Implementing Dopant-Free Hole-Transporting Layers and Metal-Incorporated CsPbI2Br for Stable All-Inorganic Perovskite Solar Cells

Mixed-halide CsPbI2Br perovskite is promising for efficient and thermally stable all-inorganic solar cells; however, the use of conventional antisolvent methods and additives-based hole-transporting layers (HTLs) currently hampers progress. Here, we have employed hot-air-assisted perovskite deposition in ambient condition to obtain high-quality photoactive CsPbI2Br perovskite films and have extended stable device operation using metal cation doping and dopant-free hole-transporting materials. Density functional theory calculations are used to study the structural and optoelectronic properties of the CsPbI2Br perovskite when it is doped with metal cations Eu2+ and In3+. We experimentally incorporated Eu2+ and In3+ metal ions into CsPbI2Br films and applied dopant-free copper(I) thiocyanate (CuSCN) and poly(3-hexylthiophene) (P3HT)-based materials as low-cost hole transporting layers, leading to record-high power conversion efficiencies of 15.27% and 15.69%, respectively, and a retention of >95% of the initial efficiency over 1600 h at 85 °C thermal stress.


S3
2000 rpm s-1. After drying at 125 °C, the TiO2 films were gradually heated to 500 °C, baked at this temperature for 15 min and then cooled to room temperature. The mesoporous deposited film was again treated with TiCl4 followed by sintering at 450 °C for 30 min.

Synthesis of the perovskite solution
CsPbI2Br preparation. The 1.2 M CsPbI2Br perovskite precursor solution was prepared by stoichiometrically mixing 0.277 g PbI2 (TCI), 0.220 g PbBr2 (Sigma) and 0.312 g CsI (Sigma), in 1ml anhydrous DMSO. This solution was stirred overnight at 75 °C and then filtered through a 0.2 µm syringe filter. InCl3:CsPbI2Br preparation. Initially 2 % InCl3 power was mixed in above CsPbI2Br solution and used as a stock solution. Desired volume of 2 % InCl3:CsPbI2Br solution was used for 0.25% InCl3 doping.
Device fabrication. The clear filtered 50 µl yellow solution was spin-coated on the top of the FTO/c-TiO2/mp-TiO2 electrode by a consecutive two-step spin coating process at 1,000 and 3,000 rpm for 10 and 30 s, respectively. Experimental details are discussed in our previous report. [S8, S9]

Synthesis of CuSCN HTM:
35 mgml -1 CuSCN HEL in diethyl sulfide (DES) solution was prepared by continuous stirring for 30 min at room temperature. The cleared solution was filtered through syringe filter and used for deposition. For CuSCN deposition, we have used dynamic spin-coating method and CuSCN solution was dropped quickly onto CsPbI2Br thin film spinning at 5000 rpm followed by 40 sec further spinning. The CuSCN deposited samples were further dried on hot plate at 100 °C for 5 min and used for rGO deposition. 1 mgml -1 well dispersed reduced graphene oxide (rGO) in chlorobenzene solution was deposited by spin coating onto the CuSCN HTM at 3000 rpm for 30 sec. Then, the substrates were transferred to a vacuum chamber subsequently evacuated to a pressure of 2×10 -6 mbar. The devices were completed by deposited 60 nm Au onto the CuSCN/rGO layer through shadow masks with an active area of 0.09 cm 2 .

Preparation of P3HT based HTM
The P3HT hole transporting material was prepared by dissolving P3HT in chlorobenzene (15 mg.ml -1 ) and used without any additive dopants. The prepared P3HT HTM solution was spin-coated on the FTO/c-TiO2/mp-TiO2/perovskite at 3,000 rpm for the 30 s on S4 preheated (100 °C) perovskite thin film quickly. Fabricated devices were further annealed at 200 °C for 5 min and then the devices were transferred to a vacuum chamber and evacuated to a pressure of 2×10 -6 mbar. For the counter electrode, a 60 nm thick Au contacts were deposited on the top of the P3HT over layer by thermal evaporation (growth rate ~0.5Å s -1 ). The active area of this electrode was fixed to 0.09 cm 2 . An active area was calculated as per gold and laser pattern cross-sectional area. The exact illumination to the active area was fixed by attaching thin metal shadow mask from the backside during measurements.
Perovskite thin film characterizations. The top-surface and cross-sectional images were recorded by a field emission scanning electron microscope (FESEM; S-4700, Hitachi). X-ray diffraction (XRD) measurements were carried out using a D/MAX Ultima III XRD spectrometer (Rigaku, Japan) with Cu Kα line of λ=1.5410 Å. A double beam spectrophotometer (Varian, CARY, 300 Conc.) in the 280-800 nm wavelength range. An inverted-type scanning confocal microscope (Picoquant, MicroTime 200, Germany) with a 100 × (oil-immersion) objective was employed to measure the fluorescence lifetime imaging (FLIM). Single-mode pulsed solid state diode laser (470 nm with ~30 ps pulse width, ~0.1 µW average power, and operated at 2.5 MHz repetition rate) was used as an excitation source. The instrumental response function of the system was ∼100 ps at FWHM. A dichroic mirror (490 DCXR, AHF), a long-pass filter (HQ500lp, AHF), a 75 µm pinhole, a band-pass filter (FB550-40, Thorlabs), and a single photon avalanche diode (PDM series, MPD) were used to collect emissions from the bare and doped CsPbI2Br samples. Time-correlated single-photon counting (TCSPC) technique was used to count the time-resolved emission signals. Typically, FLIM images of an 80 μm × 80 μm perovskite sample area consisting of 200 × 200 pixels, were recorded using the time-tagged time-resolved (TTTR) data acquisition method. The acquisition time of each pixel was 2 ms per pixel. Photoluminescence lifetime images and their exponential fittings for the obtained fluorescence decays, extracted from the FLIM images, were performed using SymPhoTime-64 software provided by the manufacturer in an exponential decay model, where I(t) is the time-dependent PL intensity, A is the amplitude, τ is the PL lifetime, and i is 3. Steady-state fluorescence spectrum 5 at a focal volume was measured for the samples by guiding the fluorescence signal to the external spectrometer (F-7000, Hitachi).

Absolute PLQY measurements
Absolute PL quantum yield spectrometer system (Hamamatsu, C9920-02/-02) was used for absolute PL Quantum Yield measurement. Absolute PL quantum yield (PLQY) measurements were performed following the procedure of de Mello and co-workers. [S10] A 405 nm laser was used to photoexcite the samples placed in an integrating sphere and a Maya pro spectrometer used to measure the signal.

XPS analysis and its fitting
A Thermo Scientific K-ALPHA + X-Ray Photoelectron spectrometer was used to perform XPS measurements using a monochromatic Al Kα X-Ray source at a take-off angle of 60 degrees.
The core level XPS spectra were recorded using a pass energy of 50 eV (resolution approximately 0.4 eV) from an analysis area of 400 μm x 400 μm. The spectrometer work function and binding energy scale were calibrated using the Fermi edge and 3 d peak recorded from a polycrystalline silver (Ag) sample prior to the commencement of the experiments.
Fitting procedures to extract peak positions from the XPS data were carried out using XPS peak 4.1/Magicplot. A Shirley background was used and the spectra were fit with a mixture of Gaussian/Lorentzian (Lorentzian = 30 %) line shapes.

TEM and HAADF-STEM
Transmission electron microscopy was performed on a high-resolution transmission electron microscopy (HRTEM) TECNAI F20 Philips operated at 200 KV. Energy-dispersive X-ray (EDX) elemental mapping was carried out in the scanning transmission electron microscopy (STEM) mode. The contrast against the background of HAADF-STEM image for mapping has been carried our background selection: ROIs mode at detector angle: 14.6 degree.

Focused Ion Beam (FIB)
Focused Ion Beam (FIB) System NX2000 used for analysis of cross sectional images of fabricated devices for different time interval.
Photovoltaic studies. The cells were illuminated using a solar simulator at AM 1.5 G for 10 s, for which the light intensity was adjusted to 1 sun intensity (100 mW cm -2 ) through the use of an NREL-calibrated Si solar cell with a KG-5 filter. The J-V curves were measured along the reverse scan direction from 1.5 V to − 0.05 V or the forward scan direction from −0.05 V to 1.5 V. The step voltage and scan speed were fixed at 10 mV and 150 mV s −1 , respectively. The J-V curves for all devices were measured by metal shadow masks with active areas of 0.3 x 0.3 =0.09 cm 2 (small-area device) and 1 × 1 cm 2 (large-area device) in size.

S6
External quantum efficiency measurement. The spectral response was taken by an spectral IPCE measurement system (K3100, McScience), which was equipped with a monochromator, a K240 XE 300 lamp source connected with K401 OLS XE300W lamp Power supply and a K102 Signal amplifier. Prior to the use of the light, the spectral response and the light intensity were calibrated using a Si-photodiode (Model: S1337-1010BQ) and InGaAs photodiode (model: G12180-050A) for 300-1100 nm and 1100-1400 nm calibration respectively.
Measurements were taken in External quantum efficiency (EQE) mode.
Device stability test. All measurements were performed on un-encapsulated cells in ambient air. The device stability was tested in air at room temperature without encapsulation and after each measurement, devices were stored in ambient condition without any encapsulation. For long term stability, devices were kept in a petri dish in ambient air without any encapsulation and J-V curves were periodically measured under AM1.5 G simulated sun light at room temperature. Maximum power point tracking (MPPT) was also carried out to verify the longterm photostability. The device is held at maximum power point during aging and kept at a constant temperature by an air stream flowing onto the devices (back side), with the device surface measuring approximately 20 °C under the white LED light was measured by IVUMstate potentiostate (Ivium Technologies B.V., Eindhoven, The Netherlands). For photostability, the fabricated devices directly illuminated under 1 sun illumination in ambient conditions and recorded photocurrent which is converted into efficiency using respective bias voltage. For thermal stability, perovskite devices were kept hot-plate in ambient conditions at 60 °C and 85 °C and taken out during measurements and reported relative humidity has been mentioned each figure captions. We have not used any special maintained accessories for any environmental conditions such as room temperature or humidity, dry-air box etc. However, we monitored devices temperature ~20 °C by continuous air-steam flow, and relative humidity was ~35% for continuous illumination testing. Table S1. Calculated hole (m * h) and electron (m * h) effective masses of CsPbI2Br, CsPb0.96Eu0.04I2Br and CsPb0.96In0.04I2Br materials along high symmetry directions.

Material
Direction  Supporting Note 1.
Initially, we have checked the thermal gravimetric analysis (TGA) of prepared perovskite precursors and optimized annealing temperature of perovskite, Fig. S1. Therefore, after hot-air process these samples were annealed at 280 °C for 10 min which results in crystallization of the intermediate phase into dense and pin-hole free black CsPbI2Br perovskite phase.

Fig. S2
shows a schematic illustration of the deposition of CsPbI2Br-based perovskite thin film deposition by our recently developed hot-air method in ambient condition. [S9] Typically, 1. which was adopted by our previous optimized conditions. [S9, S10] The use of DMSO led to the formation of the intermediate PbX2-DMSO-CsI complex phase which limits the initial number of nuclei centers being controlled by the specific solvent evaporation rate.

ln
From equation S1 we can easily calculate the expected increase in open-circuit voltage for a given PLQY ratio.

S14
The above equation S1 can be modified to calculate the bandgap-independent loss from the radiative limit. By setting PLQYdoped sample to be unity, i.e., in the ideal case, and then inputting the measured PLQY (PLQYabsolute), we obtain 1 25.7 ln rad absolute QFLS meV PLQY where PLQYabsolute is expressed in measured absolute value. images were analyzed using a three-exponential decay model: τavg, images constructed on the basis of the averaged photoluminescence lifetimes; τ1, τ2, and τ3 indicate images constructed on the basis of the fast, mid-range, and slow-components, respectively; τ1 + τ2 + τ3, overlay images.

Supporting Note 4:
The time-resolved photoluminescence (TRPL) decay profile data was fitted to a triexponential function of the form: where, τ1, τ2 and τ3 are first, second and third order decay time, A1, A2 and A3 are respective weight factors of each decay channel. The average recombination lifetimes < τavg > for S15 respective samples were calculated from lifetime values and weight fraction amplitude values (%) using the following equation: The non-radiative fast-decay lifetime (τ1 and τ2) and radiative slow-decay lifetime (t3) originated from the quenching of charge carriers and free charge carriers before the charge collection, respectively. S16  Figure S6. Survey spectra and XPS fittings for the Cs 3d, Pb 4f, I 3d and Br 3d core levels for the controlled CsPbI2Br sample. The counts and envelope have been offset to make the fittings clearer. Full details of peak positions can be found in Table S5. Figure S7. XPS fittings for the Cs 3d, Pb 4f, I 3d, Br 3d and Eu 3d core levels for the controlled CsPb0.95Eu0.05I2Br sample. The counts and envelope have been offset to make the fittings clearer.
Full details of peak positions can be found in Table S5.
S19 Figure S8. XPS fittings for the Cs 3d, Pb 4f, I 3d, Br 3d, In 3d and Cl 2p core levels for the controlled InCl3:CsPbI2Br sample. The counts and envelope have been offset to make the fittings clearer. Full details of peak positions can be found in Table S5.

Supporting Note 5:
The analysis of Eu peaks is little tricky due to its Eu 2+ and Eu 3+ oxidation states. We observed two prominent peaks in both groups denoted by trivalent Eu 3+ and divalent Eu 2+ and a small structure in the 3d5/2 region denoted by ''mult''. This additional peak assignment was done by Cho et al. based on their inelastic energy loss and surface and bulk plasmon. [S14] Furthermore, the lower BE peak-shift in I 3d and Br 3d core-levels confirms the Eu 3+ -Eu 2+ ion pair working as a redox shuttle which terminates the degradation. [S15] The presence of the Eu 3+ element is reflected from the doublet peak appears at 1133.92 and 1163.6 eV respectively for Eu 3+ 3d5/2 and Eu 3+ 3d3/2 core levels. While divalent Eu 2+ is revealed from doublet peaks at Eu 2+ 3d5/2 (1124.62 eV) and Eu 2+ 3d3/2 (1155.45 eV) core levels, Fig. 2d.  Figure S9. XPS fittings for the Cs 3d, Pb 4f, I 3d, Br 3d, In 3d and Cl 2p core levels for the controlled InCl3:CsPbI2Br sample. The counts and envelope have been offset to make the fittings clearer. Full details of peak positions can be found in Table S6.          S39 Figure S23. Phase degradation of exposed CsPbI2Br-based perovskite thin films stored under a moisture-rich atmosphere. Note: Humidity was recorded in the XRD chamber at room temperature. Figure S24. Structural refinement of δ-thin film after 6 hours of ambient storage. Figure S25. Decoupled orthorhombic (eorth) and tetragonal (etet) spontaneous strain components for the three different systems under investigation. The joining lines are merely a guide for the eye, with improved long-term stability running left to right.

Supporting Note 6:
The degenerate symmetry-breaking distortions [S16] are thus divided into the tetragonal (etet) and orthorhombic (eorth) strains, manifesting the β-phase and γ-phase, respectively. These quantities are calculated relative to an undistorted cubic unit cell, a0, which is estimated by taking the cube root of the normalized unit cell volume. It follows that the spontaneous strain components are defined as: 1 = ( − 0 )/ 0 , 2 = ( − 0 )/ 0 and 3 = ( − 0 )/ 0 , where a, b and c are the normalized lattice parameters of the CsPbI2Br-based orthorhombic phase. The separate strain components contributing to the lattice distortions are then given by: ℎ = 2 − 1 and = (2 3 − 2 /√3. A factor of √3 is included here to ensure that the two strains are on the same scale. S42 Figure S26. STEM images of CsPbI2Br, InCl3:CsPbI2Br and CsPb0.95Eu0.05I2Br-based devices recorded after (~10 hours) and 14 days aging.