Inelastic transport In molecular junctions from first principles

Abstract

This work is dedicated to development of a first-principle approach to study electron-vibration interactions on quantum transport properties. In the first part we discuss a general implementation for inelastic transport calculations based on maximally localized Wannier functions and non-equilibrium Green's functions. Our approach is designed to determine inelastic transport properties such as differential conductances, inelastic tunneling spectroscopies and nonequilibrium vibrational populations. Our approach is first applied to benzene molecular junctions connected to cumulene and carbon nanotube electrodes. In these examples, we discuss the role of the multichannel effect and of parity selection rules on the polarity of conductance steps, and the appearance of a non-monotonic behavior in the vibrational population. In the second part, we extend our formalism to study the effect of the electron-vibration interactions on the local current distribution. Using non-equilibrium Green's functions, we derive an expression for the local distribution of the inelastic current. Applying this to the benzene-cumulene junction, we show that the electron-vibration interaction can lead to a locally inverted current direction and the formation of loop currents. In the third part, we present a comprehensive study of the elastic and inelastic transport properties of carbon nanotube-zigzag graphene nanoribbon junctions, as realized in recent experiments, focusing on the local current distribution over the junctions. We calculate the local distribution of the elastic current to visualize the current injection pattern from the CNT electrodes to the ZGNRs and the current path inside the ZGNRs. For inelastic transport properties, we find a similarity in the IETS peaks and the corresponding vibrational configurations for the CNT/ZGNR/CNT junctions with different widths. As observed in the benzene-cumulene junction, we find that the inelastic current emerges from a complex network that includes loop currents. Our method and implementation can be generalized to other types of interactions, and is not limited to the electron-vibration interactions. Thus our work will be a starting point to understand the role of different and diverse interaction effects on quantum transport, using realistic predictive first-principle calculations.