In the realm of renewable energy, perovskite solar cells have garnered significant attention from researchers and industry for their exceptional performance, with commercialization appearing imminent. The state-of-the-art polycrystalline perovskite solar cells boast a power conversion efficiency (PCE) of over 26%, comparable to that of single crystalline Si solar cells, while also being cost-effective.
However, despite their promising features, polycrystalline perovskites are not without their flaws. They suffer from defects that cause loss of charge carriers and degrade structural stability, ultimately reducing photovoltaic efficiency and device stability. To tackle this issue, the research community has been actively exploring various defect passivation strategies. Notably, deep-level defects are considered detrimental due to their role as primary non-radiative recombination centers. On the other hand, shallow-level defects, such as the dominant iodide vacancy (VI+) in FAPbI3, have typically been seen as less problematic, thought to allow efficient radiative recombination of charge carriers1. However, the shallow-level defect can form polarons that activate non-radiative processes, thereby degrading the photovoltaic performance of perovskite solar cells (PSCs)2.
 While recent studies have started to pay more attention to the issues posed by shallow-level defects, their primary focus has been on their migration rather than their impact on carrier dynamics3–5. Consequently, a comprehensive understanding of how shallow-level defects affect charge carrier recombination dynamics in perovskites remains elusive.
 In our work, we propose a reappraisal of the perceived benign nature of shallow-level defects in perovskites. Based on theoretical and experimental evidence, we present a complex non-radiative recombination mechanism initiated by the formation of the dominant shallow-level defect VI+, which eventually evolves to deep-level defect VI0 (Figure 1). This insight underscores the necessity of addressing shallow-level defects related to non-radiative energy transfer in perovskites, which have not received as much attention as deep-level defects in this community.
Figure 1. a, The evolution of shallow-level defect VI+ to deep-level defect VI0. b, Schematic of carrier dynamics in the presence of VI+ that eventually facilitates non-radiative recombination.
Moreover, we demonstrate a new concept of defect passivation that employs perovskite polytypes. A typical approach so far has been to introduce an external chemical reagent. However, the external reagent (additive or dopant) can directly impact the crystalline quality of the perovskite during crystal growth, possibly degrading structural integrity.
 In contrast to prior approaches, our work does not rely on the external stabilizers. Instead, we employ a polytype of perovskite with a corner-sharing component (6H) that effectively suppresses the formation of the shallow-level defect VI+ in FAPbI3. Polytypes consist of identical constituents but differ in the atomic stacking sequences. Therefore, using desirable polytypes can achieve high crystallinity of polycrystalline FAPbI3 while maintaining material homogeneity and integrity due to high lattice coherency between the polytypes. Unlike the 2H polytype FAPbI3, which consists of a 100% face-sharing component (δ-phase), the 6H polytype includes a 66% corner-sharing component that can coherently integrate with the 3C-FAPbI3 consisting of a 100% corner-sharing component (Figure 2). This integration occurs by intervening the halide at a potential site for shallow-level defect VI+ formation. As a result, the construction of 3C/6H hetero-polytypic perovskites allows for effective defect screening and suppression of non-radiative recombination while maintaining material homogeneity and integrity.
Figure 2. a, Different polytypes of 3C, 2H, and 6H. b, the construction of 3C/6H hetero-polytypic perovskites due to high lattice coherency of corner-sharing component.
The employment of the hetero-polytypic perovskite achieved a PCE of 24.13% for unit cells and significantly improved film and device stability with extra surface passivation. Moreover, we demonstrated a perovskite solar module with a PCE of 22.57 % (active area: 27.8 cm2), and the module was certified at Newport Photovoltaic Testing and Calibration Laboratory in the US showing a PCE of 21.44%. Furthermore, we demonstrated a low-temperature processed module with a PCE of 22.96% (active area: 23.2 cm2). The module efficiency is on the highest level reported for an area > 20cm2.
Figure 3. a, b, J-V curves of perovskite solar module (a) and low-temperature processed perovskite solar module (b) using the hetero-polytypic perovskite with surface passivation.
 The promising results, alongside our understanding of the importance of shallow-level defect engineering may pave the way for further advancements not only in perovskite solar cells but also various perovskite optoelectronic devices in terms of efficiency and stability. We believe that this work will spark considerable interest in this community of perovskite optoelectronics, encouraging discussions about the impact of the perovskite polytypes and the shallow-level defect on optoelectronic properties and applications.
Â
Kovalenko, M. V. et al. Properties and potential optoelectronic applications of lead halide perovskite nanocrystals. Science 358, 745–750 (2017).
Mahata, A. et al. From Large to Small Polarons in Lead, Tin, and Mixed Lead–Tin Halide Perovskites. J. Phys. Chem. Lett. 10, 1790–1798 (2019).
Chen, B. et al. Imperfections and their passivation in halide perovskite solar cells. Chem. Soc. Rev. 48, 3842–3867 (2019).
Tan, S. et al. Shallow Iodine Defects Accelerate the Degradation of α-Phase Formamidinium Perovskite. Joule 4, 2426–2442 (2020).
Wang, T. et al. Deep defect passivation and shallow vacancy repair via an ionic silicone polymer toward highly stable inverted perovskite solar cells. Energy Environ. Sci. 15, 4414–4424 (2022).