arXiv subjects

Shang Zhu

Publications and source records attributed to Shang Zhu.

25 records · Page 2Linked to original sources

arXiv2024

Open-Source Fermionic Neural Networks with Ionic Charge Initialization

Finding accurate solutions to the electronic Schrödinger equation plays an important role in discovering important molecular and material energies and characteristics. Consequently, solving systems with large numbers of electrons has become increasingly important. Variational Monte Carlo (VMC) methods, especially those approximated through deep neural networks, are promising in this regard. In this paper, we aim to integrate one such model called the FermiNet, a post-Hartree-Fock (HF) Deep Neural Network (DNN) model, into a standard and widely used open source library, DeepChem. We also propose novel initialization techniques to overcome the difficulties associated with the assignment of excess or lack of electrons for ions.

cs.LG↗

arXiv2023

Differentiable Modeling and Optimization of Battery Electrolyte Mixtures Using Geometric Deep Learning

Electrolytes play a critical role in designing next-generation battery systems, by allowing efficient ion transfer, preventing charge transfer, and stabilizing electrode-electrolyte interfaces. In this work, we develop a differentiable geometric deep learning (GDL) model for chemical mixtures, DiffMix, which is applied in guiding robotic experimentation and optimization towards fast-charging battery electrolytes. In particular, we extend mixture thermodynamic and transport laws by creating GDL-learnable physical coefficients. We evaluate our model with mixture thermodynamics and ion transport properties, where we show improved prediction accuracy and model robustness of DiffMix than its purely data-driven variants. Furthermore, with a robotic experimentation setup, Clio, we improve ionic conductivity of electrolytes by over 18.8% within 10 experimental steps, via differentiable optimization built on DiffMix gradients. By combining GDL, mixture physics laws, and robotic experimentation, DiffMix expands the predictive modeling methods for chemical mixtures and enables efficient optimization in large chemical spaces.

physics.chem-ph↗

arXiv2023

Open Source Infrastructure for Differentiable Density Functional Theory

Learning exchange correlation functionals, used in quantum chemistry calculations, from data has become increasingly important in recent years, but training such a functional requires sophisticated software infrastructure. For this reason, we build open source infrastructure to train neural exchange correlation functionals. We aim to standardize the processing pipeline by adapting state-of-the-art techniques from work done by multiple groups. We have open sourced the model in the DeepChem library to provide a platform for additional research on differentiable quantum chemistry methods.

cs.LG↗

arXiv2023

Quantifying quantum coherence of multiple-charge states in tunable Josephson junctions

Coherence and tunneling play central roles in quantum phenomena. In a tunneling event, the time that a particle spends inside the barrier has been fiercely debated. This problem becomes more complex when tunneling repeatedly occurs back and forth, and when involving many particles. Here we report the measurement of the coherence time of various charge states tunneling in a nanowire-based tunable Josephson junction; including single charges, multiple charges, and Cooper pairs. We studied all the charge tunneling processes using Landau-Zener-Stückelberg-Majorana (LZSM) interferometry, and observed high-quality interference patterns. In particular, the coherence time of the charge states was extracted from the interference fringes in Fourier space. In addition, our measurements show the break-up of Cooper pairs, from a macroscopic quantum coherent state to individual particle states. Besides the fundamental research interest, our results also establish LZSM interferometry as a powerful technique to explore the coherence time of charges in hybrid devices.

cond-mat.mes-hall↗

arXiv2023

Gate-tunable negative differential conductance in hybrid semiconductor-superconductor devices

Negative differential conductance (NDC) manifests as a significant characteristic of various underlying physics and transport processes in hybrid superconducting devices. In this work, we report the observation of gate-tunable NDC outside the superconducting energy gap on two types of hybrid semiconductor-superconductor devices, i.e., normal metal-superconducting nanowire-normal metal and normal metal-superconducting nanowire-superconductor devices. Specifically, we study the dependence of the NDCs on back-gate voltage and magnetic field. When the back-gate voltage decreases, these NDCs weaken and evolve into positive differential conductance dips; and meanwhile they move away from the superconducting gap towards high bias voltage, and disappear eventually. In addition, with the increase of magnetic field, the NDCs/dips follow the evolution of the superconducting gap, and disappear when the gap closes. We interpret these observations and reach a good agreement by combining the Blonder-Tinkham-Klapwijk (BTK) model and the critical supercurrent effect in the nanowire, which we call the BTK-supercurrent model. Our results provide an in-depth understanding of the tunneling transport in hybrid semiconductor-superconductor devices.

cond-mat.mes-hall↗

arXiv2022

Dynamics of the lithium metal electrodeposition: Effects of a gas bubble

Rechargeable lithium metal batteries have been widely investigated recently, driven by the global trend for the electrification of transportation. Understanding the dynamics of lithium metal electrodeposition is crucial to design safe and reliable lithium metal anodes. In this study, we developed a grand potential-based phase-field model to investigate the effect of a static gas bubble, which forms due to the complicated internal side reactions, on the dynamics of the dendrite growth during electrodeposition. It is observed that with the presence of a gas bubble, the dendrite growth is largely accelerated, due to the accumulation of lithium ions on the far side of the bubble away from the anode surface, which could serve as an ion "reservoir" for the dendrite growth, leading to the bending/tilting of the lithium dendrites toward the bubble. Meanwhile, the effects of the bubble size and distance to the anode are further studied, demonstrating that the larger the bubble size and the closer to the anode, the longer the lithium dendrites grow. We hope this study could serve as an example to exploit the effect of extrinsic factors on the dendrite growth dynamics.

cond-mat.mtrl-sci↗

arXiv2022

AutoMat: Accelerated Computational Electrochemical systems Discovery

Large-scale electrification is vital to addressing the climate crisis, but several scientific and technological challenges remain to fully electrify both the chemical industry and transportation. In both of these areas, new electrochemical materials will be critical, but their development currently relies heavily on human-time-intensive experimental trial and error and computationally expensive first-principles, meso-scale and continuum simulations. We present an automated workflow, AutoMat, that accelerates these computational steps by introducing both automated input generation and management of simulations across scales from first principles to continuum device modeling. Furthermore, we show how to seamlessly integrate multi-fidelity predictions such as machine learning surrogates or automated robotic experiments "in-the-loop". The automated framework is implemented with design space search techniques to dramatically accelerate the overall materials discovery pipeline by implicitly learning design features that optimize device performance across several metrics. We discuss the benefits of AutoMat using examples in electrocatalysis and energy storage and highlight lessons learned.

cond-mat.mtrl-sci↗