An Overview of Lead, Tin, and Mixed Tin–Lead‐Based ABI3 Perovskite Solar Cells

Over a short period of approximately 10 years, metal‐halide‐perovskite‐based photovoltaics have demonstrated unprecedented improvements in solar cell performance beyond the various major photovoltaic semiconductor materials such as organic, cadmium telluride, and copper indium gallium selenide. With this, the current focus lies on the commercialization of perovskite solar cell technology and the issues encountered while ensuring and balancing high efficiency, stability, and eco‐friendliness in the photovoltaic community. This article reviews prominent developments in perovskite‐based photovoltaic power generation based on the ABI3 structure, describing the current state and understanding of state‐of‐the‐art solar cell drives. Accordingly, methods to improve the efficiency and long‐term operational stability, lead toxicity, nonlead perovskites, bandgap optimization, and tandem solar cells are discussed. Prospects and views on future research considering the feasibility of perovskite technology commercialization are provided.


Introduction
Since the invention of crystalline silicon solar cells in the 1950s, [1] research on modern photovoltaic devices has led to the development of numerous photovoltaic concepts based on various light-absorbing materials such as semiconducting inorganic compositions (copper indium gallium selenide [CIGS], CdTe), quantum dots, organic semiconductors, or dyes. [2] Nevertheless, photovoltaic cells based on these new concepts have encountered several hurdles in exceeding the efficiency (26.7%) previously demonstrated by silicon materials. However, approximately a decade ago, metal halide perovskites (MHPs) of APbI 3 (ABX 3 type) emerged as promising light-absorbing materials for light-harvesting applications. [3] MHPs demonstrate superior properties in photovoltaics [4] and optoelectronics. [5] Furthermore, they are commercially promising owing to their immense potential for scaled-up largearea vacuum-free fabrication. [6] Moreover, since the first report on perovskite solar cells (PeSCs) in 2009, they have undergone extensive improvements, leading to an increase in their power conversion efficiency (PCE) from an initial value of approximately 3% to 25.7%. [3a,7] This development took only a decade to achieve the benchmark efficiency of 25% compared to crystalline silicon solar cells. [1] Ideal perovskites with the ABX 3 formula have a simple cubic structure consisting of eight corner-sharing BX 6 4À octahedral sublattices forming a central void for A-site positioning, where A and B are cations, and X corresponds to a halide anion. [8] Generally, the A-site is occupied by an inorganic cation of Cs or organic molecules of CH 3 NH 3 þ (MA þ ) and NH 2 CHNH 2 þ (FA þ ). Moreover, the B-site is mostly occupied by metal cations of Pb 2þ or Sn 2þ from group IV-A, whereas X typically represents Cl À , Br À , or I À . MHPs are a class of ionic crystals (soft nature) and possess defect tolerance, facilitating easy synthesis and contributing to highly efficient photovoltaics. [9] However, they can also cause ion migration in the lattice, reducing operational reliability and limiting the commercial applications of PeSCs. [10] Accordingly, numerous methods have been implemented to overcome these issues with either A-or X-site modifications and additive controls, consequently improving the efficiency and complementing operational stability. [10b,11] After achieving over 25% efficient cell-scale PeSCs, the next phase of commercially motivated research-in addition to the large-area submodules scale-is mainly focused on the stability of PeSCs under continuous light soaking conditions. [10a] Typically, PeSCs must withstand light and thermal stresses during their operational lifetime to be commercialized; hence, light stability studies on PeSCs must be prioritized. However, such studies are rather limited, hindering the commercialization of PeSCs. By contrast, studies on the thermal behavior and structural stability of PeSCs under light, heat, or humidity are among the fastgrowing areas of research. [11a,12] MHP group-based solar cells include lead-based, lead-free, and mixed lead-based PeSCs. Lead-based PeSCs have demonstrated significant enhancements in their efficiency with time. However, an environmentally benign technological perspective indicates that Pb is toxic; therefore, it is fatal to the environment.
Consequently, numerous efforts have been dedicated to overcome this issue by introducing environmentally friendly and less toxic alternatives, such as Sn instead of Pb. Such replacements are good from an environmental perspective; however, they result in lower efficiencies owing to multiple issues: one being the oxidation state of Sn, which is challenging to address. Snbased PeSCs have been considered the most plausible candidates among lead-free PeSCs. However, despite substantial research, related progress has been gradual. The current efficiency of Sn-based PeSCs is approximately 14.81%, [13] and ideal replacement candidates for Pb in MHPs are yet to be discovered. Therefore, extensive research efforts are currently devoted to the search of the most suitable alternative to Pb. Until then, Pb-based perovskite should be incorporated in MHPs for commercial applications, provided that Pb can be recycled from to mitigate its environmental footprint; moreover, Pb has demonstrated unprecedented performance in MHPs compared to other mixed or Pb-free PeSCs. [14] This review discusses state-of-the-art PeSCs. Furthermore, the commercialization of highly efficient and stable Pb-, Pb-free, and Sn-Pb-based PeSCs is discussed. Moreover, Pb recycling in Pb-based solar panels is briefly reviewed; finally, prospects are presented.

APbI 3 Lead-Based PeSCs
Most of the PeSCs studies have focused on lead-based APbX 3 structures such as MAPbI 3 or FAPbI 3 . With regard to solar cell applications, narrow bandgaps are of primary consideration to harvest maximum incident sunlight. Since the first report of halide PeSC using MAPbI 3 light absorbing layer by Miyasaka group, [3b] the PCE of MAPbI 3 cell was 3.8% and currently achieved the highest efficiency of 22.28%. [15] To form or maintain a thermodynamically stable cubic phase, Cs, MA, and FA cations located at the A-site of APbI 3 require high energy. For example, the alpha phases of MAPbI 3 or FAPbI 3 , which require relatively low energy, are thermodynamically stable at 100 or 150°C, respectively, whereas CsPbI 3 requires much higher energy (above 300°C) to perovskite alpha-phase transition from nonperovskite delta phase (Figure 1f,g). CsPbI 3 exhibits high thermal stability because gasification of its constituents is not observed until reaching the melting point of 481°C, whereas in the case of MAPbI 3 , MAI vaporization occurs at about 150°C or higher, so its phase stability becomes difficult to control. [16] In the case of FAPbI 3 , it is thermally decomposed at 332°C, and the organic part, FA, is decomposed with reaction and gasified with the form of sym-triazine, formamidine, or ammonia. [17] These tasks led researchers to have motivations to 1) explore perovskite with better chemical and physical properties and stability by mixing A-site and/or modulating X-site, and 2) improve efficiency and stability with passivation strategy while freezing the phase at room temperature by introducing various additives.

A-Site Modification
In APbI 3 perovskite structures such as CsPbI 3 or FAPbI 3 , photoactive phases include α (cubic), β (tetragonal for CsPbI 3 , orthorhombic for FAPbI 3 ), or γ (orthorhombic for CsPbI 3 , tetragonal for FAPbI 3 ), and those are apparently black colored, which are known as perovskite black phases with different bandgaps. Concerning the structural formability of ABX 3 perovskite, appropriate monovalent cations at the A-site can be predicted using the tolerance factor (t) equation proposed by Goldschmidt, [18] t = (r A þ r B )/[2 1/2 (r B þ r X )], where r A , r B , and r X are the ionic radii in A-, B-, and X-sites, respectively. The ideal cubic structure can be formed with a tolerance factor of 1. Several studies show that chemical formulas with tolerance factors ranging from 0.8 to 1.0 can form ABX 3 perovskite structures at room temperature. [7a,19c] The tolerance factor of APbI 3 with Cs, MA, and FA at position A lies between about 0.8 and 1.0, with Cs extremely close to 0.8, FA at around 1.0, and MA in between them. In contrast, APbI 3 with other monovalent cation(s) (e.g., Na þ , K þ , Rb þ ) that are outside this range forms photoinactive nonperovskites, so-called δ-phase or yellow phase with an orthorhombic or hexagonal structure, which are not suitable for solar cells due to their wide bandgaps.
In the APbI 3 perovskite structures, the black phase is thermodynamically metastable at room temperature, whereas the yellow phase is stable due to the lowest free energy. In other words, as long as the black phase is formed, it is kinetically transformed over time into a yellow phase. Morronnier et al. studied the evolution of crystal structure with the temperature change of CsPbI 3 . [20] According to that study, CsPbI 3 in gamma phase transforms to alpha phase reaching at 645 K and retains as a stable phase. Conversely, when the temperature is gradually cooled to room temperature, the phase transition from alpha occurs in the order of beta (510 K) and gamma (325 K), and in turn, it is stabilized as a delta phase at room temperature (Figure 1a,c-e). This phase transition can be further accelerated as the perovskite is exposed to moisture or thermal energy. As one alternative to resolve this structural instability, a method of mixing A-site ions can be suggested. Mixing or multiple substitutions on A-site cations are accompanied by alteration in energy level as well as stability of the perovskite. An attempt to create a mixed A-site composition results in a change in the bandgap values yielded by the type and size of the ions involved. In this case, it was believed that A-site cation does not contribute to the band edge in the band structure. Recently, a rather contrasting study suggested that the orbital state(s) of A-site ions contributes to the electronic structure at the band edge. [21] Beyond the initial study of PeSCs by constructing A-site with single cations (e.g., MAPbI 3 ), in the case of double, triple, and multication mixed perovskite that mixes several cations on the A-site (e.g., (Cs,MA,FA)PbI 3 ) for devices, those have been shown to be effective in improving stability and efficiency, and have formed the mainstream of research. MA and Cs in FAPbI 3 -based perovskite improve crystal stability of perovskite. [22] In the mixed cation strategy, it was reported that the mixing cations in A-site successfully form a perovskite alloy with enhanced properties.
Some recent studies report from a different point of view that the mixed cations do not form an alloy in all the cation mixing but are phase-separated or exist in a state of microcrystals nearby the majorly grown perovskite grains. It revealed that when using representative cations of Cs, MA, and FA together, alloying is feasible, but other alkali ions, such as K and Rb, form a nonperovskite phase and exist as a mixed phase, or on the surface of the formed tiny crystals, which can act as a passivating species for which Kubicki et al. studied by using solid-state NMR analysis. [23] value gradually decreases. The bandgap ranges of MAPbX 3 , FAPbX 3 (1.48-3.02 eV), and CsPbX 3 (1.75-2.90 eV) differ, and when perovskites are prepared by lowering the dimensionality of the material, the bandgap values increase somewhat for all the perovskite compositions. CsPbI 3 has the highest bandgap among them (1.73, 1.53, and 1.48 eV for CsPbI 3 , MAPbI 3 , and FAPbI 3 , respectively) as shown in Figure 1b, but it has the advantage of being all inorganic that can be considered as a priority in the APbI 3 family when high stability is required. MAPbI 3 is intrinsically unstable while operating as solar cells due to the ease of volatilized MA at a specific temperature. According to the Shockley-Queisser (SQ) detailed balance limit, the ideal bandgap is between 1.1 and 1.4 eV to achieve a maximum PCE between 32.74% and 32.91%, and the most suitable material is FAPbI 3 perovskite. [24] For the application of the same PeSCs for specific applications such as hydrogen or carbon capture from water or carbon dioxide, an open-circuit voltage (V OC ) value of about 1.5 V or higher is recommended; therefore, bromide perovskites such as CsPbBr 3 (%1.6 V) [25] can be selected as a suitable photoactive material (discussed in a later section).
Another benefit of mixing the halide is that a higher energy value is required for the phase transition from the black to yellow phases of the iodide perovskites, which can enhance phase stability. [26] A strategy of halide mixing is beneficial for multijunction tandem solar cells. Perovskite and silicon-perovskite tandem solar cell may overcome the SQ limit by taking advantage of multiple energy bandgaps. To accomplish this, it is necessary to locate wide-bandgap perovskite with bandgaps greater than 1.63 eV that can be used as a front cell, and narrow-bandgap perovskite or silicon with bandgaps less than 1.12 eV as a bottom or rear cell. [27] However, the nonuniform distribution of halide causes large V OC losses and compositional instability due to halide segregation induced by light, heat, electric fields, etc., which makes their application limited. [28] Hoke et al. prepared perovskite thin films with the MAPb(Br x I 1Àx ) 3 composition (x = 0.6), and experimentally demonstrated that phase segregation into iodide-rich minority and bromide-enriched majority domains occurs significantly when light is irradiated; here, the minority domain acts as a recombination trap center. It seems that this phase separation induces changes in the optical properties and reduces the electrical bandgap, and lowers the V OC . As a method to solve this problem, the Cl addition suppresses the atomic migration by the electric field applied during the operation of the device. Yin et al. calculated that the formation energy of CsPbI 3Àx Cl x (111 meV) is higher than that of CsPbI 3Àx Br x (40 meV) and CsPbBr 3Àx Cl x (21 meV). [29] The introduction of Cl ions into the APbI 3 perovskite lattice affects their electronic structure [30] and suppresses deep trap states due to the higher formation energies of the Pb Cl antisite defects. [31] Cl is added to the perovskite film using various additives, including ammonium chloride, [32] metal chloride, [33] and chloride-based ionic liquid. According to PeSC studies using Cl additives, Cl evaporates easily during thermal annealing of perovskite thin films, and therefore, Cl does not necessarily remain within the perovskite lattice, and contributes significantly to the improved quality of APbI 3 thin films or the passivation effect for halide defects in those films. [34] Mahmud et al. introduced alkyl ammonium species to cover the perovskite film so that the Cl ion does not evaporate but rather diffuses into the perovskite structure at 100°C, a relatively mild annealing temperature. Xu et al. demonstrated that the triple-halide PeSC was found to be highly stable with the addition of Cl, which prevented phase separation under 100 solar illuminations. At 60°C, the translucent top cell showed less than 4% degradation after 1000 h of operation at MPP under 0.77 sun illumination in an N 2 atmosphere. This top cell was integrated with silicon bottom cells, resulting in a PCE of 27% in a monolithic tandem with two terminals, each with an area of one square centimeter. Jeong et al. introduced formate anion as pseudohalide into α-FAPbI 3 to passivate defects on perovskite. From DFT calculation, the formate anion was discovered to have stronger binding affinity toward iodide vacancy than other anions (Cl À , Br À , I À , BF 4 À ), which can effectively eliminate defect density elevating the device PCE to 25.6% (certified 25.2%). There are still many highly efficient PeSCs with mixed cation and anion perovskite, but some reconsider its mixed cation and mixed anion strategies for the stability of perovskite. For those reasons, future studies are expected to focus on additives that can be effectively used for single-cation-and anion-based perovskite.

Passivation
Nearly all materials contain defects during the synthesis and the subsequent processing of them. [35] In semiconductor materials and their optoelectronic devices, efforts to remove defects for the synthesis of flawless materials are important considerations for the fabrication of high-efficiency devices. As an example, the effective suppression of detrimental defects in perovskite photovoltaics enhances the efficiency and long-term stability of the devices. Although lead halide perovskites have a unique characteristic of defect tolerance, it is not entirely immune from defects that can deteriorate their optoelectronic properties. In halide perovskite materials, defects are readily formed on the surface and grain boundaries due to their low formation energies and low activation energies for the migration of halide-ion vacancies. [36] The presence of defects in perovskite films can exert an influence on the photovoltaic characteristics of PeSCs, such as V OC , short-circuit current density ( J SC ), fill factor (FF), and PCE. With this regard, numerous attempts to reduce defects have been reported in many research reports on perovskite materials and devices. In recent years, a variety of effective passivation approaches have been developed for the reduction of defects in thin film PeSCs ( Table 1). Lead halide PeSCs have a theoretical efficiency limit of approximately 31%, [37] and among the attempts to break the record efficiency, passivating the perovskite surface and grain boundaries with various chemicals is currently considered the most effective method to eliminate defects. [38] In addition to improving the efficiency of perovskite devices, this method often leads to additional benefits such as long-term operational stability.
Octylammonium (OA) cation which has a long alkyl chain is the most used passivation layer in PeSCs. Hydrophobicity of OA obstructs undesired electron migration and H 2 O infiltration. [39] The Seok group reported that full-armor perovskite grain is formed by OAI with preferential orientation, resulting in the enhanced device stability (Figure 2a). [38a] In the following research, the Seok group deposited OAI on perovskite surface to make 2D passivating layer, and the perovskite showed significantly increased humidity resistance. [39] As OA passivation has been currently being studied, Yang group suppressed negative work function shift by introducing tosylate anion instead of I À .
[38b] Likewise, various types of octylammonium sulfonate salts were reported, and camphorsulfonate anion showed better performance than the tosylate anion with suppressed iodide migration and thus reduced hysteresis ( Figure 2b). [38c] Besides the OA salts, in a recent study, Zhu et al. effectively passivated the perovskite surface by 3-(aminomethyl)pyridine (3-APy) treatment, inducing reaction between 3-APy and FA cation on the surface. This reaction resulted in more n-type surface, which raised the p-i-n device efficiency (25.49%) to state-of-the-art levels. [40] In addition, studies on improving stability by using 2D perovskite as passivation layer have been carried out. [41] The 2D perovskite passivation layer induces excellent stability by stabilizing the interface between perovskite and the charge transport layer and thus resulting in protecting perovskite ( Figure 2c). Besides coating an additional passivation layer on the perovskite layer, passivate-charge transport layer (p-CTL) that increases the extraction of charge and reduces recombination is also used. [42] p-CTL suppresses the decomposition of perovskite, [43] enhances the crystallinity of perovskite, [44] and improves humidity stability with encapsulating [45] by replacing materials with stability issues in solar cells such as poly(3,4-ethylenedioxythiophene):polystyrene sulfonate (PEDOT:PSS) or spiro-MeOTAD.

Lead Toxicity Issues
Lead-based PeSCs have been studied in a recent decade due to their efficiency comparable to crystal silicon solar cells and the advantages of low energy payback time (EPBT; EPBT is the time which is needed to generate a similar amount of energy that is consumed to fabricate a photovoltaic device) within 0.5 years through a simple process and more details on this would be discussed in Section 5. [7b,14a,46] However, commercialization is far due to low heat and moisture instability and lead toxicity problems. [47] There are many indicators that lead is harmful to the human body and the environment enough to be selected as one of the ten chemicals of public health concern by the WHO. [48] Pb has the greatest effect on the nervous system among "2-step" indicates sequential deposition of BX 2 solution followed by AX solution; b) () is certified value. [38c] Copyright 2023, Elsevier Ltd. c) Adapted with permission. [41] Copyright 2022, American Association for the Advancement of Science.
www.advancedsciencenews.com www.advenergysustres.com many human organs. Adults may have cognitive performance decline, and children with soft tissues are sensitive to even low levels of Pb, which may be led to behavioral problems, learning deficits, and low IQ. [49] Old and middle-aged people could be caused by increased blood pressure and anemia, and pregnant women could have a miscarriage, and males could be reduced fertility. [50] Especially lead salts comprising halide perovskites have high solubility in water. The solubility product of PbI 2 is on the order of 10 À8 , whereas that of other common heavy metal compounds used in solar cells such as CdS, PbS, and CdTe is 10 À27 -10 À34 . [51] Because it is easy to affect land and aquatic creature, attention to lead leakage problems requires immediate attention. [52] However, there are other perspectives on this; Cahen group investigated Pb loss to the environment through the worst-case scenario of the damaged PeSCs exposed to rain; they estimated it to be far from catastrophic for the environment. [53] Even if a large solar electrical power generating plant is destroyed, lead could be remediation because it is strongly adsorbed by the topsoil. The harmfulness of lead is negligible from 0.01% to about 1% through the life cycle assessment (LCA), which represents influencing indicators at five stages: raw material extraction, synthesis of the starting materials, manufacturing, using, and dismantlement. [54] In addition, lead may be preferred over tin because the cumulative energy requirement of lead-based perovskite is less than tin-based perovskite. [54,55] Baxter group suggested that lead from perovskite accounts for only 1.1% of 1.6 million metric tons of lead consumed in the U.S. domestic market annually, and that lead emissions can be reduced by 2-4 times if perovskite replaces electricity in the U.S. [56] They also introduced a new life cycle metric called toxicity potential payback time (TPPBT), in which the minimum service life of offsetting conventional energy required for PeSCs to overcome both their manufacturing burden and the risk from the release of lead. TPPBT of PeSCs takes 2 years and it is 20 times lower than the typical electrical mixes in the U.S. As a result, as lead in perovskite negatively affects the soil only in extreme scenarios, the management of lead emissions may result in little environmental impact. [57] In fact, toxic substances such as asbestos, hydrazine, sulfuric acid, and benzene are used and managed in large quantities in the industry because they are effective and have no alternatives, so lead cannot be excluded from commercialization only for risk. [58] However, as the risk of exposure to lead cannot be completely ignored, attempts to manage and recycle lead are needed instead of reducing the use of lead. The easiest way to manage PeSCs is to minimize lead leakage with encapsulation. However, both physical and chemical encapsulation are not perfect because they can be destroyed by external pressure, heat, and UV rays. Therefore, many studies have been conducted on encapsulation strategies for stability and minimizing lead leakage. [59] In addition to reducing lead leakage in the device, research is also being conducted to recycle the used device. [60] Park et al. reported that 99.97% of lead from PeSCs was recycled by forming lead iodide through an adsorbent. [61] They introduced iron-decorated hydroxyapatite (HAP/Fe) hollow composite, which modulates surface charge for higher Pb-adsorption capacity and imparts magnetic properties for the separation of Pbadsorbed HAP/Fe from nonaqueous solution on the level of 15 parts per billion of Pb, thus recycling 99.97% of Pb ions in the form of PbI 2 . Bo Chen et al. proposed using carboxylic acid cation-exchange resin to adsorb and separate Pb from the solution, and then, release adsorbed lead ions in the form of precipitated PbI 2 by ion-exchange process, with 99.2% recycling efficiency. [62] Priya group showed more than 20% of PeSCs by recycling almost 100% of all layers of devices. [63] Lisha Xie et al. recovered lead from used lead acid batteries, 1) via desulfurization reaction of sodium carbonate with lead sulfate (PbSO 4 ), which converted PbSO 4 into lead carbonate, 2) PbO 2 from the lead acid batteries was reduced to PbO using H 2 O 2 , 3) acetic acid was also added to the solution to convert lead into clear solution, and 4) finally HI was added into the filtrate to precipitate yellow powder of PbI 2 with %83% recycling yield. [64] The PbI 2 synthesized via this method showed the efficiency of 20.45% for inverted PeSCs with high air stability. The advantage of recycled PeSCs is that it not only prevents lead contamination issues but also reduces the manufacturing cost and time of PeSCs, further reducing the payback period. [65] According to the analysis carried out by Xueyu Tian et al., recycling of the PeSCs plays an even more important role and leads up to 72.6% reduction in EPBT and 71.2% reduction in greenhouse gas (GHG) emission factor (final EPBT of the reused perovskite solar module can be 0.09 years with 13.35 g CO 2 equivalent per kWh in comparison to 1.3-2.4 years EPBT and 22.08-38.11 g CO 2 equivalent per kWh for silicon-based solar modules). [14a] Nonetheless, as lead is toxic to human body and the environment, it is likely to be subject to strict legal regulations that may hinder its commercialization. Many researchers have focused on developing lead-free or lead-partially substituted perovskites in an effort to overcome this limitation; however, their efficiency is still relatively low. Specifically, lead-free perovskites, such as tin perovskites, fall short of commercialization requirements due to prerequisites such as low solar cell efficiency, which is caused by the poor chemical stability of tin. In the following sections, we will examine the research status and applications of lead-free and less-lead PeSCs, including approaches for improving their characteristics, stability, and efficiency.

Tin(II)-Based PeSCs
Among numerous alternatives to lead-based perovskite, including tin-, germanium-, bismuth-based perovskites, and double perovskites, tin-based perovskites such as CsSnI 3 , MASnI 3 , and FASnI 3 seem to be the most promising candidates for designing a nontoxic photoactive layer in a solar cell. In 2012, Kanatzidis et al. and Shum et al. first demonstrated that CsSnI 3 perovskite can act as a light absorber in dye-sensitized [66] and Schottky-type solar cells. [67] The first tin-based PeSCs using MASnX 3 , which have a conventional (n-i-p) structure, were reported by the Snaith group and the Kanatzidis group to achieve PCEs of 6.4% [68] and 5.73%, [69] respectively. Like lead, tin is also one of the group IV elements and the divalent cation of Sn 2þ has an ionic radius of 118 pm, [70] almost the same size as Pb 2þ cation (119 pm). These similarities attribute to tin-based perovskite displaying the closest optoelectrical and crystallographic properties to lead-based perovskite. Tin-based perovskites have a direct bandgap in the range of 1.2-1.4 eV, which has a great potential to exhibit high performance with respect to the SQ limit (about 33%). [24b,71] Furthermore, without extrinsic factors (related to grain boundaries, energetic disorder, impurities, and so on), the charge carrier mobility of tin-based perovskites can excel that of lead-based perovskites because the lighter element, tin, has smaller Fröhlich interaction associated with its higher longitudinal optical (LO) phonon frequencies. [72] However, as shown in Figure 3b, atomic energy levels from Sn and Pb result in different energy levels of perovskites. [73] The conduction band minimum (CBM) of perovskite is determined by the hybridization between the metal pand halide s-orbitals, and the valence band maximum (VBM) is determined by the metal sand halide p-orbitals. Due to the higher lying s-orbital states of Sn compared to s-orbital states of Pb, both the VBM and the CBM are upshifted by replacing Pb with Sn. [74] This energy-level mismatch makes it difficult to choose adequate charge transport materials. But more than that, as tin has four valence electrons, the divalent Sn 2þ cation is readily oxidized to tetravalent Sn 4þ , while heavier element Pb 2þ is stable due to the stable 6s 2 electrons with a stronger inert pair effect. [75] Tin-based perovskites suffer from their high background carrier density and high defect density caused by the oxidized Sn 4þ . In addition, the high Lewis acidity of Sn 2þ accelerates crystallization, resulting in poor film quality with higher defect density (Figure 3a). Severe degradation of tin-based perovskites can be promoted by O 2 , H 2 O, and Sn 4þ in proposed mechanism as depicted in Figure 3c. [76] These problems became a serious barrier to their applications in the first stage of tin-based PeSCs. Along with significant demand for nontoxic PeSCs, a decade of research has improved performance of tin-based PeSCs through various types of engineering approaches [77] to suppress Sn oxidation, [78] control perovskite growth, [79] and improve charge extraction. [80] In 2021, the highest PCE of 14.81% (certified 14.03%) was achieved. [13] The following sections will describe the mainstream of research topics (additives, composition modification, device architecture, and 2D/3D heterostructure) to overcome the issues of tin-based PeSCs.

Additives
Self-doping effect of tin-based perovskites induced by oxidation of Sn 2þ to Sn 4þ derives high conductivity, i.e., metallic behavior, restricting usage as solar cell materials. [81] In this regard, additive engineering becomes essential to remove the self-doping effect and the additives even provide better crystallizing conditions in the solution, which can improve quality and stability of perovskite films. In 2014, Mathews et al. demonstrated that the addition of SnF 2 reduces Sn vacancies and carrier density to inhibit metallic behavior and V OC loss through Shockley-Read-Hall mechanism (Figure 4a,b). [82] It was discovered that SnF 2 induces excellent morphology and stability, [83] which led SnF 2 being used as a major additives for tin-based PeSCs. Furthermore, it was found that tin halides such as SnCl 2 , [84] SnBr 2 , [85] and SnI 2 [86] exhibited a similar effect in tin-based PeSCs. However, an excess amount of these materials can induce separate phases, which  Some organic antioxidants are especially efficient with tinbased perovskite as an additive with reducing power. In addition to preserving the oxidation state of Sn 2þ , antioxidant additives (e.g., crystallization delay, antifloating defects, etc.), they also play a major role in improving the quality of the perovskite thin film. [88] For example, phenylhydrazine hydrochloride (PHCl) was reported to show self-repairing ability reduced trap states, better energy-level alignment, and improved stability. [89] In the following research, by incorporating Br À with PHCl, 13.4% of PCE was achieved with suppressed Br À segregation and increased stability. [90] In another report, gallic acid eliminated phase separation induced by SnCl 2 . [91] The reducing property of these materials greatly improved air stability. Particularly, the devices with gallic acid maintained %80% of the initial efficiency after 1000 h in ambient air. In addition, one of Mashima's reagents, [92] 1,4bis(trimethylsilyl)-2,3,5,6-tetramethyl-1,4-dihydropyrazine, was demonstrated to scavenge Sn 4þ ions effectively. [93] Especially, SnF 2 was reduced rather than SnI 2 to form Sn 0 nanoparticles in this case, resulting in Sn 4þ -free perovskites. Recently, a bi-linkable reductive cation, formamide, was reported by Chen et al., which improved the perovskite quality with increased grain size, preferential growth, better crystallinity, reduced trap-state density, and longer charge carrier lifetime. [94] The device showed greatly improved stability, which retained 90% of its initial PCE in N 2 atmosphere for one year. Besides, Mora-Seró et al. used a combination of dipropylammonium iodide and an inorganic reducing agent, sodium borohydride, to demonstrate the synergetic effect. The device showed higher performance than the sole additive cases and greatly improved operational stability with maintained 96% of initial PCE after 1300 h at N 2 atmosphere (Figure 6a). [95] Liquid additives can be used in the perovskite precursor solution as well. Although dimethyl sulfonate (DMSO) was found to oxidize Sn 2þ , [96] DMSO has been the most popular liquid additive to tin-based perovskite as well as lead-based perovskite. In particular, DMSO forms Lewis acid-base adducts with Sn 2þ , which slows crystallization, so that high-quality films with enlarged grain size can be easily obtained without leaving any residue. [97] Having a similar role in forming Sn(I/F) 2 -TMA complex, trimethylamine (TMA) was reported as another liquid additive used in a 2-step deposition method. [98] The 2-step deposition method has been rarely studied because 2-propanol (IPA), the main solvent for the AX solution, is destructive to tin-based perovskite. [99] Nevertheless, the TMA enhanced the device PCE from 4.34% to 7.09% with significantly improved film coverage.

Composition Modification
Modifying the composition of tin-based perovskite can give major changes to the properties such as lattice stability, thermodynamical stability, crystallinity, and optical bandgap. [104] Plenty of composition engineering methods have already been verified in lead-based perovskites. [105] To apply the verified composition engineering to tin-based perovskites, the lattice stability, closely Effect of SnF 2 used as an additive, c) composition (pseudohalides), d) paradigm change from conventional structure to inverted architecture, and e) 2D/3D structured perovskite described as the PEAI ratio increases. a,b) Adapted with permission. [82a] Copyright 2014, Wiley-VCH. c) Left: Adapted with permission. [119] Copyright 2020, American Chemical Society. c) Right: Adapted with permission. [120] Copyright 2020, American Chemical Society. d) Reproduced with permission. [80] Copyright 2021, American Chemical Society. e) Reproduced with permission. [133] Copyright 2017, American Chemical Society. related to Goldschmidt tolerance factor, should be first considered. Due to the similar radius of Sn 2þ and Pb 2þ , the Goldschmidt tolerance factor is calculated nearly the same for Sn-and Pb-based perovskites; thus, the compositional engineering associated with the A-or X-site atoms is similar. [106] In the case of bandgap energy, the bandgap of both lead-and tin-based perovskites increases as the ionic radius of the X-site anions decreases (E g : I < Br < Cl). However, in the case of A-site cations, the bandgap increases for tin-based perovskites with a larger radius (E g : FA > MA ≥ Cs) while it decreases for lead-based perovskites. [106a,107] As AsnI 3 already has bandgaps in the optimal range for solar cells at around 1.3 eV, it is not necessary to tune the bandgaps for solar cell performance. Rather, most of the research on perovskite composition has focused on lattice stability and film quality. Just mixing widely used kinds of perovskite such as CsSnI 3 , MASnI 3 , or MASnBr 3 into FASnI 3 precursor solution, superior perovskite films can be obtained with reduced pinhole concentration, decreased carrier recombination, larger grain size, and enhanced stability. [106] Guanidinium (GA) can occupy A-site in tin-based perovskite as well, whereas it is not stable in lead-based perovskite. [108] It has been reported that GA þ can be substituted by 20% in the FASnI 3 structure and the PCE of the solar cell exhibited as 9.6%. [109] However, these alloys are potentially segregated into each phase, which deteriorates operation and long-term stability. [110] As with MAPbX 3 perovskites, the volatility of MA destabilizes MASnX 3 perovskites. [111] Owing to the phase segregation issues, composition modulation seems to have been limited to a low ratio below 10% recently. In 2020, by incorporating 10 mol% ethylammonium iodide (EAI), 1 mol% ethylenediammonium diiodide (EDAI 2 ), and 5 mol% germanium in FASnI 3 -based PeSCs with ethylenediamine (EDA) passivation, high PCE of 13.24% was achieved. [112] In addition, a following research incorporated 1 mol% ethylenediammonium dibromide (EDABr 2 ) instead of EDAI 2 and achieved a PCE of 14.23% in 2022 (Figure 6c). [113] The EDABr 2 was demonstrated to possess better capability on passivation of deep-level defects due to the synergetic roles EDA 2þ cation and Br À anion. Monovalent polyatomic anions having a similar size to iodide (I À ), namely, pseudohalides, are also used to substitute halides in X-site. [114] It is well known that SCN À , [115] BF 4 À , [116] PF 6 À [117] and HCOO À[3e] act as pseudohalide in lead-based perovskites. However, unlike polyatomic organic cations in the A-site, there have been rare applications of pseudohalides to tin-based PeSCs. [118] As SCN À and BF 4 À have been reported to compose stable tin-based perovskite with high content, more than 60% of total X anions (Figure 4c), pseudohalides are expected to have an important role to overcome the instability of tin-based perovskite in further study. [119,120]

Device Architecture
As originated from solid-state dye-sensitized solar cells, the conventional structure of PeSCs is an n-i-p structure, where "n" is n-type electron transport material (ETM) such as metal oxides (TiO 2 , SnO 2 , etc.), "i" is intrinsic light absorber, i.e., perovskite, and "p" is p-type hole transport material (HTM) such as spiro-MeOTAD. [3a,66] Organic materials, e.g., C 60 , phenyl-C61-butyric acid methyl ester (PCBM), and indene-C 60 bisadduct (ICBA), are hardly used in n-i-p architecture due to the problem from processing procedure because it can be washed out while coating perovskite precursor solution. [121] While n-i-p structured leadbased PeSCs using metal oxides as electron transport layer (ETL) show the highest PCE of 25.8% (certified 25.5%) among all kinds of single-junction PeSCs, [122] n-i-p structured tin-based PeSCs show inferior performance because oxygen vacancies on the metal oxides accelerate the oxidation of Sn 2þ making poor interfacial contact, and the energy-level mismatch between the charge transport layer and perovskite interrupts favorable charge extraction. [77b,80,123] In addition, dopants required to improve the electrical properties of HTMs like spiro-MeOTAD and PTAA which are the most efficient HTMs for lead-based PeSCs, [124] also deteriorate tin-based perovskites. [125] The record PCE of tinbased PeSCs with n-i-p architecture remains at 9.06%, [126] which has a large gap with that of p-i-n architecture, 14.81%, [13] despite efforts to introduce functional interlayers that prevent ion migration [127] and relax the residual strain on the perovskite film.
Due to the limitation of n-i-p PeSCs, inverted structure (i.e., p-i-n) has been studied in much detail. For p-i-n structure, PEDOT:PSS and fullerene derivatives (C 60 , PCBM, and ICBA) are generally used as HTM and ETM, respectively. [80] With easy processability, remarkable stability, and superior optoelectronic properties, PEDOT:PSS is used in most p-i-n devices with high performance. [13,128] For ETM, there is a tendency to adopt C 60 for 3D FASnI 3 and ICBA for 2D/3D PEA x FA 1Àx SnI 3 as shown in Table 2. It may be concerned with energy-level difference, that is, CBM of FASnI 3 increases with the incorporation of PEA toward suitable value with ICBA possessing shallower lowest unoccupied molecular orbital (LUMO) energy level ( Figure 5). [99,129] In addition, because of the energy-level difference, ICBA exhibits higher V OC but lower J SC than those of C 60 . The recent study reported in 2022 showed that 2D/3D structured tin-based PeSCs with ICBA reached V OC over 1 V by incorporating GASCN and the obtained PCE was 13.79%. [130] Most recently, Zhu et al. developed a pyridine-functionalized fullerene derivative (C 60 -Bpy) to improve the interface between perovskite layer and C 60 ETL. The C 60 -Bpy bonded more strongly to the perovskite surface and complemented energy-level alignment, resulting in the achieved device PCE of 14.14% with reduced surface defects, suppressed nonradiative recombination, and higher electron extraction capability ( Figure 6d). [128b] For the late effort to seek new hole transport layer (HTL) instead of PEDOT:PSS, modified energy level of P-SnO 2 and T-SnO 2 was deposited under and on the perovskite layer, respectively, by plasma treatment, which showed increased PCE to 14.09% with efficiently reduced trap density and fast electron transport (Figure 6e). [131] 3.1.4. 2D/3D Heterostructure In recent research, 2D/3D heterostructured tin-based PeSCs have shown the highest performance with their superior stability and V OC . [13a,128] As stability issues are more critical for tin-based perovskite, 2D/3D heterostructure with hydrophobic A-site cation has become a key to minimize the efficiency loss by impeding degradation from outside of the perovskite grains. Alike lead-based perovskites, 2D/3D heterostructured tin-based perovskites commonly adopt bulky organic cations as A-site cations such as butylammonium (BA), [132] phenethylammonium (PEA), and its derivatives (Figure 4e). [13] Especially, as Ning et al. achieved PCE of 5.94% with PEA 2 FA nÀ1 Sn n I 3nþ1 in 2017, PEA has been popularly used in most of the reports about 2D/3D tin-based PeSCs, as it induces perpendicular growth of the perovskite domain. [133] By incorporating modified PEA, namely, 4-fluoro-phenethylammonium bromide, He et al. achieved the highest PCE of 14.81% (certified 14.03%) among tin-based PeSCs with effectively suppressed Sn 2þ oxidation by 2D/3D microstructure. [13] The quality and orientation of 2D/3D perovskite films were enhanced by additives like EAI, aminoguanidine hydrochloride (NH 2 GACl), and guanidinium thiocyanate (GASCN). [130] Most recently, in 2022, Mi et al. used trimethylthiourea as an additive to perovskite precursor solution which induced highly oriented perovskite film and achieved 14.2% (certified 14.0%) with great robustness maintaining 85% of initial PCE in ambient air (75% relative humidity) (Figure 6f ). [128a]

Germanium(II)-Based PeSCs
Another group IV element, a lighter element, germanium (Ge) also can construct perovskite instead of Pb or Sn. However, as a much lighter element, germanium-based perovskites have much different properties compared with Pb 2þ and Sn 2þ analogous. Germanium-based perovskites are even more unstable than www.advancedsciencenews.com www.advenergysustres.com the tin-based perovskites because Ge 2þ is more likely to oxidize to stable Ge 4þ and the much smaller radius (73 pm) is not adequate for the ABX 3 type of perovskite in terms of Goldschmidt tolerance factor. The energy levels are also much different (e.g., CBM of FAGeI 3 is À3.15 eV while that of FASnI 3 and FAPbI 3 is À3.79 [13] and À4.0 eV, respectively), and they have larger optical bandgap over 1.6 eV. [134] Due to this discrepancy, although many theoretical studies have been conducted on germanium-based perovskite, the application of germanium-based perovskites in solar cells has been rarely reported. [135] In 2015, the first germanium-based PeSCs with a photoactive layer of CsGeI 3 , MAGeI 3 , or FAGeI 3 were reported. [134] Figure 7 illustrates that the bandgaps of MAGeI 3 and FAGeI 3 are too large to absorb light efficiently, resulting in very poor performance of less than 0.20% of PCE in all fabricated PeSCs, including CsGeI 3 solar cells. In 2018, the PCE of germanium-based PeSCs rose to 0.57% by substituting 10% of iodide with bromide. [136] These terrible performances seem to be caused by poor crystal growth, dreadful surface morphology, and instability of germanium-based perovskites rather than the energy-level mismatch. [134] Due to the performance issues, germanium-based perovskites are currently not used alone, but mostly in a mixture with tin-based perovskite. [128b,137] 4. Less-Lead PeSCs

Sn-Pb PeSCs
Lead-free perovskites have alleviated the toxicity problem of leadbased ones; however, due to their low efficiency and stability, the performance of lead-free PeSCs lags behind that of their leadbased counterparts in real-world applications. Tin-lead perovskites, created by mixing lead with tin in perovskite structures, can be a compromise between mitigating toxicity and achieving high efficiency and stability. [138] However, due to the different crystallization rates of Sn and Pb perovskites, Sn-based Figure 5. LUMO-level difference of fullerene derivatives p-i-n structure: C 60 , [198] PCBM, [199] ICBA, and ICTA.
[129a] Figure 6. The latest research topics for tin-based PeSCs. a) Operational stability improvement by dipropylammonium iodide and sodium borohydride. Performance enhancement by b) N-doped graphene oxide additive to perovskite and HTL with remarkable stability; c) EDABr 2 incorporation; d) C 60 -Bpy interface between perovskite and ETL; e) P-SnO 2 as HTL and T-SnO 2 as protection layer; f ) trimethylthiourea as additive to 2D/3D perovskite. a) Reproduced under the terms of the CC-BY license. [95] Copyright 2022, The Authors, Published by Elsevier Ltd. b) Adapted with permission. [102] Copyright 2022, Wiley-VCH. c) Reproduced with permission. [113]  perovskites crystallize before Pb-based perovskites. This results in nonuniform growth, creating numerous traps in the film that interfere with carrier movement; Chen group, Liu group, and Guo group have attempted to control the crystallization rates of both materials, [139] and promote vertical crystal growth. [140] The improved crystallinity reduces the residual stress (or strain) of perovskite, increasing the quality of the device. [141] Interestingly, as lead and tin share the octahedral cage of perovskite, the bandgap narrows to 1.2-1.3 eV (bowing effect) and the value is lower than that of lead-or tin-based perovskites. Hayase group and Kanatzidis group published the first Sn-Pb PeSCs in 2014 with efficiency values of 4.18% and 7.37%, respectively. [142] Recently, the Hayase group exceeded it by 23.2%, [138c] and currently the highest efficiency of 23.6% has been achieved by the Wakamiya group. [138d] Until now, attempts have been made to enhance the efficiency of Sn-Pb PeSCs similar to those of Sn-and Pb-based PeSC studies (Table 3): additive for antioxidation, [143] n-doping of self-p-doped perovskite, [144] and formation of 2D perovskite on the surface. [138b,145] The following section will describe the unique characteristics and potential applications of Sn-Pb PeSCs.

B-Site Modulation
In alloying Sn and Pb perovskites, a unique change in bandgap is observed (Figure 8a), also known as the bandgap bowing effect which lowers the bandgap below that of each pure composition. [146] The bandgap bowing of alloy perovskite is contributed by energy-level difference and lattice strain. Mainly, these band edge formations result from the energy-level mismatches between the atomic orbitals of Sn and Pb. [73] As mentioned in the Sn-based perovskite section, VBM and CBM of Sn perovskite are upshifted (Figure 8b). [74] Those mismatches lead to the narrow bandgap of Sn-Pb perovskites. As an indirect effect, smaller Sn leads to tilt of the octahedron and lattice contraction and this affects the bandgap value. [147] As discussed in earlier sections, changes in the A-and/or X-site ions affect the bandgap values in lead-and tin-based perovskites. Depending on the organic cation and halide, the bandgaps of Sn-Pb mixed perovskites range from 1.17 to 1.5 eV, with the lowest bandgap observed in all the ABX 3 perovskites when the ratio of Sn is 40%-70% to B element (Sn/Pb). [142b,148] Changing Sn content enables controlling the bandgap as well as decreasing the defect and suppressing the oxidation. Snaith group revealed that Sn content between 0.5% and 20% of the metal content in perovskite results in defects, whereas Sn content between 30% and 50% restores optoelectronic quality (Figure 8c). [149] A small concentration of Sn incorporation generated trap sites and reduced photoconductivity, photoluminescence lifetime, and photoluminescence quantum efficiency. Sargent group showed that the Sn content of less than 30% produced deep-level traps. [150] On the other hand, 50% Sn mixed alloy  [134] Copyright 2015, Royal Society of Chemistry. d-f ) Reproduced with permission. [136] Copyright 2018, American Chemical Society.
www.advancedsciencenews.com www.advenergysustres.com exhibited longer carrier lifetimes and enhanced defect tolerance with free of deep traps (Figure 8d). The degree of oxidation is changed according to the content of Sn. Angelis group suggests Sn-poor conditions promote Sn oxidation because it acts as a dopant, so Sn 4þ is promptly formed on the perovskite surface. [151] As compared to Pb, Sn can be easily oxidized, resulting in severe oxidation [152] and more defects. [153] However, it has been determined that content of around 50% of Sn is ideal for enhancing the quality and stability of perovskite films. The bandgap of Sn-Pb perovskites with 50% of Sn content closes to 1.2 eV could theoretically achieve sufficiently high efficiency (%32.74%); therefore, the advantage of low bandgap results in other versatile applications such as photodetectors and tandem solar cells.
Recently, many studies have been reported on photodetectors using Sn-Pb perovskite due to their broad absorption spectrum (%1000 nm). Due to the fact that the absorption spectrum of Pb-based perovskite is limited from 300 to 800 nm, these photodetectors have difficulties in detecting near-infrared (NIR) light. [154] Organic bulk-heterojunction (BHJ) layers absorbing the NIR region are employed onto perovskite film to extend the absorption spectrum to 1000 nm. [155] Still, Pb-based perovskite photodetectors combined with BHJ have low detectivity at NIR region. Fortunately, Sn-Pb-based perovskite absorbing NIR light allows higher detectivity and sensitivity due to reduced leakage current. [156] Another utilization is for perovskite tandem solar cells. The perovskite tandem solar cells typically divide into two parts: a bottom wide-bandgap solar cell absorbing high-energy photons and a top narrow-bandgap solar cell absorbing low-energy photons. To date, organic, [157] CIGS, [158] and Si [159] solar cells are usually used for top cells. Considering high efficiency, low manufacturing cost, and solution processability, Sn-Pb-based perovskite is a potential candidate for top cells, [144b,160] further discussion is provided in the following section.

Tandem Solar Cells
The advantage of bandgap tuning of perovskite is that it offers possibilities for application in tandem devices. In terms of SQ limit, a two-bandgap tandem solar cell can reach up to 45% efficiency, whereas a single-bandgap solar cell has about 33% efficiency. [161] In Figure 9a, the theoretical bandgap diagram of tandem devices is shown. [162] Based on the assumption that all incident light is fully absorbed by each subcell, a 0.  For series connection, the polarity (n-i-p or p-i-n) of the bottom (or front) cell and the top (or back) cell should be the same, and the structure is defined as n-i-p or p-i-n depending on the polarity. Sn-Pb PeSCs with n-i-p structures have a lower PCE than with p-i-n structures; thus, more research has been conducted on PTSCs with p-i-n structures (Figure 9d). However, Pb-based PeSCs that are used as wide-bandgap cells are able to achieve higher PCE and V OC in the n-i-p structure (Figure 9e). To fabricate a high-efficiency tandem device, researchers are conducting studies on Sn-Pb PeSCs with n-ip structure and Pb PeSCs with p-i-n structure which have lower performance. In the case of Sn-Pb PeSCs with n-i-p structure, Grätzel group controlled grain uniformity through chemical vapor deposition to obtain large and uniform film for high efficiency [163] and the Hayase group introduced a passivation layer between the metal oxide and SnI 2 to mitigate the defects. [164] Parallel to this, p-i-n structured PeSCs have been studied for application to tandem solar cells and now it has reached PCEs exceeding 26% (Figure 9f; Table 4). [138b,165] Even though the efficiency is far lower than the theoretical value, high V OC of PTSCs could be applied for versatile applications such as water splitting. [166] Water splitting, which has recently emerged as a topic of increased interest in green  [73] Copyright 2018, American Chemical Society. c) Adapted with permission. [149] Copyright 2020, Royal Society of Chemistry. d) Adapted with permission. [150] Copyright 2021, American Chemical Society. hydrogen production, is a process in which water decomposes into clean O 2 and H 2 using incident photons on PTSCs. The green hydrogen can be produced through carbon-free process by thermoelectric, pyroelectric, triboelectric, and photoelectric power without any pollutant production during the water decomposition. [167] Among them, the device using solar energy contains  photoelectrochemical (PEC) and photovoltaic-electrochemical (PV-EC) systems as shown in Figure 10.
The PEC system, which utilizes incident solar energy and carries out redox reactions in one device, has a simple structure and low device price, but needs external bias voltage and has lower solar-to-hydrogen (STH) efficiency (η STH = [J SC Â(1.23 V) Â η F / P Total ] AM1.5G , where η F is Faradaic efficiency) than PV-EC device so that H 2 production is not cost-effective in PEC systems. [168] In contrast, PV-EC device, where battery and catalyst are separated, could utilize two mature technologies immediately. The energy difference between before and after the water splitting is 1.23 V, but typically higher voltage is required to overcome the activation energy barrier for the process.
When a voltage higher than the activation energy is applied, H 2 is generated, and the amount of conversion to hydrogen is determined by current flow produced from PV part. The first perovskite PV-EC was demonstrated by the Grätzel group. [169] Water splitting was performed through two series-connected MAPbI 3 solar cells, but STH efficiency was 12.3% due to low J SC . To increase STH efficiency, they have used the silicon/perovskite tandem device rather than series-connected device; therefore, low J SC loss allowed STH efficiency of 18.7%. [170] Theoretical calculations indicate that water splitting through silicon/perovskite tandem cells could achieve a STH efficiency of 25% and a levelized cost of hydrogen less than 3 $ kg À1 . [171] Presently, STH efficiency reached 21.32% with perovskite/silicon tandem solar cells, [172] and this makes us believe the potential to exceed 20% of the STH efficiency with PTSCs. As the performance of Sn-Pb PeSCs improves, the best V OC and J SC of PTSCs achieved 2 V and 16 mA cm À2 , respectively. [165a] Assuming that the voltage of the device exceeds activation energy, the STH efficiency of 19.68% could be obtained with those V OC and J SC values. [166] Ongoing research on Sn-Pb PeSCs and PTSCs is expected to result in further improvements in device performance and STH efficiency.

A-Site Modification
For high efficiency of over 20% in Sn-Pb PeSCs, perovskite composition with mixed A cation should be adapted as shown in Figure 11a. FA and MA are mainly used, and a small amount of Cs is occasionally added. The figure shows the efficiency trend of ASn 0.5 Pb 0.5 I 3 PeSCs, and the highest efficiency is shown when FA-based perovskite contains 30 mol% of MA. The Podraza group has found the reason that mixed cation Sn-Pb PeSCs show higher efficiency is related to Urbach Energy (E U ). They measured the E U of perovskites with various cation compositions by photothermal deflection spectroscopy. E U of perovskite thin film and V OC deficit show a direct relationship with the reduction of the E U occurring with a beneficial decrease in the V OC deficit. Consequently, they show that FA and MA mixed in a similar proportion have the lowest E U and V OC deficit (Figure 11b,c). [173] Inorganic Cs is a well-known material that significantly improves photo-and moisture stability of perovskite film. [160,174] Jen group demonstrated that Sn-Pb-based perovskites that have partially replaced MA þ or FA þ with Cs þ retard crystallization rate to facilitate homogenous film formation. [175] As a result, device stability and performance were enhanced, particularly in compositions containing high concentrations of Sn (Figure 11d-f ).
The interesting feature of Sn-Pb perovskite has no regularity in the change of bandgap by the mixed A-site. APbI 3 becomes a larger bandgap as the smaller A-site materials is dominant, [176] while ASnI 3 has the opposite tendency. [177] However, in the case of Sn-Pb perovskite, there is no clear regularity because there is a difference in the degree of distortion and orbital binding of the perovskite lattice by the mixed B-site (Figure 12a

Other Less-Lead PeSCs
Sn-Pb perovskites are considered an alternative to lead perovskites, but another alternative such as Ge may be considered because it is not toxic and theoretically provides the same performance as lead PeSCs. Raman et al. reported the simulation results demonstrating that Ge-Pb-based PeSCs employing a MA(Pb,Ge)I 3 light absorber can theoretically achieve up to 30% efficiency. [178] In spite of this, as the Ge-Pb perovskite crystal is unstable (tolerance factor > 1), [175,179] studies on Ge-Pb devices have been limited. [180] Another simulation result suggested the potential for stable and efficient devices with ternary B-site mixed cation including Pb, Sn, and Ge. [181] There are experimental studies demonstrating the effective substitutes for less-lead perovskites, such as In, [182] Cu, [183] and Zn. [184] It is therefore recommended that further research be conducted on Sn-Pb PeSC as well as effective alternatives that are commercially viable.

Environmental Footprint of Lead in Lead-Based PeSCs
Due to the unprecedented growth of the population in the world, fulfilling their concomitant energy demands is a much-needed area of research now. To offset GHG such as CO 2 emissions to control global warming, the search for alternative green energy sources, especially solar energy, is a vastly growing area of research. To put a competitive energy harvesting technology, such as PeSCs, into market dominant silicon (monocrystalline and polycrystalline)-based solar cells or other photovoltaic technologies, such as CIGS, CdTe, GaAs, etc., EPBT, GHG emissions, and human health toxicity factors are among fundamental features to consider before the deployment of third-generation PeSCs into the market. [185] These factors, among others, are chief indicators in order to assess the feasibility of new solar technology. To determine these abovementioned critical factors, a life cycle analysis (LCA) previously mentioned is carried out which is a detailed framework to assess the environmental as well as energy impacts of the newly developed solar technology. [186] EPBT is less than 0.3 years for perovskite PV technology, which is based on the lab-scale modules using spiro-MeOTAD  [173] Copyright 2022, American Chemical Society. (d-f ) Adapted with permission. [175] Copyright 2016, Royal Society of Chemistry.  as HTL and gold as the top electrode. [46] This approach differs from the actual industrial coating techniques where the use of spiro-MeOTAD is not viable due to its complex synthesis and material consumptions and gold is also not feasible due to its scarcity; therefore, based on more industrially viable processes which include either gravure printing or spray coating, EPBT is still almost 0.5 years (Figure 13a). [186,187] This scenario implies that PeSCs will have their unique position in the global energy mix in the near future due to their quick EPBTs. Unfortunately, these EPBTs are based on very stable perovskite solar modules.
Recently, a detailed study on the perovskite solar modules is carried out by Sara Pescetelli et al., in which perovskite solar modules (4.5 m 2 ) have shown nearly 80% stability of their performance in outdoor realistic conditions over the period of nine months, and one of the major findings of this work underlines the improvements needed in the area of encapsulation processes/materials and lamination protocols. [188] As absorber materials for PeSCs are soluble in organic solvents, therefore, serious environmental and health concerns would arise in the future once manufacturing of PeSCs on an industrial scale starts apart from lead toxicity issues. Unlike organic solar cells, where halogenated solvents are used for solubilizing photoactive materials, the main absorber materials of PeSCs are soluble in halogen-free solvents (such as DMSO, DMF, etc.). Nevertheless, these solvents have serious health and environmental concerns if used on an industrial scale. Rosaio Vidal et al. carried out LCA of PeSCs to evaluate their effect on human health by assessing human health toxicity factor, and their environmental impacts from global warming/particulatematter aspect. [189] Among eight different solvents studied in their study (which includes DMSO, DMF, GBL, THF, DMAC, NMP, DMI, and DMPU), DMSO is found to be the most viable solvent with a minimum footprint on an industrial scale (Figure 13b). [189] Although DMF is the most commonly used solvent in combination with DMSO, due to their high Gutmann's donor number for the fabrication of efficient PeSCs on lab scale, [3e,190] ECHA (European Chemical Agency) has classified DMF in the category of the list of substances of very high concern and toxic to reproduction, and DMF also has highest DALY (disability-adjusted life year; a time-based measure that combines years of life lost due to premature mortality and years of life lost due to time lived in states of less than full health, or years of healthy life lost due to disability according to the Department of Data and Analytics and Delivery for Impart at WHO). [189,191] Therefore, DMSO is a most suitable solvent for PeSC fabrication on industrial scale according to their assessment. To address the inherent toxicity of lead in lead-based PeSCs and the use of most greener solvent (after water), Haijin Li et al. used ethanol solvent to fabricate Pb-free perovskite PVs to make dense and homogenous perovskite microstructure and achieved V OC of 0.84 V, but other PV parameters were rather poor and thus low efficiency was observed. [192] Zhong, Vaynzof, and Park groups changed the antisolvent used for perovskite or solvent used for HTL to an eco-friendly solvent and showed more than 20% efficiency. [14b,193] In 2022, the Seok group used a solvent of perovskite as ethanol to achieve 25.1% efficiency. [194] To commercialize PeSCs on an industrial scale, one needs to find a greener solvent which has minimum environmental footprint and has high donor number to make efficient PeSCs, and render enhanced long-term light soaking stability.

Conclusion and Outlook
The unprecedented developments in PeSCs over the last decade have attracted significant attention and resulted in considerable improvements in their efficiencies ( Figure 14) compared to those of other photovoltaic technologies. Furthermore, PeSCs are considered among the most promising solar cell materials for commercialization in traditional and emerging solar cell groups. Addressing the intrinsic instability of lead-based perovskites and the toxicity associated with lead is essential before selecting a commercially viable solar cell candidate. This dilemma has been a source of contention for perovskite researchers since Figure 13. a) EPBT of PeSCs in comparison to commercial solar energy technologies without recycling. b) LCA of solvents used in the production of PeSCs via their different utilization routes. a) Reproduced with permission. [186] Copyright 2021, Wiley-VCH. b) Reproduced with permission. [189] Copyright 2020, Springer Nature.
www.advancedsciencenews.com www.advenergysustres.com its inception. Notably, the current research on PeSCs is focused on overcoming instability and toxicity, in addition to improving the efficiency. Therefore, this review first discusses research on lead-based APbI 3 materials-which form the basis of the history of PeSCs-and high-efficiency solar cells with improved stability.
Most studies have noted that stability and efficiency enhancements of perovskites are associated with improved performance of devices. The most effective method is to mix the A-site and X-sites of APbX 3 perovskites and passivate the perovskite surface with various additives. Nevertheless, the problem of lead toxicity remains unsolved. Moreover, compromising with environmental and human hazards owing to the lack of an unparalleled leap in the efficiency of leadbased PeSCs in a short period will be challenging. Two proposed options are discussed in this review. The first option considers quantifying and managing the hazards of perovskites to prevent leakage into the environment and recycling lead from waste solar cells.
Essentially, alternatives to lead-based PeSCs may be lead-free or less-lead PeSCs. Among nonlead perovskites, tin halide perovskites are the most promising candidates. However, further research is necessary because their efficiency is approximately half that of lead-based perovskites owing to the prompt oxidation of tin. The research on tin halide PeSCs approximately matches the research strategy of lead halide solar cells owing to the similarity between the properties of Pb and Sn. Future research will heavily rely on absolutely fixing the divalent tin oxidation number. Moreover, a high-efficiency solar cell with a stable tin oxidation number is expected to perform comparably or even better than lead-based perovskites, which would eliminate concerns regarding environmental hazards. A comparison between lead-and nonlead-based PeSCs suggests that tin-lead PeSCs present compromised indicators of efficiency and eco-friendliness, which are the two advantages of these two PeSCs. Tin-lead perovskites have a lower bandgap than pure lead-or tin-based perovskites, which can be used in tandem cell configurations to produce hydrogen through water decomposition. Finally, the assessment factors and indicators for the commercialization of solar cells and the possibility of halide PeSCs approaching eco-friendly models with reduced carbon emissions are discussed.  15,200] FA, [122,201] and Cs. [195,202] b) Record PCEs of Sn-based PeSCs: MA, [68,203] FA, [82b,109,112,113,204] and Cs. [82a,84À86,205]  www.advancedsciencenews.com www.advenergysustres.com