Multilayer TiO 2 –Nanotube Formation by Two-Step Anodization J. M. Macak, * S. Albu, D. H. Kim, I. Paramasivam, S. Aldabergerova, and P. Schmuki ** ,z Department of Materials Science, University of Erlangen–Nuremberg, D-91058 Erlangen, Germany In this work we report on the growth of a closely stacked double layer of a self-organized TiO 2 nanotubes. For that we first anodize Ti in acidic electrolyte containing hydrofluoric acid to form thin nanotube layers. Afterwards we start a second anodization in a different electrolyte, a glycerol/NH 4 F mixture. This procedure allows us to grow the second layer directly underneath the first one. From scanning electron microscopy and transmission electron microscopy investigations we revealed that the second growth occurs via the tube bottoms of the first layer. These stacked multilayers generate new possibilities to vertically tailor the properties of the self-organized TiO 2 nanotube layers. © 2007 The Electrochemical Society. ͓DOI: 10.1149/1.2737544͔ All rights reserved. Manuscript submitted February 5, 2007; revised manuscript received April 8, 2007. Available electronically May 15, 2007. Formation of porous alumina based on anodic oxidation of alu- minium has been investigated and well understood already for many years. 1,2 But only about one decade ago, extremely ordered and self-organized porous alumina structures could be formed using a set of specific electrochemical conditions using optimized potential, temperature, electrolyte composition, etc. 3-7 In 1999, Zwilling et al. 8 showed that Ti can also be converted to highly ordered nanotubes ͑in contrast to alumina nanopores͒ using self-assembly during an- odic oxidation. Since then there have been many efforts to tailor the morphology of the TiO 2 nanotube toward enabling potential applications. 9-13 Later on, other valve metals have shown the ability to form nanotubes. 14-16 In all these works, fluoride-anion-containing electrolytes were used to selectively dissolve the anodized metals under anodic bias applied for several hours, leaving nanotubular layers on their surfaces. Although there is still very little work done on Ti and other metals compared to Al, reports showing significant improvements in length 9-13 and tube diameter 17,18 have been re- cently published by our group as well as by others. 19,20 Typically, the diameter of tubes is controlled by the applied anodization voltage 17 and various lengths can be obtained using different elec- trolytes. All these anodic TiO 2 nanotubes have already been used for a wide range of different applications, such as those based on the semiconductive nature of TiO 2 . 21 Namely, dye-sensitization, 22 doping, 23-25 photocatalysis, 26 electrochromism, 27 and photolysis 28 have been demonstrated, as well as others based on catalysis 29 or sensing. 30 Due to high biocompatibility of TiO 2 , other reports tar- geted growth of a hydroxyapatite layer on the nanotubes 31 as well as their formation on Ti alloys. 32,33 It has also been shown that the structure of the anodized tubes is always amorphous and can be converted by annealing to anatase 21,34,35 or, e.g., BaTiO 3 36 or Ba͑Sr͒TiO 3 37 upon hydrothermal alkali treatment. Additionally, by electrochemical deposition into the tubes, 38 properties such as the magnetic behavior of nanotube layers can be modified. 39 In the present work we show that even multistacks of TiO 2 nano- tubes can be grown directly by a two-step anodization process. Experimental Titanium foils ͑0.1 mm, 99.6% purity, Advent Materials͒ were degreased by sonication in acetone, isopropanol, and methanol prior to electrochemical experiments, afterward rinsed with deionized ͑DI͒ water, and finally dried in nitrogen stream. The samples were pressed together with a Cu plate against an O-ring in an electro- chemical cell ͑1cm 2 exposed to the electrolyte͒ and anodized at 20Vin1MH 2 SO 4 electrolytes containing hydrofluoric acid ͑HF͒ ͑0.16 M͒ for 2 h to grow a 500 nm thick TiO 2 nanotube layer. After this the nanotube layers were rinsed and dried and a second anod- ization step was performed in glycerol electrolytes containing NH 4 F ͑0.27 M͒ at 20 V for several hours. In some cases ͑to achieve longer tubes͒, substrates for the first layer were grown in ͑NH 4 ͒ 2 SO 4 /NH 4 F mixtures. For the electrochemical experiments, a high-voltage po- tentiostat Jaissle IMP 88 and a conventional three-electrode configu- ration with a platinum gauze as a counter electrode and a Haber– Luggin capillary with Ag/AgCl ͑1MKCl͒ as a reference electrode were used. All electrolytes were prepared from reagent-grade chemi- cals. Selected samples were sonicated in the ultrasonic bath ͑power output 100 W͒. A scanning electron microscope ͑Hitachi FE-SEM S4800͒, a transmission electron microscope ͑CM 30 T/STEM͒, and a high-resolution transmission electron microscopy ͑HRTEM͒ Phil- ips CM 300 UT were used for the morphological and structural characterization of the TiO 2 nanotubular layers. Results and Discussion Figures 1A-1C show examples of two types of arrays of self- organized TiO 2 nanotubes used in this study and a double-layer structure consisting of these two types. The first type of layer ͑layer 1, Fig. 1A͒ consists of 500 nm long tubes formed in a mixture of 1MH 2 SO 4 and 0.16 M HF at 20 V during 2 h with a diameter of approx. 100 nm. 40 The second type of layer ͑layer 2, Fig. 1B͒ con- sists of longer tubes formed in glycerol electrolyte containing 0.27 M NH 4 F at 20 V. To form the multilayer structure of Fig. 1C, the growth of the two different types of tubes is combined in a two-step anodization process. For this, a titanium sheet is anodized to make the short tubes as described above. Then, after rinsing and drying, the sample is anodized in the glycerol electrolyte at 20 V. From Fig. 1C it is clear that the dimensions of the second layer are the same as for nanotubes formed directly in this electrolyte. 12 Fig- ure 1D shows polarization curves ͑recorded with a sweep rate of 0.1 V/s͒ and current transients recorded for the first and second layer growth. For the first anodization in HF electrolytes leading to short tubes ͑Fig. 1A͒, the current reaches a maximum during the potential sweeping; this situation is typical for a self-organization sequence involving the formation of a compact oxide and irregular pores prior to ordered pore formation. 10,40 For the second anodiza- tion in glycerol solutions leading to long and smooth tubes, the same sequence occurs, but the tube self-organization kinetics is much slower and therefore the drop in the current appears in the anodiza- tion process later, in this case in the potentiostatic phase of the anodization process. 41 At the end of the anodization experiments, stirring of the electrolytes was applied for 15 min. The resulting current behavior is shown as an inset in Fig. 1D. Just after the beginning of the stirring ͑angular speed ϳ150 rpm͒, the currents in both cases started to increase, and as soon as the stirring was termi- nated, they returned to their original values. This can be ascribed to the fact that the steady-state current density is diffusion limited. The * Electrochemical Society Student Member. ** Electrochemical Society Active Member. z E-mail: schmuki@ww.uni-erlangen.de Electrochemical and Solid-State Letters, 10 ͑7͒ K28-K31 ͑2007͒ 1099-0062/2007/10͑7͒/K28/4/$20.00 © The Electrochemical Society K28 Downloaded 21 Jul 2008 to 159.226.100.134. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp reason for the smaller increase in the glycerol electrolyte can be ascribed to a significantly longer diffusion path ͑longer tubes͒ and to the high viscosity of the electrolyte. In order to evaluate the growth of the second layer after the first layer has been formed, we performed detailed scanning electron microscopy ͑SEM͒ investigations of the interfaces between the lay- ers. Figure 2 shows a scheme of the formation of second-layer growth, which is based on the SEM observation shown in Fig. 3. We first grow an initial nanotube layer in the acidic electrolyte ͑Fig. 2a͒. Afterwards we clean and dry it, immerse it in the glycerol electro- lyte, and start the second anodization process. After several minutes, there are very small channels or holes etched in the bottoms of the first tube layer ͑Fig. 2b͒. The etching takes place preferentially at the bottom of the tubes. From the images in Fig. 3A and 3B that were taken from sample anodized for about 20 min in the glycerol elec- trolyte ͑to form the second layer͒, it is evident that the etching front penetrates the bottoms and new and somewhat irregularly distrib- uted tubes are formed ͑Fig. 2c͒. The new tubes are at this very moment competing for available space and current. Growth of some tubes is terminated after a while, because there is not sufficient space available. After about 40 min, the newly formed tubes are already self-organized ͑Fig. 2d͒, with only small variations in diam- eter ͑40 ± 10 nm͒ as shown also in Fig. 3C. From the images in Fig. 3A and 3B it can be seen that the width of the channels drawn in Fig. 2c is in the range between 15 and 25 nm. The number of chan- nels is typically between 3 and 5 and the number of the newly formed tubes from the first layer ͑one bigger tube͒ is3or4,as shown also in Fig. 2d. From Fig. 1A it can be seen that the original outer diameter of the tubes ͑layer 1͒ was about 130 nm and the newly formed tubes ͑layer 2͒ in Fig. 1B have outer diameters of Figure 1. SEM top-view images of the tubes used for the growth of ͑A͒ the first and ͑B͒ the second layer and ͑C͒ their interface after two-step anodization. The first tubes are 500 nm long with a diameter of 100 ± 10 nm, and the second tubes have a diameter of 40 ± 10 nm and length dependent on the anodization time; ͑D͒ polarization curves ͑left part, sweep rate 0.1 V/s͒ and current transients ͑right part͒ recorded for anodization of Ti sample at 20 V during the first layer growth ͑in acidic electrolyte͒ and during the second layer growth ͑in glycerol electrolyte͒. ͑Insets͒ Magnification of the current transients during electrolyte strirring introduced for 15 min ͑angular speed ϳ150 rpm͒. Figure 2. Schematic drawing demonstrating four steps in the formation of the second tubular layer during the second anodization step. K29Electrochemical and Solid-State Letters, 10 ͑7͒ K28-K31 ͑2007͒ K29 Downloaded 21 Jul 2008 to 159.226.100.134. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp 50 ± 10 nm. Using a longer tube geometry ͑more than 500 nm͒ as the upper nanotube layer is possible, but wetting of such tubes after drying becomes a more severe issue. Remarkably, there is no sig- nificant chemical dissolution apparent of either the first layer or the second layer. One can expect that the electrolyte also filled the space between the nanotubes; however, it seems that the gaps of the first tube layer not the most reactive sites under present conditions, as we did not observe any nanotube growth there. To obtain more insight into the formation of the second layer, we performed some additional characterization by TEM. Figure 4 shows TEM images of the second-layer nanotubes ͑A͒ and a single tube ͑B͒ of the first layer, with several channels at the bottom de- scribed and shown in Fig. 2 and 3. From Fig. 4A one can clearly see that the second-layer tubes are smooth and without ripples on the walls. Furthermore, the presence of open channels for mass and current flow from the first to the second layer is confirmed by Fig. 4B. Figure 4C shows a high-resolution TEM image of the tube bot- tom of the second layer. Clearly, an amorphous structure is present as confirmed by selected area diffraction pattern ͑SAED͒, shown in the inset. Figure 5 shows the influence of anodization time on the thickness of the second layer. After 10 min of anodization there are only short Figure 3. SEM images of the interface between both TiO 2 nanotubular lay- ers ͑A͒ in the top- and ͑B͒ in the cross-sectional view showing channels at the bottoms of the first layer; ͑C͒ comparison between the tube outer diam- eters during the transition from one ͑outer diameter approximately 130 nm͒ to the other type ͑approx. 70 nm͒. Space occupied originally by one tube is typically used for three tubes of the second layer. Figure 4. ͑A͒ TEM images of bundles of smooth TiO 2 nanotubes from the second layer; ͑B͒ single tube depicted from the top showing etched bottom with channels; ͑C͒ HRTEM image of the tube bottom of the second type showing amorphous structure. Inset in c shows SAED with diffuse rings matching amorphous structure. Figure 5. Dependence of the second-layer thickness on the anodization time during the second anodization step shown ͑a͒ as a sequence of SEM cross- sectional images and ͑b͒ as a plot. Linear growth with constant of 300 nm/h is achieved within first 12 h. K30 Electrochemical and Solid-State Letters, 10 ͑7͒ K28-K31 ͑2007͒K30 Downloaded 21 Jul 2008 to 159.226.100.134. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp tubes; no visible alteration of the tube bottoms has occurred. After approximately 30 min of anodization there is already a second tube layer formed with a thickness of about 100 nm. The tubes keep growing and the tube layer thickness increases with time as shown in Fig. 5b. In the early stages ͑first 12 h͒, the thickness increases linearly, and for longer times a deviation to an apparently slower growth rate is observed. After 24 h of anodization the second tube layer thickness is reaching 5 ␮m. From the cross-sectional SEM images we can estimate the growth rate in the early stages of ap- proximately 50 nm per 10 min ͑300 nm per h͒. This means that the growth of nanotubes with the same length is somewhat slower com- pared to tubes that were grown directly, 12,41 i.e., without the pres- ence of the upper tube layer. For example, in our previous work we were able to grow 7 ␮m tubes within 13 h under the same electro- chemical conditions as used here. 12 This can be described to ham- pered diffusion of the chemical species through the narrow channels between the first and second layer. Multilayer structures, as formed here, may have significant applications, for example, in size- selective reactive filtration or Bragg-stack structures. 42,43 Conclusions The results of the present work show that multilayer stacks of highly ordered and self-organized TiO 2 nanotubular layers can be grown by two-step anodization under different electrochemical con- ditions. From SEM evaluation we revealed that the growth of tubes of the second layer starts at the bottom of the first tubes by narrow channels being formed in the early stages. Further, we show that the thickness of these second layers can be in the range of several mi- crometers, depending on anodization time. Clearly, the diameter of the tubes corresponds to the formation conditions of the individual nanotube layers. The growth of the second layer in length is some- what slower than for its isolated formation. The feasibility to form two distinct layers may be exploited in view of optical properties or for size-selective reactions. Acknowledgments The authors acknowledge DFG for financial support. Hans Rollig and Martin Kolacyak are acknowledged for valuable technical help. University of Erlangen assisted in meeting the publication costs of this article. References 1. F. Keller, M. S. Huntler, and D. L. Robinson, J. Electrochem. Soc., 100,411 ͑1953͒. 2. J. W. Diggle, T. C. Downie, and C. W. Goulding, Chem. Rev. (Washington, D.C.), 69, 365 ͑1969͒. 3. H. Masuda and K. Fukuda, Science, 268, 1466 ͑1995͒. 4. G. E. Thompson, Thin Solid Films, 297, 192 ͑1997͒. 5. O. Jessensky, F. Müller, and U. Gösele, J. Electrochem. Soc., 145, 3735 ͑1998͒. 6. S. Ono, M. Saito, and H. Asoh, Electrochim. Acta, 51, 827 ͑2005͒. 7. W. Lee, R. Ji, U. Gösele, and K. Nielsch, Nat. Mater., 5, 741 ͑2006͒. 8. V. Zwilling, E. Darque-Ceretti, and A. Boutry-Forveille, Electrochim. Acta, 45, 921 ͑1999͒. 9. J. M. Macak, K. Sirotna, and P. Schmuki, Electrochim. Acta, 50, 3679 ͑2005͒. 10. L. Taveira, J. M. Macak, H. Tsuchiya, L. F. P. Dick, and P. Schmuki, J. Electro- chem. Soc., 152, B405 ͑2005͒. 11. J. M. Macak, H. Tsuchiya, and P. Schmuki, Angew. Chem., Int. Ed., 44, 2100 ͑2005͒. 12. J. M. Macak, H. Tsuchiya, L. Taveira, S. Aldabergerova, and P. Schmuki, Angew. Chem., Int. Ed., 44, 7463 ͑2005͒. 13. S. Albu, A. Ghicov, J. M. Macak, and P. Schmuki, Phys. Status Solidi (RRL), 1, R65 ͑2007͒. 14. H. Tsuchiya, J. M. Macak, I. Sieber, and P. Schmuki, Small, 1, 722 ͑2005͒. 15. H. Tsuchiya and P. Schmuki, Electrochem. Commun., 7,49͑2005͒. 16. I. Sieber, B. Kannan, and P. Schmuki, Electrochem. Solid-State Lett., 8, J10 ͑2005͒. 17. S. Bauer, S. Kleber, and P. Schmuki, Electrochem. Commun., 8, 1321 ͑2006͒. 18. H. Tsuchiya, J. M. Macak, L. Taveira, E. Balaur, A. Ghicov, K. Sirotna, and P. Schmuki, Electrochem. Commun., 7, 576 ͑2005͒. 19. Q. Cai, M. Paulose, O. K. Varghese, and C. A. Grimes, J. Mater. Res., 20, 230 ͑2005͒. 20. K. S. Raja, M. Misra, and K. Paramguru, Electrochim. Acta, 51, 154 ͑2005͒. 21. R. Beranek, H. Tsuchiya, T. Sugishima, J. M. Macak, L. Taveira, S. Fujimoto, H. Kisch, and P. Schmuki, Appl. Phys. Lett., 87, 243114 ͑2005͒. 22. J. M. Macak, H. Tsuchiya, A. Ghicov, and P. Schmuki, Electrochem. Commun., 7, 1138 ͑ 2005͒. 23. A. Ghicov, J. M. Macak, H. Tsuchiya, J. Kunze, V. Heaublein, L. Frey, and P. Schmuki, Nano Lett., 6, 1080 ͑2006͒. 24. J. M. Macak, A. Ghicov, R. Hahn, H. Tsuchiya, and P. Schmuki, J. Mater. Res., 21, 2824 ͑2006͒. 25. J. H. Park, S. Kim, and A. J. Bard, Nano Lett., 6,24͑2006͒. 26. J. M. Macak, M. Zlamal, J. Krysa, and P. Schmuki, Small, 3, 303 ͑2007͒. 27. A. Ghicov, H. Tsuchiya, R. Hahn, J. M. Macak, and A. G. Munoz, Electrochem. Commun., 8, 528 ͑2006͒. 28. O. K. Varghese, M. Paulose, K. Shankar, G. K. Mor, and C. A. Grimes, J. Nanosci. Nanotechnol., 5, 1158 ͑2005͒. 29. J. M. Macak et al., Electrochem. Commun., 7, 1417 ͑2005͒. 30. O. K. Varghese, D. Gong, K. G. Ong, and C. A. Grimes, Sens. Actuators B, 93, 338 ͑2003͒. 31. H. Tsuchiya, J. M. Macak, L. Muller, J. Kunze, F. Muller, S. P. Greil, S. Virtanen, and P. Schmuki, J. Biomed. Mater. Res., 77A, 534 ͑2006͒. 32. H. Tsuchiya, J. M. Macak, A. Ghicov, and P. Schmuki, Small, 2, 888 ͑2006͒. 33. J. M. Macak, H. Tsuchiya, L. Taveira, A. Ghicov, and P. Schmuki, J. Biomed. Mater. Res. , 75A, 928 ͑2005͒. 34. O. K. Varghese, D. Gong, M. Paulose, C. A. Grimes, and E. Dickey, J. Mater. Res., 18, 156 ͑2003͒. 35. J. M. Macak, S. Aldabergerova, A. Ghicov, and P. Schmuki, Phys. Status Solidi A, 203, R67 ͑2006͒. 36. N. T. Padture and X. Wei, J. Am. Ceram. Soc., 86, 2215 ͑2003͒. 37. X. Wei, A. L. Vasiliev, and N. P. Padture, J. Mater. Res., 20, 2140 ͑2005͒. 38. J. M. Macak, B. G. Gong, M. Hueppe, and P. Schmuki, Adv. Mater. (Weinheim, Ger.), Accepted. 39. V. M. Prida, M. Hernández-Veléz, K. R. Pirota, A. Menéndez, and M. Vasquéz, Nanotechnology, 16, 2696 ͑2005͒. 40. R. Beranek, H. Hildebrand, and P. Schmuki, Electrochem. Solid-State Lett., 6, B12 ͑2003͒. 41. J. M. Macak and P. Schmuki, Electrochim. Acta, 52, 1258 ͑2006͒. 42. H. Tsuchiya, M. Hueppe, T. Djenizian, and P. Schmuki, Surf. Sci., 547, 268 ͑2003͒. 43. G. Vincent, Appl. Phys. Lett., 64, 2367 ͑1994͒. K31Electrochemical and Solid-State Letters, 10 ͑7͒ K28-K31 ͑2007͒ K31 Downloaded 21 Jul 2008 to 159.226.100.134. Redistribution subject to ECS license or copyright; see http://www.ecsdl.org/terms_use.jsp . Multilayer TiO 2 –Nanotube Formation by Two-Step Anodization J. M. Macak, * S. Albu, D. H. Kim, I. Paramasivam, S been re- cently published by our group as well as by others. 19,20 Typically, the diameter of tubes is controlled by the applied anodization voltage 17 and