalization error in DFT is particularly problematic and one which requires solutions, such as self-interaction corrections, to move forward. We highlight the intersecting fields of spin-ordered nanoscale molecular magnets, electron transport, and coulomb blockade and highlight cases where self-interaction corrected methodologies can improve our predictive power in this emerging field.