One-stone-for-two-birds strategy to attain beyond 25% perovskite solar cells

Organic-inorganic hybrid perovskite solar cells (PSCs) have attracted significant attention from researchers since their first demonstration in 2009 due to the low-cost solution processing, tunable bandgap, high absorption coefficient, low recombination rate, and high mobility of charge carriers. The power conversion efficiency (PCE) of single-junction PSCs has rapidly increased from 3.8% to a certified value of 25.7%. The long-term operational stability of unencapsulated PSCs has also exceeded 1000 h in full sunlight after the interface engineering of device structures and molecular passivation of the perovskite layer. Therefore, PSCs represent a promising next-generation photovoltaic technology.

Over the past few years, it has been demonstrated that the elimination of deep-level defects, which act as detrimental nonradiative recombination centers, is critical for realizing high-performance solar cells. To date, iodide anions (I−) vacancy defects constitute the majority of non-radiative recombination centers in the FAPbI3 perovskite layer that are very difficult to mitigate. These defects are mainly caused by iodine losses during film fabrication or the I− ion migration during device operation, which produces deep-level defects and directly leads to the degradation of photoelectric properties. Furthermore, the I− ions desorbed from the inorganic framework can be easily oxidized to I0 species, which initiate chemical chain reactions accelerating the formation of deep-level defects in perovskite layers. This phenomenon is detrimental for devices operating under complex conditions (including high temperature, continuous light illumination, electrical bias, or their combination) and negatively affects the long-term stability of operational PSCs. Therefore, inhibiting the formation and migration of I vacancy defects is critical for stabilizing photoactive perovskite layers and preserving their good photoelectric properties.

Several attempts have been made to passivate anion vacancy defects through the introduction of additives such as quaternary ammonium halide anions and cations, caffeine and theobromine, CsI-DB21C7 complex, naphthalene-1,8-dicarboximide and perylene-3,4-dicarboximide. However, these additives typically passivate undercoordinated Pb2+ ions after anions have escaped the crystal lattice and do not fundamentally prevent vacancy formation or migration. The authors outline five desired characteristics for effective additives: strong bonding to the Pb–I framework to pin anions; localization of escaped anions to suppress migration; minimal steric hindrance to preserve charge transport; stability under thermal stress and illumination; and growth-control capability to promote perovskite crystallization. Prior sacrificial agents such as phosphonopropionic acid and amino-based ligands either degrade thermally or cannot form sufficiently strong bonds for anion pinning, and their bulky organic layers can hinder charge extraction, limiting their suitability for high-efficiency stable solar cells.

Strategy and materials: The study introduces an amidino-based ligand, 3-amidinopyridine (3AP), delivered as iodized 3-pyridinecarboximidamide (3API), to achieve in situ anion fixation and undercoordinated-Pb passivation during perovskite crystallization. 3AP is designed to strongly coordinate with the Pb–I framework via its amidino and pyridyl –NH functionalities, yielding short interlayer spacing and robust anchoring.

Computational analysis: Density functional theory (DFT, CASTEP, PBE-GGA, PW-USPP; Ecut 340 eV) was used to model ligand adsorption on the FAI-terminated FAPbI3 (001) surface, calculate adsorption energies, iodide vacancy formation energies, and iodide vacancy migration barriers. Supercells included ligands 3AP, 3AMP, 3AMPY, BA, and PEA. Transition state searches employed complete linear synchronous transit/quadratic synchronous transit methods. Adsorption energy, defect formation energy, and migration barriers were calculated from total energy differences using standard expressions provided.

Spectroscopy and coordination evidence: Coordination between 3AP and Pb–I framework was probed with 1H NMR (chemical shift changes of amidino and ring protons upon mixing 3AP with PbI2 in DMSO‑d6), FTIR (shifts of –NH and ring –CH modes), and XPS (Pb 4f binding energy shifts to lower BE indicating electron-rich Pb environment). These corroborate strong hydrogen bonding/coordination and anion stabilization at surfaces/grain boundaries.

Film and device fabrication: Planar n‑i‑p PSCs with architecture FTO/c‑TiO2/FAPbI3/Spiro‑OMeTAD/Au were fabricated. FAPbI3 precursor (1.5 M, DMF:DMSO 4:1) with MACl additive contained 0–8 mol% 3APX (X = I, Br, Cl); optimal at 4 mol% 3API. Spin-coating at 6000 rpm for 50 s with diethyl ether antisolvent drip at 10 s; anneal 150 °C for 10 min; PEAI post-treatment (20 mM in IPA). Hole transport: Spiro‑OMeTAD; Au top electrode by thermal evaporation. For thermal stability studies, PTAA replaced Spiro‑OMeTAD and organic passivation was removed.

Characterization techniques: GIWAXS to assess phase formation and orientation; SEM (plan-view and cross-section) and AFM for morphology and roughness; TOF‑SIMS for 3AP distribution; UV‑Vis absorption, steady-state PL, TRPL for optoelectronics; SCLC (FTO/c‑TiO2/perovskite/PCBM/Ag) to extract trap density and electron mobility; C–V (Mott-Schottky) for built-in potential; UPS for energy level alignment; TPC/TPV for charge extraction and recombination kinetics; KPFM for surface potential; in situ PL during spin/anneal to monitor crystallization and stacking defects; laser-scanned time-resolved PL with photocurrent mapping on working devices to correlate PL and local photocurrent; temperature-dependent conductivity (188–333 K) to extract ion migration activation energy via Nernst–Einstein relation.

Stability testing: XRD aging of films under ambient dark storage (25 °C, 30–40% RH); device thermal storage at 85 °C in N2 (unencapsulated); ambient shelf-life (25 °C, 30–40% RH) for 5000 h (unencapsulated); encapsulated device operational stability via MPP tracking under white LED 100 mW cm−2 following ISOS‑L‑1 protocol.

• Stronger ligand–perovskite coordination: DFT adsorption energy for 3AP on FAPbI3 (001) was 3.135 eV, higher than 3AMP (2.784 eV), 3AMPY (2.988 eV), BA (2.968 eV), and PEA (2.498 eV).
• Vacancy suppression: Iodide vacancy formation energy increased to >3.812 eV with 3AP (vs 1.075 eV for MAPbI3, 1.422 eV for FAPbI3; higher than 3AMP 3.574 eV, 3AMPY 3.791 eV, BA 3.716 eV, PEA 3.584 eV). Vacancy migration barrier increased from 0.737 eV (control) to 1.467 eV with 3AP (higher than others: 0.86–1.29 eV).
• Experimental ion migration: Activation energy from temperature-dependent conductivity increased from 0.109 eV (control) to 0.198 eV (3AP), surpassing 3AMP (0.177 eV), 3AMPY (0.187 eV), BA (0.178 eV), and PEA (0.162 eV).
• Device performance: Optimal 4 mol% 3API yielded PCE 25.3% (certified 24.8%), Voc 1.181 V, Jsc 26.04 mA cm−2, FF 82.21%. Control achieved PCE 22.76%, Voc 1.123 V, Jsc 24.94 mA cm−2, FF 0.81. Certified EQE-integrated Jsc was 25.97 mA cm−2 (−0.8% vs Jsc). Light-intensity Voc slope reduced from 1.60 kBT/q (control) to 1.26 kBT/q (3AP), indicating reduced nonradiative recombination.
• Film structure and morphology: GIWAXS showed suppression of undesired 6H phase and promotion of 3C phase with 3AP. Average grain size increased (~1.0 → 1.4 μm); surface roughness decreased (23.3 → 17.0 nm). 3AP uniformly distributed (TOF‑SIMS).
• Recombination dynamics: PL intensity increased; TRPL lifetime extended from 1.2 μs (control) to 5.5 μs (3AP), indicating suppressed nonradiative recombination.
• Transport properties (SCLC): Trap density reduced from 1.43 × 10^16 to 4.76 × 10^15 cm−3; electron mobility increased from 0.139 to 0.339 cm^2 V−1 s−1.
• Built-in potential and energetics: Vbi and Voc increased (Voc from 1.108 to 1.160 V via C–V). UPS showed VBM/CBM shifted from −5.46/−3.92 eV to −5.55/−4.01 eV (better alignment with ETL). TPC/TPV showed faster extraction (4.25 vs 9.38 μs) and longer recombination lifetime (11.2 vs 3.07 μs). KPFM indicated higher electronic chemical potential (less n-type surface), facilitating hole extraction.
• Crystallization and stacking defects: In situ PL during annealing showed 3API retards crystallization and suppresses stacking-defect formation (PL intensity trends). On-device PL and photocurrent mappings revealed larger grains, stronger PL, and higher local photocurrent; fraction of pixels with photocurrent >16 nA increased from 43% to 55% with 3API.
• Stability: Ambient dark film aging (74 days) showed suppressed δ‑FAPbI3/PbI2 formation with 3AP. Devices (unencapsulated) at 85 °C in N2 retained 83% PCE after 500 h vs 52% for control. Ambient unencapsulated (25 °C, 30–40% RH) retained ~92% after 5000 h vs ~80% for control. Encapsulated MPP under 100 mW cm−2 (ISOS‑L‑1): 3AP devices lost ~5% after 510 h; control lost 16% after 240 h.
• Generality: Replacing I− with Br− or Cl− in 3APX preserved performance gains (PCE ~24.6–24.7%) and defect suppression. Benefits extended to FA0.9Cs0.1PbI3 and FA0.92MA0.08PbI3 compositions.

The study addresses the critical challenge of halide vacancy formation and migration in FAPbI3 by implementing an amidino-based ligand (3AP) that strongly coordinates with the Pb–I framework. This “one-stone-for-two-birds” approach simultaneously fixes anions (suppressing vacancy formation and mobility) and passivates undercoordinated Pb sites. DFT results substantiate stronger adsorption, higher vacancy formation energies, and elevated migration barriers with 3AP relative to common large organic ligands. Spectroscopic signatures (NMR, FTIR, XPS) confirm strong hydrogen bonding/coordination to the perovskite lattice. These microscopic effects translate into macroscopic device improvements: reduced trap density, prolonged carrier lifetime, enhanced mobility, better energy-level alignment, faster charge extraction, and lower nonradiative recombination. Film formation is modulated to suppress stacking defects and favor the 3C perovskite phase, contributing to larger grains and smoother films. Consequently, devices achieve a PCE of 25.3% (certified 24.8%) with markedly enhanced thermal, shelf-life, and operational stability, and the approach generalizes across halides and compositions. Collectively, the findings provide a clear mechanistic link between molecule–perovskite coordination, defect suppression, and improved photovoltaic performance and stability.

This work presents a chemically simple, effective strategy to surpass 25% efficiency and enhance stability in FAPbI3 perovskite solar cells by using an amidino-based ligand (3AP) to achieve in situ anion fixation and Pb-passivation during crystallization. The strong 3AP–perovskite coordination increases iodide vacancy formation energies and migration barriers, suppresses nonradiative recombination, improves charge transport and energy-level alignment, and mitigates stacking defects. Devices reach 25.3% PCE (certified 24.8%) and exhibit superior thermal, ambient, and operational stability. The approach is compatible with different halides and perovskite compositions, indicating broad applicability to perovskite optoelectronics and scalability for large-area fabrication. Future research may explore module-scale implementation, long-term outdoor testing under IEC/ISOS protocols, ligand structure optimization for further defect control, and integration with tandem and semitransparent architectures.