Stabilized Wide Bandgap MAPbBrxI3–x Perovskite by Enhanced Grain Size and Improved Crystallinity

W2270564514 · 수집 2026-10-04 18:12:23

T11878 Solid-state spectroscopy and crystallography

CoT 라벨링

라벨링 전

PDF 파일

저장된 PDF 없음

기본 정보

DOI 제목 발행년도 저널 URL 페이지 열기 PDF URL PDF 열기 발행기관 피인용수 라이선스

추가 정보 초록

The light instability of CH3NH3PbIxBr3–x is one of the biggest challenges for its application in tandem solar cells. Here we show that an improved crystallinity and grain size of CH3NH3PbIxBr3–x films could stabilize these materials under one sun illumination, improving both the efficiency and stability of the wide-bandgap perovskite solar cells. Methylammonium lead trihalide perovskite (MAPbX3, where MA is methylammonium, and X is a halide)-based solar cells have been intensively investigated recently,1 with the demonstrated certified solar power conversion efficiency (PCE) exceeding 20%.2 To further boost the PCE to beyond the Schockley–Queisser limit,3 tandem structured solar cells have been investigated based on integrating MAPbX3 and low bandgap solar cells.4-8 However, the efficiency of the two-terminal integrated perovskite-silicon tandem cells is still low. The best reported efficiency of 13.7% for this type of tandem cells is far smaller than the individual cells yet, partially due to the limited performance of the mixed-halide perovskite MAPbBrxI3–x solar cell employed in this structure.5 The mixed-halide perovskite MAPbBrxI3–x is still one of the most promising candidates as the wide-bandgap light absorber for the tandem application to match the bandgap of silicon, considering its continuously tunable bandgap from 1.6 eV to 2.3 eV with different bromide incorporation ratio.7, 9, 10 However, the application of MAPbBrxI3–x based solar cells has been reported to confront with one big challenge of intrinsic light instability.11 The MAPbBrxI3–x materials on mesoporous scaffold were shown to be unstable under illumination with a photo-excited phase-separation into two phases, one iodine-rich phase and one iodine-poor phase.11, 12 The lower bandgap phase thus acts as the charge traps, which was hypothesized to be responsible to the severely reduced device open circuit voltage and device PCE for the mixed-halide perovskite devices.6, 13, 14 In this manuscript, we report that the mixed-halide perovskite, MAPbBrxI3–x—with an optical bandgap of 1.70–1.75 eV—are stable under illumination with the improved film microstructures. The application of a non-wetting hole transporting layer was found to increase the grain size dramatically and stabilize the MAPbBrxI3–x grains, which also improved the PCE of the wide bandgap MAPbBrxI3–x perovskite devices to 16.6%, the highest reported value for wide bandgap perovskite solar cells. It has been calculated that a bandgap between 1.70–1.76 eV is optimal for the top cell to get a PCE of >30% in the tandem devices with c-Si in the detailed balance study, which was targeted in this study.15 Interdiffusion method was applied to fabricate phase homogeneous, pin-hole free MAPbBrxI3–x thin films by the reaction of PbI2 with MAIy:MABr1–y mixed-halide organic precursors. The bandgap was tuned by controlling the MABr percentage in the MAIy:MABr1–y organic precursors.16, 17 The best PCE achieved with the interdiffusion formed MAPbBr0.6I1.4 planar-heterojunction structure solar cells was 13.1%, in which the hole transport layer was poly(3,4-ethylenedioxythiophene) polystyrene sulfonate (PEDOT:PSS).10 In this work, we further increase the efficiency of the device based on the similar perovskite composition to 16.6% by employing poly[bis(4-phenyl)(2,4,6-trimethylphenyl)amine] (PTAA) as the hole transport layer, which has a non-wettability to the perovskite precursors.18 Figure 1a shows the current density (J)–voltage (V) curves of the optimized MAPbBrxI3–x cells with a planar-heterojunction structure of ITO/PTAA/MAPbBrxI3–x/[6,6]-phenyl C61-butyric acid methyl ester (PCBM)/C60/9-dimethyl-4,7-diphenyl-1,10-phenanthroline (BCP)/Al,16, 19, 20 in which the device performance parameters for MAPbBr0.5I2.5 (black curves) can be derived as a Jsc of 18.3 mA cm−2, an open circuit voltage (Voc) of 1.16 V, a fill factor (FF) of 78.2%, and a PCE = 16.6%. We further increased Br concentration in methylammonium halide precursor up to 64% (or a MAI:MABr weight ratio of 1:1.3)—the optical bandgap of MAPbBrxI3–x was increased to 1.75 eV. We assign the perovskite with this composition as MAPbBr0.8I2.2 based on the lattice parameter derived from the XRD pattern. The device parameters for this wider-bandgap MAPbBr0.8I2.2 extracted from the red curves in Figure 1a are Voc = 1.21 V, Jsc = 15.8 mA cm−2, FF = 77.9%, and PCE = 14.9%. In addition to the higher PCE achieved in this study, the wide bandgap MAPbBrxI3–x devices studied in this work also showed improved photo-stability, in striking contrast to what was reported previously.9, 11, 12 The steady-state photocurrent output at the maximum power output point of the cells are shown in Figure 1b and the applied bias was labeled, which shows almost constant output over 30 min testing under one sun illumination. The steady-state photocurrents also directly confirmed the steady-state PCEs, or stabilized PCE, which consists with those derived from the J–V curves. In order to find out the mechanism for the improved solar cell PCE and photo-stability for these devices, we fabricated the control cells with PEDOT:PSS as the hole transport layer and all other layers fabricated with exactly the same protocol. It should be noted that no solvent annealing was applied, which is different from our previous study,21 because the perovskite films on top of PTAA do not need a solvent annealing process to form large grains. The typical J–V curves for these structured devices is shown Figure 1c which showed a smaller FF and larger photocurrent-hysteresis when the scanning direction was flipped. As shown in Figure 1d, the photocurrent at the maximum-power output point dropped with illumination time rapidly at first ten minutes, which explains the large hysteresis for this device. Then the photocurrent output saturated at –6 mA cm−2, which agrees with the J–V curve collected after the light-stability test (red curve in Figure 1c). The maximum power output reduced from 14.9% to 3.9% after illumination at one sun for 1000 s. Since the cells in Figure 1a and Figure 1b were fabricated with the same composition perovskite and procedures, the photovoltaic performance difference must be caused by the different hole transport layers applied. PEDOT:PSS has been used in planar heterojunction structured perovskite solar cells at the initial stage of our research;16, 19 however, the Voc of the devices with PEDOT:PSS HTL is generally smaller than best reported devices. PTAA was employed in this work with the initial intention of reducing the Voc loss due to its lower HOMO than that of PEDOT:PSS.13, 22 We did observed increase of the Voc to 1.17 V from 1.01 V for MAPbBr0.5I2.5 devices when replacing PEDOT:PSS with PTAA. The Voc increase of 0.16 V is larger than the work function difference of 0.04 eV between PTAA and PEDOT:PSS measured by KPFM, indicating additional contribution from reduced charge recombination.10 We ascribe the morphology change of the perovskite films to the increased Voc and improved device stability. Another piece of evidence was found to support the microstructure of MAPbBr0.8I2.2 determines its photo-stability by comparing the MAPbBr0.8I2.2 films with different thickness on PTAA HTL. As shown in Figure 3a, the device with a 320 nm MAPbBr0.8I2.2 film, which has large and crystalline grains, was stable under one sun illumination for 33 min, while the devices with thicker MAPbBr0.8I2.2 active layer showed quickly degraded photocurrent under illumination. The optical bandgaps of the MAPbBr0.8I2.2 films with different thicknesses are the same, as shown by the absorption spectra in Figure 3b, which excluded the contribution from the different composition to the different degradation behavior of these films. This can be explained by that fact that excess organic precursor is always applied to deplete PbI2 in interdiffusion methods. Further optical and material structure study confirmed the influence of microstructure on the grain photo-stability. The photo-induced phase separation generates a low-bandgap phase which should cause photocurrent contribution from this new phase. Figure 3c shows the external quatum efficciency (EQE) spectra of the stable cell and unstable one before and after exposed to illumination. As expected, there is no change of EQE from the stable device with 320 nm active layer. It should be noted that the EQE spectra were plotted with logarithm operation. The highest EQE value is 90.5% from the device with 340 nm thick active layer at the wavelength of 374 nm. For the unstable device with 540 nm MAPbBr0.8I2.2, the post-illumination EQE reduced significantly for above bandgap excitation, while increased for below bandgap excitation (720–750 nm), which can be assigned to the photocurrent contribution from the new low bandgap phase. The very low EQE from this new phase indicates the strong trapping effect of it. Another notable finding is the photocurrent dropped more at the shorter wavelength (350 nm to 550 nm) range, which indicates the phase separation predominately occurs in the film region close to the PTAA side. To increase the confidence of the conclusion, we investigated the light stability on the nude MAPbBr0.8I2.2 thin films with X-ray diffraction (XRD) measurement and photoluminescence (PL) measurement, which were previously employed to identify the phase separation in this type of materials.9, 11, 12, 24 Figure 4a compares the XRD pattern with normalized intensity for the stable MAPbBr0.8I2.2 film (320 nm) and the unstable one (530 nm) before exposing them to strong illumination. Although only a small amount of bromide was introduced, a cubic Pm-3m group is identified in the MAPbBr0.8I2.2 XRD pattern with the characteristic peaks of (100), (110), (111), (200), (210). The peak positions are identical, which agrees with our earlier claim that the composition of these films does not change with the MAPbBr0.8I2.2 polycrystalline microstructure. However, the increased (100) peak intensity compared to other peaks for the stable MAPbBr0.8I2.2 suggests that the thinner MAPbBr0.8I2.2 polycrystalline film is better oriented than the thicker one, which gives new microstructure difference between these two compositional identical thin films in addition to the crystallinity and grain size. The random oriented MAPbBr0.6I2.4 generally contains larger-angle grain boundaries, which would enhance the halide migration to assist the phase segregation. By looking at the (200) peak pre- and post-illumination of the two films (Figure 4b,c), we clearly observed a peak splitting from the unstable MAPbBr0.8I2.2 film after it was exposure to three sun illumination by a 532 nm laser beam for 20 min, while the 320 nm film had no change under the same illumination condition. Gaussian fitting was applied to the post-illumination (200) peak (Figure 4d), from which we can see a small new peak appearing at 2θ of 28.5°, the (200) peak position for MAPbBr0.6I2.4—which agrees with the previous study by McGehee's group.11 This MAPbBr0.8I2.2 composition might have the lowest Gibbs free energy, which drives the formation of this phase automatically assisted by photoexcitation. To exclude the possibility that the peak splitting comes from the perovskite decomposition, we kept the illuminated thick films in the dark for two hours, and found the peak splitting can be reversed (blue curve in Figure 4c), which is again consistent to the previous results.11 Finally, the light stability of the MAPbBr0.6I2.4 films with different thickness is also supported by the time integrated PL spectra on the nude perovskite films, as shown in Figure 4e,f. During the five-minute in situ observation of the thick films illuminated by one sun intensity 532 nm laser, a new PL peak originated from the lower bandgap phase gradually appeared after first 50 sec illumination, while the peak of the PL from the thin film showed no change. In summary, we achieved highly efficient and stable planar heterojunction devices based on the wide bandgap MAPbBr0.5I2.5. The light stability of the good wider-bandgap MAPbBr0.8I2.2 device is demonstrated by a steady photocurrent output at maximum power output point over 30 minute under one sun illumination. Microstructure difference between the photo-stable and photo-unstable devices is presented by the cross-section SEM images of the MAPbBr0.6I2.4 active layer: the spatial homogeneous polycrystalline with large sized grains and the stacking layered polycrystalline with small sized grains, respectively. The PL and EQE spectral change, accompanied with XRD pattern comparison between the MAPbBr0.8I2.2 thin films with two different microstructures, indicate the enhanced crystallinity and grain size are favorable to retain the homogeneous phase for the mixed halide perovskite during the photoexcitation, thus maintain a stable photocurrent output under the device working condition. The phase instability for mixed halide perovskite was studied with the MAPbBrxI3–x infiltrate into porous TiO2 scaffold,10 whose microstructure is comparable to the thin film formed on PEDOT:PSS surface or with concentrated precursors in this manuscript. We demonstrate that the microstructure and crystallinity of MAPbBrxI3–x are crucial to achieve the stable wide bandgap perovskite solar cells. This explains why some large bandgap MAPbBr0.8I2.2 built on mesoporous scaffold device were not stable. Further studies on the role of the grain boundary's area and grain orientation in phase separation of mixed halide perovskite may contribute to a deeper understanding. This work demonstrates the potential of mixed halide perovskite to stay a reliable homogeneous phase in photovoltaic working condition as the wide bandgap light absorber in tandem application. PbI2 and MAI1–xBrx Precursor Preparation: MAI was synthesized using the method described by Michael M. Lee et al. MABr was synthesized by the reaction of methylamine with a concentrated aqueous solution of hydrobromic acid (23.5 mL, 36.5 wt% in water, Alfa Aesar) at 0 °C for 2 h with constant stirring under nitrogen atmosphere followed by a crystalized, purification and dry process which was the same as the preparation of MAI. MAI1–xBrx precursor was prepared by mixing MAI and MABr in 2-propanol for 1:1.3 weight blend ratio and 62 mg ml−1 concentration. PbI2 precursor was prepared by dissolving PbI2 in DMF with the different concentration 500 mg ml−1, 600 mg ml−1 and 700 mg ml−1. Film and Photovoltaic Device Fabrication: First, hole transport layer was prepared by spin-coating 0.25 wt% PTAA (toluene) solution doped with 1 wt% F4-TCNQ at 4000 rpm for 25 sec, and the as-prepared film thermally annealed at 100 ºC for 10 min. We have already demonstrated that F4-TCNQ is a suitable dopant to PTAA.25 To fabricate 320 nm, 480 nm, and 550 nm thick MAPbI1–xBrx films, 500 mg ml−1, 600 mg ml−1, and 700 mg ml−1 PbI2 DMF precursor solutions were spun coated on the hole transport layer at 6000 rpm for 35 sec. Then, the as-prepared PbI2 film was dried in a hotplate at 100 °C for 5 min, and followed by spin coating the 62 mg ml−1 MAI1–xBrx precursor on top of the PbI2 layer at 6000 rpm for 35 sec with the subsequent 75 °C 15 min thermal annealing. This process allows the bromide inter-diffuse into the perovskite structure. Then we increased the annealing temperature to 100 °C, lasting 90 min before cooling down to the room temperature. For photovoltaic devices, 2 wt% solution of PCBM in DCB was spun onto the annealed perovskite film and followed by additional 60-min annealing. Then the device was completed by sequence depositing 20 nm C60, 8 nm BCP and 100 nm Al. Device Characterization: The photocurrent curves were measured under simulated AM 1.5G irradiation (100 mW cm−2) using a Xenon-lamp-based solar simulator (Oriel 67005, 150 W Solar Simulator). A Schott visible-color glass-filtered (KG5 color-filtered) Si diode (Hamamatsu S1133) was used to calibrate the light intensity before photocurrent measurement. The device area was 7.5 mm2. EQE was measured with a Newport QE measurement kit by focusing a monochromatic beam of light onto the devices. Film Characterization: XRD measurements were performed with the PANalytical Empyrean Diffractometer with Bragg–Brentano parafocusing geometry, the 3 kW Cu Kα source, and the PIXcel 3D detector. The scan rate is 0.02 s per step and 0.5 s per step with an angular range of 10–60. The single path absorption was measured using an Evolution 201 UV–visible spectrometer (Thermo Scientific) with the scan rate 1 nm per step and 0.5 s per step in the range of 350–900 nm. The PL spectrum was measured by iHR320 Photoluminescence Spectroscopy at room temperature. A 532 nm green laser with an intensity of 100 mW cm−2 from Laserglow Technologies was used as excitation source in PL measurement. PL integration time is 5 secs and the slit is 0.12 mm. PL intensity was collected every 10 sec continuously for 30 min. The cross-section SEM images were taken from the Quanta 200 FEG Environmental Scanning Electron Microscope (ESEM) using a field-emission gun (FEG) electron source to scan the gold coated cross-section area morphology. We are thankful for financial support from Energy Efficiency and Renewable Energy (EERE) SunShot Initiative at Department of Energy under Award DE-EE0006709.

본문

PDF 본문 자동 추출 실패: [download] HTTP 403 Forbidden (2026-10-05 00:30:45)

활용가능성

이력 0건

아직 편집 이력이 없습니다. 저장할 때마다 이전 내용이 여기에 백업됩니다.

Stabilized Wide Bandgap MAPbBr<i><sub>x</sub></i>I<sub>3–<i>x</i></sub> Perovskite by Enhanced Grain Size and Improved Crystallinity

CoT 라벨링

PDF 파일

이력 0건