{"title":"Machine learning analysis of photocatalytic CO2 reduction on perovskite materials","authors":"İrem Gülçin Zırhlıoğlu, Ramazan Yıldırım","doi":"10.1016/j.materresbull.2025.113436","DOIUrl":null,"url":null,"abstract":"<div><div>A dataset containing 328 samples extracted from 66 experimental articles on photocatalytic CO<sub>2</sub> reduction over perovskite materials was constructed and analyzed using machine learning. Random forest algorithm was used to predict total product yield in gas and liquid phase separately; decision tree algorithm was also utilized to deduce heuristic rules for high performance. Unavailable band gaps were also predicted using a linear regression trained by available data. Random forest models for both phases were quite successful. R<sup>2</sup> and RMSE for liquid phase were 0.96 and 0.21, respectively for training (0.84 and 0.36 respectively for testing); for the gas phase, R<sup>2</sup> and RMSE were 0.91 and 0.22 respectively for training (0.87 and 0.24 respectively for testing). The testing accuracy of decision tree models (0.88 % for gas and 0.73 % for liquid phases) were also reasonably high. The perovskite synthesis method was the most important descriptors for both RF and DT models.</div></div>","PeriodicalId":18265,"journal":{"name":"Materials Research Bulletin","volume":"188 ","pages":"Article 113436"},"PeriodicalIF":5.3000,"publicationDate":"2025-03-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Materials Research Bulletin","FirstCategoryId":"88","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0025540825001448","RegionNum":3,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q2","JCRName":"MATERIALS SCIENCE, MULTIDISCIPLINARY","Score":null,"Total":0}
引用次数: 0
Abstract
A dataset containing 328 samples extracted from 66 experimental articles on photocatalytic CO2 reduction over perovskite materials was constructed and analyzed using machine learning. Random forest algorithm was used to predict total product yield in gas and liquid phase separately; decision tree algorithm was also utilized to deduce heuristic rules for high performance. Unavailable band gaps were also predicted using a linear regression trained by available data. Random forest models for both phases were quite successful. R2 and RMSE for liquid phase were 0.96 and 0.21, respectively for training (0.84 and 0.36 respectively for testing); for the gas phase, R2 and RMSE were 0.91 and 0.22 respectively for training (0.87 and 0.24 respectively for testing). The testing accuracy of decision tree models (0.88 % for gas and 0.73 % for liquid phases) were also reasonably high. The perovskite synthesis method was the most important descriptors for both RF and DT models.
期刊介绍:
Materials Research Bulletin is an international journal reporting high-impact research on processing-structure-property relationships in functional materials and nanomaterials with interesting electronic, magnetic, optical, thermal, mechanical or catalytic properties. Papers purely on thermodynamics or theoretical calculations (e.g., density functional theory) do not fall within the scope of the journal unless they also demonstrate a clear link to physical properties. Topics covered include functional materials (e.g., dielectrics, pyroelectrics, piezoelectrics, ferroelectrics, relaxors, thermoelectrics, etc.); electrochemistry and solid-state ionics (e.g., photovoltaics, batteries, sensors, and fuel cells); nanomaterials, graphene, and nanocomposites; luminescence and photocatalysis; crystal-structure and defect-structure analysis; novel electronics; non-crystalline solids; flexible electronics; protein-material interactions; and polymeric ion-exchange membranes.