Effects of annealing on optical properties of Zn-implanted ZnO thin films

Effects of annealing on optical properties of Zn-implanted ZnO thin films

Journal of Alloys and Compounds 458 (2008) 569–573 Effects of annealing on optical properties of Zn-implanted ZnO thin films S.W. Xue a,b , X.T. Zu a...

915KB Sizes 0 Downloads 38 Views

Journal of Alloys and Compounds 458 (2008) 569–573

Effects of annealing on optical properties of Zn-implanted ZnO thin films S.W. Xue a,b , X.T. Zu a,∗ , L.X. Shao b , Z.L. Yuan a , W.G. Zheng d , X.D. Jiang d , H. Deng c a

Department of Applied Physics, University of Electronic Science and Technology of China, Chengdu 610054, PR China b Department of Physics, Zhanjiang Normal College, Zhanjiang 524048, PR China c School of Microelectronics and Solid-state Electronics, University of Electronic Science and Technology of China, Chengdu 610054, PR China d Laser Fusion Research Center, Chinese Academy of Engineering Physics, Mianyang 621900, PR China Received 25 February 2007; received in revised form 11 April 2007; accepted 15 April 2007 Available online 21 April 2007

Abstract Zn-ion-implantation to a dose of 1 × 1017 ions/cm2 was performed on ZnO thin films deposited on fused silica glass substrates by the sol–gel technique. After ion implantation, Zn-implanted ZnO films were annealed in air at different temperatures from 500 to 900 ◦ C. The effects of ion implantation and post-thermal annealing on the structural and optical properties of the ZnO films were investigated by X-ray diffraction (XRD), photoluminescence (PL) and optical absorption measurements. XRD reveals that diffraction peaks recover at ∼700 ◦ C. Optical absorption measurements show that the absorption edge blueshifts when the annealing temperature is below 600 ◦ C while redshifts when the annealing temperature exceeds 600 ◦ C. Urbach energy decreases with increasing the annealing temperature from 500 to 600 ◦ C while increases from 600 to 900 ◦ C. PL results showed that both near band edge (NBE) excitonic UV emission and defect related deep level emission (DLE) increased with increasing annealing temperatures from 500 to 900 ◦ C. © 2007 Elsevier B.V. All rights reserved. Keywords: Thin films; Sol–gel process; Optical properties

1. Introduction Zinc oxide (ZnO), a wide bandgap semiconductor, has received increasing attention in the research community due to its potential applications in optoelectronic devices in the blue and ultraviolet region. In recent years, many techniques have been employed to synthesize high-quality ZnO films, such as radio frequency (RF) magnetron sputtering, chemical vapor deposition, pulsed laser deposition, plasma-assisted molecular beam epitaxy (MBE) and the sol–gel process [1–5]. In order to develop materials for special application, doped ZnO films have been fabricated and investigated by many groups. Ion implantation, especially in the semiconductor industry, is a widespread tool for doping semiconductors. It can be used to fabricate doped ZnO films. Advantages of ion implantation are the lateral selectivity of sample area and doping depth as well as an accurate dose control. In principle, every element from the periodic system can be implanted, which makes it a multi-



Corresponding author. Tel.: +86 28 83201939; fax: +86 28 83201939. E-mail address: [email protected] (X.T. Zu).

0925-8388/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.jallcom.2007.04.239

purpose and flexible tool for doping semiconductors. However, undesirable side effects are the implantation-induced damage and its effects on the structural and optical electrical properties of the semiconductor. Zhao et al. have investigated the recombination mechanisms of deep level emission (DLE) in ZnO by adding excess Zn through ion implantation [6], but further studies on the influences of ion implantation and post-thermal annealing on the structural and optical properties of Zn-implanted ZnO films are not done. In this work, we report Zn ion implantation into ZnO films by the sol–gel process. The influences of ion implantation and post-thermal annealing on the structural and optical properties of Zn-implanted ZnO films are investigated. 2. Experimental 2.1. Preparation of ZnO films and Zn-ion-implantation ZnO thin film deposited on sapphire substrates was prepared by the sol–gel method. Zinc acetate dehydrate (Zn(CH3 COO)2 ·H2 O) was used as a starting material. Monoethanolamine (MEA) and 2-methoxyethanol were used as stabilizer and solvent, respectively. Zinc acetate dihydrate was first dissolved

570

S.W. Xue et al. / Journal of Alloys and Compounds 458 (2008) 569–573

Fig. 1. TRIM simulation of the depth profile of implanted Zn atoms for ZnO films. in a mixture of 2-methoxyethanol and MEA solution at room temperature. The molar ratio of MEA to zinc acetate (Zn(CH3 COO)2 ) was maintained at 1.0 and the concentration of zinc acetate was 0.35 M. The solution was stirred at 60 ◦ C for 2 h to yield a clear and homogeneous solution, which served as the coating solution after cooling to room temperature. The coating was usually made 24 h after the solution was prepared. The solution was dropped onto fused silica glass substrates, which were rotated at 4000 rpm for 30 s. After depositing by spin coating, the films were dried at 300 ◦ C for 10 min over a hot plate to evaporate the solvent and remove organic residuals. The procedures from coating to drying were repeated twelve times until the thickness of the sintered films were approximately 800 nm which was measured by an OLYMPUS BX51 interference/transmission microscopy. The films were then put into a quartz tube furnace and annealed in air at 500 ◦ C for 1 h. The as-prepared films were subjected to ion implantation. Zn ions were implanted into the as-prepared films at 56 keV to a dose of 1 × 1017 ions/cm2 . Fig. 1 shows the depth profile of implanted Zn atoms simulated using the code TRIM 96. The projected range (Rp ) of the atoms was about 26.7 nm. After ion implantation, Zn-implanted films were annealed in air at different temperatures from 500 to 900 ◦ C for 1 h in a quartz tube furnace.

2.2. Measurements

Fig. 2. AFM images of as-prepared ZnO films.

films exhibit the hexagonal wurtzite structure (JCPDS 36-1451) without preferred orientation. 3.2. Influences of ion implantation and post-thermal annealing 3.2.1. Structural properties To investigate the effects of annealing on the structural properties of Zn-implanted ZnO films, the films were annealed in air at different temperatures from 500 to 900 ◦ C for 1 h after ion implantation. Fig. 4 shows XRD patterns of ZnO films after ion implantation and annealing at different temperatures. We think that Zn nanoparticles or clusters may form in ZnO films after so high-dose Zn-ion-implantation [8,9], but no evidence of any other secondary phases and impurities are found in the XRD patterns. It can be seen that all films still show hexagonal wurtzite structure after ion implantation and annealing. As ZnO is a radiation-hard material [10], it still remains crystalline after high-dose ion implantation. However, the intensities of diffraction peaks are distinctly decreased by ion implanta-

To characterize the structural and optical properties of the as-prepared and Zn-implanted films, the crystal orientation was investigated using a PHILIPS X’PERT PRO MPD X-ray diffractometer (XRD) with the radiation source of Cu K␣ . Optical absorbance was measured using a SHIMADZU UV2550 spectrophotometer. Room temperature (RT) photoluminescence spectra (PL) were recorded by a SHIMADZU RF5301PC spectrophotometer with excitation wavelength of 345 nm.

3. Results and discussion 3.1. Characterization of as-prepared ZnO films Fig. 2 shows typical AFM images of as-prepared films deposited on fused silica glass substrates by the sol–gel process. The minimum root-mean square of surface roughness is about 5.2 nm. The average grain size is ∼65 nm. Compared with RF sputtered ZnO films [7], the surface of as-prepared films is very smooth. X-ray diffraction patterns of as-prepared ZnO films are shown in Fig. 3. It shows that as-prepared ZnO

Fig. 3. XRD pattern of as-prepared ZnO films.

S.W. Xue et al. / Journal of Alloys and Compounds 458 (2008) 569–573

Fig. 4. XRD patterns of ZnO films after ion implantation and post-thermal annealing at different temperatures.

tion. The decreases of diffraction peaks can be attributed to ion-beam induced lattice disorder. Thermal annealing has great influences on the structural properties of Zn-implanted films. Results show that diffraction peaks increase with increasing annealing temperature from 500 to 700 ◦ C. ZnO films exhibit c-axis preferred orientation after annealing at 500 ◦ C for 1 h. The other peaks increase dramatically with further annealing below 700 ◦ C. However, the intensities of all peaks decrease with increasing annealing temperature when it exceeds 700 ◦ C. This may be due to the thermal induced lattice disorder at high temperature. It is noted that significant recovery of the structural properties of Zn-implanted ZnO films occurs after annealing at 700 ◦ C. 3.2.2. Optical properties Fig. 5 shows the influences of ion implantation and thermal annealing on the optical absorption of ZnO films. It can be seen from Fig. 5 that the optical absorption was enhanced distinctly in the visible region after Zn-ion-implantation. Subsequent anneal-

Fig. 5. Optical absorption spectra of ZnO films after ion implantation and postthermal annealing at different temperatures.

571

ing also has evident influence on the optical absorption of Zn-implanted ZnO films. It is found that the optical absorption in the visible region decreases with increasing annealing temperature from 500 to 600 ◦ C. However, the optical absorption in the visible region increases dramatically with increasing annealing temperature from 600 to 900 ◦ C, which is similar to that reported in our earlier studies on the annealing of ZnO:Al films [10]. It is known that the formation of nanoparticles/clusters during highdose ion implantation may effectively affect optical absorption. Several authors [11,12] have observed evident surface plasmon peaks in optical absorption spectra after high-dose ion implantation, which is attributed to the effect of nanoparticles/clusters formed during ion implantation. In our experiment, there is no clear evidence of plasmon peaks in the optical absorption spectra. The results of XRD and optical absorption spectra suggest that Zn nanoparticles/clusters seem to be not easy to form in ZnO films during Zn-ion-implantation. The underlying mechanism is not clearly understood at present. The influence of Zn nanoparticles/clusters on the optical absorption can be neglected. It is known that the changes of carrier concentration may also affect the optical absorption. It is known that non-doped ZnO is n-type conducting oxide. We have expected that Zn-ion-implantation may increase the carrier concentration in ZnO films by adding excess Zn interstitials (Zni ). However, recent studies reveal that Zni is a fast diffuser, and is unlikely stable in n-type ZnO [13]. We have measured the resistivity after ion implantation and thermal annealing with the four-point probe method. All films exhibit high resistivity. We think that the changes of carrier concentration cannot evidently affect the optical absorption in our experiments. By ruling out the influences of Zn nanoparticles/clusters and carrier concentration, we think that the observed behaviors of optical absorption may be mainly due to the changes of defect concentrations in ZnO films after ion implantation and thermal annealing. As discussed in the structural analysis, many defects may be produced during ion implantation and high-temperature annealing. These defects may serve as trapped electron centers and can effectively affect optical absorption [14,15]. It is believed that ion-beam induced defects enhance the optical absorption. These defects are gradually annealed out during subsequent annealing from 500 to 600 ◦ C, which results in the decrease of optical absorption. However, thermal induced defects increase dramatically during the following annealing stage between 700 and 900 ◦ C. This results in the increase of the optical absorption again. It is noted that the optical absorption in the visible region nearly recovers after annealing at 600 ◦ C, indicating that most of the ion-beam-induced defects are annealed out. This is essentially in agreement with XRD analysis. In addition, Zn-ion-implantation and post-thermal annealing also have evident influences on the optical bandgaps of ZnO films as shown in the inset in Figs. 5 and 6. The optical bandgaps are determined from the conventional method [16]. It can be seen from Fig. 6 that the optical bandgap redshifts after Zn-ionimplantation. Subsequent annealing between 500 and 700 ◦ C makes the optical bandgap blueshift while redshift between 700 and 900 ◦ C. We also attribute the observed behaviors of the optical bandgaps mainly to the changes of defect concentrations in ZnO films. As mentioned above, many defects are produced

572

S.W. Xue et al. / Journal of Alloys and Compounds 458 (2008) 569–573

Fig. 6. Dependence of optical bandgaps of ZnO films on ion implantation and post-thermal annealing.

during Zn-ion-implantation. They may cause the redshift of the optical bandgap. Most of the ion-beam induced defects are annealed out by the subsequent annealing from 500 to 600 ◦ C. This results in the buleshift of bandgap due to the decrease of defect concentration in ZnO films. During the following annealing stage between 600 and 900 ◦ C, thermal induced defects at high temperatures increase dramatically, which leads to the redshift of the optical band edge again due to the increase of defect concentration with annealing temperatures. Ion implantation and high-temperature annealing may cause structural disorder in ZnO films. This can also be reflected in the optical absorption band tail and can be characterized by empirical Urbach rule. The slope of the Urbach tail (Urbach energy) characterizes the structural disorder in the material and can be used to evaluate the defect concentrations in ZnO films. It is known that absorption coefficient α(λ) near the band edge shows an exponential dependence on photon energy: [17]   hν α(λ) = α0 exp (1) E0

Fig. 7. Absorption coefficients of ZnO films after ion implantation and postthermal annealing at different temperatures.

Fig. 8. Dependence of ln[α(λ)] on photon energy.

where E0 is the Urbach energy. α0 is a constant. Thus, a plot of ln[α(λ)] versus photon energy should be linear and Urbach energy can be obtained from the slope. The dependence of α(λ) and ln[α(λ)] on wavelength are given in Figs. 7 and 8, respectively. α(λ) is obtained from the transmittance spectra (not shown here) using a simple model [18]: T = exp[−α(λ)d], where T is the optical transmittance and d is the thickness of ZnO films. Urbach energy can be calculated from the reciprocal gradient of the linear portion of the curves in Fig. 8 and is shown in Fig. 9. It can be seen that Urbach energy increases after Zn-ion-implantation and decreases with increasing annealing temperature from 500 to 600 ◦ C. This indicates that many defects are produced during Zn-ion-implantation and are gradually annealed out by the subsequent annealing at 500–600 ◦ C. Further annealing at 600–900 ◦ C makes Urbach energy increase again, indicating that thermal induced defects increase with increasing annealing temperatures. Urbach energy dependence on annealing temperatures agrees well with the analysis above. Fig. 10 shows RT PL spectra of ZnO films after ion implantation and annealing at different temperatures from 500 to 900 ◦ C.

Fig. 9. Urbach energy as a function of annealing temperatures.

S.W. Xue et al. / Journal of Alloys and Compounds 458 (2008) 569–573

573

implantation annealing, ion implantation induced non-radiative defects can be effectively annealed out, which can lead to the recovery of the structural and optical properties of Zn-implanted ZnO films. XRD intensities recover at ∼700 ◦ C. Optical absorption in the visible region nearly recovers at ∼600 ◦ C. With increasing annealing temperature, both NBE and DLE increase. Acknowledgements This work was supported by the Program for New Century Excellent Talents in University (NCET-04-0899), the Ph.D. Funding Support Program of Education Ministry of China (20050614013), the NSAF Joint Foundation of China (10376006) and Special Foundation for University Subject Construction of Department of Education of Guangdong Province under Project No. [2006] 112. Fig. 10. RT PL spectra of ZnO films after ion implantation and post-thermal annealing at different temperatures (excitation wavelength: 345 nm).

The shape of all the spectra, similar to those reported by others, is featured by a near band edge (NBE) excitonic UV emission and a defect related deep level emission in the visible region [19,20]. It can be seen that NBE and DLE are decreased after Zn-ion-implantation. The decreases are mainly due to the ion-beam induce defects which serves as non-radiative centers [21]. Compared with Ref. [21], it is noted that PL emission is not completely extinguished in our experiment. We have observed similar results in Ge-implanted ZnO films deposited on glass slides by sol–gel process. We think that this difference may be due to different implantation conditions and properties of the material used in the experiments (single crystal ZnO was used in Ref. [21]). Subsequent annealing has pronounced influences on the emission properties. Both NBE and DLE increase with increasing annealing temperatures from 500 to 900 ◦ C. 4. Conclusion Zn-implanted ZnO films are prepared on fused silica glass substrates. Zn-ion-implantation to a dose of 1 × 1017 ions/cm2 has caused structural and optical changes to ZnO thin films. After Zn-ion-implantation, ZnO films still remains single phase and hexagonal wurtzite structure. XRD intensities and PL emissions are decreased by Zn-ion-implantation. Through post-

References [1] T.L. Yang, D.H. Zhang, J. Ma, et al., Thin Solid Films 326 (1998) 60. [2] Y. Zhou, P.J. Kelly, A. Postill, et al., Thin Solid Films 447 (2004) 33. [3] T.M. Barnes, J. Leaf, C. Fry, C.A. Wolden, J. Crystal Growth 274 (2005) 412. [4] J. Mass, P. Bhattacharya, R.S. Katiyar, Mater. Sci. Eng. B 103 (2003) 9. [5] V. Musat, B. Teixeira, E. Fortunato, et al., Surf. Coat. Technol. 180 (2004) 659. [6] Q.X. Zhao, P. Klason, M. Willander, et al., Appl. Phys. Lett. 87 (2005) 211912. [7] H.W. Kim, N.H. Kim, Mater. Sci. Eng. B 103 (2003) 297. [8] X. Xiang, X.T. Zu, J.W. Bao, et al., J. Appl. Phys. 98 (2005) 073524. [9] X. Xiang, X.T. Zu, S. Zhu, et al., Nucl. Instrum. Methods Phys. Res. B 250 (2006) 192. [10] S.W. Xue, X.T. Zu, W.G. Zheng, et al., Phys. B 382 (2006) 201. [11] K. Kono, S.K. Arora, N. Kishimoto, et al., Nucl. Instrum. Methods Phys. Res. B 206 (2003) 291. [12] X. Xiang, X.T. Zu, S. Zhu, et al., Appl. Phys. Lett. 84 (1) (2004) 52. [13] A. Janotti, C.G. Van de Walle, J. Crystal Growth 287 (2006) 58. [14] J.H. Mackey, H.L. Smith, A. Halperin, J. Phys. Chem. Solids 27 (1966) 1759. [15] A.J. Cohen, G.G. Janezic, Phys. Stat. Sol. A 77 (1983) 619. [16] N. Kenny, C.R. Kannewurf, D.H. Whitmore, J. Phys. Chem. Solids 27 (1966) 1237. [17] F. Urbach, Phys. Rev. 92 (1953) 1324. [18] K.H. Kim, K.C. Park, D.Y. Ma, J. Appl. Phys. 81 (1997) 7764. [19] X.D. Gao, X.M. Li, W.D. Yu, et al., Mater. Res. Bull. 40 (2005) 1104. [20] R.J. Hong, J.D. Shao, H.B. He, et al., J. Crystal Growth 290 (2006) 334. [21] I. Sakaguchi, D. Park, Y. Takata, et al., Nucl. Instrum. Methods Phys. Res. B 206 (2003) 153.