Usage of machine learning to find suitable catalysts for ethanol oxidation reaction

• Main Author: Toh, Desmond Hong Yao
• Other Authors: -
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2023
• Subjects: Engineering::Materials
• Online Access: https://hdl.handle.net/10356/165783
• Institution: Nanyang Technological University

Ethanol oxidation reaction involves the conversion of ethanol into acetaldehyde and acetic acid. However, catalyst design and optimisation has been a challenge due to the complexity of the reactions. It involves multiple steps and intermediate species that makes the identification of desired reaction pathway challenging and the selectivity is influenced by many factors including the environment, condition in which the reaction took place, composition of catalyst and the intermediates. Thus, this requires a comprehensive understanding of the reaction mechanism which includes identifying their reaction pathways and intermediate species. Advances in technology allows us to use machine learning techniques to identify the reaction intermediates and predict potentially favourable reaction pathway. In recent studies, machine learning has been employed to predict suitable catalysts and optimisation of the reaction conditions which helps to identify the correlations between parameters and catalyst performance, providing a deeper insights into the mechanisms of the reactions and thus, able to develop a more efficient catalyst for ethanol oxidation reaction. In this project, multi alloy elements nanoparticles will be synthesized and the X-ray diffraction results and other reaction conditions and parameters will be taken in account when doing machine learning to predict suitable catalyst and conditions for ethanol oxidation reactions. Machine learning techniques, specifically decision tree regressor and linear regression will be used to find the top feature importance and understand the parameters that affect the production of a single solution and yield percentage in the reaction.