Controlled Growth of Porous InBr3: PbBr2 Film for Preparation of CsPbBr3 in Carbon-Based Planar Perovskite Solar Cells

Due to the low solubility of CsBr in organic solvents, the CsPbBr3 film prepared by the multi-step method has holes and insufficient thickness, and the light absorption capacity and current density of the perovskite film hinder the further improvement in the power conversion efficiency (PCE) of CsPbBr3 solar cells. In this study, we introduced InBr3 into the PbBr2 precursor solution and adjusted the concentration of PbBr2, successfully prepared PbBr2 with a porous structure on the compact TiO2 (c-TiO2) substrate to ensure that it fully reacted with CsBr, and obtained the planar carbon-based CsPbBr3 solar cells with high-quality perovskite film. The results reveal that the porous PbBr2 structure and the increasing PbBr2 concentration are beneficial to increase the thickness of the CsPbBr3 films, optimize the surface morphology, and significantly enhance the light absorption capacity. Finally, the PCE of the CsPbBr3 solar cells obtained after conditions optimization was 5.76%.


Introduction
From the perspective of the preparation process, a very unique advantage of perovskite solar cells (PSCs) is that they have better processability [1][2][3][4]. Whether an organic-inorganic hybrid or all-inorganic perovskite solar cell, the film quality of the perovskite active layer (including morphology, grain shape, grain size, number of grain boundaries, whether there are pinholes, etc.) determines the PCE and durability of PSCs [5][6][7][8][9]. Generally, the prepared perovskite film should have a high-density microstructure, large grain size, low grain boundary density, and no pinholes or other voids in the film, to improve electron/hole mobility, achieve reduction in radiation recombination, and other purposes [1,[10][11][12]. Both the preparation method of the perovskite film and the device structure (planar or mesoporous structure) can exert a significant impact on the quality and photoelectric performance of the perovskite film. In the process of preparing perovskite films by the two-step method, mesoporous substrates are often used to promote the PCE of PSCs. This is because the mesoporous layer formed by oxide nanoparticles has a higher specific surface area, and a large number of pores promote the improvement in the effective diffusivity of the lead halide precursor solution [13], so denser lead halide nanosized grains can be generated faster in the mesoporous layer, and the conversion process of lead halide to perovskite can be accelerated [1,14]. However, during this conversion process, due to some unconverted lead halide residues in the mesoporous layer and the possible formation of random crystals in the perovskite film [14,15], the efficiency of PSCs may be greatly affected. A method for fully converting the lead halide is a critical prerequisite for controlling the growth of perovskite films, or ensuring the perfect formation of crystals and enhancing the efficiency of PSCs [16][17][18].
Introducing a porous lead halide film into a planar structure is an effective method to replace the mesoporous structure in the two-step method [1]. The porous structure in the lead halide film can enlarge the contact area between the lead halide and other precursor solutions and increase the reaction rate, and it can reduce space-expansioninduced defects and residual stress during the expansion of the film [5]. Liu et al. [19] spin-coated a solution of PbI 2 in DMF onto a substrate, then prepared a PbI 2 film with a porous structure by adjusting the standing time. Ostwald ripening of the PbI 2 crystallites can result in larger crystals and increase the void space between grains. The increase in porosity in the PbI 2 film effectively promotes the transformation process of perovskite and reduces the residual amount of PbI 2 . The resulting CH 3 NH 3 PbI 3 PSCs with inverted planar heterojunction have an efficiency of 15.7%, and there is almost no hysteresis. Dong et al. [20] confirmed that the porosity in PbI 2 can affect the phase composition and film morphology of CsPbI 2 Br. Compared with other anti-solvents, the PbI 2 (DMSO) film treated with green ethanol has higher porosity and randomly distributed crystals, which provide more diffusion paths and contact areas for the CsBr precursor, and create an essential prerequisite for ultimately obtaining CsPbI 2 Br film with high purity, high thickness, large grains, and full coverage. Although the CsPbBr 3 film prepared by the multi-step method exhibits excellent moisture and thermal stability [21][22][23][24], it is also prone to problems such as low film coverage, thin film thickness, and the presence of impurity phases (such as CsPb 2 Br 5 ) [5,25,26]. As mentioned above, a porous PbBr 2 film with higher porosity can provide more diffusion paths for the CsBr precursor and increase the contact area between the two, which ultimately provides a favorable condition for the formation of CsPbBr 3 with high purity phase and high coverage [27][28][29]. Zhao et al. [21] prepared porous PbBr 2 films on SnO 2 substrates by precisely controlling the crystallization temperature. The porous structure of PbBr 2 not only allows the effective diffusion of CsBr solution to gain high-purity CsPbBr 3 , but also offers enough space for the growth of CsPbBr 3 grains under stress-free conditions, resulting in high-quality CsPbBr 3 film with larger grain size and lower grain boundary density. In our previous work, we prepared porous PbBr 2 films on mesoporous TiO 2 (m-TiO 2 ) by introducing InBr 3 into the PbBr 2 precursor solution, and achieved In 3+ or In cluster doping with CsPbBr 3 , thereby improving the growth quality of CsPbBr 3 films and enhancing the efficiency of CsPbBr 3 solar cells with mesoporous structure [30].
At present, it is a highly attractive choice to use carbon electrodes, which are low cost and highly stable, to construct CsPbBr 3 PSCs without hole transport layer (HTL). Carbon, which matches the work function of perovskite, can effectively transport holes, and avoid the instability of traditional organic HTL (such as Spiro-MeOTAD) and the low conductivity of metal oxide HTL. At the same time, compared with Au electrodes, the carbon electrode is lower cost, and does not diffuse into the perovskite and cause the device performance to deteriorate as does the Ag electrode [31][32][33]. On the basis of our previous research results, in this work, we introduced InBr 3 into a PbBr 2 precursor solution and adjusted the concentration of PbBr 2 to prepare a high-quality CsPbBr 3 film with high thickness, high density, and no holes on c-TiO 2 , in order to further simplify the structure of the CsPbBr 3 PSCs and optimize the growth quality of the CsPbBr 3 film. When the concentration of InBr 3 is 0.21 M and PbBr 2 concentration is 1.3 M, the prepared carbon-based CsPbBr 3 solar cells with a planar structure show the best 5.76% PCE with a short circuit current density (J SC ) of 6.52 mA/cm 2 , an open circuit voltage (V OC ) of 1.29 V, and a fill factor (FF) of 0.68.

Device Fabrication
All the following processes were carried out in an air atmosphere. The c-TiO 2 was spin-coated on pretreated FTO substrates using 0.2 M titanium isopropoxide in ethanol at 5000 rpm for 30 s and dried for 10 min at 120 • C, then the substrates were annealed at 500 • C for 1 h. Perovskite films were synthesized by a multistep solution-processing method. We added 0.21 mmol of InBr 3 to 1 mL DMF solution of 1.0, 1.1, 1.2, 1.3, or 1.4 M PbBr 2 , which was stirred under 90 • C. Afterward, the mixed solution was spin-coated on the FTO/c-TiO 2 substrate at 2000 rpm for 30 s and heated to 90 • C for 30 min to obtain InBr 3 :PbBr 2 films with a porous structure. Then, the methanol solution of CsBr (0.07 M) was spin-coated on InBr 3 :PbBr 2 film at 5000 rpm for 30 s and continuingly heated at 250 • C for 5 min. This step was performed 6 times. After the prepared sample was soaked in isopropanol for 30 min and annealed at 250 • C for 15 min, the blade coating method was applied to coat carbon paste onto the perovskite films and heated at 100 • C for 10 min to form the carbon back electrode with the effective area of 0.09 cm −2 .

Characterization
The X-ray diffraction (XRD) patterns of the synthesized sample were investigated using an X-ray diffractometer (Cu Kα radiation, λ = 1.5418 Å, Rigaku D/max2500, Tokyo, Japan). The morphologies of the synthesized films were recorded by a scanning electron microscope (SEM, FEI MAGELLAN 400, FEI, Hillsboro, OR, USA), and the system was connected to an energy-dispersive X-ray spectroscopy (EDS). X-ray photoelectron spectroscopy (XPS) was performed by an X-ray photoelectron spectrometer system (K-Alpha, Thermo Fisher Scientific, Waltham, MA, USA) equipped with a monochromatic Al Kα X-ray source (1486.6 eV) operating at 100 W. The steady-state photoluminescence (PL) and the time-resolved photoluminescence (TRPL) decay spectrums of perovskite films were collected on a photoluminescence spectrometer (FLS980, Edinburgh Instruments, Livingston, UK) with a 473 nm excitation source. The absorption spectrum was measured by a UV-Vis spectrometer (UV-3600, Shimadzu, TKY, Japan) in the range of 200 to 800 nm. The current-voltage (J-V) characteristics were measured by a solar cells test system (XP3000, Sanyou, China) equipped with a Keithley 2400 source meter. The external quantum efficiency (EQE) was tested by an EQE measured system (QTest Station 1000A, CROWNTECH, INC., Macungie, PA, USA) under DC mode. Figure 1 displays the top-view SEM images prepared without and with introducing 0.21 M InBr 3 in PbBr 2 precursors of different solubility. When the PbBr 2 precursor concentration was 1.0 M without introducing InBr 3 , the PbBr 2 film produced was uniform, dense, crack-free, and had a smooth surface, and there were pores of different sizes in the film (Figure 1a). In contrast, after the introduction of InBr 3 into the 1.0 M PbBr 2 precursor solution, as shown in Figure 1b, the InBr 3 :CsPbBr 3 film had a porous structure, the crystal grains were mostly block structures with obvious size differences, and there were relatively small gaps between the crystal grains. This indicates that InBr 3 can transform the PbBr 2 grown on the c-TiO 2 substrate from a dense film to a relatively low-density porous film. Compared with the porous PbBr 2 film formed on m-TiO 2 , the morphology of the PbBr 2 film prepared on c-TiO 2 changed significantly regardless of the introduction of InBr 3 , which reveals that both the TiO 2 substrate and InBr 3 can directly affect the morphology of the PbBr 2 film and determine the quality of the subsequent CsPbBr 3 film growth. With the further increase in the PbBr 2 concentration (1.1~1.4 M), although the films prepared under various conditions were still porous films, the morphology and porosity of the films did Nanomaterials 2021, 11, 2408 4 of 12 not illustrate obvious differences. The columnar and cubic crystal grains grouped into clusters, with gaps between the clusters, which provide space for the subsequent growth and free expansion of CsPbBr 3 crystals [13,34]. Based on the cross-sectional SEM images of the corresponding PbBr 2 films ( Figure S1, Supplementary Materials), it can be concluded that the thickness of the film (330, 400, 450, 500, and 550 nm), the grain size, and the gap between the crystal grains (clusters) all exhibit an increasing trend with the increase in the PbBr 2 concentration, which provides the pivotal condition to enhance the diffusion rate of the CsBr precursor and increase the contact area and reaction efficiency between CsBr and PbBr 2 , and subsequently guarantee the production of high-quality CsPbBr 3 films with high thickness, high purity, high density, and large crystal grains.

Results and Discussion
PbBr2 film prepared on c-TiO2 changed significantly regardless of the introduction InBr3, which reveals that both the TiO2 substrate and InBr3 can directly affect the morph ogy of the PbBr2 film and determine the quality of the subsequent CsPbBr3 film grow With the further increase in the PbBr2 concentration (1.1~1.4 M), although the films p pared under various conditions were still porous films, the morphology and porosity the films did not illustrate obvious differences. The columnar and cubic crystal gra grouped into clusters, with gaps between the clusters, which provide space for the sub quent growth and free expansion of CsPbBr3 crystals [13,34]. Based on the cross-sectio SEM images of the corresponding PbBr2 films ( Figure S1, Supplementary Materials), it c be concluded that the thickness of the film (330, 400, 450, 500, and 550 nm), the grain si and the gap between the crystal grains (clusters) all exhibit an increasing trend with increase in the PbBr2 concentration, which provides the pivotal condition to enhance t diffusion rate of the CsBr precursor and increase the contact area and reaction efficien between CsBr and PbBr2, and subsequently guarantee the production of high-quality C bBr3 films with high thickness, high purity, high density, and large crystal grains. To validate the influence of porous lead bromide films made by introducing InBr3 the morphologies of the CsPbBr3 films, the top-view and cross-sectional SEM images lustrated in Figure 2 were examined. When the PbBr2 precursor concentration was 1.0 without introducing InBr3, that is, when the PbBr2 film was a flat film, the thickness of t obtained CsPbBr3 film was about 290 nm, the film showed poor uniformity and covera and many holes in the film (Figure 2a). With the further increase in PbBr2 concentrati (1.1~1.4 M), the thickness of each perovskite film was 310, 330, 350, and 370 nm, resp tively ( Figure S2, Supplementary Material). When 0.21 M InBr3 was introduced into t PbBr2 precursor solution, with the increase in the PbBr2 concentration (1.0~1.3 M), the C bBr3 film showed a satisfying growth relationship with c-TiO2, and the film thickness s nificantly increased to approximately 430, 470, 500, and 540 nm, respectively, which veals that the porous structure of the PbBr2 film is the main factor affecting the increase the thickness of the CsPbBr3 film. Meanwhile, the coverage of the CsPbBr3 film gradua rose with the increase in the PbBr2 concentration, with the pores in the film gradua decreasing, the average size of the crystal grains increasing, and the density and u formity of the perovskite film significantly improving. In particular, when the PbBr2 co centration was 1.3 M, the surface of the CsPbBr3 film was smooth and dense with alm no holes. As mentioned above, the porous structure of the PbBr2 film can increase To validate the influence of porous lead bromide films made by introducing InBr 3 on the morphologies of the CsPbBr 3 films, the top-view and cross-sectional SEM images illustrated in Figure 2 were examined. When the PbBr 2 precursor concentration was 1.0 M without introducing InBr 3 , that is, when the PbBr 2 film was a flat film, the thickness of the obtained CsPbBr 3 film was about 290 nm, the film showed poor uniformity and coverage, and many holes in the film (Figure 2a). With the further increase in PbBr 2 concentration (1.1~1.4 M), the thickness of each perovskite film was 310, 330, 350, and 370 nm, respectively ( Figure S2, Supplementary Material). When 0.21 M InBr 3 was introduced into the PbBr 2 precursor solution, with the increase in the PbBr 2 concentration (1.0~1.3 M), the CsPbBr 3 film showed a satisfying growth relationship with c-TiO 2 , and the film thickness significantly increased to approximately 430, 470, 500, and 540 nm, respectively, which reveals that the porous structure of the PbBr 2 film is the main factor affecting the increase in the thickness of the CsPbBr 3 film. Meanwhile, the coverage of the CsPbBr 3 film gradually rose with the increase in the PbBr 2 concentration, with the pores in the film gradually decreasing, the average size of the crystal grains increasing, and the density and uniformity of the perovskite film significantly improving. In particular, when the PbBr 2 concentration was 1.3 M, the surface of the CsPbBr 3 film was smooth and dense with almost no holes. As mentioned above, the porous structure of the PbBr 2 film can increase the contact area between the CsBr precursor and PbBr 2 and make the two fully react, thereby making the process of converting PbBr 2 to CsPbBr 3 faster and more complete. With the increase in the PbBr 2 concentration, that is, as the gap between the PbBr 2 crystal grain clusters gradually enlarges, more uniform perovskite nuclei with a narrower size distribution are generated in the porous PbBr 2 film in a shorter time. Small grains with unfavorable growth orientation transfer their monomers through the solid-state grain boundary to at least a few larger grains with decent growth orientation, until smaller grains disappear to generate large grains with the largest surface and interface area, and finally producing a large-grain CsPbBr 3 film that completely covers the substrate and has no holes [1,35]. However, as the concentration of the PbBr 2 precursor containing InBr 3 was further increased to 1.4 M, the thickness of the formed CsPbBr 3 film further increased to about 580 nm, but the surface morphology of the CsPbBr 3 film began to deteriorate, and holes appeared in the film again, which can adversely affect the J SC of CsPbBr 3 solar cells [36].
contact area between the CsBr precursor and PbBr2 and make the two fully react, there making the process of converting PbBr2 to CsPbBr3 faster and more complete. With t increase in the PbBr2 concentration, that is, as the gap between the PbBr2 crystal gra clusters gradually enlarges, more uniform perovskite nuclei with a narrower size dist bution are generated in the porous PbBr2 film in a shorter time. Small grains with un vorable growth orientation transfer their monomers through the solid-state grain boun ary to at least a few larger grains with decent growth orientation, until smaller gra disappear to generate large grains with the largest surface and interface area, and fina producing a large-grain CsPbBr3 film that completely covers the substrate and has holes [1,35]. However, as the concentration of the PbBr2 precursor containing InBr3 w further increased to 1.4 M, the thickness of the formed CsPbBr3 film further increased about 580 nm, but the surface morphology of the CsPbBr3 film began to deteriorate, a holes appeared in the film again, which can adversely affect the JSC of CsPbBr3 solar ce [36]. To investigate the effect of porous PbBr2 film on the phase, structure, and eleme composition of the perovskite, the XRD and EDS mappings were employed to charact ize the CsPbBr3 films. The XRD patterns demonstrated in Figure 3a reveal that all CsPb films are cubic structures (PDF# 54-0752) [21,37], which reflects that the introduction InBr3 and the concentration change of the PbBr2 film will not change the phase of CsPbB When the PbBr2 precursor concentration was 1.0 M without introducing InBr3, a stro diffraction peak belonging to the PbBr2-rich CsPb2Br5 phase appeared at 2θ of 11.7°. contrast, after introducing InBr3 into the 1.0 M PbBr2 precursor, the diffraction peak b longing to CsPb2Br5 phase disappeared (CsPb2Br5 will not be completely consumed), b the diffraction peak belonging to Cs4PbBr6 phase appeared at 2θ of 12.7°. This is becau after the porous PbBr2 film with low thickness is fully contacted with the CsBr soluti and is completely consumed, the excess CsBr and CsPbBr3 continue to react and form CsBr-rich Cs4PbBr6 phase. On the contrary, in the planar structure of PbBr2, the cont area and the reaction efficiency between CsBr and PbBr2 are relatively low, and thus Pb is not fully consumed. After the introduction of InBr3, as the concentration of the Pb precursor gradually increased from 1.1 to 1.4 M, the intensity of the diffraction peaks b longing to the CsPb2Br5 phase in the XRD patterns of each sample showed an upwa trend, and no obvious diffraction peak belonging to Cs4PbBr6 phase was observed. T cross-sectional EDS mapping of the CsPbBr3 film shown in Figure 3b proves that Cs, P Br, and In are evenly distributed in the analyzed area, and there is no Pb or In gatheri in a certain area. To investigate the effect of porous PbBr 2 film on the phase, structure, and element composition of the perovskite, the XRD and EDS mappings were employed to characterize the CsPbBr 3 films. The XRD patterns demonstrated in Figure 3a reveal that all CsPbBr 3 films are cubic structures (PDF# 54-0752) [21,37], which reflects that the introduction of InBr 3 and the concentration change of the PbBr 2 film will not change the phase of CsPbBr 3 . When the PbBr 2 precursor concentration was 1.0 M without introducing InBr 3 , a strong diffraction peak belonging to the PbBr 2 -rich CsPb 2 Br 5 phase appeared at 2θ of 11.7 • . In contrast, after introducing InBr 3 into the 1.0 M PbBr 2 precursor, the diffraction peak belonging to CsPb 2 Br 5 phase disappeared (CsPb 2 Br 5 will not be completely consumed), but the diffraction peak belonging to Cs 4 PbBr 6 phase appeared at 2θ of 12.7 • . This is because after the porous PbBr 2 film with low thickness is fully contacted with the CsBr solution and is completely consumed, the excess CsBr and CsPbBr 3 continue to react and form a CsBr-rich Cs 4 PbBr 6 phase. On the contrary, in the planar structure of PbBr 2 , the contact area and the reaction efficiency between CsBr and PbBr 2 are relatively low, and thus PbBr 2 is not fully consumed. After the introduction of InBr 3 , as the concentration of the PbBr 2 precursor gradually increased from 1.1 to 1.4 M, the intensity of the diffraction peaks belonging to the CsPb 2 Br 5 phase in the XRD patterns of each sample showed an upward trend, and no obvious diffraction peak belonging to Cs 4 PbBr 6 phase was observed. The cross-sectional EDS mapping of the CsPbBr 3 film shown in Figure 3b proves that Cs, Pb, Br, and In are evenly distributed in the analyzed area, and there is no Pb or In gathering in a certain area. Nanomaterials 2021, 11, x FOR PEER REVIEW 6 of 12 Afterward, we coated the prepared CsPbBr3 films with commercial carbon paste, and dried them to obtain the planar HTL-free PSCs with the structure of FTO/c-TiO2/CsP-bBr3/carbon; the cross-sectional SEM image of the device is shown in Figure 4a. In order to further explore the effect of PbBr2 porous film on the device performance, Figure 4b Table 1. When the PbBr2 concentration was gradually increased from 1.0 to 1.4 M without introducing InBr3 into the precursor solution, the efficiency of each CsPbBr3 cell was at a relatively low level, and there was little difference in the results. It is noteworthy that the efficiency of the CsPbBr3 cell obtained under the condition of a PbBr2 concentration of 1.2 M was the highest, namely 1.79%, while the efficiency of the other cells was between 1.5% and 1.7%. After the introduction of InBr3 into the PbBr2 precursor solution, the efficiency of all PSCs was significantly improved. As the PbBr2 concentration increased from 1.0 to 1.3 M, JSC, VOC, FF, and PCE all showed a trend of increasing with the increase in PbBr2 concentration. When the PbBr2 concentration was 1.3 M, the champion device with a PCE of 5.76% showed a JSC of 6.52 mA/cm 2 , a VOC of 1.29 V, and an FF of 0.68. When the PbBr2 concentration was further increased to 1.4 M, the efficiency of the device decreased to 4.97%, and JSC, VOC, and FF also illustrated a slight decline, caused by the slight deterioration in the morphology of CsPbBr3. The statistics of photovoltaic parameters collected from 15 devices ( Figure S3, Supplementary Materials) and the average photovoltaic parameters with a small standard deviation (Table S1, Supplementary Materials) demonstrate that the InBr3:CsPbBr3 PSCs synthesized with different PbBr2 concentrations obtained in this work have decent reliability and reproducibility. Figure 4g displays the EQE and the corresponding integrated photocurrent densities to verify the validity of the CsPbBr3 devices based on InBr3. The photoresponse edges of these five devices are all around 540 nm. When the PbBr2 concentration is 1.3 M and the maximum EQE value of the champion device is 83%, the EQE values in the light absorption wavelength range of 350-525 nm are significantly enhanced, which proves that the device has better charge injection ability and collection capacity. In addition, the integrated photocurrent density of each device is closer to the corresponding JSC, and the mismatch is less than 5%, which signifies the reliability of the J-V curves. Afterward, we coated the prepared CsPbBr 3 films with commercial carbon paste, and dried them to obtain the planar HTL-free PSCs with the structure of FTO/c-TiO 2 /CsPbBr 3 / carbon; the cross-sectional SEM image of the device is shown in Figure 4a. In order to further explore the effect of PbBr 2 porous film on the device performance, Figure 4b Table 1. When the PbBr 2 concentration was gradually increased from 1.0 to 1.4 M without introducing InBr 3 into the precursor solution, the efficiency of each CsPbBr 3 cell was at a relatively low level, and there was little difference in the results. It is noteworthy that the efficiency of the CsPbBr 3 cell obtained under the condition of a PbBr 2 concentration of 1.2 M was the highest, namely 1.79%, while the efficiency of the other cells was between 1.5% and 1.7%. After the introduction of InBr 3 into the PbBr 2 precursor solution, the efficiency of all PSCs was significantly improved. As the PbBr 2 concentration increased from 1.0 to 1.3 M, J SC , V OC , FF, and PCE all showed a trend of increasing with the increase in PbBr 2 concentration. When the PbBr 2 concentration was 1.3 M, the champion device with a PCE of 5.76% showed a J SC of 6.52 mA/cm 2 , a V OC of 1.29 V, and an FF of 0.68. When the PbBr 2 concentration was further increased to 1.4 M, the efficiency of the device decreased to 4.97%, and J SC , V OC , and FF also illustrated a slight decline, caused by the slight deterioration in the morphology of CsPbBr 3 . The statistics of photovoltaic parameters collected from 15 devices ( Figure S3, Supplementary Materials) and the average photovoltaic parameters with a small standard deviation (Table S1, Supplementary Materials) demonstrate that the InBr 3 :CsPbBr 3 PSCs synthesized with different PbBr 2 concentrations obtained in this work have decent reliability and reproducibility. Figure 4g displays the EQE and the corresponding integrated photocurrent densities to verify the validity of the CsPbBr 3 devices based on InBr 3 . The photoresponse edges of these five devices are all around 540 nm. When the PbBr 2 concentration is 1.3 M and the maximum EQE value of the champion device is 83%, the EQE values in the light absorption wavelength range of 350-525 nm are significantly enhanced, which proves that the device has better charge injection ability and collection capacity. In addition, the integrated photocurrent density of each device is closer to the corresponding J SC , and the mismatch is less than 5%, which signifies the reliability of the J-V curves.  Furthermore, XPS was utilized to study the influence of InBr3 on the electronic of CsPbBr3 films at a PbBr2 concentration of 1.3 M. As displayed in Figure 5a, com  Furthermore, XPS was utilized to study the influence of InBr 3 on the electronic state of CsPbBr 3 films at a PbBr 2 concentration of 1.3 M. As displayed in Figure 5a, compared Nanomaterials 2021, 11, 2408 8 of 12 with the pure CsPbBr 3 , in addition to Cs 3d, Pb 4f, and Br 3d, In 3d was found in the XPS spectra for InBr 3 : CsPbBr 3 film. According to Figure 5b, In 3d 5/2 and In 3d 3/2 peaked at 444.9 eV and 452.3 eV, respectively, and the high-resolution core-level binding energies of Cs 3d, Pb 4f, and Br 3d all shifted toward higher values compared with those in the pure CsPbBr 3 , which indicates that part of the In(III) in the porous PbBr 2 was doped into the perovskite lattice in the process of transforming CsPbBr 3 , and caused the chemical state of [PbBr 6 ] 4− octahedron to change [38]. The incorporation of In 3+ or In cluster not only improves the spatial symmetry of the CsPbBr 3 lattice structure, but also reduces vacancy defects, which is beneficial to the extraction and transfer process of the charge, and finally significantly optimizes the J SC of the device [30,[38][39][40]. improves the spatial symmetry of the CsPbBr3 lattice structure, but also reduces vacancy defects, which is beneficial to the extraction and transfer process of the charge, and finally significantly optimizes the JSC of the device [30,[38][39][40]. In the optical property test, the devices with the FTO/c-TiO2/CsPbBr3 structure were further characterized by a UV-visible spectrophotometer and PL. In Figure 6a, as expected, the absorption edges of all perovskite films are at around 533 nm, which is basically the same as the EQE test result. The light absorption capacity of the films shows a trend of first increasing and then decreasing with increasing PbBr2 concentration, which is consistent with the change law of the perovskite morphology and PCE. In other words, the increase in the thickness of the CsPbBr3 film and the optimization of the surface morphology can enhance the light absorption rate of the perovskite layer and reduce the light loss, which is conducive to generating more electrons to improve JSC. Using Tuac's equation, we can calculate the corresponding band gap (Eg) patterns of the CsPbBr3 film; the curves of α 2 versus the photon energy hν, as shown in Figure 6b; and the Eg of the prepared CsPbBr3 to be about 2.35 eV. Regardless of whether InBr3 was introduced into the PbBr2 precursor, the solubility of PbBr2 did not meaningfully change the Eg of CsPbBr3. Figure S4 Figure 6c, the PL peaks for both the CsPbBr3 films were still at about 523 nm, which indicates that the change in the concentration of PbBr2 and the introduction of InBr3 do not change the band position. However, the perovskite prepared after the introduction of InBr3 into the PbBr2 precursor solution showed more observable strong quenching, which proves that the incorporation of In 3+ or In cluster and the improvement of perovskite film growth quality are beneficial to faster electron injection and collection, and reduce the recombination of electrons and holes [41,42], which explains why the JSC and VOC of the PSCs were significantly improved after the introduction of InBr3 in Figure 4b-f. Figure 6d illustrates the TRPL decay curves, and the relevant detailed pa- In the optical property test, the devices with the FTO/c-TiO 2 /CsPbBr 3 structure were further characterized by a UV-visible spectrophotometer and PL. In Figure 6a, as expected, the absorption edges of all perovskite films are at around 533 nm, which is basically the same as the EQE test result. The light absorption capacity of the films shows a trend of first increasing and then decreasing with increasing PbBr 2 concentration, which is consistent with the change law of the perovskite morphology and PCE. In other words, the increase in the thickness of the CsPbBr 3 film and the optimization of the surface morphology can enhance the light absorption rate of the perovskite layer and reduce the light loss, which is conducive to generating more electrons to improve J SC . Using Tuac's equation, we can calculate the corresponding band gap (E g ) patterns of the CsPbBr 3 film; the curves of α 2 versus the photon energy hν, as shown in Figure 6b; and the E g of the prepared CsPbBr 3 to be about 2.35 eV. Regardless of whether InBr 3 was introduced into the PbBr 2 precursor, the solubility of PbBr 2 did not meaningfully change the E g of CsPbBr 3 . Figure S4 Figure 6c, the PL peaks for both the CsPbBr 3 films were still at about 523 nm, which indicates that the change in the concentration of PbBr 2 and the introduction of InBr 3 do not change the band position. However, the perovskite prepared after the introduction of InBr 3 into the PbBr 2 precursor solution showed more observable strong quenching, which proves that the incorporation of In 3+ or In cluster and the improvement of perovskite film growth quality are beneficial to faster electron injection and collection, and reduce the recombination of electrons and holes [41,42], which explains why the J SC and V OC of the PSCs were significantly improved after the introduction of InBr 3 in Figure 4b-f. Figure 6d illustrates the TRPL decay curves, and the relevant detailed parameters of the corresponding CsPbBr 3 films are listed in Table S2. The specific lifetime of the carriers can be determined by a bi-exponential formula I = A 1 e −(τ−τ 0 )/τ 1 + A 2 e −(τ−τ 0 )/τ 2 (A 1 and A 2 are the relative amplitude, respectively; τ 1 and τ 2 are the fast and slow decay time, respectively) [43], and the average lifetimes (τ ave ) is calculated by the formula τ ave = A 1 τ 2 1 + A 2 τ 2 2 /(A 1 τ 1 + A 2 τ 2 ) [44]. Significantly decreased lifetimes of τ 1 , τ 2 , and τ ave of the InBr 3 :CsPbBr 3 film (1.35, 3.06 and 2.49 ns, respectively) compared with the control CsPbBr 3 (2.65, 5.61, and 3.95 ns, respectively) were observed, representing lower trap states in the InBr 3 :CsPbBr 3 film and an accelerated electron transport process. This phenomenon is brought about by the enhancement in the growth quality of InBr 3 :CsPbBr 3 film and the reduction in the defects of Pb 2+ and Br − after partial substitution of Pb 2+ by In 3+ or In cluster [25,39]. This is conducive to improving the V OC and fill factor (FF) of the devices [45].
Nanomaterials 2021, 11, x FOR PEER REVIEW 9 ( ) ⁄ (A1 and A2 are the relative amplitude, respectively; τ1 and τ2 are the fast slow decay time, respectively) [43], and the average lifetimes (τave) is calculated by formula τ = ( τ + τ ) ( τ + τ ) ⁄ [44]. Significantly decreased lifetimes o τ2, and τave of the InBr3:CsPbBr3 film (1.35, 3.06 and 2.49 ns, respectively) compared the control CsPbBr3 (2.65, 5.61, and 3.95 ns, respectively) were observed, represen lower trap states in the InBr3:CsPbBr3 film and an accelerated electron transport pro This phenomenon is brought about by the enhancement in the growth qualit InBr3:CsPbBr3 film and the reduction in the defects of Pb 2+ and Brafter partial substitu of Pb 2+ by In 3+ or In cluster [25,39]. This is conducive to improving the VOC and fill fa (FF) of the devices [45].

Conclusions
By introducing InBr3 into a PbBr2 precursor solution, we successfully prepared pl carbon-based CsPbBr3 solar cells with higher perovskite film quality in this work. introduction of InBr3 can transform the dense PbBr2 film originally grown on c-TiO2 a porous structure. On this basis, appropriately increasing the PbBr2 concentration in DMF precursor solution can enhance the thickness of the CsPbBr3 film and optimize surface morphology without changing the phase of CsPbBr3, which can effectively al ate the low thickness and poor quality of the CsPbBr3 film in the multi-step process. to the increase in film thickness and the incorporation of In 3+ or In cluster, the ligh sorption capacity of the CsPbBr3 film is significantly improved. Therefore, the effici

Conclusions
By introducing InBr 3 into a PbBr 2 precursor solution, we successfully prepared planar carbon-based CsPbBr 3 solar cells with higher perovskite film quality in this work. The introduction of InBr 3 can transform the dense PbBr 2 film originally grown on c-TiO 2 into a porous structure. On this basis, appropriately increasing the PbBr 2 concentration in the DMF precursor solution can enhance the thickness of the CsPbBr 3 film and optimize the surface morphology without changing the phase of CsPbBr 3 , which can effectively alleviate the low thickness and poor quality of the CsPbBr 3 film in the multi-step process.
Due to the increase in film thickness and the incorporation of In 3+ or In cluster, the light absorption capacity of the CsPbBr 3 film is significantly improved. Therefore, the efficiency and EQE of the manufactured devices are significantly optimized. As InBr 3 is introduced into the current flooding fluid and the PbBr 2 concentration is 1.3 M, the small-area CsPbBr 3 solar cell has the best PCE, which is 5.76%, and the maximum EQE value is 83%. This work provides a certain reference for the preparation of high-quality CsPbBr 3 thin films, simplifying the structure and developing planar HTL-free CsPbBr 3 solar cells.

Data Availability Statement:
The data is available on reasonable request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.