Recent Advances in Nanostructured Inorganic Hole-Transporting Materials for Perovskite Solar Cells

Organic-inorganic halide perovskite solar cells (PSCs) have received particular attention in the last decade because of the high-power conversion efficiencies (PCEs), facile fabrication route and low cost. However, one of the most crucial obstacles to hindering the commercialization of PSCs is the instability issue, which is mainly caused by the inferior quality of the perovskite films and the poor tolerance of organic hole-transporting layer (HTL) against heat and moisture. Inorganic HTL materials are regarded as promising alternatives to replace organic counterparts for stable PSCs due to the high chemical stability, wide band gap, high light transmittance and low cost. In particular, nanostructure construction is reported to be an effective strategy to boost the hole transfer capability of inorganic HTLs and then enhance the PCEs of PSCs. Herein, the recent advances in the design and fabrication of nanostructured inorganic materials as HTLs for PSCs are reviewed by highlighting the superiority of nanostructured inorganic HTLs over organic counterparts in terms of moisture and heat tolerance, hole transfer capability and light transmittance. Furthermore, several strategies to boost the performance of inorganic HTLs are proposed, including fabrication route design, functional/selectively doping, morphology control, nanocomposite construction, etc. Finally, the challenges and future research directions about nanostructured inorganic HTL-based PSCs are provided and discussed. This review presents helpful guidelines for the design and fabrication of high-efficiency and durable inorganic HTL-based PSCs.


Introduction
Nowadays, the energy supply of our society is mainly reliant on the combustion of fossil fuels, which is limited by the serious CO 2 emissions, inferior energy conversion efficiency and non-renewable nature of fossil fuels [1,2]. Thus, the exploration and utilization of renewable and clean energy resources have received rapidly increasing interest to meet the ever-grown energy demand and resolve environmental issues [3][4][5]. Among various renewable energies, solar energy has attracted extensive attention because of its abundance and clean nature. At present, the utilization of solar energy can be mainly categorized into photovoltaic (solar cells), photocatalysis and photothermal technologies [6][7][8][9][10]. Among them, solar cells are considered the most attractive ways to convert solar energy into electricity directly [11][12][13][14][15][16][17]. Nowadays, the commercialized photovoltaic devices are silicon solar cells, showing high power conversion efficiencies (PCEs) and superb operational stability [18,19]. Nevertheless, silicon solar cells suffer from high production costs, complex fabrication processes and toxic by-products [20]. Thus, it is necessary to develop other alternative solar cells with low cost, high PCE and environmental friendliness [21][22][23]. Perovskite solar cells (PSCs) as one of the third-generation solar cells have received particular In addition to the inferior stability of organic−inorganic halide perovskites under humid and high−temperature conditions, the instability issue hindering the practical applications of PSCs is also mainly caused by the poor heat/moisture tolerance and molecular structural instability of state−of−the−art organic HTLs [40][41][42]. In PSCs, the HTL is a p−type semiconductor to transport holes from the perovskite film to the metal electrode. In addition, HTL also functions as a barrier to block the electron transfer from perovskite film and to avoid the direct contact between the perovskite layer and the electrode in order to reduce carrier recombination [33,43,44]. Therefore, an excellent HTL material should have high hole extraction/mobility and suitable highest occupied molecular orbit (HOMO)/valence band maximum (VBM) with the HOMO of the perovskite layer to achieve a proper energy level alignment [45][46][47]. At present, the most commonly used HTL is 2,2′,7,7′−Tetrakis (N,N−di−p−methoxyphenyl−amine)−9,9′−spirobifluorene (Spiro−OMeTAD), showing suitable energy level alignment with perovskite layer and the metal electrode [48,49]. In addition to the inferior stability of organic-inorganic halide perovskites under humid and high-temperature conditions, the instability issue hindering the practical applications of PSCs is also mainly caused by the poor heat/moisture tolerance and molecular structural instability of state-of-the-art organic HTLs [40][41][42]. In PSCs, the HTL is a p-type semiconductor to transport holes from the perovskite film to the metal electrode. In addition, HTL also functions as a barrier to block the electron transfer from perovskite film and to avoid the direct contact between the perovskite layer and the electrode in order to reduce carrier recombination [33,43,44]. Therefore, an excellent HTL material should have high hole extraction/mobility and suitable highest occupied molecular orbit (HOMO)/valence band maximum (VBM) with the HOMO of the perovskite layer to achieve a proper energy level alignment [45][46][47]. At present, the most commonly used HTL is 2,2 ,7,7 -Tetrakis (N,N-di-p-methoxyphenyl-amine)-9,9 -spirobifluorene (Spiro-OMeTAD), showing suitable energy level alignment with perovskite layer and the metal electrode [48,49]. However, Spiro-OMeTAD must be modified by lithium salts or 4-tert-butylpyridine to exhibit high hole mobility, which unfortunately increases the hygroabsorption capability, leading to inferior moisture stability [50,51]. In addition, most of the widely used organic HTLs for PSCs, including poly (3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS), poly(triarylamine) (PTAA), suffer from the same drawback of poor thermal stability [52,53].
Thus, inorganic HTLs with high intrinsic structural and thermal stability are regarded as highly promising alternatives to replace organic counterparts for stable PSCs [54,55]. Particularly, the nanostructure construction is also reported to effectively enhance the hole transfer capability of inorganic HTLs [56,57]. For instance, a number of nanostructured inorganic HTLs have been designed and developed for PSCs such as metal oxide/sulfide−based nanostructures, including NiOx nanoparticles (NPs), CuIn0.5Ga0.5S2 NPs and Cu2ZnSnS4 NPs, CuSCN NPs, CuGaO2−CuSCN nanocomposite, etc., due to the relatively superb chemical stability, high hole mobility, tunable band gap and low cost [37,[57][58][59]. Figure 2 shows the PCEs evolution of PSCs with nanostructured inorganic HTLs since the first report in 2014. The PCEs of nanostructured inorganic HTL−based PSCs rapidly increased from 9.11% to 21.84% during the past 8 years [60][61][62][63][64][65][66][67]. Nanostructured inorganic HTLs have been widely used in both normal and inverted PSCs. Particularly, nanostructured inorganic HTLs have attracted increasing interest in inverted PSCs due to the excellent light transmission capability and superior surface morphology, which is beneficial for the further deposition of high−quality perovskite films [68,69]. However, the PCEs of nanostructured inorganic HTL−based PSCs are still far away from the cells with organic counterparts, which requires further investigations. In this paper, the recent advances in the design and fabrication of nanostructured inorganic HTLs for PSCs are summarized and discussed by highlighting the prerequisites of HTLs and the superiority of nanostructured inorganic HTLs over organic counterparts. Furthermore, several relevant strategies to boost the performance of inorganic HTLs are presented, including fabrication route design, functional/selectively doping, morphology control, nanocomposite construction, crystal structure regulation and bilayer HTL construction. At last, the remaining problems and future research directions about nanostructured inorganic HTL−based PSCs are also discussed. This review aims to present some insights for the future development and fabrication of high−efficiency and durable inorganic HTL−based PSCs.

The Crucial Roles of HTLs in PSCs
The HTLs play an important role in PSCs to effectively extract holes from the perovskite film and to transport holes to the metal electrode in normal PSCs [70,71]. In addition, HTLs also function as a barrier to prevent the direct contact of the perovskite film and metal electrode, reducing the potential recombination of electrons and holes in normal PSCs [72,73]. Particularly, the HTL also can significantly affect the grain sizes and grain boundary amount of the perovskite layer in inverted PSCs, which strongly influences the PCEs and long-term stability of PSCs [74,75]. Therefore, an ideal HTL should have the following prerequisites [41,57,76,77]. First, HTLs should have high hole mobility to transfer photogenerated holes to the electrode, which is beneficial for the achievement of a high fill factor (FF). It has been reported that when the hole mobility of HTLs was enhanced from 10 −6 to 10 −4 cm 2 V −1 s −1 , FF values of corresponding PSCs can be boosted from 0.22 to 0.80 [78]. Second, HTLs should display superior hydrophobicity, (photo)chemical and thermal stability to enhance the durability of PSCs. Third, the low cost of HTLs is also required to promote large-scale production. Forth, a suitable energy level alignment between HTLs and perovskite film is required to reduce the carrier recombination and the hole accumulation at the perovskite film/HTL interface. Fifth, the high film quality of HTLs is also required to improve the coverage of the perovskite film and to promote the deposition of high-quality perovskite film in the normal and inverted PSCs, respectively. The ideal HTL for PSCs should have no interfacial reaction with other functional layers, such as the perovskite film and the metal electrode [79,80]. The chemical corrosion of metal electrodes on the organic HTLs can be inhibited in inorganic HTL-based PSCs due to the more stable interface between inorganic HTL and the metal electrode [81,82]. Until now, the PCEs of inorganic HTL-based PSCs are still lower than those of organic HTL-based PSCs mainly due to the detrimental interfacial reaction between inorganic HTLs and perovskite film, which reduced the hole extraction capability and increased the carrier recombination rate at interfaces, leading to serious open circuit voltages (V OC ) loss and accelerated decomposition of perovskite film [83,84]. For example, Ni 3+ cations in NiO x HTL can react with the A-site cationic salts in the perovskite precursors, which reduces the amount of Ni 3+ cations in NiO x , leading to lower conductivity of NiO x HTL and reduced PCEs of PSCs [85]. In order to reduce the interfacial reaction between the HTL and the perovskite film, Zhang et al. introduced a 2-methyl-1-aziridinepropionate (SaC-100) layer between NiO x and perovskite film to achieve a physical separation [86]. The introduction of the SaC-100 layer not only effectively inhibited the redox reactions at the HTL/perovskite film interface but also increased the amount of Ni 3+ cations in NiO x HTL, which efficiently reduced the interfacial defect amount and improved the conductivity of NiO x HTL. Consequently, SaC-100 modified cells exhibited a much higher PCE of 20.21% than that of the unmodified PSC (17.54%). In addition, self-assembled monolayers (SAMs) are also widely employed to reduce interfacial reactions in PSCs due to the ultra-thin, uniform and high-quality film of SAMs, which can effectively passivate interfacial defects, improve interface contact and reduce the V OC loss in PSCs [87]. For instance, Mangalam et al. modified the interface between NiO x HTL and perovskite film by using 4-bromophenylphosphonic acid-based SAMs, which increased the V OC from 0.978 V to 1.029 V, effectively improving the PCEs of PSCs [88]. Up to now, organic HTLs based on polymer small molecules, including PEDOT:PSS, Spiro-OMeTAD, etc., have been widely used in PSCs, while inorganic HTLs such as NiO x , CoO, ZnCo 2 O 4 and Fe 3 O 4 have received increasing attention during the past 5 years [39,57,[89][90][91].

Advantages and Disadvantages of Organic HTLs
At present, the most commonly used HTL in PSCs is organic Spiro-OMeTAD, exhibiting relatively high PCEs. Nevertheless, Spiro-OMeTAD suffered from several drawbacks such as high cost, complex synthesis process, poor stability, etc. Furthermore, the PSC with pristine Spiro-OMeTAD delivered a low PCE of 9.7% because of the inferior hole mobility and conductivity induced by the disordered molecule structure of pristine Spiro-OMeTAD [92]. The hole mobility of Spiro-OMeTAD was remarkably enhanced by adding lithium bis(trifluoromethanesulfonyl)imide, 4-tert-butylpyridine and cobalt(III) tris(bis(trifluoromethyl-sulfonyl)imide) to achieve a high PCE of 19.7% [93]. However, the introduction of these additives also brings some unstable factors, which reduce the stability of PSCs [92,94]. For instance, the addition of lithium salts led to inferior moisture stability of PSCs due to the hygroscopic nature, while a detrimental reaction between 4-tert-butylpyridine and halide perovskite accelerated the decomposition of perovskite film Nanomaterials 2022, 12, 2592 5 of 26 during operation [95]. Very recently, Li et al. employed polymethyl methacrylate (PMMA) to modify Spiro-OMeTAD to boost the PCE and moisture stability of PSCs by reducing the interaction between Spiro-OMeTAD/perovskite layers and water [96]. As a result, the cell with PMMA-modified Spiro-OMeTAD produced an attractive PCE of 21.08%, which was much better than the unmodified counterparts. Furthermore, the PCE of the cell with PMMA-modified Spiro-OMeTAD maintained 77% of the initial value after storing in the air with relative humidity (RH) of 40% for 80 days, while the pristine cell deteriorated to 47% of the initial PCE under the same conditions. Besides additive engineering, some researchers have also modified Spiro-OMeTAD by regulating the functional groups in the molecule structure [97,98]. For instance, Jeon et al. tailored the position of methoxy group (para, ortho and meta) in the benzene ring connected with N atom in Spiro-OMeTAD, which significantly affected the electrical and optical properties [97]. It was found that the methoxy group located in the ortho position in Spiro-OMeTAD increased the lowest occupied molecular orbital (LUMO) value of pristine Spiro-OMeTAD while the methoxy group at meta and para positions reduced the LUMO value of pristine Spiro-OMeTAD. Consequently, the cell with ortho-Spiro-OMeTAD exhibited a higher PCE of 16.7% than those of pristine Spiro-OMeTAD (15.2%), meta-Spiro-OMeTAD (13.9%) and para-Spiro-OMeTAD (14.9%) due to the promoted electron blocking and enhanced hole transfer. PEDOT:PSS is a commonly used organic HTL in inverted PSCs due to its superior optical transparency and high conductivity [99,100]. It has been reported that a superb PCE of 20.1% was obtained by using such PEDOT:PSS HTL in inverted PSCs [101]. Nevertheless, the hydrophilic and acidic nature of PEDOT:PSS led to inferior durability of PSCs [102]. Moreover, the V OC of corresponding cells are always less than 0.95 V, which was caused by the energy level mismatch between perovskite film and PEDOT:PSS layer [103]. In summary, although organic HTLs displayed superior hole mobility, the inherent instability under high-temperature and humid conditions remarkably limited further applications in PSCs. Consequently, it is essential to seek intrinsic durable alternative HTLs (e.g., inorganic materials) to achieve high-efficiency and durable PSCs [104][105][106].

The Superiority of Inorganic HTLs and the Importance of Nanostructure Construction
Recently, inorganic materials have emerged as new-generation HTLs to replace organic counterparts due to several distinct advantages such as superior hole collection/migration capability, high conductivity, stability and light transmittance, easy adjustment of energy levels, low cost, simple fabrication, etc. [56][57][58]69]. Until now, several inorganic materials have been utilized as HTLs for PSCs, including NiO x , CuO x , CuS, MoS 2 , VO x , WO 3 , CuSCN, etc. [41,[55][56][57][58][59]. Since the first investigation of inorganic HTLs for PSCs, numerous efforts have been devoted to boosting the PCEs of inorganic HTL-based PSCs [56,64,68,107]. Fortunately, the obtained PCEs of inorganic HTL-based PSCs are now comparable to those of their organic counterparts. In 2014, Wang et al. firstly employed the spin coating method to fabricate NiO x HTL in mesoporous PSCs to obtain a PCE of 9.51% [108]. In addition to the development of fabrication methods, interface engineering and functional doping have also been employed to boost the PCEs of inorganic HTL-based PSCs [109,110]. For instance, Yip et al. modified NiO x /MAPbI 3 interface with diethanolamine (DEA) to improve the charge extraction capability and to reduce the charge recombination rate, and a high PCE of 18.1% was obtained [109]. As for the functional doping, Xiang et al. co-doped NiO x by lithium (Li + ) and magnesium (Mg 2+ ) as HTLs for PSCs [110]. The cell with such co-doped NiO x exhibited a superior PCE of >20%, which was assigned to the improved conductivity and more proper energy level alignment induced by Li + and Mg 2+ doping, respectively. Until now, the solution-based routes are the most widely used methods to fabricate inorganic HTLs for PSCs because of their excellent universality and cost-effectiveness.
Nevertheless, the bulk inorganic HTLs suffered from large particle sizes, which were very difficult to be well dispersed in the solution. As a result, many pinholes and cracks were formed in the solution-processed bulk inorganic HTLs, leading to increased carrier recombination and an inferior coverage on the perovskite layer, which strongly affected the PCEs and stability of corresponding PSCs. On this basis, nanostructure construction is introduced to develop inorganic HTLs with nanoparticles and special morphologies to enhance the PCEs of PSCs. Several promising nanostructured inorganic HTLs such as NiO x NPs, CoO NPs, Fe 3 O 4 NPs, Cr/CuGaO 2 -CC/NiO x nanocomposites, NiCo 2 O 4 NWs, etc., have been designed and developed to boost the PCEs of PSCs due to smaller grain sizes and better dispersion in solution than the bulk counterparts, leading to more compact and flatter hole-transporting films [57,90,91,[111][112][113]. In addition, nanostructured inorganic HTLs not only effectively reduced the reflection loss caused by light scattering and improved the light capturing efficiency but also increased the interfacial contact area between HTL and perovskite film, which significantly improved the carrier extraction rate and suppressed the charge recombination. In this review, the recent advances in the design and development of nanostructured inorganic HTLs for PSCs are summarized and discussed. Several strategies, including fabrication route design, functional/selectively doping, morphology control, nanocomposite construction, crystal regulation and bilayer HTL design, are also proposed, which are discussed in the following section.

Fabrication Route Design
For nanostructured inorganic HTL-based PSCs, the film quality and surface morphology of nanostructured inorganic HTLs play a vital role in achieving high PCEs and stability of PSCs, which are closely related to the fabrication methods of HTLs. Therefore, the design and optimization of the fabrication routes for nanostructured inorganic HTL-based PSCs have received increasing attention. At present, NiO x is one of the most investigated nanostructured inorganic HTLs in PSCs because of its high light transmittance, good chemical stability, wide band gap (3.6-4.0 eV) and proper energy level alignment with perovskite film [65,114]. Furthermore, it has also been reported that the fabrication route for NiO x HTL strongly influences the performance of corresponding PSCs [115]. Generally, nanostructured NiO x HTLs are fabricated by spray pyrolysis, which suffers from high-temperature annealing, leading to inferior cell performance [65,115]. Therefore, numerous efforts have been carried out for the design of fabrication routes for nanostructured NiO x HTLs such as sol-gel, pulsed laser deposition (PLD), solvothermal, etc. [114,115]. For instance, in 2014, Zhu et al. prepared NiO NCs-based HTLs for MAPbI 3 -based PSCs by sol-gel route [60]. It was found that polyhedral NiO nano single crystals were formed on the NiO film surface, enabling an intimate and large interface area connection with MAPbI 3 film, which significantly improved the cell efficiency. More specifically, the uniform NiO NCs also promoted the lead iodide (PbI 2 ) deposition in the two-step perovskite film fabrication processes, while large MAPbI 3 grains were formed to construct a highly textured MAPbI 3 /Methyl [6,6]-phenyl-C 61 -butyrate (PCBM) interface with the help of uniform NiO NCs surface, which remarkably increased the electron-hole diffusion length. In addition, NiO NCs also acted as effective light scattering centers to enhance the sunlight absorption capability of PSCs. Based on the above merits induced by NiO NCs as HTLs, the inverted PSC delivered a PCE of 9.11%. Afterward, Park et al. employed PLD to fabricate ordered nanostructured NiO film, showing superior optical transparency and optimized nanostructure by tailoring the deposition time and oxygen partial pressures [61]. Inverted PSCs with planar heterojunction of MAPbI 3 /PCBM and 30 nm-thick NiO HTL exhibited a superior PCE of 17.3%, which was assigned to the remarkably suppressed recombination rate and improved extraction capability of charge carriers as well as the high light transmittance of 95% of such optimized NiO HTL. In 2018, Tang et al. employed a solvothermal route to synthesize high-crystallinity NiO NCs by using olamine ligands to tailor the nucleation and growth processes of NiO NCs to obtain superb colloidal stability in toluene [116]. The PSC with 55 nm thick NiO NCs-based HTL generated a superior PCE of 15.47% to the cell with conventional organic PEDOT:PSS HTL (11.42%) due to the much-improved light transmittance. Furthermore, the NiO NCs-based unencapsulated cell maintained 65% of the primary PCE after 96 h's storage in ambient conditions at 25 • C, while no PCE was retained for PEDOT:PSS-based counterparts under the same conditions.
In addition to the design of new fabrication methods for nanostructured NiO x HTLs, the modification of the widely used sol-gel process has also been utilized to fabricate high-performance NiO x HTL-based PSCs. For instance, Tong et al. developed a new soft base precipitation method to fabricate NiO x HTLs by replacing NaOH with ammonia in the sol-gel process [117]. The influences of ammonia contents (from 1 to 6, molar ratio) on the crystallization behavior and particle sizes of NiO x NPs were systematically investigated. After optimizing the ammonia content (molar ratio of 1:3), the cell with a configuration of ITO/NiO x NPs/FASnI 3 /PCBM/Bathocuproine (BCP)/Ag generated a superior PCE of~8% due to the improved light transmittance and film quality of NiO x HTL. In another work, Zhang et al. synthesized highly monodispersed NiO x NPs by sol-gel method based on polymer network precipitation (PNP), as depicted in Figure 3a [118]. During the dispersion of NiO x NPs in solution, 1-ethyl-3-methylimidazolium diethyl phosphate ionic liquid (EMDP IL) was added as the surfactant to obtain a highly dispersed NiO x colloid solution. As a result, PSCs with such NiO x -PNP-EMDP IL HTL delivered high PCEs of 20.92% and 19.17% on rigid and flexible substrates, as displayed in Figure 3b,c, which were much higher than the cells with pristine NiO x HTL (18.05% and 16.50%), respectively. Furthermore, the PCE stability of flexible PSCs was also tested, with results shown in Figure 3d. NiO x -PNP-EMDP IL HTL-based device retained 60% of the initial PCE after 5000 bending cycles, while no PCE was retained for the cell with pristine NiO x HTL under the same conditions. Zhang et al.'s study provides a new strategy for the fabrication of flexible and large-area PSC with high-quality NiO x NPs as HTLs.  Besides the above−mentioned fabrication methods, room temperature solution−processing is also regarded as a facile and effective strategy to design and develop nanostructured NiOx HTLs. For instance, Zhang et al. used a room−temperature solution treatment to fabricate defect−free nanostructured NiOx HTLs [62]. The flexible and rigid PSCs with such NiOx film exhibited PCEs of 14.53% and 17.6%, respectively. Chen et al. systematically compared the performance of NiOx NPs obtained by room−temperature (L−NiOx) and high−temperature (H−NiOx) treatments for application as HTLs for mixed Sn−Pb perovskite−based PSCs [119]. It was found that the L−NiOx film displayed a more Besides the above-mentioned fabrication methods, room temperature solution-processing is also regarded as a facile and effective strategy to design and develop nanostructured NiO x HTLs. For instance, Zhang et al. used a room-temperature solution treatment to fabricate defect-free nanostructured NiO x HTLs [62]. The flexible and rigid PSCs with such NiO x film exhibited PCEs of 14.53% and 17.6%, respectively. Chen et al. systematically compared the performance of NiO x NPs obtained by room-temperature (L-NiO x ) and high-temperature (H-NiO x ) treatments for application as HTLs for mixed Sn-Pb perovskite-based PSCs [119]. It was found that the L-NiO x film displayed a more compact/smoother morphology and better energy level matching with Sn-Pb perovskite layer than those of H-NiO x . Consequently, the cell with L-NiO x HTL produced a much higher PCE of 18.77% than the H-NiO x -based device (15.85%) due to the enhanced film quality of HTL and promoted hole extraction from the perovskite layer to HTL. Furthermore, the PCE of the cell with L-NiO x HTL retained 96% of the initial value after storing in the N 2 atmosphere for 50 days, which was assigned to the high-quality perovskite and HTL films, low interfacial defect density and inhibited self-oxidation of Sn 2+ in Sn-Pb perovskites. In another work, Wang et al. various kinds of benzoic acid (BA)-based SAMs to passivate NiO x NPs-based HTLs at HTL/perovskite interface by room temperature treatment as depicted in Figure 4a [120]. Based on the top-view SEM images of perovskite films deposited on NiO x NPs-based HTLs after modification by BA-based SAMs in Figure 4b, 4-bromobenzoic acid (Br-BA) played the best role in the surface defect passivation of perovskite film as evidenced by the largest grain sizes and pinhole-free morphology. As depicted in Figure 4c, the cell with Br-BA modified NiO x HTL displayed a superior PCE of 18.4% with negligible hysteresis effect, which was much higher than that of the cell with pristine NiO x NPs as HTL (15.3%) owing to the better crystallinity, suppressed redox reactions between interfaces, reduced amount of grain boundaries and larger grain sizes of perovskite layer after the modification of Br-BA SAMs. Moreover, the cell with Br-BA modified NiO x HTL showed better stability than the unmodified PSCs after storing in ambient conditions with an RH of 30% for 15 days, as depicted in Figure 4d, which was mainly due to the reduced trap-assisted recombination, increased energy level alignment between NiO x HTL and perovskite film, the tailored surface wettability of HTL and superior quality of perovskite film. than the unmodified PSCs after storing in ambient conditions with an RH of 30% for 15 days, as depicted in Figure 4d, which was mainly due to the reduced trap−assisted recombination, increased energy level alignment between NiOx HTL and perovskite film, the tailored surface wettability of HTL and superior quality of perovskite film. Reproduced from [120], with permission from the ChemSusChem, 2017.

Functional/Selectively Doping
Although the design of advanced fabrication methods can effectively reduce the aggregation of NiOx NPs and/or its dispersion in solution, the low intrinsic conductivity of pristine NiOx is still much lower than the organic counterparts, which was assigned to the large ionization energy of Ni vacancies [3,77]. In order to tackle this problem, functional/selectively doping (e.g., Co 2+ , Ni 3+ , Fe 3+ and Cu + ) has been reported as a facile and effective strategy to boost the conductivity of nanostructured NiOx as HTLs for PSCs. [63,65,[121][122][123] For instance, Kaneko et al. reported the Co doping into NiOx NPs prepared by chemical precipitation at different Co doping concentrations (0-10 mol.%) [121]. It was found that 0.5 mol.% Co−NiOx NPs exhibited the best surface morphology while higher light transmittance was achieved at the Co doping amount of 0.5-2 mol.%. In addition, the conductivity of pristine NiOx was remarkably increased from 3.83 to 6.20 × 10 −6 S cm −1 at a Co doping concentration of 5 mol.% due to the increased amount of high−valence Ni 3+ cations while excessive Co doping amount (10 mol.%) reduced the conductivity of NiOx. A maximum PCE of 14.5% was achieved by the Co−NiOx−based

Functional/Selectively Doping
Although the design of advanced fabrication methods can effectively reduce the aggregation of NiO x NPs and/or its dispersion in solution, the low intrinsic conductivity of pristine NiO x is still much lower than the organic counterparts, which was assigned to the large ionization energy of Ni vacancies [3,77]. In order to tackle this problem, functional/selectively doping (e.g., Co 2+ , Ni 3+ , Fe 3+ and Cu + ) has been reported as a facile and effective strategy to boost the conductivity of nanostructured NiO x as HTLs for PSCs [63,65,[121][122][123]. For instance, Kaneko et al. reported the Co doping into NiO x NPs prepared by chemical precipitation at different Co doping concentrations (0-10 mol.%) [121]. It was found that 0.5 mol.% Co-NiO x NPs exhibited the best surface morphology while higher light transmittance was achieved at the Co doping amount of 0.5-2 mol.%. In addition, the conductivity of pristine NiO x was remarkably increased from 3.83 to 6.20 × 10 −6 S cm −1 at a Co doping concentration of 5 mol.% due to the increased amount of high-valence Ni 3+ cations while excessive Co doping amount (10 mol.%) reduced the conductivity of NiO x . A maximum PCE of 14.5% was achieved by the Co-NiO x -based PSC at a Co doping amount of 1 mol.%, which may be assigned to the balance between surface morphology, light transmittance and conductivity of Co-NiO x HTL. Similar results were also reported by Chandrasekhar et al. about the development of Fe doped NiO x HTLs at different concentrations of 0-2 mol.% [122]. The cell with a configuration of ITO/Fe-NiO x /MAPbI 3 /PCBM/BCP/Ag at Fe doping amount of 1 mol.% exhibited the highest PCE of 17.59%, 14% higher than that of the cell with NiO x HTL, which was mainly attributed to the increased conductivity and work function of NiO x induced by Fe doping at a proper concentration.
In another work, He et al. systematically investigated the influences of Cu doping amount on the film quality, light transmittance and hole mobility of NiO x NPs as HTLs for PSCs [63]. As depicted in Figures 5 and 6a,b, the optimized 2 mol.% Cu doped NiO x HTL displayed the best surface morphology and light transmittance (>80%) as well as increased work function, which was beneficial for the hole transporting between perovskite film and the HTL. Furthermore, the 2 mol.% Cu doping also significantly improved the hole mobility of NiO x from 3.05 × 10 −3 to 1.05 × 10 −2 cm 2 V −1 s −1 . Consequently, the 2 mol.% Cu doped NiO x -based PSCs with active areas of 0.1 and 1.08 cm 2 exhibited higher PCEs of 18.66% and 15.10% than those of NiO x NPs-based PSCs (15.47% and 11.51%, respectively) as depicted in Figure 6c,d. In addition, the 2 mol.% Cu doped NiO x -based flexible cells also displayed excellent PCE stability for 500 bending cycles ( Figure 6e). As depicted in Figure 6f, it is found that 2 mol.% Cu doped NiO x -based PSC maintained 86% of the initial PCE after 30 days' storage under ambient conditions with an RH of 40% at 25 • C, while the undoped NiO x -based PSC only retained 75% of the primary PCE under the same conditions. More importantly, no PCE was retained for the PEDOT:PSS-based PSC after 7 days' storage as depicted in Figure 6f, highlighting the crucial roles of Cu-doped NiO x HTL in boosting the cell stability. Similarly, Chen et al. prepared HTL films by spin-coating Cu-doped NiO NPs-based ink at room temperature to achieve a high PCE of > 20% without any post-treatment, which was mainly due to the co-existence of Cu + and Cu 2+ cations after Cu doping in NiO to improve the hole extraction [123]. In another work, He et al. systematically investigated the influences of Cu doping amount on the film quality, light transmittance and hole mobility of NiOx NPs as HTLs for PSCs [63]. As depicted in Figures 5 and 6a,b, the optimized 2 mol.% Cu doped NiOx HTL displayed the best surface morphology and light transmittance (>80%) as well as increased work function, which was beneficial for the hole transporting between perovskite film and the HTL. Furthermore, the 2 mol.% Cu doping also significantly improved the hole mobility of NiOx from 3.05 × 10 −3 to 1.05 × 10 −2 cm 2 V −1 s −1 . Consequently, the 2 mol.% Cu doped NiOx−based PSCs with active areas of 0.1 and 1.08 cm 2 exhibited higher PCEs of 18.66% and 15.10% than those of NiOx NPs−based PSCs (15.47% and 11.51%, respectively) as depicted in Figure 6c,d. In addition, the 2 mol.% Cu doped NiOx−based flexible cells also displayed excellent PCE stability for 500 bending cycles ( Figure 6e). As depicted in Figure 6f, it is found that 2 mol.% Cu doped NiOx−based PSC maintained 86% of the initial PCE after 30 days' storage under ambient conditions with an RH of 40% at 25 ℃, while the undoped NiOx−based PSC only retained 75% of the primary PCE under the same conditions. More importantly, no PCE was retained for the PE-DOT:PSS−based PSC after 7 days' storage as depicted in Figure 6f, highlighting the crucial roles of Cu−doped NiOx HTL in boosting the cell stability. Similarly, Chen et al. prepared HTL films by spin−coating Cu−doped NiO NPs−based ink at room temperature to achieve a high PCE of > 20% without any post−treatment, which was mainly due to the co−existence of Cu + and Cu 2+ cations after Cu doping in NiO to improve the hole extraction [123].   Besides monovalent and divalent cations, the doping of trivalent cations can also be used to remarkably boost the conductivity of NiOx, especially for the trivalent lanthanide cations with partially filled 4f orbitals and empty 5d orbitals [124]. For instance, Bao et al. synthesized samarium−doped nickel oxide (Sm:NiOx) NPs by a chemical precipitation route and deposited them as HTLs for PSCs [124]. It was found that the Sm 3+ doping suppressed the formation energy of Ni vacancies and increased the amount of Ni vacancies to increase the hole density. Thus, the conductivity of NiOx was improved remarkably by Sm 3+ doping while the work function was enlarged in Sm:NiOx HTL, which promoted the hole extraction and reduced the charge recombination at the HTL/perovskite interface. Thus, the cell with Sm:NiOx HTL exhibited a superior PCE of 19.16% to that of the NiOx−based cell (17.23%). After optimizing the compositions of perovskite film, high PCEs of 20.71% and 17.95% were achieved for rigid and flexible PSCs based on Sm:NiOx HTL, respectively. Particularly, Sm:NiOx also showed promises for application in large−area PSCs. Sm:NiOx−based PSC with active areas of 1 and 16 cm 2 produced attractive PCEs of 18.51% and 15.27%, respectively, which were among the highest PCEs of NiOx−based PSCs with large areas. Bao et al.'s work emphasizes a new doping strategy for the development of high−quality NiOx HTLs in conjunction with revealing the doping effect in depth, which can provide some guidelines for the fabrication of high−efficiency and flexible PSCs in the future. As mentioned above, high−valence Ni 3+ cations are beneficial for the achievement of the high conductivity of NiO HTL [125]. Thakur et al. synthesized Ni 3+ doped NiO NWs (Ni 3+ −NiO NWs) by hydrothermal method, as depicted in Figure 7a [125]. Firstly, the precursor solution composed of urea, chloroacetic acid, deionized water and hydrochloric acid is prepared and various concentrations of NiCl2·6H2O (1, 3 and 5 mM) are then dissolved in the precursor solution. Then, FTO/NiO was immersed into the precursor solution, which was further hydro- Besides monovalent and divalent cations, the doping of trivalent cations can also be used to remarkably boost the conductivity of NiO x , especially for the trivalent lanthanide cations with partially filled 4f orbitals and empty 5d orbitals [124]. For instance, Bao et al. synthesized samarium-doped nickel oxide (Sm:NiO x ) NPs by a chemical precipitation route and deposited them as HTLs for PSCs [124]. It was found that the Sm 3+ doping suppressed the formation energy of Ni vacancies and increased the amount of Ni vacancies to increase the hole density. Thus, the conductivity of NiO x was improved remarkably by Sm 3+ doping while the work function was enlarged in Sm:NiO x HTL, which promoted the hole extraction and reduced the charge recombination at the HTL/perovskite interface. Thus, the cell with Sm:NiO x HTL exhibited a superior PCE of 19.16% to that of the NiO x -based cell (17.23%). After optimizing the compositions of perovskite film, high PCEs of 20.71% and 17.95% were achieved for rigid and flexible PSCs based on Sm:NiO x HTL, respectively. Particularly, Sm:NiO x also showed promises for application in large-area PSCs. Sm:NiO xbased PSC with active areas of 1 and 16 cm 2 produced attractive PCEs of 18.51% and 15.27%, respectively, which were among the highest PCEs of NiO x -based PSCs with large areas. Bao et al.'s work emphasizes a new doping strategy for the development of high-quality NiO x HTLs in conjunction with revealing the doping effect in depth, which can provide some guidelines for the fabrication of high-efficiency and flexible PSCs in the future. As mentioned above, high-valence Ni 3+ cations are beneficial for the achievement of the high conductivity of NiO HTL [125]. Thakur et al. synthesized Ni 3+ doped NiO NWs (Ni 3+ -NiO NWs) by hydrothermal method, as depicted in Figure 7a [125]. Firstly, the precursor solution composed of urea, chloroacetic acid, deionized water and hydrochloric acid is prepared and various concentrations of NiCl 2 ·6H 2 O (1, 3 and 5 mM) are then dissolved in the precursor solution. Then, FTO/NiO was immersed into the precursor solution, which was further hydrothermally treated at 200 • C for 0.5 h to realize the hydrothermal growth of NPs. NiCl 2 was hydrolyzed into Ni 3+ -rich Ni (OH) 2 by reducing chemicals including ammonium cyanate, ammonia and cyanic acid decomposed by urea, which was further annealed at 500 • C to obtain Ni 3+ -NiO NWs. After optimizing the NiCl 2 ·6H 2 O concentration (1 mM), Ni 3+ -NiO NWs HTL displayed the best surface morphology. In addition, the VBM value of Ni 3+ -NiO NWs (0.83 eV) was smaller than that of pristine NiO film (0.88 eV), as depicted in Figure 7b,c due to the porous structure of Ni 3+ -NiO NWs with Ni atoms exposed at the edge, which effectively promoted the hole transport. As a result, the cell with Ni 3+ -NiO NWs HTL produced an excellent PCE of 17.75%, which was much better than the pristine NiO-based cell (11.5%), as depicted in Figure 7d.  Figure 7b,c due to the porous structure of Ni 3+ −NiO NWs with Ni atoms exposed at the edge, which effectively promoted the hole transport. As a result, the cell with Ni 3+ −NiO NWs HTL produced an excellent PCE of 17.75%, which was much better than the pristine NiO−based cell (11.5%), as depicted in Figure 7d. In addition to NiOx, Sb2S3 is also regarded as a promising hole−transporting material for PSCs because of the high carrier density, superior hole mobility, wide band gap, etc. For instance, Mohamadkhan et al. synthesized Sb2S3 and Cu doped Sb2S3 (Cu3SbS4) NCs by an oil amine−assisted heating method [126]. Cu3SbS4 NCs−based PSCs exhibited a higher PCE of 13.0% that of the Sb2S3−based cell (8.2%) due to the increased conductivity and carrier density, improved charge transport/extraction capability of Cu3SbS4 NCs induced by Cu doping. Azam et al. systematically investigated the influences of Ga doping In addition to NiO x , Sb 2 S 3 is also regarded as a promising hole-transporting material for PSCs because of the high carrier density, superior hole mobility, wide band gap, etc. For instance, Mohamadkhan et al. synthesized Sb 2 S 3 and Cu doped Sb 2 S 3 (Cu 3 SbS 4 ) NCs by an oil amine-assisted heating method [126]. Cu 3 SbS 4 NCs-based PSCs exhibited a higher PCE of 13.0% that of the Sb 2 S 3 -based cell (8.2%) due to the increased conductivity and carrier density, improved charge transport/extraction capability of Cu 3 SbS 4 NCs induced by Cu doping. Azam et al. systematically investigated the influences of Ga doping amount on the performance of Cu(In 1−x Ga x )S 2 (CIGS x , x = 0-0.75) NPs as inorganic HTLs for PSCs [127]. It was found that with the increase in Ga doping concentration, CIGS displayed a decreased valence band edge and an increased conduction band edge, which improved the electron-blocking capability and suppressed the carrier recombination rate. The cell with optimized CIGS 0.5 HTL showed an attractive PCE of 15.6%, which was much better than the cell with pristine CuInS 2 (10.85%). In addition, the cell with CIGS 0.5 HTL retained 70% of the primary PCE after aging in the dark conditions with an RH of 50% for 90 days, while the cell with Spiro-OMeTAD HTL only retained <50% of the original PCE under the same conditions.

Morphology Control
It is well known that the morphologies of inorganic HTLs as substrates for the deposition of perovskite film have significant impacts on the quality and grain sizes of perovskite film, which further influences the photovoltaic performance of PSCs. On this basis, some advanced morphologies of inorganic HTLs, including NPs, NWs, nanorods (NRs) and nanosheets (NSs), have been synthesized by controlling the fabrication parameters and designing new preparation methods. For instance, Liu et al. prepared NiO NSs by a solution extraction method at room temperature for the application as HTLs for carbon electrode-based PSCs [128]. It was found that NiO NSs-based PSC produced a high PCE of 14.2%, which was 15% higher than that of NiO NPs-based counterparts due to the improved hole collection efficiency, longer charge lifetime and better contact with perovskite film. Remya et al. prepared WO 3 nanostructures with different morphologies, including WO 3 NPs and WO 3 NSs, by a solution treatment route, which were employed as alternatives to organic HTLs in PSCs [129]. It was found that the cell with WO 3 NSs HTL exhibited a higher PCE of 7.76% than that of WO 3 NPs-based counterparts (6.83%) due to the enhanced hole collection efficiency and hole mobility induced by the NSs-like morphology.
Recently, Ramachandran et al. employed a simple electrochemical deposition method to fabricate nanostructured CuSCN with various morphologies (e.g., NRs and NWs) as HTLs for carbon-based PSCs by tailoring the applied potential and deposition time as depicted in Figure 8a-c [130]. It was found that the applied potential strongly affected the morphologies of CuSCN while nanoplates, NWs and NRs, were formed at an applied potential of −350, −400 and −450 mV. The cell with CuSCN NRs (−450 mV) as HTL exhibited a superior PCE of 12.42% with negligible hysteresis as depicted in Figure 8d, which was much higher than those of CuSCN NWs (10.99%) and CuSCN nanoplates (8.40%) due to the unique morphology, enhanced hole mobility and better energy level alignment. Furthermore, the cell with CuSCN NRs as HTL exhibited much superior stability to those of CuSCN NWs and CuSCN nanoplates under ambient conditions with an RH of >60% at room temperature, as depicted in Figure 8e.
Besides the precisely control of applied potential, Ramachandran et al. also optimized the electrochemical deposition method to further boost the PCEs of CuSCN HTL-based PSCs by adding tritionx-100 as surfactants to prepare CuSCN NWs with different morphologies, including void-free NWs and core-shell NWs, as depicted in Figure 9 [131]. It was found that core-shell CuSCN NWs-based PSCs fabricated at an applied potential of −400 mV and a deposition time of 3 min generated a superior PCE of 16.99% with negligible hysteresis, which was much higher than the cell with void-free CuSCN NWs as HTL (10.57%) due to the fact that core-shell CuSCN NRs provided fast channels for charge transfer and suppressed the charge recombination rate. Furthermore, the cell with core-shell CuSCN NRs as HTL maintained 98% of the primary PCE under ambient conditions with an RH of >60%, while CuSCN only retained 87% of the initial PCE under the same conditions.  In addition to the morphology control of simple metal oxides (e.g., NiO, WO3) and CuSCN as inorganic HTLs for PSCs, Li et al. also synthesized mesoporous complex metal oxides such as NiCo2O4 nanostructures with different morphologies as HTLs for PSCs by the simple hydrothermal method through tailoring the reaction time [132]. As depicted in Figure 10a-g, NiCo2O4 NSs, NiCo2O4 NS−NWs and NiCo2O4 NWs were obtained by hydrothermal method at reaction time 90, 120 and 150 min, respectively. It was found that  In addition to the morphology control of simple metal oxides (e.g., NiO, WO3) and CuSCN as inorganic HTLs for PSCs, Li et al. also synthesized mesoporous complex metal oxides such as NiCo2O4 nanostructures with different morphologies as HTLs for PSCs by the simple hydrothermal method through tailoring the reaction time [132]. As depicted in Figure 10a-g, NiCo2O4 NSs, NiCo2O4 NS−NWs and NiCo2O4 NWs were obtained by hydrothermal method at reaction time 90, 120 and 150 min, respectively. It was found that In addition to the morphology control of simple metal oxides (e.g., NiO, WO 3 ) and CuSCN as inorganic HTLs for PSCs, Li et al. also synthesized mesoporous complex metal oxides such as NiCo 2 O 4 nanostructures with different morphologies as HTLs for PSCs by the simple hydrothermal method through tailoring the reaction time [132]. As depicted in Figure 10a- (Figure 10h,i), which was assigned to the higher conductivity and hole mobility, promoted charge extraction and hole transport capability to reduce the interfacial carrier recombination. In addition, the one-dimensional (1D) structure of NiCo 2 O 4 NWs provided additional channels for charge transport and suppressed the charge recombination at perovskite/HTL interface. However, it should be noted that the short-circuit current density (J sc ) value of NiCo 2 O 4 NWs-based cell was lower than the cells based on NiCo 2 O 4 NSs, NiCo 2 O 4 NS-NWs HTLs, which was attributed to the larger thickness of NiCo 2 O 4 NWs film due to the increased hydrothermal reaction time, leading to a reduced light transmittance. Thus, the trade-off between light transmittance and high-performance morphologies should be further investigated to boost the PCEs of NiCo 2 O 4 HTL-based PSCs.
Nanomaterials 2022, 12, x FOR PEER REVIEW 15 kinds of NiCo2O4 with different morphologies (Figure 10h,i), which was assigned to higher conductivity and hole mobility, promoted charge extraction and hole trans capability to reduce the interfacial carrier recombination. In addition, one−dimensional (1D) structure of NiCo2O4 NWs provided additional channels charge transport and suppressed the charge recombination at perovskite/HTL interf However, it should be noted that the short−circuit current density (Jsc) value of NiC NWs−based cell was lower than the cells based on NiCo2O4 NSs, NiCo2O4 NS−N HTLs, which was attributed to the larger thickness of NiCo2O4 NWs film due to the creased hydrothermal reaction time, leading to a reduced light transmittance. Thus, trade−off between light transmittance and high−performance morphologies should further investigated to boost the PCEs of NiCo2O4 HTL−based PSCs.

Nanocomposite Construction
As high−performance HTLs for PSCs, single−phase inorganic materials may fully meet the requirements of HTL, including high conductivity and hole mobility, perior light transmittance and excellent energy level alignment with other lay

Nanocomposite Construction
As high-performance HTLs for PSCs, single-phase inorganic materials may not fully meet the requirements of HTL, including high conductivity and hole mobility, superior light transmittance and excellent energy level alignment with other layers. [111,115] Thus, nanocomposite construction is regarded as an effective strategy to tackle the abovementioned problem to achieve high-efficiency and durable inorganic HTL-based PSCs. [133] Up to now, various nanocomposites have been designed and developed as HTLs for PSCs, such as metal/metal iodide, oxide/CuSCN, oxide/metal selenide, metal oxide/oxide and oxide/carbonate, etc. [66,67,[133][134][135][136][137]. For instance, Cao et al. synthesized Cu@CuI nanocomposites by an in situ iodination route by using Cu NWs/FTO as the precursors [133]. More specifically, Cu NWs were firstly spin-coated on the FTO substrate, which was further treated in I 2 /ethanol solution for 3 s as depicted in Figure 11a,b. The Cu@CuI nanocomposite HTL-based PSC (Figure 11c) exhibited a superior PCE of 18.8% to those of the cells with CuI (17.5%) and PEDOT:PSS (16.9%) HTLs as depicted in Figure 11d, which was assigned to the boosted extraction and transfer capability of charge carriers to suppress the interfacial recombination. As depicted in Figure  NWs/FTO as the precursors [133]. More specifically, Cu NWs were firstly spin−coated on the FTO substrate, which was further treated in I2/ethanol solution for 3 s as depicted in Figure 11a,b. The Cu@CuI nanocomposite HTL−based PSC (Figure 11c) exhibited a superior PCE of 18.8% to those of the cells with CuI (17.5%) and PEDOT:PSS (16.9%) HTLs as depicted in Figure 11d, which was assigned to the boosted extraction and transfer capability of charge carriers to suppress the interfacial recombination. As depicted in Figure   In addition to metal/metal iodide, oxide/CuSCN and oxide/metal selenide nanocomposites have also been designed and developed as inorganic HTLs for PSCs [134][135][136]. For instance, Lee et al. prepared CuCaO2 NPs/CuSCN as the nanocomposite HTL to boost the PCE and stability of PSCs [134]. Firstly, CuCaO2 NPs were modified by amino silane to reduce the accumulation to obtain a uniform NPs−based film. After this, CuSCN solution was then coated on CuCaO2 NPs−based film while CuSCN was penetrated into CuCaO2 NPs−based film, which was beneficial to reducing the roughness of CuCaO2 NPs−based film and improving the hole extraction capability of CuCaO2 NPs−based HTL. As a result, the cell with CuCaO2 NPs/CuSCN nanocomposite HTL generated a superb PCE of 16.7%, which was 8% higher than that of the CuSCN HTL−based cell due In addition to metal/metal iodide, oxide/CuSCN and oxide/metal selenide nanocomposites have also been designed and developed as inorganic HTLs for PSCs [134][135][136]. For instance, Lee et al. prepared CuCaO 2 NPs/CuSCN as the nanocomposite HTL to boost the PCE and stability of PSCs [134]. Firstly, CuCaO 2 NPs were modified by amino silane to reduce the accumulation to obtain a uniform NPs-based film. After this, CuSCN solution was then coated on CuCaO 2 NPs-based film while CuSCN was penetrated into CuCaO 2 NPs-based film, which was beneficial to reducing the roughness of CuCaO 2 NPs-based film and improving the hole extraction capability of CuCaO 2 NPs-based HTL. As a result, the cell with CuCaO 2 NPs/CuSCN nanocomposite HTL generated a superb PCE of 16.7%, which was 8% higher than that of the CuSCN HTL-based cell due to the remarkably reduced trap density (40%). Furthermore, the thermal and moisture stability of CuCaO 2 /CuSCN nanocomposite HTL-based device has also significantly improved. The PCE of the CuCaO 2 /CuSCN-based cell still maintained 85% of its original PCE after 400 h's storage at 85 • C and 85% RH. More recently, Kim et al. successfully synthesized Cu 2 O-CuSCN nanocomposite by using a facile one-step deposition route to boost the PCE and stability of CuSCN-based PSCs [135]. The Cu 2 O-CuSCN nanocomposite-based cell generated a higher PCE of 19.2% than that of the CuSCN-based device (17.7%) due to the improved hole mobility and significantly reduced trapping sites. The Cu 2 O-CuSCN nanocomposite-based PSC still retained 90% of the initial PCE after storing in harsh conditions (85 • C and 85% RH) for 720 h due to the much-reduced interface degradation between perovskite layer and CuSCN HTL. Ramachandran et al. fabricated various CuSCN/CuI nanocomposite-based HTLs by electrodeposition method through tailoring the electrochemical potential and deposition time for carbon-based PSCs [136]. The cell based on the optimized CuSCN/CuI nanocomposite HTL with a thickness of 142 nm (electrochemical potential of −400 mV and deposition time of 2 min) presented a superb PCE of 18.82%, which was much higher than the CuSCN HTL-based device (16.66%) due to the enhanced light transmittance, improved surface morphology and enlarged crystal sizes of PSCs.
The rational design of nanocomposite HTL may also extend the light absorption range of PSCs from 300-800 nm to 800-1000 nm to remarkably boost the PCEs of PSCs [66]. In 2020, Davood et al. designed core-shell structured NiO@GeSe (25 nm for core and 50 nm for shell) nanocomposite as HTLs for PSCs [66]. As the shell, GeSe displayed an absorption spectrum of 800-1000 nm, which was beneficial to extending the light absorption range of pristine PSCs. As the core, NiO provided fast channels for hole transport to improve the hole mobility and reduce the carrier recombination. Consequently, the NiO@GeSe HTL-based cell with optimized film thickness of 120 nm produced an excellent PCE of 20.29%, which was 40% higher than the cell with GeSe NRs HTL (14.53%). In addition to NiO@GeSe core-shell HTL, oxide/oxide core-shell HTLs have also been constructed to boost the PCEs of NiO-based PSCs. Farzaneh et al. synthesized core-shell CuO@NiO nanospheres by a solvothermal route for the application as HTLs for PSCs [137]. The core-shell CuO@NiO-based cell generated a higher PCE of 10.11% than that of the pristine NiO-based device (8.58%), which was mainly due to the better coverage of CuO NPs over NiO nanospheres to reduce the cracks of NiO nanospheres, improved conductivity and the reduced non-radiative recombination at perovskite film/HTL interface. Very recently, Ge et al. synthesized Co 3 O 4 -SrCO 3 nanocomposite by heat treatment as HTLs for PSCs, as depicted in Figure 12a [67]. As depicted in Figure 12b,c, in such nanocomposite, Co 3 O 4 with a narrow band gap was used to improve the hole migration rate while SrCO 3 with a wide band gap was employed to selectively block the electrons and reduce carrier recombination. Furthermore, a cross-gap type I heterojunction was formed in Co 3 O 4 -SrCO 3 nanocomposite with uniform and compact morphology, which effectively enhanced the conductivity and hole mobility of HTL as well as reduced the charge recombination between the perovskite film and HTL. Consequently, MAPbI 3 -based PSC with Co 3 O 4 -SrCO 3 nanocomposite HTL produced a superb PCE of 21.84%, which was much higher than the cells with Co 3 O 4 (15.47%), SrCO 3 (8.08%) and NiO x (18.41%) as HTLs (Figure 12d,e). As depicted in Figure 12f

Crystal Structure Regulation and Bilayer HTL Design
In addition to the above−mentioned strategies for the design and fabrication of inorganic HTLs, crystal structure regulation and bilayer HTL design have also been reported to boost the PCE and stability of inorganic HTL−based PSCs, although the relevant literature is limited until now. For instance, Heidariramsheh et al. successfully synthesized Cu2SnS3 (CTS) NPs with different crystal structures (zincblende and wurtzite) as HTLs for PSCs [138]. More specifically, zincblende CTS and wurtzite CTS were synthesized by the oil amine solvent heating method using sulfur and thiourea as sulfur sources, respectively. The CTS NPs were then dispersed in non−polar chloroform solution to obtain CTS inks, which were spin−coated on perovskite layer as HTL. As depicted in Figure 13, Spiro−OMeTAD HTL displayed has the smoothest surface while zincblende−CTS exhibited the roughest surface, which was due to the NPs agglomeration in the HTL inks. In addition, the wurtzite−CTS and zincblende−CTS displayed much better hydrophobic capability than that of Spiro−OMeTAD due to the addition of lithium salts in Spiro−OMeTAD, which may be beneficial for the achievement of high moisture tolerance. The cell with wurtzite−CTS HTL exhibited a higher PCE of 13.1% than that of zincblende−CTS−based cell (7.87%), which was assigned to the more uniform and flattened wurtzite−CTS film, monodispersed wurtzite−CTS NPs and better energy level alignment. As for the moisture stability, the cell with wurtzite−CTS HTL retained ~90% of the primary PCE after 1200 h's storage under ambient conditions with an RH of >45%, which was much better than that of Spiro−OMeTAD−based cell with a PCE retention ratio of ~60% under the same conditions due to the remarkably enhanced hydrophobic capability of wurtzite−CTS HTL.
As mentioned above, CuSCN and CuI have been extensive as inorganic HTLs for PSCs [133][134][135]. CuSCN displayed a wide band gap of 3.4-3.9 eV and low hole mobility of 0.1 cm 2 V −1 s −1 , while CuI exhibited a narrow band gap of 3.1 eV and superior hole mobility of 40 cm 2 V −1 s −1 . [133,135] Thus, it is crucial to fully utilize the merits of CuSCN and CuI to further boost the PCEs of inorganic HTL−based PSCs. In 2021, Ramachandran et al. fabricated nanostructured CuSCN/CuI bilayer HTL by electrodeposition method for carbon−based inverted PSCs [139]. The PSC with CuSCN/CuI bilayer HTL (Figure 14a,b)

Crystal Structure Regulation and Bilayer HTL Design
In addition to the above-mentioned strategies for the design and fabrication of inorganic HTLs, crystal structure regulation and bilayer HTL design have also been reported to boost the PCE and stability of inorganic HTL-based PSCs, although the relevant literature is limited until now. For instance, Heidariramsheh et al. successfully synthesized Cu 2 SnS 3 (CTS) NPs with different crystal structures (zincblende and wurtzite) as HTLs for PSCs [138]. More specifically, zincblende CTS and wurtzite CTS were synthesized by the oil amine solvent heating method using sulfur and thiourea as sulfur sources, respectively. The CTS NPs were then dispersed in non-polar chloroform solution to obtain CTS inks, which were spin-coated on perovskite layer as HTL. As depicted in Figure 13, Spiro-OMeTAD HTL displayed has the smoothest surface while zincblende-CTS exhibited the roughest surface, which was due to the NPs agglomeration in the HTL inks. In addition, the wurtzite-CTS and zincblende-CTS displayed much better hydrophobic capability than that of Spiro-OMeTAD due to the addition of lithium salts in Spiro-OMeTAD, which may be beneficial for the achievement of high moisture tolerance. The cell with wurtzite-CTS HTL exhibited a higher PCE of 13.1% than that of zincblende-CTS-based cell (7.87%), which was assigned to the more uniform and flattened wurtzite-CTS film, monodispersed wurtzite-CTS NPs and better energy level alignment. As for the moisture stability, the cell with wurtzite-CTS HTL retained~90% of the primary PCE after 1200 h's storage under ambient conditions with an RH of >45%, which was much better than that of Spiro-OMeTAD-based cell with a PCE retention ratio of~60% under the same conditions due to the remarkably enhanced hydrophobic capability of wurtzite-CTS HTL.
As mentioned above, CuSCN and CuI have been extensive as inorganic HTLs for PSCs [133][134][135]. CuSCN displayed a wide band gap of 3.4-3.9 eV and low hole mobility of 0.1 cm 2 V −1 s −1 , while CuI exhibited a narrow band gap of 3.1 eV and superior hole mobility of 40 cm 2 V −1 s −1 . [133,135] Thus, it is crucial to fully utilize the merits of CuSCN and CuI to further boost the PCEs of inorganic HTL-based PSCs. In 2021, Ramachandran et al. fabricated nanostructured CuSCN/CuI bilayer HTL by electrodeposition method for carbon-based inverted PSCs [139]. The PSC with CuSCN/CuI bilayer HTL (Figure 14a,b) generated a high PCE of 15.58% (Figure 14c) due to the improved energy level alignment, increased hole mobility and reduced sheet resistance. Furthermore, the cell with CuSCN/CuI bilayer HTL maintained 85% of the original PCE after aging under ambient conditions with RH of >60% for 1400 h. In addition, Ramachandran et al. further optimized the performance of bilayer HTL-based carbon-based inverted PSCs by tailoring the positions of CuSCN and CuI layers as well as modifying the electrodeposition method through controlling the electrochemical potential and deposition time to achieve the optimal thickness and transparency [140]. After optimizing electrochemical potential (−400 mV) and deposition time (2 min), the carbon-based PSC with CuI/CuSCN bilayer HTL (Figure 14d,e) produced a superior PCE of 20.35% (Figure 14f) (Figure 14c) due to the improved energy level alignment, increased hole mobility and reduced sheet resistance. Furthermore, the cell with CuSCN/CuI bilayer HTL maintained 85% of the original PCE after aging under ambient conditions with RH of >60% for 1400 h. In addition, Ramachandran et al. further optimized the performance of bilayer HTL−based carbon−based inverted PSCs by tailoring the positions of CuSCN and CuI layers as well as modifying the electrodeposition method through controlling the electrochemical potential and deposition time to achieve the optimal thickness and transparency [140]. After optimizing electrochemical potential (−400 mV) and deposition time (2 min), the carbon−based PSC with CuI/CuSCN bilayer HTL (Figure 14d,e) produced a superior PCE of 20.35% (Figure 14f) [140]. After optimizing electrochemical potential (−400 mV) and deposition time (2 min), the carbon−based PSC with CuI/CuSCN bilayer HTL (Figure 14d,e) produced a superior PCE of 20.35% (Figure 14f)

Conclusions and Perspectives
As a key component in PSCs, high-performance HTLs should meet several requirements, including high light transmittance, hole extraction capability and hole mobility, superior moisture and thermal stability, proper energy level alignment with the perovskite layer and low cost. At present, organic HTLs have been widely used in PSCs, showing superior PCEs. Nevertheless, organic HTLs suffered from poor stability and high cost, which have become one of the largest obstacles to the large-scale applications of PSCs. As an important alternative to organic HTLs, nanostructured inorganic materials have also received increasing attention as HTLs for PSCs, which not only significantly improved the moisture/thermal stability but also reduced the raw material costs of PSCs. In this review, the recent advances in the development and fabrication of nanostructured inorganic HTLs for PSCs are reviewed by emphasizing the superiority of nanostructured inorganic HTLs over organic counterparts. In addition, several strategies to boost the PCE and stability of inorganic HTLs are also proposed and discussed. Table 1 lists the performance comparison of PSCs with nanostructured inorganic HTLs designed by functional/selectively doping. As can be observed, functional/selectively doping is widely employed to boost the hole mobility of NiO x and the PCEs of corresponding PSCs. Among various cations, Sm 3+ doping is reported to be highly effective in boosting the PCEs of NiO x -based PSCs. A highly attractive PCE of 20.71% was achieved by Sm:NiO xbased PSC, which was assigned to the unique electronic properties of trivalent Sm 3+ cations with partially filled 4f orbits and empty 5d orbits, significantly reducing the formation energy of Ni vacancies and improving the conductivity. Table 2 shows the performance comparison of PSCs with nanostructured inorganic HTLs with different morphologies. For example, nanostructured CuSCN HTLs with different morphologies (e.g., NWs, NRs) are obtained by adjusting the parameters of electro-deposition method, such as the deposition time and applied potential. Among various CuSCN HTLs with different morphologies, CuSCN NWs-based PSCs exhibited the highest PCE due to the formation of core-shell CuSCN/MAPbI 3 NRs during the perovskite film deposition process, which effectively enhanced the hole extraction/transport capability and increased the interaction between perovskite film and HTL. Table 3 depicts the performance comparison of PSCs with inorganic nanocomposite HTLs. As can be seen, the PSC with Co 3 O 4 -SrCO 3 nanocomposite HTL produced the highest PCE of 21.84%, which was mainly attributed to the improved hole migration rate by Co 3 O 4 and selectively electron-blocking capability by SrCO 3 .    [67] Although numerous efforts have been devoted to the development and fabrication of nanostructured inorganic HTL-based PSCs during the past 5 years, some challenges and problems still remain, such as low conductivity, NPs aggregation in the HTL ink and thick nanostructured inorganic HTL. Thus, the design of new materials and fabrication routes are crucial for the further performance enhancement of nanostructured inorganic HTL-based PSCs in the future. As the low conductivity, trivalent lanthanide metal ions have been proved to be effective in enhancing the conductivity of NiO x . However, such enhancement is limited due to the low doping amount of lanthanide cations and the fixed atomic environment of simple oxides. Perovskite oxides (ABO 3 ) with fully/mostly occupied lanthanide cations at A-site, such as nanostructured SmCoO 3 , should be highly attractive as inorganic HTLs for PSCs. In addition, various perovskite oxides can be designed to achieve desired performance as HTLs for PSCs due to their excellent compositional and structural flexibility. As for the challenges of NPs aggregation and thick HTL film, some advanced deposition routes, such as atomic layer deposition (ALD), can be utilized to fabricate nanofilm-based inorganic HTLs for PSCs [141]. More specifically, ALD is conducive to the formation of nanofilms with uniform morphology and controllable thickness due to the self-limiting growth mechanism of ALD. Besides the above existing problems, the hole extraction/transport mechanism of nanostructured inorganic HTLs still remains unclear, which requires further investigations in the future (e.g., advanced characterization techniques and theoretical calculations).