Highly Efficient 2D/3D Mixed-Dimensional Cs2PbI2Cl2/CsPbI2.5Br0.5 Perovskite Solar Cells Prepared by Methanol/Isopropanol Treatment

All-inorganic perovskite solar cells are attractive photovoltaic devices because of their excellent optoelectronic performance and thermal stability. Unfortunately, the currently used efficient inorganic perovskite materials can spontaneously transform into undesirable phases without light-absorption properties. Studies have been carried out to stabilize all-inorganic perovskite by mixing low-dimensional perovskite. Compared with organic two-dimensional (2D) perovskite, inorganic 2D Cs2PbI2Cl2 shows superior thermal stability. Our group has successfully fabricated 2D/3D mixed-dimensional Cs2PbI2Cl2/CsPbI2.5Br0.5 films with increasing phase stability. The high boiling point of dimethyl sulfoxide (DMSO) makes it a preferred solvent in the preparation of Cs2PbI2Cl2/CsPbI2.5Br0.5 inorganic perovskite. When the perovskite films are prepared by the one-step solution method, it is difficult to evaporate the residual solvent molecules from the prefabricated films, resulting in films with rough surface morphology and high defect density. This study used the rapid precipitation method to control the formation of perovskite by treating it with methanol/isopropanol (MT/IPA) mixed solvent to produce densely packed, smooth, and high-crystallized perovskite films. The bulk defects and the carrier transport barrier of the interface were effectively reduced, which decreased the recombination of the carriers in the device. As a result, this effectively improved photoelectric performance. Through treatment with MT/IPA, the photoelectric conversion efficiency (PCE) of solar cells prepared in the N2 atmosphere increased from 13.44% to 14.10%, and the PCE of the device prepared in the air increased from 3.52% to 8.91%.


Introduction
Organic-inorganic hybrid perovskites exhibiting attractive optoelectronic performance characteristics, such as significant light extinction, excellent defect tolerance, and superior charge transport, are among the most favorable photovoltaic materials [1][2][3]. These hybrid perovskite solar cells' (PSCs) certified PCE has reached 25.7% [4,5]. Despite their extraordinary achievements, their stability remains a major challenge for commercialization. Considering the decomposable property of CH 3 NH 3 + (MA + ) and CH(NH 2 ) 2 + (FA + ), hybrid perovskites tend to be unstable at high temperatures, generally suffering from the issue of thermal instability [6][7][8]. Replacing the organic cation with inorganic Cs + is regarded as one effective approach to increase thermal stability [9]. The tolerance temperature of inorganic cesium-based perovskite can be up to 400 • C [10]. Nowadays, many studies have been devoted to obtaining efficient and stable inorganic PSCs, and the PCE has been increased up to 21.35% [11].
Zhang and co-workers [24] introduced the ethylenediamine (EDA) cation to form the 2D EDAPbI 4 perovskite, which significantly stabilized the phase of α-CsPbI 3 crystallites and passivated the surface defects. Jiang and co-workers [25] formed quasi-2D perovskites by introducing phenylethylammonium (PEA) into inorganic perovskite, and this suppressed the harmful phase transition, resulting in a PCE of 12.4%. However, the mixed-dimensional structure is mainly based on organic 2D perovskite, which is accompanied by a decrease in thermal stability. In contrast, 2D Cs 2 PbI 2 Cl 2 exhibits excellent thermodynamic structure stability [26]. Kanatzidis and co-workers [26] proved that Cs 2 PbI 2 Cl 2 is suitable for solar cell absorption materials owing to its wide band gap of 3.04 eV, strong ultraviolet (UV) response, and long-term stability. Xu and co-workers [27] made a theoretical prediction that 2D Cs 2 PbI 2 Cl 2 displays ultrahigh carrier mobility of 9.39 × 10 3 cm 2 /V/s and confirmed its very promising photoelectric performance. Our group [28] successfully fabricated high-quality 2D/3D mixed-dimensional Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 films, which significantly improved the phase stability, and achieved a considerable increase in efficiency of 15.09%. These results show the potential application prospects of the mixed-dimensional Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 perovskite structure.
As we know, DMSO has satisfactory halide coordination ability and solubility, so it is the preferred solvent for preparing Cs 2 PbI 2 Cl 2 /CsPbI 2.5 B r0.5 perovskite. However, its high boiling point (189 • C) [29] makes the residual DMSO difficult to evaporate from the films during the one-step solution method, resulting in rough surface morphology and tending to create defects. Due to the resulting defect density, which is considerable, a large number of carrier recombination losses limits the performance of the device [30,31]. Therefore, it is required to adjust the crystallization kinetics to improve the quality of the Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 films. In reference to the research on antisolvent strategy, the use of an antisolvent with high DMSO solubility and a low boiling point is an effective strategy to remove residual solvent molecules from the prefabricated film and form high-quality perovskite film [32].
Here, a mixed solution of methanol (MT) and isopropanol (IPA), which has a weak electron donor ability [33], low boiling point, and poor solubility with halides [34], was used to assist the preparation of Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 perovskite films. By inducing rapid nucleation and crystallization of the perovskite, the MT/IPA mixed solution can improve the morphology of the films, resulting in densely packed, smooth, and high-crystallized perovskite films. The bulk defects and the carrier transmission barrier of the interface could be obviously reduced, which is valuable for decreasing the recombination of the carriers in the device. As a result, the photoelectric performance is improved, and the PCE of Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 solar cell devices increases from 13.44% to 14.10%. This method can also be applied to the preparation in the air.

Perovskite Solar Cell Fabrication
The TiO 2 compact layer was prepared on the substrates using the spraying method. The substrates were pre-heated to 90 • C, and then perovskite was spin coated at a rate of 3000 rpm for 50 s, followed by annealing for 10 min at 325 • C. For the MT/IPA treatment, the mixed solution of MT/IPA was dropped about 15 s before the end of the spin coating. After that, the 2,2 ,7,7 -Tetrakis(N,N-p-dimethoxyphenylamino)-9,9 -spirobifluorene (spiro-OMeTAD) was spin coated at a rate of 4000 rpm for 20 s. Thermal evaporation was then used to deposit 65 nm of gold (Au) to finalize the fabrication of the solar cells with an active area of 0.12 cm 2 .

Characterization
An X-ray diffractometer (XRD, SmartLab, Rigaku, Tokyo, Japan) was used to measure the X-ray diffraction (XRD) spectra. The diffraction angles between 5 • and 50 • were scanned at a speed of 10 • per min. An SU8010 scanning electron microscope (SEM) (Hitachi, Tokyo, Japan) was used to obtain the SEM images. Atomic force microscope (AFM) images were recorded by Agilent5500 AFM (Agilent, Santa Clara, CA, USA). Ultraviolet-visible (UV-vis) spectra were recorded by a UV-vis spectrophotometer (UV2450, Shimadzu, Kyoto, Japan). The steady-state photoluminescence (PL) spectra and the time-resolved photoluminescence (TRPL) spectra were recorded by a spectrophotometer (F900, Edinburgh Instruments, Livingston, UK). Films prepared for PL and TRPL were deposited on glass without conducting layers. Electrochemical impedance spectroscopy (EIS) was obtained by the electrochemical workstation (Zahner, Kansas City, MO, USA). A quantum efficiency measurement system (Enli Technology, Kaohsiung, Taiwan) was used to measure the incident photon-to-electron conversion efficiency (IPCE). A source meter (Keithley 2400) with a sunlight simulator (Xes-300T1, SANEI Electric, Tokyo, Japan) was used to measure the current density-voltage (J-V) curves and space charge limited current (SCLC). For SCLC measurement, the dual-electron transport layer devices were prepared by replacing the hole transport layer (HTL) with an electron transport material, [6,6]-phenyl C 61 butyric acid methyl ester (PCBM).

Results and Discussion
Firstly, CsPbI 2.5 Br 0.5 (3D) and Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 (2D/3D) perovskite films were prepared using the traditional one-step solution method. In order to form the 2D/3D films, the precursor solution for 2D/3D perovskite was prepared by adding 0.075 mol% CsCl and PbI 2 to a DMSO solution containing 1 M CsPbI 2.5 Br 0.5 at a molar ratio of 2:1 [28]. XRD spectra interpretation was carried out in order to demonstrate the production of 2D Cs 2 PbI 2 Cl 2 in the film and successfully generate mixed-dimension perovskite, as shown in Figure 1a. Around 14.6 • and 29.3 • , there are two significant diffraction peaks that are apparent in both 3D and 2D/3D films, assigned to (100) and (200) crystal planes of the cubic CsPbI 2.5 Br 0.5 . No visible shift in the diffraction peaks assigned to cubic phases is observed in the 2D/3D film. Rather, the peaks are exactly the same as those in the 3D film. Furthermore, the XRD spectra have been locally magnified around 9.3 • and 28.2 • . The indexed tiny diffraction peaks that are hard to spot are the (002) and (006) crystal planes of Cs 2 PbI 2 Cl 2 appear in the 2D/3D films, which indicates the formation of 2D Cs 2 PbI 2 Cl 2 phase. Moreover, accompanied by the formation of the Cs 2 PbI 2 Cl 2 , there was an apparent increase in the (100) and (200) peaks of the CsPbI 2.5 Br 0.5 . This indicates that the introduction of Cs 2 PbI 2 Cl 2 results in a significant enhancement in the crystallinity of 2D/3D film.
In order to investigate the effect that the mixed solution of methanol/isopropanol (MT/IPA) has on films that have not been annealed, a measurement of the light absorption was carried out. Figure S1  increased, so did the absorption intensity of the unannealed films, indicating that MT promotes the oligomer to transform into the perovskite structure. The devices with the typical planar structure were fabricated in order to demonstrate how the MT/IPA-treated method affects the performance of devices, and to explore an optimal volume ratio of MT/IPA that maximizes device efficiency. The J-V curves of the devices prepared by these films are shown in Figure S2 (Supplementary Materials). The device that was treated with an MT/IPA volume ratio of 50% had the best PCE result: 14.10%. Therefore, the 50% MT/IPA-treated devices were further studied to clarify the effect of treatment with MT/IPA mixed solution. In order to investigate the effect that the mixed solution of methanol/isoprop (MT/IPA) has on films that have not been annealed, a measurement of the light absor was carried out. Figure S1 (Supplementary Materials) shows the UV-vis spectra of va volume ratios of MT/IPA-treated unannealed films. When the volume ratio of MT increased, so did the absorption intensity of the unannealed films, indicating that MT motes the oligomer to transform into the perovskite structure. The devices with the ty planar structure were fabricated in order to demonstrate how the MT/IPA-treated me affects the performance of devices, and to explore an optimal volume ratio of MT/IPA maximizes device efficiency. The J-V curves of the devices prepared by these film shown in Figure S2 (Supplementary Materials). The device that was treated wit MT/IPA volume ratio of 50% had the best PCE result: 14.10%. Therefore, the 50% MT treated devices were further studied to clarify the effect of treatment with MT/IPA m solution.
Investigations were carried out in order to further explore the differences tha curred in the morphology of the 2D/3D films after being treated with MT/IPA. Figure are the SEM images. These images demonstrate that the boundaries of grains in untr film are apparent, and that there are large gaps between the grains. In contrast, the g in MT/IPA-treated film are denser, which can facilitate carrier transport in the film. M over, it was discovered that the average grain size of the untreated film was about nm, while the grains in films that were treated with MT/IPA were reduced to abou nm. It has been identified that the decrease of the perovskite crystal size can bring a the reduction of surface Gibbs free energy, which is helpful in improving stability The AFM images of both untreated and MT/IPA-treated films are shown in Figure   Investigations were carried out in order to further explore the differences that occurred in the morphology of the 2D/3D films after being treated with MT/IPA. Figure 2a,b are the SEM images. These images demonstrate that the boundaries of grains in untreated film are apparent, and that there are large gaps between the grains. In contrast, the grains in MT/IPA-treated film are denser, which can facilitate carrier transport in the film. Moreover, it was discovered that the average grain size of the untreated film was about 1000 nm, while the grains in films that were treated with MT/IPA were reduced to about 550 nm. It has been identified that the decrease of the perovskite crystal size can bring about the reduction of surface Gibbs free energy, which is helpful in improving stability [35]. The AFM images of both untreated and MT/IPA-treated films are shown in Figure 2c,d. The untreated film has a root means square (RMS) roughness of 38.03 nm, whereas the RMS of the MT/IPAtreated film is only 15.72 nm. The roughness of the film has been greatly reduced by MT/IPA treatment. The high boiling point of DMSO and its strong coordination with the precursor make it difficult for it to evaporate uniformly during annealing. However, the MT/IPA solution causes the rapid removal of DMSO from the precursor solution, which means that the saturation of the solution increases rapidly, resulting in uniform perovskite nucleation. This ensures that DMSO can evaporate completely during the subsequent annealing process, allowing the grains to grow at a uniform speed and leaving a smooth surface. Owing to the smooth surface, which is valuable for interface contact, carrier transport effectively improves. Furthermore, the dense grain surface can reduce the invasion of water into the film, thus improving the humidity stability of the perovskite. Hence, treatment with MT/IPA during the Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 crystallization process is helpful when preparing high-quality and stable perovskite solar films. Figure 3a shows the XRD spectra of untreated and MT/IPA-treated films. The diffraction peaks stand at about 14.6 • and 29.3 • in both of the films, assigned to the (100) and (200) crystal planes of the CsPbI 2.5 Br 0.5 , indicating that the MT/IPA treatment has not changed the composition of the CsPbI 2.5 Br 0.5 . The diffraction peaks of the cubic perovskite CsPbI 2.5 Br 0.5 of the MT/IPA-treated film are greatly increased compared to those of the un-treated film, indicating the MT/IPA treatment enhanced the crystallinity of the perovskite. According to Figure 3b, the UV-vis absorption spectra demonstrate that the untreated film has the same absorption intensity as the MT/IPA-treated film. The absorption edges are both around 680 nm, and the corresponding band gap is 1.82 eV, providing further evidence that the perovskite composition has not changed. Additionally, both of the films have equal absorption intensity. These results show that the MT/IPA treatment has no noticeable effect on the absorption properties of the films.   The effect of MT/IPA treatment on the photoluminescence properties of the films was further investigated. According to Figure 3c, the emission peaks shown in the PL spectra are consistent with the absorption edge results that were previously obtained in the UV-vis absorption spectra. Both the peaks of the untreated and MT/IPA-treated films are at about 680 nm. The MT/IPA-treated film has a higher PL peak than the untreated one, indicating that the perovskite films treated by MT/IPA have fewer defects. Moreover, as shown in Figure 3d, the TRPL was measured to obtain the decay curves of the PL peaks' intensities. The spectra were fitted using the biexponential function. In addition, the parameters that describe the decay curves are shown in Table 1. τ 1 represents the rapid attenuation component, which is generally considered to be caused by the fast recombination of the film surface interface [36]. τ 2 represents the slow attenuation component, which is generally thought to be caused by the recombination of the carriers in the film. The weights of τ 1 and τ 2 are given by the values of A 1 and A 2 , respectively. The weighted average τ ave of τ 1 and τ 2 is used to parameterize the carrier lifetime. According to Table 1, the τ ave of Nanomaterials 2023, 13, 1239 6 of 11 the untreated perovskite film is 8.58 ns, while that of MT/IPA-treated films is 18.23 ns. The longer carrier lifetime further confirms that fewer defects exist in the treated films, indicating that the carrier non-radiative recombination of films treated with MT/IPA is effectively suppressed. The effect of MT/IPA treatment on the photoluminescence properties of the film further investigated. According to Figure 3c, the emission peaks shown in the PL s are consistent with the absorption edge results that were previously obtained in th vis absorption spectra. Both the peaks of the untreated and MT/IPA-treated films about 680 nm. The MT/IPA-treated film has a higher PL peak than the untreated o dicating that the perovskite films treated by MT/IPA have fewer defects. Moreo shown in Figure 3d, the TRPL was measured to obtain the decay curves of the PL intensities. The spectra were fitted using the biexponential function. In addition, t rameters that describe the decay curves are shown in Table 1. τ1 represents the ra tenuation component, which is generally considered to be caused by the fast recom tion of the film surface interface [36]. τ2 represents the slow attenuation component, is generally thought to be caused by the recombination of the carriers in the film weights of τ1 and τ2 are given by the values of A1 and A2, respectively. The weight erage τave of τ1 and τ2 is used to parameterize the carrier lifetime. According to Table τave of the untreated perovskite film is 8.58 ns, while that of MT/IPA-treated films i  The effect of MT/IPA treatment on carrier recombination in the device was further characterized. As shown in Figure 4a, the dual-electron transport layer devices were tested by SCLC. It was found that the untreated film has a trap-filling limit voltage (V TFL ) of 0.41V, while the V TFL of the MT/IPA-treated film is reduced to 0.29 V. The following formula demonstrates that the trap state density (N t ) is proportional to the V TFL [37].    The reduced trap-filling limit voltage indicates a decrease in defects. The detailed result of the calculation is that the N t of the MT/IPA-treated film (1.32 × 10 16 cm −3 ) is much lower in comparison to that of the untreated film (1.87 × 10 16 cm −3 ), indicating that the MT/IPA treatment helps reduce the defects of the film and decreasing the recombination rate. Figure 4b shows the EIS of the untreated and MT/IPA-treated devices. The equivalent circuits were fitted by the EIS Nyquist diagram, and the parameters are shown in Table 2. The untreated device has a series resistance (R s ) of 22.39 Ω, while the R s of the MT/IPA-treated device is reduced to 15.34 Ω. The recombination resistance (R rec ) of the untreated device is 3480 Ω, and that of the MT/IPA-treated device is 6497 Ω. Lower series resistance and higher recombination resistance mean favorable carrier transport and reduced recombination, which is significant for the improvement of device performance. Besides, the MA/IPA-treated film has a minor bulk defect and a smooth surface morphology, both of which contribute to a reduction in the interface recombination loss that occurs within devices by facilitating hole transportation.   Figure 5c,d, that the major factor for improving device efficiency is significant increase in V oc . This is mainly because the MT/IPA-treated devices have fewer defects and good interfacial contact with the perovskite layer and HTL. The long-term stability of the untreated and MT/IPA-treated solar cells in N 2 is shown in Figure S3 (Supplementary Materials). It can be seen that both devices exhibit excellent stability. In our study, the untreated solar cell retained 83% of the initial PCE, while the MT/IPA-treated solar cell demonstrated better stability, maintaining 87% of the initial PCE after 2400 h. The improved stability should be attributed to the reduced defect density and the densely packed crystals in the MT/IPA-treated film [11]. treated sample has excellent photoelectric conversion efficiency. Integrating the contr tions of photons of different wavelengths, the integrated current density of 14.83 mA/ for the untreated device and 15.19 mA/cm 2 for the MT/IPA-treated device are basic consistent with their J-V curves. It can be concluded from the distribution box diagr of Jsc and Voc of the devices, shown in Figure 5c,d, that the major factor for improving vice efficiency is significant increase in Voc. This is mainly because the MT/IPA-tre devices have fewer defects and good interfacial contact with the perovskite layer HTL. The long-term stability of the untreated and MT/IPA-treated solar cells in N shown in Figure S3 (Supplementary Materials). It can be seen that both devices exh excellent stability. In our study, the untreated solar cell retained 83% of the initial P while the MT/IPA-treated solar cell demonstrated better stability, maintaining 87% of initial PCE after 2400 h. The improved stability should be attributed to the reduced de density and the densely packed crystals in the MT/IPA-treated film [11]. To better verify the effect of the MT/IPA treatment on devices, the 2D/3D dev without treatment and with MT/IPA treatment were prepared in air with a relative midity of 10-20%. Figure 5e shows the XRD spectra of the films prepared in the air. H the diffraction peaks of the cubic CsPbI2.5Br0.5 of MT/IPA-treated film are significa To better verify the effect of the MT/IPA treatment on devices, the 2D/3D devices without treatment and with MT/IPA treatment were prepared in air with a relative humidity of 10-20%. Figure 5e shows the XRD spectra of the films prepared in the air. Here, the diffraction peaks of the cubic CsPbI 2.5 Br 0.5 of MT/IPA-treated film are significantly higher than those of the untreated film. As evidenced by the XRD results, the poor crystallinity of the untreated film could have been due to water molecules and impurities in the air that affected the crystallization process. In contrast, the MT/IPA solution could accelerate the evaporation of DMSO, resulting in rapid nucleation and crystallization of perovskite, thus minimizing the influence of air exposure. As a result, the MT/IPA treatment was beneficial in improving the photoelectric performance of devices prepared in the air. The PCE of the devices employed by MT/IPA treatment was increased from 3.52% to 8.91%, as shown in Figure 5f.

Conclusions
In conclusion, the MT/IPA-treated method was employed to regulate the crystallization kinetics of Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 perovskite film. The MT/IPA mixed solution has a weak electron donor ability, low boiling point, and poor solubility with halides, which can accelerate the evaporation of DMSO, thus facilitating uniform nucleation and growth within films. By employing the MT/IPA-treated method, the bulk defects and the carrier transmission barrier of the interface were decreased. This was valuable for the reduction of the recombination of the carriers in the device. As a result, the PCE of Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 solar cells increased from 13.44% to 14.10%. This study proves that the MT/IPA-treated method is beneficial to the preparation of densely packed, smooth, and high-crystallized Cs 2 PbI 2 Cl 2 /CsPbI 2.5 Br 0.5 perovskite films and can be applied to their preparation in the air. Data Availability Statement: The data are available from the authors on request.

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