Thermally stable transparent conducting and highly infrared reflective Ga-doped ZnO thin films by metal organic chemical vapor deposition

Optical Materials 33 (2011) 768–772

Contents lists available at ScienceDirect

Optical Materials journal homepage: www.elsevier.com/locate/optmat

Thermally stable transparent conducting and highly infrared reflective Ga-doped ZnO thin films by metal organic chemical vapor deposition J.L. Zhao a, X.W. Sun a,b,⇑, H. Ryu c, Y.B. Moon d a

Department of Applied Physics, College of Science, Tianjin University, 92 Weijin Road, Tianjin 300072, People’s Republic of China School of Electrical & Electronic Engineering, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Singapore c Department of Nano Systems Engineering, Center for Nano Manufacturing, Inje University, Obang-dong, Gimhae, Gyeongnam 621-749, Republic of Korea d THELEDS Co., Ltd., Yongin-si, Gyeonggi-do 449-871, Republic of Korea b

a r t i c l e

i n f o

Article history: Received 17 July 2010 Received in revised form 9 October 2010 Accepted 13 December 2010 Available online 22 January 2011 Keywords: Ga-doped ZnO (GZO) Transparent conductive oxide (TCO) Thermal stability Infrared reflective coating Metal organic chemical vapor deposition (MOCVD)

a b s t r a c t Highly transparent conductive Ga-doped ZnO (GZO) thin films have been prepared on glass substrates by metal organic chemical vapor deposition. The effect of Ga doping on the structural, electrical and optical properties of GZO films has been systematically investigated. Under the optimum Ga doping concentration (4.9 at.%), c-axis textured GZO film with the lowest resistivity of 3.6  104 X cm and high visible transmittance of 90% has been achieved. The film also exhibits low transmittance (<1% at 2500 nm) and high reflectance (>70% at 2500 nm) to the infrared radiation. Furthermore, our developed GZO thin film can well retain the highly transparent conductive performance in oxidation ambient at elevated temperature (up to 500 °C). Ó 2010 Elsevier B.V. All rights reserved.

1. Introduction Transparent conducting oxide (TCO) thin films have a large variety of applications in optoelectronic industry, such as transparent electrodes for light-emitting diodes (LEDs), liquid crystal displays (LCDs), photodetectors and solar cells [1–5]. TCO films can also show high reflectance to infrared radiation due to the resonance plasma effect induced by high density carriers, which is promising for the infrared/heat reflective coating for energy saving glasses/ windows [6–8]. Currently the most developed TCO technology for practical applications is based on indium-tin-oxide (ITO). However, ITO will be difficult to meet the fast growing demand in the optoelectronics industry due to the scarcity of indium. As a result, indium-free TCOs have recently attracted much attention as a substitute for ITO [9,10]. Doped ZnO is one of the most promising candidates for replacing ITO [10–21]. ZnO has a direct wide band gap of 3.4 eV, which can be used for optoelectronic devices operating in the blue to UV region. The most important advantage of ZnO over ITO is the abundance of Zn element and the lower cost. To increase its electrical conductivity, ZnO is usually doped with group III elements including B, Al, Ga or In as effective donors. Al-doped ZnO (AZO) [7,16–18] and Ga-doped ZnO (GZO) [8,10–16] are the most ⇑ Corresponding author at: School of Electrical & Electronic Engineering, Nanyang Technological University, Nanyang Avenue, Singapore 639798, Singapore. E-mail address: [email protected] (X.W. Sun). 0925-3467/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.optmat.2010.12.008

widely developed TCOs for transparent electrode and infrared insulator applications. Furthermore, previous efforts suggest that GZO has better conduction stability than AZO because Ga is less reactive and more resistant to oxidation than Al [16]. Many deposition techniques have been employed to deposit ZnO based thin films, including magnetic sputtering [5,10–13,16–18], metal organic chemical vapor deposition (MOCVD) [1,2,15,19–23], pulsed laser deposition (PLD) [4,14,16], molecular beam epitaxy (MBE) [24], e-beam evaporation [25], chemical spray pyrolysis [26], sol–gel [27], etc. Among these methods, sputtering and MOCVD are preferred for practical use as the two are easily scalable for large area deposition with high film quality. Furthermore, MOCVD can also enable on-line glass coating (integrating the deposition process with floating glass production line) [28], which is very promising for development of low cost large area TCO glasses. In this paper, we shall report a systematically study on the Ga doping dependency, electrical conduction, transmittance and infrared reflectance properties for GZO thin films by MOCVD. Low resistivity, high visible transparency, high infrared reflectance, and high thermal stability have been achieved in the optimized GZO film. 2. Experimental procedure The GZO thin films were deposited by a home-made showerhead injector MOCVD system on amorphous glass substrates. Trimethylgallium (TMGa), dimethylzinc (DMZn) and oxygen were

J.L. Zhao et al. / Optical Materials 33 (2011) 768–772

769

used as precursors and nitrogen was employed as the carrier gas for the metal organics. The flow of N2 and O2 was controlled at 500 and 200 standard cubic centimeter per minute (sccm). The flow of TMGa and DMZn were controlled at 0–5 sccm and 2 sccm, respectively, where the Ga content in the film was adjusted by varying the TMGa flow. The GZO films were grown at a relatively low temperature of 300 °C with a chamber pressure of 25 Torr. The thickness of ZnO film is about 600 nm. To study the thermal stability, the GZO film was also deposited on quartz substrate under the aforementioned growth conditions, which was further annealed in air ambient at temperature of 400–600 °C with the duration of 2 h. The Ga content in GZO films was determined by energy dispersive X-ray spectroscopy (EDX) analysis attached to a JSM- 5600LV scanning electronic microscope (SEM). The film crystal quality was investigated by X-ray diffraction (XRD) with a Siemens X-ray diffractometer at 45 kV and 40 mA under h/2h scan mode. The surface morphology was characterized by atomic force microscopy (AFM, DIGITAL INSTRUMENTS NanoScope IIIa). The electrical properties including resistivity, carrier concentration and Hall mobility was measured by van der Pauw method using a Hall effect measurement system (HL5500PC) with the magnetic field strength of 0.326T. The optical transmittance and reflectance properties were obtained using a UV–VIS-NIR spectrophotometer (PERKIN ELMER, Lambda 950). 3. Results and discussion Fig. 1 shows the XRD patterns (h/2h) of GZO thin films with various Ga doping concentration. All the films show well defined wurtzite ZnO phase with no other impurity crystal phases. Undoped ZnO film exhibits a strong c-axis texture with only (0 0 2) diffraction in XRD scan. With the doping of Ga, some other peaks from different planes emerge. Here we introduce a texture coefficient T(hkl) to quantitatively represent the degree of preferred orientation (texture), which is defined as [29].

T ðhklÞ ¼

n Im ðhklÞ 1 X Im ðhklÞ  I0 ðhklÞ n 1 I0 ðhklÞ

Fig. 1. XRD h/2h scan of GZO thin films with various Ga content. All the peaks belong to wurtzite ZnO phase.

ð1Þ

where Im(hkl) is the measured relative intensity of the reflection from the (hkl) plane, I0(hkl) is that from the same plane in a standard reference sample (JCPDS 36-1451), and n is the total number of reflection peaks from the film. In present analysis, n = 6 since 6 major directions are involved (100, 002, 101, 102, 110, and 103). For the undoped ZnO film, the T(0 0 2) reaches the maximum value of 6 since only (0 0 2) peak detected in XRD. The dependence of (0 0 2) diffraction intensity, full width at half maximum (FWHM), and the (0 0 2) texture coefficient (T(0 0 2)) on the Ga content is shown in Fig. 2. The GZO thin films with Ga content below 6 at.% exhibit relatively good crystal quality with strong (0 0 2) preferred orientation (T(0 0 2) > 5.5). The FWHM increases with Ga doping, which indicates that the crystallite size along (0 0 2) plane becomes smaller. With the further increase of Ga concentration (>6 at.%), the (0 0 2) diffraction intensity significantly decreased by more than 10 times. Meanwhile the (0 0 2) texture degrades significantly with the increasing Ga content (>6 at.%). The degradation of the crystal quality in high density Ga doped films originates from the residual stress, distortion and dislocations owing to the difference in ionic radii between Zn (0.074 nm) and Ga (0.062 nm) [30,31]. The 4  4 lm2 AFM images for GZO films with various Ga contents are shown in Fig. 3. The films show densely packed polycrystalline structures. The crystallite size decreases with the increasing Ga doping concentration from 0 to 7.2 at.%, which is in good agreement with the XRD FWHM analysis. The crystallite size of GZO film

Fig. 2. Dependence of (0 0 2) texture coefficient T(0 0 2), (0 0 2) diffraction intensity and FWHM on the Ga content of GZO thin films.

with the highest Ga concentration of 10.5 at.% is not uniform, and the crystallite shape is not very clear due to the poor crystal quality. The surface root-mean-square roughness (rms) becomes higher with the increase of Ga concentration. The reason is that high concentration Ga doping degrades the (0 0 2) texture and thus the surface is composed of different oriented crystallites. Relative smooth and uniform GZO film can be obtained under a moderate Ga doping concentration. The rms is below 10 nm for the 600 nm-thick film containing 4.9 at.% Ga. Fig. 4 shows the dependence of resistivity, carrier concentration and Hall mobility of GZO films on the Ga content. Undoped ZnO shows n-type conductivity due to the intrinsic donor defects such

770

J.L. Zhao et al. / Optical Materials 33 (2011) 768–772

Fig. 3. 4  4 lm2 AFM images of GZO films with different Ga doping concentration (at.%).

Fig. 4. Dependence of resistivity (solid square), carrier concentration (solid triangle) and Hall mobility (open triangle) on the Ga content of GZO films. Optimized Ga content window and the best Ga doping concentration were indicated in the picture.

as oxygen vacancy (VO), but the resistivity is too high (0.2 X cm) to be used as transparent conductor. With the Ga doping, the resistivity significantly decreases and the carrier concentration increases because the GaZn acts as a more effective donor than VO in ZnO. The resistivity is below 8  104 X cm for GZO films with the Ga content in range of 4–7.2 at.%, which provides an useful Ga doping window for producing GZO based TCOs. The lowest resistivity achieved in this study is 3.6  104 X cm (carrier con-

centration of 1.7  1021 cm3, mobility of 10.1 cm2 V1 s1), which was realized in the film containing 4.9 at.% Ga. It is also found that with the further increase of Ga content (>7.2 at.%), the resistivity increases while the carrier concentration and mobility decrease. As discussed previously, the crystallinity degrades significantly with high content Ga doping and the high density structural defects can trap and scatter the free electrons released by GaZn donor, which results in the decrease of carrier concentration and mobility in the high Ga content film. The resistivity achieved in this study is lower than some previously reported values for the ZnO based TCO films by MOCVD [1,2,15]. Some reports [19,20] presented a lower resistivity for the Ga-doped ZnO films on glass, but the transmittance (85%) of their films is lower than our GZO film (90%). Fig. 5 shows UV–VIS-NIR transmittance as well as reflectance spectra of GZO films with different Ga content. The inset in Fig. 5 illustrates a photo image of GZO film deposited on a 5  5 cm2 glass, which shows good transparency. The averaged transmittance at 400–800 nm for GZO films is 90% (exclude the glass substrate). The undoped ZnO is also highly transparent in NIR range, while the transparency decreases and the reflectance increases with the Ga doping. The 4.9 at.% Ga contained film shows the lowest NIR transmittance (<1% at 2500 nm) and highest NIR reflectance (>70% at 2500 nm). According to the Drude theory [5,32], the infrared transmittance cut-off behavior is due to the forbidden propagation of photons with energy  hx < h  xp in materials, where  h is the Plank’s constant, x is the frequency of a photon, and xp is the resonance plasma frequency determined by the following equation,

J.L. Zhao et al. / Optical Materials 33 (2011) 768–772

771

up to 2500 nm (the upper limit of our existing spectrometer), we expect that the reflectance will be even higher in the 3–40 lm mid-infrared region (solar heat range) [32], which is promising for the application of IR-reflective and sunshade coatings for energyefficient windows. In the near UV region (<400 nm), the GZO films exhibit a sharp absorption edge corresponding to the near-band-edge excitation. As a direct band gap semiconductor, the absorption coefficient (a) obeys the following relation for high photon energies (hm > Eg) [33],

a2 ¼ Aðhm  Eg Þ

Fig. 5. Transmittance and reflectance spectra of GZO films with various Ga doping concentration (at.%). The inset illustrates a photo of the GZO film on a 5  5 cm2 glass, which shows good transparency.

x2p ¼ ne e2 =e0 e1 me  c2

ð2Þ

c ¼ e=ðme lÞ

ð3Þ

where e0 is the permittivity of free space, e1 is the high frequency dielectric constant, me is the effective electron mass, ne is the free electron concentration, and l is the electron mobility. Usually xp  c [5], and thus,

x2p  ne e2 =e0 e1 me

ð4Þ

Correspondingly, the resonance wavelength kp is determined as

kp ¼ 2pc0 =xp

ð5Þ

where c0 is the light velocity in free space. In order to achieve high IR reflective GZO films, we can increase xp or decrease kp to shift the cut-off wavelength into the NIR region. According to Eqs. (4) and (5), increasing carrier concentration ne can effectively shift the kp to shorter wavelength, which indicates that higher IR reflective performance is expected in the GZO film with higher carrier concentration. Fig. 6 shows the transmittance and reflectance at 2500 nm as functions of carrier concentration of the GZO film with 4.9 at.% Ga. The transmittance decreases, while reflectance increases with the increase of carrier concentration, which is in good agreement with the aforementioned theoretical analysis. The GZO film with Ga content of 4.9 at.% has the highest carrier concentration (1.7  1021 cm3), and thus the highest infrared reflectance (70% at 2500 nm). Although we only give the reflectance spectra

Fig. 6. Dependence of the transmittance (solid square) and reflectance (solid triangle) at 2500 nm on the carrier concentration of GZO films.

ð6Þ

where Eg is the optical band gap of thin film, and A is an constant. The a2  hm curves for the GZO films with different Ga content are plotted in Fig. 7. Optical band gap Eg can be evaluated by extrapolation of the linear part to ‘a = 0’. It can be seen that the optical band gap shifts to higher energy with the increase of the carrier concentration in GZO films (inset of Fig. 7), which is due to the well-known Burstein–Moss (BM) shift [34,35] in heavily doped semiconductors. The thermal stability of conductivity was investigated using GZO films (4.9 at.% Ga) on quartz substrates. Fig. 8 shows the electrical properties of GZO films annealed at various temperatures from 400 to 600 °C. The as-grown film (300 °C sample in Fig. 8) exhibits a resistivity of 3.8  104 X cm, which is similar to the value of the aforementioned sample on glass. The GZO film annealed at a higher temperature shows higher resistivity and lower carrier concentration. ZnO based films can absorb the oxygen atoms under high temperature oxidation environment. The absorbed oxygen prefers to stay at grain boundaries [12,36], and thus induces the acceptor-like interface states, which will trap the free electrons and reduce the carrier concentration. Furthermore, the absorbed oxygen also can reduce the oxygen vacancy density in the film, which also reduces the carrier concentration in the film. However, it is found that our GZO films have relative good thermal stability. The resistivity of films annealed at temperatures below 500 °C shows little change. The film annealed at 500 °C still exhibits a low resistivity of 6  104 X cm, which is only 1.6 times higher than the resistivity of as-grown film. It is also found that the annealed film retains high transmittance (90%) at visible range. The GZO film developed in this study has better thermal stability than the previously reported ZnO based TCOs (up to 400 °C) [13,37]. The previous thermal stability investigations are based on the AZO/GZO by sputtering and PLD methods under a high vacuum environment (oxygen pressure <0.1 Torr). In comparison, our GZO film was deposited by MOCVD at a relative high oxygen pressure (7 Torr). Therefore, it is suggested that our film exhibit better

Fig. 7. Plot of a2 vs. hm for GZO films with various doping concentration (at.%). The inset shows the effect of the carrier concentration on the optical band gap derived from the a2 vs. hm curve.

772

J.L. Zhao et al. / Optical Materials 33 (2011) 768–772

Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.optmat.2010.12.008. References

Fig. 8. Electrical properties, including resistivity (solid square), carrier concentration (solid triangle) and Hall mobility (open triangle) of GZO thin films annealed at various temperatures. The as-grown sample is shown as 300 °C. Our sample retains high conductive performance at high temperature (up to 500 °C).

stoichiometry with less oxygen vacancies, which results in better resistance to oxygen absorption under high temperature. 4. Conclusions In summary, highly transparent conductive GZO thin films have been realized by metal organic chemical vapor deposition at a low deposition temperature of 300 °C. The Ga doping concentration was optimized at 4.9 at.%, which was employed to achieve c-axis textured film with the lowest resistivity of 3.6  104 X cm and transmittance up to 90% averaged at the visible region. The GZO film was also found to be high reflective (>70% at 2500 nm) to the infrared radiation. Moreover, the GZO film has been proved to have excellent thermal stability (up to 500 °C). The proposed reason for the improved thermal stability is ascribed to the reduced oxygen defects in the film deposited at relative high oxygen pressure environment. Our developed high quality GZO thin films are promising for the transparent electronics (especially for devices operating at higher temperature) and energy-efficient window applications. Acknowledgement This research is supported by the MKE (the Ministry of Knowledge Economy), Korea, under the ITRC(Information Technology Research Center) support program supervised by the NIPA (National IT Industry Promotion Agency) (NIPA-2010-C10901021-0015). The work is also supported by the National Natural Science Foundation of China (NSFC) (Project No. 61006037).

[1] T.-Y. Park, Y.-S. Choi, J.-W. Kang, J.-H. Jeong, S.-J. Park, D.M. Jeon, J.W. Kim, Y.C. Kim, Appl. Phys. Lett. 96 (2010) 051124. [2] M.L. Addonizio, C. Diletto, Sol. Energy Mater. Sol. Cells 92 (2008) 1488. [3] J.C. Bernèdea, Y. Berredjem, L. Cattin, M. Morsli, Appl. Phys. Lett. 92 (2008) 083304. [4] J.J. Berry, D.S. Ginley, P.E. Burrows, Appl. Phys. Lett. 92 (2008) 193304. [5] J.K. Sheu, M.L. Lee, C.J. Tun, S.W. Lin, Appl. Phys. Lett. 88 (2006) 043506. [6] I. Hamberg, C.G. Granqvist, J. Appl. Phys. 60 (1986) R123. [7] L. Gong, Z. Ye, J. Lu, L. Zhu, J. Huang, X. Gu, B. Zhao, Vaccum 84 (2010) 947. [8] Q.-B. Ma, Z.-Z. Ye, H.-P. He, Y. Luo, L.-P. Zhu, J.-Y. Huang, Y.-Z. Zhang, B.-H. Zhao, ChemPhysChem 9 (2008) 529. [9] T. Minami, Thin Solid Films 516 (2008) 5822. [10] T. Minami, Thin Solid Films 516 (2008) 1314. [11] S. Liang, X. Bi, J. Appl. Phys. 104 (2008) 113533. [12] O. Nakagawara, Y. Kishimoto, H. Seto, Y. Koshido, Y. Yoshino, T. Makino, Appl. Phys. Lett. 89 (2006) 091904. [13] J.A. Sans, G. Martínez-Criado, J. Pellicer-Porres, J.F. Sánchez-Royo, A. Segura, Appl. Phys. Lett. 91 (2007) 221904. [14] S.-M. Park, T. Ikegami, K. Ebihara, Thin Solid Films 513 (2006) 90. [15] Y.-C. Huang, Z.-Y. Li, L.-W. Weng, W.-Y. Uen, S.-M. Lan, S.-M. Liao, T.-Y. Lin, Y.H. Huang, J.-W. Chen, T.-N. Yang, C.-C. Chiang, J. Vac. Sci. Technol. A 27 (2009) 1260. [16] J. Nomoto, M. Konagai, K. Okada, T. Ito, T. Miyata, T. Minami, Thin Solid Films 518 (2010) 2937. [17] B. Bayraktaroglu, K. Leedy, R. Bedford, Appl. Phys. Lett. 93 (2008) 022104. [18] J.H. Lee, C.-Y. Chou, Z. Bi, C.-F. Tsai, H. Wang, Nanotechnology 20 (2009) 395704. [19] Y. Li, G.S. Tompa, S. Liang, C. Gorla, Y. Lu, J. Doyle, J. Vac. Sci. Technol. A 15 (1997) 1063. [20] B. Hahn, G. Heindel, E. Pschorr-Schoberer, W. Gebhardt, Semicond. Sci. Technol. 13 (1998) 788. [21] N. Nishimoto, T. Yamamae, T. Kaku, Y. Matsuo, K. Senthilkumar, O. Senthilkumar, J. Okamoto, Y. Yamada, S. Kubo, Y. Fujita, J. Cryst. Growth 310 (2008) 5003. [22] J.D. Ye, S.L. Gu, S.M. Zhu, S.M. Liu, Y.D. Zheng, R. Zhang, Y. Shi, H.Q. Yu, Y.D. Ye, J. Cryst. Growth 283 (2005) 279. [23] Y.-C. Huang, Z.-Y. Li, H.-H. Chen, W.-Y. Uen, S.-M. Lan, S.-M. Liao, Y.-H. Huang, C.-T. Ku, M.-C. Chen, T.-N. Yang, C.-C. Chiang, Thin Solid Films 517 (2009) 5537. [24] H.Y. Liu, V. Avrutin, N. Izyumskaya, M.A. Reshchikov, Ü. Özgür, H. Morkoç, Phys. Status Solidi (RRL) 4 (2010) 70. [25] D.R. Sahu, J.–L. Huang, Sol. Energy Mater. Sol. Cells 93 (2009) 1923. [26] M.A. Kaida, A. Ashour, Appl. Surf. Sci. 253 (2007) 3029. [27] K.-M. Lin, Y.-Y. Chen, J. Sol–Gel Sci. Tech. 51 (2009) 215. [28] R. Gordon, J. Non-Cryst. Solids 218 (1997) 81. [29] J.L. Zhao, X.M. Li, S. Zhang, C. Yang, X.D. Gao, W.D. Yu, J. Mater. Res. 21 (2006) 2185. [30] W.J. Park, H.S. Shin, B.D. Ahn, G.H. Kim, S.M. Lee, K.H. Kim, H.J. Kima, Appl. Phys. Lett. 93 (2008) 083508. [31] J.L. Zhao, X.W. Sun, S.T. Tan, IEEE Trans. Electron Dev. 56 (2009) 2995. [32] V.A. Dao, T. Le, T. Tran, H.C. Nguyen, K. Kim, J. Lee, S. Jung, N. Lakshminarayan, J. Yi, J Electroceram 23 (2009) 356. [33] J.L. Zhao, X.M. Li, J.M. Bian, W.D. Yu, X.D. Gao, J. Cryst. Growth 276 (2005) 507. [34] E. Burstein, Phys. Rev. 93 (1954) 632. [35] T.S. Moss, Proc. Phys. Soc. London, Sect. B 67 (1954) 775. [36] T. Minami, T. Kuboi, T. Miyata, Y. Ohtani, Phys. Status Solidi (a) 205 (2008) 255. [37] B.D. Ahn, J.H. Kim, H.S. Kang, C.H. Lee, S.H. Oh, K.W. Kim, G. Jang, S.Y. Lee, Thin Solid Films 516 (2008) 1382.