Sn‐Based Perovskite Halides for Electronic Devices

Abstract

Because of its less toxicity and electronic structure analogous to that of lead, tin halide perovskite (THP) is currently one of the most favorable candidates as an active layer for optoelectronic and electric devices such as solar cells, photodiodes, and field‐effect transistors (FETs). Promising photovoltaics and FETs performances have been recently demonstrated because of their desirable electrical and optical properties. Nevertheless, THP's easy oxidation from Sn²⁺ to Sn⁴⁺, easy formation of tin vacancy, uncontrollable film morphology and crystallinity, and interface instability severely impede its widespread application. This review paper aims to provide a basic understanding of THP as a semiconductor by highlighting the physical structure, energy band structure, electrical properties, and doping mechanisms. Additionally, the key chemical instability issues of THPs are discussed, which are identified as the potential bottleneck for further device development. Based on the understanding of the THPs properties, the key recent progress of THP‐based solar cells and FETs is briefly discussed. To conclude, current challenges and perspective opportunities are highlighted.

Tin halide perovskite (THP) is a promising material for active layer in optoelectronic and electric devices, due to its less toxicity and analogous electronic structure to that of lead. This review provides a basic understanding of the material, including lattice structure, energy band, doping and chemical instability, and presents key progress and perspectives of THP‐based high performance solar cells and transistors.

1. Introduction

To date, metal halide perovskite (MHP) can be considered as a promising candidate for various next‐generation optoelectronics and electronics, such as solar cells,^{[ 1 ]} field‐effect transistors (FETs),^{[ 2 ]} light‐emitting diodes,^{[ 3 ]} photodetectors,^{[ 4 ]} energy storage,^{[ 5 ]} lasers,^{[ 6 ]} memories,^{[ 7 ]} and piezoelectric applications.^{[ 8 ]} Such prompt progress of MHPs in a broad spectrum of the scientific community is due to not only its impressive optoelectronic properties but also promising cost‐effective, high‐throughput, and low‐temperature manufacturing processes.^{[ 9 ]} The bulky three‐dimensional (3D) form of MHPs comprises of a corner‐sharing BX₆ octahedral network with the chemical stoichiometry ABX₃, where A is an organic or inorganic monovalent cation, B is a divalent metal ion, and X is a halide anion. The B‐X‐B configuration in a BX₆ octahedron has a bonding angle of 180°, inducing a large dispersion of both the conduction and valence bands originating from super degeneracy. This electronic structure implies that MHPs possess intrinsically outstanding and desirable electrical and optoelectronic properties such as high charge‐carrier mobility, tunable bandgaps, large optical absorption coefficient, long exciton diffusion length, and ambipolar charge transport.^{[ 10} , ¹¹ , ¹² , ¹³ , ^{14 ]} Remarkable performances from MHP‐based FETs and MHP‐based solar cells (PSCs) (power conversion efficiency (PCE) > 25%) imply the potential scopes for developing future functional optoelectronic devices.

Despite these exciting properties and advances, most MHP devices today are manufactured using Pb as the divalent metal in the ABX₃ structure, the toxicity issue of Pb remains a long‐lasting debate. The concentration of immediately dangerous to life or health of Pb is 100 mg m⁻³, which means there is a risk of human exposure to Pb in the fabrication and handling of perovskite devices.^{[ 15 ]} Moreover, our biological systems might be in great jeopardy due to the water‐soluble properties of Pb. Thus, to prevent any perturbing issues related to Pb, the European Union has publicly announced that exposures of up to 1000 ppm are allowed for all electronic devices.^{[ 15 ]} Systematic encapsulation protocol and numerous encapsulation approaches have been given serious consideration to inhibit any possible leakage of Pb; however, it is difficult to ensure how safe it is when it comes to device fabrication as well as long‐term storage. Even with a sophisticated encapsulation procedure, there is no guarantee that potential Pb leakage can be prevented. Therefore, controlling, handling, and securing safety during the manufacturing of any related Pb‐based MHPs devices remain a huge challenge.

Apart from the encapsulation method, recent attention has been devoted to replacing Pb with low toxicity and non‐toxicity cations. Several potential low‐toxic and chemically compatible materials such as Sn,^{[ 16 ]} Bi,^{[ 17 ]} and Ge^{[ 18 ]} have been proposed to replace Pb, not only to reduce its toxicity but also to maintain the unique optoelectronic properties of perovskite. Among these materials, eco‐friendly material Sn has been widely utilized in various promising optoelectronic devices, including solar cells and FETs. Similar to that of Pb, Sn has an inactive outer shell s orbital, which is vital to realizing the distinctive electrical and optical properties of MHP.^{[ 19 ]} The effects of Sn substances may vary, and it is relatively safe for a human when the hydrogen bonds grow longer in Sn substances.^{[ 20 ]} Still, the influences and their adverse effect on human health have raised some eyebrows, there is a big question regarding its use.

The first Sn‐based perovskite (tin halide perovskite; THP) was successfully developed by Fisher et al. in 1974,^{[ 21 ]} followed by a systematic study from Donaldson's research group.^{[ 22} , ^{23 ]} Later, the first hybrid THP was proposed by Yamada et al.^{[ 24} , ²⁵ , ^{26 ]} in which Mitzi and colleagues expanded the work and conducted a study in the context of dimensional reduction, and demonstrated the first structural properties under pressure.^{[ 27 ]} These initial discoveries stimulate researchers to devote huge attention to THP‐based devices until now. In 2014, Kanatzidis et al. reported the first Sn‐based perovskite solar cells (TPSCs).^{[ 28 ]} Subsequently, numerous studies have utilized Sn instead of Pb for various type of optoelectronics devices; for instance, the latest breakthrough in TPSCs have witnessed the PCE of 14.8%, whereas the highest hole and electron mobilities of 55 and 2.1 cm² V⁻¹ s⁻¹ for THP‐based FETs have been demonstrated.^{[ 20 ]} The rapid development in THP‐based optoelectronics devices makes it urgently necessary to review and assess their potential for several optoelectronic and electronic applications. In this regard, earlier published reviews have highlighted the progress and potential of THP's applications with solar cells,^{[ 29} , ³⁰ , ³¹ , ^{32 ]} photodetectors,^{[ 33 ]} light‐emitting devices, and radiation detectors.^{[ 34 ]} With intense research for various applications, THP‐based electronic devices show encouraging performances. However, the stability of the THP‐based electronic devices is still poor as compared to the Pb‐based ones. In this article, we aim to provide a comprehensive understanding of THP and highlight the impressive latest results for their functional device applications. First, we discuss the structure, chemistry, and thin‐film formation of THP. Second, we introduce the structure‐electrical property relationship of THP with energy band structure and electrical doping through intrinsic and extrinsic doping. Third, we review the recent progress of TPSCs. In this section, we also discuss the various chemical bonding and chemical structure of THP and their effect on the fabrication of TPSCs. Fourth, we discuss recent progress in THP‐based FETs, including two‐dimensional (2D) perovskite, hybrid 2D/3D perovskite, and 3D perovskite. Finally, we review the prospects for THP‐based PSCs and FETs, aiming to motivate the scientific community and better understand the development of future optoelectronic devices (Scheme  1 ).

Scheme 1.

Summary of the THP‐based perovskite materials, and their specific engineering for device performance enhancement of PSCs and FETs.

2. THP

2.1. Structure of THP

THPs consist of ABX₃ stoichiometry (Figure  1a). To maintain the neutrality of the structure, the sum of the oxidation states of the A and B cations must be 3⁺, thus requiring a 3⁻ charge, provided by the three monovalent anions is required. A stoichiometry outside this restriction forms other dimensions of perovskite or non‐perovskite structures. The chemical and structural versatility of perovskites offers many opportunities for research and applications. The inorganic sublattice is composed of [BX₆]⁴⁻ octahedra, where B is a divalent metal such as Pb²⁺, Sn²⁺, and Ge²⁺ or a mixture of monovalent and trivalent metals (e.g., Ag⁺ and In³⁺) and X is a halide: I⁻, Br⁻, or Cl⁻. By permutation of the A‐site cations and anions, a huge number of THPs are possible.^{[ 35 ]} Besides the widely used Cs and Rb, 13 protonated amines can be incorporated into the perovskite framework, as A‐site cations and eight anions, to obtain different types of perovskites. Additional structures can also be achieved through doping with various compounds. Nevertheless, to realize stable THPs, much theoretical understanding and fundamental calculation are necessary. At room temperature, the amine cations (e.g., methylammonium (MA), formamidinium (FA)) in the perovskite cavity are generally disordered, and the systems exhibit a remarkable spectrum of ferroelectric and multiferroic characteristics when cooled.^{[ 36 ]} However, to realize an ideal perovskite structure, the cations must be well‐matched with the anions. The Goldschmidt model provides an ideal method to determine the stable perovskite structure.^{[ 37 ]}

$$t=\frac{R_a+R_x}{\sqrt{2}(R_b+R_x)}$$ (1)

where R_a , R_b , and R_x are the ionic radii of the A, B, and X sites, respectively. Hence, the tolerance factor (t) assesses whether the A‐cation can fit within the framework of corner‐sharing octahedra. Experimental data suggest that t values have to be in the range between 0.8 and 1.0 to maintain a 3D perovskite structure. If the t values are within 0.9 to 1.0, an ideal cubic phase is expected. For the largest value of R_b and R_x (R _Pb = 1.19 Å, R _I = 2.20 Å) within the optimum t limit (0.8 < t < 1), the ionic radius for A‐site cation to form perovskite framework was calculated at 1.60–2.60 Å.^{[ 38} , ^{39 ]} Hence, MA (2.17 Å) and FA (2.53 Å) are the most suitable organic cations for developing effective perovskite compounds.^{[ 40} , ^{41 ]} Incorporation of A‐cation beyond the above size limit could reduce the dimensionality of perovskite from 3D to 2D and 1D.^{[ 42 ]} Interestingly, when divalent B‐cation is replaced from large Pb to small Sn cation, the values of t increase and approach toward 1. One of the remarkable aspects making the crystal structures of perovskites versatile is the structural flexibility of organic groups (Figure 1b). A t value close to 1 can be achieved by replacing a larger monovalent A‐cation (FA⁺) with a smaller cation (Cs⁺) too. The doping of FASnI₃ with Cs could tune the t downward from 1.04 to 1.02.^{[ 43} , ^{44 ]} THPs have high electrical conductivity (σ) because of the marginal crossing of Sn 5s and Sn 5p bands near the Brillouin zone boundary (due to the large dispersion of Sn 5s hybridized band) point ({½, ½, ½} 2π/a).^{[ 45 ]} Relativistic effects (e.g., spin‐orbital coupling, effective masses of electrons and holes, orbital contraction, or inert pair effect)^{[ 46 ]} are stronger in Pb than Sn, which leads to stabilization of 6s ² orbital of Pb relative to 5s ² of Sn.

Figure 1.

a) ABX₃ structure of MHP, b) size of various cations and anions, and c) tolerance factor of MHPs with a variation of cations.Adapted with permission.^{[ 43 ]} Copyright 2017, American Association for the Advancement of Science.

2.1.1. MASnI₃

MASnX₃ perovskites were studied intensively because of their similarity with MAPbI₃. MASnI₃ shows a cubic (Pm3m) phase at 295 K; however, a phase transition from cubic to lower symmetry tetragonal (I4/mcm) occurs at 275 K.^{[ 47} , ^{48 ]} Because the SnI₆ octahedra tilt about the vertical axis, the tetragonal structure corresponds to a √2a × √2a × √2a supercell extension of the cubic lattice. By lowering the temperature to 108–114 K, another phase transition from tetragonal to orthorhombic (Pbn21 or Pbnm) can be obtained.^{[ 49 ]} In the adjacent planes, the SnI₆ octahedra tilt in the same direction around the c–axis (Figure  2 ). The A‐cation has no rotational flexibility in this arrangement. Similarly, holes become more localized, extending the lifetime of charge carriers.^{[ 50 ]} According to the first principle calculations, this phase has extremely low conductivity,^{[ 51 ]} which makes them an ideal candidate for low‐temperature electronics.

Figure 2.

Rotation of adjacent octahedral layers along the c–axis according to crystal phase. a) Cubic, b) tetragonal, and c) orthorhombic phase. a–c) Reproduced under the terms of the Creative Commons CC‐BY license.^{[ 52 ]} Copyright 2022, SciELO Brazil.

2.1.2. FASnX₃

At room temperature, MASnI₃ and FASnI₃ are isostructural. Because their symmetry does not match the O_h site symmetry of the cubic perovskite structure, the FA molecules are orientationally disordered.^{[ 53 ]} Because of the steric effect of the FA cation, the bond length of Sn‐I is 3.158 Å, whereas, in the MASnI₃ system, the Sn—I bond length is 3.121 Å. FASnI₃ forms a pseudocubic Amm2 crystal structure at 340 K. Interestingly, a phase transition to orthorhombic phase with Imm2 space group can be observed below 180 K for the FASnI₃ system.^{[ 28 ]} However, clear conscientious experiments below 100 K to understand the crystal structure of FASnI₃ have not been reported yet. Nevertheless, gradual cooling to room temperature revealed some phase transition properties for FASnI₃ perovskites. The first phase transition upon cooling from room temperature was observed between 225 K and 250 K in cubic Pm3m to tetragonal P4/mbm, and a further phase transition to orthorhombic Pnma structure occurs between 125 and 150 K was confirmed by Schueller et al.^{[ 54 ]} Another study showed that a phase transition from a cubic room temperature to a tetragonal phase is possible at 255 K and a second tetragonal phase below 155 K.^{[ 55 ]} However, it must be considered that the variation of the crystal phases has been identified with either single crystals or polycrystalline films, which might have an impact. Figure  3 shows the phase transitions of FASnI₃. At room temperature, the cubic phase contains rotationally disordered FA molecules. When the temperature is deduced to 255 K, an in‐plane rotation of the SnI₆ octahedra occurs. A further deduction of temperature to 155 K depicts the fully ordered FA molecules. At this stage, instead of one, there are two distinct layers of SnI₆ octahedra with different rotation angles. The inclusion of a pseudohalide SCN partially with the FASnI₃ system can change the orthorhombic phase to monoclinic and triclinic structures.^{[ 56 ]}

Figure 3.

FASnI₃ undergoes two phase transitions in the temperature range from 300 to 85 K. a,b) The cubic phase at room temperature contains fully rotationally disordered FA molecules. c,d) The structure becomes tetragonal at 255 K because of an in‐plane rotation of the SnI₆ octahedra. The orientation of the FA molecules is two‐fold disordered due to mirror planes (shaded light blue). e,f) Doubling of the c‐lattice parameter occurs at 155 K, as the mirror planes perpendicular to the c–axis are removed and the FA molecules become fully ordered. There are now two distinct layers of SnI₆ octahedra with different rotation angles. a–f) Reproduced with permissio.^{[ 55 ]} Copyright 2020, American Chemical Society.

2.1.3. CsSnX₃

When compared with MA⁺ or FA⁺, Cs⁺ is smaller in size which leads to rotation and tilt of SnI₆ octahedral. This tilt influences the structural stability of the CsSnI₃ at room temperature. Interestingly, at a higher temperature (500 K), the CsSnI₃ perovskite converts to a black cubic (B‐α) phase (Figure  4a). Upon gradual decrease of temperature to 380 K, the black cubic phase transforms into low symmetry black tetragonal phase (B‐β) (Figure 4b) phase. At 300 K, the phase transition to the black orthorhombic (B‐γ) phase is stable at room temperature.^{[ 57 ]} In this aspect, the B‐γ phase of CsSnI₃ perovskite is most suitable for device applications. Nonetheless, studies suggest that exposure to ambient air or organic solvents in this phase tends to transform into a double‐chain structure as shown in Figure 4d. The formation of such a double‐chain structure is susceptible to air and often decomposes to Cs₂SnI₆, which has poor optoelectric and electric properties making it unsuitable for photovoltaic and FET applications. Kubicki et al. performed a detailed solid‐state nuclear magnetic resonance (ssNMR) study to compare the structures of MASnBr₃, FASnBr₃, CsSnBr₃, MASnI₃, FASnI₃, and CsSnI₃ perovskites.^{[ 58 ]} For every I/Br and Br/Cl mixed ratio, the ssNMR results showed that iodide–bromide and bromide–chloride mixtures can form a stable phase. Iodide–chloride mixes, conversely, produced phase‐segregated mixtures of phases, despite being somewhat miscible.

Figure 4.

The four polymorphs of CsSnI₃. On the top row, the three black perovskite structures are shown: a) cubic B‐α, b) tetragonal B‐β, and c) orthorhombic B‐γ. d) The bottom row shows the yellow crystal where the tin halide octahedral networks are fragmented into 1D chains. In all four images, the green spheres represent the Cs atom, whereas the purple polyhedra represent the octahedral perovskite cage formed by the bonding of the Sn (steel blue) and I (dark purple) atoms. a–d) Reproduced with permission.^{[ 59 ]} Copyright 2015, The American Physical Society.

2.2. Band Structure and Electrical Properties

Sn exhibits a comparable electronic configuration to Pb, as both elements possess s orbital lone‐pairs (Figure  5a).^{[ 48} , ⁵⁹ , ⁶⁰ , ⁶¹ , ^{62 ]} Sn and Pb have similar ionic radii and optoelectronic properties, such as direct bandgaps in the visible spectrum and relatively low electron/hole effective mass.^{[ 63 ]} However, THP exhibits a different electronic band structure from that of lead halide perovskites (LHPs). The bandgap of THP is ≈1.4 eV, which is smaller than the bandgap of LHP (≈1.5 eV).^{[ 64 ]} The bandgap of THP approaches Shockley–Queisser optimal bandgap for single‐junction solar cells, and under ideal fabrication conditions, TPSCs are expected to reach a theoretical PCE up to 33.4%. Additionally, the lone‐pair state of Sn 5s is more reactive than Pb 6s because the strong lanthanide shrinkage effect of Pb makes them an inert electron pair.^{[ 65 ]} Because of the absence of this effect, Sn becomes an easily oxidized element with two active Sn 5s electrons (Figure 5b). This active electron pair increases the Lewis acidity of SnI₂ over PbI₂ and speeds up the reaction of SnI₂ with Lewis base starting materials for perovskite such as MAI and FAI. In the following section, the formation of the energy band structure in THPs will be discussed in terms of orbital overlaps and unique properties of Sn.

Figure 5.

Electronic band structure of THPs. a) Electronic structure of Sn and Pb atoms. b) Schematics of s state valence electrons in SnI₂ and PbI₂. c) Energy band schematics of MASnI₃ and MAPbI₃. d) GW + SOC calculated band structure of MASnI₃. e) GW + SOC calculated electronic DOSs and optical absorption spectra (inset) of MASnI₃ and MAPbI₃. d,e) Adapted with permission.^{[ 48 ]} Copyright 2014, Nature Publishing Group. f) Optical bandgaps of THPs and LHPs FA_{1‐ x} Cs _x SnI₃ (0 ≤ x ≤ 1) as a function of Cs content. g) Perovskite lattice diagrams (top row) and corresponding energy level diagrams (bottom row) of lattice contraction (left panel), normal lattice (center panel), and octahedral tilting (right panel). f,g) Reproduced with permission.^{[ 66 ]} Copyright 2017, American Chemical Society.

2.2.1. Energy Band Structure

The energy band structure is the key to analyzing the optoelectronic and electronic properties of semiconductor materials. Through combined theoretical and experimental approaches, the band structures of THP can be examined. The valence band maximum (VBM) of THP is composed of the antibonding between Sn 5s and I 5p orbitals with its main contribution from I 5p. Conduction band minimum (CBM) is generated from the antibonding of Sn 5p and I 5p orbitals with dominant contribution by Sn 5p orbital (Figure 5c).^{[ 48} , ^{60 ]} Umari et al. investigated the electronic band structure of MASnI₃ by GW + Spin‐orbital coupling (SOC) method in comparison with its Pb counterpart.^{[ 48 ]} The band structure of MASnI₃ showed a direct bandgap of 1.1 eV (1.67 eV for MAPbI₃ at the same point), comparable with the bandgap of ≈1.3 eV obtained experimentally (Figure 5d).^{[ 67} , ^{68 ]} MASnI₃ exhibited more robust s—p antibonding coupling near the VBM because of the shallower and more active Sn 5s lone‐pair states than those of Pb 6s. Consequently, the VBM of MASnI₃ is more dispersive and locates ≈0.7 eV higher than that of MAPbI₃ (Figure 5e). The calculated average hole effective mass of MASnI₃ (0.13 m₀) is also smaller than that of MAPbI₃ (0.25 m₀). In the case of CBM, the Sn 5p orbital is shallower and less dispersive than Pb 6p orbitals due to much weaker SOC, and therefore, the CBM of MASnI₃ is less dispersive and positions at 0.2 eV higher than that of MAPbI₃. Thus, the calculated average electron effective mass of MASnI₃ (0.28 m₀) is larger than that of MAPbI₃ (0.19 m₀).

The band structure of THP is dependent not only on B‐site cation but also on A‐site cation by modulating its molecular size. Prasanna et al. demonstrated that by reducing the ionic radius of A‐site cation, the bandgap of THP decreases in the following order: FASnI₃ (1.41 eV), MASnI₃ (1.30 eV), and CsSnI₃ (1.25 eV), and a reverse trend in LHPs (Figure 5f).^{[ 66 ]} In the case of LHPs, the reduction of A‐site cation radius causes [PbI₆]⁴⁻ octahedra to tilt. Consequently, the Pb‐I‐Pb angle is reduced, and the orbital overlap between Pb and I is shortened, pushing VBM to a deeper energy level and widening the bandgap. For THPs, because of the smaller size of Sn 2p than that of Pb 2p, [SnI₆]⁴⁻ octahedron simply contracts instead of tilting, pushing the energy band to a shallower level, and ultimately shortens the bandgap (Figure 5g). However, the bandgap tuning with A‐site cation may only be possible in 3D perovskite structures. In 2D perovskites, the electronic state of organic cation is located deep in valence band (VB). Hence, the organic content merely participates in structural stabilization and electrostatic charge balance and does not directly contribute to electronic properties.

THPs have both advantages and disadvantages for optoelectronic applications. For example, THPs exhibit a reduced bandgap and higher absorption coefficient in the visible spectrum than LHPs. Hence, THPs are more suitable for efficient single‐junction solar cells. However, the high energy of Sn 5s ² states easily breaks the Sn—I bonds and increases the density of Sn vacancies (V _Sn). The formation of V _Sn and thus intrinsic doping in THPs will be discussed in the next section. In summary, the energy levels of Sn 5s and Sn 5p orbitals and CBM and VBM of THPs are located higher than those of Pb 6s orbitals and LHPs. The higher energy level of Sn 5s than Pb 6s contributes to weaker Sn—I bonding than Pb—I bonding, which increases the density of V _Sn in the thin film.^{[ 69 ]} The weak Sn—I bond also facilitates its reaction with oxygen and moisture to generate Sn—O and H—I bonds.^{[ 70 ]} Due to these subtle but critical differences in electronic band structures, THPs and LHPs exhibit different optoelectric properties and ambient stability. THPs show high hole mobility and conductivity due to smaller hole effective mass and low formation energy of V _Sn, whereas LHPs show better electron transport and ambient stability as a result of their smaller electron effective mass, deep Pb 6p orbitals, and stronger SOC effect.^{[ 69} , ^{71 ]} Hence, THPs are more suitable for p‐type electric devices such p‐channel FETs.

Electrical doping can provide additional electrons or holes into the lattice to further tune the Fermi level positions and carrier concentrations. Electrical doping within the lattice can generally be divided into two methods: intrinsic doping and extrinsic doping. Intrinsic doping is induced by self‐generated defects such as cation or anion vacancy within the constituents, whereas extrinsic doping is induced by the introduction of impurity atoms or molecules in or near the crystal lattice. Each method is discussed in the following sections.

2.2.2. Intrinsic Doping

Intrinsic doping in THPs can be induced by defect formation in the ABX₃ crystal lattice, such as vacancies (V _A, V _B, and V _X), interstitials (A _i, B _i, and X _i), and substitutions (A _B, B _A, A _X, X _A, B _X, and X _B) (Figure  6a).^{[ 72 ]} Density functional theory (DFT) calculations are often used to predict the range of tuned Fermi level (E _F) after defect formation and to calculate the transition level of each defect within the bandgap. For example, Xu et al. demonstrated that V _Sn is located closely below the VBM as a shallow trap.^{[ 73 ]} V _Sn is the dominant trap with low formation enthalpies, especially under Sn‐poor conditions. Hence, SnI₂‐based perovskite films exhibit high hole density above 10¹⁷ cm⁻³ due to easy V _Sn formation. The V _Sn formation originates from the strong antibonding between Sn 5s‐I 5p: the stronger the antibonding is, the lower the formation energy of V _Sn becomes (Figure 6b,c).^{[ 73} , ^{74 ]} The strength of Sn s‐I p antibonding was measured by crystal orbital Hamilton population (pCOHP) and showed that antibonding strength has a negative correlation with ionic radii of A‐site cation (Figure 6d). The increase in ionic radii of A‐site cation further perturbs the lattice and increases the Sn—I bond length. The optical bandgap is also affected by the extent of orbital overlap between Sn and I ions.^{[ 60} , ^{76 ]} The large size A‐site cation weakens Sn s—I p antibonding and increases in energy barrier of V _Sn formation. For example, DFT calculations performed on ASnI₃ (A = Cs, MA, and FA) demonstrated a high level of p‐doping under I‐rich and Sn‐poor conditions.^{[ 73} , ⁷⁴ , ^{77 ]} Larger size of FA⁺ than MA⁺ resulted in a longer Sn—I bond. Consequently, Sn 5s—I 5p antibonding coupling in FASnI₃ was weaker than that of MASnI₃, resulting in higher V _Sn formation energy for FASnI₃ (Figure 6e). Hence, the conductivity of FASnI₃ can be regulated from p‐type to intrinsic under Sn‐rich conditions. In comparison, MASnI₃ has high conductivity regardless of the condition. Furthermore, Ke et al. used bulky ethylenediammonium cation to modify the physical properties of MASnI₃ and FASnI_3. ^{[ 78} , ^{79 ]} The high energy state and instability of Sn 5s ² states facilitated the oxidation of Sn²⁺ to Sn⁴⁺. Therefore, THPs are significantly unstable in ambient air.^{[ 80 ]} The undercoordinated I⁻ ions may result in deep trap states within the bandgap because of the easy loss of Sn²⁺. Deep trap states are detrimental as non‐radiative carrier recombination centers and result in the loss of photogenerated carriers. Additionally, the deep traps degrade charge transport in FETs and provide large hysteresis in FET operation. To compare the intrinsic properties of THPs, Table  1 lists previously reported carrier concentrations and Hall mobilities.

Figure 6.

Intrinsic doping. a) Schematic of substitutional (B_x ), vacancy (V _B), and interstitial (X _i) intrinsic defects in ABX₃ perovskite crystal lattice. Reproduced with permission.^{[ 72 ]} Copyright 2021, Nature Publishing Group. b) Calculated defect formation energy as a function of Fermi energy at Sn‐rich condition (left) and Sn‐poor condition (right). c) Calculated transition energy levels for various intrinsic defects in CsSnI₃. b,c) Reproduced with permission.^{[ 73 ]} Copyright 2014, American Chemical Society. d) Negative associations between antibonding strength of Sn 5s‐I 5p atomic orbitals (red: measured at −2 eV below VBM via pCOHP method) and size of A‐site cations (blue) or bandgaps of corresponding Sn‐based iodide perovskites. e) Schematics of the bond length impacts on Sn 5s—I 5p antibonding coupling, also regulated by the size of A‐site cation (top), and partial charge densities around VBM level of MASnI₃ (bottom, left) and FASnI₃ (bottom, right). d,e) Adapted with permission.^{[ 74 ]} Copyright 2017, Royal Society of Chemistry. f) Calculated formation energy of Sn vacancy [ΔE(V _Sn)] at a fixed Fermi level, as a function of Sn chemical potential (µ _Sn). Reproduced with permission.^{[ 75 ]} Copyright 2014, Wiley‐VCH.

Table 1.

Intrinsic properties of THPs: carrier concentration and Hall mobility

Material	n [cm−3]	µ Hall (p‐type) [cm2 V−1 s−1]	µ Hall (n‐type) [cm2 V−1 s−1]	T [K]	Process	Measurement	Ref.
CsSnI3	8.73 × 1014	520	536	R.T.	Pressed pellet	Hall measurement	[28]
CsSnI3	1.00 × 1017	585		R.T.	Pressed powder	Hall measurement	[57]
CsSnI3	1.00 × 1019	24.5		R.T.	Thin film	Hall measurement	[90]
CsSnI3/BiI3	1.00 × 1017	26.9		R.T.	Thin film	Hall measurement	[90]
CsSnI3	3.00 × 1016	54		R.T.	Thin film	Hall measurement	[20]
CsSn x Pb1‐ x I3/SnF2	3.00 × 1015	486		R.T.	Thin film	Hall measurement	[20]
Cs2SnI6	1.00 × 1014		310	R.T.	Polycrystalline pellet	I–V/Hall measurement	[91]
Cs2SnI6	1.50 × 1016		79	R.T.	Thin film	Hall measurement	[92]
Cs2SnI6	6.00 × 1016		2.9	R.T.	Thin film	Hall measurement	[93]
Cs2SnI6	9.10 × 1018	20.2		R.T.	Nanocrystal solution dropcast	I–V measurement	[94]
MASnI3	2.00 × 1019	50		R.T.	Pressed pellet	Hall measurement	[83]
MASnI3	9.00 × 1017	200		250	As‐grown crystal	Hall measurement	[84]
MASnI3	7.94 × 1014	322		R.T.	Pressed pellet	Hall measurement	[28]
MASnI3	2.8 × 1017	25		R.T.	Thin film	Hall measurement	[95]
MASn(I/Br/Cl)3	2.2 × 1015	301		R.T.	Thin film	Hall measurement	[95]
(4Tm)2FASn2I7	5.49 × 1018	1.06		R.T.	Thin film	Hall measurement	[96]
FASnI3/(PEA)2SnI4	1.20 × 1016	0.21		R.T.	Thin film	Hall measurement	[97]
HC(NH2)2SnI3	8.38 × 1013		103	R.T.	Pressed pellet	Hall measurement	[28]

n: Carrier concentration, µ _Hall: Hall mobility, T: Temperature

The accuracy of DFT calculations should be tested as it depends on several factors such as exchange–correlation functionals, the secondary‐order phase constraints, and the supercell sizes.^{[ 81 ]} Specifically, a large size supercell is necessary to omit defect–defect interaction. Although DFT calculation is an important method to determine the compositional limits of a specific material under thermodynamic equilibrium conditions, it should be verified experimentally. Intrinsic defects can be experimentally confirmed by adjusting the ratio of starting materials in various film deposition methods. The experimental results on THPs are consistent with the theoretical calculations explained above. For example, strong p‐doping is observed in ASnX₃ perovskite due to the oxidation of Sn²⁺ into Sn⁴⁺ and easily formed V _Sn during synthesis or post‐degradation.^{[ 47} , ⁸² , ⁸³ , ^{84 ]} The density of V _Sn can be modulated in Sn‐rich conditions, such as the addition of SnF₂ in precursor solutions. Although the excess Sn²⁺ does not incorporate into the crystal lattice, the SnF₂ addition can increase the formation energy of V _Sn and reduce overall hole density (Figure 6f).^{[ 75} , ⁸⁵ , ⁸⁶ , ⁸⁷ , ⁸⁸ , ^{89 ]}

2.2.3. Extrinsic Doping

Extrinsic doping can be achieved by introducing impurity atoms or molecules in or near the perovskite lattice. Figure  7a illustrates the schematic for potential extrinsic dopants in place of each atomic site in ABX₃. The introduction of different A‐site cations in ASnX _y is discussed in the previous section where the lattice contraction changes Sn—I bond length, the strength of antibonding, and the overall bandgap. The B‐site extrinsic doping in place of Sn²⁺ is especially important because the inorganic octahedral cage and its network connection theoretically have the most contribution to the electronic properties of organic–inorganic perovskites.^{[ 60 ]} Extrinsic doping of B‐site can be generally divided into two types: homovalent and heterovalent doping. Despite the significant amount of B‐site atomic doping studies for LHPs, doping studies on THPs are still premature.^{[ 98} , ⁹⁹ , ¹⁰⁰ , ^{101 ]} To select an appropriate atomic substitutional dopant, the ionic size must be suitable to avoid host lattice distortion.

Figure 7.

Extrinsic doping. a) Schematics of extrinsic dopants in A‐, B‐, and X‐site of ABX₃ crystal structure. b) Schematic illustration of 2D single‐layered (C₈H₁₇NH₂)₂Sn_{1‐ x} Mn _x Br₄ (x = 0.25). c) Orbital‐projected partial DOS. b,c) Reproduced with permission.^{[ 102 ]} Copyright 2020, Royal Society of Chemistry. d) Schematic representation of CsSnI₃ crystal lattice engineering with Bi incorporation: formation of diffusion energy barrier, tin vacancy suppression, and strain engineering. e) Correlation of strain and Bi substitution on the energy barrier for Sn²⁺ diffusion. d,e) Reproduced with permission.^{[ 90 ]} Copyright 2021, American Chemical Society.

Mn²⁺ is one of the suitable candidates for homovalent doping of Sn²⁺. A previous study examined Mn²⁺ doping into octylammonium tin bromide and resulted in a red shift of PL spectrum, showing orange‐red light (Figure 7b).^{[ 102 ]} A series of measurements showed that Mn²⁺ atoms were bound to the lattice. X‐ray diffraction peaks showed a positive shift in the (001) signals, indicating Mn alloy formation with Sn.^{[ 99 ]} Additionally, according to the calculated electronic band structure and density of state (DOS), the top of VB is predominately occupied by Br orbitals and the bottom of the CB is dominated by Sn and Mn orbitals (Figure 7c). Mn orbitals are mostly located in CB enabling photogenerated electrons to jump into Mn orbitals for efficient energy transfer. Hence, the close orbital positions of Mn and Sn generate a harmonious yet competitive relationship between self‐trapped exciton emission and Mn d‐d‐transition emission. Another example candidate for homovalent doping of Sn²⁺ is Zn²⁺. Bowman et al. observed that slight addition of ZnI₂ into the precursor solution increased carrier lifetimes, photoluminescence (PL) quantum efficiencies, and film stability in the air by reducing the formation of tin‐rich clusters, which are susceptible to oxidation.^{[ 103 ]}

One of the earliest heterovalent doping studies by Takahashi et al. was based on artificially incorporating Sn⁴⁺ within the THP structure. The Sn⁴⁺ incorporation induced additional V _Sn and increased conductivity.^{[ 104 ]} Another heterovalent dopant candidate is Bi³⁺, which Zhang et al. incorporated into (PEA)₂SnBr₄. The Bi³⁺ incorporation resulted in a slight blue shift of the PL spectrum and reduced emission from the low energy tail.^{[ 105 ]} Zhou et al. also incorporated Bi³⁺ into the crystal lattice of CsSnI₃. The authors demonstrated that the air instability of black phase CsSnI₃ can be greatly lessened by Bi³⁺ dopant by restricting the direct conversion of γ‐CsSnI₃ into 0D Cs₂SnI₆.^{[ 90 ]} The improved stability was also revealed through DFT calculations. The lattice contraction suppressed V _Sn formation and increased the energy barrier between the transformation of γ‐CsSnI₃ to Cs₂SnI₆ (Figure 7d,e). Therefore, the introduction of an appropriate extrinsic dopant can achieve further modulate optoelectronic properties. However, heterovalent doping may not always work for all types of THPs. A combined theoretical and experimental study on the incorporation of monovalent metal cation (M⁺) into (PEA)₂SnI₄ revealed that the calculated formational energy of M⁺ doped‐(PEA)₂SnI₄ was too high and resulted in destabilized octahedral cage layer.^{[ 106 ]} Instead, experimental results showed that monovalent metal iodide existed separately within the film along the grain boundaries but was not incorporated into the lattice. Thus, monovalent metal iodides as additives enhanced the device's performance as a grain boundary passivating agent.^{[ 106} , ^{107 ]} Interestingly, molecular doping, which controls the amount of charge by bringing dopant molecules closer to the perovskite lattice, is also being studied. A more in‐depth study is needed on whether this molecular doping in THP occurs through the same charge transfer as in organic semiconductors.

The effect of electrical doping of THPs is necessary to expand our knowledge of the optoelectronic and electronic properties and their tunability. Extensive theoretical and experimental studies are required to visualize the effect of intrinsic or extrinsic atomic movement on the lattice structure and the device's performance. Although doping research on LHPs is widely investigated, THPs have yet been fully explored.

3. Chemical Oxidation of THP

Generally, most THP films for solar cells and FETs are fabricated by spin‐coating process. To achieve good film quality, each precursor component, such as MAI, FAI, and SnX₂ must be highly soluble in the chosen solvent. However, selecting the solvent to match the solubility of various components and to obtain high optoelectronic and electrical properties remains a challenge. For example, the solubility of Sn²⁺ compounds differs according to their halide component.^{[ 108 ]} SnCl₂ and SnF₂ are highly soluble in polar protic solvents such as water but are not suitable for making THP films. For the fabrication of high‐quality THP films, polar aprotic solvents are preferred because of their chemical coordination ability with all components. Because of this early research on THPs, the polar aprotic solvent N,N‐dimethylformamide (DMF) was used to fabricate THP films, despite the fact that the resulting film had poor surface morphology and limited performance when used in solar cells.^{[ 68 ]} In a follow‐up study, Hao et al. achieved better THP films by replacing DMF with dimethyl sulfoxide (DMSO) because DMSO intermediate phase controls the crystallization kinetics of the THP films.^{[ 109 ]} However, Saidaminov et al. found the irreversible redox reaction forms between dimethylsulfide (DMS) and Sn⁴⁺ in DMSO solution using nuclear magnetic resonance (¹H NMR) and X‐ray absorption near‐edge spectroscopy (Figure  8a–c).^{[ 110 ]} In another study reported at about the same time, Pascual et al. confirmed that Sn⁴⁺ tends to be prominent when a solution of FASnI₃ in DMSO was heated at 100°C for 30 min.^{[ 111 ]} Eventually, the presence of Sn⁴⁺ in the precursor was identified as the main cause of defects and induces p‐doping in each THP film, which undesirably affects the photovoltaic performance.^{[ 112 ]} Hence, the double‐sided effects of DMSO must be carefully engineered to benefit from forming the intermediate phase with perovskite precursors and improving crystallite quality by slowing down crystallization while compensating for the accelerated Sn oxidation. In a recent study, Girolamo et al. identified 16 non‐sulfoxide solvents, 12 of which can form stable perovskite precursors at 100°C for FASnI₃. The 12 solvents identified containing either amide, diamide, carbamate, or protic bifunctional groups were able to form SnI₂‐solvent complexes and retard Sn oxidation to some extent. However, only mixture of N,N‐diethylformamide:DMPU solvent system showed some photovoltaic performances.^{[ 113 ]}

Figure 8.

Sn oxidation in THP precursors. a) Images of fresh and aged FASnI₃ solutions in DMSO solvent (left) and in DMF solvent (right). ¹H NMR spectra of FASnI₃ solutions in DMSO solvent b) before and c) after thermal treatment (120°C, 10 min). The peak at 2.07 ppm indicates the presence of DMS. a–c) Reproduced with permission.^{[ 110 ]} Copyright 2020, American Chemical Society. d) Image of Sn oxidation of Sn²⁺ to Sn⁴⁺ in THP solution in ambient air (top), and reduction of Sn⁴⁺ to Sn²⁺ with the addition of metallic Sn powder. e) Schematic representation of Sn vacancy formation due to presence of Sn⁴⁺ in mixed Sn‐Pb perovskite precursor solution (top) and suppression of Sn vacancy formation in Sn‐reduced precursor solution with the absence of Sn⁴⁺. (d,e) Reproduced with permission.^{[ 128 ]} Copyright 2019, Nature Publishing Group.

The issue of Sn²⁺ oxidation into Sn⁴⁺ not only occurs from the chemical interaction with the solvent, but also from air exposure.^{[ 114} , ^{115 ]} Upon the exposure, the highly active 5s ² electrons of Sn²⁺ allow easy oxidation into Sn⁴⁺ to form V _Sn in the perovskite lattice, leaving shallow traps close to the valence band. The facile oxidation and surplus of Sn⁴⁺ lead to high density p‐type doping. This phenomenon is a critical issue for TPSCs, as oxidized perovskite layers tend to show reduced carrier lifetime and result in higher non‐radiative carrier recombination rate. Together with these dual aspects, TPSCs result in low photovoltaic performance. The easy oxidation of Sn²⁺ is also detrimental to FETs, as excess p‐type doping prohibits reaching an OFF‐state with extremely high hole density. Thus, to reduce the Sn⁴⁺ contents, several approaches have been explored, such as addition of tin halide‐derivatives (SnX₂)_, ^{[ 75} , ⁸⁷ , ¹¹⁶ , ¹¹⁷ , ¹¹⁸ , ¹¹⁹ , ^{120 ]} metallic Sn powder,^{[ 67} , ^{121 ]} hydrazine,^{[ 122} , ¹²³ , ^{124 ]} and other Lewis base reducing reagents.^{[ 125} , ¹²⁶ , ^{127 ]}

The fluoride in SnF₂ increases the Sn²⁺ content of the THP films by reducing the amount of Sn⁴⁺ by stabilizing the oxidation caused by DMSO.^{[ 129 ]} The higher electronegativity of F or Cl can interact with the adjacent Sn²⁺ in the perovskite lattice to prevent the Sn²⁺ from losing its electron pair. Sn(CH₃COO)₂ has the same effect as SnF₂ and can prevent Sn²⁺ oxidation and yield even higher photovoltaic response in TPSCs.^{[ 130 ]} Nevertheless, note that excess SnF₂, SnCl₂, or Sn(CH₃COO)₂ might induce phase segregation in THP films. Hence, an optimal amount of such additives for the fabrication of high‐quality THP films is much necessary. Lin et al. added metallic Sn powder as a reducing agent into an already oxidized, Sn⁴⁺ filled solution.^{[ 128 ]} The authors specifically chose Sn powder because it is insoluble in precursor solution on its own but still contributes to the perovskite lattice once oxidized by Sn⁴⁺ to generate Sn²⁺. Before the film fabrication, the leftover Sn residues were removed by filtering the solution. In the absence of Sn powder, the as‐prepared yellow perovskite precursor solution rapidly turned orange‐red when exposed to ambient air (Figure 8d). This color change indicates Sn²⁺ oxidation into Sn⁴⁺ in the solution. As previously discussed, conventional SnF₂ additive has certain limitations in preventing Sn²⁺ oxidation, and it does not sufficiently reduce Sn⁴⁺ back to Sn²⁺ once oxidized. Conversely, metallic Sn readily reduces Sn⁴⁺ into Sn²⁺ via comproportionation reaction.

$$\mathrm{Sn}+\mathrm{Sn}\left(\mathrm{IV}\right)\to 2\mathrm{Sn}\left(\mathrm{II}\right)(E_0=0.29\mathrm{V},\mathrm{\Delta }{G}_0<0)$$ (2)

After adding metallic Sn powder into oxidized perovskite precursor solution, the orange‐red solution returned to yellow even in ambient air, indicating the successful reduction of Sn⁴⁺ into Sn²⁺. The addition of Sn powder can suppress the formation of V _Sn by reducing the presence of Sn⁴⁺, forming a Sn⁴⁺free precursor (Figure 8e). This method enhances the air stability of the precursor solution itself and reduces V _Sn inside the grains of the respective THP films, which has been successfully demonstrated in THP‐based FETs in a following study by Zhu et al.^{[ 131 ]}

In an alternate approach, Nakamura et al. introduced a novel Sn⁴⁺ scavenger method by adding a tetramethyldihydropyrazine derivative (TM‐DHP) to form in situ metallic tin nanoparticles. When 1 mol% of TM‐DHP was added to the perovskite precursor solution of FA_0.75MA_0.25SnI₃ in DMSO with 10 mol% of SnF₂, the TM‐DHP reacts with the SnF₂ rapidly and retards to Sn (0) nanoparticles. The subsequent THP film showed better film coverage with larger grains and impressive certified photovoltaic response.^{[ 132 ]} Also, Song et al. introduced hydrazine as a strong reducing reagent through vapor atmosphere, ultimately suppressing Sn⁴⁺ formation.^{[ 124 ]} The highly volatile hydrazine can be easily introduced and removed without thermal treatment at high temperatures. Following this report, many studies utilized hydrazine‐based materials to prohibit Sn²⁺ oxidation.^{[ 122} , ^{123 ]} Alternative reducing agents with Lewis base groups, such as P—O, S—O, and C—O bonds can also form coordination interaction with Sn²⁺ in THP, inhibiting further oxidation.^{[ 125} , ^{126 ]} For example, Tai et al. introduced hydroxybenzene sulfonic acid along with SnCl₂ additive treatment as an anti‐oxidant in TPSC, showing effective control of the oxidation through aging time.^{[ 127 ]}

THP devices typically used encapsulation methods for reliable operation in ambient air. The THP encapsulation methods include the coating of polymer layers, such as poly(methyl methacrylate) or Cytop, and physical methods, such as glass encapsulation.^{[ 133} , ¹³⁴ , ¹³⁵ , ^{136 ]} However, these encapsulation processes can undoubtedly increase production difficulties and manufacturing costs. Based on the chemical understanding, we would like to address the following key facts that could be considered for the fabrication of THP films:

• 1)Solvent choice: To prevent the Sn⁴⁺ oxidation in the precursor solvent, a solvent system with low Sn²⁺ oxidation should be considered. This can be achieved by a single solvent or by mixing several solvents.
• 2)Sn⁴⁺ scavenger: To reduce the Sn⁴⁺ content in the THP films, techniques for removing Sn⁴⁺ from THP films should be evaluated.
• 3)High surface coverage: Poor film coverage occurs due to the fast film formation rate and rapid oxidation of Sn²⁺ to Sn⁴⁺. High film coverage can be achieved through proper solvent selection, additive addition, and control of the composition ratio of precursors.

To solve these issues, many approaches have been introduced for the development of THP‐based devices. In the following sections, we review the key development that addressed these issues for the development of TPSCs and FETs.

4. Application of THP for Solar Cells

4.1. Operation Mechanism of TPSCs

Generally, using PCE, the performance of solar cells is evaluated. The PCE of a solar cell can be obtained from the current–voltage (J–V) curves. Under solar light irradiation with a power intensity of P _in. The J–V curves provide vital information on short‐circuit current density (J _SC), open‐circuit voltage (V _OC), and the fill factor (FF), which reveals the PCE. Briefly, PCE can be obtained by the following equation:

$$\mathrm{PCE}=\frac{J_{\mathrm{SC}}\times V_{\mathrm{OC}}\times \mathrm{FF}}{P_{\mathrm{in}}}.$$ (3)

At short‐circuit conditions, the photogenerated charge carrier flows inside the solar cell, and under illumination, the total current density can be measured by incident photon flux density J_photon (λ):

$$J_{\mathrm{SC}}=q\underset{\lambda \min }{\overset{\lambda \max }{\int }}{J}_{\mathrm{photon}}\left(\lambda \right)\mathrm{IPCE}\left(\lambda \right)d\lambda$$ (4)

where λ _max and λ _min represent the maximum and minimum wavelength of the absorbed photons by the absorber layer, respectively, and IPCE is defined as the incident monochromatic photon‐to‐electron conversion efficiency. The IPCE of a solar cell can be described by the following equation:

$$\mathrm{IPCE}\left(\lambda \right)=\alpha \left(\lambda \right){\eta }_{\mathrm{c}}\left(\lambda \right)\left[1-R\left(\lambda \right)\right]$$ (5)

where α(λ) is the light‐harvesting efficiency and ηc(λ) is the charge collection efficiency and R(λ) is the total reflectivity of the solar cell. From the above equations, it is eminent that all these aspects must be considered to obtain a high J _SC. To obtain high light‐harvesting efficiency α(λ), the bandgap of THP film is the most critical because the photons can be absorbed only when their energy is higher than the bandgap. Additionally, for efficient photon absorption, the THP film must be thick enough with smooth surface morphology. The charge collection efficiency, ηc(λ) is determined by the carrier diffusion length and the charge‐carrier extraction capability of the adjacent charge transport layers. As mentioned in the previous sections, the THP films go through the rapid oxidation process resulting in severe V _Sn, which can contribute to the lower α(λ), ηc(λ) and result in lower J _SC than the theoretical limit.

Another critical parameter for obtaining high PCE in TPSCs is the V _OC, which can be described from the ideal photodiode equation:

$$V_{\mathrm{OC}}=\frac{nkt}{q}\ln \frac{J_{ph}}{J_0}+1$$ (6)

where J_ph is the reverse saturation current, which is determined by the IPCE spectrum. Therefore, to obtain high V _OC, increased absorption of incident photons is necessary. To increase the photon absorption, the THP layer in the TPSC should have low bulk and surface defects. Unfortunately, the bulk and surface defects in THPs are severe due to the rapid oxidation of the Sn²⁺ states. Additionally, the V _OC of the TPSCs can be correlated to the quasi‐fermi level splitting (QFLS). Mostly, QFLS can be obtained using PL quantum yield (PLQY):

$$V_{\mathrm{OC}}=\frac{\mathrm{QFLS}}{q}=\frac{1}{q}\left(E_{\mathrm{Fe}}-E_{\mathrm{Fh}}\right)$$ (7)

$$\mathrm{QFLS}={\mathrm{QFLS}}_{\mathrm{rad}}+\mathrm{kTln}\left(\mathrm{PLQY}\right)$$ (8)

where E _Fe corresponds to the electron quasi‐fermi level and E _Fh corresponds to the hole quasi‐fermi level. To obtain a minimal difference between the E _Fe and E _Fh, the energetic match between the THP layer and the adjacent charge transport layers is very critical. The CBM and the VBM must match well with the adjacent electron and hole charge‐carrier layers, respectively, to obtain a high V _OC. The QFLS_rad is the radiative limit of the THP layer that is associated with the V_OC in zero non‐radiative recombination.^{[ 137 ]} According to Equation (8), a high PLQY can assist in obtaining high V _OC. Therefore, a high PLQY of the THP layer and a suitable charge‐carrier layers with matched energy levels are a probable solution to obtaining high V _OC for TPSCs.^{[ 138 ]}

The FF, as shown in equation 3 is an influential parameter to obtain a high PCE of TPSCs. The FF is mostly determined by the charge flow within the TPSC and charge transport layers. Usually, the FF correlates with the balance between the series resistance (R _S) and shunt resistance (R _sh) of TPSCs. To obtain a high FF, R _s should be low, and R _sh should be as large as possible to lower the power loss of the device.^{[ 139 ]} From a theoretical understanding, it is understandable that for achieving high‐performing TPSCs, it is important to consider all aspects including the THP layer fabrication with low defects and respective charge‐carrier layers with matched energy levels. Some approaches have been summarized in the following sections to obtain a high photovoltaic response.

4.2. Structures of TPSCs

Among the existing structures for PSCs applications, TPSCs are fabricated with mesoporous and planar (n‐i‐p) and inverted planar (p‐i‐n) designs. In an n‐i‐p structure, the THP layer is sandwiched between a compact TiO₂ and a mesoporous TiO₂ layer, which acts as the electron transport layer (ETL) and a hole transport layer (HTL) with a metal electrode (usually Ag and Au) (Figure  9a). The compact TiO₂ layer prevents direct contact between the two selective contacts besides being an electron transport material, whereas the mesoscopic layer serves as a basis for the nucleation and development of the THP layer, promoting charge transfer and electron collection.^{[ 140} , ^{141 ]} To enable the injection of electrons, the conduction band energy level of the ETL layer must be lower than the THP layer. In this concern, apart from TiO₂, other metal oxides such as SnO₂ and Nb₂O₅ can be a suitable alternative, although they have seen limited success in TPSCs. However, the presence of oxygen vacancies at the ETL in this structure might accelerate the Sn oxidation. Conversely, the HTL layer should extract the photogenerated holes from the THP layer effectively while blocking the electrons. Generally, in Pb‐based PSCs, C₈₁H₆₈N₄O₈ (Spiro‐OMeTAD) has been widely adopted to fabricate high‐performing PSCs. However, Spiro‐OMeTAD as the HTL requires additional dopants such as lithium bis(trifluoromethanesulfonyl) imide (Li‐TFSI), 4‐tert‐Butylpyridine to extract hole efficiently from the THP layer. However, the presence of such dopants has a high influence on dissolving the THP layer. Although dopant‐free HTLs may be a probable solution for TPSC fabrication, Zhu et al. first discovered that the p‐i‐n structure is more suitable for TPSC fabrication.^{[ 142 ]} The p‐i‐n structure offers the advantage of depositing a dopant‐free ETL layer on top of the perovskite (Figure 9b). TPSC fabricated with the p‐i‐n structure exhibits less hysteresis, high PCE, and light soaking stability due to the shorter diffusion length and higher charge mobility of the THP film.^{[ 48} , ¹¹⁸ , ^{143 ]} In the p‐i‐n structure of TPSCs, the THP layer is fabricated on top of an HTL layer. Generally, poly(3,4‐ethylenedioxythiophene (PEDOT:PSS)^{[ 144 ]} is applied as an HTL layer with a VB level of −5.20 eV, which matches well with the VB of that THP layer.

Figure 9.

Structures of PSCs a) n‐i‐p structure and b) p‐i‐n structure.

To reduce the V _OC loss and obtain high PCE, it is important to choose suitable HTL and ETLs according to the CB and VB. Conversely, the adjacent ETL layer should have a CBM level of −4.50 eV or lower. In this aspect, [6,6]‐Phenyl C₆₁ butyric acid methyl ester (PCBM) (CB = −3.90 eV)‐, C₆₀ (CB = −4.50 eV)‐, and ICBA (−3.70 eV)‐based TPSCs have shown reproducible performances with impressive PCEs (Figure  10 ). Nevertheless, note that in the respective TPSCs, the CB level of the corresponding THP was tuned by either structural modification, additive engineering, or cation displacement. Therefore, for obtaining A high PCE, a higher or lower shift of the CB or VB is expected when the THP layer is modified, and appropriate energy matching HTL or ETL is required.

Figure 10.

Energy level diagram of THPs with ETL and HTL.

4.3. Progress of TPSCs

4.3.1. MASnX₃‐Based Solar Cells

The early days of TPSCs applications were conducted using MASnI₃ as the perovskite absorber due to the structural similarity with the Pb‐based counterpart of MAPbI₃. The energy bandgap of MASnI₃ is 1.30 eV, which is perfect for single p–n junction solar cells. The first MASnI₃‐based PSC was reported with a mesoporous structure and showed a PCE of 6.4%.^{[ 68 ]} The MASnI₃ layer is deposited by the one‐step spin‐coating from a DMF precursor. The MASnI₃ films showed high mobility and conductivity, but a substantial dark carrier concentration appeared due to severe self‐doping due to the rapid oxidation of Sn²⁺ to Sn⁴⁺. To reduce the dark carrier concentration and induce slower crystallization, DMSO was studied as a solvent to generate the SnI₂·3DMSO intermediate phase.^{[ 109 ]} The DMSO‐based MASnI₃ films exhibited a smoother surface morphology (Figure  11 ) and high surface coverage, although their respective PSC showed significantly low PCE. Song et al. highlighted the importance of the vapor atmosphere during MASnI₃ film fabrication and introduced hydrazine vapor to further improve the surface morphology of the spin‐coated MASnI₃ films and reduced the Sn⁴⁺/Sn²⁺ ratios by 20%.^{[ 124 ]} However, the respective TPSC showed a low PCE of 3.89% in an n‐i‐p structured PSC. The low performance of the TPSCs was attributed to the lower V _OC of the devices due to the bulk defects in the THP layer. Kim et al. suggested that the large V _OC loss of MASnI₃‐based TPSCs is due to the surface recombination at the interface rather than bulk recombination of the perovskite, which originates from the Sn²⁺ oxidation.^{[ 145 ]} A further study confirmed that the introduction of toluene instead of the diethyl ether as the antisolvent could not retard the recombination of the MASnI₃ perovskite due to unintentional hole doping with the nonpolar solvents.^{[ 146 ]} Ke et al. proposed an alternate method by introducing ethylenediammonium (en) in the MASnI₃ framework with a hypothesis on hollow MASnI₃ perovskite. These {en}MASnI₃ films showed a much lower electron‐hole recombination ratio because of the high surface coverage on top of the mesoporous TiO₂ layer and the respective TPSCs attained a PCE of 6.63%.^{[ 78 ]} A‐cation exchange approach through a two‐step deposition process was proposed to reduce the background carrier density of the MASnI₃ films. In the first step, a hydrazinium tin iodide (N₂H₅SnI₃) layer was deposited by a simple spin‐coating technique, which was followed by a transformation into MASnI₃ in a MA gas atmosphere through an organic cation displacement approach.^{[ 147 ]} The two‐step deposition method reduced the electron‐hole recombination of the MASnI₃ film and resulted in dense and uniform MASnI₃ film with micrometer‐sized grains and high crystallization and resulted in PSCs with PCE > 7%. Besides the solution‐processed mesoporous TPSCs, vapor‐assisted PSCs were evaluated by a low‐temperature vapor‐assisted solution process but showed poor photovoltaic performance.^{[ 148 ]}

Figure 11.

Surface morphology observed by SEM of CH₃NH₃SnI₃ perovskite layer on mesoporous TiO₂ from different solvents, a) DMF, b) NMP, and c,d) DMSO with different magnifications. a–d) Reproduced with permission.^{[ 109 ]} Copyright 2017, American Chemical Society.

In an alternate attempt to reduce the bulk defect and obtain high V _OC, ion exchange/insertion reactions between solid‐state SnF₂ and gaseous methylammonium iodide was adopted for the preparation of MASnI₃ films. Contrary to the traditional approach, the p‐i‐n was adopted where PEDOT:PSS was applied as the HTL. The respective MASnI₃ films showed a highly uniform, pinhole‐free coverage with excess SnF₂, and the respective film yielded a low content of Sn⁴⁺. This high‐quality perovskite film enables the realization of a PCE of 7.78%.^{[ 149 ]} Despite numerous efforts, the PCE of MASnI₃‐based TPSCs is still low compared with other TPSCs. Table 1 summarizes the notable performances of MASnI₃‐based PSCs (Table  2 ).

Table 2.

Development of photovoltaic performances of MASnI₃‐based PSCs

Structure	Method	J SC [mA cm−2]	V OC [V]	FF	PCE [%]	Ref.
n‐i‐p	Solvent‐mediated crystallization by DMSO	21.40	0.32	46	3.15	[109]
n‐i‐p	Hydrazine vapor atmosphere	19.92	0.377	51	3.89	[124]
n‐i‐p	Hollow({en}MASnI3) perovskite as absorber	24.28	0.42	63.72	6.63	[78]
n‐i‐p	Cation exchange by HASnI3 with MASnI3	22.91	0.486	64	7.13	[147]
p‐i‐n	Ion exchange/insertion reactions between solid‐state SnF2 and gaseous methylammonium iodide	20.68	0.57	0.66	7.78	[149]

4.3.2. FASnX₃‐Based Solar Cells

Significant progress was achieved for TPSCs when MA cation is replaced by FA. FA cation can stabilize the crystal structure more than the MA and lead to higher photovoltaic performances.^{[ 78} , ^{79 ]} Furthermore, the p‐i‐n structure is more suitable for FASnX₃‐based TPSC fabrication. Zhu et al. compared the performance of n‐i‐p and p‐i‐n structures for FASnI₃‐based TPSCs. Their findings highlight that the photovoltaic performance of p‐i‐n structured FASnI₃‐based TPSCs (PCE = 7.09%) outperforms the n‐i‐p structured TPSCs (PCE = 4.34%) due to favorable energetic matching with their respective charge transport layers (ETLs and HTLs).^{[ 142 ]} Since then, almost all FASnI₃‐based TPSCs studies have been performed utilizing the p‐i‐n structure. Hence, we only discuss here the development of FASnI₃‐based TPSCs for p‐i‐n structure.

Coadditive Engineering

Coadditive engineering has shown promising aspects to retard the Sn²⁺ oxidation when introducing coadditives with SnF₂ as they tend to form a complex within the perovskite system, which can encapsulate the perovskite grains and modify the crystal structure. In the coadditive engineering approach, SnF₂ is often used as the main additive and “other” coadditives were added. For simplicity of discussion, only “other” coadditives are discussed where mentioned. To retard the oxidization of Sn²⁺, a few reducing agents such as Catechin,^{[ 150 ]} ethylenediammonium dihypophosphite,^{[ 151 ]} and hypo‐phosphoric acid^{[ 152 ]} have been evaluated, but they failed to show a high PCE. Hydrazinium chloride showed a remarkable performance to reduce the Sn²⁺ oxidation by 20% (Figure  12a,b). Besides the reduction capability of the hydrazinum, Cl⁻ contributes to achieving a pinhole‐free smooth surface morphology. The corresponding TPSCs showed 5.4% PCE and retained 65% of their initial performance after 1000 h in the N₂ environment.^{[ 122 ]} It must be noted that by utilizing hydrazinium chloride, the shelf‐life stability for a more extended period was first observed for TPSCs. By utilizing the antioxidant properties of hydrazine, other hydrazine alternatives, such as hydrazine dihydrochloride,^{[ 153 ]} trihydrazine dihydriodide (THDH),^{[ 154 ]} and phenylhydrazine hydrochloride (PHCl) were evaluated for TPSCs. Trihydrazine dihydriodide coadditive engineered FASnI₃ TPSCs showed improved PCE of 8.48%.^{[ 154 ]} PHCl as a coadditive showed a robust capacity to reduce the Sn²⁺ oxidation and shifted the VBM by 0.984 eV (Figure 12c), which influenced obtaining a high J _SC of 23.54 mA cm⁻² and a remarkable V _OC of 0.76 V (Figure 12d). The presence of a hydrazino group and a hydrophobic phenyl group provided a bifunctional benefit to retard the surface defects and matched VBM with the PEDOT:PSS. The corresponding TPSCs showed PCE of 11.4% and showed stable photovoltaic performance in an N₂ environment for 110 days.^{[ 155 ]} Interestingly, the authors noted that the incorporation of PHCl induced the self‐repairing ability of the FASnI₃ films. Trace amount (1 mol%) of gallium acid (GA) and an excess of SnCl₂ can lead to the formation of an amorphous GA–SnCl₂ complex and result in the self‐encapsulated phase of the FASnI₃ perovskite, which increases the stability of TPSCs. The large bandgap of GA‐SnCl₂ based FASnI₃ prohibits the transfer of both charge carriers from FASnI₃ to the adjacent charge transport layers in TPSCs. The FASnI₃‐based TPSCs fabricated by this complex showed a PCE of 9.03% with stable photovoltaic behavior without any degradation for 1500 h in N₂ environment.^{[ 156 ]} Consecutively, Trimethyl thiourea (3T), with structural features of both a Lewis base and H‐bond donor can influence the fabrication of high‐quality FASnI₃ films.^{[ 157 ]} The unique bifunctional properties of 3T as a co‐additive can successfully enhance the morphology and texture of FASnI₃ films by spreading and joining individual crystal grains. With such attributes, the charge carrier lifetime of FASnI₃ perovskite was increased up to 123 ns leading to low V _OC loss (only 0.2 V lower than the theoretical limit) in the respective TPSCs. Due to the high V _OC of 0.92 V, J _SC of 20.4 mA cm⁻² and FF of 0.76, the 3T‐FASnI₃‐based TPSCs showed a high PCE of 14.05%. It is noticeable that even with staggering V _OC, the J _SC remained low, which was attributed to the low perovskite film thickness rather than the charge‐carrier diffusion length. However, even increasing the film thickness did not result in higher J _{SC ,} rather resulted in lower V _{OC .}

Figure 12.

Coadditive engineering of FASnI₃. a) High‐resolution XPS spectra (Sn 3d) of FASnI₃ + SnF₂ and b) FASnI₃ + SnF₂ + N₂H₅Cl (2.5 mol%) films a,b) Reproduced with permission.^{[ 122 ]} Copyright 2017, American Chemical Society. c) Potential energy diagrams (energies in eV with respect to vacuum) of PEDOT:PSS, C₆₀, FASnI₃ with 0.0%, 2.5%, 5.0%, and 10.0% of PHCl. d) The J–V curves of the best device with 5.0% of PHCl measured using forward and reverse scan modes under 100 mW cm⁻² AM 1.5G irradiation c,d) Reproduced with permission.^{[ 155 ]} Copyright 2020, Wiley‐VCH.

Controlled Crystallization

Numerous studies on TPSCs development highlight the rapid oxidation of Sn²⁺ in FASnI₃‐based perovskite due to the rapid crystallization of the perovskite layer. Even if the precursor solution is heat treated at 100 °C in a glovebox filled with PPM level oxygen before spin‐coating and antisolvent treatment, the organocation FA⁺ and iodide ion (I⁻) are easily volatilized and are not properly integrated into the perovskite structure. This phenomenon induces rapid oxidation and results in high crystal defects of the FASnI₃ film. To decelerate the crystallization process, Jokar et al. highlighted ethylenediammonium diiodide (EDAI₂) in trace amounts can lower the crystallization rate and promote the growth of highly crystalline FASnI₃ films more slowly. During the slow crystallization process, EDAI₂ can retard the V _Sn inside the FASnI₃ crystal by decelerating crystal growth through a kinetic balance between nucleation and crystal growth. The 1 mol% EDAI₂‐FASnI₃‐based TPSCs can achieve a PCE of 8.5% and exhibit stable performance for 1400 h in an N₂ environment.^{[ 158 ]} The π‐conjugated Lewis‐base molecule 2‐cyano‐3‐[5‐[4‐(diphenylamino)phenyl]‐2‐thienyl]‐propenoic acid (CDTA) shows electron‐accepting properties to reduce the oxidation and binding to the SnI₂ at the precursor stage. This strong electron‐accepting behavior contributes to a more stable Lewis adduct during the nucleation process. The CDTA‐based FASnI₃ films suppressed the in‐plane and out‐of‐plane rotation of [SnI₆]⁴⁻ and reduced the carrier recombination of the FASnI₃ films. Consequently, the corresponding TPSCs showed a high PCE of 10.32%, which is attributed to a 0.13 V increase in V _OC and a 10% increase in FF compared to the pristine TPSC. Controlling the crystallization via electron‐accepting small molecules of CDTA attained the first time observation of light soaking stability of the TPSCs at maximum power point tracking (MPPT) for 1000 h.^{[ 159 ]} Another molecule poly(vinyl alcohol) (PVA) with a high density of hydroxyl groups can make a strong O—H—I⁻ hydrogen bond with the FASnI₃ system leading to nucleation sites which induce lowered crystallization rate. As a result, the FASnI₃ perovskite crystals are more oriented and can inhibit the migration of the iodide ions to the adjacent charge transport layers. With these attributes, the V _OC value is increased from 0.55 V (for pristine TPSC) to 0.63 V (PVA based TPSC) and resulted in long‐term stable photovoltaic performance with a PCE of 8.9%.^{[ 160 ]} By taking the advantage of the slower nucleation of the H—I bond and inducing highly crystalline films, a bifunctional compound, hydroxylamine hydrochloride (HaHc) has recently been applied to reduce electronic defects of the FASnI₃ perovskites. The hydroxyl group in the HaHc formed a hydrogen bond with iodide ion in the FASnI₃ perovskite system to slow the crystallization, and the Cl⁻ ion coordinated with the under coordinated Sn²⁺ ions of the FASnI₃ structure. With these bifunctional aspects of the HaHc, the corresponding HaHc‐FASnI₃ based THP films showed lower trap assisted recombination, bimolecular recombination, and trap density as compared to the pristine FASnI₃. With such beneficial improved electronic properties, the V _Sn was reduced and improved V _OC up to 0.676 V for the HaHc‐FASnI₃ based TPSCs was observed with 500 h light soaking stability.^{[ 161 ]} In terms of V _Sn reduction, the H—I bond inclusion seems to be an effective approach, although the resultant TPSCs performance is still lower when compared to coadditive engineering approaches. Meng et al. proposed that the crystallization of FASnI₃ perovskite is more vulnerable to the solution air surface. By using a tailormade fluorinated organic cation, pentafluorophen‐oxy‐ethylammonium iodide (FOEI) in the FASnI₃ precursor solution, the surface energy of the solution/air surface was reduced, which simultaneously served as a template and resulted in highly crystalline films. When the FOEI‐FASnI₃ films were measured using depth‐dependent grazing incident X‐ray diffraction measurements at different incident angles (Figure  13a,b), the perpendicular growth at (100) plane was retarded significantly (Figure 13c). The prepared FASnI₃‐FOEI films showed very low surface roughness and threefold increased carrier lifetime as compared to the pristine FASnI₃ film. As a result, the FASnI₃‐FOEI based TPSCs showed improved V _OC of 0.67 V, a J _SC of 21.59 mA cm⁻², and a FF of 0.75, yielding a certificated PCE of 10.16%.^{[ 162 ]}

Figure 13.

Depth‐dependent grazing incident X‐ray diffraction patterns of the FASnI₃ and FASnI₃‐FOEI perovskite films after antisolvent dripping a) without or b) with thermal annealing at different incident angles. The ×5 in (a) left and (b) left means that the original intensity is multiplied by 5 in this image to clearly show the tendency. c) Possible scenarios of the surface‐controlled growth of the FASnI₃–FOEI perovskite films. a–c) Reproduced with permission.^{[ 162 ]} Copyright 2020, Elsevier.

Amine complex CH₃NH₃I·3CH₃NH₂ (MAI·3MA) was evaluated as potential to further stabilize the structure of the FASnI₃ perovskite by Dai et al. When compared to the conventionalFASnI₃:xSnF₂ precursor, the MAI·3MA: FASnI₃:xSnF₂ precursor can induce chemical stability owing to due to the extra donor electron groups in the amine complex. Due to the complex formation, the crystallization rate from precursor to THP film is significantly reduced resulting in highly crystalline THP films with low Sn²⁺ oxidation and non‐radiative recombination centers. As a result, the V _Sn is reduced and the corresponding TPSC can result in V _OC of 0.65 V, leading to a PCE of 9.53% with a light soaking stability of 1000 h.^{[ 163 ]} However, in a recent breakthrough study, Jiang et al. highlighted the importance of SnI₂ complex for regulating the crystal growth of FASnI₃ perovskite.^{[ 164 ]} Typically SnI₂ complexes are formed using two step synthesis process (TSS). However, SnI₂ complexes prepared by TSS method result in SnI₂ segregation. The segregation causes unsolvated edge‐sharing SnI₂ clusters which limit the crystal modulation of the THP growth. To promote smooth, compact FASnI₃ layer, and to retard the SnI₂ segregation, a one‐step synthesis (OSS) of SnI₂. (DMSO) _x colloidal complexes were developed, where in situ, tin metal and I₂‐DMSO in solution complex are obtained. The SnI₂.(DMSO) _x synthesized by the OSS method can be well dispersed to obtain SnI₂ precursor without the formation of any edge‐sharing clusters. With these beneficial aspects, the stable SnI₂. (DMSO) _x colloidal complexes can assist to bind with the cations of FAI and phenylethylamonium bromide (PEABr) and promote the growth of highly crystalline and smooth THP film. As a result, the electron diffusion length of the THP film was improved to 290 ± 20 nm (versus pristine = 210 ± 20 nm). As a result, the corresponding TPSCs show a high V _OC of 0.91 V, J _SC of 20.6 mA·cm⁻², and a FF of 0.77; leading to a PCE of 14.63%. It must be noted that by this approach TPSCs showed first time photovoltaic performance toward 15% PCE.

Template‐Assisted Growth

Throughout the THPs progress the solvent to film conversion rate has been highlighted as a potential source for Sn²⁺ oxidation. In recent reports, template‐assisted seed growth has shown more success to fabricate high‐quality FASnI₃ film for TPSCs application. In a seeded growth (SG) approach proposed by Cao et al., a layer of perovskite film is first deposited by an antisolvent method, and the same step is repeated. In this method, the first spin‐coated FASnI₃ layer acts as a template to promote the growth of uniform THP layer with low grain boundaries and highly crystalline perovskite films with compact and large grains and a longer carrier lifetime. TPSCs fabricated by the SG method showed a V _OC of 0.49 V, which is far below other conventional reported methods.^{[ 165 ]} To facilitate template assisted growth, poly(ethylene‐co‐vinyl acetate) (EVA) self‐sealing polymer with the ability of complex formation with SnI₂ was tested and verified. The C=O groups of EVA act as a powerful Lewis acid–base and forms a SnI₂ complexation with uncoordinated tin atoms in perovskite grains, which influences the improvement of the grain size, optimized grain orientation, and decrease the surface defects of FASnI₃ films. Furthermore, The FASnI₃.EVA based THPs show a self‐encapsulation effect which can repeal the moisture and oxygen ingression and result in humidity (60%) stable TPSCs.^{[ 166 ]}

Treatment of FASnI₃

Since the FASnI₃ surface is more prone to Sn²⁺ oxidation and leads to extreme surface defects, the post‐treatment method was considered a possible solution. Chowdhury et al. introduced a postdeposition vapor annealing method (PDVA), where after the antisolvent treatment, the FASnI₃ films were treated in MACl enclosed vapor system. The PDVA method induced the growth of high‐quality, defect‐free FASnI₃ film and showed a PCE of 8.48% when implemented in TPSCs.^{[ 167 ]} A solution approach for post‐treatment by bidentate amine‐edamine was evaluated by Kamaruddin et al. Before the antisolvent treatment, edamine treatment was introduced on top of the FASnI₃ films. The addition of edamine which has a free electron pair acts as an electron donating group and facilitates the co‐ordination with the under‐coordinated Sn²⁺ of the FASnI₃ system (Figure  14a,b). As a result, the respective THP films showed ideal crystallographic orientation near the grain boundaries and improves the surface morphology with improved grain size. With such beneficial aspects of the edamine post‐treatment, the TPSCs J _SC was improved up to 23.09 mA cm⁻² with V _OC of 0.60 and FF of 0.73, leading to a PCE of 10.18%^{[ 168 ]} Recently, Liu et al. reported the highest certified PCE of 11.73% (by an authorized public test center) with 1000 h light soaking stability under MPPT conditions by the pretreatment method of FASnI₃ film (Figure 14c,d) In this method, prior to the annealing of the THP film, right after the dripping of antisolvent n‐propylammonium iodide (PAI) from a mixed solvent of chloroform and DMSO was spin‐coated on top of the pre‐nucleated FASnI₃ film.^{[ 169 ]} The PAI treatment facilitated the recrystallization of the FASnI₃ films by forming an intermediate PAI‐FASnI₃ phase. Although DMSO itself is known to dissolve the perovskite layer, however it was claimed that 1% v of DMSO with 100% v chloroform can successfully assist to form highly oriented FASnI₃ crystals. With these extraordinary features, the non‐radiative recombination centers of the FAsnI₃ film were reduced leading to high electron diffusion length. As a result, the V _Sn of the THP film was reduced and respective TPSCs showed a high V _OC of 0.73 V, leading to a PCE of 11.22%.

Figure 14.

Post‐treatment–narrow XPS spectra after 10 s of argon etching of a) 0 mm edamine and b) 0.05 mm edamine‐passivated FA_0.98EDA_0.01SnI₃ samples. a,b) Reproduced with permission.^{[ 168 ]} Copyright 2021, American Chemical Society. c) Pretreatment—The J–V curves of FASnI₃ and TG‐FASnI₃‐based TPSCs. d) The stability test of the corresponding device under simulated AM 1.5 G (100 mW cm⁻²) operating at MPPT c,d) Reproduced with permission.^{[ 169 ]} Copyright 2021, Royal Society of Chemistry.

Ionic Liquid

Ionic liquids (ILs) are well known for their very low vapor pressure, making them ideal candidates for slower crystallization. When a small amount is added to the perovskite precursor, it tends to remain in the resulting perovskite film after all the solvent has been evaporated during the annealing process. With controlled crystal growth, the surface defects at the perovskite grain boundaries can be reduced effectively. Additionally, ILs can form a dipole at the perovskite/charge transport layer interface, which can improve the energy level alignment due to the reduced interfacial energy barriers.^{[ 170 ]} Lin et al. introduced 1‐butyl‐3‐methylimidazolium bromide (BMIBr) as IL to induce Ostwald ripening effect of FASnI₃ films. During the thermal annealing of perovskite films, BMIBr IL domains generated by a lower melting point act as Ostwald ripening agents for dissolving the perovskite grains. The FASnI₃ films fabricated by inclusion of BMIBr showed larger grain size with fewer grain boundaries which eventually reduces the defect states. The successful suppression of defect states reduces the V _Sn of the FASnI₃ films and the corresponding TPSCs showed V _OC up to 0.68 V, with J _SC of 19.63 mA cm⁻² and FF of 0.72 (PCE = 10.09%).^{[ 171 ]} In an alternate approach, IL formamidine acetate (FAAc) improved the FASnI₃‐based TPSCs performance up to 9.96%. The FAAc contains CH₃COO⁻ a group, which can form an intermediate phase and slow the nucleation of the FASnI₃ film. Even after the thermal annealing treatment, the FAAc remained within the perovskite system and retarded the cationic vacancies of the respective THP film, helping to retain 82% of the initial PCE during the 1500 h light soaking stability test.^{[ 172 ]} n‐butylammonium acetate (BAAc) IL showed coordination with specific O…Sn chelating bonds and N‐H…X hydrogen bonds in the FASnI₃ system retarded the Sn²⁺ oxidation in the precursor solution. Contrary to the other reports, the BAAc can suppress precursor Sn²⁺ oxidation even when heated at 100°C for 2 h and can form highly crystalline FASnI₃ films with fewer surface defects. The BAAc‐FASnI₃‐based TPSCs exhibited a PCE of 10.4% and maintained 80% of their initial photovoltaic performance for up to 100 h at 85°C.^{[ 173 ]} Note that this is the first report on TPSCs showing stable photovoltaic performance at elevated temperatures.

Charge Extraction Engineering

Due to the less charge extraction ability at the perovskite/ETL interface, the V _OC of the TPSCs is often low.^{[ 174 ]} Abdoul‐Sahkour et al. highlighted the importance of the charge collection loss of the FASnI₃/ETL interface and introduced diaminomaleonitrile (DAMN) in the FASnI₃ system to collect the photogenerated charges more effectively. The DAMN with two cyano groups which perform as an electron‐withdrawing group can successfully extract electrons effectively from the FASnI₃ layer and transfer them to the adjacent C₆₀ ETL, which was confirmed by a 42% increment of electron mobility. Consequently, the V _OC of the DAMN‐FASnI₃‐based TPSC improved by 0.5 V compared with the pristine FASnI_3. Due to the better charge management in the FASnI₃ system, the DAMN‐FASnI₃‐based TPSC showed stable photovoltaic behavior at MPPT conditions for 300 h.^{[ 175 ]} In a recent report, Chen et al. highlighted the importance of the charge transport at the HTL/FASn_0.9Ge_0.1I₃ interface.^{[ 176 ]} To promote efficient charge transport at the HTL/FASnI_3, the FASnI₃ perovskite was doped with 1 mol% GeI₂. Unlike the usual dopants or additives which contribute to the surface of the perovskite layer, the GeI₂ induces in situ formation of a thin (≈3 nm) amorphous interfacial GeO₂ layer at the NiO _X /FASn_0.9Ge_0.1I₃ interface. The GeO₂ layer prevents the Sn²⁺ oxidation at the interface, provides more robust mechanical bonding, and promotes effective charge transportation in the TPSC. As a result, the corresponding TPSCs showed a high PCE of 10.43%.

Cationic Displacement

Cationic or anionic displacement of FASnI₃ can tune the bandgap, push the VBM to a more shallower region, increase the light‐harvesting capability of the THP films and provide a basis for high‐performance TPSC fabrication.^{[ 177 ]} Compositional engineering by cationic displacement for FASnI₃‐based TPSCs was first evaluated by Zhao et al. where MA cation was partially introduced in the FASnI₃ system. The increasing content of MA cation in the FASnI₃ perovskite system shifts the VBM level and induced a favorable VBM level with the adjacent PEDOT:PSS HTL (Figure  15b). The (FA)_0.75(MA)_0.25SnI₃‐based TPSCs showed PCE of 8.12%, which were higher than the MASnI₃ (PCE = 4.29%)‐ and FASnI₃ (PCE = 6.60%)‐based TPSCs. The improved photovoltaic response was attributed to the improved V _OC by 0.15 V compared with the MASnI₃ or FASnI₃.^{[ 178 ]} Liu et al. discussed the importance of the appropriate antisolvent for the fabrication of FA _X MA _X SnI₃ films and revealed that chlorobenzene (CB) is the more suitable antisolvent. When compared with the other antisolvents of diethyl ether and toluene, CB as an antisolvent could assist the formation of a dense, compact FA_0.75MA_0.25SnI₃ layer and the respective TPSCs showed 9.06% PCE.^{[ 179 ]}

Figure 15.

a) Schematic illustration of the device structure. b) Band alignment diagram. c) Cross‐sectional scanning electron microscope (SEM) image of a completed device (scale bar: 500 nm). a–c) Reproduced under the terms of the Creative Commons CC‐BY license.^{[ 178 ]} Copyright 2017, The Authors. Published by Wiley‐VCH.

In an alternate approach by vacuum‐assisted self‐healing of defects in FA_0.75MA_0.25SnI₃ perovskite films to suppress the nonradiative recombination, losses were reported by Wan et al. Before the antisolvent treatment of the pre‐annealed FA_0.75MA_0.25SnI₃ perovskite films were placed under high vacuum of 5 × 10⁻⁴ Pa. The introduction of vacuum treatment before the annealing in order to reduce the trap density reduced the nonradiative recombination and promoted interfacial charge transfer in the respective TPSCs leading to a V _OC of 0.631 V and FF of 0.75 (PCE of 10.3%).^{[ 180 ]} Inclusion of Diethylammonium iodide (DEAI) as a cation in the (DEA _x FA_{1– x} )_0.98EDA_0.01SnI₃ perovskite system can shift the VBM to a much deeper level (from −4.97 to −5.15 eV) and reduce the deep trap states of the respective THP film with low defect densities. With such favorable VBM level, the hole transportation between the PEDOT:PSS and the (DEA _x FA_{1– x} )_0.98EDA_0.01SnI₃ layer is more efficient and the V _OC loss is reduced in the respective TPSC. With an optimal 10% DEA in this mixed cation‐based TPSC a high V _OC of 0.67 V (versus pristine = 0.59 V) was attained.^{[ 181 ]} By introducing a trace amount of ethylammonium (EA) as the cation to the FASnI₃ systems, a noteworthy improvement in TPSCs was observed by Nishimura et al. The FA_0.98EA_0.01SnI₃ based THP showed a shift of 0.30 eV in VBM and 0.15 eV shift in CBM. The favorable shift of both ends facilitated obtaining a high V _OC of 0.62 V while maintaining a high J _SC of 23.05 mA cm⁻² leading to a PCE of 10.80%. The inclusion of EA as a cation could successfully repress the trap states of the THP layer which increased charge carrier mobility. Further surface treatment of the TPSCs with EDAI elevated the V _OC up to 0.84 V and reached a PCE of 13.24%.^{[ 182 ]} However, after EDAI surface treatment, the J _SC decreased to 20.32 from 23.05 mA cm⁻², which highlights the double‐sided effect of the surface treatment process for mixed cation THP perovskites. To overcome the drawbacks of the reduced J _SC, the same research group recently introduced trimethylsilyl bromide (Me₃SiBr) as a surface passivation agent instead of EDAI.^{[ 183 ]} Upon surface treatment with 0.15 mm Me₃SiBr on top of the FA_0.98EA_0.01SnI₃ perovskite, the surface morphology was improved, which showed lower defect densities, larger grains, and higher carrier density. As a result, the J _SC of 24.11 mA cm⁻² can be maintained even after surface treatment. However, the V _OC of the respective TPSC was reduced to 0.70 V, leading to a PCE of 12.22%. Additionally, the Me₃Si⁺ remains as the counter cation on the surface and forms a hydrophobic protective layer which protects the FA_0.98EA_0.01SnI₃ film from oxidation which showed stable photovoltaic performance in a N₂ glovebox for 92 days of storage in N₂ filled glovebox by maintaining 80% of its initial efficiency.

Halide Exchange

By mixing the halide content of the THP, the bandgap can be tuned further. Weber et al. partially replaced the I with Br and reported that MA_0.75FA_0.15PEA_0.1Sn(Br_0.33 I₇₇)₃ can increase the bandgap from 1.29 to 1.50 eV. Although a significant improvement in TPSCs performance was expected, their fabricated TPSCs showed a low PCE of 4.63%.^{[ 184 ]} The bandgap of FASnI₃ films was further enhanced to 1.61 eV by the addition of GuBr by Chen et al. with an optimized composition of GA_0.06(FA_0.8Cs_0.2)_0.94SnI₂Br, which showed matched energy level alignment with the adjacent charge transport layer and reduced the trap density (both surface and bulk). The best performing TPSC showed V_OC  of 0.64 V, J _SC of 15.16 mA cm⁻², and FF of 0.72 after aging.^{[ 185 ]} In a recent report, Br mixed perovskite system of EDA_0.01(GA_0.06(FA_0.8Cs_0.2)_0.94)_0.98SnI₂Br, the PCE was further enhanced up to 8.66%.^{[ 186 ]} However, the improved photovoltaic performance of these wide bandgap TPSCs was attributed to the influence of replacement of PEDOT:PSS by a monolayer of 2PACZ rather than the influence of the wide bandgap THP layer (Table  3 ).^{[ 186 ]}

Table 3.

Summary of photovoltaic performances of FASnI₃‐based PSCs

Coadditive engineering
Additive	Coadditive	J SC [mA cm−2]	V OC [V]	FF	PCE [%]	Ref.
SnF2	N2H5Cl	17.63	0.455	0.673	5.4	[122]
SnF2	5‐AVAI	18.89	0.592	0.623	7.0	[187]
SnCl2	KHSQA	17.64	0.552	0.694	6.76	[127]
SnF2	THDH	22.12	0.54	0.71	8.48	[154]
SnF2	PHCl	23.5	0.76	0.64	11.43	[155]
SnCl2	GA	19.75	0.64	0.714	9.03	[156]
SnF2	3T	20.4	0.92	0.76	14.3	[157]

Controlled crystallization
Additive	Compound for crystallization control	J SC [mA cm−2]	V OC [V]	FF	PCE [%]	Ref.
SnF2	EDAI2	21.3	0.583	0.718	8.9	[158]
SnF2	CDTA	21.83	0.64	0.739	10.32	[159]
SnF2	PVA	20.37	0.632	0.693	8.92	[160]
SnF2	HaHc	19.40	0.676	0.70	9.18	[161]
SnF2	FOEI	21.59	0.670	0.75	10.84	[162]
SnF2 and NH4SCN	SnI2·(DMSO) x adduct	20.6	0.91	0.77	14.63	[164]

Template‐assisted growth
Additive	Compound for template‐assisted growth	J SC [mA cm−2]	V OC [V]	FF	PCE [%]	Ref.
SnF2	EVA	22.80	0.523	0.646	7.72	[166]
SnF2	SG‐FASnI3	22.52	0.490	0.662	7.30	[165]

Ionic liquid
Additive	Ionic liquid	J SC [mA cm−2]	V OC [V]	FF	PCE [%]	Ref.
SnF2	BMIBr	19.86	0.70	0.723	10.09	[171]
SnF2	FAAc	23.20	0.59	0.727	9.96	[172]
SnF2	BAAc	22.20	0.65	0.716	10.40	[173]

Cation displacement
Additive/method	Perovskite compound	J SC [mA cm−2]	V OC [V]	FF	PCE [%]	Ref.
HAT method	(FA)0.75(MA)0.25SnI3	19.40	0.55	0.670	7.20	[188]
SnF2	(FA)0.75(MA)0.25SnI3	21.20	0.61	0.627	8.12	[178]
Solvent engineering	(FA)0.75(MA)0.25SnI3	24.30	0.55	0.673	9.06	[179]
SnF2 + TM‐DHP	(FA)0.75(MA)0.25SnI3	22.00	0.76	0.690	11.5	[132]
Vacuum‐Assisted growth	(FA)0.75(MA)0.25SnI3	21.62	0.63	0.755	10.3	[180]
Surface treatment with EDAI2	FA0.98EA0.01SnI3	20.32	0.84	0.78	13.24	[182]
Surface treatment with Me3SiBr	FA0.98EA0.01SnI3	24.11	0.70	0.72	12.22	[183]

4.3.3. CsSnX₃‐Based Solar Cells

Because of the absence of hydrophilic organic cations of MA and FA, CsSnX₃ perovskites offer unique advantages in terms of structural and thermal stability. With an exciton binding energy of 12–18 meV, CsSnX₃‐based perovskites are comparable with LHPs.^{[ 61} , ¹⁸⁹ , ^{190 ]} However, when implemented in TPSCs, CsSnX₃ has shown lower performances when compared with FASnI₃‐based TPSCs. The low performance of CsSnX₃‐based TPSCs is attributed to the V _Sn, which generates severe crystal defects during the formation state of films. Distinctively, when the Sn content is increased, the p‐type conductivity reduces because of the reduction of accepter defects, which can influence the increase of shunt resistance and result in high‐performance TPSCs.^{[ 73 ]} To experimentally address these phenomena, Kumar et al. reported that adding 20% SnF₂ can decrease the carrier density from 10¹⁹ to 10¹⁷ cm⁻³, which increases the formation energy of V _Sn.^{[ 75 ]} Consequently, the TPSCs fabricated in n‐i‐p configuration showed a PCE of 2.02% with a V _OC of 0.24 V. An improved V _OC of 0.55 V with a PCE of 2.76% was obtained in p‐i‐n structured TPSCs when excess SnI₂ was added to prepare the CsSnI₃ films.^{[ 191 ]} In an alternate study, Song et al. evaluated the role of excess SnI₂ in hydrazine vapor atmosphere and highlighted the impact on the surface morphology and crystallinity. The inclusion of excess SnI₂ induced homogenous CsSnX₃ film with smooth surface morphology without influencing the crystallinity. With an optimized CsI/SnI₂ molar ratio of 0.4, a PCE of 4.8% was attained with n‐i‐p structured PSCs.^{[ 120 ]} A systematic investigation of the addition of excess SnX₂ (X = Cl, Br, I, and F) revealed that SnCl₂ is the most suitable compound to obtain smooth surface morphology and maintain the B‐γ phase of CsSnI₃ for TPSCs applications with PCE of 3.56%. However, the partial replacement of Cl and I was not observed; rather, an ultrathin layer of SnCl₂ layer formed on top of the perovskite layer and provided assistance to achieve stability.^{[ 192 ]}

Although the introduction of SnX₂ additives in the CsSnI₃ framework induced some photovoltaic response but the PCE was significantly lower than 5%. The SnX₂ additives failed to address the critical Sn²⁺ oxidation even in an N₂ atmosphere with a trace amount of oxygen (1 ppm). To increase the photovoltaic response, taking for example from FASnI₃ perovskites, the coadditive strategy was adopted for CsSnI₃ perovskite films. The addition of 2‐aminopyrazine with 20% SnF₂ remarkably suppressed the Sn²⁺ oxidation, which was attributed to the electron‐donating capability of the amino group in the aminopyrazine and the formation of the aminopyrazine–SnF₂ complex. Upon spin‐coating, the precursor followed by antisolvent and thermal annealing at 100°C resulted in smooth CsSnI₃ perovskite films with compact grains. Although the V _Sn of the respective TPSC was not reduced which can be observed from the low V _OC of 0.40 V.^{[ 193 ]} A higher PCE of 7.50% was attained by utilizing the lone electron pairs of −NH and −CO functional groups of methylenebis(acrylamide) (MBAA) (Figure  16a), which can coordinate with Sn²⁺ and prevent oxidation and eventually reduce the V _Sn of the B‐γ CsSnI₃ films.^{[ 194 ]} Remarkably, the TPSCs showed stability for 120 h under continuous 1 sun illuminations. The lone electron pairs of –NH and two —CO units can form a trigeminal coordination bonding with Sn²⁺, which reduces the defect density of the CsSnI₃ film. With such properties, when phthalimide (PTM) was used as a coadditive with SnF₂, the electron density of CsSnI₃ was successfully passivated, and smooth pinhole‐free surface morphology was observed (Figure 16b). The respective TPSCs showed the highest PCE of 10.1% for CsSnI₃‐based TPSCs with V _OC of 0.64 V, J _SC of 21.81 mA cm⁻², and FF of 0.721.^{[ 195 ]}

Figure 16.

SEM and HRTEM images of a) CsSnI₃‐MBAA and b) CsSnI₃‐PTM. J–V curves of the device fabricated with c) CsSnI₃‐MBAA and d) CsSnI₃‐PTM. a,c) Adapted with permission.^{[ 194 ]} Copyright 2021, American Chemical Society. b,d) Adapted with permission.^{[ 195 ]} Copyright 2021, American Chemical Society.

In an alternate approach, B‐site engineering of the CsSnI₃ was evaluated to tune VBM and CBM. To maintain the concept of low toxicity, Ge element with a similar electronic structure to Sn was considered and evaluated for CsSnI₃. DFT calculations revealed that 50% of Ge inclusion in the CsSnI₃ can induce a direct bandgap with VBM and CBM can induce a high absorption coefficient and low exciton binding energy.^{[ 196 ]} However, Ge incorporation into the precursor and thin film fabrication may not be a feasible approach for the fabrication of CsSn_0.5Ge_0.5I₃‐based perovskites. Chen et al. introduced a CsSn_0.5Ge_0.5I₃ perovskite powder synthesis method by solid‐state reaction in an evacuated Pyrex glass tube at 450°C.^{[ 197 ]} The as‐synthesized CsSn_0.5Ge_0.5I₃ perovskite powder showed high PL intensity, a bandgap of 1.50 eV, and upshifted VBM. Interestingly, when the CsSn_0.5Ge_0.5I₃ powders were evaporated to fabricate perovskite films, upon exposure to air, the formation of a Ge⁴⁺‐rich native‐oxide layer on the surface of the perovskite film was observed. The native‐oxide layer effectively protected the film surface, suppressed the charge recombination, and enhanced the holes extraction at the perovskite/HTL interface in an n‐i‐p‐structured TPSC. The respective TPSCs fabricated with CsSn_0.5Ge_0.5I₃ achieved a PCE of 7.11% with V _OC of 0.63 V, J _SC of 18.61 mA cm⁻², and FF of 0.606 with high ambient air stability of 500 h in an N₂ environment.^{[ 197 ]}

Apart from the typical solution processing methods, the evaporation‐assisted solution (EAS) method and the vacuum flash‐assisted solution processing (VASP) method have been evaluated for the fabrication of CsSnI₃ films. The EAS method combines the solution process followed by vacuum evaporation of the perovskite films. In the EAS method, first, a precursor containing SnI₂/SnF₂ is deposited on top of the substrate by one step of spin coating and annealed at 100°C. Before annealing, CsI is thermally evaporated at 150°C for 10 min to obtain CsSnI₃ films. The n‐i‐p‐structured TPSCs fabricated using EAS‐assisted CsSnI₃ showed a PCE of 2.23%.^{[ 198 ]} The fabrication of CsSnI₃ films was evaluated using the VASP method, where the CsSnI₃ precursor solution is first deposited on top of the indium tin oxide (ITO) substrate and immediately transferred to the vacuum chamber to remove the solvent. The introduction of such a fast solvent removal technique improved the crystallization and surface properties but attained only a PCE of 3.8%.^{[ 199 ]} Yin et al. proposed a sequential vapor deposition strategy, which combined the advantages of both the vapor deposition and the surface passivation. Thiosemicarbazide (TSC) with S‐C‐N functional groups was introduced as a passivator in addition to the SnI₂ and CsI evaporation source. The vapor deposition was performed sequentially by depositing SnI₂, TSC, and CsI sequentially and annealing at 120°C to form a (B‐γ) CsSnI₃ with large grain size and uniform coverage. The fabrication of extraordinary (B‐γ) CsSnI₃ films prepared using PASVD was attributed to the strong electrostatic attachment and coordination interaction, which leads to electron cloud density and reduced the V _Sn. When TPSCs were fabricated in a p‐i‐n configuration, the V _OC enhancement from 0.47 to 0.63 V was observed which is ascribed to the reduced deep level trap‐state density of the successfully passivated THP films. Interestingly, by adoption of this approach, a PCE > 8.0% was noted for CsSnI₃‐based TPSCs with stable photovoltaic performance up to 500 h (Table  4 ).^{[ 200 ]}

Table 4.

Development of photovoltaic performances of CsSnI₃‐based PSCs

Structure	Method	J SC [mA cm−2]	V OC [V]	FF	PCE [%]	Ref.
n‐i‐p	SnF2‐MBAA	24.85	0.45	0.670	7.5	[194]
p‐i‐n	SnF2‐Thiosemicarbazide	19.7	0.63	0.661	8.2	[200]
p‐i‐n	SnF2‐Phthalimide	21.81	0.64	0.721	10.1	[195]

4.3.4. 2D/3D Solar Cells

2D/3D THP films exhibited a superior crystal quality when compared with 3D THP films due to the decreased background ions and carriers. To further improve the MASnI₃ film quality, the conventional 2D/3D approach was examined. Perovskites with a Goldschmidt tolerance factor (t) between 0.8 and 1.0 show a photoactive black phase, which can induce high‐quality MASnI₃ film. Ji et al. adopted EA cation into MASnI₃ to induce a preferential orientation perpendicular to the substrate resulting in a 2D/3D MASnI₃‐based PSC with PCE of 9.24%.^{[ 201 ]} The (EA _X MA_{(0.98− X )}2D/3D)SnI₃ films could successfully passivate the hole traps, induce preferential crystallization rate, produce a uniform and smooth surface, and suppress the trap‐assisted nonradiative recombination. Shao et al. introduced a lower concentration of PEAI (0.16 M) to FASnI₃ to obtain 2D/3D structures and verified that the crystallinity of 3D THPs is significantly better when compared with the 3D FASnI₃‐based THPs, which increased the V _OC of the respective TPSCs by 0.67 V when compared with the 3D FASnI₃‐based TPSC and showed PCE of 9%.^{[ 202 ]} In follow‐up work, the authors evaluated the importance of stoichiometry of 2D/3D PEA_0.08FA _x SnI₃ (where x refers to the concentration of the FA cation in the case of fixed PEA concentration). It was observed the concentration of FA cation significantly impacts the crystallinity of the 2D/3D films.^{[ 203 ]} A hierarchical 2D‐quasi‐2D–3D structure using NH₄SCN as an additive was proposed to regulate the crystal growth of PEA _x FA_{1‐ x} SnI₃ THP films by Wang et al.^{[ 204 ]} With regulated crystal growth, the quasi‐2D/3D THP films showed higher carrier mobility and reduced the carrier density resulting in a PCE of 9.41% in the respective TPSCs. They hypothesize that a thin layer of 2D layer of PEA₂SnI₄ on top of the 3D FASnI₃ film can provide resilience against ambient air and moisture. Because of this, the respective TPSCs can maintain 90% of their initial PCE even after 600 h storage in air. Chen et al. added a bulky divalent organic cation, 4‐(aminomethyl)‐piperidinium (4‐AMP) to FASnI₃ system. Interestingly, the addition of 15 mol% of 4‐AMP with FASnI₃ did not influence 2D/3D formation but rather remained at the surface and grain boundaries. Interestingly, this concentration was not enough to create detectable low‐dimensional phases.^{[ 205 ]} To further elevate the PCE, Jiang et al. proposed that for the development of 2D/3D aside from the THP layer the adjacent charge carriers play a pivotal role. They replaced the ETL (PCBM) with ICBA and HTL (NiO _X ) with PEDOT:PSS in the PEA _x FA_{1‐ x} SnI₃+NH₄SCN‐based TPSCs and increased the V _OC up to 0.94 V, leading to a remarkable PCE of 12.4%.^{[ 206 ]} Such high V _OC was attributed to the matched energy level of the HTL/perovskite and perovskite/ETL interface.^{[ 206 ]} By using the same HTL and ETLs, the best performing TPSCs with 14.81% were reported with 4‐fluoro‐phenethylammonium cations (FPEABr) in FASnI₃ system. It is well known that the fluorinated cations can create stronger Van der Waals interaction or hydrogen bonding reducing the Sn²⁺ oxidation and creating a 2D microstructure to influence the 3D phase growth. Additionally, fluorinated compounds provide the unique characteristics of hydrophobicity due to the presence of fluorine atoms, which resulted in the high PCE‐based TPSCs with longer stability.^{[ 207 ]} Despite the significant efforts by numerous research groups, the V _OC loss of TPSCs was minimized up to 0.45 V. Wang et al. proposed a vacuum treatment method to establish a 2D/3D stratified heterojunction which can enable charge separation at the interface. Prior to the 2D/3D stratified heterojunction formation, guanidinium thiocyanate (GuaSCN) was introduced as a co‐additive to improve the crystallinity of the 2D perovskite phase. By the synergetic effects of the vacuum treatment and GuaSCN, the respective THP films showed high hole mobility, passivated the surface traps, and prolonged carrier lifetime up to 140 ns (the highest carrier lifetime for TPSCs). The energy level diagram of the pure quasi‐2D perovskite (n = 2) film shows a similar valence band level E _V (−5.08 eV) to 3D FASnI₃ film (−5.10 eV) while its conduction band level E _C (−3.26 eV) is much higher than that of FASnI₃ (−3.71 eV), which can prohibit electron transfer across the quasi‐2D/3D mixed phase. With such beneficial aspects, the TPSCs fabricated by this method with PEA₂FA _{n − 1}Sn _n I_{3 n + 1}(n = 10) showed a high V _OC of 1.01 V, which minimizes the V _OC loss of any type of TPSCs to 0.39 V (Table  5 ).^{[ 208 ]}

Table 5.

Notable progress of photovoltaic performances of 2D/3D TPSCs

Perovskite compound	2D material	J SC [mA cm−2]	V OC [V]	FF	PCE [%]	Ref.
FASnI3	Evaporated PEAI	20.07	0.47	0.74	6.98	[209]
FASnI3	PEABr treatment	22.64	0.54	0. 64	7.86	[210]
FASnI3	PEAI0.08	24.10	0.525	0.71	9.0	[202]
FASnI3	PEAI‐ NH4SCN	22.00	0.61	0.70	9.41	[204]
FA3Sn4I13	BA0.5PEA0.5I	21.82	0.60	0.67	8.82	[211]
FASnI3	PEACl0.12	22.06	0.59	0.69	9.1	[212]
FASnI3	PEASCN0.15	17.40	0.94	0.75	12.4	[206]
FA n − 1Sn n I3 n + 1 (n = 10)	PEAI2	20.32	1.01	0.67	13.79	[208]
FASnI3	FPEABr0.10	24.91	0.84	0.76	14.81	[207]

4.4. Upscaling of TPSCs

Despite the progress in TPSCs in terms of photovoltaic performance and stability, note that most of the performances are evaluated with a small size aperture area of 0.02–0.1 cm². To realize the scalability of all types of solar cells, authorized public test centers recommend evaluating the photovoltaic behavior with an aperture area of 1 cm² or more.^{[ 213 ]} However, because of the difficulty of fabricating THPs over a large aperture area, TPSCs with an aperture area greater than 1 cm² have been very rarely reported. Chowdhury et al. highlighted that only a traditional approach may not be suitable for scaling of TPSCs.^{[ 167 ]} After the antisolvent treatment on the FASnI₃ film, the PDVA process assisted by methylammonium chloride was introduced. The subsequent FASnI₃ films fabricated by this method depicted uniform, continuous, pinhole‐free, and highly crystalline morphology, which leads to a longer PL carrier lifetime over a large aperture area. The respective TPSCs fabricated by this PDVA method showed PCE of 6.33% with J _SC = 19.59 mA cm⁻², V _OC = 0.53 V, and FF = 0.61 with a stable performance after 200 h of storage in an N₂ atmosphere (Figure  17a).^{[ 167 ]} A conjugated organic cation containing a large volume of amines‐3‐phenyl‐2‐propen‐1‐amine (PPA) was tested for the fabrication of cm²‐sized TPSC by Ran et al.^{[ 214 ]} The addition of PPA to FASnI₃ could promote the grain size, passivate the grains, reduce trap density, extract photogenerated charges, and induce preferential orientation of the perovskite film. Consequently, a PCE of 7.08% with photovoltaic parameters of V _OC of 0.56 V, J _SC of 17.57 mA cm⁻², and FF of 0.72 (Figure 17b). Compared with the previously reported 1 cm² TPSC based on the PDVA process, the conjugated PPA introduction within the FASnI₃ framework showed a higher FF due to the better charge management of the TPSC.^{[ 214 ]} In a recent report, instead of adopting the one‐step antisolvent method, a two‐step method was evaluated as a potential route to solve the scaling issue of TPSCs. In this method, the SnI₂ layer with a 10% SnF₂ in DMSO was first deposited by a spin‐coating method followed by the deposition of FAI in a 2‐methyl‐2‐butanol solvent. The implementation of two‐step deposition successfully induced coverage over a large surface area with lower defects and showed PCE of 10.09% with J _SC of 19.96 mA cm⁻², V _OC of 0.77 V, and FF values and 0.667% (Figure 17c).^{[ 215 ]} Compared with the previously reported 1 cm²‐sized PSCs, the TPSCs fabricated by the two‐step method showed a higher V _OC of ≈0.21 V, which was attributed to the steric hindrances of 2‐methyl‐2‐butanol to hydroxyl introduced at the second step. By lowering the crystallization process, a uniform surface morphology over a large surface area was achieved which increased the PL lifetime and influenced the photovoltaic performance.

Figure 17.

Development of 1 cm²‐sized TPSCs—a) PDVA processed first 1 cm² TPSC. Reproduced with permission.^{[ 167 ]} Copyright 2019, Wiley‐VCH. b) Conjugated organic cation based 1 cm² TPSC. Reproduced with permission.^{[ 214 ]} Copyright 2019, Elsevier. c) Two‐step deposition based on highest performance 1 cm² TPSC. Reproduced with permission.^{[ 215 ]} Copyright 2022, American Chemical Society.

4.5. Challenges and Future Perspective

Recent reports show various approaches to overcome some key bottlenecks of the TPSCs development. Nevertheless, obtaining all aspects of an efficient TPSC such as J _SC, V _OC, and FF remains a challenge. The literature review presented above shows that when a high V _OC is achieved, the J _SC remains low and when a favorable J _SC is achieved the V _OC remains low. With such characteristics, the PCE of TPSCs is now in the range of ≈14%, which is significantly lower than the 33% of the theoretical limit. Although numerous researches have shown promising aspects to creating a balance between the photovoltaic parameters by the approaches mentioned above, the real intrinsic mechanism to achieve high photovoltaic performances remains unknown. As can be observed, there is a significant lack of understanding of various THP material properties. Especially for high‐performing FASnI₃ materials, many physical and electrical properties remain unknown. Therefore, apart from TPSC development, future research should be focused on understanding the THP material's optoelectronic properties, chemical behavior, and film properties. Since the complexity of TPSCs fabrication by solution‐processable methods was also highlighted, other alternative fabrication methods must be evaluated.

5. Application of THP for FETs

5.1. Operation Mechanism of FETs

FETs are the fundamental units in microelectronics for amplifying and switching electronic signals through modulating the electric field.^{[ 128} , ^{218 ]} A typical FET is constructed with four main components: gate, dielectric, active layer, and electrodes (Figure  18a). The dielectric layer separates the gate electrode from the active semiconducting layer. The most commonly used dielectric in FETs is silicon dioxide (SiO₂).^{[ 219} , ^{220 ]} However, it shows a high driving voltage because of the low dielectric constant, when the thickness is reduced, the charge carriers penetrate thin layers of SiO₂, resulting in a high gate leakage current, which is a critical fault of device downscaling.^{[ 221 ]} Hence, materials with high dielectric constant may replace SiO₂ in the future for the downscaling of FETs. The active layer is where charge‐carrier transport takes place and determines the field‐effect conductivity of the devices. For the n‐type active layer, amorphous metal oxides, especially indium gallium zinc oxide (IGZO), have succeeded in commercialization for backplane TFT of the large‐area organic light‐emitting display.^{[ 222} , ^{223 ]} For the p‐type active layer, various materials such as silicon (Si) and other inorganic III‐V semiconductors have been developed until now but suffer from the high cost and complexity of fabrication.^{[ 224} , ^{225 ]} Although the search for alternative semiconductors beyond inorganics has taken its place in organic materials with low cost, convenience, and scalability of fabrication, its low carrier transport mobility still limits the performance.^{[ 226} , ^{227 ]} Recently, MHP stood out as a promising semiconductor candidate for FET active layer due to suitable features, such as high charge‐carrier transport mobility and solution processability (Figure 18b).^{[ 228} , ²²⁹ , ²³⁰ , ^{231 ]} The charge‐carrier transport begins when a bias voltage is applied to the gate and metal contacts, which are also called electrodes. Metals that have suitable work functions when matched with the active layer can be used as source and drain electrodes. The distance between two electrodes is called the channel length L and the width is called channel width W.

Figure 18.

Structure and operational mechanism of FETs. a) Cross‐sectional schematic of a typical FET. Reproduced with permission.^{[ 216 ]} Copyright 2016, Wiley‐VCH. b) FET structure of hybrid perovskite FET. Reproduced with permission.^{[ 217 ]} Copyright 2016, Nature Publishing Group. c) Schematic illustrations of the four types of common FET configurations: (from left) bottom‐gate bottom‐contact (BGBC), bottom‐gate top‐contact (BGTC), top‐gate bottom‐contact (TGBC), and top‐gate top‐contact (TGTC). d) Schematic representations of FET operating regions: i) initial state, ii) linear region, iii) depletion region, and iv) saturation region.

As illustrated in Figure 18c, there are typically four types of transistor configurations depending on the positions of contact electrodes and gate: bottom‐contact bottom‐gate (BCBG), bottom‐contact top‐gate, top‐contact top‐gate (TCTG), and top‐contact bottom‐gate.^{[ 232} , ^{233 ]} The FET performance may vary significantly based on its fabricated configuration due to differences in interface contacts, particularly one between the active layer and the contacts, and the other between the active layer and the dielectric layer. Bottom‐gate devices can achieve a good interface between the active layer and the dielectric layer and high‐quality surface. However, the active layer is directly exposed to the environment and susceptible to degradation. In comparison, top‐gate devices can reduce potential degradation from exposure; however, the active layer may be influenced by the deposition of dielectric or contact layers.^{[ 234 ]}

In a typical FET, a bias voltage is applied to the gate and drain electrodes, whereas the source electrode is grounded. Gate voltage (V _g) creates a potential difference between gate and source, which is a driving electric force for charge carriers to accumulate at the channel between source and gate electrode in enhancement mode devices. Drain voltage generates a potential difference between source and drain that induces the accumulated charge carriers at the channel to transport from source to drain. Gate and drain voltage are applied simultaneously and form a controllable current between source and drain. Thus, the magnitude of applied gate voltage can change the carrier density participating in transport through the channel, as well as the total current. The magnitude of applied source–drain voltage affects the charge‐carrier concentration gradient at the channel.^{[ 218 ]}

The operation of FET with polycrystalline or amorphous semiconducting films can be divided into four main regions, as illustrated in Figure 18d: i) The device state before any bias voltage is applied. ii) The applied positive(negative) V _g drives electrons(holes) at the insulator/active layer interface. Some of the induced charge carriers fill in the trap states during the transport, the others are mobile carriers. The carrier density at the channel is related to the magnitude of V _g and capacitance (C) of the insulator. When there are sufficient mobile carriers at the active channel, V _g should be higher than V _Th (V _g − V _Th > 0). If only V _g is applied without source–drain voltage, the charge carriers at the channel are uniformly distributed. iii) When a small source–drain voltage (V _DS << V _g) is applied, a linear gradient of carrier concentration is formed from source to drain electrode. V _DS will be increased to the point of V _DS = V _g − V _Th, where a pinch‐off point exists at the transport channel adjacent to the drain (source) electrode. This is also called the depletion region, at which the difference between V _g and V _DS matches V _Th. At this point, the electric field is not sufficient to induce additional mobile carriers to accumulate at the channel near the drain electrode. iv) When V _DS is further increased (V _DS > V _g − V _Th), the same effect is enhanced, whereas the transport channel length is reduced. This is also called the saturation region, where the carriers can be driven across the narrow depletion zone and form a space‐charge‐limited current. Further boost of V _DS expands the depletion region, which cannot substantially increase source–drain current I _DS and saturates at a certain level called I _D,sat.

Some of the important parameters for quantifying the FET performance include field‐effect mobility (µ _FE), on/off current ratio, threshold voltage (V _Th), and subthreshold swing (SS). µ _FE can explain the carrier transport capacity and can be extracted from the linear region of the transfer curve. On/off current ratio characterizes the ability to control channel current with V _g and can be calculated by the ratio of I _DS at on‐ and off‐state. V _Th describes the interface charge traps and SS shows the ability to switch on and off the device.^{[ 235} , ^{236 ]}

As discussed above, FETs are building blocks for modern electronic technologies. Through the discovery of promising semiconductor materials, innovative device structures, and electric circuit applications, FETs have successfully obtained a highly mature level of fabrication. Among the incumbent and emerging semiconductor materials, THPs offer excellent optoelectronic properties, including low effective mass, high carrier mobility, long charge‐carrier lifetime, and defect tolerance character.^{[ 237} , ²³⁸ , ^{239 ]} These attractive properties enabled the development of high‐performance, highly reliable THP‐based FETs over the last two decades, which will be discussed in the next section.

5.2. Recent Progress in THPs in FETs

5.2.1. 2D THP: (PEA)₂SnI₄

2D layered THPs are highly stable with long‐chain organic bulky ligand in between the inorganic slabs to suppress ion migration and moisture penetration. Additionally, this family of materials is especially attractive for allowing versatile structure and tunable optoelectronic properties through compositional engineering of organic cations.^{[ 241} , ^{242 ]} Because of these advantages, the first‐ever made perovskite transistor was based on 2D THP than any other 3D perovskites or LHPs. Almost two decades ago, Kagan et al. used (PEA)₂SnI₄ as the channel layer for BGBC FET, as illustrated in Figure  19a.^{[ 240 ]} The device showed pure p‐channel characteristics with mobility of µ _h = 0.62 cm² V⁻¹ s⁻¹ and a reasonable on/off current ratio of 10⁴ at room temperature (Figure 19b,c). This work highlighted the potential for fabricating MHP‐based FETs with a cost‐effective solution process. Through spin‐coating, the perovskite layer showed a high orientation in the (001) direction, which favors hole transport in the 2D crystal plane parallel to the device channel (Figure 19d). Additionally, the microstructures of perovskite film, including crystallinity and grain boundaries, are critical for device performance. Mitzi et al. introduced (PEA)₂SnI₄‐based BGBC FETs fabricated through melt process with silicon and polyimide substrates.^{[ 243 ]} The melt‐processed channel layer exhibited considerably enlarged grain size relative to spin‐coated films. Consequently, the fabricated device showed both improved saturation and linear regime hole mobilities of 2.6 and 1.7 cm² V⁻¹ s⁻¹, respectively, at room temperature. Despite the initial success in 1999, research based on (PEA)₂SnI₄ accelerated only in the recent few years encouraged by remarkable advances in solar cells and LEDs. Recently, the performance of 2D THP‐based FETs remarkably developed through precise engineering of film morphology, device structure/interface engineering, crystal formation, and organic spacer synthesis.

Figure 19.

2D (PEA)₂SnI₄‐based FETs. a) Schematic illustration of a BGBC FET structure with a layered (PEA)₂SnI₄ channel. b) Corresponding transfer characteristics at V _D = −100 V. c) Corresponding extracted hole mobility. d) XRD spectrum of a layered (PEA)₂SnI₄ thin film. a–d) Reproduced with permission.^{[ 240 ]} Copyright 1999, American Association for the Advancement of Science. e) Schematic representation of a TGTC FET structure with (PEA)₂SnI₄ semiconducting channel, SAM‐treated substrate, and MoO _x hole‐injection layer. f) Corresponding FET output curves. g) Corresponding transfer characteristics. e–g) Reproduced with permission.^{[ 238 ]} Copyright 2016, Wiley‐VCH.

Matsushima et al. first fabricated (PEA)₂SnI₄ by vacuum deposition on top of octadecyltricholorosilane (OTS)‐treated Si/SiO₂ substrates.^{[ 244} , ^{245 ]} By optimizing the growth rate and temperature of OTS on substrates, the film morphology was much improved and the fabricated FETs exhibited hole mobility of µ _h = 0.78 cm² V⁻¹ s⁻¹ and an on/off current ratio of 10⁵ at room temperature. The same group proposed a significant development in solution‐processed (PEA)₂SnI₄ FETs by applying a series of optimization processes, including a self‐assembled monolayer (SAM)‐treatment on substrates and the addition of MoO _x as a hole‐injection layer between the channel and the electrode (Figure 19e).^{[ 238 ]} Specifically, NH₃I‐SAM treatment induced more well‐developed perovskite crystallites and reduced the amount of residual starting materials and ions. Also, TGTC configuration was formed with Cytop as a dielectric, which reduced resistance for charge carriers to travel through an alternating small‐bandgap inorganic layer and a large‐bandgap organic layer, and diminished hole trap density near the dielectric/perovskite interface. The optimized transistor exhibited high hole mobility of µ _h = 15 cm² V⁻¹ s⁻¹ and a high on/off current ratio of 10⁶ at room temperature (Figure 19f,g). The authors also investigated the intrinsic µ _h of (PEA)₂SnI₄‐based FET by increasing L to reduce the contact resistance (R _c) between the perovskite semiconductor and source/drain electrodes while using the same TCTG device configuration.^{[ 246 ]} With a negligible contribution of R _c relative to the total resistance, the hole mobility increased and saturated to a value of 26 cm² V⁻¹ s⁻¹.

The performance of polycrystalline film devices is often intrinsically limited by grain boundaries, where carrier transport is slowed down or prohibited. To solve this issue, Matsushima et al. fabricated a BGBC transistor with exfoliated crystals of (PEA)₂SnI₄.^{[ 247 ]} Due to the reduced structural disorder and amount of grain boundaries, the fabricated device exhibited tremendously high hole mobility of 40 cm² V⁻¹ s⁻¹ at room temperature. Despite such high performance, the restrictions on reproducing good quality crystals led to less than 1% yield of an operating transistor. Furthermore, their efforts in discovering the unique nature of (PEA)₂SnI₄ expanded to fabricating n‐type TGTC FETs.^{[ 248 ]} By inserting C₆₀ as a buffer layer in between the semiconductor and the low‐work‐function metal (Al) electrodes, the authors presented an n‐type transistor with electron mobility of µ _e = 2.1 cm² V⁻¹ s⁻¹ and an on/off current ratio of 10⁴ at room temperature. This work presented the possibility of ambipolar charge transport with (PEA)₂SnI₄ with a newly designed device structure.

Additionally, interface engineering was often approached to enhance the performance of (PEA)₂SnI₄ transistors. Zhang et al. fabricated a solution‐processed (PEA)₂SnI₄ transistor with PVA and cross‐linking poly(4‐vinylphenol) (CL‐PVP) as the dielectric and ITO as the gate electrode. The optimized device exhibited hole mobility of 0.28 and 0.33 cm² V⁻¹ s⁻¹ in the forward and reverse voltage scan, and an on/off current ratio of 10³ at room temperature.^{[ 249 ]} The negligible hysteresis was highlighted due to the beneficial properties of PVA/CL‐PVP dielectric, including its high quality, good compatibility with perovskite, and suppressed ion migration in perovskite film.

Although various approaches have enhanced the device performance of (PEA)₂SnI₄ FETs, the fragility of the material and other unexplored factors limited the overall operating device yield. Recently, Zhu et al. demonstrated a systematic study from starting materials to film and finally to FETs. The authors provided a set of universal methods for highly reproducible, high‐performance (PEA)₂SnI₄ FET with hole mobility of 3.51 cm² V⁻¹ s⁻¹ and an on/off current ratio of 10⁶.^{[ 131 ]} The methods included self‐passivation of grain boundaries by using excess PEAI, control of grain crystallization by adding Lewis‐base adducts, and passivation of iodide vacancy through oxygen treatment (Figure  20a). The passivation of grain boundaries and enlargement of grain size enabled the reproducibility and reliability of FET performance (Figure 20b). Notably, the authors fabricated the first perovskite‐based complementary inverters with optimized p‐channel (PEA)₂SnI₄ FET and n‐channel IGZO FETs, which showed a high gain of over 30 with an excellent noise margin (Figure 20c). Following this work, a wide range of solvent and additive engineering techniques have been applied to (PEA)₂SnI₄ FETs.^{[ 106} , ¹⁰⁷ , ²⁵⁰ , ^{251 ]}

Figure 20.

A systematic study on 2D (PEA)₂SnI₄ FET development. a) Optimization of starting materials: i) self‐passivation by adding slightly excess PEAI, ii) anti‐oxidation using metallic Sn powder, and iii) effect of halide element. b) Grain crystallization engineering through an adduct approach. c) Device fabrication, trace oxygen doping effect on FETs, and the first perovskite‐IGZO FET‐based complementary inverter. a–c) Reproduced with permission.^{[ 131 ]} Copyright 2020, Wiley‐ICH.

(PEA)₂SnI₄ is also one of the first THPs to be studied as a channel layer for a phototransistor.^{[ 252 ]} Phototransistors are a special type of transistors, as well as being categorized into photodetectors, with the capability to modulate channel conductance through electrical bias and irradiated photons.^{[ 253 ]} THPs are an especially attractive material for phototransistors, owing to their large optical absorption coefficient, long exciton diffusion length, and high charge carrier mobility.^{[ 254} , ²⁵⁵ , ^{256 ]} The detailed mechanism of the operation of halide perovskite phototransistors and photodetectors is thoroughly covered in a few previous review papers.^{[ 33} , ²⁵⁶ , ²⁵⁷ , ^{258 ]} Throughout this work, some of the important achievements made in THP‐based phototransistors will be highlighted for several THPs. Zhu et al. significantly improved the photo detecting ability of (PEA)₂SnI₄‐based phototransistor by incorporating binary solvent of DMF and ethyl acetate (EA).^{[ 250 ]} The antisolvent addition smoothened the film formation through nucleation and consecutive oriented grain ripening, allowing efficient charge transport through full‐coverage film. Through amplified transistor function and photogate properties, the DMF/EA incorporated (PEA)₂SnI₄ phototransistor showed photoresponsivity of 1.6 × 10⁵ A W⁻¹ and remarkably high detectivity of 3.2 × 10¹⁷ Jones.

5.2.2. 2D THPs: Synthesis of New Organic Spacers

Although extensive efforts were made to achieve high‐performance (PEA)₂SnI₄‐based FETs, most of these techniques focused on device engineering rather than the perovskite material itself. One of the earliest attempts in molecular engineering of 2D THPs was by Mitzi et al., who reported a study on tuning the electronic properties of (PEA)₂SnI₄ by fluorination of PEA ligand.^{[ 27 ]} By substituting a hydrogen atom on the phenyl ring of PEA with a fluorine atom, the authors composed m‐fluorophenylethyl ammonium tin iodide (m‐FPEA)₂SnI₄ (m = 2, 3, and 4, representing the position of the fluorine atom) (Figure  21a). The different positions of fluorinated atoms induced subtle changes in Sn—I—Sn bond angle and fabricated BGBC FET hole mobilities: 156.41° and 0.56 cm² V⁻¹ s⁻¹ (m = 4), 154.21° and 0.51 cm² V⁻¹ s⁻¹ (m = 3), and 153.31°, and 0.24 cm² V⁻¹ s⁻¹ (m = 2). (Figure 21b). Although the device performance did not show evident improvement, this study was one of the earliest approaches to associate cation molecular engineering with electronic devices.

Figure 21.

2D organic spacer synthesis and FET fabrication. a) (4‐FPEA)₂SnI₄ crystal structure, viewed along the (011) axis. b) Corresponding BGBC FET transfer characteristics. a,b) Reproduced with permission.^{[ 27 ]} Copyright 2001, American Chemical Society. c) Chemical structure of 4Tm. d) Side‐view of (4Tm)₂SnI₄ crystal structure. e) Corresponding FET transfer characteristics. f) Chemical structure of TT. g) Side‐view of (TT)₂SnI₄ crystal structure. h) Corresponding FET transfer characteristics. i) PL images of (TT)₂SnI₄ thin films after thermal annealing treatment at 155°C for 4 min (top) and 15 min (bottom) (scale bar = 20 µm). j) Schematic representation of carrier transport in small grain (top) and large grain (bottom)‐based (TT)₂SnI₄ thin film. c–j) Reproduced with permission.^{[ 259 ]} Copyright 2021, American Chemical Society.

Recently, Gao et al. proposed a breakthrough in novel organic spacer synthesis for 2D THPs.^{[ 260 ]} The authors modulated the highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital energy levels of the ligands, which altered the overall quantum well structure as well as the charge injection. Inspired by organic semiconductors, polythiophene, for example, the authors featured linear π‐conjugated oligothiophene ligand, 2‐(3‴,4′‐dimethyl‐[2,2′:5′,2″:5″,2‴‐quaterthiophen]‐5‐yl)ethan‐1‐ammonium, also known as 4Tm (Figure 21c). The novel 4Tm ligand exhibited a suitable HOMO level that matched the work function of Au better than PEA or BA and reduced contact resistance to enhance hole injection. Also, 4Tm was incorporated into 2D THP layered structure (4Tm)₂SnI₄ and applied as a p‐channel layer to BGTC FETs, exhibiting hole mobility of µ _h = 2.32 cm² V⁻¹ s⁻¹ (Figure 21d,e). This performance was achieved without additional treatments, which highlighted the excellence of the material itself. Moreover, the incorporation of 4Tm increased grain size up to a micrometer scale, which reduced charge scattering at grain boundaries. Furthermore, the device held remarkable air stability for up to 30 days, in contrast to the survival of pristine (PEA)₂SnI₄‐based FET for less than 1 day. The air stability greatly increased due to the hydrophobic and bulky properties of the π‐conjugated spacer, which can restrict oxygen and moisture penetration. The strong intermolecular interactions between the staggering conjugated 4Tm ligands also increased the intrinsic stability of the material.

The same group extended their studies to synthesizing thienothiophene derivative cation, 2‐(4′‐methyl‐5′‐(5‐(3‐methylthiophen‐2‐yl)thieno[3,2‐b]thiophen2‐yl)‐[2,2′‐bithiophen]‐5yl) ethan‐1‐ammonium, also known as TT (Figure 21f).^{[ 259 ]} Their champion device of (TT)₂SnI₄ exhibited hole mobility of µ _h = 9.35 cm² V⁻¹ s⁻¹ and an on/off current ratio of 10⁶ (Figure 21g,h). The performance enhancement was induced by extended π‐conjugation and increased planarity of the TT ligand, which slowed down the nucleation process and significantly enlarged grain size to almost a millimeter scale (Figure 21i,j). Their variety of works in molecular engineering of organic spacers extended the potential application of highly stable 2D THPs in electronic devices.

5.2.3. 2D/3D Hybrid THPs

From the first works of THPs to the recent extension of studies, 2D THPs demonstrated remarkable enhancement in device performance and air stability. Despite its progress, the fundamental intrinsic limits of 2D perovskites are yet to be solved. Due to the quantum and dielectric confinement effect in 2D perovskites, the charge transport is mostly confined in the corner‐sharing inorganic octahedral cage layer, whereas the charge carrier perpendicular to the inorganic layer is strongly restricted. Contrary to 2D perovskites, 3D perovskites have a higher potential for faster charge transport in absence of the quantum and dielectric confinement effect. However, 3D THPs are notorious for high trap density and high self‐p‐doping. Until the works of Shao et al., there were no reports that used 3D THPs as the semiconducting layer for FETs. The authors successfully reduced the hole density of 3D FASnI₃ by incorporating a small amount of 2D (PEA)₂SnI₄.^{[ 97 ]} The addition of (PEA)₂SnI₄ greatly enhanced the crystallinity and orientation of the 3D perovskite phase, which was critical for FET application (Figure  22a,b). The highly crystalline 2D/3D film exhibited a much smaller amount of trap states than its 3D counterpart. The reduced trap states were particularly Sn vacancies, which are the dominant source of high p‐doping in 3D Sn‐based perovskite (Figure 22c). Through the control of crystallinity and orientation, along with reduced trap density, (PEA)₂SnI₄‐incorporated FASnI₃‐based FET exhibited hole mobility of 0.21 cm² V⁻¹ s⁻¹ and an on/off current ratio of 10⁴ (Figure 22d).

Figure 22.

2D/3D hybrid FASnI₃/(PEA)₂SnI₄ FET. Schematic representation of phase distribution and orientation in a) only 3D film. b) 2D/3D film. Transfer characteristics of BGBC FET with c) only 3D FASnI₃ as an active layer d) 2D/3D hybrid (PEA)₂SnI₄/FASnI₃ as an active layer with optimized thickness. a–d) Reproduced with permission.^{[ 97 ]} Copyright 2021, Wiley‐VCH. e) Schematic illustration of grain morphologies in cross‐sectional view of 2D (top‐left), 3D (top‐right), 2D/3D core–shell (bottom‐left) and 2D/3D mixture (bottom‐right) Sn‐based perovskite thin films. f) Transfer characteristics of 2D/3D core–shell FET with 2D/3D ratio of 1:9 and vacuum treatment. e,f) Reproduced under the terms of the Creative Commons CC‐BY license.^{[ 261 ]} Copyright 2022, The Authors. Published by Wiley‐VCH.

Recently, Kim et al. also reported a study on the incorporation of 2D (PEA)₂SnI₄ into 3D FASnI₃ from an alternative perspective.^{[ 261 ]} The authors presented a new concept of 2D–3D core–shell structure, where the 3D core is fully isolated by the 2D counterpart. The suggested structure provided two major benefits: independent control of V _Th by 2D component and significantly improved grain boundary resistance by the 2D/3D interface. The 2D (PEA)₂SnI₄ FETs suffer from abnormally saturated on‐current caused by large series resistance at the grain boundaries, whereas 3D THP‐based FET exhibits the excessive carrier density that only permits on‐state (Figure 22e). This 2D–3D core–shell simultaneously solves both issues by first controlling the on‐state by 2D component and forming a smoother match between 2D/3D lattices. By adding SnF₂ to facilitate the formation of core–shell structure and vacuum treatment, (PEA)₂SnI₄‐FASnI₃‐based FET showed hole mobility of 25 cm² V⁻¹ s⁻¹ and an on/off current ratio of 10⁶ (Figure 22f). Additionally, the authors combined the optimized p‐channel FET with n‐channel IGZO FET to fabricate CMOS with a gain of over 200. Moreover, Shen et al. proposed a synthesis of 2D incorporated 3D Cs‐based THP crystal. A millimeter‐sized (PEA)₂CsSn₂I₇ crystal was applied as a channel layer to BCBG FET, which exhibited hole mobility of 34 cm² V⁻¹ s⁻¹ at a low temperature of 77 K.^{[ 262 ]}

In another perspective, FASnI₃ itself has high hole concentration and hole mobility due to self p‐doping nature of 3D THP perovskite, which is outstanding to be used as a phototransistor. Recently, Liu et al. developed a high‐performance FASnI₃‐based phototransistor with a high responsivity of ≈10⁵ A W⁻¹ at a low operating voltage.^{[ 263 ]} The authors also incorporated FASnI₃/PEDOT:PSS heterojunction due to the enhanced photogate effect from the heterojunction, reaching a maximum responsivity of 2.6 × 10⁶ A W⁻¹.^{[ 264 ]}

5.2.4. 3D THP: CsSnI₃

Groundbreaking development in p‐type tin perovskite‐based transistors was recently accomplished by our group through developing a 3D CsSnI₃‐based channel layer.^{[ 20 ]} The fabricated champion device achieved a record performance with hole mobility over 50 cm² V⁻¹ s⁻¹ and on/off current ratios exceeding 10⁸ (Figure  23a–e). These values meet the requirements for the backplane of high‐end displays and integrated logic circuits. Additionally, the optimized CsSnI₃‐based TFTs highlighted high operational stability and reproducibility (Figure 23f). Using bias–stress test, the authors discovered a strikingly fast performance recovery within 1 min after the measurement. This unique feature of CsSnI₃‐based TFTs is a drastic improvement from the commercialized p‐type polycrystalline Si or n‐type metal oxide‐based TFTs, which require hour‐long recovery time or extra thermal treatments.^{[ 265 ]} The authors developed a series of methods to accentuate the fast carrier transport and compensate for the fundamental self‐p‐doping issues of 3D THPs. The methods include SnF₂ incorporation in CsI‐rich precursor and a small portion of SnI₂ substitution with PbI₂. SnF₂ incorporation has already demonstrated efficient suppression of Sn²⁺ oxidation and control of crystallization in THPs.^{[ 75} , ⁸⁶ , ^{89 ]} Additionally, CsSnI₃ solution with slight excess CsI improved film morphology, as well as electrical performance. Moreover, due to the lower Lewis acidity of Pb²⁺ compared with Sn²⁺, a small amount of Pb substitution in the precursor slowed down the conversion to the perovskite phase, ultimately improving the film crystalline quality (Figure 23g).^{[ 266} , ²⁶⁷ , ²⁶⁸ , ²⁶⁹ , ^{270 ]} Hall‐effect measurements were conducted to further evaluate the electrical properties of the film. CsI‐rich films demonstrated lower hole concentration but significantly increased Hall mobilities due to improved film uniformity/crystallinity and reduced defect density, such as V _Sn (Figure 23h,i). This series of techniques can be a stepping stone toward the development of high‐performance p‐type THP‐based transistors and complementary electronics.

Figure 23.

3D CsSnI₃‐based FET. a) Schematic illustration of BGTC TFT structure. b) µ _FE and I _on/I _off of CsSnI₃‐based TFTs with different molar ratios of CsI/SnI₂. c) µ _FE and I _on/I _off of CsI‐rich (x = 1.25) precursors with various ratios of Pb substitution. d) Transfer characteristics of optimized CsSnI₃‐based TFT (x = 1.25, 10 mol% Pb substitution, and 7 mol% of SnF₂). e) Corresponding output curves. f) Transfer curves of one optimized CsSnI₃‐based TFT under negative bias–stress measurement for various durations and the recovery behavior at V _GS = V _DS = −40 V. g) XRD patterns of perovskite films deposited from precursors with different ratios of CsI/SnI₂ and that from CsI‐rich (x = 1.25) precursor with 10 mol% Pb substitution and 7 mol% SnF₂. Hall measurement derived parameters, including h) hole concentration of thin films deposited from precursors with different CsI/SnI₂ ratios (without SnF₂) and that from optimized CsI‐rich (x = 1.25), 10 mol% Pb substitution, and 7 mol% SnF₂ precursor and i) corresponding Hall mobility. a–i) Reproduced with permission.^{[ 20 ]} Copyright 2022, Nature Publishing Group.

5.2.5. 3D THP: MASnI₃

The first demonstration of p‐channel perovskite transistors based on 3D MASnI₃ was recently reported as shown in Figure  24a.^{[ 95 ]} By rationalizing the effects of halide (I/Br/Cl) anion engineering on the improvement of film morphology and suppression of tin/iodide vacancy, the authors developed high‐performance hysteresis‐free TFTs with a high hole mobility of 20 cm² V⁻¹ s⁻¹ and on/off ratio exceeding 10⁷ (Figure 24b). The optimized device demonstrated a threshold voltage of 0 V, which is an ideal enhancement mode where no applied bias voltage is required to turn off the transistor. This behavior is a highly desirable trait for simplifying circuit design and minimizing power consumption in electronic applications.^{[ 236} , ^{271 ]} The authors revealed that contrary to previous studies on LHPs, ion migration had a negligible contribution to hysteresis in THP‐based transistors, but minority carrier traps induced by iodine vacancy V _I were the primary cause. The deep electron traps induced by V _I defects were significantly reduced by Br/Cl co‐substitution with strong binding affinities to V _I sites and stabilized the overall lattice structure (Figure 24c–e). Furthermore, operational stability is another crucial figure of merit for practical applications. Through on/off switching stability and bias–stress stability test, the I‐pristine device suffered from serious carrier trapping, whereas the I/Br/Cl optimized device showed much‐improved stability, almost comparable to those of stable organic and amorphous silicon‐based FETs (Figure 24f,g).^{[ 239} , ^{272 ]} By further combining the optimized p‐channel perovskite FETs with n‐channel IGZO FETs, the authors developed monolithically integrated high‐gain complementary inverters and highlighted their high compatibility and processability for electronic applications.

Figure 24.

3D MASnI₃‐based FET: Removal of hysteresis through anion engineering. a) Transistor structure. b) Transfer characteristics with different perovskite channel layers. c) Transfer characteristics of I‐pristine and I/Br/Cl transistors measured at various scan speeds. d) I 3d _3/2 core level XPS spectra. e) Calculated binding affinities of halide anions with V _I‐site in MASnI₃ and MASn(I/Br)₃. f) On/off switching stability of I‐pristine and I/Br/Cl transistors. g) Variation of V _Th under bias–stress. a–g) Reproduced with permission.^{[ 95 ]} Copyright 2022, Nature Publishing Group.

5.2.6. Sn²⁺‐Free THP: Vacancy‐Ordered Double Perovskite

Despite the extensive efforts to improve the air stability of THP film, its application in an inert atmosphere remains limited.^{[ 69} , ^{273 ]} An alternative approach to developing an air‐stable THP device is to modulate the perovskite material composition itself. Vacancy‐ordered halide double perovskites are attractive alternatives for their toxic lead and unstable Sn²⁺‐based counterparts.^{[ 91} , ⁹³ , ^{274 ]} Despite the early discovery of this composition and continuous theoretical studies, the application on electronic devices was rare until a recent report by Liu et al.^{[ 92 ]} The authors demonstrated a synergetic solution process method to modulate the film quality of Cs₂SnI₆ and its feasibility in FETs (Figure  25a,b). The incorporation of slight excess SnI₄ in a mixed DMF/DMSO solvent‐based precursor generated improved crystallinity, uniform morphology, and high electrical properties (Figure 25c). Also, the authors used Mn²⁺ doping to further tune the electrical properties for TFT application and achieved electron mobility of µ _e = 1.2 cm² V⁻¹ s⁻¹ and an on/off current ratio of 10⁴ (Figure 25d,e). Most importantly, Cs₂SnI₆‐based TFT exhibited ambient air stability with reliable operation for more than 1 week, which is an impressive improvement from previous 2D THP‐based TFTs (Figure 25f). The dramatic enhancement in air stability benefited from the attractive chemical compositions of Cs₂SnI₆, including strong Sn—I covalent bonding and a stable Sn tetravalent state.^{[ 275} , ^{276 ]} Thus, the authors highlighted Cs₂SnI₆ as a promising alternative material for high‐performance and high‐stability THP‐based transistors. Furthermore, the authors expanded the potential electronic applications of n‐channel Cs₂SnI₆ TFT by combining it with p‐channel (PEA)₂SnI₄ TFT to develop the first all‐perovskite complementary inverter with a high‐gain voltage over 38 (Figure 25g–i).

Figure 25.

Sn²⁺‐free perovskite FET. a) The crystal lattice structure of vacancy‐ordered double perovskite Cs₂SnI₆. b) TFT structure and corresponding cross‐sectional TEM image. c) X‐ray diffraction peak patterns of Cs₂SnI₆ as a function of the CsI:SnI₄ ratio in the precursor solution. d) Transfer characteristics of Cs₂SnI₆ TFT with various ratios of Mn²⁺ doping. e) Summary of electron mobility µ _e and I _on/I _off of TFTs. f) Normalized µ _FE of Sn²⁺‐based (PEA)₂SnI₄ TFT and Sn(IV)‐based Cs₂SnI₆ TFTs as a function of ambient air exposure time. g) Schematic diagram of CMOS inverter with combined p‐channel (PEA)₂SnI₄ TFT and n‐channel Mn²⁺‐doped Cs₂SnI₆ TFT. Corresponding h) voltage transfer and i) gain voltage curves at different V _DDs. a–i) Reproduced with permission.^{[ 92 ]} Copyright 2022, Cell Press.

5.3. Summary and Future Perspective

From the first‐ever reported (PEA)₂SnI₄‐based transistor to recent success in CsSnI₃‐based transistor, two decades of research in THPs have shown a striking increase in device performance, along with efforts to enhance its stability. Table  6 and Figure  26 illustrate the extension of studies on THP‐based FETs with major milestones from 2D to 3D materials, achieved by multiple research groups. The majority of the reported devices are solution‐processed, and only a few used other methods, such as exfoliated crystals grown from solution or melting process from grown crystals. Only one report has used vacuum deposition to fabricate the THP layer. Each of these methods has its advantages and disadvantages. For example, solution process is a simple and cost‐effective fabrication method, advantageous for industrial applications. However, the solution process has relatively low reproducibility because the quality of the solution has a critical effect on the fabricated film and patterning issue. The recent achievement of high‐performance THP‐based FETs with hole mobility as high as over 50 cm² V⁻¹ s⁻¹ by Liu et al. was also fabricated via a solution process. Through this method, the work highlighted a great potential in the industrial application of THP‐based p‐type transistors, with comparable performance with commercialized n‐type transistors. The next step in the development of THP‐based FETs will be to greatly increase the reproducibility of high performances through vacuum deposition. As shown through the success of the OLED industry, vacuum deposition can fabricate highly reproducible devices in mass production. THP devices may follow the pathway toward commercialization with such high performance and reproducibility.

Table 6.

Summary of device performance of THP‐based transistors

Material	Dielectric/contact	Structure	Deposition/treatment	µ h [cm2 V−1 s−1]	µ e [cm2 V−1 s−1]	I on/I off	T	Ref.
(PEA)2SnI4	SiO2/Au	BGBC	–	0.62		≈104	R.T.	[240]
PEASnI4	Cytop/Au	TCTG	NH3I‐SAM, MoO x HIL	15		≈106	R.T.	[238]
PEASnI4	SiO2/Au	BGTC	Vacuum deposition, OTS‐SAM	0.78		≈105	R.T.	[244]
PEASnI4	Cytop/Al	TCTG	NH3I‐SAM, C60 EIL		2.1	≈104	R.T.	[248]
(PEA)2SnI4	SiO2/Au	BGTC	MoO x HIL	7.9		≈107	R.T.	[135]
(PEA)2SnI4	SiO2/Au	BGBC	Single crystal, MoO x HIL	40		≈106	R.T.	[247]
(PEA)2SnI4	PVA/Cl‐PVP/Au	BGTC	Polymer dielectric	0.28/0.33		102≈103	R.T.	[249]
(PEA)2SnI4/PVP	Cl‐PVP/Au	BGTC	Polymer dielectric	0.31		≈103	R.T.	[277]
(PEA)2SnI4/PVP, PEO	Cl‐PVP/Au	BGTC	Polymer mixed with precursor	0.013	0.0068	≈104	R.T.	[278]
(PEA)2Sn x Pb1‐ x I4	PVA/Cl‐PVP/Au	BGTC	Sn/Pb ratio	0.02 (x = 0.7)		≈102	R.T.	[279]
(PEA)2SnI4/CNT	SiO2/Au	BGTC	CNT mixed with precursor	1.51		≈105	R.T.	[280]
(PEA)2SnI4	SiO2/Au	BGTC	Oxygen treatment	3.51		≈106	R.T.	[131]
(PEA)2SnI4/CB/EA	SiO2/Au	BGTC	Binary solvent	3.8		≈105	R.T.	[250]
(PEA)2SnI4/Urea	SiO2/Au	BGTC	Urea mixed with precursor	4		≈105	R.T.	[251]
(PEA)2SnI4/CuI	SiO2/Au	BGTC	CuI mixed with precursor	2.61		≈106	R.T.	[106]
(4‐MeO‐PEA)2SnI4	SiO2/Polyimide/Au	BGBC	Melt process	2.6		≈106	R.T.	[243]
(m‐FPEA)2SnI4	SiO2/Pd	BGBC	–	0.2 ≈ 0.6		≈105	R.T.	[27]
(PEA)2CsSnI7	SiO2/Cr/Au	BGBC	Single crystal	34		–	77 K	[262]
(4Tm)2SnI4	SiO2/Au	BGTC	–	2.32		≈106	R.T.	[260]
(TT)2SnI4	SiO2/Au	BGTC	–	10		≈106	R.T.	[259]
(STm)2SnI4	SiO2/Au	BGTC	–	1.52		≈106	R.T.	[281]
FASnI3/(PEA)2SnI4	PMMA/Al2O3/Au	TGBC	–	0.21		≈104	R.T.	[97]
FASnI3/(PEA)2SnI4	SiO2/Pt	BGBC	Vacuum treatment	25		≈106	R.T.	[261]
MASnI3	HfO2/Au	BGTC	I/Br/Cl anion engineering	20		≈107	R.T	[95]
Cs2SnI6	SiO2/Au	BGTC	Mn2+ doping		1.2	≈104	R.T.	[92]
CsSnI3	SiO2/Au	BGTC	Small Pb substitution	55		≈108	R.T.	[20]

Channel layers are solution‐processed unless mentioned otherwise, µ _h: Hole mobility, µ _e: Electron mobility, I _on/I _off: On/off current ratio, T: Temperature

Figure 26.

Development of THP‐based FETs over the past two decades.

6. Conclusion

To conclude, we reviewed the structural property relationship, device performance, and stability of several THPs (MASnI₃, FASnX₃, and CsSnX₃) and presented their current issues and perspective opportunities. Thus far, despite the high device performance of the potential of THPs, there has been no report showing the ambient stability for the commercialization level. Therefore, to further improve our standing on device performance and its stability, we must indeed pay attention to overcoming several challenges including i) pure and non‐oxidized initial chemicals and non‐oxidizing solvents, ii) fully understanding of film formation dynamics, iii) precise control of cation and anion defects in the film, iv) suppression of Sn²⁺ oxidation, v) commercial‐grade device operation stability in the air ambient, and vi) better energy band alignment. The vital difference between THPs and its Pb‐based counterparts originated from the instability of Sn²⁺ in 3D structure and the tendency to realize V _Sn. Thus, to aid our discussion, we hence also review the mechanism governing Sn oxidation, as well as the efficient route to reduce Sn⁴⁺ and the encapsulation approach to alleviating the poor stability, especially in ambient air. It is understood that the under‐performance of THP‐based devices is due to the formation of V _Sn; thus, a greater understanding of the Sn oxidation would lead to better‐performing THP‐based devices. To date, numerous techniques have been introduced to reduce such effects, namely, the use of additives to act as reducing agents, control crystallization, partial ion substitution, and reduced dimensionality, and we indeed believe that an important breakthrough will be reported in the near future. Henceforward, finding new, stable, and effective solvents for THP‐based devices, having a deep understanding of the Sn chemistry via film formation, interfacial engineering to better match the energy bands of THPs, and stabilizing Sn²⁺ through device fabrication procedures are greatly important. Alternatively, the implementation of high‐performing devices using vacuum‐deposited THPs can be another breakthrough to fundamentally escape from the issues of various solubility of precursors and additives or stable and effective solvents. Last, although huge progress has recently been made in THP‐based optoelectronics devices, a deep understanding of the photo‐physics and photo‐chemistry processes is hardly obtained. Many approaches have been introduced to improve the performance of those devices including employing reducing agents, morphological control, compositional engineering, device engineering as well as interfacial engineering, but the impact of these approaches on the structural properties and device performance requires more systematic investigation. This will ultimately improve not only device performance but also device stability, which is critical to commercialization.

Conflict of Interest

The authors declare no conflict of interest.

Biographies

Towhid H. Chowdhury is a researcher who works with the Department of Chemical Engineering, Pohang University of Science & Technology (POSTECH). Before joining POSTECH, he used to work with National Institute for Materials Science (NIMS), Japan (2018–2021), and Kyoto University, Japan (2021) as a postdoctoral researcher. He has a keen research interest in the development of low toxic perovskite semiconductors for optoelectronics and electrics applications.

Youjin Reo received her B.S. degree in Chemical Engineering from the Pohang University of Science and Technology (POSTECH) in 2020. She is currently a Ph.D. candidate in the Department of Chemical Engineering at POSTECH under the supervision of Prof. Yong‐Young Noh. Her research interests include the development of perovskite‐based semiconductors through solution‐process and thermal deposition, as well as their applications in transistors and circuits.

Abd. Rashid Mohd Yusoff received his Ph.D from Prof. Ivo Hümmelgen's group in the Departamento de Física at Universidade Federal do Paraná, Brazil. Then, he worked as a Postdoc Fellow, followed by Research Professor in Prof. Jin Jang's group at the Department of Information Display, Kyung Hee University. In 2018, he joined the Sêr SAM group at Swansea University as a Senior Research Fellow with a focus on perovskite photovoltaics until 2020. He is now at Pohang University of Science and Technology as Assistant Research Professor focusing on halide field effect transistor and light emitting diode.

Yong‐Young Noh is chair professor of the Department of Chemical Engineering, Pohang University of Science and Technology (POSTECH). He received Ph.D. in 2005 from GIST and then worked at the Cavendish Laboratory in Cambridge, UK, as a postdoctoral associate. Afterward, he worked at ETRI, as a senior researcher, at Hanbat National University as an assistant professor, and at Dongguk University as an associate professor. His research interest is the field of developing printable semiconductors including organic, carbon nanotubes, perovskite, metal halide and metal oxide for field‐effect transistors, photodiodes, and light‐emitting diodes.

Chowdhury T. H., Reo Y., Yusoff A. R. B. M., Noh Y.‐Y., Sn‐Based Perovskite Halides for Electronic Devices. Adv. Sci. 2022, 9, 2203749. 10.1002/advs.202203749