Unnatti Sharma

* Preferred pathway prediction for catalytic CO2RR  - Modeled thermocatalytic CO2RR using RING, generating an exhaustive reaction network to produce mono-carbon products.  - Developed complete pipeline to perform adsorption energy based Reaction Network Flux Analysis to predict catalysts’ optimal pathway preference and product profile.  - Developed ML model to calculate scaling relation coefficients for adsorption on mono-metallic catalysts for hydrocarbons containing up to two carbons. * Site geometry as a descriptor for catalyst selectivity  - Designed a robust graph-based tool to analyze active ensemble geometry on intermetallic surfaces.  - Developed an interactive platform which can screen 80,000 catalyst surfaces for user-specified site geometries.  - Discovered correlations between adsorption energy, site geometry and material properties which helped in catalyst design. * Catalyst design for selective semi-hydrogenation of ethylene  - Performed high-throughput screening across cheminformatics databases like Materials Project and AFLOW.  - Designed catalyst surfaces using AI-based adsorbate placement and adsorption energy calculations.  - Presented over 50,000 new potential candidate materials for further selective catalysis research.  - Mentored an undergraduate researcher on symmetric surface energy calculation.

 Developed Cognitive Artificial Intelligence Optimal Control strategies for in-house plant design.  Beat traditional MPC based methods in speed of response using concepts from Deep Neural Networks and Metaheuristics