Performance evaluation of WS2 as buffer and Sb2S3 as hole transport layer in CZTS solar cell by numerical simulation

This study reports on performance enhancement of a Cu2ZnSnS4 solar cell introducing Sb2S3 as hole transport layer (HTL) along WS2 as buffer layer. We have investigated photovoltaic (PV) characteristics by utilizing SCAPS‐1D. A comparative analysis on PV performances between conventional CZTS/CdS and proposed Ni/Sb2S3/CZTS/WS2/FTO/Al solar cells is presented. It is revealed that “spike like” band structure at the CZTS/WS2 interface having smaller conduction band offset makes it potential alternative to commonly used CdS buffer. This report also evaluates that the Sb2S3 as an HTL inserted at the rear of CZTS enhances performances by reducing carrier recombination at back interface with appropriate band alignment. The impacts of thickness, carrier concentration of different layers, and bulk defect density in CZTS as well as the interface defects on cell outputs are analyzed. The influences of temperature, work function, and cell resistances are also examined. Optimum absorber thickness of 1.0 μm along doping density of 1017 cm−3 is selected. A maximum efficiency of 30.63% is achieved for the optimized CZTS cell. Therefore, these results suggest that Sb2S3 as HTL and WS2 as buffer layer can be employed effectively to develop highly efficient and low‐cost CZTS solar cells.

silicon (Si)-based solar cell had started its journey in 1954 with conversion efficiency about 6%. 6 The best Si-based PV device with conversion efficiency greater than 24% has been developed in recent years. [7][8][9] As the first generation Si solar cells are expensive due to high price of Si wafer and costly processing technology, the second generation thin-film solar cell (TFSC) has been introduced. 10,11 Among many other thin-film semiconducting materials like CdTe, CIGS and so forth have become most popular owing to their high absorption ability, proper band gap, and low-cost fabrication. 7,[12][13][14][15][16] The most recent TFSCs based on the CdTe and CIGS absorbers declare maximum efficiencies of 22.1% and 23.35%, respectively. 7 Nonetheless, the main limitations along the journey of CdTe and CIGS absorbers are that the materials such as Cd and Te are toxic, environmentally hazardous, and scarce elements as well as the In and Ga in CIGS are rare and expensive. [17][18][19] In this context, a chalcogenide material CZTS has drawn scientists attention as a promising alternative absorber in the TFSC technology. [20][21][22][23] Earth abundant raw material and nontoxic nature of the CZTS enable it as a precise absorber candidate for the PV technology to replace other costly and toxic absorbers. The CZTS absorber exhibits several advantages such as direct energy band gap in the range from 1.4 to 1.6 eV, large light absorption coefficient (>10 4 cm −1 ), economical fabrication process, and high stability, which facilitate it as a very reliable candidate of the TFSC for the commercial production. [24][25][26][27][28][29][30] An efficiency of 6.7% has been obtained for the experimentally fabricated CZTS-based solar cell. 27 Another research work reports the conversion efficiency of 8.4% for the TFSC employing an earth-abundant CZTS absorber. 31 A practical PCE of 9.2% for the CZTS solar cell with Zn 1-x Cd x S buffer layer is recorded by Sun et al. 32 A certified efficiency of the heterojunction CZTS device architecture including Mo/MoS 2 /CZTS/CdS/ZnO/ITO/MgF 2 has been achieved to be ∼11% with a high V oc 0.73 V by Yan et al. 33 However, the insufficient enhancement in the PCE compared to the existing TFSCs is still open to conduct further research on the CZTS solar cells.
There are numerous endeavors have also been taken numerically to develop highly efficient CZTS-based TFSCs by remodeling the heterojunction structures. [34][35][36][37][38] The previous numerical approach reports 13.41% conversion efficiency for SLG/Mo/CZTS/CdS/ZnO:Al heterojunction TFSC. 36 It is noticed in the previous studies that the CdS buffer has been extensively employed in the conventional CZTS PV devices to attain the stable performances. However, the toxic nature of Cd appeals for interdiction of CdS to be used as the buffer layer. 34 An optoelectronic simulation investigation on a silver-mixed CZTS solar cell with CZTS/ACZTS/CdS heterojucntion reports the PCE of 17.59% along V oc of 0.94 V. 35 An efficiency of 23.72% is calculated theoretically for the CZTS-based solar cell comprising glass/Mo/CZTS/CdS/ZnO/FTO heterojunction by Khattak et al. 37 With a heterojunction Mo/CZTSe/CZTS/CdS/ZnO/FTO thin-film PV device inserting a CZTSe BSF layer, a conversion efficiency of 22.03% is obtained numerically. 38 Besides, it is environmentally perilous and has detrimental impact of CdCl 2 treatment in fabricating the CdS films. 39,40 Therefore, it is crucial to replace the commonly used CdS buffer with an environment friendly, nontoxic and cost-effective material. Several buffer materials having less or no toxicity with the CZTS absorber have been utilized in the previous studies to improve the PV performances. [41][42][43][44][45][46] The PCE of 21.19% has been obtained for a CZTS/Zn(O,S) structure. 42 The overall efficiency of 16.24% for 1.5-μm CZTS absorber is measured for the device structure of Mo/CZTS/ZnSe/i-ZnO/Al-ZnO, where the ZnSe as a buffer layer is introduced. 44 Lately, an excellent efficacy of 23.69% with a structure of Mo/CZTS/MoS 2 /ZnO/front contact has been evaluated by a numerical approach, where the toxic CdS buffer is replaced by the less toxic MoS 2 material in the heterojunction. 45 There are still possibilities intended to further development of the outputs of CZTS PV cells according to the Shockly-Queisser limit. 47 In the previous works, remarkable improvement in the PV performances of TFSCs is realized by swapping the conventional toxic CdS buffer layer with nontoxic buffer materials. 48,49 In the present research, we have employed the tungsten disulfide (WS 2 ) as a buffer layer substitute to the traditional CdS buffer with the CZTS absorber. Due to excellent optoelectronic and electro-chemical properties, the WS 2 as a buffer layer alternative to CdS buffer has previously been utilized in the CIGS and CdTe TFSCs. 40,50,51 It possesses better conductivity, higher carrier mobility, and favorable energy band gap of 2.1 eV that aligns properly with the CZTS absorber layer. 40,[50][51][52] In addition, the WS 2 forms a spike like band structure with the CZTS absorber at front side, whereas a "cliff like" band arrangement is created at CZTS/CdS interface. The spike type band alignment prevents non-radiative electron-hole recombination at the front interface, thus boost the performances of the PV devices. Also a lattice mismatch, which is an interface controlling factor in a heterojunction, is minor at the CZTS/WS 2 interface than the CZTS/CdS interface. Therefore, a substantial enhancement of the PV outputs of CZTS-based solar structure using WS 2 buffer layer is expected.
Nevertheless, a considerable back surface recombination loss is another challenge in order to achieve the best PV efficiency. It has been inspected that incorporating an hole transport layer (HTL) amid the CZTS layer and the back metal contact can minimize surface recombination loss at the interface, thus enhance the conversion efficiency remarkably. [53][54][55][56] In this regard, we suggest to use the p + Sb 2 S 3 semiconducting material as an HTL with the CZTS absorber to minimize the carrier recombination loss at the interface. The Sb 2 S 3 material has important optoelectronic characteristics like suitable band gap (∼1.7 eV), superior absorption coefficient (∼10 5 cm −1 ), earth-abundant in constituents, less toxicity, and low-cost fabrication process. [57][58][59][60][61][62] The suitable band configuration between the Sb 2 S 3 and CZTS due to low VBO and sufficiently large CBO at the Sb 2 S 3 /CZTS interface ensures smoother hole transportation and opposition of electron from the absorber, respectively. Furthermore, the lattice mismatch between the Sb 2 S 3 HTL and CZTS absorber is observed to be sufficiently low which may reduce the interface defects and minority carrier recombination. Consequently, the overall PV performances may enhance by introducing the proposed Sb 2 S 3 at the back of CZTS absorber.
In the present work, the proposed novel PV device of Ni/Sb 2 S 3 /CZTS/WS 2 /FTO/Al has been modeled and simulated by using the SCAPS-1D software. The outputs of CZTS-based solar cell are explored comprehensively by investigating the influences of thickness, carrier concentration, defects, operating temperatures, work function of back metallic contact, and cell resistances.

DEVICE CONFIGURATION AND NUMERICAL SIMULATION
The CZTS PV devices have been designed and simulated by using the SCAPS-1D simulator. 63,64 The salient features of the SCAPS-1D software program include the measurements of current density-voltage (J-V) curves, spectral response, capacitance-frequency (C-f), and capacitance-voltage (C-V) of the PV devices. The simulator operates by solving fundamental semiconductor equations such as electron and hole transport equations, continuity, and Poisson equations. Figure 1A,B shows the schematic illustrations of the anticipated solar model of Ni/Sb 2 S 3 /CZTS/WS 2 /FTO/Al and the corresponding energy band diagram, respectively. The simulated structure comprises the CZTS as an active layer with the Sb 2 S 3 as an HTL and the WS 2 as a buffer layer, respectively. In addition, the FTO as a transparent conducting oxide (TCO) layer is introduced. The metal front and back contacts of Al and Ni, respectively, have been used. The impacts of thickness, acceptor density, absorber defect, interface defects, operating temperature, back metal contact work function, and cell resistances have been studied for the CZTS PV structure by using the SCAPS-1D simulation tool. This investigation is conducted under an illumination of AM1.5G with a photon intensity of 100 mW/cm 2 . The simulation parameters for various layers along the interface defects are shown in Tables 1 and 2, respectively. The parameters are carefully chosen based on the existing theoretical and experimental measurements. 20,37,40,51,55,61,[65][66][67][68][69][70][71] At room temperature, the thermal velocity of both electron and hole is anticipated to be 10 7 cm/s. [72][73][74] The absorption coefficient of each layer is managed from the corresponding equation stated as α = A α (hν−E g ) 1/2 , herein A α is chosen to has the value of 10 5 . 75 For both front and back metal contacts, the surface recombination velocity of electron and hole is set to be 10 7 cm/s. For each layer simulation, the Gaussian energetic distribution with single acceptor/donor type of defect is selected.

Defect type Neutral Neutral
Capture cross-section of electrons (cm 2 ) 10 −19 10 −19 Capture cross-section of holes (cm 2 ) 1 0 −19 10 −19 Reference for defect energy level E t Above the highest E v Above the highest E v Energy with respect to reference (eV) 0.6 0.6 Total density (cm −2 ) 10 11a 10 11a a This is a variable filed.

Enhancement of PV performances of CZTS solar cell
The most challenging task in designing Cd-free solar cell is to select an appropriate buffer layer alternative to toxic CdS without compromising the overall PV cell performances. Figure 2A elucidates the J-V curves of the different CZTS solar cells. Here, the conventional CZTS solar cells employing CdS and WS 2 buffer layers, respectively, and the proposed heterojunction CZTS solar cell with CdS and WS 2 buffers and various HTLs are demonstrated. The PV performance parameters of the CZTS-based TFSC incorporating the Sb 2 S 3 as an HTL and the CdS as a buffer layer are extracted from the J-V characteristics demonstrated in Figure 2A. The evaluated outputs for various CZTS PV structures are also given in Table 3. As can also be observed, the cell efficiency of 21.35% with the structure of CZTS/CdS is obtained, whereas the improved conversion efficiency of the Sb 2 S 3 /CZTS/CdS heterojunction PV device is measured to be 25.82%. Therefore, the introduction of HTL into the conventional CZTS/CdS solar cell significantly increases the photo-conversion efficiency. In addition, it is seen that the CZTS solar cell with WS 2 buffer has superior J-V characteristics than the typical CZTS/CdS cell. The enhanced PCE of 23.55% is estimated for the heterojunction CZTS/WS 2 solar cell. The performance parameters of Ni/SnS/CZTS/WS 2 /FTO/Al, Ni/CuS/CZTS/WS 2 /FTO/Al, and Ni/Sb 2 S 3 /CZTS/WS 2 /FTO/Al solar cells are evaluated, as shown in Table 3. It is evident from the Table 3 that the CZTS solar cell employing the Sb 2 S 3 HTL outperformed the other two configurations with SnS and CuS HTLs, respectively. Good energy level alignment and low lattice mismatch with the CZTS absorber are critical for HTL selection because they aid in the regeneration of the absorber including the extraction and transportation of light generated carriers. Low VBO at the HTL/absorber interface ensures optimum energy level alignment. The CBO and VBO values for SnS, CuS, and Sb 2 S 3 HTLs with the CZTS absorber are also provided in Table 3. Also, the lattice mismatch between the absorber and corresponding HTL (SnS or CuS or Sb 2 S 3 ) is calculated and the outcomes are represented in Table 4. From Tables 3 and 4, it can be realized that the Sb 2 S 3 HTL with the CZTS absorber meets both the concerns of small VBO and low lattice mismatch with satisfactory conversion efficiency of 30.63% as compared to the efficiencies of 30.1785% and 29.24% found for the cell with SnS and CuS HTLs, respectively. Figure 2B illustrates estimation of spectral responses of cells configured with WS 2 and CdS buffers in terms of external quantum efficiency (EQE) as a function of incident wavelength. It is also found EQE is superior in the CZTS PV device with WS 2 buffer and Sb 2 S 3 HTL at long wavelength. These performances augmentation in the modified cell with WS 2 buffer can be justified by interface optimization at absorber/buffer through the pertinent conduction band alignment. Using the energy level diagram in Figure 1B Figure 3A,B. It is observed that the conduction band energy of the CZTS is higher than that of the CdS and this forms negative cliff-like band structure (ΔE c : −0.35 eV), as manifested in Figure 3A. The "cliff" CBO causes lesser band bending, thus permitting holes at the interface. These holes result in diminished performance by recombining with electrons at the absorber/buffer interface. The activation energy (E A ) for carrier recombination is calculated using the absorber energy band gap (E g.absorber ) and the CBO of absorber/buffer layers as E A = E g.absorber −|CB offset|. It has been discussed in the former studies that the interface carrier recombination will be occurred at E A < E g.absorber . [76][77][78][79] In Figure 3A, small E A is observed at the CZTS/CdS interface. Thus, a low hole barrier is formed for the recombination at the active layer surface as well as the cross-interface recombination between the buffer and absorber. 78,80 Figure 3A indicates the enhanced carrier recombination due to the low hole barrier shown by solid arrows, and the electron and hole recombination is indicated with dashed arrow in the figure. The electron and hole recombination will be increased greatly with the small E A and thus diminish the PV outputs of the CZTS solar cell with CdS buffer. Conversely, in Figure 3B, the energy band diagram indicates that the CZTS absorber conduction band lies underneath than the WS 2 buffer. This position of the conduction bands creates a spike-like positive CBO, ΔE c of +0.15 eV at the CZTS/WS 2 interface as seen in the Figure 3B. The high hole barrier is formed for the large E A found at the CZTS/WS 2 interface. It expresses that the CB bends upward by +0.15 eV from CZTS to WS 2 at the CZTS/WS 2 interface. This large band bending comparing to the CZTS/CdS interface ensures smoother electron transportation from the absorber to the buffer towards the front contact while preventing the holes at the interface, thus the "spike" like CBO creates a hole barrier by inducing absorber inversion, and hence the values of J sc and FF are increased. Therefore, the reduction of carrier recombination current at the interface would initiate the improvement in the solar cell output parameters. Moreover, the interface quality is improved in CZTS/WS 2 structure compared to the CZTS/CdS cell as the WS 2 film has lower lattice mismatch than the CdS buffer with CZTS absorber layer, as revealed in Table 4.
After buffer layer optimization, further enhancement in the solar cell output is predicted by the insertion of Sb 2 S 3 HTL between the CZTS absorber and the back electrode. It is detected that the low VBO and sufficiently large CBO at the Sb 2 S 3 /CZTS interface assist effortless hole conduction and electron blocking from the CZTS to the back contact, respectively. There is small energy barrier for holes resulted by band bending at the Sb 2 S 3 /CZTS interface with the low VBO means. Hence, the effective hole density in the absorber is decreased by minimizing the probability of electron-hole recombination to ensure the greater photo generation and collection. Also, as the lattice mismatch between the CZTS and Sb 2 S 3 layers is considerably low (as mentioned in Table 4), no dangling bond will be created to eventually reduce traps centers for outgoing holes. The lattice mismatch defined by = 2|a s −a e |/(a s + a e ) 55 has been evaluated using the lattice parameters of buffers, HTLs, and the CZTS absorber reported in the previous studies. 55 Figure 2A. There is restriction of long wavelength absorption in the CZTS absorber which is rooted in its band gap. By inserting the Sb 2 S 3 as HTL with different band gap, the absorption at longer wavelength is feasible. Therefore, the spectral response at longer wavelengths of the device with the heterojunction Sb 2 S 3 /CZTS/WS 2 structure is improved than the CZTS/WS 2 device, which is evident in the Figure 2B. The minority carrier reduction owing to high recombination at the Ni/CZTS interface may result in the deficient performance of the standard CZTS TFSC. A strong built-in potential induced at the p + -type HTL/p-type absorber high-low junction inhibits the minority electron recombination loss at the back surface, thereby enhancing the PV performances outstandingly. 55

Impacts of HTL thickness and doping concentration on cell performances
In this study, we have analyzed the behavior of cell outputs by changing HTL thickness and doping concentration. Figure 4A represents the change of V oc , FF, J sc , and efficiency in terms of HTL thickness. The HTL thickness is changed from 0.01 to 0.2 μm for evaluation of outputs. The corresponding PV outputs have shown nearly constant with the HTL thickness. We have also studied the PV outputs deviation by shifting the Sb 2 S 3 HTL acceptor carrier concentration from 10 12 to 10 21 cm −3 to determine the optimum value. The deviation of PV outputs in terms of the HTL doping concentration is shown in Figure 4B. J sc shows almost constant with increasing the carrier concentration of HTL. However, J sc is decreased a little while increasing the carrier concentration from 10 16 to 10 21 cm −3 . FF is hardly enhanced with increasing the carrier concentration which is almost negligible. A similar variation in both the efficiency and V oc with respect to the HTL doping concentration is observed in Figure 4B. Nonetheless, a slight rise in both V oc and efficiency is found beyond the doping density of 10 18 cm −3 . The minor enhancement of V oc and efficiency at the high carrier concentration may be referred to the decreasing of recombination at the back surface. In this work, the optimum values for the thickness and carrier concentration of HTL have been selected to be 0.1 μm and 10 19 cm −3 , respectively.

Impacts of thickness and doping concentration of WS 2 buffer on cell performances
The outputs of anticipated CZTS solar cell employing WS 2 buffer and Sb 2 S 3 HTL are explored by changing the thickness and carrier concentration of WS 2 buffer layer in this section. The simulated V oc , J sc , FF, and efficiency as a function of the WS 2 buffer layer thickness in the range from 0.01 to 0.1 μm are represented in Figure 5A. It can be depicted from the figure that all the PV parameters (except FF) are decreased slightly on varying the WS 2 buffer thickness. V oc is declined initially for the buffer layer thickness up to 0.04 μm and then is stable. A minor progress in J sc in the beginning is noticed and then it is deteriorated from 0.02 μm on further growing the buffer thickness. Initially, FF decreases and then increases gradually beyond the buffer thickness of 0.04 μm. In Figure 5A, the efficiency decreases originally and becomes almost constant from the thickness 0.04 μm on rising the thickness. These results have excellent consistency with the solar cell outputs on changing the buffer thickness reported in the previous studies. 87,88 The minor current due to poor photo-generated carriers with the thicker buffer results this PV behaviors. 87,88 A reduced extent of short wavelength photons may derive to the absorber with increasing the buffer thickness. [88][89][90] Moreover, the recombination of few photo-induced electrons and holes produced during photon absorption by the absorber with defects at buffer/absorber interface before going to front contact results in the slight diminution in efficiency at the thick buffer layer. Therefore, a thin buffer layer is inhibited to realize noteworthy PV outputs. Thus the thickness of the buffer layer is optimized to be 0.05 μm. Figure 5B shows the effect of changing buffer concentration from 10 12 to 10 21 cm −3 on the cell performances. From the figure, we can see almost similar value of J sc for donor concentration up to 10 16 cm −3 and then it is decreased slightly with increasing the doping density of WS 2 buffer. Both the FF and efficiency have declined simply from 88.29% to 88.2% and 30.67% to 30.61%, respectively, with the growing the doping concentration of WS 2 from 10 12 to 10 21 cm −3 . A minimal variation (nearly unchanged) in V oc is noticed for the doping concentration ranging from 10 15 to 10 21 cm −3 and prior 10 15 cm −3 . In this work, the carrier concentration of WS 2 buffer layer is selected to be 10 17 cm −3 as optimal value.

Impacts of absorber thickness and doping concentration on PV performances
The thickness and carrier concentration of absorber are crucial factors in designing a heterojunction TFSC. Figure 6A displays the calculated V oc , J sc , FF, and PV efficiency as a function of the absorber thickness of the suggested PV cell  91 V oc shows an inverse nature with increasing the absorber thickness. The enhancement in V oc with decreasing the CZTS thickness may be owing to the reduction of back surface carrier recombination loss in the CZTS TFSC with HTL. Comparable V oc outcomes with decreasing the absorber thickness lower than 1.0 μm are also realized in the former investigations on TFSCs. 55,68,73,85,92 The output characteristics of proposed solar cell with respect to increasing doping density from 10 12 to 10 21 cm −3 is illustrated in Figure 6B. It is initially noticed that the carrier concentration of CZTS absorber layer has limited impact on the cell efficiency. For the doping concentration ranged from 10 12 -10 15 cm −3 , PCE is increased marginally. However, the cell efficiency is enlarged remarkably for the doping concentration from 10 16 to 10 21 cm −3 . A maximum efficiency of 37.55% has been found at doping concentration of 10 21 cm −3 . FF is noticed to be unchanged up to the carrier concentration of 10 15 cm −3 and then is started to increase rapidly. The influence of changing the carrier concentration on J sc is found to be insignificant. Highest J sc of 32.05 mA/cm 2 is measured at the doping density of 10 18 cm −3 . A slight variation in V oc is perceived for the low carrier concentration. On the other hand, the doping concentration beyond 10 17 cm −3 has remarkable consequence on the V oc . With the doping density in the range from 10 17 to 10 21 cm −3 , the V oc is enhanced from 1.08 to 1.30 V. To attain the outstanding cell performances with low fabrication cost, the optimum thickness of 1.0 μm and the doping concentration of 10 17 cm −3 are adjusted in the next numerical calculations. The doping density optimized for the CZTS in this work is in well compliance with the carrier concentration obtained experimentally in the previous work. 65

F I G U R E 7
Effects of absorber defect density on efficiency of proposed CZTS solar cell with HTL

Impacts of bulk defect density of absorber on PV performances
The defect in the absorber is one of the most significant aspects to acquire outstanding PV output. The optoelectronic properties of absorber in the TFSC may be altered owing to the existence of defect states. Therefore, to realize a highly efficient CZTS solar cell, the quality of the absorber material needs to be evaluated. In this study, the impact of the absorber defects on the efficiency is analyzed and the simulated output parameter is demonstrated in Figure 7. The absorber defect density is varied from 10 12 to 10 17 cm −3 with the parameters fixed at the optimum values of different materials. The efficiency shows a regular decrement for the defect density ranged from 10 12 to 10 16 cm −3 . A sudden drop in efficiency is noticed at the defect density beyond 10 16 cm −3 . It is revealed that the PV performance drops at the very high absorber defect density. This degradation in the PV output at the high defect density is may be due to sufficient carrier recombination loss by localized energy centers in the heterojunction. Thus, the absorber material having low defect levels should be considered to achieve the best PV outputs. Herein, to optimize the device performances, the optimal defect density of the anticipated CZTS solar structure with HTL have been selected to be 10 15 cm −3 .

Impacts of interface defects on cell performances
During fabrication process owing to some structural defects, the interfacial defect is developed in the heterojunction PV device. As a result, it is essential to test the effect of interfacial defect on the solar cell characteristics. In this simulation, we have explored the characteristics of the PV performances by varying the defect densities at the HTL/absorber and absorber/buffer interfaces. Figure 8A demonstrates the outcomes of the proposed CZTS cell regarding the CZTS/WS 2 interface defect. The defect density at the CZTS/WS 2 interface is shifted from 10 10 to 10 18 cm −2 . J sc is approximately constant outcome till 10 17 cm −2 and then decreases. V oc and efficiency change in a similar manner with enhancing the interface defect density, as revealed in Figure 8A. As further provided in the figure, the FF starts to decrease swiftly when defect density exceeds 10 13 cm −2 at the CZTS/WS 2 interface. Figure 8B shows the aftermaths as a function of Sb 2 S 3 /CZTS interface defects of the suggested CZTS cell. At this interface, the defect density is also varied from 10 10 to 10 18 cm −2 . There has been a significant decrement of V oc and J sc at the defect density > 10 15 cm −2 at Sb 2 S 3 /CZTS interface. FF saturates in the range from 10 10 to 10 13 cm −2 , as shown in Figure 8B. The conversion efficiency is remained constant up to 10 14 cm −2 with the increment of defect density at Sb 2 S 3 /CZTS interface and afterward declined drastically. The maximum PCE greater than 30.63% is determined for the interface defect density below 10 12 cm −2 . Herein, the defect density at both interfaces is used to be 10 11 cm −2 to accomplish the better PCE in the CZTS thin-film PV device with HTL.
In the previous study, it has been discussed that the current collection at longer wavelength affects significantly for the absorber/buffer interface defect density beyond 5 × 10 11 cm −2 . 93 On the other hand, lower sensitivity of short wavelength collection for the interface defects below 1 × 10 11 cm −2 is found in their investigation. 93,94 Furthermore, it has been reported that the performance of the CZTS solar device may be improved by minimizing the interface recombination loss for the CZTS/buffer interface defect density in the range from 10 8 to 10 12 cm −2 . 95 Therefore, the interface defect density of 10 11 cm −2 chosen in this work to attain the better device performance has excellent consistency with the realistic defect density at the CZTS/buffer interface reported in the previous experimental researches. [93][94][95]

Impacts of temperature and rear electrode work function on cell performances
The working temperature plays a vital part in determining the PV performances of the solar cell. Herein, the temperature has been altered in the range from 273 to 475 K to recognize the thermal durability of the proposed CZTS-based solar cell, while the optimized values of physical parameters discussed above for the various layers are used in this numerical calculation. The impact of functioning temperature on V oc , J sc , FF, and efficiency of the proposed CZTS-based TFSC with HTL is depicted in Figure 9A. In Figure 9A, it is clearly seen that V oc , FF, and efficiency deteriorates with increasing the operating temperature. At high temperature, more electron-hole pairs are generated which boost up recombination rates in between energy bands. 73 The internal recombination rates of the carriers surge the reverse saturation current, thus reducing the V oc . Slight rise in J sc is also found with the increase of temperature. This behavior is well consistent with the results obtained in the previous works. 55,61,72,73,88 Furthermore, the enhanced short-circuit current with expanding the functional temperature may be attributed to the reduction in energy band gap of the semiconductors in the heterojunction TFSC. 73,[96][97][98] The combined outcomes of V oc and J sc result in the declination of PCE and FF. FF drops from 90 to 78% with the increase of effective temperature from 273 to 475 K, respectively. The cell efficiency is obtained to be 32.05% and 21.27% at 273 and 475 K, respectively. The simulation has also been performed to explore the impact of the work function on the outputs of the proposed PV structure. The back electrode work function is varied from 4.6 to 5.6 eV. The simulated PV parameters dependent on the work function are displayed in Figure 9B. It is noticed that V oc , FF, and efficiency increase linearly with the increment of work function up to 5.0 eV and then are saturated. In case of J sc , constant result is obtained with the increase of back contact metal work function. Similar behaviors for J sc are also found in the previous reports of TFSCs. 61,88 The higher work function of back metal contact decreases the barrier height, consequently ensuring smoother transportation of majority charge carriers (holes) from the absorber. As a result, the cell efficiency is improved notably. 55,61,88 These results suggest that the rear electrode with work function >5.0 eV is required to realize a highly efficient CZTS PV device. By taking into account of expense and constituent efficacy, in the present study, the Ni as a back metal contact with work function of 5.15 eV is chosen. 99

Impacts of series and shunt resistances on cell performances
Both series (R s ) and shunt (R sh ) resistances have significant roles in determining the PV features of solar cells. Consequently, to alleviate the cell performance, the effects of R s and R sh need to be recognized. In this regard, the simulation is conducted on the proposed CZTS solar cell with Sb 2 S 3 HTL and WS 2 buffer by changing the R s and R sh , while the parameters optimized in the above sections are used for various physical layers. The R s has been varied in the range of 0 to 6 Ω-cm 2 with R sh set at 10 5 Ω-cm 2 to observe the impact of R s on the PV performance parameters, and the simulated results are presented in Figure 10A. This is apparent from the figure that the V oc and J sc have nearly no effect on R s , while the efficiency falls from 30.62% to 24.90%. FF is limited greatly from 88.17% to 71.66% by the large value of R s and this aspect solely contributes in lowering the efficiency. 74 The PV outputs as a function of R sh are also demonstrated in Figure 10B. With R s fixed at 0.5 Ω-cm 2 , R sh is diverted from 30 to 10 7 Ω-cm 2 . It has been found that all the performance characteristics are boosted up to the value of 10 3 Ω-cm 2 and then become constant. This is because high value of R sh prevents power loss in the PV cell. Thus, it can be concluded that both R s and R sh are fundamental to the PV performance of the solar cell and need to be kept at low and high values, respectively. 73 (A) (B)

CONCLUSIONS
In this work, we have designed and studied a novel heterojunction PV device structure of Ni/Sb 2 S 3 /CZTS/WS 2 /FTO/Al using the SCAPS-1D simulator, where Ni and Al are engaged as back and front metallic electrodes, respectively. The WS 2 as a benign and earth-abundant buffer substitute to the toxic CdS buffer in the conventional CZTS solar cell is introduced for the first time. The PV performances of the CZTS solar cell can be improved by forming an appropriate band alignment at CZTS/WS 2 interface with more suitable CBO and lower lattice mismatch than CZTS/CdS interface. To further increase the cell performances, p + type Sb 2 S 3 as an HTL is incorporated between the CZTS absorber and Ni back contact for smooth holes transportation, thus minimizing the carrier recombination loss at the absorber/back contact interface significantly. Moreover, the small lattice mismatch and favorable VBO between Sb 2 S 3 HTL and CZTS absorber enhance the PV outputs of proposed CZTS solar cell. Herein, the impacts of thicknesses and doping concentrations of different layers on the cell outputs are scrutinized. The thermal stability of the proposed CZTS solar cell with Sb 2 S 3 HTL and WS 2 buffer layer has also been researched. In addition, the interfacial quality of both Sb 2 S 3 /CZTS and CZTS/WS 2 interfaces of the sketched solar cell is examined. In this article, the carrier and defect densities for the CZTS absorber are optimized at 10 17 and 10 15 cm −3 , respectively. The PV outputs including V oc of 1.08 V, J sc of 32.05 mA/cm 2 , FF of 88.19%, and PCE of 30.63% of the designed CZTS PV device are estimated with the optimized thicknesses of 0.1 μm for Sb 2 S 3 HTL, 1.0 μm for CZTS absorber, and 0.05 μm for WS 2 buffer layer, respectively. Therefore, it is believed that the findings evaluated in this work will offer significant information to scientists for fabricating highly competent, stable, and cost-effective Cd-free CZTS-based TFSCs.