Inverse Design of a Graphene-Based Quantum Transducer via Neuroevolution

TL;DR

This paper presents an inverse design framework combining neural networks, genetic algorithms, and tight-binding calculations to optimize graphene-based quantum transducers for valleytronics, achieving high purity and robustness.

Contribution

It introduces a novel non-linear optimization method for designing nanoelectronic devices with large configuration spaces, specifically applied to graphene valleytronics.

Findings

Achieved high valley current separation with ~96% and ~94% purity.

Optimized devices show higher merit and robustness compared to geometry-based designs.

Converged to synthesizable devices within a few thousand evaluations.

Abstract

We introduce an inverse design framework based on artificial neural networks, genetic algorithms, and tight-binding calculations, capable to optimize the very large configuration space of nanoelectronic devices. Our non-linear optimization procedure operates on trial Hamiltonians through superoperators controlling growth policies of regions of distinct doping. We demonstrate that our algorithm optimizes the doping of graphene-based three-terminal devices for valleytronics applications, monotonously converging to synthesizable devices with high merit functions in a few thousand evaluations (out of $\simeq 2^{3800}$ possible configurations). The best-performing device allowed for a terminal-specific separation of valley currents with $\simeq 96$\% ($\simeq 94\%)$ $K$ ($K'$) valley purity. Importantly, the devices found through our non-linear optimization procedure have both higher merit function and higher robustness to defects than the ones obtained through geometry optimization.