Bioinspired “cage traps” for closed-loop lead management of perovskite solar cells under real-world contamination assessment

Perovskite solar cells (PSCs) are promising for low-carbon energy due to high efficiency and scalable fabrication, but intrinsic instability of volatile organic cations and soft lattices leads to degradation and formation of toxic Pb-containing compounds (PbI2, Pb, PbO) under moisture, light, and heat. This raises environmental sustainability and health concerns due to potential Pb leakage. Existing encapsulation strategies mitigate but do not eliminate Pb diffusion, especially upon device breakage. There is also a lack of realistic, ecosystem-based assessments of Pb leakage from in-service and end-of-life PSCs. This study aims to develop a bioinspired, multifunctional capture-and-shield layer to minimize Pb leakage under real-world stressors, elucidate its synergistic chemical and physical adsorption mechanisms, and establish a closed-loop Pb management route including adsorption and recycling, validated in actual environmental media (Yellow River water and soil).

Prior efforts to mitigate Pb leakage include physical encapsulants such as polyurethane, polyisobutylene, graphene, and Al2O3 (hot-press or ALD), which can reduce ingress but have limited Pb capture capacity when devices crack, and may increase fabrication complexity and cost. Chemical adsorption approaches using functional materials, e.g., semi-transparent Pb adsorbents on the illumination side, are constrained by transparency-thickness tradeoffs that limit adsorption capacity and durability. Li et al. integrated a polymer mixture of cation exchange resin (CER) and UV resin, achieving ~90% Pb sequestration via sulfonic acid–Pb2+ exchange, but raised concerns about secondary pollution from solid CER waste and lack of sustainable end-of-life strategies. Overall, conventional laboratory leakage tests oversimplify conditions by omitting complex organics, heavy metals, and microorganisms, potentially misrepresenting real-world environmental risks. A need exists for robust on-device solutions with high capture capacity on the backside (where leaked Pb tends to flow) and for standardized real-world leakage assessments and closed-loop recycling strategies.

Design and synthesis: Inspired by spider web chemistry and structure, a mesoporous amino-grafted carbon net (BCT) was synthesized by grafting ethylenediamine (EDA) onto an acid-activated mesoporous matrix (MM) to form functional groups (–CONH–R–NH2). Three stages: (I) initial MM; (II) ultrasound-assisted structural separation (90 min) to increase pore formation and dispersion; (III) surface chemical reconfiguration with EDA at 40 °C for 150 min, followed by filtration, washing (ethanol:water 1:1), and vacuum drying. Characterization: FTIR confirmed –COOH in MM and successful EDA grafting (–NH2, –CONH signatures). XPS showed N1s components for –NH2 and –CONH and binding energy shifts upon Pb interaction. DFT calculations (VASP, PBE-GGA, PAW, 400 eV cutoff, 4×4×1 k-mesh) computed Pb binding energies for –COOH, –NH2, –CONH, and –CONH–R–NH2. Morphology and porosity: SEM, 3D optical profilometry, and BET assessed pore evolution, roughness, and surface area; contact angles assessed wettability; capillary adsorption modeled via capillary rise and Young’s equation. Adsorption kinetics: Batch adsorption without stirring using 0.3 g adsorbent in 60 mL perovskite-derived Pb solution (167 ppm). ICP-OES measured Pb concentration over 1–180 min. Langmuir isotherm and pseudo-second-order kinetic models quantified equilibrium capacity (qe), distribution coefficient (Ka), and rate constants (K2). Pb leakage tests: Devices (16.1 mm × 16.1 mm) with structures ITO/SnO2/FA-based perovskite/Spiro-OMeTAD/Au were prepared by a two-step perovskite deposition. Encapsulation variants: unencapsulated (U-PSC), glass encapsulated (G-PSC), BCT-encapsulated on backside (BG-PSC, BCT mixed with conductive carbon paste 2:1 and hot-pressed onto glass). Hail impact simulated per ASTM E1038; subsequent tests included heavy rainfall flushing (immersion in DI water up to 6 h), acid rain (HNO3 pH 4.2, 10 mL/h for 1 h at 45°), alkaline conditions (pH 10 and 11), and complete breakage scenarios. Real-world assessments: End-of-life devices tested in Yellow River water (30 mL, 6 h) and soils (buried; sampling days 1–30 at 5 cm depth). Soil Pb contamination evaluated via Igeo, Pi, and PN indices. Device performance and stability: J–V and EQE measured for U-PSC and BG-PSC; ISOS-like stability under AM1.5G at 60 ± 5 °C, and storage at 25 °C under 50% and 85% RH. Closed-loop Pb management: Four steps—(I) Pb precipitation, (II) Pb adsorption by BCT, (III) Pb desorption via protonation in HNO3 (pH 1), yielding Pb(NO3)2, and (IV) PbI2 recovery by adding NaI and reusing as precursor. Recovered PbI2 characterized by XRD, Raman, FTIR, ICP-OES for Pb/I stoichiometry; PSCs refabricated and benchmarked versus commercial 99.99% PbI2.

Mechanism and materials: EDA grafting formed –CONH–R–NH2 groups; FTIR and XPS shifts indicated strong chelation with Pb2+. DFT binding energies (eV): –COOH 3.17, –NH2 2.87, –CONH 3.04, –CONH–R–NH2 3.30, confirming enhanced multi-site chelation. Morphology and adsorption physics: Pores increased from ~13 to ~36 to ~52 per 100 µm² across stages I–III; BET surface area increased to 108.44 m² g⁻¹ (BCT), more than double MM. Contact angle decreased (BCT 38.5° vs MM 58.9°), favoring capillary adsorption. Adsorption performance: In batch tests, BCT captured 88.24% Pb within 5 min and >98.60% at equilibrium; MM reached 42.57% at 180 min. Equilibrium capacities qe: BCT 32.934 mg g⁻¹ vs MM 14.220 mg g⁻¹. Distribution coefficient Ka: BCT 14134.76 mL g⁻¹ vs MM 148.27 mL g⁻¹. Pseudo-second-order fits had R² > 0.99; calculated qe (mg g⁻¹): BCT 32.48, MM 14.13; rate constants K2: BCT 0.0690 g⁻¹ (mg/min)⁻¹ vs MM 0.0306. Stage contributions: Stage I (0–5 min, chemical) adsorption 29.44 mg g⁻¹ for BCT vs 9.60 for MM; Stage II (5–120 min, physical) lower increments (BCT 2.84; MM 4.30 mg g⁻¹). Overall adsorption efficiency from perovskite-degraded water: BCT reduced Pb from >100 ppm to near 0 ppm; MM to ~20 ppm; EDS showed higher Pb at% in BCT (0.4 At%) vs MM (0.1 At%). Device-level Pb leakage mitigation: Heavy rainfall flushing after hail: U-PSC reached ~3.45 ppm in 120 min; G-PSC 1.4 ppm after 6 h (>40% leakage); BG-PSC only 4.8 ppb (Pb sequestration 99.86%, below WHO drinking water guideline). Complete breakage scenario: leaked Pb (ppm) after 6 h—U-PSC 7.84, G-PSC 5.57, resin-encapsulated R-PSC 4.03, BG-PSC 0.876; corresponding sequestration efficiencies: 0.00%, 26.79%, 48.59%, and 88.83%. Acid rain (pH 4.2): U-PSC 11.6 ppm, G-PSC 5.09 ppm, BG-PSC 0.965 ppm (91.68% sequestration). Alkaline water: pH 10—U-PSC 7.82 ppm, G-PSC 2.58 ppm, BG-PSC 0.523 ppm (93.31%); pH 11—U-PSC 6.27 ppm, G-PSC 2.17 ppm (34.61% leakage rate), BG-PSC 0.461 ppm (92.64%). Real-world media: Yellow River water leakage rates normalized to U-PSC = 100%: G-PSC 61.36%, BG-PSC 8.40%. In DI water: G-PSC 40.56%, BG-PSC 0.14%. Increased leakage in river water vs DI: BG-PSC +8.26%, G-PSC +20.82%. Soil contamination indices (Yellow River soil): For BG-PSC, average Igeo −0.62977 (No contamination), Pi 0.970456 (No contamination), PN 0.9969 (Still clean). U-PSC and G-PSC showed higher contamination categories (e.g., U-PSC PN 2.1256, moderate to severe). Device performance and stability: BCT encapsulation did not degrade initial PCE or EQE; distributions overlapped for U-PSC and BG-PSC. Under AM1.5G, 60 ± 5 °C for 360 h, normalized PCE retained: BG-PSC 81.49% vs U-PSC 43.65% and G-PSC 36.62%. Storage at 25 °C, 50% RH for 1000 h: BG-PSC retained 92% vs U-PSC 56%, G-PSC 62%. At 25 °C, 85% RH for 600 h: BG-PSC retained 98.40% vs U-PSC 47.92%, G-PSC 79.22%. BCT showed measurable moisture adsorption (mass increase 560.5 to 566.9 mg at 50% RH, 24 h). Closed-loop Pb management and recycling: Adsorption rate up to 99.77% in steps I–II; desorption via protonation in HNO3 released captured Pb as Pb(NO3)2; PbI2 recovered with 96.07% yield. Recovered PbI2 matched commercial 99.99% PbI2 by XRD and Raman; FTIR showed no extra impurity peaks. ICP-OES indicated minor stoichiometric deviation (relative Pb 103%, I 100.3%; I:Pb = 1.93449). PSCs fabricated with recycled PbI2 achieved PCE 22.08% (Voc 1.11 V, Jsc 25.87 mA cm⁻², FF 76.89%) vs commercial PbI2 PCE 22.37% (Voc 1.13 V, Jsc 25.66 mA cm⁻², FF 77.03%); average PCEs over 10 devices: 20.93% (recycled) vs 21.02% (fresh); standard difference factor 0.42%.

The bioinspired BCT addresses the central challenge of preventing Pb leakage from PSCs while maintaining device performance and enhancing stability. Mechanistically, multi-site chelation (–CONH, –NH2) with high binding energies, together with increased porosity, wettability, and capillary-driven uptake, enables rapid and high-capacity Pb adsorption. Device-level tests show that BCT on the backside is particularly effective under realistic damage pathways where leaked Pb migrates downward, outperforming front-side encapsulation approaches even under catastrophic breakage and extreme weather (hail, acid rain, alkaline water). Real-world evaluations in Yellow River water and soils demonstrate that laboratory-only tests underestimate environmental complexity; despite some reduction in adsorption efficacy in river water, BG-PSC maintained low leakage and reduced soil contamination to background/no-contamination levels. The encapsulation is compatible with PSC operation and confers substantial stability benefits by adsorbing and isolating moisture. The closed-loop management, including desorption via protonation and PbI2 recovery, demonstrates a practical pathway to recycle Pb with high purity and negligible performance penalty upon reuse, aligning environmental safety with circular manufacturing.

This work introduces a spider web-inspired mesoporous amino-grafted carbon net (BCT) that synergistically combines chemical chelation and physical capillary adsorption to capture Pb and shield PSCs from environmental ingress. The material delivers rapid, high-capacity Pb adsorption and exceptional device-level Pb sequestration across severe stress scenarios, with validated performance in realistic water and soil environments. BCT encapsulation maintains photovoltaic efficiency and substantially enhances operational stability under light and humidity. A closed-loop Pb management process enables desorption and high-yield recovery of PbI2, which is reusable without compromising device performance. These advances provide a robust framework for greener, sustainable industrialization of PSC technology and establish a benchmark for real-world Pb leakage assessment and mitigation.

Adsorption efficacy decreases in complex natural waters (e.g., Yellow River water), likely due to organics, heavy metals, microorganisms, and silt, leading to modestly higher Pb leakage than in DI water. The recycled PbI2 exhibits a slight iodine deficiency (I:Pb ≈ 1.93449), indicating minor stoichiometric deviation that may require rectification. Tests focus on specific environmental conditions and device architectures; broader deployment scenarios may involve additional challenges not covered here.