Abstract
The photocatalytic CO2 conversion into value-added chemicals such as CH4 and CO is a highly promising sustainable approach to meet rising energy demands while mitigating atmospheric CO2 levels. This study investigates the prediction and optimization of CH4 and CO production using tree-based machine learning (ML) models applied to g-C3N4/TiO2 nanotube arrays (TNTAs) photocatalysts for gas-phase CO2 conversion. To predict the photoconversion rates of CO2 into CH4 and CO, several tree-based ML algorithms, including AdaBoost, Bagging, CatBoost, Decision Tree, Extra Trees, Gradient Boosting, HistGradientBoosting, LightGBM, RandomForest, and XGBoost were employed. The production of CH4 and CO (µmol/cm2) were designated as the target outputs, while input parameters included catalyst exposed surface area, initial concentration of CO2, feed pressure, light power, and irradiation time. The performance of ML algorithms was appraised using five statistical metrics. Bayesian optimization was employed to fine-tune the hyperparameters of the machine learning model algorithms. Among the evaluated models, CatBoost (CB) performed the most accurately, with R2 = 0.9887 (training) and R2 = 0.9883 (test) for CH4 production and R2 = 0.9885 (training) and R2 = 0.9874 (test) for CO production. Feature importance analysis and SHAP plots highlighted the significant influence of irradiation time and catalyst exposed surface area on the production efficiency of both products. Additionally, the input parameters were systematically optimized using CB model predictions, which were validated against experimental data and achieved almost similar prediction. These results support the effectiveness of ML-directed modeling in maximizing CO2 conversion efficiency and revealing the potential of data-driven strategies in steering photocatalytic technology.
Data availability
Data will be made available on request to corresponding authors.
References
Crake, A. et al. Titanium dioxide/carbon nitride nanosheet nanocomposites for gas phase CO2 photoreduction under UV-visible irradiation. Appl. Catal. B Environ. 242, 369–378. https://doi.org/10.1016/j.apcatb.2018.10.023 (2019).
Xu, Q. et al. Recent advances in solar-driven CO2 reduction over g-C3N4-based photocatalysts. Carbon Energy. 5 https://doi.org/10.1002/cey2.205 (2023).
Sim, L. C., Yee, P. L., Leong, K. H., Aziz, A. A. & Hossen, M. A. CO2 photoreduction to hydrocarbons and oxygenated hydrocarbons (INC, 2024). https://doi.org/10.1016/B978-0-443-19235-7.00017-8
Liu, L., Wang, S., Huang, H., Zhang, Y. & Ma, T. Surface sites engineering on semiconductors to boost photocatalytic CO2 reduction. Nano Energy. 75, 104959. https://doi.org/10.1016/j.nanoen.2020.104959 (2020).
Lu, Y. et al. Advances and roles of oxygen vacancies in semiconductor photocatalysts for solar-driven CO2 reduction. Surf. Interfaces. 53, 104957. https://doi.org/10.1016/j.surfin.2024.104957 (2024).
Hossen, M. A. et al. Recent progress in TiO2-Based photocatalysts for conversion of CO2 to hydrocarbon fuels: A systematic review. Results Eng. 16, 100795. https://doi.org/10.1016/j.rineng.2022.100795 (2022).
Farooq, N. et al. Recent trends of Titania (TiO2) based materials: A review on synthetic approaches and potential applications. J. King Saud Univ. - Sci. 36, 103210. https://doi.org/10.1016/j.jksus.2024.103210 (2024).
Hossen, M. A. et al. A comprehensive review on advances in TiO2 nanotube (TNT)-Based photocatalytic CO2 reduction to Value-Added products. Energies 15 https://doi.org/10.3390/en15228751 (2022).
Yao, H. et al. Metal-organic framework [NH2-MIL-53(Al)] functionalized TiO2 nanotube photoanodes for highly stable and efficient photoelectrochemical cathodic protection of nickel-coated Mg alloy. J. Mater. Sci. Technol. 182, 67–78. https://doi.org/10.1016/j.jmst.2023.09.038 (2024).
Ikreedeegh, R. R., Hossen, M. A., Tahir, M. & Aziz, A. A. A comprehensive review on anodic TiO2 nanotube arrays (TNTAs) and their composite photocatalysts for environmental and energy applications: Fundamentals, recent advances and applications. Coord. Chem. Rev. 499, 215495. https://doi.org/10.1016/j.ccr.2023.215495 (2024).
Hossen, M. A. et al. Optimization of anodizing parameters for the morphological properties of TiO2 nanotubes based on response surface methodology. Next Mater. 2, 100061. https://doi.org/10.1016/j.nxmate.2023.100061 (2024).
Li, W. et al. Recent progress in g-C3N4–Based materials for remarkable photocatalytic sustainable energy. Int. J. Hydrogen Energy. 47, 21067–21118. https://doi.org/10.1016/j.ijhydene.2022.04.247 (2022).
Alam, K. M. et al. Low bandgap carbon nitride nanoparticles incorporated in Titania nanotube arrays by in situ electrophoretic anodization for photocatalytic CO2 reduction. Chem. Eng. J. 456, 141067. https://doi.org/10.1016/j.cej.2022.141067 (2023).
Hossen, M. A., Ikreedeegh, R. R., Aziz, A. A., Zerga, A. Y. & Tahir, M. Carbon-based nanomaterials (CNMs) modified TiO2 nanotubes (TNTs) photo-driven catalysts for sustainable energy and environmental applications: A comprehensive review. J. Environ. Chem. Eng. 12, 114088. https://doi.org/10.1016/j.jece.2024.114088 (2024).
Liu, Y. et al. Investigating the impact of pretreatment strategies on photocatalyst for accurate CO2RR productivity quantification: A machine learning approach. Chem. Eng. J. 473, 145255. https://doi.org/10.1016/j.cej.2023.145255 (2023).
Kim, C. M., Jaffari, Z. H., Abbas, A., Chowdhury, M. F. & Cho, K. H. Machine learning analysis to interpret the effect of the photocatalytic reaction rate constant (k) of semiconductor-based photocatalysts on dye removal. J. Hazard. Mater. 465, 132995. https://doi.org/10.1016/j.jhazmat.2023.132995 (2024).
Nguyen, H. N., Tran, Q. T., Ngo, C. T., Nguyen, D. D. & Tran, V. Q. Solar energy prediction through machine learning models: A comparative analysis of regressor algorithms. PLoS One. 20, 1–23. https://doi.org/10.1371/journal.pone.0315955 (2025).
Zhang, Y. & Xu, X. Machine learning band gaps of Doped-TiO2 photocatalysts from structural and morphological parameters. ACS Omega. 5, 15344–15352. https://doi.org/10.1021/acsomega.0c01438 (2020).
Mikolajczyk, A. et al. Visible-light photocatalytic activity of rare-earth-metal-doped TiO2: experimental analysis and machine learning for virtual design. Appl. Catal. B Environ. 346, 123744. https://doi.org/10.1016/j.apcatb.2024.123744 (2024).
Li, X. et al. Combining machine learning and high-throughput experimentation to discover photocatalytically active organic molecules. Chem. Sci. 12, 10742–10754. https://doi.org/10.1039/d1sc02150h (2021).
Tao, Q. et al. Machine learning aided design of perovskite oxide materials for photocatalytic water splitting. J. Energy Chem. 60, 351–359. https://doi.org/10.1016/j.jechem.2021.01.035 (2021).
Yurova, V. Y., Potapenko, K. O., Aliev, T. A., Kozlova, E. A. & Skorb, E. V. Optimization of g-C3N4 synthesis parameters based on machine learning to predict the efficiency of photocatalytic hydrogen production. Int. J. Hydrogen Energy. 81, 193–203. https://doi.org/10.1016/j.ijhydene.2024.07.245 (2024).
Jose, A., Devijver, E., Jakse, N. & Poloni, R. Informative training data for efficient property prediction in Metal-Organic frameworks by active learning. J. Am. Chem. Soc. 146, 6134–6144. https://doi.org/10.1021/jacs.3c13687 (2024).
Liu, Q. et al. Data-driven for accelerated design strategy of photocatalytic degradation activity prediction of doped TiO2 photocatalyst. J. Water Process. Eng. 49, 103126. https://doi.org/10.1016/j.jwpe.2022.103126 (2022).
Ahmed, Y. et al. Optimizing photocatalytic dye degradation: A machine learning and metaheuristic approach for predicting methylene blue in contaminated water. Results Eng. 25, 103538. https://doi.org/10.1016/j.rineng.2024.103538 (2025).
Salahshoori, I., Namayandeh Jorabchi, M., Baghban, A. & Khonakdar, H. A. Integrative analysis of multi machine learning models for Tetracycline photocatalytic degradation with MOFs in wastewater treatment. Chemosphere 350, 141010. https://doi.org/10.1016/j.chemosphere.2023.141010 (2024).
Javed, M. F. et al. Evaluation of machine learning models for predicting TiO2 photocatalytic degradation of air contaminants. Sci. Rep. 14, 1–25. https://doi.org/10.1038/s41598-024-64486-7 (2024).
Zhou, J., Wu, Z., Jin, C. & Zhang, J. X. J. Machine learning assisted dual-functional nanophotonic sensor for organic pollutant detection and degradation in water. Npj Clean. Water. 7 https://doi.org/10.1038/s41545-023-00292-4 (2024).
Mohammadzadeh kakhki, R. & Mohammadpoor, M. Machine learning-driven approaches for synthesizing carbon Dots and their applications in photoelectrochemical sensors. Inorg. Chem. Commun. 159, 111859. https://doi.org/10.1016/j.inoche.2023.111859 (2024).
Haghshenas, Y. et al. Predicting the rates of photocatalytic hydrogen evolution over cocatalyst-deposited TiO2 using machine learning with active photon flux as a unifying feature. EES Catal. 2, 612–623. https://doi.org/10.1039/d3ey00246b (2023).
Saadetnejad, D., Oral, B., Can, E. & Yıldırım, R. Machine learning analysis of gas phase photocatalytic CO2 reduction for hydrogen production. Int. J. Hydrogen Energy. 47, 19655–19668. https://doi.org/10.1016/j.ijhydene.2022.02.030 (2022).
Özsoysal, S., Oral, B. & Yıldırım, R. Analysis of photocatalytic CO2 reduction over MOFs using machine learning. J. Mater. Chem. A. 12, 5748–5759. https://doi.org/10.1039/d3ta07001h (2024).
Zhang, C., Liu, J., Huang, X., Chen, D. & Xu, S. Multistage polymerization design for g-C3N4 nanosheets with enhanced photocatalytic activity by modifying the polymerization process of melamine. ACS Omega. 4, 17148–17159. https://doi.org/10.1021/acsomega.9b01510 (2019).
Ikreedeegh, R. R., Hossen, M. A. & Tahir, M. Noble-Metal-Free Modified TiO2 Nanotube Arrays (TNTAs) for Efficient Photocatalytic Reduction of CO2 to CO Under Visible Light, ChemistrySelect 9 1–11. (2024). https://doi.org/10.1002/slct.202403536
Shan, Y., Wu, Q., Yuan, H. & Liu, W. Develop machine learning-based model and automated process for predicting liquid heat capacity of organics at different temperatures. Fluid Phase Equilib. 584, 114132. https://doi.org/10.1016/j.fluid.2024.114132 (2024).
Hosseinzadeh, A. et al. Machine learning-based modeling and analysis of PFOS removal from contaminated water by nanofiltration process. Sep. Purif. Technol. 289, 120775. https://doi.org/10.1016/j.seppur.2022.120775 (2022).
Casas, A., Rodríguez-Llorente, D., Rodríguez-Llorente, G., García, J. & Larriba, M. Machine learning screening tools for the prediction of extraction yields of pharmaceutical compounds from wastewaters. J. Water Process. Eng. 62, 105379. https://doi.org/10.1016/j.jwpe.2024.105379 (2024).
Yang, L. & Shami, A. On hyperparameter optimization of machine learning algorithms: theory and practice. Neurocomputing 415, 295–316. https://doi.org/10.1016/j.neucom.2020.07.061 (2020).
Hanifi, S., Cammarono, A. & Zare-Behtash, H. Advanced hyperparameter optimization of deep learning models for wind power prediction. Renew. Energy. 221, 119700. https://doi.org/10.1016/j.renene.2023.119700 (2024).
Li, J., Liu, X., Wang, H., Sun, Y. & Dong, F. Prediction and interpretation of photocatalytic NO removal on g-C3N4-based catalysts using machine learning. Chin. Chem. Lett. 35, 108596. https://doi.org/10.1016/j.cclet.2023.108596 (2024).
Wu, Z., Luo, J., Rincon, D. & Christofides, P. D. Machine learning-based predictive control using noisy data: evaluating performance and robustness via a large-scale process simulator. Chem. Eng. Res. Des. 168, 275–287. https://doi.org/10.1016/j.cherd.2021.02.011 (2021).
Adán, C., Marugán, J., Sánchez, E., Pablos, C. & Van Grieken, R. Understanding the effect of morphology on the photocatalytic activity of TiO2 nanotube array electrodes. Electrochim. Acta. 191, 521–529. https://doi.org/10.1016/j.electacta.2016.01.088 (2016).
Bamola, P. et al. Effect of nanotube diameter on the photocatalytic activity of bimetallic AgAu nanoparticles grafted 1D-TiO2 nanotubes. J. Mater. Sci. Mater. Electron. 32, 1427–1444. https://doi.org/10.1007/s10854-020-04914-2 (2021).
Hossen, M. A. et al. Enhanced photocatalytic CO2 reduction to CH4 using novel ternary photocatalyst RGO/Au-TNTAs. Energies 16 https://doi.org/10.3390/en16145404 (2023).
Ikreedeegh, R. R. & Tahir, M. Photocatalytic CO2 reduction to CO and CH4 using g-C3N4/RGO on Titania nanotube arrays (TNTAs). J. Mater. Sci. 56, 18989–19014. https://doi.org/10.1007/s10853-021-06516-7 (2021).
Arora, I., Chawla, H., Chandra, A., Sagadevan, S. & Garg, S. Advances in the strategies for enhancing the photocatalytic activity of TiO2: conversion from UV-light active to visible-light active photocatalyst. Inorg. Chem. Commun. 143, 109700. https://doi.org/10.1016/j.inoche.2022.109700 (2022).
Ishaq, T. et al. Recent strategies to improve the photocatalytic efficiency of TiO2 for enhanced water splitting to produce hydrogen. Catalysts 14 https://doi.org/10.3390/catal14100674 (2024).
Sim, L. C. et al. In situ growth of g-C3N4 on TiO2 nanotube arrays: construction of heterostructures for improved photocatalysis properties. J. Environ. Chem. Eng. 8 https://doi.org/10.1016/j.jece.2019.103611 (2020).
Zhang, F., Liu, J., Yue, H., Cheng, G. & Xue, X. Construction of g-C3N4 nanoparticles modified TiO2 nanotube arrays with Z-scheme heterojunction for enhanced photoelectrochemical properties. J. Mater. Sci. 58, 2676–2688. https://doi.org/10.1007/s10853-022-07730-7 (2023).
Xin, S. et al. Enhanced visible light photoelectrocatalytic degradation of o-chloronitrobenzene through surface plasmonic Au nanoparticles and g-C3N4 co-modified TiO2 nanotube arrays photoanode. Appl. Catal. B Environ. 323, 122174. https://doi.org/10.1016/j.apcatb.2022.122174 (2023).
Hossen, M. A. et al. Experimental and AI-driven enhancements in gas-phase photocatalytic CO2 conversion over synthesized highly ordered anodic TiO2 nanotubes. Energy Convers. Manag. 327, 119544. https://doi.org/10.1016/j.enconman.2025.119544 (2025).
Ahmed, Y. et al. A novel interpretable machine learning and metaheuristic-based protocol to predict and optimize Ciprofloxacin antibiotic adsorption with nano-adsorbent. J. Environ. Manage. 370, 122614. https://doi.org/10.1016/j.jenvman.2024.122614 (2024).
Albahri, A. S. et al. A systematic review of trustworthy and explainable artificial intelligence in healthcare: assessment of quality, bias risk, and data fusion. Inf. Fusion. 96, 156–191. https://doi.org/10.1016/j.inffus.2023.03.008 (2023).
Hancock, J. T. & Khoshgoftaar, T. M. CatBoost for big data: an interdisciplinary review. J. Big Data. 7 https://doi.org/10.1186/s40537-020-00369-8 (2020).
Javed, M. F., Fawad, M., Lodhi, R., Najeh, T. & Gamil, Y. Forecasting the strength of Preplaced aggregate concrete using interpretable machine learning approaches. Sci. Rep. 14, 1–28. https://doi.org/10.1038/s41598-024-57896-0 (2024).
Subba, S. & Chatterjee, S. Machine learning-driven determination of key absorber layer parameters in perovskite solar cells. Mater. Today Commun. 42, 111113. https://doi.org/10.1016/j.mtcomm.2024.111113 (2025).
Delavari, S., Amin, N. A. S. & Ghaedi, M. Photocatalytic conversion and kinetic study of CO and CH4 over nitrogen-doped Titania nanotube arrays. J. Clean. Prod. 111, 143–154. https://doi.org/10.1016/j.jclepro.2015.07.077 (2016).
Li, H., Gao, Y., Xiong, Z., Liao, C. & Shih, K. Enhanced selective photocatalytic reduction of CO2 to CH4 over plasmonic Au modified g-C3N4 photocatalyst under UV–vis light irradiation. Appl. Surf. Sci. 439, 552–559. https://doi.org/10.1016/j.apsusc.2018.01.071 (2018).
Ikreedeegh, R. R. & Tahir, M. Facile fabrication of well-designed 2D/2D porous g-C3N4–GO nanocomposite for photocatalytic methane reforming (DRM) with CO2 towards enhanced Syngas production under visible light. Fuel 305, 121558. https://doi.org/10.1016/j.fuel.2021.121558 (2021).
Fu, J., Jiang, K., Qiu, X., Yu, J. & Liu, M. Product selectivity of photocatalytic CO2 reduction reactions. Mater. Today. 32, 222–243. https://doi.org/10.1016/j.mattod.2019.06.009 (2020).
Kumar, D. P. et al. Highly stable and durable ZnIn2S4 nanosheets wrapped oxygen deficient blue TiO2(B) catalyst for selective CO2 photoreduction into CO and CH4. J. Colloid Interface Sci. 651, 264–272. https://doi.org/10.1016/j.jcis.2023.07.197 (2023).
Zhang, Y. et al. Modulation of Fe–MOF via second-transition metal ion doping (Ti, Mn, Zn, Cu) for efficient visible-light driven CO2 reduction to CH4, Sep. Purif. Technol. 336, 126164. https://doi.org/10.1016/j.seppur.2023.126164 (2024).
Acknowledgements
The authors highly acknowledged the support from the Chittagong University of Engineering and Technology (CUET), Bangladesh and Universiti Malaysia Pahang Al-Sultan Abdullah (UMPSA), Malaysia. Nadeem A Khan, extend their appreciation to the deanship of research and graduate studies at King Khalid University for supporting the work through a large research project under grant number RGP.2/67/46.
Author information
Authors and Affiliations
Contributions
Md. Arif Hossen : Conceptualization, Methodology, Data collection, Data analysis, Writing–original draft, Writing – review & editing. Meherunnesa Prima: Data curation, Methodology, Software, Visualization, Writing – review & drafting. Azrina Abd Aziz : Funding acquisition, Project administration, Writing– review & editing. Nadeem A Khan : Writing – review & editing. Yunus Ahmed : Conceptualization, Software, Supervision, Visualization, Writing – review & editing.
Corresponding authors
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
About this article
Cite this article
Hossen, M.A., Prima, M., Aziz, A.A. et al. Bridging photocatalysis and artificial intelligence to maximize CH4 and CO production from CO2 reduction using synthesized g-C3N4/TNTAs photocatalysts. Sci Rep (2026). https://doi.org/10.1038/s41598-026-36838-y
Received:
Accepted:
Published:
DOI: https://doi.org/10.1038/s41598-026-36838-y
