Passivating defects and optimizing interfaces to boost the comprehensive performance of carbon-cathode hole-transport-layer-free CsPbI₂Br solar cells via an ionic liquid†

Wenxuan Li‡ , Xueyan Ma‡, Hai Liu, Hongyan Cheng, Zhengjun Meng, Xiaoyang Liu, Guodong Wan, Zhe Gao, Yali Li*, Yujun Fu, Deyan He and Junshuai Li*
LONGi Institute of Future Technology, School of Materials & Energy, Lanzhou University, 222 South Tianshui Road, Lanzhou, Gansu 730000, China. E-mail: liyli@lzu.edu.cn; jshli@lzu.edu.cn

First published on 15th July 2025

---

---

1. Introduction

The power conversion efficiency (PCE) of hybrid perovskite solar cells (PSCs) has now reached parity with that of silicon solar cells.^1–3 However, the industrialization of hybrid PSCs is critically dependent on material stability, with core challenges primarily focused on structural and photothermal stability.⁴ While moisture-induced structure degradation can be effectively mitigated through encapsulation techniques, the inherent tendency of hybrid perovskite structures to undergo phase transition and operational photothermal instability remain significant bottlenecks.^5–8 All-inorganic CsPbX₃ perovskites effectively eliminate volatile organic components (e.g., MA⁺, FA⁺), leading to enhanced photothermal stability. Among these materials, the mixed-halide CsPbI₂Br offers great advantages for commercial application owing to its favorable Goldschmidt tolerance factor (∼0.85), stable photoactive phase at room temperature, appropriate bandgap (1.91 eV), and exceptional environmental resilience.^9–12

Carbon-cathode hole-transport-layer-free (HTL-free) CsPbI₂Br PSCs effectively address the issues of metal electrode corrosion of the HTL and excessive costs by simultaneously replacing the traditional noble metal electrodes and organic HTLs with a carbon electrode.^13–17 However, the PCE of this kind of device (theoretically approaching 20%)¹⁸ is limited by poor crystalline quality, high defect density, and suboptimal interfacial contact of CsPbI₂Br.^19–22 Although traditional solid-state additives enhance crystallization, their limited dispersibility and slow diffusion in the precursor solutions lead to inefficient defect passivation.^23–28 In contrast, ionic liquid (IL) additives exhibit high solubility and dynamic diffusion in the precursor solutions due to their dual ion characteristics and tunable intermolecular forces, providing a novel concept for crystallization regulation and defect passivation in CsPbI₂Br.²⁹

This study systematically elucidates the synergistic regulatory mechanism of 1-butyl-3-methylimidazolium trifluoromethanesulfonate ([BMIM][OTF]), a bifunctional IL additive, on the growth of CsPbI₂Br and carrier transport, based on carbon-cathode HTL-free CsPbI₂Br PSCs. Theoretical and experimental characterizations reveal that the BMIM⁺ cations spontaneously accumulate at both the top and buried interfaces of CsPbI₂Br through the steric hindrance effect. The N within an imidazole ring serves as a negative charge center, forming a coordination bond with uncoordinated Pb²⁺ in CsPbI₂Br or Sn⁴⁺ in SnO₂, significantly reducing the interface defect density. Additionally, the overall positive charge characteristic allows BMIM⁺ to adsorb on the SnO₂ surface via an electrostatic interaction, inhibiting the formation of vacancies and regulating the crystallization of CsPbI₂Br. Meanwhile, a hydrophobic barrier is constructed by the alkyl chain of BMIM⁺ on the CsPbI₂Br surface, enhancing its moisture resistance. The trifluoromethyl group (–CF₃) in OTF⁻ has a strong electron-withdrawing effect, which evenly distributes the negative charge of O in the sulfonate group (–SO₃⁻) of OTF⁻ through the C–S bond, facilitating the formation of robust Pb–O bonds to passivate uncoordinated Pb²⁺ at grain boundaries. Furthermore, incorporation of [BMIM][OTF] reduces the work function (WF) of CsPbI₂Br by 0.25 eV, optimizing the energy-level alignment with the electron transport layer (ETL, i.e., SnO₂), lowering the interface charge injection barrier and thus significantly improving carrier extraction.

Driven by these mechanisms, the devices modified with [BMIM][OTF] can deliver a PCE of 15.39%, representing a 30.4% improvement over the control device that exhibits a PCE of 11.80%, setting one of the highest records among similar devices. Long-term stability tests show that the unencapsulated devices maintain 94% and 78% of their corresponding initial PCEs after being stored in Ar for 3500 h at 20 °C and 2500 h at 60 °C, significantly surpassing the control devices, which maintain only 57% and 32% of their initial efficiencies under the corresponding conditions. This work provides a generalized additive engineering strategy for defect control in carbon-cathode HTL-free CsPbI₂Br solar cells.

2. Results and discussion

The optimal molar ratio of IL to CsPbI₂Br was determined to be 0.005:1. All CsPbI₂Br films and the related devices, with detailed preparation and characterization in the ESI,† were fabricated at this ratio unless otherwise specified.

To examine the interaction mechanisms between [BMIM][OTF] and CsPbI₂Br, 1-butyl-3-methylimidazolium methane sulfonate ([BMIM][MS]) was also included for analysis, and the chemical structures are shown in Fig. S1.† The density functional theory (DFT) calculations of the electrostatic potential (ESP) distribution reveal that OTF⁻ exhibits a higher charge density due to the stronger electron-withdrawing of –CF₃ in OTF⁻ (see Fig. S2a and b†). Through the inductive effect transmitted via the C–S bond, the negative charge of –SO₃⁻ is delocalized across the entire anion, resulting in a more uniform distribution of negative charge within OTF⁻. This charge delocalization enhances stable adsorption of uncoordinated Pb²⁺ at O. As shown in Fig. 1a and b, the adsorption energy of OTF⁻ with uncoordinated Pb²⁺ is −1.724 eV, which is 30% lower than that of MS⁻ (−1.318 eV), confirming that OTF⁻ demonstrates superior coordination bonding capability and passivation efficacy to uncoordinated Pb²⁺.

The Fourier transform infrared (FTIR) spectra in Fig. 1c and d indicate that upon PbI₂ incorporation, the SO characteristic peaks of MS⁻ shift from 1193 and 1160 cm⁻¹ to 1183 and 1151 cm⁻¹, while those of OTF⁻ shift from 1226 and 1162 cm⁻¹ to 1216 and 1154 cm⁻¹. This phenomenon arises due to Pb²⁺ acting as a Lewis acid and forming a coordination bond with O, which carries lone-pair electrons in –SO₃⁻. During the coordination process, the electron density in the Pb–O bond shifts toward Pb²⁺, thereby reducing the electron density and the vibrational frequency of SO, as reflected by the aforementioned red shift of the FTIR signals.³⁰ X-ray photoelectron spectroscopy (XPS) was performed to elucidate the evolution of interfacial electronic structures. As depicted in Fig. 1e and f, for the OTF⁻-added sample, the Pb 4f_5/2 and Pb 4f_7/2 binding energies decrease from 143.15 and 138.28 eV to 142.78 and 137.88 eV, whereas the MS⁻-added sample exhibits smaller reductions of 0.26 and 0.28 eV. The downward shift in Pb 4f binding energies indicates the increased electron density around Pb, consistent with the FTIR results.³¹ The larger shift observed for OTF⁻ highlights its stronger electron cloud modulation capability and interaction with uncoordinated Pb²⁺. Additionally, the stable positions of the I 3d peaks confirm the absence of interaction between the anion and I vacancy (V_I).

Driven by the steric effect, BMIM⁺ autonomously diffuses to the top and buried interface of CsPbI₂Br, exhibiting unique dual functionality. The ESP calculation identifies the unprotonated N in the imidazole ring of BMIM⁺ as a region of significant negative potential (see Fig. S2c†), enabling it to act as a Lewis base for coordinating with Sn⁴⁺/Pb²⁺. Furthermore, the adsorption of BMIM⁺ at Sn⁴⁺/Pb²⁺ neutralizes the local electric field through electrostatic effects, thereby inhibiting the migration of I⁻ in CsPbI₂Br or O²⁻ in SnO₂ and reducing the formation of V_I and O vacancy (V_O). Adsorption energy calculations in Fig. 2a and b reveal strong BMIM⁺ adsorption on SnO₂ (−1.852 eV) and CsPbI₂Br (−1.565 eV), demonstrating effective passivation to uncoordinated Sn⁴⁺ and Pb²⁺.

The FTIR spectra (Fig. 2c and d) reveal a red shift in the CN stretching vibration signal of BMIM⁺ from 1572 to 1566 or 1562 cm⁻¹ upon mixing with SnO₂ or PbI₂, confirming a strong coordination between N and Sn⁴⁺/Pb²⁺, which thus reduces the electron density and weakens the CN bond strength.³² To further validate the interaction between BMIM⁺ and SnO₂, XPS measurement was conducted on the exposed SnO₂ layer after removing CsPbI₂Br using a UV-curable photoresist (see Fig. S3† for the peeling-off process). As shown in Fig. 2e, the Sn 3d_3/2 and 3d_5/2 binding energies decrease from 495.66 and 487.24 eV to 495.47 and 486.94 eV, signifying an increased electron cloud density around Sn due to the coordination with the unprotonated N in BMIM⁺.^33–35

Fig. 2f shows the O 1s spectra, revealing that both SnO₂ films have an asymmetric broad signal that can be separated into two peaks. The peak around 532 eV represents the lattice O, and the other peak around 533.5 eV indicates V_O. After BMIM⁺ modification, the reduced V_O defects are indicated by the increased intensity ratio of the lattice O to V_O.^36,37 This can be attributed to three mechanisms. The first one is the interaction of BMIM⁺ with uncoordinated Sn⁴⁺ on the SnO₂ surface that strengthens the Sn–O bond, thereby inhibiting the migration of O²⁻ to form V_O. The second one is that BMIM⁺ adsorbs onto the SnO₂ surface via the van der Waals interaction and forms a conformal protective layer that blocks O diffusion and mitigates O loss under thermal or environmental stress. Additionally, the positively charged BMIM⁺ compensates for local charge imbalance caused by V_O through electrostatic interactions, increasing the thermodynamic energy barrier for V_O formation.

To elucidate the effects of IL on the microstructure and lattice stress of CsPbI₂Br, we systematically investigated the regulatory mechanisms of IL additives on surface morphology, crystallinity, and residual stress. As shown in Fig. 3a–c, the scanning electron microscopy (SEM) analysis reveals that IL incorporation significantly increases grain size while reducing the grain boundary density. The average grain size of the pristine CsPbI₂Br is about 250 nm, whereas the [BMIM][OTF]-added film exhibits a 96% increase to 489 nm with a narrow grain size distribution, outperforming the [BMIM][MS]-added film (442 nm). As illustrated in Fig. 3d–f, the atomic force microscopy (AFM) characterization demonstrates that the [BMIM][OTF]-added film has a significant 51% reduction in the root mean square (RMS) surface roughness, from 22.4 to 10.9 nm. Moreover, the [BMIM][OTF]-added layer exhibits vertically aligned grains with fewer grain boundary cracks compared to the pristine one (see the cross-sectional SEM images in Fig. S4†). These results indicate that [BMIM][OTF] possesses excellent capabilities in regulating crystallization, optimizing crystal growth orientation, and flattening the surface of CsPbI₂Br, which potentially aids interfacial electron transfer.³⁸

As shown in Fig. 3g, the X-ray diffraction (XRD) patterns of IL-added films exhibit no evident shift in the (100) and (200) diffraction peaks, indicating that neither [BMIM][MS] nor [BMIM][OTF] alters the perovskite crystal structure. The full width at half maximum (FWHM) of the (200) peak for the [BMIM][OTF]-added film decreases by 44%, i.e., from 0.32 to 0.18°, compared to that of the control one (see Fig. S5†). This reduction reflects enhanced crystallinity, consistent with SEM and AFM observations.³⁹ Grazing-incidence XRD (GIXRD) was employed to evaluate the influence of [BMIM][MS] and [BMIM][OTF] on residual stress within CsPbI₂Br. As illustrated in Fig. 3h, i and S6,† the (200) diffraction peaks of the pristine film shift to lower angles with increasing incidence angle ω (from 0.5 to 1.5°), accompanied by a significant expansion in d-spacing, indicating severe residual tensile stress. In contrast, the [BMIM][OTF]-added film shows minimal variations in d-spacing across different incidence angles, confirming effective release of residual tensile stress.⁴⁰

Fig. 4a shows the ultraviolet-visible (UV-vis) absorbance spectra, which demonstrate the unaltered absorption edge at 650 nm, except for the limited absorbance enhancement, upon IL addition. It thus suggests that IL additives do not affect the optical bandgap of CsPbI₂Br (E_opt = 1.91 eV, see Fig. S7† for the Tauc plots). To explore the impact of IL on the carrier transport mechanism, ultraviolet photoelectron spectroscopy (UPS) was employed to measure the WF of CsPbI₂Br. As depicted in Fig. 4b, [BMIM][OTF] reduces the WF from 3.69 to 3.44 eV, effectively optimizing the interfacial energy-level alignment and facilitating efficient dissociation and extraction of photogenerated carriers accordingly.^41,42

To further verify carrier extraction dynamics, the steady-state and time-resolved photoluminescence (SSPL/TRPL) spectra were obtained from a glass/ITO/SnO₂/CsPbI₂Br architecture, with the optical excitation applied through the perovskite layer. As shown in Fig. 4c, both pristine and IL-added cases have characteristic emission peaks at 650 nm, corresponding to the 1.91 eV bandgap of CsPbI₂Br. However, upon incorporation of [BMIM][OTF], the significant reduction in the SSPL intensity compared to that of the pristine case indicates pronounced photoluminescence quenching. This finding suggests improved carrier transfer from the IL-added CsPbI₂Br to SnO₂.⁴³ The TRPL decay curves presented in Fig. 4d further confirm the reduced average carrier lifetime (τ) from 16.01 to 4.28 ns (see Table S1†). This indicates that [BMIM][OTF] inhibits carrier recombination and substantially enhances the interface charge extraction efficiency.⁴⁴

According to the Shockley–Read–Hall theory, defect states act as carrier trapping centers that mediate non-radiative recombination, resulting in an inverse proportionality between τ and defect density (N_t). To quantitatively determine N_t, the trap-filling-limit voltage (V_TFL) was measured through space charge-limited current (SCLC) characterization. The relationship between V_TFL and N_t is governed by N_t = 2εV_TFL/(eL²) (ε represents the permittivity of CsPbI₂Br, e is the elementary charge, and L is the thickness of the CsPbI₂Br layers). As illustrated in Fig. 4e and f, modification with [BMIM][OTF] results in a reduced V_TFL for the electron-only configuration (ITO/SnO₂/CsPbI₂Br/PCBM/carbon) from 1.71 to 0.68 V and for the hole-only one (ITO/NiO_x/CsPbI₂Br/P3HT/carbon) from 1.62 to 0.73 V, corresponding to reductions in N_t by approximately 40% and 55%, respectively. This significant decline in N_t arises from both the formation of Pb–O bonds between OTF⁻ and uncoordinated Pb²⁺ as well as passivation of BMIM⁺ at defects. To verify the efficacy of [BMIM][OTF] and [BMIM][MS] in passivating defects, PL analysis was conducted for the CsPbI₂Br layer on bare glass. Fig. 4g shows an evidently increased PL intensity for the [BMIM][OTF]-added layer, suggesting a marked N_t reduction and a corresponding decrease in non-radiative recombination.⁴⁵ Furthermore, TRPL measurements were performed on this structure to measure τ of photogenerated carriers. Following the incorporation of [BMIM][MS] and [BMIM][OTF], the τ extends from 31.46 to 58.09 and 96.31 ns respectively (see Fig. 4h and Table S2†), confirming reduced defect densities in IL-added CsPbI₂Br layers.⁴⁶

To investigate the dependence of the photovoltaic performance of carbon-cathode HTL-free CsPbI₂Br PSCs on the concentration of IL additives, devices with the configuration of ITO/SnO₂/CsPbI₂Br/carbon (see Fig. S8†) were fabricated by adding [BMIM][MS] or [BMIM][OTF] to the CsPbI₂Br precursor solutions at molar ratios of 0.25%, 0.50%, and 0.75%. As demonstrated in Fig. S9,† the current density–voltage (J–V) curves reveal an optimal concentration of 0.50% for each IL additive. Fig. 5a shows that the champion device modified with [BMIM][OTF] delivers a leading PCE of 15.39%, accompanied by a short-circuit current density (J_sc) of 14.92 mA cm⁻², an open-circuit voltage (V_oc) of 1.333 V, and a fill factor (FF) of 0.774. This represents a 30.4% increment over the control device with a PCE of 11.80%, making it one of the highest PCEs reported for carbon-cathode HTL-free CsPbI₂Br solar cells with PCEs over 15% as summarized in Table S3.† To ensure process reproducibility, 15 independent control and modified devices were fabricated in each batch. The statistical analysis of key parameters is summarized in Fig. S10 and Table S4,† showing a reduced PCE standard deviation from ±0.44% to ±0.10%. Additionally, the notably narrowed distributions of V_oc, J_sc, and FF for the [BMIM][OTF]-added devices highlight improved repeatability for this additive strategy.

The external quantum efficiency (EQE) spectra, along with their integrated J_sc curves (see Fig. 5b), further demonstrate the reduced non-radiative recombination and thus the improved collection efficiency for photogenerated carriers of the IL-added devices.⁴⁷ The integrated J_sc rises from 14.18 to 14.70 and 14.78 mA cm⁻², aligning with the J_sc change trends in the illuminated J–V curves. Furthermore, the hysteresis index decreases from 23.4% for the control device to 7.0% for the [BMIM][MS]-added device and further to 4.2% for the [BMIM][OTF]-added one (see Fig. S11 and Table S5†), due to a reduction in defects and enhancement in interfacial carrier transport efficiency in IL-added devices.

The electrochemical impedance spectroscopy (EIS) analysis (Fig. 5c and S12†) reveals that the [BMIM][OTF]-added device has the highest recombination resistance (R_rec) of 49 kΩ and the lowest series resistance (R_s) of 4.4 Ω. The highest R_rec and lowest R_s suggest the smallest leakage current and ohmic losses because of the efficient separation and transport of electron–hole pairs due to the enhanced internal electric field and reduced recombination in the [BMIM][OTF]-added device, respectively.^48,49 Light-intensity-dependent measurements were performed to explore the relationship between J_sc, V_oc, and incident light intensity. As illustrated in Fig. 5d, the exponential factor (α), related to bimolecular recombination and the space-charge effect, increases from 0.949 for the control device to 0.971 for the [BMIM][MS]-added device and further to 0.977 for the [BMIM][OTF]-added one, nearing the ideal value of 1, demonstrating a reduction in N_t for IL-added devices.⁵⁰ The relationship between V_oc and light intensity is described by the ideality factor (n_id), as depicted in Fig. 5e. After incorporating ILs, n_id decreases from 1.89 for the control device to 1.40 and 1.36 for IL-added devices, confirming that [BMIM][MS] and [BMIM][OTF] effectively passivate defects in CsPbI₂Br, thereby reducing non-radiative recombination and suppressing defect-induced charge trapping. This transition changes the charge transport mechanism from a recombination-dominated process in control devices to a diffusion-controlled process in IL-added devices, ultimately increasing device performance.⁵¹ The built-in voltage (V_bi) estimated from capacitance–voltage measurements (see Fig. 5f) increases from 1.19 V for the control device to 1.29 V for the [BMIM][OTF]-added device, supporting V_oc change trends observed from illuminated J–V results.⁵²

We further checked the impact of IL addition on device stability, and the normalized PCE changes of the unencapsulated devices are presented in Fig. 5g and h. It has been noted that the [BMIM][OTF]-added solar cell retains over 94% of its initial PCE after being stored in Ar at 20 °C for 3500 h, while the PCE of the control device drops to 57%. After being stored in Ar at 60 °C for 2500 h, the [BMIM][OTF]-added device maintains more than 78% of its initial performance, compared to 32% for the control one. To understand the improved stability, we measured the water contact angle of CsPbI₂Br (see Fig. S13†). The contact angle increases from 40.3° for the control sample to 57.8° and 63.4° when adding optimal [BMIM][MS] and [BMIM][OTF] into CsPbI₂Br, respectively, indicating enhanced hydrophobicity of IL-added CsPbI₂Br. This can be attributed to the oriented arrangement of the hydrophobic alkyl chains of BMIM⁺ on the CsPbI₂Br surface, thus forming a hydrophobic protective layer that prevents water penetration. Additionally, defect passivation by anions at grain boundaries further inhibits moisture penetration through defect channels.

3. Conclusions

This work reports the synergistic optimization of comprehensive performance for carbon-cathode HTL-free CsPbI₂Br solar cells using a bifunctional IL additive, i.e., [BMIM][OTF]. Thanks to the steric hindrance effect, BMIM⁺ accumulates at both the top and buried interfaces of CsPbI₂Br, thus passivating the corresponding uncoordinated Pb²⁺ in CsPbI₂Br and Sn⁴⁺ in SnO₂ by the N atoms in the imidazole rings. The hydrophobic alkyl chain of BMIM⁺ enhances the moisture resistance of CsPbI₂Br. Meanwhile, OTF⁻ can enter the grain boundaries and passivate the corresponding uncoordinated Pb²⁺. Accompanied by improved crystallinity and reduced surface roughness of the CsPbI₂Br layers, as well as optimized interface energy-level alignment with the SnO₂ ETL due to [BMIM][OTF] addition, a leading PCE of 15.39% among the carbon-cathode HTL-free CsPbI₂Br solar cells is achieved, representing a 30.4% increment compared with a PCE of 11.80% for the control device. In addition, [BMIM][OTF]-added devices exhibit reduced J–V hysteresis and notably improved long-term stability. This simple bifunctional IL additive strategy provides a new approach for boosting the comprehensive performance of carbon-cathode HTL-free CsPbI₂Br solar cells.

Data availability

The data supporting this article have been included as part of the ESI.†

Author contributions

Wenxuan Li and Xueyan Ma: methodology, data curation, data analysis, manuscript writing – original draft. Hai Liu: data curation. Hongyan Cheng: formal analysis. Zhengjun Meng: methodology. Xiaoyang Liu: visualization. Guodong Wan: simulation. Zhe Gao: validation. Yali Li: project administration, supervision, resources. Yujun Fu: resources, visualization. Deyan He: supervision, resources. Junshuai Li: supervision, conceptualization, resources, writing – review & editing. All authors edited and revised the manuscript.

Conflicts of interest

There are no conflicts of interest to declare.

References

1. X. Liu, H. Tong, Y. Li, G. Wan, W. Li, Z. Gao, H. Ma, Q. Liu, Y. Fu, D. He, Z. Li and J. Li, J. Power Sources, 2024, 614, 235046 CrossRef CAS .
2. X. Liu, H. Tong, Y. Li, W. Li, G. Wan, Q. Liu, Y. Fu, D. He, Z. Li and J. Li, Small, 2024, 21, 2410160 CrossRef PubMed .
3. The national renewable energy laboratory (NREL), https://www2.nrel.gov/pv/cell-efficiency (accessed April 2025).
4. M. Saliba, Science, 2018, 359, 388–389 CrossRef CAS PubMed .
5. Y. Zhou and Y. Zhao, Energy Environ. Sci., 2019, 12, 1495–1511 RSC .
6. B. W. Park and S. I. Seok, Adv. Mater., 2019, 31, 1805337 CrossRef PubMed .
7. J. Song, H. Liu, W. Pu, Y. Lu, Z. Si, Z. Zhang, Y. Ge, N. Li, H. Zhou, W. Xiao, L. Wang and M. Sui, Energy Environ. Sci., 2022, 15, 4836–4849 RSC .
8. G. Dastgeer, S. Nisar, M. W. Zulfiqar, J. Eom, M. Imran and K. Akbar, Nano Energy, 2024, 132, 110401 CrossRef .
9. J. Song, H. Xie, E. L. Lim, A. Hagfeldt and D. Bi, Adv. Energy Mater., 2022, 12, 2201854 CrossRef CAS .
10. L. Yan, Y. Li, S. Li, X. Sun, Y. Li, X. Han, M. Huang and X. Tao, J. Mater. Chem. A, 2022, 10, 14799–14809 RSC .
11. V. M. Goldschmidt, Trans. Faraday Soc., 1929, 25, 253–283 RSC .
12. M. Wang, F. Cao, M. Wang, K. Deng and L. Li, Adv. Mater., 2021, 33, 2006745 CrossRef CAS PubMed .
13. X. Duan, J. Duan, N. Liu, J. Li, J. Dou, X. Zhang, Q. Guo, Y. Wang, Z. Wang, Y. Zhao, C. Jiang, J. Li and Q. Tang, SusMat, 2024, 4, e233 CrossRef CAS .
14. J. Li, J. Duan, Q. Guo, Z. Qi, X. Duan, H. Li, S. Geng, N. Liu, B. He, P. Yang and Q. Tang, Adv. Funct. Mater., 2023, 33, 2308036 CrossRef CAS .
15. C. Kang, S. Xu, H. Rao, Z. Pan and X. Zhong, ACS Energy Lett., 2023, 8, 909–916 CrossRef CAS .
16. G. Zhang, J. Zhang, Z. Yang, Z. Pan, H. Rao and X. Zhong, Adv. Mater., 2022, 34, 2206222 CrossRef CAS PubMed .
17. H. Liu, X. Ma, W. Li, G. Wan, X. Liu, Y. Li, Y. Fu, D. He and J. Li, Prog, Photovoltaics Res. Appl., 2025, 0, 1–9 Search PubMed .
18. G. D. Wan, Y. L. Li, X. Y. Liu, W. X. Li, H. Liu, M. B. Du, Q. M. Liu, Y. J. Fu, D. Y. He and J. S. Li, Phys. Scr., 2025, 100, 025020 CrossRef CAS .
19. W. Li, H. Tong, Y. Li, X. Liu, G. Wan, X. Ma, H. Liu, Z. Gao, Y. Fu, D. He, Z. Li and J. Li, Small, 2024, 20, 2406784 CrossRef CAS PubMed .
20. N. Liu, J. Duan, H. Li, L. Ma, B. Wang, J. Li, X. Duan, Q. Guo, J. Dou, S. Geng, Y. Liu, C. Zhang, Y. Liu, B. He, X. Yang and Q. Tang, Adv. Energy Mater., 2024, 14, 2402443–2402453 CrossRef CAS .
21. H. Li, J. Duan, C. Zhang, N. Liu, L. Ma, X. Duan, J. Dou, Q. Guo, B. He, Y. Zhao and Q. Tang, Angew. Chem., Int. Ed., 2024, 64, e202419061 CrossRef PubMed .
22. S. Geng, J. Duan, N. Liu, H. Li, X. Zhu, X. Duan, Q. Guo, J. Dou, B. He, Y. Zhao and Q. Tang, Angew. Chem., Int. Ed., 2024, 63, e202407383 CrossRef CAS PubMed .
23. Y. Xu, F. Liu, R. Li, Y. Jing, Q. Chen, X. Chen, C. Deng, Z. Du, W. Sun, J. Wu and Z. Lan, Chem. Eng. J., 2023, 461, 141895 CrossRef CAS .
24. Y. Wang, Y. Li, Z. Gao, Q. Chen, W. Liu, Y. Fu, Q. Liu, D. He and J. Li, ACS Appl. Mater. Interfaces, 2023, 15, 53558–53567 CrossRef CAS PubMed .
25. H. Li, J. Duan, N. Liu, L. Ma, J. Dou, X. Zhang, Q. Guo, Y. Zhao, B. He and Q. Tang, J. Mater. Chem. A, 2024, 12, 23067–23075 RSC .
26. X. Yao, Z. Qi, P. Yang, J. Li and W. Yang, Chem. Eng. J., 2023, 470, 144315 CrossRef CAS .
27. X. Yao, Y. Li, Y. Wang, Z. Gao, Q. Liu, Y. Fu, D. He and J. Li, ACS Appl. Nano Mater., 2022, 5, 10064–10072 CrossRef CAS .
28. X. Ma, H. Tong, W. Li, H. Liu, X. Liu, G. Wan, Y. Li, Y. Fu, Z. Li, D. He and J. Li, Nano Lett., 2025, 25, 8851–8858 CrossRef CAS PubMed .
29. F. Wang, D. Duan, M. Singh, C. M. Sutter-Fella, H. Lin, L. Li, P. Naumov and H. Hu, Energy Environ. Mater., 2022, 6, e12435 CrossRef .
30. H. Liu, Y. Gao, F. Xu, X. Zhang, A. Ullah, L. Xu, S. Zhang, J. Wang, S. De Wolf and H. L. Wang, Adv. Funct. Mater., 2024, 34, 2315843 CrossRef CAS .
31. L. Liu, S. Huang, Y. Lu, P. Liu, Y. Zhao, C. Shi, S. Zhang, J. Wu, H. Zhong, M. Sui, H. Zhou, H. Jin, Y. Li and Q. Chen, Adv. Mater., 2018, 30, 1800544 CrossRef PubMed .
32. X. Ge, Z. Huang, B. Shi, P. Wang, Z. Liu, Y. Gao, X. Du, Y. Zhao and X. Zhang, Adv. Funct. Mater., 2024, 35, 2417493 CrossRef .
33. J. Cheng, H. Zhang, S. Zhang, D. Ouyang, Z. Huang, M. K. Nazeeruddin, J. Hou and W. C. H. Choy, J. Mater. Chem. A, 2018, 6, 23865–23874 RSC .
34. W. Zhang, S. Pathak, N. Sakai, T. Stergiopoulos, P. K. Nayak, N. K. Noel, A. A. Haghighirad, V. M. Burlakov, D. W. Dequilettes, A. Sadhanala, W. Li, L. Wang, D. S. Ginger, R. H. Friend and H. J. Snaith, Nat. Commun., 2015, 6, 10030 CrossRef CAS PubMed .
35. R. Wang, J. Xue, K.-L. Wang, H.-K. Wang, Y. Luo, D. Fenning, G. Xu, S. Nuryyeva, T. Huang, Y. Zhao, J. L. Yang, J. Zhu, M. Wang, S. Tan, I. Yavuz, K. N. Houk and Y. Yang, Science, 2019, 366, 1509–1513 CrossRef CAS PubMed .
36. Y. Yang, Y. Wang and S. Yin, Appl. Surf. Sci., 2017, 420, 399–406 CrossRef CAS .
37. J. Chen, X. Zhao, S. G. Kim and N. G. Park, Adv. Mater., 2019, 31, 1902902 CrossRef PubMed .
38. W. Li, Y. Li, Z. Gao, G. Wan, X. Liu, Y. Fu, Q. Liu, D. He and J. Li, J. Mater. Chem. C, 2024, 12, 1421–1429 RSC .
39. C. Xie, J. Li, Y. Wang, Q. Chen, W. Liu, X. Yao, Y. Fu, Q. Liu, Q. Chen, D. He and Y. Li, IEEE J. Photovoltaics, 2022, 12, 948–953 Search PubMed .
40. Z. Li, C. Jia, Z. Wan, J. Cao, J. Shi, J. Xue, X. Liu, H. Wu, C. Xiao, C. Li, M. Li, C. Zhang and Z. Li, Nat. Commun., 2025, 16, 1771 CrossRef CAS PubMed .
41. Z. Gao, Y. Li, W. Li, Y. Wang, G. Wan, X. Liu, Y. Fu, Q. Liu, D. He and J. Li, ACS Appl. Electron. Mater., 2023, 5, 6221–6227 CrossRef CAS .
42. Z. Wang, T. Xu, N. Li, Y. Liu, K. Li, Z. Fan, J. Tan, D. Chen, S. Liu and W. Xiang, Energy Environ. Sci., 2024, 17, 7271–7280 RSC .
43. Y. Wang, C. Xie, X. Yao, Q. Chen, W. Liu, Y. Fu, Q. Liu, J. Li, Y. Li and D. He, J. Materiomics, 2022, 8, 358–365 CrossRef .
44. J. Zhang, F. Gao, Z. Wang, Y. Li, L. Lang, T. Zhou, R. Li, F. Yang, Q. Tian and S. Liu, Nano Energy, 2025, 135, 110633 CrossRef CAS .
45. Y. Wang, J. Li, Q. Chen, W. Liu, Z. Gao, Y. Fu, Q. Liu, D. He and Y. Li, ACS Appl. Energy Mater., 2023, 6, 4584–4592 CrossRef CAS .
46. S. Tai, S. Wan, B. Fan, X. Xiong, H. Fu, Y. Ma and Q. Zheng, Nano Energy, 2025, 135, 110670 CrossRef CAS .
47. Q. Ma, Y. Wang, L. Liu, P. Yang, W. He, X. Zhang, J. Zheng, M. Ma, M. Wan, Y. Yang, C. Zhang, T. Mahmoudi, S. Wu, C. Liu, Y.-B. Hahn and Y. Mai, Energy Environ. Sci., 2024, 17, 1637–1644 RSC .
48. X. Yao, L. Yin, Y. Wang, W. Liu, C. Xie, Q. Liu, Y. Fu, Y. Li, J. Li and D. He, Sci. China Mater., 2022, 65, 896–903 CrossRef CAS .
49. Z. Wu, W. Zhang, C. Xie, L. Zhang, Y. Wang, Y. Zhang, Q. Liu, Y. Fu, Y. Li, J. Li and D. He, IEEE J. Photovoltaics, 2020, 10, 1353–1358 Search PubMed .
50. Y. Zhao, C. Wang, T. Ma, L. Zhou, Z. Wu, H. Wang, C. Chen, Z. Yu, W. Sun, A. Wang, H. Huang, B. Zou, D. Zhao and X. Li, Energy Environ. Sci., 2023, 16, 2080–2089 RSC .
51. C. Lu, X. Li, H. Yuan, W. Zhang, X. Guo, A. Liu, H. Yang, W. Li, Z. Cui, Y. Hu and J. Fang, Chem. Eng. J., 2024, 480, 147267 CrossRef CAS .
52. Q. Zhou, B. Liu, Y. Chen, D. Ma, X. Han, D. He, Z. Zhang, H. Yang, L. Ding, J. Feng, J. Yi, C. Chen and J. Chen, Adv. Funct. Mater., 2024, 34, 2315064 CrossRef CAS .

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta04556h

‡ These authors contributed equally.

This journal is © The Royal Society of Chemistry 2025