Introduction

Renewable energy has been increasing its share in the electricity grid due to the target of net-zero emissions, alongside the rapid growth of solar power1. To further reduce the levelized cost of electricity (LCOE) of solar technologies, metal halide perovskite solar cells (PSCs) are emerging candidates in next-generation photovoltaics (PVs). Of all types, inverted PSCs (IPSCs), with a positive-intrinsic-negative (p-i-n) structure, exhibit many benefits, such as diversified material choices, negligible hysteresis, and growing power conversion efficiencies (PCEs, see in Fig. 1a and Table 1)2,3,4,5,6. When incorporated into tandem cells, the inverted configuration has fewer parasitic absorption losses than the mesoporous or regular (n-i-p) counterparts7, conferring the PCEs (Fig. 1b) of perovskite/Si (approaching 30%) and perovskite/perovskite (26.4%) surpassed the single crystal Si solar cell (26.1%) or even its theoretical maximum (29.4%)8,9. Thanks to the high tolerance to high-energy radiation than Si10,11,12, perovskite/Cu(In, Ga)Se2 (CIGS) tandem cells (with a PCE of 24.2%, higher than CIGS of 23.4%) have a promising future in space application.

Fig. 1: The efficiency summary of IPSCs.
figure 1

Evolution of a Small-area single-junction cells. b Tandem cells. c Minimodules. Note that the inset symbol in (c) is proportional to the actual aperture areas (AAs). Aside from the separately labeled, the PCEs and AAs are all certified, and the details can be found in Table 1.

Full size image
Table 1 The evolution of IPSCs, consisting of small-area single-junction cells, tandem cells, and high-performance perovskite minimodules (over 14% PCEs).
Full size table

In terms of mass production, the IPSCs are compatible with upscaling manufacturing that can construct the whole device by scalable solution-based processing or vapor-phase evaporation (increased PCEs and areas of minimodules are shown in Fig. 1c)13. Coupled with low-temperature attributes (normally below 130 °C), the IPSCs are suitable for flexible substrates and roll-to-roll processes, contributing to wearable, portable and ultra-light electronics14,15. In addition to these advantages, the stability of IPSCs has improved significantly in recent years, with enhanced robustness against operational stressors16. For example, IPSCs could maintain 96.8% of the initial PCEs after 1200 h at 65 °C17 or have no efficiency loss after 1000 h (40 °C in a nitrogen atmosphere)3 during the maximum power point tracking (MPPT). Under open-circuit condition (~0.76 Sun in ambient air at 85 °C), IPSCs showed a T95 (retained 95% of the post-burn-in efficiencies) of 1200 h16. More encouragingly, IPSCs have passed the industrial damp-heat test most recently6. Thus, a timely review focusing on the stability progress is highly necessary considering the rapid development of IPSCs.

In this Review, we aim to discuss the current challenges and future opportunities of IPSCs based on rigid substrates and hybrid three-dimensional (3D) lead perovskite, albeit the all-inorganic perovskites18, 2D perovskites19, and tin-based perovskites20,21 are studied elsewhere. We first review the advance of stability research conducted on the IPSCs, starting from each functional layer, such as electrodes, charge transport layers, perovskite light absorbers, and the interfaces thereof, to the entire devices under environmental or operational stressors. The following section lists a series of strategies to improve the stability of IPSCs, including small-area devices, tandem cells, and minimodules. A better understanding of these approaches would provide insights into their commercialization pathways. The next focus is on the stability assessment of IPSCs, in which the details of MPPT are compared. The final section offers our concerns about the current stability protocols and perspectives for future research and development.

Stability of each functional layer and interface

To make a good knowledge of stability in IPSCs, we would like to disassemble the device into two parts: functional layers and interfaces. As shown in Fig. 2a, a conventional IPSC is generally composed of six functional layers and five interfaces. In a typical fabrication process, a p-type hole transport layer (HTL), a perovskite light absorber, and an n-type electron transport layer (ETL) are deposited in order. Also, an additional buffer layer is required between an ETL and a metal electrode. These functional layers have diversified material choices as illustrated in Fig. 2b, most of which are coming from commercial candidates and can be prepared by solution processing with low temperatures. In addition to intrinsic properties (physical, chemical, or structural issues), the stability of functional layers and interfaces can be also evaluated by the performance stability of the fabricated devices. For entire devices, there are many factors, also called stressors, influencing the device lifetime as summarized in Fig. 2c. Based on these multiple stressors, the material stability of each functional layer and the operational stability of IPSC devices will be emphasized in the following discussion.

Fig. 2: Overview of IPSCs.
figure 2

a Schematic configuration and corresponding materials for functional layers. Note that HTL denotes the hole transport layer. ETL denotes the electron transport layer. ITO denotes indium tin oxide and FTO denotes fluorine-doped tin oxide. The IPSCs based on the exhibited materials showed both high efficiency and long stability. b Energy level diagram of functional layers in high-performance IPSCs. c The stressors affecting the device lifetime, including environmental and operational stressors.

Full size image

Electrode layer

The electrodes (the bottom and top) are typically made of transparent conductive oxide (TCO) and metal contact, respectively. The bottom TCO not only supports the thin-film deposition but also acts as the anode in IPSCs. For TCO electrodes, indium tin oxide (ITO) and fluorine-doped tin oxide (FTO) coated glass are the two main types of rigid substrates. Unlike FTO with higher resistance of temperature, the ITO is more frequently used in IPSCs due to the better transmittance at a given conductivity51 and low-temperature requirement for solution-processed HTLs. Although the ITO is relatively stable (lower thermal expansion coefficients52) compared to other functional layers, there is still potential to improve the device stability by modifying the ITO. Our recent work discloses that replacing ITO with single-walled carbon nanotubes (SWCNTs) can further improve the environmental and mechanical stability of the fabricated devices53. The top electrode (cathode) is normally evaporated by metals, such as Ag, Cu, Au, Al54,55,56. These materials are widely used thanks to their good electric conductivity whilst chemical-stability issues are also encountered. Al is less stable than other candidates in the air according to the reactivity order. Ag and Au could go through the charge transport layer to form “metal halide” with the perovskites because of the ion-diffusion process57,58. Cu can be oxidized by moisture and oxygen and the products of Cu(OH)2 or CuOx would also react with the perovskite56. Based on the performance data, Au-based IPSC devices show excellent operational stability under 85 °C16,59, but the high cost will inhibit its wide application. From an economic perspective, Cu is a favorable candidate for large-area perovskite devices (see in Table 1). Moreover, non-metal materials (like carbon60, ITO58, Al-doped zinc oxide (AZO)61, and indium zinc oxide (IZO)3), are useful for thermal stability although the devices would sacrifice efficiency due to the increased series resistance. Besides, these transparent electrodes are essential in the fabrication of perovskite tandem cells while the complex processing may inhibit their application in large modules.

Hole transport layer (HTL)

Although HTL-free structure has been reported62, the HTL is indispensable for highly stable IPSCs16,25. The most famous hole transport material, 2,2′,7,7′-Tetrakis[N,N-di(4-methoxyphenyl)amino]-9,9′-spirobifluorene (spiro-OMeTAD), cannot be applied in IPSCs because of its solubility in perovskite solvents63. Nevertheless, there are many other candidates, which are divided into inorganic and organic categories. For inorganic HTLs, nickel oxide (NiOx,) achieved a higher device efficiency than other materials26. However, in terms of the chemical property, the NiOx is a photo-catalyst and prone to be Ni2O364, indicating less stability under light and oxygen65. Also, the NiOx device showed inferior stability induced by electrical bias66 or thermal heating (85 °C testing in nitrogen)16. This might be ascribed to the high defects concentration of NiOx that the Ni≥3+ would interact with the perovskite, leading to severe interfacial recombination67,68. Therefore, the organic HTLs are essential to be introduced, yielding the champion efficiency of 24.3% (certified) in the IPSCs27. Poly (3,4-ethylene dioxythiophene):poly(styrenesulfonate) (PEDOT: PSS) is the first HTL used in IPSCs22 but its acidic and hygroscopic properties deteriorate the device durability69. Thus, more organic polymers were studied, such as poly[N,N′-bis(4-butylphenyl)-N,N′-bis(phenyl)benzidine] (poly-TPD)70, poly(bis(4-phenyl)(2,4,6-trimethylphenyl)amine) (PTAA)63, and poly[3-(4-methylamine carboxylbutyl)thiophene] (P3CT-N)5, all of which showed excellent device efficiencies and performance stability in lab-scale devices or minimodules. In recent years, self-assembled organic materials have emerged increasingly, like [2-(9H-carbazol-9-yl)ethyl]phosphonic acid (2PACz)38, poly[3-(6-carboxyhexyl)thiophene-2,5-diyl] (P3HT-COOH)71, 4′-(3,6-bis(2,4-dimethoxyphenyl)-9H-carbazol-9-yl)-[1,1′-biphenyl]-4-carboxylic acid (EADR04)72, and [4-(3,6-dimethyl-9H-carbazol-9-yl)butyl]phosphonic acid (Me-4PACz)9, with excellent performance stability in single junction6 and tandem cells9,35.

Perovskite light absorber

The hybrid three-dimensional (3D) lead perovskite with a formula of APbX3 (A denotes organic or inorganic cations and X denotes halide anions) is still the mainstream in IPSCs. To fabricate perovskite thin films, various deposition methods have been created while solution processing is still one of the best choices for future commercialization. Aside from processing in extremely low-humidity and oxygen conditions (like in a glove box, dry box, or vacuum chamber), the growth of perovskite thin film can be conducted in ambient conditions with controlled humidity. The most recent study displayed a certified PCE of 23.5% based on the annealing process in the air (relative humidity, RH: 30–40%)5. However, perovskite stability is still the key issue since the as-crystalized perovskite will decompose in the presence of water and oxygen73. Also, due to the ionic and polycrystalline nature, the perovskite film inevitably contains numerous defects that further accelerate the aging process. For example, halide vacancies could trap oxygen to form superoxide radicals under illumination74, which is detrimental to the perovskite film.

Electron transport layer (ETL) and buffer layer

The fullerene ETLs are predominant for high-performance devices, mainly consisting of C60 and [6,6]-phenyl-C61-butyric acid methyl ester (PC61BM, abbreviated as PCBM in the following section). Thanks to the Lewis-acid property, these two fullerene materials can passivate electron-rich defects in perovskite75. With complementary passivation, double-fullerene layers (PCBM and C60) were frequently adopted to suppress traps76, retarding the intrusion of water and oxygen to some extent. The recent IPSCs based on mere PCBM or C60 ETL also showed excellent operational stability6,16. This may be due to the ‘scavenging effect’ of PCBM77. On top of ETL, introducing a buffer layer can reduce the interfacial barrier at ETL/electrode78. Bathocuproin (BCP) is the most common use in IPSCs, yielding the highest PCE25 and longest operational stability16,79. The BCP deposition is compatible with the solution or evaporation techniques. However, thermal evaporation could cause crystalization, so the solution-processed BCP devices showed better thermal stability50. Furthermore, BCP can be replaced by other candidates to improve device durability, such as metal acetylacetonates80, metal-organic frameworks81, atomic-layer-deposition (ALD) SnO23, Zn-doped SnO2 (ZTO)31, and organometallic complexes82.

Interfaces

Interfaces have multifunction in IPSCs, involving energy-level alignment, charge dynamics (extraction, transfer, and recombination), defect passivation, and ion migration83. Thus, the imperfect contacts will induce interfacial losses and impair the device performance. From bottom to top, the first interface is located between TCO and HTL. The HTLs have a tight adhesion and low thickness (normally less than 10 nm for polymer and 1–3 nm for monolayer small molecules)2,38, leading to challenges to study TCO/HTL interface. Also, the buffer layer (also with a thin thickness of several nanometers)3 can be regarded as an interlayer between ETL and electrode so the interfaces between ETL/buffer layer and buffer layer/metal, are less studied. Overall, most studies focus on the adjacent interfaces of perovskite (HTL/perovskite and perovskite/ETL), because these two interfaces are prone to have high trap densities in comparison with high-quality perovskite films84,85. The buried interface between HTL and perovskite was found to cause most deep traps due to the sub-microscale imperfections or inhomogeneities85,86. Through lift-off or peel-off strategies, clear voids are observed at this interface25. These voids will be enlarged after several-hour light soaking, leading to device aging. In comparison with HTL/perovskite interface, the perovskite/ETL interface has severe non-radiative recombination when using fullerene ETLs87. In addition, the ion diffusion from perovskite to metal electrode (bilateral process) is the other factor affecting the device stability.

Strategies for improving the stability of IPSCs

To improve the stability of IPSCs, many strategies have been investigated, such as materials engineering of charge transport layers (CTLs) or electrodes, compositional engineering of perovskites, and interface engineering of whole devices. In this section, we will discuss these approaches in more detail. The above layer-by-layer discussion indicates that the intrinsic stability of IPSCs is closely related to each functional layer, in particular for the CTLs. For bottom HTLs, organic materials are more beneficial for stability based on the present device performances5,6,16. This could be attributed to their unique properties. Specifically, the cross-linking process of polymer HTLs enhances the resistance of solvent, moisture, and mechanical stress88. The self-assemble process between small organic molecules and TCO increases the cohesion and adhesion toughness (low fracture energy, Gc, J m–2)72,89. Thus, both polymer and small organic-molecule HTLs show excellent operational stability5,6,16. For top ETLs, modified C60 and PCBM enabled the devices to improve moisture stability. For example, the cross-linkable silane with C60 (C60-SAM)90 or PCBM-derived electrolyte (PCBB-3N-3I)91 enhances the water resilience but it still needs to combine with the pure C60 or PCBM in the fabrication of devices. For pure ETLs, non-fullerene substitutes were found to improve thermal stability. Small rylene diimide molecules, like N,N′-Bis(1-indanyl)naphthalene-1,4,5,8-tetracarboxylic diimide (NDI-ID)92, and N-type conjugated polymer, like poly(2,2′-bithiazolothienyl-4,4′,10,10′-tetracarboxydiimide) (PDTzTI)93, achieved PCEs over 20% with better thermal stability than PCBM based devices. Hence, the material design of ETLs is the breakthrough for device stability although the C60 and PCBM are irreplaceable for the state-of-the-art high-performance IPSCs.

In addition to CTLs, the perovskite light absorber is still the key component. Finding means to increase the intrinsic stability of perovskite bulk film will prolong the device lifetime. Compositional engineering (tunning APbX3 or introducing additives) is a straightforward method, in which the A sites can affect the perovskite stability. Formamidinium (FA, CH3(NH2)2+) or cesium (Cs+) is known to be more thermal stable than methylammonium (MA, CH3NH3+). The mixed A-site perovskites, such as MAFA, CsFA, CsMAFA, could further increase the phase, photo, moisture, and thermal stability compared to the pure FA perovskite94. Therefore, the multi-cation compositions constitute perovskite light absorbers in most stable IPSCs2,3,6,16. Moreover, non-stoichiometric perovskite precursors also showed advantages. Apart from excess PbI2 (0.5–1 mol%2), recent studies found that adding AX in perovskite precursor can suppress redox reactions between the perovskite and NiOx68 and enhance the photostability of FAxCs1-xPbI3 perovskite film50. To further improve the operational stability, a series of ligands and additives were introduced in perovskite precursors, such as alkylamine3 and ionic liquid salts (Fig. 3a)16,59. According to the device data, the ligands design would be a useful aspect to enhance the perovskite stability.

Fig. 3: Typical organic structures for stable IPSCs.
figure 3

a Chemicals added in perovskite. b Chemicals applied at interfaces between the perovskite and charge transport layers. Note that the chemicals of the top row (PMMA, PS, PEAI, F-PEAI) in (b) can be applied in different interfaces, such as the HTL/perovskite and perovskite/ETL.

Full size image

The optimized interfaces could further delay the device degradation and extend the operational lifetime. Thanks to facile processing, interface engineering is widely adopted in IPSCs. As summarized in Table 2, numerous passivation agents (modulators or passivators, also seen in Fig. 3b) were introduced in the fabricated devices, such as post-treatments that resided on the perovskite surface. In addition to physical methods (microwave annealing95 and mechanically peeling off96), introducing hydrophobic insulating polymer (electron can transport from perovskite to ETL through tunneling effect) could significantly increase water or moisture resistance97. Also, the reacted capping layer (sulfidation5 or oxidation17 with excess lead source) can improve the operational stability since it significantly inhibited ion migration or diffusion under illumination. Even for the encapsulated devices, this ion-diffusion process is inevitable and thus many barrier layers were introduced between the buffer layer and metal electrode, such as Cr/Cr2O357,59, bismuth (Bi)98, benzotriazole (BTA)99. Moreover, a recent study found that two-dimensional (2D) perovskite layers (Ruddlesden-Popper phase, post-treated by oleylammonium iodide) can prolong the performance stability6. Although 2D perovskite has low charge-carrier mobility and short diffusion length100, the hydrophobic large cations could enable the 2D perovskite as the moisture or oxygen barrier. Also, the 2D perovskite can passivate surface defects and block ion migration, contributing to one of the best efficiencies (24.3%) and passing the standard industrial damp-heat test6. However, the 2D perovskite is highly dependent on the fabrication method since the oleylamine added in the perovskite precursor did not show any 2D component3. Hence, the detailed formation process of 2D perovskite merits further investigation.

Table 2 The summary of instability origins and corresponding strategies to improve the stability of IPSCs (based on rigid substrates).
Full size table

To enhance device stability, a large number of engineering methods have been developed in the IPSCs. However, more attention should be paid to utilizing these approaches. Some interfacial layers are useful for device efficiency but detrimental to operational stability (like LiF9). This indicates that the passivation materials may have the opposite impact on efficiency and stability. Introducing an interlayer or agent may cause new interfaces to form inside the devices and thus increase uncertainties with the time elapsed. Long-term illumination or other stressors could induce a new reaction between the passivation agents and perovskite or CTLs. Therefore, we should be cautious of material selections in the preparation of IPSCs and be more patient in verifying the operational stability even if an obvious increase in the initial PCE.

Many of these stability approaches, albeit impressive and useful, are based on lab-scale devices and normally require precise control in the preparation process (e.g., the 2D capping layer). Besides, it is questionable whether these methods can be scaled up in the fabrication of large-area modules. Additional factors of a specific upscaling deposition method will bring new challenges to use these stability approaches. For inverted minimodules, the stability is also affected by the entire design, like the laser scribing (P1-P2-P3) interconnection13. Thus, a standard geometry is helpful for all researchers, rather than only defining the geometrical fill factor (GFF) for different patterns. Aside from the aforementioned functional layers, an anti-reflection coating (such as magnesium fluoride layer (MgF2)46 and polydimethylsiloxane (PDMS)25) was additionally introduced in recent high-performance cells or minimodules to increase the PCE (due to the relatively lower current density (JSC) than that of the mesoporous and regular devices). However, the stability of these anti-reflective layers is yet to be investigated in IPSCs. Being exposed to the incident light directly, the anti-reflective layer requires more evidence to confirm its survival with long-term stressors.

Stability assessment of the entire device

Standardized testing protocols are prerequisites for commercialization. The stability assessments of IPSCs can learn from two mature standards: The International Summit on Stability of Organic Solar Cells (ISOS) and the International Electrotechnical Commission (IEC). The former provides various stability measurements for lab-scale devices and the latter (like IEC 61215:2016) is an industrial qualification. The IPSCs can refer to ISOS procedures, which judge the device stability at various levels, ranging from dark storage to operational stability120. To reflect real-world lifetime, operational stability is encouraged to be reported121. Of all types, maximum power point tracking (MPPT) is a typical measurement, which can simulate the operational performance by considering electrical bias, heat (self-heating effect under light source), and illumination simultaneously120,122. To perform the practical measurements, the detailed aging conditions need to be noted, involving the light source, bias, UV filter, temperature, atmosphere, encapsulation, device structure, initial PCE, and operational time. After years of endeavor, the operational stability of the IPSCs has improved significantly (Table 3). The MPPT results, as shown in Fig. 4a–c, indicate that the small-area single-junction device displays only 2% drops after 1500 h. The perovskite/Si tandem cell shows a PCE of around 28% after 300 h. The minimodule still maintained 85% of initial PCE after 1000 h. In addition to ISOS procedures, a suite of IEC 61215:2016 standards is complementary for device stability even though it looks stringent for the present IPSCs. Encouragingly, the encapsulated IPSCs have passed the damp-heat test (85 °C/85%RH)6, which inspires us to spend more effort on stability study and eventually push the IPSCs into the PV market.

Table 3 Comparisons of MPPT in high-performance IPSCs based on rigid substrates.
Full size table
Fig. 4: The summary of operational stability of IPSCs based on MPPT results.
figure 4

a Small-area single-junction cells. b Tandem cells. c Minimodules. Note that all these devices are based on rigid substrates. The initial efficiency of sample 1 in (a) is unclear so we use the claimed champion PCE of 25% for reference. For each point, the data were extracted from the representative works (see details in Table 3). LED denotes light-emitting diode. AM 1.5 indicates air mass 1.5. T80 denotes the lifetime at 80% of the initial efficiency.

Full size image

For stability assessments, the current protocols are not sufficient under certain conditions, such as under ultraviolet (UV) light and high-level radiation exposure. According to the IEC 61215 requirement, the minimum standard for UV illumination is 15 kW h m−2 (irradiation from 280 to 400 nm)123. However, the UV-light stability of IPSCs is seldom reported. For most MPPT measurements, the illumination source normally used LEDs or a cut-off filter (at 35072 or 4203nm). These lead to scarce knowledge of device stability under UV light. Thus, it is essential to investigate the UV-light stability in IPSCs since it can help to de-couple degradation mechanisms. In addition to the UV-irradiation, a high-radiation environment induced by high-charge and high-energy particles also need to be considered. The research found that NiOx is a problematic layer in perovskite/CIGS tandem cells under high-proton irradiation (68 MeV)124. However, such acute exposures cannot reflect the real conditions in the space station, Moon, Mars or even beyond, so a standard assessment for radiation degradation should be added in IPSCs. Aside from the UV or high-radiation stressor affecting the device intrinsic stability, external stability, such as encapsulation processing, also needs to be considered. Most encapsulations have so far used the methods directly from the previous PV experience, but the PSC degradation can be induced by these methods, like the electrical feedthrough (oxygen and water can diffuse) or the encapsulant itself (which may react with perovskite)125. The efficient encapsulations for IPSCs with long-term robustness are urgently needed. To commercialize IPSCs, real-world operation or field tests are indispensable and the IPSCs have shown promising durability. For example, the IPSC devices (initial PCEs are over 23%) show an obvious recovery after pausing 12 h in the 500 h MPPT (ISOS-L-3)26. Also, IPSC minimodules show superior tolerance to the shading effect (no permanent damage to the whole module when performing over 50 cycles of shading/de-shading for subcells)47. These characteristics indicate the excellent performance stability of IPSCs in real-world conditions.

Outlook

With the rapidly advancing device efficiency, the IPSC community faces an urgency for an in-depth understanding of the degradation mechanism. Only by resolving stability issues properly, the IPSCs would participate in the global PV market. To further develop the IPSCs toward commercialization, there are some suggestions are listed as follows.

  1. a.

    We encourage researchers to report the intrinsic stability of IPSCs (unencapsulated in the air with monitored temperature and humidity, or unencapsulated in inert conditions) when performing long-term operational measurements. This would offer a good opportunity to understand the underlying degradation mechanism. In addition, we appeal to give more descriptions of the fabrication process. For example, it needs to notify the lab conditions, such as chemistry laboratory or cleanroom. For the cleanroom, the classifications (by using International or US federal standards) should be given. Besides, the details of the used chemicals (e.g., the lot number) need to be considered since the material impurity is one of the unstable issues. We believe that these efforts will increase the reproducibility and development of the whole IPSC community.

  2. b.

    In terms of the solution-processed devices, the atmosphere effect, no matter in a glove box, dry box, or even in a fume hood, is a crucial factor for the final performance stability. For mass production, the fabrication processes will inevitably involve a large volume of solvents. Even if introducing the so-called green or environmentally friendly solvents, these solvents will affect perovskite growth and film stability. Therefore, we must supervise the solvent influence in the actual workshop (e.g., increase the sensor to detect and track the real-time solvent concentration).

  3. c.

    The stability measurement (e.g., the operational stability) needs to adopt standard facilities (e.g., consistently or uniformly approved light sources and generally accepted calibration method). To benefit inter-laboratory comparison, the details (e.g., the spectrum) of light sources should be given, in case of a mismatch from different suppliers. Due to the cost-effectiveness (low price with a several-year lifespan), the white LED is still a typical illumination source and its intensity calibration is also quite straightforward by comparing the device JSC with a standard solar simulator. Apart from white-LED stability, the UV or Infrared (IR) stability can be separately studied by using the particular UV or IR illumination since AM 1.5 spectrum contains these parts. Thus, the detailed wavelength and intensity of UV or IR light are required. We thus appeal to display the operational stability, at least to show the performance stability under a white LED (minimum requirement). To approach industrial standards, the assessment of operational stability could combine the damp-heat test and MPPT in future studies.

Further research regarding the durability of devices for industrial commercialization would be worthwhile. With the collective efforts of academia and industry, there will be a bright prospect for IPSCs. For future applications, the inverted perovskite tandem modules are most promising. For example, the perovskite/Si modules could replace Si in solar farms and roof-integrated panels whilst the perovskite/CIGS modules can be used in spacecraft and interstellar probes. Associating with the advance of inverted perovskite modules, these products could compensate for power consumption in our daily life.