1 Introduction

Current developments of semiconductor thin films have shown extensive attention in optoelectronic devices. Among the optoelectronic devices, the photodetector is one of the prominent devices with extensive applications in communication and automation. The II–VI compound semiconductors have shown enormous potential for photodetector device fabrication. The II–VI semiconductors such as ZnS and ZnSe have advantages in photodetector applications [1], [2]. ZnS and ZnSe are nontoxic wide bandgap semiconductors with high UV and visible absorption. The presence of sulfur atoms in the ternary ZnSxSe1−x exhibits the wideband bap, which influences the response in the UV region. The tuning of the bandgap of the films can be achieved with the composition variation [3]. In our previous work [4], we deposited ZnSxSe1−x thin films on a glass substrate using binary source materials ZnS and ZnSe for photodetector application. The structural and optical characteristics of the films were highly affected by the composition parameter ‘x’. As composition changed from x = 0.0 to 1.0, the bandgap was tuned from 2.5 eV to 3.5 eV a systematic shift was observed in the position of XRD peak. The carrier concentration and resistivity of the films were highly influenced by the composition. Also, the photo response of the film has shown superior properties for the composition x = 0.4. The ZnSxSe1−x film with x = 0.4 can be used as a photodetector, as observed from structural, electrical, and photo response studies.

In device applications, silicon-based thin film heterojunctions have several advantages. Silicon-based heterojunctions are low-cost systems, and they can be easily integrated into optical devices. The performance of silicon with II–VI semiconductors thin films have been explored tremendously with its reliability and superior characteristics [5]. ZnSxSe1−x/Si heterojunction shows good quantum efficiency and photo response in the UV-Visible region. M.Das and team [6] studied the individual and enhanced UV to visible light sensing property of metal-semiconductor-metal (MSM)-based ZnS–PS (Porous Silicon):p–Si hybrid heterostructure by the response and recovery time in 365 and 400 nm, respectively. E. Coskun et al.[7], studied the device properties of n–ZnSe/p–Si heterostructure, and the diode parameters like ideality factor and series resistance were found to be 3.12 and 416 ohm, respectively. H.K. Kaplan et al.[8], have fabricated ZnS/Si heterostructure by thermionic vacuum arc technique. The series resistance and the built-in potential were calculated as 3.6 kΩ and 0.7 V, respectively. S. Venkatachalam et al. [9] have fabricated ZnSe/p-Si Schottky diodes, and from the I–V measurements, the evaluated ideality factor was found in the range of 2.8–3.01. The built-in potential and barrier height were calculated from C–V measurements as 0.61 eV and 1.457 eV, respectively. Yi-Lin Sun et al. [10] synthesized UV detection hybrid graphene/cadmium-free ZnSe/ZnS quantum dots photoresistors. They attained a high responsivity above 103 AW− 1 in the wavelength region 405 nm with a response time of 0.52 µs. Arka Dey and team [11] studied the light-induced charge transport property of nanostructured ZnS-based Schottky diode; the rectification ratio (on/off ratio) was found to be 28.2 and 70 at 2 V for dark and illumination. The Schottky diode parameters like ideality factor, barrier height, and series resistance were evaluated for measuring conditions.

Among physical preparation techniques, the thermal co-evaporation technique has superior characteristics. The composition ‘x’ of the films can be easily controlled in the adopted technique. In this method, the film grows under a vacuum, which reduces the oxidation and defects in the film [12][13]. Also, the semiconductor heterojunction can be fabricated at a low-cost from the thermal co-evaporation technique [14][15]. Even though ZnSSe is one of the cadmium-free chalcogenides, the preparation and characterization of the p–Si/ZnSxSe1−x heterostructures have not been analyzed in detail. The properties of heterostructure and its photodetector behavior are yet to be investigated. In the current project, the ZnSxSe1−x films were synthesized on Si substrate (p–Si) by the thermal co-evaporation technique. The structural, optical, and electrical properties of the films have been studied for different compositions. The optimized Si/ZnSSe heterostructure was fabricated for photodetector applications has been analyzed.

1.1 Experimental details

ZnSxSe1−x films were grown on p-Si substrate by thermal co-evaporation method, where the deposition was carried out inside 12-inch diameter vacuum chamber with a diffusion pump backed by a rotary pump (Make: Mansha Vacuum Equipments Pvt. Ltd. Bangalore, India). The source materials ZnS and ZnSe (Alfa Aesar, 99.999% purity) were taken according to the required composition in 200 A molybdenum (Mo) boats and placed in the vacuum chamber. The 500 μm thick Boron doped Si wafer was taken as substrate, and before the deposition, the wafer was cleaned by standard methods. To remove the oxide layer from the polished surface of the wafer dil. HF acid solution, Methanol, and distilled water were used. The source and substrate holder were maintained at an optimized distance to obtain the homogeneous film, i.e., about 16 cm, and films were grown with the deposition rate of 0.2–0.3 nm/sec. The base pressure of 2 × 10− 6 mbar was achieved, and the Si wafer was kept at room temperature throughout the deposition. After the deposition of the p–Si/ZnSxSe1−x heterojunctions (dimension 1 cm × 1 cm), the film’s thickness was measured by a stylus profilometer (Model: Brucker DektakXT). The HRXRD was used to quantify the X-ray Diffractogram of the films. To record emission and RAMAN (RT and LT) spectra, LABRAM Horiba spectrometer was used. The electrical measurements were done by Keithley 2450 source meter by varying temperature and illumination conditions.

2 Result and discussion

2.1 Structural analysis

The grazing incidence X-ray Diffraction (GIXRD) studies have been supported to study the structural properties of the ZnSxSe1−x (0.0 ≤ x ≤ 1.0) films grown on p–Si (100) substrates. Figure 1a represents the GIXRD patterns of p–Si/ZnSxSe1−x (0.0 ≤ x ≤ 1.0) heterostructures with a grazing incidence angle 2°. The GIXRD patterns of the x = 1.0 (binary ZnS) films exhibited a sharp, intense peak at 28.57°, corresponding to (111) lattice reflection planes of cubic zinc blende ZnS phase and a less intense peak around 56° corresponding to (311) plane of the ZnS. Similarly, GIXRD patterns of the x = 0.0 (binary ZnSe) films showed a prominent peak at 27.22° related to (111) lattice reflection planes of cubic zinc blende ZnSe phase and less intense peaks around 45° and 55° matching with (220) and (311) planes of the ZnSe, respectively [16]. All the films showed polycrystalline nature, and the Bragg’s angle for the highly intense peak was shifted to lower angles i.e., from 28.57° to 27.22°, which confirms Se atoms substitute S atoms as composition ‘x’ varied from 1.0 to 0.0. The structural properties like interplanar distance (d), lattice parameter (a), the crystallite size (D), strain (ε), and dislocation density (δ) were calculated using the formula [15][17],

Fig. 1
figure 1

a X-Ray Diffractogram, b PL spectra of p–Si/ ZnSxSe1-x (0 ≤ x ≤ 1) heterostructures

Full size image
$$2d\,sin\theta\,=\,n\lambda$$
(1)
$$\frac1{d^2}\,=\,\frac{h^2\,+\,k^2\,+\,l^2}{a^2}$$
(2)
$$D\,=\,\frac{0.9\lambda}{\beta\,cos\theta}$$
(3)
$$\varepsilon\,=\,\frac{\beta\,cot\theta}4$$
(4)
$$\delta\,=\,\frac1{D^2},$$
(5)

Here, θ is Bragg’s angle, λ is the radiation wavelength, (h, k, l) is the miller indices, β is the peak width. From Table 1, it is confirmed that the structural constraints of the p–Si/ZnSxSe1−x heterostructures is influenced by the composition ‘x’. As composition parameters vary from 1.0 to 0.0, the interplanar distance ‘d’ and lattice parameter ‘a’ were shifted from 3.12 Å to 3.27 Å and 5.41 Å to 5.66 Å. The value of the ‘d’ and ‘a’ are increased as composition varied from 1.0 to 0.0 due to the ionic radii of the Se (190 pm) atom being larger than the S (180 pm) atom. The D was found to be in the range of 19 to 12 nm, and extreme compositions have a larger crystallite size. The random variation in structural parameters might be because of the inconsistency in the proportion of lattice strains in ZnSSe (difference in the lattice constant of ZnS and ZnSe) individual composition, which develops the possibility of random movement of the atoms [18]. Due to the variation in the composition and strain present in the films, the crystallite size was varied with ‘x’. The strain is formed due to the lattice mismatch between silicon and ZnSxSe1−x (0 ≤ x ≤ 1). Also, crystallographic defects present in the films can contribute to it. Strain is undeniable in thin films, which can emerge because of a mismatch in the thermal coefficient between films and the substrate. The dislocation density is characterized as the length of separation lines present in the unit volume of the crystal, giving the details about variation of the stoichiometry, which influences the volumetric development of the films. Dislocation density and strain are influenced by composition ‘x’. Bigger crystallites reduce the defects in the crystal lattice, which minimizes the strain and dislocation density in the films. Also, the improvement in the crystallites leads to the increment in the mobility and reduction in the resistivity. Microstructural properties such as higher crystallites, minimum strain, and dislocation density are essential for photodetector devices. So, the structural properties of these films have been considered to achieve favorable optical and electrical characteristics. The ε and δ got in the range of 10− 3 and 1015 m− 2 depending on a subsequent change in the size of crystallites. Among the ternary heterostructures, x = 0.8 exhibited good crystalline properties with low structural defects, which is important for the effective performance of the photodetector. Figure 2 depicts the SEM image of p–Si/ZnSxSe1−x heterostructures. It confirms that films were crack free, uniform, and dense.

Table 1 Structural properties of p-Si/ZnSxSe1-x heterostructures
Full size table
Fig. 2
figure 2

SEM micrographs of p-Si/ ZnSxSe1-x (0 ≤ x ≤ 1) heterostructures

Full size image

2.2 Photoluminescence analysis

Photoluminescence spectra were recorded and analyzed to investigate the energy states formed in the material. Figure 1b depicts the Gaussian deconvoluted PL spectra of p–Si/ZnSxSe1−x films recorded with an excitation wavelength of 325 nm (3.82 eV). The PL spectra of the binary ZnS (x = 1.0) have two peaks around 405 and 503 nm; these emission peaks are less than the band edge of the ZnS films, which correspond to the defect levels emission (DLE). These defects involve only small groups of atoms known as point defects, including interstitials, vacancies, stacking faults, and surface states plentiful on the surface of the nanostructures. Based on previous articles, a peak at 400 nm is ascribed to transitions from sulfur vacancy-related donor to valence band edge, which is also observed in the present work [19][20]. The peak around 503 nm in both the binaries (x = 0 and x = 1) corresponds to the defects related to Zinc interstitials in the films [21]. The spectra of ZnSxSe1−x films with x ranging from 0.2 to 0.8 is a composite of ZnS and ZnSe-based defect states. In this range (x = 0.2 to 0.8), the peaks around ~ 400 nm and ~ 500 nm are due to the sulfur vacancy and Zinc interstitials in the deposited films, respectively. The Zinc interstitials and sulfur vacancy peaks are common in all the heterostructures of p–Si/ZnSxSe1−x (x = 0 to1), i.e., the samples are zinc-rich, and sulfur and selenium concentrations are less than expected. This is common in II–VI chalcogenides grown by PVD, which may be assigned to the difference in vapor pressure of constituent elements [13]. The films with higher Zn content showed n-type conductivity [4]. Therefore, the Fermi level in ZnSSe approaches to the conduction band [22].

2.3 Raman analysis

The Raman spectra of p–Si/ZnSxSe1−x (0.0 ≤ x ≤ 1.0) heterostructure with Lorentzian fit was recorded with 532 nm excitation are shown in Fig. 3. From Fig. 3, the mode at 520 cm− 1 corresponds to Si vibrational mode in all the heterostructures. It is possible to observe the phonon modes of the deposited substrate if the grown ZnSxSe1−x films are thin (thickness around 250 nm) and transparent at excitation source radiation [23]. The excitation wavelength and optical density are related, and longer excitation radiation diffuses deeper into the films [13]. The penetration depth is higher for materials with high transparency; the thickness of all the films is around 250 nm, which could be the reason for observing the Raman mode of Si substrate. For x = 0 (i.e., ZnSe), the Raman mode at 250 cm− 1 is allocated as LO mode of ZnSe [24]. For binary ZnS, a peak around 269 cm− 1 is related to TO mode of cubic ZnS [25]. Due to the strain, the shifting of the Raman modes in ZnS and ZnSe was observed from the bulk phonon modes. In ternary samples, as ‘x’ increases from 0.0 to 1.0, the Raman mode shifts from 250 cm− 1 to 269 cm− 1, analogous to pure ZnS and ZnSe. The mode around 304 cm− 1 in all ternary ZnSxSe1 − x corresponds to the (LO + TA) mode of the ZnSe [26], which is very nearby (LA + TA) to ZnS [27]. The Raman mode of ternary ZnSxSe1−x is a combination of ZnS and ZnSe phonon modes, and it verifies the modulation in the composition of the ternary alloys [3].

Fig. 3
figure 3

Raman spectra of p–Si/ ZnSxSe1-x (0 ≤ x ≤ 1) heterostructures

Full size image

LT Raman spectroscopy investigates the chemical structure and temperature-dependent strain established in the films. The LT Raman spectra for p–Si/ZnS0.8Se0.2 heterostructures from 300 K to 120 K is shown in Fig. 4a. At 300 K, TO (ZnS) and Si phonon modes are observed at 252.24 and 520 cm− 1. The presence and position of the Raman modes also depend on the excitation source wavelength. The crystal lattice parameters are highly sensitive to ambient temperature. The redshift and blueshift in the phonon modes are mainly because of the tensile and compressive stress in the films[28]. As temperature decreases from 300 K to 120 K, due to the tensile stress, the TO (ZnS) Raman modes move to the lower wavenumber side.

Fig. 4
figure 4

a Low-temperature Raman spectra of p–Si/ZnS0.8Se0.2 heterostructure, b Peak positions of TO(ZnS) at various temperatures

Full size image

Based on the anharmonic model with respect to temperature, the Raman shift ω(T) can be linearly fitted [14],

$$\omega\left(T\right)\,=\,\omega_0\,+\,C\left[\frac2{e^x-1}\right],$$
(6)

where x = \(\frac{hc{\omega }_{0}{10}^{2}}{2T{k}_{B}}\), ω0 and C are the Raman band at zero kelvin and the tunable parameter, respectively. The anharmonic model for purely quasi-harmonic input is based on the three-phonon scattering process. In Fig. 4b, the dashed line shows the linear fitted data (R2 = 0.97) gained from the anharmonic model. The dot represents the experimental value from 300 K to 120 K. The experimental and theoretical statistics are in line with each other and trailing thermal anharmonic behavior, and 95% of the values are within the two standard deviations of the mean. The fitting parameters obtained for TO (ZnS) are ω0 = 242.2 cm−1 and C = 3.72 cm−1.

2.4 Electrical analysis

The p–Si/ZnSxSe1−x (0.0 ≤ x ≤ 1.0) heterojunction demonstrates a good rectification behavior. The temperature-reliant I–V curve of p–Si/ZnS0.8Se0.2 heterostructure in the forward bias is shown in Fig. 5a. To calculate the diode parameters and to analyze the main carrier conduction mechanism over the junction, temperature-dependent I–V measurements were achieved. As temperature increases, the value of forward current increases in all heterostructures because of the inhomogeneous trap distribution and effect of trap levels localized at the intersection [29]. According to TE model, the current (I) and voltage (V) can be expressed as [23],

Fig. 5
figure 5

a I–V characteristics (inset: device structure), b Semi-logarithmic I–V plot) at different temperatures for p–Si/ZnS0.8Se0.2 heterostructure

Full size image
$$I\,=\,I_S\left[\exp\left(\frac{qV}{\eta kT}\right)\right],$$
(7)

where \(I_S\,=\,AA^\ast T^2\exp\left(\frac{-q_B}{kT}\right)\), q, k, A*, A, and T are the saturation current in the reverse direction through the junction, the charge of an electron, Boltzmann constant, Richardson constant, active area of the device, and temperature, respectively. Eq. (7) can be expressed as[23],

$$lnI\,=\,\frac{qV}{\eta kT}\,+\,lnI_S$$
(8)

The ideality factor (η) was estimated by the slope of I–V plot of the heterojunction (Fig. 5b), and the barrier height (ΦB) was calculated by[23],

$$\phi_B\,=\,\frac{kT}q\ln\left(\frac{AA^\ast T^2}{I_S}\right)$$
(9)

The obtained parameters are depicted in Table 2. We can observe the deviation from the linearity in a semilogarithmic I–V plot after a bias voltage 0.5 V, due to the presence of interface states and series resistance. Table 2 shows that as the temperature increases η reduces, and ΦB improves. The increase in the temperature induces the charge carriers, which acquire sufficient energy to cross the high barrier height. η for the ideal diode should be one; the variation from this value (greater than one) implies nonideal performance and inhomogeneity across the junction. The higher ideality factor may be because of leakage current, free carriers recombination in the depletion region, and lattice mismatch [30]. The barrier height variation is strongly dependent on temperature, demonstrating that carriers moving across the interface are stimulated by temperature.

Table 2 Variation of diode parameters with temperature for p-Si/ZnS0.8Se0.2 heterostructure
Full size table

2.5 Photosensitivity

To study the photosensitivity, I–V characteristics of p–Si/ZnSxSe1−x heterojunction were studied under dark and illumination conditions. Compared to all the heterojunctions, films with x = 0.8 had shown maximum photoresponse at constant bias. Figure 6a evaluates I–V measurements of x = 0.8 heterostructure for different wavelength illumination conditions. The origin of dark current may be due to the generation-recombination of carriers, tunneling, leakage, and diffusion currents. Under illumination conditions, the rectifying behavior remains unaltered, and the photocurrent improves because of the creation of electron-hole pairs and the rise in the mobility of the free charge carriers.

Fig. 6
figure 6

a Illumination dependent I–V characteristics, b Variation of photosensitivity with wavelength, c semilogarithmic I–V plot at different illumination wavelengths for p–Si/ZnS0.8Se0.2 heterostructures

Full size image

The photosensitivity (S) of p–Si/ZnSxSe1−x (0.0 ≤ x ≤ 1.0) heterojunctions were calculated by [4],

$$S\,=\,\frac{I_{Light}\,-\,I_{Dark}}{I_{Dark}},$$
(10)

Here Ilight and Idark are current at illumination and dark conditions. Figure 6b depicts the photosensitivity (S). The heterostructure with composition x = 0.8 showed good photoresponse with maximum sensitivity at 635 nm. Photoresponse is highly dependent on absorbance as it increases the electron-hole pair upon incident radiation. ZnSxSe1−x (0 ≤ x ≤ 1) thin films are highly transparent than other chalcogenides. If the incident photon wavelength is equivalent to the absorption edge of the films or the defect states in the films, the mobility of the free charge carrier increases, leading to high photocurrent in the region of wavelength [14]. The sample x = 0.8 showed a larger grain size and fewer crystal structural defects, which are essential factors in showing very high photosensitivity. The diode parameters of p–Si/ZnS0.8Se0.2 were calculated using TE model. Figure 6c denotes the semilogarithmic I–V plot under dark and illumination for p–Si/ ZnS0.8Se0.2 heterojunction. The η and ΦB showed converse relation between them upon illumination of wavelength are presented in Table 3. Bilgin [31] investigated that the value of the ideality factor decides the conduction mechanism in the diode. η = 1 to 2, η = 2, and η > 2 are known for tunnelling, generation-recombination, and leakage current mechanisms. The heterostructure is exposed to the illumination, then the photo-excited charge carriers surrounded by grain boundaries lead to a decrease in the ΦB; thus helping transporter movement with the minimum resistive environment, and barrier height decreases as the illumination power increases. [14].

Table 3 Variation of diode parameters with illumination wavelength for p–Si/ZnS0.8Se0.2 heterostructure
Full size table

3 Conclusion

In the current work, ZnSxSe1−x (0.0 ≤ x ≤ 1.0) thin films were grown on p-Si (100) by thermal co-evaporation technique using ZnS and ZnSe as the source materials. The GIXRD study of the ZnSxSe1−x (0 ≤ x ≤ 1) films exhibited a cubic phase with a prominent (111) orientation. The higher crystallinity was found for x = 0.8. PL measurements confirmed the presence of zinc interstitials and sulfur vacancy in the films. The low-temperature Raman spectra of x = 0.8 samples were analyzed using the anharmonic model. In temperature-dependent I–V measurements, the ideality factor was found to decrease from 4.22 to 2.13 with the temperature rise from room temperature to 320 K. Also, the barrier height increased with the temperature. The p–Si/ZnSxSe1−x heterojunction with x = 0.8 has exhibited significant photoresponse at 635 nm.