Device fabrication and characterization

FTO substrate (8 Ω per square, Nippon Sheet Glass) was sequentially cleaned with detergent, deionized water, acetone, and alcohol under sonication. Dry substrate was then immersed in a 25 mM TiCl₄ aqueous solution for 60 min at 70 °C and washed with deionized water and ethanol, followed by annealing at 500 °C for 60 min in muffle to form a c-TiO₂ blocking layer. The CsPbI₂Br layer was fabricated by spin-coating the precursor solution on the FTO/c-TiO₂ substrate via a two-step process with 500 rpm for 3 s, and 3500 rpm for 30 s. Subsequently, these films were annealed at 43 °C for 2 min, and then at 160 °C for 10 min⁴⁵. P3HT (15 mg/mL in chlorobenzene, with 45 µL Li-TFSI solution (2 mg in 1 mL acetonitrile) and 10.2 µL t-BP) was spin-coated onto the CsPbI₂Br perovskite film at 2500 rpm for 25 s as a hole transporting layer. Finally, the device was finished by evaporating Ag or Au layers²⁹.

FTIR spectroscopy, Raman spectroscopy (Raman) and Carbon-13 nuclear magnetic resonance (¹³C-NMR) were performed by FTIR Nicolet 6700, Laser Raman InVia Reflex and Bruker Avance 500, respectively. The surface morphology and roughness of films were characterized by field emission scanning electron microscopy (HITACHI S4800) and atomic force microscopy (AFM, Veeco/DI). The XPS measurements were performed in ESCALAB 250Xi (Mg anode, 250 W, 14 kV), and the binding energy of the C 1s peak at 284.8 eV was taken as an internal pristine. XRD spectra of the prepared perovskite films were measured using powder X-ray diffraction (PXRD, Bruker Advance D8 X-ray diffractometer, Cu Kα radiation, 40 kV). Time-of-flight secondary-ion mass spectrometry (ToF-SIMS VI, IONTOF GmbH, Muenster, Germany) elemental depth profiling was used to probing the distribution of elements. UV–Vis spectra were collected using a Cary 500 UV–Vis–NIR spectrophotometer in air ambient environments. The steady-state PL and TRPL spectra of perovskite films were acquired by Fluorolog-3-p spectrophotometer and Endinburgh FLS890 spectrometer in air at room temperature, respectively. PL excitation wavelength was 380 nm. Solar cells were illuminated by a solar light simulator (Solar IV-150A, Zolix) and light intensity was calibrated by a standard Newport calibrated KG5-filtered Si reference cell. The J–V curves of devices were measured with Keithley 2400 digital sourcemeter under AM 1.5G irradiation (100 mW cm⁻²) at a scan rate of 0.15 V s⁻¹ (voltages scan range: 0.3–1.5 V, voltage step of 10 mV) in air ambient environments. Devices were masked with a metal aperture to define the active area to be 0.0625 cm². The steady-state photocurrent output of the best-performing devices was measured by biasing the device at MPP in air ambient environments. EIS were measured out using an electrochemical workstation (Parstat 2273, Princeton) in the frequency range of 1 MHz and 1 Hz under different positive bias voltages at dark conditions in air. The EQE was carried out on a Newport-74125 system (Newport Instruments) in air.

The CsPbI₂Br films and unsealed devices were stored in air ambient environments with 15 ± 3% relatively humidity for long-term humidity test. White LED light illumination of 100 mW cm⁻² for illumination stability tests of unencapsulated devices in nitrogen glovebox. The unencapsulated devices were baked on the hot plate at 85 °C in nitrogen glovebox for thermal stability test.