Proposal Number: CTS-0080823 Principal Investigator: Can Erkey Institution: University of Connecticut
The goal of this project is to develop a methodology based on computational chemistry and mathematical programming coupled with computational phase equilibria to design homogeneous catalysts that are highly active, selective, and can be readily recovered from the reaction mixture. A model reaction will be the hydroformylation reaction catalyzed by rhodium phosphine complexes. To avoid computationally intensive quantum chemical calculations, the PIs plan to screen homogeneous catalysts using a neural network model based on selectivity data obtained from data on several different ligands. Catalyst ligand characteristics will be related to selectivity. Through model predictions additional catalysts can be screened without using ab initio calculations. Solvent effects on catalytic reactions will also be modeled using several approaches. The PIs will develop a model for predicting distribution coefficients for a fluorous solvent and supercritical carbon dioxide. These efforts will complement experimental work currently being carried out in homogeneous catalysis. This work may lead to an improved methodology for catalyst design and accelerate the use of homogeneous catalysts in industrial processes.
