{"title":"Hybrid modeling with supervised learning and SCAPS simulations for performance analysis of Cs0.17FA0.83PbI3−xBrx perovskite solar cells","authors":"Subham Subba, Joy Sarkar, Suman Chatterjee","doi":"10.1016/j.mseb.2025.118482","DOIUrl":null,"url":null,"abstract":"<div><div>In this study, we integrate numerical simulations with supervised learning to predict the performance of Cs<sub>0.17</sub>FA<sub>0.83</sub>PbI<span><math><msub><mrow></mrow><mrow><mn>3</mn><mo>−</mo><mi>x</mi></mrow></msub></math></span>Br<span><math><msub><mrow></mrow><mrow><mi>x</mi></mrow></msub></math></span>-based perovskite solar cells. A dataset of 3,240 points was generated using SCAPS by varying absorber thickness (<span><math><mi>t</mi></math></span>), defect density (<span><math><msub><mrow><mi>N</mi></mrow><mrow><mi>t</mi></mrow></msub></math></span>), and donor density (<span><math><msub><mrow><mi>N</mi></mrow><mrow><mi>d</mi></mrow></msub></math></span>) across six bromine compositions (x = 0, 0.5, 1, 1.5, 2, and 2.5). Four Random Forest models were trained to predict power conversion efficiency (PCE), open-circuit voltage (Voc), short-circuit current density (Jsc), and fill factor (FF), achieving test R<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span> values consistently above 0.99. The corresponding RMSE values were 0.197%, 0.008 V, 0.317 mA/cm<span><math><msup><mrow></mrow><mrow><mn>2</mn></mrow></msup></math></span>, and 1.181% for PCE, Voc, Jsc, and FF, respectively. To validate generalization, the models were tested on Cs<sub>0.17</sub>FA<sub>0.83</sub>PbI<sub>2.6</sub>Br<sub>0.4</sub> composition, showing strong agreement with simulations. Feature correlation analysis identified <span><math><msub><mrow><mi>N</mi></mrow><mrow><mi>t</mi></mrow></msub></math></span> and <span><math><msub><mrow><mi>N</mi></mrow><mrow><mi>d</mi></mrow></msub></math></span> as key performance factors. This approach can be extended to other perovskite compositions, device configurations, transport layers, and alternative ML techniques for improved generalization.</div></div>","PeriodicalId":18233,"journal":{"name":"Materials Science and Engineering: B","volume":"321 ","pages":"Article 118482"},"PeriodicalIF":3.9000,"publicationDate":"2025-06-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Materials Science and Engineering: B","FirstCategoryId":"88","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0921510725005069","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
In this study, we integrate numerical simulations with supervised learning to predict the performance of Cs0.17FA0.83PbIBr-based perovskite solar cells. A dataset of 3,240 points was generated using SCAPS by varying absorber thickness (), defect density (), and donor density () across six bromine compositions (x = 0, 0.5, 1, 1.5, 2, and 2.5). Four Random Forest models were trained to predict power conversion efficiency (PCE), open-circuit voltage (Voc), short-circuit current density (Jsc), and fill factor (FF), achieving test R values consistently above 0.99. The corresponding RMSE values were 0.197%, 0.008 V, 0.317 mA/cm, and 1.181% for PCE, Voc, Jsc, and FF, respectively. To validate generalization, the models were tested on Cs0.17FA0.83PbI2.6Br0.4 composition, showing strong agreement with simulations. Feature correlation analysis identified and as key performance factors. This approach can be extended to other perovskite compositions, device configurations, transport layers, and alternative ML techniques for improved generalization.
期刊介绍:
The journal provides an international medium for the publication of theoretical and experimental studies and reviews related to the electronic, electrochemical, ionic, magnetic, optical, and biosensing properties of solid state materials in bulk, thin film and particulate forms. Papers dealing with synthesis, processing, characterization, structure, physical properties and computational aspects of nano-crystalline, crystalline, amorphous and glassy forms of ceramics, semiconductors, layered insertion compounds, low-dimensional compounds and systems, fast-ion conductors, polymers and dielectrics are viewed as suitable for publication. Articles focused on nano-structured aspects of these advanced solid-state materials will also be considered suitable.