Thin Solid Films 497 (2006) 20 – 23 www.elsevier.com/locate/tsf

Transparent conductive ZnO:Al thin films deposited on flexible substrates prepared by direct current magnetron sputtering Z.L. Pei, X.B. Zhang, G.P. Zhang, J. Gong, C. Sun, R.F. Huang, L.S. Wen *,1 Institute of Metal Research, Chinese Academy of Science, Shenyang 110016, PR China Received 22 November 2004; received in revised form 16 September 2005; accepted 16 September 2005 Available online 24 October 2005

Abstract In this paper, we report different methods to reduce the sheet resistance of ZnO:Al (AZO) films on flexible substrates without degrading the optical transmittance in the visible range. Under proper bias, AZO films deposited on Al2O3-buffered flexible substrates showed a significant decrease of sheet resistance when compared with those deposited on bare polymer. The films with resistivity as low as 8.4
10 4 V cm and the optical transmittance about 80% have been obtained by improved methods. By calculating the Al doping efficiency and the mean free path of electrons, ionized impurity scattering was considered to be the dominant factor for the transport mechanism of carriers. D 2005 Elsevier B.V. All rights reserved. PACS: 73.61.-r; 78.66.-w; 68.55.-a Keywords: Zinc oxide; TCO; Buffer layer; Sputtering

1. Introduction Al-doped ZnO thin films deposited on rigid glass substrates have been extensively studied in recent years because they combine attractive properties with high visible transparency and electrical conductivity [1 –3]. As is well known, glass is too heavy and brittle to be easily deformed, especially for certain applications, such as smart card, electronic map and flat panel display where flexibility and lightweight are needed [4,5]. Therefore transparent conducting films deposited on flexible polymer substrates could overcome these problems and it’s necessary to investigate the characteristic of the oxide thin films deposited on polymer substrates. To our knowledge, few papers have been published for ZnO:Al (AZO) films deposited on flexible substrates [6,7] compared to In2O3:Sn (ITO) films. Due to the poor thermal stability of organic material, the films should be deposited at low temperature. However, the low temperature process is not in favor of depositing high quality films, the resistivity of AZO films on polymer is quite high to be adopted. In order to overcome these problems, the approach follow in * Corresponding author. Tel.: +86 24 83978232; fax: +86 24 23843446. E-mail addresses: [email protected] (Z.L. Pei), [email protected] (C. Sun), [email protected], [email protected] (L.S. Wen). 1 Tel.: +86 24 83978235; fax: +86 24 23843446. 0040-6090/$ - see front matter D 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.tsf.2005.09.110

this work was: (1) Before deposition, the argon plasma surface treatment was applied to boost the reactivity of the polymer and improve the sticking and adhesion of the deposited particles. (2) An inorganic buffer layer was sandwiched between the AZO film and polymer substrate. (3) A bias supply could give the additional energy to the sputtering particles and rule out the weak bonding atoms condensed on the substrate [8]. Some authors [9] suggested that proper bias bombardment could improve the film crystallinity and be beneficial to the film growth in the low temperature deposition. In the present work, an attempt was made to improve the electrical properties of AZO thin films deposited on organic substrates. The dependence of the electrical, structural and optical properties on AZO films deposited on the bufferedsubstrate was investigated. Finally, the transport mechanism of the free electrons will be discussed. 2. Experimental details The AZO thin films were deposited by a direct current (DC) reactive magnetron sputtering system from a Zn:2.0 wt.% Al alloy target (purity: 99.99%) onto unheated polymer substrates. The polymeric substrate employed was polyethylene terephalate (PET) from LuckyFilm Co., LTD with a standard thickness of 125 Am, which was ultrasonically cleaned by isopropyl

Z.L. Pei et al. / Thin Solid Films 497 (2006) 20 – 23

alcohol and blown dry with N2 gas before they were introduced into the deposition chamber. Before each run, the target was pre-sputtered in a pure argon atmosphere for 3 min in order to remove the natural surface oxide layer of the target. The polymer substrates were not intentionally heated during deposition. The substrate temperature can increase to about 80 -C due to the continuous bombardment of energetic particle. The distance between the target and the substrate was set at 65 mm. A parallel experiment for AZO films deposited on hard glass substrate was also carried out for comparison. High purity Ar and O2 were introduced through a mass flow controller after the vacuum chamber was evacuated to below 3
10 3 Pa. Prior to AZO deposition, the Al2O3 thin film (about 40 nm) was deposited on the substrate using pulse magnetron technology (see Fig. 1). The main objective of the thin buffer layer is to prevent the degradation of the polymer during the deposition, as well as to prevent diffusion or chemical reactions at the polymer/film interface. The experimental parameters for both AZO film and Al2O3 buffer layer are listed in Table 1. The sheet resistance (R g) of the samples was measured with a four-point probe and the resistivity of the films was calculated (q = R gId). The film thickness (d) was determined by Ambios Technology XP-2 Surface profilometer. An X-ray diffractometer (XRD) with Cu K a radiation (k = 0.154056 nm) was used to identify the crystalline phase of the films. The films were cross-sectioned by a focused-ion-beam (FIB, FEI Nova NanoLab 200, combined with a thermal field emission SEM) at an ion beam current of 0.14 nA and then further cleaned by a fine beam current of 10 pA at a 30 kV accelerating voltage. The optical transmittance and reflectance of the films were measured with an ultraviolet-visible-near infrared spectrophotometer (Hitachi U-2800). 3. Results 3.1. Electrical and optical properties Fig. 2 shows the electrical sheet resistance of AZO films deposited on bare PET, Al2O3-buffered PET and glass

21

Table 1 Deposition parameters for Al2O3 buffer and AZO films Deposition parameter

Al2O3 buffer

AZO film

Power mode Total gas pressure (Pa) Oxygen ratio (%) Bias voltage (V) Plasma treatment (min)

pulse 0.6 15 40 6

DC 0.6 6~10 40 0

substrates as a function of oxygen partial pressure at room temperature, and a fixed total gas pressure of 0.6 Pa. The reactive sputtering process was carried out in the transition mode related to the target surface state. For all samples Ar plasma surface treatment process was made and the power was kept at 80 W in order to avoid the surface damage. Extensive work [10] has already been performed on the treatment of polymers, which is one of the physical and chemical methods of modifying polymer surface. The polymer was cleaned in an ultrasonic bath with isopropanol before treatment. It is found from Fig. 2 that the overall sheet resistance for bare polymer was relatively high compared to that of glass substrate. As is well known, the flexible polymeric substrate is very sensitive to moisture and oxygen and the PET substrate has a tendency of absorbing and permeating much more oxygen gas and moisture than glass substrate [11]. During deposition, these inexhaustible elements could flow out and diffuse into the films, which would work as acceptors or scattering centers in AZO films to deteriorate the electrical properties. So it is necessary to prevent the diffusing process for prepared high quality AZO films. One of the most effective methods was to deposit a buffer layer between the film and the substrate. Transparent Al2O3 thin films on PET substrate have been widely studied as a gas barrier material [12] and the experimental results have proved the idea. From Fig. 2 it can be seen that the overall sheet resistance of AZO film on Al2O3-buffered PET substrate showed remarkable decrease than that of AZO film on bare PET substrate. Fig. 3 shows the transmittance spectra of AZO films grown on Al2O3-buffered PET (solid line) and bare PET (dashed line) in the range of 200 –800 nm. These samples were prepared at room temperature, 0.6 Pa total pressure and 7.5% oxygen ratio.

Sheet resistance (Ω /sq.)

800 AZO / Glass AZO / PET AZO / Al2O3 / PET

600

400

200

0 6

7

8

9

10

O2 concentration (%)

Fig. 1. SEM image (with 52- tilt) of a section of an AZO film deposited on Al2O3-buffered PET substrate patterned using FIB.

Fig. 2. Influence of the oxygen partial pressure on the sheet resistance for different substrate types. The substrate deposition temperature was room temperature and the total pressure was kept at 0.6 Pa.

22

Z.L. Pei et al. / Thin Solid Films 497 (2006) 20 – 23 1.0

Transmittance

0.8

0.6 Bare polymer AZO / PET AZO / Al2O3 / PET

0.4

0.2

0.0 200

400

600

800

Wavelength (nm) Fig. 3. The transmittance spectra of AZO films grown on Al2O3-buffered PET (solid line) and bare PET (dashed line) in the visible range.

The Al2O3 films were about 40 nm. An average transmittance of about 80% for Al2O3-buffered PET in the visible range was obtained. Compared to AZO films on bare PET, the decrease of transmittance in the blue region and the increasing of transmittance in the red region for AZO film on Al2O3buffered PET were also observed. The wave-like modulation in the transparency may be due to the thickness effect. The additional Al2O3 thin layer does not much degrade the film transparency. 3.2. Structural characterization Fig. 4 shows XRD curve of AZO films deposited on Al2O3buffered and bare PET substrates at a total gas pressure of 0.6 Pa, oxygen ratio of 7.5%. The AZO films with and without Al2O3 buffer layers were all dominated by the (002) texture and no Zn, Al and Al2O3 characteristic peaks were detected. It can be also noted from Fig. 4, with Al2O3 buffer layer, AZO film shows higher diffraction intensity than AZO film does on bare PET. This behavior could be related to the film structure and the interface effect between the AZO and Al2O3 buffer. The slight shift of (002) diffraction angle could be attributed to the crystal lattice deformation resulting from doping or stress

in the films compared with (002) diffraction angle in the Joint Committee on Powder Diffraction Standards file no. 5664[13]. SEM (available in the FIB apparatus) was also used to investigate the morphology of AZO films under different conditions. The SEM imaging conditions are at an operating voltage of 10 kV and beam current of 0.13 nA. The oxygen concentration (characterized by ratio of oxygen partial pressure to total pressure) was kept at 7.5% during deposition. Samples deposited at 0 V bias have spherical crystallites with an average grain size of 35 nm, as shown in Fig. 5(a). The film was quite homogeneous and no cracks or peeling off was found. Fig. 5(b) shows the surface morphology of the AZO film deposited at 40 V bias voltage. The slightly tilted elliptical crystallites with grain sizes between 45 and 65 nm were observed on the sample surface. An increase in crystallite size would cause an increase in Hall mobility, which is related to grain scattering [14]. 3.3. Scattering mechanism of carriers In order to relate the fundamental electronic transport behavior to the dependence of resistivity, a scattering mechanism should be emphasized. Refer to the In2O3:Sn (ITO) film deposited at low temperature, it shows high conductivity and non-crystal structure, which the grain scattering can be neglected. As for the un-doped polycrystalline ZnO film, the grain boundary plays an important role in determining the characteristic of carrier scattering [15]. But for Al-doped ZnO films, the dopant atoms (Al) are substitutionly incorporated at Zn sites in the ZnO lattice. Under the assumption that every dissolved Al atom provides one free electron, the doping

2.5x105

ZnO (002)

Intensity (arb.units)

3.0x105

2.0x105 AZO / Al2O3- buffered PET

1.5x105

1.0x105 30

AZO / PET 35

40

45

50

2θ (degree) Fig. 4. X-ray diffraction patterns for AZO films deposited on Al2O3-buffered PET and bare PET substrates. The oxygen concentration was kept at 7.5% during deposition.

Fig. 5. Surface morphology of AZO films deposited at different bias voltages (a) 0 V, (b) 40 V. The oxygen concentration was kept at 7.5% during deposition.

Z.L. Pei et al. / Thin Solid Films 497 (2006) 20 – 23

efficiency, g, is defined as the ratio of the number of free electrons (n) in the films to that of Al doped atoms (n), the specific density of ZnO crystal is 5.67 g cm 3): n ,56% g¼ ½ n That is to say that about 60% of these Al atoms in the film is effectively doped to work as donors. Thus together the oxygen vacancies, the excess Al atoms incorporated in the films may segregate into grain boundary or tend to form neutral Al-based defect complexes in form of Al2O3, which not only not contribute the free electrons, but also act as scattering centers to deteriorate the electrical property of AZO thin films. Comparing the electron mean free paths [16] with the grain sizes of the AZO films:   1=3 h 3n l¼ l; e p Where l is about a few nanometers, much smaller than the grain size (about 40– 60 nm in our films). So it can be concluded that intra-grain scattering, namely ionized impurity scattering seems to play a dominant role. Additionally in our experiment, as the oxygen pressure increased from 7.0% to 12%, we found that the intensity of the (002) diffraction peak increased and the FWHM of the (002) peak decreased, indicating that the crystalline and grain sizes relating to grain scattering were improved. However the conductivity of the films was significantly decreased inversely as the increase of oxygen partial pressure. Therefore for highly degenerated Al-doped ZnO films, ionized impurity scattering dominated the transport mechanism of free electrons [17] and grain boundaries scattering plays a subordinate role. Until now many discussions [18 –20] of the mechanism of the ionized impurity scattering are not able to describe the experimental data very well, based on the classic theories of Conwell-Weisskopf and Brook-Herring-Dingle [21,22], even if the non-parabolic band structure is taken into account. In view of the fact that the resistivities of TCO films have reached a limit of around 10 4 V cm, further attempts will be needed to surmount this limit. 4. Conclusions In summary, transparent conductive ZnO:Al (AZO) films have been deposited successfully by DC reactive magnetron sputtering onto flexible PET substrates. By applying a thin

23

Al2O3 buffer layer and proper bias, the overall sheet resistance of AZO films was remarkable reduced. The transmittance spectrum showed few changes for AZO films on Al2O3buffered or bare PET substrates. By calculating the Al doping efficiency and the mean free path of electrons, ionized impurity scattering was concluded to be the major factor in determining the scattering mechanism. All the AZO films deposited on Al2O3-buffered PET showed a resistivity of 8.4
10 4 V cm with a visible transmittance of about 80%. Acknowledgements This work was supported by National Natural Science Foundation of China (Grant No. 50172051). References [1] G. Granqvist, Appl. Phys. A57 (1993) 19. [2] K. Ellmer, F. Kudella, R. Mientus, R. Schieck, S. Friechter, Thin Solid Films 247 (1994) 15. [3] B. Szyszka, Thin Solid Films 351 (1999) 164. [4] Y. Yang, MRS Bull. 22 (1997) 31. [5] S. Forrest, P. Burrows, M. Thompson, IEEE Spectrum 37 (2000) 29. [6] T.L. Yang, D.H. Zhang, J. Ma, H.L. Ma, Y. Chen, Thin Solid Films 326 (1998) 60. [7] E. Fortunato, P. Nunes, D. Costa, D. Brida, I. Ferreira, R. Martins, Vacuum 64 (2002) 233. [8] H.L. Ma, X.T. Hao, J. Ma, Y.G. Yang, S.L. Huang, F. Chen, Q.P. Wang, D.H. Zhang, Surf. Coat. Technol. 161 (2002) 58. [9] N. Danson, I. Safi, G.W. Hall, R.P. Howson, Surf. Coat. Technol. 99 (1998) 147. [10] Y. Leterrier, Prog. Mater. Sci. 48 (2003) 1. [11] S.K. Park, J.I. Han, W.K. Kim, M.G. Kwak, Thin Solid Films 397 (2001) 49. [12] B.M. Henry, A.G. Erlat, A. McGuigan, C.R.M. Grovenor, G.A.D. Briggs, Y. Tsukahara, T. Miyamoto, N. Noguchi, T. Niijima, Thin Solid Films 382 (2001) 194. [13] 1976 Powder Diffraction Data Joint Committee on Powder Diffraction Standards, Pennsylvania, USA, p29. [14] H. Kim, C.M. Gilmore, J. Appl. Phys. 86 (1999) 6451. [15] F.R. Blom, F.C.M. Van de pol, G. Bauhuis, Th.J.A. Popma, Thin Solid Films 204 (1991) 365. [16] L. Eckertova, Phys. Thin Films, Plenum Press, New York, 1997, p. 180. [17] T. Minami, H. Sato, K. Ohashi, T. Tomofuji, S. Takata, J. Cryst. Growth 117 (1992) 370. [18] N. Taga, O.Y. Shigesato, I. Yasui, M. Kamei, T.E. Haynes, J. Appl. Phys. 80 (1996) 978. [19] R.B.H. Tahar, T. Ban, Y. Ohya, Y. Takahashi, J. Appl. Phys. 83 (1998) 2631. [20] K. Ellmer, J. Phys. D: Appl. Phys. 34 (2001) 3097. [21] E. Conwell, V.F. Weisskopf, Phys. Rev. 83 (1951) 879. [22] H. Brooks, Phys. Rev. 69 (1946) 258.