Hydrogen bond-mediated pseudo-halide complexation for stable and efficient perovskite precursors and solar cells

Abstract

The deprotonation of organic cations and oxidation of halide ions in perovskites are major degradation factors causing irreversible stability and efficiency loss in devices. To address these issues, we designed the 3-mercaptobenzoic acid (3-MBA) additive, which facilitates spontaneous deprotonation due to its carboxyl group and enables hydrogen bonding with formamidinium (FA⁺). Adding 3-MBA to the perovskite precursor solution inhibits both deprotonation of organic cations and oxidation of halide ions, thereby enhancing the stability of perovskite precursors and films at elevated temperatures. This approach also improves perovskite crystallinity and passivates halide-related defects through covalent bonding with uncoordinated lead. As a result, 3-MBA-treated inverted (p–i–n) solar cells achieve a power conversion efficiency (PCE) of 24.3%. Moreover, the unencapsulated 3-MBA-treated devices show impressive thermal stability with a T₉₈ lifetime after 1740 hours at 85 °C under nitrogen conditions. Additionally, 140-day-aged perovskite precursors containing 3-MBA retain over 96% of their initial efficiency.

---

Broader context

Fabricating solution-based perovskite solar cells (PSCs) with high efficiency and long-term stability is essential for their commercialization. Using MA-based cationic additives has proven to be one of the most effective methods for enhancing the crystallinity of perovskite materials, leading to cutting-edge performance. However, the deprotonation of these organic cations and the oxidation of halide ions lead to significant decomposition of perovskite materials. These degradations cause the unfixed tolerance factor and alter stoichiometry, resulting in poor stability of perovskite precursors and films. Herein, we developed a multifunctional additive that interacts with organic cations via strengthened hydrogen bonding complexation. We discovered that the strength of the interaction is crucial for preserving perovskite cations in both precursors and films at high temperatures. Additionally, the additive effectively inhibits halide oxidation, improves perovskite crystallinity, and passivates halide-related defects, resulting in significantly improved device performance and thermal stability. This study offers valuable guidelines for designing multifunctional additives to advance the efficiency and stability of perovskites involved with preserving ion components toward the commercialization of perovskite devices.

Introduction

Since their introduction in 2009,¹ metal halide perovskite solar cells (PSCs) have seen significant advances in the development of high-performing photovoltaic devices. The power conversion efficiency (PCE) of metal halide PSCs has dramatically risen, moving from an initial 3.8% to a certified efficiency of 26.1%.^2–4 To date, the predominant method for fabricating high-performance, low-defect-density metal halide perovskite films involves a solution-based process incorporating organic cations such as formamidinium (FA⁺) and methylammonium (MA⁺).^5–7 Additionally, methylammonium chloride (MACl) has recently been used as an additive in this process to promote the development of highly crystalline films and boost efficiency in both inverted (p–i–n) and regular (n–i–p) PSCs.^2,4,6 While the incorporation of MA⁺ components has significantly stabilized the perovskite phase and enhanced reproducibility,^5,8 it has been discovered that using MA⁺ cations poses substantial challenges for the prolonged stability of both the perovskite precursor and films. The low acid dissociation constant (pK_a) of MA⁺, which is prone to deprotonation,⁹ leads to the degradation of both the perovskite film and solution.^10–12 A series of reactions consume small cations (e.g., FA⁺ and MA⁺) and generate large cations (e.g., MFA⁺ and DMFA⁺), as illustrated in Fig. S1 (ESI†).^10,13 In addition to the deprotonation of organic cations, iodides (I⁻) in perovskite precursors easily oxidize to iodine (I₂) during aging.^14–16 These processes are considered major degradation factors, causing irreversible stability and efficiency loss in devices.¹⁷

Recent investigations have suggested that incorporating functional additives into the perovskite precursor can suppress both the irreversible dissipation of small cations (e.g., amine–cation reaction, decomposition and volatilization of deprotonated cations) and the oxidation of iodide.^16,18–20 This can be achieved by mitigating the deprotonation of organic cations and consuming I₂, resulting in a stabilized perovskite precursor. While these additive approaches are useful for controlling unfavorable reactions at the precursor solution stage, effectively preserving ion components such as organic cations and halide ions in perovskite precursors and further stabilizing them in perovskite films at elevated temperatures remain significant challenges. Long-term durability of perovskite devices under continuous illumination, which subjects them to high thermal stress,^21–23 is crucial for their commercialization. However, organic cations (e.g., FA⁺ and MA⁺) are easily deprotonated at high temperatures,²⁴ leading to their degrading decomposition and volatilization and resulting in poor device stability and performance.^16,25

Therefore, advanced additive design strategies that go beyond simply controlling unfavorable reactions are highly needed to stabilize perovskite compositions in both precursors and films at high temperatures. To achieve this, it is essential to thoroughly investigate the interactions involving key elements in organic cations. From this perspective, while hydrogen bonds are weaker than ionic or covalent bonds, they have been essential in stabilizing perovskite materials.^26–29 Recently, the incorporation of hydrogen bond-interactive additives through interfacial and defect engineering has emerged as a promising approach to address challenges like hydration and thermally induced deprotonation of perovskite compositions under air conditions.^30,31 However, this understanding is primarily based on solid states, especially surfaces exposed to air, and further studies are needed to expand and enhance the application of hydrogen bonds in both solid and solution phases. These studies should focus on modulating interaction strength and preserving perovskite components under various conditions. Taking these considerations, this study presents an all-in-one additive strategy designed to preserve perovskite components in perovskite precursors and films simultaneously. For this purpose, we designed the 3-mercaptobenzoic acid (3-MBA) additive, which facilitates spontaneous deprotonation due to its carboxyl group and enables strong hydrogen bonding interaction with FA⁺ through its electron-donating properties provided by a meta-positioned thiol (–SH) functional group in the aromatic molecule. With these benefits, we found that adding 3-MBA to the perovskite precursor solution inhibits both the deprotonation of organic cations and the oxidation of halide ions in perovskite precursors and films. Additionally, this approach improves perovskite crystallinity and passivates halide-related defects through covalent bonding with uncoordinated lead. As a result, 3-MBA-treated inverted (p–i–n) solar cells achieve a PCE of 24.3% with negligible hysteresis. Moreover, the unencapsulated devices exhibited remarkable long-term thermal stability under continuous heating at 85 °C in a nitrogen environment, maintaining 98.4% of their initial efficiency after 1740 hours. The ambient stability at 25 °C and a relative humidity (RH) of 35 ± 5% were maintained for 2240 hours with virtually no efficiency decay. In precursor aging results, the 3-MBA-modified PSCs made from 140-day-aged perovskite precursor solutions exhibit more than 96% of their initial PCE compared to those made from fresh perovskite precursor solutions.

Results and discussion

Characterization of perovskite precursors containing 3-MBA

In this study, we selected the triple-cation perovskite Cs_0.05(FA_0.85MA_0.15)_0.95Pb(I_0.85Br_0.15)₃, introducing a small amount of caesium to stabilize the FAPbI₃-based perovskite phase, as a precursor to understanding the interaction of organic cations. We introduced 3-MBA into the perovskite solution as the target additive molecule and chose benzoic acid (BA) for comparison,³² with their chemical structures shown in Fig. S2 (ESI†). After that, we discovered that hydrogen bonding interactions can occur between the deprotonated additives, deprotonated 3-MBA (D-3-MBA) and deprotonated BA (D-BA), and FA⁺ in perovskite precursors as shown in Fig. 1a. In addition, density functional theory (DFT) calculations revealed that selecting the meta-positioned regioisomer (3-MBA) was crucial for enhancing the hydrogen bonding strength compared to ortho- and para-regioisomers (2-MBA and 4-MBA), as described in Fig. S3 and Note S1 (ESI†).

Fig. 1 Schematic illustration and experimental investigations of hydrogen bonding-mediated complexation. (a) Hydrogen bonding complexation of D-3-MBA:FA⁺ and D-BA:FA⁺. (b) ¹H-NMR spectra of FAI and the FAI + 3-MBA mixture (DMSO-d₆ solvent). (c) Evolution of pH value depending on the PbI₂ incorporation in 3-MBA solution (DMF/DMSO 4 : 1 mixture solvent) and band area ratio of ν(COO⁻) to ring breathing peak (α peak). (d) Binding energy between additives and NH₂⁺ of FA⁺ and the corresponding average hydrogen bonding length. (e) FTIR spectra for FAI, 3-MBA, and mixed FAI and 3-MBA. (f) Schematic illustration of a perovskite precursor and crystal stabilization caused by two types of D-3-MBA-mediated complexation; hydrogen bonding is marked in sky blue, and covalent bonding is marked in brown.

Fig. 1b and Fig. S4 (ESI†) show proton nuclear magnetic resonance (¹H-NMR) spectra. The four proton peaks of the amine in FA⁺, which were originally at 8.81 ppm, split into two distinct peaks after the addition of 3-MBA and BA, indicating different chemical environments (H_a and H_b, respectively). These findings suggest the presence of hydrogen bonds between deprotonated additives and FA⁺ in certain complexes (Fig. 1a).¹¹ When additives containing a carboxyl group were added, similar splitting was observed in the perovskite precursor solutions (Fig. S5, ESI†). Additionally, we noted a shift of the C–H peak of FA⁺ (7.78 ppm) into a multiplet, further suggesting the potential for hydrogen bonding between the carboxylate and FA⁺.³³

DFT calculations, presented in Fig. S6 and S7 (ESI†), reveal that BA and 3-MBA are spontaneously deprotonated by dimethylformamide (DMF) with the aid of PbI₂. Consistently, upon contact with PbI₂, the pH levels significantly decreased without noticeable differences between BA and 3-MBA, as shown in Fig. 1c and Fig. S8 (ESI†). Using Raman spectroscopy, we confirmed the spontaneous deprotonation of the carboxyl group facilitated by PbI₂, as evidenced by an increase in the carboxylate stretching mode band area (at 1418 cm⁻¹) after adding PbI₂ to 3-MBA (Fig. 1c and Fig. S9, ESI†).³⁴ Importantly, DFT calculations reveal that D-3-MBA preferentially interacts with FA⁺ through hydrogen bonding rather than coordination bonding with PbI₂ (Fig. S10, S11, and Note S2 ESI†). These indicate that D-3-MBA can facilitate complex formation with FA⁺ through the hydrogen bonding in perovskite precursors containing PbI₂ (Fig. 1a).¹¹

The electron-donating –SH group on the aromatic ring typically triggers a push–pull intramolecular charge transfer, further promoting π-electron delocalization.^35,36 Consequently, DFT calculations have shown that this effect intensifies intramolecular charge separation in D-3-MBA, resulting in a higher dipole moment of 11.7 D compared to that of D-BA. This difference is attributable to the presence of the electron-donating –SH group, as shown in Fig. S2 (ESI†). Therefore, the oxygen in the carboxylate group of D-3-MBA, having a more negatively charged distribution, can induce strong hydrogen bond interactions with organic cations such as FA⁺ (Fig. S12, ESI†). The hydrogen bond length for the D-3-MBA to FA⁺ interaction was found to be 1.53 Å, which is shorter than 1.66 Å observed for the D-BA to FA⁺ interaction, indicating a stronger hydrogen bond strength in the D-3-MBA and FA⁺ complex (Fig. 1d). The bonding nature is further corroborated by the charge density difference (FA⁺ ↔ additives) and the crystal orbital overlap population analysis (H 1s of FA⁺ ↔ O 2p_x of additives) in Fig. S12 (ESI†).

Fourier transform infrared (FTIR) spectroscopy analyses were further conducted to assess the intensity of the hydrogen bonding (Fig. 1e and Fig. S13, ESI†). The stretching vibrational peak of NH₂⁺ in FA⁺ at 3335.3 cm⁻¹ and the C=N peak of FA⁺ at 1692.4 cm⁻¹ exhibited significant shifts of 9.7 cm⁻¹ and 2.1 cm⁻¹, respectively, compared to the shifts observed with BA, which were 0.9 cm⁻¹ and 1.1 cm⁻¹, respectively. This distinct trend suggests that the addition of D-3-MBA reinforces hydrogen bonding with FA⁺ more than D-BA does,⁶ as supported by theoretical calculations. This suggests that the hydrogen bonding complex that forms between D-3-MBA and FA⁺ can act as a stabilizer in the perovskite solution, enhancing the retention capacity of organic cations (Fig. 1f).

To investigate the stabilizing effects of D-3-MBA on perovskite precursors, we tracked the ¹H-NMR spectra of various precursor solutions (without additives, with BA, and with 3-MBA) over 55 days to understand the behavior of organic cations in the precursor solutions. Subsequently, we quantified the percentages of residual organic cations in the precursor solutions relative to fresh solutions by integrating the corresponding ¹H signal peaks of each cation (Fig. S14 and S15, Tables S1–S4, ESI†). The detailed calculation method for determining the percentage of each cation is provided in Note S3 (ESI†). For the pristine perovskite precursor solution, there was a noticeable decrease in smaller organic cations (FA⁺/MA⁺) and a significant increase in larger organic cations (MFA⁺/DMFA⁺) over the aging period (Fig. 2a). This observation aligns with the addition–elimination reactions depicted in Fig. S1 (ESI†), confirming the anticipated trends. Notably, the more rapid loss of MA⁺ compared to FA⁺ in the pristine precursor can be attributed to the inherently more volatile nature of MA⁰ compared to FA⁰.^2,10 In contrast, adding BA and 3-MBA has been shown to enhance the stability of perovskite precursor solutions, as demonstrated in Fig. 2b and c. Considering the crucial role of the proton in maintaining perovskite solution stability (see Note S4, ESI†), the similar pH ranges observed for the solutions with 3-MBA and BA after adding PbI₂ indicate comparable proton donation capabilities, as described in Fig. S8 (ESI†). Therefore, we can confirm that protons donated by additives effectively reduce the deprotonation of MA⁺, leading to a decrease in the excessive dissipation of MA⁺ compared with the pristine perovskite solution (Fig. 2a–c). On the other hand, the BA-added perovskite precursor solution showed a gradual reduction in MA⁺ concentration over time (Fig. 2b), whereas the 3-MBA-added perovskite precursor solution nearly preserved its original organic content under the same aging conditions (Fig. 2c). This indicates that while the hydrogen bonding strength between D-BA and FA⁺ is insufficient to effectively suppress the irreversible reaction, the enhanced hydrogen bonding in the D-3-MBA and FA⁺ complex successfully reduces the irreversible addition–elimination reactions involving both cationic FA⁺ and MA⁰. This helps maintain the original organic contents, which is clearly demonstrated in the long-term aged precursor after 140 days, as shown in Fig. S15 and Table S5 (ESI†). These results imply that in addition to proton donation, forming a complex through hydrogen bonding between the additive and FA⁺ is essential for stabilizing the perovskite precursor solution.

Fig. 2 Characterization of stabilized 3-MBA treated perovskite precursors and films. Evolution of organic cations of (a) control and (b) BA- and (c) 3-MBA- treated perovskite solutions over time. (d) Evolution of the PbI₂ peak ratio (up) and α-FAPbI₃ phase peak shift (down) obtained from XRD patterns for control, BA-, and 3-MBA-treated films during thermal aging at 150 °C in a nitrogen environment. (e) UV-vis absorption spectra of solutions of pure FAI (1 M) and FAI (1 M) + additional 1 mol% of 3-MBA with aging time. (f) UV-vis absorption spectra representing the evolution of iodine peak intensity while continuously adding 3-MBA (0.5 to 2.0 v/v%).

A recent study has shown that the thiol (–SH) functional group in additives can cause iodine (I₂) to reduce to iodide (I⁻), thereby aiding in the stabilization of organic cations.³⁷ By further evaluating the aging of a thiophenol-added perovskite precursor solution (Fig. S16, ESI†), we found that the stabilization of organic cations through hydrogen bonding complexation surpasses the effect of iodine reduction, as evidenced by comparisons between the results in Fig. 2c and Fig. S15 and Table S1–4 and 6 (ESI†). To verify the potential impact of hydrogen bonding interactions between organic cations and additives on the thermal stability of perovskite films,⁶ we conducted additional assessments of their thermal degradation at 150 °C under a nitrogen atmosphere (Fig. S17, ESI†). The thermal decomposition of the α-phase in perovskite films during thermal aging is primarily driven by the desorption of volatile organic cations, such as FA⁺ and MA⁺, and leads to the generation of Pb(Br_xI_1−x)₂.^31,38 In this regard, the 3-MBA-added film exhibits more robust characteristics than both the pristine and BA-added films during thermal aging (Fig. 2d, up and Fig. S17, ESI†).

Importantly, a distinctive trend in the α-phase (100) peak shift, related to the organic components in the perovskite films, is observed (Fig. 2d, down). Initially, the MA component primarily volatilizes, causing the peak to move toward an FA-dominant α-phase,^39,40 with this variation being consistent across both pristine and additive-added films. By 120 h, both pristine and BA-treated films exhibit a shift in the opposite direction, towards the Cs-dominant phase peak,⁴⁰ due to the prevalent volatilization of the FA organic component.³¹ Crucially, films treated with 3-MBA showed a less pronounced shift, marking a clear difference compared to both pristine and BA-added films. The reduced thermal degradation identified in 3-MBA-added films reinforces the possibility of improving the thermal stability of PSCs through 3-MBA modification. This improvement is attributed to the formation of stronger hydrogen bonds with FA⁺ cations and further advocates the effective preservation of organic cations within both the perovskite precursors and films as illustrated in Fig. 1f.

To understand the impact of precursor aging on the perovskite film, we tracked the X-ray diffraction (XRD) patterns (Fig. S18, ESI†). For perovskite films derived from the aged control precursor solution, the intensity of the α-phase (100) facet gradually diminished along with the distinct appearance of the δ-phase peak at 2θ ≈ 11.5° over 55 days of aging. However, the perovskite film prepared from the aged BA-added perovskite solution exhibits a somewhat mitigated decrease in the intensity of the α-phase accompanied by a slight increase in the δ-phase peak intensity. Notably, control solution-based films exhibited the α-phase uniquely shifted in the direction of the lower angle (Fig. S19, ESI†). The angle shift reflects lattice expansion within the perovskite structure due to the intervention of organic by-products (e.g., MFA⁺ and DMFA⁺). This is mainly due to the larger ionic radius of MFA⁺ (2.79 Å) compared to the dominant cation, FA⁺ (2.53 Å). Moreover, these by-products trigger a high tolerance factor above 1.0, which in turn suppresses the crystallization of the photoactive α-phase.¹⁸ Thereby, the notable degradation of the aged control solution-based films is attributed to the dissipation of small cations along with excessive accumulation of by-products. On the other hand, 3-MBA-added perovskite films, even after aging for 55 days, still exhibit a highly crystalline perovskite phase and show an unchanged lattice angle for the α-phase (100) peak. These trends are well matched with the changes in optical properties of ultraviolet-visible (UV-vis) absorption spectroscopy for control, BA, and 3-MBA treated perovskite films (Fig. S20, ESI†). The scanning electron microscopy (SEM) images show the top surfaces of perovskite films, shown in Fig. S21 (ESI†). Notably, after 55 days of aging, the control films show significant morphological degradation, characterized by large pinholes and merged grains. We attribute the formation of large pinholes to iodine gas generated by iodide oxidation, while the merged grains may be influenced by the formation of δ-FAPbI₃, primarily driven by MFA⁺ production.^10,19,41 This assumption is further supported by observations of the 3-MBA-treated perovskite film, which maintains its initial morphology, presumably due to the stabilizing effect of 3-MBA on the perovskite precursor.

Besides stabilizing organic cations, another pathway leading to the degradation of the perovskite precursor and film involves the oxygen or light-induced oxidation of I⁻, which forms I₂ impurities.^17,42 These result in non-stoichiometric and deep-defect centers that capture electrons.¹⁷ Meanwhile, a thiol can oxidize to disulfide via a redox reaction with I₂, resulting in a bond formation with another thiol, a process known as the thiol-disulfide exchange reaction (see Fig. S22, ESI†).⁴³ To substantiate the hypothesis, we performed Raman analysis of the 3-MBA film with and without injection of I₂. As shown in Fig. S23 (ESI†), after adding excess I₂, the S–H bending vibrational peak (913 cm⁻¹) and S–H stretching peaks (2543 and 2567 cm⁻¹) completely disappeared, while a distinct peak corresponding to the S–S disulfide stretching peak (542 cm⁻¹) emerged, providing a clear evidence of a thiol-disulfide oxidation reaction between 3-MBA and I₂.⁴⁴ After that, we investigated UV absorption spectra of solutions containing pure FAI, FAI with 3-MBA (Fig. 2e and Fig. S24 and Note S5, ESI†). In pure FAI solution, the I₂ characteristic peak at 365 nm gradually increased along with the solution turning pale yellow during aging, indicating that the I⁻ oxidation occurs immediately upon exposure to oxygen.¹⁵ In contrast, the FAI solution with 3-MBA shows negligible changes in its absorption spectra and maintains its transparent color over time. Furthermore, adding pure 3-MBA to the aged FAI solution gradually reduced the I₂ peak (Fig. 2f). This confirms that a thiol-disulfide exchange reaction can occur instantly between 3-MBA and I₂, allowing 3-MBA-treated precursors to retain their initial yellow color upon aging (Fig. S25, ESI†).⁴³ Moreover, perovskite films produced from 3-MBA-added precursor solutions nearly maintain their PL intensity after 55 days of aging (Fig. S26, ESI†). This observation further supports the idea that adding 3-MBA suppresses the generation of I₂ defect sites and retains its passivation ability, even in the presence of dimers (Fig. S27, ESI†). Overall, 3-MBA allows the stabilized precursor to preserve the initial stoichiometry of the perovskite composition, whereas perovskite films with BA exhibit a significant decline in PL intensity over the same aging period. This reduction in PL intensity for BA-added films could be due to the formation of numerous I₂ defect sites, serving as non-radiative charge recombination centers, a phenomenon observed alongside a few pinholes (Fig. S21, ESI†).

Film properties of 3-MBA-treated perovskite films

The effect of 3-MBA addition on perovskite film morphology was analysed using SEM images. As shown in Fig. 3a, the 3-MBA-modified sample exhibits a notably uniform and dense morphology, featuring fewer PbI₂ flakes and fewer grain boundaries compared to the control film. In addition, the average grain size in the perovskite film increased from 248.9 nm in the control to 309.9 nm in the 3-MBA modified samples (Fig. S28, ESI†). This improvement is further supported by corresponding cross-sectional SEM images (Fig. S28, ESI†), which notably show aligned grains without distinct grain boundaries in the vertical direction, a feature that is advantageous for promoting effective charge-carrier transport. The XRD patterns of the 3-MBA-added film also exhibit improved intensity of the α-phase (100) at 14.2° and significantly reduced PbI₂ peaks at 12.6°, compared to those of the control and BA-added films (Fig. 3c). The suppression of residual PbI₂ can be attributed to the slower crystallization rate induced by 3-MBA, allowing the grain crystals to more effectively consume adjacent PbI₂ during maturation.¹¹ We identified the formation of an intermediate phase between 3-MBA and PbI₂ (Fig. S29, ESI†). In contrast, BA does not form a PbI₂-mediated intermediate phase due to its weaker coordination bonding. This suggests that the strong coordination binding between 3-MBA and PbI₂ facilitates perovskite growth.^45–47 X-ray photoelectron spectroscopy (XPS) analysis further revealed that 3-MBA strongly interacts with uncoordinated Pb²⁺. For the Pb 4f spectra of the control perovskite film, the binding energies are identified at 142.68 eV for Pb 4f_5/2 and 137.81 eV for Pb 4f_7/2 (Fig. 3d and Fig. S30, ESI†). A significant shift is observed in the 3-MBA-treated perovskite film, with the Pb characteristic peaks shifting down to 142.48 eV (Δ = 0.2 eV) for Pb 4f_5/2 and 137.62 eV (Δ = 0.19 eV) for Pb 4f_7/2, marking a notable shift compared to the BA-treated perovskite film.⁴⁸ This indicates an electron-rich environment in the PbI₆⁴⁻ octahedra due to strong coordination with D-3-MBA, which is further evidenced by the FTIR spectra (Fig. S31, ESI†). The distinct shift upon adding 3-MBA indicates possible coordination bonding with PbI₂.¹⁸ Atomic force microscopy analysis was performed to assess the surface morphology and properties of the perovskite film. With the addition of 3-MBA, the average surface roughness decreased from 17.3 nm to 14.6 nm (Fig. S32, ESI†), indicating that the surface of the 3-MBA-treated perovskite film became more uniform and homogeneous, as further supported by top-view SEM images (Fig. 3a). The Kelvin probe force microscopy (KPFM) results indicate that 3-MBA causes a substantial increase in surface potential compared to the control, suggesting a Fermi level (E_F) closer to the vacuum level (Fig. 3e, Fig. S33 and Note S6, ESI†). This observation is consistent with the ultraviolet photoelectron spectroscopy (UPS) findings reported in Fig. S34 (ESI†), with the resulting band diagram shown in Fig. S35 (ESI†). Considering that this shift can result in the beneficial charge extraction ascribed to the favorable band alignment, it is evident that adding 3-MBA effectively enhances interfacial charge transport and improves contact through a smoother surface.^49,50

Fig. 3 Film properties for the modified perovskite films. (a) and (b) Top-view SEM images of control and 3-MBA-modified perovskite films. (c) XRD patterns of control and BA- and 3-MBA-treated films. (d) XPS spectra for Pb 4f of control and BA- and 3-MBA-treated films. (e) Statistical surface potential distributions obtained via KPFM images. (f) Absolute-steady state PL spectra and the corresponding PLQY (inset). (g) Optimized geometry of the D-3-MBA absorbed PbI₂-terminated (100) surface with iodide vacancy (left) and the FAI-terminated (100) surface (right). (h) Binding energy of different passivation agents anchored on the FAI-terminated (100) perovskite surface. (i) Calculated density of states (DOS) of the iodide vacancy-containing PbI₂-terminated surface with (red) and with D-3-MBA (green).

Fig. 3e also demonstrates that the surface potential in the 3-MBA-treated film exhibits a uniform distribution, contrasting with the inhomogeneous surface potential observed in the control film. This suggests a reduction in defect-induced charge recombination following 3-MBA treatment. To examine the passivation effect of 3-MBA on defect sites, we conducted photoluminescence quantum yield (PLQY) measurements and time-resolved photoluminescence (TRPL) analysis (Fig. 3f and Fig. S36, S37 and Note S7, ESI†). The perovskite film treated with 3-MBA exhibits a significantly enhanced PL intensity, with a notable improvement in PLQY (4.3%), surpassing both the control film (2.7%) and the BA-treated film (3.2%).⁴⁷ This result also supports the notion of defect passivation by 3-MBA, corroborating the findings related to the surface potential distribution shown in Fig. 3e. Furthermore, TRPL decay curves show that 3-MBA-treated films have an increased carrier lifetime, with lifetimes of 1.78 μs for BA-treated and 1.83 μs for 3-MBA-treated films compared to 1.54 μs for the control. The enhanced passivation ability of 3-MBA can be attributed to the strong coordination of D-3-MBA, particularly its carboxylate group, which is capable of passivating positively charged defects, primarily iodide vacancies, predominantly found at the grain boundaries and surfaces of the perovskite films.^32,51

In this perspective, using DFT calculations, we further evaluated the passivation effect of D-3-MBA by positing the presence of an iodide vacancy defect on the perovskite surface (Fig. S38, ESI†). We found that the D-3-MBA ligand adsorbed on the iodide vacancy exhibited strong Pb–O coordination bonding and weak hydrogen bonding between the –SH group and I⁻, indicating beneficial passivation of iodide vacancies (Fig. 3g, left). The presence of the hydrogen bonding is further confirmed by notable shifts in the I and –SH characteristic peaks in the XPS and FTIR spectra, respectively (Fig. S30 and S31, ESI†).

As depicted in Fig. 3i, D-3-MBA induced significant electron redistribution between the iodide vacancy and the incorporated D-3-MBA, forming coordinated bonds and thereby effectively passivating the iodide vacancy, which indicates the presence of shallow-level trap states near the conduction band minimum reported as non-radiative charge recombination centers.⁵¹ Furthermore, D-3-MBA exhibited a higher binding affinity on the FAI-terminated (100) surface (Fig. 3g, right) compared to D-BA (Fig. 3h and Fig. S39, ESI†). This is attributed to the dual functionality of the additional –SH group, which acts as both an electron-donating and anchoring group, enabling favorable interaction in the form of D-3-MBA:FA⁺. This demonstrates that D-3-MBA can form two types of complexes, including covalent bonding with uncoordinated Pb²⁺ and hydrogen bonding with FA⁺, effectively passivating iodide vacancies and organic cations (Fig. 1f). Notably, the primary coordination site of D-3-MBA is the carboxylate group, which passivates iodide vacancies. Similarly, thiophenol, with its thiol functional group, also coordinates with uncoordinated Pb²⁺, effectively passivating defect states, as shown in Fig. S40 (ESI†). However, further investigation into the effects of thiophenol on perovskite films revealed that its treatment leaves a significant amount of unreacted PbI₂ in bulk films (Fig. S40b),⁵² which accelerates metal-mediated degradation under light exposure. Further details on the effects of thiophenol on perovskite films are provided in Note S8 (ESI†). These findings suggest that while thiophenol offers some stabilization benefits, it also has detrimental effects on perovskite stability. In contrast, 3-MBA emerges as a more ideal additive, providing effective stabilization for both perovskite precursors and films.

Photovoltaic performance

To demonstrate the aforementioned benefits of 3-MBA in PSCs, we fabricated inverted PSCs using the device structure FTO/NiO_x/N719/Cs_0.05(FA_0.85MA_0.15)_0.95Pb(I_0.85Br_0.15)₃/PEAI/PC₆₁BM/BCP/Ag. The current density–voltage (J–V) curves and performance parameters of these devices are presented in Fig. S41 and S42 (ESI†). The untreated control device exhibited a PCE of 20.43%, while the 3-MBA-treated device (target) showed impressive progress in PCE, reaching 22.48%, which is attributed to significant improvements in open-circuit voltage (V_oc) and fill factor (FF) (Fig. S42 and Table S8, ESI†). Additionally, the target PSCs exhibit a notably improved hysteresis index (HI) of 0.9%, compared with the control, which has an HI of 2.0% (Fig. S41, ESI†). We also tracked the steady-state PCE for both control and target devices under AM 1.5G illumination (Fig. S43, ESI†). The corresponding values for the control and target devices are 19.66% and 21.43%, respectively, aligning well with the PCEs obtained from the J–V curves. Furthermore, the integrated photocurrent density for the target PSC, calculated as 23.00 mA cm⁻² from the external quantum efficiency (EQE) spectra, corroborates the J–V measurement results (Fig. S43, ESI†). To demonstrate the universality of the effect on 3-MBA, we further fabricated inverted PSCs using a Cs_0.05FA_0.92MA_0.03PbI_2.97Br_0.03 perovskite composition, exhibiting a narrower bandgap of 1.54 eV (Fig. S44, ESI†), employing the same device configuration as previously described in Fig. S41 (ESI†). The target device shows an increase in V_oc from 1.10 V to 1.151 V and in FF from 80.8% to 82.5%, compared to the control device. This results in a PCE of 24.31% for the target device with negligible hysteresis, compared to a PCE of 22.61% for the control device (Fig. 4a, Fig. S45 and Table S9, ESI†). The target device also maintains a steady-state PCE of 24.0% and an integrated J_sc of 24.95 mA cm⁻² from the EQE spectra (Fig. 4b and c), aligning well with the PCE and short-circuit current density (J_sc) values obtained from the J–V curve. To gain further insight into the mechanisms behind the improvement of photovoltaic parameters in PSCs, we measured the ideality factor of the PSCs (Fig. 4d and Note S9, ESI†). The ideality factor (n) of 1.20, closer to 1 in the target device compared to 1.64 in the control device, indicates that 3-MBA treatment significantly suppresses trap-assisted non-radiative recombination in the device. The built-in potential (V_bi) of PSCs was further estimated through Mott–Schottky measurements (Fig. 4e). An increased V_bi for the target device (1.02 V) indicates an accelerated driving force for charge carrier separation and transport, aligning well with the enhanced V_oc. Fig. 4f and Fig. S46 (ESI†) show that the target PSC exhibits a longer average transient photovoltage (TPV) decay lifetime and a shorter average transient photocurrent decay lifetime compared to the control PSC (Note S10, ESI†). These findings further underscore the beneficial impact of 3-MBA on trap passivation and improved charge transport. In summary, the notable improvement in photovoltaic performance of devices modified with 3-MBA can be attributed to chemical passivation that mitigates detrimental recombination in both the bulk and interface, improvement of film quality along the longitudinal direction, and facilitation of charge extraction. The long-term stability of both unencapsulated control and target PSCs was monitored under ambient conditions at a temperature of 25 °C and a RH of 35 ± 5%, following the ISOS-D1 protocol, where ISOS represents the International Summit on Organic PV Stability.

Fig. 4 Photovoltaic characteristics and stability of the control and target devices. (a) Current density–voltage (J–V) curves of the control and 3-MBA modified (target) PSCs. (b) EQE curves and integrated J_sc of the target PSCs. (c) The steady state of PCE for the target devices. (d) V_ocversus light intensity plots. (e) Mott–Schottky plots of the control and target devices with a frequency at 10 kHz. (f) TPV curves of control and target PSCs. (g) Long-term stability test when stored at 85 °C in a nitrogen environment; we used 6–16 individual unencapsulated devices for the stability test. Initial average PCEs are 18.0% and 18.5% for control and target devices, respectively. All data are presented as average values ± standard deviation. (h) Variation of average PCE statistics for PSC aged precursors (∼140 days) with and without 3-MBA. All data are presented as average values ± standard deviation.

We employed devices with the configuration ITO/MeO-2PACz/Perovskite/C₆₀/SnO₂/Ag, where SnO₂ was applied as a buffer layer using atomic layer deposition to impede ion migration to adjacent layers such as the ETL and electrode (e.g., Ag),^53,54 and the perovskite composition was Cs_0.05(FA_0.85MA_0.15)_0.95Pb(I_0.85Br_0.15)₃. We confirmed that the 3-MBA modification strategy is also effective in this device configuration (Fig. S47 and Table S10, ESI†). Fig. S48 (ESI†) demonstrates that the target PSCs maintained their initial average efficiency of 19.31% without notable decay after 2240 h. In contrast, the control PSCs declined to 87.5% of their initial average PCE of 18.60% after 1740 h. The superior stability of the 3-MBA-treated PSCs can be attributed to the regulated crystallinity of the perovskite film, suppressed PbI₂ residue, and effective defect passivation, particularly of iodide vacancies. The hydrophobicity of the perovskite film following 3-MBA treatment also contributes to its resistance to moisture invasion,⁵⁵ as shown in Fig. S49 (ESI†). This corresponds with the results of the storage stability test conducted under harsh humid conditions (ambient temperature of 25 °C and RH of 85 ± 5%), detailed in Fig. S50 (ESI†). Operational stability of the unencapsulated devices was also assessed under continuous 1-sun equivalent white LED light illumination at maximum power point tracking in humid air (RH of 70 ± 5%) (Fig. S51, ESI†). The control device showed an efficiency of 76.9% of the initial value along with an initial burn-in effect over 42 h. In contrast, the target device showed a 92.7% initial efficiency after 180 h of testing, demonstrating improved light stability.

The thermal stability test was further assessed at 85 °C in a nitrogen environment using the ISOS-D-2I@85 °C protocol (Fig. 4g). The target devices modified by 3-MBA maintained 98.4% of their initial average efficiency after 1740 h, outperforming the retention rate of 66.6% for the control devices. Moreover, we also conducted the ISOS-D2I test under humid air (85 °C and RH of 70 ± 5%) (Fig. S52, ESI†). The control devices drastically degraded to 76.7% of their initial average efficiency within 72 h. In contrast, the target devices showed 93.5% of their initial value after 96 h.

Lastly, we evaluated the PCE distribution of the PSCs, both with and without 3-MBA, depending on the storage time of the perovskite precursor solution (Fig. 4h). PSCs modified with 3-MBA maintain over 96% of their initial efficiency even after a storage time of 140 days. In contrast, a drastic decrease in efficiency is observed in the control PSCs, primarily attributed to decreased J_sc and FF (Fig. S53, ESI†). This is due to the presence of a bright yellow phase and large cavities in the films, as observed in Fig. S21 and S53 (ESI†). In conclusion, the improvements brought about by 3-MBA modification signify significant progress in integrated stabilization systems for all solution-based processes, impacting both perovskite films and precursors. This study suggests that an additive modification approach emphasizing the formation of hydrogen bonds between the additive and organic cations could be a promising route to commercializing PSCs, by providing outstanding photovoltaic performance and sustainable precursor stability.

Conclusions

In this study, we presented an integrated stabilizing approach to manage both perovskite precursors and films through additive modification by deprotonated anion complexation. To achieve this, we designed the 3-MBA additive, which facilitates spontaneous deprotonation due to its carboxyl group and enables strong hydrogen bonding interaction with FA⁺ through its electron-donating properties provided by a meta-positioned thiol functional group in the aromatic molecule. This unique feature of 3-MBA inhibits both the deprotonation of organic cations and the oxidation of halide ions after the addition of perovskite precursors. In addition to stabilizing perovskite precursors, the strong coordination bonding between D-3-MBA and uncoordinated lead improves film crystallinity, effectively reduces halide-related defects, and enhances charge transfer. As a result, inverted PSCs achieved a remarkable PCE of 24.31% along with impressive thermal and precursor aging stability. We believe that this study provides valuable guidelines for designing multifunctional additives to enhance the performance and stability of both perovskite films and precursors, offering a promising path for commercialization.

Author contributions

All authors discussed the results and commented on the manuscript. T. Y. conceived the research idea, designed most of the experiments, and wrote the manuscript. J. C. supervised the overall research and manuscript writing. B. J. M. performed DFT calculations and supported the manuscript description. S. C., S. -K. K., and J. L. contributed to device fabrication. S. C., S. -K. K., S. H., G. S., H. J. K., J. Y. P., H. N. Y., H. R. Y., and E. J. L. were involved in all the experimental parts. Y. L. and D. -H. N. carried out Raman analysis. G. L. and S. J. L. carried out absolute photoluminescence quantum yield analysis. W. L., S. K., S. Y., and J. L. carried out time-resolved photoluminescence and transient photovoltage/photocurrent decay analysis. D. -H. K. and Y. K. contributed to research supervision.

Data availability

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

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This research was supported by the National Research Council of Science & Technology (NST) grant by the Korean government (MSIT) (No. CAP 23071-300) and the DGIST R&D Program of the Ministry of Science and ICT (21-CoE-ET-01). This work was also financially supported by the Korea Institute of Science and Technology (KIST) Institutional Program (2E33323).

References

1. A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am. Chem. Soc., 2009, 131, 6050–6051.
2. J. Park, J. Kim, H.-S. Yun, M. J. Paik, E. Noh, H. J. Mun, M. G. Kim, T. J. Shin and S. I. Seok, Nature, 2023, 616, 724–730.
3. Z. Huang, Y. Bai, X. Huang, J. Li, Y. Wu, Y. Chen, K. Li, X. Niu, N. Li, G. Liu, Y. Zhang, H. Zai, Q. Chen, T. Lei, L. Wang and H. Zhou, Nature, 2023, 623, 531–537.
4. 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.
5. M. Kim, G.-H. Kim, T. K. Lee, I. W. Choi, H. W. Choi, Y. Jo, Y. J. Yoon, J. W. Kim, J. Lee, D. Huh, H. Lee, S. K. Kwak, J. Y. Kim and D. S. Kim, Joule, 2019, 3, 2179–2192.
6. L. E. Lehner, S. Demchyshyn, K. Frank, A. Minenkov, D. J. Kubicki, H. Sun, B. Hailegnaw, C. Putz, F. Mayr, M. Cobet, G. Hesser, W. Schöfberger, N. S. Sariciftci, M. C. Scharber, B. Nickel and M. Kaltenbrunner, Adv. Mater., 2023, 35, 2208061.
7. K. Odysseas Kosmatos, L. Theofylaktos, E. Giannakaki, D. Deligiannis, M. Konstantakou and T. Stergiopoulos, Energy Environ. Mater., 2019, 2, 79–92.
8. Y. Zhang, G. Grancini, Y. Feng, A. M. Asiri and M. K. Nazeeruddin, ACS Energy Lett., 2017, 2, 802–806.
9. M. Wang, Z. Shi, C. Fei, Z. J. D. Deng, G. Yang, S. P. Dunfield, D. P. Fenning and J. Huang, Nat. Energy, 2023, 8, 1229–1239.
10. X. Wang, Y. Fan, L. Wang, C. Chen, Z. Li, R. Liu, H. Meng, Z. Shao, X. Du, H. Zhang, G. Cui and S. Pang, Chem, 2020, 6, 1369–1378.
11. F. Li, X. Deng, Z. Shi, S. Wu, Z. Zeng, D. Wang, Y. Li, F. Qi, Z. Zhang, Z. Yang, S.-H. Jang, F. R. Lin, S. W. Tsang, X.-K. Chen and A. K. Y. Jen, Nat. Photonics, 2023, 17, 478–484.
12. Q. Dong, W. Shang, X. Yu, Y. Yin, C. Jiang, Y. Feng, J. Bian, B. Song, S. Jin, Y. Zhou, L. Wang and Y. Shi, ACS Energy Lett., 2022, 7, 481–489.
13. T. Zhu, D. Zheng, M.-N. Rager and T. Pauporté, Sol. RRL, 2020, 4, 2000348.
14. N. Wu, T. Yang, Z. Wang, Y. Wu, Y. Wang, C. Ma, H. Li, Y. Du, D. Zhao, S. Wang, P. Liu, W. Huang, X. Ren, S. Liu and K. Zhao, Adv. Mater., 2023, 35, 2304809.
15. S. Chen, X. Xiao, H. Gu and J. Huang, Sci. Adv., 2021, 7, eabe8130.
16. H. Meng, K. Mao, F. Cai, K. Zhang, S. Yuan, T. Li, F. Cao, Z. Su, Z. Zhu, X. Feng, W. Peng, J. Xu, Y. Gao, W. Chen, C. Xiao, X. Wu, M. D. McGehee and J. Xu, Nat. Energy, 2024, 9, 536–547.
17. Z. Wang, Q. Tian, H. Zhang, H. Xie, Y. Du, L. Liu, X. Feng, A. Najar, X. Ren and S. Liu, Angew. Chem., Int. Ed., 2023, 62, e202305815.
18. M. Li, H. Li, Q. Zhuang, D. He, B. Liu, C. Chen, B. Zhang, T. Pauporté, Z. Zang and J. Chen, Angew. Chem., Int. Ed., 2022, 61, e202206914.
19. Z. Li, Z. Xing, H. Peng, X. Meng, D. Li, L. Huang, X. Hu, T. Hu and Y. Chen, Adv. Energy Mater., 2022, 12, 2200650.
20. C. Tian, T. Wu, Y. Zhao, X. Zhou, B. Li, X. Han, K. Li, C. Hou, Y. Li, H. Wang and Q. Zhang, Adv. Energy Mater., 2024, 14, 2303666.
21. Y. Yang, S. Cheng, X. Zhu, S. Li, Z. Zheng, K. Zhao, L. Ji, R. Li, Y. Liu, C. Liu, Q. Lin, N. Yan and Z. Wang, Nat. Energy, 2024, 9, 37–46.
22. R. Azmi, D. S. Utomo, B. Vishal, S. Zhumagali, P. Dally, A. M. Risqi, A. Prasetio, E. Ugur, F. Cao, I. F. Imran, A. A. Said, A. R. Pininti, A. S. Subbiah, E. Aydin, C. Xiao, S. I. Seok and S. De Wolf, Nature, 2024, 628, 93–98.
23. J. Suo, B. Yang, E. Mosconi, D. Bogachuk, T. A. S. Doherty, K. Frohna, D. J. Kubicki, F. Fu, Y. Kim, O. Er-Raji, T. Zhang, L. Baldinelli, L. Wagner, A. N. Tiwari, F. Gao, A. Hinsch, S. D. Stranks, F. De Angelis and A. Hagfeldt, Nat. Energy, 2024, 9, 172–183.
24. J. Xu, A. Maxwell, Z. Song, A. S. R. Bati, H. Chen, C. Li, S. M. Park, Y. Yan, B. Chen and E. H. Sargent, Nat. Commun., 2024, 15, 2035.
25. K. Liu, S. Rafique, S. F. Musolino, Z. Cai, F. Liu, X. Li, Y. Yuan, Q. Bao, Y. Yang, J. Chu, X. Peng, C. Nie, W. Yuan, S. Zhang, J. Wang, Y. Pan, H. Zhang, X. Cai, Z. Shi, C. Li, H. Wang, L. Deng, T. Hu, Y. Wang, Y. Wang, S. Chen, L. Shi, P. Ayala, J. E. Wulff, A. Yu and Y. Zhan, Joule, 2023, 7, 1033–1050.
26. K. L. Svane, A. C. Forse, C. P. Grey, G. Kieslich, A. K. Cheetham, A. Walsh and K. T. Butler, J. Phys. Chem. Lett., 2017, 8, 6154–6159.
27. A. Binek, F. C. Hanusch, P. Docampo and T. Bein, J. Phys. Chem. Lett., 2015, 6, 1249–1253.
28. M. Zhu, C. Li, B. Li, J. Zhang, Y. Sun, W. Guo, Z. Zhou, S. Pang and Y. Yan, Mater. Horiz., 2020, 7, 2208–2236.
29. W. Luo, G. AlSabeh and J. V. Milić, in Photochemistry: Volume 50, ed. S. Crespi and S. Protti, The Royal Society of Chemistry, 2022, vol. 50, pp. 346–370.
30. S. Du, H. Huang, Z. Lan, P. Cui, L. Li, M. Wang, S. Qu, L. Yan, C. Sun, Y. Yang, X. Wang and M. Li, Nat. Commun., 2024, 15, 5223.
31. J. Xu, A. Maxwell, Z. Song, A. S. R. Bati, H. Chen, C. Li, S. M. Park, Y. Yan, B. Chen and E. H. Sargent, Nat. Commun., 2024, 15, 2035.
32. Q. Wang, Y. Chen, X. Chen, W. Tang, W. Qiu, X. Xu, Y. Wu and Q. Peng, Adv. Mater., 2024, 36, 2307709.
33. W. T. M. Van Gompel, R. Herckens, G. Reekmans, B. Ruttens, J. D’Haen, P. Adriaensens, L. Lutsen and D. Vanderzande, J. Phys. Chem. C, 2018, 122, 4117–4124.
34. D. Shin, H. Seo and W. Jhe, ACS Cent. Sci., 2020, 6, 2079–2087.
35. J. P. Caradonna, E. W. Harlan and R. H. Holm, J. Am. Chem. Soc., 1986, 108, 7856–7858.
36. J. Liu, J. Chen, L. Xie, S. Yang, Y. Meng, M. Li, C. Xiao, J. Zhu, H. Do, J. Zhang, M. Yang and Z. Ge, Angew. Chem., Int. Ed., 2024, 63, e202403610.
37. W. Hu, J. Yang, C. Yang, X. Xiao, C. Wang, Z. Cui, Q. Gao, J. Qi, M. Xia, Y. Su, A. Mei and H. Han, J. Energy Chem., 2023, 90, 32–39.
38. C. Shi, Q. Song, H. Wang, S. Ma, C. Wang, X. Zhang, J. Dou, T. Song, P. Chen, H. Zhou, Y. Chen, C. Zhu, Y. Bai and Q. Chen, Adv. Funct. Mater., 2022, 32, 2201193.
39. Y. Zhang, S.-G. Kim, D. Lee, H. Shin and N.-G. Park, Energy Environ. Sci., 2019, 12, 308–321.
40. A. Nur'aini, S. Lee and I. Oh, J. Electrochem. Sci. Technol., 2022, 13, 71–77.
41. L. Chen, M. Hu, S. Lee, J. Kim, Z.-Y. Zhao, S.-P. Han, M. S. Lah and S. I. Seok, J. Am. Chem. Soc., 2023, 145, 27900–27910.
42. D. G. Lee, D. H. Kim, J. M. Lee, B. J. Kim, J. Y. Kim, S. S. Shin and H. S. Jung, Adv. Funct. Mater., 2021, 31, 2006718.
43. G. J. Corban, C. D. Antoniadis, S. K. Hadjikakou, N. Kourkoumelis, V. Y. Tyurin, A. Dolgano, E. R. Milaeva, M. Kubicki, P. V. Bernhardt, E. R. T. Tiekink, S. Skoulika and N. Hadjiliadis, Heteroat. Chem., 2012, 23, 498–511.
44. P. Bazylewski, R. Divigalpitiya and G. Fanchini, RSC Adv., 2017, 7, 2964–2970.
45. F. Qiu, H. Cheng, P. Mao, W. Bi, S. Xing, B. Wang and Y. Zhong, ACS Energy Lett., 2024, 9, 1115–1124.
46. Y. Huang, T. Liu, B. Wang, J. Li, D. Li, G. Wang, Q. Lian, A. Amini, S. Chen, C. Cheng and G. Xing, Adv. Mater., 2021, 33, 2102816.
47. R. Chen, J. Wang, Z. Liu, F. Ren, S. Liu, J. Zhou, H. Wang, X. Meng, Z. Zhang, X. Guan, W. Liang, P. A. Troshin, Y. Qi, L. Han and W. Chen, Nat. Energy, 2023, 8, 839–849.
48. Z. Huang, Z. Ma, C. Deng, T. Yu, G. Li, Z. Du, W. You, J. Yang, Y. Chen, Y. Li, S. Hou, Q. Yang, Q. Zhang, H. Du, Y. Li, H. Shu, Q. Liu, C. Peng, Y. Huang, J. Yu, Y. Lin, K. Sun and W. Long, Adv. Energy Mater., 2023, 2, 2302769.
49. Y. Duan, J. Wang, D. Xu, P. Ji, H. Zhou, Y. Li, S. Yang, Z. Xie, X. Hai, X. Lei, R. Sun, Z. Fan, K. Zhang, S. Liu and Z. Liu, Adv. Funct. Mater., 2024, 34, 2312638.
50. C. Deng, J. Wu, Y. Yang, Y. Du, R. Li, Q. Chen, Y. Xu, W. Sun, Z. Lan and P. Gao, ACS Energy Lett., 2023, 8, 666–676.
51. J. Xu, H. Chen, L. Grater, C. Liu, Y. Yang, S. Teale, A. Maxwell, S. Mahesh, H. Wan, Y. Chang, B. Chen, B. Rehl, S. M. Park, M. G. Kanatzidis and E. H. Sargent, Nat. Mater., 2023, 22, 1507–1514.
52. B. P. Finkenauer, Y. Zhang, K. Ma, J. W. Turnley, J. Schulz, M. Gómez, A. H. Coffey, D. Sun, J. Sun, R. Agrawal, L. Huang and L. Dou, J. Phys. Chem. C, 2023, 127, 930–938.
53. S. M. Park, M. Wei, J. Xu, H. R. Atapattu, F. T. Eickemeyer, K. Darabi, L. Grater, Y. Yang, C. Liu, S. Teale, B. Chen, H. Chen, T. Wang, L. Zeng, A. Maxwell, Z. Wang, K. R. Rao, Z. Cai, S. M. Zakeeruddin, J. T. Pham, C. M. Risko, A. Amassian, M. G. Kanatzidis, K. R. Graham, M. Grätzel and E. H. Sargent, Science, 2023, 381, 209–215.
54. S. M. Park, M. Wei, N. Lempesis, W. Yu, T. Hossain, L. Agosta, V. Carnevali, H. R. Atapattu, P. Serles, F. T. Eickemeyer, H. Shin, M. Vafaie, D. Choi, K. Darabi, E. D. Jung, Y. Yang, D. B. Kim, S. M. Zakeeruddin, B. Chen, A. Amassian, T. Filleter, M. G. Kanatzidis, K. R. Graham, L. Xiao, U. Rothlisberger, M. Grätzel and E. H. Sargent, Nature, 2023, 624, 289–294.
55. L. Yang, Y. Xiao, G. Han, Y. Chang, Y. Zhang, W. Hou, J.-Y. Lin and H. Li, Org. Electron., 2020, 81, 105681.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4ee02793k

---