synthesis of three kinds of gan nanowires through ga2o3 films’ reaction with ammonia
TRANSCRIPT
![Page 1: Synthesis of three kinds of GaN nanowires through Ga2O3 films’ reaction with ammonia](https://reader031.vdocuments.us/reader031/viewer/2022020402/5750200e1a28ab877e98cb3f/html5/thumbnails/1.jpg)
ARTICLE IN PRESS
1386-9477/$ - se
doi:10.1016/j.ph
�Correspondi+86 0531 61800
E-mail addre
Physica E 27 (2005) 32–37
www.elsevier.com/locate/physe
Synthesis of three kinds of GaN nanowires through Ga2O3
films’ reaction with ammonia
Zhihua Dong, Chengshan Xue�, Huizhao Zhuang, Shuyun Wang, Haiyong Gao,Deheng Tian, Yuxin Wu, Jianting He, Yi’an Liu
The Functional Materials Laboratory, Institute of semiconductors, Shandong Normal University,
East Culture Road, Ji’nan 250014, China
Received 2 August 2004; received in revised form 18 October 2004; accepted 22 October 2004
Available online 28 November 2004
Abstract
A new method was employed to obtain GaN nanowires (NWs). In this method, SiC films were deposited with radio
frequency (r.f.) magnetron sputtering onto silicon substrates and annealed at high temperature, and then Ga2O3 films
were deposited on top of the SiC intermediate layers and annealed in NH3 atmosphere. SiC layer was used to reduce
thermal and lattice mismatch between GaN and Si, and improve NWs’ quality. After Ga2O3 films reacted with NH3, a
great quantity of GaN NWs with the shape of birch trunks and stalactites were found by transmission electron
microscopy (TEM). At the same time, a few very even and uniform pillarlike NWs were observed. The electron
diffraction patterns (EDP) show that birch trunk-shaped and pillarlike NWs are all single-crystalline structures. These
NWs were also analyzed with the assistance of X-ray diffraction (XRD), Fourier transformed infrared spectra (FTIR)
and high-resolution transmission electron microscopy (HRTEM) to show their properties.
r 2004 Elsevier B.V. All rights reserved.
PACS: 68.65.�k; 71.55.Eq; 61.45.+w
Keywords: GaN; Intermediate layer; Stalactite-shaped; Birch trunk-shaped; Pillarlike
1. Introduction
As a kind of important semiconductor, GaNattracted extensive attention for decades. GaN
e front matter r 2004 Elsevier B.V. All rights reserve
yse.2004.10.003
ng author. Tel.: +86 0531 6182624; fax:
17.
ss: [email protected] (C. Xue).
nanowires (NWs), nanorods and nanobelts havepromising device applications in 1-D systems dueto their excellent properties. Many groups haveprepared GaN on the nanoscale. Chia-Chun Chenet al. produced high-quality GaN NWs usingvapor–liquid–solid (VLS) technique [1]. JoushuaGoldberger et al. formed single-crystal GaNnanotubes with the assistance of template [2].
d.
![Page 2: Synthesis of three kinds of GaN nanowires through Ga2O3 films’ reaction with ammonia](https://reader031.vdocuments.us/reader031/viewer/2022020402/5750200e1a28ab877e98cb3f/html5/thumbnails/2.jpg)
ARTICLE IN PRESS
Fig. 1. The anneal system, also the reaction system.
Z. Dong et al. / Physica E 27 (2005) 32–37 33
H. Li et al. fabricated bamboo-shaped GaNnanorods in 2002 [3]. Other methods such asthermal chemical vapor deposition (CVD) [4],oxide-assisted technique [5], sublimation [6], arcdischarge [7], laser ablation [8], and pyrolysis [9]were also utilized to produce GaN NWs.Currently, Hoche et al. reported the preparationof well-ordered substrate-adhered GaN NWs withdiffraction-mask-projection excimer-laser patterning[10]. A novel method was carried out to preparesingle-crystalline GaN NWs in our laboratory [11].Only several simple appliances were used in thismethod, so it simplified GaN NWs’ preparationtechnique greatly. However, the NWs we got are noteven and smooth, and a big thermal mismatchbetween GaN and silicon substrate will restrict GaNNWs’ application in devices. In order to avoid thisdefect and improve GaN’s quality, we used SiC asthe intermediate layer.
2. Experimental details
In our experiment, we formed GaN NWs byself-assembling of Ga2O3 films in their reactionwith NH3. The films were deposited by magnetronsputtering on the SiC intermediate layer on siliconsubstrates.The first step is the process of forming SiC
intermediate layers. SiC films were deposited tosilicon substrates with a JCK-500A radio fre-quency (r.f.) magnetron sputtering machine. CleanSi (1 1 1) wafers were employed as substrates. SiCwafers with the purity of more than 99.999% wereused as targets. The conditions of sputtering wereas follows: the background pressure was4� 10�4 Pa; the distance between targets andsubstrates was 8 cm; the pressure of Ar(X99.99%) was 2 Pa; the sputtering power was150W and the frequency was 13.56MHz. After15min, SiC films with a thickness of about 50 nmwere obtained.Subsequently, SiC films were put into the anneal
system (Fig. 1). The furnace was kept at 950 1C. N2
(X99.999%, with a flow of 1000ml/min) wasintroduced into the oven chamber to protect thesamples. SiC intermediate layers with the samethickness formed at high temperature after 15min.
The second step is to deposit Ga2O3 films andsynthesize GaN NWs. The SiC-covered waferswere sputtered with Ga2O3 (99.999%) for 90minunder the same conditions and 500 nm thick filmsconsisting of Ga2O3 nanoparticles were obtained.Then the samples were placed into the reactionsystem without the least delay. N2 was introducedinto the system for 5min to expel the air.Subsequently, NH3 (99.999%) with a flux of500ml/min was introduced into the system. Thereaction lasted for 5, 10 and 15min, respectively.And the corresponding samples were named A, Band C. As our observation, Ga2O3 began toconvert into GaN at about 800 1C. The reactionequation can be expressed as Ga2O3+2NH3 ¼ 2GaN+3H2Om.After these processes, three kinds of GaN NWs
were obtained. We analyzed the samples withRigaku D/max-rB X-ray diffraction (XRD) meter(Tokyo, Cu Ka, l ¼ 1:54178 (A; 2y mode, 301–501)at room temperature to specify their crystallinestructure. Nicolet710 Fourier transformed infraredspectra (FTIR) meter was used to measure thesamples’ chemical states. Transmission electronmicroscopy (TEM) (Hitachi H-800) and high-resolu-tion transmission electron microscopy (HRTEM)(GAM2010, JEOL company) were carried out atroom temperature to measure the samples’ morphol-ogy and microstructure, respectively.
3. Results and discussion
3.1. XRD analysis
The XRD pattern of sample C is depicted inFig. 2. Peaks were found at 2y ¼ 32:31; 34.51 and
![Page 3: Synthesis of three kinds of GaN nanowires through Ga2O3 films’ reaction with ammonia](https://reader031.vdocuments.us/reader031/viewer/2022020402/5750200e1a28ab877e98cb3f/html5/thumbnails/3.jpg)
ARTICLE IN PRESS
30 35 40 45 500
500
1000
1500
Inte
nsi
ty (
a. u
.)
2 theta (degree)
GaN
SiC
(002
)
(101
)
(100
)
(111
)
(102
)
_β
Fig. 2. XRD pattern of sample C.
Fig. 3. FTIR image of sample C measured from 400 to
1300 cm�1.
Z. Dong et al. / Physica E 27 (2005) 32–3734
36.71 in correspondence with the hexagonalwurtzite GaN (lattice constants a ¼ 0:3186 nmand c ¼ 0:5178 nm). The peak (0 0 2) (at2y ¼ 34:51) has the highest intensity. That isdifferent from bulk GaN, whose 100% intensitypeak is peak (1 0 1). That fact results from theGaN NWs’ growth preference of plane (0 0 2). Theregular arrangement in every GaN NW causes 002XRD peak to be the most pronounced. A singleNW’s microstructure will be showed in theHRTEM plot later. There is only one peak ofthe SiC at 2y ¼ 35:71 ðd ¼ 0:251 nmÞ with lowintensity and big width in this plot. This peak isattributed to b� SiC (1 1 1) [12]. It indicates thatthe intermediate layer has (1 1 1) growth orienta-tion. The reason of its low intensity and wide full-width of the half-maximum (FWHM) possibly isthat the SiC layer is ultra-thin and of low-gradecrystalline quality.
3.2. FTIR analysis
The FTIR results of sample C are given inFig. 3. We can see five bands at 450.19, 568.01,611.18, 738.43, and 1058.65 cm�1 in the spectrumof Fig. 3. In the light of the reported band Ga–Nstretching vibration absorption in hexagonal GaNat 560.45 [13], the band at 568.01 cm�1 should be aGa–N stretching vibration absorption peak. Thisalso illustrates that the GaN we produced has ahexagonal structure. The absorption bands at450.75 cm�1 are corresponding to Si–O bending[14]. Si–O resulted from the oxygenation process
when the samples were placed in the air. The narrowabsorption band at 611.18 cm�1 is the local vibrationof the substitutional carbon in the Si crystal lattice[15]. And the 738.43 cm�1 peak is corresponding toSi–C bending absorption band [16]. This shows thatSiC has grown into the crystal in the high-temperature process. The band at 1058.65 cm�1 isthe Si–O bending band [17]. Si–O came from SiO2
formed on the surface of silicon wafers.
3.3. TEM analysis
NWs with three kinds of shapes were found inthe samples synthesized with different reactiontime when we observed them in TEM.A single stalactite-shaped nanowire, which
prevails in sample A, is shown in Fig. 4(a). Theaverage diameter of this nanowire is about 230 nm.There are so many sawtooth structures on itssurface. Selected area electron diffraction (SAED)indicates its bad crystalline quality.Birch trunk-shaped NWs prevail in sample C.
These straight NWs have some black rings(possibly the thicker part), which make themlook like a birch trunk. These NWs are muchthinner than A, with diameter ranging fromabout 50 to 100 nm. An example with a diameterabout 80 nm is showed in Fig. 4(c). Distinctly,it is smooth and even, with several black rings.The inset is its SAED pattern, whereð1 1 0 0Þ; ð1 0 1 0Þ and ð0 1 1 0Þ diffraction spotspresent. This indicates that it is single-crystalhexagonal GaN.
![Page 4: Synthesis of three kinds of GaN nanowires through Ga2O3 films’ reaction with ammonia](https://reader031.vdocuments.us/reader031/viewer/2022020402/5750200e1a28ab877e98cb3f/html5/thumbnails/4.jpg)
ARTICLE IN PRESS
Fig. 4. TEM images of all kinds of GaN NWs; (a) stalactite-
shaped NW from sample A; (b) NW from sample B; (c) birch
trunk-shaped NW from sample C; (d) pillarlike NW from
sample C.
Z. Dong et al. / Physica E 27 (2005) 32–37 35
Fig. 4(b) depicts two NWs prevailing in sampleB, proving that their shape is just intermediate ofsamples A and B. Their diameter is about 100 nm,also between two other kinds. They should beclose to single crystal in light of their SAED image.A pillarlike NW is showed in Fig. 4(d), which is
also the longest one, inclining from the bottom tothe top. It is extremely straight, even and smoothwith a diameter about 50 nm. Compared to birchtrunk-shaped NW, there are no black rings on itssurface. Its same SAED pattern with the birchtrunk-shaped shows its single crystalline structure.Since all these NWs formed at the same
conditions (only with different time), stalactite-shaped NW must be a previous state of birchtrunk-shaped NW in its growth. The same relationbetween birch trunk-shaped and pillarlike NW canbe deduced according to this rule. But the
mechanism of these NWs’ evolution is not preciseyet. We presume that atoms in NWs’ surfaces willsublime and diffuse as the heating goes on. Hence,the NWs become thinner and thinner, and moreand more smooth at the same time. So thestalactite-shaped NWs become birch trunk-shapedNWs, and finally very smooth and uniformNWs—pillarlike NWs. The further evidences ofour assumption are now being pursued.
3.4. HRTEM analysis
The HRTEM images of three kinds of NWswere given in Fig. 5. Figs. 5(a) and (b) are theimages tested with lower amplification. There aretwo kinds of NWs in Fig. 5(a), birch trunk-shapedand pillarlike NWs. These NWs were all fromsample C. There is a stalactite-shaped NW in thecenter of Fig. 5(b). The backbone-shaped structureis very clear.Figs. 5(c)–(e) are higher-amplification images of
three kinds of NWs. They are corresponding tostalactite-shaped, birch trunk-shaped and pillar-like NW accordingly.There are lots of blurry stripes with approximate
orientation in Fig. 5(c). No obvious latticestructure is present. This is possibly due to theNWs’ large thickness. Comparatively, Fig. 5(d)has an excellent lattice structure with more clearorientation and little defects. It is a single-crystalGaN incontrovertibly.The lattice is more perfect in Fig. 5(e). Its
orientation is most clear and uniform. Wemeasured the plane distance according to the scaleattached. The distance between the two stripes is0.259 nm, which is corresponding to the planedistance of GaN (0 0 2). The orientation that theNW runs along is showed in Fig. 5(e). Theinclination between this axis and [0 0 2] is about111. In view of the XRD peaks of SiC, we canconclude that the (1 1 1) SiC makes (0 0 2) GaNgrow on it preferentially in our experiments.
4. Conclusion
We have synthesized GaN NWs on SiC inter-mediate layer over silicon substrates through the
![Page 5: Synthesis of three kinds of GaN nanowires through Ga2O3 films’ reaction with ammonia](https://reader031.vdocuments.us/reader031/viewer/2022020402/5750200e1a28ab877e98cb3f/html5/thumbnails/5.jpg)
ARTICLE IN PRESS
Fig. 5. HRTEM images of GaN NWs: (a), (b) the images with lower amplification; (c) high-resolution image of stalactite-shaped NW;
(d) high-resolution image of birch trunk-shaped NW; (e) high-resolution image of pillarlike NW, which shows the orientation of NW’s
axis and its plane distance (0.259 nm) and its crystalline orientation of [0 0 2].
Z. Dong et al. / Physica E 27 (2005) 32–3736
reaction between Ga2O3/SiC films and NH3. Thereaction time affected NWs’ configuration. Threekinds of GaN NWs, stalactite-shaped, birch trunk-
shaped and pillarlike NWs, were obtained. Thelast two are all single crystal. Pillarlike NWs haveperfect lattice structure and (0 0 2) growth
![Page 6: Synthesis of three kinds of GaN nanowires through Ga2O3 films’ reaction with ammonia](https://reader031.vdocuments.us/reader031/viewer/2022020402/5750200e1a28ab877e98cb3f/html5/thumbnails/6.jpg)
ARTICLE IN PRESS
Z. Dong et al. / Physica E 27 (2005) 32–37 37
preference. The axes they run along have ainclination of about 111 with [0 0 2]. In ourexperiments, SiC (1 1 1) makes GaN (0 0 2) growfastest.
Acknowledgements
The authors would like to thank the NationalNatural Science Foundation of China (no.90201025,90301002).
References
[1] C.-C. Chen, C.-C. Yeh, Adv. Mater. 12 (2000) 738.
[2] J. Goldberger, R. He, Y. Zhang, et al., Nature 422 (2003)
599.
[3] H. Li, J.Y. Li, M. He, X.L. Chen, Z. Zhang, Appl. Phys. A
74 (2002) 561.
[4] X. Chen, J. Xu, R.M. Wang, D. Yu, Adv. Mater. 15 (2003)
419.
[5] W.S. Shi, Y.F. Zheng, N. Wang, C.S. Lee, S.T. Lee, Chem.
Phys. Lett. 345 (2001) 377.
[6] J.Y. Li, X.L. Chen, Z.Y. Qiao, Y.G. Cao, Y.C. Lan,
J. Cryst. Growth 213 (2000) 408.
[7] W. Han, P. Redlich, F. Ernst, M. Ruhle, Appl. Phys. Lett.
76 (2000) 652.
[8] X. Duan, C.M. Lieber, J. Am. Chem. Soc. 122 (2000) 188.
[9] W.-Q. Han, A. Zettl, Appl. Phys. Lett. 80 (2002) 305.
[10] T. Hoche, R. Bohme, J.W. Gerlach, et al., Nano Lett. 4 (5)
(2004) 895.
[11] X.M. Wang, L. Yang, C.M. Wang, C.S. Xue, Rare Metal
Mater. Eng. 33 (2004) 670.
[12] Y.M. Lei, Y.H. Yu, C.X. Ren, et al., Thin Solid Films 365
(2000) 53.
[13] J.-H. Boo, C. Rohr, W. Ho, J. Cryst. Growth 189/190
(1998) 439.
[14] P. Bettotti, G. Das, G. Mariotto, L. Pavesi, Appl. Phys.
Lett. 83 (2003) 749.
[15] Y. Sun, T. Miyasato, et al., J. Appl. Phys. 84 (1998) 6451.
[16] F. Demichelis, G. Crovini, C.F. Pirri, E. Tresso, G.
Amato, U. Coscia, G. Ambrosone, P. Rava, Thin Solid
Films 241 (1994) 274.
[17] N. Laidani, R. Capelletti, M. Elena, L. Guzman, G.
Mariotto, A. Miotello, P.M. Ossi, Thin Solid Films 223
(1993) 114.