Hierarchical Structures
Novel Three-Dimensional Nanoporous Alumina as a Template for Hierarchical TiO2 Nanotube Arrays Daoai Wang,* Lianbing Zhang, Woo Lee, Mato Knez, and Lifeng Liu* Porous anodic aluminium oxide (AAO), made by electrochemical anodization of aluminium, has been extensively investigated in the past two decades for its widespread applications in the field of nanoscience and nanotechnology. Besides its use as photonic crystals,[1,2] sensors,[3] bio-separators[4] and superhydrophobic supports,[5,6] porous AAO is very likely the most commonly used hard template, which allows fabrication of a broad spectrum of 1D nanostructures.[7–20] In general, the anodization of aluminium has to be carried out under relatively mild conditions with a chemically or electrochemically polished mirror-finished Al surface. Any harsh conditions, e.g., an elevated temperature and/or an increased concentration of electrolyte, a high voltage or a rough Al surface, could cause a very high local current flow, eventually leading to a catastrophic ‘breakdown’ of the substrate.[21–23] So far, a few attempts have been made in order to avoid a breakdown during an anodization.[24–29] For instance, Chu et al. found that aged sulfuric acid helps to improve the critical anodizing potential for a breakdown.[24] Lee et al. demonstrated that a pre-anodized oxide layer (>400 nm) is able to provide uniform pore nucleation sites and therefore effectively suppresses the breakdown during hard or mild anodization in phosphoric acid (H3PO4) at a high voltage (e.g., 195 V).[26,27] Nevertheless, a very smooth, mirror-finished aluminium surface as well as a low anodizing temperature are still needed in order to avoid failure of anodization under harsh conditions. Particularly, anodization of Al in H3PO4 at
Dr. D. Wang, Dr. L. Zhang, Prof. M. Knez, Dr. L. Liu Max Planck Institute of Microstructure Physics Weinberg 2, D-06120 Halle, Germany E-mail:
[email protected];
[email protected] Dr. L. Zhang, Prof. M. Knez CIC nanoGUNE Consolider Tolosa Hiribidea 76, 20018 Donostia-San Sebastian, Spain Prof. W. Lee Korea Research Institute of Standards and Science (KRISS) Yuseong 305-340 Daejeon, South Korea Prof. M. Knez Ikerbasque Basque Foundation for Science Alameda Urquijo 36-5, 48011 Bilbao, Spain Dr. L. Liu International Iberian Nanotechnology Laboratory (INL) 4715-330 Braga, Portugal DOI: 10.1002/smll.201201784 small 2013, 9, No. 7, 1025–1029
a high voltage still remains challenging. Any surface pits or defects will likely result in a burnt Al foil. We recently found that anodizing Al in phosphoric acid (H3PO4) can be accomplished at a high voltage (195 V) using an extremely rough, chemically etched microstructured aluminium foil. Even under conditions as harsh as 10 wt% H3PO4 and room temperature (∼22 °C), we did not observe any signs of breakdown during the entire anodization process. The as-anodized Al exhibits a hierarchical micro-/nanostructured surface with 3D distributed nanopores. Moreover, the dimensions of the microstructures and nanopores can be independently tuned by the chemical etching process and the anodization conditions, respectively. The resulting hierarchical 3D AAO can be used as template to fabricate a number of micro-/nano-structures with controllable morphology, offering a generic route for the preparation of hierarchical architectures. Furthermore, the micrometer-sized Al terraces on the surface of the microstructured Al foils show very interesting phenomena upon anodization. For instance, it allowed observation of the evolution of nanopores at the edges of an Al micro-terrace during anodization. Microstructuring Al surface was achieved by chemically etching the as-received Al foils (99.999%, Goodfellow) in a mixed solution of CuCl2 and HCl at room temperature. Figure 1 shows representative scanning electron micrographs (SEM, JEOL 6701F) of the microstructured Al surface etched in different solutions for the same etching duration (5 min). It is evident that the etching exclusively results in the formation of numerous interconnected and hierarchically arranged micrometer-sized terraces on the Al surface. The average size of these terraces can be well tuned from several micrometers to a few tens of nanometers by adjusting the ratio of HCl to CuCl2. Obviously, the etching treatment greatly increases the effective anodizing area as compared to planar Al. Figures 2 reveals the top-view and cross-sectional view SEM micrographs of the 3D hierarchical nanopore structure obtained by anodizing microstructured Al foils in 10 wt% H3PO4 at 10 °C. More information about the sample appearance and morphology as well as the anodizing currenttime transients is presented in Figures S1 and S2 in the Supporting Information. Upon anodization, a hierarchical 3D nanoporous architecture that combines both micrometer-sized terraces and nanometer-sized channels on each terrace is obtained. Interestingly, the microstructured Al provides a unique platform to explore the 3D evolution of oxide nanopores during anodization.[30–32] In the traditional
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
wileyonlinelibrary.com
1025
communications
D. Wang et al.
Additionally, two differing nanopore growth modes were observed near the edges of the terraces: (i) Near the concave edges, where two adjacent terraces connect, branched nanopores are observed, as visibly enhanced with black lines in Figure 2d. (ii) Around the convex edges, the nanopores initially grow perpendicularly to the terrace faces but subsequently bend, as shown in Figure 2e. This bending is a result of the interaction of the electric fields at the anodization front when the bottoms of nanopores evolving from two neighboring facets are approaching each other. We believe that the unusual growth modes of the nanopores are closely related to the electric field distribution near the edges and the buildup of internal stress of the micro-cubes. It is inferred that the growth of nanopores near the vertexes of micro-terraces should be more complex. However, it is practically difficult to examine the overview morphology of this region. It is worth mentioning that according to our experience, the probability of failure (i.e., breakdown) in anodizing Al in phosphoric acid is as high as 20–30%. This applies even under optimal conditions with a pre-anodization treatment,[26,27] namely, using an electropolished mirror-finished Al surface at 195 V in 1 wt% H3PO4 at Figure 1. SEM micrographs of microstructured Al surfaces obtained by etching. The etching 1 °C. In stark contrast, we never suffered was conducted at room temperature in a mixed solution containing 0.2 M CuCl2 and HCl with varying concentrations (a) 0 M, (b) 0.5 M, (c) 1 M, (d) 2 M, (e) 4 M, and (f) 6 M. The etching from breakdown during anodizing microstructured Al, even under harsh conditime was 5 min in all cases. tions. Obviously, the microstructured Al has a much higher tolerance against the way to prepare porous alumina, the nucleation and growth of critical conditions. In fact, we found that the microstructured nanopores usually occur two-dimensionally on an extended Al was also able to sustain without breakdown upon anodiflat Al surface. As a result, the nanopores grow vertically to zation in 0.3 M sulfuric acid at an anodic voltage as high as the Al surface, giving rise to a close-packed array of straight 190 V (Figure S3, Supporting Information). The high breakand parallel nanochannels. For the microstructured Al, it is down resistance can be ascribed to the following virtues obvious that the distribution of the electric field governing of microstructured Al: (i) As revealed by exponentially the growth of nanopores must be rather complicated. How- decreasing current profiles in I-t curves (Figure S1), 3D ever, extensive SEM examinations demonstrated that the hierarchical micrometer-scale features on the Al surface are vertical growth mode of nanopores still applies to all exposed highly effective for accumulating stress in anodic oxide. The Al facets regardless of their orientations (Figure 2c). It is also stress-driven retardation of oxidation kinetics may positively noteworthy that in some places the Al residing in the interior contribute to suppressing the breakdown of anodic oxide. of micro-cubes was not completely oxidized. An example is The present stress relevance of oxidation kinetics may also shown in Figure 2c as an area marked with a black line. This explain the effective suppression of breakdown during hard observation is consistent with the anodization behavior of Al anodization of Al with an oxide layer.[26,34] (ii) Microstrucin a confined micrometer-sized trench,[33] and implies that the tured Al has a much larger surface area in comparison to a accumulation of compressive stress during anodization plays flat Al foil. The effective anodizing current density of microan important role in governing the oxidation kinetics. Previ- structured Al samples, even if a high anodic voltage is applied, ously, Lee et al. observed long Al nanopillars at the irregular is not as high as it’s supposed to be for the flat Al foil, and cell junctions of porous alumina formed by malonic acid- most likely is lower than the threshold current density that based hard anodization.[34] We believe that the formation of may cause a breakdown. This also indicates that it is the curAl nanopillars is also associated with stress-induced retarda- rent density (i.e. the effective electric field strength across tion of anodic oxidation under high electric field. the barrier oxide), but not the anodic voltage that governs
1026 www.small-journal.com
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2013, 9, No. 7, 1025–1029
Templated Hierarchical TiO2 Nanotube Arrays
well-defined hemi-spherical ends is clearly seen. Figure 3c displays a magnified SEM micrograph showing the morphology of the convex edge of the terrace, from which many branched nanotubes are observed. These branched nanotubes should be the replica of the branched nanopores formed near the concave edges of the Al terraces during anodization (Figure 2d). Obviously, the TiO2-ALD process reproduced the structure of the parent template with a high fidelity. In Figure 3d, a broken section of the sample is shown, in which the tubular structure is apparent. Furthermore, transmission electron microscopy (TEM, JEOL-1010) investigations also confirmed the tubular nature of the fabricated nanostructures. Both straight and branched nanotubes can be observed by TEM, as shown in Figure 3e,f. The measured thickness of the tube walls is approximately 20 nm. The inset of Figure 3f exhibits an electron diffraction (ED) pattern of the annealed TiO2 nanotubes, indicating that the nanotubes are polycrystalline after thermal annealing. Besides TiO2 nanotubes, we also demonstrated that hierarchical 3D polystyrene nanorod arrays can be fabricated by polymer wetting using the as-obtained 3D AAO as templates (Supporting Information, Figure S5). We performed initial tests on the phoFigure 2. (a,b) Top-view SEM micrographs of the as-anodized 3D hierarchical nanopore tocatalytic activity of the as-fabricated structures (10 wt% H3PO4, 10 °C, 195 V). (c) Cross-sectional SEM image showing growth of hierarchical 3D TiO2 nanotube arrays vertical nanopores on the micro-terraces of Al. The enclosed area represents the remained non-anodized Al due possibly to the internal stress (d) Branched nanopores evolved at the towards photo-degradation of methylene concave edges visibly enhanced by black lines. (e) Bent nanopores at the convex edges, blue (MB), a model pollutant frequently where the black arrows indicate the growth direction of nanopores. (f) 3D hierarchical nano- used to evaluate photocatalysts.[36] For forest structures after pore widening treatment (5 wt% H3PO4, 30 °C, 1 h). comparison, we also tested the photocatalytic performance of a flat ALD-TiO2 film the anodization behavior.[35] These merits may reasonably and a TiO2 nanotube array fabricated by ALD using a conexplain why the microstructured Al can be sustained during ventional porous AAO membrane as template (2D TiO2 nananodization in H3PO4 under such harsh conditions. otube arrays). Figure 4 shows the degradation profile of the Analogous to traditional porous AAO membranes, the MB solution as a function of time upon UV illumination. The as-prepared hierarchical 3D porous alumina can also be used results show that among the prepared samples the 3D TiO2 as template to fabricate hierarchical micro-/nano-architec- nanotube arrays exhibit the best photocatalytic performtures that cannot be easily realized by conventional physical ance towards degradation of MB owing to their large surface and chemical routes. As a proof-of-concept, we demonstrate area and the enhanced light-harvesting capability possibly in this communication the fabrication of hierarchical 3D resulting from their hierarchical structure. titania (TiO2) nanotube arrays (Figure 3, see also Figure S4, In summary, we report a new strategy to fabricate porous Supporting Information) using the as-anodized 3D porous alumina with a hierarchical micro-/nano-porous morphology AAO as a template. Figure 3 represents SEM micrographs without suffering from breakdown. The surprisingly high of a free-standing hierarchical 3D TiO2 nanotube array. The tolerance of microstructured aluminium to critical anodizaarray was prepared by atomic layer deposition (ALD) of tion conditions is ascribed on the one hand, to the accumuTiO2 onto the as-anodized 3D AAO membrane, followed by lated stress in the anodic oxide which retards the anodization thermal annealing treatment in air at 450 °C for 2 h in order kinetics and thereby suppresses breakdown, and on the other to obtain a crystalline phase. From Figure 3a we can see that hand, to the large surface area of microstructured aluminium the replicated hierarchical architecture remains upon thermal that helps to reduce the effective anodizing current density: a annealing. Figure 3b is a magnified SEM image of the terrace governing factor of breakdown. These 3D AAO membranes surface, in which a vertically aligned array of nanotubes with can be utilized as templates to fabricate hierarchical 3D TiO2 small 2013, 9, No. 7, 1025–1029
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.small-journal.com
1027
communications
D. Wang et al.
nanotube arrays and polystyrene nanorod arrays by using atomic layer deposition and polymer wetting, respectively. Given the fact that a general method for preparing hierarchical micro-/nano-structures is presently still absent, our approach reported here represents an important advance in this regard and will have a great impact on the development of many applications such as solar cells, batteries and photo-/electro-catalysts, where the use of hierarchical structures is preferred.
Experimental Section Preparation of AAO membranes: The conventional AAO membranes were prepared by a two-step anodization process, as reported previously.[37] In order to obtain hierarchical 3D porous AAO membranes, the highly pure Al foils were first chemically etched in a mixture of CuCl2 (0.2 M) and HCl (0.5–2 M) at room temperature for 5 min. Afterwards, the microstructured Al substrates were directly anodized at 195 V in 1–10 wt% H3PO4 solution for 3 h at various temperatures ranging from 1 °C to room temperature. The free-standing hierarchical 3D porous AAO was obtained by selectively etching the remaining Al substrate away in a mixed aqueous solution of CuCl2 (0.2 M) and HCl (4 M). Figure 3. (a–d) SEM micrographs of the hierarchical 3D TiO2 nanotube arrays fabricated by Atomic Layer Deposition (ALD) of TiO2: The ALD of TiO2 using 3D porous AAO as a template. (e,f) Typical TEM images of the TiO2 nanotubes deposition of TiO2 was carried out in a comannealed at 450 °C in air for 2 h. Inset of (f): electron diffraction pattern. mercial ALD chamber (Savannah 100, Cambridge Nanotechnology Inc.) utilizing Ar both as precursor carrier and purge gas at a pressure of 0.2 Torr. Titanium tetra-isopropoxide (TIP) and deionized H2O were used as precursors. The pulse durations for TIP and H2O were 1 s and 1.3 s, respectively, and the purge time for both precursors was 60 s. The deposition was performed at 150 °C for 800 cycles, which resulted in film thickness of 20 nm. Three different substrates, namely, a bare Si wafer, a conventional porous AAO membrane and a hierarchical 3D porous AAO membrane, were used in order to obtain a compact TiO2 film, a 2D TiO2 nanotube array and a 3D TiO2 nanotube array, respectively. For photocatalytic studies, all as-deposited samples were annealed at 450 °C in air for 2 h to obtain crystalline anatase TiO2. Characterization: The morphology and structure of the samples were examined by a field emission scanning electron microscope (JEOL, JSM-6701F) and a transmission electron microscope (JEOL, JEM-1010). The photocatalytic activity of the samples was evaluated by degrading 5 × 10−5 M methylene blue aqueous solution under the irradiation of a UV light (9 W). All the samples were installed Figure 4. Photocatalytic performance of the hierarchical 3D TiO2 into a home-made cell where only 1 cm2 area was exposed to the nanotube arrays, 2D TiO2 nanotube arrays and a compact TiO2 film analyte solution and UV light for investigation. The concentration of towards the degradation of methylene blue under UV illumination. The TiO2 was deposited using ALD, followed by a thermal treatment in air at MB in the analyte solution was monitored by measuring the max450 °C for 2 h. The deposited TiO2 was 20 nm thick for all samples (with imum absorption of the methylene blue at 668 nm with a micro-UVan empirical deposition rate of 0.25 Å/cycle for 800 cycles). Vis spectrometer (NanoDrop 1000, Thermo Scientific).
1028 www.small-journal.com
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
small 2013, 9, No. 7, 1025–1029
Templated Hierarchical TiO2 Nanotube Arrays
Supporting Information Supporting Information is available from the Wiley Online Library or from the author.
[1] J. Choi, Y. Luo, R. B. Wehrspohn, R. Hillebrand, J. Schilling, U. Gösele, J. Appl. Phys. 2003, 94, 4757. [2] H. Masuda, M. Ohya, H. Asoh, M. Nakao, M. Nohtomi, T. Tamamura, Jpn. J. Appl. Phys. 1999, 38, 1403. [3] O. K. Varghese, D. W. Gong, M. Paulose, K. G. Ong, C. A. Grimes, E. C. Dickey, J. Mater. Res. 2002, 17, 1162. [4] A. Thormann, N. Teuscher, M. Pfannmoller, U. Rothe, A. Heilmann, Small 2007, 3, 1032. [5] W. Wu, X. Wang, D. Wang, M. Chen, F. Zhou, W. Liu, Q. Xue, Chem. Commun. 2009, 9, 1043. [6] X. Wang, X. Liu, F. Zhou, W. Liu, Chem. Commun. 2011, 47, 2324. [7] B. B. Lakshmi, C. J. Patrissi, C. R. Martin, Chem. Mater. 1997, 9, 2544. [8] B. Pulamagatta, M. Y. E. Yau, I. Gunkel, T. Thurn-Albrecht, K. Schroter, D. Pfefferkorn, J. Kressler, M. Steinhart, W. H. Binder, Adv. Mater. 2010, 23, 781. [9] M. Steinhart, S. Zimmermann, P. Goring, A. K. Schaper, U. Gösele, C. Weder, J. H. Wendorff, Nano Lett. 2005, 5, 429. [10] L. F. Liu, W. Lee, R. Scholz, E. Pippel, U. Gösele, Angew. Chem. Int. Ed. 2008, 47, 7004. [11] X. D. Chen, S. Yeganeh, L. D. Qin, S. Z. Li, C. Xue, A. B. Braunschweig, G. C. Schatz, M. A. Ratner, C. A. Mirkin, Nano Lett. 2009, 12, 3974. [12] W. Lee, H. Han, A. Lotnyk, M. A. Schubert, S. Senz, M. Alexe, D. Hesse, S. Baik, U. Gösele, Nat. Nanotechnol. 2008, 3, 402. [13] L. F. Liu, E. Pippel, R. Scholz, U. Gösele, Nano Lett. 2009, 9, 4352. [14] N. Peranio, E. Leister, W. Tollner, O. Eibl, K. Nielsch, Adv. Funct. Mater. 2012, 22, 151. [15] K. Pitzschel, J. M. M. Moreno, J. Escrig, O. Albrecht, K. Nielsch, J. Bachmann, ACS Nano 2009, 3, 3463.
small 2013, 9, No. 7, 1025–1029
[16] L. Liu, S. H. Yoo, S. A. Lee, S. Park, Nano Lett. 2011, 11, 3979. [17] L. D. Qin, S. Park, L. Huang, C. A. Mirkin, Science 2005, 309, 113. [18] Y. Qin, A. L. Pan, L. F. Liu, O. Moutanabbir, R. B. Yang, M. Knez, ACS Nano 2011, 5, 788. [19] D. Wang, L. Liu, Y. Kim, Z. Huang, D. Pantel, D. Hesse, M. Alexe, Appl. Phys. Lett. 2011, 98, 243109. [20] Y. Yang, L. F. Liu, F. Guder, A. Berger, R. Scholz, O. Albrecht, M. Zacharias, Angew. Chem. Int. Ed. 2011, 50, 10855. [21] S. Ikonopisov, Electrochim. Acta 1977, 22, 1077. [22] J. M. Albella, I. Montero, J. M. Martínez-Duart, Electrochim. Acta 1987, 32, 255. [23] S. Ono, M. Saito, H. Asoh, Electrochem. Solid-State Lett. 2004, 7, 21. [24] S. Z. Chu, K. Wada, S. Inoue, M. Isogai, A. Yasumori, Adv. Mater. 2005, 17, 2115. [25] S. Y. Zhao, K. Chan, A. Yelon, T. Veres, Adv. Mater. 2007, 19, 3004. [26] W. Lee, R. Ji, U. Gösele, K. Nielsch, Nat. Mater. 2006, 5, 741. [27] W. Lee, R. Scholz, K. Nielsch, U. Gösele, Angew. Chem. Int. Ed. 2005, 44, 6050. [28] G. D. Sulka, A. Brzozka, L. Liu, Electrochim. Acta 2011, 56, 4972. [29] W. Lee, K. Schwirn, M. Steinhart, E. Pippel, Z. Scholz, U. Gösele, Nat. Nanotechnol. 2008, 3, 234. [30] A. Yin, R. S. Guico, J. Xu, Nanotechnology 2007, 18, 035304. [31] R. Zakeri, C. Watts, H. Wang, P. Kohli, Chem. Mater. 2007, 19, 1954. [32] O. Kopp, M. Lelonek, M. Knoll, Electrochim. Acta 2009, 54, 6594. [33] S. H. Park, Y. Lee, J. K. Lee, K. B. Kim, Electrochem. Solid-State Lett. 2006, 9, D31. [34] W. Lee, K. Nielsch, U. Gösele, Nanotechnology 2007, 18, 475713. [35] W. Lee, J. C. Kim, U. Gösele, Adv. Funct. Mater. 2010, 20, 21. [36] D. Wang, Y. Liu, B. Yu, F. Zhou, W. Liu, Chem. Mater. 2009, 21, 1198. [37] L. Liu, W. Lee, Z. Huang, R. Scholz, U. Gösele, Nanotechnology 2008, 19, 335604.
© 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Received: July 25, 2012 Published online: December 3, 2012
www.small-journal.com
1029
