
Web Release Date: December 4,
Free-Standing Porous Anodic Alumina Templates for Atomic Layer Deposition of Highly Ordered TiO2 Nanotube Arrays on Various Substrates
Institute of Materials Research and Engineering, 3 Research Link, Singapore 117602
Received: August 30, 2007
In Final Form: October 9, 2007
Abstract:
Free-standing porous anodic alumina (PAA) templates are used for fabrication of dense, uniform, highly
ordered, and well-aligned arrays of TiO2 nanotubes on various substrates, including glass, Si, and flexible
polyimide films, via atomic layer deposition (ALD). A large-area free-standing PAA template (>4.5 cm2) is
first attached onto the substrate with support of a protective layer of poly(methyl methacrylate) (PMMA)
applied on its top. UV-ozone treatment is then employed to completely remove the PMMA layer, keeping
the large-area template intact. TiO2 nanotube arrays with atomic layer control of wall thickness are prepared
via ALD on the template. The TiO2 nanotube arrays with varying wall thickness show size-dependent optical
adsorption behaviors. Anatase TiO2 nanotubes are formed by annealing in air above 300
C. Our method can
easily create highly ordered TiO2 nanotube arrays on a wide range of substrates without causing chemical
contamination and changing surface morphology on the substrates. The prepared TiO2 nanotube arrays show
good contact with the substrate, which is of great importance in device applications.
The fabrication of TiO2 nanotube arrays directly on substrates
are highly sought after in applications such as sensing,1-3
Several methods such as hydrothermal synthesis,11-13
Compared with the method of using a free-standing PAA
template as a mask for nanostructure fabrication, the PAA-substrate method has an advantage of having intimate contact
between the template and the substrate, which allows for both
vapor- and solution-based fabrication.24-26
In this paper, we demonstrate the fabrication of dense, uniform, highly ordered, and well-aligned arrays of TiO2 nanotubes on various substrates, over large areas, via ALD on free-standing PAA templates. In comparison with a PAA-substrate method, this method can easily fabricate highly ordered TiO2 nanotube arrays on various substrates (Si, glass, and flexible polyimide substrates are demonstrated) without changing the surface morphology and producing chemical contaminations on the underlying substrate. The large-area free-standing PAA (>4.5 cm2) template can be easily transferred and attached onto the substrates due to a PMMA protective layer. The adhesion of the template and the substrate are improved by pretreating the substrate with mild O2 plasma and applying acetone to the template after the attachment. The ALD TiO2 nanotube arrays prepared by this free-standing PAA template also show intimate contact with the substrate and demonstrate size-dependent optical adsorption behaviors due to the precise wall-thickness control offered by the ALD technique.
Attaching Free-Standing PAA onto Substrates. Free-standing PAA templates were prepared by two-step anodization
of high purity (99.999%) Al foils (Goodfellow Cambridge) as
published previously.22 Briefly, the electrochemically polished
Al foil was anodized in 0.3 M oxalic acid at room temperature
for 6-10 h at 40 V. After removal of the as-prepared PAA
film with a mixture of 3.5 vol % H3PO4 and 45 g/L CrO3 at
~60
C, a second anodization at 2
C was performed. The
desired template thickness depends on the anodization time, with
a rate of ~40 nm/min. The pores of the as-prepared template
were widened in 5 wt % H3PO4 for ~50 min before applying
a protective poly(methyl methacrylate) (PMMA) layer. This
layer was subject to drying at 120
C for ~30 min prior to
separating the template from the Al foil with 0.1 M CuSO4 and
10 wt % HCl. The barrier layer at the bottom of the template
was etched in 5 wt % H3PO4 solution for another 30-40 min.
The template was then carefully attached to three different
substrates: p-type Si (100) with 50 nm SiO2 layer, glass, and
flexible polyimide (Dupont-Kapton) film. The substrates were
pretreated with mild O2 plasma (80 sccm O2, 80 mTorr, 150
W) for ease of attaching PAA to them. The PMMA protective
layer was removed by UV-ozone in a dry stripper (Samco UV-1) at 200
C for ~30 min. Finally, drops of acetone were added
to the attached template to further reduce any voids.
ALD of TiO2 Nanotube Arrays. Free-standing PAA templates attached on the substrates were deposited with TiO2 films
in a home-built ALD setup at room temperature. The substrates
were alternately exposed to vapors of TiCl4 (Merck,
99%)
and deionized H2O at a base pressure of 1 × 10-3 Torr. Both
precursors have a 2-s exposure time and a 30-s interval between
the two exposures. Reactive ion etching (RIE, Oxford Plasmalab
80, 55 sccm CHF3, and 5 sccm O2) was employed to etch the
TiO2 overlay on top of the template. The nanotube arrays were
then released from the template by immersion in 1 M KOH for
~15 min followed by a rinse in deionized water.
Characterization Methods. Field emission scanning electron
microscopy (FESEM, JEOL-6700F) was used to characterize
morphology of the free-standing PAA template and TiO2
nanotube array. Individual TiO2 nanotubes and cross-sectional
arrays of as-deposited TiO2 nanotubes in the free-standing PAA
template were studied by high-resolution transmission electron
microscopy (HRTEM, Philips CM300). The arrays of TiO2
nanotubes were scrapped off from the Si substrate and dispersed
in ethanol and onto a TEM grid. The cross-sectional TEM
sample was prepared by bonding the top surfaces of 2 pieces
of 3 mm × 1 mm samples with epoxy glue. The sample was
cured and subsequently polished using graphite lapping film
until about 10-50
m thick. The sample was dimpled and
finally ion milled until transparent. X-ray diffractograms of the
TiO2 film on Si were measured with a Bruker D8 GADDS XRD.
UV-vis spectrum for TiO2 nanotube arrays on glass substrate
were performed using a Shimadzu UV-3101PC instrument.
A schematic of the fabrication process is shown in Figure 1. A large-area free-standing PAA template supported by a PMMA layer is attached onto the substrate. Subsequently, the protective layer is removed to form a through-hole PAA template on the substrate for TiO2 ALD. The contact between the template and the substrate is improved by pretreating the substrate with O2 plasma and addition of acetone to the attached template. TiO2 films are deposited via ALD on the template at room temperature. The TiO2 overlay on top of the PAA template is dry etched by reactive ion etching, and the template is removed by wet chemical etch to finally release highly ordered TiO2 nanotube arrays on the substrate.
Figure 2 demonstrates the versatility of our method by attaching the free-standing PAA template on various substrates. In the PAA-substrate method, various substrates show different anodization behaviors at the end of anodization, and great cares need to be taken to avoid formation of pores and pits on the substrate, or worse, template detachment from the substrate.25,26 However, the free-standing PAA method is substrate-friendly and shows no damages to the substrate. As shown in Figure 2a, the free-standing PAA template with an average area >4.5 cm2 are firmly attached on the substrates such as glass (left), Si (middle), and flexible polyimide films (right). To make a large-area free-standing PAA template, a thin film of PMMA was used as a support layer for the ultrathin (~300 nm) and brittle template.30,32 UV-ozone treatment was applied to remove this protective layer, and the free-standing template is still intact after the removal. Voids between the template and substrate are not desirable in this process. To avoid these voids, substrates such as glass and polyimide films were pretreated with mild O2 plasma to modify surfaces hydrophilicity before attachment. In addition, drops of acetone were added onto the attached template as capillary forces produced during the evaporation of acetone can further reduce the voids. The attachment of free-standing PAA template can also be further extended to polymer substrates such as flexible polyimide film. Direct anodization of Al on the flexible film is however difficult to be achieved.24 Unlike the PAA-substrate method, in principle, our method can be applied to any substrates, though only three types of substrates were illustrated here. Furthermore, the highly ordered pores in the free-standing PAA template are easily achieved by a well-established two-step anodization, and no special cares are required as compared with the fabrication of highly ordered PAA-substrate.22,23 As shown in Figure 2b, the free-standing PAA template via two-step anodization shows hexagonally packed pores with a diameter of ~65 nm and interpore distance of ~110 nm. The shape, size, and interpore distance can be well-controlled by varying the anodization conditions as reported elsewhere.22 Although the PAA-substrate method has numerous advantages, the fabrication of free-standing PAA template is much easier, simpler, and more flexible.
The free-standing PAA template attached on the Si substrate after removal of the PMMA protective layer is shown in Figure 3a (oblique-view SEM image). The sample is cleaved to reveal the cross-sectional structure on the Si substrate. The template is in good contact with the substrate after cleavage, and no voids are observed along the interface between the template and the substrate, indicating firm contact between them. The free-standing PAA template is estimated to be ~300 nm thick. It is apparent in the SEM image that the surface morphology of the Si substrate remains unmodified, unlike the PAA-substrate method that will produce anodized pores and pits on the substrate at the end of Al anodization.27 The image also shows that the PMMA protective layer has been completely removed by UV-ozone treatment, whereas the template remains intact. Using a PMMA protective layer to facilitate the transfer of ultrathin PAA films has been previously reported,30,32 but in these reports, removal of the protective layer by acetone rinse tends to cause cracks and detachment of the template. Our dry-etching-based method overcomes this drawback and makes it possible to fabricate intact free-standing PAA templates on substrates over large areas. After the removal of the protective layer from the template, it was then subject to TiO2 ALD. As shown in Figure 3b, the pores are significantly reduced after 150-cycle ALD at a growth rate of 1.0 Å/cycle, and the template is still firmly in contact with the substrate. This is attributed to the conformal coating of ALD not only on in-plane surfaces but also along walls of PAA pores. The TiO2 overlayer can be removed by RIE etching, and the free-standing PAA template is easily removed by chemical etching. After these processes, the TiO2 nanotube arrays on Si substrate are fully released from the template, as shown in Figure 3c. The dimensions of the TiO2 nanotube can be finely tuned by correspondingly varying the height and interpore distance of the template in Al anodization,22 while the nanotube wall thickness is determined by the number of ALD cycles. The TiO2 nanotube arrays in Figure 3c are highly ordered, dense, and well-aligned, inheriting the patterns of the free-standing PAA template. The cross section of the well-aligned TiO2 nanotube arrays was illustrated in Figure 3d.
The morphologies of the TiO2 nanotubes were examined under TEM by scrapping the nanotubes from the Si substrate (Figure 4a) and cleaving the sample for cross sectional imaging (Figure 4b). As shown in Figure 4a, the nanotube wall is uniform in thickness along the length of the nanotube and its bottom is closed due to ALD also occurring on the underlying substrate. The bottom layers of the nanotubes are direct evidence of good contacts between the nanotube arrays and the substrate. The intimate contact between them is vital for device applications. Free-standing PAA templates are widely used as masks for line-of-sight deposition such as sputtering and evaporation. In these processes, materials can be deposited only on the top and bottom surfaces of the template. As such, small voids between the template and the substrate may not have any impact for the formation of nanoparticle arrays. However, these voids can seriously affect the contact between nanotube arrays and the substrate because ALD is a surface-limiting and conformal film-coating technique. Micrometer voids between the interfaces cannot be completely filled by ALD. To further check the contact, the interface between as-deposited TiO2 nanotube arrays and its substrate is examined with cross-sectional TEM imaging. The TEM image in Figure 4b reveals that the as-deposited nanotube arrays in the free-standing PAA template are in good contact with the Si substrate (a thin layer of SiO2 (~50 nm) is deposited on the Si substrate). This might be attributed to the conformal film coating of ALD and the good contact between the free-standing template and the substrate. With the pretreatment of the substrate with mild O2 plasma and the use of acetone to attach the template to the substrate, we are able to achieve close contact between the template and the substrate. Therefore, ALD might be a convenient and promising method as nanoglues for bonding free-standing PAA templates on substrates.33
Figure 5 shows the XRD spectrum of ~30 nm TiO2 film
grown by ALD on thermally oxidized Si(100). There are no
sharp diffraction peaks for the as-deposited film, films annealed
at 100 and 200
C for 1 h in air. This is in good agreement
with previous studies since amorphous films are deposited
at room temperature and postdeposition annealing is needed
to improve crystallinity.34,35
C and above for 1 h in air, diffraction
peaks corresponding to anatase TiO2 were apparent. The
crystallinity of the ALD TiO2 film is crucial in many applications.
Figure 5 XRD spectra of ~30-nm ALD TiO2 film on Si (100). The
as-deposited ALD film is amorphous, and anatase peaks were seen after
annealing at 300 C and above for 1 h in air.
|
The optical absorption behaviors of the as-deposited ALD TiO2 nanotube arrays in the free-standing PAA template on transparent glass substrate were investigated by UV-vis spectroscopy. By variation of the ALD cycles, the normalized absorption spectra of the nanotubes demonstrate a red shift with increasing nanotube wall thickness, as shown in Figure 6. A tunable red-shift in the adsorption edge was observed with increasing ALD cycles from 0 to 100 cycles as similar to previous reports.26 Although reasons for this phenomenon are still controversial,26,36 it is clear that the precise wall-thickness control offered by ALD gave rise to tunable optical adsorption properties of the nanotube arrays. The ability to control nanotubes' electronic structures provided by ALD is very useful in applications such as photocatalysis and photovoltaics, where wall thickness affects photoconversion efficiency.
| Figure 6 UV-vis spectra of highly ordered ALD TiO2 nanotube arrays on the glass substrate. A tunable red-shift to lower energies is shown with increasing the thickness of nanotube walls. |
In summary, we have demonstrated the fabrication of highly ordered TiO2 nanotube arrays on various substrates via ALD on free-standing PAA templates attached on the substrate. As compared to the PAA-substrate method, this is a simple method to create highly ordered nanotube arrays on substrates over large areas without affecting the surface morphology and causing chemical contamination on the substrate. As no anodization of the substrate is involved, this method can be used for fabrication of nanotubes on a wide range of substrates, such as Si, glass, and flexible polyimide films, as demonstrated. The conformal film coating and precise wall-thickness control by ALD give rise to dense, uniform, well-aligned, and dimension-tunable TiO2 nanotube arrays that show size-dependent optical adsorption behaviors and have an intimate contact with substrates. This is also a general method that can fabricate nanotube arrays of various materials on a wide range of substrates.
* To whom correspondence should be addressed. Phone: (65) 6872 7526. Fax: (65) 6772 7744. Email: h-gao@imre.a-star.edu.sg.
1. Varghese, O. K.; Gong, D. W.; Paulose, M.; Ong, K. G.; Dickey,
E. C.; Grimes, C. A. Adv. Mater. 2003, 15, 624-627.
2. Mor, G. K.; Carvalho, M. A.; Varghese, O. K.; Pishko, M. V.;
Grimes, C. A. J. Mater. Res. 2004, 19, 628-634.
3. Shyue, J. J.; Cochran, R. E.; Padture, N. P. J. Mater. Res. 2006,
21, 2894-2903.
4. Uchida, S.; Chiba, R.; Tomiha, M.; Masaki, N.; Shirai, M.
Electrochemistry 2002, 70, 418-420.
5. Mor, G. K.; Varghese, O. K.; Paulose, M.; Shankar, K.; Grimes,
C. A. Sol. Energy Mater. Sol. Cells 2006, 90, 2011-2075.
6. Mor, G. K.; Shankar, K.; Paulose, M.; Varghese, O. K.; Grimes,
C. A. Nano Lett. 2006, 6, 215-218.
7. Zhu, K.; Neale, N. R.; Miedaner, A.; Frank, A. J. Nano Lett. 2007,
7, 69-74.
8. Adachi, M.; Murata, Y.; Harada, M.; Yoshikawa, S. Chem. Lett.
2000, 942-943.
9. Kemell, M.; Pore, V.; Tupala, J.; Ritala, M.; Leskela, M. Chem.
Mater. 2007, 19, 1816-1820.
10. Chu, S. Z.; Inoue, S.; Wada, K.; Li, D.; Haneda, H.; Awatsu, S. J.
Phys. Chem. B 2003, 107, 6586-6589.
11. Chen, Q.; Zhou, W. Z.; Du, G. H.; Peng, L. M. Adv. Mater. 2002,
14, 1208.
12. Yao, B. D.; Chan, Y. F.; Zhang, X. Y.; Zhang, W. F.; Yang, Z. Y.;
Wang, N. Appl. Phys. Lett. 2003, 82, 281-283.
13. Kasuga, T.; Hiramatsu, M.; Hoson, A.; Sekino, T.; Niihara, K.
Langmuir 1998, 14, 3160-3163.
14. Quan, X.; Yang, S. G.; Ruan, X. L.; Zhao, H. M. Environ. Sci.
Technol. 2005, 39, 3770-3775.
15. Gong, D.; Grimes, C. A.; Varghese, O. K.; Hu, W. C.; Singh, R.
S.; Chen, Z.; Dickey, E. C. J. Mater. Res. 2001, 16, 3331-3334.
16. Zheng, Q.; Zhou, B. X.; Bai, J.; Cai, W. M.; Liao, J. S. Prog. Chem.
2007, 19, 117-122.
17. Zhao, J. L.; Wang, X. H.; Chen, R. Z.; Li, L. T. Solid State
Commun. 2005, 134, 705-710.
18. Hoyer, P. Langmuir 1996, 12, 1411-1413.
19. Sander, M. S.; Cote, M. J.; Gu, W.; Kile, B. M.; Tripp, C. P. Adv.
Mater. 2004, 16, 2052.
20. Shin, H. J.; Jeong, D. K.; Lee, J. G.; Sung, M. M.; Kim, J. Y. Adv.
Mater. 2004, 16, 1197.
21. Tian, Z. R. R.; Voigt, J. A.; Liu, J.; McKenzie, B.; Xu, H. F. J.
Am. Chem. Soc. 2003, 125, 12384-12385.
22. Li, F. Y.; Zhang, L.; Metzger, R. M. Chem. Mater. 1998, 10, 2470-2480.
23. Masuda, H.; Fukuda, K. Science 1995, 268, 1466-1468.
24. Chong, A. S. M.; Tan, L. K.; Deng, J.; Gao, H. Adv. Funct. Mater.
2007, 17, 1629.
25. Rabin, O.; Herz, P. R.; Lin, Y. M.; Akinwande, A. I.; Cronin, S.
B.; Dresselhaus, M. S. Adv. Funct. Mater. 2003, 13, 631-638.
26. Sander, M. S.; Tan, L. S. Adv. Funct. Mater. 2003, 13, 393-397.
27. Chen, P. L.; Kuo, C. T.; Tsai, T. G.; Wu, B. W.; Hsu, C. C.; Pan,
F. M. Appl. Phys. Lett. 2003, 82, 2796-2798.
28. Tian, M. L.; Xu, S. Y.; Wang, J. G.; Kumar, N.; Wertz, E.; Li, Q.;
Campbell, P. M.; Chan, M. H. W.; Mallouk, T. E. Nano Lett. 2005, 5,
697-703.
29. Lau, K. H. A.; Tan, L. S.; Tamada, K.; Sander, M. S.; Knoll, W.
J. Phys. Chem. B 2004, 108, 10812-10818.
30. Gao, H.; Gosvami, N. N.; Deng, J.; Tan, L. S.; Sander, M. S.
Langmuir 2006, 22, 8078-8082.
31. Crouse, D.; Lo, Y. H.; Miller, A. E.; Crouse, M. Appl. Phys. Lett.
2000, 76, 49-51.
32. Lei, Y.; Chim, W. K. J. Am. Chem. Soc. 2005, 127, 1487-1492.
33. Jung, H. Y.; Jung, S. M.; Gu, G. H.; Suh, J. S. Appl. Phys. Lett. 2006, 89.
34. Aarik, J.; Aidla, A.; Uustare, T.; Sammelselg, V. J. Cryst. Growth
1995, 148, 268-275.
35. Kumagai, H.; Matsumoto, M.; Toyoda, K.; Obara, M.; Suzuki, M.
Thin Solid Films 1995, 263, 47-53.
36. Yuwono, A. H.; Xue, J. M.; Wang, J.; Elim, H. I.; Ji, W.; Li, Y.;
White, T. J. J. Mater. Res. 2003, 13, 1475-1479.