Pengfei Wanga,
Nianyao Chaia,
Chang Wanga,
Jingchen Huaa,
Fuzhi Huanga,
Yong Penga,
Jie Zhonga,
Zhiliang Ku*ab and
Yi-bing Chengac
aState Key Laboratory of Advanced Technologies for Materials Synthesis and Processing, International School of Materials Science and Engineering, Wuhan University of Technology, 122 Luoshi Road, Wuhan, Hubei, P. R. China. E-mail: zhiliang.ku@whut.edu.cn
bHubei Key Laboratory of Low Dimensional Optoelectronic Material and Devices, Hubei University of Arts and Science, 296 Longzhong Road, Xiangyang, Hubei Province, P. R. China
cDepartment of Materials Science and Engineering, Monash University, Wellington Road, Clayton, VIC 3800, Australia
First published on 16th April 2019
Despite the impressive photovoltaic performance with a power conversion efficiency beyond 23%, perovskite solar cells (PSCs) suffer from poor long-term stability, failing by far the market requirements. Although many efforts have been made towards improving the stability of PSCs, the thermal stability of PSCs with CH3NH3PbI3 as a perovskite and organic hole-transport material (HTM) remains a challenge. In this study, we employed the thermally stable (NH2)2CHPbI3 (FAPbI3) as the light absorber for the carbon-based and HTM-free PSCs, which can be fabricated by screen printing. By introducing a certain amount of CsBr (10%) into PbI2, we obtained a phase-stable CsxFA1−xPbBrxI3−x perovskite by a “two-step” method and improved the device power conversion efficiency from 10.81% to 14.14%. Moreover, the as-prepared PSCs with mixed-cation perovskite showed an excellent long-term stability under constant heat (85 °C) and thermal cycling (−30 °C to 85 °C) conditions. These thermally stable and fully-printable PSCs would be of great significance for the development of low-cost photovoltaics.
The typical carbon-based PSCs mainly consist of three mesoporous inorganic layers (TiO2, ZrO2 and carbon), which can be prepared by screen printing. Obviously, the non-use of organic HTM and Au electrode makes the carbon-based PSCs not only low-cost but also stable. Since the first report of carbon-based PSCs,14 more and more researchers have shown their interest on such PSCs with a special mesoporous structure.3,6,15–20 However, most of them have still focused on how to enhance the PCE of the device. Several strategies including solvent engineering,18 post-treatments17 and adding the halide component19,21–23 into the perovskite precursor have been developed to optimize the crystallization of the MAPbI3 perovskite. Up to now, the champion PCE of carbon-based PSCs with the MAPbI3 perovskite has exceeded 16%.19 Although there is an obvious gap in terms of the PCE value between carbon-based PSCs and the traditional one with Au electrode, the carbon-based PSCs actually have huge industrialization potential because of their easy scaling-up character.20 However, as we mentioned above, the MAPbI3 perovskite has intrinsic stability problems. Therefore, introducing a thermally stable FAPbI3 perovskite into the carbon-based PSCs would be of great significance.
Herein, we comprehensively investigated the FAPbI3 perovskite as a light absorber for high-performance carbon-based PSCs. By introducing a certain amount of CsBr into the PbI2 precursor, we found that the phase transition of the FAPbI3 perovskite from the α phase to the δ phase can be successfully restrained. Thus, the CsxFA1−xPbBrxI3−x-based device achieved a PCE value up to 14.14%, which is much higher than that of the FAPbI3 version (10.81%). Importantly, the mixed-cation CsxFA1−xPbBrxI3−x perovskite showed a pleasant thermal stability both under a constant temperature of 85 °C and thermal cycling condition (−30 °C to 85 °C). To the best of our knowledge, this is the first study on the thermal cycling performance of the carbon-based PSCs, which would provide a certain reference value for their practical application.
Fig. 1 (a) The schematic of a fully printable HTM-free mesoscopic PSCs with carbon CE; (b) the schematic of a two-step sequential deposition method. |
The microstructures of the as-prepared devices can be observed using scanning electron microscopy (SEM) images of the cross-section. We can clearly find that each layer of the device showed well-defined boundaries and a uniform thickness (Fig. 2a). Through the EDS mapping, the triple layers from the top to the bottom can be identified as carbon, ZrO2, and TiO2. For the perovskite, Pb, I, Br, and Cs have a uniform distribution in the carbon/ZrO2/TiO2 layers, indicating a good filling of perovskite in the mesopores.
Fig. 2 (a) The cross-sectional view of the SEM image of the carbon-based PSCs; (b) the corresponding EDS mapping. |
To study the influence of the Cs cation and Br ion in the crystal structure of the FAPbI3 perovskite, we introduced CsI, PbBr2, and CsBr with different mole ratios into the PbI2 solution. X-ray diffraction (XRD) measurements were used to identify the crystal structure of the as-prepared perovskite samples. The pure FAPbI3 exhibited a series of diffraction peaks at 2θ = 13.8°, 20°, 24.2°, 28°, 31°, 40°, and 42.7°, corresponding to the (111), (120), (021), (222), (231), (240), and (333) crystal planes of the α-FAPbI3 perovskite (Fig. 3). Moreover, a small diffraction peak at 2θ = 11.7° can be detected, indicating the existence of δ-FAPbI3.25 With the increase in Br, the δ-FAPbI3 showed a declining trend. Moreover, when the Br ratio came to 10%, there was no obvious peak of δ-FAPbI3. Similarly, the introduction of Cs could also restrain the generation of δ-FAPbI3.
To further investigate the effect of Br and Cs on the stability of the FAPbI3 perovskite, we stored all the samples in ambient air at 25 °C at a relative humidity (RH) of 50% and measured their XRD patterns every few days (7 days) to record the changes. As shown in Fig. S1a–h,† after 7 days, the peak of δ-FAPbI3 emerged from all of the samples except the one with 10% of CsBr. Moreover, after 14 days, the FAPbI3–10% CsBr sample still maintained the black α phase, while other samples had an obvious increase in the δ phase.
Hence, we can conclude that though both Cs and Br could inhibit the transformation of the FAPbI3 perovskite from α phase to δ phase, only the coexistence of Cs and Br could actually enhance the stability of the FAPbI3 perovskite (Fig. 3).
The optical properties of the FAPbI3-based perovskite films with different content of Br and Cs were measured by the UV-vis spectra and steady-state photoluminescence (PL) spectra. As shown in Fig. 4a, with the increase in the content of Br, the absorption edge of the samples had an obvious blue-shift, which is in good agreement with the theoretical calculations.7 Moreover, all the samples exhibited a slight enhancement of light absorption from 650 nm to 750 nm. For the samples with Cs, a slight blue-shift can be observed, manifesting the limited effect of Cs on the band gap of the FAPbI3 perovskite. Unlike Br, the introduction of Cs could significantly enhance the light absorption of the FAPbI3 perovskite from 650 nm to 750 nm. Note that the FAPbI3 perovskite with both Br and Cs showed the strongest absorbance from 650 nm to 750 nm among all the samples (Fig. 4c). Steady-state PL spectra measurements were also obtained for the perovskite samples loaded in the ZrO2/glass substrate. The emission peak of the Br-based FAPbI3 perovskite shifted from 807 nm to 798 nm as the Br ratio increased from 5% to 15% (Fig. 4b). Furthermore, the blue shifting of the Cs-based FAPbI3 perovskite was more inconspicuous (from 809 nm to 805 nm as the Cs ratio increased from 5% to 15%) than that of Br (Fig. 4d). These results are in good agreement with the UV-vis results.
Fig. 4 UV-vis and PL spectra of the FAPbI3-based films with different amounts of (a and b) Br and (c and d) Cs. |
Carbon-based PSCs with these different FAPbI3-based perovskites were measured under standard AM 1.5 illumination. After optimization (see Tables S1 and S2†), we found that the device with 10% Br or 10% Cs possessed higher PCE than that with other ratios. To ensure that the improvement of PCE is repeatable, four batches of PSCs (each consists of 15 devices) were fabricated using pristine FAPbI3, FAPbI3 with 10% of Br, FAPbI3 with 10% of Cs and FAPbI3 with 10% of Br–Cs, respectively. The detailed J–V performances are presented in Fig. S2,† and we found that the average PCE value of the devices was improved by introducing Br and Cs into FAPbI3.
To find out how Br and Cs affect the performance of the FAPbI3-based PSCs, we compared the champion devices with 10% Br, 10% Cs and 10% Cs–Br. As shown in Fig. 5a, the device with pristine FAPbI3 perovskite possessed an open circuit voltage (Voc) of 929 mV, a short circuit current density (Jsc) of 20.53 mA cm−2 and a fill factor (FF) of 0.567, yielding an overall PCE of 10.81%. By introducing 10% of Br, the FAPbBr0.1I2.9-device exhibited a slightly higher Voc (958 mV) and FF (0.593) than that of the pristine one, which resulted in a PCE of 11.53%. The Cs0.1FA0.9PbI3-device showed similar J–V parameters with the FAPbBr0.1I2.9-based one. Interestingly, the Cs0.1FA0.9PbBr0.1I2.9-device showed obvious enhancement in Voc (1018 mV), Jsc (22.16 mA cm−2) and FF (0.627) in comparison to the other device and exhibited a champion PCE of 14.14% (see Table 1 for the detail parameters). The steady current output of the device was maintained at 18.5 mA cm−2 for 2 min (see Fig. S3†). External quantum efficiency (EQE) measurement was used to verify the improvement of Jsc. As shown in Fig. 5b, the Cs0.1FA0.9PbBr0.1I2.9-device had a much higher EQE value than other ones in the range from 500 nm to 800 nm. The integrated Jsc of Cs0.1FA0.9PbBr0.1I2.9-device reached 22.0 mA cm−2, which is highly consistent with the J–V results.
Composition | Voc (mV) | Jsc (mA cm−2) | FF | PCE(%) |
---|---|---|---|---|
Pristine | 929 | 20.53 | 0.567 | 10.81 |
10% Br | 958 | 20.29 | 0.593 | 11.53 |
10% Cs | 970 | 20.91 | 0.578 | 11.72 |
10% Br–Cs | 1018 | 22.16 | 0.627 | 14.14 |
Time-resolved photoluminescence (TRPL) was performed with an excitation wavelength of 780 nm to investigate the recombination behaviour between the electron and hole in the FAPbI3-based perovskite. In comparison to the pristine FAPbI3 perovskite, other samples with Br or Cs showed a longer electron life time (see Table S3† for the detail fitting parameters), which can be attributed to the better crystal quality. We deduced that the trap state density in pristine FAPbI3 is pretty high because of the phase transition. By introducing Br and Cs into the crystal, the structure of the FAPbI3 perovskite became more stable, and as a result, the trap state density was reduced. Apparently, the Cs0.1FA0.9PbBr0.1I2.9 crystal possessed the best phase stability and the lowest trap state density. Moreover, the J–V hysteresis of the Cs0.1FA0.9PbBr0.1I2.9-device remained at a low level (Fig. 5d), which also can provide evidence for our deduction.26
The long-term thermal stability measurements of FAPbI3-based PSCs were performed in an environmental chamber (Weiss SC3 600 MHG). At a constant temperature of 85 °C and 20% relative humidity (RH) condition, the device with the pristine FAPbI3 perovskite lost >60% (from 100% to 35.8%) of the initial PCE value after 500 h (Fig. 6a). However, the device with the Cs0.1FA0.9PbBr0.1I2.9 perovskite showed a pleasant long-term stability and maintained 90% of its initial PCE value (from 100% to 90.2%). At a RH 50% condition, the Cs0.1FA0.9PbBr0.1I2.9-device also showed a strong moisture resistance (see Fig. S4†). To further evaluate the long-term stability of the device under practical thermal stress, we also performed the thermal cycling test on the devices. First, the encapsulated devices were loaded in the environmental chamber and stabilized at 25 °C. Then, the chamber was heated up to 85 °C for 25 min, and chilled down to −30 °C for 25 min. After 100 cycles, the Cs0.1FA0.9PbBr0.1I2.9 device had no obvious decay in the PCE value, manifesting its superior stability under thermal stress.
Fig. 6 The thermal stability measurements of FAPbI3-based PSCs at (a) a constant temperature of 85 °C and 20% RH in dark condition; (b) thermal cycling from −30 °C to 85 °C. |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ra00043g |
This journal is © The Royal Society of Chemistry 2019 |