As a new generation of solar cells, organic–inorganic hybrid metal halide perovskite solar cells (PSCs) have sparked great attention due to their astonishing boosted power conversion efficiencies from 3.8% to 25.7% in the last few years.1–41. J. Jeong, M. Kim, J. Seo, H. Lu, P. Ahlawat, A. Mishra, Y. Yang, M. A. Hope, F. T. Eickemeyer, M. Kim, Y. J. Yoon, I. W. Choi, B. P. Darwich, S. J. Choi, Y. Jo, J. H. Lee, B. Walker, S. M. Zakeeruddin, L. Emsley, U. Rothlisberger, A. Hagfeldt, D. S. Kim, M. Gratzel, and J. Y. Kim, Nature 592, 381 (2021). https://doi.org/10.1038/s41586-021-03406-52. M. Kim, J. Jeong, H. Lu, T. K. Lee, F. T. Eickemeyer, Y. Liu, I. W. Choi, S. J. Choi, Y. Jo, H. B. Kim, S. I. Mo, Y. K. Kim, H. Lee, N. G. An, S. Cho, W. R. Tress, S. M. Zakeeruddin, A. Hagfeldt, J. Y. Kim, M. Gratzel, and D. S. Kim, Science 375, 302 (2022). https://doi.org/10.1126/science.abh18853. Q. Jiang, J. Tong, Y. Xian, R. A. Kerner, S. P. Dunfield, C. Xiao, R. A. Scheidt, D. Kuciauskas, X. Wang, M. P. Hautzinger, R. Tirawat, M. C. Beard, D. P. Fenning, J. J. Berry, B. W. Larson, Y. Yan, and K. Zhu, Nature 611, 278 (2022). https://doi.org/10.1038/s41586-022-05268-x4. A. Kojima, K. Teshima, Y. Shirai, and T. Miyasaka, J. Am. Chem. Soc. 131, 6050 (2009). https://doi.org/10.1021/ja809598r All these brilliant achievements could be attributed to their excellent optoelectrical properties, such as long free carrier lifetime, superior charge carrier diffusion length, and low trap state density.5–75. D. Shi, V. Adinolfi, R. Comin, M. Yuan, E. Alarousu, A. Buin, Y. Chen, S. Hoogland, A. Rothenberger, K. Katsiev, Y. Losovyj, X. Zhang, P. A. Dowben, O. F. Mohammed, E. H. Sargent, and O. M. Bakr, Science 347, 519 (2015). https://doi.org/10.1126/science.aaa27256. S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza, and H. J. Snaith, Science 342, 341 (2013). https://doi.org/10.1126/science.12439827. Z. Guo, Y. Wan, M. Yang, J. Snaider, K. Zhu, and L. Huang, Science 356, 59 (2017). https://doi.org/10.1126/science.aam7744 Beyond the impressive evolution, there always remained puzzling issues, the so-called hysteresis effect,88. H. J. Snaith, A. Abate, J. M. Ball, G. E. Eperon, T. Leijtens, N. K. Noel, S. D. Stranks, J. T. Wang, K. Wojciechowski, and W. Zhang, J. Phys. Chem. Lett. 5, 1511 (2014). https://doi.org/10.1021/jz500113x and its feasibly entangled origin ion migration.9–119. D. H. Kang and N. G. Park, Adv. Mater. 31, e1805214 (2019). https://doi.org/10.1002/adma.20180521410. J. Liu, M. Y. Hu, Z. H. Dai, W. X. Que, N. P. Padture, and Y. Y. Zhou, ACS Energy Lett. 6, 1003 (2021). https://doi.org/10.1021/acsenergylett.0c0266211. S. Ravishankar, S. Gharibzadeh, C. Roldán-Carmona, G. Grancini, Y. Lee, M. Ralaiarisoa, A. Asiri, N. Koch, J. Bisquert, and M. Nazeeruddin, Joule 2, 788 (2018). https://doi.org/10.1016/j.joule.2018.02.013 The hysteresis effect described the divergence of the photocurrent and photovoltage dependence when conducting the photoelectronic performance evaluation in different bias voltage-sweeping directions. Therefore, precise estimation of the actual device performance becomes a complicated task, since the photoelectronic performance is strongly dependent on test conditions including pre-condition, such as electronic or light bias,12,1312. F. Wu, B. Bahrami, K. Chen, S. Mabrouk, R. Pathak, Y. Tong, X. Li, T. Zhang, R. Jian, and Q. Qiao, ACS Appl. Mater. Interfaces 10, 25604 (2018). https://doi.org/10.1021/acsami.8b0729813. G. A. Nemnes, C. Besleaga, V. Stancu, D. E. Dogaru, L. N. Leonat, L. Pintilie, K. Torfason, M. Ilkov, A. Manolescu, and I. Pintilie, J. Phys. Chem. C 121, 11207 (2017). https://doi.org/10.1021/acs.jpcc.7b04248 and test parameters, for instance, bias-voltage sweeping rate (SR) or range.1414. N. E. Courtier, J. M. Cave, J. M. Foster, A. B. Walker, and G. Richardson, Energy Environ. Sci. 12, 396 (2019). https://doi.org/10.1039/C8EE01576GThe ion migration is regarded as the most possible origination of the hysteresis effect, not only because of the coincidence that the ionic diffusions were also strongly correlated with the external electrical bias and light irradiation15–1715. I. Zarazua, G. Han, P. P. Boix, S. Mhaisalkar, F. Fabregat-Santiago, I. Mora-Sero, J. Bisquert, and G. Garcia-Belmonte, J. Phys. Chem. Lett. 7, 5105 (2016). https://doi.org/10.1021/acs.jpclett.6b0219316. G. Xia, B. Huang, Y. Zhang, X. Zhao, C. Wang, C. Jia, J. Zhao, W. Chen, and J. Li, Adv. Mater. 31, e1902870 (2019). https://doi.org/10.1002/adma.20190287017. X. M. Lian, L. J. Zuo, B. W. Chen, B. A. Li, H. T. Wu, S. Q. Shan, G. Wu, X. G. Yu, Q. Chen, L. W. Chen, D. R. Yang, D. Cahen, and H. Z. Chen, Energy Environ. Sci. 15, 2499 (2022). https://doi.org/10.1039/D2EE01097F but also because several direct observations that build up their causality.18,1918. S.-G. Kim, C. Li, A. Guerrero, J.-M. Yang, Y. Zhong, J. Bisquert, S. Huettner, and N.-G. Park, J. Mater. Chem. A 7, 18807 (2019). https://doi.org/10.1039/C9TA07595J19. H. S. Kim, I. H. Jang, N. Ahn, M. Choi, A. Guerrero, J. Bisquert, and N. G. Park, J. Phys. Chem. Lett. 6, 4633 (2015). https://doi.org/10.1021/acs.jpclett.5b02273 These results could be adopted to understand the efficacy that plenty of efforts succeed in hysteresis inhibition via ionic migration passivation. For instance, by interfacial engineering with the help of some additives, such as fullerene derivatives20,2120. J. X. Li, X. C. Meng, Z. Q. Huang, R. Y. Dai, W. P. Sheng, C. X. Gong, L. C. Tan, and Y. W. Chen, Adv. Funct. Mater. 32, 2105917 (2022). https://doi.org/10.1002/adfm.20210591721. Y. Zhong, M. Hufnagel, M. Thelakkat, C. Li, and S. Huettner, Adv. Funct. Mater. 30, 1908920 (2020). https://doi.org/10.1002/adfm.201908920 and alkali cations,22–2422. M. Abdi-Jalebi, Z. Andaji-Garmaroudi, S. Cacovich, C. Stavrakas, B. Philippe, J. M. Richter, M. Alsari, E. P. Booker, E. M. Hutter, A. J. Pearson, S. Lilliu, T. J. Savenije, H. Rensmo, G. Divitini, C. Ducati, R. H. Friend, and S. D. Stranks, Nature 555, 497 (2018). https://doi.org/10.1038/nature2598923. C. Li, A. Wang, L. Xie, X. Deng, K. Liao, J.-A. Yang, T. Li, and F. Hao, J. Mater. Chem. A 7, 24150 (2019). https://doi.org/10.1039/C9TA08130E24. Y. Zhao, I. Yavuz, M. Wang, M. H. Weber, M. Xu, J. H. Lee, S. Tan, T. Huang, D. Meng, R. Wang, J. Xue, S. J. Lee, S. H. Bae, A. Zhang, S. G. Choi, Y. Yin, J. Liu, T. H. Han, Y. Shi, H. Ma, W. Yang, Q. Xing, Y. Zhou, P. Shi, S. Wang, E. Zhang, J. Bian, X. Pan, N. G. Park, J. W. Lee, and Y. Yang, Nat. Mater. 21, 1396 (2022). https://doi.org/10.1038/s41563-022-01390-3 the hysteresis effect was reduced or eliminated accompanied by an enhancement in photoelectronic performance. Though great achievements had been reached, an underlying scientific problem has not been satisfactorily elucidated. What or which ions diffusion behavior was responsible for the hysteresis effect? The answer would undoubtedly disclose the key to conquering the hysteresis effect.Accordingly, various techniques have been developed to characterize the ion migration properties trying to find any hints correlated with the hysteresis effects, for instance, the electrochemical impedance spectra helped to assign low frequency capacitance components to migrated ion related.1919. H. S. Kim, I. H. Jang, N. Ahn, M. Choi, A. Guerrero, J. Bisquert, and N. G. Park, J. Phys. Chem. Lett. 6, 4633 (2015). https://doi.org/10.1021/acs.jpclett.5b02273 Microscopic fluorescence imaging2525. C. Li, A. Guerrero, S. Huettner, and J. Bisquert, Nat. Commun. 9, 5113 (2018). https://doi.org/10.1038/s41467-018-07571-6 and time-flight secondary ion mass spectrometry2626. Y. Liu, A. V. Ievlev, N. Borodinov, M. Lorenz, K. Xiao, M. Ahmadi, B. Hu, S. V. Kalinin, and O. S. Ovchinnikova, Adv. Funct. Mater. 31, 2008777 (2021). https://doi.org/10.1002/adfm.202008777 directly observed the ion migration under electric field. In addition, the activation energies for ionic migration were also estimated via density functional theory (DFT) calculation, suggesting that the most movable species are iodide (I−) and methylamine ions (MA+).2727. C. Eames, J. M. Frost, P. R. Barnes, B. C. O'Regan, A. Walsh, and M. S. Islam, Nat. Commun. 6, 7497 (2015). https://doi.org/10.1038/ncomms8497 Despite extensive investigations being carried out, the direct correlations between migrated ions and the hysteresis effect were hardly established, which was subject to the complexity of the many-body interactions between the mobile ions and the charge carriers, as any of the individual ones affect each other. The crucial breakthrough would be localized on the distinguishment of the two dynamic components and the observations of their behaviors.In our previous work,2828. S. Yuan, H. Y. Wang, F. G. Lou, X. L. Wang, Y. Wang, Y. J. Qin, X. C. Ai, and J. P. Zhang, J. Phys. Chem. C 126, 3696 (2022). https://doi.org/10.1021/acs.jpcc.1c10049 a circuit-switched transient photoelectric technique (cs-TPT) was established and acted as a powerful tool to monitor the behaviors of the migrated ions. Herein, the interfacial ion accumulation process resulting from ion migration was decoupled from the slow photovoltage establishment stage. Furthermore, different contribution fractions that dominate the photovoltage from various migrated ions and charge carriers were exposed. Most importantly, the strong dependence of the hysteresis effect on the migration of MA+ within the perovskite layer and the accumulation of MA+ at the interface was revealed by taking the electric coupling between MA+ and charge carriers into account. Inspired by this finding, significant hysteresis factor reduction was achieved through the inhibition of cation migration via a fullerene derivative modification. Our work proved that cation management might be a targeted strategy to control the hysteresis effect in perovskite solar cells.As above discussed, the hysteresis effect was closely related to ionic migration and the corresponding interfacial ion accumulations. Two main points of view described the ionic dynamics based on an entire solar device. One believed that the movable ions stayed as a neutralized condition in bulk when fabricated, the ionic migration and successive interfacial ion accumulation would be triggered under illumination, in which the ion induced electrostatic potential would constitute the overall photovoltage with the build-in potential.2929. R. Gottesman, P. Lopez-Varo, L. Gouda, J. A. Jimenez-Tejada, J. G. Hu, S. Tirosh, A. Zaban, and J. Bisquert, Chem 1, 776 (2016). https://doi.org/10.1016/j.chempr.2016.10.002 However, the other holds the point that the ions would be migrated to the interface to compensate for the built-in potential when fabricated, while the weakened built-in potential under illumination would facilitate the backfilling process of the accumulated ions.1010. J. Liu, M. Y. Hu, Z. H. Dai, W. X. Que, N. P. Padture, and Y. Y. Zhou, ACS Energy Lett. 6, 1003 (2021). https://doi.org/10.1021/acsenergylett.0c02662 These antithesis debates call for an answer to how the migrated ions were accumulated at the interface and their potential influence on the charge carrier behaviors. Therefore, in this work, the ionic behaviors were observed in detail via a home-built transient optoelectronic technique. Figure 1(a) illustrated the schematic apparatus of cs-TPT. All the resulting voltage responses will be represented via a sampling resistance and collected via a parallel-connected oscilloscope. The analog switch and the time-delay generator (DG535) are the two key components in the setup. The analog switch was adopted to switch the sampling resistance between 50 Ω and 1 MΩ to modulate a short-circuit and an open-circuit condition in different test processes. While the DG535 is the commander to independently regulate the dwell time of the laser illumination and the switch timing of the analog switch, which also defines the time zero and decides the repeat frequency of the experiment.The detailed time sequence of the experiment was provided in Fig. 1(b), in which three different stages could be readily observed. First, the so-called open circuit voltage build-up process (OCVB), during which the device is illuminated by a continuous laser. If no further operation is conducted and in ideal condition, the accumulated photogenerated charge carriers would result in the splitting of the quasi-Fermi level of the corresponding interface and finally leads to the increase in the photovoltage (Vph). Eventually, the Vph would reach a steady state if the dwell time of illumination is long enough, which is presented as the dashed line in green. In addition, the dwell time is defined as tbuild for further convenient discussion. Second, the so-called time-resolved charge extraction (TRCE) process is used to exhaust the photogenerated charge carriers.3030. N. W. Duffy, L. M. Peter, R. M. G. Rajapakse, and K. G. U. Wijayantha, Electrochem. Commun. 2, 658 (2000). https://doi.org/10.1016/S1388-2481(00)00097-7 When the device has already been illuminated for a designed tbuild, the analog switch is swiftly switched to short circuit accompanied by the shutdown of the laser illumination. Theoretically, no free photogenerated charge carriers would survive after TRCE. The last and the most critical process associated with the migrated ions is the voltage recovery process. It is achieved by returning the analog switch to open circuit condition after TRCE, and a considerable voltage recovery behavior could be observed, which is attributed to a polarized induced trap state (PITS) according to our previous work.2828. S. Yuan, H. Y. Wang, F. G. Lou, X. L. Wang, Y. Wang, Y. J. Qin, X. C. Ai, and J. P. Zhang, J. Phys. Chem. C 126, 3696 (2022). https://doi.org/10.1021/acs.jpcc.1c10049 As shown in Fig. S1, the PITS describes the mechanism that the voltage recovery originates from the de-trapping and interfacial transfer process of the trapped charge carriers of interfacial accumulated ions. Therefore, the time constant of the rising part, define as τr, could be borrowed to evaluate the strength of interaction between the accumulated ions and the charge carriers. Meanwhile, the intensity of the voltage recovery could be used as a metric to estimate the intensity of ion accumulation and is defined as Vrmax.A target perovskite solar cell with a classic architecture as illustrated in Fig. S2(a) (FTO/SnO2/MAPbI3/Spiro-OMeTAD/Au) was named as “control device” for convenience. The current density–voltage (J–V) curve was displayed in Fig. S2(b), and a substantial hysteresis effect could be easily observed. Figure 2(a) provided the OCVB measurements of the control device, it could be found that the photovoltage took almost 10 s to reach the maximum (the open-circuit voltage, VOC). Three rise stages could be roughly distinguished: a steep increase in the Vph to ∼0.4 V within ∼20 μs, followed by a slow process from 0.4 to 0.9 V in a hundred-millisecond timescale, and finally, another slower increase lasts a few seconds to gradually reach the steady VOC. The fastest process was the contribution from the interfacial accumulation of the photogenerated free carriers, which could find proof according to a similar rise time constant of about ∼15 μs from the same measurement on a commercial silicon-based device [Fig. 2(b)], as no interference from ions could be involved and only free carriers could contribute to the photovoltage.3131. T. Y. Yang, G. Gregori, N. Pellet, M. Gratzel, and J. Maier, Angew. Chem., Int. Ed. 54, 7905 (2015). https://doi.org/10.1002/anie.201500014 Meanwhile, the abnormal slow processes were generally associated with the ion migration1010. J. Liu, M. Y. Hu, Z. H. Dai, W. X. Que, N. P. Padture, and Y. Y. Zhou, ACS Energy Lett. 6, 1003 (2021). https://doi.org/10.1021/acsenergylett.0c02662 and could be attributed to corresponding mobile species according to literature.3232. F. G. Lou, S. Yuan, X. L. Wang, H. Y. Wang, Y. Wang, Y. J. Qin, X. C. Ai, and J. P. Zhang, Chem. Phys. Lett. 796, 139570 (2022). https://doi.org/10.1016/j.cplett.2022.139570To uncover the accumulation behaviors of the migrated ions, the cs-TPT measurement was conducted during OCVB. Figure 2(c) provides the entire cs-TPT process accompanied by the OCVB. A voltage recovery process with a duration of about 10−3 s was observed, followed by a slow decay in the time magnitude of 101 s, and the detailed profiles could be achieved via varying tbuild. As shown in Fig. 2(d), two main features could be perceived. First, the Vrmax becomes enhanced and conspicuous as tbuild is prolonged, implying an expanded ion accumulation triggered by illumination.1515. I. Zarazua, G. Han, P. P. Boix, S. Mhaisalkar, F. Fabregat-Santiago, I. Mora-Sero, J. Bisquert, and G. Garcia-Belmonte, J. Phys. Chem. Lett. 7, 5105 (2016). https://doi.org/10.1021/acs.jpclett.6b02193 Second, the voltage recovery rise process becomes slower as tbuild increases, and a more obvious support could be found as a normalized view in Fig. 2(e), suggesting a stronger bonding interaction between the accumulated ions and the charge carriers. Since the interfacial charge transfer process would complete within 10−9–10−6 s;3333. J. J. Shi, Y. M. Li, Y. S. Li, D. M. Li, Y. H. Luo, H. J. Wu, and Q. B. Meng, Joule 2, 879 (2018). https://doi.org/10.1016/j.joule.2018.04.010 therefore, the rate determining step relies on the de-trapping process which represents the interaction strength between these two components (as illustrated in Fig. S1).More details could be found in Fig. 2(f), where the dependences of Vrmax and τr against Vph were exhibited, respectively. A monotonic increasing Vrmax could be observed as the Vph increases, indicating that the interfacial accumulated ions become notable and influential in photovoltage build-up process, especially at high voltage. The charged accumulated ions would directly participate in or at least contribute to the photovoltage, which has already been aware by other colleagues.2929. R. Gottesman, P. Lopez-Varo, L. Gouda, J. A. Jimenez-Tejada, J. G. Hu, S. Tirosh, A. Zaban, and J. Bisquert, Chem 1, 776 (2016). https://doi.org/10.1016/j.chempr.2016.10.002 Furthermore, three different parts could be roughly recognized: at low Vph (Vrmax was particularly small. Then, at a median Vph range (0.6–0.9 V), Vrmax boosted to a plateau region and followed by another sharp soar appeared at a high Vph region (>0.9 V). According to the previous discussion about OCVB, the last two processes could be attributed to migrated ion related. In addition, the first Vrmax increasing region (Vph: 0.6–0.9 V) would be assigned to an I− dominant domain owing to its lower activation energy for ion migration.2727. C. Eames, J. M. Frost, P. R. Barnes, B. C. O'Regan, A. Walsh, and M. S. Islam, Nat. Commun. 6, 7497 (2015). https://doi.org/10.1038/ncomms8497 Since plenty of works have proven that I− is much easier to migrate than other ions. In addition, our previous work has demonstrated that the migration timescale of I− and MA+ would be about 10−3 and 101 s, respectively.3232. F. G. Lou, S. Yuan, X. L. Wang, H. Y. Wang, Y. Wang, Y. J. Qin, X. C. Ai, and J. P. Zhang, Chem. Phys. Lett. 796, 139570 (2022). https://doi.org/10.1016/j.cplett.2022.139570 Therefore, it is reasonable to believe that I− would accumulate at the interface first and be preemptive to influence the photovoltage. Consequently, the last Vrmax increasing region (Vph: > 0.9 V) would belong to MA+ dominant domain. According to the assignment, the τr–Vph dependence would be much easier to be understood. As the time constant of the rising process could be adopted to describe the bonding strength; thus, a relatively smaller τr at the Vph region about 0.6–0.9 V indicates a weaker interaction strength while the slower de-trapping dynamics at high Vph region (>0.9 V) shows a strong interaction strength dominated by MA+. This could be attributed to a deeper trap state that might be formed by MA+ rather than I−.3434. A. Buin, P. Pietsch, J. Xu, O. Voznyy, A. H. Ip, R. Comin, and E. H. Sargent, Nano Lett. 14, 6281 (2014). https://doi.org/10.1021/nl502612m By inhibiting MA+ migration, the trap state can be reduced [see Fig. S4, the space-charge-limited current (SCLC) result]. Hereafter, we could elucidate the different dynamic behaviors of different migrated species and their potential contributions to photovoltaic performance.More importantly, the ion migration was designed to be suppressed by a fullerene derivative (C60 pyrrolidine tris-acid, CPTA) aiming at elucidating the potential correlation with the hysteresis effect. Fullerene derivative was adopted since it has widely been reported to possess the ability to suppress ion migration by forming an interfacial barrier layer by physical isolation3535. Y. C. Shih, L. Wang, H. C. Hsieh, and K. F. Lin, ACS Appl. Mater. Interfaces 10, 11722 (2018). https://doi.org/10.1021/acsami.8b03116 or electrostatic repulsion. Meanwhile, fullerene derivative can penetrate into the bulk of the perovskite layer and then passivate the trap state at grain boundary2121. Y. Zhong, M. Hufnagel, M. Thelakkat, C. Li, and S. Huettner, Adv. Funct. Mater. 30, 1908920 (2020). https://doi.org/10.1002/adfm.201908920 which acts as the pathway of ion migration.3636. Y. Shao, Y. Fang, T. Li, Q. Wang, Q. Dong, Y. Deng, Y. Yuan, H. Wei, M. Wang, A. Gruverman, J. Shield, and J. Huang, Energy Environ. Sci. 9, 1752 (2016). https://doi.org/10.1039/C6EE00413J Figure S3 provided the device architecture and the chemical structure of CPTA, the J–V characterization exhibited significant inhibition of the hysteresis effect. Meanwhile, the introduction of CPTA also brought other positive effects, including passivating trap state (Fig. S4), improving photoluminescence lifetime and intensity (Fig. S5 and Table S1). So, the CPTA device provides an ideal sample for further discussion. Different from the control device, the OCVB curve of the CPTA device jumped to almost 0.6 V in the initial few microseconds and reached the voltage plateau in about several hundred milliseconds without the slow rising part associated with MA+ ions, as provided in Fig. 3(a). A similar accelerated voltage build-up tendency was also observed with the help of another fullerene derivative.3535. Y. C. Shih, L. Wang, H. C. Hsieh, and K. F. Lin, ACS Appl. Mater. Interfaces 10, 11722 (2018). https://doi.org/10.1021/acsami.8b03116 The Vrmax–Vph curve shows a more intuitive view [Fig. 3(b)], the MA+ dominant domain disappears at high voltage region, implying inhibited migration and interfacial accumulation of the MA+ ion. In Fig. 3(c), it could be found that the two devices exhibited almost the same voltage recovery dynamics when Vph+, which suggests a stronger interaction between the electrons and MA+ rather than I−. According to the assignments and analysis, the quantitative contributions of different ion species to photovoltage were evaluated and summarized as shown in Figs. S6 and 3(d). Apparently, in CPTA devices, the contribution from MA+ almost vanished owing to the inhibition of the MA+ migration; consequently, the proportion of free charge carriers rises to 64.5%.The hysteresis effect was manifested via a sweep rate dependent measurement to unveil the potential inner connection with the ion migration properties. As shown in Fig. 3(e), the inset showed the step sweeping model for J–V measurement and defined the sweeping rate (SR) as dV/dt. Meanwhile, the hysteresis factor (HF) was calculated by the following formula: HF = (Areverse − Aforward)/Areverse, in which A represents the integral area of the corresponding J–V curve with an integral range from 0 to VOC, and the subscript indicates different sweep directions. Thus, a higher HF strands for a serious hysteresis effect. An ion–electron coupling model was adopted to understand the SR dependent hysteresis effects, in which two aspects should be considered: one is the sweep direction, and the other is the sweep rate. The hysteresis effect originated from different electron bonding strengths corresponding to different ion distributions. Accordingly, during a reverse sweep process, a significant ion accumulation has already been established at a high applied voltage and will decrease during the entire sweep process, and vice versa. This means that the ion accumulation will be established gradually when performing forward sweep measurements. The ion distribution could hardly be the same though the identical voltage was applied during each sweep process. Thus, the photocurrent varies in different sweep directions, especially at the maximum power point, leading to the hysteresis effect. It is readily understood that the mismatch of the ion distributions would be reduced if the ion migration could be inhibited, so was shown in Figs. 3(e) and 3(f) that the CPTA-modified device exhibited a smaller HF than the control device, this could be credit to the inhibition of the MA+ as achieved by CPTA modification.On the other hand, ion migration is a dynamic process; therefore, the sweep rate also plays an important role. Significant hysteresis effects would be observed only when the SR and ion mobility are comparable, so the HF–SR curve will display a Gaussian-like distribution,1414. N. E. Courtier, J. M. Cave, J. M. Foster, A. B. Walker, and G. Richardson, Energy Environ. Sci. 12, 396 (2019). https://doi.org/10.1039/C8EE01576G i.e., the ions got sufficient time to migrate to and accumulate at the interface if the SR is too low, or the ions hardly respond to voltage changes if the SR is too high. Consequently, the hysteresis effect would be regardless of sweep direction unless the SR resonant with the ion mobility. The control device showed a half-Gaussian-like HF-SR tendency in Fig. 3(e), indicating a serious hysteresis effect compared to the CPTA modified device, especially in a SR region of about 0.3–2 V/s [Fig. 3(f)]. This appearance was assigned to the MA+ domain region as has been confirmed in our previous work.3232. F. G. Lou, S. Yuan, X. L. Wang, H. Y. Wang, Y. Wang, Y. J. Qin, X. C. Ai, and J. P. Zhang, Chem. Phys. Lett. 796, 139570 (2022). https://doi.org/10.1016/j.cplett.2022.139570 Moreover, this SR highly overlaps with the range of the common J–V test (such as 1 V/s). The extensively reduced HF of CPTA modified device in this region suggests that the cations would be the critical ionic species that determine the hysteresis effect.

In summary, ion migration and interfacial ion accumulation processes as well as their impacts on the photovoltage of perovskite solar devices were characterized by the home-built cs-TPT measurements. For the prototypical planar perovskite, I−-coupled and MA+-coupled charge carriers contribute to 36.2% and 12% of the overall photovoltage output, respectively. A fullerene derivative CPTA was adopted to effectively inhibit the migration of MA+; as a result, the MA+ related photovoltage component was thoroughly eliminated. Comparing the migration time constants of MA+ and I− with the sweep rate of the J–V characterization, it was established that the hysteresis effect is predominantly determined by MA+ instead of I− migration which strongly interacts with charge carriers following interfacial MA+ accumulation. This interpretation helps understanding why the CPTA treatment dramatically suppressed the hysteresis effect. Our work provides the answer to the specific ions that contribute to the photovoltage and paves the way to conquer the hysteresis effect.

See the supplementary material for experimental section; PITS model; device structure and characterization; SCLC curve; and TRPL and PL characterization.

This work has been supported by the National Key R&D Program of China (Grant No. 2018YFA0208701), the National Natural Science Foundation of China (Grant Nos. 22203103, 22103096, and 21973112), and the Fundamental Research Funds for the Central Universities and the Research Funds of Renmin University of China (Grant Nos. 22XNKJ08 and 22XNKJ17).

Conflict of Interest

The authors have no conflicts to disclose.

Author Contributions

Shuai Yuan and Yiyi Li performed cs-TPT measurements; Feige Lou fabricated the PSC devices; Jian-Ping Zhang and Xi-Cheng Ai conducted funding acquisition; Shuai Yuan, Hao-Yi Wang, Yi Wang, and Xi-Cheng Ai contributed the data analysis and result interpretation; Shuai Yuan and Hao-Yi Wang drafted the manuscript; Yi Wang provided revisions to the manuscript; and all authors discussed the results and commented on the manuscript.

Shuai Yuan: Data curation (equal); Investigation (equal); Methodology (equal); Writing – original draft (equal). Feige Lou: Methodology (equal). Yiyi Li: Methodology (equal). Hao-Yi Wang: Supervision (equal); Writing – review & editing (lead). Yi Wang: Writing – review & editing (supporting). Xi-Cheng Ai: Funding acquisition (equal); Supervision (equal). Jian-Ping Zhang: Funding acquisition (equal); Supervision (equal).

The data that support the findings of this study are available from the corresponding authors upon reasonable request.

REFERENCES

1. J. Jeong, M. Kim, J. Seo, H. Lu, P. Ahlawat, A. Mishra, Y. Yang, M. A. Hope, F. T. Eickemeyer, M. Kim, Y. J. Yoon, I. W. Choi, B. P. Darwich, S. J. Choi, Y. Jo, J. H. Lee, B. Walker, S. M. Zakeeruddin, L. Emsley, U. Rothlisberger, A. Hagfeldt, D. S. Kim, M. Gratzel, and J. Y. Kim, Nature 592, 381 (2021). https://doi.org/10.1038/s41586-021-03406-5, Google ScholarCrossref, ISI2. M. Kim, J. Jeong, H. Lu, T. K. Lee, F. T. Eickemeyer, Y. Liu, I. W. Choi, S. J. Choi, Y. Jo, H. B. Kim, S. I. Mo, Y. K. Kim, H. Lee, N. G. An, S. Cho, W. R. Tress, S. M. Zakeeruddin, A. Hagfeldt, J. Y. Kim, M. Gratzel, and D. S. Kim, Science 375, 302 (2022). https://doi.org/10.1126/science.abh1885, Google ScholarCrossref3. Q. Jiang, J. Tong, Y. Xian, R. A. Kerner, S. P. Dunfield, C. Xiao, R. A. Scheidt, D. Kuciauskas, X. Wang, M. P. Hautzinger, R. Tirawat, M. C. Beard, D. P. Fenning, J. J. Berry, B. W. Larson, Y. Yan, and K. Zhu, Nature 611, 278 (2022). https://doi.org/10.1038/s41586-022-05268-x, Google ScholarCrossref4. A. Kojima, K. Teshima, Y. Shirai, and T. Miyasaka, J. Am. Chem. Soc. 131, 6050 (2009). https://doi.org/10.1021/ja809598r, Google ScholarCrossref, ISI5. D. Shi, V. Adinolfi, R. Comin, M. Yuan, E. Alarousu, A. Buin, Y. Chen, S. Hoogland, A. Rothenberger, K. Katsiev, Y. Losovyj, X. Zhang, P. A. Dowben, O. F. Mohammed, E. H. Sargent, and O. M. Bakr, Science 347, 519 (2015). https://doi.org/10.1126/science.aaa2725, Google ScholarCrossref, ISI6. S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. Alcocer, T. Leijtens, L. M. Herz, A. Petrozza, and H. J. Snaith, Science 342, 341 (2013). https://doi.org/10.1126/science.1243982, Google ScholarCrossref, ISI7. Z. Guo, Y. Wan, M. Yang, J. Snaider, K. Zhu, and L. Huang, Science 356, 59 (2017). https://doi.org/10.1126/science.aam7744, Google ScholarCrossref, ISI8. H. J. Snaith, A. Abate, J. M. Ball, G. E. Eperon, T. Leijtens, N. K. Noel, S. D. Stranks, J. T. Wang, K. Wojciechowski, and W. Zhang, J. Phys. Chem. Lett. 5, 1511 (2014). https://doi.org/10.1021/jz500113x, Google ScholarCrossref, ISI9. D. H. Kang and N. G. Park, Adv. Mater. 31, e1805214 (2019). https://doi.org/10.1002/adma.201805214, Google ScholarCrossref10. J. Liu, M. Y. Hu, Z. H. Dai, W. X. Que, N. P. Padture, and Y. Y. Zhou, ACS Energy Lett. 6, 1003 (2021). https://doi.org/10.1021/acsenergylett.0c02662, Google ScholarCrossref11. S. Ravishankar, S. Gharibzadeh, C. Roldán-Carmona, G. Grancini, Y. Lee, M. Ralaiarisoa, A. Asiri, N. Koch, J. Bisquert, and M. Nazeeruddin, Joule 2, 788 (2018). https://doi.org/10.1016/j.joule.2018.02.013, Google ScholarCrossref12. F. Wu, B. Bahrami, K. Chen, S. Mabrouk, R. Pathak, Y. Tong, X. Li, T. Zhang, R. Jian, and Q. Qiao, ACS Appl. Mater. Interfaces 10, 25604 (2018). https://doi.org/10.1021/acsami.8b07298, Google ScholarCrossref13. G. A. Nemnes, C. Besleaga, V. Stancu, D. E. Dogaru, L. N. Leonat, L. Pintilie, K. Torfason, M. Ilkov, A. Manolescu, and I. Pintilie, J. Phys. Chem. C 121, 11207 (2017). https://doi.org/10.1021/acs.jpcc.7b04248, Google ScholarCrossref14. N. E. Courtier, J. M. Cave, J. M. Foster, A. B. Walker, and G. Richardson, Energy Environ. Sci. 12, 396 (2019). https://doi.org/10.1039/C8EE01576G, Google ScholarCrossref15. I. Zarazua, G. Han, P. P. Boix, S. Mhaisalkar, F. Fabregat-Santiago, I. Mora-Sero, J. Bisquert, and G. Garcia-Belmonte, J. Phys. Chem. Lett. 7, 5105 (2016). https://doi.org/10.1021/acs.jpclett.6b02193,