Speaker
Description
The machine-learned force field (MLFF) method embedded within ab-initio molecular dynamics (AIMD) in VASP was employed to investigate the bulk, surface and collector’s adsorption properties on the PtAs₂ (100) surface at 300K. This approach was used to gain deeper insight into the physicochemical behaviour of the sperrylite mineral (PtAs₂). AIMD-MLFF bulk training gave a lattice parameter of 5.977 Å, while a 4×4× supercell yielded a value 5.991 Å. These lattice parameters were in good agreement with the experimental lattice parameter of 5.970 Å. From the Ab-initio calculations the (100) surface was identified as the most stable, with surface energy of 1.00 J/m². The MLFF-trained (100) surface yielded a higher surface energy of 1.76 J/m², while the supercell model produced a value of 1.67 J/m². Among the collectors studied, dibutyl-dithioarsenate (DBDTAs) was found to be the most favourable based on the ab-initio calculations, with adsorption energy of –330.738 kJ/mol. The AIMD-MLFF approach gave similar adsorption energy of –398.888 kJ/mol for the trained system and –499.426 kJ/mol for the applied force-field. These findings indicated that the AIMD-MLFF approach was a valuable and efficient tool for simulating large-scale systems, providing results both consistent with both density functional theory (DFT) calculations.
| Apply for student award at which level: | MSc |
|---|---|
| Consent on use of personal information: Abstract Submission | Yes, I ACCEPT |