-
为了摆脱对化石燃料的过度依赖和发展绿色低碳的新能源。加速无铅无机卤化物钙钛矿在太阳能电池中的应用，开发具合适带隙的新材料至关重要。然而，传统实验与密度泛函理论（DFT）计算都需要考虑耗时、成本和准确率等问题。本研究利用了深度学习与DFT计算的协同策略，通过深度神经网络（DNN）模型精准预测55种无机卤化物双钙钛矿材料的带隙宽度。基于1181种双钙钛矿材料构建的数据库，本研究系统对比了五种机器学习模型：随机森林回归（RFR）、梯度提升回归（GBR）、支持向量回归（SVR）、极限梯度提升回归（XGBR）及深度神经网络（DNN）模型。结果表明，DNN凭借其强大的非线性映射与高维特征自动提取能力，在带隙预测中展现出卓越准确性（测试集MAE降至0.264 eV，R2提升至0.925）与泛化性能。利用这种策略，成功筛选出四种有前景的无机双钙钛矿候选材料： Cs2GaAgCl6、Cs2AgIrF6、Cs2InAgCl6、Cs2AlAgBr6。其中，Cs2AgIrF6和Cs2AlAgBr6的性能尤为突出，他们的带隙分1.36eV和1.20eV。其模拟效率分别达到23.71%（VOC=0.94 V,JSC=31.19 mA/cm2,FF=80.81%）和22.37%（VOC=0.78 V,JSC=36.73 mA/cm2,FF=77.66%），较高的开路电压和填充因子表明其具有优异的载流子分离效率与较低的器件内部非辐射复合损失。在本次研究中利用的深度学习与DFT计算的协同策略，加速了DFT数据的解析与规律挖掘，为高性能、高稳定性的环保无铅钙钛矿太阳能电池的理性设计提供了新思路。Accelerating the application of lead-free inorganic halide perovskites in solar cells necessitates the development of novel perovskite materials with suitable bandgap widths, high stability, and environmental friendliness. This represents a crucial pathway for driving photovoltaic technology innovation and reducing reliance on conventional fossil fuels. However, traditional material development paradigms heavily depend on trial-and error experimental screening or pure density functional theory (DFT) calculations, which incur significant time and material costs.
To address these challenges, this study innovatively proposes and implements an efficient screening strategy based on the synergy between deep learning and DFT calculations. By constructing a database containing 1181 inorganic halide double perovskite materials, we systematically trained and compared the performance of five mainstream machine learning models for the bandgap prediction task: Random Forest Regression (RFR), Gradient Boosting Regression (GBR), Support Vector Regression (SVR), eXtreme Gradient Boosting Regression (XGBR), and a Deep Neural Network (DNN) model. Results demonstrate that the DNN model, leveraging its powerful nonlinear mapping capability and advantage in automatic high-dimensional feature extraction, achieved exceptional prediction accuracy on the test set, with the Mean Absolute Error (MAE) significantly reduced to 0.264 eV and the coefficient of determination (R2) reaching 0.925. Its performance was markedly superior to other compared models, highlighting the immense potential of deep learning in predicting complex material properties.
Using this optimized DNN model, this study successfully screened four promising inorganic double perovskite candidates from 55 potential materials: Cs2GaAgCl6, Cs2AgIrF6, Cs2InAgCl6, and Cs2AlAgBr6. Among them, Cs2AgIrF6 and Cs2AlAgBr6 performed particularly well, with predicted bandgaps of 1.36 eV and 1.20 eV, respectively. This range ideally matches the requirement for efficient light absorption in solar cells. Further device performance simulations revealed that the solar cell based on Cs2AgIrF6 achieved a simulated power conversion efficiency (PCE) of 23.71%, with an open-circuit voltage (VOC) of 0.94 V, a short-circuit current density (JSC) of 31.19 mA/cm2, and a fill factor (FF) of 80.81%. Cs2AlAgBr6 also exhibited a simulated efficiency of 22.37%, corresponding to VOC=0.78 V, JSC=36.73 mA/cm2, and FF=77.66%. Notably, both materials demonstrated high open-circuit voltages and fill factors, clearly indicating excellent carrier separation efficiency and significantly reduced nonradiative recombination losses within these materials.
In summary, this study successfully validates the significant efficacy of the deep learning-DFT synergistic strategy in accelerating the discovery of novel lead-free perovskite materials. This method not only substantially enhances the efficiency of DFT data analysis and the depth of pattern mining, overcoming some bottlenecks associated with traditional highthroughput calculations, but more importantly, it provides a practical and highly innovative approach for the rational design of high-performance, stable, and environmentally friendly lead-free perovskite solar cells, holding positive implications for advancing green, low-carbon energy technologies.-
Keywords:
- Perovskite /
- Deep Learning /
- Density Functional Theory (DFT) /
- Solar Cell
-
[1] Smal T, Wieprow J 2023 ENERGIES 16 1605
[2] Ingram T, Wieczorek-Kosmala M, Hlavácek K 2023 Energies 16 702
[3] Kannan N, Vakeesan D 2016 Renew. Sust. Energy Rev 62 1092
[4] Liang Z, Zhang Y, Xu H, Chen W, Liu B, Zhang J, Zhang H, Wang Z, Kang D-H, Zeng J, Gao X, Wang Q, Hu H, Zhou H, Cai X, Tian X, Reiss P, Xu B, Kirchartz T, Xiao Z, Dai S, Park N-G, Ye J, Pan X 2023 Nature 624 557
[5] Li Y, Ru X, Yang M, Zheng Y, Yin S, Hong C, Peng F, Qu M, Xue C, Lu J, Fang L, Su C, Chen D, Xu J, Yan C, Li Z, Xu X, Shao Z 2024 Nature 626 105
[6] Razzaq A, Allen T G, Liu W Z, Liu Z X, De Wolf S 2022 JOULE 6 514
[7] Urbaniak A, Czudek A, Dagar J, Unger E L 2022 Sol. Energy Mater. Sol. Cells 238 111618
[8] Xie G S, Li H, Fang J, Wang X, Peng H C, Lin D X, Huang N S, Gan L, Li W J, Jiang R X, Bu T L, Huang F Z, He S S, Qiu L B 2025 Angew. Chem. Int. Ed 64 e202501764
[9] Chen Y H, Feng Z J, Pal A, Zhang J C 2022 Phys. Status Solidi A 219 2200018
[10] Kwon N, Lee J, Ko M J, Kim Y Y, Seo J 2023 Nano Convergence 10 28
[11] Ge Y S, Zheng L K, Wang H B, Gao J, Yao F, Wang C, Li G, Guan H L, Wang S X, Cui H S, Ye F H, Shao W L, Zheng Z M, Yu Z X, Wang J H, Xu Z X, Dai C J, Ma Y H, Yang Y, Guan Z Q, Liu Y, Wang J B, Lin Q Q, Li Z Y, Li X, Ke W J, Grätzel M, Fang G J 2025 Nat. Photonics 19 170
[12] Ahn N, Choi M 2024 Adv. Sci 11 2306110
[13] Sun Y Q, Li F M, Zhang H, Liu W Z, Wang Z H, Mao L, Li Q, He Y L, Yang T, Sun X G, Qian Y C, Ma Y Y, Zhang L P, Du J L, Shi J H, Wang G Y, Han A J, Wang N, Meng F Y, Liu Z X, Liu M Z 2025 Nat. Commun 16 5733
[14] Geng L, Ma Y Y, Sun Y Q, Cai Z L, Lan L, Ma H C, Zhang H, Mao L, Li F M, Liu M Z 2025 Adv. Mater 37 2419018
[15] Zhang T, Cai Z H, Chen S Y 2020 ACS Appl. Mater. Interfaces 12 20680
[16] Murugan S, Lee E C 2023 Materials 16 5275
[17] Cao Y, Zhang Y Q, Bao Y N, Shi L L, Cao G Y, Qin L L, Wang C L, Li X F, Yang Z H 2025 Nano Energy 142 111228
[18] Zhou P P, Xiao X Z, Zhu X Y, Chen Y P, Lu W M, Piao M Y, Cao Z M, Lu M, Fang F, Li Z N, Jiang L J, Chen L X 2023 Energy Storage Mater 63 102964
[19] Al-Sabana O, Abdellatif S O 2022 Optoelectron. Lett 18 148
[20] Hariharan A, Ackermann M, Koss S, Khosravani A, Schleifenbaum J H, Köhnen P, Kalidindi S R, Haase C 2025 Adv. Sci 12 2414880
[21] Wang Y X, Li Y, Tang Z C, Li H, Yuan Z L, Tao H G, Zou N L, Bao T, Liang X H, Chen Z Z, Xu S H, Bian C, Xu Z M, Wang C, Si C, Duan W H, Xu Y 2024 Sci. Bull 69 2514
[22] Usoltsev A N, Korobeynikov N A, Novikov A S, Plyusnin P E, Kolesov B A, Fedin V P, Sokolov M N, Adonin S A 2020 Inorg. Chem 59 17320
[23] Chen Y X, Zhang L F, Wang H, Weinan E 2021 J. Chem. Theory Comput 17 170
[24] Absike H, Baaalla N, Attou L, Labrim H, Hartiti B, Ez-zahraouy H 2022 Solid State Commun 345 114684
[25] Jebari H, Ouichou H, Hamideddine I, Boudad L, Tahiri N, El Mansouri A, El Bounagui O, Taibi M, Ez-Zahraouy H 2023 Comput. Theor. Chem 1220 113993
[26] Wan X Y, Zhang Y H, Lu S H, Wu Y L, Zhou Q H, Wang J L 2022 Acta Phys. Sin 71 177101
[27] Hu W G, Zhang L 2023 Mater. Today Commun 35 105841
[28] Lu S H, Zhou Q H, Ouyang Y X, Guo Y L, Li Q, Wang J L 2018 Nat. Commun 9 3405
[29] Yang C, Chong X Y, Hu M Y, Yu W, He J J, Zhang Y L, Feng J, Zhou Y Y, Wang L W 2023 ACS Appl. Mater. Interfaces 15 40419
[30] Zhang R Q, Motes B, Tan S, Lu Y L, Shih M C, Hao Y L, Yang K R, Srinivasan S, Bawendi M G, Bulovic V 2025 ACS Energy Lett 10 1714
[31] Jain A, Ong S P, Hautier G, Chen W, Richards W D, Dacek S, Cholia S, Gunter D, Skinner D, Ceder G, Persson K A 2013 APL Mater 1 011002
[32] Kresse, Furthmuller 1996 Phys. Rev. B 54 11169
[33] Segall M D, Philip J D L, Probert M J, Pickard C J, Hasnip P J, Clark S J, Payne M C 2002 J. Phys.: Condens. Matter 14 2717
[34] Ming C, Yang K, Zeng H, Zhang S B, Sun Y Y 2020 Mater. Horiz 7 2985
[35] Huck P, Gunter D, Cholia S, Winston D, N'Diaye A T, Persson K 2016 Concurr. Comput 28 1982
[36] Li W B, Li Y H, Zhang Z L, Gao P 2023 J. Energy Chem 84 347
[37] Al-Daraghmeh T M, Zayed O, Zelai T, Saba S, Mustafa G M, Hakami O, Albalawi H, Bouzgarrou S, Mahmoud Z, Mahmood Q 2023 J. Solid State Chem 322 124003
[38] Du K-z, Meng W, Wang X, Yan Y, Mitzi D B 2017 Angew. Chem. Int. Ed 56 8158
[39] White J, Power S D 2023 Sensors 23 6077
[40] Chen Z M, Wang J, Li C J, Liu B Q, Luo D X, Min Y G, Fu N Q, Xue Q F 2024 J. Mater. Chem. C 12 15444
[41] Nasir I M, Khan M A, Yasmin M, Shah J H, Gabryel M, Scherer R, Damasevicius R 2020 Sensors 20 6793
[42] Wang Q Y, Wang Z, Mizuguchi K, Takao T 2025 Sci. Adv 11 eadt2624
[43] Wenzel J, Matter H, Schmidt F 2019 J. Chem. Inf. Model 59 1253
[44] Hornik K, Stinchcombe M, White H 1989 Neural Netw 2 359
[45] Beniaguev D, Segev I, London M 2021 Neuron 109 2727
[46] Smajic A, Grandits M, Ecker G F 2022 J. Cheminform 14 54
[47] Zhu S, Yu C, Hu J 2024 Appl. Soft Comput 154 111337
[48] Sun Y T, Ding S F, Zhang Z C, Jia W K 2021 Soft Comput 25 5633
[49] Browne 2000 J. Math. Psychol 44 108
[50] Yates L A, Aandahl Z, Richards S A, Brook B W 2023 Ecol. Monogr 93 e1557
[51] Baldi P, Brunak S, Chauvin Y, Andersen C A, Nielsen H 2000 Bioinformatics 16 412
[52] Guo H, Lorin E, Noxon C, Yang X 2025 J. Sci. Comput 103 76
[53] Huang J, Lei T, Siron M, Zhang Y, Yu S, Seeler F, Dehestani A, Quan L N, Schierle-Arndt K, Yang P 2020 Nano Lett 20 3734
[54] Deng X, Zhang Z, Zhang Z, Wu Y, Song H, Li H, Luo B 2024 Adv. Sci 11 2407751
[55] Sharma S, Ward Z D, Bhimani K, Sharma M, Quinton J, Rhone T D, Shi S F, Terrones H, Koratkar N 2023 ACS Appl. Mater. Interfaces 15 18962
[56] Liang Y, Li F, Cui X, Stampfl C, Ringer S P, Yang X, Huang J, Zheng R 2025 Sci. Adv 11 eads7054
[57] Ashfaq A, Tahir S, Mushtaq S, Alqurashi R S, Haneef M, Almousa N, Rehman U u, Bonilla R S 2023 Mater. Today Commun 35 106016
[58] Wang S, Chen C, Shi S, Zhang Z, Cai Y, Gao S, Chen W, Guo S, Abduryim E, Dong C, Guan X, Liu Y, Liu G, Lu P 2024 J. Energy Chem 95 271
[59] Yuan S, Ullah M, Tighezza A M 2025 Front. Energy 19 334
[60] Sharma S, Ward Z D, Bhimani K, Sharma M, Quinton J, Rhone T D, Shi S F, Terrones H, Koratkar N 2023 ACS Appl. Mater. Interfaces 15 18962
[61] Adekoya A H, Snyder G J 2024 Adv. Funct. Mater 34 2403926
[62] Wang R, Jia Y L, Zhang Y, Ma X J, Xu Q, Zhu Z X, Deng Y H, Xiong Z H, Gao C H 2020 Acta Phys. Sin 69 038501
[63] Nyayban A, Panda S, Chowdhury A 2023 Phys. B Condens. Matter 649 414384
[64] Wang T, Zhou D, Wang Y, Li W, Wang Y, Yang H, Song R, Wang E, Song R, Fang Y, Zhou S, Bai X, Song H 2025 Adv. Funct. Mater 35 2504436
[65] Wei Y Q, Xu L, Peng Q M, Wang J P 2019 Acta Phys. Sin 68 158506
[66] van Zeist W-J, Ren Y, Bickelhaupt F M 2010 Sci. China Chem 53 210
[67] Lu X Q, Hu R J, Zhu Y B, Song K P, Qin W 2023 NPG Asia Mater 15 38
[68] Jain A, Voznyy O, Sargent E H 2017 J. Phys. Chem. C 121 7183
[69] Bartel C J, Trewartha A, Wang Q, Dunn A, Jain A, Ceder G 2020 npj Comput. Mater 6 97
[70] Sun Q D, Yin W J 2017 J. Am. Chem. Soc 139 14905
[71] Bartel C J, Sutton C, Goldsmith B R, Ouyang R H, Musgrave C B, Ghiringhelli L M, Scheffler M 2019 Sci. Adv 5 eaav0693
[72] Fedorovskiy A E, Drigo N A, Nazeeruddin M K 2020 Small Methods 4 1900426
[73] Wang M, Hu H, Lin S, Yang P, Yang D 2023 Chin.Sci. Bull 68 3146
[74] Emery A A, Wolverton C 2017 Sci. Data 4 170153
[75] Li Y W, Na G R, Luo S L, He X, Zhang L J 2021 Acta Phys.-Chim. Sin 37 2007015
[76] Gwandu B U, Nawawi Y H, Faruk S, Zauro A S 2024 Equity J. Sci. Technol 10 10
[77] Hao L, Zhou M, Song Y, Ma X, Wu J, Zhu Q, Fu Z, Liu Y, Hou G, Li T 2021 Solar Energy 230 345
[78] Li X, Li J, Wu S, Li Y, Peng C, Wu M, Wu J, Lin J, Ma X, Huang S 2022 Solar Energy 247 315
[79] Lin C X, Liu G L, Xi X, Wang L, Wang Q Q, Sun Q Y, Li M X, Zhu B J, de Lara D P, Zai H C 2022 Materials 15 8142
[80] Shen H, Li X, Zhang X, Zhou H, Zhang H, Liu X, Zhang M, Wu J, Xiang Z, Fang W 2023 Solar Energy 253 240
[81] Li S, Zhou B, Zhao Q, Yang X, Xie Z, Duan Z, Zhao E, Hu Y 2025 Acta Phys. Sin 74 127701
[82] Hunde B R, Woldeyohannes A D 2023 Mater. Today Commun 34 105420
计量
- 文章访问数: 9
- PDF下载量: 0
- 被引次数: 0
下载:
