Computational study of oxide stoichiometry and variability in the Al/AlOx/Al tunnel junction

Aluminium tunnel junctions are key components of a wide variety of electronic devices. These superconducting tunnel junctions, known as Josephson Junctions (JJ's) are one of the main components of superconducting qubits, a favourite qubit technology in the race for working quantum computers. In this simulation study our JJ configurations are modelled as two aluminium electrodes which are separated by a thin layer of amorphous aluminium oxide. There is limited understanding of how the structure of the amorphous oxide barrier affects the performance and shortcomings of JJ systems. In this paper we present a computational study which combines molecular dynamics, atomistic semi-empirical methods (Density Functional Tight Binding) and non-equilibrium Green's function to study the electronic structure and current flow of these junction devices. Our results suggest that the atomic nature of the amorphous barrier linked to aluminum-oxygen coordination sensitively affects the current-voltage ( ) characteristics, resistance and critical current. Oxide stoichiometry is an important parameter that can lead to variation in resistance and critical currents of several orders of magnitude. The simulations further illustrate the variability that arises due to small differences in atomic structure across amorphous barriers with the same stoichiometry, density and barrier length. Our results also confirm that the charge transport through the barrier is dominated by metallic conduction pathways.