A machine learning based electronic property predictor of Cu2SnS3 thin film synthesized by ultrasonic spray pyrolysis
Source abstract
Abstract
Copper tin sulphide (CTS) is an important material used as an absorber layer for low cost solar cells. Variation of the elemental constituents of CTS and the substrate temperature cause a significant alteration in its electronic property, which may cause unpredictable change in the device efficiency. In view of the same, an electronic property predictor for CTS is developed in this work based on real-time experimental data for accurately predicting the resistivity, mobility, carrier concentration and bandgap with variation in the atomic constituents of CTS and the substrate temperature. Seven machine learning (ML) algorithms are in-built in the predictor. Prediction performance is carried out with all these ML models on the basis of mean absolute percentage error (MAPE), coefficient of determination (R 2 ) and root mean square error (RMSE). Decision Tree is found to be the best in accuracy towards predicting the resistivity (MAPE - 1.634 Ωcm, R 2 - 0.93 and RMSE = 0.049Ωcm), XGBoost in predicting the mobility (MAPE - 0.567 cm 2/Vs, R 2 - 0.409, RMSE- 0.429 cm 2/Vs) and Random Forest in predicting the bandgap (MAPE- 0.017 eV, R 2 - 0.863, RMSE - 0.051 eV) and the logarithmic values of carrier concentration (MAPE - 0.006, R 2 - 0.715, RMSE - 0.562). The results so obtained are from limited number of 199 data samples and addition of more data can further boost the accuracy. This know-how of electronic property thus becomes a very crucial factor in accurately determining the absorption coefficient of a particular CTS sample and its role in enhancing the solar cell efficiency.
Source: expertdata · Quality: candidate unverified · Source record ↗
Research context
OpenAlex topics: Chalcogenide Semiconductor Thin Films · Gas Sensing Nanomaterials and Sensors · Machine Learning in Materials Science
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
References (31)
Show source-provided references
1. Solar energy technology and its roles in sustainable development10.1093/ce/zkac023 ↗
2. 24·5% Efficiency silicon PERT cells on MCZ substrates and 24·7% efficiency PERL cells on FZ substrates10.1002/(SICI)1099-159X(199911/12)7:6<471::AID-PIP298>3.0.CO;2-7 ↗
3. The path to 25% silicon solar cell efficiency: history of silicon cell evolution10.1002/pip.892 ↗
4. Elongated nanostructures for radial junction solar cells10.1088/0034-4885/76/10/106502 ↗
5. Potential environmental risk of solar cells: current knowledge and future challenges
6. Optimization and fabrication of low cost Cu2SnS3/ZnS thin film heterojunction solar cell using ultrasonic spray pyrolysis10.1016/j.optmat.2021.111838 ↗
7. J.J. Chaudhari, S. Patel, U.S. Joshi, Optimization of low cost, non toxic, earth abundant p-type Cu2SnS3 thin film for Photovoltaic application, in: IOP Conf. Ser. Mater. Sci. Eng., 2016. https://doi.org/10.1088/1757-899X/149/1/012165.10.1088/1757-899X/149/1/012165 ↗
8. Preparation and characterization of Cu2 SnS3 ternary semiconductor nanostructures via the spray pyrolysis technique for photovoltaic applications10.1088/0031-8949/85/03/035603 ↗
9. Improved Jsc by increasing the absorber layer thickness of monoclinic-dominated Cu2SnS3 thin film solar cells fabricated on flexible mo foil10.1002/solr.202100743 ↗
10. Optoelectrical and structural characterization of Cu2SnS3 thin films grown via spray pyrolysis using stable molecular ink10.1016/j.solener.2021.05.088 ↗
11. Cu2SnS3 and Cu4SnS4 thin films via chemical deposition for photovoltaic application10.1149/1.3384660 ↗
12. Photovoltaic characteristics of thin films of Cu2SnS310.1016/0165-1633(87)90019-0 ↗
13. Cu2Sn1-xGexS3 (x = 0.17) thin-film solar cells with high conversion efficiency of 6.0%10.7567/APEX.6.045501 ↗
14. Fabrication of Cu2SnS3 thin film solar cells via a sol-gel technique in air10.1016/j.physb.2021.413613 ↗
15. Fabrication of Cu2SnS3 thin-film solar cells with power conversion efficiency of over 4%10.7567/JJAP.54.08KC06 ↗
16. Thin film solar cell using earth abundant Cu2SnS3 (CTS) fabricated through spray pyrolysis: Influence of precursors
17. Effect of copper concentration on CTS thin films for solar cell absorber layer and photocatalysis applications
18. Temperature dependent growth of Cu2SnS3 thin films using ultrasonic spray pyrolysis for solar cell absorber layer and photocatalytic application10.1088/2053-1591/ab3928 ↗
19. Machine learning tools to predict hot injection syntheses outcomes for II-VI and IV-VI quantum dots10.1021/acs.jpcc.0c05993 ↗
20. Machine learning assisted analysis, prediction, and fabrication of high-efficiency CZTSSe thin film solar cells10.1002/adfm.202303459 ↗
21. Machine learning with knowledge constraints for process optimization of open-air perovskite solar cell manufacturing10.1016/j.joule.2022.03.003 ↗
22. Ridge regression: biased estimation for nonorthogonal problems10.1080/00401706.2000.10485983 ↗
23. Ridge regression ensemble of machine learning models applied to solar and wind forecasting in Brazil and Spain10.1016/j.apenergy.2022.118936 ↗
24. Regression shrinkage and selection via the lasso10.1111/j.2517-6161.1996.tb02080.x ↗
25. Regression shrinkage and selection via the lasso: a retrospective
26. A small review and further studies on the LASSO
27. Regularization and variable selection via the elastic net
28. Deisenroth · 2020
29. https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.ElasticNet.html.
30. https://scikit-learn.org/stable/modules/generated/sklearn.tree.DecisionTreeRegressor.html.
31. https://xgboost.readthedocs.io/en/stable/parameter.html.