Stoichiometric gradient rebalancing achieves surface reconstruction and bulk homogenization in high-performance vapor-deposited perovskite solar cells †

Changyu Duan ^a, Yichen Dou ^a, Shenghan Hu ^a, Xinyu Deng ^a, Meichen Liu ^a, Mengjun Liu ^a, Guijie Liang *^b, Yong Peng ^a, Yi-Bing Cheng ^ac and Zhiliang Ku *^a
aState Key Laboratory of Advanced Technology for Materials Synthesis and Processing, Wuhan University of Technology, Wuhan 430070, Hubei Province, China. E-mail: zhiliang.ku@whut.edu.cn
^bHubei Key Laboratory of Low Dimensional Optoelectronic Material and Devices, Hubei University of Arts and Science, Xiangyang 441053, Hubei Province, China. E-mail: guijie-liang@hbuas.edu.cn
^cXianhu Laboratory of the Advanced Energy Science and Technology Guangdong Laboratory, Foshan 528200, Guangdong Province, China

First published on 21st June 2025

---

---

Introduction

In recent years, perovskite solar cells (PSCs) have been rapidly developed, with power conversion efficiencies (PCEs) now rivaling those of crystalline silicon cells,^1,2 positioning PSCs as a promising next-generation low-cost photovoltaic technology.³ However, the transition of perovskite technology from laboratory research to industrial-scale production still faces significant challenges, particularly in the reproducible fabrication of high-quality perovskite films over large areas and at high throughput.⁴ Compared with conventional solution processing, vapor deposition techniques offer notable advantages, including uniform film thickness, the absence of solvent residues, and compatibility with a variety of substrates.^5,6 These features make vapor deposition a highly promising way for large-scale PSC manufacturing and have attracted increasing attention in recent years.^7,8

In 2013, Snaith et al. first reported the fabrication of CH₃NH₃PbI_3−xCl_x perovskite films via dual-source vapor deposition, achieving a PCE exceeding 15% in a planar heterojunction structure.⁹ Shortly after, Qi et al. successfully employed chemical vapor deposition to fabricate large-area FAPbI₃ films, achieving a PCE of 9.0% in a mini-module with an active area of 12 cm².¹⁰ More recently, Yi et al. employed sequential vapor deposition to achieve a remarkable PCE of 24%.¹¹ In multiple studies, vapor-deposited technology has also demonstrated excellent device stability,¹² batch-to-batch consistency,¹³ and potential for large-area fabrication.^14–16 Despite these advantages, vapor-deposited PSCs still lag compared to their solution-processed counterparts, primarily due to the insufficient crystal quality of the perovskite films, which limits further performance improvements. Recent studies have revealed that defects such as iodine vacancies commonly form during the growth of vapor-deposited films, at grain boundaries and within grains, acting as non-radiative recombination centers.¹⁷ To address this issue, various strategies have been explored to enhance crystal quality, such as introducing intermediate phases to promote crystal growth,¹⁸ employing annealing to regulate crystallization kinetics,¹⁹ and modifying the film composition by incorporating mixed-cation and mixed-halide materials.²⁰ While these approaches have improved device performance to some extent, a thorough understanding of the growth mechanisms behind vapor deposition and their precise control is still lacking. Thus, gaining deep insight into the evolution of film formation and crystallization during vapor deposition and establishing a clear relationship between film growth mechanisms and crystal quality are essential for enabling efficient, controllable, and scalable perovskite fabrication.

Unlike solution-based methods, where nucleation and crystallization are governed by solvent coordination and fluid dynamics,²¹ vapor-deposited perovskites exhibit distinct growth mechanisms that introduce additional complexity. The transport behavior of precursors, especially the diffusion of organic species, plays a crucial role in crystallization and film formation.²² This is especially critical in widely used two-step vapor deposition methods, where organic halides (e.g., FAI and FACl) must diffuse through the pre-deposited inorganic layer (e.g., PbI₂). The rate and uniformity of this diffusion strongly influence the crystallization pathway. Studies have shown that modifying the crystallinity of the PbI₂ precursor film can alleviate diffusion barriers, enhance the transport of organic salts, and facilitate perovskite formation.^23,24 For example, Zhang et al. obtained high-quality, fully converted perovskite films by controlling the deposition temperature to enhance diffusion pathways for organic salts.²⁵ Nevertheless, most of the current research has focused on tuning the apparent diffusion process, while systematic understanding is still lacking regarding how diffusion-dominated growth intrinsically affects the crystal quality, defect characteristics, and compositional distribution of vapor-deposited perovskite films. These fundamental questions remain largely unanswered. Moreover, few targeted strategies have been developed to mitigate structural or compositional defects induced by diffusion limitations. Therefore, there is an urgent need for investigations to clarify the relationship between diffusion-crystallization behavior and defect evolution and to develop precise control strategies that can improve crystal quality and enhance device performance.

In this work, we reveal for the first time that the diffusion-limited nature of vapor-deposited perovskite crystallization leads to vertical compositional gradients within the film. To address this, we propose a simple yet effective stoichiometric gradient rebalancing (SGR) strategy. By precisely depositing a PbI₂/PbCl₂ layer on the surface of the as-deposited perovskite film followed by a brief secondary reaction, we compensate for the excess organic component gradient and achieve compositional uniformity throughout the film from surface to bottom. Systematic characterization studies show that SGR-treated films exhibit near-ideal I/Pb ratios, significantly improved compositional homogeneity, stabilized crystal structures, reduced defect state densities, and enhanced crystal quality. Device tests further confirm that this strategy significantly enhances photovoltaic performance, achieving a PCE of 22.45% for small-area cells and 19.92% for mini-modules (5 × 5 cm²), representing advanced performance in vapor-deposited PSCs. In addition, after 500 hours of continuous operation at the maximum power point, the SGR device maintained an initial efficiency of more than 80%. This work not only provides an effective approach to enhance the performance of vapor-deposited PSCs, but also offers important insights into their crystallization behavior and defect formation mechanisms, paving the way for the further development and practical application of vapor-deposited perovskites.

Results and discussion

The perovskite films in this work were fabricated via a two-step vapor deposition method, which has been extensively optimized in our previous studies (Fig. 1a).^26–28 An inorganic PbI₂ layer was first thermally evaporated onto SnO₂/FTO glass substrates, followed by controlled heating in an organic halide atmosphere to induce vapor–solid reactions for perovskite formation. The organic vapor was supplied by a separate FAI-coated glass substrate. Details of the fabrication are provided in the ESI.† The film growth mechanism, as illustrated in Fig. 1b, involves organic halide adsorption onto the PbI₂ surface, followed by inward diffusion and top-down crystallization. Cross-sectional SEM images obtained at different reaction stages clearly reveal this top-down crystallization growth mode (Fig. 1c). Specifically, this crystallization growth mode comprises two key components: the top-down diffusion of organic halides within the precursor film and the perovskite grain growth resulting from the reaction between organic halides and lead halide. Our previous work has demonstrated that a slower diffusion process acts as the rate-determining step in crystallization growth.²⁹ Driven by concentration gradients, this diffusion may lead to undesirable non-stoichiometric surface enrichment and vertical compositional gradients in the final film.

To investigate the impact of this diffusion-limited top-down growth mode on vapor–solid reaction films, we systematically analyzed the vertical composition gradients in the final perovskite films. Time-of-Flight Secondary Ion Mass Spectrometry (TOF-SIMS) was employed to map the distribution of FA⁺ cations and Pb elements across the pristine perovskite film (Fig. 1d and S1a†).^25,30 The FA⁺ profile exhibited a distinct top-down concentration gradient with significant surface enrichment, while Pb signals remained uniformly distributed throughout the film. The Sn⁺ signal detected in the bottom region originated from the SnO₂/FTO substrate. To corroborate these findings, XPS depth profiling was conducted to quantify the I/Pb ratio at varying depths (Fig. 1e).^31,32 The pristine vapor–solid reaction film exhibited severe iodine excess at the surface (I/Pb = 3.46), with gradually decreasing ratios at 200 nm (3.14) and 400 nm depths (3.06), confirming a vertical iodine gradient. These TOF-SIMS and XPS results collectively demonstrate that the diffusion-dominated crystallization mechanism inherently induces non-stoichiometric surface enrichment and vertical compositional gradients of organic halides in vapor–solid reaction perovskite films.

To further explore the impact of compositional nonuniformity on the perovskite crystal structure, grazing-incidence X-ray diffraction (GIXRD) was performed at different depths.^33,34 As shown in Fig. 1f and S2a,† the (100), (110), and (210) diffraction peaks of the vapor–solid reaction perovskite films exhibited noticeable shifts to lower angles with an increasing grazing incidence angle, indicating lattice expansion and structural distortion due to the compositional gradient. Such crystalline inhomogeneity can adversely affect device performance.³⁵ Notably, no secondary phases (unreacted FAI) were detected even at minimal incidence angles (0.1°), suggesting that the excess organic halides were incorporated into the perovskite lattice rather than existing as separate molecular phases. This incorporation leads to deviation from the stoichiometric composition.

To address the stoichiometric imbalance introduced by the growth mechanism, we developed a Stoichiometric Gradient Rebalancing (SGR) strategy (Fig. 2a). This approach involves vapor deposition of a precise amount of lead halide onto the surface of gradient-stoichiometry perovskite films, followed by a re-reaction to compensate for excess organic halide and redistribute the stoichiometric gradient. Fig. 2b shows a comparison of the microstructural morphology of the pristine vapor–solid reaction perovskite film before and after PbI₂ deposition.

Precise control of this process is essential for achieving high-quality crystalline films, maintaining stoichiometric balance and avoiding excess PbI₂. PbI₂ layers of varying thicknesses (1, 5, 15, and 30 nm) were deposited onto perovskite films, followed by a re-reaction at 160 °C for 5 min to ensure a full reaction.²⁸ The morphological analysis (Fig. S3†) confirmed that thin PbI₂ layers formed uniform and complete surface coverage, which would be beneficial for future large-area film compositional control. After the re-reaction (Fig. S4†), films with 1 and 5 nm PbI₂ exhibited a complete reaction with no residual PbI₂, whereas excessive deposition (15 and 30 nm) led to unreacted PbI₂ aggregates on the surface, potentially hindering interfacial charge transport.^36,37 Consistent with these observations, XRD analysis (Fig. S5†) revealed enhanced diffraction intensity, indicating improved crystallinity. The strongest diffraction signal was observed for the 5 nm PbI₂ condition, while residual PbI₂ peaks at 12.6° appeared in films with 15 and 30 nm PbI₂. Further XPS analyses (Fig. S6†) were conducted to investigate the I/Pb ratio of the re-reacted films, providing insights into surface stoichiometry. The results indicated that increasing PbI₂ deposition shifted the surface composition from iodine-rich (I/Pb > 3) to lead-rich (I/Pb < 3), with the 5 nm PbI₂ film exhibiting an optimal I/Pb ratio of 2.98. Photovoltaic devices were also fabricated using the relevant perovskite films to evaluate their photovoltaic performance (Fig. S7†). These precise adjustments ensured that the SGR strategy achieved optimal performance in the microstructure, crystal quality, stoichiometric control, and PCE.

Previous studies have demonstrated that chlorine enhances ion diffusion and accelerates crystallization.^11,38,39 Based on these findings, a controlled amount of PbCl₂ was introduced into the optimized PbI₂ deposition to further facilitate compositional rebalancing and surface reconstruction. To precisely optimize the incorporation ratio of PbCl₂ and evaluate its impact on carrier dynamics, photoluminescence (PL) spectra were measured (Fig. S8a†). With increasing Cl content, PL intensity initially increased, indicating suppression of nonradiative recombination,⁴⁰ but excessive Cl incorporation led to intensity reduction, likely due to defect formation from excessive Cl incorporation into the lattice. Time-resolved PL (TRPL) measurements (Fig. S8b and Table S1†) showed an increase in the carrier lifetime from 148 ns to 410 ns upon optimized Cl incorporation, further supporting the beneficial role of Cl in improving crystal quality. Space-charge-limited current (SCLC) measurements were conducted to assess trap density (Fig. S9†).^18,34 The optimal PbI₂/PbCl₂ ratio (5:1) resulted in the lowest trap-filled limit voltage (V_TFL = 0.17 V), compared to 0.30 V for pure PbI₂ SGR films. Corresponding device performance evaluations further confirmed this optimal PbI₂/PbCl₂ ratio (Fig. S10†). These findings collectively demonstrate that Cl⁻ incorporation enhances compositional rebalancing and improves crystal quality, likely due to the smaller ionic radius of Cl⁻ facilitating ion exchange during rebalancing.

Reaction time and temperature are also crucial parameters in the stoichiometric gradient rebalancing strategy, as they influence ion diffusion, re-reaction rates, and crystallization dynamics. To precisely control the SGR process, it is necessary to optimize both the temperature and duration of the re-reaction. Fig. S11† presents the relationship between the average efficiency of SGR devices and reaction conditions, both with and without PbCl₂. In the absence of PbCl₂, the optimal conditions for maximum device efficiency are 160 °C and 3 minutes. Upon introducing PbCl₂, the optimal temperature and duration decrease to 140 °C and 1 minute, respectively, leading to a 0.3% enhancement in PCE. This result indicates that the incorporation of chlorine reduces the energy barrier for compositional rearrangement, accelerates ion exchange rates, and further highlights the critical role of chlorine in the SGR process.

With the SGR strategy thoroughly optimized, we further evaluated its effectiveness in addressing the vertical stoichiometric imbalance issue in vapor-deposited perovskites. ToF-SIMS was employed to analyze the distribution of FA⁺ ions and Pb elements in SGR films. As shown in Fig. 2c and S1b,† after sufficient film reconstruction, FA⁺ no longer accumulates at the surface with a top-down concentration gradient. Instead, it is uniformly distributed throughout the film. Furthermore, depth-profiling XPS analysis reveals that (Fig. 2d), compared to the pristine film, the SGR film exhibits an improved I/Pb ratio at the surface, which approaches the theoretical I/Pb ratio of 3, reaching 2.98. At depths of 200 nm and 400 nm, the I/Pb ratios are 3.01 and 3.03, respectively, indicating that the FA-enriched surface in the pristine film is effectively rebalanced and that the excess composition gradient within the film is also mitigated (Fig. S12†). The corresponding XPS survey spectra (Fig. S13†) show, in addition to the above-mentioned Pb and I signals, an N1s peak attributed to FA⁺ cations. Additionally, GIXRD measurements were conducted to evaluate the effect of the SGR strategy on perovskite crystal structures at different depths (Fig. 2e). A detailed comparison of GIXRD patterns between pristine and SGR-treated films is presented in Fig. S2.† The SGR-treated film exhibits smaller diffraction peak shifts with an increasing incidence angle, suggesting a more homogeneous crystal structure along the vertical direction. In contrast, the pristine film, due to compositional excess, shows lattice distortions along the vertical direction (Fig. S14†). Furthermore, the increased diffraction intensity in SGR films suggests improved crystallinity, which can be attributed to the redistribution of excess compositional gradients within the crystals (Fig. S15†). These compositional and structural analyses confirm that precise control over the stoichiometric gradient significantly improves the compositional and structural homogeneity in vapor-deposited perovskites, addressing issues of top-down compositional and structural inhomogeneity.

To further understand the SGR strategy, we conducted an in-depth investigation of film properties. First, ultraviolet photoelectron spectroscopy (UPS) was utilized to examine the effect of surface reconstruction on film energy levels (Fig. 3a and S16†).⁴¹ Combining these results with the optical bandgap measured using UV-vis absorption spectroscopy (Fig. S17†), we constructed an energy-level diagram of the three types of perovskite films, as shown in Fig. 3b. The results indicate that compared to the stoichiometric imbalanced pristine film, the valence bands of SGR-treated films with pure PbI₂ and PbI₂–PbCl₂ shift upward, approaching the energy level of the hole transport layer (spiro-OMeTAD).^42,43 This surface energy level adjustment enhances energy-level alignment, which is beneficial for improving the open-circuit voltage (V_oc) of perovskite solar cells.⁴⁴ The SGR strategy also induces noticeable changes in film morphology. SEM images (Fig. 3c) reveal that, after recrystallization through the re-reaction process, SGR films exhibit smoother, more uniform grains with an increased average size of 620 nm, compared to 302 nm in pristine perovskite films. We also observe that the SGR films exhibit a broader grain size distribution, which is likely associated with Ostwald ripening during the recrystallization process.⁴⁵ In this process, smaller grains decrease in size, whereas larger grains increase in size. Water contact angle measurements were further conducted to assess the hydrophobicity of the films. The pristine film, which has an organic-rich surface, shows stronger water affinity with a contact angle of 56.42°, while the SGR film with an optimized stoichiometric ratio exhibits a contact angle of 81.67°, indicating enhanced hydrophobicity. This improved surface hydrophobicity contributes to enhanced moisture stability in perovskite films and devices.⁴⁶

Grazing incidence wide-angle X-ray scattering (GIWAXS) was used to investigate the crystalline structural changes in the films (Fig. 3d and e). The Debye–Scherrer rings located at q = 0.99, 1.40, 1.71, and 1.98 Å⁻¹ correspond to the (100), (110), (111), and (200) planes of the α-phase perovskite, respectively.^11,15 The intensity distribution along the azimuthal angle for the (100) ring in pristine and SGR films is presented in Fig. S18.† Compared to pristine perovskites, SGR perovskites exhibit stronger diffraction intensities at all angles and display out-of-plane orientation, whereas pristine perovskites show a random orientation with uniformly distributed diffraction signals. To further assess film crystallinity, XRD with multiple tilt angles was performed (Fig. 3f and g).⁴⁷ The shift in diffraction peak positions was analyzed to evaluate lattice spacing variations, which reflect internal stress within the crystal. The pristine perovskite exhibits peak shifts toward lower angles with increasing tilt angles, indicating tensile stress parallel to the substrate, which may be attributed to the top-down stoichiometric gradient. In contrast, after the SGR treatment, peak shifts are significantly reduced, indicating that the compositional gradient is mitigated and the internal stress is relieved. A quantitative assessment of stress using the 2θ − sin²(φ) relationship further supports this observation,⁴⁸ as shown in Fig. 3h. The fitted slope for the SGR film is −0.033, whereas for the pristine film, it is −0.167, providing quantitative evidence of stress relaxation.

Improvements in the surface composition and crystal quality may lead to reduced defect densities and extended carrier lifetimes.^49,50 As shown in Fig. 4a, PL measurements on glass substrates indicate significantly enhanced PL intensity for SGR films, suggesting suppression of non-radiative recombination.⁴⁰ Furthermore, TRPL measurements show that the carrier lifetime in SGR films increases from 20 ns to 410 ns,⁵¹ improving by an order of magnitude (Fig. 4b and Table S2†). This enhancement is likely due to the elimination of excess stoichiometric imbalances, reducing defect density and thereby lowering carrier recombination rates. In addition, SCLC measurements reveal that the V_TFL for electron-only SGR devices is significantly lower (0.17 V) compared to pristine devices (0.53 V), with an estimated trap state density reduced from 9.88 × 10¹⁵ cm⁻³ to 3.17 × 10¹⁵ cm⁻³ (Fig. 4c).^18,34 These findings highlight the positive impact of the SGR strategy in improving carrier lifetime, suppressing recombination, and reducing trap state density.

Transient absorption (TA) spectroscopy was further employed to investigate the carrier dynamics of the perovskite films. All perovskite films were deposited on non-conductive glass substrates. Fig. 4d and e present the two-dimensional pseudo-color mapping of ΔA (changes in optical density with and without pumping) as a function of pump-probe delay time and probe wavelength. A negative ΔA signal at 780 nm, corresponding to ground-state bleaching (GSB), was observed in both samples, which is attributed to the transition of charge carriers from the valence band maximum to the conduction band minimum.⁵² The perovskite films on glass substrates exhibit three primary carrier dynamic processes: hot carrier relaxation, carrier diffusion within the film, and carrier recombination. Carrier recombination leads to the recombination of excited electrons and holes back to the ground state, reducing the number of excited-state carriers and consequently weakening the GSB signal. Compared to the pristine film, the SGR film exhibits a slower decay of the GSB signal (Fig. 4f), indicating a lower density of defect states that capture or scatter carriers. This result provides further evidence that the SGR strategy improves carrier dynamics in the perovskite films.

To evaluate the photovoltaic performance enhancement enabled by the SGR strategy, we fabricated perovskite solar cells. The device architecture is illustrated in Fig. 5a, which consists of FTO/SnO₂/perovskite/spiro-OMeTAD/Au. Benefiting from the improved crystal quality and enhanced surface charge transport induced by the elimination of the gradient excess components, the champion efficiency of the SGR device reached 22.45%, with an open-circuit voltage (V_oc) of 1.14 V, a short-circuit current density (J_sc) of 24.29 mA cm⁻², and a fill factor of 81.4%. In contrast, the pristine device exhibited a maximum efficiency of only 20.20%, with a V_oc of 1.11 V, a J_sc of 23.72 mA cm⁻², and a fill factor of 77.0% (Fig. 5b). This efficiency is consistent with our previous reports,¹⁴ representing an advanced level for vapor-deposited perovskites (Table S3†). The external quantum efficiency (EQE) spectra and integrated current densities of the SGR and pristine devices are shown in Fig. 5c. The integrated J_sc increased from 23.27 mA cm⁻² (pristine device) to 23.86 mA cm⁻² (SGR device), which is consistent with the J–V measurements. To assess the operational stability at the maximum power point (MPP), stabilized power output (SPO) measurements were conducted (Fig. 5d). Under continuous illumination (AM 1.5 G) for 180 s, the SGR device exhibited a stable output current density of 21.97 mA cm⁻² and an SPO efficiency of 21.53%, whereas the pristine device only achieved a current density of 19.60 mA cm⁻² and an SPO efficiency of 18.42%. To confirm the reliability of the efficiency enhancement induced by the SGR strategy, we statistically analyzed the PCE distribution of 30 pristine and SGR devices. As shown in Fig. 5e, the efficiency of SGR devices was significantly improved, demonstrating the critical role of the SGR strategy in enhancing device performance. These results demonstrate the crucial role of the SGR strategy in enhancing device efficiency, highlighting the importance of balancing the stoichiometric gradient and achieving an ideal stoichiometric ratio in vapor–solid reaction perovskites.

One significant advantage of the vapor–solid reaction is its compatibility with large-area perovskite film fabrication.^10,27,28 The SGR strategy, developed based on vapor deposition, is also easily applicable to large-area perovskite fabrication.^5,6 Therefore, we successfully implemented this strategy in perovskite modules. As shown in Fig. 5f, a 5 × 5 cm² perovskite module achieved a champion efficiency of 19.92%, highlighting the effectiveness of the SGR strategy for large-area applications. Additionally, we performed multi-point SEM characterization across the large-area films to confirm their uniformity (Fig. S19†).

In addition to enhancing photovoltaic performance, the improved surface reconstruction and crystal uniformity also contribute to improving device stability. As depicted in Fig. 5g, the humidity stability of the devices was evaluated in an environmental chamber (∼25 °C, 60 ± 5% RH) following the ISOS-D-1 protocol.⁵³ After 580 hours of aging, the unencapsulated pristine device retained only 72% of its initial efficiency, whereas the SGR device maintained 93%. This significant improvement in humidity stability is attributed to the removal of excess organic components on the surface and within the bulk, which modifies the hydrophobicity of the film. Furthermore, the thermal stability of the devices was also assessed (Fig. 5h). Encapsulated devices were aged in a nitrogen-filled glovebox at 85 °C (ISOS-D-2I). After 1080 hours, the SGR device retained 84% of its initial efficiency, whereas the pristine device degraded to 51% within 790 hours. To further evaluate operational stability (Fig. 5i), MPP tracking was performed under 1-sun-equivalent white LED illumination (∼25 °C, ambient atmosphere; ISOS-L-1 protocol). After 500 h of continuous operation, the SGR device maintained ∼81% of its initial efficiency, whereas the pristine device decayed to ∼57% after 220 hours. These results demonstrate that the SGR strategy effectively enhances device stability.

Conclusion

In this work, we addressed the inherent stoichiometric gradient issue in vapor–solid reaction-based perovskites, which arises from the top-down crystallization growth mode, by designing a simple and scalable stoichiometric gradient rebalancing (SGR) strategy. Through this strategy, we successfully reconstructed both the surface and bulk of vapor–solid reaction perovskites. The reconstructed perovskites exhibited a uniform top-down stoichiometric composition and crystal structure, leading to defect passivation, improved interfacial energy alignment, and enhanced stability. As a result, we achieved a champion PCE of 22.45% in small-area devices (0.148 cm²) and 19.92% in 5 × 5 cm² modules, representing an advanced level for vapor–solid reaction perovskites. This study, rooted in crystallization growth mechanisms, reveals and resolves the intrinsic stoichiometric inhomogeneity in vapor–solid reaction perovskites, providing deep insights into the growth mechanisms and crystal regulation of vapor-deposited perovskites. Our findings pave the way for future advancements in this field.

Data availability

All the data supporting the findings of this study are available within this article and its ESI.† Any additional information can be obtained from the corresponding authors upon request. Source data are provided with this paper.

Author contributions

Changyu Duan: conceptualization, methodology, formal analysis, writing – original draft. Yichen Dou: validation, formal analysis. Shenghan Hu: investigation, formal analysis. Xinyu Deng: investigation, formal analysis. Meichen Liu: formal analysis. Mengjun Liu: methodology. Guijie Liang: resources, writing – review & editing, funding acquisition. Yong Peng: resources, writing – review & editing, funding acquisition. Yi-Bing Cheng: funding acquisition, supervision. Zhiliang Ku: conceptualization, writing – review & editing, supervision, funding acquisition.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors acknowledge the financial support of the Joint Foundation for Innovation and Development of Hubei Natural Science Foundation (2023AFD032) and Hubei Provincial Natural Science Foundation of China (2023AFA010).

References

1. H. Chen, C. Liu, J. Xu, A. Maxwell, W. Zhou, Y. Yang, Q. Zhou, A. S. R. Bati, H. Wan, Z. Wang, L. Zeng, J. Wang, P. Serles, Y. Liu, S. Teale, Y. Liu, M. I. Saidaminov, M. Li, N. Rolston, S. Hoogland, T. Filleter, M. G. Kanatzidis, B. Chen, Z. Ning and E. H. Sargent, Science, 2024, 384, 189–193 CrossRef CAS PubMed .
2. M. Tao, Y. Wang, K. Zhang, Z. Song, Y. Lan, H. Guo, L. Guo, X. Zhang, J. Wei, D. Cao and Y. Song, Joule, 2024, 8, 3142–3152 CrossRef CAS .
3. Y. Rong, Y. Hu, A. Mei, H. Tan, M. I. Saidaminov, S. I. Seok, M. D. McGehee, E. H. Sargent and H. Han, Science, 2018, 361, eaat8235 CrossRef PubMed .
4. P. Zhu, C. Chen, J. Dai, Y. Zhang, R. Mao, S. Chen, J. Huang and J. Zhu, Adv. Mater., 2024, 36, 2307357 CrossRef CAS PubMed .
5. Y. Jiang, S. He, L. Qiu, Y. Zhao and Y. Qi, Appl. Phys. Rev., 2022, 9, 021305 CAS .
6. T. Abzieher, D. T. Moore, M. Roß, S. Albrecht, J. Silvia, H. Tan, Q. Jeangros, C. Ballif, M. T. Hoerantner, B.-S. Kim, H. J. Bolink, P. Pistor, J. C. Goldschmidt, Y.-H. Chiang, S. D. Stranks, J. Borchert, M. D. McGehee, M. Morales-Masis, J. B. Patel, A. Bruno and U. W. Paetzold, Energy Environ. Sci., 2024, 17, 1645–1663 RSC .
7. J. Ávila, C. Momblona, P. P. Boix, M. Sessolo and H. J. Bolink, Joule, 2017, 1, 431–442 CrossRef .
8. H. Li, M. Liu, M. Li, H. Park, N. Mathews, Y. Qi, X. Zhang, H. J. Bolink, K. Leo, M. Graetzel and C. Yi, iEnergy, 2022, 1, 434–452 Search PubMed .
9. M. Liu, M. B. Johnston and H. J. Snaith, Nature, 2013, 501, 395–398 CrossRef CAS PubMed .
10. M. R. Leyden, Y. Jiang and Y. Qi, J. Mater. Chem. A, 2016, 4, 13125–13132 RSC .
11. H. Li, J. Zhou, L. Tan, M. Li, C. Jiang, S. Wang, X. Zhao, Y. Liu, Y. Zhang, Y. Ye, W. Tress and C. Yi, Sci. Adv., 2022, 8, eabo7422 CrossRef PubMed .
12. H. A. Dewi, J. Li, H. Wang, B. Chaudhary, N. Mathews, S. Mhaisalkar and A. Bruno, Adv. Funct. Mater., 2021, 31, 2100557 CrossRef CAS .
13. J. Borchert, I. Levchuk, L. C. Snoek, M. U. Rothmann, R. Haver, H. J. Snaith, C. J. Brabec, L. M. Herz and M. B. Johnston, ACS Appl. Mater. Interfaces, 2019, 11, 28851–28857 CrossRef CAS PubMed .
14. Y. Wang, J. Chen, Y. Zhang, P. Lv, J. Pan, M. Hu, W. L. Tan, Z. Ku, Y.-B. Cheng, A. N. Simonov and J. Lu, Adv. Mater., 2024, 36, 2412021 CrossRef CAS PubMed .
15. D. Jiang, Z. Liu, J. Li, H. Cao, Y. Qian, Z. Ren, S. Zhang, Y. Qiu, C. Zhang, J. Wei, L. Yang and S. Yin, Joule, 2024, 8, 1161–1175 CrossRef CAS .
16. Z. Tao, Y. Song, B. Wang, G. Tong and L. Ding, J. Semicond., 2024, 45, 040201 CrossRef CAS .
17. Z. Wu, Y. Wang, M. Dai, D. Lin, J. Cai, X. Wu, P. Liu and W. Xie, Sol. RRL, 2024, 8, 2300952 CrossRef CAS .
18. G. Tong, J. Zhang, T. Bu, L. K. Ono, C. Zhang, Y. Liu, C. Ding, T. Wu, S. Mariotti, S. Kazaoui and Y. Qi, Adv. Energy Mater., 2023, 13, 2300153 CrossRef CAS .
19. J. Feng, Y. Jiao, H. Wang, X. Zhu, Y. Sun, M. Du, Y. Cao, D. Yang and S. Liu, Energy Environ. Sci., 2021, 14, 3035–3043 RSC .
20. L. Gil-Escrig, C. Dreessen, F. Palazon, Z. Hawash, E. Moons, S. Albrecht, M. Sessolo and H. J. Bolink, ACS Energy Lett., 2021, 6, 827–836 CrossRef CAS PubMed .
21. L. Chao, T. Niu, W. Gao, C. Ran, L. Song, Y. Chen and W. Huang, Adv. Mater., 2021, 33, 2005410 CrossRef CAS PubMed .
22. Z. Zhao, H. Cao, J. Li, H. Zhu, L. Yang and S. Yin, Appl. Phys. Express, 2018, 11, 105501 CrossRef .
23. Y. Wu, A. Islam, X. Yang, C. Qin, J. Liu, K. Zhang, W. Peng and L. Han, Energy Environ. Sci., 2014, 7, 2934–2938 RSC .
24. Z. Li, K. Chen, Z. Cheng, D. Lin, J. Wang, T. Shi, W. Xie and P. Liu, Phys. Chem. Chem. Phys., 2020, 22, 981–984 RSC .
25. Q. Xu, B. Shi, Y. Li, J. Liu, Y. Li, Z. SunLi, P. Liu, Y. Zhang, C. Sun, W. Han, Q. Huang, D. Zhang, H. Ren, X. Du, Y. Zhao and X. Zhang, Adv. Mater., 2024, 36, 2308692 CrossRef CAS PubMed .
26. L. Luo, Y. L. Zhang, N. Y. Chai, X. Deng, J. Zhong, F. Z. Huang, Y. Peng, Z. L. Ku and Y. B. Cheng, J. Mater. Chem. A, 2018, 6, 21143–21148 RSC .
27. C. Duan, J. Zhong, S. Hu, Y. Dou, J. Lu, Y.-B. Cheng and Z. Ku, Adv. Funct. Mater., 2024, 34, 2313435 CrossRef CAS .
28. Y. Wang, P. Lv, J. Pan, J. Chen, X. Liu, M. Hu, L. Wan, K. Cao, B. Liu, Z. Ku, Y.-B. Cheng and J. Lu, Adv. Mater., 2023, 35, 2304625 CrossRef CAS PubMed .
29. C. Duan, S. Hu, Y. Dou, X. Deng, X. Jiang, Y. Peng, G. Liang, Y.-B. Cheng and Z. Ku, J. Energy Chem., 2025, 108, 263–271 CrossRef CAS .
30. Q. Guesnay, F. Sahli, K. Artuk, D. Turkay, A. G. Kuba, N. Mrkyvkova, K. Vegso, P. Siffalovic, F. Schreiber, H. Lai, F. Fu, M. Ledinský, N. Fürst, A. Schafflützel, C. Bucher, Q. Jeangros, C. Ballif and C. M. Wolff, Adv. Energy Mater., 2024, 14, 2303423 CrossRef CAS .
31. J. Zhao, J. Li, X. Liu, L. J. Bannenberg, A. Bruno and T. J. Savenije, ACS Appl. Energy Mater., 2022, 5, 7049–7055 CrossRef CAS .
32. M. Roß, S. Severin, M. B. Stutz, P. Wagner, H. Köbler, M. Favin-Lévêque, A. Al-Ashouri, P. Korb, P. Tockhorn, A. Abate, B. Stannowski, B. Rech and S. Albrecht, Adv. Energy Mater., 2021, 11, 2101460 CrossRef .
33. X. Liu, X. Jiang, Y. Yin, J. Zhang, H. Tian, J. Guo, X. Guo and C. Li, Energy Environ. Sci., 2024, 17, 6058–6067 RSC .
34. S. Li, X. Xu, X. Wang, N. Huang, J. Fang, D. Lin, Y. Shao, J. Zhou, A. K. K. Kyaw, S. He and L. Qiu, Angew. Chem., Int. Ed., 2025, 64, e202421174 CrossRef CAS PubMed .
35. Z. Liang, Y. Zhang, H. Xu, W. Chen, B. Liu, J. Zhang, H. Zhang, Z. Wang, D.-H. Kang, J. Zeng, X. Gao, Q. Wang, H. Hu, H. Zhou, X. Cai, X. Tian, P. Reiss, B. Xu, T. Kirchartz, Z. Xiao, S. Dai, N.-G. Park, J. Ye and X. Pan, Nature, 2023, 624, 557–563 CrossRef CAS PubMed .
36. Q. Chen, H. Zhou, T.-B. Song, S. Luo, Z. Hong, H.-S. Duan, L. Dou, Y. Liu and Y. Yang, Nano Lett., 2014, 14, 4158–4163 CrossRef CAS PubMed .
37. Y. Gao, H. Raza, Z. Zhang, W. Chen and Z. Liu, Adv. Funct. Mater., 2023, 33, 2215171 CrossRef CAS .
38. 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 .
39. S. Ngqoloda, C. J. Arendse, S. Guha, T. F. Muller, S. C. Klue, S. S. Magubane and C. J. Oliphant, Sol. Energy, 2021, 215, 179–188 CrossRef CAS .
40. W. H. M. Remmerswaal, B. T. van Gorkom, D. Zhang, M. M. Wienk and R. A. J. Janssen, Adv. Energy Mater., 2024, 14, 2303664 CrossRef CAS .
41. D. Lin, J. Fang, S. Li, Z. Zhan, H. Li, X. Wang, G. Xie, D. Wang, N. Huang, H. Peng, W. Xie, L. K. Ono, Y. Qi and L. Qiu, Adv. Sci., 2025, 12, 2407380 CrossRef CAS PubMed .
42. X. Liu, J. Xu, C. Zhao, F. Shen, P. Zou, Y. Wang, B. Ji, X. Guan, X. Pan and J. Yao, ACS Energy Lett., 2025, 2084–2092,  DOI:10.1021/acsenergylett.5c00305 .
43. J. Li, H. A. Dewi, H. Wang, J. Zhao, N. Tiwari, N. Yantara, T. Malinauskas, V. Getautis, T. J. Savenije, N. Mathews, S. Mhaisalkar and A. Bruno, Adv. Funct. Mater., 2021, 31, 2103252 CrossRef CAS .
44. Z. Yi, W. Wang, R. He, J. Zhu, W. Jiao, Y. Luo, Y. Xu, Y. Wang, Z. Zeng, K. Wei, J. Zhang, S.-W. Tsang, C. Chen, W. Tang and D. Zhao, Energy Environ. Sci., 2024, 17, 202–209 RSC .
45. J. Zhang, J. Wu, S. Langner, B. Zhao, Z. Xie, J. A. Hauch, H. A. Afify, A. Barabash, J. Luo, M. Sytnyk, W. Meng, K. Zhang, C. Liu, A. Osvet, N. Li, M. Halik, W. Heiss, Y. Zhao and C. J. Brabec, Adv. Funct. Mater., 2022, 32, 2207101 CrossRef CAS .
46. C. Xu, Z. Liu and E.-C. Lee, J. Mater. Chem. C, 2021, 9, 679–686 RSC .
47. T. Xue, B. Fan, K.-J. Jiang, Q. Guo, X. Hu, M. Su, E. Zhou and Y. Song, Energy Environ. Sci., 2024, 17, 2621–2630 RSC .
48. X. Chen, W. Cai, T. Niu, H. Wang, C. Liu, Z. Zhang, Y. Du, S. Wang, Y. Cao, P. Liu, W. Huang, C. Ma, B. Yang, S. Liu and K. Zhao, Energy Environ. Sci., 2024, 17, 6256–6267 RSC .
49. M. Chen, Z. Dai, N. Yan, Y. Cao, Y. Yuan, J. Zhang, D. Qi, L. Meng, S. Liu and J. Feng, J. Mater. Chem. C, 2024, 12, 10540–10547 RSC .
50. C. Kim, K. Kim, Y. Kim, N. Tsvetkov, N. J. Jeon, B. J. Kang and H. Min, Energy Environ. Sci., 2024, 17, 8582–8592 RSC .
51. C. Geng, K. Zhang, C. Wang, C. H. Wu, J. Jiang, F. Long, L. Han, Q. Han, Y.-B. Cheng and Y. Peng, Nano-Micro Lett., 2024, 17, 8 CrossRef PubMed .
52. R. Zhi, C.-Q. Yang, M. U. Rothmann, H.-Q. Du, Y. Jiang, Y.-Y. Xu, Z.-W. Yin, Y.-P. Mo, W. Dong, G. Liang, U. Bach, Y.-B. Cheng and W. Li, ACS Energy Lett., 2023, 8, 2620–2629 CrossRef CAS .
53. M. V. Khenkin, E. A. Katz, A. Abate, G. Bardizza, J. J. Berry, C. Brabec, F. Brunetti, V. Bulović, Q. Burlingame, A. Di Carlo, R. Cheacharoen, Y.-B. Cheng, A. Colsmann, S. Cros, K. Domanski, M. Dusza, C. J. Fell, S. R. Forrest, Y. Galagan, D. Di Girolamo, M. Grätzel, A. Hagfeldt, E. von Hauff, H. Hoppe, J. Kettle, H. Köbler, M. S. Leite, S. Liu, Y.-L. Loo, J. M. Luther, C.-Q. Ma, M. Madsen, M. Manceau, M. Matheron, M. McGehee, R. Meitzner, M. K. Nazeeruddin, A. F. Nogueira, Ç. Odabaşı, A. Osherov, N.-G. Park, M. O. Reese, F. De Rossi, M. Saliba, U. S. Schubert, H. J. Snaith, S. D. Stranks, W. Tress, P. A. Troshin, V. Turkovic, S. Veenstra, I. Visoly-Fisher, A. Walsh, T. Watson, H. Xie, R. Yıldırım, S. M. Zakeeruddin, K. Zhu and M. Lira-Cantu, Nat. Energy, 2020, 5, 35–49 CrossRef .

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d5ta03102h

This journal is © The Royal Society of Chemistry 2025