p-Phenylenediaminium iodide capping agent enabled self-healing perovskite solar cell

Organic–inorganic hybrid perovskite solar cells (PSCs) offer excellent optoelectronic properties but suffer from instability under ambient conditions, hindering practical deployment. Additionally, highly efficient devices often rely on costly, doped organic hole-transport layers (HTLs), such as spiro-OMeTAD with Li-TFSI and TBP, which can compromise stability and increase cost. Inorganic CuSCN is a promising low-cost HTM with higher hole mobility and suitable energy levels, but typical deposition solvents can attack perovskite and result in incomplete coverage and textured grains, leading to weak perovskite/HTL interactions. Interface engineering with 2D perovskite layers on 3D films has been shown to passivate defects, improve interfacial energetics, and enhance device stability. This work investigates whether using p-Phenylenediaminium iodide (PDAI) to in-situ form an ultra-thin 2D (PDA)₂PbI₄ layer atop (FAPbI₃)₀.₈₅(MAPbBr₃)₀.₁₅ can simultaneously enhance performance, stability, reduce hysteresis, and impart self-healing in PSCs using CuSCN as a low-cost HTL.

Prior studies have demonstrated the benefits of forming 2D perovskite interface layers on 3D perovskites to passivate defects and enhance stability and performance. For example, alkylammonium halides forming 2D layers improved moisture tolerance; engineering (HOOC(CH₂)₃NH₃)₂PbI₄ at the MAPbI₃/spiro-OMeTAD interface yielded ultra-stable devices with 14.6% PCE. In-situ growth of PEA₂PbI₄ on 3D perovskite increased Fermi-level splitting under illumination, boosting Voc and PCE. Benzylamine treatments on FAPbI₃ induced 2D layers, delivering 19.2% efficiency and improved stability at ~50% RH. Phenyl groups and alkyl chains have also served as effective passivating ligands. Compared with a related report using 5-AVAI at a (FAPbI₃)₀.₈₈(CsPbBr₃)₀.₁₂/CuSCN interface (champion PCE 16.75%, Jsc 21.93 mA cm⁻², Voc 1.068 V), the present study achieves a higher Voc, attributed to composition and interface effects. These works collectively motivate exploring PDAI-derived 2D layers to improve the perovskite/CuSCN interface and device stability.

Materials: FAI was synthesized by reacting formamidinium acetate with hydroiodic acid; MABr was synthesized from methylammonium solution and hydrobromic acid; PDAI was obtained by adding hydroiodic acid to p-phenylenediamine at 0 °C, followed by purification. CuSCN HTM powder was synthesized via reduction of Cu(II) to Cu(I) with Na₂S₂O₃, followed by precipitation with KSCN, washing, and drying. Device fabrication: FTO substrates were patterned and cleaned. A compact TiO₂ blocking layer (bl-TiO₂) was spin-coated from TTIP/ethanol and sintered at 500 °C (1 h). A mesoporous TiO₂ (mp-TiO₂) layer was spin-coated from TiO₂ paste diluted in ethanol and sintered at 500 °C (30 min). The perovskite precursor (FAI 1 M, MABr 0.2 M, PbI₂ 1.1 M, PbBr₂ 0.2 M in DMF:DMSO 4:1 v/v) was spin-coated (1000 rpm 10 s, then 3000 rpm 40 s), with 500 µL ethyl acetate dripped before the end of spinning, and annealed at 130 °C for 30 min. PDAI in 2-propanol (3, 5, 7, 10, or 15 mg mL⁻¹) was spin-coated onto (FAPbI₃)₀.₈₅(MAPbBr₃)₀.₁₅ at 3000 rpm for 20 s and annealed at 100 °C for 10 min to in-situ form a 2D (PDA)₂PbI₄ capping/interface layer. CuSCN HTL was deposited by spin-coating 40 mg mL⁻¹ CuSCN in dipropyl sulfide/acetonitrile (DPS/ACN) at 5000 rpm for 20 s, then annealed at 60 °C for 15 min. Au (50 nm) was thermally evaporated as the top electrode. Fabrication was conducted under ambient conditions except for Au evaporation. Architecture: FTO/bl-TiO₂/mp-TiO₂/(FAPbI₃)₀.₈₅(MAPbBr₃)₀.₁₅/ultra-thin 2D (PDA)₂PbI₄/CuSCN/Au. Devices: Sets A₁–F₁ used CuSCN HTL; A₂–F₂ had no CuSCN. A₁: control without PDAI; B₁–F₁: PDAI concentrations 3, 5, 7, 10, 15 mg mL⁻¹. Characterization: XRD (PANalytical X'Pert Pro, Cu Kα), SEM (Philips XL30), UV–Vis (PerkinElmer Lambda 25), steady-state PL, J–V (Keithley 2400, AM1.5G 100 mW cm⁻², scan rate 50 mV s⁻¹, masked 0.09 cm² aperture), IPCE (IRASOL, IPCE-015). Stability: Unencapsulated samples tested in 90 ± 5% RH at 27 ± 2 °C (UV–Vis every 10 d) and thermal cycling at 85 °C (15 ± 5% RH) for 30 min followed by 1 h cooling, with J–V recorded each cycle. Self-healing: Exposure to hot water vapor followed by ambient healing; color, XRD, and J–V monitored.

• Optimal PDAI concentration: 5 mg mL⁻¹ post-treatment produced enlarged grains, compact grain boundaries, reduced trap density, and improved charge extraction; excessive PDAI (≥10–15 mg mL⁻¹) left unreacted residues, increased trap-assisted recombination, light scattering, and raised series resistance.
• Device performance (with CuSCN): At 5 mg mL⁻¹ (device C₁), average PCE 16.10% with Jsc 21.45 mA cm⁻², Voc 1.09 V, FF 70.21%. A champion efficiency of 17.52% is reported (Supplementary Table S9). Low PDAI (3 mg mL⁻¹, B₁) increased Jsc and Voc versus control (A₁), but FF initially declined due to incomplete 2D coverage. Very high PDAI (15 mg mL⁻¹, F₁) reduced PCE to 9.60%.
• IPCE: Device C₁ exhibited higher IPCE than A₁ from 350–750 nm without changing absorption onset; integrated currents were 21.73 mA cm⁻² (C₁) versus 18.77 mA cm⁻² (A₁), matching J–V.
• Hysteresis: After 30 days (storage at 15 ± 5% RH, dark), hysteresis index (HI) values: A₁ = −0.219, B₁ = −0.205, C₁ = −0.027, D₁ = 0.036, E₁ = −0.044, F₁ = 0.049. PDAI (5 mg mL⁻¹) markedly reduced hysteresis.
• Stability: • Thermal cycling (6 cycles at 85 °C, 15 ± 5% RH): normalized PCE retained after 6 cycles: A₁ 54.93%, B₁ 62.99%, C₁ 90.23%, D₁ 79.97%, E₁ 69.96%, F₁ 60.01%. Device C₁ showed partial recovery during cycling (self-healing behavior). • High humidity (90 ± 5% RH): After 5 h, unencapsulated A₁ retained 58.02% of initial PCE; C₁ and D₁ retained 99.01% and 93.99%, respectively. Film-level tests showed that control films converted to PbI₂ over 30 days, while C₁ films exhibited minimal further degradation after 20 days. • Low humidity (15 ± 5% RH): Over 1440 h, the PDAI-treated device (C₁) showed essentially unchanged PCE; A₁ retained 72.27%.
• Photoluminescence: 3–7 mg mL⁻¹ PDAI increased PL intensity relative to control, indicating reduced non-radiative recombination; PL was quenched upon adding TiO₂ and CuSCN, with stronger quenching when the 2D interface was present, consistent with improved hole extraction.
• Self-healing: Under hot water vapor exposure, PDAI-passivated films and devices (notably C₁) exhibited reversible color changes and recovery of J–V characteristics upon drying. XRD revealed a PbI₂ peak at 12.7° during exposure that disappeared after healing, indicating restoration of the perovskite phase.

Engineering an ultra-thin 2D (PDA)₂PbI₄ layer via PDAI post-treatment at the perovskite/CuSCN interface effectively passivates grain boundaries and surface defects, leading to larger grains, fewer traps, reduced recombination, and improved charge extraction. The 2D layer enhances interfacial contact and energetics with CuSCN, facilitating hole transfer and boosting Voc, Jsc, and FF at the optimal PDAI concentration. The reduced hysteresis (HI ≈ 0.03) stems from suppressed ion migration and decreased interfacial trap density; the 2D layer provides a physical barrier that raises activation energy for ion diffusion, while PDAI infiltrates GBs to passivate iodide-rich traps. Moisture and thermal stability improvements arise from the hydrophobic nature of the 2D capping layer and hydrogen-bond-mediated crosslinking across perovskite grains, which limits A-site cation migration and water penetration. The reversible formation/removal of PbI₂ under water vapor exposure and subsequent drying evidences a self-healing mechanism that restores the original perovskite structure and device performance, addressing a key barrier to PSC stability, particularly when paired with a low-cost, inorganic CuSCN HTL.

The study demonstrates that post-treatment with 5 mg mL⁻¹ p-Phenylenediaminium iodide induces in-situ formation of an ultra-thin 2D (PDA)₂PbI₄ interface layer on (FAPbI₃)₀.₈₅(MAPbBr₃)₀.₁₅, enabling efficient, stable, and self-healing perovskite solar cells with CuSCN as a low-cost HTL. The optimized devices achieved average PCE of 16.10% (champion 17.52%), with enhanced Jsc, Voc, FF, markedly reduced hysteresis, and excellent moisture and thermal stability, including near-complete retention at high humidity over hours and negligible degradation over 1440 h at low humidity. The 2D layer passivates defects, improves interfacial charge extraction, inhibits ion migration, and provides a hydrophobic barrier, collectively enabling rapid recovery after humidity-induced degradation. These findings introduce a viable path toward cheap and extra-stable PSCs with outstanding self-healing ability, supporting progress toward commercialization.

• Excess PDAI (≥10–15 mg mL⁻¹) leads to unreacted diammonium trapped at grain boundaries, increased series resistance, enhanced trap-assisted recombination, degraded performance, and reduced stability.
• The majority of stability tests were performed on unencapsulated small-area devices (0.09 cm²) under controlled humidity and temperature; long-term operational stability under continuous illumination and in fully packaged modules was not reported.
• CuSCN processing can be challenging due to solvent–perovskite interactions; although mitigated by the 2D interlayer, broader compatibility across processing conditions was not explored here.