A Simulation Study of n-ZnO/Perovskite/p-Cu2SnS3 Based Self-Powered Photodetector: Design and optimisation

ABSTRACT  Perovskite has been widely used in the field of optoelectronics owing to its superb optoelectronic properties. The simplest fabrication along with low-temperature handling of perovskite-based planner structures (p-i-n or n-i-p) has drawn their potential impact in organic–inorganic lead halide optoelectronic devices. The poor conductivity, stability, and high processing cost of absorber materials hinder the optical performance of the device. The sandwiching of the perovskite as absorber materials between two inorganic materials working as electron transport layer (ETL) and hole transport layer (HTL) may be a possible solution. An active absorber perovskite (CH3NH3PbI3) assisted by two n-ZnO and p-CTS materials working as an ETL and HTL, respectively, has been used to maximize the overall performance of the proposed device. The resultant optimized device shows the maximum responsivity of ∼0.35 (A/W) for a broad visible spectrum (300–900 nm), while the EQE (%) reported in the range of 25–80%.


Introduction
Photodetectors are devices that convert photon energy into electrical energy, and manifest as a photocurrent.In the current scenario, photodetectors offer a wide range of applications including medical diagnosis, aviation, target recognition, and missile warning [1,2,3].The detection of photon energy via photodetector structures is divided into three phases, which are classified as absorption of light, generation of photocarriers, and finally the transportation of generated charge carriers not taking part in the combination process [4].Recently, self-powered photodetector seeks a strong interest in optoelectronic industries due to their advantages in device miniaturization and low power consumption.The detection of photon energy in self-powered photodetectors results from the existence of an internal built-in electric field [3,5,6].The process which is used to describe the mechanism of a self-powered photodetector/solar cell is called photovoltaic [7,8].
Perovskite-based photodetectors have drawn a potential impact on optoelectronics industries, due to their excellent electrical and optical properties.Perovskite materials are widely used as active absorber materials in photodetector structures and offer economic fabrication feasibility.In contrast to the initial perovskite-based research in microstructures, researchers have widely focused on planner structures, i.e. p-i-n or n-i-p, which offers structural simplicity and lowtemperature processing [9][10][11].
Apart from the selection of potential absorber materials, the role of electrons as well as hole transport materials also play an impact role while designing efficient photodetection structures.In contrast to the selection of electron transport layer (ETL) materials, zinc oxide (ZnO) has played its potential impact due to its high chemical stability and low toxicity over other materials like TiO 2 and SnO 2 and offers high binding energy over other traditional materials like GaN [12,13].The high exciton binding energy of ZnO makes this a suitable substitution for traditional materials at room temperature.Theoretically, ZnO offers strong radiation hardness, low cost, large band gap (∼3.37 eV), and mobility up to 155 cm 2 /V-s [14].ZnO is widely used as an n-transport layer or ETL along with a strong contender in the list of UV detection materials [15,16].
Copper tin sulphide (Cu 2 SnS 3 ) with moderate band gap material having band gap tunability in the range of visible to NIR (0.93-1.5 eV) [17] has also sought the attention of researchers in optoelectronic industries as hole transport layers (HTLs) [18,19].Recently, CTS materials have proved their worth as a potential absorber (absorption coefficient: 10 4 cm −1 ) [17] for optoelectronic applications with a maximum achieved efficiency of 33% [20].The strong efficiency of CTS motivated the researchers to explore this material and its hybrid structure to further optimise the performance [19,21,22].Recently, p-CTS have been explored as a hole transport layer due to its low toxicity and earth-abundant nature over other similar kinds of materials like PbS and HgTe along with good hole mobility [19,22,23].
In this work, a self-powered p-CTS/perovskite/n-ZnO and a p-i-n planner structure have been investigated, designed, and analysed by SCAPS-1D simulation software.A perovskite material having excellent light absorption and good carrier mobility has been used to realise a highly efficient photodetector structure [11].The simple film preparation process and impressive physical properties of perovskite also lead to opportunities to realise low-cost, industry-scalable photodetectors [11].The poor stability and toxicity of the absorber material have been improved by sandwiching it between two highly stable, non-toxic, and low-cost solutionprocessed p-CTS and n-ZnO materials over other toxic materials like PbS, HgTe, and TiO 2 [14,17,19,22,23].In the results, this paper reports a simulation and optimisation study of a highly stable, low-cost processed, non-toxic, and highly efficient p-i-n perovskite photodetector for visible range applications.
The optical performance of the proposed structures has been investigated by varying the thickness of absorber materials as well as the thickness of ETL and HTL.All the optical parameters of the device have been carried out under an inbuilt standard light spectrum AM 1.5 G (1000 W/m 2 ).The resultant proposed structure shows the maximum responsivity of 0.35 A/W for an optimised absorber thickness.The optical performance of the device has also been studied under different conditions like changing the band gap of absorber materials and changing trapped charge density in absorber materials.The photodetector based on perovskite material shows a broad response and cover almost visible and partial UV region.The simulated results provide a promising research direction to further broaden the perovskite-based optoelectronics device.

Device structure and modeling
The proposed self-powered photodetection structure incorporated in this manuscript has been modelled and simulated by SCAPS-1D (3.3.09).The above software is developed by the Department of Electronics and Information Systems of Gent University (Ghent, Belgium).SCAPS-1D software is widely used for simulating the thin film solar cell and photodetector structure and for studying their optoelectronic property under different optical as well as physical parameters [6].Recently, many researchers have explored photodetector structures simulated in SCAPS-1D environments [6,24,25].The obtained simulated results have shown good agreement with experimental results.Kaifi et al. [25] have explored a perovskite-based self-powered photodetector with optical responsivity in the range of 0-0.35 A/W for a 300-nm absorber thickness, which holds a good agreement with the experimental results obtained from perovskite-based structures [26,27].Similarly, some ZnO, as well as CTS-based structures, has also been explored by Benami et al. [28] and Islam et al. [21] in the SCAPS environment, respectively, and claim the high accuracy in the simulated results in comparison to experimental findings.
Fundamentally, the whole simulation setup is based on basically three sets of equations; they are Poisson's equation, hole continuity, and electron continuity under the constraint of boundary conditions.All three equations are given below [6,[29][30][31].
The schematic of the proposed perovskite-based selfpowered photodetector structure is shown in Figure 1(a).The band bending of the proposed device structure under thermal equilibrium along with transportation of effective charge carriers, i.e. electron and hole, is shown in Figure 1(b).The band bending of the proposed hybrid structure with band gap energy and thickness of proposed materials is shown in Figure S1.All the physical parameter sets for the simulation study of perovskite-based photodetector structure are listed in Table 1.Table 2 shows the defect parameters at the interface of two materials.In the proposed device structure, a perovskite material has been used as an active absorber of photon energy, while ZnO and CTS are used as an ETL and hole transport layer.Both the electron and hole transport layers effectively involve in the structure to carry the generated electron-hole pair and result in this structure being an efficient photodetector.The fluorine-doped tin oxide (FTO) layer is employed as a transparent conductive oxide layer [6].The proposed device is illuminated from the FTO side to record all the optical characteristics of the proposed device.

Results and discussion
This section is dedicated to the results and discussion of the proposed structure under different parameter variations.All the possible variations in the device parameters have been performed and discussed in this section

Device performance
The important figure of merits (FOMs) of any photodetector structures used to define their performance are current density vs. voltage or J-V curve under dark and illumination, responsivity, and external quantum efficiency (EQE).The proposed device has been illuminated with a 1.5 G 1 sun inbuilt light spectrum with an optical power density of 1000 W/m 2 used to illuminate the device from the FTO/ glass side.The responsivity and the EQE (%) of the proposed device structure have been measured from the following two equations for an obtained optical current density (J ph ) and incident optical power density (P opt ).Here, λ denotes the wavelength of incident photon energy [19,23,39,40].

Influence of absorber (perovskite) thickness
In the photodetector structure, the thickness of absorber materials plays an important role to determine the performance of the device.Figure 2(a) shows the J-V response of the proposed device structure for a variable absorber thickness, while the other parameters are kept constant as listed in Table 1 and Table 2.In the proposed device, the parameters of the used materials (Perovskite, ZnO, and CTS) like the thickness, trap charge densities, and band gap have been varied within permitted values to explore the optimised values of these parameters.In this study, the J-V response of the device has been recorded by varying the thickness of absorber material in the range of 100-1000 nm.The J-V response of the device shows the increment in optical current density with the increment in the absorber material thickness and reaches the maximum at 300 nm for zero volts applied bias.The sufficient current in the proposed structure for a zero potential confirms the self-powered operation of device.Further increasing the thickness of perovskite materials shows no contribution in the enhancement of optical current density.The optical current density of the device starts decreasing as the absorber materials thickness is further enhanced from 400 to 1000 nm.The J-V response of the proposed perovskite structure holds a good agreement with the existing literature [41,42].As per the band gap modelling (see Figure 1(b)) of the proposed device), it has been observed that at the interface of perovskite/ZnO, the conduction band energy of perovskite material is higher than the conduction band energy of ZnO materials.The difference in the conduction band energy also called conduction band offset would promote the photogenerated electrons towards the FTO electrode.Similarly, the valence band offset of perovskite/CTS will promote the photogenerated holes in the absorber material to the Ag electrode.As seen in the band gap modelling of the proposed device structure, the CTS materials will work as a hole transport layer, while the electron remains trapped within the CTS due to the high barrier potential for the electron generated in CTS materials.Despite the sufficient absorption of NIR light by CTS material, the trapping of an electron in CTS results in a negligible optical response in NIR regions.
The external quantum efficiency of the proposed structure has also been explored by varying the thickness of perovskite materials as shown in Figure 2(b).The device shows the efficient optical efficiency for the visible spectrum, i.e. from 400 to 750 nm, under self-powered mode.The efficient quantum efficiency of the device for the visible spectrum results from the strong absorption of perovskite materials [43,44].The quantum efficiency of the proposed device holds a good agreement with some fabricated as well as simulated perovskite-based photodetector structures [26,33,41,42].The external quantum efficiency of the device reported a maximum for the optimised perovskite thickness of 300 nm.The EQE (%) of the device initially increases with increasing in perovskite thickness and reaches a maximum and then starts decreasing.The increase in the EQE (%) of the proposed structure with increasing absorber thickness results from the sufficient availability of absorber materials to capture the incident photon light [4].The low thickness of absorber materials results in transparent behaviour of light with absorber materials and reduce the number of photons generated in absorber materials.On the other hand, when the absorber thickness becomes very high, the materials' absorption coefficient starts decreasing as per the well-known relation between absorption coefficient and material thickness [4].The other reason for decreasing EQE (%) at higher absorber thickness is due to the existence of more trapped states in the absorber materials in comparison to low thickness.The trapped states in the materials also degrade the mobility of photo-charge carriers and reduce the overall EQE (%) of the proposed device at a higher thickness value of the perovskite layer.Figure 2(c) shows the responsivity of the device related to the EQE (%) of the proposed structure as per Equations ( 4) and (5).The responsivity of the device also reported a maximum of ∼0.35 A/W for 300 nm absorber thickness and start decreasing beyond this optimised thickness.The explanation of the responsivity of the proposed device will be similar to the EQE (%) as shown in Figure 2(b)

Influence of p-CTS and n-ZnO layer thickness
The optical performance of the device depends not only on absorber materials but also on the materials work as ETL and HTL, which carry and transport the photogenerated carriers.For the obtained optimised thickness of absorber materials, this study relates the variation of ETL and HTL layer thickness to the optical responses of the device.Figure 2(d) shows the EQE (%) of the device under the variation of ZnO layer thickness from 1 to 100 nm.The EQE (%) of the device measured maximum for a 10 nm thickness of the ZnO layer.The device shows a strong variation in EQE (%) for the UV (300-400 nm) region, while the EQE (%) of the device remains almost constant for the visible spectrum.The variation in EQE (%) of the device for UV region by keeping the high ZnO thickness leads to more trap states and also reduces the absorption of ZnO in the UV spectrum.The trap states as well as the low absorption coefficient for a higher value of ZnO thickness reduces the EQE (%) of the device in the UV spectrum, while the EQE (%) of the device for visible regions remains unchanged.On the other hand, when the thickness of ZnO is kept very low, the absorption of photon energy is mostly done at the surface of ZnO materials, which leads to the recombination of photogenerated carriers with dangling bonds on ZnO surfaces [4].The recombination of photocarriers reduces the external quantum efficiency of the device for the UV region.
The performance of the device was further investigated by varying the thickness of CTS material work as an HTL for photogenerated holes under optical illumination.In Figure 2(e), the EQE (%) of the device is measured by varying the CTS layer thickness from 1 to 100 nm.The device shows no dependency of EQE (%) on the CTS thickness.No change in the responsivity results from no active participation of CTS materials as a photon absorber.The two common regions for constant EQE (%) are the following: CTS offers here very low absorption in comparison to the other two materials, i.e.ZnO and perovskite.However, the other regions are the band bending CTS and perovskite at the interface under thermal equilibrium.The sufficient barriers for the electron moving from CTS to perovskite materials trap them within the material and restrict their participation in the overall photocurrent of the device.Therefore, the role of CTS is to transport the holes optically generated in ZnO and absorber materials.
Figure 2(f) shows the optimised current density-voltage variation for all the optimised thicknesses of the absorber, ETL, and HTL as discussed above.The J-V response of the device was measured under dark as well as illumination conditions with an optical density of 1000 W/m 2 .The photon-to-dark current density ratio known as the sensitivity of the device was found to be much larger in comparison to the previously reported photodetector structures [33,41].The used proposed combination in SCAPS software enhanced the overall response of the device.The inset image in Figure 2(f) shows the optimised EQE (%) of the device for a broad spectrum ranging from UV to NIR.The device shows the maximum response for the visible band, which results from the maximum absorption of perovskite materials for these regions.The maximum EQE (%) of the device is reported to be 80% for 550 nm illumination, which also confirms the band gap of perovskite materials corresponding to this wavelength.

Influence of band gap of p-CTS (HTL) and n-ZnO (ETL)
In this subsection, the band bending of the proposed structure has been optimised by varying the band gap of CTS and ZnO materials for a fixed band gap of the perovskite material.The band bending at the interface of CTS and perovskite materials is also shown in Figure S2.The EQE (%) of the device for a variable band gap of CTS materials has been studied and shown in Figure 3(a).The band gap of CTS materials varied in the permissible limit as reported by various researchers in their reports.The EQE (%) of the device increases as the band gap of CTS materials increases and reaches a maximum for the band gap value near 1.5 eV.Further increasing the band gap of CTS material leads to the degradation in the EQE (%) of the device.The increasing EQE (%) for the range of the CTS band gap from 0.1 to 1.5 eV could be explained by the band bending between CTS and perovskite materials.As the band gap of CTS materials increases, the valence band offset between perovskite and CTS decreases and reaches a minimum of 1.35 eV band gap value of CTS.On other hand, the conduction band offset between CTS and perovskite is a decreasing function with the increasing value of CTS materials band gap up to 2 eV.When the band gap of CTS is 0.1 eV, the large valence band offset value allows more photogenerated holes to move towards CTS from perovskite and contribute to the photocurrent.At the same time, large conduction band offset between CTS and perovskite results in barriers for electrons moving from CTS to perovskite and starting trapping within CTS materials.The trapped electron in CTS disturbs the charge neutrality which leads to more photogenerated hole (originated in perovskite) recombination in CTS and degrades the EQE (%) of the device for the low band gap of CTS.In another way, as the CTS band gap reaches 1.35 eV, the EQE (%) of the device is maximised due to very small valence and conduction band offset for holes and electrons, which minimise the trapping of holes in the CTS materials and allow some electron transfer from CTS to perovskite materials due to tunnelling or thermal emission process.As the band gap of CTS reaches 2 eV, the holes from perovskite materials will observe strong barriers and be unable to cross the barriers and leading to the degradation in the EQE (%) of the proposed device.
The variation in the band gap of ZnO within the allowed range has also been performed to measure the EQE (%) of the device.Figure 3(b) shows the EQE (%) of the device for the variable band gap of ZnO.The variable band gap of ZnO with the fixed band gap value of perovskite materials is shown in Figure S3.As the band gap of ZnO materials varies from 3.1 to 3.6 eV, the conduction and valence band offset between perovskite and ZnO always remains in favourable condition for the transportation of electrons from perovskite to ZnO and for the holes from ZnO to perovskite.The unaffected movements of photogenerated carriers will not affect the EQE (%) of the device by varying the bandgap of ZnO material.

Influence of the concentration of defect density
This subsection is dedicated to the study of trap/defect state density on the EQE (%) of the device.Figure 3(c) shows the variation of defect density states in absorber materials from the range of 10 13 cm −3 to 10 17 cm −3 .It is seen that the quantum efficiency of the device is decreasing with increasing the defect density state in the perovskite material.The decreasing quantum efficiency of the device with increasing defect density results from the creation of large trap states in the absorber materials.As the trap density states increase in the materials, the probability of trapping photogenerated carriers in the absorber material will start increasing.The trapping of photogenerated carriers in the absorber material prevents these carriers from contributing to photocurrent and decreases the quantum efficiency of the proposed device.

Absorption coefficient of the proposed materials
This subsection shows the absorption coefficient of the proposed materials used in this study.The absorption coefficient of perovskite (CH 3 NH 3 PbI 3 ), ZnO, CTS, and FTO as a function of wavelength is shown in Figure 3(d).All the absorption parameters have been included from the previously reported absorption coefficient of the materials used to simulate photodetector structures [43,[45][46][47].

Results comparison analysis
This section is dedicated to the summary of all the findings/results of the proposed device structure.The optical characteristics of the device have been compared with some perovskite-based photodetector structures.The performance comparison of the device with some existing experimental as well as simulated results is shown in Table 3.

Conclusions
This paper reports a self-powered photodetector structure based on perovskite material.A planner p-i-n structure has been designed and simulated with an intrinsic (i)absorber material sandwiched between an n-ETL and p-HTL materials.The optical performance of the device has been optimised with the thickness, bandgap, and trap density of absorber, ZnO, and CTS materials.The optical performance of the device has been simulated in terms of the optical responsivity, external quantum efficiency, and J-V response under a light spectrum AM1.5 G 1 sun with an optical power density of 1,000 W/m 2 .The proposed device shows the maximum visible band (400-750) responsivity of ∼0.35 A/W while, the EQE (%) of the device for the same visible band is reported up to 80%.The obtained results of the proposed structure matched well with the previously reported experimental and simulated data of perovskite material-based photodetectors with an almost similar optical response for a broad visible spectrum.The studied mechanism of light-induced trap formation along with the thickness-dependent optical characteristics may be crucial in improving the performance and stability of perovskite-based photodetector structures.The proposed device structure with new nontoxic (CTS), highly stable, and good carrier mobility (ZnO) material may lead to a potential impact on perovskite-based optoelectronics devices.The proposed perovskite-based device structure may be used in several applications such as in photochromic, electrochromic, image storage, switching, filtering, and surface acoustic wave signal processing.[42] Experimental 300-800 0.39 0-81 MoO 3 /BHJ/Perovskite/TiO 2 [48] Experimental 300-800 0.40 0-82 FTO/CH 3 NH 3 PbI 3−x Cl x /Au [25] Simulated 450-1,000 0.30-0.65 65-100 (SnO 2 )/Cs 2 AgBiBr 6 /P 3 HT/Au [49] Simulated 350-600 NA 0-50

Figure 1 .
Figure 1.(a) A schematic of the proposed n-ZnO/i-CH 3 NH 3 PbI 3 /p-CTS-based photodetection structure.(b) The energy band diagram of the proposed structure under thermal equilibrium along with photocarrier transportation mechanism.

Figure 2 .
Figure 2. (a) J-V characteristic curves for varying the absorber material (perovskite) layer thickness.(b) Quantum efficiency variation with respect to the thickness of perovskite.(c) The responsivity of the device in (A/W) as a function of perovskite layer thickness.(d) Quantum efficiency of the device as a function of the layer thickness of ZnO.(e) Quantum efficiency of the proposed device for variable CTS layer thickness.(f) J-V response of the device under dark and illumination conditions (@1000 W/m 2 ) for an optimised thickness of perovskite, ZnO, and CTS materials.The Inset image shows the optimised efficiency of the proposed device.

Figure 3 .
Figure 3. (a) The quantum efficiency of the device for variable band gap of CTS material.(b) The quantum efficiency of the device for variable band gap of ZnO material.(c) The quantum efficiency of the proposed device as a function of trap density of absorber, i.e. perovskite.(d) The absorption spectra of the proposed device materials, i.e. perovskite, ZnO, CTS, and FTO.

Table 1 .
Parameters set for the simulation of perovskite-based photodetector structure in SCAPS-1D.

Table 2 .
Defect parameters at the ZnO/CH 3 NH 3 PbI 3 and CH 3 NH 3 PbI 3 /CTS interface used in this simulation.