COMPUTER-AIDED DESIGN OF HYDROGEN STORAGE MATERIALS BASED ON SILICON-CARBIDE SYSTEMS VIA A HIERARCHAL SIMULATION PLATFORM
Abstract
Hydrogen energy is a promising energy resource for reducing greenhouse gas emissions. To realize the US DOE target of 6-7 wt.% uptakes, various types of hydrogen storage materials have been widely explored for using the hydrogen energy industrially, though all the efforts have yet to reach the DOE target in industrial practice. Silicon-carbide nanotubes (SiCNTs) are a typical nanomaterial studied for that purpose owing to their enhanced molecular interactions due to their heteropolar binding nature compared with carbon nanotubes (CNTs). But, SiCNTs exhibit smaller adsorption energies than desired, so several attempts have been made to tune the energy by doping metal into pristine SiCNTs, though proper tuning schemes have yet to be established. This project aims to establish a "hierarchical simulation" platform for exploring hydrogen storage materials by doping various metals into SiC systems with various morphologies. The platform consists of not only ab initio electronic structure simulations, but also molecular dynamics (MD) simulations with neural network potentials (NNPs). The former is responsible for achieving the accuracy of simulations, but not practical for investigating actual situations (e.g., finite temperature/pressure and various morphologies) due to its computational cost. In contrast, the MD approach is appropriate for more realistic simulations, but requires accurate potentials. The NNPs can be more accurate than conventional potentials due to their functional reproducibility by learning more accurate ab initio simulation data. Based on our establishing platform, we can realize more accurate and realistic simulations at finite pressure/temperature. The resultant uptake values would be more appropriate for a direct comparison with existing experiments, thereby giving new insight into computer-aided materials design for hydrogen storage materials based on silicon-carbide systems.
Document Details
- Document Type
- DoD Grant Award
- Publication Date
- Apr 20, 2023
- Source ID
- FA23862214065
Entities
People
- Kenta Hongo
Organizations
- Air Force Office of Scientific Research
- Japan Advanced Institute of Science and Technology
- United States Air Force