Probing photocatalysis via artificial intelligence
The generation of renewable energy through photocatalysis is an attractive option to utilize the abundantly available solar radiation for a sustainable future. Photocatalysis is considered a subset of heterogeneous catalysis which is inherently a complex phenomenon and is particularly challenging from an informatics perspective.1 Catalyst design can be aided by process modelling, but is often hampered by the limited understanding of intricate interaction between catalyst surface, physicochemical properties and reaction environment. Fortunately, the modern advancements in artificial intelligence techniques have heavily facilitated the modelling and prediction of complex high-dimensional data. In principle, machine learning models exhibit exceptional ability to ‘learn’ from a given set of experimental data, and predict the process response without knowledge of physical and chemical laws governing the system.2 Such technique has been successfully utilised for designing various heterogeneous catalysts, including ammoxidation of propylene catalyst, methane oxidative decoupling catalyst, propane ammoxidation catalyst, as well as analysis and prediction of nitric oxide over zeolites.3 The approach is frequently employed for the modelling of electrochemical processes, however, the potential of this technique to probe photocatalytic/photooxidative systems remains largely unexplored.
In a broad perspective, the project aims to elucidate the quantitative structure-property-activity relationship, coupled with process conditions, for photocatalytic hydrogen generation system. The specific objective is data collection from literature to aid formulation of machine learning models, and further to be corroborated with experimental data from in-house photocatalysis system. The validated model will provide insights on the importance of catalysts as well as reaction attributes that govern the overall performance of the process.
The students selected for this project will be given opportunity to work in Particles and Catalysis Research Group (PartCat) at School of Chemical Engineering, under the supervision of Scientia Professor Rose Amal. The students will have access to state-of-the-art experimentation facilities and computational tools, and will gain necessary expertise facilitating their career in industry or academic research. Further information can be obtained by contacting Professor Rose Amal (firstname.lastname@example.org).
1. Medford AJ, Kunz MR, Ewing SM, Borders T and Fushimi RR, Extracting knowledge from data through catalysis informatics. ACS Catalysis (2018).
2. Khataee A and Kasiri M, Artificial neural networks modeling of contaminated water treatment processes by homogeneous and heterogeneous nanocatalysis. Journal of Molecular Catalysis A: Chemical 331:86-100 (2010).
3. Serra JM, Corma A, Chica A, Argente E and Botti V, Can artificial neural networks help the experimentation in catalysis? Catalysis Today 81:393-403 (2003).
Suitable for: Chemical Engineers and Industrial Chemistry student. Project maybe available to do during summer semester.
Level of difficulty: Challenging