Numerical development of lead-free Cs2TiI6-based perovskite solar cell via SCAPS-1D

Because of the toxicity and stability concerns, commercialization of lead-based perovskite solar cells (PSCs) is limited. Solar cells made entirely of Ti-based all-inorganic perovskite could be a viable answer to these issues. This paper is a theoretical paper on a perovskite solar cell (PSC) based on Cs2TiI6 using all-inorganic charge transport materials. We proposed a high performance perovskite solar cell (PSC) according to variables such as charge transport materials and its optimal thicknesses, absorber thickness, absorber defect density and interface defect density and working temperature. The optimal absorber thickness, Hole transport layer (HTL) thickness, and Electron transport layer (ETL) thickness are 500 nm, 50 nm, and 10 nm, respectively. After analyzing the other factors, we ended up with a high-performance PSC with a power conversion efficiency of 22.5% at room temperature and 22.84% at 270 K. These results are useful for the conception and manufacture of PSCs.


Introduction
Organic-inorganic perovskite materials have been extensively explored in recent years as an alternative to silicon-based solar cells for improving device efficiency, and it are a potential alternative to silicon-based solar cells because of its lower manufacturing costs [1,2]. Since the 2012 breakthroughs [3,4], it is now generally agreed that the halide perovskite solar cells could have a significant practical influence in the future generation solar cells. The active material of a typical perovskite solar cell was an organic-inorganic halide material [5,6]. Jena et al. developed the perovskite solar cell (PSC) by employing an organic-inorganic lead halide as an absorber instead of an organic dye [7]. From 2009 to 2019, the PCE (power conversion efficiency) of a typical organic-inorganic hybrid PSC (perovskite solar cell) has risen from 3.8% to over 25%. Despite the excellent optoelectronic properties of organic-inorganic hybrid PSCs, they display three major problems. For starters, it reduce the shelf life of perovskite compounds. Second, the presence of organic cations creates instability, and third, they use a lead (Pb)-based absorber, which is environmentally damaging [8]. As a result, all inorganic lead-free halide-based PSCs are now introduced [9]. As with organic-inorganic hybrid absorber, organic charge transport become unstable in oxygen, sunlight, and humidity [10,11]. Chen and his colleagues developed Cs2TiBr6 thin film Perovskites with a 3.28 % solar efficiency [12].
Because of its potential light harvesting properties, acceptable bandgap, good optical absorption, high stability, and benign defect property, Cs2TiI6 absorber material is shown to be a suitable alternate. The Cs2TiI6 absorber can be more resistant to environmental stress and have better tolerance [12],13]. Ahmed et al. reported an FTO/SnO2/Cs2TiBr6/MoO3/Au structure, with a stable PCE up to 11.49% [14]. Ahmad and his colleagues suggested an unique perovskite solar cell configuration of Au/CdTe/Cs2TiI6/TiO2/ITO, which has a PCE of 15.06 % [15].
In this study, we propose an all-inorganic PSC with a new combination of FTO/ZnO/Cs2TiI6/MoO3/Au using (SCAPS 1D-3.3.10). We analyzed the effects of various inorganic charge transport materials to find the optimum material for Cs2TiI6-based PSC, e.g., p-MoO3, p-CuI, p-NiO, n-TiO2, n-SnO2, n-ZnO, n-CdS as HTL (Hole Transport Layer) and ETL (Electron Transport Layer) and optimized its thicknesses for achieving optimum performance. We also studied the effect of defect density of Cs2TiI6, defect density of the interface, and the effect of operating temperature on device performance. Upon comparing the performance parameters, e.g., short circuit Jsc (current density), Voc (open-circuit voltage), PCE (power conversion efficiency), and FF (Fill Factor), we have obtained an optimized high-performance PSC with a novel combination of FTO/ZnO/Cs2TiI6/MoO3/Au, that may aid in designing eco-friendly Ti-based perovskite solar cell for future technologies.

Theoretical simulation
FTO/ ETL/perovskite/HTL/Au is a unidimensional ni-p planar heterojunction composition in our modeled PSC structure (Fig. 1). In this structure, we used inorganic lead-free Cs2TiI6 as PAL (perovskite active layer), inorganic n-TiO2, n-SnO2, n-ZnO, and n-CdS as ETL (electron transport layer), and inorganic p-MoO3, p-CuI, and p-NiO as HTL (hole transport layer). For validation of our simulation, we first simulated the FTO/TiO2/Cs2TiBr6 (200 nm)/P3HT/Au structure using the SCAPS 1D simulation software and then compared it to the experimental results [12]. HTL and ETL thicknesses of 30 nm are determined by calibrating simulation data with experimental data.  FTO and Au have work functions of 4.4 eV and 5.1 eV, respectively. We used the following settings for all defect layers: At 300K, the thermal velocity of electrons and holes is 10 7 cm/s, the energy distribution is Gaussian, the characteristic energy is 0.1 eV, and the energy level relative to the reference is 0.6 eV [18]. We provided absorber/ETL and absorber/HTL interface defects with hole capture cross section of 10 -18 cm 2 and 10 -19 cm 2 , electron-capture cross section of 10 -19 cm 2 and 10 -18 cm 2 respectively, and interface defect density of 10 13 cm -3 . We took a constant illumination of 1000 W/m 2 at AM 1.5G, a continuous temperature of 300 K, a series resistance of 1 Ω-cm 2 , and a shunt resistance of 4200 Ωcm 2 to perform the simulation.   Table 2 shows the performance comparison between the simulated and experimental structure. The simulation results are clearly in agreement with the relevant experimental results. This also means that the simulation parameters are nearly identical to those of a real device.

Results and discussion
In this paper, analyzing several HTMs (Hole Transport Materials) and ETMs (Electron Transport Materials) by optimizing its thicknesses and, also by optimizing the density of PAL defects and interface defects, we propose a new structure FTO/ZnO/Cs2TiI6/ MoO3/Au.

HTL optimization
Instead of using organic P3HT as an HTM, we now select the most suited HTM and calculate its optimal thickness using inorganic CuI, MoO3, and NiO. Table 3 shows the input values for several HTMs.    The Valance Band Offset (VBO) for all HTMs is negative, as seen in Fig. 4. When VBO is negative, there is no barrier to photo-generated holes flowing toward the back electrode, hence Jsc is nearly constant. Jsc is found to be 11.52 mA/cm 2 for NiO, 11.53 mA/ cm 2 for CuI and MoO3. However, the increase in negative VBO causes an increase in interface recombination, which reduces Voc [23]. We find Voc is 0.97 V, 0.69 V, and 0.75 V for MoO3, CuI, and NiO respectively. For MoO3, CuI, and NiO, the Voc is 0.97 V, 0.69 V, and 0.75 V, respectively. When it is increased, negative VBO decreases FF from 83.88 % to 81.92 %. MoO3 has the highest PCE = 9.4 % because it has the lowest negative VBO, i.e. best band alignment with Cs2TiI6. PCE declines to 7.08 % when negative VBO increases in NiO. CuI, on the other hand, has the largest negative VBO, indicating poor band alignment, and lowest PCE of 6.42 % [23].

Fig. 5. PCE for several HTLs
The structure of MoO3 has the best PCE, as shown in Fig. 5, so we picked MoO3 to optimize the thickness. We changed the HTL thickness from 20 nm to 50 nm to optimize the HTL thickness taking in mind the regular feasibility of the perovskite solar cell construction process [24], and its effect is shown in Fig. 6. It can be found that with the increase in the HTL thickness, the values of Jsc decreases gradually. Voc increases when the thickness is smaller than 35 nm and then it remained almost unchanged. The FF and PCE of the device first increase when the HTL thickness is less than 50 nm and then decrease slightly with further increasing the HTL thickness. This is because when the HTL is too thin, it may lead to low shunt resistance and current leakage, thus resulting in a low FF [25]. However, when the HTL is too thick, the series resistance will increase, leading to a reduction in FF. The structure using MoO3 as HTM provides the best performance at 50 nm having Voc  0.97 V, Jsc = 11.54 mA/cm 2 , FF = 84.06%, and PCE = 9.46 %. Therefore, we select MoO3 as the optimum HTM for Cs2TiI6 based PSC with an optimized thickness of 50 nm.

Absorber thickness optimization
We have optimized the absorber thickness by using optimized MoO3 as HTL. We have varied the absorber thickness from 300-520 nm using inorganic TiO2, SnO2, ZnO, and CdS as ETMs underneath the perovskite. Tables 1 and 3 provide the input parameters of the ETMs. Fig. 7 displays the optimization of the absorber thickness for different ETMs. It shows that the use of TiO2 and ZnO provides both the best results than other ETMs, with a PCE of 12.01%, but using ZnO gives a better Jsc than using TiO2. The other parameters; Voc and FF are almost the same for all ETMs. So we chose ZnO as ETL, with optimal absorber thickness of 500 nm. The reason for this better PCE of ZnO can be elicited from Fig. 8.   Fig. 7. PCE curve against thickness variation of the Cs2TiI6 absorber for different ETMs.  ZnO and, 16.301709 mA/cm 2 for SnO2 due to the variation in its quantum efficiency (QE) which is shown in Fig. 9. As Jsc varies, the PCE of structure using different ETMs also varies. Fig. 9. Quantum efficiency curves for various ETMs at optimal absorber thickness.

ICEGC'2021
From Fig. 9, we find the CdS, TiO2, SnO2, and ZnO exhibit as much as 98.75 %, 98.72 %, 98.69%, and 98.72 % QE respectively in the visible range. Therefore, ZnO possesses Jsc of 16.402363 mA/cm 2 and thus the best PCE. Therefore, we select the optimized Perovskite Active Layer (PAL) thickness to be 500 nm for ZnO as ETL.

ETL thickness optimization
We optimized the ZnO thickness by varying it from 10 nm to 55 nm as ETL using the optimal HTL and PAL thicknesses. As shown in Fig. 10, there was no significant improvement in the output parameters when the ETL thickness was varied. As a result, we choose the optimized thickness of 10 nm with the highest PCE 12.02 %, Voc 0.9805 V, Jsc 16.402430 mA/cm 2 , and FF 74.71%.

Absorber defect density optimization
We have studied the impacts of defect density by varying it from 10 11 cm -3 to 10 16 cm -3 using selected ETM and HTM, as well as optimal HTL, PAL, and ETL thicknesses. The J-V curves at various defect densities are seen in Fig. 11. The Jsc of PSC degrades from 16.50 mA/cm 2 for 10 12 cm -3 to 16.26 mA/cm 2 for 10 16 cm -3 . The increase in the recombination rate as defect density increases is the source of this degradation [26]. When the defect density is less than 10 12 cm -3 , however, it remains almost constant. As a result, we chose a defect density of 10 12 cm -3 , with a Voc of 0.9886 V, and Jsc of 16.50 mA/cm 2 . Fig. 11. J-V curve comparison for various absorber defect densities 3.5 Optimization of the defect density at the interface J-V characteristics for different defect densities at the interfaces for values between 10 4 cm -3 and 10 9 cm -3 are shown in Fig. 12. Voc saturates at defect densities below 10 5 cm -3 , as shown in this diagram. Voc, on the other hand, decreases from 1.63V to 1.4V above 10 5 cm -3 to 10 9 cm -3 . As a result, at 10 5 cm -3 , we choose the optimum interface defect with Voc of 1.63 V, Jsc of 16.50 mA/cm 2 , FF of 83.42 % and PCE of 22.50 %.

Optimized device
We achieved a high performance PSC with a new combination of FTO/ZnO/CsTiI6/MoO3/Au after optimizing HTL, PAL, and ETL. The new structure's schematic is shown in Figure 13.  Table 6 compares the performance of different lead-free PSC studies. It can be seen that our device produces comparatively promising results, it's eco-friendly because it is made entirely of inorganic Tibased perovskite materials, and that it is expected to be highly stable in the environment due to the use of all inorganic charge transport layers.

Working Temperature effect
To understand the effect of temperature on our new device, and the suitable temperature to get the best PCE, we changed the working temperature from 270 K to 360 K. Fig. 16 shows that as the working temperature rises from 270K to 360 K, the Voc decreases from 1.6585 V to 1.5790 V. This is due to the fact that when the temperature rises, the saturation current and recombination rate increase. Additionally, due of the thermal generation of carriers and the reduction of the band gap, Jsc increases from 16.503854 mA/cm 2 to 16.504397 mA/ cm 2 [33]. The FF decreases from 83.43 % to 81.63 % as the rate of increase in Jsc is less than the rate of decrease in Voc, and the PCE reduces from 22.84 % to 21.27 %. The overall performance of the PSC degrades when PCE decreases [34]. So the device gives a better performance for 270K, it can be used especially in cold regions.