Investigation on the Stability and Efficiency of MAPbI3 and MASnI3 Thin Films for Solar Cells

Hybrid organic–inorganic halides are considered as outstanding materials when used as the absorber layer in perovskite solar cells (PSCs) because of its efficiency, relieve of fabrication and low‐cost materials. However, the content of lead (Pb) in the material may origin a dramatic after effect on human's health caused by its toxicity. Here, we investigate replacing the lead in MAPbI3 with tin (Sn) to show its influence on the growth of the film nucleation and stability of the solar device based on MASnI3. By analysing the manufactured perovskite films by scanning electron microscopy (SEM), transmission electron microscopy (TEM), X‐ray diffraction (XRD), UV–visible absorption, photoluminescence (PL) and atomic force microscopy (AFM), the properties of the thin films when lead is replaced by tin are reported. The simulation run for the case of MAPbI3 is reported, where Voc = 0.856 V, Jsc = 25.65 mA cm−2, FF = 86.09%, and ETA = 18.91%, and for MASnI3, Voc = 0.887 V, Jsc = 14.02 mA cm−2, FF = 83.72%, and ETA = 10.42%. In perovskite‐based devices using MASnI3 as absorber, it was found to be more stable despite of its lower efficiency, which could be improved by enhancing the bandgap alignment of MaSnI3. The results of this paper also allow the development of a new, reliable production system for PSCs.

DOI: 10.1002/pssa.202100664 Hybrid organic-inorganic halides are considered as outstanding materials when used as the absorber layer in perovskite solar cells (PSCs) because of its efficiency, relieve of fabrication and low-cost materials. However, the content of lead (Pb) in the material may origin a dramatic after effect on human's health caused by its toxicity. Here, we investigate replacing the lead in MAPbI 3 with tin (Sn) to show its influence on the growth of the film nucleation and stability of the solar device based on MASnI 3 . By analysing the manufactured perovskite films by scanning electron microscopy (SEM), transmission electron microscopy (TEM), X-ray diffraction (XRD), UV-visible absorption, photoluminescence (PL) and atomic force microscopy (AFM), the properties of the thin films when lead is replaced by tin are reported. The simulation run for the case of MAPbI 3 is reported, where V oc ¼ 0.856 V, J sc ¼ 25.65 mA cm À2 , FF ¼ 86.09%, and ETA ¼ 18.91%, and for MASnI 3, V oc ¼ 0.887 V, J sc ¼ 14.02 mA cm À2 , FF ¼ 83.72%, and ETA ¼ 10.42%. In perovskite-based devices using MASnI 3 as absorber, it was found to be more stable despite of its lower efficiency, which could be improved by enhancing the bandgap alignment of MaSnI 3 . The results of this paper also allow the development of a new, reliable production system for PSCs.
First, the prepared solutions of MAPbI 3 and MASnI 3 were dropped onto fluorine-dope tin oxide (FTO) glass substrates. The FTO glass was spun at 4000 rpm for 50 s during the dripping of the perovskite solution and the addition of toluene as antisolvent. Later, the samples were annealed at 60 C for 5 min, followed by an annealing at 100 ºC for 10 min. Figure 1 shows a diagram of the elaboration process carried out.

Characterization Techniques
The structural properties of the thin films were examined by XRD with RIGAKU Ultima IV diffractometer using Cu Kα radiation (λ ¼ 1.5418 Å). SEM was used to examine the surface morphology of the perovskite layers using 1.5 kV at different magnifications. AFM measurements were performed using Nano Surf with a voltage cell from À1.5 to 1.5 V at a scan rate of 0.5 Hz. TEM (JEO-JEM-1010) analysis was carried out with 2.5 kV at different magnifying tools. UV-vis absorption analysis was made with a Si charge couple device, and PL emission was excited with a 405 nm semiconductor laser. The output parameters of the solar cell, namely, in the short-circuit current density ( J sc ), the open-circuit voltage (V oc ), the fill factor (FF), and the efficiency (η), were assessed by SCAPS simulator.

Results and Discussion
The XRD scanning of MAPbI 3 and MASnI 3 thin films ( Figure 2) reveals various diffractions peaks located at 14 , 28 , and 52 matching the characteristics peaks of the perovskite materials (110), (220), and (303) respectively. Previously published structures of MAPbI 3 [9] are in good agreement with the patterns of this work. Furthermore, it was observed that the replacement of Pb by Sn leads to an increase in the intensity of the peak (110). All perovskite films form a tetragonal crystal structure with the space group I4/mcm.     www.advancedsciencenews.com www.pss-a.com The (110) XRD peak is wider for MASnI 3 than for MAPbI 3 , as revealed by the values of full width at half maximum (FWHM) for this XRD peak (FWHM (110) ¼ 0.23 for MASnI 3 and FWHM (110) ¼ 0.13 for MaPbI 3 ), which means bigger crystallite sizes for MAPbI 3, according to the well-known Scherrer equation. [10] The calculated grain size, dislocation density, lattice strain, and effective lattice strain data are represented in Table 1. The grain size for MASnI 3 (403 nm) thin films was found to have a substantial bigger grain size than MAPbI 3, which was 302 nm. So, according to XRD and AFM analysis, when Pb is replaced by Sn, grain sizes increase even if crystallite sizes decrease. Further, dislocation density and effective lattice strain have been calculated to have an idea about scarcity and deformations of the grains in the film. Figure 3a,b shows SEM images of MAPbI 3 and MASnI 3 , and the surface morphology of both thin films was smooth and contained random grain boundaries. Furthermore, MASnI 3 showed improvement of the surface morphology like previously reported investigations. [11,12] It can be clearly observed from the homogeneity of the surface that MASnI 3 had large grain size around 403 nm, whereas for MAPbI 3 film, the grain size was about 302 nm. Also, the addition of antisolvent toluene, which evaporates the solvent and therefore creates supersaturation to speed up the crystallization process, resulted in excellent film development. Moreover, temperature annealing treatment had a significant impact on the surface of the films. Results indicate that when annealing temperature increases, the crystal as well as the grain size grow considerably. Figure 3c,d shows the AFM analysis of MAPbI 3 and MASnI 3 over a 2 μm Â 2 μm region. Computed roughness of MAPbI 3 was 37.0 nm, and when Pb was replaced by Sn, roughness was 46.7 nm ( Table 1). The surface of MASnI 3 is rougher due to the huge size of hills and troughs.
TEM examinations of MASnI 3 and MAPbI 3 thin films with lattice fringe spacings of 0.28 and 0.75 nm, respectively, corresponding to (110) or (220) of the tetragonal perovskite phases, are shown in Figure 3a,c. Figure 3 exhibits the selected-area electron diffraction (SAED) spectrum and revealed that MAPbI 3 and MASnI 3 thin films are polycrystalline; validated findings ( Figure 4). [13][14][15] PL measurement was recorded in the range of 500-1100 nm, as represented in Figure 5. The PL peak intensity in the region of 700-900 nm is in good agreement with the previously reported study of MAPbI 3 . [16] The measured intensity of the PL peak in MAPbSn 3 films is about 30% higher than in MAPbI 3 films.
It is suggested that tin can be placed at the optimal level for absorbing light. This finding also boosts the enhancement crystallinity of the perovskite thin film. The UV-vis absorption spectrum of MAPbI 3 and MASnI 3 was obtained between 400 and 900 nm. The optical bandgap was calculated through the Tauc www.advancedsciencenews.com www.pss-a.com plot for the absorbance spectrum, from the equation (Ahv) 2 ¼ B (hvÀE g ). An optical bandgap of 1.60 eV and 1.62 was found for MAPbI 3 and MASnI 3 films respectively (Figure 5c,d). [17][18][19] Table 2 summarizes the optical bandgap and PL emission peaks according to PL and UV-vis absorption measurements. The energy difference between the edge of the optical absorption and the energy of the PL emission, known as Stokes shift, is found to be higher for MASnI 3 (40 meV) than for MAPbI 3 (10 meV), meaning that the bottom of the conduction band is more filled with electrons in MASnI 3 films.

Degradation Study
Environmental components like oxygen and humidity have a significant impact on the photovoltaic stability of perovskite solar devices. The degradation mechanism of methylammoniumbased perovskites when exposed to the environment has been attributed to a reduction of these compounds in PbI 2 , CH 3 NH 2 , and HI. To evaluate the degradation of MAPbI 3 and MaSnI 3 samples, we performed XRD, SEM, and PL measurements on fresh and 4 weeks-aged samples, kept under 60% of humidity and in the dark ( Figure 6).
For the 4 weeks-aged MAPbI 3 samples, XRD patterns reveal lesser intensity of characteristic peaks ( Figure 6a) and a reduction of the intensity of PL emission (Figure 6c). In the case of 4 weeksaged MASnI 3 samples, both XRD peaks and PL emission decrease in intensity, but this reduction in intensity is minor for MASnI 3 than for MAPbI 3 samples (Figure 6d,e). This fact means that MASnI 3 samples are more stable than MAPbI 3 ones. Similar results are reported in literature. [20][21][22][23][24] Further, SEM images (Figure 6b,e) support this finding. New grain boundaries, as presented in Figure 6b, appear in aged MAPbI 3 films. In the case of aged MASnI 3 films, some new defects consisting of pinholes and changes in the surface morphology with respect to no fresh samples (Figure 6e) are observed in SEM images. As a   www.advancedsciencenews.com www.pss-a.com result, the degradation process seems to be different for both types of films.

Device Manufacture and Numerical Simulation
For the study of the solar cell performance, we performed the simulation of MAPbI 3 and MASnI 3 -based Perovskites solar cells using SCAPS simulator software. The structure of solar spiro OMeTAD/MA(Pb/Sn)I 3 /TiO 2 /FTO used in the simulation is shown in the diagram (Figure 7). The simulation was run two times for OMeTAD/MAPbI 3 / TiO2/FTO and OMeTAD/MASnI 3 /TiO 2 /FTO proposed models separately. [25,26] Figure 8 and 9 show the J-V and P-V characteristics curves, and from these curves we observed that MASnI 3based solar cell is less efficient as compared with MAPbI 3 based solar cell. The bandgap of MAPbI 3 is more favorable for light absorption as compared with the other one.
The results of simulation are summarized in Table 3, where the photovoltaic characteristic parameters V oc (open-circuit voltage), J sc (short-circuit current density), FF, and ETA (conversion efficiency) are shown. When the simulation was run for the case of MAPbI 3 , we find the V oc of 0.856 V, J sc of 25.65 mA cm À2 , FF of 86.09%, and ETA of 18.91%. For MASnI 3 , we report V oc of 0.888 V, J sc of 14.02 mA cm À2 , FF of 83.72%, and ETA of 10.42%. It is worthy to notice that there is about 9% difference in the efficiencies of both devices. For Sn-based PSCs, this efficiency could be improved by enhancing the bandgap alignment of MASnI 3 with respect the hole transport layer and by adjusting film thickness.
To better show the differences of the photovoltaic parameters of MAPbI 3 and MASnI 3 , a bar graph displaying the four photovoltaic parameters for both proposed devices is shown in Figure 10. This comparison shows a well-defined difference between both devices' parameters.

Conclusion
In summary, the use of Sn in spite of Pb to increase the stability and decrease toxicity of perovskite solar cells (PSCs) has been demonstrated. Results show a higher intensity of the characteristic peaks of perovskite when using Sn. Replacement of Pb by Sn has a significant impact on the crystallization process of    www.advancedsciencenews.com www.pss-a.com