Origin of VOC in Perovskite Solar Cells by Faradaic Junction Model

Origin of VOC in Perovskite Solar Cells by Faradaic Junction Model | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Origin of VOC in Perovskite Solar Cells by Faradaic Junction Model Wenjun Luo, Mengfan Xue, Zhoujun Li, Qiong Wang, Yihong Chen, and 8 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6295095/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 19 Oct, 2025 Read the published version in Science China Chemistry → Version 1 posted You are reading this latest preprint version Abstract Perovskite solar cells have attracted much interest due to the very fast increasement of power conversion efficiency (PCE) as well as the low-cost solution processing, excellent absorption coefficient, long charge carrier diffusion length. To date, further improvement of PCE in perovskite solar cells is mainly limited by the open-circuit voltage (V OC ). In previous studies, the origin of V OC is usually explained by the energy band alignment theory. However, in some experiments, the V OC does not change apparently when the band positions of electron or hole transport layers are changed. Therefore, the energy band alignment theory is not suitable to explain the origin of V OC in perovskite solar cells, and it’s desirable to reveal the origin of V OC to further improve the PCE of perovskite solar cells. Here, we develop a new method to simulate the interface charge transfer process in the all solid-state devices in working conditions by connecting perovskite/ETL half device with perovskite/HTL half device under simultaneous illumination. Reversible reduction reactions of Pb 2+ /Pb 0 and oxidation reactions of I - /I 2 are observed on MAPbI 3 surfaces under illumination, respectively. Moreover, the oxidation processes of Spiro-OMeTAD are also investigated by in situ FTIR and NMR. We find that the V OC comes from the difference between the electrode potentials of Pb 2+ /Pb 0 and X - /X 2 (X=I, Br) on perovskite surfaces. The results suggest that the origin of V OC does not come from the energy band alignment but from the electrode potential alignment, which can offer new perspectives to improve the V OC of perovskite solar cells. Physical sciences/Chemistry/Photochemistry/Solar cells Physical sciences/Materials science/Materials for energy and catalysis/Solar cells Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Introduction A typical ABX 3 perovskite consists of methylammonium (MA + ), formamidinium (FA + ) or Cs + ions on the A site, Pb 2+ or Sn 2+ ions on the B site, and halide ions (I - , Br - , Cl - ) on the X site 1,2 . Since 2009, perovskite solar cells have been a very promising technique for solar energy conversion due to its low cost and high power conversion efficiency (PCE), close to commercial Si solar cells 3-5 . The PCE of a solar cell is usually determined by open-circuit voltage (V OC ), short-circuit current (I SC ) and fill factor (FF). To date, I SC and FF of perovskite solar cells have been close to the theoretical limits. Therefore, it is the predominant way to increase the V OC to further improve the PCE 6 . In previous studies, great efforts, such as surface passivation, ion doping and crystal quality optimization, have been made to improve the V OC . A V OC of 1.26 V is obtained in a perovskite solar cell with a bandgap of 1.60 eV, in which a V OC loss (E g /q - V OC ) is 0.34 V 7 . Wide-bandgap perovskites (E g > 1.7 eV) are usually used in multi-junction solar cells by adjusting the ratio of halide ions (I - /Br - ) on the X site 8,9 . However, a much higher V OC loss of 0.69 V is observed in the perovskite solar cell with a bandgap of 2.0 eV, which limits the applications of wide-bandgap perovskites in multi-junction perovskite solar cells 10 . In a typical perovskite solar cell, both electron transport layer (ETL) and hole transport layer (HTL) are introduced to improve the separation efficiency of the photo-generated carriers in the perovskite light absorber. According to energy band alignment theory, after the perovskite light absorber is contacted to charge transport layers, new equilibrium of the fermi level is established between the perovskite and the charge transport layers. When the perovskite is illuminated, the V OC in the perovskite solar cell is given by the difference between the quasi-Fermi levels of electrons at the electron transport layer and the quasi-Fermi level of holes at the hole transport layer 11,12 . Therefore, the V OC of the perovskite solar cells depend on the band positions of ETL and HTL, and the regulation of energy levels of charge transport layers can enhance the V OC of the perovskite solar cell in some studies 13-16 . However, in some experiments, the V OC is similar even when ETL with different conduction bands or HTL with different highest-occupied molecular orbitals (HOMO) are used 1,17 . Moreover, in the TiO 2 /perovskite/carbon solar cell, the difference between the conduction band of TiO 2 (-4.0 eV~ -4.2 eV) and the work function of carbon (-5.0 eV) is about 0.8~1.0 eV. However, the V OC of CsPbBr 3 and MAPbBr 3 solar cells is about 1.5~1.6 V. A large discrepancy of ~0.6 V can be observed 18-20 . These results are not consistent with the theoretical V OC by the band alignment theory. Therefore, it is significant to understand the origin of V OC in perovskite solar cells. Different from electron transfer in a conventional solar cell model, ion migration is observed in perovskite solar cells 21,22 . A new coupled electron and ion transfer model, which describes chemical reactions at semiconductor interfaces (faradaic junction model), has been proposed to understand interface charge transfer at semiconductor/liquid interface recently 23-26 . Different from built-in electric field in a physical junction, the driving force for charge separation in a faradaic junction is the difference of electrode potentials (molar Gibbs free energy) at the interface. Is it possible that the faradaic junction charge transfer mechanism also exists in such all-solid-state devices? Moreover, in previous studies, faradaic reactions are only observed on the surfaces of oxides, such as TiO 2 , WO x and MnO x , but not on the surfaces of non-oxides 25,26 . To date, it is still a key challenge to characterize the faradaic reactions and interface charge transfer in all-solid-state perovskite solar cells, especially in working conditions. In this work, we developed a new method to simulate the charge transfer in a simulated full device by connecting a perovskite/TiO 2 half device with a perovskite/Spiro-OMeTAD half device under simultaneous illumination. By in situ XPS, in situ FTIR and other characterization methods, we found that the faradaic reactions indeed exist on the surface of perovskites, TiO 2 and Spiro-OMeTAD, as well as the interfaces between perovskite/TiO 2 and perovskite/Spiro-OMeTAD under illumination. Moreover, we measured electrode potentials of perovskites, TiO 2 and Spiro-OMeTAD to understand the V OC of perovskite solar cells. The results suggest that faradaic junction model is more suitable to describe the charge transfer in perovskite solar cells than the band alignment theory, which offers new concepts to improve V OC and PCE of perovskite solar cells by adjusting electrode potentials of perovskites. Charge transfer at MAPbI 3 /TiO 2 interface A TiO 2 layerwas firstly prepared on FTO substrates and then a MAPbI 3 layer was prepared on the surface of the TiO 2 layer. In situ X-ray Photoelectron Spectroscopy (XPS) was used to investigate the charge transfer at MAPbI 3 /TiO 2 interface under illumination. In a normal solar cell, the perovskite layer is usually dense and covers the TiO 2 bottom layer completely, which hinders the detection of XPS signal of the TiO 2 layer. In order to detect the XPS signal of MAPbI 3 and TiO 2 at the same time, a sparse MAPbI 3 layer was prepared on the FTO substrate in this study. The as-prepared sparse MAPbI 3 layer indicates the same crystal structure and optical property with a dense MAPbI 3 (Figure S1 and S2). To further investigate the charge transfer at the interface between the perovskite and the contact layer, we constructed TiO 2 /dense MAPbI 3 and TiO 2 /sparse MAPbI 3 heterojunctions (Figure S3). The linear sweep voltammetry (LSV) curves were measured and the results are shown in Figure S4. The two heterojunctions indicate also the same onset potentials under illumination, which suggest that similar interface charge behaviour happens in the dense and sparse perovskites. The XPS peaks of Pb 2+ and I - are observed on the surface of sparse MAPbI 3 in the dark, and no apparent valence state change are observed under illumination (Figure S5). When the sparse MAPbI 3 layer is prepared on TiO 2 to form a MAPbI 3 /TiO 2 half device (Figure S3a), the XPS results suggest that Pb 2+ /I - and Ti 4+ /O 2- are observed on the surface of the MAPbI 3 and TiO 2 in the dark (Figure 1a and b, Figure S6), respectively. When the MAPbI 3 /TiO 2 is illuminated, new peaks of Pb 0 at 141.2 eV and 136.1 eV and Ti 3+ at 456.8 eV and 463.1 eV appear (Figure 1a and b), which suggest that Pb 2+ and Ti 4+ are reduced into lower valence states. The observed reduction of Pb 2+ and Ti 4+ in the half device suggests that more photo-generated electrons transfer to the surface in the MAPbI 3 /TiO 2 than in the single MAPbI 3 . Moreover, since no XPS signal of Sn is detected in the MAPbI 3 /TiO 2 , the XPS peaks of O come from the TiO 2 layer but not the FTO substrate (Figure S6a). The content of adsorbed H 2 O on the surface of TiO 2 decreases remarkably under illumination, which possibly donate ions during the reduction reaction of Ti 4+ to Ti 3+ (Figure S6b). Similar results are also observed in some previous reports 25,27 . For the ions of I - , no obvious XPS change is observed under illumination (Figure 1c). Therefore, in the MAPbI 3 /TiO 2 half device, the photo-generated electrons in the MAPbI 3 layer reduce Pb 2+ on the surface of MAPbI 3 and Ti 4+ on the surface of TiO 2 into Pb 0 and Ti 3+ , respectively (Figure 1d). In order to investigate the charge transfer process in a perovskite full device in working conditions, we develop a new method to simulate the full device (TiO 2 /MAPbI 3 /Spiro-OMeTAD) in short circuit conditions by connecting the MAPbI 3 /TiO 2 half device with the MAPbI 3 /Spiro-OMeTAD half device by indium wires (MAPbI 3 /TiO 2 + MAPbI 3 /Spiro-OMeTAD). The MAPbI 3 /Spiro-OMeTAD half device was prepared by spin-coating Spiro-OMeTAD HTLlayer on FTO substrates and then preparing a sparse MAPbI 3 layer on the surface of the Spiro-OMeTADlayer (Figure S3b). The two half devices are illuminated at the same time and in situ XPS signals are collected. Under illumination, the ratio of Pb 0 /Pb 2+ reduces from 24% in the MAPbI 3 /TiO 2 half device to 7% in the simulated full device, and the ratio of Ti 3+ /Ti 4+ reduces from 43% in the MAPbI 3 /TiO 2 half device to 22% in the simulated full device (Figure 1e and f). There are two possible reasons for the lower ratios of Pb 0 /Pb 2+ and Ti 3+ /Ti 4+ in the simulated full device under illumination. One is that less photo-generated electrons transfer from MAPbI 3 to TiO 2 , the other is that photo-reduced products of Pb 0 and Ti 3+ are oxidized by photo-generated holes from the MAPbI 3 /Spiro-OMeTAD half device. The ions of I - on the surface MAPbI 3 are not photo-oxidized in the half device, in contrast, obvious photo-oxidation of I - to I 2 are observed in the simulated full device (Figure 1g). The results suggest that more photo-generated holes transfer to the surface of MAPbI 3 and oxidize I - . Therefore, the lower ratios of Pb 0 /Pb 2+ and Ti 3+ /Ti 4+ in the simulated full device under illumination comes from the oxidization by photo-excited holes from the MAPbI 3 /Spiro-OMeTAD half device (Figure 1h). The reversible process leads to long-term stability of a perovskite full device under illumination. Charge transfer at MAPbI 3 /Spiro-OMeTAD interface Similarly, charge transfer at MAPbI 3 /Spiro-OMeTAD interface was investigated by in situ XPS and the results are shown in Figure 2a and b. The peaks of I - and Pb 2+ are observed on the surface of MAPbI 3 in the dark. When the MAPbI 3 /Spiro-OMeTAD is illuminated, new peaks at 620.3 eV and 631.8 eV appear, which are assigned to I 2 . The result suggests that I - is oxidized to I 2 on the surface of MAPbI 3 under illumination. However, no apparent change of Pb 2+ is observed in the half device under illumination (Figure 2b). Since the in situ XPS method is difficult to investigate the change details of organic such as Spiro-OMeTAD under illumination (Figure S7), in situ FTIR was used to investigate the oxidation process of Spiro-OMeTAD under illumination (Figure S8a). The raw data of FTIR spectra of FTO/TiO 2 /MAPbI 3 /Spiro-OMeTAD is shown in Figure S9. The peaks at 1605 cm -1 and 1574 cm -1 are assigned to skeletal stretching vibration peaks of benzene ring 28,29 . When the device is illuminated in open circuit conditions, vibration peaks of benzene ring become stronger and some new peaks at 1162, 1137 and 1118 cm -1 appear, which are identified as the in-plane bending vibration of C-H on the benzene ring. In order to analyze the change FTIR spectra of Spiro-OMeTAD under illumination more clearly, the FTIR spectra of the device in the dark were used as the base line and the results are shown in Figure 2c. The peak intensity of benzene ring and C-H increases over time under illumination. The results suggest that oxidation reactions of Spiro-OMeTAD under illumination happen on the benzene ring. The FTIR peaks do not change on single Spiro-OMeTAD and MAPbI 3 under illumination (Figure S10), which suggests that the oxidation of Spiro-OMeTAD comes from I 2 on the surface of MAPbI 3 under illumination. Charge transfer process at MAPbI 3 /Spiro-OMeTAD interface under illumination is shown in Figure 2d. However, the photo-induced oxidation reactions in the MAPbI 3 /Spiro-OMeTAD will also lead to irreversible corrosion of MAPbI 3 under long-time illumination. Similar to the investigation of the reduction reactions in Figure 1e-g, the simulated full device (MAPbI 3 /Spiro-OMeTAD+ MAPbI 3 /TiO 2 ) was also constructed and the oxidation reaction was measured by in situ XPS. Under illumination, the ratio of I 2 /I - reduces from 9% in the MAPbI 3 /Spiro-OMeTAD half device to only 1% in the simulated full device (Figure 2a and e). The Pb 2+ ions on the surface of MAPbI 3 are not photo-reduced in the half device, but obvious photo-reduced of Pb 2+ to Pb 0 are observed in the simulated full device (Figure 2b and f). The results also suggest more photo-generated electrons transfer to the surface of MAPbI 3 and reduce Pb 2+ . Therefore, the lower ratio of I - /I 2 in the simulated full device also comes from the reduction of I 2 by photo-excited electrons from the MAPbI 3 /TiO 2 half device. In order to further investigate the reversibility in Spiro-OMeTAD during charge transfer process, structure changes of the Spiro-OMeTAD in short circuit conditions were also measured by in-situ FTIR (Figure S8b) and the results are shown in Figure 2g. The peak intensities of both skeletal stretching vibration of benzene ring (1605 and 1574 cm -1 ) and C-H of the benzene ring (1162, 1137 and 1118 cm -1 ) are much lower in the short circuit than in the open circuit condition (Figure 2c and g, Figure S11). To investigate the repeated cycling processes of redox reactions, in situ FTIR was used to characterize the charge transfer at MAPbI 3 /Spiro-OMeTAD interface (Figure S12). When MAPbI 3 /Spiro-OMeTAD is illuminated, the peak intensity of benzene ring and C-H of Spiro-OMeTAD increases over time, which implies the accumulation of oxidation of Spiro-OMeTAD. When the sample was stored in the dark for 60 min, the benzene ring vibration peak and C-H vibration peak of Spiro-OMeTAD can be recovered to the initial state. And when the sample was illuminated again, the benzene ring vibration peak and C-H vibration peak of Spiro-OMeTAD can also gradually increase again. Therefore, the redox reactions are reversible and accumulation of I 2 can be avoided if the hole can be transferred through the external circuit. Charge transfer processes in the simulated device under illumination is shown in Figure 2h. The reversible redox process in the simulated device of the MAPbI 3 /Spiro-OMeTAD is similar to that of MAPbI 3 /TiO 2 , leading to long-term stability in the perovskite full device under illumination. Moreover, in order to investigate the interface charge transfer rate in perovskite solar cells, we also measured the fs-TAS of MAPbI 3 and MAPbI 3 /Spiro-OMeTAD and the results are shown in Figure S13 and Table S1. The redox chemical reaction process happens from picoseconds to nanoseconds, which are close to the values in other redox reactions in photolyases and photocatalysts 30,31 . Though the charge transfer in the simulated full device has been investigated, it is not completely the same to that in the real full device. The simulated full device is similar to a parallel plate capacitor, with positive charges accumulated on one plate and negative charges accumulated on the other. Photo-charging processes happen in the simulated full device under illumination, which leads to Pb 0 and I - ions accumulated on one side and I 2 and Pb 2+ accumulated on the other (Figure S14). In a real full device, the redox reactions of Pb 2+ /Pb (2-x)+ (0<x≤2) and I - /I -1+x (0<x≤1) happen and no Pb 0 and I 2 accumulate at the interfaces, which leads to no obvious Pb 2+ ions and I - ions diffusion into the electron or hole transport layers and maintains the long-term stability of the device (Figure S14b, Figure S15). Chemical oxidation process of Spiro-OMeTAD Since I - ions of MAPbI 3 is oxidized into I 2 by photo-generated holes, the Spiro-OMeTAD is possibly oxidized by I 2 directly. Therefore, we further investigate the details of oxidation reactions of Spiro-OMeTAD by I 2 in the dark. The dark oxidation reactions were carried out by exposing the Spiro-OMeTAD in I 2 vapor for different time and the results are shown in Figure 3a. Before reactions, the peak at 1605 cm -1 and 1574 cm -1 of benzene ring on Spiro-OMeTAD are observed. The intensities of the two peaks increase over exposure time, and the new peaks of C-H (1162, 1137 and 1118 cm -1 ) appear, which are similar to the oxidation reaction of Spiro-OMeTAD in the simulated full device under illumination (Figure 2 c and g). Therefore, in the full device, the Spiro-OMeTAD is oxidized by I 2 , which comes from the oxidation of I - by photo-generated holes in MAPbI 3 .By XPS measurement, both I - ions and I 2 are observed on the surface of Spiro-OMeTAD after exposure to I 2 vapor in the dark, which suggest that I 2 is reduced into I - during the oxidation process of Spiro-OMeTAD (Figure 3b). From UV-Vis spectra, a new absorbance shoulder around 430~550 nm appear after exposure to I 2 vapor in the dark (Figure 3c). According to the previous study, the shoulder comes from the oxidation of Spiro-OMeTAD 32 . To further identify the accurate structure change during oxidation of Spiro-OMeTAD, 1 H NMR spectra was also used and the raw data is shown in Figure S16-18. Before reactions, the proton signal at 3.71, 6.17-6.18, 6.68-6.70, 6.80-6.86, 7.47-7.49 ppm are assigned to the methoxy group and benzene rings of Spiro-OMeTAD (Figure 3d, Figure S16) 33,34 . After exposure to I 2 vapor for different time in the dark, no new proton signals are observed. The results suggest that no methoxy group is substituted by I - during the oxidation process of Spiro-OMeTAD. However, the integral of the proton signal at 6.80-6.86 ppm decreases from 8 to 6 over time, which suggests that the proton on the ring with methoxy groups is substituted by I - step by step. Since the amino groups are electron-donating groups, I - substitutes ortho protons on the phenylamino 35 . The oxidation process of Spiro-OMeTAD is shown Figure 3e. Spiro-OMeTAD can be oxidized by I 2 in the dark, which is the same oxidation process in the full device under illumination. Faradaic junction charge transfer mechanism According to above results and analysis, we propose a charge transfer mechanism in a full device of FTO/TiO 2 /MAPbI 3 /Spiro-OMeTAD/Au in the dark (Figure 4a) and under illumination (Figure 4 b-d). The ion distribution of the device is shown in Figure S15. We have observed fast and reversible faradaic reactions of Pb 2+ /Pb (2-x)+ , I - / I -1+x , Ti 4+ /Ti 3+ and Spiro-OMeTAD at the interface in perovskite solar cells. The interface charge transfer happens by faradaic redox reactions, which are simplified as “faradaic junction model” in our previous studies 23-26 . By faradaic junction model, Pb 2+ and I - in MAPbI 3 are reduced and oxidized into Pb (2-x)+ (0<x≤2) and I -1+x (0<x≤1) by photo-generated electrons and holes under illumination (Figure 4b). Then, Pb (2-x)+ further reduces Ti 4+ in TiO 2 ETL and Pb (2-x)+ is re-oxidized into Pb 2+ . On the other side, I -1+x oxidizes Spiro-OMeTAD HTL and then is reduced to I - (Figure 4c). The two redox reactions of Pb 2+ /Pb (2-x)+ and I - /I -1+x are interface transfer mediators of photo-generated electrons and holes. In working conditions, both reduced TiO 2 and oxidized Spiro-OMeTAD can partly be recovered to initial states by charge transfer in the external circuit, which leads to good cycle stability (Figure 4d). Since interface charge transfer in FTO/TiO 2 /MAPbI 3 /Spiro-OMeTAD/Au comes from fast and reversible faradaic reactions of Pb 2+ /Pb (2-x)+ , I - / I -1+x , Ti 4+ /Ti 3+ and Spiro-OMeTAD, the perovskite solar cells are faradaic junction devices. Since some similar organic substances have also been used in organic solar cells, it is possible that a faradaic junction model can also provide plausible understanding on interface charge transfer processes in organic semiconductor solar cells 36 . New understanding on the origin of V OC by electrode potential alignment Following our previous study, the electrode potential is a quantitative thermodynamic descriptor for charge transfer in faradaic junction model 27 . The photovoltage of a MAPbI 3 solar cell depends on the electrode potentials of the two surface redox reactions of Pb 2+ /Pb (2-x)+ and I - /I -1+x . In order to quantitatively understand the V OC of perovskite solar cells, electrode potentials of MAPbI 3 , TiO 2 , and Spiro-OMeTAD were measured in the electrolyte of 0.1 M n-Bu 4 NPF 6 in CH 2 Cl 2 and the results are shown in Figure 5a-d 37 . All the potentials were calibrated with reference to the ferrocene/ferrocenium (Fc/Fc + ) redox couple (Figure S19) 38,39 . The electrochemical reduction and oxidation of MAPbI 3 are observed in the CV curves, respectively (Figure 5a and b). In order to obtain the accurate onset potentials, i-t curves were measured at different potentials and the MAPbI 3 were characterized by XRD and SEM. When the potential is more negative than -1.09 V vs. Fc/Fc + , MAPbI 3 is irreversibly reduced into Pb 0 (Figure S20). When the potential is more positive than 0.18 V vs. Fc/Fc + , the surface of MAPbI 3 is corroded apparently (Figure S21). According to the previous study, the corrosion of MAPbI 3 possibly comes from the irreversible oxidation of I - into I 2 on the surface(Figure S22) 40 . When the potential is at the range of -1.09 V and 0.18 V vs. Fc/Fc + , MAPbI 3 is reversible and stable. Therefore, the onset reduction corrosion potential and oxidation corrosion potential of MAPbI 3 is -1.09 V and 0.18 V vs. Fc/Fc + , respectively. Moreover, CV curves of TiO 2 ETL and Spiro-OMeTAD HTL suggest that the electrode potential windows are -1.49 V~-1.06 V vs. Fc/Fc + and -0.15 V~1.31 V vs. Fc/Fc + , respectively (Figure 5c and d). The energy band positions of MAPbI 3 were also measured by UV-Vis and UPS spectra (Figure S23). Figure 5e indicates the bulk band positions of MAPbI 3 and surface electrode potentials of MAPbI 3 , TiO 2 , and Spiro-OMeTAD in the dark. When MAPbI 3 is illuminated (Figure 5f), the photo-generated electrons in the conduction band of MAPbI 3 transfer to the MAPbI 3 /TiO 2 interface to reduce Pb 2 to Pb (2-x)+ (0<x≤2), the limiting potential of Pb 2+ /Pb 0 (x=2) is -1.09 V vs. Fc/Fc + . When Pb (2-x)+ further reduces Ti 4+ , this is a faradaic junction, in which isoenergetic charge transfer is observed in previous studies 25,26 . According to above discussion (Figure S11 and Table S1), this faradaic junction charge transfer is rapid enough to prevent voltage loss when current flows. Therefore, the electrode potential of TiO 2 after reduction is also -1.09 V vs. Fc/Fc + . Since TiO 2 is not oxidized easily, the photo-generated holes cannot transfer to the MAPbI 3 /TiO 2 interface, but to the MAPbI 3 /Spiro-OMeTAD interface to oxidize I - into I -1+x (0<x≤1), which further oxidize Spiro-OMeTAD, this is also an isoenergetic charge transfer process in a faradaic junction. The electrode potential of Spiro-OMeTAD after oxidation is at 0.18 V vs. Fc/Fc + . Therefore, a V OC of a MAPbI 3 solar cell can be quantitatively calculated by the difference between the electrode potentials of Pb 2+ /Pb (2-x)+ (0<x≤2) and I - /I -1+x (0<x≤1), and their limiting potentials are those of the redox reactions of Pb 2+ /Pb 0 (x=2)and I - /I 2 (x=1), which is 1.27 V in the TiO 2 /MAPbI 3 /Spiro-OMeTAD solar cell. In order to verify the validity of the prediction by the faradic junction model, a series of FTO/TiO 2 /MAPbI 3 /Spiro-OMeTAD/Au solar cells were constructed and the performances were measured (Figure S24). Figure 5g indicates a V OC of 1.09 V, which is a little lower than 1.27 V by faradaic junction model. To further confirm the universality of the faradic junction model in other wide bandgap perovskite solar cells, a MAPbBr 3 solar cell was also constructed. Similar faradaic junction charge transfer process is observed in a MAPbBr 3 solar cell (Figure S25-31) and a V OC of 1.78 V is predicted by faradaic junction model, which is higher than our experimental value of 1.41 V (Figure S32 and S33). In previous studies, the highest V OC of 1.26 V and 1.65 V have been obtained in MAPbI 3 -based and MAPbBr 3 -based solar cells, which are very close to the theoretical V OC by faradaic junction model 7,41 . The V OC loss is still observed in MAPbI 3 and MAPbBr 3 solar cells, which comes from the inconsistence between the bandgap and the electrode potential difference of a perovskite, slow interface charge transfer rate and back chemical reactions. In order to decrease the V OC loss, the electrode potentials of Pb 2+ /Pb (2-x)+ (0<x≤2) and I - /I -1+x (or Br - /Br -1+x , 0<x≤1) can be adjusted by changing the ratio of Pb 2+ and I - (or Br - ) on MAPbI 3 (or MAPbBr 3 ) surface following Nernst equation, which cannot be predicted by energy band alignment theory. Moreover, the V OC loss can also be decreased by changing transporting materials and suppressing the back chemical reactions, such as I -1+x + xe - →I - (or Br -1+x + e - →Br - ) at MAPbI 3 (MAPbBr 3 )/Spiro-OMeTAD interface and Pb (2-x)+ + xh + →Pb 2+ at MAPbI 3 (MAPbBr 3 )/TiO 2 interface (Figure S34) 42 . Moreover, we prepared perovskites with different A-site ions and investigated charge transfer behaviour of the mixed cation perovskites. Figure S35 indicates the XRD patterns, UV-Vis spectra and SEM images of FAPbI 3 and CsFAMAPbI 3 , respectively. The V OC of MAPbI 3 , CsFAMAPbI 3 and FAPbI 3 solar cells are 1.09 V, 1.04 V and 1.01 V, respectively (Figure S36a). The electrode potentials of Pb 2+ /Pb 0 and I - /I 2 of the three perovskites were measured and the results are shown in Figure S36 b-i. The electrode potentials of Pb 2+ /Pb 0 in MAPbI 3 , CsFAMAPbI 3 and FAPbI 3 are -1.09 V, -1.06 V and -1.04 V vs. Fc/Fc + , respectively, while the electrode potentials of I - /I 2 of the three perovskites are very close (0.18 V vs. Fc/Fc + ). The results suggest that A-site cations in the perovskites have influences on the reduction potential of Pb 2+ , but not the oxidation potential of I - . The predicted V OC of MAPbI 3 , CsFAMAPbI 3 and FAPbI 3 solar cells by faradaic junction model are 1.27 V, 1.24 V and 1.22 V, which are in good agreement with the experimental values. Therefore, faradaic junction theory can explain the behaviour of mixed cation perovskite solar cells. As for the mixed halide (I/Br) perovskites, photo-induced phase segregation process makes halide perovskites unstable, which leads to more complex the interface reactions 22,43 . Therefore, the mixed halide perovskites are not discussed here. In order to compare with the conventional energy band alignment theory, we also measured the energy band positions of TiO 2 ETL and Spiro-OMeTAD HTL and the results are shown in Figure S37 and S38. Schematic diagrams for interface charge transfer by energy band alignment theory are shown in Figure S39. The difference between the conduction band of TiO 2 and the HOMO of Spiro-OMeTAD is 1.14 eV. However, the quasi-fermi level of electrons and holes cannot be the same with the conduction band and the HOMO, and an energy loss between the fermi level and the energy band position is usually more than 0.1 eV 44 . Therefore, the V OC of 1.04 V can be predicted in both MAPbI 3 and MAPbBr 3 solar cells by energy band alignment theory. However, the experimental values of V OC in MAPbI 3 and MAPbBr 3 solar cells are 1.09 V and 1.41 V, which exceed the V OC values based on energy band alignment theory. Although variations in energy band positions exist, the conduction band of TiO 2 is reported as about -4.0 ~ -4.2 eV, and the HOMO of Spiro-OMeTAD is about -5.0 ~ -5.2 eV 1,45,46 . In the TiO 2 /perovskite/Spiro-OMeTAD solar cell, the V OC is predicted as about 0.8 ~ 1.2 V by the energy band alignment theory. However, the value is still much lower than the experimental value (1.4 V) of MAPbBr 3 solar cells, which cannot come from variations of energy levels. Therefore, the energy band alignment theory fails to understand the V OC in MAPbBr 3 solar cells. In contrast, by faradaic junction model, a V OC of 1.78 V is predicted for MAPbBr 3 cells by the electrode potential alignment, which is more reasonable. Moreover, the effect of energy band positions of charge transport layers on the V OC of perovskite solar cells is also considered. We used electron transport layers of TiO 2 and SnO 2 to construct MAPbI 3 solar cells. Though the difference of conduction bands of TiO 2 and SnO 2 is usually more than 0.2 eV (Figure S40a) 1,46 , a similar V OC of 1.09 V is obtained in the TiO 2 /MAPbI 3 /Spiro-OMeTAD and SnO 2 /MAPbI 3 /Spiro-OMeTAD (Figure S40b) solar cells. In contrast, both TiO 2 and SnO 2 show similar electrode potential window (Figure S40c). Therefore, not energy band alignment theory but electrode potential alignment in faradaic junction model is suitable to explain the V OC of perovskite solar cells. The origin of V OC of perovskite solar cells comes from the difference between the electrode potentials of Pb 2+ /Pb (2-x)+ (0<x≤2) and I - /I -1+x (or Br - /Br -1+x , 0<x≤1), and their limiting potentials are those of the redox reactions of Pb 2+ /Pb 0 (x=2)and I - /I 2 (or Br - /Br 2 , x=1). The understanding on the origin of V OC of perovskite solar cells is intrinsically different from energy band alignment theory (Figure S41). Based on faradaic junction model, to further increase the V OC of perovskite solar cells, major efforts should be focused on enlarging the electrode potential difference by shifting Pb 2+ /Pb 0 more negative and I - /I 2 (or Br - /Br 2 )more positive, and suppressing the back chemical reactions of I -1+x + xe - →I - (or Br -1+x + e - →Br - ) and Pb (2-x)+ + xh + →Pb 2+ 42,47 . Conclusion In summary, we found the faradaic reactions exist on the surfaces of perovskites (Pb 2+ + 2e - ↔ Pb 0 , 2I - + 2h + ↔ I 2 ), as well as the interfaces between perovskite/TiO 2 and perovskite/Spiro-OMeTAD (Pb 0 + 2Ti 4+ ↔ Pb 2+ + 2Ti 3+ , I 2 + Spiro-H + ↔ I - + Spiro-I - + H + ) by connecting a perovskite/ETL half device with a perovskite/HTL half device to simulate an all-solid-state device under illumination. Moreover, the faradaic oxidation reactions of Spiro-OMeTAD were also investigated in detail by in situ FTIR and NMR, which suggest that faradaic reactions happen not only on the surface of inorganic substances, but also on organics substances. Based on faradaic junction model, the V OC in perovskite solar cells was predicted by the difference between the electrode potentials of Pb 2+ /Pb (2-x)+ (0<x≤2) and I - /I -1+x (or Br - /Br -1+x , 0<x≤1), and their limiting potentials are those of the redox reactions of Pb 2+ /Pb 0 (x=2)and I - /I 2 (or Br - /Br 2 , x=1) on the surface of MAPbI 3 (MAPbBr 3 ). By enlarging the electrode potential difference between Pb 2+ /Pb 0 and I - /I 2 (or Br - /Br 2 ) and suppressing the back chemical reactions of I -1+x + xe - →I - (or Br -1+x + e - →Br - ) and Pb (2-x)+ + xh + →Pb 2+ , the V OC of perovskite solar cells can be further increased. Faradaic junction model can not only explain the origin of V OC in perovskite solar cells, but also provide guidance for the improvement of organic solar cells and organic light-emitting diodes. Declarations Acknowledgements This work was supported by the National Natural Science Foundation of China (22425202, 22279052, 22372067, 51902121, and 22372078) and the National Key Research and Development Program of China (2024YFB4609400, 2021YFF0500501). Author contributions W. L. supervised the project, proposed the concept and designed the experiments. M. X. carried out sample preparation, characterization and electrochemistry measurement; Y. C. conducted the fs-TAS measurements and analyzed the data. W. L. and M. X. analyzed the data and wrote the paper. All authors discussed the results and gave comments on the manuscript. Conflict of interest The authors declare no competing interests. Supplementary data Supplementary Information is available for this paper. Correspondence and requests for materials should be addressed to Wenjun Luo. References Jena, A. K., Kulkarni, A. & Miyasaka, T. Halide perovskite photovoltaics: Background, status, and future prospects. Chem. Rev. 119 , 3036-3103, (2019). Zhou, Y., Herz, L. M., Jen, A. K. Y. & Saliba, M. Advances and challenges in understanding the microscopic structure-property-performance relationship in perovskite solar cells. Nat. Energy 7 , 794-807, (2022). Best research-cell efficiencies. NREL , https://www.nrel.gov/pv/cell-efficiency.html. Zhang, H., Pfeifer, L., Zakeeruddin, S. M., Chu, J. & Grätzel, M. Tailoring passivators for highly efficient and stable perovskite solar cells. Nat. Rev. Chem. 7 , 632-652, (2023). Kojima, A., Teshima, K., Shirai, Y. & Miyasaka, T. Organometal halide perovskites as visible-light sensitizers for photovoltaic cells. J. Am. Chem. Soc. 131 , 6050-6051, (2009). Guo, Z., Jena, A. K., Kim, G. M. & Miyasaka, T. The high open-circuit voltage of perovskite solar cells: A review. Energy Environ. Sci. 15 , 3171-3222, (2022). Liu, Z. et al. Open-circuit voltages exceeding 1.26 V in planar methylammonium lead iodide perovskite solar cells. ACS Energy Lett. 4 , 110-117, (2019). He, R. et al. Improving interface quality for 1-cm 2 all-perovskite tandem solar cells. Nature 618 , 80-86, (2023). Wang, R. et al. Prospects for metal halide perovskite-based tandem solar cells. Nat. Photonics 15 , 411-425, (2021). Wang, Z. et al. Suppressed phase segregation for triple-junction perovskite solar cells. Nature 618 , 74-79, (2023). Xia, J., Sohail, M. & Nazeeruddin, M. K. Efficient and stable perovskite solar cells by tailoring of interfaces. Adv. Mater. 35 , 2211324, (2023). Zhang, T. et al. Ion-modulated radical doping of Spiro-OMeTAD for more efficient and stable perovskite solar cells. Science 377 , 495-501, (2022). Lin, Y. et al. Matching charge extraction contact for wide-bandgap perovskite solar cells. Adv. Mater. 29 , 1700607, (2017). Xu, F., Zhang, M., Li, Z., Yang, X. & Zhu, R. Challenges and perspectives toward future wide-bandgap mixed-halide perovskite photovoltaics. Adv. Energy Mater. 13 , 2203911, (2023). Caprioglio, P. et al. On the relation between the open-circuit voltage and quasi-fermi level splitting in efficient perovskite solar cells. Adv. Energy Mater. 9 , 1901631, (2019). Stolterfoht, M. et al. The impact of energy alignment and interfacial recombination on the internal and external open-circuit voltage of perovskite solar cells. Energy Environ. Sci. 12 , 2778-2788, (2019). Belisle, R. A., Jain, P., Prasanna, R., Leijtens, T. & McGehee, M. D. Minimal effect of the hole-transport material ionization potential on the open-circuit voltage of perovskite solar cells. ACS Energy Lett. 1 , 556-560, (2016). Dong, C. et al. Carbon-based all-inorganic perovskite solar cells: Progress, challenges and strategies toward 20% efficiency. Mater. Today 50 , 239-258, (2021). Liang, Y. et al. Achieving high open-circuit voltages up to 1.57 V in hole-transport-material-free MAPbBr 3 solar cells with carbon electrodes. Adv. Energy Mater. 8 , 1701159, (2017). Li, Y. et al. Lattice modulation of alkali metal cations doped Cs 1−x R x PbBr 3 halides for inorganic perovskite solar cells. Solar RRL 2 , 1800164, (2018). Liu, Y. et al. Direct observation of photoinduced ion migration in lead halide perovskites. Adv. Funct. Mater. 31 , 2008777, (2020). Frolova, L. A. et al. Reversible Pb 2+ /Pb 0 and I - /I 3 - redox chemistry drives the light‐induced phase segregation in all-inorganic mixed halide perovskites. Adv. Energy Mater. 11 , 2002934, (2021). Chen, X. et al. Reversible charge transfer and adjustable potential window in semiconductor/faradaic layer/liquid junctions. iScience 23 , 100949, (2020). Yin, Z. et al. Mildly regulated intrinsic faradaic layer at the oxide/water interface for improved photoelectrochemical performance. Chem. Sci. 11 , 6297-6304, (2020). Chen, M. et al. Faradaic junction and isoenergetic charge transfer mechanism on semiconductor/semiconductor interfaces. Nat. Commun. 12 , 6363, (2021). Wang, P. et al. Photovoltage memory effect in a portable faradaic junction solar rechargeable device. Nat. Commun. 13 , 2544, (2022). Xue, M. et al. Bipolarized intrinsic faradaic layer on a semiconductor surface under illumination. Natl. Sci. Rev. 10 , nwac249, (2023). Du, G., Yang, L. & Zhang, J. Light soaking induced halide doping of evaporated spiro‐ometad in perovskite solar cells. Laser Photonics Rev. 17 , 2200475, (2022). Trchová, M., Šeděnková, I., Tobolková, E. & Stejskal, J. FTIR spectroscopic and conductivity study of the thermal degradation of polyaniline films. Polym. Degrad. Stabil. 86 , 179-185, (2004). Maestre-Reyna, M. et al. Visualizing the DNA repair process by a photolyase at atomic resolution. Science 382 , eadd7795, (2023). Hoffmann, M. R., Martin, S. T., Choi, W. & Bahneman, D. W. Environmental applications of semiconductor photocatalysis. Chem. Rev. 95 , 69-96, (1995). Xi, H. et al. Performance enhancement of planar heterojunction perovskite solar cells through tuning the doping properties of hole-transporting materials. ACS Omega 2 , 326-336, (2017). Li, Z. et al. Acid additives enhancing the conductivity of Spiro‐OMeTAD toward high-efficiency and hysteresis-less planar perovskite solar cells. Adv. Energy Mater. 7 , 1601451, (2016). Jeong, M. et al. Stable perovskite solar cells with efficiency exceeding 24.8% and 0.3-V voltage loss. Science 369 , 1615-1620, (2020). Haruna, K., Alenaizan, A. A. & Al-Saadi, A. A. Density functional theory study of the substituent effect on the structure, conformation and vibrational spectra in halosubstituted anilines. RSC Adv. 6 , 67794-67804, (2016). Bakulin, A. A. et al. The role of driving energy and delocalized states for charge separation in organic semiconductors. Science 335 , 1340-1344, (2012). Samu, G. F. & Janáky, C. Photocorrosion at irradiated perovskite/electrolyte interfaces. J. Am. Chem. Soc. 142 , 21595-21614, (2020). Gagne, R. R., Koval, C. A. & Lisensky, G. C. Ferrocene as an internal standard for electrochemical measurements. Inorg. Chem. 19 , 2854-2855, (2002). DuBose, J. T. & Kamat, P. V. Probing perovskite photocatalysis. Interfacial electron transfer between CsPbBr 3 and ferrocene redox couple. J. Phys. Chem. Lett. 10 , 6074-6080, (2019). Mathew, P. S., Samu, G. F., Janáky, C. & Kamat, P. V. Iodine (I) expulsion at photoirradiated mixed halide perovskite interface. Should I stay or should I go? ACS Energy Lett. 5 , 1872-1880, (2020). Zhu, H. et al. Efficient and stable large bandgap MAPbBr 3 perovskite solar cell attaining an open circuit voltage of 1.65 V. ACS Energy Lett. 7 , 1112-1119, (2022). Cao, D. et al. Cathodic shift of onset potential for water oxidation on a Ti 4+ doped Fe 2 O 3 photoanode by suppressing the back reaction. Energy Environ. Sci. 7 , 752-759, (2014). Hoke, E. T. et al. Reversible photo-induced trap formation in mixed-halide hybrid perovskites for photovoltaics. Chem. Rev. 6 , 613-617, (2015). Butler, M. A. & Ginley, D. S. Prediction of flatband potentials at semiconductor-electrolyte interfaces from atomic electronegativities. J. Electrochem. Soc. 125 , 228-232, (1978). Hu, Z. et al. The impact of atmosphere on energetics of lead halide perovskites. Adv. Energy Mater. 10 , 2000908, (2020). Yuan, Y., Yan, G., Hong, R., Liang, Z. & Kirchartz, T. Quantifying efficiency limitations in all-inorganic halide perovskite solar cells. Adv. Mater. 34 , 2108132, (2022). Zhang, J. et al. A high-voltage solar rechargeable device based on a CoPi/BiVO 4 faradaic junction. J. Mater. Chem. A 10 , 1802-1807, (2022). Methods Materials CH 3 MH 3 I (MAI), CH 3 MH 3 Br (MABr), PbI 2 , PbBr 2 , Spiro-OMeTAD (99.8%), 4-tert-butylpyridine (tBP) and LiTFSI were purchased from Xi’an Polymer Light Technology. Titanium(diisopropoxide) bis(2,4-pentanedionate) in isopropanol (75%), N, N-dimethylformamide (DMF, 99.8%), dimethyl sulfoxide (DMSO, 99.8%) and tin (IV) oxide (SnO 2 ) colloid solution (15% in H 2 O colloidal dispersion) were purchased from Alfa Aesar. Anhydrous ethanol (99.8%) and Br 2 was purchased from Sinopharm Chemical Reagent. Ethyl acetate (anhydrous, 99.8%), acetonitrile (99.8%), dimethyl sulfoxide-d6 (DMSO-d6), and ferrocene (Fc) were purchased from Sigma-Aldrich. I 2 was purchased from Aladdin. Tetrabutylammonium hexafluorophosphate (n-Bu 4 NPF 6 )) and dichloromethane (DCM, CH 2 Cl 2 ) were purchased from Macklin. All materials were used without further purification. Preparation of samples FTO substrates (1.5 cm*1.5 cm) were cleaned sequentially by deionized water, acetone, and anhydrous ethanol. Before the film deposition, FTO substrates were treated by the UV-ozone for 30 min. Preparation of TiO 2 films A compact TiO 2 (c-TiO 2 ) layer was prepared by spin-coating at 3000r for 30s, and then transferred to the muffle furnace and annealed at 480 o C for 40 min with the heating rate of 2 o C/min 48 . The precursor solution of c-TiO 2 was prepared by the mixture of titanium(diisopropoxide) bis(2,4-pentanedionate) in isopropanol and anhydrous ethanol (v : v = 1 : 10). Preparation of SnO 2 films A SnO 2 layer was prepared by spin-coating at 4000r for 30s, and then annealed at 150 o C for 30 min on a hotplate in air. The precursor solution of SnO 2 was prepared by the mixture of SnO 2 colloid solution and deionized water (v : v = 1 : 4). Preparation of MAPbI 3 films All of the perovskite films were prepared by one-step method in nitrogen-filled glovebox 19,49 . The precursor solution for MAPbI 3 was obtained by dissolving 461 mg PbI 2 and 159 mg MAI in the mixture of DMF and DMSO (1 mL, v : v = 4 : 1). The solution was spin-coated on the different substrates at 1500r for 10s and 4000r for 40s, respectively. A drop of 500 μL ethyl acetate as antisolvent was poured on the film at 15 s after the starting of the spin-coating procedure. The MAPbI 3 films were then heated on a hotplate at 100 o C for 10 min and dense MAPbI 3 films were prepared. The sparse MAPbI 3 films were prepared with the same method but without the treatment of ethyl acetate antisolvent. Preparation of MAPbBr 3 films The precursor solution for MAPbBr 3 was obtained by dissolving 367 mg PbBr 2 and 112 mg MABr in the mixture of DMF and DMSO (1 mL, v : v = 4 : 1). The solution was spin-coated on the different substrates at 4000r for 40s. A drop of 500 μL ethyl acetate as antisolvent was poured on the film at 20 s after the starting of the spin-coating procedure. The MAPbBr 3 films were then heated on a hotplate at 100 o C for 10 min and dense MAPbBr 3 films were prepared. Preparation of FAPbI 3 films The precursor solution for FAPbI 3 was obtained by dissolving 880 mg PbI 2 , 309.5 mg FAI and 42.54 mg MACl in the mixture of DMF and DMSO (1 mL, v : v = 3 : 1) 50 . The solution was spin-coated on the different substrates at 5000r for 30s. A drop of 750 μL diethyl ether as antisolvent was poured on the film at 10 s after the starting of the spin-coating procedure. The FAPbI 3 films were then heated at 150 o C for 20 min on a hotplate in air and dense FAPbI 3 films were prepared. Preparation of CsFAMAPbI 3 films The CsI solution was prepared by dissolving 75 mg of CsI in 500 μL of DMSO solution 51 . The precursor solution for CsFAMAPbI 3 was obtained by dissolving 472 mg PbI 2 , 243 mg FAI, 14 mg MABr, 65.6 mg PbBr and 44 μL CsI solution in the mixture of 681 μL DMF and 170 μL DMSO. The solution was spin-coated on the different substrates at 3000r for 10s and 6000r for 30s, respectively. A drop of 500 μL chlorobenzene as antisolvent was poured on the film at 30 s after the starting of the spin-coating procedure. The CsFAMAPbI 3 films were then heated at 100 o C for 1 hour on a hotplate and dense CsFAMAPbI 3 films were prepared. Preparation of Spiro-OMeTAD films All of the Spiro-OMeTAD films in this study were doped with Li-TFSI and tBP. The precursor solution was prepared by dissolving 80 mg Spiro-OMeTAD, 28.8 μL tBP and 17.8 μL Li-TFSI (520 mg/mL in acetonitrile) in 1 mL chlorobenzene 52 . The Spiro-OMeTAD films were spin-coated on different substrates at 3000r for 30s in nitrogen-filled glovebox. The films were kept in dry air at room temperature for oxidation. Preparation of Au films A 100 nm thick Au layer was deposited by thermal evaporation. Characterization of samples The crystal structures of the samples were characterized by X-ray diffraction (XRD, Rigaku D/MAX-Ultima III). The morphologies of the samples were investigated by scanning electron microscope (SEM, Gemini 500) with an accelerating voltage of 10 kV. The energy-dispersive X-ray spectroscopy (EDS) was carried out by Gemini 500 equipped with an Oxford EDS system. The optical absorption was obtained by UV-Visible spectrophotometer (Shimadzu, UV 2550; Varian, Cary 50). Photoluminescence (PL) measurements were performed on a Varian Cary Eclipse. The band positions were measured by Ultraviolet Photoelectron spectroscopy (UPS, Thermofisher Escalab 250Xi, He I hv = 21.22 eV) The X-ray photoelectron spectroscopy (XPS, Thermofisher Escalab 250Xi) were performed with an Al Kα X-ray source. The binding energy of the C 1s peak at 284.6 eV was used to calibrate the XPS data. The samples of MAPbI 3 /TiO 2 and MAPbI 3 / Spiro-OMeTADwere illuminated under a Xe lamp (300 W) for 5 min and were characterized by in situ XPS. For the simulated full device, a MAPbI 3 /TiO 2 half device was connected with a MAPbI 3 /Spiro-OMeTAD half device by indium wires, and both of the two half devices were are illuminated under a Xe lamp (300 W). The Fourier Transform Infrared Spectroscopy (FTIR, Shimadazu IR Prestige-21) was performed with a resolution of 8 cm -1 . In situ FTIR was performed in the dark and under the illumination of a 405 nm-laser light (100 mW) for 20 min. In open circuit conditions, FTO/TiO 2 /MAPbI 3 /Spiro-OMeTAD was illuminated. In short circuit condition, the two ends of FTO/TiO 2 /MAPbI 3 /Spiro-OMeTAD/Au were connected by indium wires, and illuminated. To further investigate the details of oxidation reactions of Spiro-OMeTAD by I 2 and Br 2 in the dark, the FTO/Spiro-OMeTAD films were exposed to I 2 or Br 2 vapor in a closed glass vial in the dark for different time. The 1 H Nuclear Magnetic Resonance (NMR) spectra was performed with DMSO-d6 as solvent on a Bruker 600’ 54 Ascend LH 600MHz. Similar to the precursor solution for Spiro-OMeTAD films, 80 mg Spiro-OMeTAD and 17.8 μL Li-TFSI (520 mg/mL in acetonitrile) were dissolved in 1mL tetrahydrofuran, and then were evaporated. The Li-TFSI doped Spiro-OMeTAD powder was obtained, which were then exposed to I 2 vapor and Br 2 (dissolved in isopropanol) in a closed glass vial in the dark for different time for NMR measurement. The femtosecond transient absorption spectroscopy (fs-TAS) measurements were performed with a femtosecond broadband pump-probe system (HARPIA, Light Conversion company. Briefly, a 200 μJ pulse, was generated at a ~100 fs pulse width at 1030 nm with a repetition rate of 75 kHz using a regeneratively amplified Yb:KGW laser (PHAROS, Light Conversion). The output of the laser beam was split into two beams by a beam splitter. The larger portion of the beam passed through an optical parametric amplifier (ORPHEUS-HP, Light Conversion,) to generate a pump beam. The small portion of the fundamental beam passed through a second harmonic nonlinear crystal (BBO) to generate 515 nm, and then the 515 nm laser beam was focused onto a sapphire plate to generate supercontinuum white light as the probe beam. The absorption change (ΔA) of kinetic traces was measured at 6369 nm with a laser pulse of 700 nm and a fluence of 246 μJ cm −2 . The excited beam spot size (diameter) was of ~0.4 mm. The data were analyzed using CarpetView software (Light Conversion). Electrochemical measurements The electrochemical measurement of films was investigated in a three-electrode cell by using an electrochemical workstation (Shanghai Chenhua, CHI 760e). The prepared films were used as the working electrodes. A nonaqueous Ag electrode was used as the pseudo-reference electrode and a Pt wire as used as a counter electrode 53 . The electrolyte was 0.1M n-Bu 4 NPF 6 in CH 2 Cl 2 . 1 mM ferrocene was used as the internal reference and all the potentials were calibrated with reference to the ferrocene/ferrocenium (Fc/Fc + ) redox couple. The vacuum energy level of Fc/Fc + is -4.9 eV. Photovoltaic measurements The current density-voltage (J-V) characteristics were measured in a nitrogen-filled glovebox by a sourcemeter (EnliTech, B2091A) under the illumination of a solar simulator (EnliTech) at the light intensity of 100 mW/cm 2 (calibrated by a crystalline Si reference solar cell). The active area was 0.11 cm 2 . References 48. Feng, J. et al. Curing the fundamental issue of impurity phases in two-step solution-processed CsPbBr 3 perovskite films. Sci. Bull. 65 , 726-737, (2020). 49. Ravishankar, S. et al. Influence of charge transport layers on open-circuit voltage and hysteresis in perovskite solar cells. Joule 2 , 788-798, (2018). 50. Hui, W. et al. Stabilizing black-phase formamidinium perovskite formation at room temperature and high humidity. Science 371 , 1359-1364, (2021). 51. Dong, Q. et al. Interpenetrating interfaces for efficient perovskite solar cells with high operational stability and mechanical robustness. Nat. Commun. 12 , 973, (2021). 52. Burschka, J. et al. Sequential deposition as a route to high-performance perovskite-sensitized solar cells. Nature 499 , 316-319, (2013). 53. Joseph, V. et al. Stable perovskite solar cells using molecularly engineered functionalized oligothiophenes as low-cost hole-transporting materials. Small 17 , 2100783, (2021). Additional Declarations There is NO Competing Interest. Supplementary Files 20250317SupplementaryInformation.pdf Supplementary Information Cite Share Download PDF Status: Published Journal Publication published 19 Oct, 2025 Read the published version in Science China Chemistry → Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6295095","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":442255987,"identity":"6e492425-9822-4255-979b-d2d431711894","order_by":0,"name":"Wenjun Luo","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAu0lEQVRIiWNgGAWjYFCCww0MDBUMzCCmBJFaDgK1nCFNC2MDA2MbhEmcFoODB9skPs67w25wgPngbR4GuzzCWg4cbJOcue0Zs8EBtmRrHobkYoJazIBapHm3HQZq4TGT5mE4kNhAnJY5IC3830jR0gC2hY04LfYHDjZbzjh2mFnyMJux5RyDZMJaJGccPnjjQ83hZL7jzQ9vvKmwI6yFQeIAmEqGRKYBQfVAwA8x1Y4YtaNgFIyCUTBCAQAyuT5mtREksgAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0000-0001-6760-6929","institution":"Nanjing University","correspondingAuthor":true,"prefix":"","firstName":"Wenjun","middleName":"","lastName":"Luo","suffix":""},{"id":442255988,"identity":"49b5f666-7c46-4a99-9d3d-71dce61f5ac7","order_by":1,"name":"Mengfan Xue","email":"","orcid":"","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Mengfan","middleName":"","lastName":"Xue","suffix":""},{"id":442255989,"identity":"074d54f0-3576-47ca-b4ab-ff864c71f5dc","order_by":2,"name":"Zhoujun Li","email":"","orcid":"","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Zhoujun","middleName":"","lastName":"Li","suffix":""},{"id":442255990,"identity":"d9bd5b98-ebc9-4ebd-a9a9-60fe39ea0c34","order_by":3,"name":"Qiong Wang","email":"","orcid":"","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Qiong","middleName":"","lastName":"Wang","suffix":""},{"id":442255991,"identity":"acddc745-e8fd-47f3-bb8b-85993847def6","order_by":4,"name":"Yihong Chen","email":"","orcid":"","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Yihong","middleName":"","lastName":"Chen","suffix":""},{"id":442255992,"identity":"9f873619-f7cd-47c7-8751-e139c42e7ef3","order_by":5,"name":"Jun Luo","email":"","orcid":"","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Jun","middleName":"","lastName":"Luo","suffix":""},{"id":442255993,"identity":"cc9c92d0-446f-4e96-889c-b8e5badc5467","order_by":6,"name":"Ruotong Bao","email":"","orcid":"","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Ruotong","middleName":"","lastName":"Bao","suffix":""},{"id":442255994,"identity":"753b8ae1-df2b-44af-9720-5f124494f1d6","order_by":7,"name":"Dongjian Jiang","email":"","orcid":"","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Dongjian","middleName":"","lastName":"Jiang","suffix":""},{"id":442255995,"identity":"9aad8e5c-b93f-4710-b22e-d75694d68442","order_by":8,"name":"Shengyao Wang","email":"","orcid":"","institution":"Shanghai Jiao Tong University","correspondingAuthor":false,"prefix":"","firstName":"Shengyao","middleName":"","lastName":"Wang","suffix":""},{"id":442255996,"identity":"44f0a77a-b03b-4a84-8994-eb312b91c3d1","order_by":9,"name":"Bing Wang","email":"","orcid":"https://orcid.org/0000-0002-8735-9248","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Bing","middleName":"","lastName":"Wang","suffix":""},{"id":442255997,"identity":"667f774a-0142-4e4e-abbb-b11658d18c2f","order_by":10,"name":"Tao Yu","email":"","orcid":"https://orcid.org/0000-0003-1981-3469","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Tao","middleName":"","lastName":"Yu","suffix":""},{"id":442255998,"identity":"584e750f-2107-4e55-9eaa-7f38dceb60af","order_by":11,"name":"Yingfang Yao","email":"","orcid":"https://orcid.org/0000-0003-4823-0094","institution":"Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Yingfang","middleName":"","lastName":"Yao","suffix":""},{"id":442255999,"identity":"e37154ce-cfc3-458d-87fb-16e3fb8a8604","order_by":12,"name":"Zhigang Zou","email":"","orcid":"https://orcid.org/0000-0003-2092-8335","institution":"College of Engineering and Applied Sciences, Nanjing University","correspondingAuthor":false,"prefix":"","firstName":"Zhigang","middleName":"","lastName":"Zou","suffix":""}],"badges":[],"createdAt":"2025-03-24 11:55:30","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6295095/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6295095/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s11426-025-3073-0","type":"published","date":"2025-10-20T00:00:00+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":80614096,"identity":"bae6784e-5084-4b2c-99a0-1ba820f786d7","added_by":"auto","created_at":"2025-04-15 08:29:37","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":356005,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePhoto-induced reduction reactions at MAPbI\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e/TiO\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e interface. \u003c/strong\u003eIn situ XPS spectra of Pb 4f (a), Ti 2p (b) and I 3d (c) in a MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e half device in the dark and under illumination; (d) a schematic diagram of interface charge transfer in a MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e half device under illumination; in situ XPS spectra of Pb 4f (e), Ti 2p (f) and I 3d (g) in a MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e half device connected with a MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device in the dark and under illumination; (h) a schematic diagram of interface charge transfer in a simulated full device (MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e half device + MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device) under illumination.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6295095/v1/5e171f3bb236c027d45416af.png"},{"id":80614594,"identity":"45c75032-ec22-4719-870d-4eedb295cf97","added_by":"auto","created_at":"2025-04-15 08:37:37","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":395671,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePhoto-induced oxidation reactions at MAPbI\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e/Spiro-OMeTAD interface. \u003c/strong\u003eIn situ XPS spectra of I 3d (a) and Pb 4f (b) in a MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device in the dark and under illumination; (c) the changes of in situ FTIR spectra of FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD in open circuit conditions in the dark and under illumination; (d) a schematic diagram of interface charge transfer in a MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device under illumination; in situ XPS spectra of I 3d (e) and Pb 4f (f) in a MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device connected with a MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e half device in the dark and under illumination; (g) the changes of in situ FTIR spectra of FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD in short circuit conditions in the dark and under illumination; (h) a schematic diagram of interface charge transfer in a simulated full device (MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device + MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2 \u003c/sub\u003ehalf device) under illumination. Spiro-OMeTAD is abbreviated as Spiro in the figures.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-6295095/v1/d834ef425fa7551e79bf64ef.png"},{"id":80614595,"identity":"0919680c-11df-4c9e-bd34-a0c5bed73f2e","added_by":"auto","created_at":"2025-04-15 08:37:37","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":334217,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eChemical oxidation processes of Spiro-OMeTAD HTL in the dark.\u003c/strong\u003e (a) FTIR spectra of FTO/Spiro-OMeTAD before and after exposure to I\u003csub\u003e2\u003c/sub\u003e vapor for different time; (b) XPS spectra of I 3d in FTO/Spiro-OMeTAD before and after exposure to I\u003csub\u003e2\u003c/sub\u003e vapor; (c) UV-Vis spectra of FTO/Spiro-OMeTAD before and after exposure to I\u003csub\u003e2\u003c/sub\u003e vapor for different time; (d) \u003csup\u003e1\u003c/sup\u003eH NMR spectra of Spiro-OMeTAD powder before and after exposure to I\u003csub\u003e2\u003c/sub\u003e vapor for different time; (e) a schematic diagram of the molecular structure and multi-step oxidation process of Spiro-OMeTAD by I\u003csub\u003e2\u003c/sub\u003e in the dark.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-6295095/v1/01d92737fc7b807c189f5653.png"},{"id":80614102,"identity":"c784bed8-67f7-40f4-bc5c-1c1348642e20","added_by":"auto","created_at":"2025-04-15 08:29:37","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":725146,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFaradaic charge transfer processes in a perovskite solar cell. \u003c/strong\u003eSchematic diagrams of faradaic charge transfer processes in a perovskite solar cell of FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD/Au in different conditions: in the dark (a), photo-induced surface reactions of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003eand I\u003csup\u003e-\u003c/sup\u003e/ I\u003csup\u003e-1+x\u003c/sup\u003e in MAPbI\u003csub\u003e3\u003c/sub\u003e (b), interface charge transfer (c) and reversible processes in working conditions (d).\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-6295095/v1/6b7ac9daf8dbd7adccb74d05.png"},{"id":80614099,"identity":"4603bf28-4563-4acc-b4b1-4c3048686171","added_by":"auto","created_at":"2025-04-15 08:29:37","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":412051,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eThe origin of V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eOC\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e in a perovskite solar cell by faradaic junction model. \u003c/strong\u003eCyclic Voltammetry (CV) curves of reduction reactions (a) and oxidation reactions (b) of MAPbI\u003csub\u003e3\u003c/sub\u003e; CV curves of TiO\u003csub\u003e2\u003c/sub\u003e (c) and Spiro-OMeTAD (d), electrolyte: 0.1M n-Bu\u003csub\u003e4\u003c/sub\u003eNPF\u003csub\u003e6\u003c/sub\u003e in CH\u003csub\u003e2\u003c/sub\u003eCl\u003csub\u003e2\u003c/sub\u003e; scan rate: 50 mV/s; (e) a schematic diagram of the bulk band positions of MAPbI\u003csub\u003e3 \u003c/sub\u003eand surface electrode potentials of MAPbI\u003csub\u003e3\u003c/sub\u003e, TiO\u003csub\u003e2\u003c/sub\u003e, and Spiro-OMeTAD in the dark; (f) a schematic diagram of interface charge transfer in TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD under illumination; (g) current density-voltage (J-V) curves of FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD/Au in the dark and under illumination.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-6295095/v1/d4a30159cc7cfa6a6700cf15.png"},{"id":97143499,"identity":"5c2e505f-33b5-4c79-a4ee-9eedd1263d0c","added_by":"auto","created_at":"2025-12-01 10:10:08","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":3549351,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6295095/v1/e2983bcb-d6fd-470e-a7af-86ceb28b52fb.pdf"},{"id":80614111,"identity":"631f3c8f-5540-4581-bb45-f912282e5a61","added_by":"auto","created_at":"2025-04-15 08:29:38","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":4232572,"visible":true,"origin":"","legend":"Supplementary Information","description":"","filename":"20250317SupplementaryInformation.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6295095/v1/1d053eeb649353e9023d015a.pdf"}],"financialInterests":"There is \u003cb\u003eNO\u003c/b\u003e Competing Interest.","formattedTitle":"Origin of VOC in Perovskite Solar Cells by Faradaic Junction Model","fulltext":[{"header":"Introduction","content":"\u003cp\u003eA typical ABX\u003csub\u003e3\u003c/sub\u003e perovskite consists of methylammonium (MA\u003csup\u003e+\u003c/sup\u003e), formamidinium (FA\u003csup\u003e+\u003c/sup\u003e) or Cs\u003csup\u003e+\u003c/sup\u003e ions on the A site, Pb\u003csup\u003e2+\u003c/sup\u003e or Sn\u003csup\u003e2+\u003c/sup\u003e ions on the B site, and halide ions (I\u003csup\u003e-\u003c/sup\u003e, Br\u003csup\u003e-\u003c/sup\u003e, Cl\u003csup\u003e-\u003c/sup\u003e) on the X site\u003csup\u003e1,2\u003c/sup\u003e. Since 2009, perovskite solar cells have been a very promising technique for solar energy conversion due to its low cost and high power conversion efficiency (PCE), close to commercial Si solar cells\u003csup\u003e3-5\u003c/sup\u003e. The PCE of a solar cell is usually determined by open-circuit voltage (V\u003csub\u003eOC\u003c/sub\u003e), short-circuit current (I\u003csub\u003eSC\u003c/sub\u003e) and fill factor (FF). To date, I\u003csub\u003eSC\u003c/sub\u003e and FF of perovskite solar cells have been close to the theoretical limits. Therefore, it is the predominant way to increase the V\u003csub\u003eOC\u003c/sub\u003e to further improve the PCE\u003csup\u003e6\u003c/sup\u003e. In previous studies, great efforts, such as surface passivation, ion doping and crystal quality optimization, have been made to improve the V\u003csub\u003eOC\u003c/sub\u003e. A V\u003csub\u003eOC\u003c/sub\u003e of 1.26 V is obtained in a perovskite solar cell with a bandgap of 1.60 eV, in which a V\u003csub\u003eOC\u003c/sub\u003e loss (E\u003csub\u003eg\u003c/sub\u003e/q - V\u003csub\u003eOC\u003c/sub\u003e) is 0.34 V\u003csup\u003e7\u003c/sup\u003e.\u0026nbsp;Wide-bandgap perovskites (E\u003csub\u003eg\u003c/sub\u003e \u0026gt; 1.7 eV) are usually used in multi-junction solar cells by adjusting the ratio of halide ions (I\u003csup\u003e-\u003c/sup\u003e/Br\u003csup\u003e-\u003c/sup\u003e) on the X site\u003csup\u003e8,9\u003c/sup\u003e. However, a much higher V\u003csub\u003eOC\u003c/sub\u003e loss of 0.69 V is observed in the perovskite solar cell with a bandgap of 2.0 eV, which limits the applications of wide-bandgap perovskites in multi-junction perovskite solar cells\u003csup\u003e10\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn a typical perovskite solar cell, both electron transport layer (ETL) and hole transport layer (HTL) are introduced to improve the separation efficiency of the photo-generated carriers in the perovskite light absorber. According to energy band alignment theory, after the perovskite light absorber is contacted to charge transport layers, new equilibrium of the fermi level is established between the perovskite and the charge transport layers. When the perovskite is illuminated, the V\u003csub\u003eOC\u003c/sub\u003e in the perovskite solar cell is given by the difference between the quasi-Fermi levels of electrons at the electron transport layer and the quasi-Fermi level of holes at the hole transport layer\u003csup\u003e11,12\u003c/sup\u003e. Therefore, the\u0026nbsp;V\u003csub\u003eOC\u003c/sub\u003e of the perovskite solar cells depend on the band positions of ETL and HTL, and the regulation of energy levels of charge transport layers can enhance the V\u003csub\u003eOC\u003c/sub\u003e of the perovskite solar cell in some studies\u003csup\u003e13-16\u003c/sup\u003e.\u0026nbsp;However, in some experiments, the\u0026nbsp;V\u003csub\u003eOC\u003c/sub\u003e is similar even when ETL with different conduction bands or HTL with different highest-occupied molecular orbitals (HOMO) are used\u003csup\u003e1,17\u003c/sup\u003e. Moreover, in the TiO\u003csub\u003e2\u003c/sub\u003e/perovskite/carbon solar cell, the difference between the conduction band of TiO\u003csub\u003e2\u003c/sub\u003e (-4.0 eV~ -4.2 eV) and the work function of carbon (-5.0 eV) is about 0.8~1.0 eV. However, the V\u003csub\u003eOC\u003c/sub\u003e of CsPbBr\u003csub\u003e3\u003c/sub\u003e and MAPbBr\u003csub\u003e3\u003c/sub\u003e solar cells is about 1.5~1.6 V. A large discrepancy of ~0.6 V can be observed\u003csup\u003e18-20\u003c/sup\u003e. These results are not consistent with the theoretical V\u003csub\u003eOC\u003c/sub\u003e by the band alignment theory. Therefore, it is significant to understand the origin of V\u003csub\u003eOC\u003c/sub\u003e in perovskite solar cells. Different from electron transfer in a conventional solar cell model, ion migration is observed in perovskite solar cells\u003csup\u003e21,22\u003c/sup\u003e. A new\u0026nbsp;coupled electron and ion transfer model, which describes\u0026nbsp;chemical reactions at semiconductor interfaces (faradaic junction model),\u0026nbsp;has been proposed to understand interface charge transfer at semiconductor/liquid interface recently\u003csup\u003e23-26\u003c/sup\u003e.\u0026nbsp;Different from built-in electric field in a physical junction, the driving force for charge separation in a faradaic junction is the difference of electrode potentials (molar Gibbs free energy) at the interface. Is it possible that the faradaic junction charge transfer mechanism also exists in such all-solid-state devices? Moreover, in previous studies, faradaic reactions are only observed on the surfaces of oxides, such as TiO\u003csub\u003e2\u003c/sub\u003e, WO\u003csub\u003ex\u003c/sub\u003e and MnO\u003csub\u003ex\u003c/sub\u003e, but not on the surfaces of non-oxides\u003csup\u003e25,26\u003c/sup\u003e. To date, it is still a key challenge to characterize the faradaic reactions and interface charge transfer in all-solid-state\u0026nbsp;perovskite\u0026nbsp;solar cells, especially in working conditions.\u003c/p\u003e\n\u003cp\u003eIn this work, we developed a new method to simulate the charge transfer in a simulated full device by connecting a perovskite/TiO\u003csub\u003e2\u003c/sub\u003e half device with a perovskite/Spiro-OMeTAD half device under simultaneous illumination. By in situ XPS, in situ FTIR and other characterization methods, we found that the faradaic reactions indeed exist on the surface of perovskites, TiO\u003csub\u003e2\u003c/sub\u003e and Spiro-OMeTAD, as well as the interfaces between perovskite/TiO\u003csub\u003e2\u003c/sub\u003e and perovskite/Spiro-OMeTAD under illumination. Moreover, we measured electrode potentials of perovskites, TiO\u003csub\u003e2\u003c/sub\u003e and Spiro-OMeTAD to understand the V\u003csub\u003eOC\u003c/sub\u003e of perovskite solar cells. The results suggest that faradaic junction model is more suitable to describe the charge transfer in perovskite solar cells than the band alignment theory, which offers new concepts to improve V\u003csub\u003eOC\u003c/sub\u003e and PCE of perovskite solar cells by adjusting electrode potentials of perovskites.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCharge transfer at MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e interface\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eA TiO\u003csub\u003e2\u0026nbsp;\u003c/sub\u003elayerwas firstly prepared on FTO substrates and then a MAPbI\u003csub\u003e3\u003c/sub\u003e layer was prepared on the surface of the TiO\u003csub\u003e2\u003c/sub\u003e layer. In situ X-ray Photoelectron Spectroscopy (XPS) was used to investigate the charge transfer at MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e interface under illumination. In a normal solar cell, the perovskite layer is usually dense and covers the TiO\u003csub\u003e2\u003c/sub\u003e bottom layer completely, which hinders the detection of XPS signal of the TiO\u003csub\u003e2\u003c/sub\u003e layer. In order to detect the XPS signal of MAPbI\u003csub\u003e3\u003c/sub\u003e and TiO\u003csub\u003e2\u0026nbsp;\u003c/sub\u003eat the same time, a sparse MAPbI\u003csub\u003e3\u003c/sub\u003e layer was prepared on the FTO substrate in this study. The as-prepared sparse MAPbI\u003csub\u003e3\u003c/sub\u003e layer indicates the same crystal structure and optical property with a dense MAPbI\u003csub\u003e3\u003c/sub\u003e (Figure S1 and S2). To further investigate the charge transfer at the interface between the perovskite and the contact layer, we constructed TiO\u003csub\u003e2\u003c/sub\u003e/dense MAPbI\u003csub\u003e3\u003c/sub\u003e and TiO\u003csub\u003e2\u003c/sub\u003e/sparse MAPbI\u003csub\u003e3\u003c/sub\u003e heterojunctions (Figure S3). The linear sweep voltammetry (LSV) curves were measured and the results are shown in Figure S4. The two heterojunctions indicate also the same onset potentials under illumination, which suggest that similar interface charge behaviour happens in the dense and sparse perovskites. The XPS peaks of Pb\u003csup\u003e2+\u003c/sup\u003e and I\u003csup\u003e-\u0026nbsp;\u003c/sup\u003eare observed on the surface of sparse MAPbI\u003csub\u003e3\u003c/sub\u003e in the dark, and no apparent valence state change are observed under illumination (Figure S5). When the sparse MAPbI\u003csub\u003e3\u003c/sub\u003e layer is prepared on TiO\u003csub\u003e2\u003c/sub\u003e to form a MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e half device (Figure S3a), the XPS results suggest that Pb\u003csup\u003e2+\u003c/sup\u003e/I\u003csup\u003e-\u0026nbsp;\u003c/sup\u003eand Ti\u003csup\u003e4+\u003c/sup\u003e/O\u003csup\u003e2-\u003c/sup\u003e are observed on the surface of the MAPbI\u003csub\u003e3\u003c/sub\u003e and TiO\u003csub\u003e2\u003c/sub\u003e in the dark (Figure 1a and b, Figure S6), respectively. When the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e is illuminated, new peaks of Pb\u003csup\u003e0\u003c/sup\u003e at 141.2 eV and 136.1 eV and Ti\u003csup\u003e3+\u003c/sup\u003e at 456.8 eV and 463.1 eV appear (Figure 1a and b), which suggest that Pb\u003csup\u003e2+\u003c/sup\u003e and Ti\u003csup\u003e4+\u003c/sup\u003e are reduced into lower valence states. The observed reduction of Pb\u003csup\u003e2+\u003c/sup\u003e and Ti\u003csup\u003e4+\u003c/sup\u003e in the half device suggests that more photo-generated electrons transfer to the surface in the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e than in the single MAPbI\u003csub\u003e3\u003c/sub\u003e. Moreover, since no XPS signal of Sn is detected in the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e, the XPS peaks of O come from the TiO\u003csub\u003e2\u0026nbsp;\u003c/sub\u003elayer but not the FTO substrate (Figure S6a). The content of adsorbed H\u003csub\u003e2\u003c/sub\u003eO on the surface of TiO\u003csub\u003e2\u0026nbsp;\u003c/sub\u003edecreases remarkably under illumination, which possibly donate ions during the reduction reaction of Ti\u003csup\u003e4+\u003c/sup\u003e to Ti\u003csup\u003e3+\u003c/sup\u003e (Figure S6b). Similar results are also observed in some previous reports\u003csup\u003e25,27\u003c/sup\u003e. For the ions of I\u003csup\u003e-\u003c/sup\u003e, no obvious XPS change is observed under illumination (Figure 1c). Therefore, in the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u0026nbsp;\u003c/sub\u003ehalf device, the photo-generated electrons in the MAPbI\u003csub\u003e3\u003c/sub\u003e layer reduce Pb\u003csup\u003e2+\u003c/sup\u003e on the surface of MAPbI\u003csub\u003e3\u003c/sub\u003e and Ti\u003csup\u003e4+\u003c/sup\u003e on the surface of TiO\u003csub\u003e2\u003c/sub\u003e into Pb\u003csup\u003e0\u003c/sup\u003e and Ti\u003csup\u003e3+\u003c/sup\u003e, respectively (Figure 1d).\u003c/p\u003e\n\u003cp\u003eIn order to investigate the charge transfer process in a perovskite full device in working conditions, we develop a new method to simulate the full device (TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD) in short circuit conditions by connecting the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e half device with the MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device by indium wires (MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e + MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD). The MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device was prepared by spin-coating Spiro-OMeTAD HTLlayer on FTO substrates and then preparing a sparse MAPbI\u003csub\u003e3\u003c/sub\u003e layer on the surface of the Spiro-OMeTADlayer (Figure S3b). The two half devices are illuminated at the same time and in situ XPS signals are collected. Under illumination, the ratio of Pb\u003csup\u003e0\u003c/sup\u003e/Pb\u003csup\u003e2+\u003c/sup\u003e reduces from 24% in the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u0026nbsp;\u003c/sub\u003ehalf device to 7% in the simulated full device, and the ratio of Ti\u003csup\u003e3+\u003c/sup\u003e/Ti\u003csup\u003e4+\u003c/sup\u003e reduces from 43% in the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u0026nbsp;\u003c/sub\u003ehalf device to 22% in the simulated full device (Figure 1e and f). There are two possible reasons for the lower ratios of Pb\u003csup\u003e0\u003c/sup\u003e/Pb\u003csup\u003e2+\u003c/sup\u003e and Ti\u003csup\u003e3+\u003c/sup\u003e/Ti\u003csup\u003e4+\u003c/sup\u003e in the simulated full device under illumination. One is that less photo-generated electrons transfer from MAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003eto TiO\u003csub\u003e2\u003c/sub\u003e, the other is that photo-reduced products of Pb\u003csup\u003e0\u003c/sup\u003e and Ti\u003csup\u003e3+\u003c/sup\u003e are oxidized by photo-generated holes from the MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device. The ions of I\u003csup\u003e-\u003c/sup\u003e on the surface MAPbI\u003csub\u003e3\u003c/sub\u003e are not photo-oxidized in the half device, in contrast, obvious photo-oxidation of I\u003csup\u003e-\u003c/sup\u003e to I\u003csub\u003e2\u003c/sub\u003e are observed in the simulated full device (Figure 1g). The results suggest that more photo-generated holes transfer to the surface of MAPbI\u003csub\u003e3\u003c/sub\u003e and oxidize I\u003csup\u003e-\u003c/sup\u003e. Therefore, the lower ratios of Pb\u003csup\u003e0\u003c/sup\u003e/Pb\u003csup\u003e2+\u003c/sup\u003e and Ti\u003csup\u003e3+\u003c/sup\u003e/Ti\u003csup\u003e4+\u003c/sup\u003e in the simulated full device under illumination comes from the oxidization by photo-excited holes from the MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device (Figure 1h). The reversible process leads to long-term stability of a perovskite full device under illumination.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCharge transfer at MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD interface\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSimilarly, charge transfer at MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD interface was investigated by in situ XPS and the results are shown in Figure 2a and b. The peaks of I\u003csup\u003e-\u003c/sup\u003e and Pb\u003csup\u003e2+\u003c/sup\u003e are observed on the surface of MAPbI\u003csub\u003e3\u003c/sub\u003e in the dark. When the MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD is illuminated, new peaks at 620.3 eV and 631.8 eV appear, which are assigned to I\u003csub\u003e2\u003c/sub\u003e. The result suggests that I\u003csup\u003e-\u003c/sup\u003e is oxidized to I\u003csub\u003e2\u003c/sub\u003e on the surface of MAPbI\u003csub\u003e3\u003c/sub\u003e under illumination. However, no apparent change of Pb\u003csup\u003e2+\u003c/sup\u003e is observed in the half device under illumination (Figure 2b). Since the in situ XPS method is difficult to investigate the change details of organic such as Spiro-OMeTAD under illumination (Figure S7), in situ FTIR was used to investigate the oxidation process of Spiro-OMeTAD under illumination (Figure S8a). The raw data of FTIR spectra of FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD is shown in Figure S9. The peaks at 1605 cm\u003csup\u003e-1\u003c/sup\u003e and 1574 cm\u003csup\u003e-1\u003c/sup\u003e are assigned to skeletal stretching vibration peaks of benzene ring\u003csup\u003e28,29\u003c/sup\u003e. When the device is illuminated in open circuit conditions, vibration peaks of benzene ring become stronger and some new peaks at 1162, 1137 and 1118 cm\u003csup\u003e-1\u003c/sup\u003e appear, which are identified as the in-plane bending vibration of C-H on the benzene ring. In order to analyze the change FTIR spectra of Spiro-OMeTAD under illumination more clearly, the FTIR spectra of the device in the dark were used as the base line and the results are shown in Figure 2c. The peak intensity of benzene ring and C-H increases over time under illumination. The results suggest that oxidation reactions of Spiro-OMeTAD under illumination happen on the benzene ring. The FTIR peaks do not change on single Spiro-OMeTAD and MAPbI\u003csub\u003e3\u003c/sub\u003e under illumination (Figure S10), which suggests that the oxidation of Spiro-OMeTAD comes from I\u003csub\u003e2\u003c/sub\u003e on the surface of MAPbI\u003csub\u003e3\u003c/sub\u003e under illumination. Charge transfer process at MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD interface under illumination is shown in Figure 2d.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eHowever, the photo-induced oxidation reactions in the MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD will also lead to irreversible corrosion of MAPbI\u003csub\u003e3\u003c/sub\u003e under long-time illumination. Similar to the investigation of the reduction reactions in Figure 1e-g, the simulated full device (MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD+ MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e) was also constructed and the oxidation reaction was measured by in situ XPS. Under illumination, the ratio of I\u003csub\u003e2\u003c/sub\u003e/I\u003csup\u003e-\u003c/sup\u003e reduces from 9% in the MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device to only 1% in the simulated full device (Figure 2a and e). The Pb\u003csup\u003e2+\u003c/sup\u003e ions on the surface of MAPbI\u003csub\u003e3\u003c/sub\u003e are not photo-reduced in the half device, but obvious photo-reduced of Pb\u003csup\u003e2+\u003c/sup\u003e to Pb\u003csup\u003e0\u003c/sup\u003e are observed in the simulated full device (Figure 2b and f). The results also suggest more photo-generated electrons transfer to the surface of MAPbI\u003csub\u003e3\u003c/sub\u003e and reduce Pb\u003csup\u003e2+\u003c/sup\u003e. Therefore, the lower ratio of I\u003csup\u003e-\u003c/sup\u003e/I\u003csub\u003e2\u003c/sub\u003e in the simulated full device also comes from the reduction of I\u003csub\u003e2\u003c/sub\u003e by photo-excited electrons from the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e half device. In order to further investigate the reversibility in Spiro-OMeTAD during charge transfer process, structure changes of the Spiro-OMeTAD in short circuit conditions were also measured by in-situ FTIR (Figure S8b) and the results are shown in Figure 2g. The peak intensities of both skeletal stretching vibration of benzene ring (1605 and 1574 cm\u003csup\u003e-1\u003c/sup\u003e) and C-H of the benzene ring (1162, 1137 and 1118 cm\u003csup\u003e-1\u003c/sup\u003e) are much lower in the short circuit than in the open circuit condition (Figure 2c and g, Figure S11). To investigate the repeated cycling processes of redox reactions,\u0026nbsp;in situ FTIR was used to characterize the charge transfer at MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD interface (Figure S12). When MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD is illuminated,\u0026nbsp;the peak intensity of benzene ring and C-H of Spiro-OMeTAD increases over time, which implies the accumulation of oxidation of Spiro-OMeTAD. When the sample was stored in the dark for 60 min, the benzene ring vibration peak and C-H vibration peak of Spiro-OMeTAD can be recovered to the initial state. And when the sample was illuminated again, the benzene ring vibration peak and C-H vibration peak of Spiro-OMeTAD can also gradually increase again. Therefore, the redox reactions are reversible and\u0026nbsp;accumulation of I\u003csub\u003e2\u003c/sub\u003e can be avoided if the hole can be transferred through the external circuit. Charge transfer processes in the simulated device under illumination is shown in Figure 2h. The reversible redox process in the simulated device of the MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD is similar to that of MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e, leading to long-term stability in the perovskite full device under illumination. Moreover, in order to investigate the interface charge transfer rate in perovskite solar cells, we also measured the fs-TAS of MAPbI\u003csub\u003e3\u003c/sub\u003e and MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD and the results are shown in Figure S13 and Table S1. The redox chemical reaction process happens from picoseconds to nanoseconds, which are close to the values in other redox reactions in photolyases and photocatalysts\u003csup\u003e30,31\u003c/sup\u003e. Though the charge transfer in the simulated full device has been investigated, it is not completely the same to that in the real full device. The simulated full device is similar to a parallel plate capacitor, with positive charges accumulated on one plate and negative charges accumulated on the other. Photo-charging processes happen in the simulated full device under illumination, which leads to Pb\u003csup\u003e0\u003c/sup\u003e and I\u003csup\u003e-\u003c/sup\u003e ions accumulated on one side and I\u003csub\u003e2\u003c/sub\u003e and Pb\u003csup\u003e2+\u003c/sup\u003e accumulated on the other (Figure S14). In a real full device, the redox reactions of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003e (0\u0026lt;x\u0026le;2) and I\u003csup\u003e-\u003c/sup\u003e/I\u003csup\u003e-1+x\u003c/sup\u003e (0\u0026lt;x\u0026le;1) happen and no Pb\u003csup\u003e0\u003c/sup\u003e and I\u003csub\u003e2\u003c/sub\u003e accumulate at the interfaces, which leads to no obvious Pb\u003csup\u003e2+\u003c/sup\u003e ions and I\u003csup\u003e-\u003c/sup\u003e ions diffusion into the electron or hole transport layers and maintains the long-term stability of the device (Figure S14b, Figure S15).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eChemical oxidation process of Spiro-OMeTAD\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSince I\u003csup\u003e-\u003c/sup\u003e ions of MAPbI\u003csub\u003e3\u003c/sub\u003e is oxidized into I\u003csub\u003e2\u003c/sub\u003e by photo-generated holes, the Spiro-OMeTAD is possibly oxidized by I\u003csub\u003e2\u003c/sub\u003e directly. Therefore, we further investigate the details of oxidation reactions of Spiro-OMeTAD by I\u003csub\u003e2\u003c/sub\u003e in the dark. The dark oxidation reactions were carried out by exposing the Spiro-OMeTAD in I\u003csub\u003e2\u003c/sub\u003e vapor for different time and the results are shown in Figure 3a. Before reactions, the peak at 1605 cm\u003csup\u003e-1\u003c/sup\u003e and 1574 cm\u003csup\u003e-1\u003c/sup\u003e of benzene ring on Spiro-OMeTAD are observed. The intensities of the two peaks increase over exposure time, and the new peaks of C-H (1162, 1137 and 1118 cm\u003csup\u003e-1\u003c/sup\u003e) appear, which are similar to the oxidation reaction of Spiro-OMeTAD in the simulated full device under illumination (Figure 2 c and g). Therefore, in the full device, the Spiro-OMeTAD is oxidized by I\u003csub\u003e2\u003c/sub\u003e, which comes from the oxidation of I\u003csup\u003e-\u003c/sup\u003e by photo-generated holes in MAPbI\u003csub\u003e3\u003c/sub\u003e.By XPS measurement, both I\u003csup\u003e-\u003c/sup\u003e ions and I\u003csub\u003e2\u003c/sub\u003e are observed on the surface of Spiro-OMeTAD after exposure to I\u003csub\u003e2\u003c/sub\u003e vapor in the dark, which suggest that I\u003csub\u003e2\u003c/sub\u003e is reduced into I\u003csup\u003e-\u003c/sup\u003e during the oxidation process of Spiro-OMeTAD (Figure 3b). From UV-Vis spectra, a new absorbance shoulder around 430~550 nm appear after exposure to I\u003csub\u003e2\u003c/sub\u003e vapor in the dark (Figure 3c). According to the previous study, the shoulder comes from the oxidation of Spiro-OMeTAD\u003csup\u003e32\u003c/sup\u003e. To further identify the accurate structure change during oxidation of Spiro-OMeTAD, \u003csup\u003e1\u003c/sup\u003eH NMR spectra was also used and the raw data is shown in Figure S16-18. Before reactions, the proton signal at 3.71, 6.17-6.18, 6.68-6.70, 6.80-6.86, 7.47-7.49 ppm are assigned to the methoxy group and benzene rings of Spiro-OMeTAD (Figure 3d, Figure S16)\u003csup\u003e33,34\u003c/sup\u003e. After\u0026nbsp;exposure to\u0026nbsp;I\u003csub\u003e2\u003c/sub\u003e vapor for different time in the dark, no new proton signals are observed. The results suggest that no methoxy group is substituted by I\u003csup\u003e-\u003c/sup\u003e during the oxidation process of Spiro-OMeTAD. However, the integral of the proton signal at 6.80-6.86 ppm decreases from 8 to 6 over time, which suggests that the proton on the ring with methoxy groups is substituted by I\u003csup\u003e-\u003c/sup\u003e step by step. Since the amino groups are electron-donating groups, I\u003csup\u003e-\u003c/sup\u003e substitutes ortho protons on the phenylamino\u003csup\u003e35\u003c/sup\u003e. The oxidation process of Spiro-OMeTAD is shown Figure 3e. Spiro-OMeTAD can be oxidized by I\u003csub\u003e2\u003c/sub\u003e in the dark, which is the same oxidation process in the full device under illumination.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFaradaic junction charge transfer mechanism\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAccording to above results and analysis, we propose a charge transfer mechanism in a full device of FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD/Au in the dark (Figure 4a) and under illumination (Figure 4 b-d). The ion distribution of the device is shown in Figure S15. We have observed fast and reversible faradaic reactions of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003e, I\u003csup\u003e-\u003c/sup\u003e/ I\u003csup\u003e-1+x\u003c/sup\u003e, Ti\u003csup\u003e4+\u003c/sup\u003e/Ti\u003csup\u003e3+\u003c/sup\u003e and Spiro-OMeTAD at the interface in perovskite solar cells. The interface charge transfer happens by faradaic redox reactions, which are simplified as \u0026ldquo;faradaic junction model\u0026rdquo; in our previous studies\u003csup\u003e23-26\u003c/sup\u003e. By faradaic junction model, Pb\u003csup\u003e2+\u003c/sup\u003e and I\u003csup\u003e-\u003c/sup\u003e in MAPbI\u003csub\u003e3\u003c/sub\u003e are reduced and oxidized into Pb\u003csup\u003e(2-x)+\u0026nbsp;\u003c/sup\u003e(0\u0026lt;x\u0026le;2) and I\u003csup\u003e-1+x\u003c/sup\u003e (0\u0026lt;x\u0026le;1) by photo-generated electrons and holes under illumination (Figure 4b). Then, Pb\u003csup\u003e(2-x)+\u003c/sup\u003e further reduces Ti\u003csup\u003e4+\u003c/sup\u003e in TiO\u003csub\u003e2\u003c/sub\u003e ETL and Pb\u003csup\u003e(2-x)+\u003c/sup\u003e is re-oxidized into Pb\u003csup\u003e2+\u003c/sup\u003e. On the other side, I\u003csup\u003e-1+x\u003c/sup\u003e oxidizes Spiro-OMeTAD HTL and then is reduced to I\u003csup\u003e-\u003c/sup\u003e (Figure 4c). The two redox reactions of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003e and I\u003csup\u003e-\u003c/sup\u003e/I\u003csup\u003e-1+x\u003c/sup\u003e are interface transfer mediators of photo-generated electrons and holes. In working conditions, both reduced TiO\u003csub\u003e2\u003c/sub\u003e and oxidized Spiro-OMeTAD can partly be recovered to initial states by charge transfer in the external circuit, which leads to good cycle stability (Figure 4d). Since interface charge transfer in FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD/Au comes from fast and reversible faradaic reactions of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003e, I\u003csup\u003e-\u003c/sup\u003e/ I\u003csup\u003e-1+x\u003c/sup\u003e, Ti\u003csup\u003e4+\u003c/sup\u003e/Ti\u003csup\u003e3+\u003c/sup\u003e and Spiro-OMeTAD, the perovskite solar cells are faradaic junction devices. Since some similar organic substances have also been used in organic solar cells, it is possible that a faradaic junction model can also provide plausible understanding on interface charge transfer processes in organic semiconductor solar cells\u003csup\u003e36\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eNew understanding on the origin of V\u003csub\u003eOC\u003c/sub\u003e by electrode potential alignment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFollowing our previous study, the electrode potential is a quantitative thermodynamic descriptor for charge transfer in faradaic junction model\u003csup\u003e27\u003c/sup\u003e. The photovoltage of a\u0026nbsp;MAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003esolar cell\u0026nbsp;depends on the electrode potentials of the two surface redox reactions of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003e and I\u003csup\u003e-\u003c/sup\u003e/I\u003csup\u003e-1+x\u003c/sup\u003e. In order to quantitatively understand the V\u003csub\u003eOC\u0026nbsp;\u003c/sub\u003eof perovskite solar cells, electrode potentials of MAPbI\u003csub\u003e3\u003c/sub\u003e, TiO\u003csub\u003e2\u003c/sub\u003e, and Spiro-OMeTAD were measured in the electrolyte of 0.1 M n-Bu\u003csub\u003e4\u003c/sub\u003eNPF\u003csub\u003e6\u003c/sub\u003e in CH\u003csub\u003e2\u003c/sub\u003eCl\u003csub\u003e2\u003c/sub\u003e and the results are shown in Figure 5a-d\u003csup\u003e37\u003c/sup\u003e. All the potentials were calibrated with reference to\u0026nbsp;the ferrocene/ferrocenium (Fc/Fc\u003csup\u003e+\u003c/sup\u003e) redox couple\u0026nbsp;(Figure S19)\u003csup\u003e38,39\u003c/sup\u003e. The electrochemical reduction and oxidation of MAPbI\u003csub\u003e3\u003c/sub\u003e are observed in the CV curves, respectively (Figure 5a and b). In order to obtain the accurate onset potentials, i-t curves were measured at different potentials and the MAPbI\u003csub\u003e3\u003c/sub\u003e were characterized by XRD and SEM. When the potential is more negative than -1.09 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e, MAPbI\u003csub\u003e3\u003c/sub\u003e is irreversibly reduced into Pb\u003csup\u003e0\u003c/sup\u003e (Figure S20). When the potential is more positive than 0.18 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e, the surface of MAPbI\u003csub\u003e3\u003c/sub\u003e is corroded apparently (Figure S21). According to the previous study, the corrosion of MAPbI\u003csub\u003e3\u003c/sub\u003e possibly comes from the irreversible oxidation of I\u003csup\u003e-\u003c/sup\u003e into I\u003csub\u003e2\u003c/sub\u003e on the surface(Figure S22)\u003csup\u003e40\u003c/sup\u003e. When the potential is at the range of -1.09 V and 0.18 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e, MAPbI\u003csub\u003e3\u003c/sub\u003e is reversible and stable. Therefore, the onset reduction corrosion potential and oxidation corrosion potential of MAPbI\u003csub\u003e3\u003c/sub\u003e is -1.09 V and 0.18 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e, respectively. Moreover, CV curves of TiO\u003csub\u003e2\u003c/sub\u003e ETL and Spiro-OMeTAD HTL suggest that the electrode potential windows are -1.49 V~-1.06 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e and -0.15 V~1.31 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e, respectively (Figure 5c and d). The energy band positions of MAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003ewere also measured by UV-Vis and UPS spectra (Figure S23).\u003c/p\u003e\n\u003cp\u003eFigure 5e indicates the bulk band positions of MAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003eand surface electrode potentials of MAPbI\u003csub\u003e3\u003c/sub\u003e, TiO\u003csub\u003e2\u003c/sub\u003e, and Spiro-OMeTAD in the dark. When MAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003eis illuminated (Figure 5f), the photo-generated electrons in the conduction band of MAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003etransfer to the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u0026nbsp;\u003c/sub\u003einterface to reduce Pb\u003csup\u003e2\u0026nbsp;\u003c/sup\u003eto Pb\u003csup\u003e(2-x)+\u003c/sup\u003e (0\u0026lt;x\u0026le;2), the limiting potential of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e0\u003c/sup\u003e (x=2) is -1.09 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e. When Pb\u003csup\u003e(2-x)+\u003c/sup\u003e further reduces Ti\u003csup\u003e4+\u003c/sup\u003e, this is a faradaic junction, in which isoenergetic charge transfer is observed in previous studies\u003csup\u003e25,26\u003c/sup\u003e. According to above discussion (Figure S11 and Table S1), this faradaic junction charge transfer is rapid enough to prevent voltage loss when current flows. Therefore, the electrode potential of TiO\u003csub\u003e2\u003c/sub\u003e after reduction is also -1.09 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e. Since TiO\u003csub\u003e2\u003c/sub\u003e is not oxidized easily, the photo-generated holes cannot transfer to the MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u0026nbsp;\u003c/sub\u003einterface, but to the MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD interface to oxidize I\u003csup\u003e-\u003c/sup\u003e into I\u003csup\u003e-1+x\u003c/sup\u003e (0\u0026lt;x\u0026le;1), which further oxidize Spiro-OMeTAD, this is also an isoenergetic charge transfer process in a faradaic junction. The electrode potential of Spiro-OMeTAD after oxidation is at 0.18 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e. Therefore, a V\u003csub\u003eOC\u003c/sub\u003e of a MAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003esolar cell can be quantitatively calculated by the difference between the electrode potentials of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003e (0\u0026lt;x\u0026le;2) and I\u003csup\u003e-\u003c/sup\u003e/I\u003csup\u003e-1+x\u003c/sup\u003e (0\u0026lt;x\u0026le;1), and their limiting potentials are those of the redox reactions of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e0\u003c/sup\u003e(x=2)and I\u003csup\u003e-\u003c/sup\u003e/I\u003csub\u003e2\u0026nbsp;\u003c/sub\u003e(x=1),\u0026nbsp;which is 1.27 V in the TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD solar cell. In order to verify the validity of the prediction by the faradic junction model, a series of FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD/Au solar cells were constructed and the performances were measured (Figure S24). Figure 5g indicates a V\u003csub\u003eOC\u003c/sub\u003e of 1.09 V, which is a little lower than 1.27 V by faradaic junction model. To further confirm the universality of the faradic junction model in other wide bandgap perovskite solar cells, a MAPbBr\u003csub\u003e3\u003c/sub\u003e solar cell was also constructed. Similar faradaic junction charge transfer process is observed in a MAPbBr\u003csub\u003e3\u003c/sub\u003e solar cell (Figure S25-31) and a V\u003csub\u003eOC\u003c/sub\u003e of 1.78 V is predicted by faradaic junction model, which is higher than our experimental value of 1.41 V (Figure S32 and S33). In previous studies, the highest V\u003csub\u003eOC\u003c/sub\u003e of 1.26 V and 1.65 V have been obtained in MAPbI\u003csub\u003e3\u003c/sub\u003e-based and MAPbBr\u003csub\u003e3\u003c/sub\u003e-based solar cells, which are very close to the theoretical V\u003csub\u003eOC\u003c/sub\u003e by faradaic junction model\u003csup\u003e7,41\u003c/sup\u003e. The V\u003csub\u003eOC\u003c/sub\u003e loss is still observed in MAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003eand MAPbBr\u003csub\u003e3\u003c/sub\u003e solar cells, which comes from the inconsistence between the bandgap and the electrode potential difference of a perovskite, slow interface charge transfer rate and back chemical reactions. In order to decrease the V\u003csub\u003eOC\u003c/sub\u003e loss, the electrode potentials of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003e (0\u0026lt;x\u0026le;2) and I\u003csup\u003e-\u003c/sup\u003e/I\u003csup\u003e-1+x\u003c/sup\u003e (or Br\u003csup\u003e-\u003c/sup\u003e/Br\u003csup\u003e-1+x\u003c/sup\u003e, 0\u0026lt;x\u0026le;1) can be adjusted by changing the ratio of Pb\u003csup\u003e2+\u003c/sup\u003e and I\u003csup\u003e-\u003c/sup\u003e (or Br\u003csup\u003e-\u003c/sup\u003e) on MAPbI\u003csub\u003e3\u003c/sub\u003e (or MAPbBr\u003csub\u003e3\u003c/sub\u003e) surface following Nernst equation, which cannot be predicted by energy band alignment theory. Moreover, the V\u003csub\u003eOC\u003c/sub\u003e loss can also be decreased by changing transporting materials and suppressing the back chemical reactions, such as I\u003csup\u003e-1+x\u0026nbsp;\u003c/sup\u003e+ xe\u003csup\u003e-\u0026nbsp;\u003c/sup\u003e\u0026rarr;I\u003csup\u003e-\u003c/sup\u003e (or Br\u003csup\u003e-1+x\u003c/sup\u003e + e\u003csup\u003e-\u003c/sup\u003e\u0026rarr;Br\u003csup\u003e-\u003c/sup\u003e) at MAPbI\u003csub\u003e3\u003c/sub\u003e (MAPbBr\u003csub\u003e3\u003c/sub\u003e)/Spiro-OMeTAD interface and Pb\u003csup\u003e(2-x)+\u003c/sup\u003e + xh\u003csup\u003e+\u0026nbsp;\u003c/sup\u003e\u0026rarr;Pb\u003csup\u003e2+\u003c/sup\u003e at MAPbI\u003csub\u003e3\u003c/sub\u003e (MAPbBr\u003csub\u003e3\u003c/sub\u003e)/TiO\u003csub\u003e2\u003c/sub\u003e interface (Figure S34)\u003csup\u003e42\u003c/sup\u003e. Moreover,\u0026nbsp;we\u0026nbsp;prepared perovskites with different A-site ions and investigated charge transfer behaviour of the mixed cation perovskites. Figure S35 indicates the XRD patterns, UV-Vis spectra and SEM images of\u0026nbsp;FAPbI\u003csub\u003e3\u003c/sub\u003e and CsFAMAPbI\u003csub\u003e3\u003c/sub\u003e, respectively. The V\u003csub\u003eOC\u0026nbsp;\u003c/sub\u003eof\u0026nbsp;MAPbI\u003csub\u003e3\u003c/sub\u003e, CsFAMAPbI\u003csub\u003e3\u003c/sub\u003e and FAPbI\u003csub\u003e3\u003c/sub\u003e solar cells are 1.09 V, 1.04 V and 1.01 V, respectively (Figure S36a). The electrode potentials of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e0\u003c/sup\u003e and I\u003csup\u003e-\u003c/sup\u003e/I\u003csub\u003e2\u003c/sub\u003e of the three perovskites were measured and the results are shown in Figure S36 b-i. The electrode potentials of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e0\u003c/sup\u003e in MAPbI\u003csub\u003e3\u003c/sub\u003e, CsFAMAPbI\u003csub\u003e3\u003c/sub\u003e and FAPbI\u003csub\u003e3\u003c/sub\u003e are -1.09 V, -1.06 V and -1.04 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e, respectively, while the electrode potentials of\u0026nbsp;I\u003csup\u003e-\u003c/sup\u003e/I\u003csub\u003e2\u003c/sub\u003e of the three perovskites are very close (0.18 V vs. Fc/Fc\u003csup\u003e+\u003c/sup\u003e). The results suggest that A-site cations in the perovskites have influences on the reduction potential of Pb\u003csup\u003e2+\u003c/sup\u003e, but not the oxidation potential of I\u003csup\u003e-\u003c/sup\u003e. The predicted V\u003csub\u003eOC\u003c/sub\u003e of MAPbI\u003csub\u003e3\u003c/sub\u003e, CsFAMAPbI\u003csub\u003e3\u003c/sub\u003e and FAPbI\u003csub\u003e3\u003c/sub\u003e solar cells by faradaic junction model are 1.27 V, 1.24 V and 1.22 V, which are in good agreement with the experimental values. Therefore, faradaic junction theory can explain the behaviour of mixed cation perovskite solar cells. As for the mixed halide (I/Br) perovskites, photo-induced phase segregation process makes halide perovskites unstable, which leads to more complex the interface reactions\u003csup\u003e22,43\u003c/sup\u003e. Therefore, the mixed halide perovskites are not discussed here.\u003c/p\u003e\n\u003cp\u003eIn order to compare with the conventional energy band alignment theory, we also measured the energy band positions of TiO\u003csub\u003e2\u003c/sub\u003e ETL and Spiro-OMeTAD HTL and the results are shown in Figure S37 and S38. Schematic diagrams for interface charge transfer by energy band alignment theory are shown in Figure S39. The difference between the conduction band of TiO\u003csub\u003e2\u003c/sub\u003e and the HOMO of Spiro-OMeTAD is 1.14 eV. However, the quasi-fermi level of electrons and holes cannot be the same with the conduction band and the HOMO, and an energy loss between the fermi level and the energy band position is usually more than 0.1 eV\u003csup\u003e44\u003c/sup\u003e. Therefore, the V\u003csub\u003eOC\u003c/sub\u003e of 1.04 V can be predicted in both MAPbI\u003csub\u003e3\u003c/sub\u003e and MAPbBr\u003csub\u003e3\u003c/sub\u003e solar cells by energy band alignment theory. However, the experimental values of V\u003csub\u003eOC\u003c/sub\u003e in MAPbI\u003csub\u003e3\u003c/sub\u003e and MAPbBr\u003csub\u003e3\u003c/sub\u003e solar cells are 1.09 V and 1.41 V, which exceed the V\u003csub\u003eOC\u003c/sub\u003e values based on energy band alignment theory. Although variations in energy band positions exist, the conduction band of TiO\u003csub\u003e2\u003c/sub\u003e is reported as about -4.0 ~ -4.2 eV, and the HOMO of Spiro-OMeTAD is about -5.0 ~ -5.2 eV\u003csup\u003e1,45,46\u003c/sup\u003e. In the TiO\u003csub\u003e2\u003c/sub\u003e/perovskite/Spiro-OMeTAD solar cell, the V\u003csub\u003eOC\u003c/sub\u003e is predicted as about 0.8 ~ 1.2 V by the energy band alignment theory. However, the value is still much lower than the experimental value (1.4 V) of MAPbBr\u003csub\u003e3\u003c/sub\u003e solar cells, which cannot come from variations of energy levels. Therefore, the energy band alignment theory fails to understand the V\u003csub\u003eOC\u003c/sub\u003e in MAPbBr\u003csub\u003e3\u003c/sub\u003e solar cells. In contrast, by faradaic junction model, a V\u003csub\u003eOC\u003c/sub\u003e of 1.78 V is predicted for MAPbBr\u003csub\u003e3\u003c/sub\u003e cells by the electrode potential alignment, which is more reasonable. Moreover, the effect of energy band positions of charge transport layers on the V\u003csub\u003eOC\u003c/sub\u003e of perovskite solar cells is also considered. We used electron transport layers of TiO\u003csub\u003e2\u003c/sub\u003e and SnO\u003csub\u003e2\u003c/sub\u003e to construct MAPbI\u003csub\u003e3\u003c/sub\u003e solar cells. Though the difference of conduction bands of TiO\u003csub\u003e2\u003c/sub\u003e and SnO\u003csub\u003e2\u003c/sub\u003e is usually more than 0.2 eV (Figure S40a)\u003csup\u003e1,46\u003c/sup\u003e, a similar V\u003csub\u003eOC\u003c/sub\u003e of 1.09 V is obtained in the TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD and SnO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD (Figure S40b) solar cells. In contrast, both TiO\u003csub\u003e2\u003c/sub\u003e and SnO\u003csub\u003e2\u003c/sub\u003e show similar electrode potential window (Figure S40c). Therefore, not energy band alignment theory but electrode potential alignment in faradaic junction model is suitable to explain the V\u003csub\u003eOC\u003c/sub\u003e of perovskite solar cells.\u003c/p\u003e\n\u003cp\u003eThe origin of V\u003csub\u003eOC\u003c/sub\u003e of perovskite solar cells comes from the difference between the electrode potentials of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003e (0\u0026lt;x\u0026le;2) and I\u003csup\u003e-\u003c/sup\u003e/I\u003csup\u003e-1+x\u003c/sup\u003e (or Br\u003csup\u003e-\u003c/sup\u003e/Br\u003csup\u003e-1+x\u003c/sup\u003e, 0\u0026lt;x\u0026le;1), and their limiting potentials are those of the redox reactions of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e0\u003c/sup\u003e (x=2)and I\u003csup\u003e-\u003c/sup\u003e/I\u003csub\u003e2\u003c/sub\u003e (or Br\u003csup\u003e-\u003c/sup\u003e/Br\u003csub\u003e2\u003c/sub\u003e, x=1). The understanding on the origin of V\u003csub\u003eOC\u0026nbsp;\u003c/sub\u003eof perovskite solar cells is intrinsically different from energy band alignment theory\u0026nbsp;(Figure S41). Based on faradaic junction model, to further increase the V\u003csub\u003eOC\u003c/sub\u003e of perovskite solar cells, major efforts should be focused on enlarging the electrode potential difference by shifting Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e0\u003c/sup\u003e more negative and I\u003csup\u003e-\u003c/sup\u003e/I\u003csub\u003e2\u003c/sub\u003e (or Br\u003csup\u003e-\u003c/sup\u003e/Br\u003csub\u003e2\u003c/sub\u003e)more positive, and\u0026nbsp;suppressing the back chemical reactions of I\u003csup\u003e-1+x\u0026nbsp;\u003c/sup\u003e+ xe\u003csup\u003e-\u0026nbsp;\u003c/sup\u003e\u0026rarr;I\u003csup\u003e-\u003c/sup\u003e (or Br\u003csup\u003e-1+x\u003c/sup\u003e + e\u003csup\u003e-\u003c/sup\u003e\u0026rarr;Br\u003csup\u003e-\u003c/sup\u003e) and Pb\u003csup\u003e(2-x)+\u003c/sup\u003e + xh\u003csup\u003e+\u0026nbsp;\u003c/sup\u003e\u0026rarr;Pb\u003csup\u003e2+\u003c/sup\u003e \u003csup\u003e42,47\u003c/sup\u003e.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn summary, we found the faradaic reactions exist on the surfaces of perovskites (Pb\u003csup\u003e2+\u003c/sup\u003e + 2e\u003csup\u003e-\u003c/sup\u003e ↔ Pb\u003csup\u003e0\u003c/sup\u003e, 2I\u003csup\u003e-\u003c/sup\u003e + 2h\u003csup\u003e+\u003c/sup\u003e ↔ I\u003csub\u003e2\u003c/sub\u003e), as well as the interfaces between perovskite/TiO\u003csub\u003e2\u003c/sub\u003e and perovskite/Spiro-OMeTAD (Pb\u003csup\u003e0\u003c/sup\u003e + 2Ti\u003csup\u003e4+\u003c/sup\u003e ↔ Pb\u003csup\u003e2+\u0026nbsp;\u003c/sup\u003e+ 2Ti\u003csup\u003e3+\u003c/sup\u003e, I\u003csub\u003e2\u003c/sub\u003e + Spiro-H\u003csup\u003e+\u003c/sup\u003e ↔ I\u003csup\u003e-\u003c/sup\u003e + Spiro-I\u003csup\u003e-\u003c/sup\u003e + H\u003csup\u003e+\u003c/sup\u003e) by connecting a perovskite/ETL half device with a perovskite/HTL half device to simulate an all-solid-state device under illumination. Moreover, the faradaic oxidation reactions of Spiro-OMeTAD were also investigated in detail by in situ FTIR and NMR, which suggest that faradaic reactions happen not only on the surface of inorganic substances, but also on organics substances. Based on faradaic junction model, the V\u003csub\u003eOC\u003c/sub\u003e in perovskite solar cells was predicted by the difference between the electrode potentials of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e(2-x)+\u003c/sup\u003e (0\u0026lt;x≤2) and I\u003csup\u003e-\u003c/sup\u003e/I\u003csup\u003e-1+x\u003c/sup\u003e (or Br\u003csup\u003e-\u003c/sup\u003e/Br\u003csup\u003e-1+x\u003c/sup\u003e, 0\u0026lt;x≤1), and their limiting potentials are those of the redox reactions of Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e0\u003c/sup\u003e(x=2)and I\u003csup\u003e-\u003c/sup\u003e/I\u003csub\u003e2\u003c/sub\u003e (or Br\u003csup\u003e-\u003c/sup\u003e/Br\u003csub\u003e2\u003c/sub\u003e, x=1) on the surface of MAPbI\u003csub\u003e3\u003c/sub\u003e (MAPbBr\u003csub\u003e3\u003c/sub\u003e). By enlarging the electrode potential difference between Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e0\u003c/sup\u003e and I\u003csup\u003e-\u003c/sup\u003e/I\u003csub\u003e2\u003c/sub\u003e (or Br\u003csup\u003e-\u003c/sup\u003e/Br\u003csub\u003e2\u003c/sub\u003e) and suppressing the back chemical reactions of I\u003csup\u003e-1+x\u0026nbsp;\u003c/sup\u003e+ xe\u003csup\u003e-\u0026nbsp;\u003c/sup\u003e→I\u003csup\u003e-\u003c/sup\u003e (or Br\u003csup\u003e-1+x\u003c/sup\u003e + e\u003csup\u003e-\u003c/sup\u003e→Br\u003csup\u003e-\u003c/sup\u003e) and Pb\u003csup\u003e(2-x)+\u003c/sup\u003e + xh\u003csup\u003e+\u0026nbsp;\u003c/sup\u003e→Pb\u003csup\u003e2+\u003c/sup\u003e, the V\u003csub\u003eOC\u003c/sub\u003e of perovskite solar cells can be further increased. Faradaic junction model can not only explain the origin of V\u003csub\u003eOC\u003c/sub\u003e in perovskite solar cells, but also provide guidance for the improvement of organic solar cells and organic light-emitting diodes.\u003c/p\u003e\n\u003cp\u003e\u003cbr\u003e\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Natural Science Foundation of China (22425202, 22279052, 22372067, 51902121, and 22372078) and the National Key Research and Development Program of China (2024YFB4609400,\u0026nbsp;2021YFF0500501).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eW. L. supervised the project, proposed the concept and designed the experiments. M. X. carried out sample preparation, characterization and electrochemistry measurement; Y. C. conducted the fs-TAS measurements and analyzed the data. W. L. and M. X. analyzed the data and wrote the paper. All authors discussed the results and gave comments on the manuscript.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupplementary data\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplementary Information is available for this paper.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCorrespondence and requests for materials\u0026nbsp;\u003c/strong\u003eshould be addressed to Wenjun Luo.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eJena, A. K., Kulkarni, A. \u0026amp; Miyasaka, T. Halide perovskite photovoltaics: Background, status, and future prospects. \u003cem\u003eChem. Rev.\u003c/em\u003e \u003cstrong\u003e119\u003c/strong\u003e, 3036-3103, (2019).\u003c/li\u003e\n\u003cli\u003eZhou, Y., Herz, L. M., Jen, A. K. Y. \u0026amp; Saliba, M. Advances and challenges in understanding the microscopic structure-property-performance relationship in perovskite solar cells. \u003cem\u003eNat. Energy\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 794-807, (2022).\u003c/li\u003e\n\u003cli\u003e\u003cem\u003eBest research-cell efficiencies. NREL\u003c/em\u003e, https://www.nrel.gov/pv/cell-efficiency.html.\u003c/li\u003e\n\u003cli\u003eZhang, H., Pfeifer, L., Zakeeruddin, S. M., Chu, J. \u0026amp; Gr\u0026auml;tzel, M. Tailoring passivators for highly efficient and stable perovskite solar cells. \u003cem\u003eNat. Rev. Chem.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 632-652, (2023).\u003c/li\u003e\n\u003cli\u003eKojima, A., Teshima, K., Shirai, Y. \u0026amp; Miyasaka, T. Organometal halide perovskites as visible-light sensitizers for photovoltaic cells. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e131\u003c/strong\u003e, 6050-6051, (2009).\u003c/li\u003e\n\u003cli\u003eGuo, Z., Jena, A. K., Kim, G. M. \u0026amp; Miyasaka, T. The high open-circuit voltage of perovskite solar cells: A review. \u003cem\u003eEnergy Environ. Sci.\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 3171-3222, (2022).\u003c/li\u003e\n\u003cli\u003eLiu, Z. et al. Open-circuit voltages exceeding 1.26 V in planar methylammonium lead iodide perovskite solar cells. \u003cem\u003eACS Energy Lett.\u003c/em\u003e \u003cstrong\u003e4\u003c/strong\u003e, 110-117, (2019).\u003c/li\u003e\n\u003cli\u003eHe, R. et al. Improving interface quality for 1-cm\u003csup\u003e2\u003c/sup\u003e all-perovskite tandem solar cells. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e618\u003c/strong\u003e, 80-86, (2023).\u003c/li\u003e\n\u003cli\u003eWang, R. et al. Prospects for metal halide perovskite-based tandem solar cells. \u003cem\u003eNat. Photonics\u003c/em\u003e \u003cstrong\u003e15\u003c/strong\u003e, 411-425, (2021).\u003c/li\u003e\n\u003cli\u003eWang, Z. et al. Suppressed phase segregation for triple-junction perovskite solar cells. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e618\u003c/strong\u003e, 74-79, (2023).\u003c/li\u003e\n\u003cli\u003eXia, J., Sohail, M. \u0026amp; Nazeeruddin, M. K. Efficient and stable perovskite solar cells by tailoring of interfaces. \u003cem\u003eAdv. Mater.\u003c/em\u003e \u003cstrong\u003e35\u003c/strong\u003e, 2211324, (2023).\u003c/li\u003e\n\u003cli\u003eZhang, T. et al. Ion-modulated radical doping of Spiro-OMeTAD for more efficient and stable perovskite solar cells. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e377\u003c/strong\u003e, 495-501, (2022).\u003c/li\u003e\n\u003cli\u003eLin, Y. et al. Matching charge extraction contact for wide-bandgap perovskite solar cells. \u003cem\u003eAdv. Mater.\u003c/em\u003e \u003cstrong\u003e29\u003c/strong\u003e, 1700607, (2017).\u003c/li\u003e\n\u003cli\u003eXu, F., Zhang, M., Li, Z., Yang, X. \u0026amp; Zhu, R. Challenges and perspectives toward future wide-bandgap mixed-halide perovskite photovoltaics. \u003cem\u003eAdv. Energy Mater.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 2203911, (2023).\u003c/li\u003e\n\u003cli\u003eCaprioglio, P. et al. On the relation between the open-circuit voltage and quasi-fermi level splitting in efficient perovskite solar cells. \u003cem\u003eAdv. Energy Mater.\u003c/em\u003e \u003cstrong\u003e9\u003c/strong\u003e, 1901631, (2019).\u003c/li\u003e\n\u003cli\u003eStolterfoht, M. et al. The impact of energy alignment and interfacial recombination on the internal and external open-circuit voltage of perovskite solar cells. \u003cem\u003eEnergy Environ. Sci.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 2778-2788, (2019).\u003c/li\u003e\n\u003cli\u003eBelisle, R. A., Jain, P., Prasanna, R., Leijtens, T. \u0026amp; McGehee, M. D. Minimal effect of the hole-transport material ionization potential on the open-circuit voltage of perovskite solar cells. \u003cem\u003eACS Energy Lett.\u003c/em\u003e \u003cstrong\u003e1\u003c/strong\u003e, 556-560, (2016).\u003c/li\u003e\n\u003cli\u003eDong, C. et al. Carbon-based all-inorganic perovskite solar cells: Progress, challenges and strategies toward 20% efficiency. \u003cem\u003eMater. Today\u003c/em\u003e \u003cstrong\u003e50\u003c/strong\u003e, 239-258, (2021).\u003c/li\u003e\n\u003cli\u003eLiang, Y. et al. Achieving high open-circuit voltages up to 1.57 V in hole-transport-material-free MAPbBr\u003csub\u003e3\u003c/sub\u003e solar cells with carbon electrodes. \u003cem\u003eAdv. Energy Mater.\u003c/em\u003e \u003cstrong\u003e8\u003c/strong\u003e, 1701159, (2017).\u003c/li\u003e\n\u003cli\u003eLi, Y. et al. Lattice modulation of alkali metal cations doped Cs\u003csub\u003e1\u0026minus;x\u003c/sub\u003eR\u003csub\u003ex\u003c/sub\u003ePbBr\u003csub\u003e3\u003c/sub\u003e halides for inorganic perovskite solar cells. \u003cem\u003eSolar RRL\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 1800164, (2018).\u003c/li\u003e\n\u003cli\u003eLiu, Y. et al. Direct observation of photoinduced ion migration in lead halide perovskites. \u003cem\u003eAdv. Funct. Mater.\u003c/em\u003e \u003cstrong\u003e31\u003c/strong\u003e, 2008777, (2020).\u003c/li\u003e\n\u003cli\u003eFrolova, L. A. et al. Reversible Pb\u003csup\u003e2+\u003c/sup\u003e/Pb\u003csup\u003e0\u003c/sup\u003e and I\u003csup\u003e-\u003c/sup\u003e/I\u003csub\u003e3\u003c/sub\u003e\u003csup\u003e-\u003c/sup\u003e redox chemistry drives the light‐induced phase segregation in all-inorganic mixed halide perovskites. \u003cem\u003eAdv. Energy Mater.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 2002934, (2021).\u003c/li\u003e\n\u003cli\u003eChen, X. et al. Reversible charge transfer and adjustable potential window in semiconductor/faradaic layer/liquid junctions. \u003cem\u003eiScience\u003c/em\u003e \u003cstrong\u003e23\u003c/strong\u003e, 100949, (2020).\u003c/li\u003e\n\u003cli\u003eYin, Z. et al. Mildly regulated intrinsic faradaic layer at the oxide/water interface for improved photoelectrochemical performance. \u003cem\u003eChem. Sci.\u003c/em\u003e \u003cstrong\u003e11\u003c/strong\u003e, 6297-6304, (2020).\u003c/li\u003e\n\u003cli\u003eChen, M. et al. Faradaic junction and isoenergetic charge transfer mechanism on semiconductor/semiconductor interfaces. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 6363, (2021).\u003c/li\u003e\n\u003cli\u003eWang, P. et al. Photovoltage memory effect in a portable faradaic junction solar rechargeable device. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e13\u003c/strong\u003e, 2544, (2022).\u003c/li\u003e\n\u003cli\u003eXue, M. et al. Bipolarized intrinsic faradaic layer on a semiconductor surface under illumination. \u003cem\u003eNatl. Sci. Rev.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, nwac249, (2023).\u003c/li\u003e\n\u003cli\u003eDu, G., Yang, L. \u0026amp; Zhang, J. Light soaking induced halide doping of evaporated spiro‐ometad in perovskite solar cells. \u003cem\u003eLaser Photonics Rev.\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 2200475, (2022).\u003c/li\u003e\n\u003cli\u003eTrchov\u0026aacute;, M., \u0026Scaron;eděnkov\u0026aacute;, I., Tobolkov\u0026aacute;, E. \u0026amp; Stejskal, J. FTIR spectroscopic and conductivity study of the thermal degradation of polyaniline films. \u003cem\u003ePolym. Degrad. Stabil.\u003c/em\u003e \u003cstrong\u003e86\u003c/strong\u003e, 179-185, (2004).\u003c/li\u003e\n\u003cli\u003eMaestre-Reyna, M. et al. Visualizing the DNA repair process by a photolyase at atomic resolution. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e382\u003c/strong\u003e, eadd7795, (2023).\u003c/li\u003e\n\u003cli\u003eHoffmann, M. R., Martin, S. T., Choi, W. \u0026amp; Bahneman, D. W. Environmental applications of semiconductor photocatalysis. \u003cem\u003eChem. Rev.\u003c/em\u003e \u003cstrong\u003e95\u003c/strong\u003e, 69-96, (1995).\u003c/li\u003e\n\u003cli\u003eXi, H. et al. Performance enhancement of planar heterojunction perovskite solar cells through tuning the doping properties of hole-transporting materials. \u003cem\u003eACS Omega\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 326-336, (2017).\u003c/li\u003e\n\u003cli\u003eLi, Z. et al. Acid additives enhancing the conductivity of Spiro‐OMeTAD toward high-efficiency and hysteresis-less planar perovskite solar cells. \u003cem\u003eAdv. Energy Mater.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 1601451, (2016).\u003c/li\u003e\n\u003cli\u003eJeong, M. et al. Stable perovskite solar cells with efficiency exceeding 24.8% and 0.3-V voltage loss. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e369\u003c/strong\u003e, 1615-1620, (2020).\u003c/li\u003e\n\u003cli\u003eHaruna, K., Alenaizan, A. A. \u0026amp; Al-Saadi, A. A. Density functional theory study of the substituent effect on the structure, conformation and vibrational spectra in halosubstituted anilines. \u003cem\u003eRSC Adv.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 67794-67804, (2016).\u003c/li\u003e\n\u003cli\u003eBakulin, A. A. et al. The role of driving energy and delocalized states for charge separation in organic semiconductors. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e335\u003c/strong\u003e, 1340-1344, (2012).\u003c/li\u003e\n\u003cli\u003eSamu, G. F. \u0026amp; Jan\u0026aacute;ky, C. Photocorrosion at irradiated perovskite/electrolyte interfaces. \u003cem\u003eJ. Am. Chem. Soc.\u003c/em\u003e \u003cstrong\u003e142\u003c/strong\u003e, 21595-21614, (2020).\u003c/li\u003e\n\u003cli\u003eGagne, R. R., Koval, C. A. \u0026amp; Lisensky, G. C. Ferrocene as an internal standard for electrochemical measurements. \u003cem\u003eInorg. Chem.\u003c/em\u003e \u003cstrong\u003e19\u003c/strong\u003e, 2854-2855, (2002).\u003c/li\u003e\n\u003cli\u003eDuBose, J. T. \u0026amp; Kamat, P. V. Probing perovskite photocatalysis. Interfacial electron transfer between CsPbBr\u003csub\u003e3\u003c/sub\u003e and ferrocene redox couple. \u003cem\u003eJ. Phys. Chem. Lett.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 6074-6080, (2019).\u003c/li\u003e\n\u003cli\u003eMathew, P. S., Samu, G. F., Jan\u0026aacute;ky, C. \u0026amp; Kamat, P. V. Iodine (I) expulsion at photoirradiated mixed halide perovskite interface. Should I stay or should I go? \u003cem\u003eACS Energy Lett.\u003c/em\u003e \u003cstrong\u003e5\u003c/strong\u003e, 1872-1880, (2020).\u003c/li\u003e\n\u003cli\u003eZhu, H. et al. Efficient and stable large bandgap MAPbBr\u003csub\u003e3\u003c/sub\u003e perovskite solar cell attaining an open circuit voltage of 1.65 V. \u003cem\u003eACS Energy Lett.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 1112-1119, (2022).\u003c/li\u003e\n\u003cli\u003eCao, D. et al. Cathodic shift of onset potential for water oxidation on a Ti\u003csup\u003e4+ \u003c/sup\u003edoped Fe\u003csub\u003e2\u003c/sub\u003eO\u003csub\u003e3 \u003c/sub\u003ephotoanode by suppressing the back reaction. \u003cem\u003eEnergy Environ. Sci.\u003c/em\u003e \u003cstrong\u003e7\u003c/strong\u003e, 752-759, (2014).\u003c/li\u003e\n\u003cli\u003eHoke, E. T. et al. Reversible photo-induced trap formation in mixed-halide hybrid perovskites for photovoltaics. \u003cem\u003eChem. Rev.\u003c/em\u003e \u003cstrong\u003e6\u003c/strong\u003e, 613-617, (2015).\u003c/li\u003e\n\u003cli\u003eButler, M. A. \u0026amp; Ginley, D. S. Prediction of flatband potentials at semiconductor-electrolyte interfaces from atomic electronegativities. \u003cem\u003eJ. Electrochem. Soc.\u003c/em\u003e \u003cstrong\u003e125\u003c/strong\u003e, 228-232, (1978).\u003c/li\u003e\n\u003cli\u003eHu, Z. et al. The impact of atmosphere on energetics of lead halide perovskites. \u003cem\u003eAdv. Energy Mater.\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 2000908, (2020).\u003c/li\u003e\n\u003cli\u003eYuan, Y., Yan, G., Hong, R., Liang, Z. \u0026amp; Kirchartz, T. Quantifying efficiency limitations in all-inorganic halide perovskite solar cells. \u003cem\u003eAdv. Mater.\u003c/em\u003e \u003cstrong\u003e34\u003c/strong\u003e, 2108132, (2022).\u003c/li\u003e\n\u003cli\u003eZhang, J. et al. A high-voltage solar rechargeable device based on a CoPi/BiVO\u003csub\u003e4\u003c/sub\u003e faradaic junction. \u003cem\u003eJ. Mater. Chem. A\u003c/em\u003e \u003cstrong\u003e10\u003c/strong\u003e, 1802-1807, (2022).\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Methods","content":"\u003cp\u003e\u003cstrong\u003eMaterials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eCH\u003csub\u003e3\u003c/sub\u003eMH\u003csub\u003e3\u003c/sub\u003eI (MAI), CH\u003csub\u003e3\u003c/sub\u003eMH\u003csub\u003e3\u003c/sub\u003eBr (MABr), PbI\u003csub\u003e2\u003c/sub\u003e, PbBr\u003csub\u003e2\u003c/sub\u003e, Spiro-OMeTAD (99.8%), 4-tert-butylpyridine (tBP) and LiTFSI were purchased from Xi’an Polymer\u0026nbsp;Light Technology. Titanium(diisopropoxide) bis(2,4-pentanedionate) in isopropanol (75%), N, N-dimethylformamide (DMF, 99.8%), dimethyl sulfoxide (DMSO, 99.8%) and tin (IV) oxide (SnO\u003csub\u003e2\u003c/sub\u003e) colloid solution (15% in H\u003csub\u003e2\u003c/sub\u003eO colloidal dispersion) were purchased from Alfa Aesar. Anhydrous ethanol (99.8%) and Br\u003csub\u003e2\u003c/sub\u003e was purchased from Sinopharm Chemical Reagent. Ethyl acetate (anhydrous, 99.8%), acetonitrile (99.8%), dimethyl sulfoxide-d6 (DMSO-d6), and ferrocene (Fc) were purchased from Sigma-Aldrich. I\u003csub\u003e2\u003c/sub\u003e was purchased from Aladdin. Tetrabutylammonium hexafluorophosphate (n-Bu\u003csub\u003e4\u003c/sub\u003eNPF\u003csub\u003e6\u003c/sub\u003e)) and dichloromethane (DCM, CH\u003csub\u003e2\u003c/sub\u003eCl\u003csub\u003e2\u003c/sub\u003e) were purchased from Macklin. All materials were used without further purification. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of samples\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eFTO substrates (1.5 cm*1.5 cm) were cleaned sequentially by deionized water, acetone, and anhydrous ethanol. Before the film deposition, FTO substrates were treated by the UV-ozone for 30 min.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of TiO\u003csub\u003e2\u003c/sub\u003e films\u0026nbsp;\u003c/strong\u003eA compact TiO\u003csub\u003e2\u003c/sub\u003e (c-TiO\u003csub\u003e2\u003c/sub\u003e) layer was prepared by spin-coating at 3000r for 30s, and then transferred to the muffle furnace and annealed at 480 \u003csup\u003eo\u003c/sup\u003eC for 40 min with the heating rate of 2\u003csup\u003e\u0026nbsp;o\u003c/sup\u003eC/min\u003csup\u003e48\u003c/sup\u003e. The precursor solution of c-TiO\u003csub\u003e2\u003c/sub\u003e was prepared by the mixture of titanium(diisopropoxide) bis(2,4-pentanedionate) in isopropanol and anhydrous ethanol (v : v = 1 : 10).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of SnO\u003csub\u003e2\u003c/sub\u003e films\u0026nbsp;\u003c/strong\u003eA SnO\u003csub\u003e2\u003c/sub\u003e layer was prepared by spin-coating at 4000r for 30s, and then annealed at 150 \u003csup\u003eo\u003c/sup\u003eC for 30 min on a hotplate in air. The precursor solution of SnO\u003csub\u003e2\u003c/sub\u003e was prepared by the mixture of SnO\u003csub\u003e2\u003c/sub\u003e colloid solution and deionized water (v : v = 1 : 4).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of MAPbI\u003csub\u003e3\u003c/sub\u003e films\u0026nbsp;\u003c/strong\u003eAll of the perovskite films were prepared by one-step method in nitrogen-filled glovebox\u003csup\u003e19,49\u003c/sup\u003e. The precursor solution for MAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003ewas obtained by dissolving 461 mg PbI\u003csub\u003e2\u003c/sub\u003e and 159 mg MAI in the mixture of DMF and DMSO (1 mL, v : v = 4 : 1). The solution was spin-coated on the different substrates at 1500r for 10s and 4000r for 40s, respectively. A drop of 500\u0026nbsp;μL ethyl acetate as antisolvent was poured on the film at 15 s after the starting of the spin-coating procedure. The MAPbI\u003csub\u003e3\u003c/sub\u003e films were then heated on a hotplate at 100 \u003csup\u003eo\u003c/sup\u003eC for 10 min and dense MAPbI\u003csub\u003e3\u003c/sub\u003e films were prepared. The sparse MAPbI\u003csub\u003e3\u003c/sub\u003e films were prepared with the same method but without the treatment of ethyl acetate antisolvent.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of MAPbBr\u003csub\u003e3\u003c/sub\u003e films\u0026nbsp;\u003c/strong\u003eThe precursor solution for MAPbBr\u003csub\u003e3\u003c/sub\u003e was obtained by dissolving 367 mg PbBr\u003csub\u003e2\u003c/sub\u003e and 112 mg MABr in the mixture of DMF and DMSO (1 mL, v : v = 4 : 1). The solution was spin-coated on the different substrates at 4000r for 40s. A drop of 500\u0026nbsp;μL ethyl acetate as antisolvent was poured on the film at 20 s after the starting of the spin-coating procedure. The MAPbBr\u003csub\u003e3\u003c/sub\u003e films were then heated on a hotplate at 100 \u003csup\u003eo\u003c/sup\u003eC for 10 min and dense MAPbBr\u003csub\u003e3\u003c/sub\u003e films were prepared.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of FAPbI\u003csub\u003e3\u003c/sub\u003e films\u0026nbsp;\u003c/strong\u003eThe precursor solution for FAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003ewas obtained by dissolving 880 mg PbI\u003csub\u003e2\u003c/sub\u003e, 309.5 mg FAI and 42.54 mg MACl in the mixture of DMF and DMSO (1 mL, v : v = 3 : 1)\u003csup\u003e50\u003c/sup\u003e. The solution was spin-coated on the different substrates at 5000r for 30s. A drop of 750\u0026nbsp;μL diethyl ether as antisolvent was poured on the film at 10 s after the starting of the spin-coating procedure. The FAPbI\u003csub\u003e3\u003c/sub\u003e films were then heated at 150 \u003csup\u003eo\u003c/sup\u003eC for 20 min on a hotplate in air and dense FAPbI\u003csub\u003e3\u003c/sub\u003e films were prepared.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of CsFAMAPbI\u003csub\u003e3\u003c/sub\u003e films\u0026nbsp;\u003c/strong\u003eThe CsI solution was prepared by dissolving 75 mg of CsI in 500\u0026nbsp;μL of DMSO solution\u003csup\u003e51\u003c/sup\u003e. The precursor solution for CsFAMAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003ewas obtained by dissolving 472 mg PbI\u003csub\u003e2\u003c/sub\u003e, 243 mg FAI, 14 mg MABr, 65.6 mg PbBr and 44\u0026nbsp;μL CsI solution in the mixture of 681\u0026nbsp;μL DMF and 170\u0026nbsp;μL DMSO. The solution was spin-coated on the different substrates at 3000r for 10s and 6000r for 30s, respectively. A drop of 500\u0026nbsp;μL chlorobenzene as antisolvent was poured on the film at 30 s after the starting of the spin-coating procedure. The CsFAMAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003efilms were then heated at 100 \u003csup\u003eo\u003c/sup\u003eC for 1 hour on a hotplate and dense CsFAMAPbI\u003csub\u003e3\u0026nbsp;\u003c/sub\u003efilms were prepared.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of Spiro-OMeTAD films\u0026nbsp;\u003c/strong\u003eAll of the Spiro-OMeTAD films in this study were doped with Li-TFSI and tBP. The precursor solution was prepared by dissolving 80 mg Spiro-OMeTAD, 28.8\u0026nbsp;μL tBP and 17.8\u0026nbsp;μL Li-TFSI (520 mg/mL in acetonitrile) in 1 mL chlorobenzene\u003csup\u003e52\u003c/sup\u003e. The Spiro-OMeTAD films were spin-coated on different substrates at 3000r for 30s in nitrogen-filled glovebox. The films were kept in dry air at room temperature for oxidation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePreparation of Au films\u0026nbsp;\u003c/strong\u003eA 100 nm thick Au layer was deposited by thermal evaporation.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCharacterization of samples\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe crystal structures of the samples were characterized by X-ray diffraction (XRD, Rigaku D/MAX-Ultima III). The morphologies of the samples were investigated by scanning electron microscope (SEM, Gemini 500) with an accelerating voltage of 10 kV. The energy-dispersive X-ray spectroscopy (EDS) was carried out by Gemini 500 equipped with an Oxford EDS system. The optical absorption was obtained by UV-Visible spectrophotometer (Shimadzu, UV 2550; Varian, Cary 50). Photoluminescence (PL) measurements were performed on a Varian Cary Eclipse. The band positions were measured by Ultraviolet Photoelectron spectroscopy (UPS, Thermofisher Escalab 250Xi, He I hv = 21.22 eV)\u003c/p\u003e\n\u003cp\u003eThe X-ray photoelectron spectroscopy (XPS, Thermofisher Escalab 250Xi) were performed with an Al Kα X-ray source. The binding energy of the C 1s peak at 284.6 eV was used to calibrate the XPS data. The samples of MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e and MAPbI\u003csub\u003e3\u003c/sub\u003e/ Spiro-OMeTADwere illuminated under a Xe lamp (300 W) for 5 min and were characterized by in situ XPS. For the simulated full device, a MAPbI\u003csub\u003e3\u003c/sub\u003e/TiO\u003csub\u003e2\u003c/sub\u003e half device was connected with a MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD half device by indium wires, and both of the two half devices were are illuminated under a Xe lamp (300 W).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe Fourier Transform Infrared Spectroscopy (FTIR, Shimadazu IR Prestige-21) was performed with a resolution of 8 cm\u003csup\u003e-1\u003c/sup\u003e. In situ FTIR was performed in the dark and under the illumination of a 405 nm-laser light (100 mW) for 20 min. In open circuit conditions, FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD was illuminated. In short circuit condition, the two ends of FTO/TiO\u003csub\u003e2\u003c/sub\u003e/MAPbI\u003csub\u003e3\u003c/sub\u003e/Spiro-OMeTAD/Au were connected by indium wires, and illuminated. To further investigate the details of oxidation reactions of Spiro-OMeTAD by I\u003csub\u003e2\u003c/sub\u003e and Br\u003csub\u003e2\u003c/sub\u003e in the dark, the FTO/Spiro-OMeTAD films were exposed to I\u003csub\u003e2\u003c/sub\u003e or Br\u003csub\u003e2\u003c/sub\u003e vapor in a closed glass vial in the dark for different time.\u003c/p\u003e\n\u003cp\u003eThe \u003csup\u003e1\u003c/sup\u003eH Nuclear Magnetic Resonance (NMR) spectra was performed with DMSO-d6 as solvent on a Bruker 600’ 54 Ascend LH 600MHz. Similar to the precursor solution for Spiro-OMeTAD films, 80 mg Spiro-OMeTAD and 17.8\u0026nbsp;μL Li-TFSI (520 mg/mL in acetonitrile) were dissolved in 1mL tetrahydrofuran, and then were evaporated. The Li-TFSI doped Spiro-OMeTAD powder was obtained, which were then exposed to I\u003csub\u003e2\u003c/sub\u003e vapor and Br\u003csub\u003e2\u003c/sub\u003e (dissolved in isopropanol) in a closed glass vial in the dark for different time for NMR measurement.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe femtosecond transient absorption spectroscopy (fs-TAS) measurements were performed with a femtosecond broadband pump-probe system (HARPIA, Light Conversion company. Briefly, a 200 μJ pulse, was generated at a ~100 fs pulse width at 1030 nm with a repetition rate of 75 kHz using a regeneratively amplified Yb:KGW laser (PHAROS, Light Conversion). The output of the laser beam was split into two beams by a beam splitter. The larger portion of the beam passed through an optical parametric amplifier (ORPHEUS-HP, Light Conversion,) to generate a pump beam. The small portion of the fundamental beam passed through a second harmonic nonlinear crystal (BBO) to generate 515 nm, and then the 515 nm laser beam was focused onto a sapphire plate to generate supercontinuum white light as the probe beam. The absorption change (ΔA) of kinetic traces was measured at 6369 nm with a laser pulse of 700 nm and a fluence of 246 μJ cm\u003csup\u003e−2\u003c/sup\u003e. The excited beam spot size (diameter) was of ~0.4 mm. The data were analyzed using CarpetView software (Light Conversion).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eElectrochemical measurements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe electrochemical measurement of films was investigated in a three-electrode cell by using an electrochemical workstation (Shanghai Chenhua, CHI 760e). The prepared films were used as the working electrodes. A nonaqueous Ag electrode was used as the pseudo-reference electrode and a Pt wire as used as a counter electrode\u003csup\u003e53\u003c/sup\u003e. The electrolyte was 0.1M n-Bu\u003csub\u003e4\u003c/sub\u003eNPF\u003csub\u003e6\u003c/sub\u003e in CH\u003csub\u003e2\u003c/sub\u003eCl\u003csub\u003e2\u003c/sub\u003e. 1 mM ferrocene was used as the internal reference and all the potentials were calibrated with reference to the ferrocene/ferrocenium (Fc/Fc\u003csup\u003e+\u003c/sup\u003e) redox couple. The vacuum energy level of Fc/Fc\u003csup\u003e+\u0026nbsp;\u003c/sup\u003eis -4.9 eV.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003ePhotovoltaic measurements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe current density-voltage (J-V) characteristics were measured in a nitrogen-filled glovebox by a sourcemeter (EnliTech, B2091A) under the illumination of a solar simulator (EnliTech) at the light intensity of 100 mW/cm\u003csup\u003e2\u003c/sup\u003e (calibrated by a crystalline Si reference solar cell). The active area was 0.11 cm\u003csup\u003e2\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eReferences\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e48.\u0026nbsp; \u0026nbsp;Feng, J. et al. Curing the fundamental issue of impurity phases in two-step solution-processed CsPbBr\u003csub\u003e3\u003c/sub\u003e perovskite films. \u003cem\u003eSci. Bull.\u003c/em\u003e \u003cstrong\u003e65\u003c/strong\u003e, 726-737, (2020).\u003c/p\u003e\n\u003cp\u003e49.\u0026nbsp; \u0026nbsp;Ravishankar, S. et al. Influence of charge transport layers on open-circuit voltage and hysteresis in perovskite solar cells. \u003cem\u003eJoule\u003c/em\u003e \u003cstrong\u003e2\u003c/strong\u003e, 788-798, (2018).\u003c/p\u003e\n\u003cp\u003e50.\u0026nbsp; \u0026nbsp;Hui, W. et al. Stabilizing black-phase formamidinium perovskite formation at room temperature and high humidity. \u003cem\u003eScience\u003c/em\u003e \u003cstrong\u003e371\u003c/strong\u003e, 1359-1364, (2021).\u003c/p\u003e\n\u003cp\u003e51.\u0026nbsp; \u0026nbsp;Dong, Q. et al. Interpenetrating interfaces for efficient perovskite solar cells with high operational stability and mechanical robustness. \u003cem\u003eNat. Commun.\u003c/em\u003e \u003cstrong\u003e12\u003c/strong\u003e, 973, (2021).\u003c/p\u003e\n\u003cp\u003e52.\u0026nbsp; \u0026nbsp;Burschka, J. et al. Sequential deposition as a route to high-performance perovskite-sensitized solar cells. \u003cem\u003eNature\u003c/em\u003e \u003cstrong\u003e499\u003c/strong\u003e, 316-319, (2013).\u003c/p\u003e\n\u003cp\u003e53.\u0026nbsp; \u0026nbsp;Joseph, V. et al. Stable perovskite solar cells using molecularly engineered functionalized oligothiophenes as low-cost hole-transporting materials. \u003cem\u003eSmall\u003c/em\u003e \u003cstrong\u003e17\u003c/strong\u003e, 2100783, (2021).\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"","lastPublishedDoi":"10.21203/rs.3.rs-6295095/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6295095/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003ePerovskite solar cells have attracted much interest due to the very fast increasement of power conversion efficiency (PCE) as well as the low-cost solution processing, excellent absorption coefficient, long charge carrier diffusion length. To date, further improvement of PCE in perovskite solar cells is mainly limited by the open-circuit voltage (V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eOC\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e). In previous studies, the origin of V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eOC\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e is usually explained by the energy band alignment theory. However, in some experiments, the V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eOC \u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003edoes not change apparently when the band positions of electron or hole transport layers are changed. Therefore, the energy band alignment theory is not suitable to explain the origin of V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eOC\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e in perovskite solar cells, and it’s desirable to reveal the origin of V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eOC \u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003eto further improve the PCE of perovskite solar cells. Here, we develop a new method to simulate the interface charge transfer process in the all solid-state devices in working conditions by connecting perovskite/ETL half device with perovskite/HTL half device under simultaneous illumination. Reversible reduction reactions of Pb\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e2+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e/Pb\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e0\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e and oxidation reactions of I\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e-\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e/I\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e are observed on MAPbI\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e3\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e surfaces under illumination, respectively. Moreover, the oxidation processes of Spiro-OMeTAD are also investigated by in situ FTIR and NMR. We find that the V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eOC\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e comes from the difference between the electrode potentials of Pb\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e2+\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e/Pb\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e0\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e and X\u003c/strong\u003e\u003csup\u003e\u003cstrong\u003e-\u003c/strong\u003e\u003c/sup\u003e\u003cstrong\u003e/X\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003e2\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e (X=I, Br) on perovskite surfaces. The results suggest that the origin of V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eOC\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e does not come from the energy band alignment but from the electrode potential alignment, which can offer new perspectives to improve the V\u003c/strong\u003e\u003csub\u003e\u003cstrong\u003eOC\u003c/strong\u003e\u003c/sub\u003e\u003cstrong\u003e of perovskite solar cells.\u003c/strong\u003e\u003c/p\u003e","manuscriptTitle":"Origin of VOC in Perovskite Solar Cells by Faradaic Junction Model","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-04-15 08:29:33","doi":"10.21203/rs.3.rs-6295095/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f307cea0-d2f6-4c36-8b06-8b42e907c0dc","owner":[],"postedDate":"April 15th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":47085128,"name":"Physical sciences/Chemistry/Photochemistry/Solar cells"},{"id":47085129,"name":"Physical sciences/Materials science/Materials for energy and catalysis/Solar cells"}],"tags":[],"updatedAt":"2025-12-01T09:30:15+00:00","versionOfRecord":{"articleIdentity":"rs-6295095","link":"https://doi.org/10.1007/s11426-025-3073-0","journal":{"identity":"science-china-chemistry","isVorOnly":true,"title":"Science China Chemistry"},"publishedOn":"2025-10-20 00:00:00","publishedOnDateReadable":"October 20th, 2025"},"versionCreatedAt":"2025-04-15 08:29:33","video":"","vorDoi":"10.1007/s11426-025-3073-0","vorDoiUrl":"https://doi.org/10.1007/s11426-025-3073-0","workflowStages":[]},"version":"v1","identity":"rs-6295095","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6295095","identity":"rs-6295095","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}