Available online at www.sciencedirect.

com

Thin Solid Films 516 (2008) 1377 1381 www.elsevier.com/locate/tsf

Dependence of the electrical and optical properties on the bias voltage for ZnO:Al films deposited by r.f. magnetron sputtering
Jae-Hyeong Lee a,, Jun-Tae Song b
a b

School of Electronics and Information Engineering, Kunsan National University, Kunsan, South Korea School of Information and Communication Engineering, Sungkyunkwan University, Suwon, South Korea Available online 31 March 2007

Abstract Aluminum-doped zinc oxide (ZnO:Al) thin films were deposited on glass, polycarbonate (PC), and polyethylene terephthalate (PET) substrates by r.f. magnetron sputtering. The substrate dc bias voltage varied from 0 V to 50 V. Structural, electrical and optical properties of the films were investigated. The deposition rate of ZnO:Al films on glass substrate initially increased with the bias voltage, and then decreased with further increasing bias voltage. It was found that the best films on glass substrate with a low as 6.2 10 4 cm and an average transmittance over 80% at the wavelength range of 500900 nm can be obtained by applying the bias voltage of 30 V. The properties of the films deposited on polymer substrate, such as PC and PET, have a similar tendency, with slightly inferior values to those on glass substrate. 2007 Elsevier B.V. All rights reserved.
Keywords: Al-doped ZnO (ZnO:Al); R.f. magnetron sputtering; Bias voltage; Polymer substrate

1. Introduction Transparent conducting oxide (TCO) films based on zinc oxide are taking a great impact because they have a number of advantages, such as non-toxicity, low cost, material abundance, relatively low deposition temperature, and high stability against hydrogen plasma compared to ITO and SnO2 films [1]. TCO films deposited on polymer substrates have many merits due to their light weight, small volume. Devices made from these films can be folded and easily carried. They can be used in plastic liquid crystal display devices, transparent electromagnetic shielding materials, flexible electro-optical devices, and unbreakable heat-reflecting mirrors [2]. In order to realize sputter deposition of the ZnO films on plastic substrate, low temperature and damage-less deposition is necessary to obtain the film with low resistivity. Many researchers have tried to produce highly conductive ZnO films at low substrate temperature [35]. However, it became difficult to obtain a film with low resistivity without substrate heating. Recently, the introduction

of H2 gas during film growth or in-situ H2 post-treatment on the undoped ZnO films were performed in order to achieve highly conductive films [3,6]. In addition, Raniero et al. [7] and Tohsophon et al. [8] reported the electro-optical behavior and the structure of gallium-doped zinc oxide (GZO) after hydrogen plasma treatment and the damp heat stability behavior of aluminum doped zinc oxide (AZO or ZnO:Al) films to investigate the chemical stability and the long-term stability. For sputtering technique, it is well known that the energetic particles such as electrons [9] and negative ion [10] have some impact on the deposited film. Also it was reported that composition of a ceramic film was much dependent on substrate bias [11] and proper ion impact on the deposited film caused improvement of film properties [12]. In this study, ZnO:Al films were deposited on glass, PC, and PET at room temperature by r.f. magnetron sputtering. The effects of bias voltage on the electrical and optical properties of the films were evaluated. 2. Experimental The ZnO:Al films were prepared by r.f. magnetron sputtering system. The polymeric substrate used in this work was PC and PET with a standard thickness of 100 m. A commercially available, sintered ceramic ZnO:Al (2 wt.% Al2O3) target with

Corresponding author. San 68, miryong-dong, Kunsan, Jeollabuk-do, South Korea (573-701). Tel.: +82 63 4694707; fax: +82 63 4694699. E-mail address: jhyi@kunsan.ac.kr (J.-H. Lee). 0040-6090/$ - see front matter 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2007.03.078

1378

J.-H. Lee, J.-T. Song / Thin Solid Films 516 (2008) 13771381

99.99% purity (Superconductive Materials Co.) of 3 inch in diameter was employed as source materials. The target-tosubstrates distance was 7 cm. Before depositing the ZnO:Al thin films, the substrates were ultrasonically cleaned in a detergent bath, followed by methanol alcohol and dried in nitrogen. The substrates were placed inside the chamber and then evacuated to a base pressure of 6.6 10 4 Pa. The gas pressure was kept at 0.3 Pa and the sputtering power during deposition was 150 W. All the films were deposited at room temperature and the target was water-cooled. The substrate surface was monitored by thermocouple during sputter deposition. Although the substrates were not intentionally heated, the surface temperature reached about 40 C. The substrate was negatively biased with d.c. source of 20 to 80 V. The film thickness was measured by a conventional stylus surface profiler. The film thickness is 200 nm and all samples kept the same thickness in this work. Because the polymer substrate, such as PET, deforms during electron microscopic investigation, atomic force microscope (AFM) was used to investigate the microstructure of films. A compositional analysis was performed using X-ray photoemission spectroscope (XPS). The sheet resistance of the samples was measured with a four-point probe and the resistivity of the film was calculated. Carrier concentration and Hall mobility were obtained from Hall-effect measurement by the Van der Pauw technique. The optical transmittance measurements were performed with a Hitachi UV-3200 spectrophotometer. 3. Results and discussion The prepared films at optimum conditions on PC or PET substrates were physically stable and had good adhesion to the substrates. No cracking or peel-off of the films was observed after deposition. Fig. 1 shows the deposition rate of ZnO:Al films deposited on glass substrate at different bias voltages. The deposition rate initially increases with the bias voltage. This is due to the Ar+ cation and clusters of the sputtered material in the plasma being

Fig. 2. X-ray diffraction patterns for films deposited at different bias voltages: (A) glass substrate; (B) PC substrate; (C) PET substrate.

Fig. 1. The dependence of the deposition rate on bias voltage for ZnO:Al films.

J.-H. Lee, J.-T. Song / Thin Solid Films 516 (2008) 13771381

1379

attracted by the substrate, resulting in a higher deposition rate [13]. As the bias voltage increases above 30 V, however, the bombardment makes the number of molecules peeled off from the film increase and the deposition rate consequently decreases. The XRD patterns of the ZnO:Al films deposited at different bias voltage are shown in Fig. 2. For glass substrate, the film deposited without bias exhibits only a peak at 2 = 34.2 which was very close to the preferred orientation of the standard ZnO crystal (2 = 34.458). This implies that the film has a hexagonal structure with the (0 0 2) preferred orientation. No Al2O3 related phase was detected from the XRD patterns, suggesting that aluminum substitutionally replaces zinc in the hexagonal lattice or aluminum segregates into the non-crystalline region in grain boundary [14]. As the bias voltage increases, the intensity of (0 0 2) peak becomes more intense and sharper while the location of the measured diffraction peak does not change significantly. This should be attributed to large amount of electrons flying to the substrate. These electrons transfer their energy to the atoms and molecules on the surface [15]. This effect will increase the diffusion mobility of these particles on the surface, which will help the films to crystallize [16]. In addition, new diffraction peak, shows at 2 = 72.2, which is associated with the (0 0 4) plane of the hexagonal phase. When the bias voltage is higher than 40 V, however, the intensity of (0 0 2) diffraction peak decreases and the (0 0 4) peak disappeared. For the PC and PET substrate, the (0 0 2) and (0 0 4) peaks exhibit for the sample prepared without the substrate bias. The intensity of (0 0 2) plane reduces with the bias voltage. Especially, it should be noted that the diffraction peak disappear when the bias voltage is 50 V. Fig. 3 shows the dependence of bias voltage on the resistivity of ZnO:Al films deposited at various substrates. As the bias voltage increases from 0 to 30 V, the resistivity decreases from 8.5 10 4 cm to 6.2 10 4 cm. These variations are attributed to improved crystallinity, increased substitutional doping and decreased interstitial atoms. When the bias voltage increases further, however, the resistivity increases. For the films deposited

Fig. 4. Carrier concentration and Hall mobility of ZnO:Al films on glass substrate as a function of bias voltage.

Fig. 3. Dependence of bias voltage on the resistivity of ZnO:Al film deposited on glass and PC substrate.

on a PC substrate, the resistivity decreases with the bias voltage and reaches a minimum with value of 1.3 10 3 cm at 20 V. Beyond 20 V, the resistivity increases rapidly with the bias voltage. Although the films on PC substrate exhibit a similar dependence on the bias voltage, the resistivity is higher than that for glass substrate, regardless of bias voltage. Although the films on PC substrate exhibit a similar dependence on the bias voltage, the resistivity is higher than that for glass substrate, regardless of bias voltage. The polymer substrates present certain challenges such as considerably lower deposition temperature and rougher surfaces as compared to glass substrates. Since the electrical conductivity of ZnO:Al films may be affected by carrier scattering from a rough surface, the ZnO:Al films on polymer substrates are more resistive than films on glass substrate. Especially, too much of energy deforms the polymer substrate due to poor heat endurance and leads to some undesirable impurities escaping from the polymer substrate, which result in a degradation of film properties [15]. For PET substrate, the resistivity of ZnO:Al films is 4.5 and 80.2 cm at the bias voltage of 30 V and 50 V, respectively. Fig. 4 illustrates the carrier concentration (N) and the Hall mobility () of the ZnO:Al films deposited on glass substrate as a function of the bias voltage. As the bias voltage increases, the carrier concentration and Hall mobility is enhanced, hence, the resistivity decreases. However, N and , after reaching a maximum, starts decreasing again with further increasing bias voltage. The dependence of carrier concentration on bias voltage was contributed by oxygen content in the film. In this work, the Zn/O ratio changed from 0.89 to 1.09 when the bias voltage of 30 V applied to the substrate. Therefore, the change of the carrier concentration could be explained by the variation of oxygen deficiency in the films. On the other hand, the increase of mobility is attributed to the crystalinity of the films, as seen in Fig. 2. A bias applied to the substrate can attract cations in the plasma to bombard the growing film. This bombardment gives an additional energy to the molecules and clusters condensed on substrate and peel off the molecules with weak bonding from the film [17]. Danson et al. [18] believed that this kind of bombardment was beneficial to film

1380

J.-H. Lee, J.-T. Song / Thin Solid Films 516 (2008) 13771381

without applying bias voltage, respectively. This difference is attributed to the lower transmittance of the polymer substrates. The average transmittance of the films slightly increases with the bias voltage, regardless of substrate types. The applied bias is beneficial to film crystallinity and can reduce the defects ratios, which increases the transmittance of the ZnO:Al film. Compared with the film deposited without bias, while the bias voltage is 30 V, the absorption edge of the films has a slightly shift to the short wavelength, indicating that the optical band gap of the films increases with the bias voltage. This can be interrupted as follows: When the bias apply to the substrate, the crystallite sizes increase leading to less grain boundary and microvoids, and thereby the carrier concentration of the samples increased because of the more clearly effect of replacement doping [13]. The widening of the band gap due to increasing carrier concentration is attributed to BursteinMoss shift [19]. The band gap shift in an n-type semiconductor, such as ZnO:Al, is given by [20] DEBM J2 3p2 Ne 2=3: 2m e ve 1

where Ne is the free-electron concentration, m ve the reduced effective mass, and e the electronic charge. However, the optical transmittance deteriorated when the higher bias voltage (50 V) applied to the substrate. 4. Conclusions ZnO:Al films were deposited on glass, PC, and PET substrates by bias r.f. magnetron-sputtering. The suitable bias can improve the quality of the films because it provides extra energy by attracting the cations to bombard the surface of the growing films. All of the obtained films were polycrystalline with the hexagonal structure and had a preferred orientation with the caxis perpendicular to the substrate. Highly transparent films with a visible transmittance above 80% and the lowest resistivity of 6.2 10 4 cm have been obtained by applying a bias voltage of 30 V. The properties of the films deposited on polymer substrate, such as PC and PET, have a similar tendency, with slightly inferior values to those on glass substrate. Acknowledgement This work has been supported by KESRI (R-2005-7-147), which is funded by MOCIE (Ministry of Commerce, Industry and Energy).
Fig. 5. Optical transmittance spectra for ZnO:Al films deposited at different bias voltages: (A) glass substrate; (B) PC substrate; (C) PET substrate.

References
[1] D. Song, A.G. Aberle, J. Xia, Appl. Surf. Sci. 195 (2002) 291. [2] D.H. Zhang, T.L. Yang, Q.P. Wang, D.J. Zhang, Mater. Chem. Phys. 68 (2001) 233. [3] S.Y. Myong, K.S. Lim, Solar Energy Mater. Solar Cells 86 (2005) 105. [4] C. Guillna, J. Herrero, Thin Solid Films 515 (2006) 640. [5] X. Yu, J. Ma, F. Ji, Y. Wang, X. Zhang, C. Cheng, H. Ma, Appl. Surf. Sci. 239 (2005) 222. [6] S. Takeda, M. Fukawa, Thin Solid Films 468 (2004) 234.

crystallinity and could replace the energy provided by a higher substrate temperature. Fig. 5 shows the optical transmittance of ZnO:Al films as a function of bias voltage. The average transmittance at the wavelength range of 500900 nm is 83.6% (glass substrate), 76.9% (PC substrate) and 75.2% (PET substrate) for the films

J.-H. Lee, J.-T. Song / Thin Solid Films 516 (2008) 13771381 [7] L. Raniero, I. Ferreira, A. Pimentel, A. Goncalves, P. Canhola, E. Fortunato, R. Martins, Thin Solid Films 511512 (2006) 295. [8] T. Tohsophon, J. Hupkes, S. Calnan, W. Reetz, B. Rech, W. Beyer, N. Sirikulrat, Thin Solid Films 511512 (2006) 673. [9] T. Lee, T. Machino, T. Sugihara, Thin Solid Films 224 (1993) 105. [10] S. Ishibashi, Y. Higuchi, Y. Ota, K. Nakamura, J. Vac. Sci. Technol. A8 (1990) 1403. [11] R.D. Bland, G.J. Kominiak, D.M. Mattox, J. Vac. Sci. Technol. 11 (1974) 671. [12] E. Kubota, Y. Shigesato, M. Igrashi, K. Suzuki, Jpn. J. Appl. Phys. 33 (1994) 4997. [13] H.-L. Ma, X.-T. Hao, J. Ma, Y.-G. Yang, S.-L. Huang, F. Chen, Q.-P. Wang, H. Zhang, Surf. Coat. Technol. 161 (2002) 58.

1381

[14] J. Isidorsson, C.G. Granqvist, Solar Energy Mater. Solar Cells 44 (1996) 375. [15] X. Hao, J. Ma, D. Zhang, Y. Yang, X. Xu, F. Chen, H. Ma, Appl. Surf. Sci. 189 (2002) 157. [16] B. Stjerna, E. Olsson, C.G. Granqvist, J. Appl. Phys. 76 (1994) 3797. [17] Z.W. Yang, S.H. Han, T.L. Yang, D.H. Lina Ye, H.L. Zhang, C.F. Ma, C.F. Cheng, Thin Solid Films 366 (2000) 4. [18] N. Danson, I. Safi, G.W. Hall, R.P. Howson, Surf. Coat. Technol. 99 (1998) 147. [19] E. Burstein, Phys. Rev. 93 (1954) 632. [20] D. Song, P. Widenborg, W. Chin, A.G. Aberle, Solar Energy Mater. Solar Cells 73 (2002) 1.