Modulation of nucleation and crystallization in PbI₂ films promoting preferential perovskite orientation growth for efficient solar cells

Abstract

Formamidinium (FA)-rich lead iodide perovskite solar cells (PSCs) are gaining popularity because of their excellent photovoltaic performance. Nevertheless, the FA-rich perovskites processed by a two-step sequential deposition method usually possess many non-radiative recombination sites, undesired crystal orientations, and too much excess lead iodide (PbI₂) that can seriously deteriorate device performance. Here, we modulate the nucleation and crystallization of the PbI₂ films by introducing pentafluoroanilinium trifluoromethanesulfonate (PFAT) into PbI₂ precursor solutions. Such modulatory PbI₂ films as porous templates allow the growth of perovskites with a preferential orientation, enlarged grains, reduced defects, and decreased PbI₂ residues, owing to the reduced Gibbs free energy of PbI₂ and the acceleration of perovskite formation. Ultimately, we derive FA-rich lead iodide PSCs with an impressive power conversion efficiency of 24.52%, which is among the highest values of two-step sequential deposition-processed PSCs. Furthermore, the unencapsulated device maintains ∼90% of its initial efficiency after 500 hours of maximum power point loading under constant light irradiation at ∼55 °C in an N₂ atmosphere. This study presents an effective way to improve the crystalline quality of perovskites by synchronously pre-modulating the morphology, nucleation, and crystallization of PbI₂ precursor films.

---

Broader context

In recent years, formamidinium (FA)-rich lead iodide perovskite solar cells (PSCs) are of high interest because of their excellent photovoltaic performance. However, the FA-rich perovskites processed by a two-step sequential deposition method usually possess many non-radiative recombination sites, undesired crystal orientation, and too much excess lead iodide (PbI₂) that can deteriorate device performance. Given the heavy dependence of the two-step method on PbI₂ films, the nucleation and crystallization of PbI₂ films can affect the formation of subsequent perovskite phases and the orientation growth of perovskite crystals. Here, we modulate the nucleation and crystallization of the PbI₂ films by using pentafluoroanilinium trifluoromethanesulfonate (PFAT) additive. The incorporation of PFAT allows the formation of large PbI₂ clusters and enhances the crystallinity of PbI₂, resulting in the reduced Gibbs free energy of PbI₂. Then the secondary growth of the large PbI₂ clusters and the early formation of the perovskite phases are promoted. Such modulatory PbI₂ films enable us to obtain preferentially oriented perovskite films, harvesting high PCEs and enhanced stability. This study presents a key perspective on the further preferential perovskite growth from the nucleation and crystallization of PbI₂ itself in a two-step approach.

Introduction

In recent years, hybrid organic–inorganic metal halide perovskite solar cells (PSCs) have arrived at a staggering power conversion efficiency (PCE) of 25.7%,¹ symbolizing a prosperous outlook of the perovskite photovoltaic field. Specifically, formamidinium-rich lead iodide (FAPbI₃) is of interest to light absorbers because its narrow bandgap is close to the Shockley–Queisser optimum.^2–9 FAPbI₃-based PSCs deliver extremely efficient PCEs, and their perovskite absorbers are generally prepared by a one-step antisolvent or a two-step sequential deposition method.^10–16 In particular, the two-step approach has excellent operability and vast potential for commercialization, thanks to the possibility of prior regulation of the lead iodide (PbI₂) templates. However, the heavy dependence of perovskite crystal growth on PbI₂ substrates and organic ammonium salt infiltration means that perovskite grains can easily become disordered and introduce unwanted non-radiative recombination sites, which drastically compromise the photovoltaic efficiency and stability of PSCs. Moreover, the orientation of perovskite crystals can ultimately affect the quality and performance of devices. Via theoretical calculations, prior reports have demonstrated that the (001) crystal plane is the optimal orientation, while conversely, the (111) plane is undesirable and accompanied by a higher defect density,^17,18 which can be also seriously influenced by the PbI₂ film quality. Therefore, the growth of PbI₂ films is crucial for the subsequent intercalation of organic ammonium salts and thus the conversion to the perovskite phases in the typical two-step sequential deposition method.^{13,17,19–22}

Several works have been reported to alter the porosity of PbI₂ and the crystallization process of perovskites for the growth of high-quality perovskite films.^{17,20,21,23–26} For example, Liu and co-workers introduced 1-naphthalenemethylammonium iodide (NpMAI) or two-dimensional (2D) (NpMA)₂PbI₄ perovskites to regulate the crystal growth in 2D/three-dimensional (3D) perovskite films.²¹ The treated PbI₂ films were mesoporous, which allowed full penetration of the ammonium salts and improved the quality of the 2D/3D perovskite films. Luo et al. applied 2D (BDA)PbI₄ (BDA is 1,4-butanediamine) perovskite seeds to PbI₂ precursor solutions. The presence of large perovskite seed crystals allowed the critical Gibbs free energy barrier required for perovskite growth to be overcome, then the optimally oriented perovskite crystals were obtained.¹⁷ Some more reports focus on the nucleation and crystallization of perovskites.^19,26–29 For instance, Han and co-workers proposed perovskite crystal arrays (PCA) with regular distribution to aid in the nucleation and growth of perovskites. The PCA supplied perovskite nuclei that can start crystallization without conquering the critical Gibbs free energy for perovskite nucleation.²⁶ However, few reports have studied in detail the effect of additives on the nucleation and crystallization of the PbI₂ films themselves in the two-step method.

Previous reports have shown that the PbI₂ (001) plane with low surface formation energy is the preferential orientation of PbI₂ films.³⁰ Objectively, the PbI₂ films exist in both amorphous domains and minute (001) crystalline clusters before the ammonium salts dropdown.^31,32 The secondary growth of PbI₂ clusters can be stimulated by regulating the critical Gibbs free energy of PbI₂.³⁰ Intriguingly, Grätzel and co-workers found that when a PbI₂ film was immersed in an MAI solution, PbI₂ (001) facet orientation was further crystallized before the formation of perovskite phases,^22,30 implying that PbI₂ with higher crystallinity before the participation of organic ammonium salts shortens the process required for the subsequent formation of perovskite phases. Therefore, promoting the secondary growth of PbI₂ clusters and improving the crystallinity of PbI₂ films play a crucial role in the subsequent growth of perovskites.

All of these motivate us to prepare porous lead iodide precursor films with larger clusters and higher crystallinity in a sequence deposition process by adding pentafluoroanilinium trifluoromethanesulfonate (PFAT) to PbI₂ solutions, which guarantees the secondary growth of PbI₂ clusters and early formation of the perovskite phases. This further promotes the preferential growth of perovskite crystals in a (001) direction and effectively suppresses an undesired (111) facet orientation. The oriented (001) growth of perovskite crystals perpendicular to the substrates promotes the transfer and extraction of photogenerated carriers and reduces interfacial recombination, ultimately delivering an impressive PCE of 24.52% with a significantly enhanced fill factor (FF) of 0.828. While the control cell only gives a PCE of 22.60% and an FF of 0.796. Moreover, after 500 hours of constant light irradiation, the PFAT-based device upholds ∼90% of its initial PCE. For comparison, the control cell has a deteriorating decline within 150 h under the same conditions. This work offers insights into the critical significance of the prior regulation of PbI₂ and the subsequent growth of perovskite crystals in a two-step sequential deposition approach.

Results and discussion

The device configuration of indium tin oxide (ITO)/Tin(IV) oxide (SnO₂)/perovskite (PFAT)/phenethylammonium iodide (PEAI)/2,2′,7,7′-tetrakis(N,N-dipmethoxyphenylamine)-9,9′-spirobifluorene (spiro-OMeTAD)/Au is shown in Fig. 1a, and the molecule structure of PFAT is displayed in Fig. 1b. Firstly, we introduced PFAT into the PbI₂ precursor solutions and set the concentrations of PFAT to 0, 0.5, 1, and 2 mg mL⁻¹, respectively. Then, we obtained the statistical distribution of device performance parameters and finally confirmed the optimal concentration of 1 mg mL⁻¹ (Fig. S1, ESI†). For the sake of description, the concentration of the variable sets applied below is defaulted to 1 mg mL⁻¹ (called Target), while the pristine films are called Control.

Fig. 1 Characterization of PbI₂. (a) Device architecture of as-fabricated solar cells. (b) Molecule structure of PFAT. (c) FTIR spectra of PFAT before and after mixing with PbI₂. Top-view SEM images of the deposited (d) PbI₂-Control and (e) PbI₂-Target (i.e., 1 mg mL⁻¹ of PFAT) annealed films. The scale bars represent 1 μm. (f) Surface tension measurements of PbI₂ films w/o and with PFAT via the OWRK method. (g) Dependence of Gibbs free energy change on cluster size of the PbI₂-Control and PbI₂-Target films. (h) Cluster size distribution in the PbI₂-Control and PbI₂-Target films, therein, the fraction of the PbI₂-Control clusters acting as nuclei is marked in blue and that of the PbI₂-Target clusters is marked in red. XRD patterns of the (i) PbI₂-Control and (j) PbI₂-Target films with various annealing times at 70 °C. (k) Evolution of FWHM for (001) planes of different PbI₂ films with various annealing times. (l) Gibbs free energy changes during perovskite formation.

Characterization of PbI₂ precursor and film properties

To begin with, we compared PbI₂ precursors with and without (w/o) PFAT to check its effect on the turbidity of the PbI₂ solutions (Fig. S2, ESI†). All the solutions show a distinct Tyndall effect, and no obvious difference can be seen in the PbI₂ solutions. To further understand the underlying chemical interaction between PFAT and PbI₂ in precursor solutions, Fourier transform infrared spectroscopy (FTIR) measurements were conducted. As shown in Fig. 1c, peaks located at 1034 and 1252 cm⁻¹ are related to symmetrical (ν_s) and antisymmetric (ν_as) stretching vibrations of –SO₃⁻ in PFAT, respectively.^33,34 After mixing with PbI₂, the corresponding peaks shift to 1038 and 1254 cm⁻¹, respectively. Besides, N–H bending vibration of –NH₃⁺ (δ_N–H (–NH₃⁺)) in PFAT appears at 1586 cm⁻¹, which is red-shifted by 42 cm⁻¹ after interacting with PbI₂. The red shift of vibrations of –SO₃⁻ and –NH₃⁺ indicates the inherent interaction between –SO₃⁻ and –NH₃⁺ in PFAT becomes weaker, which means the interaction of PFAT with PbI₂.³⁵

We then performed scanning electron microscopy (SEM) and atomic force microscopy (AFM) to study the effect on the morphology of the resulting PbI₂ films. For the two-step sequential deposition method, spin-coated PbI₂ films usually undergo an annealing process at 70 °C for about 1 min.^17,36 It should be noted that this brief annealing process is indispensable because the as-deposited unannealed PbI₂ films present various-size PbI₂ crystals (Fig. S3a and b, ESI†). Moreover, AFM tests show that the unannealed pristine PbI₂ film is very rough, and the addition of PFAT reduces the roughness (Fig. S4a and b, ESI†). As can be seen in Fig. S5a and b (ESI†), both the pristine and PFAT-processed PbI₂ films present sharp protrusions at the top surface. All of these can hinder organic ammonium salts intercalate into the PbI₂ plates and lead to inhomogeneous perovskite grains. Meanwhile, failure to volatilize the redundant solvents can inhibit the subsequent growth of perovskite grains. As shown in Fig. 1d and e, we can see that the morphology of the annealed PbI₂ film becomes much more porous after appropriate PFAT treatment. For comparison, the pristine PbI₂ film is relatively denser, which is not conducive to the penetration of ammonium salts. The much highly porous PbI₂ thin film can lead to the formation of favorable α-FAPbI₃.¹¹ We speculate that the formation of the much more porous PbI₂ film may arise from the pre-aggregation induced by an intermediate phase.²¹ The statistics of cluster sizes (Fig. S6, ESI†) show that the PbI₂ clusters are significantly larger after the PFAT treatment. This facilitates the subsequent cluster growth, which will be mentioned later. The corresponding AFM tests reflect the same trend (Fig. S7a and b, ESI†). Incidentally, the brief annealing process for the PbI₂ films reduces the surface roughness. Cross-sectional SEM images present that a porous trend is already present in the direction perpendicular to the substrate in the unannealed PFAT-processed film (Fig. S8a and b, ESI†). After annealing, the PbI₂-Target film is more porous and looser in the longitudinal direction compared to the PbI₂-Control film (Fig. S9a and b, ESI†), which provides more channels for the penetration of the ammonium salts. When excess PFAT is added, the pores become enlarged and are accompanied by deterioration of the homogeneity of PbI₂ crystals (Fig. S10, ESI†), which is detrimental to the subsequent growth of perovskite grains.

The crucial thermodynamic factors affecting PbI₂ nucleation are Gibbs free energy (ΔG*) and the critical cluster size (N*), which are determined by the interfacial term and the volume term (the following eqn (1) and (2)).³⁰

here, η_i and Γ_i are the shape factor and the interfacial free energy per unit interfacial area for the ith area, respectively. μ₁ and μ₂ are the respective chemical potentials of the bulk amorphous and crystalline phases, respectively. We assume that the volume term is unchanged, therefore the smaller interfacial term brings the lower ΔG* and N*.³⁰ Then, surface tension measurements were performed to observe the effect of PFAT on the interfacial term of the annealed PbI₂ films via the Owens–Wendt–Rabel–Kaelble (OWRK) method.³⁵ The contact angles (θ) of four types of standard liquids whose surface tensions (σ_L) including their dispersive (σ^d_L) and polar (σ^p_L) components are known were measured (Fig. S11, S12, Table S1 and Note S1, ESI†) to confirm the surface tensions (σ) of the PbI₂ films. As shown in Fig. 1f, the total σ decreases from 59.9 to 58.3 mN m⁻¹ after incorporating PFAT, which implies slightly lower ΔG* and N* (ΔG*_Target < ΔG*_Control, N*_Target < N*_Control) (Fig. 1g). From the PbI₂ cluster distribution (Fig. S6, ESI†), we know that the PbI₂-Target annealed film has larger PbI₂ clusters relative to the PbI₂-Control annealed film. Combined that the critical N*_Target is smaller than N*_Control, more PbI₂ clusters are able to surpass the Gibbs free energy barrier directly and undergo the secondary nucleation and the further crystallization for the PbI₂-Target film in the subsequent interaction with the ammonium salts, as shown in Fig. 1h.³⁰ The further growth of the already relatively large PbI₂ clusters in the PbI₂-PFAT film is beneficial for the formation of perovskites, as will be discussed later.

Apart from the nucleation, the PFAT addition also has an obvious effect on the crystallinity of PbI₂ films. To further understand the impact of the PFAT addition on the crystallinity of PbI₂ films, X-ray diffraction (XRD) was carried out. We tested XRD patterns of the PbI₂ films with various annealing times at 70 °C to analyze the change in crystallinity, as presented in Fig. 1i and j. Given the variation of the crystallinity of PbI₂, the full width at half-maximum (FWHM) and peak intensity evolution of the PbI₂ (001) plane with various annealing times are recorded in Fig. 1k and Fig. S13 (ESI†), respectively. Initially, the pristine PbI₂ film shows a strong peak intensity and a narrow FWHM. In combination with the aforementioned results shown in Fig. S4 and S5 (ESI†), the unannealed pristine PbI₂ film shows a lamellar structure with many prominent spikes, which confirm the strong crystallinity of the initial PbI₂ film. However, the unsmooth surface of the unannealed PbI₂ film can hinder the penetration of the ammonium salts and destroy the favorable growth of perovskite grains. The slightly higher FWHM value and the much lower peak intensity in the PbI₂-Target film are probably due to the complexation of PFAT with part of PbI₂. During the subsequent 20 s of annealing, the (001) peak intensity of the PbI₂-Control film decreases sharply, while that of the PbI₂-Target film remains almost unchanged, even though the FWHMs of both films increase. This is because the rapid evaporation of the residual solvents leads to the partial destruction of the inherent crystallinity for the pristine PbI₂ film. While this phenomenon is alleviated by the interaction of PFAT with PbI₂ in the PFAT-processed PbI₂ film. Accompanied by further thermal annealing, the peak intensity of the control PbI₂ film no longer changes owing to the solvents being evaporated at the beginning of the annealing, while the peak intensity of the target one slightly decreases as a result of the evaporation of the remaining solvent dragged by PFAT. All the final annealed PbI₂ films have smoother morphology compared to the initial unannealed PbI₂ films (Fig. S4 and S7, ESI†), which creates favorable conditions for the subsequent interaction of the ammonium salts, even with the compromise of the reduced crystallinity. Moreover, the PFAT-treated film maintains a lower FWHM value relative to the untreated one, which implies a higher crystallinity of the annealed PFAT-processed PbI₂ film. This is very crucial for perovskite growth, since the formation of the perovskite phases does not occur immediately after the contact of the ammonium salts with PbI₂, but that further crystallization of PbI₂ takes place first.^22,30 In other words, the higher the crystallinity of the annealed PbI₂ films, the earlier the formation of the perovskite phases proceeds.

The interaction of the ammonium salts with the PbI₂ is usually followed by further crystallization of the as-deposited PbI₂ (named Process I) before intercalation of the ammonium salts and formation of the perovskite phases (named Process II).^22,30,31 The whole processes (Process I + II) follows Ostwald's step rule, where the reaction gradually stabilizes from a metastable state at the beginning, accompanied by a decrease in Gibbs free energy.²² In that way, the formation of a higher-crystallinity PbI₂ film with lower Gibbs free energy before the interaction with the ammonium salts facilitates the faster completion of Process I and then, an earlier start of Process II, which leads to adequate interaction and larger perovskite grains, as shown schematically in Fig. 1l. This indicates that the crystallinity regulation of PbI₂ films by PFAT is beneficial to the growth of perovskite crystals. Considering the secondary nucleation of PbI₂ clusters in the PbI₂-PFAT film mentioned earlier (Fig. 1g and h), more and more highly crystalline PbI₂ forms and promotes the early formation of perovskites. Thus, through the introduction of PFAT, the modulation of both nucleation and crystallization in PbI₂ films promotes perovskite formation.

Morphology and photoelectrical properties of perovskite films

The XRD evolution of the deposited perovskite films with various thermal annealing times (0, 3, 9, and 13 min) was further explored. As shown in Fig. 2a and b, for the control film, the (001) peak intensity of perovskite does not change much with the annealing time increases but is accompanied by a noticeable increase in the PbI₂ (001) peak intensity and an undesirable perovskite (111) orientation. The enhancement of the PbI₂ (001) peak intensity can be attributed to the partial decomposition of the perovskite films with the volatilization of the ammonium salts during the annealing process.^37–39 In contrast, the peak intensity of PbI₂ in the target film does not change much, while the perovskite (001) peak intensity increases significantly and (111) plane diffraction peak decreases. It can be visualized that the ratio of perovskite (001)/(111) for the target film rises sharply with the annealing time increases, while that for the control one remains almost constant (Fig. 2c). In previous reports, the (001) facet is benign in perovskite growth and has a higher photoluminescence (PL) intensity, while the (111) facet tends to have a higher defect state density.^17,18,40 The higher (001)/(111) ratio implicates that the preferential orientation of perovskite crystals can be modulated by doping PFAT, which helps to obtain perovskite films with higher quality and fewer defects. Remarkably, when we compare the change in the ratio of the perovskite (001) plane and PbI₂ (001) plane for the two sets of films (Fig. 2d), it can be found that the ratio increases rapidly in the first 3 minutes of annealing for the target film. This is because flat and porous PbI₂ films with higher crystallinity have a facile and adequate intercalation with the ammonium salts and convert more rapidly to perovskite phases, which confirms the above analysis of the section on PbI₂ films. However, the pristine over-dense PbI₂ film deteriorates the formation of perovskite phases. In the subsequent annealing, the slowing down of the perovskite (001)/PbI₂ (001) ratio for both films may be associated with the partial volatilization of the ammonium salts induced by thermal annealing. Moreover, we can also observe that the residual PbI₂ in perovskite films is greatly reduced with the incorporation of PFAT (Fig. 2a, b, and d).⁴¹

Fig. 2 Characterization of perovskite films. XRD patterns of the (a) control and (b) target perovskite films with varied annealing time. Evolution of (c) the perovskite (001)/(111) ratio and (d) the perovskite (001)/PbI₂ (001) ratio in different perovskite films. Top-view and inserted cross-sectional SEM images of perovskite films (e) before and (f) after PFAT incorporation. All scale bars are represented as 1 μm. KPFM images of perovskite films (g) w/o and (h) with PFAT. (i) Energy level diagram of different perovskite films.

We then performed top-view and cross-sectional SEM measurements to verify the effect of PFAT on the growth of perovskite grains. As depicted in Fig. 2e and f, for the control perovskite film, the grains are dispersed and heterogeneous in size because of the penetration difficulty of the ammonium salts. On the contrary, the grains are much larger and interconnected for the target film, owing to the thorough penetration of the ammonium salts and the adequate interaction with PbI₂, which reduce perovskite grain boundaries and further diminish non-radiative recombination. From the cross-sectional SEM images (insets of Fig. 2e and f), one can see that the PFAT-treated perovskite film has larger grains in size and volume and grows perpendicular to the substrate, while the untreated film is irregular in the longitudinal direction. The preferred vertical growth orientation of the PFAT-treated perovskite film is conducive to charge transport and extraction. Consistent with the XRD results, the SEM images also present that the residual PbI₂ in perovskite films is greatly reduced with the incorporation of PFAT, especially for that at the grain boundaries (Fig. 2e and f). In addition, the reduction of excess PbI₂ within perovskites would benefit the stability of perovskite films and devices.⁴¹

AFM and Kelvin probe force microscopy (KPFM) were utilized to test the surface roughness and potential of perovskite films. For both the perovskite films w/o and with PFAT, there is no significant change in roughness (Fig. S14a and b, ESI†), yet the surface potential of the two samples differs considerably (Fig. 2g and h). The surface potential of the target film (−0.145 V) is increased by 222 mV with respect to that of the control one (−0.367 V). The apparently higher surface potential implies a downward shift in the Fermi energy level of the PFAT-treated perovskite film surface, which would facilitate carrier extraction.⁴² A highly oriented pyrolytic graphite (HOPG) was used as a standard sample to calibrate the values of Fermi energy levels of perovskite films (Fig. S15, ESI†). Based on the KPFM results, the Fermi energy levels of the control and target films are estimated to be around −3.994 and −4.216 eV (see ESI† Note S2 for details), respectively. Ultraviolet photoelectron spectroscopy (UPS) was then employed to validate Fermi energy levels and valence band maximums (VBMs) of the two films (Fig. S16, ESI†). The energy level diagram was drawn on the basis of the UPS results, as shown in Fig. 2i. The Fermi levels of the control and target films are determined to be 4.01 and 4.41 eV as well as VBMs of −5.45 and −5.66 eV, respectively, which nearly coincide with the KPFM results. The deeper VBM of PFAT-tailored perovskite contributes to higher open-circuit voltage (V_OC) of devices,^42–45 which can be evidenced later.

The optical properties of perovskite films were then investigated by ultraviolet-visible (UV-vis) absorption spectra (Fig. S17, ESI†). The markedly enhanced absorption of the target perovskite film can be ascribed to its better orientation and extended grain size and volume (insets in Fig. 2e and f). In addition, compared to that of the control one, the bandgap of the target one becomes slightly wider (Fig. S18, ESI†), which may be attributed to the presence of the large cation in PFAT. Transient absorption (TA) spectroscopic measurements were then employed to distinguish the carrier transport dynamics in the control and target perovskite films, as shown in Fig. 3a and b. Upon excitation of a sample with a 400 nm pump light, electrons in the ground state absorb photons and jump to the excited state, reducing the number of electrons in the ground state. The reduced absorption of a probe light by the excited sample allows a negative ΔA signal to be detected, i.e., a ground-state bleaching (GSB) signal.^46,47 There is a slight blue shift in the GSB peak position of the target film (784.2 nm) compared to that of the control one (785.8 nm) because of the slightly wider bandgap (Fig. S18, ESI†). The variation of the GSB peaks with decay time is shown in Fig. S19 (ESI†), and the fitted data are summarized in Table S2 (ESI†). The carrier lifetime τ₂ increases from 7613.3 ps to 11293.4 ps after PFAT addition, indicating decreased carrier recombination in the target film.^41,48 Steady-state photoluminescence (PL) and time-resolved photoluminescence (TRPL) were also measured to compare carrier transport dynamics in different perovskite films. The PFAT-processed perovskite film shows a stronger PL emission and longer PL decay lifetime than the control film, owing to the suppression of non-radiative recombination (Fig. 3c and d and Table S3, ESI†). The slight blue shift conforms to the change in bandgap obtained from the absorption results (Fig. S18, ESI†). For both glass/SnO₂/perovskite and glass/perovskite/spiro-OMeTAD structures, the stronger PL quenching of the target film illustrates better electron and hole extraction ability relative to that of the control one (Fig. S20, ESI†). We also employed a confocal laser scanning fluorescence microscopy (CLSM) to examine the PL mapping of the samples. As viewed in Fig. 3e and f, for the pristine perovskite film, the uneven and weak PL distribution implies poor crystalline quality. After PFAT addition, much more homogeneous and brighter PL emissions and larger perovskite grains observed in the target film confirm the significantly improved crystallinity and reduced non-radiative recombination of the sample.⁴¹

Fig. 3 Optoelectronic properties of perovskite films. Pseudo color maps of TA spectra for the (a) control and (b) target perovskite films upon excitation at 400 nm. (c) Steady-state PL, and (d) TRPL spectra of the control and target perovskite films deposited on glass substrates. PL mapping images of the perovskite films under CLSM (e) w/o and (f) with PFAT treatment. The scale bars represent 1 μm. (g) Pb 4f, (h) I 3d, (i) N 1s, and S 2p spectra of perovskite films w/o and with PFAT incorporation.

Underlying mechanism explanation

X-Ray photoemission spectroscopy (XPS) was utilized to explore the chemical interaction on the surface of perovskite films. The full XPS spectra of different perovskite films are displayed in Fig. S21 (ESI†). The detailed Pb 4f and I 3d spectra of the target film simultaneously show a slight shift of the peak position toward smaller binding energy (Fig. 3g and h), implying the interaction of PFAT with Pb-I octahedra. The –SO₃⁻ in PFAT interacts with the uncoordinated Pb,^49,50 while –NH₃⁺ donates a lone pair of electrons to positively charged iodine vacancies for defect healing.^51,52 The attenuated Pb⁰ peaks imply the reduction of residual PbI₂, which can further decompose into harmful Pb⁰ impurities by light excitation.⁴¹ The elevated ratio of C–N/C=N in the N 1s peaks and the appearance of the S 2p peaks in Fig. 3i signify that both the cation (with C–N) and anion (with S element) of PFAT are present in the perovskite films after PFAT processing, rather than volatilizing with thermal annealing.

With the above characterizations, we construct schematic diagrams to describe potential mechanisms of the underlying perovskite crystal growth (Fig. 4). Generally, after annealing at 70 °C, PbI₂ films have both amorphous domains and PbI₂ clusters.^30–32 For the control sets (Fig. 4a–c), initially, the PbI₂ clusters are small and the crystallinity is relatively low. When the ammonium salt solutions are dropped on the PbI₂-Control films with low crystallinity and small crystalline clusters, clusters whose sizes do not reach the critical N*_Control cannot grow further, and PbI₂ in the amorphous domains is transported to the larger clusters (cluster sizes > N*_Control) to promote their secondary growth. Then, only a small amount of PbI₂ clusters reach a critical crystallinity and the ammonium salts begin to interact with them. The slow consumption of PbI₂ leads to increased residual PbI₂ and the formation of an undesirable perovskite (111) plane, which eventually brings about small grains and many defects. In contrast, for the target sets (Fig. 4d–f), the PbI₂ clusters are larger and have relatively higher crystallinity in the PbI₂-Target films initially. When the ammonium salts start to contact PbI₂, more PbI₂ clusters undergo their secondary nucleation because of N*_Target < N*_Control (Fig. 1g and h), which has been mentioned earlier. The large clusters that reach the critical crystallinity start to interact with the ammonium salts to form perovskite phases, and the slightly smaller clusters consume amorphous domains to achieve the desired crystallinity. As the interaction continues, more PbI₂ clusters begin to be converted to perovskites. Then, more PbI₂ is consumed and a more optimal orientation of a perovskite (001) facet is obtained in the target ones, owing to the higher crystallinity of the PbI₂ films with large clusters. Note that when the PbI₂ films are in contact with the ammonium salts, the whole thermodynamic system is metastable at the beginning and spontaneously tends to a steady state until the final conversion to perovskites, accompanied by a reduction of Gibbs free energy of the whole system, which follows Ostwald's step rule. In addition, the –NH₃⁺ and –SO₃⁻ terminal groups in PFAT can further passivate inherent defects in perovskite films.^13,35,53 The –NH₃⁺ passivates positively charged iodine vacancies by delivering a lone pair of electrons, and the –SO₃⁻ can interact with undercoordinated Pb²⁺ ions.

Fig. 4 Underlying mechanism explanations. Schematic diagrams of the interaction of dripped organic ammonium salts with the deposited PbI₂ films (a–c) w/o and (d–f) with PFAT and the formation of perovskite crystals. For simplicity, the ammonium salts are replaced by FAI. (a) The small clusters with low crystallinity. (b) A small number of clusters start the secondary growth and convert to perovskite phases, accompanied by the formation of undesired perovskite (111) facets and PbI₂ residue. (c) The perovskite film with small grains and an unwanted (111) orientation. (d) The large clusters with high crystallinity. (e) The conversion of perovskite phases occurs and clusters with sizes larger than N*_Target commence the secondary growth and then interact with the ammonium salts. (f) The perovskite film with large grains and a preferential orientation.

Device performance, defect characterization, and stability studies

The operation performance statistics of the control and PFAT-processed devices are presented in Fig. 5a. Obviously, PFAT doping leads to a significant increase in the V_OC and FF of PSCs, which promotes an effective increment in PCE. The average (champion) PCE, V_OC, FF, and the short-circuit current density (J_SC) for the target devices without antireflective films are 22.77 ± 0.23% (23.13%), 1.163 ± 0.005 V (1.171 V), 81.69 ± 0.14% (81.86%), and 23.98 ± 0.20 mA cm⁻² (24.28 mA cm⁻²), respectively, while those for the control cells without antireflective films are 22.10 ± 0.22% (22.42%), 1.143 ± 0.002 V (1.146 V), 80.55 ± 0.54% (81.16%), and 24.00 ± 0.24 mA cm⁻² (24.33 mA cm⁻²), respectively. Polydimethylsiloxane (PDMS) antireflective films were also pasted on the glass substrates of the PFAT-processed PSCs to reduce the reflection light on the glass side, thus significantly increasing the J_SC of the devices (Fig. S22, ESI†). The best-performing PFAT-treated PSC with a PDMS antireflective coating gives an impressive PCE of 24.52% with a much-enhanced V_OC of 1.164 V, an FF of 82.84%, and a J_SC of 25.43 mA cm⁻² (Fig. 5b), which is one of the highest PCEs in two-step sequential deposition-processed PSCs so far. The data of the forward and reverse sweeps of the corresponding device is also ordered (Fig. S23 and inset therein, ESI†). Depending on the external quantum efficiency (EQE) spectrum (Fig. 5c), the integrated current density of the target device is 25.27 mA cm⁻², which coincides with the J_SC obtained from the J–V curve (Fig. 5b). Steady power output (SPO) test (Fig. 5d) shows that the PFAT-processed device maintains a steady-state PCE of approximately 24.3% at a bias voltage of 0.99 V for 200 s. For comparison, the relevant parameters for the best-performing control cell are placed in Fig S24 (ESI†).

Fig. 5 Photovoltaic performance and stability of PSCs. (a) Statistical distribution charts of the photovoltaic parameters for the control and PFAT-processed devices. (b) J–V curve of the best-performing target device, which includes an antireflective film to increase the J_SC. (c) EQE spectrum and integrated curve of the best-performing target cell. (d) SPO test of the best-performing target device measured at a constant bias voltage of 0.99 V within 200 s. SCLC measurements for the electron-only (e) control and (f) target devices. (g) Mott–Schottky curves for different PSCs. The dependence of (h) J_SC and (i) V_OC on light intensity for devices. (j) PCE decay of the best-performing control and target PSCs in a dry cabinet (∼25 °C, 0–10% relative humidity) in the dark. (k) MPP tracking of the unencapsulated cells under constant light exposure at ∼55 °C in an N₂-filled glovebox.

Space-charge-limited current (SCLC) characterization was conducted to compare the trap density (N_t) of electrons and holes (Fig. 5e and f and Fig. S25a and b, ESI†) for different samples by using electron-only and hole-only devices, respectively.^54,55 For the electron-only devices, the trap-filled limited voltage (V_TFL) of the target cell is around 1.00 V, which can be calculated as N_t = 3.6 × 10¹⁵ cm⁻³, whereas that of the control one is around 1.33 V (N_t = 4.8 × 10¹⁵ cm⁻³). As for the hole-only devices, the derived N_t equals 1.4 × 10¹⁵ cm⁻³ for the control device (V_TFL = 0.39 V), while the N_t of the PFAT-tailored cell is only 1.0 × 10¹⁵ cm⁻³ (V_TFL = 0.28 V). The reduced density of defect states for both electrons and holes for the target device can be ascribed to a preferred crystalline orientation, lowered defects, and enlarged grains. We then measured Mott–Schottky curves (Fig. 5g) to investigate the variation of the built-in electric field in PSCs, which can be estimated by the following formula:

The meaning of each parameter was introduced in our previous work.⁵⁰ The intercept of the tangent line with the horizontal coordinate is the built-in potential (V_bi). The target device exhibits a higher V_bi (0.91 V) compared to that of the control one (0.81 V), which suggests a stronger driving force to separate electrons and holes, thus favoring carrier extraction.⁵⁶ Electrochemical impedance spectra (EIS) were also tested in the dark to evaluate charge transfer kinetics (Fig. S26, ESI†). As compared to the control device, the PFAT-based PSC demonstrates a slightly higher series resistance (R_S) and a lower charge transfer resistance (R_tr) owing to the insulating behavior of PFAT and the vertical orientation of perovskites, respectively, which is reflected in the slightly decreased J_SC and the increased FF and PCE of solar cells.^49,50,57 To acquire the power exponent (α) and ideal factor (n_id), we tested the dependence of J_SC and V_OC on light intensity for PSCs, respectively, as indicated in Fig. 5h and i. The PFAT-processed PSC possesses a α of 0.990 and a n_id of 1.32, which are closer to 1 as compared to the α (0.986) and n_id (1.41) of the pristine cell, confirming the decreased non-radiative recombination in the target devices.

We then tested and compared the long-term storage and light stability of the control and target devices. The devices were placed in a dry cabinet (∼25 °C, 0–10% relative humidity) in the dark for long-term storage stability measurements (Fig. 5j). The initial PCEs of the control and target devices were 22.60% and 24.52%, respectively. After 61 days of storage, the efficiency of the target group remained above 90% of its initial efficiency, while the efficiency of the control group fell below 90% of its initial efficiency within 35 days, demonstrating much better storage stability after the PFAT treatment. Maximum power point (MPP) tracking at an operating temperature of ∼55 °C was utilized to further monitor the long-term photostability of PSCs in an N₂ atmosphere. As shown in Fig. 5k, after 500 hours of constant light exposure, the target cell maintained ∼90% of its initial efficiency, in contrast to the control PSC, which dropped to ∼75% of its initial PCE after only 136 hours of constant light illumination. All these results indicate that the PFAT treatment improves the operation and photostability of the devices, which can be attributed to the enhanced crystalline quality, decreased surplus PbI₂, and reduced defects of as-fabricated perovskite films.

Conclusion

In summary, by using PFAT to modulate the morphology, nucleation, and crystallinity of PbI₂ films, we obtained preferentially oriented perovskite films with a largely reduced amount of residual PbI₂. The incorporation of PFAT allows for the effective intercalation of organic ammonium salts. The secondary nucleation of larger PbI₂ clusters and the enhanced crystallinity of PbI₂ films enable the organic salts to interact with PbI₂ earlier. Besides, the multi-functional groups in PFAT can passivate inherent defects in perovskite films through Lewis acid-base interactions. All of these reasons contribute to high-quality perovskite films with enlarged grains, reduced defects, and favored crystal growth of perovskites perpendicular to the substrates. The homogeneous growth of perovskites ultimately affords solar cells with a striking PCE of 24.52%, accompanied by significantly improved storage and MPP stability. This work provides a way to achieve favorable growth of perovskites and optimize the operational performance and stability of devices by pre-modulating PbI₂ precursor films.

Experimental

Materials and reagents

PbI₂ (99.9985%) and lithium bis(trifluoromethanesulfonyl)imide (Li-TFSI) were bought from Tokyo Chemical Industry (TCI, Japan). PFAT was purchased from Macklin. PEAI was purchased from Xi’an Polymer Light Technology. Formamidinium iodide (FAI), methylammonium iodide (MAI), and spiro-OMeTAD (99%) were bought from Advanced Election Technology CO., Ltd. Methylammonium chloride (MACl) was purchased from Sigma Aldrich. SnO₂ colloid solution (15 wt%) was bought from Alfa-Aesar. Hydrogen peroxide aqueous solution (H₂O₂, 30%) was purchased from Sinopharm Chemical Reagent Co., Ltd. N,N-dimethylformamide (DMF, 99.8%), dimethyl sulfoxide (DMSO, 99.9%), isopropanol (IPA) (99.7%), chlorobenzene (CB 99.8%), 4-tert-butylpyridine (tBP), and acetonitrile (ACN, 99.9%) were also purchased from Sigma-Aldrich.

Device fabrication

SnO₂ colloid solution was diluted with DI water and H₂O₂ aqueous solution (1 : 4 : 1, v : v : v) and stirred at room temperature for 1 hour. The mixed aqueous solution was spin-coated onto the glass/ITO substrates at 4000 rpm for 30 s and then the substrates were annealed at 180 °C for 25 min in ambient air. To prepare the precursor solutions for the perovskite layers, 1.4 M PbI₂ solutions (DMF/DMSO, v : v = 9 : 1) with different concentrations of PFAT (0, 0.5, 1, and 2 mg mL⁻¹) were stirred at 70 °C for 6 hours, and organic ammonium salt solutions (FAI : MACl : MAI = 60 mg : 10 mg : 10 mg in 1 mL IPA) were stirred at room temperature for 6 h. In an N₂-filled glove box, PbI₂ solutions were spin-coated onto the deposited SnO₂ layers at 1500 rpm for 30 s and immediately transferred to a hot plate for annealing at 70 °C for 1 min. After cooling, 100 μL of organic ammonium salt solutions were then rapidly poured down and spin-coated at 1800 rpm for 30 s. The substrates were transferred to a hot table in ambient air (30–40% relative humidity) to anneal at 150 °C for 15 min. For the passivation treatment, a PEAI solution (2 mg mL⁻¹ in IPA) was deposited on the cooled-down perovskite films at 5000 rpm for 30 s without further annealing in an N₂-filled glove box. A stock spiro-OMeTAD solution consisting of 72.3 mg in 1 mL of CB with 28.8 μL of tBP and 17.5 μL of Li-TFSI solution (520 mg of Li-TFSI powder in 1 mL of ACN) was then coated on the modified-perovskite films at 3000 rpm for 30 s. After at least 8 hours of oxidation in a dry cabinet, 60 nm of Au was thermally evaporated at 2 × 10⁻⁴ Pa. PDMS antireflection films were pasted on the glass substrates to increase light collection.

Characterizations

FTIR spectra were obtained by a NICOLET 5700 FTIR Spectrometer in the wavenumber range of 900–4000 cm⁻¹. The morphology of the films was visualized by a JSM 6700F field emission scanning electron microscope (FE-SEM) with an accelerating voltage of 15 kV. XRD patterns were attained by an X-ray diffraction instrument (SmartLab-3kW, Rigaku, Japan) with a scan speed of 5 degrees min⁻¹. TA spectra were obtained from an Ultrafast System HELIOSTM TA spectrometer. AFM and KPFM images were observed by a Bruker Dimension Icon XR equipment in a Scanasyst-Air and Peak force KPFM Mode. UV-vis absorption spectra were measured by a SHIMADZU mini 1280 UV-visible spectrophotometer. Steady-state PL and TRPL of the perovskite films were recorded by a HORIBA Jobin Yvon IBH Ltd with a 481 nm laser and a time-correlated single-photon counting module. PL mapping was captured by a confocal laser scanning fluorescence microscopy (CLSM) (TCS SP8, Leica). A 552 nm laser was applied as the PL excitation source to detect and collect the PL signals at 700–800 nm, and the scale range collected on the perovskite films was 10 μm × 10 μm. UPS and XPS measurements were carried out by an ESCALAB 250Xi (Thermo Fisher Scientific, USA). J–V curves, SCLC, and steady-state power output measurements were scanned and recorded by a Keithley 2400 Source Meter with a Newport 94023A solar simulator. Before each time J–V measurement, the light intensity of the solar simulator was calibrated by a National Renewable Energy Laboratory-certified and KG5-filtered crystalline silicon solar cell. The mismatch factor was 1.0. The active region (0.1188 cm²) of the devices was covered by a mask with an aperture area of 0.0557 cm². EQE spectra were collected by a QE system (Enli Technology Co., Ltd). Electrochemical impedance spectra (EIS) were tested by a CHI760E electrochemical workstation (Shanghai Chenhua Instruments, Inc). MPP tracking data measured under light illumination (a white LED light source with 70 mW cm⁻²) were recorded with a multi-channel testing system (Suzhou D & R Instruments). The unencapsulated cells were placed in an N₂-filled glove box and the operational temperature is about 55 °C.

Author contributions

Conceptualization, W. S.; methodology, W. S.; investigation, W. S.; formal analysis, W. S., H. W., F. Y., C. W., C. W., H. C., K. D., Y. G., and T. W.; funding acquisition, W. K., and G. F.; project administration, G. F.; resources, W. S.; validation, W. S.; visualization, W. S.; writing – original draft, W. S.; writing – review & editing, W. S., W. K., and G. F.; supervision, W. K., and G. F.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was supported by the National High Technology Research and Development Program (grant number 2015AA050601), the National Natural Science Foundation of China (grant numbers 12134010 and 12174290), the Natural Science Foundation of Hubei Province, China (grant number 2021CFB039), the Shenzhen Fundamental Research Program (grant number JCYJ20190808152609307), and Guangdong Basic and Applied Basic Research Foundation (grant number 2022A1515011931). We thank the Core Facility of Wuhan University for SEM and FTIR analysis.

References

1. NREL, Best research-cell efficiencies, (accessed September, 2022), https://www.nrel.gov/pv/cell-efficiency.html.
2. M. Kim, J. Jeong, H. Lu, K. Lee Tae, T. Eickemeyer Felix, Y. Liu, W. Choi In, J. Choi Seung, Y. Jo, H.-B. Kim, S.-I. Mo, Y.-K. Kim, H. Lee, G. An Na, S. Cho, R. Tress Wolfgang, M. Zakeeruddin Shaik, A. Hagfeldt, Y. Kim Jin, M. Grätzel and S. Kim Dong, Science, 2022, 375, 302–306.
3. J. Jeong, M. Kim, J. Seo, H. Lu, P. Ahlawat, A. Mishra, Y. Yang, M. A. Hope, F. T. Eickemeyer, M. Kim, Y. J. Yoon, I. W. Choi, B. P. Darwich, S. J. Choi, Y. Jo, J. H. Lee, B. Walker, S. M. Zakeeruddin, L. Emsley, U. Rothlisberger, A. Hagfeldt, D. S. Kim, M. Gratzel and J. Y. Kim, Nature, 2021, 592, 381–385.
4. H. Lu, Y. Liu, P. Ahlawat, A. Mishra, W. R. Tress, F. T. Eickemeyer, Y. Yang, F. Fu, Z. Wang, C. E. Avalos, B. I. Carlsen, A. Agarwalla, X. Zhang, X. Li, Y. Zhan, S. M. Zakeeruddin, L. Emsley, U. Rothlisberger, L. Zheng, A. Hagfeldt and M. Grätzel, Science, 2020, 370, eabb8985.
5. H. Min, M. Kim, S.-U. Lee, H. Kim, G. Kim, K. Choi, H. Lee Jun and I. Seok Sang, Science, 2019, 366, 749–753.
6. G. Kim, H. Min, S. Lee Kyoung, Y. Lee Do, M. Yoon So and I. Seok Sang, Science, 2020, 370, 108–112.
7. Y. Zhang, Y. Li, L. Zhang, H. Hu, Z. Tang, B. Xu and N. G. Park, Adv. Energy Mater., 2021, 11, 2102538.
8. Z. Liu, P. Liu, M. Li, T. He, T. Liu, L. Yu and M. Yuan, Adv. Energy Mater., 2022, 12, 2200111.
9. C. Liu, J. Zhang, L. Zhang, X. Zhou, Y. Liu, X. Wang and B. Xu, Adv. Energy Mater., 2022, 12, 2200945.
10. C. Chen, Y. Lou, K. Wang, Z. Su, C. Dong, J. Chen, Y. Shi, X. Gao and Z. Wang, Adv. Energy Mater., 2021, 11, 2101538.
11. P. Ahlawat, A. Hinderhofer, A. Alharbi Essa, H. Lu, A. Ummadisingu, H. Niu, M. Invernizzi, M. Zakeeruddin Shaik, M. I. Dar, F. Schreiber, A. Hagfeldt, M. Grätzel, U. Rothlisberger and M. Parrinello, Sci. Adv., 2021, 7, eabe3326.
12. S. Chen, X. Xiao, B. Chen, L. Kelly Leah, J. Zhao, Y. Lin, F. Toney Michael and J. Huang, Sci. Adv., 2021, 7, eabb2412.
13. Z. Qin, Y. Chen, X. Wang, N. Wei, X. Liu, H. Chen, Y. Miao and Y. Zhao, Adv. Mater., 2022, 34, 2203143.
14. X. Luo, Z. Shen, Y. Shen, Z. Su, X. Gao, Y. Wang, Q. Han and L. Han, Adv. Mater., 2022, 34, 2202100.
15. J. Lu, T. Yang, T. Niu, N. Bu, Y. Zhang, S. Wang, J. Fang, X. Chang, T. Luo, J. Wen, Y. Yang, Z. Ding, K. Zhao and S. Liu, Energy Environ. Sci., 2022, 15, 1144–1155.
16. X. Zhu, S. Yang, Y. Cao, L. Duan, M. Du, J. Feng, Y. Jiao, X. Jiang, Y. Sun, H. Wang, S. Zuo, Y. Liu and S. Liu, Adv. Energy Mater., 2022, 12, 2103491.
17. C. Luo, G. Zheng, F. Gao, X. Wang, Y. Zhao, X. Gao and Q. Zhao, Joule, 2022, 6, 240–257.
18. L. Zhang, Y. Liu, X. Ye, Q. Han, C. Ge, S. Cui, Q. Guo, X. Zheng, Z. Zhai and X. Tao, Cryst. Growth Des., 2018, 18, 6652–6660.
19. J. Wang, L. Yuan, H. Luo, C. Duan, B. Zhou, Q. Wen and K. Yan, Chem. Eng. J., 2022, 446, 136968.
20. W. Chen, S. Liu, Q. Li, Q. Cheng, B. He, Z. Hu, Y. Shen, H. Chen, G. Xu, X. Ou, H. Yang, J. Xi, Y. Li and Y. Li, Adv. Mater., 2022, 34, 2110482.
21. T. Zhou, Z. Xu, R. Wang, X. Dong, Q. Fu and Y. Liu, Adv. Mater., 2022, 34, 2200705.
22. A. Ummadisingu and M. Grätzel, Sci. Adv., 2018, 4, e1701402.
23. X. Duan, X. Li, L. Tan, Z. Huang, J. Yang, G. Liu, Z. Lin and Y. Chen, Adv. Mater., 2020, 32, 2000617.
24. Y. Zhao, X. Zhang, X. Han, C. Hou, H. Wang, J. Qi, Y. Li and Q. Zhang, Chem. Eng. J., 2021, 417, 127912.
25. Z. Liu, K. Deng, Y. Zhu, M. Wang and L. Li, Solar RRL, 2018, 2, 1800230.
26. Z. Shen, Q. Han, X. Luo, Y. Shen, T. Wang, C. Zhang, Y. Wang, H. Chen, X. Yang, Y. Zhang and L. Han, Energy Environ. Sci., 2022, 15, 1078–1085.
27. W. Zhang, J. Du, W. Zhang, Y. Chu, A. Mei, Y. Rong, X. Gao, H. Han and Y. Hu, J. Energy Chem., 2022, 69, 253–260.
28. Y. Zhang, Y. Wang, X. Yang, L. Zhao, R. Su, J. Wu, D. Luo, S. Li, P. Chen, M. Yu, Q. Gong and R. Zhu, Adv. Mater., 2022, 34, 2107420.
29. F. Wang, P. Wai-Keung Fong, Z. Ren, H.-L. Xia, K. Zhou, K. Wang, J. Zhu, X. Huang, X.-Y. Liu, H. Wang, Y. Shi, H. Lin, Q. Zhu, G. Li and H. Hu, J. Energy Chem., 2022, 73, 599–606.
30. A. Ummadisingu, L. Steier, J. Y. Seo, T. Matsui, A. Abate, W. Tress and M. Gratzel, Nature, 2017, 545, 208–212.
31. H. A. Harms, N. Tétreault, N. Pellet, M. Bensimon and M. Grätzel, Faraday Discuss., 2014, 176, 251–269.
32. F. Liu, F. Sommer, C. Bos and E. J. Mittemeijer, Int. Mater. Rev., 2007, 52, 193–212.
33. K. Liu, Q. Liang, M. Qin, D. Shen, H. Yin, Z. Ren, Y. Zhang, H. Zhang, P. W. K. Fong, Z. Wu, J. Huang, J. Hao, Z. Zheng, S. K. So, C.-S. Lee, X. Lu and G. Li, Joule, 2020, 4, 2404–2425.
34. J. Zhu, Y. Qian, Z. Li, O. Y. Gong, Z. An, Q. Liu, J. H. Choi, H. Guo, P. J. Yoo, D. H. Kim, T. K. Ahn, G. S. Han and H. S. Jung, Adv. Energy Mater., 2022, 12, 2200632.
35. Y. Li, Z. Chen, B. Yu, S. Tan, Y. Cui, H. Wu, Y. Luo, J. Shi, D. Li and Q. Meng, Joule, 2022, 6, 676–689.
36. Q. Jiang, Z. Chu, P. Wang, X. Yang, H. Liu, Y. Wang, Z. Yin, J. Wu, X. Zhang and J. You, Adv. Mater., 2017, 29, 1703852.
37. F. Ye, J. Ma, C. Chen, H. Wang, Y. Xu, S. Zhang, T. Wang, C. Tao and G. Fang, Adv. Mater., 2020, 33, 2007126.
38. H. Wang, F. Ye, J. Liang, Y. Liu, X. Hu, S. Zhou, C. Chen, W. Ke, C. Tao and G. Fang, Joule, 2022, 6, 1–16.
39. 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.
40. D. Kim, J.-H. Yun, M. Lyu, J. Kim, S. Lim, J. S. Yun, L. Wang and J. Seidel, J. Phys. Chem. C, 2019, 123, 14144–14151.
41. J. Liang, X. Hu, C. Wang, C. Liang, C. Chen, M. Xiao, J. Li, C. Tao, G. Xing, R. Yu, W. Ke and G. Fang, Joule, 2022, 6, 816–833.
42. J. Wang, H. Ma, A. Wang, Z. Li, J. Dong, Y. Liu, S. Gao, S. Yan, X. Chen, Y. Li, Z. Wu, W. Xu, F. Liu, F. Wang, W. Huang and T. Qin, Matter, 2022, 5, 1–16.
43. R. Wang, J. Xue, K. Wang, Z. Wang, Y. Luo, D. Fenning, G. Xu, S. Nuryyeva, T. Huang, Y. Zhao, L. Yang Jonathan, J. Zhu, M. Wang, S. Tan, I. Yavuz, N. Houk Kendall and Y. Yang, Science, 2019, 366, 1509–1513.
44. H. Zhu, Y. Ren, L. Pan, O. Ouellette, F. T. Eickemeyer, Y. Wu, X. Li, S. Wang, H. Liu, X. Dong, S. M. Zakeeruddin, Y. Liu, A. Hagfeldt and M. Grätzel, J. Am. Chem. Soc., 2021, 143, 3231–3237.
45. Q. Jiang, Z. Ni, G. Xu, Y. Lin, P. N. Rudd, R. Xue, Y. Li, Y. Li, Y. Gao and J. Huang, Adv. Mater., 2020, 32, 2001581.
46. H. Liu, Z. Zhang, Z. Su, W. Zuo, Y. Tang, F. Yang, X. Zhang, C. Qin, J. Yang, Z. Li and M. Li, Adv. Sci., 2022, 9, 2105739.
47. J. Qing, X.-K. Liu, M. Li, F. Liu, Z. Yuan, E. Tiukalova, Z. Yan, M. Duchamp, S. Chen, Y. Wang, S. Bai, J.-M. Liu, H. J. Snaith, C.-S. Lee, T. C. Sum and F. Gao, Adv. Energy Mater., 2018, 8, 1800185.
48. J. Zhang, J. Qin, M. Wang, Y. Bai, H. Zou, J. K. Keum, R. Tao, H. Xu, H. Yu, S. Haacke and B. Hu, Joule, 2019, 3, 3061–3071.
49. Y. Cai, J. Cui, M. Chen, M. Zhang, Y. Han, F. Qian, H. Zhao, S. Yang, Z. Yang, H. Bian, T. Wang, K. Guo, M. Cai, S. Dai, Z. Liu and S. Liu, Adv. Funct. Mater., 2021, 31, 2005776.
50. W. Shao, H. Wang, F. Ye, C. Wang, C. Liu, S. Zhou, C. Tao and G. Fang, J. Power Sources, 2022, 520, 230900.
51. Y. Zhang, Y. Wang, L. Zhao, X. Yang, C.-H. Hou, J. Wu, R. Su, S. Jia, J.-J. Shyue, D. Luo, P. Chen, M. Yu, Q. Li, L. Li, Q. Gong and R. Zhu, Energy Environ. Sci., 2021, 14, 6526–6535.
52. T. Li, S. Wang, J. Yang, X. Pu, B. Gao, Z. He, Q. Cao, J. Han and X. Li, Nano Energy, 2021, 82, 105742.
53. L. Yang, J. Feng, Z. Liu, Y. Duan, S. Zhan, S. Yang, K. He, Y. Li, Y. Zhou, N. Yuan, J. Ding and S. Liu, Adv. Mater., 2022, 34, 2201681.
54. Z. Xiong, X. Chen, B. Zhang, G. O. Odunmbaku, Z. Ou, B. Guo, K. Yang, Z. Kan, S. Lu, S. Chen, N. A. N. Ouedraogo, Y. Cho, C. Yang, J. Chen and K. Sun, Adv. Mater., 2021, 34, 2106118.
55. S. Zhang, L. Zhang, Q. Tian, X. Gu, Y. Du, K. Zhao and S. Liu, Adv. Energy Mater., 2022, 12, 2103007.
56. Y. Wang, W. Li, Y. Yin, M. Wang, W. Cai, Y. Shi, J. Guo, W. Shang, C. Zhang, Q. Dong, H. Ma, J. Liu, W. Tian, S. Jin, J. Bian and Y. Shi, Adv. Funct. Mater., 2022, 32, 2204831.
57. L. Zhu, X. Zhang, M. Li, X. Shang, K. Lei, B. Zhang, C. Chen, S. Zheng, H. Song and J. Chen, Adv. Energy Mater., 2021, 11, 2100529.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d2ee03342a

---