Photoresponse in sequentially stacked antimony selenide thin films

Antimony selenide (Sb2Se3), a binary semiconducting compound has widespread research attention due to its excellent optoelectronic properties in the visible region and usefulness in applications such as solar cells, photosensors and photoelectrodes. The presented study explores the thickness dependent photoresponse in Sb2Se3 thin films, prepared by reactive selenization of antimony films having thickness values of ∼938 nm and ∼1879 nm when stacked second time. Growth orientation along [001] direction was achieved through carefully optimized selenization conditions to enable favourable charge transport in anisotropic Sb2Se3. Predominant Sb2Se3 formation was inferred from x-ray diffraction, Raman spectroscopy, secondary electron microscopy and energy-dispersive X-ray analyses. High optical absorption coefficient values of about 1 × 105 cm−1 and 5.7 × 104 cm−1 were observed for ∼938 nm and ∼1879 nm thick Sb2Se3 thin films. Further, the optoelectronic properties were elucidated through current–voltage and transient photoresponse measurements under dark and illumination conditions. The measurements were done under zero and different bias voltages. Sb2Se3 films having∼ 938 nm thickness exhibited self-driven photoresponse with a responsivity of 4.3×10−8 A W−1 and detectivity of 3.5 × 106 jones respectively, under AM 1.5 G illumination conditions.


Introduction
Antimony selenide (Sb 2 Se 3 ) as a binary V-VI semiconducting compound, has been gaining increasing attention in recent years due to its significant material [1], optical [2], electrical [3] and photoelectric properties [4], in addition to its environmentally benign constituents and stability under ambient conditions [5].Additionally, the direct band gap (E g ) of 1.1−1.3eV in the NIR region, large absorption coefficient (of the order of 10 5 cm −1 in the visible region), high carrier mobility (about 10 cm 2 V −1 s −1 ) [1,6], and broad spectral response from ultraviolet region to near-infra red region makes it an optimal choice for various applications such as optical storage [7], visible-to-near-infrared photodetectors [8] and photovoltaic applications [9].
In technologically matured commercialized silicon solar cells, the minority carrier lifetime of the carriers is about ∼ 1-10 μs [10].However, the established absorbers for thinfilm solar cells, namely, Copper Indium Gallium Selenide (CIGS) and Cadmium Telluride (CdTe) show carrier lifetime of ∼20 ns and ∼1.7 ns respectively [11,12].In the case of quaternary Cu 2 ZnSn(S,Se) 4 (CZTSSe), the recombination lifetime is even lower, which is less than 15 ns [13].Compared to the above thin film absorber materials, Sb 2 Se 3 demonstrates a notably longer carrier lifetime of approximately 70 ns which is considered advantageous for photovoltaic applications, as it allows for more efficient charge carrier collection, reduced recombination losses, and potentially improves the device performance.
From crystallography perspective, the low symmetry in the Sb 2 Se 3 unit cell, as a result of lattice distortion plays a crucial role in stabilizing a 5s 2 lone pair in an antibonding position, strategically blocking a bonding direction [17].Therefore, Sb 2 Se 3 possesses a quasi-one-dimensional ribbon-like morphology.The ribbons, denoted as (Sb 4 Se 6 ) n , align along the [001] direction through strong covalent Sb-Se bonds.In contrast, in the [100] and [010] directions, the (Sb 4 Se 6 ) n ribbons are bound together by weak van der Waals forces [18].This direction-dependent bonding in Sb 2 Se 3 imparts robust anisotropic optoelectronic properties.Due to this strong in-plane anisotropic characteristics such as linear dichroism [19], Sb 2 Se 3 extends its applications in polarized photodetection, and polarimetric imaging [20] applications as well.
Photoresponse studies of the Sb 2 Se 3 thin films deposited on the glass substrates measured with a silver metal contact will provide insights into the charge carrier transport properties enabled by the ohmic contact between the p-type Sb 2 Se 3 with an electron affinity of ∼ 4.08 eV [31] and silver with work function of ∼4.26 eV [32].With that perspective, we investigated the structural, optical, morphological and photoresponse behavior of Sb 2 Se 3 thin films prepared via reactive selenization of antimony films having two different thicknesses.

Materials and methodology
Glass substrates, measuring 1 × 1 cm, were cleaned by immersing them in chromic acid for 24 h.Following this, the substrates were sonicated sequentially in de-ionized water, isopropyl alcohol, and acetone, each for 15 min.Following the sonication, the glass substrates were subjected to plasma cleaning for 10 min to eliminate any dust accumulation and to enhance adhesion between the substrates and the subsequent antimony layer coating.
2.1.Preparation of Sb 2 Se 3 thin film 60 mg of elemental antimony (Sb) powder was loaded into the molybdenum evaporation boat in a thermal evaporation unit, with a source-to-substrate distance of 5 cm.The evaporation was done at ∼3 × 10 −5 mbar, without any intentional heat supplied to the substrates using ∼ 60 A current, until the complete evaporation of source powder to obtain antimony films.
Afterward, the thermally evaporated Sb films were selenized at 375 °C in a three-zone furnace using 70 mg of elemental Se (5 N pure, acquired from Alfa Aesar) as a vapor source.The distance between the thin film and the selenium powder in the graphite box was maintained at 3 cm.Argon gas was supplied at a rate of 250 SCCM and the graphite box was inserted once the set temperature of 375 °C was reached.The selenization was allowed for 30 min and then the films were cooled down to room temperature and taken out.

Preparation of stacked Sb 2 Se 3 thin film
With an aim to double the thickness of Sb 2 Se 3 thin films, Sb deposition and subsequent selenization cycle was repeated sequentially as presented in the figure 1.The resultant films were obtained through first Sb deposition/ first selenization/second Sb deposition/second selenization.This stacking configuration was chosen to increase thickness, in view of inducing improved inter diffusion of Se and Sb.The thickness values were inferred to be ∼938 nm (referred as S1) and ∼1879 nm (referred as S2) respectively for single and two-time deposition of Sb 2 Se 3.

Characterization techniques
Structural characterization to know the crystallinity and the phase formation in the grown films was done using x-ray diffraction (XRD) (Bruker D8 Advance x-ray Diffractometer) with CuK α radiation (λ = 0.15406 nm).The XRD instrument was operated at 40 kV and 40 mA, and data were acquired over a range from 10°to 70°.Rietveld refinement of the experimental data was performed using GSAS II software.Raman spectral data revealing lattice vibrations, crystallinity and phase information were obtained using a HORIBA LabRam HR Evolution MicroRaman Spectrometer.This spectrometer employed an internal HeNe laser with an incident wavelength of 633 nm.Scanning electron microscope (SEM) with energy-dispersive X-ray (EDX) analysis from Thermo Scientific Apero S provided insights into the surface morphology, cross-section, thickness, elemental distribution (along line and areal sections) and composition of the films.The optical transmittance and absorption in the films were measured using a UV-Vis-NIR spectrophotometer (UV-3600 plus series) to determine the absorption coefficient and band gap values.Photoluminescence (PL) of the films was measured using a spectrofluorometer (Horiba Fluorolog-QM) with an excitation wavelength of 450 nm at room temperature.Current-voltage (I-V) and transient photoresponse (I-t) measurements were done under dark and illumination conditions.For I-t measurements, white light illumination power density values were varied from 20 to 100 mW cm −2 in a Class AAA solar simulator (Newport Corporation) and the current values were measured using Keithley 2400 source meter.The AAA-class solar simulator used for the photoresponse of the samples in this study is equipped with a Xenon lamp with a power rating of 1000 W. It has a uniform output beam of 6 × 6 inches and a beam uniformity of 2%, with a beam divergence of < ± 3°.Besides, the effect of applied bias on photoresponse was measured at two different biasing voltage values of 3 and 5 V under AM1.5 G illumination conditions.

XRD analysis
Typical XRD patterns in figure 2(a) reveal the polycrystalline nature of both S1 and S2 Sb 2 Se 3 films, as evidenced by distinct sharp peaks.The primary peaks at the 2θ values of 45.5°, 31.1°,28.2°and 23.6°are due to (002), ( 221), (211), and (101) planes respectively.All the peaks in the XRD patterns match with the standard JCPDS data (card no.015-0861) for Sb 2 Se 3 , confirming the formation of single-phase Sb 2 Se 3 with an orthorhombic crystal structure.Notably, no discernible peaks belonging to elemental Sb or Se, or any secondary phases were observed.Both the films demonstrate a predominant orientation along the (002) plane, which is highly favorable for charge carrier transport, due to its intrinsic quasi-one-dimensional crystal structure [33].It is noteworthy that the dominance of the (002) orientation in the S2 is significantly higher, which is evident through the suppression of intensity in other peaks, as compared to S1.The (hk0) planes such as (120), (240), and (340) are in very low intensities and are barely detectable which is commonly observed in the Sb 2 Se 3 thin films prepared by the various physical and chemical-based methods [21,[34][35][36].In this presented method, the films exhibit a [001] orientation, indicating that they align perpendicular to the substrates, in contrast to (hk0) and (hk1) orientations, where the films align parallel or at a specific tilted angle to the substrate as indicated in the figure 2(b).To quantitatively assess this degree of orientation preference in both the films, the texture coefficient (TC) was calculated using the formula proposed by Harris et al [37].
where N is the number of diffraction peaks, I(hkl) and I 0 (hkl) are the intensity values from the experimental and standard data.TC (hkl) is the texture coefficient of the plane defined by Miller indices (hkl).Four prominent reflections corresponding to (211), ( 221), ( 301) and (002) were considered for the texture coefficient calculation and the calculated values are presented in figure 2(c).As evident from the graph (211), ( 221) and (301) planes exhibited lower texture coefficient values compared to the (002) plane.Also, a higher texture coefficient value of 3.7 was obtained for the S2, than for S1, which is 3.1, indicating the prominence of [001] orientation and improved crystallinity in the S2 films having higher thickness.This process facilitates the robust growth of Sb 2 Se 3 films onto pre-existing layers, resulting in enhancement in vertical orientation growth and superior crystalline quality when compared to its thinner counterpart.
Further, to estimate the quality of the films prepared, microstructural properties such as crystallite size, lattice strain and dislocation density were calculated as follows and the results are presented in table 1.
The crystallite size ( ) D was calculated using the Debye-Scherrer equation: Where the Scherrer constant, denoted as K, is commonly assigned a value of 0.94, λ is the wavelength of the radiation used ∼1.54 Å, and θ corresponds to the diffraction angle of the peak.The other microstructural parameters namely the dislocation density (δ) and the lattice strain (ε) of the films are calculated using the equations (3) and (4).The S2 Sb 2 Se 3 film exhibited lower lattice strain and dislocation density compared to S1 suggesting improved quality of the films.This effect is attributed to two key factors namely the implementation of heat treatment twice, which results in enhanced crystalline properties, and secondly, the increment in the thickness of the S2 films [7].
Rietveld refinement was performed, considering the Pbnm space group.As shown in figures 2(d) and (e), a fine consistency between the experimental and theoretical profiles was observed and is also reflected in the χ 2 values for both samples, 1.08 and 1.53 respectively.The obtained crystallographic parameters are tabulated in table 1.The calculated lattice parameter values and the cell volume are consistent with the reported literature values [38,39].

Raman analysis
The Raman analysis was performed to determine and resolve any unidentified phases or elemental residues present in the prepared films.The deconvoluted spectra with identified Raman modes in the range 100-300 cm −1 are presented in figure 3. A 633 nm laser with a 1% excitation power from 19 mW laser power was used as an excitation source.For both spectra, the scan was performed for 30 s with 12 accumulations, at room temperature.The primary structural units are made up of the Sb-Se stretching A g mode, which is characterized by the strongest Raman peak at 190 cm −1 in both films [40].The Raman peak located at 102 cm −1 is attributed to the A g mode of Sb 2 Se 3 [41].The peak at 118 cm −1 is identified as the A g mode and is associated with the Se−Se bending.The A 2u mode due to Sb-Sb bond has been identified at 153 cm -1 and 212 cm -1 [42].The observation of these vibrational modes indicates the Sb 2 Se 3 phase without any secondary Sb 2 O 3 phase [43] as indicated in figure S1 (Supplementary Information) or elemental Sb or Se in both films.No apparent shifts or differences were observed among S1 and S2 films.

SEM analysis
The surface morphology and cross-section images of both the Sb 2 Se 3 thin films are presented in figures 4(a)-(d).The grains in both Sb 2 Se 3 thin films exhibit a compact and homogeneously distributed structure, with welldefined grain boundaries.Significantly there are no observable cracks, pinholes, or delamination in either film, which is indicative of the good film quality.A notable distinction is observed in the S2 thin film, which has undergone selenization treatment twice.From the corresponding SEM image (figure 4(c)) the influence of heat treatment on grain boundaries is visible.The grains in the S2 film tend to merge with adjacent grains, indicating a consumptive effect.Contrastingly, the S1 films, as depicted in figure 4(a), do not show this effect of apparent grain consumption.Also, discernable voids are observed at the backside of the S2 film, contrary to the S1 films.This observation suggests a potential separation of selenium from the structure of the S2 film, likely occurring during the second heat treatment process [44].In addition, the thicknesses observed from the cross-section analysis are ∼ 938 nm and 1879 nm respectively.
The elemental composition of the films, represented by the Se/Sb ratio, was determined through surface and cross-sectional EDS analyses.The ideal stoichiometric ratio of Se/Sb in Sb 2 Se 3 is 1.50 [45].For S1, a close to ideal stoichiometric ratio of 1.43 was observed on the surface and at the cross-section, this ratio decreased to 1.04, implying a potential limitation in the formation of Sb 2 Se 3 at the back side of the films despite surface stoichiometry nearing ideal conditions.In contrast, for S2, consistent Se/Sb ratios of 1.32 were observed both at the surface and the cross-section.This uniformity suggests the complete intermixing of selenium throughout the film, most likely due to the dual selenization process.
Statistical plots in figures 5(a) and (b) provide data from various locations, incorporating both line and areal EDS scans conducted across the surface and cross-section of the sample, revealing substantial deviation of composition on the surface and cross section in S1 while indicating no significant variance between the surface and cross-section in S2.
The elemental distribution for both the films obtained on the surface, appears uniform and is presented in figure S2.The homogeneous distribution of both elements in S2 is depicted in figure 6, demonstrating no zonation of elemental composition.This observation indicates the congruent mixing of Sb and Se through this method.However, in S2 films, a reduced Se percentage was observed, indicating the need for supplementing more Se during selenization.

UV-analysis
The optical properties of both the S1 and S2 films, including transmittance, absorption coefficient, and band gap, were analysed, and are presented in figure 7. The optical transmittance spectra obtained in the 400 nm − 900 nm wavelength range are significantly low, attributing to strong photon absorption.Beyond 900 nm, the transmittance sharply increased with the wavelength and in concurrence with the reported values [46].
The optical absorption coefficient of the S1 and S2 films were calculated using the relation.

( ) ( ) a = -T t ln 5
Where T represents the transmittance, and t denotes the thickness of the prepared samples.As a semiconductor material, the Sb 2 Se 3 exhibits a pronounced ability to absorb photon energy due to the presence of lone pairs of  electrons from both Sb and Se, leading to a high density of states [21].The absorption coefficient encounters a rapid, sharp surge in response to photon energy, ranging from 1 eV to 1.5 eV.This elevated absorption behavior persists until 3.5 eV, demonstrating a high affinity for absorption in Sb 2 Se 3 throughout the entire visible region, indicating its suitability for photovoltaic and broadband photo detection applications.It is crucial to observe that the experimental data has confirmed that the optical absorption coefficient of the S1 film is 1× 10 5 cm −1 , while for the S2, an absorption coefficient of 5.7× 10 4 cm −1 was observed.Further, the optical band gap value E g was determined through the plots of (αhν) 2 versus hν.The calculated value of E g was approximately 1.18 eV for the S1 and 1.3 eV for the S2 films closely aligning with the reported band gap value for Sb 2 Se 3 thin films [47].

Photoluminescence studies
Photoluminescence studies were employed to probe and analyze the correlation between stoichiometric deviations and specific defects and understand the radiative recombination of both the S1 and S2 films.The PL yield for the S1 was observed to be in higher order than the S2.In S1 films, the higher PL intensity observed indicates reduced/lesser non-radiative recombination sites than the S2 films.It is inferred from the crosssectional compositional analysis that the stoichiometric ratio of Se/Sb layers is lesser than ideal 1.5, exhibiting a slightly antimony-rich condition, which has led to the enhancement of non-radiative recombination.
The PL spectra in the figure 8 show peaks corresponding to band edge emission and other sub band gap defect-related emissions, which are attributed to arise from the fabrication method, indicative of nonuniformity in the selenization of antimony films.The defect emission bands observed at various places such as 1.05 eV, 1.11 eV 1.14 eV, 1.17 eV, 1.24 eV, and 1.26 eV in both the samples are attributed to arise from the defect states in the films.It has been reported that the deep defects would be higher at the antimony-rich compositional conditions, giving rise to defect-related peaks in the PL spectra [48,49].

Photoresponse studies
To explore the optoelectronic properties of the fabricated films, silver (Ag) contacts were thermally deposited onto the sides of Sb 2 Se 3 films on glass substrates.This study is indented to understand the intrinsic photoresponse properties of deposited Sb 2 Se 3 thin films.These films were subsequently studied under at various illuminated power densities, including AM 1.5 G light illumination.Initially, the investigation was conducted without applying any external voltage bias, i.e., at 0 V condition.However, the observed response was very minimal, the measurements were subsequently extended under applied bias voltages of 3 and 5 V.All these studies were performed under room temperature.The schematic representation of the measurement is provided in figure 9.
The I-V curves for both Sb 2 Se 3 films, measured in the dark and under light illumination at different power densities, are presented in figures 10(a) and (b).The symmetric I-V curves confirm the ohmic nature of the contact and the current show a linear correlation with the applied bias voltage.Further, the gradual increase in photocurrent (I ph ), I ph = I light −I dark is evident as the power density of incident light rises, indicating enhanced generation of photogenerated carriers [50].
Stability and response rate are crucial factors for assessing the photo-detecting properties and to gauge these attributes, the time-resolved photoresponse of the detector was examined by manually switching the illumination source on/off at an interval of 10 s under a zero-bias condition.The corresponding I-t graphs presented in figures 10(c) and (d) demonstrate rapid and notable photoresponse for all power densities and exhibited minimal deviation, with nearly identical responses after multiple cycles within a 100 s hold time, indicating robustness and reproducibility.
The rise time (τ rise ) and decay time (τ decay ) are quantified as the duration required in the response current of the photo detector for transition from 10% to 90% of its maximum level and return to 10%, respectively.A single on and off cycle of the rise and decay time curve is presented in figure 11(a) for S1.The calculated rise and   Where, R s is the responsivity, and q is the charge of an electron.The calculated R s values of both the Sb 2 Se 3 thinfilms under different intensities are presented in the figure 11(c) and range between 1.9 × 10 −8 to 4.3 × 10 −8 A W −1 and 4.3 × 10 −9 to 7.9 × 10 −9 A/W for the S1 and S2 respectively.The calculated D * of the films lies between 2.3 × 10 6 to 3.5 × 10 6 Jones, and 1.0 × 10 6 to 1.3 × 10 6 Jones, for the S1 and S2 respectively.
The estimated on and off ratios are 398 and 53 for the S1 and S2 respectively, with S1 exhibiting superior values.It is important to note that comparing the performance parameters such as I ph , R s , on and off ratio and D * requires consideration of the the active area of the device as well and hence difficult in direct comparison with the reported literature values [30].
However, in this study the comparative analysis of optoelectronic characteristics, of prepared Sb 2 Se 3 films derived from the photoresponse properties, determines the self-biased photodetecting attributes primarily due to their photovoltaic behaviour.Notably, the S1 films outperformed the S2 films, indicating that the optimal thickness for enhanced optical and optoelectronic properties of Sb 2 Se 3 lies within the range of 850-950 nm.This observation denotes that the film thicknesses contribute significantly to the favorable performance of Sb 2 Se 3 in self-powered photodetection applications.
Nevertheless, under the influence of a specific applied voltage such as (3 V and 5 V) as presented in figure 12, the I ph increased with an increase in applied voltage due to the enhanced carrier drift velocity and the corresponding decrease in carrier transit time [51].It is to be mentioned that in the case of applied bias, the I ph of the S2 surpassed that of the S1.
Since the photocurrent generation involves photon absorption, photo-carrier generation, and subsequent extraction and transport, despite low photon absorption in S2 films, efficient carrier extraction, facilitated by the vertically oriented structure of S2, is believed to significantly enhance photocurrent transport.The enhanced texture coefficient of [001] orientation, particularly along the S2, facilitates a more facile movement of the generated charge carriers under the significant applied bias voltage [26].This suggests that, despite the higher thickness, under the influence of voltage bias, the stacked configuration enables much favourable carrier transport and extraction, due to the higher texture coefficient of the preferred orientation.The rise and decay time of the S1 and S2 lies in the range of 0.25 s as represented in figure 13(a) and has also improved compared to the self-bias analysis, due to the impact of applied voltage and is comparable with the values reported for Sb 2 Se 3 based nanowires in the literature [52].Similar comparative analysis of the samples S1 and S2 under the applied voltages (3V and 5V) pertaining to the photocurrent and the responsivity are given in figures 13(b) and (c).
The calculated detectivity of the S1 and S2 films were in the order of 3.1-3.8×10 5Jones under the applied bias voltages 3 V and 5 V, which is one order lesser than the detectivity obtained for the films when operated at zero bias.This is attributed to the impact of dark current that prevails in the films because of applied voltage.Semi logarithmic dark I-V characteristics of S1 and S2 are presented in the figures 14(a) and (b).

Conclusion
This study has comprehensively explored the structural, compositional, optical, and photoresponse characteristics of Sb 2 Se 3 thinfilms, prepared by reactive selenization of antimony films having thickness values of ∼938 nm and ∼1879 nm when stacked second time.Additionally, critical properties, including responsivity, detectivity, rise and decay time revealed that the thinfilms having ∼938nm are well-suited for the photo sensing applications under self-biased conditions, and have the potential in the field of optoelectronics.Under applied bias conditions, the thicker films having ∼1879 nm showed fair photoresponse behaviour.The studies conducted affirm that the method chosen for Sb 2 Se 3 film preparation resulted in films with potential optoelectronic characteristics that are comparable to other reported methods and has the viability for large area applications.

Figure 1 .
Figure 1.Schematic illustration of the Sb evaporation and subsequent selenization employed for the preparation of Sb 2 Se 3 thin films.

Figure 2 .
Figure 2. Typical XRD spectra of (a) S1 and S2 Sb 2 Se 3 films (b) Schematic representation of [001]oriented films with respect to the substrate and (c) Texture coefficient analysis of the S1 and S2 films (d) and (e) Rietveld refinement for S1 and S2 films respectively.

Figure 5 .
Figure 5. Statistical plots of Se/Sb ratios in S1and S2 films (a) combined surface and cross-section data, (b) separated surface and cross-section.

Figure 6 .
Figure 6.(a) Cross sectional SEM image of S2, cross sectional elemental mapping of S2 using EDS, (b) overlapped Sb and Se, (c) Sb and (d) Se.

Figure 9 .
Figure 9. Schematic representation of Sb 2 Se 3 thin films on glass substrates with Ag top contact and illumination source depicting the configuration used in photoresponse measurements.

Figure 10 .
Figure 10.(a) and (b) I-V characteristics of S1 and S2 films under dark and white light illumination conditions at various power densities; (c) and (d) (I-t) response cycles of S1 and S2 under different white light power densities at zero bias.

Figure 11 .
Figure 11.(a) Representative plot of current versus time showing τ rise and τ decay of the S1 film at 100 mW cm −2 power density at zero bias voltage (b) plot of τ rise and τ decay for both S1 and S2 films at different power densities.(c) Responsivity as a function of illumination power densities.

Figure 12 .
Figure 12.I-t response cycles of S1 and S2 films under (a) 3 V and (b) under 5 V applied bias voltages.

Figure 13 .
Figure 13.(a) Representative plot of I-t showing τ rise and τ decay of the S1 at 100 mW cm −2 illumination (5 V) (b) Calculated photocurrent values as a function of applied bias voltage for S1 and S2 films.(c) R s values as a function of applied bias voltage for S1 and S2 Sb 2 Se 3 films.

Table 1 .
Microstructural properties and crystallographic parameters of the prepared Sb 2 Se 3 films.