2020/07/31 by Kevin Ryczko, Pierre Darancet, Isaac Tamblyn
Engineering · Materials Science · Mathematics · Physics and Astronomy · #2D Materials and Applications #Acoustics #Advanced Memory and Neural Computing #Artificial intelligence #Artificial neural network #Computer science #Graphene #Graphene research and applications #Inverse #Inverse problem #Materials science #Mathematical analysis #Mathematics #Nanotechnology #Neuroevolution #Physics #Quantum #Quantum mechanics #Transducer #cond-mat.mes-hall
paper · pdf · doi:10.1021/acs.jpcc.0c06903
openalex publication_date 2020/11/18 · arxiv created 2021/02/24 · arxiv updated 2021/02/25 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/05
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 ≃2 3800 possible configurations). The best-performing device allowed for a terminal-specific separation of valley currents with ≃96% (≃94%) K ( K ′) valley purity. Importantly, the devices found through our non-linear optimization procedure have both a higher merit function and higher robustness to defects than the ones obtained through geometry optimization.