Perovskite solar cells (PSCs) have attracted much attention due to their low cost, high efficiency and easy processing. Recently, You et al. achieved a record efficiency of 26.1% (certified 25.6%) in PSCs with excellent stability^[1]. Excess PbI₂ in perovskite films was converted into inactive (PbI₂)₂RbCl to avoid its detrimental effect to device stability, while maintaining its positive effect to device efficiency.

However, scientists found that excess PbI₂ in perovskite films could cause severe stability issues of PSCs especially under illumination, despite increased efficiency^{[6, 7]}. PbI₂ tended to decompose into metallic Pb⁰ and I₂ under illumination, yielding Pb⁰ defects and thus decreasing device efficiency^[8]. I₂ product could further promote perovskite decomposition, even destroy charge-transporting layer^{[9, 10]} and corrode metal electrode^{[11, 12]}, accelerating device degradation. In addition, excess PbI₂ can absorb adjacent MA⁺ (or FA⁺) and I⁻ during perovskite aging, yielding MA⁺ and I⁻ vacancies^[13]. Theses vacancies can act as channels for ion migration in perovskite films, also leading to device degradation.

Most highly efficient PSCs were achieved with excess PbI₂ in either perovskite bulk or film surface. For example, Seo et al. fabricated PSCs with >25% efficiency by using perovskite precursor solution with excess PbI₂^[2]. Besides, You et al. achieved planar PSCs with efficiency close to 20% in 2016 through annealing perovskite films at high temperature to induce its partial decomposition and form excess PbI₂ spontaneously^[3]. The increased efficiency in PSCs with excess PbI₂ resulted from PbI₂ passivation. Owing to the larger bandgap of PbI₂, a favorable type Ⅰ band alignment was established between PbI₂ and perovskite, forming energy barriers to block both electron and hole transport to grain boundaries (GBs) and repel them into perovskite grains^{[4, 5]}. As a result, potential carrier recombination induced by defects at GBs was effectively inhibited, thus improving device efficiency. In addition, the excess PbI₂ at perovskite/charge-transport layer interface can block possible interfacial recombination, further enhancing device efficiency.

To reduce the negative effect of excess PbI₂, Zhu et al. reported a chemical polishing method to optimize perovskite films. They introduced excess PbI₂ during perovskite growth to obtain high-quality perovskite films. Then they washed away excess PbI₂ on perovskite surface through its reaction with ammonium salts in the polishing agents (Fig. 1(a))^[14], avoiding PbI₂-induced perovskite degradation. The PSCs exhibited over 24% efficiency with good light and air stability. In addition, Luo et al. reported a ligand-modulated method (Fig. 1(b)) to regulate the shape and distribution of excess PbI₂ in perovskite films via introducing cetyltrimethylammonium bromide (CTAB)^[15]. CTAB caused PbI₂ into 2D nanosheet, vertically sticking among perovskite grains. As a result, over 22% efficiency was obtained. Besides, Li et al. modulated excess PbI₂ with ionic liquid [BMIM]X (Fig. 1(c))^[16]. [BMIM]X could interact with PbI₂ and form supramolecular structures, which could relax lattice distortion and release residual tensile strain in perovskite films, thus improving device stability.

Recently, You et al. managed excess PbI₂ through RbCl doping in perovskite films^[1]. RbCl reacted with PbI₂ to form (PbI₂)₂RbCl (Fig. 2(a)). (PbI₂)₂RbCl promoted the formation and stabilization of black FAPbI₃ phase through Pb-Cl bonding. On the other hand, (PbI₂)₂RbCl was inactive and could not react with FA⁺ or I⁻ of perovskite grains, thus inhibiting the formation of FA⁺ or I⁻ vacancies and suppressing ion migration (Fig. 2(b)). Besides, (PbI₂)₂RbCl is an intercalated compound and is more favorable to block ion migration in perovskite. As a result, a record efficiency of 26.1% (certified 25.6%) was realized in PSCs with good stability (Fig. 2(c)). The devices retained 96% of initial efficiency after aging in N₂ for 1000 h and 80% of initial value after aging at 85°C for 500 h.

In short, future efforts on PSCs will focus on improving the module efficiency and stability for real commercialization.