Scanning gate microscopy (SGM) is a valuable scanning probe technique for characterizing electronic transport in mesoscopic systems. However, the interpretation of the method is often limited by many experimental challenges. In this work, we propose an empirically-constrained optimization approach based on swarm search and Green’s functions to extract more information from SGM measurements. The approach is applied to a quantum point contact fabricated on an InAlAs/InGaAs/InAlAs quantum well, and the results indicate that the corresponding SGM could be generated, in a weak approximation, by a fluctuating background potential with features with radii in the order of 20 to 30 nm and a correlation length of 5.7 nm. Our method represents a data-driven tool for estimating solutions for inverse problems in mesoscopic physics, and can be used to generate estimates for the potential landscape experienced by free electrons in mesoscopic systems.
Read full abstract