Machine learning facilitates the development of interconnecting layers for perovskite/silicon heterojunction tandem solar cells with proof-of-concept efficiency &gt; 38%

Xuejiao Wang; Guanlan Chen; Ying Liu; Guangyi Wang; Wei Han; Jin Wang; Pengfei Liu; Jilei Wang; Shaojuan Bao; Bo Yu; Ying Liu; Xinliang Chen; Shengzhi Xu; Ying Zhao; Xiaodan Zhang

doi:10.1088/1674-4926/25050011

J. Semicond. > Just Accepted

ARTICLES

Machine learning facilitates the development of interconnecting layers for perovskite/silicon heterojunction tandem solar cells with proof-of-concept efficiency > 38%

Xuejiao Wang^{1, 2, 3, 4, 5, §}, Guanlan Chen^{1, 2, 3, 4, 5, §}, Ying Liu^{1, 2, 3, 4, 5}, Guangyi Wang^{1, 2, 3, 4, 5}, Wei Han^{1, 2, 3, 4, 5}, Jin Wang^{1, 2, 3, 4, 5}, Pengfei Liu^{1, 2, 3, 4, 5}, Jilei Wang⁶, Shaojuan Bao⁶, Bo Yu⁷, Ying Liu⁷, Xinliang Chen^{1, 2, 3, 4, 5}, Shengzhi Xu^{1, 2, 3, 4, 5}, Ying Zhao^{1, 2, 3, 4, 5,} and Xiaodan Zhang^{1, 2, 3, 4, 5,}

+ Author Affiliations

1. Institute of Photoelectronic Thin Film Devices and Technology, Nankai University, Tianjin 300350, P. R. China
2. Tianjin Key Laboratory of Efficient Utilization of Solar Energy, Tianjin 300350, P. R. China
3. State Key Laboratory of Photovoltaic Materials and Cells, Tianjin 300192, P. R. China
4. Engineering Research Center of Thin Film Photoelectronic Technology of Ministry of Education, Tianjin 300350, P. R. China
5. Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin 300072, P. R. China
6. Anhui Soltrend New Energy Technology Co., Ltd, Anhui 230000, P. R. China
7. Yingli Energy Development Co., Ltd, Baoding 071002, P. R. China
^§Xuejiao Wang and Guanlan Chen contributed equally to this work and should be considered as co-first authors.

Author Bio:

Xuejiao Wang received the B.Sci. degree in Optoelectronic Information Science and Engineering from Hebei University, Baoding, China, in 2018. Then she received the M. Sc. Degree in Optical Engineering from Hebei University, Baoding, China, in 2021. She is currently pursuing the D.Eng. degree in Nankai University, Tianjin, China. Her research interests include the silicon heterojunction solar cells and perovskite/silicon tandem solar cells.

Guanlan Chen received the B.Eng. degree in Communication Engineering from Tongji University, Shanghai, China, in 2020. He is currently pursuing the D.Eng. degree with the College of Electronic Information and Optical Engineering, Nankai University, Tianjin, China. His research interests include the silicon film and transparent conductive oxide(TCO) in silicon heterojunction solar cells and perovskite/silicon tandem solar cells.

Ying Zhao, a professor at Nankai University, is the director of the State Key Laboratory of Photovoltaic Materials and Cells, the director of China Photovoltaic Committee of the Chinese Renewable Energy Society and vice Chairman of the International Advisory Committee of the International Photovoltaic Science and Engineering Conference (PVSEC).

Xiaodan Zhang, a professor at Nankai University, is the director of the Institute of Optoelectronic Thin Film Devices and Technology, the deputy director of the State Key Laboratory of Photovoltaic Materials and Cells. Her current research interests include perovskite/crystalline silicon tandem solar cells, perovskite solar cells, novel silicon solar cells utilizing oxide-based carrier transport layers, photo/electrocatalytic materials, solar cell-driven photoelectrochemical water splitting and CO₂ reduction.

Corresponding author: Ying Zhao, zhaoygds@nankai.edu.cn; Xiaodan Zhang, xdzhang@nankai.edu.cn

DOI: 10.1088/1674-4926/25050011CSTR: 32376.14.1674-4926.25050011

Abstract: As the development of single-junction solar cells reaches a bottleneck, tandem solar cells have emerged as a critical pathway to further enhance power conversion efficiency. Among them, monolithic perovskite/silicon heterojunction tandem solar cells are currently the fastest-growing technology, achieving the highest efficiencies at relatively low costs. The interconnecting layer, which connects the two sub-cells, plays a crucial role in tandem cell performance. It collects electrons and holes from the respective sub-cells and facilitates recombination and tunneling at the interface. Therefore, the properties of the interconnecting layer are pivotal to the overall device performance. In this work, we applied statistical analysis and machine learning algorithms to systematically analyze the interconnecting layer. A comprehensive dataset on interconnecting layer parameters was established, and predictive modeling was performed using Lasso linear regression, random forest, and multilayer perceptron (a type of neural network). The analysis revealed key feature importance for experimental parameters, providing valuable insights into the application of interconnecting layers in perovskite/silicon heterojunction tandem solar cells. The final optimized interconnecting layer can achieve a proof-of-concept efficiency of 38.17%, providing guidance and direction for the development of monolithic perovskite/silicon tandem solar cells.

Key words: perovskite/silicon heterojunction tandem solar cells, interconnecting layer, machine learning

References

[1]	Wu H, Ye F, Yang M, et al. Silicon heterojunction back-contact solar cells by laser patterning. Nature, 2024, 635(8039), 604 doi: 10.1038/s41586-024-08110-8
[2]	Richter A, Hermle M, Glunz S W. Reassessment of the limiting efficiency for crystalline silicon solar cells. IEEE J Photovolt, 2013, 3(4), 1184 doi: 10.1109/JPHOTOV.2013.2270351
[3]	Kerr M J, Cuevas A, Campbell P. Limiting efficiency of crystalline silicon solar cells due to Coulomb-enhanced Auger recombination. Progress Photovoltaics, 2003, 11(2), 97 doi: 10.1002/pip.464
[4]	Albrecht S, Saliba M, Correa Baena J P, et al. Monolithic perovskite/silicon-heterojunction tandem solar cells processed at low temperature. Energy Environ Sci, 2016, 9(1), 81 doi: 10.1039/C5EE02965A
[5]	Werner J, Niesen B, Ballif C. Perovskite/silicon tandem solar cells: Marriage of convenience or true love story? –an overview. Adv Materials Inter, 2018, 5(1), 1700731 doi: 10.1002/admi.201700731
[6]	Sahli F, Werner J, Kamino B A, et al. Fully textured monolithic perovskite/silicon tandem solar cells with 25.2% power conversion efficiency. Nat Mater, 2018, 17(9), 820 doi: 10.1038/s41563-018-0115-4
[7]	Yan L L, Li Y X, Shi B, et al. Reducing electrical losses of textured monolithic perovskite/silicon tandem solar cells by tailoring nanocrystalline silicon tunneling recombination junction. Sol Energy Mater Sol Cells, 2022, 245, 111868 doi: 10.1016/j.solmat.2022.111868
[8]	Turkay D, Artuk K, Chin X Y, et al. Synergetic substrate and additive engineering for over 30\%-efficient perovskite-Si tandem solar cells. Joule, 2024, 8(6), 1735 doi: 10.1016/j.joule.2024.04.015
[9]	Mariotti S, Köhnen E, Scheler F, et al. Interface engineering for high-performance, triple-halide perovskite-silicon tandem solar cells. Science, 2023, 381(6653), 63 doi: 10.1126/science.adf5872
[10]	Al-Ashouri A, Köhnen E, Li B, et al. Monolithic perovskite/silicon tandem solar cell with >29% efficiency by enhanced hole extraction. Science, 2020, 370(6522), 1300 doi: 10.1126/science.abd4016
[11]	Aydin E, Ugur E, Yildirim B K, et al. Enhanced optoelectronic coupling for perovskite/silicon tandem solar cells. Nature, 2023, 623(7988), 732 doi: 10.1038/s41586-023-06667-4
[12]	Liu J, He Y C, Ding L, et al. Perovskite/silicon tandem solar cells with bilayer interface passivation. Nature, 2024, 635(8039), 596 doi: 10.1038/s41586-024-07997-7
[13]	Werner J, Walter A, Rucavado E, et al. Zinc tin oxide as high-temperature stable recombination layer for mesoscopic perovskite/silicon monolithic tandem solar cells. Appl Phys Lett, 2016, 109(23), 233902 doi: 10.1063/1.4971361
[14]	Chin X Y, Turkay D, Steele J A, et al. Interface passivation for 31.25%-efficient perovskite/silicon tandem solar cells. Science, 2023, 381(6653), 59 doi: 10.1126/science.adg0091
[15]	Golobostanfard M R, Othman M, Turkay D, et al. Bifacial perovskite/silicon heterojunction tandem solar cells based on FAPbI₃-based perovskite via hybrid evaporation-spin coating. Nano Energy, 2024, 131, 110269 doi: 10.1016/j.nanoen.2024.110269
[16]	Liu Y M, Tan X Y, Liang J, et al. Machine learning for perovskite solar cells and component materials: Key technologies and prospects. Adv Funct Materials, 2023, 33(17), 2214271 doi: 10.1002/adfm.202214271
[17]	Saqib M, Shoukat U, Soliman M M, et al. Machine learning assisted designing of hole-transporting materials for high performance perovskite solar cells. Chem Phys, 2025, 589, 112515 doi: 10.1016/j.chemphys.2024.112515
[18]	Zhang L, He M, Shao S F. Machine learning for halide perovskite materials. Nano Energy, 2020, 78, 105380 doi: 10.1016/j.nanoen.2020.105380
[19]	Jiang K L, Ma Z P, Lan J, et al. Screening of functional small molecules via modified machine learning strategy toward efficient all-inorganic perovskite solar cells. ACS Appl Mater Interfaces, 2024, 16(24), 31833 doi: 10.1021/acsami.4c06226
[20]	Gok E C, Yildirim M O, Haris M P U, et al. Predicting perovskite bandgap and solar cell performance with machine learning. Sol RRL, 2022, 6(2), 2100927 doi: 10.1002/solr.202100927
[21]	Li J X, Pradhan B, Gaur S, et al. Predictions and strategies learned from machine learning to develop high‐performing perovskite solar cells. Adv Energy Mater, 2019, 9(46), 1901891 doi: 10.1002/aenm.201901891
[22]	Karade V C, Sutar S S, Shin S W, et al. Machine learning assisted analysis, prediction, and fabrication of high-efficiency CZTSSe thin film solar cells. Adv Funct. Materials, 2023, 33(41), 2303459
[23]	Li X L, Hou Z F, Gao S S, et al. Efficient optimization of the performance of Mn²⁺-doped kesterite solar cell: Machine learning aided synthesis of high efficient Cu₂(Mn, Zn)Sn(S, Se)₄ solar cells (solar RRL 12∕2018). Sol RRL, 2018, 2(12), 1870237 doi: 10.1002/solr.201870237
[24]	Zhu C W, Liu W, Li Y Y, et al. Key factors governing the device performance of CIGS solar cells: Insights from machine learning. Sol Energy, 2021, 228, 45 doi: 10.1016/j.solener.2021.09.031
[25]	Bandaru N, Enduri M K, Venkata Reddy C, et al. Aspects of effectiveness and significance: The use of machine learning methods to study CuIn_1-xGa_xSe₂ solar cells. Sol Energy, 2023, 263, 111941 doi: 10.1016/j.solener.2023.111941
[26]	Zhou J K, Jesper Jacobsson T, Wang Z, et al. Data-driven tunnel oxide passivated contact solar cell performance analysis using machine learning. Adv Mater, 2024, 36(14), e2309351 doi: 10.1002/adma.202309351
[27]	Liu Y M, Ahmadpour M, Adam J, et al. Modeling multijunction solar cells by nonlocal tunneling and subcell analysis. IEEE J Photovolt, 2018, 8(5), 1363 doi: 10.1109/JPHOTOV.2018.2851308
[28]	Shrivastav N, Madan J, Pandey R, et al. Investigations aimed at producing 33% efficient perovskite-silicon tandem solar cells through device simulations. RSC Adv, 2021, 11(59), 37366 doi: 10.1039/D1RA06250F
[29]	Mohamad I S, Doroody C, Alkharasani W M, et al. Elucidating the effects of interconnecting layer thickness and bandgap variations on the performance of monolithic perovskite/silicon tandem solar cell by wxAMPS. Materials, 2023, 16(11), 4106 doi: 10.3390/ma16114106
[30]	He Y Z, Chen H Z , Wang S Q, et al. Unveiling recombination in top cells: SCAPS-1D simulations for high-efficiency bifacial planar perovskite/silicon tandem solar cells. Sol Energy, 2024, 282, 112921 doi: 10.1016/j.solener.2024.112921
[31]	Liu W Z, Liu Y J, Yang Z Q, et al. Flexible solar cells based on foldable silicon wafers with blunted edges. Nature, 2023, 617(7962), 717 doi: 10.1038/s41586-023-05921-z
[32]	Wu T J, Liu Z L, Lin H, et al. Free-standing ultrathin silicon wafers and solar cells through edges reinforcement. Nat Commun, 2024, 15(1), 3843 doi: 10.1038/s41467-024-48290-5
[33]	Li Y, Ru X N, Yang M, et al. Flexible silicon solar cells with high power-to-weight ratios. Nature, 2024, 626(7997), 105 doi: 10.1038/s41586-023-06948-y
[34]	Allen T G, Ugur E, Aydin E, et al. A practical efficiency target for perovskite/silicon tandem solar cells. ACS Energy Lett, 2025, 10(1), 238 doi: 10.1021/acsenergylett.4c02152
[35]	Green M A, Dunlop E D, Yoshita M, et al. Solar cell efficiency tables (version 65). Prog Photovolt: Res Appl, 2025, 33(1), 3 doi: 10.1002/pip.3867

Fig. 1. (Color online) Commonly used ICLs: (a) nanocrystalline silicon thin film prepared by PECVD as ICL^[6] (Reproduced with permission. Copyright 2018, Sahli et al, Springer Nature), (b) ITO as ICL^[10] (Reproduced with permission. Copyright 2020, Albrecht et al, American Association for the Advancement of Science), (c) IZO as ICL^[11] (Reproduced with permission. Copyright 2023, De Wolf et al, Springer Nature), (d) silicon thin film and ITO as ICL together^[14] (Reproduced with permission. Copyright 2023, Ballif et al, American Association for the Advancement of Science).

DownLoad: Full-Size Img PowerPoint

Fig. 2. (Color online) Workflow.

DownLoad: Full-Size Img PowerPoint

Fig. 3. (Color online) Heatmaps of the Pearson correlation coefficient analysis between the four output features in (a) the dataset of Si-ICL, (b) the dataset of TCO-ICL, and (c) the dataset of Si+TCO-ICL.

DownLoad: Full-Size Img PowerPoint

Fig. 4. (Color online) Heatmaps of the Pearson correlation coefficient analysis between the numerical variables in the input features and the output in (a) the dataset of Si-ICL, (b) the dataset of TCO-ICL, and (c) the dataset of Si+TCO-ICL.

DownLoad: Full-Size Img PowerPoint

Fig. 5. (Color online) Tandem solar cell performance predicted by different models based on the dataset of Si-ICL: (a–c) V_oc, (d–f) FF, (g–i) J_sc, (j–l) PCE. Models used include Lasso, RF and NN. The dashed line represents a perfect fitting, while the solid red line represents the actual fitting result.

DownLoad: Full-Size Img PowerPoint

Fig. 6. (Color online) Feature importance for the dataset of Si-ICL based on the RF model. (a) V_oc, (b) FF, (c) J_sc, (d) PCE.

DownLoad: Full-Size Img PowerPoint

Fig. 7. (Color online) Optimal simulation results for perovskite/SHJ tandem solar cells.

DownLoad: Full-Size Img PowerPoint

Table 1. wxAMPS simulation parameters for each layer^[28−30].

Material properties	Top cell					Interconnecting layer			Bottom cell
Material properties	IZO	SnO₂	C60	PSC	Me-4PACz	ITO	n-nc-SiO_x:H	p-nc-SiO_x:H	n-nc-SiO_x:H	i-a-Si:H	n-Si	i-a-Si:H	p-nc-SiO_x:H
Dielectric constant	10	3.3	4.2	10	10	11.9	11.9	11.9	11.9	11.9	11.9	11.9	11.9
Bandgap (eV)	3.8	3.5	1.7	1.68	3.3	3.8	1.8–2.6	1.8–2.6	2.3	1.74	1.12	1.74	2.2
Electron affinity (eV)	4.3	4.1	3.9	3.9	2.8	4.3	3.8	3.8	3.8	3.9	4.05	3.9	3.8
CB effective density (/cm³)	1e¹⁸	2.2e¹⁸	8e¹⁹	1e¹⁹	1e¹⁹	1e¹⁸	1e²⁰	1e²⁰	1e²⁰	1e²⁰	2.8e¹⁹	1e²⁰	1e²⁰
VB effective density (/cm³)	1e¹⁹	1.8e¹⁹	8e¹⁹	1e¹⁹	1e¹⁹	1e¹⁹	1e²⁰	1e²⁰	1e²⁰	1e²⁰	1.04e¹⁹	1e²⁰	1e²⁰
Electron mobility (cm²/Vs)	60	50	0.08	20	6	50	20	20	20	20	1400	20	20
Hole mobility (cm²/Vs)	10	30	0.0035	2	24	25	5	5	5	5	450	5	5
N_D (/cm³)	5e²⁰	1e¹⁷	1e¹⁷	1e¹⁷	/	5e²⁰	1e²¹	/	1e¹⁹	/	5e¹⁶	/	/
N_A (/cm³)	/	/	/	/	1e¹⁸	/	/	1e²¹	/	/	/	/	1e¹⁹

DownLoad: CSV

Table 2. Details of the extracted features.

Features	Features type	Feature space
ST–Silicon wafer thickness	Numerical variables	150–280 μm
NT–n-type silicon thin film thickness	Numerical variables	0–50 nm
PT–p-type silicon thin film thickness	Numerical variables	0–50 nm
NRH–Hydrogen dilution ratio of n-type silicon thin film	Numerical variables	70–120
PRH–Hydrogen dilution ratio of p-type silicon thin film	Numerical variables	180–240
ND–Doping ratio of n-type silicon thin film	Numerical variables	0.3–2.2
PD–Doping ratio of p-type silicon thin film	Numerical variables	0.5–2.5
NO–Oxygen doping ratio of n-type silicon thin film	Numerical variables	0–1.4
PO–Oxygen doping ratio of p-type silicon thin film	Numerical variables	0–1
TS–Active area of tandem solar cell	Numerical variables	1–65.1 cm²
SM–Surface morphology of silicon wafer	Categorical variables	2 (texture, plane)
PSC–The type of perovskite solar cell structure	Categorical variables	2 (n-i-p, p-i-n)
PSCH–The perovskite solar cell Hole Transport Layers	Categorical variables	12
TMO–The type of TCO	Categorical variables	7
OS–Doping ratio of ITO targets	Numerical variables	0.03–15.6
WF–Work function of TCO	Numerical variables	4.3–5.27
OT–The thickness of TCO ICL	Numerical variables	1–110
OSR–Sheet resistance of the TCO	Numerical variables	15–∞
OME–Electron mobility of the TCO	Numerical variables	2–100 cm²/(V•s)
OMH–Hole mobility of the TCO	Numerical variables	5–25 cm²/(V•s)
V_oc	Numerical variables	1.37–2.09 V
J_sc	Numerical variables	12–20.5 mA/cm²
FF	Numerical variables	23–86.9%
PCE	Numerical variables	5–35.4%
Note: For categorical variables, the feature space represents the number of feature categories, while for numerical variables it represents the range of variation.

DownLoad: CSV

Table 3. One-way ANOVA results of dataset of Si-ICL.

Categorical variables	Numerical Variables	F Value	P Value
PSCH	V_oc	3530.267259	5.483809e^–159
PSCH	J_sc	15.73319	0.0000004421083
PSCH	FF	694.53997	8.995908e^–92
PSCH	PCE	1264.88677	1.077701e^–115

DownLoad: CSV

Table 4. One-way ANOVA results of dataset of TCO-ICL.

Categorical variables	Numerical Variables	F Value	P Value
TMO	V_oc	4.851974	9.88e^–3
TMO	J_sc	3.233722	4.38e^–2
TMO	FF	4.872835	9.69e^–3
TMO	PCE	1.297649	2.78e^–1
PSCH	V_oc	237.973744	5.86e^–53
PSCH	J_sc	3.321445	5.32e^–3
PSCH	FF	58.700255	5.36e^–29
PSCH	PCE	33.91421	4.20e^–21

DownLoad: CSV

Table 5. The specific parameters of the ICL in optimal device.

Materials properties	Si thin film	TCO
Bandgap (eV)	2.5	3.8
Thickness (nm)	20	3
Refractive index (@ 950 nm)	2.2	1.5
Note: the refractive index (@ 950 nm) of n-type doping layer of SHJ solar cells is 2.95, which forms a graded refractive index matching structure with ICL.

DownLoad: CSV

Table 6. Summary of the highest PCE including both experimental and simulation achieved for perovskite/SHJ tandem solar cells.

V_oc (V)	FF (%)	J_sc (mA/cm²)	PCE (%)	References
2.120	90.30	22.17	42.45	Theoretical PCE limit^[34]
2.077	87.69	20.96	38.17	Practical PCE target in this work
2.061	87.50	20.94	37.76	Practical PCE target^[34]
1.996	83.60	20.71	34.60	LONGi tandem cell^[35]
1.966	83.00	20.76	33.89	LONGi tandem cell^[12]

DownLoad: CSV

[1]	Wu H, Ye F, Yang M, et al. Silicon heterojunction back-contact solar cells by laser patterning. Nature, 2024, 635(8039), 604 doi: 10.1038/s41586-024-08110-8
[2]	Richter A, Hermle M, Glunz S W. Reassessment of the limiting efficiency for crystalline silicon solar cells. IEEE J Photovolt, 2013, 3(4), 1184 doi: 10.1109/JPHOTOV.2013.2270351
[3]	Kerr M J, Cuevas A, Campbell P. Limiting efficiency of crystalline silicon solar cells due to Coulomb-enhanced Auger recombination. Progress Photovoltaics, 2003, 11(2), 97 doi: 10.1002/pip.464
[4]	Albrecht S, Saliba M, Correa Baena J P, et al. Monolithic perovskite/silicon-heterojunction tandem solar cells processed at low temperature. Energy Environ Sci, 2016, 9(1), 81 doi: 10.1039/C5EE02965A
[5]	Werner J, Niesen B, Ballif C. Perovskite/silicon tandem solar cells: Marriage of convenience or true love story? –an overview. Adv Materials Inter, 2018, 5(1), 1700731 doi: 10.1002/admi.201700731
[6]	Sahli F, Werner J, Kamino B A, et al. Fully textured monolithic perovskite/silicon tandem solar cells with 25.2% power conversion efficiency. Nat Mater, 2018, 17(9), 820 doi: 10.1038/s41563-018-0115-4
[7]	Yan L L, Li Y X, Shi B, et al. Reducing electrical losses of textured monolithic perovskite/silicon tandem solar cells by tailoring nanocrystalline silicon tunneling recombination junction. Sol Energy Mater Sol Cells, 2022, 245, 111868 doi: 10.1016/j.solmat.2022.111868
[8]	Turkay D, Artuk K, Chin X Y, et al. Synergetic substrate and additive engineering for over 30\%-efficient perovskite-Si tandem solar cells. Joule, 2024, 8(6), 1735 doi: 10.1016/j.joule.2024.04.015
[9]	Mariotti S, Köhnen E, Scheler F, et al. Interface engineering for high-performance, triple-halide perovskite-silicon tandem solar cells. Science, 2023, 381(6653), 63 doi: 10.1126/science.adf5872
[10]	Al-Ashouri A, Köhnen E, Li B, et al. Monolithic perovskite/silicon tandem solar cell with >29% efficiency by enhanced hole extraction. Science, 2020, 370(6522), 1300 doi: 10.1126/science.abd4016
[11]	Aydin E, Ugur E, Yildirim B K, et al. Enhanced optoelectronic coupling for perovskite/silicon tandem solar cells. Nature, 2023, 623(7988), 732 doi: 10.1038/s41586-023-06667-4
[12]	Liu J, He Y C, Ding L, et al. Perovskite/silicon tandem solar cells with bilayer interface passivation. Nature, 2024, 635(8039), 596 doi: 10.1038/s41586-024-07997-7
[13]	Werner J, Walter A, Rucavado E, et al. Zinc tin oxide as high-temperature stable recombination layer for mesoscopic perovskite/silicon monolithic tandem solar cells. Appl Phys Lett, 2016, 109(23), 233902 doi: 10.1063/1.4971361
[14]	Chin X Y, Turkay D, Steele J A, et al. Interface passivation for 31.25%-efficient perovskite/silicon tandem solar cells. Science, 2023, 381(6653), 59 doi: 10.1126/science.adg0091
[15]	Golobostanfard M R, Othman M, Turkay D, et al. Bifacial perovskite/silicon heterojunction tandem solar cells based on FAPbI₃-based perovskite via hybrid evaporation-spin coating. Nano Energy, 2024, 131, 110269 doi: 10.1016/j.nanoen.2024.110269
[16]	Liu Y M, Tan X Y, Liang J, et al. Machine learning for perovskite solar cells and component materials: Key technologies and prospects. Adv Funct Materials, 2023, 33(17), 2214271 doi: 10.1002/adfm.202214271
[17]	Saqib M, Shoukat U, Soliman M M, et al. Machine learning assisted designing of hole-transporting materials for high performance perovskite solar cells. Chem Phys, 2025, 589, 112515 doi: 10.1016/j.chemphys.2024.112515
[18]	Zhang L, He M, Shao S F. Machine learning for halide perovskite materials. Nano Energy, 2020, 78, 105380 doi: 10.1016/j.nanoen.2020.105380
[19]	Jiang K L, Ma Z P, Lan J, et al. Screening of functional small molecules via modified machine learning strategy toward efficient all-inorganic perovskite solar cells. ACS Appl Mater Interfaces, 2024, 16(24), 31833 doi: 10.1021/acsami.4c06226
[20]	Gok E C, Yildirim M O, Haris M P U, et al. Predicting perovskite bandgap and solar cell performance with machine learning. Sol RRL, 2022, 6(2), 2100927 doi: 10.1002/solr.202100927
[21]	Li J X, Pradhan B, Gaur S, et al. Predictions and strategies learned from machine learning to develop high‐performing perovskite solar cells. Adv Energy Mater, 2019, 9(46), 1901891 doi: 10.1002/aenm.201901891
[22]	Karade V C, Sutar S S, Shin S W, et al. Machine learning assisted analysis, prediction, and fabrication of high-efficiency CZTSSe thin film solar cells. Adv Funct. Materials, 2023, 33(41), 2303459
[23]	Li X L, Hou Z F, Gao S S, et al. Efficient optimization of the performance of Mn²⁺-doped kesterite solar cell: Machine learning aided synthesis of high efficient Cu₂(Mn, Zn)Sn(S, Se)₄ solar cells (solar RRL 12∕2018). Sol RRL, 2018, 2(12), 1870237 doi: 10.1002/solr.201870237
[24]	Zhu C W, Liu W, Li Y Y, et al. Key factors governing the device performance of CIGS solar cells: Insights from machine learning. Sol Energy, 2021, 228, 45 doi: 10.1016/j.solener.2021.09.031
[25]	Bandaru N, Enduri M K, Venkata Reddy C, et al. Aspects of effectiveness and significance: The use of machine learning methods to study CuIn_1-xGa_xSe₂ solar cells. Sol Energy, 2023, 263, 111941 doi: 10.1016/j.solener.2023.111941
[26]	Zhou J K, Jesper Jacobsson T, Wang Z, et al. Data-driven tunnel oxide passivated contact solar cell performance analysis using machine learning. Adv Mater, 2024, 36(14), e2309351 doi: 10.1002/adma.202309351
[27]	Liu Y M, Ahmadpour M, Adam J, et al. Modeling multijunction solar cells by nonlocal tunneling and subcell analysis. IEEE J Photovolt, 2018, 8(5), 1363 doi: 10.1109/JPHOTOV.2018.2851308
[28]	Shrivastav N, Madan J, Pandey R, et al. Investigations aimed at producing 33% efficient perovskite-silicon tandem solar cells through device simulations. RSC Adv, 2021, 11(59), 37366 doi: 10.1039/D1RA06250F
[29]	Mohamad I S, Doroody C, Alkharasani W M, et al. Elucidating the effects of interconnecting layer thickness and bandgap variations on the performance of monolithic perovskite/silicon tandem solar cell by wxAMPS. Materials, 2023, 16(11), 4106 doi: 10.3390/ma16114106
[30]	He Y Z, Chen H Z , Wang S Q, et al. Unveiling recombination in top cells: SCAPS-1D simulations for high-efficiency bifacial planar perovskite/silicon tandem solar cells. Sol Energy, 2024, 282, 112921 doi: 10.1016/j.solener.2024.112921
[31]	Liu W Z, Liu Y J, Yang Z Q, et al. Flexible solar cells based on foldable silicon wafers with blunted edges. Nature, 2023, 617(7962), 717 doi: 10.1038/s41586-023-05921-z
[32]	Wu T J, Liu Z L, Lin H, et al. Free-standing ultrathin silicon wafers and solar cells through edges reinforcement. Nat Commun, 2024, 15(1), 3843 doi: 10.1038/s41467-024-48290-5
[33]	Li Y, Ru X N, Yang M, et al. Flexible silicon solar cells with high power-to-weight ratios. Nature, 2024, 626(7997), 105 doi: 10.1038/s41586-023-06948-y
[34]	Allen T G, Ugur E, Aydin E, et al. A practical efficiency target for perovskite/silicon tandem solar cells. ACS Energy Lett, 2025, 10(1), 238 doi: 10.1021/acsenergylett.4c02152
[35]	Green M A, Dunlop E D, Yoshita M, et al. Solar cell efficiency tables (version 65). Prog Photovolt: Res Appl, 2025, 33(1), 3 doi: 10.1002/pip.3867

Search

GET CITATION

shu

Export: BibTex EndNote

Article Metrics

Article views: 111 Times PDF downloads: 27 Times Cited by: 0 Times

History

Received: 08 May 2025 Revised: Online: Accepted Manuscript: 15 May 2025

Article Navigation > Journal of Semiconductors > 2025 > Accepted Manuscript

Xuejiao Wang, Guanlan Chen, Ying Liu, Guangyi Wang, Wei Han, Jin Wang, Pengfei Liu, Jilei Wang, Shaojuan Bao, Bo Yu, Ying Liu, Xinliang Chen, Shengzhi Xu, Ying Zhao, Xiaodan Zhang. Machine learning facilitates the development of interconnecting layers for perovskite/silicon heterojunction tandem solar cells with proof-of-concept efficiency > 38%[J]. Journal of Semiconductors, 2025, In Press. doi: 10.1088/1674-4926/25050011 ****X J Wang, G L Chen, Y Liu, G Y Wang, W Han, J Wang, P F Liu, J L Wang, S J Bao, B Yu, Y Liu, X L Chen, S Z Xu, Y Zhao, and X D Zhang, Machine learning facilitates the development of interconnecting layers for perovskite/silicon heterojunction tandem solar cells with proof-of-concept efficiency > 38%[J]. J. Semicond., 2025, accepted doi: 10.1088/1674-4926/25050011

Citation:

X J Wang, G L Chen, Y Liu, G Y Wang, W Han, J Wang, P F Liu, J L Wang, S J Bao, B Yu, Y Liu, X L Chen, S Z Xu, Y Zhao, and X D Zhang, Machine learning facilitates the development of interconnecting layers for perovskite/silicon heterojunction tandem solar cells with proof-of-concept efficiency > 38%[J]. J. Semicond., 2025, accepted doi: 10.1088/1674-4926/25050011

Citation:

PDF( 7826 KB)

Machine learning facilitates the development of interconnecting layers for perovskite/silicon heterojunction tandem solar cells with proof-of-concept efficiency > 38%

DOI: 10.1088/1674-4926/25050011

CSTR: 32376.14.1674-4926.25050011

Xuejiao Wang^{1, 2, 3, 4, 5, §},
Guanlan Chen^{1, 2, 3, 4, 5, §},
Ying Liu^{1, 2, 3, 4, 5},
Guangyi Wang^{1, 2, 3, 4, 5},
Wei Han^{1, 2, 3, 4, 5},
Jin Wang^{1, 2, 3, 4, 5},
Pengfei Liu^{1, 2, 3, 4, 5},
Jilei Wang⁶,
Shaojuan Bao⁶,
Bo Yu⁷,
Ying Liu⁷,
Xinliang Chen^{1, 2, 3, 4, 5},
Shengzhi Xu^{1, 2, 3, 4, 5},
Ying Zhao^{1, 2, 3, 4, 5,},
Xiaodan Zhang^{1, 2, 3, 4, 5,}

1.
Institute of Photoelectronic Thin Film Devices and Technology, Nankai University, Tianjin 300350, P. R. China
2.
Tianjin Key Laboratory of Efficient Utilization of Solar Energy, Tianjin 300350, P. R. China
3.
State Key Laboratory of Photovoltaic Materials and Cells, Tianjin 300192, P. R. China
4.
Engineering Research Center of Thin Film Photoelectronic Technology of Ministry of Education, Tianjin 300350, P. R. China
5.
Collaborative Innovation Center of Chemical Science and Engineering (Tianjin), Tianjin 300072, P. R. China
6.
Anhui Soltrend New Energy Technology Co., Ltd, Anhui 230000, P. R. China
7.
Yingli Energy Development Co., Ltd, Baoding 071002, P. R. China

More Information

Xuejiao Wang received the B.Sci. degree in Optoelectronic Information Science and Engineering from Hebei University, Baoding, China, in 2018. Then she received the M. Sc. Degree in Optical Engineering from Hebei University, Baoding, China, in 2021. She is currently pursuing the D.Eng. degree in Nankai University, Tianjin, China. Her research interests include the silicon heterojunction solar cells and perovskite/silicon tandem solar cells
Guanlan Chen received the B.Eng. degree in Communication Engineering from Tongji University, Shanghai, China, in 2020. He is currently pursuing the D.Eng. degree with the College of Electronic Information and Optical Engineering, Nankai University, Tianjin, China. His research interests include the silicon film and transparent conductive oxide(TCO) in silicon heterojunction solar cells and perovskite/silicon tandem solar cells
Ying Zhao, a professor at Nankai University, is the director of the State Key Laboratory of Photovoltaic Materials and Cells, the director of China Photovoltaic Committee of the Chinese Renewable Energy Society and vice Chairman of the International Advisory Committee of the International Photovoltaic Science and Engineering Conference (PVSEC)
Xiaodan Zhang, a professor at Nankai University, is the director of the Institute of Optoelectronic Thin Film Devices and Technology, the deputy director of the State Key Laboratory of Photovoltaic Materials and Cells. Her current research interests include perovskite/crystalline silicon tandem solar cells, perovskite solar cells, novel silicon solar cells utilizing oxide-based carrier transport layers, photo/electrocatalytic materials, solar cell-driven photoelectrochemical water splitting and CO₂ reduction
Corresponding author: zhaoygds@nankai.edu.cn; xdzhang@nankai.edu.cn
Received Date: 2025-05-08
Available Online: 2025-05-15

Abstract

Abstract

As the development of single-junction solar cells reaches a bottleneck, tandem solar cells have emerged as a critical pathway to further enhance power conversion efficiency. Among them, monolithic perovskite/silicon heterojunction tandem solar cells are currently the fastest-growing technology, achieving the highest efficiencies at relatively low costs. The interconnecting layer, which connects the two sub-cells, plays a crucial role in tandem cell performance. It collects electrons and holes from the respective sub-cells and facilitates recombination and tunneling at the interface. Therefore, the properties of the interconnecting layer are pivotal to the overall device performance. In this work, we applied statistical analysis and machine learning algorithms to systematically analyze the interconnecting layer. A comprehensive dataset on interconnecting layer parameters was established, and predictive modeling was performed using Lasso linear regression, random forest, and multilayer perceptron (a type of neural network). The analysis revealed key feature importance for experimental parameters, providing valuable insights into the application of interconnecting layers in perovskite/silicon heterojunction tandem solar cells. The final optimized interconnecting layer can achieve a proof-of-concept efficiency of 38.17%, providing guidance and direction for the development of monolithic perovskite/silicon tandem solar cells.
- perovskite/silicon heterojunction tandem solar cells,
- interconnecting layer,
- machine learning

Supplementary materials to this article can be found online at https://doi.org/10.1088/1674-4926/24050011.
^§Xuejiao Wang and Guanlan Chen contributed equally to this work and should be considered as co-first authors.

FullText(HTML)

References(35)