The application of low-temperature processed metal oxide electron transport layers in flexible perovskite solar cells

Abstract

Flexible perovskite devices have garnered significant attention due to their promising applications in wearable electronics and portable energy systems. Metal oxide electron transport layers (ETLs), such as SnO₂ and ZnO, processed at low temperatures, are particularly well-suited for fabrication processes that require compatibility with flexible substrates. This review presents recent advancements in the low-temperature preparation of SnO₂ and ZnO, followed by a summary of their latest applications, categorized into additive and interface engineering. Additionally, other metal oxide ETMs treated at low temperatures are briefly discussed. Finally, the review concludes with an analysis of the challenges and limitations associated with perovskite solar cells (PSCs) that incorporate low-temperature processed metal oxide ETLs.

---

Jianghao Tian

Jianghao Tian is currently a master’s student in Electronic Information at the University of Electronic Science and Technology of China(UESTC), supervised by Prof. Huajing Zheng. He received his bachelor's degree from the Nanjing University of Information Science and Technology in 2023. His current research focuses on the effects of low-temperature-fabricated electron transport layers on the performance of perovskite solar cells.

Pu Fan

Pu Fan is currently a researcher at the Yangtze River Delta Research Institute (Huzhou) of the UESTC. He received his PhD from the UESTC under the supervision of Prof. Junsheng Yu and conducted postdoctoral research at Tsinghua University and the Zhejiang Tsinghua Institute of Flexible Electronics. His research focuses on the performance and stability of novel organic and perovskite solar cells, as well as near-infrared detection, organic electroluminescence, and thin-film transistors.

Huajing Zheng

Huajing Zheng is currently the Vice President of the Guangdong Institute of Electronic Information Engineering at the UESTC. He received a Bachelor of Science degree from Sichuan University and Doctor of Science degree from Lanzhou University. He completed his postdoctoral research at the UESTC in 2009. He had served as a visiting scholar and visiting professor at the Hong Kong University of Science and Technology and the Shizuoka Institute of Science and Technology in Japan. His research focuses on light-emitting display devices and flexible photovoltaic devices.

Ding Zheng

Ding Zheng is currently a professor at the UESTC. He received his Bachelor of Engineering (2013) and Doctor of Engineering (2019) degrees from the UESTC, supervised by Prof. Junsheng Yu. He subsequently served as a postdoctoral researcher and then as a research assistant professor in the groups of Tobin J. Marks and Antonio Facchetti at Northwestern University. In 2023, he became a senior R&D engineer at First Solar. He conducts research on the integration of flexible wearable technologies and the development of organic/perovskite photovoltaic devices.

Junsheng Yu

Junsheng Yu is currently a professor at the UESTC and director of the Sichuan Provincial Key Laboratory of Display Science and Technology. He received his PhD from Tokyo University of Agriculture and Technology in 2001 and then conducted postdoctoral research at Osaka University. Since 2003, he has conducted research at the University of Arizona and Georgia Institute of Technology. His research interests include luminescent materials and devices, organic solar cells, organic thin-film transistors, sensitive electronics and sensors, and flexible optoelectronic devices.

---

1. Introduction

The photovoltaic effect has built a bridge between light and electricity. Solar cells, a product of the photovoltaic effect, are slowly developing a new energy way in the interweaving of light and electricity. In 2009, Miyasaka's team introduced the first Perovskite Solar Cell (PSC),¹ achieving a photoelectric conversion efficiency (PCE) of 3.8%. This breakthrough injected new vitality into the photovoltaic industry. Extensive research has been conducted on PSCs, which has increased the conversion efficiency of PSCs to 26.95% (ref. 2) in just 16 years (2009–2025). It should be noted that the highest single junction efficiency of the organic solar cell (OSC), which was discovered in 1986,³ is 20.2%,⁴ and the highest efficiency of the dye-sensitized solar cell(DSSC), which was discovered in 1991,⁵ is 14.3%.⁶ The rapid development of PSCs is evident.

Fig. 1 shows the typical structure and working principle of the n–i–p PSCs. The ETL is known as a key functional layer in PSC devices. The main task of the ETL is to transport electrons and prevent the transport of holes.⁷ To achieve high performance, the ETL's energy band should be matched with the perovskite layer's; specifically, the highest valence band of the ETL should be lower than that of the perovskite layer for blocking transport of holes into the ETL.⁸ Moreover, the transmittance and electron mobility should be high, so that the light will be absorbed by the active layer and more electrons will be collected by the electrode.

Fig. 1 The typical structure and working principle of n–i–p PSCs.

Nowadays, many types of ETLs have been used in different PSCs. C₆₀, as one of the most famous fullerene materials, has been used in ETLs for a long time because of its high electron mobility and energy level alignment, and researchers can greatly improve the performance of devices by adding other functional groups to form new fullerene materials.^9–11 Phenyl-C₆₁ butyric acid methyl ester (PCBM) is also a widely used ETL in many studies.^12–16 Because PCBM has low carrier mobility and conductivity,¹² researchers have improved PSC efficiency by doping various materials into the PCBM layer, such as oleamide,¹³ SnS₂,¹⁴ CoSe¹⁶etc. ICBA is another fullerene derivative that can act as an ETL. Its LUMO (Lowest Unoccupied Molecular Orbital) is higher than that of PCBM, so the V_oc (open circuit voltage) can be improved.^17,18 Fullerene materials and their derivatives are one of the main materials for ETLs. Other materials have also been reported in the last decade. Graphene has high electron mobility, and its work function is suitable for FTO.^19,20 Small molecule materials,²¹ carbon nanotubes,²² amino functionalized carborane derivative CB-NH₂,²³ 3D molecules based on perylene diimide (PDI)(TPE-PDI₄),²⁴etc. can also be used as ETLs.

With the fast development of flexible electronic equipment and solar cells, people tend to combine the two to produce portable and wearable flexible solar cell equipment. Flexible solar cells can be achieved by continuous roll-to-roll technology, which has faster manufacturing speed and lower cost, and their applications are more extensive, from civilian to military, and flexible photovoltaic devices can complete the corresponding work well.²⁵ However, flexible PSCs require a flexible substrate, which is not resistant to high temperature. Take polyethylene terephthalate (PET) and polyethylene naphthalate (PEN) as examples, and their glass transition temperatures are about 80 °C and 120 °C, respectively.²⁶ Additionally, the roll-to-roll technology also has strict requirements for low temperature.²⁷ This means that the traditional high-temperature preparation process is not suitable for flexible equipment, so it is very important to develop and improve the preparation process at low temperature, even at room temperature. At present, many low-temperature processed ETLs have been applied to flexible perovskite solar cells. In 2016, a new electron transport material (ETM), amorphous Bi₂S₃ (a-Bi₂S₃), was prepared by Li et al.²⁸ at room temperature through a simple thermal evaporation process. Bi₂S₃ has inherent high electron concentration and carrier mobility, which is much higher than those of PCBM, showing high electron collection efficiency and low carrier recombination loss. Cadmium sulfide (CdS) has excellent electron extraction and transmission characteristics, which help to reduce the hysteresis of solar cells and improve efficiency.²⁹ WS₂ can also be deposited by solution-treated low-temperature technology or the radio frequency (RF) sputtering method,³⁰ and the optimized solar cell structure (FTO/WS₂/CsSnI₃/rGO/Pt) showed photovoltaic performance with a power conversion efficiency of 31% from the solar cell capacitance simulator (SCAPS-1D).³¹ Yoon et al. used C₆₀ (fullerene) treated in a vacuum at room temperature as the ETL, achieving a photoelectric conversion efficiency of 19.1%.³² Other materials like TiS₂,³³ In₂S₃,³⁴ ZrSnO₄,³⁵ and mesoporous graphene³⁶ have also been reported as low-temperature ETLs in recent studies.

Metal oxide (MO) materials like TiO₂/TiO_x, ZnO, SnO₂/SnO_x, WO₃, NiO_x, Al₂O₃, etc. have many advantages and the potential to be the ETLs of efficient PSCs. MO usually has a high electron affinity, which can effectively transport electrons and inhibit the transmission of holes, so as to reduce carrier recombination and improve the efficiency of the cells. MO also has good broadband transparency,³⁷ and its chemical and thermal stability is high, so that the stability of PSCs will be improved. As we can see from the energy level of MO in Fig. 2, the conduction band minimum (CBM) level of the MO material is generally lower than that of the perovskite, facilitating electron extraction from the perovskite layer into the electron transport layer. Additionally, its valence band maximum (VBM) level is lower than that of the active layer, thereby blocking and suppressing holes. This fulfills the requirements for an excellent ETL solution as previously mentioned. As a very popular ETM in the progress of PSCs, TiO₂ has attracted a lot of attention. Some research progress of TiO₂ is shown in Table 1. As we can see, researchers used different ways for improving the performances of TiO₂-based PSCs by optimizing fabrication technology, doping, adding a modification layer, and so on. It is worth noting that metal oxides with a mesoporous structure tend to have better performance as ETLs, for the reason that the mesoporous structure can provide a channel for perovskite molecules to fully come into contact with the electron transport layer, which is equivalent to increasing the contact area between the two functional layers and greatly increasing the electron transport efficiency. Moreover, it can provide good mechanical properties.³⁸ However, as shown in Table 1, a traditional TiO₂ ETL is usually prepared by the spin coating method, which usually requires two high-temperature sintering processes;⁵⁴ so the TiO₂ ETL prepared by high-temperature technology(mainly above 400 °C) is not suitable for flexible PSCs. Moreover, the high temperature process will also increase the preparation difficulty of PSCs and is not conducive to the stability of devices. The inherent excellent properties of inorganic metal oxides cannot make people ignore this ETL material. Therefore, it is of great significance to study low-temperature metal oxides used as ETLs in PSCs, which can not only promote the progress of solar cell performance but also be conducive to the development of wearable photovoltaic devices.

Fig. 2 Conduction band (CB) and valence band (VB) of metal oxide materials, and comparison with that of perovskite materials.

Table 1 Some research progress of the TiO₂ ETL

Year	Fabrication	Optimal methods and materials	TiO2 structure	Champion PCE	Reference
2014	Thermal oxidation sputtering	—	—	15.07%	39
2015	Spin-coating	Modify with rGO (reduced graphene oxide)	Mesoporous	18%	41
2016	Sputtering	—	Anatase	12.5%	40
2016	Anodic oxidation	—	Nanorod	15.2%	40
2016	Spin-coating	Interface modification with ionic liquid	—	19.62%	42
2016	Spin-coating	La2O3 as the interface modification layer	Mesoporous	15.81%	43
2017	Atomic layer deposition method (ALD) and spin coating technology	Bilayer ETL structure	Anatase	16.5%	44
2018	Hydrolysis-pyrolysis method	Nb-doping ETL	Nanoparticle	18.88%	45
2018	Spin-coating	Interface modification with graphene quantum dots (GQDs)	Mesoporous	20.45%	46
2019	Spin-coating	—	Nanoparticle	18.72%	47
2020	Spin-coating with acid-treated	Acid-treated amorphous TiO2 as the buffer layer	Anatase	17.8%	48
2020	Radio frequency (RF) magnetron sputtering	Li^+-doping ETL	Rutile	24.23%	49
2020	Electrochemical anodic oxidation	TiCl4-doping ETL	Leaves and needles like (LNT)	9%	50
2021	Spin-coating	Li2CO3-doping ETL	Mesoporous	25.28%	51
2021	Liquid phase deposition (LPD)	2D TiS2 as the interface modification layer	Nanograss	18.73%	52
2024	Chemical bath deposition (CBD)	Embed Ti0.936O2 nanocrystals into TiO2	—	25.5%	53

---

In this review, we will focus on the frequently used low-temperature MO ETLs—SnO₂ and ZnO, and their inherent properties, traditional preparation methods, current research progress, and application in PSCs will be described in detail. We found that the functional groups or functional atoms provided by different materials play key roles in the doping or interface modification of the ETL. This review will pay more attention to the microscopic mechanism of different materials and make some inferences and comparisons. We also summarized and compared the preparation methods and modification methods of SnO₂ and ZnO, and drew a table. In the table, we counted the commonly used modified materials and main preparation methods, compared the use of different materials on different ETLs, and the research on different low-temperature preparation methods. We also conducted an in-depth analysis of the fabrication processes for different MO ETLs, with particular emphasis on their compatibility with roll-to-roll processes and large-area fabrication techniques. Later, we will present the recent application of TiO₂ in a low-temperature process. In addition, we will introduce several excellent low temperature metal oxides like Al₂O₃, ZrO₂, In₂O₃, etc. It is worth mentioning that at the end of this review, we compared the traditional preparation methods that need high temperature with the current reported low temperature methods, and formed an image to make an intuitive comparison, so as to clarify the advantages of these low temperature preparation methods. Finally, prospects of MO ETLs prepared at low temperatures will be discussed.

2. Tin oxide

2.1 Basic properties and advantages

Tin oxide (SnO₂) is a common MO ETL in PSCs. Compared with TiO₂, SnO₂ has many advantages. Firstly, from Fig. 2, SnO₂ has a deeper conduction band and valence band,⁵⁵ so the ETL will have better electron extraction capability, and the recombination of electrons and holes will also be reduced. Secondly, the bandgap is wide (3.6–4.0 eV), so that more light will be able to pass through the ETL to reach the perovskite layer and the efficiency will be improved. Thirdly, the high bulk electron mobility (>240 cm² (V⁻¹ s⁻¹))⁵⁶ will help improve electron transmission capability. Fourthly, the temperature required for preparation is usually low (≤200 °C), sometimes even room temperature. Additionally, it has the characteristics of stable chemical properties, diverse preparation methods, and strong environmental stability. All these properties make SnO₂ a promising ETM for PSCs.

2.2 Fabrication methods of SnO₂

The methods of depositing SnO₂ are various. Generally speaking, the preparation temperature is not too high, which is suitable for preparation on a flexible substrate. The main technologies are summarized here.

2.2.1 Solution process method. The solution process method is the most widely used in the deposition of SnO₂. This method has a low cost and a relatively simple process, and can be used for the preparation of large area devices. There are many kinds of solution process methods according to a previous study.

The sol–gel method is one of the primary solution preparation methods.^57–59 For obtaining SnO₂ films with high quality, the gels need to be formed from metal alkoxide precursors through hydrolysis and condensation reactions, and drying and heat treatment are also essential. In this way, Wang et al. prepared a SnO₂ nanocrystalline ETL with a small particle size and a large specific surface area.⁵⁷ This is a SnO₂/Sn composite ETL made by controlling the annealing temperature and can help to achieve 8.7% PCE. However, in the traditional wet synthesis, a high temperature is required to improve the crystallinity, so another team proposed a new method to prepare the SnO₂ ETL by the sol–gel method at low temperature (below 80 °C).⁵⁹ As shown in Fig. 3a, this method omits the 150 °C high-temperature step and realizes the oxidation of Sn⁺ by refluxing SnCl₄·2H₂O alcohol solution. The reflux process makes O₂ and H₂O in the atmosphere participate in the oxidation of Sn²⁺, reducing the crystallization barrier of SnO₂. Because of colloidal aging at room temperature, the unreacted precursor solvent or organic solvent is volatilized, and the dispersion and uniformity of the whole solution are also improved. As shown in Fig. 3b, the light absorption ability is enhanced, and finally, a steady-state efficiency of 18.48% is achieved.

Fig. 3 (a) Schematic diagram of the new route to SnO₂ nanocrystals (NCs) at low temperature, (b) the SnO₂ sol before and after aging at room temperature for 24 h;⁵⁹ Copyright 2017, Elsevier. (c) Schematic illustration of the ligand exchange procedure for TBAOH-capped SnO₂ nanoparticles;⁶³ Copyright 2020, American Chemical Society. (d) Sketch processes for the preparation of a control SnO₂ ETL with the sol–gel solution; schematic illustration of one atomic layer deposition cycle for a SnO₂ ETL; a bilayer SnO₂ ETL preparation process combining ALD and spin-coating techniques, respectively.⁶⁶ Copyright 2023, John Wiley and Son.

Thermal decomposition is also a useful solution method. Ke et al. spin-coated SnCl₂·2H₂O precursor solution at room temperature and then conducted annealing at 180 °C to prepare SnO₂ ETLs directly.⁶⁰ This involves dehydration of the solution and hydrolysis of the metal salt, resulting in the formation of a tin oxide nanocrystalline film. Due to the special structure of the nanocrystal, this SnO₂ ETL film is compact, uniform, and has strong antireflection ability. Yang's group also used the SnCl₂·2H₂O dissolved in ethanol as a precursor solution, but they aged it by vigorously stirring for 1 hour at room temperature in an ambient atmosphere, like the research mentioned before; in this way, it may be easier for the solution to crystallize to form an ETL film, and finally achieve the champion PCE of 20.2%.⁶¹ This is the typical operation technique of this method, and the precursor solution can also be SnCl₄ mixed with ethanol.

Zhu et al. used another solution method—hydrothermal method—to make SnO₂ nanocrystals, which can produce highly crystallized SnO₂ nanocrystals and meanwhile help maintain effective electron transport in thicker films.⁶² We deem that the sealed environment and constant temperature provided by the hydrothermal method make the solution more stable and uniform, the film growth will be more orderly, and the film morphology and photoelectric properties will be better. To improve the dispersibility of SnO₂ nanoparticles in ethanol, Lee and co-workers proposed a ligand exchange method.⁶³ Generally, this is a solvothermal method for preparing SnO₂ nanoparticles. In detail, TBAOH replaces the insulating OA molecules on the surface of SnO₂ nanoparticles (Fig. 3c), and thus, a more uniform and compact film can be obtained; this benefits from the large volume and appropriate polarity of the TBAOH material, which can prevent the aggregation of nanoparticles, and the hydroxyl (–OH) in it can also adjust the energy level of SnO₂ to obtain a better energy level structure. Tu's team used the hydrothermal method and they achieved a PCE of 7.81% with the CH₃NH₃PbI₃ active layer.⁸³ In the same year, Zhu et al. also used this method for making a SnO₂ ETL, and they achieved 18.8% PCE with the use of this ETL in inverted PSCs.⁸⁴ We conducted a series of comparative analyses and found that the latter uses a p–i–n type device structure. In order to prevent the SnO₂ solution from damaging the perovskite (MAPbI₃), a layer of C₆₀ was added between the two functional layers. This avoids direct contact between SnO₂ and the perovskite, and also provides an additional hole blocking layer to suppress carrier recombination and optimize the device energy level structure. This seems to be an important reason for its improved efficiency.

The above methods all involve hydrolysis reactions, and in essence, they all chemically generate SnO₂ through reaction and heat treatment. Currently, tin oxide dispersion solutions can be purchased directly on the market and can be directly applied in the spin-coating process by direct dilution. After annealing at 150 °C to evaporate the diluted liquid, the SnO₂ ETL film is obtained by physical volatilization. Based on the above analysis, we will temporarily categorize the preparation methods into chemical and physical preparation for comparison. Chemical preparation methods produce denser and more uniform films, but they require consideration of the sufficiency of the hydrolysis reaction and the formation of by-products in a water–oxygen environment, both of which significantly impact film quality. The process is more complicated, too. A simple and direct physical preparation process can obtain an effective ETL, but its surface morphology is generally unsatisfactory, and particle aggregation is inevitable. Therefore, we believe that the development of simulated self-assembled materials (SAMs) based on SnO₂ by introducing some anchoring functional groups can be explored in depth as a research topic to prepare ETLs with a simple spin-coating process and excellent morphology.

2.2.2 Atomic layer deposition (ALD). The ALD method has attracted extensive attention because of its ability to prepare ultra-thin and compact films. This method has the ability to control the atomic layer precisely, so that a high-quality film can be achieved.⁵⁵ This is attributed to its working process. Generally speaking, the precursor reactant is first placed on the substrate to react with the chemical groups and then form a monolayer compound. After flushing the reaction chamber with inert gas, another precursor is introduced to react with the first compound layer to form a new one, and so on until an ideal film layer is achieved.

Xiao et al. found that the performance of the device based on the ALD SnO₂ ETL is significantly higher than that of the ETL-free one, and after modifying the ETL with a fullerene SAM, the PCE of the PSC can be improved to 17.25%.⁶⁴ This is due to the special properties of the SAM. By combining the anchoring group with the SnO₂ ETL, it is very likely that an effective dipole moment is formed, so as to improve the energy level of the ETL and make it fully match that of the perovskite layer. In addition, with the addition of a SAM fullerene material, its end functional group can capture additional oxygen vacancies or excess metal ions, which can greatly reduce the defect state between perovskite and the ETL, thus reducing the probability of carrier recombination, and alleviating the hysteresis effect of the whole device. However, traditional ALD also has some problems. For example, insufficient temperature (<150 °C) will lead to the appearance of amorphous SnO₂, which will affect the electrical performance of the ETL. Additionally, an incomplete precursor reaction will lead to an increase in defect states in the ETL, affecting carrier transport. Plasma-assisted atomic layer deposition (PAALD) for the SnO₂ ETL is an innovation because, with the help of plasma, higher reaction barriers can be overcome, thereby reducing defects. Moreover, this method can further optimize the band structure between the ETL and perovskite layers, and reduce the preparation temperature. As shown in Kuang and co-workers’ research, though the efficiency of 50 °C SnO₂ is slightly lower than that of 200 °C (former 16.2 ± 0.7%; later 16.1 ± 1%), other key properties are similar, which means that this method can be used in flexible PSCs well.⁶⁵ In 2023, Zhang et al. used two different methods for preparing the SnO₂ ETL—ALD and sol–gel—and combined them to form a double-layer ETL device (as Fig. 3d shows).⁶⁶ The unique conformal effect of ALD-SnO₂ can effectively regulate the roughness of FTO substrates, improve ETL quality, and induce the growth of the PbI₂ crystal phase, thereby developing the crystallinity of perovskite layers. Finally, 23.86% PCE and good device stability are achieved. Additionally, it is worth noting that the post-annealing of ALD is essential to the performance of the interface between the ETL and perovskite layer. As the post-annealing temperature increases, the crystal structure of the film gradually becomes orderly, and the content of Sn⁴⁺ is also increasing, resulting in the valence band shift of the SnO₂ thin film gradually decreasing, indicating a decrease in its bandgap, and the reduction of defect states will also make the energy levels near the valence band more concentrated.⁶⁷

2.2.3 Chemical bath deposition (CBD). Immersing the substrate in the required chemical solution allows the substances on its surface to react with the solution to form thin film particles, which then grow into a continuous film. This is a typical CBD process, which is applicable to different electrodes and can be reused.^68–70

Yoo and co-workers prepared a SnO₂ ETL by the CBD method, and a certification efficiency of 25.2% was obtained by optimizing this device.⁷¹ This process mentioned in the report can be seen as two stages: Sn⁺ is dissolved in a strong acidic environment (pH = 1–1.5) and hydrolyzed to form the Sn(OH)⁺ intermediate. The Sn(OH)⁺ intermediate is further oxidized in solution and reacts with OH⁻ provided by urea decomposition to form Sn(OH)₄, which is then dehydrated to form SnO₂. This is stage A. Next, the pH of the solution will be increased, and different Sn intermediates will be formed in the solution at the same time. This is stage B. After condensation and dehydration reactions, a mixture of Sn₆O₄(OH)₄ and SnO compounds will be formed from the intermediates. The SnO₂ film formed in stage A-ii (pH = 1.5) completely covers FTO and has better performance. This also shows a characteristic of CBD, that is, it is necessary to strictly control the reaction time and environment to obtain relatively high-quality film materials.

Jeong et al. also used CBD to deposit the SnO₂ ETL, and by using Ga(acac)₃ in the P3HT HTL, the device exhibited a PCE of up to 24.6%.⁷² It is worth mentioning that the traditional CBD method typically only produces relatively dense SnO₂ films, making it difficult to form mesoporous structures with strong mechanical stability. Wang's group proposed a method to optimize the preparation of SnO₂ in CBD.⁷³ One key part of the study is adding thioglycolic acid (TGA) into the CBD solution, which has a sulfhydryl group (–SH) that can form a strengthening bond with SnO₂ nanoparticles, and a weakly acidic carboxyl group (–COOH). The former can prevent excessive aggregation or deposition of nanoparticles, and the latter can adjust the nucleation rate of SnO₂. After that, through pre-aging treatment, TGA molecules were consumed spontaneously and the aggregation of SnO₂ nanoparticles was promoted for forming a mesoporous structure. The approach used in this study is quite interesting. Basically, it involves modifying SnO₂ through doping. However, unlike typical doping (simply blending two different substances), this study incorporates the dopant directly as a reactant in the SnO₂ formation reaction, thereby altering the inherent properties of the ETL. The use of specific substances blended with the ETL precursor or solution to induce a reaction could provide new insights into doping processes and could even lead to the development of new ETMs based on mainstream ETLs.

2.2.4 Magnetron sputtering deposition. Magnetron sputtering deposition is another widely used method to prepare ETLs for solar cells. Qiu et al. used Ar as the sputtering gas and O₂ as the reaction gas for preparing SnO₂ films by magnetron sputtering at room temperature.⁷⁴ Ar becomes plasma under the action of an electric field, and then bombards the target under the action of a magnetic field to form SnO₂ vapor, which condenses on the substrate to form a thin film. The prepared ETL is thin and uniform, and has ideal conductivity; this method is suitable for the manufacture of large-area devices. There are various influencing factors of this operation, and the gas environment is one of the key parts to determine the quality of sputtered films, because it directly determines the purity and quality of the film. According to previous research, more O₂ participating in the work can reduce the oxygen vacancies in the film and make more Sn⁴⁺ oxidized, which makes the grain size of the film larger and more compact.⁷⁵ Sputtering power is another crucial factor; a lower sputtering power makes the kinetic energy of the sputtered SnO₂ material lower, which effectively inhibits the crosstalk between ions. In addition, the material has enough time to diffuse, so the roughness will be reduced.⁷⁶ Sputtering time cannot be ignored too for the reason that it can directly affect the thickness of films; if the ETL is too thin (sputtering time is too short), it may not be able to effectively extract and transmit electrons, and if the ETL is too thick (sputtering time is too long), it may increase the resistance, so finding a suitable sputtering time is very important.⁷⁷

For preparation by sputtering, attention should be paid to the damage of high-energy particles to other functional layers (similar to breaking through the film surfaces). This can be alleviated by sputtering in steps (first at low speed and then at high speed) or by adding a sacrificial layer.

2.2.5 Other methods. Other methods for preparing a SnO₂ ETL at low temperature are also mentioned in past research, and different methods have their own characteristics and advantages and can be used in specific applications. Take oxygen plasma-activated electron beam (E-beam) evaporation technology as an example; this method can be used to prepare thin films at room temperature, which is very friendly to flexible devices that require low-temperature processes. The main principle of this method is that a high-energy electron beam bombards the target and evaporates it on the substrate to form a deposition. Moreover, by introducing oxygen plasma during the evaporation process, the stoichiometric ratio of SnO₂ thin films can be effectively adjusted based on the oxidation mode.⁷⁸ Jun's team used the electrochemical deposition method to prepare ETLs.⁷⁹ By adjusting the deposition time, current density, and electrolyte concentration, the thickness and composition of the film can be precisely controlled. The flexible films with a large area can also be achieved by this method. Yu's team⁸⁰ and Cao's team⁸¹ have successively used the combustion method to prepare SnO₂ ETLs. This method can be used to adjust the thermal characteristics and treatment temperature of SnO₂ at a low temperature of 150 °C. It is energy-saving and environmentally friendly, and is compatible with the roll-to-roll printing process. Compared with the sol–gel method, SnO₂ films prepared via the combustion method exhibit a more uniform surface because the high heat flow in the combustion process will make the reactants react quickly and uniformly. SnO₂ nanosheets prepared by low-temperature electrospray exhibit a highly porous and interconnected nanoflower structure, which is conducive to the complete penetration and electron transport of the perovskite absorption layer.⁸² Additionally, other methods like chemical vapor deposition(CVD),^85,86 thermal evaporation,^87,88 slot-die coating,⁸⁹etc., have also been reported in previous research studies. Apparently, different methods have their own advantages, and people can change the details of these methods to satisfy the study’s requirements.

Through literature reading and experimental comparison, we believe that among many preparation methods, the solution method (especially the spin coating method) has broader application prospects, low cost, and a simple and easy-to-understand preparation process, and is also suitable for the preparation of flexible devices and large-sized devices. Other preparation methods require complex reaction processes and harsh reaction conditions, while some require expensive preparation equipment, and the requirement of high-energy particles or plasma will also increase the cost and difficulty of preparation. For the solution method, in addition to directly purchasing SnO₂ dispersions, the spontaneous formation of SnO₂ through appropriate reactions, such as chemical interactions between different substances in solution and annealing processes, also has corresponding research value. Given the growth environment of SnO₂, these studies require attention to the key role of the reaction range and the detection of ETL components to minimize the formation of by-products. Alleviating the surface roughness of the film prepared by the solution method is also an important research direction.

In the process of sorting out the film preparation methods, we found that the working environment or working principle of many methods is roughly similar. For example, CBD, the electrochemical deposition method, and the hydrothermal method all put the substrate in solution for film preparation; magnetron sputtering, E-beam evaporation, and thermal evaporation all seem to involve target heating or target bombardment. In order to clarify the similarities and differences of these methods, we designed Table 2 to summarize and distinguish them. Sometimes researchers combine different methods in their studies to prepare two different ETLs, which can also achieve excellent performances. The best approach is to select the method that is most suitable for the preparation of the entire device.

Table 2 The comparison of different preparation methods

---

2.3 Modifications of the SnO₂ ETL

Metal oxide, as mentioned above, has many excellent electrical and optical properties that other materials do not possess, such as high conductivity, high chemical stability, wide band gap, etc. However, when they work with other materials like perovskite or ITO/FTO, the interactions between the microscopic particles will have a great impact on the overall performance. For example, the metal ions in the MO ETL may be extended to the perovskite layer to destroy the perovskite crystal structure, and the ions in perovskite, such as I⁻, may be oxidized to form an intermediate to destroy the interface morphology. Furthermore, a single MO material cannot meet people's needs for higher performance ETLs sometimes. So, it's necessary to make some changes to the MO ETL layer. Additive engineering and interface modification are used to accomplish such a task.

2.3.1 Additive engineering. Additive engineering mainly includes two types: doping and physical incorporation. The main function of doping is to adjust the energy band of the material to obtain a more ideal electrochemical performance.⁹⁰ Incorporating special trace materials with ETM solution to alter the corresponding electrochemical properties of ETLs is also an effective modification method. There are many dopants or blended substances, such as organic, inorganic, small molecules, cations and so on. Here, we will summarize these methods for SnO₂ ETLs.
2.3.1.1 Metals and metal cations. Metals and metal cations are ideal additives. They can change the band structure of the original ETL to optimize the electrical limit. Some can also improve the interface performance and reduce defects.

Park and co-workers used metal Li to dope SnO₂ (Li:SnO₂) at a low temperature (185 °C) and achieved a flexible and wearable device.⁹¹ Li⁺ replaces Sn⁴⁺ and acts on the lattice of SnO₂, thus reducing its energy level position, and the defect states in the film are reduced accordingly, thus promoting the growth of grains and bringing fewer charge defects, so the conductivity is also improved. What needs to be noted is that different doping concentrations can be used to prepare thin films with different grain sizes as shown in Fig. 4a.⁹² Zhuang et al. realized that ion diffusion induced double-layer doping by directly adding LiOH to SnO₂ colloidal dispersion solution.⁹³ As shown in Fig. 4b, a large number of Li⁺ ions diffuse into the SnO₂/perovskite interface and the perovskite layer, forming a gradient concentration distribution, promoting the carrier to move from the high concentration region to the low concentration region, thus improving the effective transport and extraction of carriers at the interface. The doped device showed higher PCE, reaching 21.31%. Zn doping can also help to improve the energy level alignment by raising the conduction band energy level, so as to reduce energy loss and improve the charge extraction efficiency. Liao's team from Huazhong University of Science and Technology used Zn to dope the SnO₂ ETL for the first time at low temperature and achieved the highest PCE of 17.78%.⁹⁵ Co doping can reduce the charge traps.⁹⁹ Nd³⁺ doping can effectively improve electron mobility and reduce surface defects, thus reducing the interfacial recombination between perovskite and the ETL.⁸¹ The bismuthene-doped SnO₂ ETL has a smoother surface, and the energy band moves up after doping, which reduces the interface resistance between the composite layer and the perovskite layer, effectively accelerates electron extraction, and does not reduce the hole blocking ability.¹⁰⁰ It is worth mentioning that all the research above found that the doping of these metal ions will make the grain size of the SnO₂ ETL larger, and then lead to the perovskite film structure with larger grains.

Fig. 4 (a) The top-view SEM images of the perovskite film of (a) pristine SnO₂, (b) 2% Li: SnO₂, (c) 3% Li: SnO₂, (d) 4% Li: SnO₂, (e) 5% Li: SnO₂, and (f) 6% Li: SnO₂, (g) grain size distributions obtained from SEM images, abd (h) XRD patterns of perovskite films;⁹² Copyright 2020, Elsevier. (b) Schematic illustration of Li⁺ ion diffusion;⁹³ Copyright 2022, Springer Nature. (c) (a) UV-vis absorbance spectra and (b) XRD patterns of perovskite films grown on SnO₂ and Al–SnO₂ substrates, (c) EQE curves and integrated current density of perovskite solar cells based on SnO₂/Al–SnO₂ and (d) J–V curves of the best-performing PSC using SnO₂/Al–SnO₂ ETLs;⁹⁴ Copyright 2017, Hao Chen et al. (d) 7.4-MTO thin films and the corresponding schematic diagrams of the energy band structures.⁹⁶ Copyright 2022, American Chemical society.

Grain size has a great impact on device stability, interface defects, and electron transmission efficiency. As for the problem that Li, Zn, Co and Nd doping will increase the grain size of the SnO₂ ETL, we conclude that they have similar ion radii and properties, so they can well replace Sn⁴⁺ or occupy its sites to form a matching lattice, thus reducing the generation of defect states, inhibiting the effect of oxygen vacancies, and increasing the chance of grain growth. For bismuthene, it mainly depends on its self-adaptive characteristics to form a lattice-matched heterojunction with SnO₂ and reduce interface mismatch and defect density, so as to realize the growth of large-size grains. Because of the high quality, uniform, and compact ETL, it provides a better growth environment for perovskite films with large grain size. Of course, excessive doping may lead to additional defect states or lattice mismatch, which can explain why the grain size increases and then decreases with the increase in doping concentration.

Quy and Bark dissolved Ni(OCOCH₃)₂·4H₂O directly in the commercial SnO₂ colloidal solution to realize the doping of Ni.⁹⁷ The roughness of Ni:SnO₂ is more, and this may because the difference between the radius of Ni⁺ and Sn⁴⁺ causes lattice distortion, or the doping of Ni ions may affect the phase transition of SnO₂. However, the perovskite layer grown on the doped ETL is more uniform, which also greatly improves the performance of the device. This may be because the doping of Ni reduces the defect states. In addition, it is mentioned that the ETL may have a suitable wetting surface, which is conducive to the formation of a smooth perovskite surface. Another useful metal material is Al. As shown in Fig. 4c, Al doping will not affect the formation and light absorption of perovskite; in contrast, it can improve the photoelectric performance of SnO₂, optimize its energy level structure, and improve the performance of the whole device.⁹⁴ Eu³⁺ doping helps to reduce the charge accumulation at the interface by adjusting the band structure of SnO₂, and reducing the trap density in perovskite films by passivating Pb vacancies and I vacancies, and improving the carrier transport efficiency.⁹⁸ Magnesium (Mg) doping can passivate oxygen vacancy defects(Vo defect), for the reason that when Mg²⁺ replaces Sn⁴⁺, they consume the additional electrons originally provided by oxygen vacancies, thereby reducing the concentration of free electrons in the material, and as shown in Fig. 4d, the energy level structure can be optimized in this way.⁹⁶

2.3.1.2 Chlorine atoms and chlorides. Chlorine atoms and chlorides are another common dopant. Since the SnO₂ precursor solutions used in a large number of studies contain Cl⁻ ions, we believe that the presence of Cl⁻ is beneficial to ETLs. In the above discussion, metal ions are almost always provided by metal salt solutions, the most commonly used of which are chlorides, nitrates, and acetates. The latter two can volatilize during annealing, but the Cl⁻ provided by chlorides does not volatilize in large quantities and may combine with the main components of the ETL. However, the research mentioned above did not elaborate on this, focusing instead on metal particles. Here, we will focus on the Cl⁻ in chlorides for a summary. Compared with metal doping, they are more involved in changing the interface and chemical properties of ETLs. Cl can successfully combine with the SnO₂ surface to form enhanced bonds, which can reduce the existence of defect states, regulate energy levels, inhibit carrier recombination and improve electron transport performance, and this is main reason for the improvement of the device. Liang and co-workers used a chlorine-doped SnO₂ ETL to achieve high V_oc (1.195 V) on planar PSCs and the PCE reached 20% in this research.¹⁰¹ Ren et al. developed a direct-contact reaction process (DCRP) and used 1′2-dichlorobenzene to prepare SnO₂–Cl. Compared with the original SnO₂ ETL, the efficiency of the SnO₂–Cl ETL is improved (17.01% to 17.81%), and the hysteresis phenomenon is significantly reduced.¹⁰²

TiCl₄ is a common material for chlorine incorporation. The electronegativity of Ti⁴⁺ being lower than that of Sn⁴⁺ indicates that Ti⁴⁺ has a relatively weaker ability to attract electrons. Consequently, its multivalency is lower than that of Sn⁴⁺. Due to its multivalency, Sn⁴⁺ is highly susceptible to oxygen vacancies, readily forming defect states such as Sn²⁺. The incorporation of Ti⁴⁺, however, enables the formation of more stable Ti–O bonds, effectively reducing oxygen vacancies. Furthermore, the addition of Ti⁴⁺ can alter the energy level structure of the ETL. By introducing a small amount of TiCl₄ into the SnO₂ precursor solution, Cai et al. effectively reduced the active layer surface roughness (from 29.3 to 23.3 nm, with the surface roughness of the ETL decreased from 16.7 to 13.9 nm).¹⁰³ This is primarily due to the reduction in defects within the ETL, which enhances its uniformity and hydrophilicity, thereby yielding superior perovskite films. Liu et al. also used TiCl₄ for dissolution in the SnO₂ NCs solution.¹⁰⁴ Different from the research above, they stirred the solution at 70 °C, so SnO₂:TiO₂ solution was prepared. From Fig. 5a, compared with other thin films, the SnO₂:TiO₂ mixed film has better surface conditions, and the device based on it achieves a PCE of up to 23.19%. Chen and co-workers regulated the performance of the ETL by introducing small molecule copper(II) chloride (CuCl₂)(a low concentration and low molecular weight chloride material).¹⁰⁵ The introduction of CuCl₂ improves the dispersion of the SnO₂ colloid, reduces the aggregation of SnO₂ nanoparticles, and improves the uniformity and quality of the ETL. A large number of Cl ions were released at the interface between the ETL and perovskite layer, which played a great role in passivating the interface defects. As shown in the diagram in Fig. 5b, SnO₂–CuCl₂ has better conductivity, and the PCE and stability of the device based on it can be improved a lot. Zhu et al. from Nanjing University¹⁰⁶ and Lin's team from the University of Tokyo¹⁰⁷ used KCl and NaCl, respectively, to dope the SnO₂ ETL, and both achieved good passivation effects and energy level optimization, finally achieving breakthroughs in device performance.

Fig. 5 (a) AFM height images of (a) TiO₂, (b) SnO₂, and (c) SnO₂:TiO₂ NCs deposited on FTO substrates;¹⁰⁴ Copyright 2023, Elsevier. (b) Conductivity of SnO₂ and SnO₂–CuCl₂ on FTO, measured in the dark, steady-state PCE at the maximum power point of different PSCs (the bias voltages are 0.90 and 0.95 V, respectively), and PCE histograms of 30 PSCs;¹⁰⁵ Copyright 2024, American Chemical Society. (c) The schematic of the distribution of SnO₂ nanocrystals in the dispersions, arrangement of SnO₂ nanocrystals in the thin films, and the crystal growth of the perovskite layer on the SnO₂ thin films with and without HP.¹¹³ Copyright 2020, John Wiley and Sons.

Many common chloride dopants are accompanied by the doping of metal ions, such as Ti⁴⁺, K⁺, Cu²⁺, Na⁺, etc. We compared and summarized the working mechanism of metal ions and Cl⁻ in the research studies and found that compared with most metal ions directly acting on Sn⁴⁺ to change the ETL energy level, Cl⁻ may participate in the defect passivation of the ETL and perovskite interface in addition to forming a chemical bond with SnO₂. Notably, when Cl⁻ binds with Pb²⁺, it can mitigate the impact of lead vacancies on the entire device. The use of metal chlorides for dual-ion doping is highly effective and provides a viable approach for doping optimization in the future.

Ammonium chloride has also been used in doping for inducing modification of the tin oxide colloid.¹⁰⁸ This modification not only improves the electron mobility of the ETL but also improves the energy band alignment with perovskite. The introduction of NH⁴⁺ and Cl⁻ helps to optimize the ETL/perovskite interface and reduce the interface defects by inhibiting the formation of deep level defects. It is worth noting that NH⁴⁺ interacts with the surface of SnO₂ to alter the zeta potential of SnO₂ colloids, leading to colloidal aggregation and the formation of a denser ETL. Lin et al. proposed a new method of ETL modification by introducing perovskite precursors (methylammonium chloride MACl or formamide chloride FACl) into the SnO₂ layer.¹⁰⁹ These two types of doping materials are common core components in the preparation of perovskite solutions. Their cations can interact with the surface of SnO₂, providing nucleation sites for the growth of perovskite crystals. Cl⁻ can effectively passivate the defect states of the ETL, making the growth of perovskite crystals more uniform and compact, and with sufficiently large grain size.

2.3.1.3 Polymers. Polymers are also widely used in the modification of ETLs. Common polymers include conjugated polymers, organic polymers, etc. The electrochemical properties of conjugated polymers can be adjusted by adjusting their microstructure. Organic polymers also have a feature that other substances do not have, that is, they are eco-friendly. These polymer materials have special functional groups that can be utilized to achieve binding and modification with SnO₂.

Wei et al. studied novel SnO₂ in a polymer matrix (SPM) as an electron transport layer (ETL).¹¹¹ Hydrophilic polyethylene glycol (PEG) molecules have an ether bond, which can form hydrogen bonds with the surface of SnO₂, allowing it to adsorb onto the surface of SnO₂. The long chains in their molecules increase the distance between particles, thereby limiting the aggregation of nanoparticles. Adding PEG to SnO₂ colloidal solution can improve the quality of the film, and enhance the hydrophilicity, so that perovskite precursor solution can completely cover the SnO₂ film without plasma treatment. The non-ionic polymer polyacrylamide (PAM) contains a large number of amide groups, and its chemical activity can lead to the formation of hydrogen bonds with SnO₂, making the entire ETL have good hydrophilicity and surface morphology, which enables researchers to obtain devices with a high conversion rate (21.61%) and stability.¹¹²

According to the fabrication process shown in Fig. 5c, the high negative charge density of biopolymers (heparin potassium, HP) helps to stabilize SnO₂ nanocrystals in the dispersion and then enhances interface contact.¹¹³ The –COO⁻ group contained in HP can interact with Sn atoms, allowing the polymer to adsorb on the surface of SnO₂. In this way, the negatively charged polymer will cause electrostatic repulsion between particles, preventing SnO₂ from agglomerating. This can explain why HP can improve the stability of the SnO₂ colloid precursor. In addition, the abundant –SO₃⁻ in HP will bind with Pb²⁺, which appears to enhance the contact between perovskite and the ETL, thereby reducing the defect states at the interface. The SnO₂ composite, which is regulated by crystalline polymer carbonitride (cPCN), provides better energy level alignment with the perovskite layer. This is because the rich π-conjugated electron system in cPCN can adjust the band structure of SnO₂ and improve the electron extraction and migration efficiency of the entire ETL.¹¹⁴

Kwon and co-workers also conducted research on the polymer-capped SnO₂.¹¹⁵ The study has shown that polymers with carboxylic acid (–COOH), phosphoric acid (–PO₃H₂), and sulfonic acid functional groups can chemically interact with oxides and have strong binding affinity. All of them can form chemical bonds with the SnO₂ surface, improve film uniformity, and reduce interface defects. Strong binding affinity is an effective means of inhibiting nanoparticle aggregation, and it can also control the growth of nanoparticles and tightly connect conductive substrates and ETLs. However, it is worth noting that the bonding strength of these three functional groups is different, and their working mechanism on ETLs is also different. Carboxylic acid groups have the strongest bonding ability, and are anchored on the surface of SnO₂ by coordination or hydrogen bonds, effectively inhibiting the agglomeration of nanoparticles. The weak phosphate group mainly forms coordination bonds with SnO₂ metal sites, reducing local thickness differences. The weakest sulfonic acid group mainly acts on the ETL by adsorption, so its device performance improvement is limited. The essence of doping and incorporating modification is actually the interaction of different functional groups, ions, molecules, etc., in the microstructure. Therefore, the choice of blending materials is actually the choice of functional groups.

2.3.1.4 Others. Other materials can also be used in the modification of the SnO₂ ETL, such as quantum dots. This material, known as “artificial atoms”, is widely used as an additive in the mixed ETL because of its excellent electrochemical properties. Mxene quantum dots (MQDs) have different functional groups on the surface, and they can form chemical bonds with the hydroxyl groups (–OH) on the surface of SnO₂, reducing the surface energy of SnO₂, which can rapidly induce the nucleation of perovskite in the solution of the perovskite precursor to form an intermediate perovskite phase.¹¹⁶ CdS quantum dots inherently have high electron mobility, and through the coordination effect between S²⁻ and SnO₂, the conductivity of ETLs is improved.¹¹⁷ Black phosphorus quantum dots (BPQDs) incorporated in the SnO₂ layer can not only improve the inherent defects of the SnO₂ layer, but also inhibit the oxidation of BPQDs, thus improving the quality and stability of ETLs.¹¹⁸ Fullerene materials are the research focus for perovskite solar cells. Zhong et al. modified SnO₂ by using fullerene derivatives (CPTA).¹¹⁹ The carboxyl groups in CPTA molecules can react chemically with the insufficiently coordinated Sn atoms on the surface of SnO₂, forming chemical bonds. This interface modification enhances the adhesion between the MAPbI₃ thin film and the SnO₂ layer. Other inorganic materials like polyoxometalate,¹¹⁰ graphdiyne(GDY),¹²⁰ CsI,¹²¹ potassium trifluoroacetate(KTFA),¹²²etc., are also used in additive engineering in the previous work.

In addition, the usage of organic compounds in ETLs is also common. Liu's team from UESTC used a very novel way to use additives.¹²³ They pre-buried volatile organic additives (HCOONH₄) in the SnO₂ ETL, and achieved in situ and overall modification of the ETL, the perovskite layer, and their interface through bottom-up penetration during thermal annealing. The SnO₂ ETL treated with HCOONH₄ showed enhanced electron extraction ability, reduced surface oxygen vacancies, and inhibited surface charge recombination. These phenomena are attributed to the bonding relationship between the amino and carboxylate functional groups and SnO₂. The flexible PSCs based on this ETL achieved a PCE of up to 22.37% and the durability is good. Cobalt amine is a kind of biological macromolecular material, which is compounded with SnO₂ and plays a role in interface modification through coordination bond and electrostatic interaction. The SnO₂@B₁₂ ETL also has good performance.¹²⁴ The PO₄³⁻ in phosphate can form a chemical bond combined with Sn to reduce the suspended bonds and the density of trapped states on the surface of SnO₂; in addition to improving the electron transport capacity, it can also enhance the adhesion between the perovskite layer and ETL and reduce unnecessary degradation.¹⁴²

2.3.2 Interface engineering. As mentioned above, interface defects seem to be a very common problem in PSCs. There are uncoordinated ion defects on the surface and interface of perovskite materials, energy level mismatch between the perovskite layer and other functional layer materials, and the structural characteristics of the material itself, which cause charge recombination at the interface.¹²⁵ Interface engineering is one of the main methods for alleviating these problems. With the help of interface engineering, we can achieve many effects: (1) improve the energy level alignment; (2) reduce the recombination of carriers; (3) promote carrier extraction; (4) improve the stability; (5) protect the environment, etc. There are many methods for interface engineering; in addition to the traditional interface passivation, morphology control, adding a buffer layer, etc., the use of multi-layer ETLs can also be regarded as interface engineering. Here, let's do some sorting.
2.3.2.1 Bilayer ETLs. Making bilayer ETLs is one of the popular directions of developing new ETLs for PSCs. The two types of ETL are often used for different materials or the same material with different preparation methods, so they have their own advantages. Combining them can improve the electrical properties, adjust the energy level, achieve better morphology of the surface, and improve the performance of the device. Broadly speaking, it is a useful way of interface engineering.

Xue et al. used a low-temperature electron beam evaporation technology to make a SnO₂/TiO₂ bilayer ETL.¹²⁶ By combining the advantages of the SnO₂ layer and the TiO₂ layer, the energy band structure can be readjusted, and the surface morphology can be improved. Dropping SnO₂ onto a TiO₂ thin film can effectively fill the gaps on the surface of the TiO₂ thin film prepared by the hydrothermal method, which is beneficial for the growth of high-quality perovskite thin films.¹²⁷ This is also another major advantage of a double-layer ETL. Dong and co-workers also mentioned that the interface synergy between SnO₂ and TiO₂ effectively blocks the path of charge recombination and energy loss, and significantly reduces the internal series resistance of the whole PSC.¹²⁸ Huo et al. constructed a SnO₂/TiO₂ gradient heterojunction (GHJ) by the two step CBD method.¹²⁹ Due to the slightly lower conduction band of SnO₂ compared to that of TiO₂, an energy band gradient is formed between the two ETLs, which is almost a unidirectional electron transport structure, effectively alleviating the problems of electron reflux and carrier recombination. In addition, SnO₂ and TiO₂ form an interpenetrating GHJ ETL, which is highly beneficial for enhancing conductivity.

SnO₂ ETLs prepared by different methods can solve the energy level mismatch between the SnO₂ layer and the perovskite layer.⁶⁷ SnO₂ thin films prepared by spin coating on FTO may exhibit issues such as particle aggregation and mismatch with perovskite energy levels. Additionally, the unevenness of FTO can also result in poor film quality. As mentioned above, ALD can precisely control the surface morphology of thin films, resulting in excellent thin film morphology, as shown in Fig. 6a. Moreover, the study suggests that the introduction of ALD-SnO₂ increases the Fermi level of the ETL. The built-in electric field created by the double-layer SnO₂ structure also helps to overcome the electron accumulation at the ETL/perovskite interface, so as to achieve higher V_oc and fill factor (FF). The efficiency of PSCs using this ETL increased from 22.09% to 23.86%. Besides, forming a dual ETL with doped SnO₂ is also effective. Guo et al. treated SnO₂ with InCl₃ to achieve simultaneous doping of the ETL and passivation of interface defects.¹³⁶ This treatment can reduce the oxygen vacancies on the SnO₂ surface, form Pb–Cl bonds, and enhance the chemical interaction between the ETL and perovskite.

Fig. 6 (a) AFM images of the surfaces of the blank, ALD SnO₂, control SnO₂, and double SnO₂ ETL capped FTO, respectively;⁶⁶ Copyright 2023, John Wiley and Sons. (b) Synthetic process of SnO₂ and ‘‘enveloped’’ SnO₂ NCs;¹⁴⁰ Copyright 2021, Royal Society of Chemistry. (c) Schematic diagram of reducing interface recombination by bandgap alignments. The energy levels of the dense a-WO_x are experimentally determined by XPS and UV absorption;¹³⁸ Copyright 2019, Elsevier. (d) Schematic illustration of NH₄F treatment on the SnO₂ surface.¹⁴¹ Copyright 2020, American Chemical Society.

Al₂O₃ can also form a double electron transport layer with SnO₂. It is worth mentioning that the UV-irradiated Al₂O₃ layer effectively enhances the wettability of the electron transport layer and provides a larger interface area, which is conducive to the uniform growth of the perovskite layer and improves the charge extraction efficiency.¹³¹ Sulfide can be used as an ETL too. An In₂S₃ and SnO₂ bilayer ETL was studied in the previous work.¹³² The introduction of the In₂S₃ layer improves the band bending of the perovskite/ETL interface and increases electron transport. CdS has higher electron mobility than SnO₂ and TiO₂. When the CdS ETL is deposited on the SnO₂ ETL, it can be used as an intermediate step to promote the electron transport from the perovskite layer to SnO₂.¹³³

Additionally, the bilayer ETL combined with C₆₀ and ALD-SnO₂ can improve the electron transport properties by increasing the density of C₆₀ or SnO₂. The conduction band offset (CBO) can indicate the mismatch of different functional layers, and an appropriate CBO value can reduce the electron transport barrier at the C₆₀/SnO₂ interface, so that the conductivity could be improved. This can also be controlled by adjusting the density of C₆₀ or SnO₂.¹³⁴

2.3.2.2 Buffer layer. Adding a buffer layer is another method of regulating interface defects. Compared to the bilayer ETL structure, it involves a wider range of materials and may be accompanied by reactions or bonding that determine interface performance at the interface. Many buffer layer materials themselves have not been applied as a separate ETL in PSCs, so we distinguish them from bilayer ETLs. However, both of their key working principles are directly modifying the interface outside the original ETL, avoiding the introduction of additional substances through doping.
2.3.2.2.1 Common buffer layers. Yang et al. used novel histamine diiodate (HADI) for the modification of the SnO₂/perovskite interface.¹³⁵ HADI can bind with Sn⁴⁺ ions and hydroxyl groups on the surface of SnO₂, which can be imagined as HADI molecules throwing a ‘boat anchor’ onto the SnO₂ film, firmly binding the HADI layer and SnO₂ layer. Due to the generation of hydrogen bonds, the conduction band of SnO₂ may shift upward, reducing the potential barrier for electron extraction. The HADI edge in contact with perovskite can react with lead and iodine vacancies to form stable chemical bonds, reducing interface defects in perovskite. In addition, due to its suitable electrical properties, HADI acts as a ‘bridge’ for electron transport, connecting SnO₂ and perovskite layers. This is the typical and main working principle of the buffer layer.

Zuo and co-workers used two different materials—CDSC and DPAH for interface modification.¹³⁹ The molecules of these two materials react with the –OH group on the surface of SnO₂ through an esterification reaction, or coordinate with Sn⁴⁺ in SnO₂, so as to fill the oxygen vacancies in SnO₂, and combine with the cation and anion defects in perovskite films through ion bond interaction, so as to realize the passivation of defects. Fig. 6b shows a novel way to modify the SnO₂ layer. Yuan et al. formed an amorphous NbO_x layer to reduce the surface defects and promote the favourable growth of perovskite.¹⁴⁰ NbO_x forms chemical bonds with oxygen atoms on the surface of SnO₂, passivating suspended components and defect states on its surface. Nb atoms react with Pb²⁺and I⁻ to passivate defects on the surface of the perovskite layer, which also leads to a dual interface passivation effect simultaneously. Zhuang and co-workers used two different ways for modifying SnO_2—doping and a buffer layer.¹⁴³ Here, let's focus on the latter. The addition of RbF between SnO₂ and perovskite did not change the electron mobility, but the presence of Rb⁺ can inhibit the ion diffusion of the perovskite layer.

C₆₀ can also be used as a buffer layer in PSCs. Zhu et al. spin-coated C₆₀ with a thickness of approximately 20–30 nm between the perovskite layer and SnO₂, and the energy level can be readjusted.¹⁴⁶ With the excellent properties of SnO₂ nanocrystals and the help of C₆₀, the electronic transmission efficiency of the PSC was greatly improved, and finally, a PCE of 18.8% was obtained. The introduction of a glycine layer can be used to adjust the interfacial stress caused by the lattice mismatch between SnO₂, and perovskite and the interface interaction between SnO₂ and perovskite is enhanced by hydrogen bonds and/or electrostatic interactions between amino acids and perovskite frameworks.¹⁴⁷

The position of the buffer layer is also not fixed; it can be between the perovskite layer and the ETL, or between the ETL and the electrode. In p–i–n PSCs, the materials in the latter application often have high resistance to water and oxygen, preventing water and oxygen in air from damaging the perovskite. In addition, some materials can also prevent metal electrode materials from leaking into the internal functional layer. In n–i–p devices, the buffer layer between the electrode and ETL may provide a smoother ETL growth platform, regulate energy levels, transfer electrons, etc. Two-dimensional carbide (MXene) can be used for modifying the FTO/SnO₂ ETL. After the introduction of the MXene layer, a better energy level arrangement of FTO/MXene/SnO₂/perovskite is realized.¹³⁷ In this structure, the working principle of the optimization of the energy level can also be attributed to the functional groups that combine with SnO₂. The device achieves 20.65% efficiency, with extremely low saturation current density and negligible hysteresis. Wang et al. used a low temperature treated amorphous tungsten oxide (a-WO_x) and SnO₂ to form a mixed ETL.¹³⁸ The a-WO_x buffer layer was prepared using vacuum evaporation with a thickness of 5–15 nm. Comparative experiments were conducted at different film thicknesses, revealing that device performance reached its optimum at 10 nm. It is worth noting that a-WO_x does not seem to undergo very complex chemical bonding with SnO₂, and its function is simple: it acts as a hole isolation barrier, as shown in Fig. 6c. Coincidentally, it also has suitable energy levels and excellent electron mobility, so its existence greatly enhances the overall performance of the device.

It should be noted that sometimes SnO₂ can be used as an intermediate layer by itself. Dang's team improved the ZTO/perovskite interface by introducing a thin SnO₂ interlayer. It was found that the Zn₁Sn₁O_x (with the thickness around 40 nm)/SnO₂ ETL device exhibited the highest power conversion efficiency (PCE) of 19.01% and better stability.¹⁴⁸ The excellent electrical properties of SnO₂ are the key reason for the progress of the device.

2.3.2.2.2 Self-growth buffer layer. Almost all the methods for obtaining the buffer layer described above are to prepare the buffer layer solution first, and then directly apply it on the functional layer by spin coating or other methods. This method of preparing a functional layer separately usually leads to a more complicated preparation process and also carries the risk of film growth failure. In fact, the buffer layer can also be obtained by reactive growth between functional layers through technical means. This method of self-growing the buffer layer avoids the above-mentioned problems. Min et al. provided a new idea for the obtention of a buffer layer through direct growth.¹⁴⁴ First, a SnO₂ thin film (Cl-bSO) was prepared by using ETL solution containing Cl⁻, and then Cl-cPP (Cl⁻ containing FAPbI₃ perovskite precursor) solution was prepared by mixing MACl and other solutions as the Cl⁻ source into FAPbI₃ precursor solution. Using Cl⁻ containing FAPbI₃ perovskite precursor solution to grow on the perovskite layer on Cl-bSO. Because Cl-bSO contains a large number of Sn–Cl bonds, when Cl⁻ in Cl-cPP comes into contact with these chemical bonds at the interface, it will react and form an FASnCl_x intermediate layer under the action of FA⁺. Due to this special synthesis method, the ETL/buffer layer/perovskite layer is almost integrated, which greatly reduces the generation of interface defect states and makes the transmission of electrons more efficient. They finally achieved the certified PCE of 25.5%. Zhou et al. immersed a SnO₂ ETL into the TiCl₄ solution, and TiCl₄ reacts with the –OH on the SnO₂ surface to form a Ti–O–Sn bond. The Cl⁻ will also participate in the whole reaction and finally form a TiO_xCl_4−2x buffer layer after the solution hydrolysis at 75 °C.¹⁴⁵ Excess ions or oxygen vacancies will participate in the whole reaction and form the buffer layer. Therefore, the defect states are also reduced, and the ETL level is optimized. The TiO_xCl_4−2x layer improves the permeability of the SnO₂ layer surface to the perovskite precursor, and helps to form a denser, nonporous perovskite film on the surface.

Sargent's team used NH₄F to modify the surface of the ETL.¹⁴¹ As shown in Fig. 6d, the NH⁴⁺ in NH₄F is weakly acidic and can react with the terminal hydroxyl group (OH_T) on the surface of SnO₂. F⁻ will replace the OH_T group and enter the defect site on the surface of SnO₂ during the reaction. The samples treated with NH₄F had higher quenching efficiency, indicating that NH₄F treatment improved the surface of SnO₂ and was conducive to the extraction of electrons from perovskite films. A PCE of 23.2% was finally achieved by this solar cell. Therefore, it can be used in interface modification engineering too.

Compared to methods such as spin coating that directly prepare buffer layers on SnO₂, self-grown buffer layers are typically extremely thin (approximately several nanometers) and grow directly on the ETL surface. We believe that thicker layers must balance properties such as energy level alignment, surface wettability, and electron transport capability. In contrast, thinner layers, due to the presence of the quantum tunneling effect,²⁷⁰ appear to focus more on enhancing the surface properties of the ETL, such as improving hydrophilicity, providing support or connection for active layer growth, altering the conductivity of the ETL, and reducing defect states. Self-growth of the buffer layer is an interesting research direction, but we believe it also raises several issues that require further exploration. First, the reaction range needs to be strictly controlled to avoid contamination of the ETL. Second, the surface morphology needs to be studied. Extensive surface defects are detrimental to PSCs, and it is necessary to examine the surface of the reactively grown buffer layer and the interface between the buffer layer and the ETL. Third, the impact of by-products on the overall device needs to be minimized.

Actually, interface engineering involves a wide range of technologies, and the boundaries between different approaches are subtle; here, we summarize the technologies of applying an obvious modified layer (ETL/middle layer/buffer layer, etc.) or a modified material for interface engineering to distinguish from doping technology. Fig. 7 shows the working principle of the main materials mentioned above for intuitive comparison.

Fig. 7 The diagram of basic modification for the SnO₂ ETL. (a) Figure reproduced with permission.¹³⁰ Copyright 2023, John Wiley and Son. (b) Figure reproduced with permission.¹⁴⁴ Copyright 2021, The Author, under exclusive licence to Springer Nature Limited. (c) Figure reproduced with permission.⁹⁶ Copyright 2022, American Chemical society. (d) Figure reproduced with permission.¹⁰¹ Copyright 2019, American Chemical Society. (e) Figure reproduced with permission.¹¹³ Copyright 2020, John Wiley and Sons.

3. Zinc oxide

3.1 Basic properties and advantages

As another common metal oxide that can be prepared by a low-temperature processing method, zinc oxide (ZnO) exhibits great performance in the applications of PSCs. The inherent nature of ZnO provides great help for high performance of devices: having (1) good light transmittance in the visible light range; (2) high electron mobility; (3) diverse nanostructure; (4) ease of modification.^149–151 Compared with SnO₂, ZnO often has higher electron mobility (205–300 cm² V⁻¹ s⁻¹ (ref. 149)), and lower processed temperature.¹⁵¹ Actually, TiO₂ was once the first generation of the MO ETM, and SnO₂ follows closely behind. Nowadays, for these excellent properties, ZnO has gained more attention, especially in flexible perovskite solar cells.

3.2 Fabrication methods of ZnO

Like the SnO₂ ETL, the ZnO ETL has many preparation methods, each of which has its own characteristics, advantages, and application scenarios. In addition, different preparation methods can also lead to different structures of ZnO, resulting in different electrochemical properties. Here, a summary and organization of different methods will be provided.

3.2.1 Solution processed method. This low-temperature processed method is also suitable for the preparation of the ZnO ETL, and its varieties are diverse, including spin coating, sol–gel method and others. The most typical method is the sol–gel method. The basic processes have been mentioned before. This method has some advantages, such as excellent stability under ambient conditions, superior light transmission, and effective layer formation.¹⁵²

The typical precursor solution for the fabrication of ZnO is zinc acetate dihydrate (ZnAc₂),^153–155 as shown in Fig. 8a. However, this could induce severe surface defects,^156,157 and the organic compounds will remain for quite a while.^158,159 Chavan et al. employed a diethylzinc(Et₂Zn) DMSO solution as the zinc source.¹⁶⁰ Et₂Zn was injected into dimethyl sulfoxide (DMSO) solution to form ZnO QDs, which were then spin-coated on FTO. In this way, the quality and purity of the film have been improved, as the surface performance. Runjhun et al. also used Et₂Zn as a zinc source, and with a wet organometallic process, successfully fabricated a ZnO QD film.¹⁶¹ By dissolving Et₂Zn and betaine (N,N,N-trimethylglycine) in tetrahydrofuran, Zn²⁺ reacts with amino and carboxyl groups in betaine to form stable intermediates. Finally, these intermediates are stirred in air and oxidized into ZnO quantum dots using H₂O and O₂ from the atmosphere. This method can be used to precisely control the structure and quality of nanomaterials by selecting different organic ligands and reaction conditions. Zheng and co-workers used another material—Zn(NO₃)₂·6H₂O as the precursor and achieved a ZnO film with low surface roughness and high electron transport efficiency.¹⁶² Zhou et al. employed the stable aqueous solution of an ammine–hydroxo zinc complex, [Zn(NH₃)_x](OH)₂ for fabricating the ZnO ETL.¹⁶³ In this way, the work function of ZnO thin films can be reduced and the V_oc can be improved. Kelly's team spin-coated ZnO nanoparticle solution on the glass substrate three times to obtain a continuous, smooth ZnO film and this can also be seen as a solution-processed method.¹⁶⁴

Fig. 8 (a) The standard sol–gel process;¹⁶⁰ Copyright 2022, John Wiley and Sons. (b) Schematic of a rotating cylinder reactor showing one cycle of ZnO S-ALD using diethylzinc and ozone as reactants;¹⁶⁹ Copyright 2015, American Vacuum Society. (c) Device structure of PSCs with a ZnO seed layer;¹⁷³ Copyright 2016, Tsinghua University Press and Springer-Verlag Berlin Heidelberg. (d) Graphical depiction of the inside of the sputtering chamber;¹⁷⁵ Copyright 2019, American Chemical society. (e) The process of the combustion method for preparing the ZnO-ETL;¹⁶² Copyright 2019, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim. (f) Representation of the synthesis route and the flowchart utilized for producing silver-doped zinc oxide nanoparticles (Ag@ZnO NPs).²⁰² Copyright 2024, Elsevier.

There are various precursors of ZnO, and we will make a brief summary here. First, the traditional acetate dihydrate often decomposes into ZnO by high-temperature annealing. The process is relatively simple, but the morphology of the ZnO film is rough, which can be easily foreseen. Next, the working process and advantages of Et₂Zn have been mentioned before, and it is often used for forming ZnO QDs. Zn(NO₃)₂·6H₂O has the same ZnO preparation process as acetate dihydrate, and with NO₃⁻, which is more volatile during annealing, it can form a high-quality polycrystalline structure, but due to incomplete decomposition and other problems, the quality of the film will also be reduced. The last one is [Zn(NH₃)_x](OH)₂, which is also decomposed by annealing to obtain ZnO. However, due to the presence of NH₃⁻, with a faster volatilization rate, the structure of the whole solution may be destroyed before the formation of large-scale ZnO, so its crystallinity should be less than that of Zn(NO₃)₂·6H₂O.

3.2.2 Atomic layer deposition. Traditionally, diethylzinc is the zinc source for ALD-ZnO.^165–168 The ALD method will activate oxygen vacancies, which are conducive to the growth of CH₃NH₃PbI₃.¹⁶⁵ The film can be grown layer by layer, forming a uniform coating without holes on a large area and with a three-dimensional structure. Pietruszka's team found that the hardness and Young's modulus of ALD ZnO thin films were higher, making them more resistant to plastic deformation and friction, thereby improving their wear resistance.¹⁶⁸ Nevertheless, the ALD process depends on the chemisorption reaction on the substrate, which requires enough dangling bonds on the substrate. So, it's difficult to form ideal films on an inert electrode like graphene. To solve this problem, Xu et al. increased the van der Waals forces on the substrate surface by introducing a molecular layer deposition (MLD) precursor—ethylene glycol (EG) onto the graphene surface, making use of ALD on it to achieve a uniform ZnO film.¹⁶⁷ In addition, spatial ALD(S-ALD) can use ozone and diethylzinc as the reactants. Sharma and co-workers used the device shown in Fig. 8b to prepare ZnO, and they found that at a slower rotation rate, a larger aperture can obtain a more uniform Zn coverage, and this method can be adapted for the roll-to-roll operation.¹⁶⁹

3.2.3 Sputtering method. This method has some advantages, such as a high deposition rate, easy control of the fabrication process, good film properties, etc.^170,171 The most widely used sputtering method is magnetron sputtering. The team from Xiamen University sputtered ZnO nanorod films with a c-axis (vertical to substrate) arrangement.¹⁷² The causes of this structure are summarized as follows: when the highly active Zn atom comes into contact with the substrate at relatively low temperature, its energy decreases sharply, and it is difficult to diffuse on the substrate surface. Therefore, ZnO particles will grow on the limited area in contact with the Zn atom, forming this c-axis structure. Li et al. used the magnetron sputtering method to prepare a ZnO seed layer on FTO.¹⁷³ After that, the c-axis preferred orientation of the seed layer was obtained by annealing in oxygen. Finally, the c-axis ZnO nanorods were formed at 90 °C by the hydrothermal method, as shown in Fig. 8c. Compared to the two c-axis ZnO fabrication methods, the length and diameter of the nanorods can be precisely controlled by the hydrothermal method. As mentioned before, the oxygen content of magnetron sputtering operation has a lot of influence, and the oxygen-enriched environment can effectively reduce the hydroxyl concentration in ZnO, thus inhibiting the decomposition of perovskite films and promoting the improvement of their crystallization and morphology.¹⁷⁴

However, magnetron sputtering films will have some traps. To solve these problems, Abhishek and co-workers proposed a new sputtering technology—high-working-pressure sputtering (HWP).¹⁷⁵ In the process of HWP, by controlling the working pressure in the sputtering chamber (Fig. 8d), the energy of high-energy particles (such as Ar) is reduced, thus reducing the number of recombination centers and defects. A PCE of 17.3% was achieved by using ZnO PSCs sputtered by HWP. Banerjee et al. used a low-temperature vacuum-based sputter method and achieved a ZnO film with good crystallinity and high transparency.¹⁷⁶ This method also revealed the application potential of ZnO in flexible equipment.

3.2.4 Hydrothermal method. Hydrothermal fabrication of ZnO thin films is one of the commonly used methods with simple operation, low cost, and fast preparation speed. Khalid's team used this method to prepare a double-layer ZnO ETL.¹⁷⁷ From the research, the density of nanorods can be controlled by adjusting the concentration of the ZnO precursor, which can affect the penetration of the perovskite absorption layer and the contact with the ETL. In their follow-up study,¹⁷⁸ they achieved different structures of ZnO with different temperatures (among them, the nanowell is formed at 80 °C, and the nanodisk at 90 °C). The effect of temperature on ZnO nanostructures seems to be very large. We deem that at higher temperatures (90 °C), the reaction speed is faster and the crystal grows faster. At the same time, it also grows on different crystal planes, which leads to disordered crystal orientation. At a lower temperature, the reaction is milder, and the crystal seems to have enough time to grow in a specific direction, thus forming a vertical structure similar to the ‘wall’. ZnO nanowells exhibit higher performance for the reason that it allows electrons to transfer to the electrode quickly in a specific direction, and improves the efficiency of electron extraction and transmission. Sekar's team studied the influence of growth time on device performance by growing ZnO nanowires on AZO.¹⁷⁹ From the result, as the growth time increases, the length of NWs significantly increases, leading to a decrease in NW density and an increase in porosity. Longer NWs may result in uneven deposition of perovskite layers, increasing short-circuit current and recombination problems in the circuit, thereby reducing device performance. Li and co-workers used the hydrothermal method to grow ZnO nanorods (NRs).¹⁸⁰ A good interface can be formed between ZnO nanorods and perovskite layers, ultimately achieving a conversion efficiency of 17.3%. Other structures like nanotubes,^182,183 nanocandles,¹⁸⁴ and nanoparticles^185,186 have also been reported in previous studies.

As we can see in this part, different nanostructures of ZnO can be formed by the hydrothermal method. Here, we make a brief summary of the various nanostructures achieved by this method in Table 3.

Table 3 The nanostructures achieved by the hydrothermal method. (a) Nanodisks and (b) nanowells. Reproduced with permission.¹⁷⁸ Copyright 2018, Springer-Verlag GmbH Germany, part of Springer Nature. (c) Nanorods. Reproduced with permission.¹⁸⁰ Copyright 2017 Elsevier B.V. (d) Nanocandles. Reproduced with permission.¹⁸⁴ Copyright 2015 Elsevier B.V. (e) Nanotubes. Reproduced with permission.¹⁸² Copyright 2018, Science China Press and Springer-Verlag GmbH Germany. (f) Nanoparticles. Reproduced with permission.¹⁸⁵ Copyright 2013, Springer Science Business Media Dordrecht. (g) Nanowires.Reproduced with permission.¹⁷⁹ Copyright 2022 by the authors

Structure	Diagram	Cause of formation	Characteristics	Ref.
Nanodisk	(a)	At 90 °C, the precursor hydrolyzes rapidly and has no constant growth direction	Disk structure may increase grain boundaries	178
Nanowell	(b)	At 80 °C, the precursor slowly hydrolyzes to form a porous sponge structure	Perovskite can leak and increase the interface contact with perovskite	178
Nanorods	(c)	After embedding the ZnO seed layer, one-dimensional growth was promoted by adjusting the pH value	Provide a direct electronic transmission path	180
Nanocandles	(d)	The doping of Au atoms changes the lattice stress, resulting in delamination or tip growth of nanorods	Due to the introduction of Au, the conductivity has been significantly improved	184
Nanotubes	(e)	After being grown into nanorods by the hydrothermal method, the more easily soluble polar [001] plane is corroded by ammonia solution	Tube like structure reduces carrier recombination and adapts to flexible substrates	182 and 183
Nanoparticles	(f)	By adding the corresponding precipitant and reacting with Zn(NO3)2	There are various preparation methods and a wide range of applications, but due to the possible production of related by-products by precipitants, there may be more impurities or defect states in the film	185 and 186
Nanowires	(g)	Growing on an AZO substrate, due to the lattice matching properties of AZO, it directly promotes the directional growth of ZnO	High electron mobility and strong light capture capability	179

---

3.2.5 Electrochemical deposition. Electrochemical deposition with fast speed, a simple fabrication process, and ability for creating large surface area is also widely used in the fabrication of ZnO.

In 2014, Zhang et al. grew nano-ZnO thin films by electrochemical deposition.¹⁸⁷ The film grown by this method has high quality and can be accurately controlled at low temperature. Then in 2015, Zhang and colleagues used electrochemical deposition technology to prepare ZnO on flexible substrates.¹⁸⁸ They also analysed the effect of different deposition times on the performance. This research deepens people's understanding of electrochemical technology and also proves the feasibility of this technology in flexible solar cells. Christian and co-worker used 0.08 M zinc nitrate (Zn(NO₃)₂) solution as the deposition solution, and zinc wire as the electrode to deposit ZnO on ITO at 60 °C.¹⁸⁹ Like Zn(NO₃)₂ solution, Zn(NO₃)₂·5H₂O is also the typical deposition material.¹⁹⁰ In 2018, Garcia's team chose to use ZnCl₂ and KCl mixed solution as the electrolyte and also achieved an ideal result.¹⁹¹

Electrochemical deposition also has some drawbacks: high energy consumption, complex post-processing, accurate control of parameters such as time and temperature, and last but not least, environmental issues.

3.2.6 Other methods. The ZnO films with high quality can also be achieved by other methods. Zheng et al. fabricated the ZnO film by the combustion method, as shown in Fig. 8e.¹⁶² Compared with the sol–gel method, the combustion synthesis method can be used to prepare ZnO thin films with high density, high crystallinity, and a self-passivated surface. This is because the combustion method can increase the temperature by a certain value in a short time, so as to quickly remove the organic ligands and other solvents. ZnO will also crystallize rapidly, resulting in a cleaner (less impurities, self-passivation) and denser film. Due to the existence of the self-passivation effect, its conductivity has also been improved. c-ZnO thin films prepared by combustion synthesis have ideal surface morphology and chemical properties, which are conducive to the growth of large-sized and highly interconnected perovskite grains. Khalid's team prepared a ZnO ETL with the electrospray method.¹⁹² Through the electrospray technology, ZnO nanostructures with specific morphology can be formed, such as nanosheets (NS) and nanoparticles (NP), which help to improve the charge transfer efficiency in perovskite solar cells. Kumar et al. chose to use chemical bath deposition for making ZnO films.¹⁹³ They applied the CBD-ZnO ETL on a flexible substrate to realize a flexible PSC. The all-inkjet-printing approach^194,195 is another useful way, especially in the fabrication of flexible PSCs, and this method forms the desired pattern directly on a flexible substrate by precisely controlling the way ink is ejected, without the need for traditional photolithography and etching steps. Tavakoli et al. achieved a PCE of 15.2% based on these ZnO/rGO QDs.¹⁸¹ By adding the surface ligand rGO to the solvent, the growth of ZnO nuclei can be effectively restricted, limiting their size within the quantum range, resulting in ZnO/rGO QDs thin films.

As mentioned before, different methods have different advantages and drawbacks, and people should use a suitable method according to the specific application. Furthermore, the development of new preparation methods can also be used as a hot direction to promote the development of photovoltaic devices.

Thus far, the primary low-temperature methods for fabricating MO ETLs have been summarized and explained. Research into low-temperature fabrication techniques aims to integrate them with roll-to-roll processes or large-area device fabrication methods to achieve commercialization. In Table 4, we analyze and discuss the compatibility of the mentioned processes with large-area fabrication techniques.

Table 4 Adaptability analysis of different preparation methods for R2R and large-area device preparation

---

3.3 Modifications of the ZnO ETL

The ZnO ETL has been widely used in PSCs in the last few decades, and their performance has achieved huge progress. However, ZnO also has some defects: ZnO and perovskite materials (especially lead-containing materials) will have serious interface recombination,¹⁹⁶ which is due to the alkalinity of ZnO,¹⁹⁷ which will affect the stability of devices. Moreover, when the temperature is too high, ZnO can promote the deprotonation of formamide cations and promote the decomposition of the perovskite layer.¹⁹⁸ ZnO has an open hexagonal close-packed lattice structure, and Zn atoms occupy only half of the tetrahedral positions, resulting in a large number of octahedral vacancies. These defects may act as charge recombination centers and leakage current paths, leading to charge transfer barriers and water penetration.¹⁹⁹ Modification of the ZnO ETL is the main means to solve these problems.

3.3.1 Additive engineering. The basic working principle of additive engineering has been mentioned above, and it is also useful for ZnO ETLs.
3.3.1.1 Metals and metal ions. For the doping of ZnO, most researchers choose to use metal materials and metal ions as dopants. Metal doping helps to improve the interface morphology, reduce the defect density, modify the energy level structure, and achieve performance improvement and breakthroughs.

Cardozo's team used Ag to dope ETLs.²⁰² They synthesized silver-doped zinc oxide (Ag@ZnO) by using a safe and sustainable by design (SSbD) method. The specific method is mixing Zn(NO₃)₂·6H₂O with AgNO₃, then adding VERDEQANT, and finally heating and decomposing into the Ag@ZnO film as shown in Fig. 8f. This method avoids the use of ethylene glycol; instead, the vegetal biocompatible polymers are used as gels, which are renewable and eco-friendly.

Cu doping leads to a decrease in the Fermi energy level of ZnO and the formation of a shallow intermediate energy level, which is better aligned with the conduction band, thus improving the charge extraction efficiency.²⁰⁰ Another transition metal—nickel was also used by Guchhait's team.²⁰¹ Ni-doped ZnO NRs have a low electronic band gap and good conductivity. Both of these metal materials mainly enter the ZnO lattice by replacing Zn²⁺, which changes the electron cloud distribution in the ZnO lattice and thus affects the electron transport characteristics and energy levels of the entire ETL.

Gantumur et al. focused on the study of tungsten-doped zinc oxide.²⁰³ Tungsten has attracted much attention due to its low reactivity and ion radius (0.064 nm) similar to Zn²⁺ (0.074 nm).^204,205 By using aloe vera leaf extract as a powerful reductant, it can effectively reduce metal ions and prevent the aggregation of ZnO:W nanoparticles. The working principle of W is also similar to Cu and Ni mentioned above, replacing Zn²⁺ to enter the lattice and improve the conductivity of the ETL. W-doped ZnO nanoparticles (NPs) were synthesized by a green synthesis method, and the whole process can be seen in Fig. 9a. The ZnO:W prepared by this method also has better crystallinity and larger grain size as shown in Fig. 9b(a) and (b), which is helpful to form a uniform perovskite layer, provide better energy level matching, and reduce the recombination of charge carriers.

Fig. 9 (a) Schematic illustration of the W-doped ZnO ETL fabrication, and the inset shows the phytochemicals' effect on the formation of W-doped ZnO NPs; (b) schematic illustration of the degradation of the PSCs with (a) pristine ZnO and (b) ZnO:W ETLs;²⁰³ Copyright 2024, American Chemical society. (c) Top-view SEM images of the CsPbI₂Br perovskite film based on (a) and (c) ZnO and (b) and (d) ZnO:Ca ETLs. AFM images of the CsPbI₂Br films based on (e) ZnO and (f) ZnO:Ca ETLs;²¹³ Copyright 2023, AIP Publishing.

Gallium ion doping also has a good effect.²⁰⁶ Because the ion radii of Ga and Zn are similar, Zn²⁺ in GZO can be replaced by Ga³⁺ without interfering with the ZnO structure. The conduction band of GZO is higher than that of CsGeI₃, which ensures that electrons can be effectively transferred from the calcium titanite layer to the GZO layer. The doping of Ca²⁺ is also the same. Because of the doping of Ca²⁺, which enhances the hydrophilicity of ZnO:Ca, the CsPbI₂Br film on the Ca²⁺-doped ZnO ETL has a more uniform surface and larger grain size, as shown in Fig. 9c, and the efficiency of ZnO:Ca based solar cells reaches 16.39%.²¹³

AZO is the new ETL which uses Al to dope ZnO; it has high electron mobility and good energy level matching with perovskite materials.²⁰⁷ Yang and co-workers used this material.²⁰⁸ We will focus on the PL spectrum test shown in Fig. 10a. Al³⁺ replacing Zn²⁺ and entering the ZnO lattice can effectively improve the crystal quality of ZnO and reduce oxygen vacancies. The peak value of PL indicates the level of carrier recombination. From the fact that a doping concentration of 7% has the lowest PL peak, it can be seen that the carrier recombination probability is the lowest at this time. However, as the doping concentration increases, carrier recombination becomes more severe, which may be due to excessive Al doping introducing new defect states or damaging the original ZnO lattice structure. Pramothkumar and co-workers used a novel method; they doped Al and Sn into ZnO.²⁰⁹ Using Al–Sn co-doped ZnO as the ETL can successfully adjust the optical band gap position and enhance the conductivity of the ETL. Research has shown that Al and Sn exhibit a synergistic effect. Replacing Zn²⁺ with Al³⁺ will introduce additional charges and improve conductivity. The substitution of Sn²⁺ will change the local electron cloud distribution, thereby altering the band gap. In addition, the quantum confinement effect introduced by doping makes the energy state of electrons more discrete, which improves the optical absorption and the generation of photogenerated carriers.

Fig. 10 (a) PL spectra of perovskite CsPbBr₃ films based on AZO films with different Al doping ratios;²⁰⁸ Copyright 2023, American Chemical society. (b) (a) 1D g-C₃N₄ additive added ZnO layer charge transport mechanism. (b) Schematic diagram of the fabricated device structure with compounds (c) and (d) the 1D g-C₃N₄ additive without and with the FTO/perovskite interface electron transportation process;²¹⁶ Copyright 2023, Elsevier. (c) SEM images of (a) pristine ZnO and (b) PbI₂:ZnO films; contact angle tests of the (c) pristine ZnO and (d) PbI₂:ZnO films.²¹⁷ Copyright 2022, John Wiley and Sons. (d) (a) Steady-state PL spectra and (b) XRD patterns of ZnO and Mg_0.1Zn_0.9O on ITO substrates;²²³ Copyright 2021, Elsevier. (e) Schematic diagrams of band alignment and charge recombination in the cells with the (a) ZnO single layer and (b) TiO₂/ZnO bilayer as the compact layers, respectively;²²⁵ Copyright 2015, Royal Society of Chemistry.

Khan's team used different Cd/Zn molar ratios for doping.²¹⁰ Cd doping can reduce the grain size, help to reduce the grain boundary, and reduce the resistance of electron transmission. Cd doping can also reduce the non-radiative recombination in the material, thus reducing the recombination loss of electron–hole pairs. Moreover, appropriate Cd concentration helps to improve the internal reflection of light in the ETL and perovskite layer, reduce the escape of light, increase the light absorption, and thus improve the J_sc.

Other metal dopants are briefly summarized below, and their main doping principles are similar to those of the above materials. If there are significant differences, we will mention them in the following text. Li doping can change the energy band structure of ZnO, improve the minimum conduction band (CBM), fill some oxygen vacancies, and reduce electron traps, so as to improve the mobility of electrons and the conductivity of devices.²¹¹ Mg-doped ZnO nanofibers have a uniform nanometer diameter and porous structure, which is conducive to providing more electron transmission channels.²¹² Mn doping reduces the alkalinity of ZnO and effectively controls the deprotonation of perovskite at the Mn: ZnO/perovskite interface. Mn⁴⁺ doping produces two free electrons in ZnO, which leads to the improvement of the conductivity of Mn:ZnO thin films.²¹⁴

3.3.1.2 Others. Other materials can also be added into the ZnO ETL for improving the performance.

In the ZnO lattice, carbon doping can replace the position of oxygen atoms or enter the interstitial positions of the lattice, introducing new electronic states, which is helpful to improve the transmission and collection of electrons. Gouda et al. used carbon-doped zinc oxide (ZnO/C) derived from metal–organic framework (MOF) as the electron transport layer in high-performance perovskite solar cells.²¹⁵ By calcining MOFs, partial carbon loss was achieved, and ZnO/C composite materials with a high specific surface area and uniform pore structure were prepared, which helps to improve the interfacial adhesion between the ETM and perovskite.

Rajaram and co-worker added 1D graphitic carbon nitride into the ZnO ETL.²¹⁶ 1D g-C₃N₄ and ZnO nanoparticles were mixed by high-energy ball milling to prepare composite materials for ETLs. The existence of additives on the ZnO layer reduces the surface defects. The 1D g-C₃N₄ additive acts as an electron transport bridge in the ZnO layer, which improves the transmission efficiency of electrons from the perovskite layer to the ETL, as shown in Fig. 10b.

Wu et al. introduced PbI₂ to increase oxygen vacancies in ZnO.²¹⁷ Due to the large ionic radius of Pb²⁺, it is difficult for it to enter the lattice of Zn²⁺, which can lead to more defects, such as oxygen vacancies, during the formation of the thin film. The increase in oxygen vacancies in ZnO not only promotes the growth of all inorganic perovskite layers, and improves their morphology and orientation order, but also forms a favorable energy level arrangement with the perovskite layer. As shown in Fig. 10c, PbI₂:ZnO thin films exhibit a smoother surface and higher hydrophilicity than the original ZnO thin films, which is conducive to the formation of a more uniform perovskite layer.

Zhang et al. explored cesium salts with functional anions such as acetate (AC⁻), fluoride (F⁻), and trifluoroacetate (TFA⁻) to regulate the deposition of ZnO films.²¹⁸ These functional anions can coordinate with Zn²⁺ and Pb²⁺ ions to achieve defect passivation of ZnO and top perovskite films. Cs⁺ ions can also reduce hydroxyl defects and form interface dipoles on the ZnO surface through Zn–O–Cs bonds, thus establishing more stable and efficient interface electron transport.

Polymers are widely used too, and we have specifically explained that the interaction between polymers and metal oxides is mainly attributed to the reaction and bonding of different functional groups with the surface of MO. This rule also applies here. Polyethylene glycol (PEG) has a strong interaction with the ZnO surface through its internal functional groups, such as hydroxyl and ether bonds, to achieve the effect of adsorption and filling oxygen vacancies, which can effectively passivate the surface defects of ZnO, increase the conductivity, improve the stability of ZnO solution, and reduce the aggregation of ZnO nanoparticles.²¹⁹ Poly(vinyl alcohol) (PVA) contains a large number of hydroxyl groups and has good hydrophilicity, which makes the preparation of the solution easier.²²⁰ In a polar solution, it can promote the dissolution of ZnO NPs and help to form high-quality and uniform films.

3.3.2 Interface engineering. Here, we also focus on the interface between the ETL and perovskite/electrode layer, and combine the interface engineering of ZnO through interface modification, using a double-layer ETL, buffer layer, etc.
3.3.2.1 Bilayer ETL. The double electron transport layer can greatly improve the electrical performance of devices and the materials for combination are various.

Mahmood's team synthesized a bilayer ZnO ETL by the low-temperature hydrothermal method.¹⁷⁷ The structure consists of horizontally arranged nanosheets (as the bottom layer) and vertically arranged nanorods (as the top layer). The former provides a basic support for the growth of thin films with a uniform surface and excellent electron transport performance, while the latter provides a ‘fast road’ for electron transport. In addition, due to the special structure of nanorods, it can also increase the contact area with the perovskite layer. In 2019, they made improvements to the device.¹⁹² They used ZnO nanosheets and ZnO nanoparticles as a bilayer ETL with polyethyleneimine (PEI) coated on them. PEI can reduce the working function, accelerate electron transfer, and improve the contact between the perovskite and the ETL. The double-layer ZnO ETL coated with PEI further improves the power conversion efficiency of the device to 16.39%, exhibiting no hysteresis phenomenon.

Mahmud and co-workers studied single, double, and triple ZnO ETLs by preparing layers with different thicknesses (25, 45, and 60 nm, respectively).²²¹ The double-layer ZnO ETL provides high transparency and low reflectivity, ensuring that more light energy is absorbed by the perovskite layer, balancing the needs of electron transport and light absorption. Compared with the double-layer device, the triple-layer ETL device greatly increases the complexity of the device preparation process due to the increase in the number of interfaces, and may also introduce more interface defects. Therefore, the effect is not as good as the double-layer device. This also reflects that blindly increasing the electron transport layer is not an effective means to improve the performance of the device. The greatly increased preparation difficulty, reduced energy level compatibility, negatively affected light-transmittance of the n–i–p device, as well as the increased interface defects or interactions, can all influence the overall device performance. The core of a double-layer ETL to improve device performance is to realize the complementarity or synergy between different ETL performances, which is definitely not a simple superposition of conductive performance. The key is to find suitable and matching ETL materials or preparation methods.

Wu et al. proposed an ETL design of ZnO/Al-doped ZnO (AZO) bilayer films by low-temperature solution treatment based on stepped band alignment.²²² The conduction band minimum (CBM) of AZO is slightly higher than that of ZnO, which is helpful for the extraction of electrons from the CH₃NH₃PbI₃ active layer.

Li et al. synthesized a ZnO/Mg-doped ZnO (Mg_xZn_1−xO) bilayer ETL by optimizing the annealing temperature of the ZnO ETL and the Mg content in Mg-doped ZnO thin films.²²³ As shown in Fig. 10d, the XRD image shows that the doping of Mg does not significantly change the lattice structure of ZnO, which makes the Mg:ZnO film better match the pure ZnO film. In addition, it is not difficult to find from the PL image that the carrier recombination of the double ETL is reduced, and the perovskite grown on it also has larger crystal cells.

TiO₂ is a typical MO ETL material. As shown in Fig. 10e, the TiO₂/ZnO bilayer may form a type II band structure, which is conducive to the transmission of electrons at the TiO₂/ZnO interface and inhibits the charge recombination at the interface of the ZnO layer and ZnO/perovskite absorption layer.²²⁵

Samavati's team used Ag-doped ZnO and ZnO to prepare a bilayer ETL.²²⁴ Ag-doped ZnO has excellent surface properties, and the surface morphology of the ETL was optimized by adding ethanol to the dispersion of Ag-doped ZnO NPs, as shown in Fig. 11a.

Fig. 11 (a) AFM images of (A) glass/ZnO/H₂O–Ag-doped ZnO 1 wt%, and (B) glass/ZnO/H₂O–ethanol mixture-Ag-doped ZnO 1 wt% bilayer ETLs;²²⁴ Copyright 2024, Springer Nature. (b) (a) Schematic diagram of the preparation of the iodine ligand-modified ZnO/PbS-TBAI ETLs, (b) schematic illustration of the replacement of oleic acid ligands by iodide ion ligands on PbS QD films by TBAI treatment;²³⁶ Copyright 2022, Elsevier. (c) Schematic illustrating the extraction and transportion of electrons in flower-like TiO₂ ETL based devices;²⁴¹ Copyright 2016, Royal Society of Chemistry. (d) (a) Schematic device architecture and charge transport pathway in conducting substrates through TiO₂ NaPAs and (b) cross-sectional SEM image of TiO₂ NaPA based PSCs.²⁴⁹ Copyright 2021, Springer Nature.

SnO₂ has many advantages, as shown in the above content, and it can also be used as a part of a double ETL. Wu and co-workers achieved a V_oc of up to 1.15 V and PCE of 19.1% by introducing PSCs prepared by using a ZnO/SnO₂ bilayer ETL.²²² The recombination process in PSCs was analyzed by steady-state photoluminescence (PL) and time-resolved photoluminescence (TRPL) measurements. It was found that the ZnO/SnO₂ bilayer ETL reduced the recombination loss during electron transport. Dkhili et al. also found that this structure, with H₂O and KOH treatment on the SnO₂ layer, exhibits better wettability on glass substrates than on PET substrates, which helps to form a more uniform and higher quality perovskite layer.²²⁶ Good wettability can promote the uniform coating of perovskite precursor solution on the surface of the ETL, thereby facilitating the formation of uniformly dense perovskite thin films.

PCBM is the traditional ETL for PSCs. Qiu et al. used PCBM to combine the ZnO ETL and achieved a device with PCE exceeding 14%.²²⁷ As an organic material, PCBM can improve energy level alignment and reduce surface recombination, while the ZnO layer acts as a physical barrier to prevent direct contact between the metal electrode and perovskite layer. Zhang et al. also found that PCBM can be used as an interface modifier, which can not only improve the thermal stability of formamide lead iodide, but also significantly reduce the current voltage hysteresis phenomenon.²²⁸

We found that a bilayer ETL can be widely used in p–i–n PSCs. Spin-coating aqueous solutions onto perovskites is undesirable, as it can severely damage the perovskite active layer (AL). Therefore, the electron transport layer (ETL) materials used in conventional p–i–n PSCs are non-aqueous solutions such as PC₆₁BM or C₆₀. In a bilayer ETL structure, these non-aqueous ETMs form a protective film for the perovskite, reducing the impact of aqueous metal oxide solutions on the active layer. In related experiments, we believe that the density of the ETL, acting as a protective film, is crucial. Defects such as pinholes can damage the density of the AL. Furthermore, energy level matching requires attention. Electrodes (such as Ag and Au) can experience corrosion and diffusion after device fabrication, which can severely compromise device stability.²⁷¹ This necessitates improving the metal oxide ETL's ability to suppress metal ion diffusion. Furthermore, doping is crucial for improving the ETLs' resistance to water and oxygen. Adding a buffer layer such as BCP introduces additional interfaces, significantly increasing fabrication complexity. Therefore, we prioritize doping in these research studies.

3.3.2.2 Buffer layer. As mentioned in the above content, the ZnO ETL and perovskite layer will react and disrupt stability.

Let's take WO₃ as a simple example to start this part. WO₃ has a more conductive band position than perovskite and has better hydrophilicity. Adding WO₃ as a buffer layer between the perovskite layer and ETL can form a good energy level arrangement between the three, making electron transfer more effective. Next, it can make the preparation of perovskite films more uniform and the film quality better.²³⁵ This is the main task of a buffer layer. In order to make the buffer layer play a more effective role, many researchers will also focus on the interaction between the buffer layer and the perovskite layer or ETL layer, and make the buffer layer well-connected with other functional layers through chemical bonding, or even change the electrochemical properties.

Cheng et al. from City University of Hong Kong used PEI to prepare a buffer layer to solve this problem.²²⁹ PEI has a high molecular weight, which enables it to form a stable and uniform coverage on ZnO without aggregation or diffusion during annealing, and its strong intermolecular force can also prevent the invasion of water and oxygen molecules and protect the perovskite layer. Additionally, the electron affinity of PEI may be more matched with ZnO NPs and perovskite layers, facilitating electron extraction and transport, and reducing electron loss at the interface. Qiu et al. focused on the ZnO ETL modified by the PCBM layer, and they found that the PCBM layer, as an organic molecule, can form a tight bond with the ZnO layer and perovskite layer, which helps accelerate the extraction and transfer of electrons, thereby improving the efficiency of the solar cell.²³⁰

Here we need to make some explanations. The combination of PCBM and ZnO has been mentioned in many studies, and they are in the form of a double-layer structure. We distinguish and summarize the ‘buffer layer’ and ‘bilayer ETL’ according to the description of the research. However, for the main working mechanism of PCBM and ZnO, we give the following opinions: when PCBM is combined with ZnO as a double ETL, the research studies will focus more on the synergy between PCBM and ZnO, such as the formation of a gradient energy level; when PCBM is used as a buffer layer, we think that the research studies tend to see the two as a whole, emphasizing the auxiliary role of PCBM, such as isolating the perovskite layer and ZnO to alleviate the degradation of materials, etc., and the main body of the ETL is still ZnO. Finally, we still believe that the added ETL is a special buffer layer. So I hope you don't have to worry too much about the distinction here, just focus on the performance brought by the material itself.

Runjhun and co-workers coated the NH₄X solution on a ZnO QD layer.²³¹ The ZnO QD based devices treated with NH₄F passivation exhibited a high PCE of 21.9%, which is currently one of the highest performances reported in the literature among ZnO-based PSCs. After NH₄F treatment, the Fermi level of ZnO thin films shifts downward, which helps improve the alignment of energy levels between ZnO and perovskite layers and promotes the transfer of electrons from perovskite layers to ZnO layers. Furthermore, NH₄F can also passivate the surface of ZnO films and reduce its defects. Murugadoss et al. from Korea University studied different concentrations of NH₄F in the application of a modification layer.²³² They found that high concentrations of NH₄F inhibited the crystallization of ZnO. At appropriate concentrations, NH₄F treatment introduces fluorine atoms, which replace the hydroxyl (–OH) and oxygen vacancies on the surface of ZnO, helping to reduce leakage current in the device.

Tulus et al. reduced the defect density on the surface of ZnO nanorods by thermally depositing Au nanoparticles (approximately 4 nm) on the surface of ZnO nanorods.²³³ This sedimentation method avoids the contamination of the ZnO surface by chemical residues. However, although Au nanoparticles improve the initial performance of the solar cell, they do not enhance the stability of perovskite solar cells under continuous illumination. In 2022, they chose to use 25 nm C₆₀ as the modification layer.²³⁴ Research has shown that the C₆₀ layer leads to the formation of PbI₂-rich and Br-rich regions in the perovskite absorption layer, which reduces composite losses and improves operational stability. The solar cells with a C₆₀/ZnO ETL exhibit less pronounced or slower electrochemical dynamics, indicating that C₆₀ helps to suppress irreversible electrochemical processes at the ETL interface.

PbS QDs have been selected as surface modification materials due to their wide spectral absorption range, tunable bandgap, multi-exciton generation effect, and low-cost preparation technology.²³⁶ Both PbS and perovskite have six coordinated Pb atoms, which can be well matched with perovskite, and the surface morphology of the perovskite layer grown on PbS is better. As shown in Fig. 11b, by further depositing TBAI on the ZnO/PbS layer, the defect state on the surface of the functional layer is greatly suppressed, and its conductivity is improved due to the coordination reaction between iodine ions in TBAI and Pb on PbS QDs.

Liu and co-workers used AZO as the ETL and chose polydopamine (PDA) to conduct interface modification.²³⁷ The dopamine group with strong adsorption in PDA can closely connect the perovskite layer and ETL, which helps to reduce defect states and carrier recombination. In addition, the hydroxyl and amino groups in PDA interact with the perovskite precursor solution to induce the vertical growth of perovskite grains and improve the quality of perovskite crystallization. By using AZO:PDA as the ETL, a champion power conversion efficiency of 21.36% was achieved with a minimal hysteresis effect.

Here, we found that, unlike SnO₂, the buffer layers of ZnO ETLs are mostly added between the perovskite and the ETL. Few studies have placed it between the electrode and the ETL. This focus is primarily on addressing the problem of ZnO's inherent alkalinity damaging the perovskite and providing a more uniform and flat platform for perovskite growth. Of course, adding a modified layer between the ETL and the electrode to achieve a more suitable energy level alignment is also a viable experimental approach.

In Table 5, we summarized the main contents of the two metal oxide ETLs and also made some comparisons. We take this as the end of the main part of our review for your reference.

Table 5 The summary and comparison of different MO contents

---

4. Other low-temperature processed MO ETLs

Besides SnO₂ and ZnO, two typical low-temperature processed MO ETMs, there are also many useful MO ETMs suitable for low-temperature fabrication of flexible PSCs. This part will provide a brief summary.

4.1 Titanium dioxide

There are three crystal phases of TiO₂: anatase, rutile, and brookite. TiO₂ prepared at low temperature generally belongs to anatase and rutile. Many researchers have prepared TiO₂ ETLs using different low-temperature techniques, and as mentioned earlier, the TiO₂ ETL has also been combined with SnO₂ and ZnO to form a bilayer ETL structure. Here, we will introduce some low-temperature processed TiO₂ ETLs.

In 2014, Yella et al. prepared a nanocrystalline rutile phase TiO₂ ETL using the CBD method with TiCl₄ solution at 70 °C.²³⁸ This method not only achieves low-temperature preparation, but also forms a tight connection between nanocrystalline rutile TiO₂ and perovskite layers with a large interface area, which helps to more effectively extract photo-generated electrons, reduce electron–hole pair recombination, and improve charge collection efficiency. TiCl₄ is hydrolyzed to produce TiO₂, and the hydrolysis reaction itself does not need too high temperatures to occur.

One year later, Kymakis's team prepared amorphous TiO₂ by dissolving titanium isopropoxide in isopropanol, adding hydrochloric acid as a stabilizer, and then spin coating on a glass/ITO substrate, followed by annealing at 150 °C for 45 minutes.²³⁹ This TiO₂ layer exhibits a highly uniform surface, without cracks, and has very low roughness. In the same year, Kim et al. prepared amorphous TiO_x layers using plasma-enhanced atomic layer deposition (PEALD) technology, and the entire deposition process was carried out at 80 °C.²⁴⁰ This ETL has been used in flexible PSCs and has achieved great stability.

In 2016, Chen et al. prepared flower-like TiO₂ (Fig. 11c) using the CBD method at 80 °C, which is also an anatase TiO₂ nanorod material.²⁴¹ A year later, Wang's team prepared TiO₂ thin films at 80 °C and used titanium oxide bis(2,4-pentanedione acid) (TOPD) as a binder for TiO₂ NPs to reduce morphological defects in the TiO₂:TOPD film.²⁴² After that, Lan's team controlled the growth of TiO₂ NRs using oleic acid (OA) ligand assistance, and then replaced the oleic acid ligand with BF₄⁻ through ligand exchange treatment to form BF₄⁻-capped TiO₂ NRs.²⁴³

In 2019, You et al. redispersed TiO₂ nanosol by using different solvents (H₂O, ethanol, DMSO, and N,N-dimethylformamide (DMF)).²⁴⁴ TiO₂ nanosol was formed by hydrolysis of TTIP at 80 °C and the condensation reaction of HNO₃ at 60 °C. DMF is the optimal solvent for forming a uniform TiO₂ ETL due to its low surface tension and high zeta potential. This solar cell achieved a PCE of 18.2%. The flexible solar cell also demonstrated good mechanical bending stability, maintaining approximately 95% of the initial PCE after 1000 repeated bending cycles.

In 2020, Jo's team adopted a simple oxygen plasma treatment method to improve the low-temperature treatment of mesoporous TiO₂ layers.²⁴⁵ Oxygen plasma treatment effectively removes organic additives from the TiO₂ layer and reduces oxygen vacancies. Oxygen plasma treatment improves the wettability of the TiO₂ layer, promotes the penetration of perovskite precursor solution in the TiO₂ layer, and thus enhances the performance of PSCs.

Sanehira et al. synthesized a Nb-doped TiO₂ ETL using a one-step low-temperature steam annealing (SA) method.²⁴⁶ Place the substrate coated with precursor solution (TiCl₄) in a polytetrafluoroethylene (PTFE) bottle, add 5 ml of water, and then heat at 125 °C for 4 hours. This low-temperature process is beneficial for precise control of the conduction band (CB) level of the ETL.

Ren et al. investigated the effect of UV ozone treatment on the efficiency and stability of the low-temperature TiO₂ ETL in planar perovskite solar cells.²⁴⁷ The low-temperature TiO₂ interface treated with UV ozone improves the work function and built-in potential of TiO₂, and can suppress the photocatalytic activity of TiO₂, reducing the decomposition of perovskite films.

Homola's team used a cold (70 °C) plasma treatment method to prepare TiO₂ for the first time.²⁴⁸ The polysiloxane adhesive was removed rapidly (1 min) by low-temperature plasma treatment and converted into amorphous silicon dioxide. The photoelectric conversion efficiency of perovskite solar cells using the plasma-treated porous TiO₂/SiO₂ photoelectric anode is about 12%.

In 2021, Pan et al. used low temperature and a feasible glancing angle deposition (GLAD) method to directly grow TiO₂ nanopillar arrays (TiO₂ NaPAs) on conductive substrates, as shown in Fig. 11d.²⁴⁹ This material can be used in flexible substrates too.

In 2022, Hong et al. introduced acetylacetone (Acac) into the solution and prepared Acac-TiO₂ NPs by the sol–gel method at low-temperature.²⁵⁰ In the same year, Liu and co-workers prepared a TiO₂ ETL with s low temperature hydrothermal method and then used acetic acid (AA) and oleic acid (OA) to modify the ETL surface.²⁵¹ The former surface shows better morphological characteristics, and finally achieved 20.15% PCE in the device modified by using AA.

Fig. 12 shows the difference between some low-temperature fabrication methods and traditional methods. Generally, from the progress summarized above, in order to achieve the low-temperature preparation of the TiO₂ ETL, many materials or means need to be involved, which greatly increases the research and development cost and process difficulty. In addition, the properties of TiO₂, such as electron mobility, are not as good as those of zinc oxide and tin oxide, which limits the development of this ETL. It is worth noting that the toughness of TiO₂ itself is poor, and its application in flexible wearable devices will be greatly limited.

Fig. 12 The difference between some low-temperature fabrication methods and traditional methods.

4.2 Others

In 2013, Carnie et al. prepared Al₂O₃ thin films at low temperature.²⁵² An Al₂O₃-perovskite layer was co-deposited by spin coating. This method combines two manufacturing steps into one and eliminates the high-temperature sintering step. After co-deposition, low-temperature heating treatment is performed at a temperature not exceeding 110 °C. This step helps with the crystallization of perovskite materials and the formation of electron transport layers. Furthermore, ALD was applied by Zhang et al. for preparing the Al₂O₃ underlayer.²⁵³ They used trimethyl aluminum (TMA) as the aluminum precursor and H₂O as the counter reactant for preparation on a substrate at 200 °C. The insulating property of Al₂O₃ can eliminate the chemical capacitance caused by the high-density sub-band gap state, resulting in higher V_oc, and the device with an aluminum support is much more stable,²⁵⁴ so Al₂O₃ is usually used as the scaffold layer,^255,256 encapsulant,^257,258 or buffer layer²⁵⁹ for modifying the ETLs or PSCs.

Nb₂O₅ has excellent electro-optical properties: high electron mobility helps to improve the transmission efficiency of electrons and reduce electron recombination; the edge position of the conduction band is well matched with the energy level structure of the perovskite material, which is conducive to the effective extraction and transmission of electrons, while hindering the reverse transmission of holes, and helping to improve the open circuit voltage (V_oc) and fill factor (FF); it has less light absorption in the whole visible and near-infrared spectral range, which means that more light energy can be absorbed by the perovskite absorption layer, thus increasing the photogenerated current.^260–262 Ye et al. studied the preparation of zinc-doped niobium oxide (Nb₂O₅) by the low-temperature solution combustion method.²⁶³ Nb₂O₅ thin films were prepared by the low-temperature (200 °C) solution combustion method. This method does not need high-temperature sintering. Nb₂O₅ and Zn-doped Nb₂O₅ based devices show better stability than traditional mesoporous TiO₂-based devices sintered at high temperature, and can maintain 80% of their initial PCE in air for up to 20 days. Huang and co-workers prepared Nb₂O₅ thin films by RF sputtering at room temperature.¹⁹⁷ The Nb₂O₅ ETL, perovskite layer, and ZnO ETL form a good step energy level.

SnO_x, as another tin oxide material, can also be prepared at low temperatures. Shang et al. proposed a mild dehydration reaction method for the synthesis of amorphous SnO_x at a low temperature below 100 °C.²⁶⁴ The synthesis process includes three steps: first, SnCl₄·5H₂O reacts with ammonia to form Sn(OH)₄, then the obtained white precipitate is dissolved in methylamine (MA) solution to form Sn(CH₃NH₂)_x(OH)₄, and finally, the amorphous SnO_x film is formed by spin coating on an ITO substrate and annealing at 80 °C. The perovskite solar cells using amorphous SnO_x as the ETL showed a maximum photoelectric conversion efficiency (PCE) of 20.4%, which was the highest efficiency achieved by the amorphous metal oxide ETL produced at less than 100 °C at that time. Wang's team prepared the SnO_x:Cl ETL by chemical deposition at low temperature (about 100 °C).²⁶⁵ They emphasized the collaborative optimization of KX (potassium halide) and SnO_x:Cl, and potassium ions not only replace A-site cations, but also occupy a large number of gap sites at grain boundaries, effectively inhibiting iodide migration and organic volatilization, which is a new way to achieve low temperature and high performance PSCs.

As an alternative to ZnO/SnO₂, In₂O₃ has high transparency, high conductivity, and low photocatalytic activity.²⁶⁶ High temperature annealing is required during the formation of traditional In₂O₃ thin films, which may damage or degrade the underlying perovskite materials. Yang et al. prepared In₂O₃ and Sn:In₂O₃ nanoparticles by a low-temperature synthesis method, which can form an effective top ETL without any post-treatment.²⁶⁷ This double-layer ETL improves the interface contact with the perovskite layer, reduces the interface defects and recombination centers, and thus reduces the non-radiation recombination loss.

CeO₂ is a new ETL material prepared without annealing treatment, which has excellent thermal and chemical stability, a wide band gap, a high dielectric constant, and excellent ionic conductivity. CeO₂ nanocrystals are prepared by the solvothermal method, which can be re-dispersed in cyclohexane to form a stable, uniform and transparent solution, so the ETL can be prepared by the spin coating method.²⁶⁸

Additionally, Guo et al. used SnO₂-modified mesoporous ZrO₂ as the electron transport layer, which was also prepared by a low-temperature solution method.²⁶⁹

5. Conclusions and prospects

5.1 Conclusions

Flexible perovskite solar cells are one of the most popular parts of the photovoltaic industry, and its ETM plays an important role in the good performance of the device. As mentioned above, the applications of SnO₂ and ZnO in flexible PSCs have great potential. In this review, the basic properties of the two materials are first summarized, including their structure characteristics, energy band, electron mobility, and optical properties. Next, we introduce many different ways for fabricating ETL films, such as ALD, sol–gel, CBD, magnetron sputtering deposition, electrochemical deposition, hydrothermal deposition, and so on. We provide brief introductions of the working principles in each method, and various productions achieved by different teams are also sorted. We find that the materials for preparation are varied, and many researchers have prepared ETLs at a very low temperature (even room temperature), which is very suitable for flexible PSCs. Then, we provide a detailed summary of the application of these ETLs, including doping and interface engineering. We mainly classify and summarize from the perspective of materials, and introduce the research results of previous researchers. At last, we also introduce other low-temperature processed MO materials like TiO₂, Al₂O₃, Nb₂O₅, CeO₂, In₂O₃, etc. They also have many prominent advantages.

5.2 Prospectives

Although the technology of low-temperature processed metal oxide ETLs has shown rapid development, there are still many problems to be solved.

5.2.1 Electron dynamics engineering needs further research. Theory is the basis of practice. In-depth study of the role of oxygen vacancies, interface defects, energy band structure, electron–hole operation mechanism, and other important concepts needs to be further conducted, in order to provide a clearer and useful theoretical basis for solar cells.

5.2.2 Structures and innovation and optimization. The limiting efficiency of single-layer perovskite solar cells is about 33%,²⁷² and the tandem structures can break this limit. Perovskite solar cells are especially suitable for tandem structures because of their excellent performance with an adjustable band gap. However, tandem structures greatly increase the difficulty of device preparation, and the device thickness will also affect its use in flexible batteries. In order to be used in flexible batteries, it is necessary to fully consider the thickness, stiffness, and other physical parameters of the film, so further research is very much needed. The bilayer ETL device achieved good properties in the previous work. In this direction, different ETM combinations can be tried. But we should consider the energy level adaptation, chemical stability, electron extraction efficiency, transparency, and other properties of the materials.

5.2.3 Materials upgrade. Material properties are fundamental to PSCs. Initially, efforts can focus on developing new ETMs, such as those based on SnO₂ and ZnO, through artificial synthesis to create self-assembled ETMs. Of course, alternatives with superior performance and low cost are also sought. Regarding doping engineering, we believe the interaction between the functional atoms or groups of the dopant and the ETL material plays a key role. Improving research on these material interactions (such as the role of hydrogen bonding and the influence of oxygen vacancies) will positively impact the selection and preparation of dopants.

5.2.4 Long term stability needs to be improved. In addition to efficiency improvement, device stability is very important for the development of solar cells. In addition to preventing water and oxygen from damaging the device, the impact of ion diffusion on stability cannot be underestimated, and it may even affect the efficiency of the device. At present, the efficiency of most PSCs has reached about 20%, and the stability of performance has become a very popular breakthrough direction. It can be tried by means of material optimization, buffer layer addition, device packaging, and so on. In addition, for flexible perovskite devices, attention should be paid to the rigidity and brittleness of perovskite,²⁷³ and the active layer is crucial to the performance of the device. After bending, the active layer will inevitably be damaged. In order to improve the stability of related devices, self-healing materials can be used to blend the active layer. Additionally, array integration methods can also be used to prepare this device.

5.2.5 Environmental protection. This is the eternal theme of the new era. Many perovskite materials used contain toxic lead, which has adverse effects on the environment and human health. Now some researchers have focused on this part;²⁷⁴ they used lead-free materials for fabrication and finally achieved good performances. From our point of view, it is also a feasible method to modify the perovskite layer and replace lead with green elements by physical and chemical means. This also places demands on the fabrication process. When preparing the MO ETL, existing methods generate some waste or harmful substances. A completely pollution-free manufacturing process is obviously a fantasy, but it is possible to reduce the emission of pollutants through material upgrading or develop recycling processes to reuse materials.

The development of flexible perovskite solar cells is still in full swing. With the joint efforts of researchers, the performance of devices will certainly be improved a lot.

Author contributions

Jianghao Tian: conceptualization, methodology, formal analysis, investigation, writing – original draft. Kun Wang: methodology, investigation. Zhipeng Zhou: methodology, investigation. Lexiang Zhang: methodology, investigation. Pu Fan: conceptualization, methodology, writing – review & editing. Huajing Zheng: writing – review & editing. Ding Zheng: conceptualization, writing – review & editing. Junsheng Yu: resources, supervision, funding acquisition, writing – review & editing, project administration. All authors have read and agreed to the published version of the manuscript.

Conflicts of interest

The authors declare no conflict of interest.

Data availability

No primary research results, software or code have been included and no new data were generated or analysed as part of this review.

Acknowledgements

This work was financially supported by the National Natural Science Foundation of China (Grant No. 62275041) and the Project of Science and Technology of Sichuan Province (Grant No. 2023ZYZYTS0016 and 2023ZYD0162). The authors were supported by the Foundation of Yangtze Delta Region Institute (Huzhou) University of Electronic Science and Technology of China (Grant No. U03220143 and U04220079); the Major Technology Project of Sichuan Province (Grant No. 2024ZDZX0030). This work was also sponsored by the Sichuan Province Key Laboratory of Display Science and Technology.

References

1. A. Kojima, K. Teshima, Y. Shirai and T. Miyasaka, J. Am. Chem. Soc., 2009, 131, 6050–6051.
2. D. I. Kutsarov, E. Rezaee, J. Lambert, W. T. Stroud, A. Panagiotopoulos and S. R. P. Silva, Adv. Mater. Technol., 2025, 2401834.
3. E. K. Solak and E. Irmak, RSC Adv., 2023, 13, 12244–12269.
4. Y. Jiang, S. Sun, R. Xu, F. Liu, X. Miao, G. Ran, K. Liu, Y. Yi, W. Zhang and X. Zhu, Nat. Energy, 2024, 9, 8.
5. B. O'Regan and M. Grätzel, Nature, 1991, 353, 737–740.
6. K. Kakiage, Y. Aoyama, T. Yano, K. Oya, J. Fujisawa and M. Hanaya, Chem. Commun., 2015, 51, 15894–15897.
7. K. Mahmood, S. Sarwar and M. Mehran, RSC Adv., 2017, 7, 17044–17062.
8. J. L. Liu, S. Li, A. Y. Mei and H. W. Han, Small Sci., 2023, 3, 2300020.
9. J. Xu, X. Shi, J. Chen, J. Luan and J. Yao, J. Solid State Chem., 2019, 276, 302–308.
10. Z. Xing, S. Li, P. Xu, H. Tian, L. Deng, Y. Yao, B. Chen, F. Xie, M. An, D. Yun, S. Xie and L. Zheng, Chem. Res. Chin. Univ., 2022, 38(1), 75–81.
11. F. Liu, B. Chen, A. Fan, Z. Chen, Y. Pan, Z. Tang, L. Deng, Z. Xing and S. Li, Adv. Funct. Mater., 2025, 35, 2419582.
12. W. Huang, P. Huang and S. Yang, Chem. Commun., 2016, 52, 13572–13575.
13. F. Xia, Q. Wu, P. Zhou, Y. Li, X. Chen, Q. Liu, J. Zhu, S. Dai, Y. Lu and S. Yang, ACS Appl. Mater. Interfaces, 2015, 7, 13659–13665.
14. P. Patil, D. Manna, U. Nakate, Y. Hahn, S. Kwon and S. Na, Chem. Eng. J., 2020, 397, 125504.
15. Y. Liu, B. Long, R. Chen, S. Huang, W. Ou-Yang and X. Chen, ACS Appl. Energy Mater., 2021, 4, 3812–3821.
16. S. Chen, S. Yang, H. Sun, L. Zhang, J. Peng, Z. Liang, Z. Wang and Z. Wang, J. Power Sources, 2017, 353, 123–130.
17. X. Jiang, F. Wang, Q. Wei, H. Li, Y. Shang, W. Zhou, C. Wang, P. Cheng, Q. Chen, L. Chen and Z. Ning, Nat. Commun., 2020, 11, 1245.
18. M. Lee, D. Kim, Y. Lee, H. Koo, K. Lee and I. Chung, ACS Appl. Energy Mater., 2020, 3, 5581–5588.
19. L. Chen, Z. Tseng and C. Wang, Mater. Sci. Semicond. Process., 2016, 56, 179–182.
20. K. Yokoyama, H. Lin, Q. Shui, X. Wang, N. Saito and Y. Matsuo, Appl. Phys. Express, 2023, 16, 081001.
21. W. Lee, C. Chen, C. Li, S. Hsiao, W. Tsai, M. Huang, C. Cheng, C. Wu and H. Lin, Nano Energy, 2017, 38, 66–71.
22. R. Ihly, A. Dowgiallo, M. Yang, P. Schulz, N. Stanton, O. Reid, A. Ferguson, K. Zhu, J. Berry and J. Blackburn, Energy Environ. Sci., 2016, 9, 1439–1449.
23. F. Ye, S. Zhang, J. Warby, J. Wu, E. Gutierrez-Partida, F. Lang, S. Shah, E. Saglamkaya, B. Sun, F. Zu, S. Shoaee, H. Wang, B. Stiller, D. Neher, W. Zhu, M. Stolterfoht and Y. Wu, Nat. Commun., 2022, 13, 7454.
24. K. Jiang, F. Wu, H. Yu, Y. Yao, G. Zhang, L. Zhu and H. Yan, J. Mater. Chem. A, 2018, 6, 16868–16873.
25. D. Yang, R. Yang, S. Priya and S. Liu, Angew. Chem., Int. Ed., 2019, 58, 4466–4483.
26. J. Liu, Z. Zhao, J. Qian, Z. Liang, C. Wu, K. Wang, S. Liu and D. Yang, Adv. Mater., 2024, 36, 2401236.
27. X. Li, H. Yu, Z. Liu, J. Huang, X. Ma, Y. Liu, Q. Sun, L. Dai, S. Ahmad, Y. Shen and M. Wang, Nano-Micro Lett., 2023, 15, 206.
28. D. Li, L. Hu, Y. Xie, G. Niu, T. Liu, Y. Zhou, L. Gao, B. Yang and J. Tang, ACS Photonics, 2016, 3, 2122–2128.
29. J. Jia, J. Wu, J. Dong, L. Fan, M. Huang, J. Lin and Z. Lan, Chem. Commun., 2018, 54, 3170–3173.
30. M. K. S. Bin Rafiq, N. Amin, H. F. Alharbi, M. Luqman, A. Ayob, Y. S. Alharthi, N. H. Alharthi, B. Bais and M. Akhtaruzzaman, Sci. Rep., 2020, 10, 1–11.
31. L. Rani Karna, R. Upadhyay and A. Ghosh, Sci. Rep., 2023, 13, 15212.
32. H. Yoon, S. M. Kang, J.-K. Lee and M. Choi, Energy Environ. Sci., 2016, 9, 2262–2266.
33. G. Yin, H. Zhao, J. Feng, J. Sun, J. Yan, Z. Liu, S. Lin and S. Liu, J. Mater. Chem. A, 2018, 6, 9132–9138.
34. B. K. Mohammed, M. K. A. Mohammed and D. S. Ahmed, Solid-State Electron., 2023, 204, 108640.
35. Y. W. Noh, I. S. Jin, S. H. Park and J. W. Jung, J. Mater. Sci. Technol., 2020, 42, 38–45.
36. S. W. Tong, J. Balapanuru, D. Fu and K. P. Loh, ACS Appl. Mater. Interfaces, 2016, 8, 29496–29503.
37. L. Lin, T. Jones, T. Yang, N. W. Duffy, J. H. Li, L. Zhao, B. Chi, X. B. Wang and G. J. Wilson, Adv. Funct. Mater., 2021, 31, 2008300.
38. M. M. Lee, J. Teuscher, T. Miyasaka, T. N. Murakami and H. J. Snaith, Science, 2012, 338, 643–647.
39. W. Ke, G. Fang, J. Wang, P. Qin, H. Tao, H. Lei, Q. Liu, X. Dai and X. Zhao, ACS Appl. Mater. Interfaces, 2014, 6, 15959–15965.
40. J. Choi, S. Song, M. T. Horantner, H. J. Snaith and T. Park, ACS Nano, 2016, 10, 6029–6036.
41. G. S. Han, Y. H. Song, Y. U. Jin, J. W. Lee, N. G. Park, B. K. Kang, J. K. Lee, I. S. Cho, D. H. Yoon and H. S. Jung, ACS Appl. Mater. Interfaces, 2015, 7, 23521–23526.
42. D. Yang, X. Zhou, R. Yang, Z. Yang, W. Yu, X. Wang, C. Li, S. Liu and R. P. H. Chang, Energy Environ. Sci., 2016, 9, 3071–3078.
43. S. F. Shaikh, H.-C. Kwon, W. Yang, H. Hwang, H. Lee, E. Lee, S. Ma and J. Moon, J. Mater. Chem. A, 2016, 4, 15478–15485.
44. H. Lu, W. Tian, B. Gu, Y. Zhu and L. Li, Small, 2017, 13, 1701535.
45. G. Xiao, C. Shi, K. Lv, C. Ying and Y. Wang, ACS Appl. Energy Mater., 2018, 1, 2576–2581.
46. D. Shen, W. Zhang, F. Xie, Y. Li, A. Abate and M. Wei, J. Power Sources, 2018, 402, 320–326.
47. D. G. Lee, M. Kim, B. J. Kim, D. H. Kim, S. M. Lee, M. Choi, S. Lee and H. S. Jung, Appl. Surf. Sci., 2019, 477, 131–136.
48. B. Wang, M. Zhang, X. Cui, Z. Wang, M. Rager, Y. Yang, Z. Zou, Z. L. Wang and Z. Lin, Angew. Chem., Int. Ed., 2020, 59, 1611–1618.
49. R. Teimouri, Z. Heydari, M. P. Ghaziani, M. Madani, H. Abdy, M. Kolahdouz and E. Asl-Soleimani, Superlattices Microstruct., 2020, 145, 106627.
50. H. Latif, Z. Azher, S. A. Shabbir, S. Rasheed, E. Pervaiz, A. Sattar and A. Imtiaz, Opt. Mater., 2020, 109, 110281.
51. M. Kim, I. Choi, S. J. Choi, J. W. Song, S. Mo, J. An, Y. Jo, S. Ahn, S. K. Ahn, G. Kim and D. S. Kim, Joule, 2021, 5, 659–672.
52. N. Alias, A. A. Umar, N. A. A. Malek, K. Liu, X. Li, N. A. Abdullah, M. M. Rosli, M. Y. A. Rahman, Z. Shi, X. Zhang, H. Zhang, F. Liu, J. Wang and Y. Zhan, ACS Appl. Mater. Interfaces, 2021, 13, 3051–3061.
53. W. Zhao, P. Guo, J. Wu, D. Lin, N. Jia, Z. Fang, C. Liu, Q. Ye, J. Zou, Y. Zhou, H. Wang and H. Wang, Nano-Micro Lett., 2024, 16, 191.
54. W. Ke, G. Fang, J. Wang, P. Qin, H. Tao, H. Lei, Q. Liu, X. Dai and X. Zhao, ACS Appl. Mater. Interfaces, 2014, 6, 15959–15965.
55. P. Wu, S. Wang, X. Li and F. Zhang, J. Mater. Chem. A, 2021, 9, 19554–19588.
56. Q. Jiang, X. Zhang and J. You, Small, 2018, 14, 40.
57. H. Wang, M. A. Sayeed and T. Wang, Aust. J. Chem., 2015, 68, 1783–1788.
58. H. B. Lee, N. Kumar, M. M. Ovhal, Y. J. Kim, Y. M. Song and J.-W. Kang, Adv. Funct. Mater., 2020, 30, 1559.
59. Q. Dong, Y. Shi, C. Zhang, Y. Wu and L. Wang, Nano Energy, 2017, 40, 336–344.
60. W. Ke, G. Fang, Q. Liu, L. Xiong, P. Qin, H. Tao, J. Wang, H. Lei, B. Li, J. Wan, G. Yang and Y. Yan, J. Am. Chem. Soc., 2015, 137, 6730–6733.
61. L. Zuo, H. Guo, D. W. deQuilettes, S. Jariwala, N. De Marco, S. Q. Dong, R. DeBlock, D. S. Ginger, B. Dunn, M. K. Wang and Y. Yang, Sci. Adv., 2017, 3, e1700106.
62. Z. Zhu, Y. Bai, X. Liu, C.-C. Chueh, S. Yang and A. K.-Y. Jen, Adv. Mater., 2016, 28, 6478–6484.
63. P. Lee, T. Wu, K. Tian, C. Li, C. Hou, J. Shyue, C. Lu, Y. Huang and W. Su, ACS Appl. Mater. Interfaces, 2020, 12, 45936–45949.
64. C. Xiao, C. Jiang, W. Ke, C. Wang, B. Gorman, Y. Yan and M. Al-Jassim, Insul. Eng. Electr. Electron, 2016, 1197–1200.
65. Y. Kuang, V. Zardetto, R. van Gils, S. Karwal, D. Koushik, M. A. Verheijen, L. E. Black, C. Weijtens, S. Veenstra, R. Andriessen, W. M. M. Kessels and M. Creatore, ACS Appl. Mater. Interfaces, 2018, 10, 30367–30378.
66. X. Zhang, Y. Zhou, M. Chen, D. Wang, L. Chao, Y. Lv, H. Zhang, Y. Xia, M. Li, Z. Hu and Y. Chen, Small, 2023, 19, 39.
67. Y. Lee, S. Lee, G. Seo, S. Paek, K. T. Cho, A. J. Huckaba, M. Calizzi, D. Choi, J. Park, D. Lee, H. J. Lee, A. M. Asiri and M. K. Nazeeruddin, Adv. Sci., 2018, 5, 1800130.
68. J. Bing, S. Huang and A. W. Ho-Baillie, Energy Technol., 2020, 8, 1901114.
69. Y. Gu, C. Ye, X. Yin, J. Han, H. Shen, J. Li, X. Hao and H. Lin, Chem. Eng. J., 2018, 351, 791–798.
70. Y. Zhou, X. Li and H. Lin, Small, 2020, 16, 1902579.
71. J. J. Yoo, G. Seo, M. R. Chua, T. G. Park, Y. Lu, F. Rotermund, Y.-K. Kim, C. S. Moon, N. J. Jeon, J.-P. Correa-Baena, V. Bulović, S. S. Shin, M. G. Bawendi and J. Seo, Nature, 2021, 590, 587–593.
72. M. J. Jeong, K. M. Yeom, S. J. Kim, E. H. Jung and J. H. Noh, Energy Environ. Sci., 2021, 14, 2419–2428.
73. Q. Zhao, D. Liu, Z. Li, B. Zhang, X. Sun, Z. Shao, C. Chen, X. Wang, L. Hao, X. Wang, C. Gao, Y. Li, P. Chen, X. Zhang, S. Pang, G. Cui and L. Wang, Chem. Eng. J., 2022, 443, 136308.
74. L. Qiu, Z. Liu, L. K. Ono, Y. Jiang, D.-Y. Son, Z. Hawash, S. He and Y. B. Qi, Adv. Funct. Mater., 2019, 29, 1806779.
75. G. Bai, Z. Wu, J. Li, T. Bu, W. Li, A. Li, F. Huang, Q. Zhang, Y.-B. Cheng and J. Zhong, Sol. Energy, 2019, 183, 306–314.
76. H. Tao, H. Wang, Y. Bai, H. Long, H. Zhao, Q. Fu and Z. Ma, J. Mater. Sci.: Mater. Electron., 2019, 30, 12036–12043.
77. H. Tao, Z. Ma, G. Yang, H. Wang, H. Long, H. Zhao, P. Qin and G. Fang, Appl. Surf. Sci., 2018, 434, 1336–1343.
78. J. Li, L. Li, W. Chen, Q. Yi and G. Zou, Nanotechnology, 2021, 32, 025606.
79. Y. Ko, Y. R. Kim, H. Jang, C. Lee, M. G. Kang and Y. Jun, Nanoscale Res. Lett., 2017, 12, 498.
80. D. Zheng, R. Qin, M. Wu and J. Yu, in Optoelectronic Materials and Devices for Sensing and Imaging, ed. Y. Jiang, X. Ma, X. Li, M. Pu, X. Feng and B. Kippelen, 2019, vol. 10843, pp. 1084315.
81. J. Jia, J. Dong, J. Wu, H. Wei and B. Cao, J. Alloys Compd., 2020, 844, 156032.
82. K. Mahmood, A. Khalid, F. Nawaz and M. T. Mehran, J. Colloid Interface Sci., 2018, 532, 387–394.
83. P. Zhou, J. Wu, Y. Tu, M. Zhen, J. Huo, Y. Wei and Z. Lan, Sol. Energy, 2016, 137, 579–584.
84. Z. Zhu, Y. Bai, X. Liu, C.-C. Chueh, S. Yang and A. K.-Y. Jen, Adv. Mater., 2016, 28, 6478–6484.
85. M. F. M. Noh, N. A. Arzaee, J. Safaei, N. A. Mohamed, H. P. Kim, A. R. Mohd Yusoff, J. Jang and M. A. Mat Teridi, J. Alloys Compd., 2019, 773, 997–1008.
86. M. Haghighi, N. Ghazyani, S. Mahmoodpour, R. Keshtmand, A. Ghaffari, H. M. Luo, R. Mohammadpour, N. Taghavinia and M. Abdi-Jalebi, Sol. RRL, 2023, 7, 2201080.
87. W. Bin, W. Huang, W. Lin and H. Lee, Crystals, 2021, 11, 1380.
88. Y. Guo, X. Yin, J. Liu, W. Chen, S. Wen, M. Que, H. Xie, Y. Yang, W. Que and B. Gao, Org. Electron., 2019, 65, 207–214.
89. M. Fieveza, P. J. S. Rana, T. M. Koh, M. Manceau, J. H. Lew, N. F. Jamaludin, B. Ghosh, A. Bruno, S. Cros, S. Berson, S. G. Mhaisalkar and W. L. Leong, Sol. Energy Mater. Sol. Cells, 2021, 230, 111189.
90. B. Roose, S. Pathak and U. Steiner, Chem. Soc. Rev., 2015, 44, 8326–8349.
91. M. Park, J. Kim, H. J. Son, C. Lee, S. S. Jang and M. J. Ko, Nano Energy, 2016, 26, 208–215.
92. Y. Huang, S. Li, C. Wu, S. Wang, C. Wang and R. Ma, Chem. Phys. Lett., 2020, 745, 137220.
93. Q. Zhuang, H. Wang, C. Zhang, C. Gong, H. Li, J. Chen and Z. Zang, Nano Res., 2022, 15, 5114–5122.
94. H. Chen, D. Liu, Y. Wang, C. Wang, T. Zhang, P. Zhang, H. Sarvari, Z. Chen and S. Li, Nanoscale Res. Lett., 2017, 12, 238.
95. H. Ye, Z. Liu, X. Liu, B. Sun, X. Tan, Y. Tu, T. Shi, Z. Tang and G. Liao, Appl. Surf. Sci., 2019, 478, 417–425.
96. S. Lan, W. T. Zheng, S. Yoon, H. U. Hwang, J. W. Kim, D. W. Kang, J. W. Lee and H. K. Kim, ACS Appl. Energy Mater., 2022, 5, 14901–14912.
97. H. V. Quy and C. W. Bark, ACS Omega, 2022, 7, 22256–22262.
98. Y. Chen, X. Zuo, Y. He, F. Qian, S. Zuo, Y. Zhang, L. Liang, Z. Chen, K. Zhao and Z. Liu, Adv. Sci., 2021, 8, 2001466.
99. A. Kumar, A. K. Al-Mousoi, M. J. Saadh, M. K. A. Mohammed, G. V. S. S. Sarma, N. Ahmad and R. Tiwari, Electron. Mater. Lett., 2023, 19, 471–482.
100. R. Xue, X. Zhou, S. Peng, P. Xu, S. Wang, C. Xu, W. Zeng, Y. Xiong and D. Liang, ACS Sustainable Chem. Eng., 2020, 8, 10714–10725.
101. J. Liang, Z. Chen, G. Yang, H. Wang, F. Ye, C. Tao and G. Fang, ACS Appl. Mater. Interfaces, 2019, 11, 23152–23159.
102. X. Ren, Y. Liu, D. G. Lee, W. B. Kim, G. S. Han, H. S. Jung and S. Liu, InfoMat, 2020, 2, 401–408.
103. W. Cai, T. Yang, C. Liu, Y. Wang, S. Wang, Y. Du, N. Wu, W. Huang, S. Wang, Z. Wang, X. Chen, J. Feng, G. Zhao, Z. Ding, X. Pan, P. Zou, J. Yao, S. Liu and K. Zhao, Angew. Chem., Int. Ed., 2023, 62, e202309398.
104. Q. Liu, F. Fei, Y. Xu, L. Gu, X. Ding, K. Wang, K. Du, S. Wang, X. Dong, L. Li, B. Li, N. Yuan and J. Ding, Org. Electron., 2023, 120, 1068.
105. Q. Chen, J. Wu, X. Liu, Y. Du, C. Deng, X. Chen, L. Sun, L. Tan, W. Sun and Z. Lan, ACS Appl. Mater. Interfaces, 2024, 16, 8949–8959.
106. P. Zhu, S. Gu, X. Luo, Y. Gao, S. Li, J. Zhu and H. Tan, Adv. Energy Mater., 2019, 1903083.
107. C. Lin, T. Murakami, M. Chikamatsu, T. Bessho, M. Furue and H. Segawa, ACS Omega, 2021, 6, 17880–17889.
108. Z. Liu, K. Deng, J. Hu and L. Li, Angew. Chem., Int. Ed., 2019, 58, 11497–11504.
109. Z. Lin, W. Zhang, Q. Cai, X. Xu, H. Dong, C. Mu and J. Zhang, Adv. Sci., 2021, 8, 2102845.
110. R. Tao, Y. Zhang, Z. Jin, Z. Sun and L. Xu, Electrochim. Acta, 2018, 264, 1017–1025.
111. J. Wei, F. Guo, X. Wang, K. Xu, M. Lei, Y. Liang, Y. Zhao and D. Xu, Adv. Mater., 2018, 30, 1805153.
112. L. Chen, X. Li, N. Zhang, L. Yu, Z. Liu, H. Liu and G. Song, Sol. Energy Mater. Sol. Cells, 2024, 272, 112907.
113. S. You, H. Zeng, Z. Ku, X. Wang, Z. Wang, Y. Rong, Y. Zhao, X. Zheng, L. Luo, L. Li, S. Zhang, M. Li, X. Gao and X. Li, Adv. Mater., 2020, 32, 2003990.
114. Z. Li, Y. Gao, Z. Zhang, Q. Xiong, L. Deng, X. Li, Q. Zhou, Y. Fang, P. Gao and G. Peng, Nano-Micro Lett., 2021, 13, 101.
115. N. Kwon and J. Seo, Korean J. Chem. Eng., 2024, 41, 2405–2414.
116. Y. Yang, Y. Liu, M. Grätzel and A. Hagfeldt, Energy Environ. Sci., 2021, 14, 3447–3454.
117. Z. Lv, L. He, H. Jiang, X. Ma, F. Wang, L. Fan, M. Wei, J. Yang, L. Yang and N. Yang, ACS Appl. Mater. Interfaces, 2021, 13, 16326–16335.
118. B. Gu, Y. Du, B. Chen, R. Zhao, H. Lu, Q. Xu and C. Guo, ACS Appl. Mater. Interfaces, 2022, 14, 11264–11272.
119. M. Zhong, Y. Liang, J. Zhang, Z. Wei, Q. Li and D. Xu, J. Mater. Chem. A, 2019, 7, 6659–6664.
120. S. Zhang, H. Si, W. Fan, M. Shi, M. Li, C. Xu, Z. Zhang, Q. Liao, A. Sattar, Z. Kang and Y. Zhang, Angew. Chem., Int. Ed., 2020, 59, 11573–11582.
121. H. Xu, Y. Miao, N. Wei, H. Chen, Z. Qin, X. Liu, X. Wang, Y. Qi, T. Zhang and Y. Zhao, Adv. Energy Mater., 2021, 12, 2103151.
122. J. Xi, J. Yuan, J. Du, X. Yan and J. Tian, Small, 2022, 18, 2203519.
123. Z. Zheng, F. Li, J. Gong, Y. Ma, J. Gu, X. Liu, S. Chen and M. Liu, Adv. Mater., 2022, 34, 2109879.
124. X. Pang, J. Huang, C. Lin, Y. Zhang, N. Cheng, W. Zi, Z.-Z. Sun, Z. Yu and Z. Zhao, Chem.–Eur. J., 2023, 29(14), e202202744.
125. J. Xia, M. Sohail and M. K. Nazeeruddin, Adv. Mater., 2023, 35(31), 2211324.
126. T. Xue, T. Li, D. Chen, X. Wang, K. Guo, Q. Wang and F. Zhang, Micromachines, 2023, 14, 1549.
127. F. Xie, D. Mei, L. Qiu, Z. Su, L. Chen, Y. Liu and P. Du, J. Electron. Mater., 2023, 52, 4626–4633.
128. Q. Dong, Y. Xue, S. Wang, L. Wang, F. Chen, S. Zhang, R. Chi, L. Zhao and Y. Shi, Sci. China Mater., 2017, 60, 963–976.
129. Y. Hou, X. Chen, S. Yang, C. Li, H. Zhao and H. G. Yang, Adv. Funct. Mater., 2017, 27, 1700878.
130. X. Zhang, Y. Zhou, M. Chen, D. Wang, L. Chao, Y. Lv, H. Zhang, Y. Xia, M. Li, Z. Hu and Y. Chen, Small, 2023, 19, 39.
131. J. Dagar, S. Castro-Hermosa, G. Lucarelli, A. Zampetti, F. Cacialli and T. M. Brown, IEEE J. Photovoltaics, 2019, 9, 1309–1315.
132. M. Hashemib, M. Minbashi, S. M. B. Ghorashib, A. Ghobadic, M. H. Ehsanid, M. Heidariramsheh and A. Hajjiah, Sol. Energy, 2021, 215, 356–366.
133. F. Mohamadkhania, S. Javadpoura and N. Taghaviniab, Sol. Energy, 2019, 191, 647–653.
134. H. M. Pham, S. D. H. Naqvi, H. Tran, H. V. Tran, J. Delda, S. Hong, I. Jeong, J. Gwak and S. Ahn, Nanomaterials, 2023, 13, 3091.
135. Y. Lu, J. Feng, Z. Liu, Y. Duan, S. Zhan, S. Yang, K. He, Y. Li, Y. Zhou, N. Yuan, J. Ding and S. Liu, Adv. Mater., 2022, 34, 2201681.
136. X. Guo, J. Du, Z. Lin, J. Su, L. Feng, J. Zhang, Y. Hao and J. Chang, Chem. Eng. J., 2021, 407, 127997.
137. Y. Wang, P. Xiang, A. Ren, H. Lai, Z. Zhang, Z. Xuan, Z. Wan, J. Zhang, X. Hao, L. Wu, M. Sugiyama, U. Schwingenschlögl, C. Liu, Z. Tang, J. Wu, Z. Wang and D. Zhao, ACS Appl. Mater. Interfaces, 2020, 12, 53973–53983.
138. F. Wang, Y. Zhang, M. Yang, L. Xue, L. Yang, L. Fan, Y. Su, J. Yang and X. Zhang, Nano Energy, 2019, 63, 103825.
139. X. Zuo, B. Kim, B. B. Liu, D. M. He, L. Bai, W. Q. Wang, C. Y. Xu, Q. L. Song, C. Y. Jia, Z. G. Zang, D. H. Lee, X. Li and J. Z. Chen, Chem. Eng. J., 2022, 431, 133209.
140. R. Yuan, B. Cai, Y. Lv, X. Gao, J. Gu, Z. Fan, X. Liu, C. Yang, M. Liu and W. Zhang, Energy Environ. Sci., 2021, 14, 5074–5083.
141. E. Hyuk Jung, B. Chen, K. Bertens, M. Vafaie, S. Teale, A. Proppe, H. Yi, Z. Tong, C. Zheng and E. H. Sargent, ACS Energy Lett., 2020, 5, 2796–2801.
142. E. Jiang, Y. Ai, J. Yan, N. Li, L. Lin, Z. Wang, C. Shou, B. Yan, Y. Zeng, J. Sheng and J. Ye, ACS Appl. Mater. Interfaces, 2019, 11, 36727–36734.
143. J. Zhuang, P. Mao, Y. Luan, N. Chen, X. Cao, G. Niu, F. Jia, F. Wang, S. Cao and J. Wang, Adv. Funct. Mater., 2021, 31, 2010385.
144. H. Min, D. Lee, J. Kim, G. Kim, K. S. Lee, J. Kim, M. J. Paik, Y. K. Kim, K. S. Kim and M. G. Kim, Nature, 2021, 598, 444.
145. Q. Zhou, J. Duan, Y. Wang, X. Yang and Q. Tang, J. Energy Chem., 2020, 50, 1–8.
146. Z. Zhu, Y. Bai, X. Liu, C.-C. Chueh, S. Yang and A. K.-Y. Jen, Adv. Mater., 2016, 28, 6478–6484.
147. J. Du, L. Feng, X. Guo, X. Huang, Z. Lin, J. Su, Z. Hu, J. Zhang, J. Chang and Y. Hao, J. Power Sources, 2020, 455, 227974.
148. M. Thambidurai, F. Shini, P. C. Harikesh, N. Mathews and C. Dang, J. Power Sources, 2020, 448, 227362.
149. Q. Zhang, C. S. Dandeneau, X. Zhou and G. Cao, Adv. Mater., 2009, 21, 4087.
150. H. Liu, Z. Huang, S. Wei, L. Zheng, L. Xiao and Q. Gong, Nanoscale, 2016, 8, 6209.
151. P. Zhang, J. Wu, T. Zhang, Y. Wang, D. Liu, H. Chen, L. Ji, C. Liu, W. Ahmad, Z. D. Chen and S. Li, Adv. Mater., 2017, 30, 1703737.
152. J. Kim, G. Kim, T. Kim, S. Kwon, H. Back, J. Lee, S. H. Lee, H. Kang and K. Lee, J. Mater. Chem. A, 2014, 2, 17291–17296.
153. X. Xu, H. Zhang, J. Shi, J. Dong, Y. Luo, D. Li and Q. Meng, J. Mater. Chem. A, 2015, 3, 19288–19293.
154. Y. Ko, Y. Kim, S. Y. Kong, S. C. Kunnana and Y. Jun, Sol. Energy Mater. Sol. Cells, 2018, 183, 157–163.
155. H. Liang, Y.-C. Hu, Y. Tao, B. Wu, Y. Wu and J. Cao, ACS Appl. Mater. Interfaces, 2019, 11, 43116–43121.
156. D. Lee, M. Wolska-Pietkiewicz, S. Badoni, A. Grala, J. Lewiński and G. De Paëpe, Angew. Chem., Int. Ed., 2019, 58, 17163.
157. Z. Liang, Q. Zhang, L. Jiang and G. Cao, Energy Environ. Sci., 2015, 8, 3442.
158. Y. Guo, X. Li, L. L. Kang, X. He, Z. Q. Ren, J. D. Wu and J. Y. Qi, RSC Adv., 2016, 6, 62522.
159. D. Segets, R. Marczak, S. Schäfer, C. Paula, J.-F. Gnichwitz, A. Hirsch and W. Peukert, ACS Nano, 2011, 5, 4658.
160. R. D. Chavan, M. Wolska-Pietkiewicz, D. Prochowicz, M. Jędrzejewska, M. M. Tavakoli, P. Yadav, C. K. Hong and J. Lewiński, Adv. Funct. Mater., 2022, 32, 2205909.
161. R. Runjhun, E. A. Alharbi, Z. Drużyński, A. Krishna, M. Wolska-Pietkiewicz, V. Skorajanc, T. P. Baumeler, G. Kakavelakis, F. Eickemeyer, M. Mensi, S. M. Zakeeruddin, M. Grätzel and J. Leśniowski, Energy Environ. Mater., 2024, 7, e12720.
162. D. Zheng, G. Wang, W. Huang, B. Wang, W. Ke, J. L. Logsdon, Z. Wang, W. Zhu, M. R. Wasielewski, M. G. Kanatzidis, T. J. Marks and A. Facchetti, Adv. Funct. Mater., 2019, 29, 1900265.
163. J. Zhou, X. Meng, X. Zhang, X. Tao, Z. Zhang, J. Hu, C. Wang, Y. Li and S. Yang, Mater. Chem. Front., 2017, 1, 802–806.
164. J. L. Yang, B. D. Siempelkamp, E. Mosconi, F. De Angelis and T. L. Kelly, Chem. Mater., 2015, 27, 4229–4236.
165. X. Dong, H. Hu, B. Lin, J. Ding and N. Yuan, Chem. Commun., 2014, 50, 14405–14408.
166. M. Najafi, V. Zardetto, D. Zhang, D. Koushik, M. S. Dörenkämper, M. Creatore, R. Andriessen, P. Poodt and S. Veenstra, Sol. RRL, 2018, 2, 1800147.
167. X. Xu, H. Wang, J. Wang, M. Muhammad, Z. Wang, P. Chen, W. Zhao, B. Kang, J. Zhang, C. Li and Y. Duan, ACS Appl. Energy Mater., 2020, 3, 4208–4216.
168. R. Pietruszka, B. S. Witkowski, S. Zimowski, T. Stapinski and M. Godlewski, Surf. Coat. Technol., 2017, 319, 164–169.
169. K. Sharma, D. Routkevitch, N. Varaksa and S. M. George, J. Vac. Sci. Technol., A, 2016, 34, 01A146.
170. P. Rong, S. Ren and Q. Yu, Crit. Rev. Anal. Chem., 2019, 49, 336–349.
171. A. Thote, I. Jeon, H.-S. Lin, S. Manzhos, T. Nakagawa, D. Suh, J. Hwang, M. Kashiwagi, J. Shiomi, S. Maruyama, H. Daiguji and Y. Matsuo, ACS Appl. Electron. Mater., 2019, 1, 389–396.
172. L. S. Liang, Z. F. Huang, L. H. Cai, W. Z. Chen, B. Z. Wang, K. W. Chen, H. Bai, Q. Y. Tian and B. Fan, ACS Appl. Mater. Interfaces, 2014, 6, 20585–20589.
173. S. Li, P. Zhang, Y. Wang, D. Liu, J. Wu, H. Sarvari, Z. D. Chen and H. Sarvari, Nano Res., 2017, 10, 1092–1103.
174. X. Yao, J. Liang, T. Li, L. Fan, B. Shi, C. Wei, Y. Ding, Y. Li, Y. Zhao and X. Zhang, Sci. China Mater., 2018, 61, 65–72.
175. A. Thote, I. Jeon, H.-S. Lin, S. Manzhos, T. Nakagawa, D. Suh, J. Hwang, M. Kashiwagi, J. Shiomi, S. Maruyama, H. Daiguji and Y. Matsuo, ACS Appl. Electron. Mater., 2019, 1, 389–396.
176. A. N. Banerjee, S. W. Joo and B. K. Min, J. Nanosci. Nanotechnol., 2014, 14, 7970–7975.
177. K. Mahmood, B. S. Swain and A. Amassian, Nanoscale, 2014, 6, 14674–14678.
178. K. Mahmood, A. Khalid, M. Hameed, F. Rehman, M. T. Mehran and R. Song, Appl. Phys. A, 2018, 124, 1–6.
179. K. Sekar, R. Nakar, J. Bouclé, R. Doineau, K. Nadaud, B. Schmaltz and G. Poulin-Vittrant, Nanomaterials, 2022, 12, 2093.
180. S. Li, P. Zhang, H. Chen, Y. Wang, D. Liu, J. Wu, H. Sarvari, Z. D. Chen and H. Sarvari, J. Power Sources, 2017, 342, 990–997.
181. M. M. Tavakoli, R. Tavakoli, Z. Nourbakhsh, A. Waleed, U. S. Virk and Z. Fan, Adv. Mater. Interfaces, 2016, 3, 1500790.
182. Q. Yu, R. Lin, L. Jiang, J. Wan and C. Chen, Sci. China Mater., 2018, 61, 1007–1011.
183. C. Chung, B. T. Tran, K. Lin, Y. Ho, H. Yu, N. Quan and E. Chang, Nanoscale Res. Lett., 2014, 9, 338.
184. Q. Yu, L. Y. Jiang and T. T. Ai, Mater. Lett., 2015, 152, 142–144.
185. G. Zhang, Y. Liu, H. Morikawa and Y. Chen, Cellulose, 2013, 20, 1877–1884.
186. S. Wirunchit and W. Koetniyom, Phys. Status Solidi A, 2023, 220, 2200364.
187. J. Zhang, P. Barboux and T. Pauporté, Adv. Energy Mater., 2014, 4, 1400932.
188. J. Zhang, E. J. Juárez-Pérez, I. Mora-Seró, B. Viana and T. Pauporté, J. Mater. Chem. A, 2015, 3, 4909–4915.
189. C. M. Pelicano and H. Yanagi, J. Energy Chem., 2018, 27, 455–462.
190. O. N. Salman, M. O. Dawood, A. K. Ali, D. S. Ahmed and K. I. Hassoon, Dig. J. Nanomater. Bios., 2017, 12, 719–726.
191. V. J. Garcia, C. M. Pelicano and H. Yanagi, Thin Solid Films, 2018, 662, 70–75.
192. K. Mahmood, A. Khalid, M. S. Zafar, F. Rehman, M. Hameed and M. T. Mehran, J. Colloid Interface Sci., 2019, 538, 426–432.
193. M. H. Kumar, N. Yantara, S. Dharani, M. Graetzel, S. Mhaisalkar, P. P. Boix and N. Mathews, Chem. Commun., 2013, 49, 11089–11091.
194. V. T. Tran, Y. Wei, H. Yang, Z. Zhan and H. Du, Nanotechnology, 2017, 28, 095204.
195. S. El Ouakili, H. Zahdi, S. Laalioui, A. Rajira, Z. Aqachmar, A. Abounadi, A. Elhichou, A. Almaggoussi and N. Rochdi, Opt. Mater., 2024, 156, 115931.
196. Y. Yuan, L. Yang, X. Peng, Q. Chen, X. Li, C. Zuo, Z. Zhou and Z. Bai, J. Mater. Sci., 2024, 59, 11454–11467.
197. W. Huang, R. Zhang, X. Xia, P. Steichen, N. Liu, J. Chu and L. Li, Nanomaterials, 2021, 11, 329.
198. J. Cao, B. Wu, R. Chen, Y. Wu, Y. Hui, B.-W. Mao and N.-F. Zheng, Adv. Mater., 2018, 30, 1705596.
199. X. Zhang, D. Zhang, T. Guo, J. Zou, J. Jin, C. Zheng, Y. Zhou, Z. Zhu, Z. Hu, Q. Cao, S. Wu, J. Zhang and Q. Tai, J. Mater. Chem. A, 2023, 11, 9616–9625.
200. K. H. Lee, R. Farheen, Z. Arshad, M. Ali, H. Hassan, M. Alshareef, A. Dahshan and U. Khalid, RSC Adv., 2024, 14, 15391–15407.
201. T. Das, R. Nag, N. K. Rana, M. Nayak, R. Paramanik, A. Bera, S. K. Saha and A. Guchhait, Energy Fuels, 2023, 37, 10642–10651.
202. O. Cardozo, A. Stingl and S. Farooq, Mater. Sci. Semicond. Process., 2024, 180, 108537.
203. M. Gantumur, M. I. Hossain, M. Shahiduzzaman, A. Tamang, J. H. Rafij, M. Shahinuzzaman, H. T. C. Tu, M. Nakano, M. Karakawa, K. Ohdaira, H. Almohamadi, M. A. Ibrahim, K. Sopian, M. Akhtaruzzaman, J. M. Nunzi and T. Taima, ACS Appl. Mater. Interfaces, 2024, 16, 36255–36271.
204. J. H. Kang, A. Song, Y. J. Park, J. H. Seo, B. Walker and K.-B. Chung, Polymers, 2020, 12, 737.
205. P.-C. Pan, H.-S. Koo, D.-X. Chen and C.-M. Chen, Crystals, 2022, 12, 1032.
206. S. Ahmed, M. Akhtaruzzaman, W. Zulhafizhazuan, Y. Yusoff, I. Alnaser, M. R. Karim, M. Shahiduzzaman and K. Sobayel, Jpn. J. Appl. Phys., 2023, 62, 092001.
207. J. Kruszyńska, J. Ostapko, V. Ozkaya, B. Surucu, O. Szawcow, K. Nikiforow, M. Hołdyński, M. M. Tavakoli, P. Yadav and M. Kot, et al. , Adv. Mater. Interfaces, 2022, 9, 2200575.
208. L. Yang, T. Xu, Z. Bai and S. Qin, J. Phys. Chem. C, 2023, 127, 7492–7500.
209. A. Pramothkumar, N. Senthilkumar, S. Pitchaiya, N. Eswaramoorthy, V. Madurai Ramakrishnan and I. V. Potheher, J. Mater. Sci.: Mater. Electron., 2023, 34, 627.
210. M. Al-Rasheidi, F. Khan, A. Al-Ahmed, S. Rehman and F. Al-Sulaiman, Opt. Mater., 2022, 126, 112144.
211. R. Bashir, M. K. Bilal and A. Bashir, Sol. Energy, 2023, 256, 67–75.
212. Z. Arshad, S. Wageh, T. Maiyalagan, M. Ali, U. Arshad, N. ul-ain, M. B. Qadir, F. Mateen, A. G. Al-Sehemi and K. Sehemi, Ceram. Int., 2022, 48, 24363–24371.
213. Y. Li, N. Liu, Z. Xu, Z. Xu, Y. Pan, J. Zhang, L. Huang, Z. Hu, Y. Zhu and X. Liu, Appl. Phys. Lett., 2023, 123, 163901.
214. M. V. Rajendran, S. Ganesan, V. S. Menon, R. K. Raman, A. Alagumalai, S. Ashok Kumar and A. Krishnamoorthy, ACS Appl. Energy Mater., 2022, 5, 6671–6686.
215. H. M. Abd El-Lateef, M. M. Khalaf, F. E.-T. Heakal, M. F. Abou Taleb and M. Gouda, Sol. Energy, 2023, 253, 453–461.
216. N. Eswaramoorthy and K. Rajaram, Diam. Relat. Mater., 2023, 136, 109917.
217. X. Wu, J. Zhang, M. Qin, K. Liu, Z. Lv, Z. Qin, X. Guo, Y. Li, J. Xu, G. Li, H. Yan and X. Lu, EcoMat, 2022, 4, e12192.
218. X. Zhang, D. Zhang, T. Guo, J. Zou, J. Jin, C. Zheng, Y. Zhou, Z. Zhu, Z. Hu, Q. Cao, S. Wu, J. Zhang and Q. Tai, J. Mater. Chem. A, 2023, 11, 9616–9625.
219. B. Liu, B. Guo and R. Balamurugan, Nanomaterials, 2020, 10, 1753.
220. P. Fan, J. Tian, K. Wang, D. Zheng, J. Yu and T. P. Russell, J. Mater. Chem. C, 2024, 12, 19219.
221. M. A. Mahmud, N. K. Elumalai, M. B. Upama, D. Wang, K. H. Chan, M. Wright, C. Xu, F. Haque and A. Uddin, Sol. Energy Mater. Sol. Cells, 2017, 159, 251–264.
222. S. Wu, M. Lin, S. Chang, W. Tu, C. Chu and Y. Chang, J. Phys. Chem. C, 2018, 122, 236–244.
223. J. Li, J. Yang, J. Ma, J. Liang, Y. Liu, X. Hu, C. Chen, W. Yang, J. Min, Q. Bao, G. Fang and C. Tao, Chem. Eng. J., 2021, 417, 129247.
224. G. Bagha, K. Samavati, H. Naffakh-Moosavy and L. F. Matin, Sci. Rep., 2024, 14, 4617.
225. X. Xu, H. Zhang, J. Shi, J. Dong, Y. Luo, D. Li and Q. Meng, J. Mater. Chem. A, 2015, 3, 19288–19293.
226. M. Dkhili, G. Lucarelli, F. De Rossi, B. Taheri, K. Hammedi, H. Ezzaouia, F. Brunetti and T. M. Brown, ACS Appl. Energy Mater., 2022, 5, 4096–4107.
227. W. Qiu, M. Buffière, G. Brammertz, U. Paetzold, L. Froyen, P. Heremans and D. Cheyns, Org. Electron., 2015, 26, 30–35.
228. J. Zhang, C. H. Tan, T. Du, M. Morbidoni, C.-T. Lin, S. Xu, J. R. Durrant and M. A. McLachlan, Sci. Bull., 2018, 63, 343–348.
229. Y. Cheng, Q. Yang, J. Xiao, Q. Xue, H. Li, Z. Guan, H. Yip and S. Tsang, ACS Appl. Mater. Interfaces, 2015, 7, 19986–19993.
230. Z. Qiu, S. Yuan, H. Gong, H. Zhang, X. Qiu, T. Luo and B. Cao, J. Am. Ceram. Soc., 2017, 100, 176–184.
231. R. Runjhun, E. A. Alharbi, Z. Drużyński, A. Krishna, M. Wolska-Pietkiewicz, V. Škorjanc, T. P. Baumeler, G. Kakavelakis, F. Eickemeyer, M. Mensi, S. M. Zakeeruddin, M. Grätzel and J. Lewiński, Energy Environ. Mater., 2024, 7, e12720.
232. V. Murugadoss, D. Y. Kang, W. J. Lee, I. G. Jang and T. G. Kim, Adv. Compos. Hybrid Mater., 2022, 5, 1385–1395.
233. T. Tulus, S. Olthof, M. Marszalek, A. Peukert, L. A. Muscarella, B. Ehrler, O. Vukovic, Y. Galagan, S. C. Boehme and E. von Hauff, ACS Appl. Energy Mater., 2019, 2, 3736–3748.
234. T. Tulus, L. A. Muscarella, Y. Galagan, S. C. Boehme and E. von Hauff, Electrochim. Acta, 2022, 433, 141215.
235. S. Tsarev and P. A. Troshin, Synth. Met., 2020, 269, 116547.
236. Z. Pang, S. Yang, Y. Sun, L. He, F. Wang, L. Fan, S. Chi, X. Sun, L. Yang and J. Yang, Chem. Eng. J., 2022, 439, 135701.
237. G. L. Liu, Y. Zhong, H. D. Mao, J. Yang, R. Y. Dai, X. T. Hu, Z. Xing, W. P. Sheng, L. C. Tan and Y. W. Chen, Chem. Eng. J., 2022, 431, 134235.
238. A. Yella, L. Heiniger, P. Gao, M. K. Nazeeruddin and M. Grätzel, Nano Lett., 2014, 14, 2591–2596.
239. N. Vaenas, D. Konios, T. Stergiopoulos and E. Kymakis, RSC Adv., 2015, 5, 107771–107776.
240. B. J. Kim, D. H. Kim, Y. Y. Lee, H. W. Shin, G. S. Han, J. S. Hong, K. Mahmood, T. K. Ahn, Y. C. Joo, K. S. Hong, N. G. Park, S. Lee and H. S. Jung, Energy Environ. Sci., 2015, 8, 916–921.
241. X. Chen, L. J. Tang, S. Yang, Y. Hou and H. G. Yang, J. Mater. Chem. A, 2016, 4, 6521–6526.
242. F. Cai, L. Yang, Y. Yan, J. Zhang, F. Qin, D. Liu, Y. B. Cheng, Y. Zhou and T. Wang, J. Mater. Chem. A, 2017, 5, 9402–9411.
243. Z. Lan, X. Xu, X. Zhang, J. Tang, L. Zhang, X. He and J. Wu, J. Mater. Chem. C, 2018, 6, 3341–3351.
244. M. S. You, J. H. Heo, J. K. Park, S. H. Moon, B. J. Park and S. H. Im, Sol. Energy Mater. Sol. Cells, 2019, 194, 1–6.
245. J. Nam, J. H. Kim, C. S. Kim, J.-D. Kwon and S. Jo, ACS Appl. Mater. Interfaces, 2020, 12, 12648–12655.
246. Y. Sanehira, N. Shibayama, Y. Numata, M. Ikegami and T. Miyasaka, ACS Appl. Mater. Interfaces, 2020, 12, 15175–15182.
247. Z. Ren, N. Wang, P. Wei, M. Cui, X. Li and C. Qin, Chem. Eng. J., 2020, 393, 124731.
248. T. Homola, J. Pospisil, M. Shekargoftar, T. Svoboda, M. Hvojnik, P. Gemeiner, M. Weiter and P. Dzik, ACS Appl. Energy Mater., 2020, 3, 12009–12018.
249. B. Pan, J. Gu, X. Xu, L. Xiao, J. Zhao and G. Zou, Nano Res., 2021, 14, 3431–3438.
250. S. Y. Hong, H. J. Lee, J. K. Park, J. H. Heo and S. H. Im, Int. J. Energy Res., 2022, 46, 22819–22831.
251. B. Liu, G. Sun, Q. Sun, Y. Lv, M. Huang and B. Qi, Mater. Sci. Semicond. Process., 2022, 138, 106303.
252. M. J. Carnie, C. Charbonneau, M. L. Davies, J. Troughton, T. M. Watson, K. Wojciechowski, H. Snaith and D. A. Worsley, Chem. Commun., 2013, 49, 7893–7895.
253. J. Zhang, A. Hultqvist, T. Zhang, L. Jiang, C. Ruan, L. Yang, Y. Cheng, M. Edoff and E. M. J. Johansson, ChemSusChem, 2017, 10, 3810–3817.
254. H. Liu, Z. Huang, S. Wei, L. Zheng, L. Xiao and Q. Gong, Nanoscale, 2016, 8, 6209–6221.
255. Y. Qiang, Y. Xie, Y. Qi, P. Wei, H. Shi, C. Geng and H. Liu, Sol. Energy, 2020, 201, 523–529.
256. C. Chen, C. Li, F. Li, F. Wu, F. Tan, Y. Zhai and W. Zhang, Nanoscale Res. Lett., 2014, 9, 457.
257. E. Y. Choi, J. Kim, S. Lim, E. Han, A. W. Y. Ho-Baillie, N. Park and N. Park, Sol. Energy Mater. Sol. Cells, 2018, 188, 37–45.
258. R. Singh, S. Ghosh, A. S. Subbiah, N. Mahuli and S. K. Sarkar, Sol. Energy Mater. Sol. Cells, 2020, 205, 110289.
259. P. P. Rajbhandari and T. P. Dhakal, J. Vac. Sci. Technol., A, 2020, 38, 032406.
260. J. Feng, Z. Yang, D. Yang, X. Ren, X. Zhu, Z. Jin, W. Zi, Q. Wei and S. Liu, Nano Energy, 2017, 36, 1–8.
261. T. Yokoyama, Y. Nishitani, Y. Miyamoto, S. Kusumoto, R. Uchida, T. Matsui, K. Kawano, T. Sekiguchi and Y. Kaneko, ACS Appl. Mater. Interfaces, 2020, 12, 27131–27139.
262. D. Shen, W. Zhang, Y. Li, A. Abate and M. Wei, ACS Appl. Nano Mater., 2018, 1, 4101–4109.
263. X. Ye, H. Ling, R. Zhang, Z. Wen, S. Hu, T. Akasaka, J. Xia and X. Lu, J. Power Sources, 2020, 448, 227419.
264. Y. Shang, T. Zhang, D. Yu, Z. Peng, W. Zhou, D. Yin and Z. Ning, ACS Appl. Mater. Interfaces, 2021, 13, 47603–47609.
265. J. Duan, Y. Yang, X. Xie, K. Zhang, H. Wan, J. Zhang, L. Tao and W. Wang, J. Mater. Res., 2022, 37, 2825–2836.
266. Q. Dong, F. Z. Liu, M. K. Wong, A. B. Djurisic, Z. W. Ren, Q. Shen, A. Ng, C. Surya and W. K. Chan, Oxide-Based Materials and Devices VII, 2016, vol. 9749, pp. 97491S.
267. B. Yang, R. Ma, Z. Wang, D. Ouyang, Z. Huang, J. Lu, X. Duan, L. Yue and W. C. H. Choy, ACS Appl. Mater. Interfaces, 2021, 13, 27179–27187.
268. D. Jia, X. Feng, J. Jia, B. Shi, Y. Wu and B. Cao, J. Solid State Chem., 2023, 317, 123661.
269. D. Guo, J. Ma, S. Lin, X. Guo, H. Huang, D. Kong, F. Xu, Y. Gao, W. Zhang, Y. Hu and C. Zhou, Appl. Phys. Lett., 2022, 120, 263502.
270. L. J. Chen, G. Wang, L. B. Niu, Y. Q. Yao, Y. X. Guan, Y. T. Cui and Q. L. Song, RSC Adv., 2018, 8, 15961–15966.
271. C. Gong, H. Li, H. Wang, C. Zhang, Q. Zhuang, A. Wang, Z. Xu, W. Cai, R. Li, X. Li and Z. Zang, Nat. Commun., 2024, 15, 4922.
272. W. Shockley and H. J. Queisser, J. Appl. Phys., 1961, 32, 510–519.
273. B. Y. Shao, X. Song, H. W. Zhu, Y. A. Bioud, W. T. Wu, M. Abulikemu, H. Saiari, S. Aqeel, I. Gereige, O. F. Mohammed and O. M. Bakr, Chem. Soc. Rev., 2025, 54(21), 9939–9977.
274. X. Qiu, Y. Jiang, H. Zhang, Z. Qiu, S. Yuan, P. Wang and B. Cao, Phys. Status Solidi RRL, 2016, 10, 587–591.

---