Electrodeposition of CdS thin-films from cadmium acetate and ammonium thiosulphate precursors

Cyclic voltammograms were separately recorded for electro-purified aqueous electrolytes containing 0.3 M Cd(CH₃COO)₂·2H₂O, 0.06 M (NH₄)₂S₂O₃ and a mixture of 0.3 M Cd(CH₃COO)₂·2H₂O and 0.06M (NH₄)₂S₂O₃. The pH and temperature of the solutions were adjusted to 3.00 ± 0.02 and ~ 85 °C respectively. The cyclic sweep is recorded with a sweep rate of 3 mVs⁻¹ within a cathodic potential range of 0.0 mV to 2000 mV. Figure 2 depicts a typical voltammogram data plotted with instructive symbols and legends.

Figure 2a depicts, due to the difference of redox potentials amongst the consisting ions, in the forward cycle of voltammogram sulphur (S) having a redox potential (E°) of + 0.449 V (with reference to standard H₂ electrode), starts depositing at the beginning at ~ 500 mV. The deposition takes place from the S precursor, according to the electro-chemical reaction Eqs. 1 and 2:

Later, Cd having an E° of − 0.403 V (w.r.t. standard H₂ electrode) starts depositing at ~ 850 mV (Fig. 2b and point A in Fig. 2c) according to the electro-chemical reaction Eq. 3:

Hence, Fig. 2c shows the co-deposition of cadmium and sulphur begins hereafter and S-rich CdS starts to deposit from ~ 850 mV onward according to the reaction Eq. 4. As higher cathodic voltage is applied, Cd starts increasingly incorporating to the layer.

Therefore, it is estimated that near-stoichiometric CdS layer can be found within a range of ~ 1300 mV to ~ 1500 mV of growth potential. From 1500 mV onward, there is a relatively higher increment in current density, which indicates the possible dissociation of water.

In the revers cycle of Fig. 2c, the negative peak at ~ 850 mV reconfirms the dissolving of any elemental Cd as well as dissociation of CdS formed on the FTO surface.

Aim of this analysis is to carry out the structural analysis of the deposited layers at different growth voltages. Philips PW X’Pert Pro diffractometer has been used with Cu-Kα monochromator having a wavelength of 1.54 Å, where the source tension and current were kept as 40 kV and 40 mA respectively. XRD spectroscopy have been carried out on 5 as-deposited (AD) and heat-treated (HT) samples grown at 1420, 1430, 1440, 1450 and 1460 mV cathodic potential. In Fig. 3a, at 2Ɵ ≈ 26.6° and ≈ 30.1° the as-deposited samples show XRD peaks corresponding to the diffraction from CdS (111) cubic/(002) hexagonal and CdS (200) cubic planes respectively. It is difficult to differentiate between CdS (111) C and (002) H plane because they coincide within 1%, therefore the XRD peak is often considered as resultant from mixed structures [23, 24]. Also, increment in the dominant peak intensity for AD CdS layers indicates the coinciding of CdS (111)C/(002)H peak with the underlying FTO (110) peak. After heat-treatment it is widely observed that CdS transforms phase from cubic/mixed to completely hexagonal CdS [20, 22]. In this work, the effect of annealing demonstrates a different trend of changes, indexing with the JCPDS reference files 89-0440 for cubic phase and 80-0006 for hexagonal phase. Figure 3b shows, after the heat treatment of the layers at 400 °C for 20 min, whereas CdS (111) C/(002) H mixed phases still stay the preferred orientation at 2Ɵ ≈ 26.6°, the slight (200) C peak disappears and CdS (110) H/(220) C mixed phase come into appearance at 2Ɵ ≈ 43.9°. Though, this does not indicate a confirmed phase transition of the material from mixed to hexagonal, but shows a tendency towards achieving the hexagonal phase which is further supported by later experiments. For the dominant planes resulting from different growth voltages, the trend shows that peak intensity keeps gradually increasing upto 1450 mV and then goes down. Figure 4 depicts this trend for both the as-deposited and heat-treated CdS samples. Along with the peak intensity, change of the CdS crystallite size due to growth voltage is shown in Table 1 for both the AD and HT samples. Crystallite size at the preferred orientation has been calculated using Scherrer’s formula given by Eq. (5).

where wavelength, λ = 1.54 Å, β = full width at half maximum (FWHM) of the diffraction peak in radian and θ = Bragg’s angle of incidence.

Table 1 shows that the as-deposited samples grown at 1450 mV has the largest crystallite size among the others. For HT samples, the trend carries on and all the HT samples show larger crystallite size than their as-deposited counterparts.

However, all CdS layers grown here on the TEC-7 glass/FTO substrates show overlaps of the dominant CdS plane with the dominant FTO (110) plane at 2Ɵ ≈ 26.6°. Since, the first batch of FTO substrates shown the preferred FTO orientation at 2Ɵ ≈ 26.6°, for reconfirmation, another CdS layer has been deposited at the best apparent growth voltage (~ 1450 mV) on another batch of substrate having a different preferred orientation. Figure 5 shows that, CdS layers grown on a substrate having a preferred orientation elsewhere, also depicts a sharp peak intensity at 2Ɵ ≈ 26.6° for both as-deposited and heat-treated samples. The AD layer demonstrates cubic/hexagonal mixed phases at preferred orientation and (200) C peak at 2Ɵ ≈ 30.1°. The HT sample still demonstrates mixed phases at preferred orientation, but (200) C peak disappears and another mixed peak appears at 2Ɵ ≈ 43.9° like the previous batch shown in Figs. 3 and 4. Alongside, there appears a slight peak at 2Ɵ ≈ 25.0° for CdS (100) hexagonal phase. This further indicates a tendency for the layers towards achieve hexagonal phase upon heat-treatment, though having possibilities to have mixed phases present in the structure.

EDX was carried out on the as-deposited and heat-treated CdS samples grown at 1430, 1440 and 1450 mV, using the compositional analysis software associated with the Quanta 650 SEM equipment, in order to obtain the chemical composition of the samples. The heat-treated samples were annealed at 400 °C for 20 min in air. Figure 6 shows the EDX spectra of the as-deposited and heat-treated sample grown at 1450 mV. Table 2 presents the atomic composition of as-dep and HT CdS grown at different voltages along with their changes in Cd/S ratio. It depicts that all the as-dep layers are S-rich in nature regardless the growth potential, whereas after heat treatment all of them shows slight comparative decrement in S percentage which resulted the HT sample grown at 1450 mV reaching stoichiometry. Despite having concerns about accuracy, EDX technique and this analysis give a qualitative idea about the composition of the CdS films and the near-stoichiometric growth potential for the process. This achievement of stoichiometry confirms the structural analysis using XRD and the achievement of highest crystallinity at 1450 mV. Any thin-film layer with only one phase tend to crystallise better during this growth process.

Photoelectrochemical cell measurements were carried out to determine the conduction type of the electrodeposited layers. The measurement is taken by dipping a glass/FTO/CdS layer along with another carbon electrode into an electrolyte prepared from 0.1 M Na₂S₂O₃ (sodium thiosulphate) aqueous solution. The voltage difference between the electrodes under dark (V_D) and illuminated (V_L) condition is measured to determine the conduction type of the semiconductor layer. This difference (V_L − V_D) represents the open circuit voltage of the solid/liquid junction and is commonly referred as the PEC signal. Polarity of the PEC signal indicates the conduction type. In this work, negative polarity of the PEC signal directs to n-type semiconductor and positive polarity directs to p-type semiconductor, whereas for metals, insulators and intrinsic semiconductors PEC signal is ideally zero. Figure 7 shows that all the CdS layers deposited at different cathodic potentials are n-type in conductivity as expected [19, 25]. HT samples show a stronger PEC signal compared to that of the AD samples. Strength of the PEC signal qualitatively indicates to the doping density and depletion region strength.

Optical absorption study has been carried out using Carry 50 scan UV–Vis spectrophotometer, for AD and HT CdS layers deposited at different growth voltages. The obtained absorption data within the wavelength of 300 nm to 1000 nm, have been used to plot an alternative Tauc plot, where square of absorbance (a²) is plotted against photon energy (hv) [15, 26]. The intersection of extrapolated tangent line derived from the curve, gives the direct band-gap (E_g) of the studied semiconductor material.

According to Figs. 8a and 9, as-deposited CdS layers show direct band-gaps ranging between ~ 2.36 and ~ 2.40 eV, where the lowest band-gap value is observed at ~ 1450 mV of growth potential. Heat-treated CdS layers depict a similar trend as per Figs. 8b and 9, but after heat-treatment, band-gap decreases overall compared to that of the AD leyers. Band-gaps of the HT leyers range between ~ 2.31 and ~ 2.36 eV. The lowest band-gap value (~ 2.31 eV) for the heat-treated samples is found for the layer grown at 1450 mV which differs slightly from the previously measuered E_g values (~ 2.42 eV) for electrodeposited CdS layers in the group [20, 22, 27]. This observation is supported by the visual appearance of the deposited layers as shown in Fig. 10. The CdS layers electrodeposited from these precursors demonstrate an orangish yellow appearance compared to the greeninsh yellow appearance of the previously electrodeposited CdS layers in our group [15, 20]. This indicates a possible shift in the typical absorption point of the layer from lower wavelength towards higher wavelength, and vice-versa for the band-gap energy.

In order to carry out the morphological characterisation of the CdS layers, SEM analysis has been done in the form of micrographs with Quanta 650 nano SEM instrument using 15.0 kV electron beam voltage and 32,000× of magnification. SEM images of glass/FTO/CdS substrates have been taken for both as-deposited and heat-treated samples grown at 1440 mV (Fig. 11a, b) and 1450 mV (Fig. 11c, d). In Fig. 11, at both growth voltages, regardless of heat-treatment, all the surfaces demonstrate agglomerations of CdS crystallites. From the SEM micrograph the size of the agglomerations are ranging approximately between ~ 200 and ~ 300 nm. Therefore, individual CdS crystallites are not visible in the SEM image as calculated above in Table 1.