Active Learning and Reinforcement Learning for Autonomous Catalyst Design in CO2 Hydrogenation

Yingru Chen

doi:10.62051/ijmsts.v2n2.08

Authors

Yingru Chen

DOI:

https://doi.org/10.62051/ijmsts.v2n2.08

Keywords:

CO2 Hydrogenation, Catalyst Design, Machine Learning

Abstract

The increasing concentration of carbon dioxide (CO2) in the atmosphere presents a significant challenge in the context of climate change, necessitating innovative strategies for greenhouse gas mitigation. CO2 hydrogenation, which converts CO2 into hydrocarbons and other valuable chemicals using hydrogen, has emerged as a promising method for addressing both carbon emissions and renewable energy production. Catalysts are crucial in enhancing the efficiency and selectivity of these reactions; however, traditional catalyst design methods often rely on laborious trial-and-error approaches, which can be inefficient and resource-intensive. This paper explores the integration of active learning and reinforcement learning as advanced methodologies for automating catalyst design specifically for CO2 hydrogenation. Active learning focuses on selecting the most informative data points to improve model predictions while minimizing experimental costs, whereas reinforcement learning optimizes decision-making processes through iterative feedback. A case study demonstrates the application of these techniques, leading to the successful identification of novel catalyst compositions that exhibit superior performance metrics. The findings highlight the potential of machine learning to revolutionize catalyst discovery, ultimately contributing to more sustainable CO2 conversion strategies.

References

[1] Zhang Y, Xu Z, Wang T, et al. Reinforcement learning-enabled autonomous optimization for CO2 hydrogenation. Nat Commun. 2020; 11(1):5812.

[2] Liu Y, Zhao T, Ju W, Shi S. Materials discovery and design using machine learning. J Materiomics. 2017; 3(3):159-177.

[3] Gao W, Chen Y, Li Y, et al. Automated catalyst design for CO2 electroreduction. Nature. 2022; 610(7931):287-293.

[4] Asif M, Yao C, Zuo Z, Bilal M, Zeb H, Lee S, Kim T. Machine learning-driven catalyst design, synthesis and performance prediction for CO2 hydrogenation. Journal of Industrial and Engineering Chemistry. 2024 Sep 21.

[5] Gu GH, Noh J, Kim S, et al. Practical deep-learning representation for fast heterogeneous catalyst screening. NPJ Comput Mater. 2019; 5(1):47.

[6] Jia X, Lynch A, Huang Y, et al. Anthropogenic CO2 reduction: from materials design to flow reactors. Mater Today. 2019; 24:15-47.

[7] Zhong M, Tran K, Min Y, et al. Accelerated discovery of CO2 electrocatalysts using active machine learning. Nature. 2020; 581(7807):178-183.

[8] Schlexer Lamoureux P, Winther KT, Garrido Torres JA, et al. Machine Learning for Computational Heterogeneous Catalysis. ChemCatChem. 2019; 11(16):3581-3601.

[9] Li Z, Wang S, Chin WS, et al. Machine learning-guided discovery and optimization of catalysts for CO2 electroreduction. Nat Catal. 2022; 5(10):858-868.

[10] Toyao T, Maeno Z, Takakusagi S, et al. Machine Learning for Catalysis Informatics: Recent Applications and Prospects. ACS Catal. 2020; 10(3):2260-2297.

[11] Cui C, Han J, Zhu X, et al. Promotional effect of surface hydroxyls on electrochemical reduction of CO2 over SnO2/Sn electrode. J Catal. 2016; 343:257-265.

[12] Gusarov S, Stoyanov SR, Siahrostami S, Kovalenko A. Deep Learning for Molecular Design and Discovery: From Chemical Reactions to Materials. ACS Cent Sci. 2021; 7(6):902-915.

[13] Medford AJ, Kunz MR, Ewing SM, et al. Extracting knowledge from data through catalysis informatics. ACS Catal. 2018; 8(8):7403-7429.

[14] Xie T, Grossman JC. Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties. Phys Rev Lett. 2018; 120(14):145301.

[15] Grajciar L, Heard CJ, Bondarenko AA, et al. Towards operando computational modeling in heterogeneous catalysis. Chem Soc Rev. 2018; 47(22):8307-8348.

[16] Pegis ML, Roberts JAS, Wasylenko DJ, et al. Standard reduction potentials for oxygen and carbon dioxide couples in acetonitrile and N,N-dimethylformamide. Inorg Chem. 2015; 54(24):11883-11888.

[17] Ulissi ZW, Tang MT, Xiao J, et al. Machine-Learning Methods Enable Exhaustive Searches for Active Bimetallic Facets and Reveal Active Site Motifs for CO2 Reduction. ACS Catal. 2017; 7(10):6600-6608.

[18] Ma X, Li Z, Achenie LEK, Xin H. Machine-Learning-Augmented Chemisorption Model for CO2 Electroreduction Catalyst Screening. J Phys Chem Lett. 2015; 6(18):3528-3533.

[19] Boes JR, Mamun O, Winther K, Bligaard T. Graph Theory Approach to High-Throughput Surface Adsorption Structure Generation. J Phys Chem A. 2019; 123(11):2281-2285.

[20] Gasper R, Shi H, Ramasubramaniam A. Adsorption of CO2 on Pt-Cu Nanoalloys: Density Functional Theory and Machine Learning Approaches. J Phys Chem C. 2017; 121(10):5612-5619.

[21] Back S, Tran K, Ulissi ZW. Toward a Design of Active Oxygen Evolution Catalysts: Insights from Automated Density Functional Theory Calculations and Machine Learning. ACS Catal. 2019; 9(9):7651-7659.

[22] Schlexer Lamoureux P, Winther KT, Garrido Torres JA, et al. Machine Learning for Computational Heterogeneous Catalysis. ChemCatChem. 2019; 11(16):3581-3601.

[23] Jinnouchi R, Asahi R. Predicting Catalytic Activity of Nanoparticles by a DFT-Aided Machine-Learning Algorithm. J Phys Chem Lett. 2017; 8(17):4279-4283.

[24] Li Z, Ma X, Xin H. Feature engineering of machine-learning chemisorption models for catalyst design. Catal Today. 2017; 280:232-238.

[25] Gu GH, Choi C, Lee Y, et al. Progress in computational and machine-learning methods for heterogeneous small-molecule activation. Nat Rev Mater. 2020; 5(8):605-621.

[26] Iglesia E. Design, synthesis, and use of cobalt-based Fischer-Tropsch synthesis catalysts. Applied Catalysis A: General. 1997 Nov 4; 161(1-2):59-78.

[27] Agrawal, A., & Choudhary, A. (2016). Materials Data Infrastructure: A New Paradigm for Data-Driven Materials Science. Nature Reviews Materials, 1(2), 1-16.

[28] Thomas JM. Design, synthesis, and in situ characterization of new solid catalysts. Angewandte Chemie International Edition. 1999 Dec 16; 38(24):3588-628.

[29] Lim XX, Low SC, Oh WD. A critical review of heterogeneous catalyst design for carbon nanotubes synthesis: Functionalities, performances, and prospects. Fuel Processing Technology. 2023 Mar 1; 241:107624.

[30] Rioux RM, Song H, Hoefelmeyer JD, Yang P, Somorjai GA. High-surface-area catalyst design: synthesis, characterization, and reaction studies of platinum nanoparticles in mesoporous SBA-15 silica. The Journal of Physical Chemistry B. 2005 Feb 17; 109(6):2192-202.

[31] Zhang Y, Zhu X, Zhang G, Shi P, Wang AL. Rational catalyst design for oxygen evolution under acidic conditions: strategies toward enhanced electrocatalytic performance. Journal of Materials Chemistry A. 2021; 9(10):5890-914.

[32] Spivey JJ, Krishna KS, Kumar CS, Dooley KM, Flake JC, Haber LH, Xu Y, Janik MJ, Sinnott SB, Cheng YT, Liang T. Synthesis, characterization, and computation of catalysts at the center for atomic-level catalyst design. The Journal of Physical Chemistry C. 2014 Sep 4; 118(35):20043-69.

[33] Spencer MS, Twigg MV. Metal catalyst design and preparation in control of deactivation. Annu. Rev. Mater. Res.. 2005 Aug 4; 35(1):427-64.

[34] Tran K, Ulissi ZW. Active learning across intermetallics to guide discovery of electrocatalysts for CO2 reduction and H2 evolution. Nat Catal. 2018; 1(9):696-703.

[35] Settles, B. (2009). Active Learning Literature Survey. Computer Science Technical Report 1648, University of Wisconsin-Madison.

[36] Agrawal, A., & Choudhary, A. (2016). Materials Data Infrastructure: A New Paradigm for Data-Driven Materials Science. Nature Reviews Materials, 1(2), 1-16.

[37] Sun T, Yang J, Li J, Chen J, Liu M, Fan L, Wang X. Enhancing Auto Insurance Risk Evaluation with Transformer and SHAP. IEEE Access. 2024 Aug 19.

[38] Lin Y, Fu H, Zhong Q, Zuo Z, Chen S, He Z, Zhang H. The influencing mechanism of the communities’ built environment on residents’ subjective well-being: A case study of Beijing. Land. 2024 Jun 4;13(6):793.

[39] Li J, Fan L, Wang X, Sun T, Zhou M. Product Demand Prediction with Spatial Graph Neural Networks. Applied Sciences. 2024 Aug 9; 14(16):6989.

[40] Liu M, Ma Z, Li J, Wu YC, Wang X. Deep-Learning-Based Pre-training and Refined Tuning for Web Summarization Software. IEEE Access. 2024 Jul 4.

[41] Chen X, Liu M, Niu Y, Wang X, Wu YC. Deep-Learning-Based Lithium Battery Defect Detection via Cross-Domain Generalization. IEEE Access. 2024 Jun 3.

[42] Wang X, Wu YC, Ji X, Fu H. Algorithmic discrimination: examining its types and regulatory measures with emphasis on US legal practices. Frontiers in Artificial Intelligence. 2024 May 21; 7:1320277.

[43] Wang X, Wu YC, Ma Z. Blockchain in the courtroom: exploring its evidentiary significance and procedural implications in US judicial processes. Frontiers in Blockchain. 2024 Apr 12; 7:1306058.

[44] Ma Z, Chen X, Sun T, Wang X, Wu YC, Zhou M. Blockchain-Based Zero-Trust Supply Chain Security Integrated with Deep Reinforcement Learning for Inventory Optimization. Future Internet. 2024 May 10; 16(5):163.

[45] Wang X, Wu YC, Zhou M, Fu H. Beyond surveillance: privacy, ethics, and regulations in face recognition technology. Frontiers in big data. 2024 Jul 3; 7:1337465.

[46] Wang X, Wu YC. Balancing innovation and Regulation in the age of generative artificial intelligence. Journal of Information Policy. 2024 Jul 2; 14.

[47] Zuo Z, Niu Y, Li J, Fu H, Zhou M. Machine Learning for Advanced Emission Monitoring and Reduction Strategies in Fossil Fuel Power Plants. Applied Sciences. 2024 Sep 19; 14(18):8442.