Unveiling the charge distribution of a GaAs-based nanoelectronic device: A large experimental dataset approach

Abstract

In quantum nanoelectronics, numerical simulations have become a ubiquitous tool. Yet the comparison with experiments is often done at a qualitative level or restricted to a single device with a handful of fitting parameters. In this work, we assess the predictive power of these simulations by comparing the results of a single model with a large experimental dataset of 110 devices with 48 different geometries. The devices are quantum point contacts of various shapes and sizes made with electrostatic gates deposited on top of a high mobility GaAs/AlGaAs two-dimensional electron gas. We study the pinch-off voltages applied on the gates to deplete the two-dimensional electron gas in various spatial positions. We argue that the pinch-off voltages are a very robust signature of the charge distribution in the device. The large experimental dataset allows us to critically review the modeling and arrive at a robust one-parameter model that can be calibrated in situ, a crucial step for making predictive simulations.