《Science(IF>63.832)》Researchers from Helmholtz-Zentrum Berlin (HZB) - The Realization of over 31% Efficient Triple-Cation Perovskite-Silicon Tandem Solar Cells by Band Alignment Regulation via Interfacial Defect Passivation
Highlights
- The researchers utilized ionic liquid to improve band alignment between triple-halide perovskite and electron transport layer C60, reducing non-radiative recombination losses and enhancing charge extraction efficiency.
- An open-circuit voltage of 2.0V was achieved in silicon tandem solar cells.
- Uniform deposition of triple-halide perovskite top layer on crystalline silicon micropyramids yielded over 31% power conversion efficiency.
Background
Triple-halide perovskite–silicon tandem solar cells have become a hot research area in photovoltaics. Mitigating charge recombination losses and facilitating charge transfer at the interface is the key to realize high-performance tandem devices. This work aims to optimize the bandgap alignment of triple-halide perovskite and silicon through interface engineering for improved power conversion efficiency.
Results
Researchers at Helmholtz-Zentrum Berlin (HZB) found that incorporating piperazinium iodide ionic liquid between triple-halide perovskite and electron transport layer C60 formed a positive dipole and modulated the bandgap, improving their band alignment and reducing non-radiative recombination losses and enhancing charge extraction. The open-circuit voltage reached 1.28 V for single junction solar cells.
Another group discovered that using two phosphonic acid modifiers enabled uniform deposition of the triple-halide perovskite absorber layer on the micropyramids of crystalline silicon substrates. This passivation effect significantly suppressed interfacial defect density and charge recombination losses.
The optimized triple-halide perovskite–silicon tandem cells finally achieved an open-circuit voltage up to 2.0 V and a certified power conversion efficiency exceeding 31%.
Methods
TheHZB team prepared triple-halide perovskite films by solution process, incorporating methylammonium iodide ionic liquid and piperazinium iodide ionic liquid successively to form a positive dipole and modulate the bandgap for better alignment with C60. They studied the interfacial charge transfer and recombination processes by transient photoelectron spectroscopy and steady-state photoluminescence, proving that the ionic liquid addition mitigated non-radiative losses and improved charge extraction.
The other group deposited p-i-n structured triple-halide perovskite/C60 films on the micropyramids of n-type silicon substrates by spray deposition and spin coating. By adding phenylphosphonic acid and fluorophenylphosphonic acid modifiers, a uniform and compact film deposition was achieved. The interfacial defect density decreased significantly and interfacial charge recombination also reduced markedly.
Conclusion
This work optimized the triple-halide perovskite–silicon interface using ionic liquid and phosphonic acid modifiers, improving bandgap alignment, reducing interfacial defect density, and suppressing charge recombination losses. The optimized triple-halide perovskite–silicon tandem cells realized an open-circuit voltage of 2.0 V and a power conversion efficiency exceeding 31%. This provides an effective approach to achieve higher-efficiency triple-halide perovskite–silicon tandem solar cells in the future.
Figure S10. a) Representative JV curves of perovskite single junctions prepared with 3Hal with and
without surface treatment. 3Hal devices without treatment show an average PCE of 18.2% and VOC of 1.161 V. Comparing these results with 3Cat perovskite with identical band gap of 1.68 eV and device structure (glass/ITO/2PACz/3Cat/C60/SnO2/Ag) the VOC is slightly improved as the 3Cat devices show an average VOC of 1.147 V (Al-Ashouri & Köhnen et al. (8)). However, the QFLS values show that 3Hal perovskite has a substantially higher VOC potential than 3Cat (QFLS of 1.26 eV for 3Cat (8) and 1.32 eV for 3Hal measured on quartz glass, as shown in Table S1). PI surface treatment provides the highest VOC. b) Box chart showing all JV parameters of devices with and without surface treatment. We attributed the spread in JSC values between devices to changes in device area for these measurements that were performed without an aperture mask. Nevertheless, JSC values calculated from the external quantum efficiency (EQE) in c) show similar current densities for all cells. In d) the derivative of the EQE calculated from (c) shows identical band gaps for all devices (1.678 eV). Thus, the band gap is not affected by surface treatment.
Figure S11. Spectrum of the lamp used to age single-junction solar cells in comparison to AM1.5G.
Figure S13. Ageing test on perovskite/silicon tandem devices using 2PACz as HTL, 3Hal and PI as
surface treatment method. a) JV of device used for ageing test. b) Ageing test was performed over 478 hours. The graph shows (from top to bottom): absolute PCE, normalised PCE (normalised to average PCE of the first 60 min), voltage at maximum power point VMPP, current density at maximum power point JMPP, relative humidity and temperature. The device retains 75.7% of the initial PCE after 478 hours. 80% of the initial PCE is reached after 347 hours. Humidity and temperature profiles show a step around 350 h caused by fail of the cooling system during analysis.
Figure S15. A) JV curve and b) EQE of perovskite/silicon tandem device fabricated using Me-4PACzbased SAM as HTL, 3Hal with PI treatment on silicon bottom cell with RDBL.
Figure S17. JV parameters of simulated tandem solar cells with different IZO thicknesses and finger
heights as a function of the number of grid fingers. The simulations were performed using a distributed SPICE model (43). The electrical parameters were taken from Ref. (8) and the JSC values were taken from optical simulations using GenPro4 (Table S5).
Figure S19. a) EQE of tandem solar cells with IZO thicknesses of 100 nm, 60 nm, 40 nm and 20 nm
with grid fingers (labelled “100, 60, 40, 20” in the graph) and 100 nm without grid (labelled “Ref”).
Note, that these devices have LiF interlayer instead of PI, 2PACz instead of Me-4PACz and no RDBL.
The integrated current densities under AM1.5g illumination are given in the legend in mA cm-2. b)
Statistics of device parameters extracted from JV measurements. The median values are given for each variation.
Figure S20. J-V measurement of a tandem solar cell comprising PI, Me-4PACz, RDBL and grid fingers
with 40 nm IZO. For the first time, a VOC of 2.0 V is achieved for perovskite/silicon tandem solar cells enabling a PCE of 32.36%.
Figure S26. Comparison of JV measurements recorded in-house (HZB; JV sweep) and at JRC-ESTI
((quasi) steady-state). The data points of JRC-ESTI are identical to the points shown in Figure S25 and Fig. 3D.
Figure S29. Deconstruction of subcell pseudo-JV and tandem pseudo-JV curves via intensity dependent photoluminescence (PL) and injection dependent electroluminescence (EL) in comparison to the JV measured under a solar simulator. The corresponding performance parameters are given in Table S6.
