Iodide manipulation using zinc additives for efficient perovskite solar minimodules

Perovskite modules lag behind silicon module efficiencies despite small-area perovskite cells exceeding 26%. Scaling requires uniform, high-quality photoactive layers and minimal cell-to-module derate. A key barrier is nonuniformity originating from ink oxidation in ambient processing, especially iodide oxidation to I2, and increased iodide interstitials when employing 2D-iodide salts (e.g., PEAI, DDAI) for passivation. Iodide interstitials introduce deep traps that degrade Voc, FF, and stability. This work targets these issues by adding zinc salts with organic anions to perovskite inks: CF3SOO− reduces I2 to I−, and Zn2+ sequesters excess I− by forming Zn–amine complexes, aiming to suppress iodide interstitials, improve film uniformity over large areas, and reduce cell-to-module efficiency losses.

Prior efforts to improve mini-modules (20–50 cm2 aperture) focused on perovskite phase stability, HTL/perovskite and perovskite/ETL interface engineering, and charge transport optimization, yielding promising gains but leaving a typical 15–20% relative cell-to-module efficiency loss. These studies generally did not address iodide interstitial defects or ink oxidation-induced nonuniformity, which can lower Isc and FF and contribute to Voc losses. The present study specifically tackles iodide oxidation and interstitial formation as drivers of nonuniformity and module derate.

• Additives: Evaluated zinc salts with organic anions (zinc formate, acetate, trifluoroacetate, trifluoromethane sulfinate Zn(OOSCF3)2, and trifluoromethane sulfonate) versus zinc halides (ZnCl2, ZnBr2, ZnI2). Target composition FA0.3MA0.7PbI3 processed by blade coating.
• Small-area device fabrication: ITO/PTAA/Perovskite/C60/BCP/Cu architecture. PTAA (3.3 mg mL−1 in toluene) bladed (150 μm gap, 20 mm s−1). Perovskite ink (1.4 M MA0.7FA0.3PbI3 in 2-ME/DMSO with additives: 23–25 mol% DMSO, 0.63 mol% BHC, 0.37 mol% 4-F-PEAI, 0.25 mol% choline chloride, 0.026 mol% LP, 1.6×10−5 mol% MAH2PO2, 0.19 mol% n-DDAI, and 0.14–0.55 mol% zinc salts relative to Pb2+) bladed (250 μm gap, 20 mm s−1) with N2 knife (20 PSI). Anneal 120 °C for 5.5 min in air. Evaporate C60 (30 nm), BCP (6 nm), and Cu (150 nm). Active area 0.08 cm2.
• Minimodule fabrication: Same stack; 20 series-connected sub-cells; P2/P3 laser scribes after 80 and 120 nm Cu; total scribe width 0.4–0.6 mm; geometric fill factor 92–94%. Aperture areas 78/84/108 cm2 (module dimensions: 6.5×12, 7.0×12, 9.0×12 cm, sub-cell width 6.5 mm). PDMS AR coating; encapsulation with PIB edge seal, optional epoxy; overall substrate 15×15 cm.
• Additive screening and optimization: Varied molar fraction of zinc salts relative to Pb (typ. 0.14–0.55 mol%). Identified Zn(OOSCF3)2 as optimal; detailed concentration series around 0.28%.
• Characterization: • JV under AM1.5G (100 mW cm−2) using Oriel Sol3A; Keithley 2400; scan rate 0.1 V s−1. EQE for spectral response. • Photoluminescence (integrating sphere) and time-resolved PL (405 nm excitation) to assess non-radiative recombination and lifetimes; PL quantum yield comparison. • Electroluminescence at ~24 mA cm−2 using integrating sphere; inferred Voc change from EL intensity. • Morphology by SEM (top and cross-section); crystallinity by XRD (Rigaku SmartLab, Cu Kα). • Defect analysis by thermal admittance spectroscopy (TAS) for trap DOS and drive-level capacitance profiling (DLCP) for spatial trap density, focusing on iodide-related deep traps (~0.35 eV). • Chemical mechanism probes: I2 reduction tests by mixing I2 (1.9 M in 2-ME) with Zn(OOSCF3)2 and/or MAI/FAI (7:3), heating at 120 °C; UV–vis to monitor I2 bleaching; proposed redox with CF3SOO− producing SO2/CF3− and reducing I2 to I−. FTIR to confirm Zn–amine complex formation (deprotonation of FA/MA); precipitation of Zn–amine complex observed at high Zn dosing (2.5 mol% to Pb). Synthesized Zn–amine complex from FAI/MAI/Zn(OOSCF3)2 in 2-ME; XRD signature near 26°.
• Stability testing: ISOS-L-1-like light soaking in air (RH 50±20%, device temp 55±5 °C) at Voc under 1-sun equivalent LED; encapsulated devices. Module shelf storage (>6 months) and continuous 1-sun for 24 h. NREL certification: stabilized max-power tracking for 360 s, ASTM G173, device temperature 25.5±2.0 °C, irradiance 1000 W m−2; aperture area 79.67 cm2±3%.

• Additive screening (small cells, active area 0.08 cm2): Control average PCE 22.22±0.69%. With zinc additives (optimized mol% vs Pb): • Zn(OOCH)2 (0.42%): 22.59±0.44% • Zn(OOCCH3)2 (0.28%): 22.03±0.48% • Zn(OOCCF3)2 (0.42%): 22.28±0.86% • Zn(OOSCF3)2 (0.28%): 23.61±0.30% (highest) • Zn(OO2SCF3)2 (0.42%): 22.83±0.40% Zinc halides (ZnCl2, ZnBr2, ZnI2) yielded efficiencies similar to control (~22.1–22.6%), underscoring the critical role of CF3SOO− organic anion.
• Optimized Zn(OOSCF3)2 concentration (0.28% vs Pb): Average Voc increased from 1.16±0.01 V to 1.18±0.00 V; FF from 0.80±0.02 to 0.82±0.01; Jsc showed slight improvement (supported by JV and EQE). Series resistance decreased (~1.1× lower) and shunt resistance increased (~1.4× higher) in champion devices.
• Photophysics: Steady-state PL intensity ~1.6× higher; PLQY increased from 0.46% to 0.66% (~1.4×); PL lifetime extended from 0.7 μs (control) to 2.0 μs (0.28% Zn(OOSCF3)2). PL peak blue-shifted by ~6 nm, indicating reduced band-tail states. EL intensity ~2.5× higher at ~24 mA cm−2; inferred Voc gain ~23.8 mV from EL.
• Morphology and structure: SEM revealed slightly smaller grains with Zn additive; XRD showed similar crystallinity to control.
• Defects: TAS indicated reduced iodide-interstitial-related trap DOS; DLCP showed lower deep trap density throughout films (trap depth ~0.35 eV), aligning with improved Voc.
• Chemical mechanism: CF3SOO− rapidly bleached I2 in solution, especially with heat (120 °C) and presence of water/FAI+MAI, consistent with spontaneous SO2/I2 redox. Zn2+ formed precipitated Zn–amine complexes (observed at 2.5 mol%) that sequester excess I−; FTIR confirmed deprotonation of FA/MA; XRD of synthesized Zn–amine complex showed strong peak near 26°.
• Stability (ISOS-L-1-like, open-circuit): Devices with Zn(OOSCF3)2 retained 81.2% of initial PCE after 1078 h; control retained 72.8% after 1039 h.
• Minimodules (blade-coated, Zn(OOSCF3)2 0.14–0.28%): Highly uniform films over 78–108 cm2. Champion aperture-area PCEs: 20.18% (78 cm2; Jsc 21.45 mA cm−2; FF 0.801), 20.18% (84 cm2; Jsc 21.99 mA cm−2; FF 0.796), 20.23% (108 cm2; Jsc 22.04 mA cm−2; FF 0.772). Averages: 19.47±0.50% (78 cm2), 19.55±0.47% (84 cm2), 19.21±0.51% (108 cm2). NREL-certified stabilized efficiency 19.60% at 79.67 cm2 aperture; stabilized active-area efficiency 20.66%. Champion stabilized aperture ~19.60% and champion JV ~20.20% (active area 21.24%) represent record efficiencies near 100 cm2. Shelf stability: >91% PCE retention after >6 months; ~90% after 24 h under 1-sun.

The study directly addresses module-scale nonuniformity and performance loss stemming from iodide oxidation and iodide interstitial defects. The CF3SOO− anion in Zn(OOSCF3)2 acts as a reducing agent to convert I2 to I− in inks, preventing oxidative degradation pathways known to accelerate perovskite deterioration. Concurrently, Zn2+ forms Zn–amine complexes with FA/MA, precipitating and removing excess I− from solution, thereby suppressing the formation of iodide interstitials—deep traps associated with nonradiative recombination and Voc loss. Despite slightly reduced grain size (which could otherwise increase recombination), passivation from the zinc additive yields significantly enhanced PL/EL, longer carrier lifetimes, increased Voc and FF, and reduced trap densities across the film depth. These defect and ink chemistry controls translate into highly uniform blade-coated large-area films and reproducible minimodule performance with minimal derate across larger apertures (78–108 cm2). The substantially improved shunt and series resistances, stability gains under light soaking, and record NREL-certified module efficiencies underscore the efficacy and scalability of the approach. The weaker effect of ZnX2 compared to Zn(OOSCF3)2 highlights the necessity of the CF3SOO− chemistry rather than Zn2+ alone.

A family of zinc additives was identified to improve perovskite device efficiency, stability, and large-area uniformity, with Zn(OOSCF3)2 delivering the strongest benefits. CF3SOO− reduces molecular iodine generated during processing/aging, while Zn2+ precipitates excess iodide via Zn–amine complex formation, collectively lowering iodide interstitial concentrations and deep trap densities. This chemistry enhances PL/EL, Voc, FF, and stability, and enables uniform blade-coated films that yield record NREL-certified minimodule efficiencies around 80–100 cm2 aperture (stabilized 19.60%). Future work could explore broader perovskite compositions and additive loadings, integration with diverse 2D passivation schemes while managing iodide balance, deeper mechanistic quantification of interstitial suppression, long-term outdoor and thermal cycling stability, and scale-up to full-size modules with further derate minimization.

• Stability assessments, while extended (∼1000 h ISOS-L-1-like and 6-month shelf), do not represent long-term outdoor operation or rigorous IEC testing (e.g., damp heat, thermal cycling, UV exposure), leaving questions about lifetime and degradation modes under field conditions.
• The approach was demonstrated on FA0.3MA0.7PbI3 with a specific additive suite; generality across other perovskite compositions (e.g., Cs-containing, Br-mixed, Sn–Pb) and device stacks requires validation.
• Grain sizes were reduced by Zn2+ addition; although passivation compensates, trade-offs with charge transport in thicker absorbers or different architectures may emerge.
• The extent of interstitial iodide reduction is inferred (TAS, DLCP) rather than directly quantified at the atomic level; in situ/operando measurements could strengthen mechanistic claims.
• Additive–ink interactions and potential impacts on long-term interface chemistry (e.g., with transport layers, encapsulants) were not exhaustively studied.