The bulk TeSeSn was prepared by the melt-quench technique. Materials (99.99% pure) Te, Se and Sn (from Aldrich, UK) were weighted (6 g total weight) according to their atomic percentage. The weighted elements were placed into a quartz glass ampoule and sealed under vacuum of 10 - 5 Torr. The sealed ampoule was heated in Heraus programmable tube furnace (type R 07115). The heating rate was approximately 4 K/m. The temperature was kept at 800˚C for 14 h and the ampoule was rocked during the melt process to ensure complete mixing and reaction. After that, the ampoule was quenched into ice-water mixture.

Thin films were prepared by thermal evaporation under vacuum of 10 - 5 Torr using the Edwards E-306 coating system. A constant evaporation rate (4 nm/sec) was used to deposit the films. The evaporation rates as well as the films thickness were controlled using a quartz crystal monitor (FTM5). TeSeSn films were annealed at different temperature (323 ≤ T_ann ≤ 423 K) for one hour under gas Nitrogen. The morphology for as-deposited and annealed films were investigated using (SEM) type JEOL JSM-T200. The crystalline phases for as-prepared and annealed films were identified using a Philips diffractometer type 1710. The X-ray diffraction patterns were analysed using some software. Peakfit program was utilized to determine and identify the peaks of the patterns. These peaks were employed using Williamson-Hall method to estimate the crystallite size and lattice strain for the samples. Chekcell program was used to obtain and refine the cell parameters [11] . The optical transmittance (T) and reflectance (R) of the as-deposited and annealed TeSeSn films were measured at room temperature using a double-beam spectrophotometer (SHIMADZU UV-2101combined with a PC) in the wavelength 300 - 900 nm.

The morphology of the samples for as-prepared and after annealing temperature was examined using SEM. The samples were gold coated before SEM examination to study the surface morphology. The scanning micrograph specimens of as-prepared and annealed at 373 K for hour are shown in Figure 1(a) and Figure 1(b). The microstructure obtained for the as-prepared TeSeSn composition is shown in Figure 1(a), it is clear that the crystalline phase is embedded in the amorphous matrix. In general, after annealing to 373 K the system reveals that the amount of the transformed crystalline phase increases and the crystallized particles decreases in size as shown in Figure 1(b).

In order to determine the crystalline phases that appeared in SEM the X-ray diffraction pattern of films was analyzed. Figure 2 shows XRD pattern of the as-deposited and annealed TeSeSn films. The Williamson-Hall method relies on the principle that the approximate formulae for size broadening, β_L, and strain broadening, β_e,

vary quite differently with respect to Bragg angle, θ:

where K is the crystallite shape constant (≈0.89), ε lattice strain and L crystallite size. First contribution varies as 1/cosθ and the other as tanθ. The first contribution of crystallite size was measured by the Scherrer method as given in Equation (1) [12] . If both contributions are present then their combined effect should be determined by convolution. The simplification of Williamson and Hall is to assume the convolution is either a simple sum or sum of squares. Using the sum of these we get:

If we multiply this equation by cosθ we get:

and comparing this to the standard equation for a straight line (m = slope; c = intercept). We see that by plotting β_totcosθ versus sinθ we obtain the strain component from the slope (4ε) and the size component from the intercept (Kλ/L). Such a plot is known as a Williamson-Hall plot [13] . The dislocation density (δ) is defined as the length of dislocation lines per unit volume of the crystal and is given by, δ =1/L².

The deduced L, δ and ε are listed in Table 1. It is observed that the strain and crystal size decrease whereas dislocation density increases with increasing the annealing temperature. The calculated values of structure parameters of TeSeSn thin films for as-prepared and different annealing temperatures are listed in Table 2.

The spectral distribution of transmittance and reflectance for as-prepared at different thickness TeSeSn films are shown in Figure 3(a) and Figure 3(b). It could be noted that as the film thickness increased, transmittance decreased which was an indication of increase in light absorption by the films. The increase of reflectance with film thickness at specific wavelengths was due to the effect of decreased transmittance [14] . The optical absorption coefficient (α) was computed from the experimentally measured values of transmittance T(λ) and reflectance R(λ) according to the following relation [15] ,

where d denotes the film thickness, R and T are the reflection and transmission coefficients, respectively.

Figure 4 shows the dependence of the absorption coefficient (α) on the incident photon energy (hν) for the as-prepared at different thickness TeSeSn films. It is observed that the values of the absorption coefficient increase with increasing both the photon energy and thickness.

According to Tauc [16] it is possible to separate three distinct regions in the absorption edge spectrum for amorphous semiconductor: the weak absorption tail which originates from defect and impurities, the exponential edge region which strongly related to the structural randomness of the system, and the high absorption region which determine the optical energy gap. In the high absorption region (α > 10⁴ cm⁻¹), the parabolic relation can be applied [10] [17] .

where B is a characteristic parameter (independent of photon energy) for respective transitions [18] , hν denotes photon energy, E_g is optical energy gap, and r is a number characterizing the transition process, having a value 1/2 for the direct allowed transition and value of 2 for the indirect transition. Figure 5 shows that the variation of (αhν)² versus (hν) for as-prepared at different thickness films is linear at the absorption edge, which confirmed

direct band gap transition in TeSeSn films. The values of optical energy gap, E_g, have been determined by extrapolating the linear portions of the respective curves to (αhν)² = 0. In the exponential edge region, the absorption coefficient is governed by the relation [19] ,

where E_e is the band tail width and hν is the photon energy, therefore, plotting the dependence of (lnα) versus (hν) should give a straight line. The inverse of the slope gives the band tail width (E_e) of the localized states at the band gap as shown in Figure 6. The optical energy gap E_g increases with increasing the thickness and the band tail width E_e decreases with increasing the film thickness as shown in Figure 7.

The increase of energy gap (E_g) and the decrease of band tail width (E_e) may be explained in terms of unsaturated bonds present in amorphous materials. It is known that unsaturated bonds are produced as a result of an insufficient number of atoms deposited on the amorphous films [20] . These bonds are responsible for the formation of some defects in thin films which in turn produce localized states in the band gap of amorphous solids. Thicker films are greater characterized by homogeneous network, which minimizes the number of defects and the localized states thereby increasing the optical gap and decreasing the band tail width.

On the other hand, the values of refractive index (n) and extinction coefficient (k) have been calculated using the following relations [10] [21] .

where α is absorption coefficient and R is reflectance, The spectral dependence of refractive index (n) and extinction coefficient (k) on the wavelength for as-prepared at different thickness TeSeSn thin films are shown in Figure 8 and Figure 9, respectively. Mostly, the value of n increased by increasing film thickness, also the value of k decreased. This decreasing indicates that the free carrier concentration changes. Hence, the disorder induced tailing decrease by increasing film thickness. This decrease is due to reducing the density of localized states near the mobility edges, and thereby reducing the extent tailing [22] .

In order to deduce the optical properties as a function of the annealing temperature, an analysis of the transmittance and reflectance spectra was done. The absorption coefficient was plotted as (αhν)² versus photon energy (hν) for TeSeSn films (100 nm thickness) at various annealing temperature ranges from (323 ≤ T_ann. ≤ 423), see Figure 10 as a representing example. Figure 11 demonstrates that the exponential behavior of the absorption edge, via Equation (7) is satisfied in our glassy films, the calculated values of optical energy gap (E_g) and localized states tail (E_e) for TeSeSn films under investigation at different annealing temperature are listed in Table 3.

It is observed from Table 3 that, the values of (E_g) decreases with increasing the annealing temperature. The existence of tin means that the states defects increase in the amorphous selenium-tellurium glass. In addition to that the annealing process increases the defect states inside the samples, which reduces the optical band gap energy [14] . Also the effect of annealing temperature on E_e for TeSeSn thin films are shown in Table 3 it’s observed that E_e increases with increasing the annealing temperature. Also, the decrease of optical energy gap and increase of E_e with the annealing temperature can be interpreted according to the Davis and Mott model [23] for amor-

phous materials, the width of the localized tail states near the mobility edges of the band gap depends on the degree of disorder and the density of defects present in the amorphous state. In particular, it is known that unsaturated bonds together with some saturated bonds, such as like dative bonds [24] , are produced as a result of an insufficient number of atoms deposited in the amorphous films [20] . Hesegawa et al. [25] showed that the unsaturated bonds are responsible for the formation of localized tail states in the band gap. The presence of a high concentration of these states is responsible for the decrease of E_g in the as-deposited films. Therefore, the drastic effect of crystalline phases on optical gap can be explained as a result of the production of surface dangling bonds around the crystallites [26] during the crystallization process. Further increase of annealing temperature results in the breaking up of the formed crystallites into smaller crystallites, thereby increasing the number of surface dangling bonds responsible for the formation of some types of defects. These defects lead to the decrease of the E_g. Wemple and Didomenico [27] [28] used a single oscillator description of the frequency-de- pendent dielectric constant to define the dispersion energy parameters E_d and E_o. The relation between the refractive index, n, and the single oscillator strength below the band gap is given by the expression [28] :

where E_o is the oscillator energy, E_d is the dispersion energy, which measures the average strength of the interband optical transition and E is the photon energy.

Figure 12 shows the relation between (n² − 1)⁻¹ and (E²) for as-prepared and after annealing. The values of E_o and E_d are directly determined from the slope, (E_o/E_d)⁻¹ and the intercept, (E_o/E_d). The obtained values of E_o and E_d for the typical TeSeSn thin films are listed in Table 3. It is observed that the values of E_o decrease with increasing the annealing temperature but the values of E_d increase with increasing the annealing temperature. The decreasing E_o and increasing E_d with the annealing temperature could be attributed to increase the rate of diffusion of atoms of the films with increasing the annealing temperature. The increase in the diffusion rate with increasing temperature gives more number of atoms at interstitial sites, thereby leading to impurity type scattering centers [29] [30] . The obtained data of refractive index n can be analyzed to obtain the high frequency dielectric constant (ε_L). The following equation can be used to obtain (ε_L) [31] :

where ε₁ is the real part of dielectric constant, ε_L is the lattice dielectric constant or (the high frequency dielectric constant), λ is the wavelength, N is the free charge carrier concentration, ε_o is the permittivity of free space (8.854 × 10⁻¹² F/m), m^* is the effective mass of the charge carrier and c is the velocity of light. The real part of dielectric constants ε₁ = n² was calculated at different values of λ. Then, the obtained values of ε₁ are plotted as a function of λ² as shown in Figure 13. It is observed that the dependence of ε₁ on λ² is linear at longer wavelengths.

The high frequency dielectric constant ε_L and the ratio N/m^* of the as-deposited and annealed films can be determined. The values of these two parameters with annealing temperature are given in Table 3. It can be seen that both the ε_L and the ratio N/m^* increases with increasing the annealing temperature.