Structural and optical properties of ZnO nanorods ...

TSF-32806; No of Pages 4 Thin Solid Films xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

Thin Solid Films journal homepage: www.elsevier.com/locate/tsf

Structural and optical properties of ZnO nanorods grown chemically on sputtered GaN buffer layers R. Nandi a, Pranav Joshi a, Devendra Singh b, Pravanshu Mohanta b, R.S. Srinivasa c, S.S. Major a,⁎ a b c

Department of Physics, Indian Institute of Technology Bombay, Mumbai 400076, India Centre for Research in Nanotechnology and Science, Indian Institute of Technology Bombay, Mumbai 400076, India Department of Metallurgical Engineering and Materials Science, Indian Institute of Technology Bombay, Mumbai 400076, India

a r t i c l e

i n f o

Available online xxxx Keywords: ZnO GaN Nanorods Chemical bath deposition Photoluminescence

a b s t r a c t ZnO nanorods were grown on 200 nm thick sputtered ZnO and GaN buffer layers on quartz substrates by chemical bath deposition. Field emission scanning electron microscopy and X-ray diffraction studies show that the ZnO nanorods on GaN buffer layer possess larger diameter and smaller lengths and are vertically misaligned, compared to those grown on ZnO buffer layer. These differences are attributed to lack of complete c-axis orientation of crystallites in GaN buffer layer, its lattice mismatch with that of ZnO and a hindered nucleation process of ZnO on GaN surface, owing to a finite nucleation barrier and limited surface diffusion. Photoluminescence spectrum of ZnO nanorods on GaN buffer layer, however, exhibits a much stronger near-band-edge luminescence and drastically suppressed defect luminescence compared to the luminescence spectrum of the nanorods grown on ZnO buffer layer. Deconvolution of the photoluminescence peaks and Raman studies indicate significant reduction of oxygen vacancies and gallium incorporation in the ZnO nanorods grown on GaN buffer layer. These observations suggest the possibility of exchange reaction mediated by the aqueous medium, particularly during the initial stages of growth. © 2013 Elsevier B.V. All rights reserved.

1. Introduction Quasi one-dimensional nanostructures, such as nanorods, nanowires and nanobelts have attracted great attention due to their potential applications in electronic as well as optoelectronic devices [1,2]. In addition, the effects due to large surface area to volume ratios and quantum confinement seen in nanostructured materials make them more attractive for these applications, compared to their bulk counterparts [3]. Among a variety of nanostructures, vertically aligned ZnO nanorods (ZnO-NRs) are being widely studied for applications in light emitters, field emitters, field effect transistors, sensors and solar cells, owing to the wide direct band gap (3.37 eV) and large exciton binding energy (60 meV) of ZnO [1,4,5]. Several growth processes such as metal organic chemical vapor deposition (MOCVD), pulsed laser deposition, thermal evaporation, and chemical bath deposition (CBD) have been used for the growth of vertically aligned ZnO-NRs. Among these growth methods, CBD offers several advantages, such as, simplicity of the equipment, low temperature, low cost and hazard-free operation. Furthermore, ZnO-NRs can be grown on a variety of substrates, such as amorphous, crystalline and polymers, by using an appropriate seed layer of ZnO on these substrates. The junction between the seed layer and ZnO-NRs has also been utilized for device fabrication [6–10]. Among the ZnO based heterostructures, ZnO/GaN heterojunctions and ⁎ Corresponding author. Tel.: +91 22 25767567; fax: +91 22 25767552. E-mail address: [email protected] (S.S. Major).

related devices have been widely investigated, owing to the small lattice mismatch (~1.8%) between ZnO and GaN [7,11]. ZnO/GaN devices are expected to have higher light output efficiency compared to other heterostructured devices, because the refractive indices of ZnO and GaN (2.0 and 2.4, respectively) are close to each other [12]. Vertically aligned ZnO-NRs are usually grown on a ZnO buffer layer (ZnO-BL), initiated by a heterogeneous nucleation process. Chua et al. [13,14] have demonstrated epitaxial growth of ZnO-NRs by CBD on high quality, MOCVD grown GaN buffer layer (GaN-BL) on sapphire. They have shown that the epitaxial nanorods exhibit intense nearband-edge luminescence and much reduced defect luminescence. Jang et al. [15] have also shown that the epitaxially grown ZnO-NRs on GaN-BL over a sapphire substrate exhibit high optical quality. In almost all these reports, thick (2–3 μm) and high structural quality GaN, grown epitaxially on sapphire has been used as the buffer layer, which appears to influence the photoluminescence of ZnO-NRs. In this investigation, ZnO-NRs have been grown on a thin (~200 nm) and polycrystalline GaN-BL, sputter deposited on amorphous quartz substrates. The structural and optical properties of ZnO-NRs have been investigated to show that ZnO-NRs of high optical quality can be grown on thin and polycrystalline GaN-BL. 2. Experimental details ZnO-NRs were synthesized by chemical bath deposition on GaN-BL over quartz substrates. All the chemicals were commercially procured

0040-6090/$ – see front matter © 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.tsf.2013.10.152

Please cite this article as: R. Nandi, et al., Thin Solid Films (2013), http://dx.doi.org/10.1016/j.tsf.2013.10.152

2

R. Nandi et al. / Thin Solid Films xxx (2013) xxx–xxx

and used without further purification. Prior to the growth of ZnO-NRs, 200 nm thick GaN-BLs were deposited on quartz substrates by rf reactive sputtering of a GaAs target in 100% N2 at 700 °C. The details of the deposition of GaN have been reported earlier [16]. Preliminary experiments showed that the growth of ZnO-NRs on less than 200 nm thick GaN-BLs was non uniform. A uniform growth of ZnO-NRs was observed on 200–500 nm thick GaN-BLs and the morphology of the nanorods was found to be quite independent of the buffer layer thickness. ZnONRs were also grown on 200 nm thick sputtered ZnO-BL for comparison. The ZnO-BL was deposited by rf reactive sputtering of zinc target in 20% O2 at 400 °C. The details of the sputter deposition of ZnO have been reported earlier [17]. The thickness of the buffer layers was measured by surface profilometry. ZnO-NRs were grown on GaN-BL and ZnO-BL by the reaction of zinc nitrate [Zn (NO3)2. 6H2O] (0.05 M) and HMT (hexamethylenetetramine) (0.125 M). The growth of ZnO-NRs was carried out at a constant temperature of ~95 °C for ~8 h. After the growth, the samples were rinsed ultrasonically in deionized water to remove residual ions and precipitates, and then dried in air. ZnO-NRs were characterized by X-ray diffraction (XRD), field emission scanning electron microscopy (FESEM), Raman spectroscopy and photoluminescence (PL) measurements at room temperature. Powder XRD studies were performed in θ–2θ geometry with a PANalytical X'Pert PRO powder diffractometer using Cu-Kα radiation (λ = 0.154056 nm). Morphological studies of the ZnO-NRs were carried out with a JEOL model JSM-7600F FESEM at 10 kV accelerating voltage. Jobin-Yvon iHR550 monochromator and Kimmon laser (λ = 325 nm) were used for room temperature PL measurements in the wavelength range of 350–800 nm. The PL spectra were deconvoluted into multiple Gaussian peaks, using a non-linear curve fitting program. Jobin-Yvon Horiba HR800 confocal micro-Raman spectrometer equipped with Ar+ laser (514.5 nm) was used for Raman studies.

Fig. 1. FESEM images of ZnO-NRs grown on (a) ZnO-BL and (b) GaN-BL.

3. Results and discussion Fig. 1 shows the FESEM images of ZnO-NRs grown on ZnO-BL and GaN-BL. The ZnO-NRs are seen as well faceted, hexagonal rods with a pyramidal tip in both cases. The ZnO-NRs grown on ZnO-BL exhibit a high degree of vertical alignment and are ~3 μm long, having diameters in the range of 100–150 nm. In contrast, the ZnO-NRs grown on GaN-BL are misoriented and do not exhibit complete vertical alignment. The average length of ZnO-NRs on GaN-BL is found to be slightly smaller (~2.5 μm) and their diameters are substantially larger (150–350 nm). It is also found that the density of ZnO-NRs is much smaller and nonuniform on GaN-BL as compared to those on ZnO-BL. These features can be explained by noting that the growth of ZnO-NRs on a ZnO-BL is essentially homoepitaxial in character. The presence of a buffer layer or a seed layer of ZnO is known [18,19] to effectively lower the nucleation energy barrier, leading to heterogeneous nucleation at a lattice matched interface with practically negligible interface energy. In contrast, significant contributions from the surface energy of ZnO nuclei and ZnO-GaN interface energy vis-a-vis the surface energy of GaN substrate, cause a finite nucleation barrier, and hence a critical size of nuclei is required for the growth of ZnO-NRs on GaN-BL. It may be noted that the (0001) surface formation energy of GaN and the interface formation energy of ZnO on GaN have been reported [20] to be nearly equal. Assuming that the rates of chemical reaction and supersaturation are the same in both cases, the limited surface diffusion (at a relatively low growth temperature of 95 °C), appears to hinder the nucleation process of ZnO-NRs on GaN-BL, resulting in a relatively smaller density of critical nuclei and the smaller average length of ZnO-NRs on GaN-BL. The smaller density and larger diameter of ZnO-NRs on GaN-BL are seen to arise from the smaller density of nuclei, since their lateral growth is not effectively suppressed [21,22]. Fig. 2 shows the XRD patterns of ZnO-BL, GaN-BL, ZnO-NRs on ZnOBL and ZnO-NRs on GaN-BL. The XRD pattern of ZnO-BL (Fig. 2(a)) shows only two peaks, corresponding to (0002) and (0004) planes of hexagonal ZnO, indicating a strong and nearly complete c-axis orientation of crystallites in ZnO-BL. The XRD pattern of GaN-BL (Fig. 2(b)) shows a strong peak corresponding to (0002) plane along with a weak peak corresponding to 1011 plane of hexagonal GaN. In comparison

Fig. 2. XRD patterns of (a) ZnO-BL, (b) GaN-BL, (c) ZnO-NRs on ZnO-BL, and (d) ZnO-NRs on GaN-BL.



to the XRD peak of ZnO-BL, the XRD peaks of GaN-BL are much broader. These differences in the XRD patterns of ZnO-BL and GaN-BL indicate that the crystallites in GaN-BL are smaller in size and exhibit a lesser degree of c-axis orientation compared to those in ZnO-BL. The XRD patterns of ZnO-NRs on ZnO-BL and GaN-BL are shown in Fig. 2(c) and (d), respectively, which also exhibit significant differences. Fig. 2(c) shows an intense and narrow (0002) peak and a low intensity (0004) peak of hexagonal ZnO, indicating that the ZnO-NRs are almost completely oriented along [0002] direction and aligned vertically. In contrast, the XRD pattern of ZnO-NRs on GaN-BL exhibits several peaks, inaddition peak, which are indexed as to thedominant (0002) 1010 ; 1011 ; 1012 and 1013 planes of hexagonal ZnO. The presence of multiple peaks in the XRD pattern of ZnO-NRs on GaN-BL is attributed to the relatively poorer vertical alignment and misorientation of the ZnO-NRs, as evidenced from the corresponding FESEM image. The misorientation and misalignment of ZnO-NRs on GaN-BL are attributed to the lack of complete c-axis orientation of crystallites in GaN-BL. It may be noted that the growth of misaligned ZnO-NRs has been reported [23,24] on ZnO-BLs, having crystallites with multiple orientations. The small lattice mismatch between ZnO and GaN may also partly contribute to the misorientation of the epitaxially grown ZnO-NRs. Fig. 3(a) and (b) shows the room temperature PL spectra of ZnO-NRs grown on ZnO-BL and GaN-BL, respectively. ZnO-NRs on ZnO-BL exhibit a sharp UV emission peak at ~376 nm, attributed to near-band-edge emission and a broad visible emission at ~588 nm, usually attributed to emissions related to intrinsic point defects in ZnO lattice, such as, oxygen vacancy (Vo), zinc vacancy (VZn), oxygen interstitial (Oi), zinc interstitial (Zni), oxygen anti-site (OZn), and zinc anti-site (ZnO) [25]. The defect emission in this case is more intense and much broader compared to the near-band-edge emission. The near-band-edge emission exhibits a full width at half maximum (FWHM) of ~13 nm with a small asymmetry due to a possible contribution at higher wavelengths. Deconvolution of this peak, indicates the presence of two components, an intense and narrow peak at ~376 nm and a low intensity, broad

Fig. 3. Room temperature photoluminescence spectra of ZnO-NRs grown on (a) ZnO-BL and (b) GaN-BL. The deconvoluted components of the photoluminescence spectra are also shown. The inset shows the magnified (×15) defect emission in the visible region (450–700 nm) along with its deconvoluted components.

3

peak at ~388 nm, as shown in Fig. 3(a). The intense and narrow peak is attributed to the near-band-edge luminescence of ZnO and the weak shoulder is attributed to transitions from zinc interstitial levels to the valence band edge [26]. Deconvolution of the broad defect emission yields three components at 544 nm, 603 nm and 654 nm, as shown in Fig. 3(a). On the basis of a comparison with literature [27–30], the green emission peak at ~544 nm is attributed to transitions involving Vo, the yellow-orange emission at ~603 nm is attributed to transitions related to Oi and the red emission band at ~654 nm is attributed to transitions associated with Zni. Fig. 3(b) shows the room temperature PL spectrum of ZnO-NRs on GaN-BL along with the components of the peaks obtained by deconvolution. This PL spectrum exhibits significantly different features compared to those seen in Fig. 3(a). The striking differences are the near complete suppression of the defect related broad emission, and nearly five-fold increase in the intensity of the near-band-edge emission. The near-band-edge emission on deconvolution yields components due to band edge emission at ~376 nm and an emission related to Zni at ~388 nm. The FWHM of the near-band-edge emission in this case, is ~19 nm and the asymmetry seen on the higher wavelength side, attributed to Zni is more pronounced. For the purpose of clarity, the inset in Fig. 3(b) shows the magnified defect emission in the visible region, along with its deconvoluted components. The components of the broad defect emission from ZnO-NRs grown on GaN-BL are thus seen at 545 nm, 600 nm and 653 nm, which is in reasonable agreement with the deconvoluted peaks in Fig. 3(a). It is also seen that the relative contribution from the green emission attributed to Vo, to the overall defect emission is slightly smaller (29%), compared to the corresponding contribution (40%) in the case of ZnO-NRs on ZnO-BL. The drastic increase in the intensity of near-band-edge emission and strong suppression of defect emission indicate that the optical quality of ZnO-NRs grown on GaN-BL is much superior to that of ZnO-NRs grown on ZnOBL. Earlier investigations on CBD grown nanorods on GaN-BL have also shown the suppression of defect related emission [13–15]. However, as noted earlier, the GaN-BLs were MOCVD grown, high epitaxial quality films on sapphire substrates, having a thickness of 2–3 μm, as against the present case, wherein the GaN-BLs are thin (200 nm thick) and polycrystalline, and are deposited by sputtering on amorphous quartz substrates. In order to investigate the origin of the superior optical quality of ZnO-NRs on GaN-BL, Raman studies have been performed. Fig. 4(a) and (b) shows the Raman spectra of ZnO-NRs on ZnO-BL and GaN-BL, respectively. All the observed peaks have been attributed to vibrations of the wurtzite structure of ZnO, in line with published literature [31]. The dominant peak at 440 cm−1 seen for ZnO-NRs on both ZnO-BL and GaN-BL is attributed to E2 (high) mode, consistent with their preferred c-axis orientation. The peaks seen at 333 cm−1 and 581 cm−1 in Fig. 4(a) are ascribed to E2 (high) – E2 (low) mode and E1 (LO) mode of ZnO, respectively. It is reported that the intensity of E1 (LO) mode of ZnO increases with increasing defects associated with oxygen vacancies or zinc interstitials [32]. Hence, the presence of this peak in the Raman spectrum of ZnO-NRs on ZnO-BL is indicative of the substantial presence of oxygen vacancies. This observation is in agreement with the corresponding PL spectrum, in which, emissions related to both oxygen vacancies and zinc interstitials are seen. The Raman spectrum of ZnO-NRs on GaN-BL exhibits additional peaks at 564 cm−1 and 733 cm−1, which are assigned to E2 (high) and A1 (LO) modes of GaN [33]. Apart from the E2 (high) and E2 (high) – E2 (low) mode peaks in the Raman spectra of ZnO-NRs on both buffer layers, certain additional interesting features are seen in the case of the ZnO-NRs on GaN-BL. The near absence of E1 (LO) mode in the spectrum of ZnO-NRs on GaN-BL indicates substantial reduction of oxygen vacancies in ZnO-NRs grown on GaN-BL, and supports the same inference drawn from PL studies. In addition, low intensity peaks at 378 cm−1 and 483 cm−1 are seen, which are attributed to A1 (TO) and A1 (2LA) modes of ZnO, respectively. The A1 (TO) mode is completely absent for ZnO-NRs on ZnO-BL and is


4


served from ZnO-NRs grown on GaN-BL, compared to those grown on ZnO-BL. These results are supported by Raman studies, which also show the presence of peaks attributable to Raman modes seen in Ga-doped ZnO. These observations suggest the incorporation of Ga in ZnO-NRs, possibly near the GaN-ZnO interface, due to solution mediated interfacial reaction between GaN and ZnO. This study has shown that the superior optical quality of ZnO-NRs grown GaN-BL does not originate from the structural quality of the buffer layer and there is possibly a role of aqueous medium in controlling defects, such as, Zn interstitials and oxygen vacancies in ZnO-NRs and incorporation of Ga at the GaN/ZnO interface. These factors appear to significantly influence the photoluminescence of ZnO-NRs grown on GaN-BL. Acknowledgments The authors gratefully acknowledge Sophisticated Analytical Instruments Facility (SAIF), IIT Bombay for providing FESEM and Raman spectroscopy facilities. References

Fig. 4. Raman spectra of ZnO-NRs grown on (a) ZnO-BL and (b) GaN-BL.

prominently seen for ZnO-NRs on GaN-BL. According to Raman selection rules, only E2 and A1 (LO) modes are allowed and A1 (TO) and E1 (TO) modes are forbidden for c-axis oriented ZnO in backscattering geometry [34]. Thus, the presence of A1 (TO) mode in the case of ZnO-NRs on GaN-BL is attributed to the misorientation of the c-axis of ZnO-NRs. The appearance of A1 (2LA) mode in the case of ZnO-NRs on GaN-BL is another interesting observation, since this peak has been reported to be prominent in Ga-doped ZnO [35]. An additional mode at 625 cm−1 is also seen for ZnO-NRs on GaN-BL, which does not belong to ZnO or GaN Raman modes. Bundesmann et al. have [36] however, reported the presence of this peak in the Raman spectrum of Ga-doped ZnO. The appearance of the peaks at 483 cm−1 and the 625 cm−1 thus indicates the possibility of an exchange reaction mediated by the aqueous medium, which leads to Ga-doping of ZnO-NRs at the GaN-ZnO interface. 4. Conclusions ZnO-NRs have been grown by chemical bath deposition on thin polycrystalline GaN buffer layers having c-axis preferred orientation, which were sputter deposited on amorphous quartz substrates. The misalignment of ZnO-NRs grown on GaN-BL, compared to ZnO-BL is attributed to limited preferred c-axis orientation of crystallites in the GaNBL as well as the small lattice mismatch between ZnO and GaN. The presence of a nucleation barrier on GaN surface along with limited surface diffusion appears to hinder the nucleation process of ZnO-NRs, resulting in a smaller density of nucleation centers, which in turn, leads to the formation of a smaller density of ZnO-NRs having larger diameters and smaller lengths, compared to the nanorods grown on ZnOBL. Despite the limited vertical alignment of ZnO-NRs on GaN-BL and their poorer structural quality, a much stronger near-band-edge luminescence along with a drastically reduced defect luminescence is ob-

[1] M.H. Huang, S. Mao, H. Feick, H. Yan, Y. Wu, H. Kind, E. Weber, R. Russo, P. Yang, Science 292 (2001) 1897. [2] Z.W. Pan, Z.R. Dai, Z.L. Wang, Science 291 (2001) 1947. [3] G. Kenanakis, M. Androulidaki, D. Vernardou, N. Katsarakis, E. Koudoumas, Thin Solid Films 520 (2011) 1353. [4] W.I. Park, G.C. Yi, Adv. Mater. 16 (2004) 87. [5] M.S. Arnold, P. Avouris, Z.W. Pan, Z.L. Wang, J. Phys. Chem. B 107 (2003) 659. [6] D.C. Kim, W.S. Han, H.K. Cho, B.H. Kong, H.S. Kim, Appl. Phys. Lett. 91 (2007) 231901. [7] X. Chen, A.M.C. Ng, F. Fang, Y.H. Ng, A.B. Djuris, H.L. Tam, K.W. Cheah, S. Gwo, W.K. Chan, P.W.K. Fong, H.F. Lui, C. Surya, J. Appl. Phys. 110 (2011) 094513. [8] A.C. Mofor, A. Bakina, U. Chejarlaa, E. Schlenker, A. El-Shaer, G. Wagner, N. Boukos, A. Travlosc, A. Waag, Superlattice. Microst. 42 (2007) 415. [9] Q. Zhang, H. Guo, Z. Feng, L. Lina, J. Zhou, Z. Lin, Electrochim. Acta 55 (2010) 4889. [10] X.W. Sun, J.Z. Huang, J.X. Wang, Z. Xu, Nano Lett. 8 (2008) 1219. [11] J.T. Yan, C.H. Chen, S.F. Yen, C.T. Lee, IEEE Photon. Technol. Lett. 22 (2010) 3. [12] M.S. Lin, C.C. Chen, W.C. Wang, C.F. Lin, S.Y. Chang, Thin Solid Films 518 (2010) 7398. [13] H.Q. Le, S.J. Chua, Y.W. Koh, K.P. Loh, Z. Chen, C.V. Thompson, E.A. Fitzgerald, Appl. Phys. Lett. 87 (2005) 101908. [14] H.Q. Le, S.J. Chua, Y.W. Koh, K.P. Loh, E.A. Fitzgerald, J. Cryst. Growth 293 (2006) 36. [15] J.M. Jang, J.Y. Kim, W.G. Jung, Thin Solid Films 516 (2008) 8524. [16] B.S. Yadav, S.S. Major, R.S. Srinivasa, J. Appl. Phys. 102 (2007) 073516. [17] S. Singh, R.S. Srinivasa, S.S. Major, Thin Solid Films 515 (2007) 8718. [18] R.B. Peterson, C.L. Fields, B.A. Gregg, Langmuir 20 (2004) 5115. [19] J. Qiu, X. Li, W. He, S.J. Park, H.K. Kim, Y.H. Hwang, J.H. Lee, Y.D. Kim, Nanotechnology 20 (2009) 155603. [20] Y.J. Hong, J. Yoo, Y.J. Doh, S.H. Kang, K. Kong, M. Kim, D.R. Lee, K.H. Oh, G. Yi, J. Mater. Chem. 19 (2009) 941. [21] J.B. Cui, C.P. Daghlian, U.J. Gibson, R. Püsche, P. Geithner, J. Appl. Phys. 97 (2005) 044315. [22] J. Song, S. Lim, J. Phys. Chem. C 111 (2007) 596. [23] H. Lee, J.H. Shin, J. Chae, J.B. Kim, T.H. Kim, K.B. Park, Electron. Mater. Lett. 9 (2013) 357. [24] S. Guillemin, V. Consonni, E. Appert, E. Puyoo, L. Rapenne, H. Roussel, J. Phys. Chem. C 116 (2012) 25106. [25] S. Xu, Z.L. Wang, Nano Res. 4 (2011) 1013. [26] N. Hana, X. Wua, L. Chai, H. Liua, Y. Chen, Sensors Actuators B 150 (2010) 230. [27] K.H. Tam, C.K. Cheung, Y.H. Leung, A.B. Djurisic, C.C. Ling, C.D. Beling, S. Fung, W.M. Kwok, W.K. Chan, D.L. Phillips, L. Ding, W.K. Ge, J. Phys. Chem. B 110 (2006) 20865. [28] A.B. Djurišić, Y.H. Leung, K.H. Tam, L. Ding, W.K. Ge, H.Y. Chen, S. Gwo, Appl. Phys. Lett. 88 (2006) 103107. [29] A.E. Morales, U. Pal, Appl. Phys. Lett. 93 (2008) 193120. [30] J. Lv, C. Liu, W. Gong, Z. Zi, X. Chen, X. Chen, K. Huang, G. He, S. Shi, X. Song, Z. Sun, Semicond. Sci. Technol. 27 (2012) 115021. [31] R. Cusco, E.A. Llado, J. Ibanez, L. Artus, J. Jimenez, B. Wang, M.J. Callahan, Phys. Rev. B 75 (2007) 165202. [32] Z.Q. Chen, A. Kawasuso, Y. Xu, H. Naramoto, X.L. Yuan, T. Sekiguchi, R. Suzuki, T. Ohdaira, J. Appl. Phys. 97 (2005) 013528. [33] V.Y. Davydov, Y.E. Kitaev, I.N. Goncharuk, A.N. Smirnov, J. Graul, O. Semchinova, D. Uffmann, M.B. Smirnov, A.P. Mirgorodsky, R.A. Evarestov, Phys. Rev. B 58 (1998) 12899. [34] Y. Zhang, F. Lu, Z. Wang, L. Zhang, J. Phys. Chem. C 111 (2007) 4519. [35] A.E. Morales, U. Pal, Curr. Appl. Phys. 11 (2011) 525. [36] C. Bundesmann, N. Ashkenov, M. Schubert, D. Spemann, T. Butz, E.M. Kaidashev, M. Lorenz, M. Grundmann, Appl. Phys. Lett. 83 (2003) 1974.
