1 Introduction

In2S3 material is an important member of III–VI group semiconductors family. It is considered as a promising candidate for optoelectronic and photovoltaic applications [1,2,3,4]. In the literature, it is found that In2S3 crystallizes into three different structural forms α, β and γ which are growth temperature dependent [5, 6]. The β-In2S3 structure is the most stable phase at room temperature. It crystallizes in a normal spinel structure with a high degree of tetrahedral and octahedral vacancy sites [7, 8]. Such structure is characterized by a direct band gap semiconductor ranging from 2 to 3.7 eV [7]. The presence of a large number of defects destroys the periodicity of the lattice and distorts the crystal field. Therefore, the semiconducting character of these materials should be strongly affected by annealing process. The latter increases the degree of ordering and hence the material physical properties [9]. To improve the stability of deposited films and to reduce the possible undesirable influence of the substrate, a thermal treatment is usually used. Such process improves crystal quality and reduces structural defects in materials. The annealing process affects the dislocations and other structural defects. Accordingly, the structure and the stoichiometric ratio of the material will change due to the motion of defects and adsorption/decomposition will occur on the surface. Such phenomenon leads to major effects on semiconductor device properties [10,11,12,13].

It is well known that In2S3 thin films are prepared using several methods such as spray pyrolysis [12,13,14,15,16,17], ultrasonic dispersion [18], chemical bath deposition [19], physical vapor deposition [20], hydrothermal [21], etc. As the spray pyrolysis method allows preparing large area of thin films at low cost, it is chosen in the present work to conduct the preparation of In2S3 thin films. Then, we have investigated the effect of vacuum heat-treatment on some physical properties of In2S3 films. The annealing time effect on structural, morphological, optical and electrical properties of the prepared films were studied and discussed.

2 Experiment

The In2S3 films were deposited on glass substrates by using spray pyrolysis technique. The aqueous solution of pulverization contains indium(III) chloride InCl3 and thiourea SC(NH2)2 as indium and sulphur precursors, respectively. Nitrogen was used as a carrier gas. The used solution composition is 10− 2 mol l− 1 of InCl3 and 2 × 10− 2 mol l− 1 of SC(NH2)2. In2S3 material is obtained according to the chemical reaction described by the equation below:

$$2{\text{ InC}}{{\text{l}}_3}+{\text{ }}3{\text{ SC}}{\left( {{\text{N}}{{\text{H}}_2}} \right)_2}+{\text{ }}6{\text{ }}{{\text{H}}_2}{\text{O}}~ \to {\text{I}}{{\text{n}}_2}{{\text{S}}_3}+{\text{ }}3{\text{ C}}{{\text{O}}_2}+{\text{ }}6{\text{ N}}{{\text{H}}_4}{\text{Cl}}$$
(1)

During the pulverization, the distance between the substrate and the jet is 20 cm. The solution and the nitrogen flows were fixed at 2 ml min− 1 and 6 l min− 1, respectively. The sprayed time was 20 min. The obtained samples were sealed under vacuum in Pyrex tubes (10− 5 Torr). Then they were heated at 340 °C for different annealing times (1, 2, 3, 4 and 5 h) according to the profile shown in Fig. 1.

Fig. 1
figure 1

Profile of annealing temperature evolution versus time

Full size image

The film crystalline structure was analyzed by X-ray diffraction (XRD) using CuKα radiation (1.54186 Å). The surface morphology was investigated by means of field emission scanning electron microscopy (FESEM) using a Zeiss FESEM Ultra PLUS microscope, equipped with an energy dispersive X-ray spectroscopy detector (EDS) for chemical microanalysis, and by atomic force microscopy (AFM) using an XE-100 instrument (Park Systems Corporation) in non-contact mode (NC-AFM). Raman spectral measurements were carried out using a Raman spectrometer (Renishawin Via Reflex Spectrometer System) with an excitation wavelength of 514 nm. The transmittance and reflectance spectra were measured using a Shimadzu UV 3101 PC (spectrophotometer), with a wavelength range from 200 to 3000 nm, equipped with an integrating sphere. The resistivity, Hall mobility and carrier concentration were obtained from Hall effect measurements using the standard Van der Pauw configuration.

3 Results and discussion

3.1 Structural characterization

3.1.1 XRD analysis

Figure 2 shows the XRD patterns of In2S3 films annealed at 340 °C for various times (ta). The films are polycrystalline and the crystallographic peaks belong to the cubic β-In2S3 phase with diffractions from (311), (400) and (511) planes, according to the JCPDS Card 32-0456. Impurity phases are not detected.

Fig. 2
figure 2

X-ray diffraction patterns of as-deposited and annealed In2S3 films at 340 °C for different times

Full size image

The peak intensities seem to change with increasing annealing time. Such observation indicates that the recrystallization process occurs during the annealing treatment.

The preferred orientations of In2S3 films were evaluated by the texture coefficient (TC), calculated from the X-ray data using the well-known formula:

$$TC(hkl)=\frac{{I(hkl)/{I_0}(hkl)}}{{\frac{1}{N}\sum {I(hkl)/{I_0}(hkl)} }}$$
(2)

where TC(hkl) is the texture coefficient of the (hkl) diffraction plane, I(hkl) is the measured or normalized intensity, I0(hkl) is the corresponding standard intensity given in JCPDS data, and N is the number of reflections. The film preferred orientation will be (hkl) plane corresponding to the higher value of TC(hkl). TC(hkl) calculated for (311), (400) and (511) peaks were summarized in Table 1. The highest TC values are obtained for (400) plane for all the films.

Table 1 Crystallite size, dislocation density and texture coefficient for different annealing times
Full size table

Therefore, the film preferential orientation is (400). The (400) diffraction peak, for the film treated for 3 h, has the highest texture coefficient.

A rough estimate of the crystallite size (D) was made from X-ray diffraction data from (400) reflection by using the Scherer formula [22]:

$$D=\frac{{0.9\lambda }}{{\beta \cos \theta }}$$
(3)

where β is the size broadening in radians obtained as \(\beta =\sqrt {\beta _{e}^{2} - \beta _{0}^{2}}\), where βe is the full width at half maximum of the diffraction peak and β0 is the instrumental broadening (1.5 × 10− 3 rd), λ is the wavelength of X-ray and θ is the Bragg angle.

The dislocation density (δ) is described as the length of dislocation lines per unit volume of the crystal [23]. The dislocation density gives information about the crystal structure. For the preferential orientation (400), such parameter can be calculated using the formula [24]:

$$\delta =\frac{1}{{{D^2}}}$$
(4)

Table 1 summarizes the calculated values of grain size and dislocation density of films.

It is observed that the grain size increases from about 39 to 49 nm with increasing of annealing time. Such observation can be related to the improvement of the film recrystallization [11, 25]. Similar results were recently reported by Avilez Garcia et al. [26]. The dislocation density decreases from 6.60 × 1010 to 4.16 × 1010 lines cm− 2 with annealing time. Such behavior can be explained by the change of the grain size with annealing time. Indeed, the smaller crystallites have a greater surface to volume ratio. Hence, the dislocation density decreases with increasing annealing time which may be due to the decrease of lattice imperfections concentration [27].

3.1.2 Raman analysis

The Raman spectra of the films are shown in Fig. 3. The presence of β-In2S3 phase modes in all the films is noticed [28]. Indeed, the active modes of β-In2S3 observed at 139, 264, 312, and 368 cm− 1 confirm the composition and the structure of the samples [29, 30]. The Raman modes are broad with low intensity. The film annealed for 3 h has a good crystallinity. The peak position and the half width of the β phase Raman bands exhibit independence upon the annealing time. However, there is a slight change in the relative intensities between the Raman bands at 264 and 312 cm− 1 that seems to be correlated to the structural changes observed by X-ray diffraction.

Fig. 3
figure 3

Raman spectra of In2S3 thin films annealed at 340 °C at different times

Full size image

3.2 Surface morphological and compositional analysis

Figure 4 illustrates 2D AFM images of In2S3 thin films annealed at 340 °C for different annealing time.

Fig. 4
figure 4

2D AFM images of as-deposited and annealed In2S3 thin films at 340 °C with different times

Full size image

The effect of annealing time on the film morphology properties such as the surface average and RMS roughness are listed in Table 2. It is observed that the average and RMS roughness values decrease with the annealing time increase up to 3 h.

Table 2 Average and RMS roughness of In2S3 thin films annealed at 340 °C for various annealing times, obtained from the AFM images
Full size table

However, the surface becomes rougher as the annealing time is increased to 4 h and 5 h. An increase of the surface roughness may be due to the structural modification produced by the increase of annealing time. The film crystallites become larger leading to the obtention of larger surface roughness [26, 31]. The films surface morphology was also studied by FESEM. The typical FESEM images for different annealing times, are depicted in Fig. 5. These images show that the morphology of the investigated films has larger number of grain size and are homogeneously distributed. It is also observed that the grains are formed in cubic structure with void space between the grains. FESEM image of In2S3 film, annealed for 1 h, shows that the particles exhibit a small size. As the annealing time increases to 3 h, the particles size increases. Thus, the particle size, of the In2S3 films, increases rapidly with increasing the annealing time. FESEM results confirm the increase of grain size increasing annealing time as revealed by XRD measurements.

Fig. 5
figure 5

FESEM images of In2S3 thin films prepared at 340 °C with different annealing times (1 h, 2 h, 3 h, 4 h and 5 h) and EDS spectrum of In2S3 films annealed at 340 °C for 3 h

Full size image

For the investigated samples, annealed for different times, the chemical composition was studied using EDS analysis. The obtained results are given in Table 3. A typical EDS graph, for selected sample, is also shown in Fig. 4. It is observed that the atomic ratio S/In is not sensible to the increase of annealing time. Its average value is in order of 1.4. Such value is lower than the molar ratio in the initial solution (S/In = 2) and the stoichiometric one (S/In = 1.5). This deviation is probably due to the volatility of sulfur. It can be concluded that the annealing time has not a significant influence on the stoichiometry of the formed In2S3 layers.

Table 3 Atomic contents in In2S3 thin films annealed for various times characterized by EDS technique
Full size table

3.3 Optical properties

3.3.1 Optical transmittance and reflectance spectra

Figure 6 shows the spectral distributions of both the transmission (T) and the reflection (R) at normal incidence with annealing time. For 1 h annealing, the transmission decreases to a saturation value of about 65% (Fig. 6a). When the annealing duration exceeds 1 h the film transparency is not remarkably affected. We point out the presence of the interference fringes due to the multitude reflection phenomena showing homogenous films. However, this decrease occurs mainly in transmission, while reflection is almost not affected (Fig. 6b). Moreover, these films exhibit a good transparency in the visible and infrared regions. The films are transparent of about 70%, and the total reflection (R) remains nearly at the same value of about 20% for all films. In general, the optical transmittance is influenced mainly by the film thickness, the surface morphology as well as the grain density and defects at the grain boundaries [11, 32].

Fig. 6
figure 6

Optical transmission (a) and reflection (b) spectra of as deposited and annealed In2S3 films treated at 340 °C as a function of annealing time

Full size image

3.3.2 Refractive index and thickness

The film thicknesses were calculated from the positions of the interference maxima and minima of transmittance spectra using a standard method. The refractive index can be calculated by using the maxima (Tmax) and the minima (Tmin) of the envelopes of the transmission spectra [33] given on Fig. 6. The expression of refractive index is given by [34]:

$$n={(N+{({N^2} - {n_0}^{2}{n_1}^{2})^{\frac{1}{2}}})^{\frac{1}{2}}}$$
(5)

where

$$N=\frac{{{n_0}^{2}+{n_1}^{2}}}{2}+2{n_0}{n_1}\frac{{{T_{\hbox{max} }} - {T_{\hbox{min} }}}}{{{T_{\hbox{max} }}{T_{\hbox{min} }}}}$$
(6)

n0, n1 are the refractive index of air and substrate, respectively.

The thickness (d) was given by [35]:

$$d=\frac{{{\lambda _1}{\lambda _2}}}{{2(n({\lambda _1}){\lambda _2} - n({\lambda _2}){\lambda _1})}}$$
(7)

The evolution of the thickness and the average refractive index of the layers with annealing time are illustrated in Fig. 7. We notice that the layers become thinner by increasing the annealing time. This fact can be explained by the grains coalescence. Also, in the region of transparency, the annealing time has not a remarkable effect on the refractive index of the samples.

Fig. 7
figure 7

Evolution of refractive index and thickness as a function of annealing times

Full size image

3.3.3 Band gap energy

The film absorption coefficient (α) was calculated using the equation:

$$\alpha = - \frac{1}{d}\ln \frac{T}{{1 - {R^2}}}$$
(8)

The analysis of the absorption coefficient, in the high absorption region is carried out to obtain the detailed information about the band gap energy. The optical band gap of the films is inferred using the following relation [11]:

$$\alpha h\nu =A{(h\nu - {E_g})^n}$$
(9)

where A is a constant, α is the absorption coefficient, hν is the photon energy, Eg is the optical band gap and the exponent (n) characterizes the nature of band transition. Indeed, n = 1/2 and 2 correspond to direct allowed and indirect forbidden transitions, respectively. Figure 8 shows the variation of (αhν)2 against (hν). The band gap energy is determined by extrapolating the straight line portion to the energy basis at α = 0. We observe that the curve has a very good straight line fit.

Fig. 8
figure 8

Dependence of (αhν)2 versus hν for as deposited and annealed of In2S3 thin films at different annealing times

Full size image

The films exhibit a direct band gap varying from 2.63 to 2.38 eV as a function of annealing time. Moreover, the band gap energy could also be influenced by the variation of particle size with annealing time as it is in the nanometer range. Though, it is not close to the Bohr radius. Therefore, in the present study, the increase of grain size by the agglomeration of particles, leads to the decrease of band gap energy of the films with increasing annealing time [36]. Indeed, as revealed by the XRD patterns and AFM images, the films undergo a crystallization process during the annealing route. Similar variation of band gap energy was observed by Revathi et al. [37]. The measured Eg values are lower than those reported in reference [38]. In the literature, large variations of the band gap of indium sulphide films were reported [39,40,41,42]. Such type of variation is generally attributed to structural effects. It was observed that the band gap energy has the highest value for 1 h heat treated thin film. Then it decreases with increasing annealing time up to 2 h. However, for samples annealed for more than 2 h, the band gap energy increases with annealing time. Such result can be related to the enhancement of the realignment in orientation and strong interaction between the substrate and vapor atoms [43].

3.4 Electrical properties

The resistivity, the carrier concentration and the mobility values of the In2S3 thin films annealed at different time were evaluated by Hall measurement system at room temperature. The corresponding values were listed in Table 4. The carrier concentration (nv) is derived from the relation \({n_v}=\frac{1}{{e.{R_H}}}\) [44], where RH is the Hall coefficient and e is the absolute value of the electron charge. The carrier mobility (µ) is determined using the relation \(\mu =\frac{1}{{e{n_v}\rho }}\), where ρ is the resistivity [45].

Table 4 Resistivity, carrier concentration and mobility of In2S3 films annealed at 340 °C as a function of annealing time
Full size table

All films exhibit n-type conductivity which means that the annealing time did not change the type of the majority carriers. Moreover, the mobility and the carrier concentration vary in the range 1.65–29.3 cm2 V−1 s−1 and 1.16×1017–2.82×1017 cm− 3, respectively. The mobility increases as a result of reconstituting and the crystallinity prosperity of the material. The carrier concentration variation is probably due to the variation of sulfur vacancies. The electrical resistivity of the films decreases from 14.47 to 1.65 Ω cm when annealing time increases. Such decrease is consistent with the results observed from XRD diffractograms. Indeed, the increase of grain size during the annealing process results in a smaller effect of grain boundary leading to a reduction of the films resistivity. The noticed reduction of the electrical resistivity is proven to be quite advantageous for photovoltaic applications.

4 Conclusions

In this work, we report the effect of annealing time on sprayed In2S3 films deposited under optimum conditions (Ts = 340 °C, S/In = 2) and annealed in vacuum at different times (1 h, 2 h, 3 h, 4 h and 5 h). Results showed that the annealing duration ameliorates the films crystallinity and their orientation. The obtained films were polycrystalline with cubic crystal structure and with preferential orientation (400). The film thickness decreases from 2.7 to 1 µm with annealing time but the refractive index is not sensitive to such factor. The grain size increases from 39 to 49 nm with the annealing time. Also, the morphology and the crystallinity of the films are strongly influenced. Raman spectroscopy confirms the presence of β-In2S3 phase for all films. The films are transparent in the visible (80%). The values of optical band gap were evaluated and were found to be ranged between 2.63 and 2.38 eV with the annealing time. The carrier concentration and the mobility vary in the ranges 1.16 × 1017–2.82 × 1017 cm− 3 and 1.65–29.3 cm2 V−1 s−1, respectively. The electrical resistivity of the films decreases from 14.47 to 1.65 Ω cm, indicating an improvement of the conductivity. Annealed films with these properties may be promising material for optoelectronic applications.