Study of thermally evaporated Sb2Se3-based substrate-configured solar cell

Antimony selenide (Sb2Se3) has emerged as a promising absorber material for thin film solar cell (TFSC) application. In this work, a (120) oriented substrate-configured Sb2Se3 based TFSC has been fabricated using the thermally evaporated Sb2Se3 thin films. Pre-synthesized bulk Sb2Se3 was used as a source material and the films were subjected to post-deposition selenization. TFSCs were fabricated in a device configuration of Glass/Mo/Sb2Se3/CdS/ITO/Ag. It was found that there is a significant increment in the power conversion efficiency (PCE) with increased Voc and Jsc in the devices, wherein the Sb2Se3 absorber films were subjected to post-deposition selenization compared to the devices made with as-deposited films. TFSC with as grown Sb2Se3 film was showing an efficiency of ∼ 1% with Voc ∼ 208 mV, Jsc∼16 mA cm−2 and fill factor (FF) ∼ 29.9%. The device with selenized Sb2Se3 films showed a power conversion efficiency of 3.38% with Voc, Jsc and FF values of 362 mV, 18.54 mA cm−2 and 50.39%, respectively. The increase in PCE for selenized films is attributed to better grain growth and suppression of selenium vacancy defects. Overall, the findings of this work demonstrate the potential prospects of Sb2Se3 as an absorber material for TFSCs applications and suggest that post-deposition selenization plays a significant role in the enhancement of device efficiency. The obtained results are contributive in the understanding and development of low-cost Sb2Se3-based TFSCs.


Introduction
In the recent years, thin film based solar cells have emerged as one of the promising alternative to commercial silicon based solar cells.Among them CdTe and Cu(In,Ga)S 2 (CIGS) based solar cells have shown significant advancements with power conversion efficiencies of ∼22.1% and ∼23.35% respectively [1].However, the concerns regarding the toxicity of cadmium and the scarcity of indium & gallium in the absorber materials, limits their ability to meet the industrial demands.As a result, chalcogenide based thin film materials are being explored as a promising, sustainable, non-toxic substitutes in place of CdTe and CIGS.For instance, the quaternary compound absorber Cu 2 SnZn(S,Se) 4 (CZTSSe) has demonstrated power conversion values upto ∼12.6%.However, the chemical complexity, possibility of formation of numerous native defects and secondary phases, hinder its potential as an viable alternative to CdTe/CIGS based solar cells.
TFSC based on antimony selenide (Sb 2 Se 3 ) has recently gained tremendous interest in the research community owing to its single phase structure [2], favorable bandgap of ∼1.1-1.3 eV and high absorption coefficient of ∼10 5 cm −1 in the visible region [3].Moreover, it consists of earth-abundant, low-cost elements making it favorable for large-scale production.Stibnite Sb 2 Se 3 has orthorhombic crystal structure and consists of (Sb 4 Se 6 ) n ribbons covalently bonded along the [001] direction.These ribbons are confined to each other by weak Vander Waal's force along the, [100] and [010] direction [3,4].Due to this, Sb 2 Se 3 shows an anisotropic behavior, where its electrical, magnetic [5] and optical properties [6] depend on growth orientation.
A considerable amount of reports are available on superstrate-configured Sb 2 Se 3 solar cells with thermally evaporated absorber layer, but very few results are available for substrate-configured solar cells.Among these reports, Yao et al [20] have shown a highest efficiency of 6.24% for superstrate configured Sb 2 Se 3 solar cells, in which the (221) oriented thermally evaporated absorber was subjected to H 2 S treatment prior to subsequent device fabrication steps.Li et al [14] fabricated substrate-configured Sb 2 Se 3 based solar cells, wherein coevaporation of Sb 2 Se 3 and Se resulted in the formation of absorber film with (221) orientation and have achieved an efficiency of 3.07%, which further improved to 4.25% upon back contact selenization.
In this current work, we fabricated substrate-configured Sb 2 Se 3 based solar cells in which Sb 2 Se 3 films were deposited by thermal evaporation from pre-synthesized bulk.After the deposition of Sb 2 Se 3 films, postdeposition selenization was performed.

Experimental procedure
2.1.Deposition of Sb 2 Se 3 film Firstly, Sb 2 Se 3 thin films were deposited on Mo films coated on glass substrates (thickness ∼ 700 nm) by thermal evaporation using pre-synthesized bulk Sb 2 Se 3 as source material.Before the deposition of Sb 2 Se 3 films, Mo films coated on glass substrates were cleaned thoroughly by sonicating the substrate in de-ionized water for 30 min.The glass substrate was having a thickness of 2 mm.Mo films coated on glass substrates is henceforth referred as Mo substrates for improving readability.Bulk source Sb 2 Se 3 was synthesized by mechanical alloying (the synthesis details are published elsewhere) [31].The source Sb 2 Se 3 was kept on a molybdenum boat.The distance between the source and substrate was kept at 5cm.Substrates were heated from room temperature at an average rate of 10 °C /min to 320 °C.The evaporation was maintained for 5 min ensuring the complete evaporation of the source material.Sb 2 Se 3 deposition was done using thermal evaporation system (Hind High Vacuum, 12A4D) operated at a base pressure of ∼ 5 × 10 -5 mbar.

Selenization of Sb 2 Se 3 film
After the deposition of films, the films were subjected to post-deposition selenization.Selenization was done in a three-zone furnace.Firstly, the samples were placed in a graphite box containing 30mg of selenium powder and placed on an unheated zone.Then, a vacuum was created inside the tube using a rotary pump.After the vacuum is reached, the rotary pump was switched off and argon gas flow was maintained at 250 sccm.Thereafter, the temperature was set at 300 °C with a ramp rate of 10 °C per minute in all the three zones of the tube.When the temperature reached 300 °C, the graphite box was pushed inside till it reaches the middle portion of the heating zone.Selenization was carried out for 30 min at a temperature of 300 °C under Ar atmosphere.After the selenization is done, the furnace is allowed to cool naturally and samples were taken out once it reached near room temperature.

Fabrication of substrate photovoltaic device:
After selenization of Sb 2 Se 3 , the CdS buffer layer having ∼50 nm thickness was deposited by chemical bath deposition using cadmium nitrate and thiourea as cadmium and sulphur source as described in literature [33].ITO layer of ∼300 nm was deposited by RF sputtering [34].Finally, the device was completed by deposition of silver contacts.The devices were separated mechanically using a diamond scriber.The complete device configuration is as follows; Glass/Mo/Sb 2 Se 3 /CdS/ITO/Ag.Two sets of devices were studied, one (i) without selenization or as-grown films and (ii) the other with selenized films.Henceforth these are referred as (ASe-A) and (ASe-S) respectively.

Characterization tools used
The phase formation of Sb 2 Se 3 films grown by thermal evaporation was confirmed by x-ray diffraction (XRD) using Bruker D8 advance, wherein CuK α was used as incident radiation having a wavelength of ∼1.54 Å with a step size of 0.02 degree.Further confirmation of phase formation was done using a micro-Raman spectrophotometer (Horiba JobinYvon HR) using an incident laser beam of wavelength 633nm acquired at 3.2% of its total power.Low laser power was used in order to avoid any localized heating.The thickness measurement was performed using an optical profilometer (MicroXAM-800, KLA TENCOR).Surface morphology and elemental composition were examined using HRSEM (Thermo scientific Apreo S).Sb 2 Se 3 solar cell device J-V measurements were studied using Keithley 2400 under dark and illumination conditions.The illumination was done with an Oriel 3A class solar simulator with an intensity of 100 mW cm −2 operated at room temperature, without an aperture (AM 1.5).Standard Si solar cell was used as reference cell for calibration.

Structural analyses
Figure 1 shows the typical XRD plot of the as grown and selenized Sb 2 Se 3 grown on Mo substrates.From the XRD data analysis, it is inferred that the films exhibited crystalline nature without any detrimental secondary phase such as Sb 2 O 3 or elemental Sb and Se.The absence of elemental selenium peaks in the obtained results suggests the complete incorporation of selenium into the films surface.All the peaks obtained from XRD were assigned to orthorhombic Sb 2 Se 3, which matches with standard JCPDS data (file no.00-015-0861).It is also noticed that the films are predominantly oriented along (120) plane, due to their thermodynamic tendency to grow along low energy planes.From XRD results, it was observed that the diffraction peak intensity values of ASe-S were more compared to that of ASe-A, indicating that selenization of films has incubated better crystalline growth.
The crystallite size D for the as grown and annealed films were calculated using Debye-Scherrer's formula Where, k is the Scherer constant (0.9), λ is the wavelength of x-ray (0.154 nm), β is the full width half maximum of dominant peak for the samples.The crystallite size of the films increased to 83.96 nm for the ASe-S films from 59 nm as observed for the ASe-A films.
It is worth noting that the device showed better efficiency when the films are oriented along [001] plane [2].
Raman spectroscopy was employed further to confirm the phase purity of the Sb 2 Se 3 films grown on Mo.Raman spectra were recorded using a 633nm laser in backscattering mode with 3.2% of its total power.As per the literature, theoretical studies have reported 30 Raman active modes for Sb 2 Se 3 .Figure 2 shows typical Raman spectra of ASe-A and ASe-S films on Mo substrate.Raman peaks were observed at 155, 191 and 211 cm −1 for ASe-A, while for ASe-S, Raman peaks were observed at 117, 155,191 and 211 cm −1 .All the vibration modes observed in the spectra can be assigned to the Sb 2 Se 3 phase [35,36].It is also seen that the Raman peaks were sharper and well resolved in ASe-S compared to ASe-A, which also agrees with the results of XRD analyses indicating selenization of the films led to increase in crystallinity.

Surface morphology and elemental composition studies
Figure 3 shows the surface morphology of ASe-A and ASe-S films on Mo substrates.Distinct particles with welldefined grain boundaries indicate well-defined growth in both films.It is also seen that the particle size got enhanced for post-selenization of the Sb 2 Se 3 films grown on Mo substrates.Post-deposition selenization plays a critical role in compensating the selenium loss that has been commonly reported to occur during the thin film growth from the Sb 2 Se 3 compound via evaporating methods due to the low vapor pressure of selenium.The loss of selenium leads to selenium vacancy and thereby generating deep recombination centres, affecting the device performance.Further, post-selenization helps in reducing the grain boundaries and helps in enhancing particle size growth.EDS analysis showed the composition of ASe-A to be selenium-poor and antimony rich with an atomic percentage of Sb ∼ 47.43% and Se ∼ 52.57%.After the selenization of the films, the elemental composition of Se increased to 58.57% and with Sb as 41.43% for ASe-S films, as indicated in table 1.

Thicknessanalysis
The thickness measurements were done using the optical profilometer for both as-deposited ASe-A and selenized ASe-S films.The as-deposited films exhibited a thickness of approximately 743 nm, while the selenized films showed a slightly higher thickness of around 825 nm.This increase in thickness is attributed to arise from  the inclusion of selenium during the selenization process, that results in lattice expansion within the Sb 2 Se 3 structure, thereby improvement in the volume of the films [20].
3.4.Current density versus voltage (J-V) studies J-V studies were done on the completed solar cells using a Keithley source meter and Oriel Sol 3A solar simulator for both the devices i,e.ASe-A and ASe-S having active area of 0.05 cm 2 and 0.08 cm 2 , respectively.Figure 4. shows the schematic diagram of a typical Sb 2 Se 3 solar cell.Figure 5 shows the typical J-V curve for the ASe-A and ASe-S devices.The measured and calculated solar cell parameters for the both devices are given in table 2.
As presented in the table 2, analysis of the J-V data shows PCE of ∼1% for ASe-A solar cells with V oc ∼ 208 mV, J sc ∼ 16 mA cm −2 and FF ∼ 29.9%.This low efficiency is attributed to arise from V se and Sb se antisite defects, a dominant defect in a Se-poor absorber [37] as a results of partial decomposition of Sb 2 Se 3 leading to the loss of selenium during thermal evaporation.With post-selenization, the power conversion efficiency has enhanced to ∼ 3.38% with a substantial increment in V oc ∼ 362 mV, J sc ∼ 18.54 mA cm −2 and FF ∼ 50.39%.This enhancement in PCE is also explainable in terms of the effect of series (R s ) and shunt (R sh ) resistance values.For a good solar cell, R s should be significantly low, whereas R sh should be very high [38].From table 2, it is seen that the R s values got reduced from 11.2 to 5.8 Ω cm 2 while R sh values increased from 24.25 to 175.51 Ω cm 2 for ASe-A and ASe-S devices, respectively.For the highest achieved efficiency of Sb 2 Se 3 , R s and R sh values obtained are 2.23 and 606.04 Ω cm 2 respectively [7].
The increment in R sh of ASe-S is due to enhanced grain growth as inferred from HRSEM images and better compositional ratio confirmed through EDS.Inclusion of Se in the Sb 2 Se 3 matrix results in suppression of defects, thereby presents a low resistance route for generated carriers to by-pass p-n junction.
It should also be noted that the growth orientation plays a vital role in the charge transport mechanism in Sb 2 Se 3 solar cells owing to its anisotropic nature.For better charge transport, the favorable growth orientation is along the [001] direction, whereas our Sb 2 Se 3 film is oriented along a less favorable (120) plane.The efficiency results of substrate-configured Sb 2 Se 3 solar cells fabricated by thermal evaporation by different groups, as reported in the literature is summarized in table 3.In all the reports, (in table 2) the absorber Sb 2 Se 3 films were prepared by thermal evaporation; however, in the case of the solar cell with the highest reported efficiency, back  contact selenization was also done and Se was co-evaporated along with Sb 2 Se 3 during the evaporation to compensate Se loss.

Conclusions
In this work, substrate-configured Sb 2 Se 3 solar cells (Glass/Mo/Sb 2 Se 3 /CdS/ITO/Ag) were fabricated, in which Sb 2 Se 3 absorber films were grown by thermal evaporation using pre-synthesized bulk.It was found that the films were near stoichiometric, single phase and having growth orientation along (120).It is observed that selenization of the as grown films improved the crystallinity, composition and enhanced the particle size.It was also found that the as-grown films were Se-poor, which was compensated by selenization.The post-deposition selenization led to an increase in V oc , J sc and FF compared to the device with as-grown Sb 2 Se 3 films showing an efficiency of 3.38%, V oc ∼ 362 mV, J sc ∼ 18.54 mA cm −2 and FF ∼ 50.39%.

Figure 1 .
Figure 1.Typical XRD spectra of ASe-A and ASe-S films on Mo substrate.

Figure 2 .
Figure 2. Typical Raman spectra of ASe-A and ASe-S films on Mo substrates.

Figure 4 .
Figure 4. Schematic diagram of a substrate configured Sb 2 Se 3 solar cell.

Figure 5 .
Figure 5. J-V characteristics of ASe-A and ASe-S solar cells.

Table 1 .
The atomic percentage composition values of ASe-A and ASe-S films.

Table 2 .
Measured and calculated solar cell parameters of ASe-A and ASe-S devices.

Table 3 .
Comparison of substrate-configured Sb 2 Se 3 solar cell device parameters from literature with the present work.