Imaging and quantifying non-radiative losses at 23% efficient inverted perovskite solar cells interfaces

Abstract

Interface engineering through passivating agents, in the form of organic molecules, is a powerful strategy to enhance the performance of perovskite solar cells. Despite its pivotal function in the development of a rational device optimization, the actual role played by the incorporation of interfacial modifications and the interface physics therein remains poorly understood. Here, we investigate the interface and device physics, quantifying charge recombination and charge losses in state-of-the-art inverted solar cells with power conversion efficiency beyond 23% - among the highest reported so far - by using multidimensional photoluminescence imaging. By doing that we extract physical parameters such as quasi-Fermi level splitting (QFLS) and Urbach energy enabling us to assess that the main passivation mechanism affects the perovskite/PCBM ([6,6]-phenyl-C₆₁-butyric acid methyl ester) interface rather than surface defects. In this work, by linking optical, electrical measurements and modelling we highlight the benefits of organic passivation, made in this case by phenylethylammonium (PEAI) based cations, in maximising all the photovoltaic figures of merit.

Fig. 1. Electrical characterisation of the samples.

a Schematic of the reference samples d Schematic of the A-cation p-i-n devices with dual interfacial modification. Photovoltaic characteristics of reference (orange), Cl-PEAI (purple) and F-PEAI (teal blue) devices. b Open circuit voltage (V_oc). c Short circuit current (J_sc). e Fill Factor (FF). f Power conversion Efficiency (PCE).

Fig. 2. Spatially averaged photoluminescence analysis in the continuous wave and time-resolved regime.

a Photoluminescence average spectra and corresponding fits acquired on the stack glass/ITO/PTAA/perovskite/PCBM/BCP samples. PL spectra were acquired on reference (orange), Cl-PEAI (purple) and F-PEAI samples (teal blue). b QFLS values extracted from PL spectra for neat perovskite, half cells and full devices compared with radiative QFLS Δµ^rad, open circuit voltage V_oc and radiative open circuit voltage V_oc^rad. TR-FLIM (spatially integrated) decays acquired at 10¹¹ph.cm⁻² fluence for c perovskite layers deposited on glass, top illumination and d full cells without top bottom electrode, top illumination. e Comparison of the fitted top surface recombination rate on full devices.

Fig. 3. Continuous wave photoluminescence imaging analysis.

Hyperspectral measurements on the stack glass/ITO/PTAA/perovskite/BCP/PCBM. quasi-Fermi level splitting (Δµ) maps for a reference sample, b Cl-PEAI and c F-PEAI samples. Urbach Energy maps for d reference sample, e Cl-PEAI and f F-PEAI samples. g Correlation between energy gap and QFLS. h correlation between Urbach Energy and QFLS. The scalebar applies to all images.

Fig. 4. Time-resolved photoluminescence imaging analysis.

a–c Map of decay time obtained on full-stacks (glass/ITO/PTAA/perovskite/BCP/PCBM) for the high fluence (1.5 × 10¹²ph.cm⁻²) acquisition. The homogeneity of the A-cations devices is comparable to the neat reference cell. d Histogram of the maps, showing the passivation effect of the layers.