CdSe thin films can be prepared by using several precursors of Cd and Se. In this work we used cadmium acetate dehydrate [Cd(CH₃COO)₂.2H₂O] (99% purity), selenium powder (Se) (99.9% purity), Triethanolamine (TEA) [N(CH₂CH₂OH)₃] (98% purity, sodium sulphite (Na₂SO₃) (98% purity) and 25% ammonium hydroxide (NH₄OH) as the precursors. Substrate cleaning is vital in the fabrication of thin films to enhance the adhesion of the films. The glass slide substrates were first washed with detergent and water, kept in hydrochloric acid (a suitable amount of HCL is taken in which commercially available glass slides were immersed vertically) for 2 h, washed in deionized water and then rinsed in acetone. Ultrasonic cleaning in deionized water for 15 min was subsequently carried out as the last step in our cleaning process. Well-cleaned glass substrates were placed in a hot oven for a few minutes to dry.

Prior to the deposition of CdSe thin films, aqueous solution of the various precursors was prepared. Cadmium acetate dihydrate [Cd(CH₃COO)₂.2H₂O] serves as the source of Cd ions while freshly prepared sodium selenosulfate (Na₂S_eSO₃) was used as the source of Se ions. For preparing its solution, 1.89 g of sodium sulphite (Na₂SO₃) was dissolved in 15 ml of water. Then 1.84 g of selenium (Se) was added to this solution. The solution was kept at the temperature of 85°C for 5 h with constant magnetic stirring. A fresh solution of sodium selenosulfate is required because the salt is unstable if allowed to stand for a long time in solution. Figure 1a shows the stages involved in the growth of CdSe thin films in which 0.5 mol.L⁻¹ of cadmium acetate dihydrate was first dissolved in a beaker with 10 mL of water. The solution was subsequently placed in a magnetic stirrer set at a low stirring speed. Five ml of triethanolamine (TEA) which serves as the complexing agent was introduced to the solution dropwise while stirring. To stabilize the pH of the solution, 10 ml of 25% aqueous ammonia was added into the beaker, followed by thorough mixing with the magnetic stirrer. Next, 15 ml of freshly prepared sodium selenosulfate solution was added, dropwise, with the pH of the solution kept constant at 10 ± 0.5.

A homemade substrate catcher was used to place the substrate centrally in the solution as shown in Figure 1a. Thin films of CdSe was deposited on the glass substrate for 6 h at fixed temperature, which was maintained by a water bath apparatus. This procedure was repeated to grow other samples at different temperatures of 40, 50, 60, 70, and 80^oC, respectively. A new precursor solution was used for each tested temperature. A control sample at room temperature was also deposited as shown in Figure 1b. After the deposition, the slide was washed with a fountain of deionized water to eliminate residues before drying them in air. It is observed that films of CdSe which are prepared at room temperature were colorless but as the bath temperature increased colorless CdSe thin film changed gradually to orange and then to dark red as shown in Figure 1b.

To obtain thin films with good crystallinity, the samples were annealed in a tube furnace at temperatures of 100, 200, and 400°C for 2 h. It was observed that after annealing, the color of the CdSe films turned to dark black consistently with previous reports (Ali et al., 2013). The set of samples that were annealed at 400°C are depicted in Figure 1c. Typical SEM image of annealed samples shown in Figure 1d indicates that the thin films comprised of well-nucleated nanoparticles.

The structural and optical characteristics of these films were investigated as a function of the annealing temperatures. Structural and phase analysis of CdSe thin films were done by using Rigaku Geiger flux instrument with Cu K_α radiations having wavelength 1.54051 Å. The XRD patterns of these films were obtained in angular range (2θ) from 10° to 80° with a step size of 0.05°. FTIR transmissions spectra of CdSe thin films synthesized at different bath temperatures and annealed at 400°C were obtained with Perkin-Elmer FTIR Spectrometer. The spectra were taken at room temperature in the range of 400–4,000 cm⁻¹. The Surface morphology of the thin films was examined with a scanning electron microscope (Jeol: JSM-6510LV), which was operated at 25 kV. The optical properties were investigated with Perkin Elmer Lambda 19 Spectrometer operated at the wavelength range from 300 to 800 nm.

The chemical composition of CdSe thin films and interface structure for interdiffusion between CdSe films and glass substrate were studied using Rutherford Backscattering Spectroscopy (RBS). The RBS was carried out by utilizing a 5 MeV pelletron tandem accelerator. The accelerator generated a mono energetic He⁺⁺ ions beam, which has energy 2.084 MeV with a beam current of 28 nA. The RBS data simulation was carried out by using ion beam analysis software (SIMNRA).

The deposition of CdSe thin film using CBD technique involves controlled precipitation of the ions in solution and subsequent nucleation of the resulting species on the substrates such as the microscopic glass slides used in this study. At optimum reaction temperature, the precipitation of CdSe occurs when the ionic product of reactants becomes greater than their solubility product. However, if the conditions for precipitation are not fully satisfied the resulting solid phase dissolves in the solution and ultimately resulting in a depleted or no deposition (Ezugwu et al., 2009; Deshpande et al., 2013a; Onyia, 2017).

The pH of the solution, the deposition temperatures and the complexing agent concentration are the key parameters that control the concentration of the free metal ions in CBD setups. By controlling the release of Cd²⁺ and Se²⁻ ions, one can obtained uniform thin films at optimum deposition conditions. In this work, triethanolamine (TEA) was used as the complexing agent to regulate the hydrolysis of cadmium metal ion in the solution. TEA was added dropwise to the cadmium salt to initiate a slow reaction in the solution and subsequent formation of complex Cd salt as shown below:

Ammonia was dissolved in water to provide OH⁻ ions which forms part of the reacting species:

The overall chemical reaction for the deposition of thin films of CdSe can be described by the following chemical equation.

To enhance the deposition rate and tune the properties of the resulting thin films of CdSe, the deposition was carried out at different temperatures by placing the solution bath on a regulated water bath. Six samples of films were fabricated on glass substrates at different temperatures of 40, 50, 60, 70, 80°C and a control sample at room temperature.

CdSe has two structural polymorphs, the crystal structure appearing as Wurtzite or cubic, zinc blende-type (Xia et al., 2010). A previous report (Kale and Lu, 2015) show that the thin films of CdSe deposited on a glass substrate by CBD is either cubic or hexagonal. In our specific case, the XRD pattern of the films deposited at different bath temperatures shown in Figure 2 revealed that the samples are mostly amorphous in nature. However, as the deposition temperature increases, a small-broadened peak starts to appear. The broad XRD peak around 25.5°C is clearly observed for the thin films synthesized at 80°C. It appears that the best bath temperatures for the thin film deposition are 70 and 80°C. Hence, the sample deposited at 80°C will be investigated further. It was observed that the thermal treatment of the samples helps to induce structural modification from amorphous to crystalline, as shown in Figure 3. It is clear from these results that the CdSe films deposited at 80°C and annealed at 100, 200, and 400°C are polycrystalline in nature and have a cubic structure with intense (111) reflection peaks. Figure 3 also shows that as annealing temperature increase from 100 to 400°C the crystallinity also increases. However, the sample which was annealed at 450°C showed an increased broadening of the (111) diffraction peak. This is possibly due to a decrease in the crystallite size at this elevated annealing temperature (see Table 1). When CdSe thin films are annealed at a temperature of 450°C, chemical degradation can occur with a possible conversion from CdSe to the oxide, CdO (Mahato and Kar, 2017).

The crystalline size of the nanocrystallites was calculated by using Scherrer's formula given below.

where “D” is crystallite size, k is constant having value 0.9, λ is the wavelength of X-rays, β is full width at half maximum, and θ is Bragg's angle. Dislocation density (δ) and microstrain (ε) of CdSe thin films were also calculated with the following equations:

The values of the Lattice constant (a), d spacing, FWHM (β), grain size (D), dislocation density (δ), and microstrain (ε) calculated from the XRD results in Figure 3 are reported in Table 1. The results show that as the annealing temperatures of thin films of CdSe increases the full width at half maximum, FWHM (β) of the films decreases. Consequently, the grain size of the samples increases from 2.23 to 4.13 nm with annealing temperatures. However, the dislocation density (δ) and microstrain (ε) of the films decreases with the increase in the annealing temperature. The intensity of the peaks also increases by increasing the annealing temperature.

Fourier transform spectroscopy is used to investigate vibrational modes of different finctional groups. Figure 4 shows the FTIR pattern of thin films of CdSe deposited at bath temperatures of (a) 50°C, (b) 70°C, (c) 80°C, and (d) film synthesized at 80^oC and then annealed at 400°C. The broad peak observed at 3,460 cm⁻¹ is assigned to O-H stretching and is due to the presence of a small quantity of water molecules in CdSe thin films. The peak at 2,990 cm⁻¹ is assigned to C-H stretching of CH₂ group (Ravindranadh et al., 2013). The peak observed at 668 cm⁻¹ is the fingerprint spectral peak of CdSe (Rani, 2014). This band confirmed the occurrence of a vibrational mode of CdSe lying between 400 and 700 cm⁻¹ (Ali et al., 2013). Figure 4 also revealed that the peak which is due to O-H stretching shows decreasing %T intensities by increasing the deposition temperatures from 50 to 80°C and by annealing at 400°C.

Scanning electron microscope (SEM) which provides an appropriate way to study the surface morphology of thin films was used to examine the surface morphology of our samples. SEM gives information about grain size, roughness, and surface morphology of thin films. Figures 5a–c show SEM micrographs and surface morphology of thin films of CdSe synthesized at different bath temperatures of 50, 70, and 80°C. From the SEM micrograph of these films, it is evident that films adhered well to the surface of the substrate and consist of many small grains that are sparsely distributed. However, the density of the nanoparticles appears to increase with the increasing temperature of the deposition bath. Figure 5d shows that post-deposition thermal annealing has significant effects on the morphology of our CdSe thin films. A comparison of Figures 5a,d indicates that the morphology of the sample annealed at 400°C is dense and compact. This can be attributed to the increase in the activation energy needed for intergrain diffusion, leading to smaller grains joining to form large grains, which in turn increases the size of the nanoparticles as shown in Figure 5e.

Figure 5e shows that the particle size also depends on the deposition temperature, increasing from 230 nm at 50°C deposition temperature to as much as 680 nm at 80°C. We observe a significant increase of the size of nanoparticles for the samples deposited at 80°C when compared to the two samples deposited at lower bath temperatures of 50 and 70°C. This observation suggests that 80°C is the ideal bath temperature in which nucleation and growth of Cd and Se ions on the substrates can proceed favorably to form nanoparticles. Once the initial nuclei are attached on the substrate, further growth from the impinging ions in the condensation is expected to proceed via the Volmer-Weber growth pattern, leading to the observed increase in both the density and size of the nanoparticles (Apeh et al., 2019).

Ultraviole-tvisible (UV-vis) spectroscopy is an optical technique used to investigate optical aborption/transmission of the sample. Figure 6 represent transmittance spectra of CdSe thin films deposited at different bath temperatures and annealed at 400°C. The figure indicates that the transmittance of thin films of CdSe is generally low, approaching 10 % at maximum. The figure shows that the transmission spectra of CdSe thin films have relatively higher transmittance (>5%) above 500 nm (in the visible region), and near-infrared region, 700–800 nm. In UV region, 350–400 nm, the transmittance of CdSe thin films decrease to a very low value (<2%), indicative of very high absorbance in the UV region of electromagnetic radiation. The result of our UV-Vis measurements shows that the transmittance of thin films of CdSe depends on the deposition and annealing temperatures. From the graphs shown in Figure 6, the transmittance decreases as deposition temperature increases, consistently with our earlier observation that the density and compactness of the samples increase with bath temperatures. The decrease in the transmittance of CdSe samples with increasing bath temperature is also consistent with the observed color changes of the films in Figure 1, from light red to dark red, indicative of greater interdiffusion of isolated grains and particle growth increase with increasing bath temperature. The transmittance of the annealed sample is also very low as compared to the un-annealed samples. After annealing, the dark red color of the thin film of CdSe changes to dark black as shown previously in Figure 1. Besides the color changes, our SEM analysis shows that post-deposition thermal annealing increases the diffusion energy for the grains to form large particles, which leads to denser films (Jamil et al., 2012). The observed color changes from dark red to dark black, the particle growth and increased density are responsible for the huge decrease in the transmittance of our CdSe thin films.

Figure 7 represent the graphs of (αhν)² vs. hν for the thin film of CdSe deposited at different bath temperatures of 50, 70, and 80°C and later annealed at 400°C. The energy bandgap of the CdSe films was determined by extrapolating the straight-line part of the plot to zero absorption making an intercept with the energy axis. The interception on the energy axis (hν) gives the value of the energy bandgap. Figure 7 shows that the graphs of the bandgap energy are linear over a wide range of energy, which confirms that CdSe thin films deposited in this work, are direct bandgap semiconductor (Bernard et al., 2004). Values of the energy band gap of as-synthesized thin films of CdSe at different temperatures decrease from 2.20 to 2.12 eV as illustrated in Table 2. El-Menyawy and Azab reported similar values for CdSe thin films prepared by the thermal evaporation technique (El-Menyawy and Azab, 2018). In our case, the bandgap energy tends to decrease with high temperature thermal annealing. In fact, after annealing at 400°C values of the energy band gap of CdSe thin film decreased further to 1.73 eV in agreement with the bulk CdSe and consistently with the observed low value of the optical transmittance with thermal annealing. The decrease in the energy band gap can be attributed to the quantum confinement in CdSe thin films due to the increase in crystallite size and a decrease in defects in these films (Deshpande et al., 2013b). Visual inspection shows that the films exhibited color changes upon thermal annealing from dark red to gloomy black in conformity to earlier work reported in the literature (Daniel et al., 2017).

The average diameter of nanocrystallites can be calculated from the blue shift of band gap of CdSe, which was examined in the samples prepared at different bath temperatures, by using the following equation (Mehta et al., 2007).

where μ is the translation mass and is equal to (m_h+ m_e), E_shift is bandgap shift and R is the radius of the crystallites. By putting the observed values of band gaps of CdSe thin films from the band gap value of bulk CdSe and putting the standard values of other parameters in the above equation, the diameter of the crystallites of CdSe can be determined. The crystallite (grain) size of CdSe, which were deposited at a temperature of 50, 70, and 80°C are found to be 2.396, 2.488, and 2.626 nm, respectively. These values are close to the grain sizes determined from the XRD diffractogram reported previously in Table 1.

Values of the bandgap, wavelength, band energy shift, the crystallite size and some of the other optical parameters of CdSe thin films calculated from the optical spectra are shown in Table 2. We observe that the thickness of CdSe thin films increases with increasing deposition temperature. The grain size of the films also increases and consequently the energy band gap decreases as the deposition temperature increases. These results show that the increase in the crystallinity of the films agrees with the SEM micrographs reported previously in Figure 5. The bandgap energies of the samples deposited at different temperatures are situated in the visible light of the solar spectrum. As such, the thin films are suitable for optoelectronic applications, especially in a solar cell where they can be used to generate photocurrent from visible light absorption.

Rutherford Back Scattering Spectroscopy (RBS) is useful technique that is used to investigate the elemental profile of Cd and Se along the thickness of the prepared thin films. Figure 8 shows the RBS spectra of thin films of CdSe which were deposited at 70 and 80°C, respectively. Apart from the Si signal arising from the glass substrate, two other peaks of Cadmium (Cd) and selenium (Se) are emanating from the thin films. From the RBS graph, it is seen that resolution of observed peaks for Cd and Se is very poor which is due to interdiffusion between CdSe thin films and glass substrate of these few nm thick films. As a result, the composition analysis of our sample could not be calculated with the help of RBS spectra due to the poor resolution of the Cd and Se peaks. However, the film thickness was calculated by the model fitting of the RBS data using software SIMNRA and has been reported in Table 2.