Wanjiang
Wang
a,
Linsong
Hou
a,
Haihua
Hu
b,
Binbin
Chang
a,
Yuqi
Yuan
a,
Ping
Lin
a,
Peng
Wang
a,
Xiaoping
Wu
a,
Xuegong
Yu
c,
Lingbo
Xu
*a and
Can
Cui
*a
aKey Laboratory of Optical Field Manipulation of Zhejiang Province, Department of Physics, School of Science, Zhejiang Sci-Tech University, Hangzhou 310018, People's Republic of China. E-mail: xlb@zstu.edu.cn; cancui@zstu.edu.cn
bHangzhou City University, Hangzhou 310015, People's Republic of China
cState Key Laboratory of Silicon Materials, School of Materials Science and Engineering, Zhejiang University, Hangzhou 310027, People's Republic of China
First published on 28th March 2024
Cs2AgBiBr6 double perovskites are considered potential alternatives to lead-based perovskites for solar cell applications owing to their advantages of stability and non-toxicity. However, the disordered arrangement of Ag/Bi octahedra leads to the formation of self-trapped excitons (STEs) and hinders carrier transport, which is a severe problem in planar solar cells due to inferior charge extraction to mesoporous solar cells. Herein, we employ ethoxylated polyethylenimine (PEIE) to regulate the growth of Cs2AgBiBr6 films by sandwiching it between an SnO2 electron transport layer and a perovskite layer. On the one hand, the introduction of PEIE delays the growth of Cs2AgBiBr6 and improves the ordering extent of the Ag/Bi octahedra. This inhibits the formation of STEs and facilitates carrier transport, which profoundly boosts the short-circuit current density. On the other hand, PEIE modulates the energy level of the SnO2 electron transport layer to match that of Cs2AgBiBr6, which reduces the open-circuit voltage loss. Consequently, the power conversion efficiency (PCE) of the planar solar cell increases by 250% after PEIE modification from 0.59% to 1.52%. In addition, the PEIE-decorated device maintains over 90% of its initial PCE for 30 days under ambient conditions, demonstrating good environmental stability.
Cs2AgBiBr6 is a typical representative of double perovskites, with a long photogenerated carrier lifetime (more than 1 μs), a small effective carrier mass, strong optical absorption in the high-energy part of the solar spectrum, etc.19,20 The research of applying Cs2AgBiBr6 on solar cells has lasted for years; however, the PCE for Cs2AgBiBr6 solar cells is typically lower than that of lead-based perovskites.21,22 This is partially attributed to the wide indirect bandgap of Cs2AgBiBr6 exceeding 2.0 eV, which limits the absorption of solar illumination.23 To solve this problem, several methods such as hydrogenation to reduce the bandgap or introduction of sensitizers to enhance light absorption have been adopted.24–26 Furthermore, the poor film quality of Cs2AgBiBr6 limits its PCE.27–30 Due to the underdeveloped fabrication methods of Cs2AgBiBr6, the defect density in Cs2AgBiBr6 is much higher than that in lead-based perovskites.31 Particularly, the cation anti-site defects would enhance the electron–phonon coupling and thus limit charge transport in Cs2AgBiBr6.32–34 In the ideal Cs2AgBiBr6 double perovskite structure, the [AgBr6]5− and [BiBr6]3− octahedra are strictly arranged alternately.35 However, in practical preparations, the octahedra disorder is energetically more favorable.36,37 This disorder of the b-site cations increases the electrostatic repulsion of the two adjacent octahedra, resulting in lattice distortion of the double perovskite. Photogenerated carriers are bound in distorted octahedra under the strong electron–phonon coupling, forming self-trapped excitons (STEs) that impede charge transport.38 To solve this problem, Abhishek Maiti et al.39 introduced phenethylammonium bromide (PEABr) into a Cs2AgBiBr6 precursor to obtain ordered films because the phenyl and ammonium groups of PEA+ can selectively coordinate with Bi3+ and Ag+, respectively. Furthermore, the anti-site defects of b-site cations are also present in other double perovskites such as double perovskite oxides.40–42 In these systems, researchers have developed a polymer-assisted deposition method (usually polyethylenimine, PEI) to improve the ordering of b-site occupations by slowing down the growth process close to thermodynamic equilibrium conditions.43–45 These successful cases demonstrate that the ordering of b-site cations could be modulated by manipulating the growth dynamics of perovskite films, which has not gained enough attention yet in the Cs2AgBiBr6 system and deserves further investigation.
Taking into account the limited charge transport capability in bulk films, most Cs2AgBiBr6 solar cells are based on a mesoporous TiO2 structure due to the abundant contact area, which reduces charge transport lengths and facilitates charge extraction.46 Although planar-type solar cells with SnO2 electron transport layer (ETL) have the advantages of a simplified fabrication process and low processing temperature, their application in Cs2AgBiBr6 systems is not successful and leads to a lower PCE than that of a mesoporous structure.47 Thus, it is highly desirable to improve the charge transport of Cs2AgBiBr6 films in planar-type solar cells. Herein, we report a different strategy of interface modification to fabricate ordered Cs2AgBiBr6 double perovskites on an SnO2 ETL for efficient planar solar cells, apart from the above-mentioned additive engineering methods. Interface modification is a widely used approach for modulating the growth of perovskite films. For example, Zhuang et al. modified the perovskite/SnO2 interface with guanidine salts of different anions, which promoted the crystallization of perovskites and improved the photovoltaic performance of PSCs.48 Gong et al. used a series of potassium salts to modify the SnO2/perovskite interface and found that the crystallization kinetics of perovskites are controlled by a compromise between the chemical interaction strength and wettability of substrates.49 However, interface modification engineering has not drawn enough attention in the Cs2AgBiBr6 system. In this work, 80% ethoxylated polyethylenimine (PEIE) was introduced between the SnO2 ETL and the perovskite layer for interfacial modulation. PEIE is a derivative of PEI with about 80% of the primary and secondary amines ethoxylated. Similar to the widely used PEI in the above-mentioned polymer-assisted deposition method, PEIE interacts with perovskite precursors to modulate film growth. PEIE has better solution processability and film formation properties than PEI, which widen the fabrication window.50,51 Moreover, PEIE is thermally more stable than PEI and is more adaptable to high-temperature processing according to Kim's report, which makes it suitable for the Cs2AgBiBr6 system.52 The introduction of PEIE delays the growth of Cs2AgBiBr6 grains, resulting in an ordered and high-quality Cs2AgBiBr6 film. Besides, PEIE modulates the energy band of SnO2 to match that of Cs2AgBiBr6. The PEIE-decorated Cs2AgBiBr6 solar cells show enhanced carrier transport and reduced open-circuit voltage (VOC) loss, with a profound PCE improvement as high as 250%.
We further investigated the structural properties and cation occupancy ordering of Cs2AgBiBr6 films through X-ray diffraction (XRD) tests. Fig. 2a shows XRD patterns of the control and PEIE samples. With the introduction of PEIE, the diffraction peak intensity of the Cs2AgBiBr6 film was enhanced. Meanwhile, the full width at half maximum (FWHM) of the (222) plane is 0.125° in the PEIE sample, which is smaller than that of the control sample (0.138°), as shown in Fig. 2b. The stronger diffraction peak and smaller FWHM indicate that PEIE enhances the crystallization of Cs2AgBiBr6 double perovskites, which is consistent with the SEM results. Moreover, the analysis of the XRD patterns is a straightforward and effective method for estimating the ordering of b-site cations in double perovskite films.57 In the case where the [AgBr6]5− and [BiBr6]3− octahedra are completely disordered, the (111) peak would disappear according to the extinction law.39 Consequently, the intensity ratio between the (111) and (022) planes (I111/I022) quantifies the ordering extent of the [AgBr6]5− and [BiBr6]3− octahedra, where a higher ratio indicates a more ordered cation arrangement.37,39,58 As shown in Fig. 2b, the I111/I022 value of the PEIE sample is 0.413, higher than that of the control sample (0.321), indicating a noticeably increased ordering of the [AgBr6]5− and [BiBr6]3− octahedra.
The variation in the film quality and cation ordering extent is also reflected by the optical properties. Fig. 2c shows the ultraviolet-visible absorption spectra of the control and PEIE samples. The spectra exhibit an absorption edge at around 600 nm with a characteristic peak at about 438 nm.59,60 The origin of this sharp characteristic peak is due to the direct gap exciton, according to Wright's review.61 The absorbance of the PEIE sample is slightly higher than that of the control sample, which is attributed to the enhanced crystallinity and improved film quality. We also calculated the indirect bandgap of Cs2AgBiBr6 films based on the Tauc plots, as shown in Fig. S2 (ESI†). The indirect bandgap of the Cs2AgBiBr6 film shows a slight increase from 2.39 to 2.45 eV, which is related to the enhanced ordering of the [AgBr6]5− and [BiBr6]3− octahedra, according to the previous report.36Fig. 2d shows the steady-state photoluminescence (PL) spectra of Cs2AgBiBr6 films. It is worth noting that the Cs2AgBiBr6 films deposit on glass substrates (with or without PEIE) to exclude the charge extraction effect of SnO2. The spectrum of the control sample is broad with a peak at about 636 nm. Such a broad peak with an FWHM of 140 nm and a large Stokes shift (defined as the difference between the energy of the absorption peak at 438 nm and the emission peak) of 881 meV (as listed in Table 1) indicate the strong electron–phonon coupling in the control sample.61 The PEIE sample shows a blue-shifted peak at about 604 nm with decreased intensity. Accordingly, the FWHM decreases to 131 nm and the Stokes shift decreases to 778 meV, as shown in Table 1. This indicates the weakened electron–phonon coupling in the PEIE sample due to increased ordering of the [AgBr6]5− and [BiBr6]3− octahedra, which is consistent with the XRD results.
Sample | Band gap (eV) | PL peak (nm) | FWHM (nm) | Stokes shift (meV) | Huang–Rhys factor (S) |
---|---|---|---|---|---|
Control | 2.39 | 636 | 140 | 881 | 20.4 |
PEIE | 2.45 | 604 | 131 | 778 | 18.1 |
To gain deeper insights into the electron–phonon coupling induced by anti-site defects of b-site cations, we calculated the Huang–Rhys factor (S) of different samples, which can be derived using the following equation:62
Stokes shift = (2S − 1)Ephonon | (1) |
In addition to modulating the growth of Cs2AgBiBr6 films, the introduction of PEIE will significantly affect the electronic structure of SnO2 because of its high-density amine groups and electron-donor capability.63,64Fig. 3 shows the ultraviolet photoelectron spectroscopy (UPS) spectra of SnO2 and PEIE-decorated SnO2. Compared with the control sample, the secondary electron cutoff binding energy (Ecutoff) of PEIE-decorated SnO2 increased from 16.66 to 16.71 eV. Meanwhile, the onset binding energy (Eonset) was reduced from 3.64 to 3.47 eV. Subsequently, the valence band maximum (EV) and conduction band minimum (EC) of SnO2 ETLs can be calculated using the following equations:
EV = Ecutoff − Eonset − 21.22 eV | (2) |
EC = Eg + EV | (3) |
Planar-type devices with the structure of ITO/SnO2/PEIE/Cs2AgBiBr6/Spiro-OMeTAD/MoO3/Ag were fabricated to study the effect of PEIE on the performance of solar cells. The schematic illustration and cross-sectional SEM image are shown in Fig. S6 (ESI†). To determine the optimal value of PEIE, we set different precursor concentrations and denote them as PEIE-x. For example, the PEIE-0.2 sample indicates that the concentration of PEIE in the precursor is 0.2 mg ml−1. Fig. 4a shows the J–V characteristic curves of the devices prepared based on different concentrations of PEIE under 1 sun illumination, with photovoltaic parameters listed in Table 2. The PCE of the control sample was 0.59% with a VOC of 0.90 V, a JSC of 1.17 mA cm−2, and a fill factor (FF) of 56.80%. The PCE for the control device is comparable with many previous reports.24,39,65 After PEIE decoration, the PCE increases and the optimal PCE is achieved at a PEIE concentration of 0.2 mg ml−1. The champion device displays a VOC of 1.04 V, a JSC of 2.59 mA cm−2, an FF of 56.20%, and a PCE of 1.52%, showing a 250% increase. 25 independent devices for each PEIE concentration were prepared to test the reproducibility. The statistical distributions of the photovoltaic parameters are shown in Fig. 4b–e. The statistical data of photovoltaic parameters are consistent with the optimal data, where PEIE decoration increases the device performance and 0.2 mg ml−1 is the best, demonstrating the effectiveness and reproducibility of such a strategy. Generally, the increase in PCE after PEIE decoration is mainly due to the improvement in JSC and VOC. On the one hand, the PEIE-modulated crystallization of Cs2AgBiBr6 film leads to an increase in the ordering of b-site cations and a reduction of anti-site defects, effectively suppressing the generation of STEs and thus improving the carrier transport process, leading to a significant increase in JSC. On the other hand, PEIE is capable of adjusting the energy level of SnO2 to make it more compatible with Cs2AgBiBr6, thereby reducing the loss of VOC. Notably, the PCE of the device begins to decline when the concentration of PEIE exceeds 0.3 mg ml−1, mainly due to the decrease in JSC. This cannot be ascribed to the degradation of Cs2AgBiBr6 film quality because the SEM images and XRD patterns show similar results, as shown in Fig. S7 (ESI†). To gain further insights into the decline of JSC, we performed conductivity measurements on ITO/SnO2/PEIE-x/Ag devices, as shown in Fig. 4f. An increase in conductivity for PEIE-decorated SnO2 was observed when the PEIE concentration was below 0.2 mg ml−1 due to the electron doping effect of PEIE.55 However, 0.3 mg ml−1 PEIE-decorated SnO2 showed a significant decrease in conductivity. This is probably due to the fact that PEIE is actually insulating, and thus excessive PEIE would block electron transport. The decreased conductivity of 0.3 mg ml−1 PEIE-decorated SnO2 thus leads to a decline in JSC as well as PCE. Since 0.2 mg ml−1 PEIE decoration leads to the best result, we will focus on this device and the control one in the following paragraph.
Device | J SC (mA cm−2) | V OC (V) | FF (%) | PCE (%) |
---|---|---|---|---|
Control | 1.17 | 0.90 | 56.80 | 0.59 |
PEIE-0.1 | 2.24 | 1.01 | 56.35 | 1.27 |
PEIE-0.2 | 2.59 | 1.04 | 56.20 | 1.52 |
PEIE-0.3 | 1.71 | 1.03 | 53.96 | 0.97 |
The photoelectric conversion ability of Cs2AgBiBr6 PSCs was evaluated by external quantum efficiency (EQE) tests on the control and PEIE samples, as shown in Fig. 5a. The shape of the EQE spectra is similar to that of the absorption spectra, with an overall increase in efficiency for the PEIE sample over the entire spectral range. This is due to the enhanced charge transmission in the PEIE sample benefited from increased cation occupancy ordering. The integrated JSC increased from 1.08 mA cm−2 in the control sample to 2.41 mA cm−2 in the PEIE sample, which is consistent with the JSC in the J–V curves and demonstrates the reliability of our J–V measurements.
Subsequently, we prepared single-electron and single-hole devices with the structures of ITO/SnO2/PEIE/Cs2AgBiBr6/PCBM/Ag and ITO/PEDOT:PSS/PEIE/Cs2AgBiBr6/Spiro-OMeTAD/Ag to analyze the defect-state density (Nt) and carrier mobility of the control and PEIE devices by the space charge limited current (SCLC) method (details in the ESI†),66 as shown in Fig. 5b and c. The calculated Nt of electrons decreases from 5.26 × 1016 cm−3 for the control device to 4.24 × 1016 cm−3 for the PEIE device, and the corresponding electron mobility increases from 0.067 to 0.074 cm2 V−1 s−1. Meanwhile, the Nt of holes decreases from 1.77 × 1017 to 1.09 × 1017 cm−3, and the hole mobility increases from 0.0013 to 0.0047 cm2 V−1 s−1. The reduction in Nt stems from the reduced grain boundaries and increased cation occupancy ordering after PEIE decoration. Due to the annihilation of electronic traps as well as the weakened STE effect, both carrier mobilities increase, which facilitate charge collection and are thus beneficial to the increase in JSC.
Electrochemical impedance spectroscopy (EIS) is an effective test method for analyzing the electrical performance and charge recombination of solar cells.67 The EIS spectra of the devices were recorded in the darkness, as shown in Fig. 5d, with the equivalent-circuit shown in the inset. The extracted recombination resistance (Rre) in the low-frequency region represents the ease of carrier recombination in the device. After PEIE decoration, the Rre increases from 12.02 to 15.12 kΩ, indicating that carrier recombination was effectively suppressed due to the reduction of antisite defects and STEs. A decrease in charge recombination would increase VOC.
To further study the enhancement of VOC, the Mott–Schottky curves with the 1/C2 plots of the control and PEIE devices were recorded (details in the ESI†),53 as shown in Fig. 5e. The built-in voltage (Vbi) increased from 0.79 to 0.91 V after PEIE decoration. The increased Vbi is beneficial for charge separation and would thus impede charge recombination, which is consistent with the improvement of VOC in device performance.
Finally, stability tests were performed for both the control and PEIE devices. Fig. 5f shows the evolution of normalized PCE over time (temperature: 25 ± 5 °C; relative humidity: 25 ± 5%). Both devices exhibit excellent stability under atmospheric conditions. The PCE of the PEIE device remains 90% after 30 days, slightly higher than that of the control device (85%). As shown in Fig. S8 (ESI†), the VOC, JSC, and FF of the control device show a similar declining trend with time, which is delayed in the PEIE device. The increased device stability should stem from the improved film quality on PEIE-decorated SnO2.
Footnote |
† Electronic supplementary information (ESI) available: Experimental sections with materials, preparation of Cs2AgBiBr6 powders and films, fabrication of solar cell devices, and characterizations; color conversions of Cs2AgBiBr6 films on different substrates during thermal treatment; Tauc plots of Cs2AgBiBr6 films; Raman spectra of Cs2AgBiBr6 films and schematic diagrams of different Raman modes; UV-vis absorption spectra of SnO2 and SnO2/PEIE films and corresponding Tauc plots; UPS spectra for Cs2AgBiBr6 films on different substrates; cross-sectional SEM and schematic structure of Cs2AgBiBr6 PSCs; top-view SEM images, average grain size, and XRD patterns of Cs2AgBiBr6 films deposited on substrates with different PEIE concentrations; stability data for VOC, JSC, and FF; fitting parameters for TRPL; equations for TRPL, SCLC, and Mott–Schottky analyses. See DOI: https://doi.org/10.1039/d4tc00219a |
This journal is © The Royal Society of Chemistry 2024 |