Perovskite solar cells based on organic-inorganic hybrid perovskite materials have become a research hotspot in photovoltaics field due to their outstanding power conversion efficiency (PCE)[1].Nonetheless, the organic cations are volatile and have rotation freedom, which is not good for photoand thermal-stability of the devices.Fortunately, these issues can be solved by all-inorganic perovskites, which consist of Cs, Pb and I (or Br)[2, 3].Moreover, the all-inorganic perovskites, such as CsPbl3, are excellent candidates as top-cell absorbers in tandem solar cells because of their suitable bandgaps and high stability.All-inorganic perovskites were first used as light absorbers in solar cells in 2015[4, 5].All-inorganic perovskite solar cells experienced rapid development in last few years, and the PCE reaches 20.4％ at the end of 2020[6].Most recently, Meng et al.pushed the PCE to ＞21.0％ (unpublished).Despite these advances, we should recognize that there still remains a big gap between the PCEs for all-inorganic perovskite solar cells and hybrid perovskite solar cells (Fig.1(a))[7, 8].The PCE for hybrid cells has reached 80％ of the theoretical limit, while the PCE for all-inorganic cells is still below 70％ of the theoretical limit[9].A detailed analysis on performance parameters of these cells suggests that the PCE for all-inorganic cells is mainly limited by the open-circuit voltage (Voc) and fill factor (FF) (Fig.1(b)).In solar cells, these two parameters correlate to non-radiative charge loss caused by defects[10].Therefore, defect engineering is the most critical approach for achieving higher PCE for all-inorganic perovskite solar cells.