November 2021 - January 2024
GitHub link https://github.com/mhan8/Metastable_ML
This project established a comprehensive public repository featuring approximately 800 structures of metastable perovskites. Each structure was carefully tuned under high pressure and strain conditions, then optimized using Density Functional Theory (DFT) calculations. The database also offers several pre-trained machine-learning models for predicting band gaps and enthalpy, providing useful resources for researchers in the field.
September 2023 - present
(Coming soon)
This project aims to establish an open-access database featuring experimental data on halide perovskites with various compositions and dimensions under high pressure. Currently, such data are scarce and scattered across literature in text and figure formats due to their limited availability and the challenges of determining exact structures. By leveraging a large language model and optical character recognition, we collect and compile these experimental data to construct a comprehensive database. This resource will facilitate model training and prediction of metastable phases. The project is a collaborative effort with the University of Florida and SLAC scientists.
September 2023 - present
(Coming soon)
This project employs active learning to expedite the identification of the optimal pressure and strain levels required for achieving target material properties. Traditionally, extensive computations and experiments were necessary to scan the entire composition and condition spaces, consuming significant time and resources. By adopting a data-driven active learning framework, we aim to design more efficient experiments and computations, thus uncovering synthetic pathways to achieve stable materials with optimal properties. This project is a collaborative effort with SLAC scientists.
April 2023 - present
(Coming soon)
This project investigates the kinetics of silicon-vacancy (SiV) and diamond formation under high pressure-temperature (P-T) conditions. SiV centers in diamond structures are gaining attention as quantum emitters for quantum information processing. I have been actively participating in and leading experiments to better understand the intermediate phases and thresholds of the SiV formation reaction using diamond anvil cells (DAC) and x-ray diffraction (XRD). This work is a collaborative effort with scientists from European XFEL, Advanced Light Source, Advanced Photon Source, and SLAC.