A Machine-Learning Model to Predict Activation Energies of Hydrogenation Reactions
AIChE Annual Meeting
2014
2014 AIChE Annual Meeting
Liaison Functions
Undergraduate Research Forum III: Classical Chemical Engineering/Other Special Topics
Monday, November 17, 2014 - 4:35pm to 4:55pm
The goal of this research is to develop a machine-learning model to predict activation energies associated to the hydrogenation of organic compounds in the presence of specific catalytic substrates. The current method of predicting activation energies involves a linear relationship with the change in Gibbs free energy of the reaction, which can be very inaccurate. Other methods include experimentation and simulation, both of which have significant cost and/or time drawbacks. This research seeks to improve on these methods to develop a means to predict activation energies quickly, accurately, and with minimal cost.