Inverse Design of a Graphene-Based Quantum Transducer via Neuroevolution

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 $\simeq 96$\% ($\simeq 94\%)$ $K$ ($K^{\prime }$) 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.

References in corpus (30)

• The electronic properties of graphene
• Semi-Supervised Classification with Graph Convolutional Networks
• scikit-image: Image processing in Python
• Control of valley polarization in monolayer MoS2 by optical helicity
• Valley polarization in MoS2 monolayers by optical pumping
• MoS_2 as an ideal material for valleytronics: valley-selective circular dichroism and valley Hall effect
• SchNet - a deep learning architecture for molecules and materials
• Valley contrasting physics in graphene: magnetic moment and topological transport
• The Valley Hall Effect in MoS2 Transistors
• Valley filter and valley valve in graphene
• Kwant: a software package for quantum transport
• Inverse design and demonstration of a compact and broadband on-chip wavelength demultiplexer
• Veselago Lens for Electrons: Focusing and Caustics in Graphene p-n Junctions
• Deep Neuroevolution: Genetic Algorithms Are a Competitive Alternative for Training Deep Neural Networks for Reinforcement Learning
• Graphene valley filter using a line defect
• Gauge field induced by ripples in graphene
• Valley filter in strain engineered graphene
• Graphene Nanobubbles as Valley Filters and Beamsplitters
• Mapping the global design space of nanophotonic components using machine learning pattern recognition
• Controlled Growth of a Line Defect in Graphene and Implications for Gate-Tunable Valley Filtering
• Current splitting and valley polarization in elastically deformed graphene
• Valley engineering by strain in Kekulé-distorted graphene
• Current flow paths in deformed graphene: from quantum transport to classical trajectories in curved space
• Pseudomagnetic fields and triaxial strain in graphene
• Local versus extended deformed graphene geometries for valley filtering
• Quantum Wires and Waveguides Formed in Graphene by Strain
• Quantized transport, strain-induced perfectly conducting modes and valley filtering on shape-optimized graphene Corbino devices
• Graphene Quantum Strain Transistors
• Valley Hall Effect and Non-Local Resistance in Locally Gapped Graphene
• Valley notch filter in a graphene strain superlattice: Green's function and machine learning approach

Cited by in corpus (4)

• Neural evolution structure generation: High Entropy Alloys
• Greedy optimization of the geometry of Majorana Josephson junctions
• Predicting magnetic edge behaviour in graphene using neural networks
• Simplifying inverse material design problems for fixed lattices with alchemical chirality