• For Contributors +
• Journal Search +
Journal Search Engine
ISSN : 1225-0562(Print)
ISSN : 2287-7258(Online)
Korean Journal of Materials Research Vol.30 No.1 pp.1-7
DOI : https://doi.org/10.3740/MRSK.2020.30.1.1

# Semiconductor CdTe-Doped CdO Thin Films: Impact of Hydrogenation on the Optoelectronic Properties

Department of Physics, College of Science, University of Bahrain, P.O. Box 32038, Kingdom of Bahrain
Corresponding author E-Mail : adakhil@uob.edu.bh (A. A. Dakhel, Univ. of Bahrain)
August 29, 2019 August 29, 2019 October 2, 2019

## Abstract

Doping or incorporation with exotic elements are two manners to regulate the optoelectronic properties of transparent conducting (TCO) cadmium oxide (CdO). Nevertheless, the method of doping host CdO by CdTe semiconductor is of high importance. The structural, optical, and electrical properties of CdTe-doped CdO films are studied for the sake of promoting their conducting parameters (CPs), including their conductivity, carrier concentration, and carrier mobility, along with transparency in the NIR spectral region; these are then compared with the influence of doping the host CdO by pure Te ions. X-ray fluorescence (XRF), X-ray diffraction (XRD), optical absorption spectroscopy, and electrical measurements are used to characterise the deposited films prepared by thermal evaporation. Numerous results are presented and discussed in this work; among these results, the optical properties are studied through a merging of concurrent BGN (redshift) and BGW (blue shift) effects as a consequence of doping processes. The impact of hydrogenation on the characterisations of the prepared films is investigated; it has no qualitative effect on the crystalline structure. However, it is found that TCO-CPs are improved by the process of CdTe doping followed by hydrogenation. The utmost TCO-CP improvements are found with host CdO film including ~ 1 %Te, in which the resistivity decreases by ~ 750 %, carrier concentration increases by 355 %, and mobility increases by ~ 90% due to the increase of Ncarr. The improvement of TCO-CPs by hydrogenation is attributed to the creation of O-vacancies because of H2 molecule dissociation in the presence of Te ions. These results reflect the potential of using semiconductor CdTe -doped CdO thin films in TCO applications. Nevertheless, improvements of the host CdO CPs with CdTe dopant are of a lesser degree compared with the case of doping the host CdO with pure Te ions.

CdTe-doped CdO , CdO

## 초록

This is an Open-Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.

## 1. Introduction

Cadmium Oxide (CdO) is a wide-bandgap degenerate semiconductor having a wide range of optoelectronic applications especially in the field of detectors and production of solar cells. CdO has a cubic structure (Fm- 3 m, 0.4695 nm),3) a direct bandgap around ~ 2.1 eV, and a resistivity of 10−2 ~ 10−4 Wcm depending on the procedure of synthesis and preparation.1,2,4) It’s transparent conducting (TC) properties are generated by natural point defects, i.e. oxygen vacancies (VO) and Cd interstitials (Cdi). Therefore, its TC properties can be controlled by controlling that point defect through doping with exotic types of impurity ions.5-7)

On the other hand, cadmium telluride (CdTe) is a p-type semiconductor of direct bandgap 1.5 eV and cubic zincblende (F-43m) structure (0.648 nm). CdTe thin films were used in intensively in photovoltaic applications.8,9)

In the present work, self-doping of CdO crystal by CdTe semiconductor is investigated as an objective to obtain a single phase of CdO doped with Te ions (CdO:Te) and compare it with previous work on CdO doped with only Te ions. Metalloid Te is a chalcogenide element with oxidation state +4. The ionic radius of Te+4 is 0.097 nm, which is approximately similar to that of Cd ions in CdO [10]. Hence, Te+4 ions could replace Cd2+ ions in CdO lattice creating a substitutional solid solution (SSS) without forming a remarkable geometrical distortion to the CdO lattice unit cell (UC).

Nevertheless, using CdTe semiconductor for doping CdO with Te and Cd ions is an implement to enhance its electronic transport characteristics. It is worth mentioning that Te-doped CdO prepared by sol-gel method was carried out in ref. 11, while it was synthesized by vacuum evaporation in ref. 12. However, doping of Te ions into other TCO’s like ZnO,13-15) In2O3, 16) and SnO217) was investigated earlier. Moreover, annealing of CdTedoped CdO samples in a hydrogen atmosphere is also examined in the present work seeking further improvement to the optoelectronic parameters of CdO films.

## 2. Experimental Procedure

Thin films of cadmium oxide (CdO) merged with various amounts of CdTe (CdO:CdTe) were prepared by physical vapour deposition (PVD) technique to avoid the inclusion of impurities in the films. The initial materials, pure CdTe and CdO (supplied by Fluka A.G. Germany) were frequently evaporated by alumina baskets (Midwest tungsten service, USA) in a vacuum chamber of residual oxygen atmosphere of pressure ~ 1.3 × 10−2 Pa. Corning 2947 glass substrates were ultrasonically cleaned in distilled water, acetone, and alcohol while silicon wafer substrates were cleaned with acetone, deionized water, and KOH solution. The Si wafers were used for X-ray fluorescence (XRF) measurements. The evaporated masses were regulated by a piezoelectric microbalance crystal sensor (type Philips FTM5) fixed close to the substrates. The as-grown films were flash annealed in the air at 400 °C for 1 h keeping samples inside the closed furnace for natural cooling to room temperature. The thickness of films which determined after annealing by a MP100-M spectrometer (Mission Peak Optics Inc., USA), was in the range of 0.10-0.30 mm. The weight ratio of Te to Cd content in each film sample was evaluated by the X-ray fluorescence (XRF) method carried with an Amptek XR- 100CR (USA) X-ray detector of energy resolution 180 eV controlled by a built-in MCDWIN 3.1 program. The crystalline structure of the prepared film samples was investigated by a Rigaku Ultima-VI X-ray diffractometer using Cu Ka radiations. The Rietveld structural analyses were performed by a built-in PDXL program. The spectral optical transmittance, T(λ) and normal reflectance, R(λ) were measured in a spectral range of 300 ~ 1,500 nm with Shimadzu UV-3600 double beam spectrophotometer in double-beam regime to rule out the substrate absorption effect. The dc-electrical conduction parameters (CPs); conductivity (σ), carrier mobility (μel), and carrier concentration (Nel), were measured with a standard Van der Pauw method using ~ 1T magnetic field.

## 3. Structural Analysis

The elemental analysis of each film sample grown on a silicon substrate was carried out by using an XRF method. In Fig. 1, the XRF spectrum of a sample manifests Cd Lband (3.13 ~ 3.53 keV), Te L-band (3.758 ~ 3.769 keV and 4.029 ~ 4.301 keV for La and Lb, respectively), and Si substrate Kα-signal (1.74 keV), and consequently, it emphasizes the purity of the prepared sample. The ratio of the integrated intensity of Cd-L band to that of Te La band was used to calculate the fractional weight ratio Te/ Cd in each film sample. For the sake of that objective, a known method of micro-radiographic analysis was utilized,18) and a pure CdTe thin film grown on Si substrates was used as a reference sample. Fig. 1 shows XRF spectra of CdTe thin film (reference) and of sample S-4. The estimated Te/Cd mass ratios in the chosen investigated film samples were ~ 1 %, 4 %, and 7 % for the sample S-1, S-2, and S-3 respectively.

Fig. 2(a) demonstrates the XRD patterns of pristine CdTe film, undoped and CdTe-doped CdO films deposited on glass substrates. The XR patterns of all investigated samples reflect the formation of a single-phase of CdO structure. The XR peaks were indexed according to the standard cubic (Fm-3m) structure.3) The relative intensity of the XR peaks exhibits that the energetically preferred [111] orientation of undoped CdO films was strongly strengthened by CdTe incorporation, especially for sample S-1 where the ratio [I(200)/I(111)] was reduced by ~ 30 %, comparing to the undoped CdO film. The change in the level of preferred orientation growth of the investigated films confirmed the incorporation and sharing of Te ions in the doping process of CdO. The reinforcement of [111] orientation for host CdO films was also seen in the case of other metallic dopants like Fe, Cr, Pt.19-21) As the development of [111] orientation of a film is considered to be a level of enhancement of its crystallinity, therefore, the carrier mobility (μ) is expected to be increased with the present doping.

Fig. 2(a) reveals that the present doping did not amendment the ordinary cubic crystalline structure of host CdO. Moreover, the XRD patterns did not show peaks arising from pure oxide or any Te related phases, which emphasizes the total doping/incorporation of Te ions into the CdO lattice, without the creation of separated Te phases or oxides within the host CdO lattice.

Table 1 presented the structural analyses of patterns of Fig. 2(a). The lattice parameter (a) is estimated by a Rietveld refinement method getting acceptable values for R-parameters (Rwp, the weighted profile factor and S, the goodness-of-fit parameter). For good fitting, the S parameter value should be between 1 and 2 (or close to 1) that indicating a good single-phase fitting.22) Fig. 2(b) demonstrates graphically the Rietveld refinements results for the as-synthesized S-3 sample, as an example of excellent single-phase fitting (S = 1.14),22) where; the blue solid line (up) is the experimental data, the red solid line (up) is the calculated pattern, and the solid pink line (down) is the intensity difference.

The calculated lattice parameter for un-doped and doped CdO powders are very close to the standard JCPDS value.3) The unit-cell volume (Vcell) of the host CdO has slightly decreased with increasing of Te % content, as shown in Table 1 and Fig. 3, attaining lowest value at around ~ 4 % doping level. Moreover, the results show the successfully merged of Te4+ ions into the crystalline lattice of CdO, thus it is possible to assume integration of Te4+ ions within CdO crystal when being replaced for Cd2+ ions forming a substitutional solid solution (SSS), since the radius of Te4+ ions is almost equal to that of Cd2+ ions. Nevertheless, the exchange of Cd2+ ions by Te4+ ions causes the disturbing of static charge balance of the host CdO unit cell. Therefore, the host crystal would be rebalanced by the creation of Ovacancies (VO) in addition to create slight shifts in ionic positions in the unit cell. That’s all together leading to the reduction of the unit-cell volume (Vcell) of CdO as shown in Fig. 3. For higher doping levels, some of added Te ions accumulate on CBs rather than real enter into host CdO unit cell.

The structural analysis including CS and internal strain were determined by using the Williamson-Hall (W-H) method, which is a part of the built-in software of the used XRD apparatus. The used W-H equation is23):

$( β h k l / tan θ h k l ) 2 = ( k / D ) ( β h k l / tan θ h k l sin θ h k l ) + ( ε / λ ) 2$
(1)

where θ is the Bragg angle, k’ ~ 0.75, D is the crystallite size (CS), βhkl is the peak width at half-maximum, ε is the strain, and l is the used X-ray wavelength. The plot of $Y = ( β h k l / tan θ h k l ) 2$ versus $X = ( β h k l / tan θ h k l / sin θ h k l )$ yields straight line, thus CS and e can be calculated for sample S-3, as shown in Fig. 2(c). The nano-CS, given in Table 1 decreased to ~ 20 nm for the highest doping level referring to the accumulation of CBs as a result of the limited solid solubility of Te ions in CdO lattice.

Fig. 4 presents the XRD patterns of hydrogenated samples; S-0-H, S-1-H, S-2-H, and S-3-H revealing that the hydrogenation did not produce any variation in the crystalline space group (SG) of CdO films. However, a slight change was occurred in the lattice parameter of the host CdO, as presented in Table 1. Hence, the hydrogenation slightly increases Vcell, especially for higher doping level (Fig. 3).

## 4. Optical Properties

The corrected transmittance, T(λ) and its wavelength reliance of undoped and doped CdO films are illustrated in Fig. 5(a). It is obvious that the investigated films were translucent in the visible region and gradually became transparent in the NIR spectral region. The experimental spectral reflectance, R(λ) of the treatise films/glass substrates was almost constant of ~ 5 %.

Some variations in the spectral distribution of R(λ) and T(λ) were observed under hydrogenation, as illustrated in Fig. 5(b), where the T(λ) curves begun damping in the NIR region. This is a result of the absorption of conduction electrons, referring to increases in the concentration of conduction electrons (Nel) by the hydrogenation.

The spectral absorbance A(λ) of the films was calculated using A(λ)=ln[(1-R(λ))/T(λ), while the spectral absorption coefficient α(λ) is related to the absorbance A(λ) by A(λ) = α(λ)d, where d is the film thickness. Therefore, the optical direct bandgap (Eg) can be estimated for each film through Tauc technique24):

$A . h υ = A o p ( h υ − E g ) 0.5$
(2)

where is the photon energy and Aop is the constant for each film. Fig.6a illustrates a graph of (A.) 2vs. ; the extrapolation of the straight-line portion to (A.)2 = 0 - line determines the bandgap (Table 2). The obtained bandgap of undoped CdO was within the known standard range of values, (2.2 ~ 2.6 eV) for films synthesized by variant techniques.4) With various CdTe doping level, the bandgap (Eg) of host CdO films varied as shown in the inset of fig.6b and Table 2, which can be explained by the followings; the doping calls for two contrary effects: the bandgap widening (BGW) or Moss-Burstein (B-M) effect25) and the bandgap-narrowing (BGN) effect. The changes in electron-lattice interactions that caused by the insertion of dopant impurity ions create impurity band tail broadening that incorporates with the conduction band producing BGN.26,27) The inset of Fig. 6(b) shows that the BGN defeated BGW in the concentration of less than ~ 4 %. However, for doping level ~ 7 %, where the part of the added Te ions accumulated on the CBs (as a result of the limited solid solubility of Te ions in CdO lattice), the BGW defeated the BGN.

Because of the hydrogenation, the optoelectronic properties of the samples changed based on the consequence of the creation of more O-vacancies along with more carrier concentration, Nel. The noticed hydrogenation effect can be studied through the inset of Fig. 6(b) and data of Nel in Table 2. The hydrogenation of the samples increased the electronic concentration, which supports the B-M effect; however, the created O-vacancies generate more tail to the conduction band that supports the BGN. Therefore, in the result, the bandgap by the hydrogenation was almost unchanged (inset of Fig. 6(b)). The creation of O-vacancies by the hydrogenation is attributed to the dissociation of H2 molecules in the presence of Te4+ metalloid ions. The formed active H ions/atoms create Ovacancies by removing structural oxygen, which can intensify the density of conduction electrons Nel.

## 5. Electrical Properties

The measured conduction parameters (TCO-CPs): σ, μel, and Nel of n-type undoped and CdTe-doped CdO films are presented in Fig. 7(a) and Table 2. The TCO-CPs of the host CdO films are evolved by the incorporation of CdTe. Fig. 7(a) disclosed that the utmost TCO-CPs improvement was found with the host CdO film included ~ 1%Te, where the resistivity decreased by 8.5 times, the carrier concentration increased by 4.5 times, and the mobility increased by 1.9 times. The resistivity slightly and almost linearly increased with increasing of doping level more than ~ 1 %, which was attributed to the gradual pile-up of dopant on the CBs and GBs. Such accumulation created a potential barrier that reduced the measured effective carrier concentration and conductivity, while the mobility remained around 14 cm2/Vs. This means that the active changes in TCO-CPs of CdTe-doped CdO are being at low dopant concentrations (around 1 %), which agree with previous results on Te-doped CdO.12)

The measured TCO CPs of hydrogenated undoped and CdTe-doped CdO films is illustrated in Fig. 7(b) and Table 2. The hydrogenation improved the TCO-CPs, especially for sample S-1-H, so that the resistivity reduced by 2.5 times and the carrier concentration by ~ 3 times, however, the mobility was reduced by 20 % due to increasing of Ncarr, comparing with non-hydrogenated films.

As mentioned above, the strongest [111] orientation for sample S-1 should have a strong effect on the mean-freepath (mfp), which defined as mfp = (h/2e)(3Nel/π)1/3μel, where h is the Planck constant and e is the electronic charge.28) The calculated value of mfp for films was in the range of 1 ~ 2 nm, with the highest value of 1.6 nm and 1.8 nm for samples S-1 and S-1-H, respectively.

## 6. Conclusions

CdO thin films merged with variant amounts of CdTe semiconductor (CdTe-doped CdO) were prepared by vacuum evaporation technique. The structural study indicates that the prepared CdTe-doped CdO films have a single phase of CdO structure. Moreover, it was found that the CdTe doping strongly consolidated the [111] preferred orientation of CdO/glass films. The optical results of CdTe doping of CdO are explained by emerging of two concurrent effects; BGN (redshift) and BGW (blueshift) as a consequence of doping processes. The hydrogenation does not affect the space group of the structure of the samples; however, it could improve the TCO-CPs. It was found that the TCOCPs of host CdO films were improved by CdTe doping but with lower scale comparing to the case of doping with pure Te ions. The utmost TCO-CPs improvements were found with host CdO film included ~ 1%Te, where the resistivity decreased by ~ 750 %, carrier concentration increased by 355 %, and the mobility increased by ~ 90 %. The improvement of TCO-CPs by hydrogenation is attributed to the creation of O-vacancies because of H2 molecules dissociation in the presence of Te ions. The hydrogenation improved the TCO-CPs, especially for the film included ~ 1%Te, so that the conductivity increased by 150% and the carrier concentration by 210%, however, the mobility was reduced by 20 % due to increase of Ncarr, comparing to non-hydrogenated films. These results clarify the possibility of using the semiconductor CdTe - doped CdO thin films in TCO applications.

## Figure

XRF spectra of CdTe reference film and S-3 film sample grown on Si wafers.

(a) XRD patterns of undoped, CdTe-doped CdO films and pristine CdTe reference film, (b) Rietveld refinements for S-3 film: the blue solid line (up) is the experimental data, the red solid line (up) is the calculated pattern, and the solid pink line (down) is the intensity difference, (c) Williamson-Hall plot for S-3 film.

Variation of unit-cell volume (Vcell) with Te% doping level for the as-prepared and hydrogenated samples.

XRD patterns of hydrogenated undoped and CdTe-doped CdO films.

(a) Transmittance, T(l) and reflectance, R(l) spectra of pristine and CdTe-doped CdO films, (b) Transmittance, T(λ) and reflectance, R(λ) spectra of hydrogenated (H) pristine and CdTedoped CdO films.

(a) Tauc plot for pristine and CdTe-doped CdO films, (b) Tauc plot for hydrogenated pristine and CdTe-doped CdO films. The inset shows the doping-level dependence of the as-prepared and hydrogenated samples.

(a) Conduction parameters (CPs) dependence on Te% doping level in the samples, (b) Conduction parameters (CPs) dependence on Te% doping level in the hydrogenated samples.

## Table

The structural parameters of the studied undoped and CdTe-doped CdO films; lattice parameter (α), volume of unit-cell (Vcell), Rietveld refinement parameters (Rwp and S), average crystallite size (CS), structural strain (εss), and Bragg angle of (111) reflection.

The measured electrical conduction parameters (CPs) (ρ, μel, and Nel) and optical band gap (Eg) for all film samples of the present work.

## Reference

1. B. G. Lewis and D. C. Paine, Mater. Res. Soc. Bull., 25, 22 (2000).
2. M. Yan, M. Lane, C. R. Kannewarf and R. P. H. Changa, Appl. Phys. Lett., 78, 2342 (2001).
3. Powder Diffraction File, Joint Committee for Powder Diffraction Studies (JCPDS) file No. 05-0640.
4. Z. Zhao, D. L. Morel and C. S. Ferekides, Thin Solid Films, 413, 203 (2002).
5. P. K. Gupta, K. Ghosh, R. Patel, S. R. Mishra and P. K. Kahol, Curr. Appl. Phys., 9, 673 (2009).
6. A. A. Dakhel, Adv. OptoElectron., 2013, 6 (2013).
7. A. A. Dakhel, Thin Solid Films, 41, 2405 (2012).
8. K. M. A. Hussain, Z. H. Mahmood, I. M. Syed, T. Begum, T. Faruqe1 and J. Parvin, Am. J. Mater. Sci. Application, 2, 91 (2014).
9. F. V. Wald, Revue de Physique Appliquee, 12, 277 (1977).
10. R. D. Shannon, Acta Crystallogr., Sect. A: Found. Adv., 32, 751 (1976).
11. S. Sonmezoglu, T. A. Termeli, S. Akın and I. Askeroglu, J. Sol-Gel Sci. Technol., 67, 97 (2013).
12. A. A. Dakhel, Sol. Energy, 84, 1433 (2010).
13. F. Jamali-Sheini, R. Yousefi, M. R. Mahmoudian, N. A. Bakr, A. Saaedi and N. M. Hyang, Ceram. Int., 40, 7737 (2014).
14. S. Sonmezoglu and E. Akman, Appl. Surf. Sci., 318, 319 (2014).
15. T. Minami, Thin Solid Films, 516, 1314 (2008).
16. T. Ratcheva, M. Nanova, L. Kinova and I. Penev, Thin Solid Films, 202, 243 (1991).
17. B. Georgieva, I. Podolesheva, G. Spasov and J. Pirov, Sensors (Basel), 14, 8950 (2014).
18. J. M. Jaklevic and F. S. Goulding, Energy Dispersion in X-ray Spectrometry, p.50, ed. H. K. Herglotz and L. S.Birks, NY: M. Dekker (1978).
19. A. A. Dakhel and H. Hamad, Int. J. Thin Films Sci. Technol., 1, 25 (2012).
20. A. A. Dakhel, Chem. Phys., 130, 398 (2011).
21. A. A. Dakhel, J. Mater. Sci.: Mater. Electron., 29, 3584 (2018).
22. L. B. McCusker, R. B. Von Dreele, D. E. Cox, D. Louer and P. Scardi, J. Appl. Cryst., 32, 36 (1999).
23. A. Khorsand Zak, W. H. Abd Majid, M. E. Abrishami and R. Yousefi, Solid State Sci., 13, 251 (2011).
24. J. Tauc and F. Abelesn ed., Optical Properties of Solids, North Holland (1969).
25. J. I. Pankove, Optical processes in semiconductors, p. 36, NY: Dover (1975).
26. Y. Z. Zhang, J. G. Lu, Z. Z. Ye, H. P. He, L. P. Zhu, B. H. Zhao and L. Wang, Appl. Surf. Sci., 254, 1993 (2008).
27. Y. Dou, R. G. Egdell, T. Walker, D. S. L. Law and G. Beamson, Surf. Sci., 398, 241 (1998).
28. M. Chen, Z. L. Pei, X. Wang, Y. H. Yu, X. H. Liu, C. Sun and L. S. Wen, J. Phys. D: Appl. Phys., 33, 2538 (2000).