Fabrication and Characterization of SnCl 2 - and CuBr-Added Perovskite Photovoltaic Devices

: Perovskite photovoltaic devices added with tin (Sn) dichloride and copper (Cu) bromide were fabricated and characterized. The thin ﬁlm devices were prepared by an ordinary spin-coating technique using an air blowing method in ambient air. A decaphenylcyclopentasilane layer was coated at the surface of perovskite layer and annealed at a high temperature of 190 ◦ C. Conversion efﬁciencies and short-circuit current densities were improved for devices added with Sn and Cu compared with the standard devices. The energy gap of the perovskite crystal decreased through the Sn addition, which was also conﬁrmed by ﬁrst-principles calculations.


Introduction
Perovskite solar cells provide photoelectric conversion in wide wavelength ranges and exhibit excellent photovoltaic properties [1][2][3][4][5][6]. Since the thin films of CH 3 NH 3 PbI 3 (MAPbI 3 ) can be formed by a spin-coating method [7,8], there is the advantages of an easy production process and low cost. Despite these merits, there is a problem that the stability of the perovskite compound is extremely low because of desorption of CH 3 NH 3 (MA), and the lead used in the perovskite layer is toxic [9][10][11]. On the other hand, the photovoltaic properties of perovskite solar cells have been reported to depend strongly on the compositions and crystal structures of the perovskite compounds [12][13][14]. Therefore, to solve these problems, studies have been conducted in which various elements are doped in a perovskite precursor solution [15][16][17][18][19][20].
Several studies on partial replacement of MA in MAPbI 3 by formamidinium (HC(NH 2 ) 2 , FA) have been reported [21][22][23][24][25][26][27]. One of these studies reported that the conversion efficiency of MAPbI 3 perovskite solar cells decreased to 50% of the initial value within one month [28]. On the other hand, it was reported that the conversion efficiency of the FA-added device was maintained about 80% of the initial value within one month after the production. The addition of FA can be expected to improve the stability of the perovskite solar cells.
There are also studies using Sn as an alternative element to Pb because of its toxicity [29][30][31][32]. MASnI 3 is unstable in the atmosphere, and the Sn-added device fabricated in a nitrogen atmosphere has an initial conversion efficiency of 5.38% [33]. When the device was held in a nitrogen atmosphere for 24 h, the conversion efficiency was reduced to 1.94%, and it was difficult to maintain the conversion efficiency [33].
The purpose of this study is to fabricate and characterize tin (Sn)-and copper (Cu)added perovskite solar cells, which were prepared by high temperature annealing at 190 • C in ambient air using decaphenylcyclopentasilane (DPPS). The addition of Sn and Cu can be expected to influence crystal growth of the perovskite grains, which would lead to increase the short-circuit current density [34][35][36]. Although the annealing temperatures for the fabrication of the MAPbI 3 perovskite layers are commonly 100 • C, this low-temperature Technologies 2022, 10, 112 2 of 12 annealing might cause instability. Since high-temperature annealing could be effective for the formation of more stable perovskites, hot air-blowing [37,38] and higher temperature annealing in ambient air using DPPS [39,40] were applied in the present work. The DPPS layer is expected to work as a protective layer and a hole transport layer on the perovskite layer. The effects of Sn-and Cu-additions to the perovskite crystals on the photovoltaic properties were investigated using light-induced current density-voltage (J-V) measurements, external quantum efficiency, X-ray diffraction (XRD), optical microscopy, and first-principle calculation.

Materials and Methods
A schematic illustration showing the fabrication processes used to fabricate photovoltaic cells is shown in Figure 1. All processes were performed in air [41,42]. F-doped tin oxide (FTO, Nippon Sheet Glass Company, Ltd., Tokyo, Japan) substrates were cleaned in an ultrasonic bath with acetone and ethanol, and dried under nitrogen gas. Subsequently, the FTO substrates were treated with an ultraviolet ozone cleaner (ASM401N, Asumi Giken, Tokyo, Japan) for 15 min. and chlorobenzene (Fujifilm Wako Pure Chemical Corporation) [39,40]. The thickness of the DPPS layer was about 10 nm. The device was annealed at 190 °C for 15 min in the ambient air.
The J-V characteristics (B2901A, Keysight, Santa Rosa, CA, USA) of the photovoltaic cells were measured when illuminated with a solar simulator (XES-301S, San-ei Electric, Osaka, Japan) at 100 mW cm −2 with an air mass 1.5. The solar cells were illuminated through the side of the FTO substrates, and the exposed area was 0.080 cm 2 . Three electrodes were measured for each device, and the maximum and minimum values were indicated as error bars. The EQEs of the solar cells were also measured by an incident photon to electron conversion efficiency measurement system (QE-R, Enli Technology, Kaohsiung, Taiwan). The microstructures of the perovskite layers were investigated by an X-ray diffractometer (D2 PHASER, Bruker, Billerica, MA, USA), while the surface morphologies of the perovskite layers were examined using an optical microscope (Eclipse E600, Nikon, Tokyo, Japan).   TiO x precursor solutions (0.15 and 0.30 M) were prepared from titanium diisopropoxide bis(acetyl acetonate) (Sigma Aldrich, Tokyo, Japan) and 1-butanol (Fujifilm Wako Pure Chemical Corporation, Osaka, Japan). Both TiO x precursor solutions were spin-coated onto the FTO substrate at 3000 rpm for 30 s (MSA100, Mikasa, Tokyo, Japan) and annealed at 125 • C for 5 min, however, the 0.30 M precursor was spin-coated twice to form a uniform layer. After that, the FTO substrate was sintered at 550 • C for 30 min to form a compact TiO 2 layer. A mesoporous TiO 2 layer was spin-coated on top of the compact TiO 2 layer at 5000 rpm for 30 s using TiO 2 paste. The TiO 2 paste was prepared using TiO 2 powder (P-25, Aerosil, Tokyo, Japan, 100 mg) and poly(ethylene glycol) (PEG, Nacalai Tesque, Kyoto, Japan, #20000, 10 mg) in distilled water (0.5 mL). This solution was mixed with acetylacetone (Fujifilm Wako Pure Chemical Corporation, 10 µL) and the surfactant Triton X-100 (Sigma Aldrich, 5 µL) for 30 min and was then allowed to stand for~24 h to suppress bubble formation within the solution. The FTO substrates with the TiO 2 were annealed at 550 • C for 30 min to form the mesoporous TiO 2 layer.
The perovskite compounds were prepared by mixing solutions of CH 3 NH 3 I (Tokyo Chemical Industries, Tokyo Japan), HC(NH 2 ) 2 I (FAI, Tokyo Chemical Industries), PbCI 2 (Sigma Aldrich), and SnCl 2 (Fujifilm Wako Pure Chemical Corporation), CuBr (Sigma Aldrich), at 70 • C. Formamidinium [HC(NH 2 ) 2 ] (FA) with a larger ionic radius than MA was added to improve the stability [43,44]. All materials were dissolved in N,N-dimethylformamide (Sigma Aldrich).  Table 1. The chemical reaction of 3MAI + PbCl 2 → MAPbI 3 + 2MACl (↑) was used in the present work [37], and the MACl was vaporized during annealing. To stabilize the perovskite structure, FA was co-added when Sn and Cu were added to the crystals. As the FAI, Sn, and Cu composition increased, the tolerance factor (t-factor) [14] increased a little, indicating that the crystals have similar stabilities from the viewpoint of t-factors. The perovskite precursor solutions were spin-coated at 2000 rpm for 60 s and applied an air-blowing method during spin-coating [37]. The perovskite precursor solution was spin-coated three times, and DPPS was added dropwise during the third spin coating (45 s after the start). The DPPS solutions were prepared from decaphenylcyclopentasilane (SI-30-15, Osaka Gas Chemical, Osaka, Japan) and chlorobenzene (Fujifilm Wako Pure Chemical Corporation) [39,40]. The thickness of the DPPS layer was about 10 nm. The device was annealed at 190 • C for 15 min in the ambient air. The hole-transport layer was deposited by spin-coating. A chlorobenzene solution (0.5 mL) of 2,2 ,7,7 -tetrakis-(N,N-di-p-methoxyphenylamine)-9,9 -spirobifluorene (spiro-OMeTAD, Fujifilm Wako Pure Chemical Corporation, 36.1 mg) was prepared by mixing it for 12 h. An acetonitrile solution (0.5 mL) of lithium bis (trifluoromethylsulfonyl) imide (Li-TFSI, Tokyo Chemical Industry) was also prepared by mixing it for 12 h. A mixture solution of the spiro-OMeTAD solution with 4-tertbutylpridine (14.4 µL, Sigma Aldrich) and Li-TFSI solution (8.8 µL) was prepared by mixing it at 70 • C for 30 min. The spiro-OMeTAD layer was deposited by spin-coating at 4000 rpm for 30 s. After that, gold (Au) thin film electrodes were deposited as electrodes by using a resistance heating type vacuum evaporation system (SVC-700TM, Sanyu Electron, Co., Ltd., Tokyo, Japan).
The J-V characteristics (B2901A, Keysight, Santa Rosa, CA, USA) of the photovoltaic cells were measured when illuminated with a solar simulator (XES-301S, San-ei Electric, Osaka, Japan) at 100 mW cm −2 with an air mass 1.5. The solar cells were illuminated through the side of the FTO substrates, and the exposed area was 0.080 cm 2 . Three electrodes were measured for each device, and the maximum and minimum values were indicated as error bars. The EQEs of the solar cells were also measured by an incident photon to electron conversion efficiency measurement system (QE-R, Enli Technology, Kaohsiung, Taiwan). The microstructures of the perovskite layers were investigated by an X-ray diffractometer (D2 PHASER, Bruker, Billerica, MA, USA), while the surface morphologies of the perovskite layers were examined using an optical microscope (Eclipse E600, Nikon, Tokyo, Japan).
The ab initio quantum calculations used the Vanderbilt ultra-soft pseudo-potentials, scalar relativistic generalized gradient approximations, and the Perdew-Burke-Ernzerhof exchange-correlation function and density functional theory (Quantum Espresso, version 5.2.1, Quantum Espresso Foundation, Cambridge, UK). The initial parameters used for the crystal were 2 × 2 × 2 supercells, the kinetic energy cut-off for the wave functions of 25 Ry, and the charge density of 225 Ry.

Results and Discussion
The photovoltaic properties of perovskite-based solar cells were investigated using the J-V curves obtained under illumination. The J-V curves of the solar cells featuring FAI 10% + Sn 2%, FAI 10% + Sn 1% + Cu 1%, FAI 10% + Sn 2% + Cu 2%, and FAI 10% + Sn 2% + Cu 4% incorporated MAPbI 3 perovskites are presented in Figure 2a. The photovoltaic parameters, namely open-circuit voltages (V OC ), short-circuit current densities (J SC ), fill factor (FF), series resistance (R s ), shunt resistance (R sh ), photoconversion efficiency (η), averaged photoconversion efficiency (η ave ) and the energy gap (E g ) of all analyzed cells are listed in Table 2. The E g values decreased by adding Sn, which also corresponded to the previous study [33].
The ab initio quantum calculations used the Vanderbilt ultra-soft pseudo-potentials, scalar relativistic generalized gradient approximations, and the Perdew-Burke-Ernzerhof exchange-correlation function and density functional theory (Quantum Espresso, version 5.2.1, Quantum Espresso Foundation, Cambridge, UK). The initial parameters used for the crystal were 2 × 2 × 2 supercells, the kinetic energy cut-off for the wave functions of 25 Ry, and the charge density of 225 Ry.

Results and Discussion
The photovoltaic properties of perovskite-based solar cells were investigated using the J-V curves obtained under illumination. The J-V curves of the solar cells featuring FAI 10% + Sn 2%, FAI 10% + Sn 1% + Cu 1%, FAI 10% + Sn 2% + Cu 2%, and FAI 10% + Sn 2% + Cu 4% incorporated MAPbI3 perovskites are presented in Figure 2a. The photovoltaic parameters, namely open-circuit voltages (VOC), short-circuit current densities (JSC), fill factor (FF), series resistance (Rs), shunt resistance (Rsh), photoconversion efficiency (η), averaged photoconversion efficiency (ηave) and the energy gap (Eg) of all analyzed cells are listed in Table 2. The Eg values decreased by adding Sn, which also corresponded to the previous study [33].  Addition of SnCl2, and CuBr to MA0.9FA0.1PbI3 perovskite affected the photovoltaic parameters, and the JSC, VOC, FF, and η values of the Sn-added cells were lower than those of the standard device. When 1%, 2%, and 4% CuBr were co-added to MAPbI3 perovskite, the η values of the cells increased.
The external quantum efficiencies (EQEs), which is the ratio of the number of charge carriers collected by the solar cell to the number of incident photons, of the devices including FAI, SnCl2, and CuBr are shown in Figure 2b. The addition of FAI or SnCl2 to the MAPbI3 perovskite expanded the region of EQE to 400-840 nm. Energy gaps (Eg) were  Addition of SnCl 2 , and CuBr to MA 0.9 FA 0.1 PbI 3 perovskite affected the photovoltaic parameters, and the J SC , V OC , FF, and η values of the Sn-added cells were lower than those of the standard device. When 1%, 2%, and 4% CuBr were co-added to MAPbI 3 perovskite, the η values of the cells increased.
The external quantum efficiencies (EQEs), which is the ratio of the number of charge carriers collected by the solar cell to the number of incident photons, of the devices including FAI, SnCl 2 , and CuBr are shown in Figure 2b. The addition of FAI or SnCl 2 to the MAPbI 3 perovskite expanded the region of EQE to 400-840 nm. Energy gaps (E g ) were calculated from the EQE spectra, and are listed in Table 2. The E g values of Sn-added and Sn/Cuadded perovskites were lower than that of the standard MAPbI 3 crystal.
The changes in η and J SC of perovskite solar cells were investigated at 25 • C and 20% humidity for about 200 days, as shown in Figure 3. For the MAPbI 3 solar cell, both η and J SC decreased after 203 days. The degradation would be due to the decrease in photocurrent by carrier recombination around defects which were formed by diffusion of MA cations and halogen anions for the long-term period. calculated from the EQE spectra, and are listed in Table 2. The Eg values of Sn-added and Sn/Cu-added perovskites were lower than that of the standard MAPbI3 crystal.
The changes in η and JSC of perovskite solar cells were investigated at 25 °C and 20% humidity for about 200 days, as shown in Figure 3. For the MAPbI3 solar cell, both η and JSC decreased after 203 days. The degradation would be due to the decrease in photocurrent by carrier recombination around defects which were formed by diffusion of MA cations and halogen anions for the long-term period. For the 10% FA and 2% Sn-added device, both η and JSC increased after 193 days. The increase in JSC and η would be due to promotion of carrier diffusion in the perovskite crystals. The co-addition of FA and Sn would facilitate the solid-phase reaction with diffusion in the perovskite layer. On the other hand, slow deterioration behavior was observed for the standard MAPbI3 device.
J-V characteristics and photovoltaic parameters of each device with the best efficiency for the time course are summarized in Figure 4 and Table 3. Sn 2% added device provided the best efficiencies in the present work. Since several devices showed small increases of efficiencies, their long-term performances were also investigated, as listed in Table 4. Although efficiencies of Sn-added devices decreased, device with Sn 2% + Cu 2% showed better stability than Sn 2% device.  For the 10% FA and 2% Sn-added device, both η and J SC increased after 193 days. The increase in J SC and η would be due to promotion of carrier diffusion in the perovskite crystals. The co-addition of FA and Sn would facilitate the solid-phase reaction with diffusion in the perovskite layer. On the other hand, slow deterioration behavior was observed for the standard MAPbI 3 device.
J-V characteristics and photovoltaic parameters of each device with the best efficiency for the time course are summarized in Figure 4 and Table 3. Sn 2% added device provided the best efficiencies in the present work. Since several devices showed small increases of efficiencies, their long-term performances were also investigated, as listed in Table 4. Although efficiencies of Sn-added devices decreased, device with Sn 2% + Cu 2% showed better stability than Sn 2% device.
Sn/Cu-added perovskites were lower than that of the standard MAPbI3 crystal.
The changes in η and JSC of perovskite solar cells were investigated at 25 °C and 20% humidity for about 200 days, as shown in Figure 3. For the MAPbI3 solar cell, both η and JSC decreased after 203 days. The degradation would be due to the decrease in photocurrent by carrier recombination around defects which were formed by diffusion of MA cations and halogen anions for the long-term period. For the 10% FA and 2% Sn-added device, both η and JSC increased after 193 days. The increase in JSC and η would be due to promotion of carrier diffusion in the perovskite crystals. The co-addition of FA and Sn would facilitate the solid-phase reaction with diffusion in the perovskite layer. On the other hand, slow deterioration behavior was observed for the standard MAPbI3 device.
J-V characteristics and photovoltaic parameters of each device with the best efficiency for the time course are summarized in Figure 4 and Table 3. Sn 2% added device provided the best efficiencies in the present work. Since several devices showed small increases of efficiencies, their long-term performances were also investigated, as listed in Table 4. Although efficiencies of Sn-added devices decreased, device with Sn 2% + Cu 2% showed better stability than Sn 2% device.   Perovskite solar cells without FA and DPPS layer were also investigated, and the photovoltaic parameters are listed in Table 5. The compositions of the perovskite compounds were CH 3 NH 3 Pb 0.95 Sn 0.05 I 3 and CH 3 NH 3 Pb 0.9 Sn 0.05 Cu 0.05 I 3 , which are denoted as Sn 5% and Sn 5% + Cu 5%, respectively. Since there was no DPPS layer, the annealing temperature of the perovskite layers was 140 • C for 10 min. For the Sn-added device, both η and J SC increased after 115 days, which agreed with the results of Figure 3, indicating the effectiveness of Sn addition.  Optical microscopy images of the present solar cells are shown in Figure 5, and the particle sizes and surface coverages of the perovskite crystals are summarized in a part of Table 6. The average particle sizes were measured from dark regions of the image, and the surface coverage was calculated from the histogram of the contrast by using an image software (Adobe, Photoshop). Lattice constants and 100 crystal orientation in Table 6 were measured from XRD data in Figure 6, which will be described later. The particle size and surface coverage of the MAPbI 3 perovskite incorporated with 2% Sn increased from 5.45 µm to 7.68 µm and from 67.95 to 93.35, respectively, after 193 days. On the other hand, for the solar cell featuring pristine MAPbI 3 perovskite, the particle size and surface coverage decreased from 6.77 µm to 6.46 µm and from 75.65 to 68.05, respectively. to 2.1. An intensity ratio of the 100 peak (I100) to 210 peak (I210) for the Sn 2% + Cu 2% added device increased further to 3.7 after 283 days. Addition of a small amount of Sn and Cu improved the crystal orientation and suppressed the grain boundaries in the perovskite layer. The crystal growth in the perovskite layer would reduce the trap density between perovskite grains and increase the JSC and η [36].    to 2.1. An intensity ratio of the 100 peak (I100) to 210 peak (I210) for the Sn 2% + Cu 2% added device increased further to 3.7 after 283 days. Addition of a small amount of Sn and Cu improved the crystal orientation and suppressed the grain boundaries in the perovskite layer. The crystal growth in the perovskite layer would reduce the trap density between perovskite grains and increase the JSC and η [36].    XRD patterns and lattice constants (a) of the standard, +Sn, and +Sn + Cu devices are shown in Figure 6 and Table 6. An intensity of the 100 peak for the Sn 2% added device increased after 200 days, and the crystal orientation of the I 100 /I 210 ratio increased from 2.4 to 2.6, as listed in Table 6. This indicates crystal growth of (100)-oriented perovskite grains after 200 days. In comparison, the I 100 /I 210 ratio for the standard device decreased from 2.5 to 2.1. An intensity ratio of the 100 peak (I 100 ) to 210 peak (I 210 ) for the Sn 2% + Cu 2% added device increased further to 3.7 after 283 days. Addition of a small amount of Sn and Cu improved the crystal orientation and suppressed the grain boundaries in the perovskite layer. The crystal growth in the perovskite layer would reduce the trap density between perovskite grains and increase the J SC and η [36]. Figure 7a,b are schematic illustrations of crystal growth of the perovskites and the ionic diffusion in the perovskite, respectively. It is considered that MA and I were desorbed by annealing the perovskite layer during device fabrication. The energy derived from these lattice defects would cause ionic diffusion, which would promote grain growth. Then, the surface coverage and crystal orientation are improved, resulting in an increase in short-circuit current density and conversion efficiency [45][46][47][48].
Sn and Cu ions would have smaller activation energy for diffusion, which could contribute to the crystal growth even by the room temperature aging, as observed in the XRD patterns of Figure 6.
The added Sn 2+ can be easily oxidized into Sn 4+ , which would act as non-radiative recombination centers, degrading solar cell performance. However, the perovskite device with a small amount of Sn showed improvement of the performance in the present work, and high temperature annealing might be related with this device stability. In addition to the device characterization, the partial density of states (PDOS), total density of states (TDOS) and electron density distribution were investigated by first-principles calculation [50][51][52][53][54][55]. The structure models of MAPbI3 and MA(FA)Pb(Sn)I3 used for the calculations are shown in Figure 8a,b, and the PDOS are shown in Figure 8c  Growth rates of perovskite grains for the devices were compared. The initial grain size of the crystal is d 0 (m), the material/temperature-dependent constant is K (m 2 s −1 ), the constant is n, and the temperature retention time is t r (s). The crystal grain size was d (m), and the constant K was calculated using the following formula d n − d n 0 = Kt r , and listed in Table 6. The n changes depending on the variable, and the n was set as 2 because of a single perovskite layer. d and d 0 values were determined from the particle sizes. The K values for the standard and Sn/Cu added systems negative and positive, respectively.
Activation energies of ionic diffusions of MA + , Pb 2+ , and I − are 0.84 eV, 2.31 eV, and 0.58 eV, respectively [49]. The Pb ion has the largest activation energy for ionic diffusion in this crystal, and crystal growth of MAPbI 3 at room temperature might be suppressed by the difficulty of the Pb diffusion. Since Sn and Cu have smaller ionic radii than Pb, the Sn and Cu ions would have smaller activation energy for diffusion, which could contribute to the crystal growth even by the room temperature aging, as observed in the XRD patterns of Figure 6.
The added Sn 2+ can be easily oxidized into Sn 4+ , which would act as non-radiative recombination centers, degrading solar cell performance. However, the perovskite device with a small amount of Sn showed improvement of the performance in the present work, and high temperature annealing might be related with this device stability.
In addition to the device characterization, the partial density of states (PDOS), total density of states (TDOS) and electron density distribution were investigated by first-principles calculation [50][51][52][53][54][55]. The structure models of MAPbI 3 and MA(FA)Pb(Sn)I 3 used for the calculations are shown in Figure 8a,b, and the PDOS are shown in Figure 8c,d, respectively. The PDOS of Sn/FA doped perovskite is different from that of the standard MAPbI 3 , especially around the conduction band. For the MAPbI 3 , the p-orbital of Pb is dominant around the conduction band, whereas the p-orbital of Sn is dominant for the MA(FA)Pb(Sn)I 3 . Bandgap energies of the MAPbI 3 and MA(FA)Pb(Sn)I 3 were calculated as 1.382 and 1.178 eV, respectively, which agree with the results obtained in the experimental system.
Calculation results of electron density distribution for the MAPbI 3 and MA(FA)Pb(Sn)I 3 are shown in Figure 8e

Conclusions
Solar cells using perovskite in the photoactive layer were fabricated by using a spincoating technique in ordinary air, and the influences on photovoltaic characteristics by adding Sn and Cu to the perovskite compounds were investigated. Addition of Sn to the MAPbI 3 perovskite crystals improved the long-term photovoltaic properties owing to the crystal growth and improvement of (100) crystal orientation of the perovskite layer. The Cu addition with Sn also enhanced the (100) orientation. The partial substitution of Pb 2+ in the perovskite crystals by Sn 2+ ions affected charge transfer via hybridization between the p orbitals of I and Sn.