A hole-transport-material-free planar solar cell of cesium lead mixed halide perovskite(CsPbIBr2) is deposited by dual source thermal evaporation for the first time, achieving an efficiency of 4.7%. The addition of iodine into the bromide lowers the band-gap resulting in wider solar spectrum absorption. Compared to the hybrid halide perovskites, CsPbIBr2demonstrates better thermal stability.
Q. Ma, Dr. S. Huang, Dr. X. Wen, Prof. M. A. Green, Dr. A. W. Y. Ho-Baillie
The emergence of organic–inorganic hybrid halides perovskite solar cells has generated enormous interests in the photovoltaic research community. Due to their excellent optical absorption, good carrier mobility and lifetime, the power conversion efficiency (PCE) of organic–inorganic hybrid halide perovskite solar cells has been rapidly improved to the independently con- firmed 20% over a short period of time.[1–5] While the state-of- the-art cells have methylammonium (MA) and formamidinium (FA) as the organic cation, inorganic cations, such as Cs and Rb are yet to be further explored as substitutes for MA or FA possibly from a device stability point of view. In particular, Cs- containing inorganic perovskites have demonstrated high electron and hole mobility: up to 2300 cm2 V s−1 and 320 cm2 V s−1, respectively. They also have the potential of better thermal and light stability compared to MA- or FA-based perovskites. Kulbak et al. have successfully demonstrated CsPbBr3 perovskite solar cells using two-step sequential solution process with over 5% PCE. More recently, Snaith et al. have reported a spin-coated stable CsPbI3 film by adding HI into the precursor solution and demonstrated a CsPbI3 perovskite solar cell with a PCE of 2.9%. These inorganic Cs perovskites are stable at 250 °C and 330 °C as reported in these work while the organic containing perovskites degrade.[7,8]
Planar heterojunction organic–inorganic halide perovskite solar cell has been proven to be an efficient and simple solar cell structure and various perovskite deposition techniques have been developed to achieve such a structure effectively.[9–12] The techniques include solution processes via spin-coating of pre- cursors in one step or in sequential steps,[11,12] vapour-assisted deposition, gas-assisted solution process, and dual source thermal evaporation.[15,16] The evaporation of methylammonium iodide (MAI) and PbCl2 precursors simultaneously has been demonstrated to be effective in achieving uniform perovskite films over a large area. In addition, this process does not involve solvents (leaving no solvent residue in the film), and is also suitable for insoluble materials deposition. However, the deposition of organic MAI is vapour-induced, i.e., nondirectional due to its relatively high vapour partial pressure. This makes the accurate control of the rate of PbCl2 deposition difficult as its rate reading is inevitably affected by the MAI sublimation during co-deposition.
Another advantage of perovskite solar cells is the ability to tune the band gap by mixing the halide anions. This applies to Cs mixed halide[18,19] as well as MA or FA lead mixed halides, enabling optimisation of solar spectrum absorption for both single-junction and tandem solar cells.
In this study, we report a dual source thermal evaporation process to deposit mixed halide CsPbIBr2 perovskite absorber with a bandgap of 2.05 eV. This material has a potential to be used in a three-junction tandem as the quality of the material and therefore voltage of the device further improves. Inorganic CsI and PbBr2 precursors are simultaneously evaporated onto a compact TiO2 layer (c-TiO2) on FTO glass substrates. Post- annealing is carried out on a hot plate in a glove box to enable the full crystallization of the CsPbIBr2 perovskite. A series of experiments investigating the effect of postanneal conditions on the crystalline structure is conducted in this work. Films achieve best quality in terms of crystallinity, thickness uniformity, and grain size uniformity when the samples are annealed at 250° for 10 min. Aiming at a simple architecture and an organic- component-free device, we fabricated a hole transport mate- rial (HTM) free planar Glass/FTO/c-TiO2/CsPbIBr2/Au solar cell, first of its kind, with a PCE of 4.7%, a short-circuit current density (JSC) of 8.7 mA cm−2, an open-circuit voltage (VOC) of 959 mV, and a fill factor (FF) of 56% under reverse scan, while PCE=3.7%,JSC =8.7mAcm−2,VOC =818mVandFF=52% under forward scan.
As described in the Experimental Section, the CsPbIBr2 sam- ples were prepared by evaporating the same molar quantity of CsI and PbBr2 onto the substrates. The chemical composition of the samples was evaluated by X-ray photoelectron spectros- copy (XPS). The atomic ratio of Pb/Cs and Br/I was estimated to be 1.1 and 2.3, respectively, which is in good agreement with the CsPbIBr2 composition. The XPS spectra are shown in Figure S1 in the Supporting Information. An Energy-dispersive X-ray spectroscopy (EDS) measurement at 15 kV was also car- ried out by a 20 μm line scan of the CsPbIBr2 film showing the atomic ratios of Pb/Cs and Br/I to be 1.2 and 1.94, respectively. The EDS spectra are shown in Figure S2 in the Supporting Information. One reason for the deviation between EDS and XPS results is the difference in accuracy between the measure- ments (XPS, ±5%, EDS, ±15%). It is also noted that the XPS carried out measures of the elemental composition of about 10 nm in depth from the surface. For bulk measurement, Ar ion etching will be required causing damage to the CsPbIBr2 film. Given the uniform column grains formed from the bottom to the surface of the CsPbIBr2 film as shown in Figure 5a, the atomic ratios of Pb/Cs and Br/I measured by XPS will be a reasonable representation for the entire film. On the other hand, the depth of EDS measurement can be about 1000 nm, providing good bulk information of the elements.
Figure 1 shows the X-ray diffraction (XRD) patterns of the CsPbIBr2 films deposited via dual source thermal evaporation on c-TiO2/ FTO glass substrates at 20 °C or 75 °C fol- lowed by a postanneal at 100 °C or 250 °C for 10 min. The main diffraction peaks at 15.05° 21.35°, and 30.25° correspond to the (110), (200), and (220) planes of the CsPbIBr2 per- ovskite orthorhombic phase. The perovs- kite crystallinity improves with the increase of substrate temperature during deposition as well as the post-annealing temperature.
The morphology of the perovskite films is studied by top view scanning electron micros- copy (SEM) as shown in Figure 2. The films without post-annealing show small crystal grains in Figure 2a,c. After post-annealing, the crystal grains grow larger and are more compact. The film deposited at 75 °C sub- strate temperature gives larger crystalline grains than that deposited at 20 °C. When the 75 °C deposited film is annealed at 250 °C, its grains grow as large as 500–1000 nm in size.
The optical transmission and reflection spectra of the CsPbIBr2 films were meas- ured using a Perkin Elmer LAMBDA 1050 UV-VIS-NIR spectrophotometer. The absorp- tion coefficient of the samples was calculated using α = t−1 ln((1−R)/T), where R and T are the reflection and transmission and t is the thickness of the samples. All of the samples annealed at different temperatures show similar absorption coefficients, over 5 × 104 cm−1 in the absorption range of 250–580 nm, which are very similar to that of CH3NH3PbI3. Figure 3a shows the absorp- tion coefficient of the sample deposited at 75 °C and annealed at 250 °C. The onset of optical bandgap edge transition is 604 nm (2.05 eV). Compared with CsPbI3 (Eg ≈1.73 eV) and CsPbBr3 (Eg ≈ 2.25 eV), the optical bandgap of the mixed halide CsPbIBr2 shows a linear relationship with the content of Br in the mixed halide Cs perovskites which is in good agreement with the Vegard’s law, see Figure S3 (Sup- porting Information).
The photoluminescence (PL) spectrum of the CsPbIBr2 shows a peak at 2.00 eV (620 nm), which is very close to the optical band edge and does not undergo shift under light soaking, see Figure S4 (Supporting Information). In the case of CH3NH3Pb(BrxI1−x)3, low energy PL features appear upon light soaking observed by Hoke et al. suggesting photo-induced halide segregation resulting in a reduction in the electronic band gap and quasi-Fermi level splitting. This causes the observed red- shift in PL and limits the achievable voltage as can be seen in MA lead mixed halide devices.[22–24] This lack of PL red-shift and halide segregation in Cs lead mixed halide perovskite is an advantage over organic lead mixed halide perovskite due to higher voltage potential and the ability to tune the band gap of the Cs perovskites by alloying different halides into the structure as also reported by Akkerman et al. and Song et al. in[18,19] respectively. Sharma et al. also reported the absence of segregation for the mixed halide with composition similar to CsPbIBr2.
Figure 3b shows the PL image of the perovskite film excited at 470 nm and detected at 620 nm. The grain size detected by the imaging is very similar to that of SEM images. The small grains that are brighter indicate higher PL efficiency while the larger and typically dimmer and dark grains have lower PL efficiency. The typical PL decay traces of a small-bright grain; a large- dim grain; and a dark grain were measured using time correlated single photon counting (TCSPC) technique as shown in Figure 3c. Using biexponential decay function y=A1 exp(−t/τ1)+A2 exp(−t/τ2)s, the PL decay trace of the large grains is fitted to determine the decay times of the fast (τ1) and slow (τ2) components for each region, while the PL decay trace of the small gains is fitted with dingle exponential function with one decay time component, as tabulated in Table 1. The presence of the fast component in the PL decay for the dim (and dark) regions of the CsPbIBr2 film indicates the presence of defect trapping. Otherwise, a typical lifetime of 9.35 and 17.7 ns found for the CsPbIBr2 film corresponds to the carrier recombination time. This lifetime is comparable with that of CH3NH3PbBr3 film on c-TiO2 (3.5 ns).
Preliminary thermal stability tests were also performed on CsPbIBr2 films which involves heat treatment at 200 °C on a hot plate in a glove box. The 6 h interval results are shown in Figure 4. The XRD patterns in Figure 4a show no detectable cells (850 mV), the higher VOC from our mixed halide devices demonstrates the potential of tunable and larger open-circuit voltage by incorporating Br content. In addition, the short-circuit current of the cell can be further improved by increasing the thickness of CsP- bIBr2 absorber to be beyond the current 190 nm.
Find this complete article in at Advanced Energy Materials
Q. Ma, Dr. S. Huang, Dr. X. Wen, Prof. M. A. Green, Dr. A. W. Y. Ho-Baillie
Australian Centre for Advanced Photovoltaics, University of New South Wales
Contributing to this perskovite solar cell research was specialised instrumentation such as:
Vigor Thin Film Glovebox Deposition System supplied by John Morris Scientific
Find out more about Thin Film Deposition technology
Contact our team of Vacuum experts
Freecall Ph. AUS 1800 251 799 and NZ 0800 651 700
NEW website: www.johnmorrisgroup.com/AU/Vacuum