Lead and Iodide Fixation by Thiol Copper(II) Porphyrin for Stable and Environmental-Friendly Perovskite Solar Cells

Open AccessCCS ChemistryRESEARCH ARTICLE1 Oct 2021Lead and Iodide Fixation by Thiol Copper(II) Porphyrin for Stable and Environmental-Friendly Perovskite Solar Cells Guo-Bin Xiao†, Lu-Yao Wang†, Xi-Jiao Mu†, Xiao-Xin Zou, Yi-Ying Wu and Jing Cao Guo-Bin Xiao† State Key Laboratory of Applied Organic Chemistry, Key Laboratory of Nonferrous Metal Chemistry and Resources Utilization of Gansu Province, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou 730000 †G.-B. Xiao, L.-Y. Wang, and X.-J. Mu contributed equally to this work.Google Scholar More articles by this author , Lu-Yao Wang† School of Materials Science and Engineering, Shanghai Jiao Tong University, Shanghai 200240 †G.-B. Xiao, L.-Y. Wang, and X.-J. Mu contributed equally to this work.Google Scholar More articles by this author , Xi-Jiao Mu† State Key Laboratory of Applied Organic Chemistry, Key Laboratory of Nonferrous Metal Chemistry and Resources Utilization of Gansu Province, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou 730000 †G.-B. Xiao, L.-Y. Wang, and X.-J. Mu contributed equally to this work.Google Scholar More articles by this author , Xiao-Xin Zou State Key Laboratory of Inorganic Synthesis and Preparative Chemistry, College of Chemistry, Jilin University, Changchun 130012 Google Scholar More articles by this author , Yi-Ying Wu Department of Chemistry and Biochemistry, The Ohio State University, Columbus, OH 43210 Google Scholar More articles by this author and Jing Cao *Corresponding author: E-mail Address: [email protected] State Key Laboratory of Applied Organic Chemistry, Key Laboratory of Nonferrous Metal Chemistry and Resources Utilization of Gansu Province, College of Chemistry and Chemical Engineering, Lanzhou University, Lanzhou 730000 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202000516 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail The formation of undersaturated lead or iodide ions and I2 on the perovskite surface can decrease the performance and stability of perovskite solar cells (PSCs). Additionally, the leakage of noxious lead limits the application of PSCs. Here, we develop a strategy for molecular modulation of a perovskite surface using thiol copper(II) porphyrin (CuP) to post-treat the perovskite film. The Raman spectra reveal that the CuP molecules anchor on the perovskite surface by a coordination interaction between the thiol-terminal of CuP and Pb ion in perovskite. DFT calculations demonstrate that the porphyrin core in CuP has a strong fixing effect on I− and I2 by the induction force and electrostatic attraction. Such a fixation was demonstrated by the shift of the UV absorption peak of I2 in solution with the addition of CuP and the decreased defects density in the CuP-treated perovskite film. Finally, the modified PSCs exhibit an improved cell performance with the best efficiency up to 21.76% (certified 20.97%). The modified PSCs acquired significantly improved stability. In addition, there is almost no lead leakage from the treated film immersed in water. This work provides a surface modulation strategy by thiol CuP treatment to fabricate efficient, stable, and environmental-friendly PSCs. Download figure Download PowerPoint Introduction Organic–inorganic hybrid perovskite materials have attracted tremendous attention owing to their low fabrication cost and intriguing optical and electronic properties.1–6 The recently reported power conversion efficiency (PCE) of perovskite solar cells (PSCs) has exceed 25%.7 However, during the fabrication of polycrystalline perovskite films under the solution-based preparation process, the inevitably formed large number of undersaturated metal or halide ions at the surface and grain boundaries (GBs) of perovskite film can act as ion migration paths.8–10 Additionally, the I− is easily oxidized to I2 under environment condition, which can further cause chemical chain reactions to result in the degradation of perovskite film.11,12 The leakage of noxious lead also limits the practical application of PSCs.13,14 Thereby, effective modulation of perovskite surface and GBs of perovskite film is vital for fabricating efficient, stable, and environmental-friendly PSCs. Recently, a series of modified approaches have been developed to address the above issues.15–21 For instance, employing sulfate or phosphate ions to treat the surface of perovskite film yielded the lead oxysalt thin capping layer, thereby remarkably improving the environmental stability of PSCs.22 The usage of phenethylammonium iodide to modify the perovskite films realized the fabrication of higher-efficiency PSCs by reducing the defects and suppressing nonradiative recombination.23 Introducing other halide anions (such as F−, Cl−, or Br−) into perovskite film increased the lattice strain relaxation of perovskite, to improve the long-term durability of PSCs.24–26 In our recent work, the ammonium/amine porphyrin or phthalocyanine macromolecules with excellent photoelectric properties was used to treat the perovskite film to obtain efficient surface and GBs modification and interfacial charge transfer, thus significantly enhancing the performance and stability of PSCs.3,12,27 Since the defects mainly locate at the top surface of the perovskite film,10,28 it is still a challenge to effectively encapsulate and modulate the perovskite surface to fabricate the efficient, stable, and environmental-friendly PSCs. Here, by optimizing the structures of porphyrin molecules, we develop a strategy for molecular modulation of the perovskite surface using a thiol copper(II) porphyrin (CuP, shown in Figure 1a) to post-treat the perovskite film (Figure 1b). The thiol was selected as the peripheral functional group to firmly bind the CuP molecules on the surface of perovskite film by forming interfacial Pb–S bonds, thus to effectively passivate the surficial uncoordinated Pb ions and prevent lead leakage.29–31 Additionally, the Cu(II) ions in CuP can efficiently tune the electron distribution of porphyrin π ring, to achieve the effective fixation of I− and I2 on the porphyrin core in CuP by the induction force and electrostatic attraction, thus reducing the defects within the perovskite film. Finally, the PSCs with CuP modification revealed the significantly enhanced cell performance, and the best efficiency was up to 21.76% (certified 20.97%).The introduction of CuP produced efficient surface modulation and encapsulation of perovskite film, which further improves the cell stability and safety. Figure 1 | (a) Structures of thiol porphyrins. (b) Scheme illustration of PSCs with CuP post-treatment to achieve the interfacial modulation. Download figure Download PowerPoint Experimental Methods Materials Chemicals: PbI2 (TCI, Tokyo, Japan), PbBr2 (TCI), CsI (Xi’an Polymer Light Technology Corp., Xi’an, China), FAI (Xi’an Polymer Light Technology Corp.), methylammonium bromide (MABr; Xi’an Polymer Light Technology Corp.), Cu(OAc)2 (Alfa Aesar, Shanghai, China), Zn(OAc)2•2H2O (Alfa Aesar), toluene (J&K, Beijing, China), chlorobenzene (J&K), N,N-dimethylformamide (J&K), dimethyl sulfoxide (J&K), 2-propanol (J&K), acetonitrile (J&K), Li-TFSI (J&K), TBP (J&K), Spiro-OMeTAD (Xi’an Polymer Light Technology Corp.). The methylammonium iodide (MAI) was synthesized and purified according to the literature.32 The thiol porphyrin was prepared by the reported method.33 The syntheses of ZnP and CuP are supplied in the Supporting Information Scheme S1. Other solvents were purchased from chemical companies and used without further purification. Solar cell fabrication Initially, the fluorine-doped tin oxide (FTO) glass substrates (2.0 cm × 2.0 cm) were etched with zinc powder and 2 M HCl to obtain specific electrode patterns, and then the substrates were sequentially cleaned with acetone, distilled water, and ethanol ultrasonic baths. The compact TiO2 layer was prepared with 0.15 M titanium tetraisopropanolate isopropanol solution by spin-coating onto the FTO substrate at a spin speed of 2500 rpm for 25 s, and then the substrate was annealed at 550 °C for 30 min. After cooling to room temperature, the substrates were dipped into TiCl4 solution (20 mM) at 70 °C for 20 min. Then, the mesoporous TiO2 layer was spin-coated with TiO2 paste (Dyesol DSL 18NR-T) in isopropanol (1∶7, mass ratio) at 5000 rpm for 25 s, and then annealed at 550 °C for 30 min. The Cs0.05(MA0.17FA0.83)0.95Pb(I0.83Br0.17)3 perovskite precursor solution was prepared with the reported ratio.34 The perovskite film was formed by spin-coating the precursor solution on the mesoporous TiO2 at 4000 rpm for 25 s, and then 0.5 mL of diethyl ether was quickly dripped on the rotating surface at 6 s, followed by annealing at 70 °C for 1 min and at 120 °C for 2 min. For the post-treatment of perovskite film, the porphyrin in toluene solution was spin-coated on the perovskite film at 3000 rpm for 20 s. Subsequently, the Spiro-OMeTAD [Spiro-OMeTAD/chlorobenzene (CB) (72 mg/mL) containing 17.5 μL lithium bis(trifluoromethanesulphonyl)imide (Li-TFSI) in acetonitrile (520 mg/mL) and 28.8 μL 4-tert-butylpyridine (TBP)] was spin-coated on the top of perovskite film at 3000 rpm for 30 s as the hole transport material (HTM). Finally, 80 nm gold electrode was deposited by thermal evaporation. Device characterization The current density–voltage (J–V) tests were recorded by using a solar simulator equipped with a 300 W collimated Xenon lamp (Newport) and Keithley 2400 source meter (Keithley, Cleveland, OH). The light intensity of the simulator was calibrated by the certified silicon solar cell to 100 mW/cm2 at air mass (AM) 1.5 G solar light condition. Incident photon-to-electron conversion efficiency (IPCE) was characterized on a computer-controlled IPCE system with a Xenon lamp, a Keithley multimeter and a monochromator. The device was calibrated with the certified silicon solar cell and the IPCE data were collected in direct current (DC) mode. The X-ray diffraction (XRD) data were acquired on Cu Kα radiation on a Rigaku RINT-2500 X-ray diffractometer (Rigaku, Tokyo, Japan). The surface morphology was recorded on a SEM-4800 field emission scanning electron microscope (FESEM; Hitachi, Tokyo, Japan). UV–vis absorption spectra were measured on a Cary-5000 UV–vis spectrophotometer (Agilent, Santa Clara, CA). The steady and time-resolved photoluminescence (PL) spectra were recorded on a Horiba FL-3 spectrometer (Horiba, United States). The cyclic voltammetry (CV) and I–V measurements for the space-charge-limited current (SCLC) analysis were prepared by an electrochemical workstation (CHI660E; ChenHua, China). The methods of theoretical calculation are provided in the Supporting Information. Surficial work function of the perovskite films was performed by the scanning Kelvin probe microscopy (SKPM) with an atomic force microscope (AFM) (Agilent SPM 5500; Agilent). Results and Discussion Perovskite surface modification with CuP by forming interfacial Pb–S bonds The synthesis of CuP is provided in the Supporting Information. For the CuP modification process in PSCs, the CuP varying in concentration from 0 to 3 mg/mL in toluene was used to post-treat the perovskite film. The optimized CuP concentration of 2 mg/mL was selected to prepare the following measurements. As shown in Supporting Information Figure S1, SEM images revealed a changed morphology of the CuP-treated perovskite film. This result was further observed in the AFM measurements. As shown in Figures 2a and 2b, the introduced CuP roughens the perovskite surface. A root mean square (RMS) analysis was performed and shown in Figures 2c and 2d. The value of the perovskite film was 6.69 nm, while the value of CuP-treated film was 5.67 nm. Since the capping layer was not observed in the SEM and AFM images, we concluded the self-assembly monolayer porphyrins may be anchored on the perovskite surface by the coordination effect of Pb from perovskite and S in porphyrin. The appearance of Cu peaks at 934.9 and 954.9 eV for the CuP-treated perovskite sample from X-ray photoelectron spectroscopy (XPS) measurement indicates the existence of CuP within the treated perovskite film (see Supporting Information Figure S2). The shift of the S peak in the modified film (Figure 2e) is evidence for the coordination between Pb in perovskite and S from CuP. Figure 2 | AFM images of perovskite films without (a and c) and with CuP (b and d) post-treatment. XPS spectra for elemental S (e) and Raman spectra (f) of CuP and perovskite with and without CuP treatment. Download figure Download PowerPoint Considering that it is difficult for the large CuP to enter the 3D perovskite lattice, the introduced CuP molecules should stay on the surface of the perovskite film. This hypothesis was reflected by the similar optical absorption spectra and XRD patterns (see Supporting Information Figure S3) of the perovskite films with and without CuP treatment. To understand how CuP functionalizes the perovskite film surface, Raman spectroscopy tests were performed and shown in Figure 2f. The appearance of new peaks for the CuP-treated perovskite sample at 415, 748, and 816 cm−1 assigned to the Pb–S bond35–37 indicates that the thiol in CuP can react with the surficial Pb ion from the perovskite film. It can be concluded that the CuP molecules anchor on the perovskite surface by forming interfacial Pb–S bonds between perovskite and CuP. The cross-sectional elemental maps of perovskite with CuP post-treatment further confirmed the CuP mainly located on the top surface of the perovskite film (see Supporting Information Figure S4). Effect of the introduced CuP on the interfacial property of perovskite film Effective modulation of the interfacial I− and I2 defects of perovskite film is vital for the fabrication of efficient and stable PSCs. Studies revealed that the I2 defects within perovskite film could be reduced or captured by aromatic rings,6 thus we speculated that the porphyrin ring would have a similar property. The Cu(II) ion with the unoccupied d orbital would tune the electron distribution of the porphyrin ring in CuP to capture and immobilize I− and I2 defects within perovskite films. To investigate the above hypothesis, the free-base porphyrin (P) and zinc porphyrin (ZnP) were also designed as control samples. Electrostatic potential images of the porphyrin molecules were prepared by density functional theory (DFT) calculation and shown in Supporting Information Figure S5. We can clearly observe that the introduction of metal ions in porphyrin cores results in an increased central charge intensity within the porphyrin core compared with the P, which probably facilitates the adsorption of I− and I2. The change of the Gibbs free energy (ΔGgas, kcal/mol) for I− and I2 adsorbed on the central porphyrin ring in different molecules bound on perovskite surfaces was also calculated (Figure 3a and Supporting Information Figure S6). As listed in Table 1, the metal porphyrin-based systems revealed a decreased ΔGgas to fix I− and I2 compared with the P. The results suggest the central metal ions can tune the electron distribution within the porphyrin core to fix I− and I2 on the central porphyrin rings. Figure 3 | (a) Schematic diagram of stabilized structure. The BOMD intermolecular interaction potential energy between porphyrins and I− (b) and I2 (c). (d) The vdW potential isosurface (blue surface, isovalue = 0.05) and attractors (yellow balls). (e) IGM isosurface (isovalue = 0.003) of I− and I2 fixed on the central porphyrin rings in different molecules. Download figure Download PowerPoint Table 1 | Variation of Gibbs Free Energy (ΔGgas, kcal/mol) of I− and I2 Fixed on the Central Porphyrin Rings in Different Molecules by DFT Calculation P ZnP CuP I− 6.47 −6.10 −12.95 I2 −1.16 −4.11 −3.91 We used the Born–Oppenheimer molecular dynamics (BOMD) method to further dynamically analyze the movement law of I− and I2 on the porphyrin molecules bound on the surface of perovskite lattices. The stable equilibrium structures were selected in the simulation process and shown in Figure 3a. The more negative interaction energy of the CuP-based system is observed (Figures 3b and 3c), indicating that the CuP-based system has the strongest fixing effect on the I− and I2 at room temperature. Generally, the van der Waals (vdW) potential reflects well the attraction of different atoms to specific atoms, and the effect of the vdW potential is dominant for the interaction between nonpolar molecules. To deeply understand the mechanism of the porphyrin system’s fixation effect on I− and I2, we conducted the vdW potential and independent gradient model (IGM) analyses of the stable structures during the BOMD processes. The mathematical definition of vdW potential is38: V vdW ( r ) = V repul ( r ) + V disp ( r ) = ∑ A ɛ AB ( R AB 0 | R A − r | ) 12 + ∑ A [ − 2 ɛ AB ( R AB 0 | R A v − r | ′ ) 6 ] where the Vrepul( r) and Vdisp( r) are the exchange mutual exclusion potential and dispersion attraction potential in vdW potential, which produce positive contribution and negative contribution, respectively. All atoms in the A cycle, ɛ, and R AB 0 are the depth and equilibrium distances of the Lennard–Jones potential between atoms; RA represents the nuclear coordinates of the A atom. In the molecular force field, the vdW parameters of different atoms may be distinct, thereby both ɛ and R AB 0 are atom dependent. In the vdW potential as defined above, the B atom is equivalent to the probe atom. The iodine atom was used as a probe in our calculation, which is in accordance with the actual situation. As shown in Figure 3d, the vdW potential distribution of I− is spherically symmetric, and the distribution of attractors is vacant perpendicular to the paper surface. This is due to the exposure of the d-orbital in the I atom. The vdW potential isosurface near the porphyrin ring has a notch partly circular, while the vdW potential isosurface of I− and I2 has a closed surface. Therefore, I− and I2 can be attracted to the porphyrin ring. There are more attractors near the porphyrin ring with a metal core, which can help fix the I− and I2. The IGM isosurface analyses are shown in Figure 3e. The green and blue isosurface represent vdW and electrostatic interaction, respectively.39 The size of the isosurface reflects the size of attraction. The CuP-based system reveals the larger attractive forces to the I− and I2 than those of P- or ZnP-based systems, indicating that the CuP bound on perovskite surface can more effectively fix the I− and I2 than the P or ZnP. The colored IGM isosurface indicates there are two types of interactions between the metal porphyrin ring and I− and I2 (Figure 3e). One part is the green vdW interaction, and the other part is the blue electrostatic interaction. To quantitatively analyze the relationship between the electrostatic attraction and vdW interaction, we used the symmetric-adapted perturbation theory (SAPT) method to quantitatively calculate the intermolecular interaction energy between CuP and I− or I2. As shown in Table 2, it was found that the main interaction energy is the inducing force (−175.30 and −181.33 kcal/mol to I− and I2, respectively). While the electrostatic interaction energy also reaches −80.44 and −89.98 kcal/mol for I− and I2, respectively, an interaction energy component that cannot be ignored. Therefore, it can be concluded that the fixing effect of I− and I2 on CuP is dominated by the induction force and supplemented by electrostatic attraction, thus effectively modulating the perovskite surface. Table 2 | Intermolecular Interaction Energy (kcal/mol) of I− and I2 Fixed on the Central Porphyrin Ring in CuP by SAPT Calculation CuP-I− CuP-I2 Electrostatics −80.44 −89.98 Exchange 33.28 45.86 Induction −175.30 −181.33 Dispersion −7.49 −6.17 To experimentally confirm the above results, the P and ZnP (Figure 1a) were synthesized and used to post-treat the perovskite films (see Supporting Information Figures S7 and S8). The synthetic processes of P and ZnP are provided in the Supporting Information. To demonstrate that CuP has an effective fixing ability on I2, the UV–vis absorption spectra for I2 in CH2Cl2/ethyl acetate (V/V = 1/1) solution with the addition of P, ZnP, and CuP were tested. As shown in Supporting Information Figure S9, the absorption peak of I2 was blue-shifted with the addition of CuP. In contrast, there was almost no change with the introduction of P or ZnP. These results indicate the CuP can effectively fix the I2. Confocal laser scanning fluorescence microscopy (CLSFM) measurements were prepared to study the microstructure characteristic relationships of perovskite films with and without treatment. As illustrated in Figures 4a and 4b and Supporting Information Figure S10, the surface of pristine perovskite film exhibited substantial local PL heterogeneity, while an obviously brighter region in the CuP-modified perovskite film was observed, suggesting that the CuP post-treatment could activate partial dark regions at the perovskite film surface, which was consistent with previous reports.40–42 The steady and transient PL measurements were performed to confirm the above results. Compared with the perovskite films with and without P or ZnP modification, the CuP-based perovskite film revealed an enhanced steady-state PL intensity (see Supporting Information Figure S11a). The obtained decay time of transient PL for the CuP-based film (181.6 ns) was also larger than those of perovskite films with and without P or ZnP treatment (36.7, 85.2, and 88.8 ns for the pure perovskite film, P- and ZnP-modified perovskite films, as shown in Figure 4c and Supporting Information Figure S11b). The SCLC tests were further used to evaluate quantitatively the defects’ densities of perovskite films. The trap-state density (nt) was calculated according to the relation: nt = (2ɛɛ0VTFL)/(eL2), where ɛ is the dielectric constant of perovskite, ɛ0 is the vacuum permittivity, L is the thickness of perovskite film, and e is the elementary charge.43–46 As shown in Figure 4d and Supporting Information Figure S12b, the nt for CuP-treated perovskite films was determined to be 1.40 × 1016 cm−3, which was lower than those of the perovskite without (4.21 × 1016 cm−3) and with P (2.71 × 1016 cm−3) and ZnP (2.18 × 1016 cm−3) treatment (see Supporting Information Figure S12b). The above results clearly confirm that introduction of CuP on the surface of perovskite film can effectively reduce the defects at the surface of perovskite film. Figure 4 | CLSFM images of perovskite films without (a) and with (b) CuP treatment. Transient PL spectra (c) and SCLC data (d) of perovskite film without and with CuP post-treatment. (e) Energy band structure of TiO2, perovskite, perovskite treated by CuP and Spiro-OMeTAD. (f) Best J–V data of PSCs with and without CuP post-treatment tested in reverse scan (RS) and forward scan (FS). Download figure Download PowerPoint The interfacial energy level distribution between perovskite and modifier is crucial for the interfacial charge extraction and transport. We employed CV to determine the energy levels of P, ZnP, and CuP. As shown in Supporting Information Figure S13, the highest occupied molecular orbital (HOMO) levels of P, ZnP, and CuP locate at −4.79, −4.88, and −4.96 eV, respectively. These results indicate the CuP HOMO is closer in energy level to the Spiro-OMeTAD HOMO (−5.22 eV) to realize the charge transfer and enhance the cell efficiency. To confirm the energy level of CuP coordinated on the surface of perovskite film, the surficial work function of the perovskite films without and with CuP modification were analyzed via SKPM measurements. As illustrated in Supporting Information Figures S14 and S15, the contact potential difference (CPD) of the pure perovskite film was ∼350 mV, while the CPD of CuP-modified perovskite film changed to ∼550 mV. The results suggest the surficial work function of perovskite modified by CuP is adjusted to −5.23 eV, slightly lower than the HOMO level of Spiro-OMeTAD (Figure 4e). This indicates the introduction of CuP on the perovskite surface can facilitate the efficient charge transfer between perovskite and Spiro-OMeTAD, probably enhancing the efficiency of PSCs. Effect of CuP on device efficiency Devices with the structure of FTO/compact TiO2/mesoporous- TiO2/Cs0.05(MA0.17FA0.83)0.95Pb(I0.83Br0.17)3/Spiro-OMeTAD/Au were fabricated to evaluate the effect of CuP on the corresponding PSCs (see Supporting Information Figure S16). To assess whether the CuP on the surface of the perovskite layer was dissolved during the deposition of the HTM layer, we employed chlorobenzene (CB) to treat the CuP-modified perovskite film. Raman spectroscopy of the films with and without CB-treatment was then conducted. As shown in Supporting Information Figure S17, the films with and without CB treatment have similar spectra. This result confirms the successful fabrication of PSCs with CuP-modified perovskite in this work. The cell performances of PSCs were then evaluated and shown in Figure 4f, Supporting Information Figure S18, and Table S1. The PSCs’ efficiencies improved as the content of CuP increased from 0 to 3 mg/mL. The best efficiency of the CuP-modified PSC (2 mg/mL post-treatment) was 21.76%, which was further certified as 20.97% cell performance (see Supporting Information Figure S19). In contrast, the best efficiency of the PSC with pure perovskite was 20.01%. A remarkably reduced hysteresis of modified PSC with CuP was also observed (Figure 4f and Supporting Information Table S2). Measurements over 30 cells modified with CuP gave an average efficiency of 21.0 ± 0.8%. The efficiencies of control PSCs averaged 19.1% ± 1.0% (see Supporting Information Figure S20). In addition, the cell performances of PSCs with P and ZnP were also tested. The results revealed that the best performances of PSCs with P and ZnP were 20.76% and 21.12%, respectively (see Supporting Information Figure S21 and Table S4), which were lower than that of PSCs with CuP. Accordingly, the introduction of CuP on the perovskite surface demonstrated effective interfacial modulation to reduce the interfacial defects, and thus remarkably improved the performance of PSCs. Effect of CuP on device stability As shown in Supporting Information Figure S22, the contact angle of a water droplet on the CuP-modified perovskite film obviously increased when compared to that of the pure perovskite film. The water-resistance properties of perovskites films with and without CuP modification were further assessed by immersing the films into water solution. As shown in Figures 5a–5c, the pure perovskite film quickly turned yellow after 2 s, but the perovskite film treated with CuP remained black after 25 s. Since the PSCs with Spiro-OMeTAD as HTM have poor moisture and thermal stability (see Supporting Information Figure S23),47 the long-term stabilities of PSCs with the structure of FTO/compact TiO2/mesoporous-TiO2/perovskite/poly(3-hexylthiophene) (P3HT)/Au were performed (see Supporting Information Figure S24 and Table S3, all PSCs without encapsulation in this work) to assess the effect of CuP on the stability of perovskite film. As shown in Figure 5d, after 3000 h with a humidity of 45%, the CuP-modified device retained 90% of the original efficiency. The reference PSC almost failed under the same conditions. These results indicate the CuP achieves the effective encapsulation of the perovskite film, thereby impeding the permeation of water into perovskite film. Figure 5 | Images of perovskite films without (left) and with (right) CuP post-treatment immersed in water solution for 2 s (a), 25 s (b), and 35 s (c). (d) Moisture stability carried out under the humidity of 45%. (e) Thermal stability measured at 85 °C in N2 environment. (f) Photostability tested at AM 1.5 G illumination in N2 atmosphere at open-circuit condition. (g) Stabilized power output of PCE at maximum power point under AM 1.5 G condition. Download figure Download PowerPoint The thermal stabilities of PSCs were further performed at 85 °C in N2 atmosphere. The device with the CuP modification reta