The Role of Transition Metals on CeO2 Supported for CO2 Adsorption by DFT and Machine Learning Analysis

• Published in: Industrial & engineering chemistry research Vol. 63; no. 25; pp. 11018 - 11029
• Main Authors: Mashhadimoslem, Hossein, Karimi, Peyman, Abdol, Mohammad Ali, Zanganeh, Kourosh, Shafeen, Ahmed, AlHammadi, Ali A., Elkamel, Ali
• Format: Journal Article
• Language: English
• Published: American Chemical Society 26.06.2024
• Subjects: adsorption; algorithms; carbon dioxide; catalysts; data collection; density functional theory; energy; Gibbs free energy; industry; Kinetics, Catalysis, and Reaction Engineering; prediction; temperature
• ISSN: 0888-5885; 1520-5045; 1520-5045
• DOI: 10.1021/acs.iecr.4c01200

Catalyst design is a field where machine learning (ML) algorithms have found many useful applications using atomistic simulation data sets. Atomistic simulation of CO2 adsorption energy using several transition metals on ceria oxide (CeO2) catalysts is the subject of our research. The density functional theory (DFT) calculation was used, and its results were applied as a data set to train several ML algorithms. Gibbs free energy (ΔG) was simulated for all TMs such as Ni, Fe, Cu, Co, Mo, Ru, Rh, Pd, Ag, Pt, Zr, and Ti and used as the goal of CO2 adsorption prediction using ML algorithms. The XGBoost algorithm provided satisfactory predictions of ΔG using different transition metals and bond energy at different adsorption temperatures on the catalysts. The effective role of Cu, Pt, Fe, and Ni in all temperature ranges regarding the increase of CO2 adsorption energy was evident. The best CO2 adsorption efficiency will be achieved by Ni at temperatures below 300 K and Cu at temperatures over 300 K. Evaluation of new catalysts using prediction findings and correlations between complex DFT simulation parameters demonstrates the potential of ML as a powerful tool for industry development.