⁎ Corresponding author. Tel.: +886 2 27376385; fax: +886 2 27376424.
E-mail address:[email protected](Y.S. Huang).
0040-6090/$– see front matter © 2009 Elsevier B.V. All rights reserved.
doi:10.1016/j.tsf.2009.11.014
Contents lists available atScienceDirect
Thin Solid Films
j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / t s f
was located in front of the target. The sputtering power supply had a maximum output of 300 W. Two separate gas lines, each equipped with a massflow controller, were used to control the Ar and O2flow rates with an accuracy of 0.1 standard cubic centimeters per minute (sccm) for both gases. The sample holder was approximately 45 mm from the target and can be heated to a maximum temperature of 550 °C. To promote uniform transfer of heat to the substrates, a thin coating of melted In was placed between the substrate and the sample holder. Care was being taken to confine all the In metal only to the back of the substrate.
The sputtering chamber was evacuated with a turbo-molecular pump and had a base pressure of∼3×10−3Pa. Reactive sputtering was carried out in a mixture of argon and oxygen. For better oxidation of reactive sputtered Ti atom under the surface of the samples, the oxygen line was extended to the substrate holder. Thus O2was introduced over the substrate into the sputtering chamber with Ar atmosphere.
The sputtering parameters for sapphire (100) and sapphire (012) were O2/Ar = 1/1 and 1/5 (rate of O2flow of 10 and 2 sccm, respectively for the two substrates) corresponding to sputtering pressures of 14 and 8 Pa, respectively. The applied rf power was 230 W and the deposition time was 180 min. The substrate temperature Ts was maintained at 400 °C throughout the entire sputtering process.
2.2. Characterization of R-TiO2nanocrystals
The TiO2NCs morphology was studied with a JEOL-JSM6500F field-emission scanning electron microscope with an accelerating voltage of 15 kV. The dimensions and growth rates of various TiO2samples were estimated according to the recorded 90° cross-sectional FESEM images.
Crystal structures and overall out-plane orientation of the samples were analyzed using a Rigaku D/Max-RC X-ray diffractometer (XRD) equipped with Cu Kα radiation source, Ni filter and scintillation counter probe SC-30 as a detector. XRD data was collected in Brag Brentano configuration with a step size 2Θ=0.01°. The chemical binding state of the TiO2samples was investigated from Ti 2p and O 1 s spectra obtained by XPS using the Al Kα1486.68 eV as a radiation source in a Thermo VG Scientific Theta Probe system under the base pressure of 10−7Pa. The Ag 3d5/2 line at 368.26 eV was the calibration reference. XPS peak positions and integrated intensities were obtained through curvefitting, using Thermo VG Scientific: Avantage v3.2 software. The Raman spectra were recorded at room temperature utilizing the back-scattering mode on a Renishaw inVia micro-Raman system with 1800 grooves/mm grating and an optical microscope with a 50× objective. The Ar+laser beam of the 514.5 nm excitation line with a power of∼1.5 mW was focused onto a spot size∼5 μm in diameter. Prior to the measurement, the system was calibrated by means of the 520 cm−1Raman peak of a polycrystalline Si.
3. Results and discussion
3.1. Structure of R-TiO2 nanocrystals on sapphire (100) and (012) substrates
3.1.1. R-TiO2on sapphire (100)
As illustrated inFig. 1(a), the FESEM images show the growth of densely populated and vertically well-aligned columnar structure with lateral size of about 55 nm and an average length of about 680 nm TiO2
NCs on sapphire (100) substrate. The typical XRD pattern of the vertically aligned rutile TiO2NCs grown on sapphire (100) depicted in Fig. 1(b) shows the preferable orientation of the nanostructures along TiO2[001] (2θ ∼62.8°). Here we observe anisotropic growth and as a result, NCs formation is restricted by the in-plane mismatch. Thus, the deposited Ti and O atoms are stacked into a 1D nanostructure along c-direction with TiO2plane formation following the substrate orien-tation. Similar growth behavior of R-TiO2, IrO2and RuO2grown on sapphire (100) by MOCVD were observed by our group earlier[31–33].
The preferable oriented growth of R-TiO2(001) along [001] can be explained by examining the TiO2and sapphire (100) planar structure at the atomic level. The unidirectional growth of the NCs is correlated to the epitaxial relation between the rutile lattice of TiO2 and the underlying single-crystal sapphire (001) substrate. The main assump-tion is that there are oxygen vacancies at the sapphire (100) surfaces.
The schematic diagrams illustrated inFig. 2show the atomic arrange-ments on TiO2(001) and sapphire (100) planes. The lattice parameters for TiO2 are a =b = 4.59 Å and c = 2.96 Å [JCPDS No.21-1276], for sapphire they are a =b = 4.76 Å and c = 12.99 Å [JCPDS No. 10-0173].
The incoming Ti atoms have sufficient mobility to minimize the lattice Fig. 1. (a) FESEM images (30º perspective- and cross-sectional-view) and (b) XRD pattern of rutile TiO2(001) NCs grown on sapphire (SA) (100) substrate by reactive magnetron sputtering.
Fig. 2. Schematic plots of the lattice relationships between vertically aligned rutile TiO2
NCs and sapphire (SA) (100) substrate: (a) R-TiO2(001); (b) SA(100); (c) R-TiO2(001) on SA(100).
4122 C.N. Yeh et al. / Thin Solid Films 518 (2010) 4121–4125
misfit and align themselves in TiO2(001) arrangement because of the similar oxygen arrangement of underlying substrates and that of TiO2
(001). Thus, the growth relationship can be described as TiO2(001) [100] // sapphire (100)[010]. These alignments produce residual stress due to mismatch values of −3.57% {[(4.59–4.76)Å/4.76 Å]} along sapphire [010], and +5.76% {[(4.59–4.34)Å/4.34 Å]} along sapphire [001] for the NCs grown on sapphire (100). Therefore, we conclude that the vertical growth of TiO2NCs on sapphire (100) substrate, on which the template for TiO2(001) planes formation is facilitated, is dictated by the c-axis directional growth together with the lattice mismatch minimizing mechanism[32].
3.1.2. R-TiO2on sapphire (012)
The micrographs of densely-packed TiO2NCs on sapphire (012) are illustrated inFig. 3(a). The FESEM images reveal the growth of titled columnar structure with lateral size of about 85 nm and an average length of about 860 nm TiO2NCs on sapphire (012) substrate.
These column-like nanostructures exhibit regularly tilted NCs with identical tilt angle (∼33°) from the normal to substrate. The possible explanation of this unique directional growth will be discussed below.
Fig. 3(b) shows typical XRD patterns of the regularly tilted TiO2NCs deposited on sapphire (012). Two peaks can be indexed as (101) and (202) diffraction planes at 2θ around ∼36.0° and ∼76.2°, respectively.
The observation indicates parallel in-plane TiO2(101) orientation. For this particular case, anisotropic growth has been observed resulting in NCs formation restricted by the in-plane mismatch mechanism. Thus, the deposited Ti and O atoms are stacked into 1D nanostructure along c-direction with TiO2plane formation following the substrate orientation with the growth relation TiO2(101) // sapphire (012).
To determine the directions of planar deposition we have to examine the atomic arrangements of appropriate surfaces.Fig. 4illustrates the schematic plots of the atoms arrangements and lattice relationships between TiO2 (101) and sapphire (012) surfaces. According to the argument of the minimization of the oxide sublattice structural mismatch, the possible NCs-substrates alignment can be described as TiO2(101)[010] // sapphire (012)[100]. The alignments mentioned
above produce directional mismatches of−3.57% {(4.59 Å–4.76 Å)/
4.76 Å} and +6.43% {(5.46 Å–5.13 Å)/5.13 Å} along TiO2[010] and TiO2
[101̅], respectively, on sapphire (012). Concluding this part of the study, we noted that the mechanisms responsible for conversion fromfilm to nanocrystal and the well-aligned directional growth of the nanocrystals are guided by internal and/or external factors. The c-directional growth mechanism is referred as the internal factor and has it origin from the anisotropy of the crystal structure and results in different growth rate along different directions of NCs. The other parameters such as deposition conditions (RF power, substrate temperature, sputtering pressure) and substrate orientations are classified as the external factors. The minimization of the oxide sublattice structural mismatch as dictated by the substrate orientation and the c-directional growth mechanism are the two main driving forces which can determine the alignment and the formation of either tilted or vertical TiO21D NCs.
The external factors of substrate orientation, substrate temperature, and sputtering pressure are responsible for conversion fromfilm to nanocrystals. It is quite apparent that these external factors have an overlapping influence on the internal factor of energetically favorable surface for the incoming atoms (c-directional growth mechanism) and initiate the preferable plane orientation of TiO2 NCs, whereby the incoming atoms will stick onto the lower energy sites.
3.2. XPS investigation
X-ray photoelectron spectroscopy is frequently used as a comple-mentary technique for assigning oxidation states and the stoichiometry of the oxides. TiO2has been reported to be relative easy to handle experimentally[34]. Although it has a bulk band gap of 3.1 eV, no charging problems occur during XPS measurements due to possible oxygen dissociation under ultra-high vacuum and/or surface carbon contamination of the samples.Fig. 5depicts slow scan XPS spectra in the vicinity of (a) C 1 s, (b) Ti 2p and (c) O 1 s regions before (curve I) and after (curve II) Ar ion bombardment. The Gaussian and Lorentzian mixing line shape after the treatment of background by Shirley function has been used in thefitting to determine the accurate peak positions.
As shown inFig. 5(a) a carbon C 1s peak at a binding energy of 284.5 ± 0.1 eV is observed before ion bombardment (curve I). The Fig. 3. (a) FESEM images (30º perspective- and cross-sectional-view) and (b) XRD
pattern of rutile TiO2(101) NCs grown on sapphire (SA) (012) substrate by reactive magnetron sputtering.
Fig. 4. Schematic plots of the lattice relationships between rutile TiO2(101) NCs and sapphire (SA) (012) substrate: (a) R-TiO2(101); (b) SA (012); (c) R-TiO2(101) on SA (012).
C.N. Yeh et al. / Thin Solid Films 518 (2010) 4121–4125 4123
presence of this peak is related to surface pollution which corresponds to the fact that the sample is exposed to air before the XPS measure-ments. The area of the C 1s peak decreases upon ion bombardments (see curve II). This signature persists even after a prolonged ion bombardment of duration much longer than the one known to be necessary for removal of the surface pollution on single crystal titanium dioxide surfaces[35].
It can be seen inFig. 5(b), before Ar ion bombardment the Ti 2p3/2
curve is composed of a single peak at a binding energy of 459.1 ± 0.1 eV,
with a fullwidth at half maximum of 1.3 ± 0.1 eV. The separation between the Ti 2p3/2and Ti 2p1/2is 5.6 ± 0.2 eV. The O 1s binding energy is 530.2 ± 0.1 eV. These results are in good agreement with that of single crystal TiO2(110)[36]. The stoichiometry is determined by the relative areas of the total Ti 2p and O 1s XPS peaks with the correction of the relative sensitivity factors. The oxygen vs. titanium ratio of 2.0 ± 0.1 is obtained. After argon bombardment, the Ti 2p3/2 peak showed a shoulder on the low binding energy side, which was evidenced by the presence of Ti3+[37], indicating that the surface has become non-stoichiometric. The result is a consequence of the removal of oxygen from the surface caused by the preferential sputtering phenomenon [38].
InFig. 5(c), two oxide states attributed to O2−and OH−species are observed from the as-deposited NCs (curve I). However, the OH− shoulder disappeared after Ar ion sputtering (curve II) and the main O 1s feature shifted its binding energy to the reference value of O2−peak [36]. This indicates that the OH−peak is just a surface contamination peak probably due to water adsorption in the air. Similar XPS spectra (peak positions and broadening parameters) were also obtained for TiO2 NCs grown on sapphire (012) substrate. However, the intensity of the O 1 s and Ti 2p core level emissions for TiO2NCs grown on sapphire (012) substrate is about 20% lower than that grown on (100) plane.
This difference in intensity is expected in view of the different surface morphology of the (001) and (101) orientated NCs.
3.3. Raman scattering investigation
Fig. 6(a) and (b) shows the Raman spectra (dashed curves) of as-deposited R-TiO2NCs samples on sapphire (100) and (012), respec-tively. For comparison purpose, the Raman spectra (solid curves) for R-TiO2(001) and (101) single crystal plates (MTI crystals, Richmond, CA, USA) are also included inFig. 6. Rutile TiO2is tetragonal and belongs to the space group D4h14with two TiO2molecules per unit cell. There are four Raman-active modes with symmetry of A1g, B1g, B2g, and Eg[39]. These four Raman-active modes of rutile TiO2single crystal were detected at 143 cm− 1(B1g), 447 cm−1(Eg), 612 cm− 1(A1g), and 826 cm−1(B2g) by Porto et al.[40]. In addition several second order Raman bands were also observed. As can be seen inFig. 6, the spectra exhibit three distinct broad peaks and a broad weak underlying continuum tailing off at higher
Fig. 5. Slow scan XPS spectra in the vicinity of (a) C 1 s, (b) Ti 2p and (c) O 1 s regions before (curve I) and after (curve II) Ar ion bombardment for R-TiO2(001).
Fig. 6. Raman spectra of (a) R-TiO2(001) NCs on sapphire (100) (dashed curve) and R-TiO2(001) single crystal plate (solid curve), and (b) R-TiO2(101) NCs on sapphire (012) (dashed curve) and R-TiO2(101) single crystal plate (solid curve).
4124 C.N. Yeh et al. / Thin Solid Films 518 (2010) 4121–4125
Raman shifts, and are in good agreement with the features corresponding to rutile TiO2[40]. The major features are the Egand A1g modes located at 445 and 608 cm−1 with full width at half maximum (FWHM) of 38 and 49 cm− 1, respectively, along with the two-phonon bands at∼237 cm−1(marked as *). The Raman spectra exhibits soft B2gmode located at around 823 cm−1and B1gmode is completely absent. The results of Raman scattering measurements confirm the rutile phase of the as-deposited NCs. NCs exhibit a slight redshift in the peak position and a slight broadening in linewidth as compared with the SC (Egat 447 cm− 1, FWHM∼35 cm− 1; A1g at 610 cm− 1, FWHM∼39 cm−1). The results indicated the formation of good quality rutile phase nanocrystalline TiO2. Relative Raman intensities of the Eg, A1g, and B2g phonon modes for the various polarization configurations for R-TiO2are similar to that for IrO2and RuO2and can be found elsewhere[41]. The selection rules maintain that Eg, A1g, and B2gmode are allowed for all polarization configurations for the (101) plane, and the Egmodes is forbidden for all configurations from the scattering of (001) planes. The appearance of Raman signal of the normally forbidden mode might indicate observation of the scattering from the other planes of the NCs as evidenced from the pyramidal shape of the nanostructures' tips.
4. Summary
Well-align densely-packed TiO2NCs have been grown on sapphire (100) and (012) substrates via by reactive radio frequency magnetron sputtering using Ti metal target. We demonstrate that reactive radio frequency magnetron sputtering is a simple method which has several advantages including better control of the growth conditions and a single deposition step to obtain the nanostructures. The results of the structural study reveal that the vertically aligned NCs were deposited on sapphire (100), while the NCs on the sapphire (012) were grown with a tilt angle of∼33° from the normal to substrates. The strong substrates effect on the TiO2NCs alignment can be explained by the effects of minimization of the oxide sublattice structural mismatch overlapped with the c-directional growth mechanism. XPS analyses reveal oxygen vs. titanium ratio of 2.0 ± 0.1 for the as-grown TiO2NCs.
The Raman spectra show the formation of good quality rutile-phase nanocrystalline TiO2.
Acknowledgements
The authors wish to acknowledge the support of the National Science Council of Taiwan under Contract No. NSC 97-2112-M-011-001-MY3.
References
[1] S.P. Albu, A. Ghicov, J.M. Macak, R. Hahn, P. Schmuki, Nano Lett. 7 (2007) 1286.
[2] X. Zhang, M. Jin, Z. Liu, D.A. Tryk, S. Nishimoto, T. Murakami, A. Fujishima, J. Phys.
Chem. C 111 (2007) 14521.
[3] A.R. Armstrong, G. Armstrong, J. Canales, R. Garcia, P.G. Bruce, Adv. Mater. 17 (2005) 862.
[4] K. Hashimoto, H. Irie, A. Fujishima, Jpn. J. Appl. Phys. Part 1 44 (2005) 8269.
[5] A. Fujishima, T.N. Rao, D.A. Tryk, J. Photochem. Photobiol. C 1 (2000) 1.
[6] L. Francioso, D.S. Presicce, M. Epifani, P. Siciliano, A. Ficarella, Sens. Actuators B 107 (2005) 563.
[7] A. Kolmakov, M. Moskovits, Annu. Rev. Mater. Res. 34 (2004) 151.
[8] O.K. Varghese, C.A. Grimes, J. Nanosci. Nanotechnol. 3 (2003) 277.
[9] C. Garzella, E. Bontempi, L.E. Depero, A. Vomiero, G. Della Mea, G. Sberveglieri, Sens. Actuators B 93 (2003) 495.
[10] C. Garzella, E. Comini, E. Tempesti, C. Frigeri, G. Sberveglieri, Sens. Actuators B 68 (2000) 189.
[11] K. Fujihara, A. Kumar, R. Jose, S. Ramakrishna, S. Uchida, Nanotechnology 18 (2007) 365709.
[12] B. Tan, Y.Y. Wu, J. Phys. Chem. B 110 (2006) 15932.
[13] B. O'Regan, M. Grätzel, Nature 353 (1991) 737.
[14] O.V. Sakhno, L.M. Goldenberg, J. Stumpe, T.N. Smirnova, Nanotechnology 18 (2007) 105704.
[15] R. Rabady, I. Avrutsky, Appl. Opt. 44 (2005) 378.
[16] A. Richel, N.P. Johnson, D.W. McComb, Appl. Phys. Lett. 76 (2000) 1816.
[17] P. Zeman, S. Takabayashi, Thin Solid Films 433 (2003) 57.
[18] G.R. Gu, Z. He, Y.C. Tao, Y.A. Li, J.J. Li, H. Yin, W.Q. Li, Y.N. Zhao, Vacuum 70 (2003) 17.
[19] S. Cerneaux, X. Xiong, G.P. Simon, Y.B. Cheng, L. Spiccia, Nanotechnology 18 (2007) 055708.
[20] D. Liu, M.Z. Yates, Langmuir 23 (2007) 10333.
[21] Z. Miao, D. Xu, J. Ouyang, G. Guo, X. Zhao, Y. Tang, Nano Lett. 2 (2002) 717.
[22] S. Qourzal, A. Assabbane, Y. Ait-Ichou, J. Photochem. Photobiol. A 163 (2004) 317.
[23] W.F. Zhang, M.S. Zhang, Z. Yin, Phys. Status Solidi A 179 (2000) 319.
[24] J.M. Wu, H.C. Shih, W.T. Wu, Chem. Phys. Lett. 413 (2005) 490.
[25] J.M. Wu, W.T. Wu, H.C. Shih, J. Electrochem. Soc. 152 (2005) G613.
[26] J.M. Wu, H.C. Shih, W.T. Wu, J. Vac. Sci. Technol. B 23 (2005) 2122.
[27] P. Evans, M.E. Pemble, D.W. Sheel, Chem. Mater. 18 (2006) 5750.
[28] L.A. Brook, P. Evans, H.A. Foster, M.E. Pemble, A. Steele, D.W. Sheel, H.M. Yates, J. Photo. Chem. A 187 (2007) 53.
[29] M.K. Lee, J.J. Huang, T.S. Wu, Semicond. Sci. Technol. 20 (2005) 519.
[30] J.J. Wu, C.C. Yu, J. Phys. Chem. B 108 (2004) 3377.
[31] C.A. Chen, Y.M. Chen, A. Korotcov, Y.S. Huang, D.S. Tsai, K.K. Tiong, Nanotechnology 19 (2008) 075611.
[32] R.S. Chen, A. Korotcov, Y.S. Huang, D.S. Tsai, Nanotechnology 17 (2006) R67.
[33] C.C. Chen, R.S. Chen, T.Y. Tsai, Y.S. Huang, D.S. Tsai, K.K. Tiong, J. Phys., Condens.
Matter 16 (2004) 8475.
[34] J.T. Mayer, U. Diebold, T.E. Madey, E. Garfunkel, J. Electron Spectrosc. Relat. Phenom.
73 (1995) 1.
[35] S. Bourgeois, P. Le Seigneur, M. Perdereau, Surf. Sci. 328 (1995) 105.
[36] S. Pétigny, B. Domenichini, H. Mostéfa-Sba, E. Lesniewska, A. Steinbrunn, S.
Bourgeois, Appl. Surf. Sci. 142 (1999) 114.
[37] W. Göpel, J.A. Anderson, D. Frankel, M. Jaenig, K. Phillips, J.A. Schäffer, G. Rocker, Surf. Sci. 139 (1984) 333.
[38] G. Betz, G.K. Wehner, Sputtering by Particle Bombardment II, in: R. Behrisch (Ed.), Topics in Applied Physics, vol. 52, Springer-Verlag, Berlin, 1983, p. 11.
[39] R. Loudon, Adv. Phys. 13 (1964) 423.
[40] S.P.S. Porto, P.A. Fleury, T.C. Damen, Phys. Rev. 154 (1967) 522.
[41] A.V. Korotcov, Y.S. Huang, K.K. Tiong, D.S. Tsai, J. Raman Spectrosc. 38 (2007) 737.
C.N. Yeh et al. / Thin Solid Films 518 (2010) 4121–4125 4125
IOP PUBLISHING NANOTECHNOLOGY
Nanotechnology 21 (2010) 035702 (7pp) doi:10.1088/0957-4484/21/3/035702