Prospect for Bismuth/Antimony Chalcohalides‐Based Solar Cells

Inorganic–organic hybrid lead halide perovskites are emerging optoelectronic materials for solar cell application. However, the toxicity concerns and poor stability largely hamper their practical applications. For these reasons, the search for “perovskite‐inspired” alternatives, having the same advantages but overcoming the drawbacks of the lead‐based one, has become an important sector in the field. Among the candidates, Bi3+ and Sb3+ containing materials are of great interest, due to their electronic structures resembling the Pb2+. Bismuth/antimony chalcohalides have been known for a long time as the potential absorber in photovoltaics, even if their performances are still low. Interestingly, pnictogen chalcohalides can be the stepping stone toward numerous quaternary compounds, including some perovskite structures. The understanding of the fundamental properties and the current limitations of both the starting ternary compounds and the final quaternary materials can allow the achievement of improved photovoltaic absorbers, stable, and efficient. In this review, the fundamental properties and device performances of many ternary pnictogen chalcohalides and the derived quaternary compounds are summarized, focusing on the different preparation strategies.


Introduction
Fabrication of stable, efficient, and scalable photovoltaic cells for commercialization has captivated much research interest.[3][4][5] Meanwhile, third-generation solar cells, and in particular Pb-based perovskite solar cells, suffer from fundamental drawbacks for commercialization, including: 1) substantial performance degradation under exposure to moisture, heating, or prolonged illumination in the ambient condition; [6,7] 2) chemical and structural instability on account of uncontrollable crystallization process; [8,9] 3) the presence of lead in the perovskite crystal causing serious pollution in the environment. [10,11]o solve these problems, the research community started exploring "lead-free perovskites" as possible alternatives.It is worth mentioning that when elemental substitution is performed, the resulting material might or might not have the characteristic perovskite structure formed by the network of corner-sharing MX 6 octahedra, even having the classic AMX 3 formula.In this work, we will distinguish between "lead-free perovskites", which present the 3D characteristic perovskite structure, and "perovskite inspired" or "perovskite derivates" for the one that does not present the perovskite crystalline structure.Lead-free perovskites can be obtained by substituting the divalent lead cation with a homovalent ion, such as Sn 2+ and Ge 2+ .These compounds (such as MASnX 3 or CsGeI 3 ) present very similar properties to the lead perovskites: narrow bandgap, and highly delocalized electronic structure that ensure good transport properties.Unfortunately, these materials show even higher instability than the Pb-based perovskite, due to the easy oxidation of Sn 2+ and Ge 2+ to the corresponding 4 + state. [12,13]ead-free perovskites and perovskite derivates can be obtained using also heterovalent ions, such as Bi 3+ , Sb 3+ , Sn 4+ , Ti 4+ , or Ag + .In this way, lead-free perovskites, such as double perovskites (Cs 2 AgBiBr 6 ) [14] or vacancy ordered double perovskites (Cs 2 SnI 6 [15] or Cs 2 TiI 6 [15] ) can be obtained.These compounds present large compositional variability, they show very good air and thermodynamic stability and low toxicity.Unfortunately, the transport properties of these material are generally worse than the lead-based perovskites, and many present indirect bandgaps.As a result, the efficiency of such photovoltaic devices is still limited, and these compounds are mainly finding applications in photodetector and sensors.18] Another possible path to obtain lead-free perovskite is the substitution of both the lead and the halide ion.Examples are the chalcogenide perovskites, such as BaZrS 3 .Despite the high stability and the low toxicity shown by these compounds, photovoltaic devices have not been reported yet due to the high temperature synthesis required by these compounds, which limit their applications. [19,20]inally, a third path is possible: a lead-free chalco-halide perovskite, such as MASbSI 2 .
This type of compounds is the less explored one, and only few compositions have been reported.The steppingstone to obtain this material is the pnictogen chalcohalide compound MChX, as expressed by the following equation: [23] Additionally, such materials have been employed in electronic devices, including room-temperature radiation detection and photovoltaic devices. [24,25]These materials are generally presented with the MChX formula, where M is a pnictogen cation (Sb 3+ , Bi 3+ ), Ch is a chalcogen anion (O 2− , S 2− , or Se 2− ) and X is a halide (Cl − , Br − , or I − ).Importantly, computational studies indicated that the trivalent metal cation presents the 5s 2 electronic configuration as Pb 2+ and can consequently produce photovoltaic materials with a defect-tolerant electronic structure similar to lead halide perovskites. [26]Additionally, metal-chalcogenide bonds have a more covalent character than metal-halide bonds, which generally translates in more stable bonds (e.g., bond dissociation energy of Pb─I, Pb─S, Bi─I, and Bi─S are 197, 398, 186, 315 kJ mol −1 respectively). [27]However, the thermodynamic stability of crystalline compounds is better expressed by the material formation energy than by the bond dissociation energy, which rather refers to molecular systems.Chalcohalides show negative formation energies, indicating that they are more thermodynamically stable than constituent elemental solids. [40]s such, chalcohalides and their derivates have been at the center of a renovated interest and in this work, we review the available literature on the bismuth/antimony chalcohalide materials, focusing on their structural and optoelectronic properties, and analyzing the photovoltaic efficiency-deposition strategy relationship.In addition, we provide an extensive overview of the possible quaternary phases that can be obtained reacting pnictogen chalcohalides with other halogen or chalcogen binary materials.The large variety of species achievable in this way present a broad range of properties and structures, among which leadfree perovskite and perovskite inspired structures can be found.Understanding the properties and challenges of the ternary compounds, and if and how they reflect in the resulting quaternary materials, is essential for the achievement of novel, efficient photovoltaic absorbers.

Crystal Structure
Bi/Sb chalcohalides have different structures depending on the elements used, as shown in Table 1, but the majority crystallizes in an orthorhombic Pnma space group (Figure 1).The structure is one-dimensional and consists of (M 2 Ch 2 X 2 ) n ribbons, resulting from distorted edge-sharing pseudo-octahedra formed by 3 M─Ch, 2 M─X bonds, and a vacant site occupied by the M s 2 lone pair.While along these ribbons covalent bonds are formed, at out-of-chain-axis directions the ribbons have a closed electronic shell and form weak van der Waals bonds to hold the chains together.The open space between the layers makes the materials susceptible to ion exchange. [28,29]or the BiChX type, the common feature of BiSX and BiSeX (X = Cl, Br, I) is the formation of chain-like structures in which Bi atoms are connected with three S or Se atoms and two halogen atoms.However, BiOX (X = F, Cl, Br, I) crystalizes in the tetragonal P4/nmm space group. [30,31]All of them share a unique layered structure, where a positive charge of [Bi 2 O 2 ] 2+ is bonded with halogen ions and Bi atom is surrounded by four O atoms and four halogen atoms.Similarly, SbSX and SbSeX (X = Br, I) are based on a pyramid structure, where five atoms (Sb with two halogens and two S or Se) form the base of a square pyramid with the third S atom on the apex.On the other hand, SbOX (X = F, Cl) has a ladder-like arrangement of Sb and O atoms.However, more irregular or unusual structures can be found, when the stoichiometric ratio is not 1:1:1.For example, the monoclinic Bi 4 O 5 I 2 is formed due to the thermal decomposition of BiOI, where the [Bi 4 O 5 ] 2− layers are separated by I − layers; [32,33] [34] Sb 2 S 19 F 12 and Sb 2 Se 10 F 12 contain S 19 2+ and the boat-shape Se 10 2+ to form a more complex structure. [34,35]he 1D nature of the MChX crystals presents across these mono-dimensional structures, causing anisotropic charge mobility.The carriers will move faster across the ribbon than across multiple ribbons. [36,37]With this condition, it is likely that higher photovoltaic efficiencies will be obtained with a rigorous control of the ribbon orientation, as it happens for the structurally and compositionally similar photovoltaic absorber Sb 2 Se 3 . [38,39]

Bandgap
The bandgap energy (E g ) of a certain semiconductor is the basic prerequisite for its implementation in specific photovoltaic applications.The bandgaps from computational and experimental studies for the twenty-seven bismuth and antimony-based compounds are summarized in (Figure 2a) and more detailed information is shown in Table 1.Through compositional engineering a wide bandgap range (from 1 to 3.5 eV) can be easily obtained, indicating the possibility of bandgap tailoring for the  ) SbSeI and SbSI. [44]Reproduced with permission. [44]Copyright 2018, The Royal Society of Chemistry.optimization of solar cells.With the replacement of the chalcogen atom with the heavier element (from oxygen to selenium), the atomic radius is larger and a clear decreasing trend in E g can be observed for both bismuth and antimony-based materials, with the compounds containing oxygen presenting bandgaps over 3 eV.It's worth noting that most of these chalcogenide materials have indirect bandgap, which is always smaller than the calculated direct bandgap. [40]However, calculations show that the indirect and direct bandgap are promisingly closed, only 40 meV in the case of BiSeBr and BiSI. [40]On the other hand, there have been demonstrated that, 1D SbSeI experiences an indirect-todirect band gap transition under a small tensile strain. [41]Only a few bismuth-based compounds are found with E g between 1.1 and 1.6 eV, which is preferred for single-junction solar cells.For example, Ganose et al. found that two candidate materials, BiSI and BiSeI, possess suitable electronic structures, with a bandgap of 1.57 and 1.29 eV respectively.Additionally, low electron effective masses, and anti-bonding states at the top of the valence band were observed, suggesting an excellent signal for higher defect tolerance. [42]Similarly, they assumed that antibonding states contain a substantial cation s contribution mixed with unoccupied cation p states at the top of the valence band, which promoted the asymmetric electron density and induced the distortion of the pseudo-octahedra responsible for a stabilized effect. [43]The antimony compounds show larger bandgap than the corresponding bismuth compounds, but overall show the same trends: bandgap reduces using heavier chalcogens as well as using heavier halides (SbSeI = 1.86 eV, SbSI = 2.11 eV, SbSBr = 2.31 eV). [26]Obviously, it is worth studying how to provide a strategy for further red shift of Sb-based chalcohalide with such a high bandgap.
The ionization potential and electron affinity have been calculated for several compositions, including SbSI, Sb-SBr, SbSeI, BiSI, and BiSeI.The electron affinity has values ranging from 5.2 to 6.4 eV, while the ionization potentials lie between 3.15 and 5.0 eV. [26,44]As it will be discussed below, such deep valence and conduction bands pose challenges in finding suitable hole/electron transport layers.

Effective Mass
[47][48] As shown in (Figure 2b), the electron or hole effective masses of most chalcohalide materials are less than the rest mass of electrons, allowing good carrier mobility.It is worth mentioning that low hole effective mass is also a good signal as a candidate for transparent conducting materials.Ran et al. calculated the effective masses and electronegativity of 36 compounds and found that the majority of the materials have higher hole effective masses than the electron one.Some bismuth-based compounds (BiSeCl, BiSeBr, BiSeI, and Bi 3 Se 4 Br) and antimonybased (SbSI and SbSeI) present both suitable bandgap and effective mass lower than the mass of electrons, resulting promising for several applications beyond photovoltaics. [26,40]The ternary antimony chalcohalides show an 1D structure as same as the bismuth-based ones, [28] but the distinguished photoferroicity of SbSI allows potential interest for anomalous photovoltaic applications, in which it shows a higher theoretical open circuit voltage (V OC ) than other semiconductors. [49]

Electrical Properties
Compared to other physical and structural properties, less attention has been paid to the electrochemical and electrical characterization of chalcohalide materials.In 1995, Park et al made SbSI, SbSeI, BiSI, and BiSeI single crystals and found the relationship between the temperature dependence of the optical energy gap at the phase transition points and the Varshni equation. [50,51]Furthermore, some investigations showed that the electrical resistivity of BiSBr, in the range from 10 3 to 10 4 Ω cm, is much lower than the one of BiSI. [52,53]On the other hand, some reports about the dielectric properties of SbSI showed the dielectric constanttemperature dependence. [49]Based on the differences in composition, SbSI glasses exhibited a change in conductivity of few orders of magnitude and slight modification of the carrier mobilities, in the range from 50 to 100 cm 2 V −1 s −1 . [49,54]Besides,  [24] Copyright 2017, Wiley-VCH.Reproduced under terms of the CC-BY license. [82]Copyright 2021, published by Multidisciplinary Digital Publishing Institute.Reproduced with permission. [23]Copyright 2019, Wiley-VCH.
Pikka et al. demonstrated that the electrical conductivity of Sb-SeBr was similar to the reported for SbSeI, which was around 10 −7 Ω −1 cm −1 . [55]Audzijonis et al. reported that SbSeI could be displayed as antiferroelectric and paraelectric phases with the change of temperature. [56]

Deposition Methods and Devices of Bismuth/Antimony-Based Chalcohalides
As mentioned previously, due to the 1D nature of these materials, the thin film morphology and the device performances are intimately linked, and the control of the first is essential for the latter.Additionally, the quaternary materials deriving from the ternary chalcohalides are often achieved through a conversion of a pre-deposited chalcohalide thin film, with the final morphology deeply affected by the starting one.For this reason, in this section, we will introduce the preparation methods of the chalcohalides films for solar cell devices focusing on the structure and morphology of chalcohalides material, besides discussing the performances of the corresponding solar cells.It is worth noticing that we will focus only on the reports in which chalcohalide materials are the light absorbers, not reporting the ones that use these materials as transport layer.

Antimony-Based Chalcohalides
Some antimony-based ternary chalcogenide materials combined with Br, I, S, Se, and Te have been experimentally studied since the 20 th century.][73][74][75] Even though SbSI has excellent optoelectronic properties, reports of SbSI as a light absorbing material have only appeared in recent years.To the best of our knowledge, the solar cells fabricated by utilizing SbSI as a light harvester were reported by Nie et al. [68] They used a three-step sequential process for depositing SbSI layers.The fabrication procedure (shown in Figure 3a) consists in spincoating SbI 3 (dissolved in DMF) onto the Sb 2 S 3 layer, which was deposited by the chemical bath deposition (CBD) method.The SbSI layer was formed by thermal annealing in Ar or N 2 gas at 150 °C for 5 min.Since SbSI was combined with the mesoporous TiO 2 (mp-TiO 2 ), the morphology of SbSI/mp-TiO 2 was similar to that of bare mp-TiO (PCPDTBT) exhibited the maximum PCE of 3.05%.The authors explained that PCPDTBT can absorb light that SbSI cannot, creating additional electron-hole pairs.The devices without encapsulation were stored in different humidity environments and kept more than 93% of their initial PCE after 15 days, demonstrating the stability of this material.
In 2020, the same group improved the fabrication process through a vapor-process (VP) treatment. [76]SbSI was formed by the reaction of the Sb 2 S 3 (deposited on mp-TiO 2 substrate) with the vapors of SbI 3 carried by argon flow.The SbSI-based solar cells and SbSI-interlayered Sb 2 S 3 solar cells were prepared by adjusting the reaction temperature and time.Compared with the solution-processed (SP) SbSI, the charge transfer of VP SbSI was improved by producing a more uniform SbSI layer with better contact with the HTL which shortened the path length from SbSI to the HTL.This process led to a record PCE of 3.62% in pure SbSI-based solar cells.In the SbSI-interlayered Sb 2 S 3 solar cells, an energetically favorable external driving force was provided by forming the TiO 2 /Sb 2 S 3 /SbSI/HTL structure, which promoted the charge transfer among each layer, and achieved an impressive PCE of 6.08%.
Around the same time as the first reported SbSI solar cell, Choi also reported a simple solution processing approach based on the SbCl 3 -thiourea (TU) molecular solution to overcome the limitations of the CBD method for preparing Sb 2 S 3 . [77]The amorphous Sb 2 S 3 (am-Sb 2 S 3 ) film was deposited on a TiO 2 blocking layer (TiO 2 -BL) substrate by spin coating SbCl 3 -TU solutions and annealing it.Then, the SbI 3 solution was spin-coated and annealed for conversion into the SbSI film.The authors mentioned that the amount of TU, the am-Sb 2 S 3 island shape, and the solvent used for the SbI 3 dissolution were three key factors in the formation of dense SbSI films under high TU conditions.It can be observed that due to the nucleation of Sb 2 S 3 , the film is more compact and flatter than with the previously reported SP, and the grain size is relatively large.However, due to a poor TiO 2 coverage of the 1D crystals achieved with this method, they used CBD method to achieve tightly packed grains.By optimizing the solution system and preparation process, the authors prepared high crystallinity and uniform SbSI films, and the device exhibits an efficiency of 0.93%.This value, lower than with mesoporous TiO 2 devices, is likely caused by the poorer charge extraction in planar structures than in bulk heterojunctions.
It is interesting to note that not only the differences in the preparation process lead to dramatic changes in the film morphology, but also the compositional variation can induce such changes.Partial incorporation of Se into SbSI deeply affected the morphology and optoelectronic properties of the obtained films.For example, Jung et al. fabricated various Sb(S 1−x Se x )I film (0 ≤ x ≤ 1) by varying the S/Se molar ratio, obtaining bandgaps from 2.05 to 1.69 eV. [78]Similarly to the previous two-step solution method, [76] they controlled the S/Se molar ratio by adjusting the molar concentration of the two solutions of SbCl 3 -TU and SbCl 3 -SeU (selenourea) in the formation process of Sb 2 (S,Se) 3 .As the molar concentration of Se in the solution increased, the Sb(S,Se)I nanorods became larger and longer.This morphological change is explained by the different thermal stability between SbCl 3 -TU and SbCl 3 -SeU, as the SeU groups are more thermally stable than the TU groups.Moreover, volatilization of the as-formed Sb 2 (S,Se) 3 decreases with increasing Se concentration, creating relatively dense films under selenium-rich conditions.Because of the undesirable conduction band offset of the electron-transporting layer (ETL) and the hole-transporting layer, the Sb(S,Se)I solar cells exhibit very poor device performance, but this work demonstrates an effective method to tune the morphology of SbSI films.
A ferroelectric-photovoltaic effect in nanowires of SbSI was presented by Mistewicz et al. for the first time, [79] and other antimony-based chalcohalides such as SbSeX (X = Br, I) and Sb(Se,Te)I, have been proposed to be promising for ferroelectric photovoltaic applications.With a few microns thick, they could theoretically reach maximum efficiency over 28%. [80,81]Similar to the two-step fabrication process of SbSI, the SbSeI solar cells were fabricated for the first time using multiple spin-coating cycles of a SbI 3 solution on Sb 2 Se 3 by Nie et al. [82] It is worth mentioning that the Sb 2 Se 3 thin film was formed by thermal decomposition after depositing [(SbL 2 Cl 2 )Cl] 2 •(CH 3 ) 2 CO], where L is N,N-dimethyl selenourea.Due to insufficient absorption and inefficient charge transfer of the obtained film, both the films were fabricated by 8 and 12 repetitive spin-coating cycles, but exhibited poor PCE.The best SbSeI device, with a bandgap of 1.67 eV and a film thickness of around 1 μm, was achieved through ten spin-coating cycles, and the cells with mp-TiO 2 and PCPDTBT exhibited a PCE of 4.10%, besides demonstrating excellent humidity, thermal, and photostability.Balakrishnan et al. carried out the transformation of Sb 2 Se 3 to SbSeI through two different synthetic routes: metal halide solution casting and halogen vapor exposure. [29]Through mechanistic studies of the products and intermediates, they elucidate the chemical reaction nature, and reveal common features of both synthetic methods.In the solution casting route, the SEM images of partially transformed SbSeI rods show that the SbI 3 precursor causes the tip of the Sb 2 Se 3 rod to split in an "unzipping" process along the long axis to form shorter and thinner SbSeI nanorods.In the halogen vapor exposure route, these results indicate that the iodination of the Sb 2 Se 3 films starts converting the Sb 2 Se 3 into SbI 3 at room temperature, which then diffuses and converts the Sb 2 Se 3 into SbSeI through the "unzipping" process when the temperature is applied.The SbSeI and Sb 2 Se 3 nanorods have a consistent 1D morphology and crystal orientation, also indicating the presence of a solid-state "topotactic" transition, that is, a reaction in which the crystalline structure of the reactant is preserved in the product.In 2021, Choi et al. proposed a much simpler onestep solution-phase precursor engineering method for SbSeI film fabrication. [83]The precursor solutions were synthesized by mixing Sol A (formed by a mixture of SbCl 3 and SeU) and Sol B (containing SbI 3 ) at different molar ratios (Figure 3b).By modifying the solvent and annealing parameters, they produced dense and coarse crystals on the CdS/FTO substrate.The whole process including solution preparation, spin coating, and annealing, greatly simplified the solution or gas phase two-step methods, suggesting that the proposed strategy can be readily applied to the fabrication of various chalcohalides.However, due to the uneven crystal morphology, the device exhibited a very poor efficiency of 0.23%.
Bi x Sb 1−x SeI systems have also been reported and investigated for solar cells.Nie et al. found that the substitution of antimony for bismuth within individual crystals was homogeneous.In fact, the Raman spectroscopy showed a two-mode behavior after the substitution, indicating the presence of covalent bonds within the structure, and TEM/SEM data confirmed the absence of nanoclustering or segregation within the crystals.Bi 1−x Sb x SI solid solutions showed that the original 1D chain structure was maintained while the bandgap could be easily tuned in the range from 1.55 to 1.86 eV (BiSI→SbSI). [84,85]Such a way of controlling the bandgap and transport properties was proven to be effective in fabricating solar cells based on the Sb 0.67 Bi 0.33 SI compound. [23]n this work, the authors deposited Sb 2 S 3 films by CBD and annealed the films at 250 °C under argon or nitrogen, then spincoated BiI 3 (dissolved in DMF) and annealed for two minutes at 250 °C to obtain Sb 0.67 Bi 0.33 SI films.The CBM and VBM of Sb 0.67 Bi 0.33 SI were determined as 4.09 and 5.71 eV.As can be seen in (Figure 3c), the crystal morphology is consistent with mp-TiO 2 in both Sb 2 S 3 and Sb 0.67 Bi 0.33 SI.By optimizing Sb 2 S 3 for different CBD times, and spin-coating different concentrations of BiI 3 , the best performing Sb 0.67 Bi 0.33 SI-based solar cell exhibited an impressive efficiency of 4.07% under standard AM 1.5 test conditions, and the unencapsulated cell showed good overall stability.This work demonstrated that the active layer can be deposited in the pores of the electrode by controlling the CBD process time, leading to sufficient infiltration of the hole conducting organic polymer into the bottom of the electrode, which greatly improves the short-circuit current density (J SC ) leading to higher PCE.
These studies have demonstrated the variety of methods available for the preparation of antimony-based chalcohalide compounds, and more attention is being paid to this material.So far, the most efficient device structures are obtained with mesoporous TiO 2 , indicating that the transport properties (or shorth carrier lifetimes) are limiting the performances of these materials.For this same reason, planar architectures still lag behind in terms of efficiency, and a more careful control over the crystal orientation, film morphology, and transport material selection needs more in-depth investigation.In this perspective, we believe that inspiration could be taken from materials with similar crystalline structure, such as Sb 2 Se 3 . Similar synthetic strategies (such as rapid thermal evaporation of the absorber powders [88][89][90][91] ) could be applied to the pnictogen chalcohalides; alternatively, the topotactic conversion of the correctly oriented Sb 2 Se 3 and Sb 2 S 3 thin films could be developed.

Bismuth-Based Chalcohalides
If the M site in MChX is completely occupied with bismuth, bismuth-based chalcohalide is obtained.This type of material has been studied for a long time.In 1965, Horák synthesized BiSI and investigated its spectral distribution of the internal photo-effect, showing that the maximum of the photo-current lies in the region of 785 nm. [92]BiSI and BiSeI both possess the same structure as SbSI with the space group of Pnma and the one-dimensional crystal growth, and both are n-type materials. [93]These two materials also possess suitable electronic properties for solar energy applications including high absorption coefficients and indirect bandgaps of 1.57 and 1.29 eV. [94]It should be noted that the difference between the direct and indirect bandgap of these two materials is small, and the effective mass of electrons along the BiChI chain direction is low.Furthermore, the presence of antibonding states at the top of the valence band, coupled with a high dielectric constant, is expected to provide an ideal highlevel defect tolerance for high-efficiency solar absorbers. [40,95]Using defect analysis, Ganose et al. revealed that both BiSeI and BiSI compounds are suited for use in p-i-n junction devices, with a predicted large spectroscopically limited maximum efficiency of 25.0% and 22.5% for BiSeI and BiSI based solar cells, respectively, competitive with the best current generation photovoltaic absorbers.However, the electronic and structural features of BiSeI and BiSI can lead to the presence of deep trap states, which is the main reason for the prior poor performance of these materials. [96]n terms of material synthesis, several studies have also reported the preparation of BiSI through solution process methods.In a typical synthesis process reported by Tiwari et al., stoichiometric BiSI films could be deposited by a single precursor solution prepared by dissolving Bi(NO 3 ) 3 •5H 2 O, TU, and NH 4 I, in the 2-methoxyethanol and acetylacetone solvent mixture. [97]iSI films were obtained after spin-coating in air and annealing at 200 °C for 5 min.The compact and homogeneous films, which displayed an optical bandgap of 1.57 eV, were composed of flake-shaped grains with sharp edges and sizes between 600 and 700 nm, oriented anti-planarly to the substrate (Figure 4a).Electrochemical impedance spectroscopy revealed n-type doping with valence and conduction band edges located at 4.6 and 6.2 eV below the vacuum level, respectively.Their BiSI solar cells with planar device structure glass/FTO/SnO 2 /BiSI/F8/Au showed a record conversion efficiency of 1.32% under AM 1.5 illumination.Sugathan et al. developed a simple one-step solution-processing route to synthesize phase-pure BiSI polycrystalline thin films. [98]hey employed BiCl 3 , NaI, and TU as the precursor materials to overcome the presence of impurity phases and the formation of BiI 3 -TU complex, which is stable and quite difficult to convert into BiSI via thermal annealing.The large electronegativity difference between Na and Cl contributes to the dissociation of the Bi─Cl bond in BiCl 3 , facilitating the reaction between sulfide and iodide and promoting the formation of BiSI.Choi and Hwang developed a two-step solution method and applied it to the fabrication of BiSI films. [99]Bi 2 S 3 was first prepared by spin-coating a mixture of Bi 2 O 3 and TU followed by the spin-coating of a BiI 3 solution to obtain BiSI films (Figure 4b).The structure, absorption, and morphology were controlled by tuning the Bi:S molar ratio of the Bi 2 O 3 -TU solution and the number of repetitions of Bi 2 S 3 depositions.Through the optimization of the experimental process, the best BiSI crystallinity was obtained with the molar ratio of Bi:S = 1:3 and with two repetitions.The morphology of as-fabricated BiSI film exhibited nanorods rather than a compact film with an optical bandgap of 1.61 eV (Figure 4c).They tried the complete device structure but did not report the efficiency.
Frutos et al. investigated the formation mechanism of BiSI nanocrystals by solution and hydrothermal methods. [100]In this chemical reaction, there is a synergistic effect among the three components involved in the synthesis of BiSI, the iodine monomers act as reactants, the crystal structure of Bi 2 S 3 acts as a self-sacrificing template, and the reaction medium, monoethylene glycol, is a coordinating solvent that plays an important role in facilitating the entry of iodine atoms into the Bi 2 S 3 structure to form BiSI. Xiong et al. developed a solvothermal method that took advantage of the 1D properties of BiSI materials to design vertically oriented BiSI nanostructures. [101]In this work, they deposited precursors of Bi 2 S 3 nanorod arrays on tungsten (W) foil, which was then immersed in an autoclave containing a solution of BiI 3 and oleic acid (OA) and heated to obtain BiSI nanorods.The synthesized BiSI nanorods show strong broadband absorption covering the entire visible region and have an indirect bandgap energy of 1.57 eV.The structure of the device is ITO/p-CuSCN/n-BiSI/W, with all the active layer nanorods vertically aligned with good contact both with the p-CuSCN and with W. This 1D nanorod array structure provides a high surface area Reproduced with permission. [96]Copyright 2019, American Chemical Society.Reproduced under terms of the CC-BY license. [99]opyright 2019, published by Multidisciplinary Digital Publishing Institute.Reproduced under terms of the CC-BY license. [99]Copyright 2019, published by Multidisciplinary Digital Publishing Institute.
at the p-n interface and significantly reduces the distance travelled by the electron-hole pairs, thus reducing losses.The assembled device exhibits a J SC of 2.73 mA cm −2 and a V OC of 0.46 V, yielding a fill factor (FF) of 52.81% and a PCE of 0.66%.The limited performances achieved are probably due to the substantial band alignment offset between absorber, transport layers, and electrodes; additionally, the large surface area at the p/n junction is likely to present trap states that hinder the voltage and current generation.
Besides, spray pyrolysis can also be used to synthesize BiSI.For example, Hahn et al. reported the preparation of polycrystalline BiSI thin films via a single-source chemical spray pyrolysis method by optimizing the deposition temperatures.The obtained films exhibited promising photoelectrochemical properties on both metal foils and FTO glass slides. [96]More specifically, the films displayed microrod morphologies consistent with the microscopic crystal structure of BiSI.The strong light absorption in the visible range and the well-crystallized layered structure enabled their excellent photoelectrochemical performance through improved electron-hole generation and separation.Further, they studied the effects of selenium doping on the optoelectronic properties of the BiSI film. [70]They turned the bandgaps of BiS 1−x Se x I by substituting selenium with sulfur through the optimization of the amounts of SeO 2 and TU in the BiSI spray pyrolysis precursor solutions.With higher Se-doping levels, the film changes from large micrometer-scale rods to smaller, cubelike structures, and decreased their apparent direct bandgap from 1.63 to 1.48 eV.The limited PCE (up to 0.25% and 0.012%) were measured for dye-sensitized solar cells (DSSC) and planar structure solar cells, respectively.Manipulation of these films' optical and electronic properties by doping remains to be further explored and may provide a possible strategy to optimize the optoelectronic and morphological properties of these materials.
Phase-pure BiSI and BiSeI produced with the ball milling method were first reported by Murtaza et al.They demonstrated the feasibility of synthesizing large quantities of phase-pure chalcohalides by mechanochemical synthesis. [102]This method, which happens at room temperature, avoids deviations from stoichiometry due to the volatilization of halides that can occur during hightemperature synthesis.The samples produced by ball milling exhibited optical properties comparable to those of materials prepared using high-temperature routes.
In addition to the ball milling method, Kunioku et al. developed a new, facile method for synthesizing BiSI and BiSeI chalcohalides nanoparticles at a significantly lower temperature than the previously reported processes. [103]In this study, BiSI and BiSeI particles were readily synthesized by simply heating BiOI particles under H 2 S or H 2 Se gas at low temperatures and short time via the substitution of anions from O 2− to S 2− (or Se 2− ).BiSBr 1−x I x solid solution was also synthesized by this lowtemperature anion substitution, obtaining continuous tailoring of bandgaps achieved by controlling the ratio between Br − and I − .Attributed to their n-type nature, all these chalcohalides loaded on FTO clearly exhibited an anodic photocurrent in an acetonitrile solution containing I − .
In addition to BiSI, other bismuth chalcohalide compositions have also been developed.Quarta et al. presented a colloidal approach to prepare orthorhombic bismuth chalcohalide nanocrystals (NCs) via the hot-co-injection of both the chalcogen and the halogen precursors to a solution of Bi-carboxylate complexes in a non-coordinating solvent. [104]Following this approach, they have synthesized colloidal BiSCl, BiSBr, BiSI, and BiSeBr NCs.What's more, the BiChX NCs can be processed from the solution phase at ambient conditions without significant degradation and showed composition-dependent bandgaps and large optical absorption coefficients across the visible spectral range.The  [105] Copyright 2021, Wiley.Reproduced with permission. [105]Copyright 2021, Wiley-VCH.fabricated photoelectrodes based on BiChX NCs can harvest the sunlight and generate an electric current with an internal quantum efficiency (EQE) above 10% across the entire visible spectral range.These findings set the colloidal BiChX NCs as promising solar absorbers for photocatalytic and optoelectronic applications.
Among these BiChX nanocrystals, only a limited number of studies reported on BiSCl, because BiSCl is not easy to be synthesized and the sample is often prepared using high-temperature processes.The complex preparation has limited the experimental studies and applications of BiSCl, and its physicochemical properties have barely been discovered.However, BiSCl has been reported for use in solar cells.by Li et al. (Figure 5a).BiSCl singlecrystalline nanofibers were synthesized by a one-pot solvothermal approach and exhibited strong light absorption in the visible, with a bandgap of 1.96 eV. [105]UPS and DFT revealed that BiSCl is a direct n-type semiconductor with the valence band maximum and conduction band minimum located at 6.04 and 4.08 eV below the vacuum level, respectively.As the reaction time increases, the fine BiSCl nanofibers grow into longer BiSCl fibers, as can be seen in Figure 5b.The BiSCl fibers prepared by solvothermal treatment for 12 h present a particle size of about 200 nm and a length of about 20 μm.The corresponding solar cells, with the BiSCl nanorods oriented vertically to the surface of a TiO 2 porous film, exhibit a PCE of 1.36% and a relatively large J SC of 9.87 mA cm −2 , revealing the potential of the BiSCl nanorod array light absorber for a new type of solar cell.
BiOX has also been largely investigated.By analyzing the energy band structure, the bandgap of P4/nmm and Cmcm phases of BiOF are 2.74 and 2.47 eV, respectively. [106]The experimental bandgaps of BiOCl, BiOBr, and BiOI crystals are 3.32, 2.85, and 1.83 eV, [70,54,59] while their corresponding calculated indirect bandgaps are located at 2.50, 2.10, and 1.59 eV, respectively. [107]he BiOX compounds have high photocatalytic activity.The bismuth, as a plasma cocatalyst in photocatalytic reactions, can be electron/hole trappers, charge transfer mediators, or oxygen vacancy coordinators.110][111][112][113] Among the BiOX materials, BiOI has been used in dyesensitized solar cells.Zhao et al. embedded hierarchical BiOI nanoplate microspheres in chitosan matrix, and obtained limited V OC and J SC . [114]In another report on BiOI dye-sensitized solar cells, Sfaelou et al. prepared solar cell composed of a BiOI/TiO 2 /FTO photoanode, a Pt/FTO cathode, and the I 3 − /I − liquid electrolyte.In this work it was shown that BiOI nanoflakes deposited by 5 and 15 SILAR (Successive Ionic Layer Adsorption and Reaction) cycles form a layer of about 600 nm, [115] but was characterized by isolated clusters of BiOI with consequent reduction of the conductivity of the film.The J SC was 3.8 mA cm −2 , far below the upper theoretical limit of 11 mA cm −2 .The V OC was 0.61 V and the FF was 45%, giving an overall efficiency PCE of 1.03%.Ahmmad reported the successful application of BiOI using in a three-phase solid-state solar cell (SSSC) for the first time. [116]Under visible light irradiation, photocurrent and photovoltage were observed for TiO 2 -BiOI-based SSSCs, even if only 0.1% efficiency was obtained.Later, Hoye et al. developed an allinorganic BiOI planar solar cell comprised of NiO x hole transport layer and ZnO electron transport layer. [117]Their highestperforming device had an integrated J SC of 7.3 mA cm −2 , almost twice the highest efficiency previously reported for BiOI.An important reason for the high performance of their devices is the densely packed morphology of the NiO x and BiOI layers, which grew by chemical vapor transport (CVT) inside a two-zone horizontal tube furnace, as well as the conformal AP-CVD ZnO layer.The PCE of BiOI devices is about 1.8%, and showed improvement in stability over perovskite solar cells.
Currently, BiSI is the main bismuth-based chalcohalide material used as the light absorbing material in solar cells and there have been many studies on the synthesis and preparation of BiSI.As shown in Figure 2a, the simulated and experimental bandgaps of chalcohalides, in addition to BiSI, bismuth-based chalcohalide like BiSeI, BiSeBr, and BiSeCl also have suitable bandgaps for use as light absorbing materials.So, further investigations on these materials are highly desirable to reveal their fundamental properties.; each represents the orthorhombic space group Pnma and the monoclinic space group C2/m, respectively.Reproduced with permission. [120]Copyright 2022, Wiley-VCH.Reproduced with permission. [124]Copyright 2018, American Chemical Society.Reproduced with permission. [125]Copyright 2021, The Royal Society of Chemistry.Reproduced with permission. [131]Copyright 2006, American Chemical Society.
Overall, the analysis of the literature reporting pnictogen chalco-halide solar cells has revealed several issues that needs to be overcome in order to achieve higher performances in photovoltaic devices and beyond.These materials present highly anisotropic properties, in relation to their 1D nature.A rigorous control of the crystal orientation in the device is therefore essential to avoid intrinsic limitation in the carrier generation, movement, and extraction.Moreover, the rod and needle like crystals generally formed by these materials pose serious challenges in terms of device preparation, such us partial covering of the transport layers, risk of short circuit and large surface/volume ratio, which increases the chances of surface trap states.The study and characterization of defects in chalcohalides is still partial, and generally dedicated to the most commonly studied composition (such as BiSI).Another challenge to overcome, is the identification of suitable transport layers presenting good energy alignment with these absorber materials.The chalcohalides have deep valence bands that does not align with the most common hole transport layers.As a consequence, poor FF and severe V OC losses are still present in the solar cells.
Material doping could represent a way to optimize the chalcohalides materials, but it has been scarcely reported.Mixed chalcogen and mixed halide compositions can provide a path for bandgap and morphological tailoring, as discussed previously.On the other hand, extrinsic doping has been much less explored and to the best of our knowledge, only few reports are present: i) Ni and V doping of SbSI, SbSeI, BiSI, BiSeI which cause a general reduction of the bandgap (excluding SbSI for Ni and BiSI for V, that showed an enlargement of the bandgap). [118]ii) Co-doping of single crystals of BiSBr and SbSBr, that lead to a reduction in the bandgap. [119]

From Metal Chalcohalide to Other Quaternary Materials
Typically, when we talk about chalcohalides we only refer to the 1D MChX materials.All the other structures obtained by adding metal chalcogenide (ACh) or metal halide (AX) binary compounds to the MChX do not show 1D structures anymore, but develop 2D and 3D crystals, among which some are perovskite structures.These materials have been less studied than the heavy pnictogen chalcohalides and most of the research focuses on the characterization of their properties.However, some recent reports have shown that these materials are suitable for photovoltaic applications and have been successfully used in the preparation of solar cells.

AMChX 2 Type Structures
These materials can be obtained through the reaction of the chalcohalides with a metal halide, as described in Equation (1-1).Remembering that the structural, optical, and electric properties of these materials are strongly dependent on the specific composition, we will discuss the AMChX 2 properties referring to CuBiSCl 2 as a representative compound of this family.CuBiSCl 2 is present an orthorhombic distorted perovskite structure with Cmcm space group.The structure of CuBiSCl 2 is shown in (Figure 6a).Cu cations sit at the centers of the dual-anion octahedra, and are bonded to two equivalent S 2− and four equivalent Cl − atoms to form a mixture of corner and edge-sharing CuS 2 Cl 4 octahedra.
Ming et al. combined computational screening and experimental synthesis to explore stable and defect-tolerant lead-free photovoltaic materials based on hybrid mixed chalcogenidehalide compounds.Their results indicated that CuBiSCl 2 is the most promising material with an experimental bandgap of 1.44 eV. [120]The unique energy band feature combined with the large dielectric constant of CuBiSCl 2 produced a high defect tolerance, besides being stable at temperatures of up to 300 °C and resistant at room temperature in air at 60% relative humidity for 25 days.They further fabricated solar cells using CuBiSCl 2 as the light absorber and, without optimizing the fabrication parameters, a PCE of 1.00% and a V OC of 1.09 V were obtained, showing that CuBiSCl 2 is a promising and stable lead-free material for photovoltaic applications.Ruck et al. provided a detail discussion of the synthesis, crystal structure, and electronic band structure of AgBiSCl 2 , which has the exact same crystal structure as CuBiSCl 2 . [121]ecently, research has been conducted on the introduction of MA + , which is largely used as organic cations in lead-based perovskite solar cells, into this type of perovskite structure.Nie et al. reported that the tolerance factor of MASbSI 2 is 0.99, close to that (1.0) of an ideal cubic perovskite structure. [122]They used a multi-step reaction method to prepare the films: Sb 2 S 3 film was first deposited by CBD method and was converted to SbSI through spin-coating of SbI 3 solution.Finally, MAI solution was spin-coated and thermally annealed to form MASbSI 2 .From the Tauc plot, the bandgap of MASbSI 2 was estimated to be 2.03 eV.The best-performing MASbI 2 -based solar cell exhibited PCE as high as 3.08%, which retained 90% after being stored in the dark ambient conditions for 15 days.This is the first report on the fabrication of solar cells using MASbSI 2 as light harvester, suggesting that it could be a viable strategy for the fabrication of stable lead-free solar cells.Besides, by using hybrid functional calculations including spin-orbit coupling, Sun et al. showed that MABiSeI 2 and MABiSI 2 possess a direct bandgap of 1.3-1.4eV, ideal for single-junction solar cells. [64]ASbSI 2 material has also attracted the interest of other researchers.Zhao et al. have demonstrated hetero-anionic perovskite MASbSI 2 with stable out-of-plane ferroelectricity and unique anion order as a long-sought ferroelectric organic-inorganic perovskite material.However, they mentioned that secondary-phases are found in experimental MASbSI 2 samples. [123]It has also been reported that a combination of experimental validation and computational analysis shows a lack of formation of the mixed sulfide and iodide perovskite phase MASbSI 2 in favor of t a mixture of binary and ternary compounds (Sb 2 S 3 and MA 3 Sb 2 I 9 ) (Figure 6b). [124]Density functional theory calculations also indicate that the quaternary MASbSI 2 perovskite phase should be less thermodynamically stable compared with binary/ternary anion-segregated secondary phases and less likely to be synthesized under equilibrium conditions.

A 2 MCh 2 X 3
Through the reaction of pnictogen metal halides with divalent metal chalcogenide compounds, quaternary compositions with 2D structure can be obtained.as schematically indicated by Equation (4-1): Kavanagh et al. employed a combination of methods to probe the static and dynamic crystal structure of Sn 2 SbS 2 I 3 . [125]n 2 SbS 2 I 3 crystallizes in the orthorhombic Cmcm space group (Figure 6c), which comprises infinite chains of (Sn 2 S 2 I 2 ) n tightlypacked perpendicularly to the chain length which forms 2D layers.Multiple of these layers can stack one on top of the other, with antimony and iodine atoms located between these layers.In this material, Sn 2+ is bonded in a 5-coordinate geometry to three equivalent S 2− and two equivalent I − atoms.Sb 3+ is bonded in an 8-coordinate geometry to two equivalent S 2− and six I − atoms.S 2− is bonded to three equivalent Sn 2+ and one Sb 3+ atom to form a mixture of distorted edge and corner-sharing SSn 3 Sb tetrahedra.This (Sn 2 S 2 I 2 ) n structural motif matches the 1D chain structures of the MChX chalcohalide family.Further, they assessed the radiative efficiency limit of this material, finding that a maximum efficiency of >30% can be obtained with film thicker than 0.5 μm.In a major breakthrough, Nie et al. used Sn 2 SbS 2 I 3 as the absorber obtaining a PCE exceeding 4%.[126] To prepare it, they spin-coated a solution of all the precursors (SbCl 3 , TU, and SnI 2 ) in DMF, and annealed the obtained film at 300 °C in Ar atmosphere.All processes were carried out in the glove box, as the Sn 2+ present in Sn 2 SbS 2 I 3 easily oxidized into Sn 4+ , presenting the same problem found in Sn-based perovskite solar cells.[127] The bandgap of Sn 2 SbS 2 I 3 is 1.41 eV, and the optimized Sn 2 SbS 2 I 3 solar cell with mesoporous TiO 2 shows a PCE of 4.04%, accompanied by a J SC of 16.1 mA cm −2 , a V OC of 0.44 V, and an FF of 57.0%.The unencapsulated devices retained 81.0% of the initial PCE after 600 h of exposure stored under ambient conditions at 80% relative humidity at room temperature in the dark.
Lead is an alternative to replace tin and Sn 2 SbS 2 I 3 ; the resulting Pb 2 SbS 2 I 3 has the same crystal structure as the Sn counterpart at room temperature. [128]In 2018, Nie et al. prepared a Pb 2 SbS 2 I 3 absorber layer for solar cells by spin-coating a solution of PbI 2 on a CBD-deposited Sb 2 S 3 , followed by thermal annealing. [129]The morphologies of all Pb 2 SbS 2 I 3 are similar to that of Sb 2 S 3 , indicating that spin-coating of the PbI 2 solution does not change the morphology and the distribution of Pb 2 SbS 2 I 3 on the mp-TiO 2 is uniform.The best-performing solar cells based on Pb 2 SbS 2 I 3 exhibited a PCE of 3.12% and showed good humidity stability over 30 days without encapsulation.Interestingly, the bismuth counterparts Pb 2 BiS 2 I 3 and Sn 2 BiS 2 I 3 have not been used in solar cells.However, Islam et al. reported that Pb 2 BiS 2 I 3 and Sn 2 BiS 2 I 3 crystallize in the Cmcm space group and exhibit direct bandgaps of 1.60 and 1.22 eV respectively. [130]learly, the use of Pb 2+ and Sn 2+ implies toxicity and air stability issues, posing the same drawbacks currently present for the far more efficient lead-and tin-based perovskites.

AMCh 2 X Type Structures
This class of MChX derivatives can be obtained by adding metal chalcogenides compounds (such as CdS or MnSe), as expressed in Equation (4-2): These materials form two structure types depending upon the combination of chalcogenide and halide anions. [131]As represented in Figure 6d through CdBiS 2 Cl, AMCh 2 X crystallizes in the orthorhombic space group Pnma when both anions belong to the same period, that is, Ch 2− /X − = S 2− /Cl − and Se 2− /Br − .With these conditions, layered-type structures are formed, with Cd 2+ cation bonded to four S 2− and two Cl − anions to form CdS 4 Cl 2 octahedra.Each slab is made of an extended array of the corner and edge-sharing Cd-centered octahedra, alternated with BiS 5 square pyramids, recalling the Ruddlesden-Popper structures of several perovskite compositions.Materials belonging to this crystal structure are: CdSbS 2 Cl, CdBiS 2 Cl, CdBiSe 2 Br, [131] MnBiS 2 Cl, [132] MnSbS 2 Cl, [133] MnBiSe 2 Br, MnSbSe 2 Br. [134]On the other hand, when the chalcogen and halide anions do not belong to the same group (i.e., Ch 2− /X − = S 2− /Br − and Se 2− /I − ) the AMCh 2 X crystallizes in the monoclinic space group C2/m, as observable in CdBiS 2 Br (Figure 6d).This crystal structure also generates layered-type structures, but unlike the structures before, the Cd-containing chalcohalide slabs in this crystal structure consist of two types of Cd octahedra: CdS 6 and CdS 2 Br 4 .Materials belonging to this crystal structure are: CdSbS 2 Br, CdBiSe 2 I, CdBiS 2 Br, [131] MnSbS 2 Br, [134] MnBiSe 2 I, [135] MnBiS 2 Br. [136]Research into both these types of materials has been limited to compound synthesis and fundamental characterization and has not been applied to solar cells.

Other Type of Quaternary AMChX Type Structures
Other structures have been prepared, such as FeBiS 2 Cl, which was crystallized in the orthorhombic space group Cmcm.39]

Comparison of the PV Performances of Lead-Free Perovskite, Perovskite-Inspired and Chalcohalides Materials
To conclude this review, we want to offer a comparison between the solar cell performances of lead-free perovskites, perovskitederivates, and chalcohalides, to provide a discussion on the viability of these compounds in the photovoltaic landscape.In Table 2, we summarize the device structure and performance of the most efficient photovoltaic devices based on double perovskite, vacancy ordered double perovskite, vacancy ordered perovskites and the compounds discussed in this review.
Observing the figure of merits of these devices it can be noted how the ternary and quaternary chalcohalides are comparable to the most efficient lead-free and perovskite-inspired solar cells.The optimization of the device architecture, interfaces, and morphology can lead to substantial improvement in the power conversion efficiencies of these devices, and efficiencies above 6% have been recorded for SbSI, very close to the best performances of Cs 2 AgBiBr 6 .Meanwhile, in terms of stability, these devices are comparable and competitive with the perovskite counterparts.Even if much less investigation has been dedicated to the quaternary chalcohalides compounds, the reported efficiencies are promising: 1% PCE for a non-optimized CuBiSCl 2 device structure (no HTL and planar architecture) is definitely encouraging.On the other hand, higher efficiencies have been achieved by MASbSI 2 , Sn 2 SbS 2 I 3 and Pb 2 SbS 2 I 3 , even if these materials have shown some stability (both thermodynamic and atmospheric) issues.
perovskites compositions (Figure 7f) present sharp peaks in the absorption profile, suggesting the presence of excitonic behavior.[142][143][144] On the other hand, the experimentally collected ternary chalcohalides absorption spectra (BiSI Figure 7a,b, SbSI Figure 7c, and SbSeI Figure 7d) show a step-like absorption without clear excitonic behavior, similarly to the quaternary chalcohalides (MASbSI 2 Figure 7g, CuBiSCl 2 Figure 7h).However, it is important to note that carrier local-ization effects could still be present in these materials (especially in the compounds having dimensionally confined crystal structures [145] ), which could explain the limited device performances so far achieved (especially for Bi).For this reason, more investigations and studies are needed to clarify the excited charges behavior in ternary and quaternary chalcohalides.

Conclusion
In this perspective, we reviewed recent advances in the application of bismuth/antimony-based ternary and quaternary chalcohalides as light-absorbing materials.We discussed the band gap, effective mass, and electrical properties of these materials, as well as various methods for the preparation of antimony-and bismuth-based chalcohalide solar cells.Thanks to the promising PCE reported, antimony-based chalcohalides are gaining increasing interest from the photovoltaic community.BiSI is currently the main bismuth-based chalcohalide used as a light absorbing material in solar cells and many studies have been carried out on its synthesis and device integration.Other bismuth-based chalcohalide also have suitable band gaps for their use as light absorbers, but limited investigation is currently available, highlighting the need for thorough research into these materials.Lastly, we discussed the 2D and 3D structures observable in the quaternary materials and the pseudo-perovskite structures (generally identified as AMChX) that can be derived from MChX.In this context we overview their crystal structure, and optical properties and summarize the reported devices based on these materials, showing their potential in expanding the elemental pool for the design of environmentally friendly and high-performance solar absorber materials.
Although bismuth/antimony-based chalcohalides have many advantages such as low effective mass and high dielectric constants, they generally present high bandgap, >1.7 eV (with many having bandgaps above 2.0 eV), which is challenging for efficient light absorption.Additionally, the indirect nature of the ternary chalcohalide bandgap limits the absorption coefficient.The maximum efficiency of single junction solar cells under one sun illumination in the standard AM 1.5 solar spectrum occurs at 33.7% for semiconductors with a bandgap of 1.34 eV. [146]An excessively high bandgap will result in photons with energy below the bandgap not being absorbed, many of which cannot be converted into current and the current density is very low.However, the bandgap of these materials is highly dependent on the compositions, hinting at composition engineering as a possible solution to increase the efficiency of metal chalcohalide single-junction solar cells.In this context, mixed cations (M), mixed chalcogen (Ch), or mixed halide (X) compositions have been scarcely reported, representing a sector where a lot of research is still needed.Even less explored, is the effect of extrinsic doping, that could be an additional strategy to tune their optoelectronic properties.On the other hand, these large bandgap materials could be used in multi-junction applications, as wide-bandgap top-cells.No literature is currently present on chalcohalide-based tandem, clearly due to the limited efficiency achieved by the MChX-based single-junctions.However, if their performances improve, these materials are potential candidates for multijunction devices.
In order to improve the power conversion efficiency, the morphology of the bismuth/antimony-based compound films needs to be optimized.The thin films forming a solar cell (especially the absorber) must have homogeneous crystals, tightly packed with low roughness to ensure good electric conduction and effective light absorption.As most bismuth/antimony-based compounds are 1D materials, the crystals grow along one axis, re-

Figure 2 .
Figure 2. a) Simulated and experimental bandgaps, b) effective masses of electron and hole for the bismuth and antimony-based chalcohalide compounds.

Figure 3 .
Figure 3. a) Schematic illustration of the two-step method for SbSI fabrication.b) Schematic illustrations of the deposition process for SbSeI and FESEM images of samples prepared using precursor solutions with different Sol A: Sol B molar ratios.c) The SEM of Sb 2 S 3 and Sb 0.67 S 0.33 I on TiO 2 substrates and cross-sectional SEM of SbSI cell devices.Reproduced with permission.[24]Copyright 2017, Wiley-VCH.Reproduced under terms of the CC-BY license.[82]Copyright 2021, published by Multidisciplinary Digital Publishing Institute.Reproduced with permission.[23]Copyright 2019, Wiley-VCH.

Figure 4 .
Figure 4. a) Suggested band alignment of the FTO/SnO 2 /SnO 2 /BiSI/F8/Au solar cell and surface morphology of flake-shaped BiSI grains.b) Schematic diagram of the BiSI fabrication via the two-step solution process.c) Cross-sectional FESEM images of samples prepared with different repetition times of fabrication of Bi 2 S 3 .Reproduced with permission.[96]Copyright 2019, American Chemical Society.Reproduced under terms of the CC-BY license.[99]Copyright 2019, published by Multidisciplinary Digital Publishing Institute.Reproduced under terms of the CC-BY license.[99]Copyright 2019, published by Multidisciplinary Digital Publishing Institute.

Figure 5 .
Figure 5. a) Schematic illustration of formation reaction mechanism of BiSCl.b) FESEM images of products synthesized by solvothermal treatment of solution at 200 °C for different reaction times.Reproduced with permission.[105]Copyright 2021, Wiley.Reproduced with permission.[105]Copyright 2021, Wiley-VCH.

Figure 6 .
Figure 6.a) Atomic structure of CuBiSCl 2 .b) Reacting route of SbSI with MAI, the proposed structure of MASbSI 2 from DFT relaxation, and the optical images of solid-state reaction product after annealing a well-mixed pellet of SbSI and MAI at 150 °C in an open vial.c) Calculated crystal structures for Cmcm polymorphs of Sn 2 SbS 2 I 3 in the conventional orthorhombic unit cell.d) Projected views of CdBiS 2 Cl and CdBiS 2 Br; each represents the orthorhombic space group Pnma and the monoclinic space group C2/m, respectively.Reproduced with permission.[120]Copyright 2022, Wiley-VCH.Reproduced with permission.[124]Copyright 2018, American Chemical Society.Reproduced with permission.[125]Copyright 2021, The Royal Society of Chemistry.Reproduced with permission.[131]Copyright 2006, American Chemical Society.

Table 1 .
Space group, simulated and experimental bandgaps, effective masses of carriers, and static dielectric constant of the 27 bismuth and antimonybased chalcohalide materials considered in the current work.

Table 2 .
Table reporting the device architecture, PCE, V oc , J sc , and FF of different lead-free perovskite, perovskite-inspired, ternary chalcohalides, and quaternary chalcohalides solar cells.