Effect of La on optical and structural properties of CdS–Se films

2.1 Sample preparation

The samples for various studies are prepared by vertically dipping the cleaned substrates of highly transparent glass plates (dimension = 75 × 24 × 2 mm) at RT for 18 hours. Prior to deposition, the substrates are cleaned with double distilled water, acetone, and ultrasonic cleaner. A 25.0 ml beaker is used as a container for reaction of chemicals. The aqueous solution taken in the beaker contains highly pure and analytical reagent grade 1 M cadmium acetate [Cd (CH₃COO)₂], appropriate ratio of thiourea [SC (NH₂)₂] to sodium selenosulphate [Na₂SeSO₃] solutions [prepared by heating elemental selenium (99.9% pure) in aqueous solution of sodium sulphite [Na₂SO₃] at 90 °C for 5 h], triethanolamine (TEA) (HOCH₂CH₂)₃N and 30% aqueous ammonium hydroxide (NH₄OH). For preparing doped films, calculated proportions of 0.01 M solutions of cadmium chloride (CdCl₂) and lanthanum oxide (La₂O₃) are also added to the original mixture. The beaker containing aqueous solution of the chemicals is well covered with another inverted bigger beaker to prevent possible ammonia loss to some extent.

In the beginning solution is stirred for few minutes and no further stirring is done during the deposition. The deposition is made in the static condition by placing glass substrates inclined vertically to the walls of the beaker. Thereafter, the substrate is removed from the beaker and treated with distilled water to wash out the uneven overgrowth of grains at the surface and dried by keeping in open atmosphere under sun light until it completely dries. This helps in achieving longer operating life of PC and PL devices. The dried film is quite adherent to the substrate surface and is irremovable. Bhushan and Chandra (2008) reported that chemically deposited films have operating life of more than 2 years. In the above, the role of TEA and ammonia solution is to adjust pH of the reaction mixture and to increase film adherence. The thickness of the film is measured by interferometry method and is found to lie between 0.76 and 0.85 μm.

The mechanism of deposition of CdS–Se films is similar to that of CdS, which is based on the slow release of Cd⁺⁺ and S⁻⁻ ions in aqueous basic bath and subsequent condensation of these ions on the substrates suitably mounted in the bath. The slow release of Cd⁺⁺ ions is achieved by dissociation of a complex species of cadmium Cd (TEA)⁺⁺. The availability of Cd⁺⁺ ions is governed by the following dissociation equilibrium:

2.2 Measuring instruments

The PL excitation is done with 365 nm line of mercury, obtained from a high pressure mercury vapour lamp filtered by Carl Zeiss interference filter. The light output is detected by an RCA-6217 photomultiplier tube operated by a highly regulated power supply. The integrated light output in the form of current is recorded by a sensitive polyflex galvanometer. The spectral studies are performed using a prism monochromator. The optical absorption spectra are recorded using a 1700 Pharmespec Shimadzu Spectrophotometer over the wavelength range 350–700 nm. The XRD and SEM studies are performed at Inter University Consortium, Indore, India using Rigaku X-ray diffractometer and JEOL JSM-5600 scanning electron microscope, respectively.

3.1 Structural analysis

3.1.1

3.1.1 XRD studies

The samples are characterized by X-ray diffraction. The XRD diffractograms of Cd (S_0.7–Se_0.3) and Cd (S_0.7–Se_0.3):CdCl₂, La films are presented in Fig. 1 and corresponding data are listed in Table 1. The assignment of diffraction lines is made by comparing with American Standards for Testing Materials (ASTM) data and also by determination of parameters like lattice interval, lattice constant and Miller indices.

Close

Figure 1

(a) X-ray diffractograms of Cd (S_0.7–Se_0.3). (b) X-ray diffractograms of Cd (S_0.7–Se_0.3):CdCl₂, La films.

Export to PPT

Table 1 XRD diffraction data of Cd (S_0.7–Se_0.3) films (preparation time = 16 h; preparation temperature ∼28 °C).

Inter planner spacing (d values)		Intensity		Miller indices (hkl)	Lattice constant
Obs. values (Å)	Rep. values (Å)	Obs. values (Å)	Rep. values (Å)		Obs. values (Å)	Rep. values (Å)
(a) Cd (S0.7–Se0.3) films
3.77	3.72	100	100	(1 0 0)h – CdSe	a = 4.35	4.30
3.37	3.36	100	100	(1 1 1)c – CdS	a = 5.837	5.818
2.96	2.90	99	40	(2 0 0)c – CdS	a = 5.920	5.818
2.35	2.45	44	25	(1 0 2)h – CdS	a = 4.137	4.14
					c = 6.31	6.71
2.07	2.06	50	80	(1 1 0)h – CdS	a = 4.137	4.135
1.839	1.833	44.3	50	(1 1 2)h – CdSe	a = 4.35	4.30
					c = 7.25	7.02
1.672	1.68	22	10	(2 2 2)c – CdS	a = 5.791	5.818
1.58	1.581	22	7	(2 0 2)h – CdS	a = 4.137	4.135
					c = 6.82	6.71
(b) Cd (S0.7–Se0.3):CdCl2 (2 ml), La (8 ml) films
4.06	3.72	40	100	(1 0 0)h – CdSe	a = 4.67	4.30
3.32	3.36	71	100	(1 1 1)c – CdS	a = 5.74	5.818
3.0	3.16	77	100	(1 0 1)h – CdS	a = 4.1	4.135
					c = 6.81	6.71
2.83	2.90	100	40	(2 0 0)c – CdS	a = 5.66	5.818
2.47	2.45	24	25	(1 0 2)h – CdS	a = 4.1	4.135
					c = 6.79	6.71
2.05	2.06	33	57	(1 1 0)h – CdS	a = 4.1	4.135
						6.71
2.16	2.15	30	82	(1 1 0)h – CdSe	a = 4.32	4.30
1.76	1.753	51	60	(3 1 1)c – CdS	a = 5.837	5.818
1.66	1.68	25	10	(2 2 2)c – CdS	a = 5.749	5.818

XRD patterns show diffraction peaks corresponding to CdS and CdSe both. The peak intensities change due to incorporation of La as impurity. CdS is found in both cubic and hexagonal phases, which arise due to difference in arrangement of atomic layers. The possible reason for this rising of cubic and hexagonal phases is atomic arrangements as ABC ABC … and AB AB … (Kittle, 1995). Mixed phases are also found as a result of random stacking of very long period (Lopez et al., 2009).

Langer et al. (1966) reported that a solid solution is a mixture of microcrystalline regions of pure CdSe and CdS, where each micro region is composed of a number of unit cells of each material with the lattice-constant of CdS stressed by surrounding CdSe and vice-versa. Such a model explains uniform shift of absorption edge with composition. These researchers found possibility of solid solution consisting of statistical distribution of CdSe and CdS with respect to their overall concentration. Present investigation witnesses the shift of absorption edge. The average particle size is evaluated using famous Debye Scherer’s Formula (Ubale et al., 2007):

(3)

$$D=0.94\lambda /{\beta }_{1/2}\cos \theta$$ where β_1/2 is the full width at half maximum (FWHM) for the peaks expressed in radians, λ – the wavelength of X-rays and θ – the Bragg’s angle. The average particle size evaluated for different samples lie in the range 5.125–7.123 nm. The shift in spectrum reported earlier is due to nano crystalline effect in addition to other effects mentioned earlier. The values of related parameters like strain ε (Senthilkumar et al., 2005) and dislocation density δ (Singh and Bhushan, 2009) obtained for cubic (1 1 1) peak of CdS are listed in Table 2.

Table 2 Values of particle size, strain and dislocation density for Cd (S–Se):CdCl₂, La films at RT (∼28 °C).

System	Particle size D (nm)	Dislocation density (×1015 lin m−2)	Strain ε (lin−2 m−4)
(Cd (S0.7–Se0.3):CdCl2 (3.00 ml), La (4.00 ml)	4.28	54.61	0.00843
Cd (S0.7–Se0.3):CdCl2 (2.00 ml), La (8.00 ml)	3.39	87.03	0.0106

3.1.2

3.1.2 SEM studies

The SEM micrographs of Cd (S_0.7–Se_0.3) and Cd (S_0.7–Se_0.3):CdCl₂, La films at a magnification of 5 k are presented in Fig. 2. In both the micrographs, large numbers of clusters comprising of small grains along with some leafy structures are seen. There is some vacant space in fig. (a) between the clusters of non-homogenous spherical grains while fig. (b) shows layered growth along with few fibers. Earlier workers (Bhushan and Shrivastava, 2007) found well developed cabbage structure, which they interpreted in terms of overlapping of different layers formed under continuous growth. Karanjai and Dasgupta (1988) found that CdS grains have greater tendency to coalesce with the increasing concentration of CdCl₂. CdCl₂ is also known to promote recrystallization of CdS grains (Mukherjee and Bhushan, 2003a,b). The average grain size is calculated using Heyne’s intercept method (Subba Rao et al., 1972) in which instead of estimating the number of grains per unit area, the grains intercepted by a theoretical line on the specimen surface are counted. In the present case, the average particle size is found to lie between 8.3 and 10.7 nm.

Close

Figure 2

(a) SEM micrographs of Cd (S_0.7–Se_0.3) film. (b) SEM micrographs of Cd (S_0.7–Se_0.3):CdCl₂, La films.

Export to PPT

3.1.3

3.1.3 Optical absorption studies

The study of optical absorption is important in understanding the behavior of semiconductor nano crystals. A fundamental property of semiconductors is band-gap energy separation between the filled valence band and the empty conduction band. Optical excitations of electrons across the band-gap are strongly allowed, producing an abrupt increase in absorption at the wavelength corresponding to the band-gap energy.

Results of optical absorption spectra recorded in the wavelength range 340–660 nm and carried out at RT for pure and doped Cd (S–Se) films are presented in Fig. 3. From these curves it is noticed that absorption is dominant mainly in blue region. As the impurities are added to pure Cd (S_0.7–Se_0.3) films, absorption peak is blue shifted. This shift is due to the quantization effect according to which the band-gap value increases with the size reduction of crystallites. The band-gap values of these materials determined from the Tauc’s plots (Fig. 4) are found as Cd (S_0.7–Se_0.3) = 2.16 eV; Cd (S_0.7–Se_0.3):CdCl₂ (2.00 ml) = 2.15 eV; Cd (S_0.7–Se_0.3):CdCl₂ (2.00 ml), La (8.00 ml) = 2.41 eV. The materials of present study are of direct band-gap nature and a slight increase in band-gap value is noticed in La doped sample, which is due to its different ionic radii as compared to that of Cd where it is substituted.

Close

Figure 3

Absorption spectra of different Cd (S_0.7–Se_0.3) films: Cd (S_0.7–Se_0.3): Cd (S_0.7–Se_0.3). CdCl₂ (2 ml), Cd (S_0.7–Se _0.3) CdCl₂ (2 ml), La (8 ml).

Export to PPT

Close

Figure 4

Tauc’s plots of different Cd (S_0.7–Se_0.3) films: Cd (S_0.7–Se_0.3): Cd (S_0.7–Se_0.3). CdCl₂ (2 ml), Cd (S_0.7–Se_0.3) CdCl₂ (2 ml), La (8 ml).

Export to PPT

The absorption coefficient (α) and the band-gap (E_g) are related by the following expression in direct band-gap materials (Dawar et al., 1990):

(4)

$$\alpha =C{(\text{h}\nu -E_{\text{g}})}^{1/2}/\text{h}\nu$$ where ‘E_g’ is the optical band-gap and ‘C’, the speed of light.

3.2 PL spectral studies

The PL spectra of different Cd (S_0.7–Se_0.3) films at varied concentrations of CdCl₂ and La are presented in Fig. 5. PL emission spectrum of CdS peaks at 516 nm (∼2.39 eV), which is quite close to the band-gap (2.43 eV) of pure CdS as observed from the absorption spectrum. Thus, the emission is attributed to edge emission of CdS. Thomas and Hopfield, (1962) correlated this emission to the excitonic transitions involving free excitons. Maximum PL emission is observed for 0.7:0.3 combinations of CdS: CdSe and hence this combination is used for PL studies. In presence of CdCl₂, the highest emission is obtained at a volume of 2.00 ml. Therefore, this concentration is used in the presence of impurities.

Close

Figure 5

PL emission spectra of different Cd (S_0.7–Se_0.3) films: (1) Cd (S_0.7–Se_0.3), (2) Cd (S_0.7–Se_0.3): CdCl₂ (2 ml), (3) Cd (S_0.7–Se_0.3): CdCl₂ (2 ml), La (4 ml), (4) Cd (S_0.7–Se_0.3): CdCl₂ (2 ml), La (6 ml), (5) Cd (S_0.7–Se_0.3): CdCl₂ (2 ml), La (8 ml), (6) Cd (S_0.7–Se_0.3): CdCl₂ (2 ml), La (10 ml), (7) Cd (S_0.7–Se_0.3): CdCl₂ (2 ml), La (12 ml).

Export to PPT

The peak positions of different films are listed in Table 3. Cd (S–Se) films with ratio of 0.7:0.3 show two broad emission peaks. The lower wavelength emission is associated to a annihilation of free excitons because of similar excitonic nature of edge emission of CdS and CdSe (Bhushan and Oudhia, 2009). The broadening of this peak is attributed to the localized excitonic state with a broad distribution of energies (Shevel et al., 1987) along with exciton-lattice coupling and scattering. The second peak is found to shift towards higher wavelength with increasing percentage of Se. This emission is associated to donor–accepter transitions formed by incorporations of cations (excess Cd in present case) introducing deep accepter levels (like S vacancies with ionization energy ∼0.03 eV) and that of anions (S or Se in present case) introducing deep accepter levels (Cd vacancies with ionization energy typically ∼1.1 eV for sulphides and ∼0.6 eV for selenides). Bhushan and Pillai (2008), through EDX measurements have already reported that in such preparation excess CdS exists as compared to its counter part. With increasing concentration of La, these peak positions get shifted towards longer wavelength side. This supports that these emissions are not due to direct transitions in energy bands of La.

However, the origin of emission peaks, in presence of La is related to transitions between the excitonic levels of pure Cd (S–Se) and the energy levels due to La, which cause slight shift in emission peaks. Further improvement in peak intensities is also observed in presence of La, which is due to energy transfer from energy levels of La. Improvement in intensities of emission peaks in presence of CdCl₂ is due to better crystallization in its presence (Mukherjee and Bhushan, 2003a, b).