Suppressing the Electron-Phonon Coupling in 2D Perovskite Cs3Sb2I9 for Lead-Free Indoor Photovoltaics

Abstract

Antimony-based perovskite-inspired materials (Sb-PIMs) are promising lead-free candidates for indoor photovoltaic application. Cs₃Sb₂I₉, in particular, with a ≈2.0 eV bandgap, is ideal for harvesting indoor white light. However, solution-processed Sb-PIMs preferentially crystallize into thermodynamically stable 0D structures, leading to strong self-trapped exciton (STE) formation, limiting device performance. Although chloride (Cl) doping can induce 2D structural transitions, it enhances Fröhlich electron-phonon coupling (EPC), creating an intrinsic trade-off. Here, we develop an anion-exchange strategy to fabricate phase-pure, Cl-free 2D Cs₃Sb₂I₉ films that suppress STE formation while enabling controlled dimensional reconstruction. This approach yields a reduced Huang-Rhys factor (from 30.7 to 21.5) and prolonged STE lifetime (8.60 to 9.19 ps). Density functional theory (DFT) calculations reveal a significant reduction in excited-state octahedral distortion (Δd = 0.898 × 10^-3 for Cs₃Sb₂I₉ vs. 5.752 × 10^-3 for Cs₃Sb₂I₆Cl₃), confirming intrinsically weaker EPC in Cl-free structures. The device achieves a power conversion efficiency (PCE) of 3.40% under AM 1.5G solar illumination and an 8.2% PCE under 1000 lux white LED conditions. alongside Long-term stability measurement confirms its environmental robustness. These results represent the highest indoor performance reported to date for Sb-based perovskite-inspired solar cells.

Keywords: 2D C3Sb2I9; Pb‐free; electron–phonon coupling; indoor photovoltaic.

Figure 1

a) Crystal structures of typical 0D and 2D Sb‐based perovskite‐inspired materials. b) Raman spectra of MA₃Sb₂I_9‐xCl_x, Cs₃Sb₂I_9‐xCl_x, and MA‐Cs₃Sb₂I_9‐xCl_x films annealed at 100 °C. c) Raman spectra of Cs₃Sb₂I_9‐xCl_x, and MA‐Cs₃Sb₂I_9‐xCl_x films annealed at 240 °C. d) Absorption spectra and optical images of MA₃Sb₂I_9‐xCl_x, Cs₃Sb₂I_9‐xCl_x, and MA‐Cs₃Sb₂I_9‐xCl_x films annealed at 100 °C. e) Absorption spectra and optical images of Cs₃Sb₂I_9‐xCl_x, and MA‐Cs₃Sb₂I_9‐xCl_x films annealed at 240 °C, along with MA₃Sb₂I_9‐xCl_x film annealed at 100 °C.

Figure 2

a) XRD patterns of MA₃Sb₂I_9‐xCl_x, Cs₃Sb₂I_9‐xCl_x and MA‐Cs₃Sb₂I_9‐xCl_x annealed at 100 and 240 °C. b,c) GIWAX patterns of (b) Cs₃Sb₂I_9‐xCl_x and (c) MA‐Cs₃Sb₂I_9‐xCl_x films annealed at 240 °C. d–g) SEM images of Cs₃Sb₂I_9‐xCl_x and MA‐Cs₃Sb₂I_9‐xCl_x films treated with anti‐solvent and annealed at (d,e) 100 °C or (f,g) 240 °C. h,i) SEM images of Cs₃Sb₂I_9‐xCl_x and MA‐Cs₃Sb₂I_9‐xCl_x films annealed at 240 °C without anti‐solvent treatment.

Figure 3

a) Room temperature PL spectra of 2D‐MA₃Sb₂I_9‐xCl_x, 2D‐Cs₃Sb₂I_9‐xCl_x, and 2D‐Cs₃Sb₂I₉ films. b‐d) Gaussian deconvolution of PL spectra for (b) 2D‐Cs₃Sb₂I_9‐xCl_x, (c) 2D‐Cs₃Sb₂I₉, and (d) 2D‐MA₃Sb₂I_9‐xCl_x films. e,f) Temperature‐dependent PL spectra of (e) 2D‐Cs₃Sb₂I_9‐xCl_x and (f) 2D‐Cs₃Sb₂I₉ films. g,h) FWHM as a function of temperature extracted from (g) 2D‐Cs₃Sb₂I_9‐xCl_x and (h) 2D‐Cs₃Sb₂I₉ films. i,j) TA spectra at different time delays for (i) Cs₃Sb₂I_9‐xCl_x and (j) 2D‐Cs₃Sb₂I₉ films. k,l) TA decays of (k) glass/2D‐Cs₃Sb₂I_9‐xCl_x and (l) glass/2D‐Cs₃Sb₂I₉ films probed at 600 nm.

Figure 4

a) Charge density plots of the VBM and CBM for 2D‐Cs₃Sb₂I₉. b,c) PDOS for (b) 2D‐Cs₃Sb₂I₉ and (c) 2D‐Cs₃Sb₂I₆Cl₃. d,e) Ground‐state optimized structures from DFT calculations for (d) 2D‐Cs₃Sb₂I₉ and (e) 2D‐Cs₃Sb₂I₆Cl₃. f,g) Relaxed excited‐state geometries for (f) 2D‐Cs₃Sb₂I₉ and (g) 2D‐Cs₃Sb₂I₆Cl₃.

Figure 5

Device configuration and photovoltaic performance of Sb‐based perovskite‐inspired solar cells. a) Schematic of the planar architecture: FTO/Nb₂O₅/PIMs/P3HT (or F4TCNQ:P3HT)/Carbon. b) Energy level alignment of each functional layer. c) J–V curves of 2D‐MA₃Sb₂I_9‐xCl_x, 2D‐Cs₃Sb₂I_9‐xCl_x, and 2D‐Cs₃Sb₂I₉ devices under AM1.5G. d) Efficiency statistics of corresponding devices. e) J–V curves of 2D‐Cs₃Sb₂I₉ devices with pristine or F4TCNQ‐doped P3HT HTL. f) Efficiency distributions with and without F4TCNQ doping. g) Comparison of AM1.5G and WLED spectra. h) J–V curve of 2D‐Cs₃Sb₂I₉ device under 1000 lux WLED. i) Indoor PCE statistics under WLED. j) PCE comparison with reported Sb/Bi‐based perovskite‐inspired indoor photovoltaics.