Toxicity of Perovskite Solar Cells

Abstract

Over the past decade, there has been significant and rapid developments in organic-inorganic hybrid perovskite solar cells (PVSCs). Despite the fact that the power conversion efficiency (PCE) of PVSCs has increased from 3.8% to 25.8%, approaching that of commercial single crystalline Si solar cells, the market is still dominated by Si-based photovoltaic (PV) technology. This can be attributed to the challenges associated with upscaling PVSCs, improving device stability, and reducing the toxicity of PVSCs, which are hurdles in commercializing perovskite PV technologies. In particular, the toxicity due to lead leakage of PVSCs makes it difficult for them to enter the market. Hence, in this article, the structure and working principle of PVSCs are first summarized. Then, the toxicity of PVSCs is discussed, including the impacts of organic solvents and perovskite precursor materials on the health and environment. In this section, examples of advanced strategies for reducing the toxicity of PVSCs are also provided. Finally, challenges and a perspective for developing nontoxic PVSCs are given.

1. Introduction

With the rapid development of the world economy, the demand for electricity in various forms is on the rise all over the world [1]. However, the issues brought by traditional fossil energy, such as resource depletion and environmental damage [2], have become increasingly severe. Therefore, developing technologies to use renewable energy as an alternative power source is a growing tendency and has attracted tremendous attention from researchers. Among these technologies, solar cells based on the well-known photovoltaic (PV) effect, which can utilize clean and pollution-free solar energy with high efficiency, are considered one of the most promising technologies. Recently, the perovskite solar cells (PVSCs) have emerged as the most promising PV technology due to various advantages, such as high absorption coefficient, bandgap tunability, long carrier lifetime, and low recombination rate, which contribute to their high-power conversion efficiency (PCE) [3]. In fact, the world record PCE of PVSCs has reached 25.8%, approaching that of the single crystalline Si solar cells [4]. Moreover, the superb property that bandgap can be tuned within the range of 1.3 to over 3.0 eV via composition engineering has made PVSC the most promising material in the tandem solar cells and building integrated photovoltaics (BIPV) devices [5,6,7]. Consequently, researchers are increasingly interested in developing and optimizing PVSCs to improve their performance and enable their widespread adoption as a clean and efficient alternative to traditional fossil energy sources.

Development of Perovskite Solar Cells (PVSCs)

As shown in Figure 1a, organic-inorganic metal halide perovskites have a crystal structure of ABX₃ where the A site is occupied by organic or inorganic cations, such as methylammonium (MA⁺), formamidinium (FA⁺), and Caesium (Cs⁺); B site is occupied by metal cations, such as Pb²⁺ and Sn²⁺; and X site is occupied by halide anions, such as I⁻, Br⁻, and Cl⁻. As depicted in Figure 1b, the PVSCs have a sandwich structure with the perovskite active layer deposited between the electron transport layer (ETL) and hole transport layer (HTL). The specific ETL and HTL materials used in these architectures can vary depending on the application and the desired device performance. Some commonly used ETL materials include metal oxides, such as titanium dioxide (TiO₂) [8], zinc oxide (ZnO) [9], and tin oxide (SnO₂) [10], and organic materials, such as fullerene derivatives [11] and conjugated polymers [12]. On the other hand, common HTL materials used include organic small molecules, such as Spiro-OMeTAD (2,2′,7,7′-tetrakis(N,N-di-p-methoxyphenylamine)-9,9′-spirobifluorene) [13], and polymers, such as poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS) [14]. During device operation, the perovskite active layer absorbs the light under illumination and generates charge carriers. Then, the photo-generated electrons and holes are extracted by ETL and HTL, respectively. With a suitable energy level alignment between different layers, the charge carriers can be transported to the corresponding electrodes and form current through the outside circuit [15].

The device architectures of PVSCs, as shown in Figure 2, include both mesoscopic and planar structures. The n-i-p mesoscopic device architecture, as depicted in Figure 2a, employs a mesoporous material (e.g., TiO₂) as a scaffold that is deposited on the conductive glass substrate. Conversely, the n-i-p planar structure, as shown in Figure 2b, only comprises a planar layer on the substrate. Similarly, Figure 2c,d show the p-i-n planar and mesoscopic structures, respectively, with a p-type HTL deposited on the surface of a conductive glass substrate. The PVSCs with a planar structure are easy to fabricate, while the mesoscopic structure requires an additional step to form a mesoporous scaffold for perovskite deposition. However, it is hard to draw a conclusion regarding which structure is better, as PVSCs with both structures have achieved high efficiency exceeding 25% [17].

In addition to their high efficiency in single-junction PVSCs, the tunability of the bandgap is one of the key features that make PVSCs a promising technology. As illustrated in Figure 3, PVSCs have potential applications in indoor PV, underwater PV, tandem PV, and BIPV. With suitable absorbance, wide-bandgap (WBG) perovskites can be used in indoor PVs that are typically powered by fluorescent lamps or white light-emitting diodes [18]. WBG perovskites with a bandgap of over 1.8 eV can also be used in underwater PVs, as they can absorb blue to yellow light [19]. In addition, WBG PVSCs have colorful appearances, making them attractive candidates for BIPVs [20]. They not only respond well to outdoor and indoor light but also provide aesthetic benefits due to their semitransparency and multiple color options. Furthermore, motivated by multijunction solar cells that combine semiconducting materials with different bandgaps, both narrow-bandgap and WBG perovskites have been explored in tandem PVs [21]. This approach allows the solar cell to capture a wider solar spectrum, increasing the overall device efficiency beyond the single-junction Shockley–Queisser limit [6]. Currently, the state-of-the-art perovskite/Si tandem PVs have achieved the highest PCE of 33.2%, showing the potential of perovskite-based PV technology [22].

The formation of a high-quality perovskite film is also critical for achieving high-performance PVSCs. Figure 4a–h shows the main methods for PVSCs fabrication from small-area devices to large-area modules. For the small-area perovskite film deposition, spin coating is a commonly used method. Figure 4i shows both the one-step and two-step spin coating methods reported previously [25]. For the one-step process, the perovskite precursor materials are prepared in one solution, then spin-coated to form a film. While for the two-step process, the PbI₂ layer is deposited on the substrate first, then the ammonium salt is coated on its top and reacts with PbI₂ to form a perovskite layer. However, spin coating is only applicable for small area devices made in laboratories. To scale up the perovskite PV devices, as shown in Figure 4b–h, other methods, such as blade coating [26,27], slot-die coating [28,29], spray coating [30], dip coating [31,32], evaporation [33,34,35,36], chemical vapor deposition(CVD) coating [37], and screen-printing [38], have been applied for device fabrication. These methods can overcome the coating size limitation of spin coating, but controlling the morphology of the perovskite films is more challenging and thus requires further improvement for the development of PVSCs [27].

To sum up, it is noticeable from Figure 5 that over the past decade, the PVSCs have experienced significant development in efficiency in comparison with Si-based PV technology. This progress can be attributed to the research achievements in different perovskite materials, device architectures, and fabrication methods [3]. However, the commercialization of perovskite PV technology is still hampered by several issues, such as upscaling perovskite thin-film fabrication processing, improving the lifetime of perovskite PV devices, and reducing the toxicity of perovskite PV technology caused by the use of organic solvents and the toxic Pb during the device fabrication [39,40,41]. Additionally, the metal halide perovskite materials tend to degrade during the operation of PVSCs, and thus the toxic Pb²⁺ may enter the surrounding soil and water systems, consequently impacting human health. Therefore, this article focuses on the toxicity of PVSCs and elaborates on the environmental and health concerns arising from the use of solvents and perovskite precursor materials. Furthermore, this work overview advanced strategies for reducing the toxicity of PVSCs, specifically with regard to solvents and Pb materials. Finally, the potential and challenges of developing nontoxic PVSCs have been discussed in detail.

2. Toxicity Due to Organic Solvents

2.1. Origin and Hazards of the Toxicity

The utilization of organic solvents in the fabrication process of PVSCs is a significant source of toxicity of this technology. These solvents are crucial for dissolving and coordinating with precursor materials during the fabrication process. However, as illustrated in Figure 6a, most of the high-efficiency PVSCs reported to date have been fabricated using volatile organic solvents with high toxicity [42,43,44,45,46,47,48]. The solvents utilized in the fabrication of PVSCs can be summarized into precursor solvents, antisolvents, and solvents for the ETL/HTL, most of which are toxic and thus represent a significant obstacle to the upscaling implementation of PVSC technology.

Generally, the precursor solvents are used to dissolve perovskite materials and are subsequently deposited to form polycrystalline films upon thermal annealing. However, the most commonly used solvents, such as N,N-dimethylformamide (DMF), dimethyl sulfoxide (DMSO), and γ-butyrolactone (GBL), are known to be hazardous to human health and the environment. Although DMSO is generally regarded as less toxic, a mixture of DMF and DMSO has gained popularity in perovskite precursor solutions to enhance the morphology of perovskite films [49]. Specifically, DMF is used to improve the dissolution of PbI₂, while DMSO coordinates with Pb²⁺ [50,51]. Therefore, the use of DMF in the fabrication process is necessary, despite its well-known toxicity.

DMF is considered the most toxic substance among all the solvents used in the fabrication process of PVSCs [52]. In the European Union, the concentration of DMF in PVSCs has been restricted to <0.3% to limit its potential harm to the environment and human health [40,53,54]. More dangerously, exposure to DMF through inhalation or skin contact may pose a potential risk to workers, as it is easily absorbed by the skin and can cause adverse health effects such as liver damage, skin irritation, anxiety, palpitations, headache, flushing, nausea, and vomiting [55,56]. Therefore, it is essential to explore alternative solvents or methods that can reduce or eliminate the use of DMF in the fabrication of PVSCs.

Antisolvents are commonly used to assist the nucleation and crystallization of perovskite films by quickly removing the host solvent [57]. However, it is noteworthy that the universally used antisolvents, such as chlorobenzene (CB) and toluene, are highly toxic and volatile. CB is a hazardous chemical that is known to be toxic to aquatic organisms, and it may have long-term adverse effects on the aquatic environment. Several studies have reported that CB poses a significant risk to aquatic life due to its persistence, bioaccumulation, and toxic properties [58]. More dangerously, exposure to CB can lead to a range of health effects in humans, including exhaustion, nausea, lethargy, headache, and irritation to the upper respiratory tract and eyes [59]. Furthermore, CB has been linked to several reported carcinogenic effects, making it a serious environmental and health concern [60]. Toluene poses significant health and environmental risks. Toluene is a volatile organic compound that can easily contaminate crops, vegetation, aquatic life, and soil fertility [61]. Exposure to toluene can lead to a range of severe health effects, including neurological, respiratory, skin, and eye damage [62]. Toluene is a known neurotoxicant, and exposure to high levels can result in headaches, dizziness, confusion, and memory loss [63]. In addition, chronic exposure to toluene has been linked to developmental abnormalities and reproductive toxicity in humans and animals [61].

In the fabrication of PVSCs, ETLs, and HTLs are critical components that perform a significant role in facilitating efficient charge transfer and collection. For the preparation of ETL and HTL, the selection of solvents can greatly impact the quality and performance of these layers [64]. For the preparation of ETLs, solvents, such as CB, dichlorobenzene (DCB), and GBL, are commonly used [65,66]. For HTLs, Spiro-OMeTAD is the most widely used HTL in n-i-p structure PVSCs. However, this material can only be dissolved in solvents, such as 2-methoxyethanol (2-ME), DMSO, DCB, toluene, and isopropanol [67,68]. These solvents have also been identified as toxic and harmful to the environment and human health, with isopropanol being a potential irritant to the respiratory tract, eyes, and skin [69,70,71].

The toxicity of the aforementioned polar non-protonic solvents for perovskites has been documented by the European Chemical Agency (ECHA), as they are listed as Substances of Very High Concern (SVHC) [72,73]. Ding et al. conducted an analysis of ten commonly used solvents with regard to their toxicity to the ecosystem, workers, and the general population [74]. Based on device performance, the solvents were classified into four categories: most widely used, high industry potential, low industry potential, and high toxicity. Figure 6b presents the results of this analysis. Notably, DMSO, ethanol, and isopropyl alcohol (IPA) are shown to have the least impact on the environment and human health while maintaining the desired device performance.

Figure 6. (a) The mainly used solvents in the fabrication of PVSCs with their pictograms, where the blue balls represent nitrogen atoms, where yellow, red, white, and gray are sulfur, oxygen, hydrogen, and carbon atoms, respectively [75,76,77,78,79,80,81,82]. (b) A comprehensive evaluation of the DMF, DMSO, Ethanol, isopropyl alcohol (IPA), 2-methoxy ethanol (2-ME), DMPU, NMP, DMAC, and GBL through the comparison of their properties, including eco-toxicity factor (ECOF), workers hazard factor (WHF), general population hazard factor (GPHF), donor number (DN), and PCE. the larger the area enclosed, the better the solvent. Reproduced with permission [74]. Copyright 2022, Royal Society of Chemistry.

Figure 6. (a) The mainly used solvents in the fabrication of PVSCs with their pictograms, where the blue balls represent nitrogen atoms, where yellow, red, white, and gray are sulfur, oxygen, hydrogen, and carbon atoms, respectively [75,76,77,78,79,80,81,82]. (b) A comprehensive evaluation of the DMF, DMSO, Ethanol, isopropyl alcohol (IPA), 2-methoxy ethanol (2-ME), DMPU, NMP, DMAC, and GBL through the comparison of their properties, including eco-toxicity factor (ECOF), workers hazard factor (WHF), general population hazard factor (GPHF), donor number (DN), and PCE. the larger the area enclosed, the better the solvent. Reproduced with permission [74]. Copyright 2022, Royal Society of Chemistry.

In order to comprehensively evaluate the impact of different stages in the production process on the environment, economy, and human health, Life cycle analysis (LCA) has been commonly used in various industries, such as the coal chemical industry and biomass power generation. Although some LCA or similar studies on PVSCs have been reported, the impact of solvents on the environment and human health in perovskite production is still unresolved [83].

Figure 7 provides the potential pathways for LCA of solvents used in perovskite production. All production pathways of solvents for the manufacturing of PVSCs involve a transport section and a solvent removal section during the drying and annealing of the thin film. These sections have the potential to emit hazardous materials to the environment. The latter sections vary in different stages of production. In the laboratory and pre-industrial stages, solvents are directly released into the environment. However, in the commercialization stage, the solvents may be condensed and either incinerated or reused to enhance energy efficiency, resulting in only a small fraction directly emitted. Another pathway in the commercialization stage involves distilling the condensed solvent and incinerating the fraction that is not distilled.

Hence, it is worth noting that the solvents used in perovskite production, especially the toxic and hazardous solvents, can have significant impacts on the environment and human health. Therefore, it is crucial to comprehensively assess the environmental and human impact of the solvents used in the fabrication of high-performance PVSCs. Although the harm of some solvents has been reported, the lack of a full LCA system to evaluate the impact of solvent toxicity on the environment and human health poses a significant challenge to the sustainable development of PVSCs. LCA can provide a useful tool to assess the potential impacts of these solvents in different production stages and to identify the most environmentally friendly and sustainable pathways for perovskite production. Future research should focus on conducting more comprehensive LCAs to evaluate the impact of solvents in perovskite production, and to develop more sustainable and environmentally friendly manufacturing processes for perovskite PV industry.

2.2. Strategies to Reduce the Toxicity

2.2.1. The “Green” Solvents

In the quest for more environmentally friendly and sustainable production of PVSCs, researchers have explored various alternatives to traditional solvents used in the fabrication process. Cui et al. have provided a comprehensive summary of the green antisolvents and green solvents for HTL/ETL [84]. Therefore, in this article, we will focus on the green solvents used as precursor solvents in the fabrication of PVSCs.

As DMF is known as the most toxic chemical in the solvent system, one effective strategy to reduce the harmful impact of DMF is finding alternative solvents that are less hazardous and nontoxic. To achieve this goal, it is vitally important to understand the relationship between various solvent systems, lead salts, and post-deposition treatments. Recently, several studies have successfully shed light on some of these factors [17,18,19]. Additionally, the search for “green solvents” with lower toxicity has been a focus of research in this area. The term “green solvents” can refer to solvents that have been modified with additives or solvent mixtures that can contribute to reducing toxicity during the fabrication of PVSCs [40]. The use of such green solvents can potentially reduce the overall environmental impact and promote sustainable development in the photovoltaic industry.

One such “green solvent” is ethyl acetate (EA), which has been utilized as an antisolvent for perovskite and solvent for the HTL to optimize the perovskite/HTM interface [67]. EA has also been employed in antisolvent treatments [85,86] and precursor solvents. Zhang [87] explored EA as additive in the precursor solution, followed by an EA antisolvent treatment to facilitate the quick removal of residual solvents and expedite the MAPbI₃ perovskite crystallization process. The effects of the EA-modified precursor treatment on the perovskite film morphology were investigated and compared with the initial precursor treatment. As shown in Figure 8a, scanning electron microscopy (SEM) images revealed that the EA-modified perovskite film had a smooth, flat, and defect-free morphology with larger grain size compared to the initial precursor treatment. Furthermore, an optimized n-i-p mesoscopic FTO/TiO₂/Perovskite/Spiro-OMeTAD/Ag device was fabricated using the EA-modified precursor treatment, which achieved a PCE of 19.53% with enhanced stability. These results demonstrated the potential of EA as a “green solvent” in the fabrication of high-performance and stable PVSCs.

To reduce the use of toxic DMF, Wang developed a solvent system composed of DMSO, 2-methylpyrazine, and 1-pentanol [91]. By optimizing the preparation conditions, a well-crystallized perovskite film with such a solvent system has been achieved, and the corresponding planar n-i-p PVSCs show a PCE of 16%, indicating the potential of such mixed solvent for perovskite fabrication.

In another study, Stancu investigated the partial replacement of DMF with less toxic N-methyl-2-Pyrrolidone (NMP) and EA [88]. The PV characteristics of the devices obtained with different solvents are shown in Figure 8b. Under the same deposition parameters, the partial replacements of NMP and EA demonstrated promising results in reducing the toxicity in the fabrication process.

Several all-green solvent systems have also been reported, which are environmentally friendly and have a lower toxicity. However, the device performance achieved with these systems is relatively low compared to conventional systems. As depicted in Figure 8c,d, Cheng [92] summarized two all-green solvent systems to fabricate all-inorganic CsPbBr₃ films. Despite improving the device stability, the champion PCE is only 8.11%, which is less than the highest PCE of 10.45% among the all-inorganic CsPbBr₃ PVSCs [93]. Another all-green solvent engineering approach for organic–inorganic hybrid perovskite layers was also reported by Cao [50]. They employed a green Lewis base solvent, triethyl phosphate (TEP), to prepare (FAPbI₃)_1−x(MAPbBr₃)_x (x = 0 or 0.05) films, followed by an antisolvent treatment with the non-polar solvent, dibutyl ether (DEE), as shown in Figure 8e. The resulting FTO/SnO₂/Perovskite/Spiro-OMeTAD/Au device achieved a PCE of 18.65% (x = 0) and 20.13% (x = 0.05), which is the highest value for all-green solvent processed PVSCs.

In summary, replacing toxic solvents with green and nontoxic solvents is an important step towards sustainable and environmentally friendly PVSCs. While the use of green solvents can reduce toxicity, it has limited effectiveness as the device performance is relatively low [74]. Therefore, it is necessary to find a strategy that can keep a balance between toxicity and device efficiency.

2.2.2. Fabrication Methods without Solvent Treatment

While solvent treatment methods using less toxic alternatives have shown some promise, several techniques have been proposed that eliminate the need for solvents altogether.

Jones developed a solvent-free solid-state reaction method that employs the low-toxicity terpineol to form perovskite precursor ink [52]. As shown in Figure 9a–c, the X-ray powder diffraction results show the perovskites can be formed by such solvent-free solid-state reactions. The MAPbI₃ perovskite film fabricated by bar casting using this ink showed enhanced stability compared to its counterpart fabricated by spin-coating, as shown in Figure 9d. However, maintaining compositional control during solid-state reactions is a challenging task.

Apart from the blade-coating method, thermal evaporation has also been explored as a viable nontoxic method to fabricate PVSCs. Li et al. have reported the successful fabrication of perovskite modules using a co-evaporation method, as shown in Figure 10a. By combining perovskite composition engineering and surface/interface engineering, the resulting perovskite module with an active area of 21 cm² exhibited a PCE of 18.13%. This value is only slightly lower than the efficiency of the perovskite single cell (1 cm²) of 19%, indicating the potential of the thermal evaporation method for developing highly efficient large-area perovskite modules [94].

Kilian B et al. demonstrated the improvement of growth and optoelectronic properties for vapor-deposited [CH(NH₂)₂]_0.83Cs_0.17PbI₃ perovskite solar cells by partially substituting PbI₂ for PbCl₂ as the inorganic precursor [33]. The current density-voltage (J-V) for 0.25 cm² devices with active layers deposited with and without PbCl₂, as shown in Figure 10b, reveals a substantial increase in all J-V figures of merit for the devices with PbCl₂. In fact, the champion device PCE improves from 16.4% for the neat iodide film to 19.3% for 20% PbCl₂ substitution.

Hang L et al. have reported a Cl-containing alloy-mediated sequential vacuum evaporation approach to fabricate perovskite films, as illustrated in Figure 10c [34]. The resulting PVSCs yield a PCE of 24.42%, 23.44%, and 19.87% for 0.1, 1.0, and 14.4 square centimeters, respectively. These results demonstrate the potential of evaporation methods for fabricating efficient and large-area PVSCs.

In conclusion, the development of solvent-free methods for fabricating perovskite solar cells has shown promising results in terms of enhancing device stability and efficiency while eliminating the use of toxic solvents. Solid-state reactions and thermal evaporation methods have emerged as viable alternatives, with some studies achieving efficiencies close to those of conventional methods. However, challenges, such as maintaining compositional control and scalability, remain to be addressed. Future research should focus on improving the reproducibility and scalability of these methods while continuing to optimize device performance. Overall, the development of non-toxic and environmentally friendly methods for perovskite solar cell fabrication is a critical step towards sustainable energy production.

3. Toxicity due to Perovskite Materials: Pb

3.1. Origin and Hazards of the Toxicity

Lead (Pb) is a well-known, highly toxic element, and the use of Pb in PVSCs has raised concerns about the potential environmental and health hazards associated with its toxicity. In the PVSCs, the perovskite tends to decompose with exposure to environmental conditions, such as exposure to humidity, oxygen, elevated temperature, UV radiation, or a combination of these factors, resulting in the formation of harmful compounds containing heavy metals that can leak into the environment and impact living organisms [95].

Taking MAPbI₃ as an example, the possible degradation pathways for perovskites are illustrated by the following (1)–(7) chemical reactions. Although reactions (1) and (2) are reversible, the perovskite material will slowly break down into its constituents and additional byproducts as the presence of oxygen and UV radiation further breaks down hydroiodic acid (HI), producing water (H₂O), hydrogen (H₂), and iodine (I₂) gases as byproducts, as shown in reactions (3)–(7) [96]. Due to their ionic nature, degraded perovskites can be easily dissolved by water, potentially contaminating the soil, and groundwater, and impacting plants and other living organisms.

CH₃NH₃PbI₃(s) ↔ PbI₂(s) + CH₃NH₃I(aq)

(1)

CH₃NH₃I(aq) ↔ CH₃NH₂(aq) + HI(aq)

(2)

4HI(aq) + O₂ ↔ 2I₂ + 2H₂O(l)

(3)

2HI(aq) ↔ H₂(g) + I₂(s)

(4)

CH₃NH₃PbI₃(s) ↔ PbI₂(s) + CH₃NH₂ ↑ + HI ↑

(5)

2I⁻ ↔ I₂ + 2e⁻

(6)

3CH₃NH₃⁺ ↔ 3CH₃NH₂↑ + 3H⁺

(7)

Although the Pb content per square meter of PVSCs is only a few hundred milligrams, the potential Pb exposure associated with the commercialization of perovskite PV devices should be treated cautiously [95]. The weight percentage of Pb in CH₃NH₃PbI₃ perovskite is calculated as 33.4% by the equation of:

$$\frac{\mathrm{atomic}\mathrm{mass}\mathrm{of}\mathrm{Pb}}{\mathrm{total}\mathrm{atomic}\mathrm{mass}\mathrm{of}C{\mathrm{H}}_3\mathrm{N}{\mathrm{H}}_3\mathrm{P}\mathrm{b}{\mathrm{I}}_3}=\frac{207.2}{619.9}=33.4\mathrm{\%}$$

Given 1 m² MAPbI₃ perovskite with 500 nm thickness, the total weight of MAPbI₃ perovskite is $m=\rho V=$ 4.16 g/cm³ × 1 m² × 500 nm = 2.08 g. Therefore, the total weight of Pb in a 1 m² MAPbI₃ perovskite solar cell is 2.08 g × 33.4% = 0.69 g.

As shown in the above calculation, although there is only 0.69 g Pb in 1 m² MAPbI₃ perovskite solar cell, once the devices are damaged during rainfall, the toxic lead can leak out, then enter the soil and underground water, causing serious environmental, and health contamination.

In terms of the environment, the contamination of soil and water with Pb is permanent and can have a significant negative impact on human, animal, and plant survival [97]. Moreover, perovskite active layers in PVSCs are easily degraded and dissolved in water to generate toxic Pb²⁺, which can easily enter plants and, thus the food chain, ultimately impacting human health [95].

For humans, in the event of lead leakage, the predominant route of human exposure is through the consumption of drinking water or food contaminated by lead seepage into the soil. Therefore, there are regulatory guidelines in place for permissible lead concentrations in water, soil, and food. For instance, in China, the maximum allowable lead content in agricultural soil is limited to 250 mg/kg [98]. The gastrointestinal, respiratory, and dermal routes of exposure are the primary routes of Pb intake. The absorption, distribution, and excretion of Pb are illustrated in Figure 11, and the percentage of overall absorption is indicated for each route of intake [95]. Furthermore, the elimination of Pb from the body is difficult, as it is primarily deposited in the skeleton, with a total amount of 90% and a long half-life of 20 to 30 years [99,100]. Pb can impair the functionality of enzymes and receptors, interfere with haem activity, and gradually form insoluble lead phosphate in bones [95]. Pb can also be excreted through breast milk, making infants more susceptible to its harmful effects [101]. Symptoms associated with Pb exposure include neurological deficiencies, nausea, vomiting, abdominal pain, headaches, seizures, muscle weakness, genotoxicity, carcinogenicity, and reproductive defects [102,103,104,105]. In particular, the usage of toxic solvents, such as DMF and DMSO, increases the bioavailability of Pb, thereby increasing the risk of absorption by dermal contact and oral ingestion, making frontline workers more vulnerable to infection [106].

3.2. Strategies to Reduce the Toxicity

3.2.1. Lead-Free Perovskite

To mitigate the toxicity of Pb in PVSCs, nontoxic elements, such as Sn, Bi, Ge, Sb, Mn, and Cu have been proposed as alternative candidates to replace Pb in ABX₃ perovskite structures [107]. Although some of these elements are toxic to some degree, the current perovskite PV community generally holds the view that these elements are nontoxic alternatives in PVSCs. Hence, massive research has been devoted to these substitutions, which not only increase the diversity of perovskite species but enhance the environmental sustainability of PVSCs. Nevertheless, lead-free perovskite devices still require further improvements.

Among the various approaches to reduce the toxicity of Pb, lead-free perovskite is currently one of the hottest research areas, which has increased the diversity of perovskite materials. Figure 12a displays the elements that have been theoretically predicted to substitute Pb in perovskite structures [108]. However, the current efficiency of PVSCs based on lead-free perovskites is still below 15%, as shown in Figure 12d [107].

Among the lead-free perovskites, the Sn-based perovskites have gained much attention due to their high-performance potential [53]. The Sn-based perovskite structures are depicted in Figure 12b,c, and they have been reported to have a certified PCE of up to 14.8%, which is the highest among reported lead-free devices (Figure 12e) [111]. However, the instability caused by the easy oxidation of the Sn remains a major hurdle in the development of Sn-based perovskites. To address this issue, Hoshi [3] proposed the use of additives to improve the stability of Sn-based devices. While significant improvement was observed, the efficiencies of these devices were very low compared to Pb-based devices. Another strategy for surmounting the poor stability of Sn-based PVSCs is the use of double perovskites, which are depicted in Figure 12f. In such a structure, nontoxic monovalent (B) and trivalent (B′) metal cations neutralize the overall charge of the perovskite lattice.

Apart from Sn, Ge-based perovskites have also emerged as a promising alternative to Pb-based perovskite due to their low toxicity [112]. Among the main Ge-based perovskites, MAGeI₃, CsGeI₃, and FAGeI₃ have been studied [113]. By altering the A-site cations, the bandgap of Ge-based perovskites can be tuned to close to the conventional lead halide perovskite of 1.5 eV, as depicted in Figure 12g. Generally, Ge-based perovskites exhibit different energy structures than Pb-based ones, so optimizing suitable ETL and HTL is crucial to improve the charge extraction efficiency in the device. However, current Ge-based devices suffer from low PCE of less than 1% and stability issues similar to those observed in Sn-based perovskites, making the future of Ge-based devices appear bleak.

Other elements, such as Bismuth (Bi), antimony (Sb) [114,115], and Divalent transition metal cations (e.g., Cu²⁺, Fe²⁺, Zn²⁺) [116], have also been considered as potential candidates for Pb-free perovskites. All of these elements can be used to tune the photovoltaic properties via various routes. However, it is essential to acknowledge that while lead-free alternatives for PVSCs have been developed, these alternative elements may still pose a certain degree of toxicity and pose risks to human health and the environment. Although the perovskite community generally considers these elements to be non-toxic and eco-friendly, it is critical to conduct a comprehensive LCA of these lead-free perovskite solar cells to determine their overall environmental impact.

3.2.2. Pb Recycling Technologies

According to the International Lead Association, Pb is known to be one of the most effectively recycled materials globally [95]. Therefore, to reduce the detrimental impact on the environment, sustainable recycling technologies at the end of PVSC life have become necessary for the commercialization of PVSCs. The main methods for dealing with these issues include selective Pb extraction recycling and in situ Pb recycling processes, as summarized in Figure 13a,b.

The selective Pb recycling process involves the dissolution of the perovskite materials in solvents, followed by the subsequent collection of lead from the resulting solution. The collected Pb can be reused in the fabrication of PVSCs, but a small loss in PCE may occur. The collection step mainly includes electrochemical deposition, precipitation, and adsorption processes. As illustrated in Figure 13c, electrochemical deposition involves the dissolution of perovskites in deep eutectic solvents (DES) [118] or molten LiCl-KCl(eutectic composition) [117] to form a mixed solvent. Subsequently, the Pb and I₂ are deposited at the cathode and graphite anode, respectively. Although a recovery rate of 100% can be achieved by electrochemical deposition [118], it is challenging to control the complex electrodeposition conditions required to obtain the desired Pb layer. Moreover, additional steps are needed to recycle Pb to PbI₂, which serves as a precursor for PVSCs.

In order to achieve a more convenient recycling of Pb, the precipitation of PbI₂ after dissolving the perovskite layer has been implemented. Polar aprotic solvents, such as DMF, GBL, and DMSO have been reported to be widely used to recycle PbI₂ in degraded PVSCs [109]. However, since the dissolution of HTMs also occurs in the first step, significant efforts have been made to improve the PCE of the recycled PVSCs using this method. Furthermore, in order to address the toxicity issues associated with the aforementioned solvents, less hazardous solvents have also been explored [119].

As a sustainable and environmentally friendly recycling of PbI₂ from PVSCs, using adsorbents that offers a high recovery rate and purity has been studied. Hydroxyapatite (HAP, Ca₁₀(PO₄)₆(OH)₂) has been used as an adsorbent in the recycling process, as it exhibits a high adsorption efficiency of 99.99% from Pb-containing solvents such as DMF. More recently, a Fe-decorated HAP (HAP/Fe) was employed for complete Pb management, achieving a 99.99% adsorption efficiency of dissolved Pb²⁺ ions in DMF, as shown in Figure 13d. Recycled PVSCs using HAP/Fe showed a PCE of 16.04%, which was similar to the PCE of fresh PVSCs at 16.65%.

Another selective recovery method that offers a nontoxic and simple approach is ion exchange. Wang [53] reported a “one-key-reset” strategy for recycling PVSCs, as depicted in Figure 14a. The PCE of recycled PVSCs was 20.08%, which was only slightly lower than the PCE of fresh PVSCs at 20.65%, even after being recycled twice. This method offers a sustainable approach to the recycling of PVSCs and paves the way for the commercialization of this technology.

As depicted in Figure 13b, in situ Pb recycling technologies aim to restore the degraded perovskite layer to its initial state. This method has the potential to reduce production costs by reusing materials. One rejuvenation strategy in hydroiodic involves immersing the entire PVSC in a methyl ammonium iodide (MAI) solution. As shown in Figure 14b, Huang et al. [121] prepared an ITO/ZnO/perovskite/Spiro-OMe TAD/Au device that was rejuvenated after 60 days using an MAI-assisted rejuvenation approach. The results showed that after the second rejuvenation process, the device had a PCE of 11.4% compared to 11.6% in its initial state.

Other in situ Pb recycling technologies have also been proposed [122,123]. These in situ recycling processes can minimize refabrication costs and time while reducing toxic Pb loss. However, as the number of rejuvenation cycles increases, the rejuvenate effect gradually decreases, which limits the ability to rejuvenate PVSCs. Moreover, in practical commercial scenarios, it is important to consider that perovskite materials are susceptible to degradation and potential leakage when exposed to outdoor environments. This poses a significant challenge for the implementation of in situ Pb recycling technologies, as the effectiveness of these methods can be greatly diminished by such degradation and leakage. Therefore, it is necessary to develop and employ protective measures and strategies to mitigate these issues and ensure the practicality and viability of in situ Pb recycling in commercial settings.

3.2.3. Device Encapsulation to Prevent the Pb Leakage

Encapsulation is considered the most promising strategy for mitigating the risk of Pb and other toxic material leakage caused by the degradation of PVSCs. External encapsulation involves the use of a layer of physical or chemical barriers, such as lead-trapping materials or chemical barriers, to surround the PVSC from the top, bottom, and edges, as illustrated in Figure 15a. This approach not only shields the PVSC against moisture and UV radiation and prevents volatile perovskite decomposition products but also can potentially mitigate Pb leakage from high-performance lead-based PVSCs [124].

Given the high toxicity of Pb and its potential environmental and health risks, preventing Pb leakage from PV devices is of utmost importance. While physical encapsulation can provide a barrier to prevent volatile decomposition products and protect the PVSCs against moisture and UV radiation, chemical encapsulation can prevent potential Pb leakage from PVSCs more efficiently [126]. Therefore, researchers have explored the use of chemical additives that can capture leaked Pb from the perovskite layer under various environmental conditions.

Two strategies have been pursued to achieve Pb capture: modification of internal functional layers or incorporation of an exterior encapsulant with Pb-chelating materials, as shown in Figure 15b,c. To be effective, an encapsulation layer with a high Pb adsorption rate and capacity is essential. In addition, from a commercial standpoint, encapsulation materials that are abundant, low-cost, and easily processable are desirable.

Materials of epoxy, SnO₂, butyl rubber, and dyads are commonly used as carrier transport layers to prevent Pb leakage. Chen [127] has integrated cation-exchange resin into carbon electrodes to prevent lead leakage. Xu [128] has effectively prevented Pb leakage by employing Pb-absorbing materials through chemical reactions between phosphonic acid groups and lead ions. Qi [129] found that encapsulating devices with epoxy resin reduces Pb leakage without affecting device performance. These studies confirm the effectiveness of chemical encapsulation in reducing Pb leakage.

3.2.4. Immobilization of the Pb

As depicted in Figure 15d,e, the immobilization of Pb refers to the incorporation of special materials into the perovskite layer or surface, which interact with Pb²⁺ ions during perovskite degradation. These materials act as additives or interlayer and surface passivation layers, respectively. Compared to device encapsulation, immobilization of Pb is not as effective in preventing Pb leakage from the device. However, materials that immobilize Pb can also improve the crystallinity of the perovskite layer and passivate defects in the perovskite film [109]. Therefore, various materials with strong electron-donating functional groups have been introduced into the perovskite layer, as summarized in Figure 15f. These functional groups include carbonyl, thiol, sulfonate, phosphate groups, etc.

Niu et al. [130] introduced a polymerizable acrylamide (AAm) monomer additive into a perovskite precursor solution, resulting in more than a 50% reduction in lead with an improved PCE of 10%. Bi et al. [131] demonstrated a PVSC modified with 3-mercaptopropyl-tri-ethoxy-silane that exhibited a PCE of 22.18%, compared to 19.87% of the initial device, while the lead leakage concentration was reduced threefold. In 2021, Li et al. [132] suggested a 4-[(trifluoromethyl) sulfonyl] aniline (4TA) material with an aniline group. When the device was immersed in pH 5.6 deionized water for 60 s, the Pb²⁺ concentration of the modified 4TA device was significantly lower than that of the control device.

4. Conclusions and Perspective

The commercialization of PVSCs has been impeded by several obstacles, among which the issue of toxicity is a major concern. Despite the rapid development of PVSCs, the toxicity of Pb-based PVSCs has not been effectively resolved, which makes it impossible to scale up their manufacture. Consequently, considerable research effort has been devoted to evaluating and resolving the toxicity of Pb-based PVSCs during their production and application.

One of the primary sources of toxicity is the solvents used in the fabrication process. While the hazards from solvents are relatively less than the toxicity of Pb, the upscale production of PVSCs will result in a significant consumption of DMF, which has adverse effects on both the environment and human health. Therefore, a comprehensive life cycle assessment is necessary to fully understand the hazards to the environment and human health for the commercialization of PVSCs. On the other hand, as DMF is the most toxic material in the precursor solvents, which can easily be absorbed through the skin and cause many adverse health effects, significant efforts have been made to replace DMF partly or completely. While the role of DMF is vital in the solvent treatment, it is nearly impossible to replace it completely. Nevertheless, solvent-free fabrication methods are considered promising but are still under research, as some PVSCs made using this method do not exhibit decent PV performance.

Another major source of toxicity is the Pb materials used in high-performance Pb-based PVSCs. The usage and leakage of Pb have dangerous effects on humans and other living creatures due to its property of easily dissolving in water and being hard to remove from the body. To mitigate the toxicity of Pb-based PVSCs, significant efforts have been made in the development of lead-free PVSCs, Pb recycling technologies, device encapsulation, and immobilization of Pb. Although the development of lead-free PVSCs is the earliest, their PCEs are still below 15%, and their stability remains a significant issue that needs to be resolved. Therefore, the assessment and reduction of the toxicity in Pb-based PVSCs are crucial for their commercialization.

Among the various methods to reduce toxicity, device encapsulation is considered the most efficient, using physical or chemical barriers to prevent the Pb leakage. Recycling technologies are essential for sustainable applications with reduced environmental effects. However, the current mainstream recycling method still involves the use of toxic chemicals, necessitating the establishment of an overall environmental impact assessment and the development of eco-friendly Pb recycling processes. In terms of Pb immobilization strategies, even though they are less effective than encapsulation, their additional advantages, such as enhanced crystallinity and passivation of defects, make them promising for further research.

In conclusion, significant progress has been made in the field of green PVSCs, and it is anticipated that research on nontoxic PVSCs will soon have a breakthrough. We look forward to further development of lead-free perovskite materials and PVSCs that meet environmental requirements. We expect continuous improvements in PVSCs to meet the needs of commercial development.