- Email: [email protected]
Introduction of LiCl into SnO2 electron transport layer for efficient planar perovskite solar cells
Chemical Physics Letters 745 (2020) 137220
Contents lists available at ScienceDirect
Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett
Research paper
Introduction of LiCl into SnO2 electron transport layer for efficient planar perovskite solar cells
T
Yinyi Huanga,1, Shina Lib,1, Chaorong Wua, Shuo Wanga, Chengyan Wanga,c,e, , Ruixin Maa,d, ⁎
⁎
a
School of Metallurgical and Ecological Engineering, University of Science and Technology Beijing, Beijing 100083, China Tianjin Research Institute for Water Transport Engineering, M. O. T. Tianjin 300000, China c Beijing Key Laboratory of Rare and Precious Metals Green Recycling and Extraction, University of Science and Technology Beijing, Beijing 100083, China d Beijing Key Laboratory of Special Melting and Preparation of High-End Metal Materials, University of Science and Technology Beijing, Beijing 100083, China e Faculty of Materials Metallurgy and Chemistry, Jiangxi University of Science and Technology, Ganzhou 341000, China b
HIGHLIGHTS
Li-doped SnO (Li:SnO ) as an effective ETL, improve the electrical properties of SnO . • Using devices are fabricated by a facile two-step deposition strategy. • Solar • Optimized equipment improved V , FF and J . Got a device with a maximum PCE close to 19%. 2
2
2
oc
sc
ARTICLE INFO
ABSTRACT
Keywords: Perovskite solar cell Tin oxide Electronic transport layer Li doping
SnO2 has recently aroused huge attention as an electron transfer material for planar halide perovskite solar cells. Nevertheless, planar structure devices exhibit significant hysteresis behavior and low optical stability for their considerable trap states and high ultraviolet transmittance. In this study, LiCl was added to the SnO2 electron transport layer. As revealed from the results, adding LiCl could enhance the mobility of SnO2 film and promote the optical stability. Lastly, Li:SnO2 devices achieved high power conversion efficiency (PCE) of over 18% and steady-state PCE of 18.35%. Besides, they displayed prominent stability storage under dry conditions.
1. Introduction In recent years, the organic-inorganic hybrid perovskite solar cell (PSC) has been leaping forward, exhibiting its high potential in the photovoltaic industry [1–3]. Power conversion efficiency (PCE) has increased to 25.2% [4]. Besides its prominent optoelectronic characteristics, halo-perovskite has been extensively studied on its ferroelectric and optical characteristics to enhance its applicability in fields (e.g., luminescence and memory devices) [5–7]. Moreover, low temperature treated halides have been recognized as the optimized light absorbing materials for high efficiency, lightweight and flexible photovoltaic devices for their remarkable mechanical flexibility [8–10]. The materials employed in charge transport layers for PSCs primarily originate from conventional organic and dye sensitized solar cells, facilitating the process of developing novel device architectures and high-performance flexible devices [11,12]. The mesoporous
structure of the PSC has been exploited typically to achieve high PCE as the perovskite layer is tightly contacted to the electron transport layer (ETL). Nevertheless, high sintering temperatures (> 450 °C) and unsatisfactory UV stability have limited their application in flexible and series high efficiency devices. To address the mentioned problems, the development of planar perovskite solar cells based on a low temperature process was an effective way. A range of metal oxides are capable of transferring carriers for their high optical transparency [13]. Metal oxides, sulfides and selenides (e.g., SnO2 and ZnO) have been studied as electron transfer materials for planar PSCs [14–16]. SnO2 exhibits a proper energy level, an ideal band gap as well as high electron mobility, ensuring higher light stability and more effective charge injection [17]. The use of SnO2 instead of TiO2 is more conducive to energy band orientation in PSC since SnO2 ECB is lower than TiO2 and can lead to higher charge injection efficiency. Thus, SnO2 has turned into a promising metal oxide to be applied in effective
Corresponding authors at: School of Metallurgical and Ecological Engineering, University of Science and Technology Beijing, Beijing 100083, China. E-mail addresses: [email protected] (Y. Huang), [email protected] (S. Li), [email protected] (C. Wu), [email protected] (S. Wang), [email protected] (C. Wang), [email protected], [email protected] (R. Ma). 1 Yinyi Huang and Shina Li contributed equally to this work. ⁎
https://doi.org/10.1016/j.cplett.2020.137220 Received 5 December 2019; Received in revised form 11 February 2020; Accepted 12 February 2020 Available online 13 February 2020 0009-2614/ © 2020 Elsevier B.V. All rights reserved.
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 1. Schematic representation of the manufacturing process for perovskite devices.
electron transport layers (ETL) in PSCs. Nevertheless, given the mentioned methods, further commercialization of such devices may be limited by the hysteresis and stability [18–20]. Accordingly, much effort has been made on enhancing the photoelectricity characteristics of SnO2. Hagfeldt et al. [21] reported a high-efficiency planar PSC (PCE = 18.4%) to treat SnO2 by atomic layer deposition (ALD). Park et al. [22] employed the zwitterionic modified SnO2 electron transport layer to cause shifts in the work function; thus, a device exhibiting high PCE and remarkable thermal stability was developed. Liu et al. [23] exploited EDTA composite SnO2 electron transport layer to achieve the record PCE of the planar-type PSCs exhibiting negligible hysteresis and long-term stability. Wang et al. [24] added carbon nanodots into SnO2 electron transport layer for efficient and UV stable planar perovskite solar cells. By doping metal at a low concentration in SnO2, the energy level was regulated. Accordingly, an appropriate alignment of the energy levels adjacent to the transparent electrode was effectively induced, and the electrical conductivity was enhanced, while a good surface coverage of the film was maintained, thereby enhancing device performance [25–27]. In this study, the manufacture of perovskite solar cells was reported with Li-doped SnO2 (Li: SnO2) as an effective ETL. The Li:SnO2 film was produced at a low temperature (150 °C), and the conductivity of the film was slightly enhanced. By varying the electrical characteristics of SnO2 through Li doping, the extraction and transfer of electrons from the ECB of the perovskite were facilitated, thereby hysteresis behavior was suppressed.
dissolved in 1 mL of isopropanol. By dissolving 72.3 mg spiro-OMeTAD in 1 mL CB and 17.5 μl Li-TFSI solution (520 mg in 1 mL ACN) and 28.8 μl TBP, The spiro-OMeTAD solution was produced. 2.3. Device fabrication ITO substrates were cleaned with detergent, deionized water and absolute ethanol for 15 min. And then use the flowing nitrogen gas to dry the glass. Then irradiated with UV-Ozone for 15 min. After that, SnO2 solution was spin-coated on the top of ITO substrate at 4000 rpm for 30 s and then annealed at 150 °C for 30 min. Then, PbI2 (1.2 M) was spin-coated on the top of SnO2 at 3000 rpm for 30 s. The MAI solution was directly coated on “wet” PbI2 layer at 3000 rpm for 30 s, followed by annealing at 100 °C for 30 min. The spiro-OMeTAD was formed at 4500 rpm for 20 s to obtain the hole transporting layer. Finally, 80 nm of gold metal was thermally evaporated on top of the device to form the back contact. 2.4. Characterization The scanning electron microscopy (SEM) images were captured under a ZEISS SUPRA55 microscope. The X-ray diffraction (XRD) patterns were collected with a Smart Lab from Rigaku at 40 kV and 150 mA based on Cu-Ka radiation (λ = 0.15405 nm). With a CHI660d electrochemical station, relevant electrochemical tests were performed. The photovoltaic performance of PSCs was recorded with a Keithley 2400 source meter under one-sun AM 1.5G (100 mW cm−2) illumination using a solar light simulator (Newport Oriel Sol3A Class A, 64023A Simulator). The external quantum efficiency (EQE) was measured in the wavelength range of 350–800 nm with a DK240 monochromator. Steady state photoluminescence (PL) was measured with FLS980 spectrometer (Edinburgh, England). The UV–vis light absorption was measured with an ultraviolet–visible (UV–vis-IR) spectrophotometer (Lambda 950).
2. Experimental 2.1. Materials Alfa Aesar provided SnO2 colloid precursor [Sn (IV) oxide, 15% in H2O colloidal dispersion], anhydrous N, N-dimethylforma-mide (DMF), anhydrous dimethyl sulfoxide (DMSO), anhy-drous isopropyl alcohol (IPA) and chlorobenzene. Indium tin oxide (ITO)-coated glass, Methylammonium iodide (MAI), and lithium bis (trifluoromethylsulfonyl) imide (Li-TFSI) were purchased from You Xuan Technology. PbI2, and spiro-OMeTAD were provided by Xi’an Polymer Light Technology Corp. 4-tertbutylpyridine (TBP) was obtained from Aldrich. Shanghai Jufeng Chemical Technology Co. provided LiCl. All commercially available materials were used as received.
3. Results and discussion A schematic diagram of a two-step process for developing a perovskite device is given in Fig. 1. As revealed from the figure, SnO2, PbI2, MAI, and Spiro-OMeTAD solutions were continuously spin-coated on an ITO substrate to develop a planar structure device. Unlike the conventional two-step process, MAI was spin-coated onto an unannealed PbI2 film. By SEM, the morphology of the SnO2 layer at a range of concentrations of LiCl was studied. Fig. 2 gives the top-view SEM images of SnO2 layers, the morphologies of SnO2 and Li:SnO2 films were almost identical, whereas 3%, 4%, 5% Li:SnO2 films exhibited smooth surfaces, while the surfaces of 0%, 2%, 6% films were uneven. The top-view SEM images of perovskite layers of LiCl at different
2.2. Materials preparation The SnO2 electron layer was fabricated by dissolving the SnO2 colloid precursor in deionized water at a volume ratio of 1:5. Subsequently, the LiCl was mixed in the SnO2 precursor solution at a ratio of 2, 3, 4, 5, and 6 mg%. By dissolving 1.2 M PbI2 in anhydrous DMF/DMSO (9:1, v/v), PbI2 solution was prepared. 60 mg MAI was 2
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 2. The Top-view SEM images of SnO2 film of (a) pristine SnO2. (b) 2% Li: SnO2. (c) 3% Li: SnO2. (d) 4% Li: SnO2. (e) 5% Li: SnO2. (f) 6% Li: SnO2.
Fig. 3. The Top-view SEM images of perovskite film of (a) pristine SnO2. (b) 2% Li: SnO2. (c) 3% Li: SnO2. (d) 4% Li: SnO2. (e) 5% Li: SnO2. (f) 6% Li: SnO2. (g) Grain size distributions obtained from SEM images. (h) XRD patterns of perovskite films.
concentrations are presented in Fig. 3. As suggested from the figure, after LiCl was added into SnO2, the grain size of the perovskite was enlarged. The grain size distribution is illustrated in Fig. 3g. With the rise in Li content, perovskite exhibited larger grain size. The ion exchange between the ETL and the perovskite layer facilitated the growth of perovskite, thereby leading to an increase in the grain size. However, too much Cl accumulated on the ETL surface, capable of introducing pinholes and defects. The crystallinity of these perovskite films was characterized and then compared by XRD. Fig. 3h displays a typical XRD pattern without any shift in peak position. The crystallinity of the perovskite films was evidently enhanced by adding LiCl into the SnO2, probably because the ion exchange at the lower interface promoted the interface contact and the perovskite growth. The crystallinity of 4% perovskite reached the peak. Subsequently, 5% and 6% perovskite crystallinity declined, since too much Cl accumulated at the interface and affected the perovskite growth. Note that the holes, defects and impurities were observed on the grain boundaries with the rise in LiCl concentration (Fig. 3f), as revealed from the results of XRD and UV (shown later). The opened grain boundary and impurities might lead to
charge leakage between the infiltrated hole and the electron transferring layers, since the rough surface and intensive grain boundaries were identified as the most efficient routes for ion migration. X-ray photoelectron spectroscopy (XPS) measurement was performed. Fig. 4a presents the XPS survey spectra of pristine SnO2 and Li: SnO2; the spectra illustrate that only Sn, C and O peaks are present on the pure SnO2 film, and the doped thin film spectra consist of Sn, C and O and Li peaks. Fig. 4b gives a partially enlarged view of the XPS spectrum. A clear Li peak is observed in Li:SnO2, whereas pristine SnO2 displays no Li peak. The binding energy (BE) of the Li 1s peak reached 47.5 eV. Fig. 4c illustrates the XPS spectra of Sn 3d peaks, the Sn states with a binding energy (BE) of 495.33 eV (pristine SnO2) for the Sn 3d3/2 peak shift to 495.20 eV upon 4% Li doping. The binding energy (BE) of 486.81 eV (pristine SnO2) for the Sn 3d5/2 peak shifts to 486.74 eV upon 4% Li doping. The mentioned peaks of Sn 3d spectra reveal a slightly positive shift, attributed to the interaction among Li, Sn and O atoms. The slightly lower O1s binding energy for the Li-doped sample revealed the linkage of the LieOeSn bond in LieSnO2 (Fig. 4d). The linkage of the SneO bond was weakened with Li addition. The 3
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 4. (a) XPS spectra of different concentrations of LiCl prepared on the ITO glass substrate. (b) XPS spectra of Li 1s peaks. (c) XPS spectra of Sn 3d peaks. (d) XPS spectra of O 1s peaks. (e) XRD patterns of SnO2 layer of different concentrations of LiCl.
mentioned variations confirmed the formation of SneOeLi structure in the Li-doped SnO2 [28]. Fig. 4e illustrates X-ray diffraction (XRD) patterns of SnO2 at various LiCl concentrations (0%, 2%, 3%, 4%, 5% and 6%). All X-ray diffraction peaks for SnO2 are indexable to the tetragonal SnO2 structure, suggesting the formation of SnO2 crystals. To clarify the effect of LiCl on the optical characteristics of the perovskite film, the ultraviolet–visible (UV–vis) spectrum was ascertained, as shown in Fig. 5a. As compared with the pristine SnO2 film, the 4% Li:SnO2 film exhibited a stronger absorption in the entire light absorption range, and the absorption edge was red-shifted to ensure that perovskite could capture low-energy photons. To quantitatively assess the effect of LiCl addition on the band gap of perovskite materials, the band gap was calculated in accordance with the Tauc curve. Fig. 5b suggests that by adding LiCl, a shrinkage of the band gap was observed [29,30]. Fig. 5c presents the transmission spectrum of the pristine SnO2 or Li:SnO2 film. As impacted by the band gap of SnO2, it exhibited low transmittance in the range of 300–350 nm and high transmittance in the range of 350–800 nm. Moreover, 3% and 4% Li:SnO2 film displayed a higher transmittance in the range of 350–550 nm. The charge transfer behaviors at the interface between ETL and perovskite was delved into by steady-state photoluminescence (PL) quenching of perovskites. Fig. 5d presents the PL spectra of five samples, namely, glass/ITO/SnO2/perovskite, glass/ITO/Li:SnO2/perovskite. Fluorescence emission of perovskite was identified at 770 nm. The 2%, 3%, 4% Li:SnO2 films exhibited noticeable PL quenching compared to the SnO2 film, indicating that the conductivity could be enhanced, and the trap state density could decline with LiCl addition. PL spectra suggested that 5% and 6% Li:SnO2 displayed less quenching than the pristine SnO2, since excessive Li aggregated at the interface, resulting in severe interface recombination [31]. Fig. 6 shows the statistics of PCE of 20 devices based on a range of doping concentrations. The increase in Voc was because Li enhanced the mobility. The rise in Jsc and FF resulted from the ion exchange at the lower interface Cl, which promoted the interface contact and up-regulated the grain size. Besides, the 4% Li:SnO2 devices exhibited the maximal photovoltaic parameters as compared with the others. The
degradation of 5% and 6% device performance might result from the aggregation of Cl on ETLs surface, which hinders the charge transport from perovskite to ETL [32]. The photocurrent hysteresis of SnO2 and Li:SnO2 devices was also tested, and the photocurrent hysteresis of the device was calculated. The J-V curves of the devices in a range of scanning directions are plotted in Fig. 7a, and the relevant J-V parameters are listed in Table 1. PCEreverse PCEforward The hysteresis index was calculated by h = [18,33]. PCEreverse The PCE of the SnO2 device in a range of scanning directions took up 16.96% and 13.33%, and the hysteresis index reached 21.40%. In different scanning directions, the PCE of the 4% Li:SnO2 device were 18.12% and 15.90%, and the hysteresis index was 12.25%. As impacted by the ion-exchange effect of Cl at the interface, the grain size of the perovskite increased, so the FF was enhanced. Accordingly, negligible high-efficiency hysteresis of the J-V value was achieved [19,34]. The steady-state power output measured at the maximal power is presented in Fig. 7b. The 4% Li:SnO2 device exhibited a stabilized efficiency of 18.35% at the maximal power point (0.919 V). It had stable output at 4000 s and prominent optical stability. To more effectively characterize the effect of LiCl on the characteristics of SnO2 films, the I-V characteristics are depicted in Fig. 7c. The SnO2 and Li: SnO2 films were measured with the sandwich structures ITO/Li:SnO2/Au and ITO/ SnO2/ Au. The current in the Li:SnO2 film is observed to be up-regulated. The direct current conductivity (σ0) is calculated by the formula I = 0 Ad 1 V , where d denotes the thickness of the film, A refers to the area of the film, and the conductivity σ0 of Li:SnO2 was enhanced from 0.079 mS cm−1 to 0.106 mS cm−1 as LiCl was added. According to the external quantum efficiency (EQE) curve (Fig. 7d), the device was identified exhibiting similar spectral shapes between 350 nm and 400 nm. The SnO2 and Li:SnO2-based devices achieved the maximal EQE values at 520 nm, 87.21% and 91.12%, respectively. The integral Jsc of the devices based on SnO2 and Li:SnO2 reached 19.77 mA cm−2 and 20.77 mA cm−2. The higher EQE of Li:SnO2 was attributed to the enhanced charge collected and transferred with Li addition [35–38]. Stability has posed a challenge to perovskite equipment. To compare the storage stability of the device, the storage stability was tested
4
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 5. (a) Absorption spectra of the perovskite layers of the different concentrations of LiCl. (b) The band gap calculation using the Tauc plot. (c) Transmittance spectra of the SnO2 layers of the different concentrations of LiCl. (d) PL of the perovskite layers of the different concentrations of LiCl.
Fig. 6. Statistics on the distribution of (a) Voc, (b) Jsc, (c) FF and (d) PCE of PSCs with different concentrations of octocrylene.
5
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 7. (a) J–V curves of the device based on SnO2 and Li: SnO2 at reverse and forward scans. (b) Stabilized output of the cell measured at a bias voltage of 0.919 V under simulated 100 mW cm−2, AM 1.5G irradiation of the device based on Li: SnO2. (c) I–V characteristics of the ITO/Li:SnO2/Au and ITO/SnO2/Au. (d) EQE spectra of the devices comprising SnO2 and Li:SnO2 as an ETL.
4. Conclusion
Table 1 Device parameters in forward and reverse scans. Jsc (mA/cm2)
Voc (V)
FF (%)
PCE (%)
Hysteresis index
0%
Reverse Forward
21.43 21.81
1.065 1.051
74.30 58.11
16.96 13.33
21.40%
4%
Reverse Forward
22.28 22.62
1.069 1.033
76.11 68.07
18.12 15.90
12.25%
In summary, Li:SnO2 was successfully prepared as an effective ETL in PSC by a low temperature process. After doping with LiCl, the enhancement in conductivity in SnO2 promoted electron injection and transfer and hindered charge recombination. The optimized equipment achieved improved Voc, FF and Jsc. A device with a maximum PCE close to 19% was developed. Moreover, the enhancement of optical stability led to a steady state output of 18.35% at 4000 s. CRediT authorship contribution statement Yinyi Huang: Validation, Writing - original draft, Writing - review & editing. Shina Li: Methodology, Data curation. Chaorong Wu: Validation. Shuo Wang: Writing - review & editing. Chengyan Wang: Project administration, Resources, Software. Ruixin Ma: Supervision. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Fig. 8. Storage stability of the corresponding devices stored under dry conditions for 40 days without encapsulation.
Acknowledgments This work was supported by the National Natural Science Foundation of China (U1802253, U1702252, 51834008), the Guangxi Innovation-driven Development Project (AA18242042-1), and the Fundamental Research Funds for the Central Universities (FRF22-TT19-001).
under dark and dry conditions (humidity, 10–15%) for 40 days, as suggested in Fig. 8. Obviously, after 40 days, the SnO2 device retained 73.63% of the initial PCE, and the Li:SnO2 device maintained 78.89% relative to its initial PCE. 6
Y. Huang, et al.
Chemical Physics Letters 745 (2020) 137220
References
[20] F.D. Giacomo, S. Razza, F. Matteocci, J. Power Sources 251 (2014) 152–156. [21] J.P.C. Baena, L. Steier, W. Tress, M. Saliba, A. Hagfeldt, Energy Environ. Sci. 8 (2015) 2928–2934. [22] K. Choi, J. Lee, H.I. Kim, C.W. Park, G. Kim, H. Choi, S. Park, S.A. Park, T. Park, Energy Environ. Sci. 11 (2018) 3238–3247. [23] D. Yang, R. Yang, K. Wang, C. Wu, X. Zhu, J. Feng, X. Ren, G. Fang, S. Priya, S. Liu, Nat. Commun. 9 (2018) 3239. [24] S. Wang, Y. Zhu, B. Liu, C. Wang, R. Ma, J. Mater. Chem. A 7 (2019) 5353–5362. [25] Q. Jiang, L. Zhang, H. Wang, Nat. Energy 2 (2019) 16177. [26] Y. Ko, Y.R. Kim, H. Jang, Nanoscale Res. Lett. 12 (2017) 498. [27] L. Huang, X. Sun, C. Li, ACS Appl. Mater. Interfaces 26 (2017) 21909–21920. [28] L. Li, Y. Chen, Z. Liu, Adv. Mater. 28 (2016) 9862–9868. [29] T. Bu, X. Liu, Y. Zhou, Energy Environ. Sci. 10 (2017) 1039. [30] N. Cheng, W. Li, M. Zhang, H. Wu, S. Sun, Curr. Appl. Phys. 19 (2019) 25–30. [31] S. Wang, Y. Zhu, B. Liu, C. Wang, R. Ma, Ionics 25 (2019) 4509–4516. [32] P. Zhu, S. Gu, H. Tan, Adv. Energy Mater. (2019) 1903083. [33] Y. Rong, Y. Hu, S. Ravishankar, H. Liu, X. Hou, Y. Sheng, Energy Environ. Sci. 10 (2017) 2383–2391. [34] M. Li, Y. Huan, X. Yan, Z. Kang, Y. Guo, Y. Li, ChemSusChem. (2017). [35] J.H. Heo, M.S. You, M.H. Chang, Nano Energy 15 (2015) 530–539. [36] D.P. Joseph, P. Renugambal, M. Saravanan, Thin Solid Films. 517 (2009) 6129–6136. [37] Y. Zhu, S. Wang, R. Ma, C. Wang, Org. Electron. 74 (2019) 228–236. [38] Z. Yang, J. Xie, V. Arivazhagan, Nano Energy 40 (2019) 345–351.
[1] A. Kojima, K. Teshima, Shirai, Y. Shirai, T. Miyasaka, J. Am. Chem. Soc. 131 (2009) 6050–6051. [2] J.K. Im, C.R. Lee, J.W. Lee, S.W. Park, N.G. Park, Nanoscale 3 (2011) 4088–4093. [3] N. Ahn, D.Y. Son, I.H. Jang, J. Am. Chem. Soc. 137 (2015) 8696–8699. [4] NREL, Efficiency chart, https://www.nrel.gov/pv/assets/images/efficiency-chart20190926.jpg, 2019. [5] Saleem, A. Saleem, M. Imran, M. Arshad, Appl. Phys. A (2019) 125–229. [6] Z.P. Wang, Q.Q. Lin, W. Bernard, Nature Energy 3 (2019) 855–861. [7] J.J. Ma, X.L. Zheng, Sol. RRL 10 (2017). [8] Y. Liu, Z. Hong, Q. Chen, Nano Lett. 15 (2015) 662–668. [9] M. Thomas, T. Thue, Adv. Energy Mater. 5 (2015) 1500569. [10] P. Shen, M.N. Yao, G.X. Wang, J. Mater. Chem. A 10 (2018) 1039. [11] E. Calabrò, F. Matteocci, A.L. Palma, Sol. Energy Mater. Sol. Cells 185 (2018) 136–144. [12] P. Docampo, J.M. Ball, M. Darwich, Nat. Commun. 4 (2013) 2761. [13] M. Park, J.Y. Kim, H.J. Son, Nano Energy 26 (2016) 208–215. [14] L. Xiong, M. Qin, C. Chen, Adv. Funct. Mater. 28 (2018) 1706276. [15] J. You, L. Meng, T. Song, Nat. Nanotechnol. 11 (2016). [16] S.S. Shin, W.S. Yang, E.J. Yeom, Chem. Lett. 7 (2016) 1845–1851. [17] W. Ke, G. Fang, Q. Liu, J. Am. Chem. Soc. 137 (2015) 6730–6733. [18] Y. Shao, Z. Xiao, C. Bi, Nat. Commun. 5 (2014) 5784. [19] N. Pellet, P. Gao, G. Gregori, Angew. Chem. Int. Ed. 53 (2014) 3151–3157.
7
Contents lists available at ScienceDirect
Chemical Physics Letters journal homepage: www.elsevier.com/locate/cplett
Research paper
Introduction of LiCl into SnO2 electron transport layer for efficient planar perovskite solar cells
T
Yinyi Huanga,1, Shina Lib,1, Chaorong Wua, Shuo Wanga, Chengyan Wanga,c,e, , Ruixin Maa,d, ⁎
⁎
a
School of Metallurgical and Ecological Engineering, University of Science and Technology Beijing, Beijing 100083, China Tianjin Research Institute for Water Transport Engineering, M. O. T. Tianjin 300000, China c Beijing Key Laboratory of Rare and Precious Metals Green Recycling and Extraction, University of Science and Technology Beijing, Beijing 100083, China d Beijing Key Laboratory of Special Melting and Preparation of High-End Metal Materials, University of Science and Technology Beijing, Beijing 100083, China e Faculty of Materials Metallurgy and Chemistry, Jiangxi University of Science and Technology, Ganzhou 341000, China b
HIGHLIGHTS
Li-doped SnO (Li:SnO ) as an effective ETL, improve the electrical properties of SnO . • Using devices are fabricated by a facile two-step deposition strategy. • Solar • Optimized equipment improved V , FF and J . Got a device with a maximum PCE close to 19%. 2
2
2
oc
sc
ARTICLE INFO
ABSTRACT
Keywords: Perovskite solar cell Tin oxide Electronic transport layer Li doping
SnO2 has recently aroused huge attention as an electron transfer material for planar halide perovskite solar cells. Nevertheless, planar structure devices exhibit significant hysteresis behavior and low optical stability for their considerable trap states and high ultraviolet transmittance. In this study, LiCl was added to the SnO2 electron transport layer. As revealed from the results, adding LiCl could enhance the mobility of SnO2 film and promote the optical stability. Lastly, Li:SnO2 devices achieved high power conversion efficiency (PCE) of over 18% and steady-state PCE of 18.35%. Besides, they displayed prominent stability storage under dry conditions.
1. Introduction In recent years, the organic-inorganic hybrid perovskite solar cell (PSC) has been leaping forward, exhibiting its high potential in the photovoltaic industry [1–3]. Power conversion efficiency (PCE) has increased to 25.2% [4]. Besides its prominent optoelectronic characteristics, halo-perovskite has been extensively studied on its ferroelectric and optical characteristics to enhance its applicability in fields (e.g., luminescence and memory devices) [5–7]. Moreover, low temperature treated halides have been recognized as the optimized light absorbing materials for high efficiency, lightweight and flexible photovoltaic devices for their remarkable mechanical flexibility [8–10]. The materials employed in charge transport layers for PSCs primarily originate from conventional organic and dye sensitized solar cells, facilitating the process of developing novel device architectures and high-performance flexible devices [11,12]. The mesoporous
structure of the PSC has been exploited typically to achieve high PCE as the perovskite layer is tightly contacted to the electron transport layer (ETL). Nevertheless, high sintering temperatures (> 450 °C) and unsatisfactory UV stability have limited their application in flexible and series high efficiency devices. To address the mentioned problems, the development of planar perovskite solar cells based on a low temperature process was an effective way. A range of metal oxides are capable of transferring carriers for their high optical transparency [13]. Metal oxides, sulfides and selenides (e.g., SnO2 and ZnO) have been studied as electron transfer materials for planar PSCs [14–16]. SnO2 exhibits a proper energy level, an ideal band gap as well as high electron mobility, ensuring higher light stability and more effective charge injection [17]. The use of SnO2 instead of TiO2 is more conducive to energy band orientation in PSC since SnO2 ECB is lower than TiO2 and can lead to higher charge injection efficiency. Thus, SnO2 has turned into a promising metal oxide to be applied in effective
Corresponding authors at: School of Metallurgical and Ecological Engineering, University of Science and Technology Beijing, Beijing 100083, China. E-mail addresses: [email protected] (Y. Huang), [email protected] (S. Li), [email protected] (C. Wu), [email protected] (S. Wang), [email protected] (C. Wang), [email protected], [email protected] (R. Ma). 1 Yinyi Huang and Shina Li contributed equally to this work. ⁎
https://doi.org/10.1016/j.cplett.2020.137220 Received 5 December 2019; Received in revised form 11 February 2020; Accepted 12 February 2020 Available online 13 February 2020 0009-2614/ © 2020 Elsevier B.V. All rights reserved.
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 1. Schematic representation of the manufacturing process for perovskite devices.
electron transport layers (ETL) in PSCs. Nevertheless, given the mentioned methods, further commercialization of such devices may be limited by the hysteresis and stability [18–20]. Accordingly, much effort has been made on enhancing the photoelectricity characteristics of SnO2. Hagfeldt et al. [21] reported a high-efficiency planar PSC (PCE = 18.4%) to treat SnO2 by atomic layer deposition (ALD). Park et al. [22] employed the zwitterionic modified SnO2 electron transport layer to cause shifts in the work function; thus, a device exhibiting high PCE and remarkable thermal stability was developed. Liu et al. [23] exploited EDTA composite SnO2 electron transport layer to achieve the record PCE of the planar-type PSCs exhibiting negligible hysteresis and long-term stability. Wang et al. [24] added carbon nanodots into SnO2 electron transport layer for efficient and UV stable planar perovskite solar cells. By doping metal at a low concentration in SnO2, the energy level was regulated. Accordingly, an appropriate alignment of the energy levels adjacent to the transparent electrode was effectively induced, and the electrical conductivity was enhanced, while a good surface coverage of the film was maintained, thereby enhancing device performance [25–27]. In this study, the manufacture of perovskite solar cells was reported with Li-doped SnO2 (Li: SnO2) as an effective ETL. The Li:SnO2 film was produced at a low temperature (150 °C), and the conductivity of the film was slightly enhanced. By varying the electrical characteristics of SnO2 through Li doping, the extraction and transfer of electrons from the ECB of the perovskite were facilitated, thereby hysteresis behavior was suppressed.
dissolved in 1 mL of isopropanol. By dissolving 72.3 mg spiro-OMeTAD in 1 mL CB and 17.5 μl Li-TFSI solution (520 mg in 1 mL ACN) and 28.8 μl TBP, The spiro-OMeTAD solution was produced. 2.3. Device fabrication ITO substrates were cleaned with detergent, deionized water and absolute ethanol for 15 min. And then use the flowing nitrogen gas to dry the glass. Then irradiated with UV-Ozone for 15 min. After that, SnO2 solution was spin-coated on the top of ITO substrate at 4000 rpm for 30 s and then annealed at 150 °C for 30 min. Then, PbI2 (1.2 M) was spin-coated on the top of SnO2 at 3000 rpm for 30 s. The MAI solution was directly coated on “wet” PbI2 layer at 3000 rpm for 30 s, followed by annealing at 100 °C for 30 min. The spiro-OMeTAD was formed at 4500 rpm for 20 s to obtain the hole transporting layer. Finally, 80 nm of gold metal was thermally evaporated on top of the device to form the back contact. 2.4. Characterization The scanning electron microscopy (SEM) images were captured under a ZEISS SUPRA55 microscope. The X-ray diffraction (XRD) patterns were collected with a Smart Lab from Rigaku at 40 kV and 150 mA based on Cu-Ka radiation (λ = 0.15405 nm). With a CHI660d electrochemical station, relevant electrochemical tests were performed. The photovoltaic performance of PSCs was recorded with a Keithley 2400 source meter under one-sun AM 1.5G (100 mW cm−2) illumination using a solar light simulator (Newport Oriel Sol3A Class A, 64023A Simulator). The external quantum efficiency (EQE) was measured in the wavelength range of 350–800 nm with a DK240 monochromator. Steady state photoluminescence (PL) was measured with FLS980 spectrometer (Edinburgh, England). The UV–vis light absorption was measured with an ultraviolet–visible (UV–vis-IR) spectrophotometer (Lambda 950).
2. Experimental 2.1. Materials Alfa Aesar provided SnO2 colloid precursor [Sn (IV) oxide, 15% in H2O colloidal dispersion], anhydrous N, N-dimethylforma-mide (DMF), anhydrous dimethyl sulfoxide (DMSO), anhy-drous isopropyl alcohol (IPA) and chlorobenzene. Indium tin oxide (ITO)-coated glass, Methylammonium iodide (MAI), and lithium bis (trifluoromethylsulfonyl) imide (Li-TFSI) were purchased from You Xuan Technology. PbI2, and spiro-OMeTAD were provided by Xi’an Polymer Light Technology Corp. 4-tertbutylpyridine (TBP) was obtained from Aldrich. Shanghai Jufeng Chemical Technology Co. provided LiCl. All commercially available materials were used as received.
3. Results and discussion A schematic diagram of a two-step process for developing a perovskite device is given in Fig. 1. As revealed from the figure, SnO2, PbI2, MAI, and Spiro-OMeTAD solutions were continuously spin-coated on an ITO substrate to develop a planar structure device. Unlike the conventional two-step process, MAI was spin-coated onto an unannealed PbI2 film. By SEM, the morphology of the SnO2 layer at a range of concentrations of LiCl was studied. Fig. 2 gives the top-view SEM images of SnO2 layers, the morphologies of SnO2 and Li:SnO2 films were almost identical, whereas 3%, 4%, 5% Li:SnO2 films exhibited smooth surfaces, while the surfaces of 0%, 2%, 6% films were uneven. The top-view SEM images of perovskite layers of LiCl at different
2.2. Materials preparation The SnO2 electron layer was fabricated by dissolving the SnO2 colloid precursor in deionized water at a volume ratio of 1:5. Subsequently, the LiCl was mixed in the SnO2 precursor solution at a ratio of 2, 3, 4, 5, and 6 mg%. By dissolving 1.2 M PbI2 in anhydrous DMF/DMSO (9:1, v/v), PbI2 solution was prepared. 60 mg MAI was 2
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 2. The Top-view SEM images of SnO2 film of (a) pristine SnO2. (b) 2% Li: SnO2. (c) 3% Li: SnO2. (d) 4% Li: SnO2. (e) 5% Li: SnO2. (f) 6% Li: SnO2.
Fig. 3. The Top-view SEM images of perovskite film of (a) pristine SnO2. (b) 2% Li: SnO2. (c) 3% Li: SnO2. (d) 4% Li: SnO2. (e) 5% Li: SnO2. (f) 6% Li: SnO2. (g) Grain size distributions obtained from SEM images. (h) XRD patterns of perovskite films.
concentrations are presented in Fig. 3. As suggested from the figure, after LiCl was added into SnO2, the grain size of the perovskite was enlarged. The grain size distribution is illustrated in Fig. 3g. With the rise in Li content, perovskite exhibited larger grain size. The ion exchange between the ETL and the perovskite layer facilitated the growth of perovskite, thereby leading to an increase in the grain size. However, too much Cl accumulated on the ETL surface, capable of introducing pinholes and defects. The crystallinity of these perovskite films was characterized and then compared by XRD. Fig. 3h displays a typical XRD pattern without any shift in peak position. The crystallinity of the perovskite films was evidently enhanced by adding LiCl into the SnO2, probably because the ion exchange at the lower interface promoted the interface contact and the perovskite growth. The crystallinity of 4% perovskite reached the peak. Subsequently, 5% and 6% perovskite crystallinity declined, since too much Cl accumulated at the interface and affected the perovskite growth. Note that the holes, defects and impurities were observed on the grain boundaries with the rise in LiCl concentration (Fig. 3f), as revealed from the results of XRD and UV (shown later). The opened grain boundary and impurities might lead to
charge leakage between the infiltrated hole and the electron transferring layers, since the rough surface and intensive grain boundaries were identified as the most efficient routes for ion migration. X-ray photoelectron spectroscopy (XPS) measurement was performed. Fig. 4a presents the XPS survey spectra of pristine SnO2 and Li: SnO2; the spectra illustrate that only Sn, C and O peaks are present on the pure SnO2 film, and the doped thin film spectra consist of Sn, C and O and Li peaks. Fig. 4b gives a partially enlarged view of the XPS spectrum. A clear Li peak is observed in Li:SnO2, whereas pristine SnO2 displays no Li peak. The binding energy (BE) of the Li 1s peak reached 47.5 eV. Fig. 4c illustrates the XPS spectra of Sn 3d peaks, the Sn states with a binding energy (BE) of 495.33 eV (pristine SnO2) for the Sn 3d3/2 peak shift to 495.20 eV upon 4% Li doping. The binding energy (BE) of 486.81 eV (pristine SnO2) for the Sn 3d5/2 peak shifts to 486.74 eV upon 4% Li doping. The mentioned peaks of Sn 3d spectra reveal a slightly positive shift, attributed to the interaction among Li, Sn and O atoms. The slightly lower O1s binding energy for the Li-doped sample revealed the linkage of the LieOeSn bond in LieSnO2 (Fig. 4d). The linkage of the SneO bond was weakened with Li addition. The 3
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 4. (a) XPS spectra of different concentrations of LiCl prepared on the ITO glass substrate. (b) XPS spectra of Li 1s peaks. (c) XPS spectra of Sn 3d peaks. (d) XPS spectra of O 1s peaks. (e) XRD patterns of SnO2 layer of different concentrations of LiCl.
mentioned variations confirmed the formation of SneOeLi structure in the Li-doped SnO2 [28]. Fig. 4e illustrates X-ray diffraction (XRD) patterns of SnO2 at various LiCl concentrations (0%, 2%, 3%, 4%, 5% and 6%). All X-ray diffraction peaks for SnO2 are indexable to the tetragonal SnO2 structure, suggesting the formation of SnO2 crystals. To clarify the effect of LiCl on the optical characteristics of the perovskite film, the ultraviolet–visible (UV–vis) spectrum was ascertained, as shown in Fig. 5a. As compared with the pristine SnO2 film, the 4% Li:SnO2 film exhibited a stronger absorption in the entire light absorption range, and the absorption edge was red-shifted to ensure that perovskite could capture low-energy photons. To quantitatively assess the effect of LiCl addition on the band gap of perovskite materials, the band gap was calculated in accordance with the Tauc curve. Fig. 5b suggests that by adding LiCl, a shrinkage of the band gap was observed [29,30]. Fig. 5c presents the transmission spectrum of the pristine SnO2 or Li:SnO2 film. As impacted by the band gap of SnO2, it exhibited low transmittance in the range of 300–350 nm and high transmittance in the range of 350–800 nm. Moreover, 3% and 4% Li:SnO2 film displayed a higher transmittance in the range of 350–550 nm. The charge transfer behaviors at the interface between ETL and perovskite was delved into by steady-state photoluminescence (PL) quenching of perovskites. Fig. 5d presents the PL spectra of five samples, namely, glass/ITO/SnO2/perovskite, glass/ITO/Li:SnO2/perovskite. Fluorescence emission of perovskite was identified at 770 nm. The 2%, 3%, 4% Li:SnO2 films exhibited noticeable PL quenching compared to the SnO2 film, indicating that the conductivity could be enhanced, and the trap state density could decline with LiCl addition. PL spectra suggested that 5% and 6% Li:SnO2 displayed less quenching than the pristine SnO2, since excessive Li aggregated at the interface, resulting in severe interface recombination [31]. Fig. 6 shows the statistics of PCE of 20 devices based on a range of doping concentrations. The increase in Voc was because Li enhanced the mobility. The rise in Jsc and FF resulted from the ion exchange at the lower interface Cl, which promoted the interface contact and up-regulated the grain size. Besides, the 4% Li:SnO2 devices exhibited the maximal photovoltaic parameters as compared with the others. The
degradation of 5% and 6% device performance might result from the aggregation of Cl on ETLs surface, which hinders the charge transport from perovskite to ETL [32]. The photocurrent hysteresis of SnO2 and Li:SnO2 devices was also tested, and the photocurrent hysteresis of the device was calculated. The J-V curves of the devices in a range of scanning directions are plotted in Fig. 7a, and the relevant J-V parameters are listed in Table 1. PCEreverse PCEforward The hysteresis index was calculated by h = [18,33]. PCEreverse The PCE of the SnO2 device in a range of scanning directions took up 16.96% and 13.33%, and the hysteresis index reached 21.40%. In different scanning directions, the PCE of the 4% Li:SnO2 device were 18.12% and 15.90%, and the hysteresis index was 12.25%. As impacted by the ion-exchange effect of Cl at the interface, the grain size of the perovskite increased, so the FF was enhanced. Accordingly, negligible high-efficiency hysteresis of the J-V value was achieved [19,34]. The steady-state power output measured at the maximal power is presented in Fig. 7b. The 4% Li:SnO2 device exhibited a stabilized efficiency of 18.35% at the maximal power point (0.919 V). It had stable output at 4000 s and prominent optical stability. To more effectively characterize the effect of LiCl on the characteristics of SnO2 films, the I-V characteristics are depicted in Fig. 7c. The SnO2 and Li: SnO2 films were measured with the sandwich structures ITO/Li:SnO2/Au and ITO/ SnO2/ Au. The current in the Li:SnO2 film is observed to be up-regulated. The direct current conductivity (σ0) is calculated by the formula I = 0 Ad 1 V , where d denotes the thickness of the film, A refers to the area of the film, and the conductivity σ0 of Li:SnO2 was enhanced from 0.079 mS cm−1 to 0.106 mS cm−1 as LiCl was added. According to the external quantum efficiency (EQE) curve (Fig. 7d), the device was identified exhibiting similar spectral shapes between 350 nm and 400 nm. The SnO2 and Li:SnO2-based devices achieved the maximal EQE values at 520 nm, 87.21% and 91.12%, respectively. The integral Jsc of the devices based on SnO2 and Li:SnO2 reached 19.77 mA cm−2 and 20.77 mA cm−2. The higher EQE of Li:SnO2 was attributed to the enhanced charge collected and transferred with Li addition [35–38]. Stability has posed a challenge to perovskite equipment. To compare the storage stability of the device, the storage stability was tested
4
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 5. (a) Absorption spectra of the perovskite layers of the different concentrations of LiCl. (b) The band gap calculation using the Tauc plot. (c) Transmittance spectra of the SnO2 layers of the different concentrations of LiCl. (d) PL of the perovskite layers of the different concentrations of LiCl.
Fig. 6. Statistics on the distribution of (a) Voc, (b) Jsc, (c) FF and (d) PCE of PSCs with different concentrations of octocrylene.
5
Chemical Physics Letters 745 (2020) 137220
Y. Huang, et al.
Fig. 7. (a) J–V curves of the device based on SnO2 and Li: SnO2 at reverse and forward scans. (b) Stabilized output of the cell measured at a bias voltage of 0.919 V under simulated 100 mW cm−2, AM 1.5G irradiation of the device based on Li: SnO2. (c) I–V characteristics of the ITO/Li:SnO2/Au and ITO/SnO2/Au. (d) EQE spectra of the devices comprising SnO2 and Li:SnO2 as an ETL.
4. Conclusion
Table 1 Device parameters in forward and reverse scans. Jsc (mA/cm2)
Voc (V)
FF (%)
PCE (%)
Hysteresis index
0%
Reverse Forward
21.43 21.81
1.065 1.051
74.30 58.11
16.96 13.33
21.40%
4%
Reverse Forward
22.28 22.62
1.069 1.033
76.11 68.07
18.12 15.90
12.25%
In summary, Li:SnO2 was successfully prepared as an effective ETL in PSC by a low temperature process. After doping with LiCl, the enhancement in conductivity in SnO2 promoted electron injection and transfer and hindered charge recombination. The optimized equipment achieved improved Voc, FF and Jsc. A device with a maximum PCE close to 19% was developed. Moreover, the enhancement of optical stability led to a steady state output of 18.35% at 4000 s. CRediT authorship contribution statement Yinyi Huang: Validation, Writing - original draft, Writing - review & editing. Shina Li: Methodology, Data curation. Chaorong Wu: Validation. Shuo Wang: Writing - review & editing. Chengyan Wang: Project administration, Resources, Software. Ruixin Ma: Supervision. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Fig. 8. Storage stability of the corresponding devices stored under dry conditions for 40 days without encapsulation.
Acknowledgments This work was supported by the National Natural Science Foundation of China (U1802253, U1702252, 51834008), the Guangxi Innovation-driven Development Project (AA18242042-1), and the Fundamental Research Funds for the Central Universities (FRF22-TT19-001).
under dark and dry conditions (humidity, 10–15%) for 40 days, as suggested in Fig. 8. Obviously, after 40 days, the SnO2 device retained 73.63% of the initial PCE, and the Li:SnO2 device maintained 78.89% relative to its initial PCE. 6
Y. Huang, et al.
Chemical Physics Letters 745 (2020) 137220
References
[20] F.D. Giacomo, S. Razza, F. Matteocci, J. Power Sources 251 (2014) 152–156. [21] J.P.C. Baena, L. Steier, W. Tress, M. Saliba, A. Hagfeldt, Energy Environ. Sci. 8 (2015) 2928–2934. [22] K. Choi, J. Lee, H.I. Kim, C.W. Park, G. Kim, H. Choi, S. Park, S.A. Park, T. Park, Energy Environ. Sci. 11 (2018) 3238–3247. [23] D. Yang, R. Yang, K. Wang, C. Wu, X. Zhu, J. Feng, X. Ren, G. Fang, S. Priya, S. Liu, Nat. Commun. 9 (2018) 3239. [24] S. Wang, Y. Zhu, B. Liu, C. Wang, R. Ma, J. Mater. Chem. A 7 (2019) 5353–5362. [25] Q. Jiang, L. Zhang, H. Wang, Nat. Energy 2 (2019) 16177. [26] Y. Ko, Y.R. Kim, H. Jang, Nanoscale Res. Lett. 12 (2017) 498. [27] L. Huang, X. Sun, C. Li, ACS Appl. Mater. Interfaces 26 (2017) 21909–21920. [28] L. Li, Y. Chen, Z. Liu, Adv. Mater. 28 (2016) 9862–9868. [29] T. Bu, X. Liu, Y. Zhou, Energy Environ. Sci. 10 (2017) 1039. [30] N. Cheng, W. Li, M. Zhang, H. Wu, S. Sun, Curr. Appl. Phys. 19 (2019) 25–30. [31] S. Wang, Y. Zhu, B. Liu, C. Wang, R. Ma, Ionics 25 (2019) 4509–4516. [32] P. Zhu, S. Gu, H. Tan, Adv. Energy Mater. (2019) 1903083. [33] Y. Rong, Y. Hu, S. Ravishankar, H. Liu, X. Hou, Y. Sheng, Energy Environ. Sci. 10 (2017) 2383–2391. [34] M. Li, Y. Huan, X. Yan, Z. Kang, Y. Guo, Y. Li, ChemSusChem. (2017). [35] J.H. Heo, M.S. You, M.H. Chang, Nano Energy 15 (2015) 530–539. [36] D.P. Joseph, P. Renugambal, M. Saravanan, Thin Solid Films. 517 (2009) 6129–6136. [37] Y. Zhu, S. Wang, R. Ma, C. Wang, Org. Electron. 74 (2019) 228–236. [38] Z. Yang, J. Xie, V. Arivazhagan, Nano Energy 40 (2019) 345–351.
[1] A. Kojima, K. Teshima, Shirai, Y. Shirai, T. Miyasaka, J. Am. Chem. Soc. 131 (2009) 6050–6051. [2] J.K. Im, C.R. Lee, J.W. Lee, S.W. Park, N.G. Park, Nanoscale 3 (2011) 4088–4093. [3] N. Ahn, D.Y. Son, I.H. Jang, J. Am. Chem. Soc. 137 (2015) 8696–8699. [4] NREL, Efficiency chart, https://www.nrel.gov/pv/assets/images/efficiency-chart20190926.jpg, 2019. [5] Saleem, A. Saleem, M. Imran, M. Arshad, Appl. Phys. A (2019) 125–229. [6] Z.P. Wang, Q.Q. Lin, W. Bernard, Nature Energy 3 (2019) 855–861. [7] J.J. Ma, X.L. Zheng, Sol. RRL 10 (2017). [8] Y. Liu, Z. Hong, Q. Chen, Nano Lett. 15 (2015) 662–668. [9] M. Thomas, T. Thue, Adv. Energy Mater. 5 (2015) 1500569. [10] P. Shen, M.N. Yao, G.X. Wang, J. Mater. Chem. A 10 (2018) 1039. [11] E. Calabrò, F. Matteocci, A.L. Palma, Sol. Energy Mater. Sol. Cells 185 (2018) 136–144. [12] P. Docampo, J.M. Ball, M. Darwich, Nat. Commun. 4 (2013) 2761. [13] M. Park, J.Y. Kim, H.J. Son, Nano Energy 26 (2016) 208–215. [14] L. Xiong, M. Qin, C. Chen, Adv. Funct. Mater. 28 (2018) 1706276. [15] J. You, L. Meng, T. Song, Nat. Nanotechnol. 11 (2016). [16] S.S. Shin, W.S. Yang, E.J. Yeom, Chem. Lett. 7 (2016) 1845–1851. [17] W. Ke, G. Fang, Q. Liu, J. Am. Chem. Soc. 137 (2015) 6730–6733. [18] Y. Shao, Z. Xiao, C. Bi, Nat. Commun. 5 (2014) 5784. [19] N. Pellet, P. Gao, G. Gregori, Angew. Chem. Int. Ed. 53 (2014) 3151–3157.
7