Design of Metastable Phases
A general computational package to predict metastable phase of inorganic materials using the concepts of machine learning, electronic structure methods and atomistic potentials. The long-term goal of this package is to provide elemental composition as input (e.g. B, HfO2 or Al0.2Cu0.8) and output a set of potential metastable phases, their properties and their plausible conditions of stability (pressure, temperature, strain profile, etc.).
Notable Publications
Nature Communications 13.1 1-12 (2022)
Chemistry of Materials 32.9, 3823-3832 (2020)
Depolymerizable Polymers
Using first principles and ML methods, study depolymerization mechanisms in a class of biodegradable plastics containing polymeric ester linkages. This will allow discovery of novel polymer systems that can potentially display similar properties as everyday polymers, such as polyethylene and polyethylene terephthalate, but are easily depolymerized into valuable monomers or other organic systems through hydrolysis reaction.
Notable Publications
Chemistry of Materials 32.24 10489-10500 (2020)
Materials Science and Engineering: R: Reports 144 100595 (2021)
Design of Self-assembling Peptides
An AI-expert to autonomously suggest peptide sequences that can self-assemble and form interesting nanostructures. Traditional methods of peptide design are inefficient and biased to a few specific amino acids. Thus, machine learning methods are being developed to overcome human biases and develop an analytical expression to guide efficient design of peptides. This work is in collaboration with Argonne National Lab, USA.
Notable Publications
Research Square [https://doi.org/10.21203/rs.3.rs-505801/v1]
