Anodization of nanoimprinted titanium: A comparison with formation of porous alumina

Article abstract of DOI:10.1016/j.electacta.2004.02.015

Electrochemically prepared ordered porous alumina has become one of the most important nanotemplate materials to date. Naturally the questions arises whether other valve metals such as Ti, Ta, Nb, V, Hf or W can also be used for fabricating ordered pore arrays. We investigate in detail the electrochemical anodization of titanium in different electrolytes and its potential and temperature dependence. It turns out that due to the semiconducting properties of titania, a mirror image of the behavior of the electrically insulating porous alumina seems to be impossible. So-called porous titania in literature corresponds to the pitting regime of aluminum where pores are created due to dielectric breakdown of titania or alumina, respectively. Below the breakdown potential of titania, only thick barrier layers can be obtained. However, by nanoimprint of titanium and successive anodization below the breakdown potential, monodomain porous titanium oxide with a pore depth of 60nm on a cm²-scale can be prepared. We discuss in detail the growth mechanism of porous structures of titanium and compare it with that of porous alumina.

Full text of DOI:10.1016/j.electacta.2004.02.015

Electrochimica Acta 49 (2004) 2645–2652
Anodization of nanoimprinted titanium: a comparison with
formation of porous alumina
a,∗,1
a,b
c,2
a
Jinsub Choi
, Ralf B. Wehrspohn , Jaeyoung Lee , Ulrich Gösele
a
Max Planck Institute of Microstructure Physics, Weinberg 2, 06120 Halle, Germany
b
Nanophotonic Materials Group, Department of Physics, University of Paderborn, Warburger Strasse 100, 33098 Paderborn, Germany
c
Water Protection Research Team, Research Institute of Industrial Science and Technology, Pohang City 790-330, South Korea
Received 9 December 2003; received in revised form 4 February 2004; accepted 14 February 2004
Abstract
Electrochemically prepared ordered porous alumina has become one of the most important nanotemplate materials to date. Naturally the
questions arises whether other valve metals such as Ti, Ta, Nb, V, Hf or W can also be used for fabricating ordered pore arrays. We investigate
in detail the electrochemical anodization of titanium in different electrolytes and its potential and temperature dependence. It turns out that due
to the semiconducting properties of titania, a mirror image of the behavior of the electrically insulating porous alumina seems to be impossible.
So-calledporoustitaniainliteraturecorrespondstothepittingregimeofaluminumwhereporesarecreatedduetodielectricbreakdownoftitania
or alumina, respectively. Below the breakdown potential of titania, only thick barrier layers can be obtained. However, by nanoimprint of tita-
2
nium and successive anodization below the breakdown potential, monodomain porous titanium oxide with a pore depth of 60 nm on a cm -scale
can be prepared. We discuss in detail the growth mechanism of porous structures of titanium and compare it with that of porous alumina.
©
2004 Elsevier Ltd. All rights reserved.
Keywords: Anodization; Nanoimprinted titanium; Alumina
1
. Introduction
ity. However, an interface layer, e.g., a hydroxyapatite (HA)
coating or porous layer, on the surface of the Ti substrate
is necessary to osseointegrate with bones [5] because native
TiO2 coatings have no ability to form a strong bond with
bony tissue due to their low surface energy.
Al, Ti, Ta, Nb, V, Hf, W are classified as “valve metals”
because their surface is immediately covered with a native
oxide film of a few nanometers when these metals are ex-
posed to oxygen containing surroundings [1,2]. Inherently,
these oxides retard the rate of reaction on the metal surface.
Therefore, they have been widely used as highly protec-
tive layers resisting corrosion. Recently, valve metals again
gained considerable interest, because some oxides of valve
metals have unique and excellent properties in optics, elec-
tronics, photochemistry and biology. For example, titanium
oxide is a promising material for photoelectrocatalysis [3]
and implants [4]. Titanium implants have been intensively
investigated due to its superior strength and biocompatibil-
Current state-of-the-art techniques for aluminum an-
odization using self-assembly of pores allow the fabrication
of monodomain and monodisperse porous alumina struc-
tures with adjustable aspect ratios [6–8]. However, porous
templates based on other valve metals have been rarely
studied since the 1970s. There were a few attempts to com-
pare anodic oxide growth on other value metals with porous
alumina [9–11]. Zwilling et al. [9] reported that a porous
titanium oxide layer is created in a chromic acid solution in
the presence of a small amount of HF, whereas only a barrier
titanium oxide layer is formed in pure chromic acid con-
trary to porous alumina. According to their results, the TiO2
barrier oxide thickness is time-independent whereas the
thickness of the porous oxide of titania increases with an-
odizing time. However, they did not mention the maximum
obtainable thickness of the anodic porous titanium oxide.
Gong et al. claimed that the final thickness of porous TiO2
structures formed in dilute HF solution is independent of
∗
Corresponding author. Tel.: +1-626-395-2207;
fax: +1-626-577-8442.
E-mail address: jinsub@caltech.edu (J. Choi).
1
Present address: Nanofabrication Group, Division of Engineering and
Applied Science, California Institute of Technology, Pasadena, CA 91125,
USA.
Present address: Fuel Cell Research Center, Korea Institute of Science
and Technology (KIST), Seoul 136-791, South Korea.
2
0
013-4686/$ – see front matter © 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.electacta.2004.02.015

2
646
J. Choi et al. / Electrochimica Acta 49 (2004) 2645–2652
the anodizing time [10]. They discussed that the dissolution
rate of the formed TiO2 oxide is very high in HF even in the
absence of an anodizing process. More recently, Beranek
et al. have shown that highly ordered porous titania can be
produced in H2SO4 electrolytes containing HF. However,
the thickness of pores is still limited up to 500 nm [11].
The understanding of the parameters ruling self-ordering
as well as nanoimprint lithography to fabricate perfectly
ordered porous alumina templates are well understood to
date [12,13]. Application of this understanding to other
valve metals would be very helpful not only to fabricate
self-ordered porous oxides grown on other valve metals,
but potentially also to improve the understanding of the
growth of porous alumina. In this work, porous anodic ox-
ide layer grown on titanium by electrochemical methods
will be studied and compared to the mechanism governing
the formation of porous alumina.
Fig. 2. Linear sweep voltammogram (LSV) of aluminum and titanium.
LSV was conducted with the sweep rate of 0.1 V/s in the range from 0
◦
to 210 V: (a) LSV of aluminum (A) 0.26 M K2HPO4 at 60 C and (B)
◦
◦
0
.1 M H3PO4 at 5 C. (b) LSV of titanium at 60 C in 4 M H3PO4. The
LSV can be divided to four regions: amorphous oxide formation (Regime
I), transformation into anatase or formation of anatase in addition to
the existing amorphous oxide (II–III), and transformation into rutile (IV)
2
. Experimental
(Refs. [16–19]).
2
.1. Preparation of polished substrates
◦
in a non-solvent cleaning fluid (Logitech Ltd., UK) at 80 C.
Then, it was cut into squares (2 cm × 2 cm) with a mechan-
ical cutter (Profiform AG, Feather Products Modellbahnen
GmbH, Switzerland). Electropolishing [14] was skipped be-
cause it would yield an inhomogeneous oxide layer on the
titanium substrate.
High purity titanium foil (99.99%) with a thickness of
.5 mm was purchased from Alfa Aesar Inc. After mounting
0
the foil on a 4-in. glass substrate, it was mechanically lapped
and polished using an alumina suspension with 0.04 ␮m
grains (OPS suspension, Struers) for 5 h. Afterwards, the me-
chanically polished Ti was released from the glass substrate
2.2. Imprint
The fabrication procedure of a master stamp was de-
scribed in detail elsewhere [7]. The master stamp consists
of Si3N4 pyramids, which are wafer-bonded to a 4-in. sil-
icon substrate. The pyramids have a height of 260 nm and
are arranged in a hexagonal lattice with a nearest neighbor
distance of 500 nm. Before indentation, dust on the surface
was removed by a strong stream of N2 gas. Subsequently,
Fig. 1. (a) Schematic diagram of the indentation of Ti under a pressure
2
of 25 kN/cm . I: imprint master stamp consisting of a hexagonal array
of pyramids of Si3N4 with a height of 260 nm and a lattice constant
of 500 nm. S: mechanically polished Ti substrate. (b) Scanning electron
microscope (SEM) image of the nanoindented surface of the Ti substrate.
Fig. 3. Current density transients as a function of anodizing time in 4 M
H3PO4 at 40 V (a) and at 210 V (b), which correspond to Regime III and
Regime IV, respectively.

J. Choi et al. / Electrochimica Acta 49 (2004) 2645–2652
2647
a piece of the stamp (1 cm × 1 cm) was placed on the me-
chanical polished Ti. The pattern on a master stamp was
transferred onto the surface of the titanium foil by applying
a pressure of 25 kN/cm using an oil press (PW, Paul-Otto
Weber GmbH, Germany) as shown in Fig. 1. The average
depth of the pits is around 50 nm.
stat (Keithley 2400) interfaced to a computer. The cell is a
two-electrode system consisting of a Pt mesh acting as the
counter electrode and Ti as the working electrode. During
the anodization, the temperature and velocity of stirring of
the electrolyte were kept constant. The morphologies of
the porous titanium oxide were characterized by a scanning
electron microscope (SEM, JEOL JSM-6300F). The thick-
ness and composition of titanium oxide formed at different
potentials were measured by Auger electron spectroscopy
(Physical Electronics, PHI 600). The depths of the pores
formed in ethanolic HF were analyzed by atomic force
microscopy (AFM, Digital Instruments 5000 microscope)
2
2
.3. Anodization and analysis
The Ti substrate was anodized under a constant cell
voltage in aqueous phosphoric acid (1–4 M H3PO4) or
ethanolic hydrofluoric acid (0.5 M HF), using a potentio-
◦
◦
◦
Fig. 4. Scanning electron microscope (SEM) images of titanium anodized in 1 M phosphoric acid at 210 V for 120 min: (a) 5 C, (b) 20 C, (c) 45 C,
(
◦
d) 60 C, (e) and (f) are enlarged view of (a) and (d), respectively. The scales in (a–d) are identical. The scale of (e) is the same with that of (f).

2
648
J. Choi et al. / Electrochimica Acta 49 (2004) 2645–2652
Table 1
Regimes of the anodic treatment of aluminium (Ref. [28])
Compact alumina structures (so
called barrier alumina oxides)
Thick porous alumina structures
with barrier layers
Aluminum etch tunnel or aluminum
dissolution
Schematic pictures
−
Type of electrolytes Very weak acidic or neutral solutions Medium or weakly acidic solutions
Strong acidic solutions containing Cl
Pore density
Thickness
No pores
Determined by applied potential
Constant with increasing anodizing time Increase with increasing anodizing time
Determined by anodization time Determined by diffusion limits of electrolytes
Breakdown potential
using standard Si3N4 tips. The scanning speed of AFM
was 0.5 Hz.
To investigate the effect of the cell voltage, linear sweep
voltammetry (LSV) with the sweep rate of 0.1 V/s was per-
formed in the range of 0 V to 210 V in 4 M phosphoric acid
◦
at a constant temperature of 60 C. To avoid the influence
3
. Results and discussion
of a reaction at the counter electrode on the LSV, a Pt wire
acting as a quasi-reference electrode was inserted to the
two-electrode cell, placed adjacent to the working electrode.
The sweep ramp has directly been superimposed on the cell
voltage.
3
.1. Pore formation above the breakdown potential of the
anodic oxide
3
.1.1. Influence of the potential
LSV of aluminum in 0.1 M H PO and 0.26 M K HPO
4
3
4
2
Titanium anodization is influenced by the growth mode,
shows the typical behavior of anodic oxides (Fig. 2a). Ini-
tially, the current is very low and then suddenly increases
close to the breakdown potential of the anodic oxide. In the
case of titanium (Fig. 2b) four different regimes can be ob-
served, indicating that there are transitions in the structure
for example, constant potential, constant current density or
a combination of these methods [15]. Here, we only focus
on the constant potential method to compare the fabrication
process with that of porous alumina.
◦
Fig. 5. SEM top views of titanium anodized for different times. Anodization (210 V, 60 C, 1 M phosphoric acid) was performed for: (a) 45 min (b)
1
20 min and (c) 1000 min (d) 3500 min.

J. Choi et al. / Electrochimica Acta 49 (2004) 2645–2652
2649
or new crystalline phases are formed in addition to the
already existing amorphous layer during the anodization.
In addition, it is observed that the current density below
the breakdown potential is much higher than in the case
of aluminum. This demonstrates the higher conductivity of
titanium oxide with respect to alumina. It has been reported
that in the case of the anodization of titanium, the oxide
exhibits a phase transition with increasing potential from an
amorphous phase into a crystalline phase such as anatase
or rutile [16–18] (brookite structure was rarely mentioned).
Moreover, it has been reported that the electronic resistiv-
ity of the rutile phase is three order of magnitude higher
◦
than that of anatase (at 100 C) [19]. We assume that the
Fig. 6. Statistic analysis of the density of pores shown in Fig. 5. Note
that the pore density in porous alumina formed at 195 V (ideal case) is
4.61 × 10
ionic resistivity of the rutile phase is higher than that of
anantase since rutile TiO2 (4.25 g/cm ) is a denser and
1
2
−2
m .
3
thermodynamically more stable phase than anatase TiO2
3
(3.9 g/cm ) [19]. The decrease in the current density in
resulting in a porous alumina structure with straight pore
channels and a narrow pore size distribution of the pore size.
Anodization of titanium was performed at 210 V in 1 M
phosphoric acid at various temperatures. In Fig. 4, we ob-
serve that porous titania structures formed at relative low
temperatures have veins of interconnected pores (or craters)
and cracks caused by mechanical stress (Fig. 4a and b). In
addition, thick rims around the craters with a diameter on
the ␮m-scale are observed in the enlarged view of the SEM
image (Fig. 4e). In contrast, pore like morphologies are cre-
ated over the entire area at higher temperature (Fig. 4c,d
and f), even though the distribution of pore sizes is relative
the regime IV of Fig. 2b can be attributed to the transi-
tion of the anatase (or anatase + amorphous) phase into
the rutile phase. The exact potentials of the phase change
depend on the anodization conditions such as electrolyte
and temperature [20,21]. There are few research works on
the anodization of titanium in phosphoric acid [22]. How-
ever, several results obtained in other electrolyte systems
might be taken into account. For example, Arsov et al.
reported that during the anodization of titanium in sulfu-
ric acid, amorphous oxides are formed below 20 V, mixed
oxide between 20 and 50 V and fully crystalline anatase
at voltages higher than 50 V. In the graph of Fig. 2b, a
sparking of the current is observed at potentials higher
than 80 V, indicating that the breakdown potential might
be around 80 V. Interestingly, porous titania structures are
only obtained at potentials where sparking is observable.
This is in agreement with the observation that pore-like
morphologies in titanium are only formed above 100 V
(sparking potential region) in sulfuric acid [23]. The break-
down potential observed during titanium of the anodization
has been proposed to be related to the crystallization which
leads to compressive stress in the oxide or oxygen evolution
[17,24,25].
Fig. 3a and b show transients of the current density as a
function of anodizing time at constant cell potentials of 40
and 210 V, respectively. For 210 V anodization, sparking is
observed from the beginning of anodization whereas there
is no sparking in the case of 40 V.
3
.1.2. Influence of the temperature
In the case of the anodization of aluminum, the tempera-
ture is typically kept below room temperature during the an-
odization to prevent the formed oxide from being dissolved
in an acidic solution. For example, if a porous alumina struc-
◦
ture is formed at 65 C in sulfuric acid, the porous alumina
thickness does not increase because the rate of dissolution
equals that of the growth of porous alumina [26,27]. Fur-
thermore, a constant temperature at the pore tips maintains a
homogenous field-enhanced dissolution over the entire area,
Fig. 7. Auger electron spectroscopy (AES) depth profiles for oxide lay-
ers grown on titanium under different anodization conditions: (a) 210 V,
120 min, (b) 210 V, 1000 min and (c) 80 V, 240 min.

2
650
J. Choi et al. / Electrochimica Acta 49 (2004) 2645–2652
Fig. 8. Schematic diagram of porous titanium oxide formation above the breakdown potential: (A) oxide growth to maximal thickness, (B) burst of oxide
by the formation of crystallites (pore formation), (C) immediate repassivation of pore tips, (D) burst of repassivated oxide, and (E) dissolution of the
formed oxide and second repassivation.
◦
broad. If the temperature is too high (T > 70 C), the rate of
oxide exists. However, in the case of aluminum dissolution
above the breakdown potential, tunnel growth competes with
the formation of new pits since unetched areas between tun-
nel pits are not insulating (category 3 of Table 1). Therefore,
the density of pits on the surface increases with Al etching
time [30].
In the case of titanium anodization, the density of pores
formed above the breakdown potential increases with time,
which is similar to the formation of etch tunnels in Al
(Fig. 5). In Fig. 6, the density of pores in titania is ana-
lyzed as a function of the quantity of charge passed during
anodization. The logarithm of pore density increases lin-
early as a function of the logarithm of the charge. In addi-
tion, it is observed that new pores inside existing pores are
created.
The thickness of the grown oxide on titanium has a max-
imum thickness which is mainly determined by the applied
potential (2.5–3 nm/V) [2,20,31] but not by the anodizing
time. In the case of alumina growth, the thickness of porous
alumina increases with anodizing time but that of the barrier
oxide is solely determined by the applied potential [27,28].
Fig. 7 shows Auger depth profiles, demonstrating that the
thickness of the anodic oxide on titanium depends on the ap-
plied potential rather than on the anodizing time. Assuming
that all the samples formed at 210 V for 120 min or 1000 min
and at 80 V for 240 min consist of the rutile phase, we ex-
pect the rate of sputtering during the Auger depth profile to
be the same. We observe that there is hardly any difference
in the oxide thicknesses with increasing anodization time
for a fixed potential (Fig. 7a and b). Fig. 7c shows that a
much thinner oxide will be produced when the anodization
is carried out at a lower potential even though the anodiza-
tion time is twice as long as for the sample in Fig. 7a.
the dissolution would be as fast as that of the oxide growth.
3
.1.3. Comparison with alumina anodization
Since titanium oxide is a n-type semiconductor (with a
high donor density in the range of 2 to 6 × 10 cm
20
−3
,
bandgap: 3.2–3.8 eV) [15], the electronic current is expected
to be significantly larger than in the case of aluminum an-
odization (bandgap: 7–9.5 eV). Note that for insulating metal
oxides, since the electronic conductivity is very low, the
ionic current density predominantly transports the charges.
Secondly, different phases are observed in anodic titanium
oxides formed in phosphoric acid, depending on the applied
potential or current density, whereas in the case of anodic
aluminum the amorphous phase can mostly be obtained in
phosphoric acid.
It is well known that electrochemical anodic treatments
of aluminum can be classified into three categories with
respect to the formation of pits, pores, or a barrier layer
(
Table 1) [28]: (1) compact alumina structures (so called
barrier oxides) in very weak acidic or neutral solutions such
as H3BO3 or ammonium tartrate; (2) thick porous alumina
structures with barrier layers at the interface Al/electrolyte
in medium or weakly acidic solutions (slightly soluble in
these electrolytes) such as H2SO4, (COOH)2, or H3PO4; (3)
aluminum etch tunnels (pits) or aluminum dissolution by
wet-etching in strong acidic or basic solutions such as HCl
or NaOH [29].
Note that porous titanium oxide structures are created
above the breakdown potential, whereas porous aluminum
oxide is ordered below the breakdown potential. Aluminum
dissolution occurs above the breakdown potential. The
breakdown potential of Al2O3 strongly depends on the
electrolyte and temperature, and is, for example, more than
5
00 V in non-dissolving electrolytes like H3BO3 but only a
3.1.4. Growth models
few mV in HCl [27,28].
The electrochemistry of growth of porous titanium oxide
at high potentials is quite different to that of porous alu-
minum oxide because porous titanium oxide is formed above
the breakdown potential. The development of a porous tita-
nia layer on the surface of titanium might be described as
follows. Initially, the thickness of a non-porous, barrier ti-
tania grows larger with increasing anodizing voltage. Then,
its structure undergoes transitions (Fig. 8A). When it is
These different growth regimes also influence the pore
density. In the case of porous alumina growth occurring be-
low the breakdown potential, pores grow continuously with
time (category 2 of Table 1). The formation of new pores on
the surface can be neglected. The formed oxide layer acts as
an insulator to the flow of an ionic current. Thus, the reac-
tions only occur at the pore bottom where a relatively thin

J. Choi et al. / Electrochimica Acta 49 (2004) 2645–2652
2651
transformed to a dense crystalline structure such as anatase
or rutile, the compressive stress in the oxide increases sig-
nificantly. Breakdown of the barrier oxide film occurs and
new pores in between crystallites are formed (Fig. 8B). This
corresponds to the occurrence of the sparking. However, the
areas of electrical breakdown of the oxide are immediately
covered with titanium oxide again (repassivation, Fig. 8C)
due to the characteristics of valve metals. When the break-
down occurs again inside the repassivated pores (Fig. 8D
and E), it looks like there is the formation of small pores
inside the pores. Since current can flow through the whole
oxide layer, the thickness of oxide can not linearly increase
with anodizing time.
3
.2. Pore formation below the breakdown potential of the
anodic oxide
Fig. 10. Atomic force microscopy (AFM) topography of a titanium oxide
array with 500 nm periodic pores. The average depth of pores is about
Since pores are randomly formed above the breakdown
potential as discussed in the previous section, the anodization
of lithographically structured titanium can not lead to titania
templates with an arrangement of pores. In this section, the
anodization of nanoindented titanium below the breakdown
60 nm.
potential is studied to fabricate porous titania templates with
monodomain pore arrangement.
In Fig. 9a, a titanium oxide array with 500 nm interpore
distance is shown which has been anodized in ethanolic
0.5 M HF at room temperature for 240 min at 10 V. Ethanol
based electrolytes were used since anodization in aqueous
HF electrolyte below the breakdown potential yields a coarse
oxide layer on the surface of titanium (Fig. 9c). Under these
conditions, the prepatterns on the Ti substrate disappear
and porous structures like those formed above the break-
down potential are not created. This originates from the
very fast dissolution rate of titanium oxide in aqueous HF.
In the case of anodization in ethanolic HF, an oxide layer
is formed which maintains the structure of the prepattern.
AFM analysis reveals an average depth of the pores of about
6
0 nm after 240 min anodization (Fig. 10). The oxide for-
mation on the surface can be confirmed by AES analysis
Fig. 11). Even though the deviation of the initial size of
(
prepatterned holes is large at the beginning of anodization,
the sizes become identical during the anodization. However,
Fig. 9. SEM images of anodization of the titanium at 10 V having a lattice
constant of 500 nm: (a) in ethanolic 0.5 M HF for 240 min, (b) enlarged
view of (a) and (c) in aqueous 0.5 M HF for 20 min.
Fig. 11. Auger electron spectroscopy (AES) depth profiles for oxide layers
grown on titanium in ethanolic 0.5 M HF at 10 V for 240 min.

2
652
J. Choi et al. / Electrochimica Acta 49 (2004) 2645–2652
Fig. 12. Schematic diagram of the formation of barrier titanium oxide below the breakdown potential: (a) imprinted Ti substrate before anodization, and
b) oxide growth in inverted prepatterns after anodization.
(
after a long time anodization (>5 h), the monodomain ti-
tania structure becomes flat and eventually disappears due
to the dissolution of the titanium oxide film. Note that the
rectangular shape of the original prepatterns on the sur-
face of the Ti substrate is changed into a more circular-like
shape during the anodization (Fig. 9b). This is also observed
during the anodization of prepatterned aluminum [7]. Note
that the areas which are not imprinted by the stamp have
no pores and stay flat even after anodization, demonstrat-
ing the formation of barrier oxide under these conditions
[4] T. Albrektsson, P.I. Branemark, H.A. Hansson, J. Lindstrom, Acta
Orthop. Scand. 52 (1981) 155.
[5] H.M. Kim, F. Miyaji, T. Kokubo, T. Nakamura, J. Biomed. Mater.
Res. 32 (1996) 409.
[6] H. Masuda, H. Yamada, M. Satoh, H. Asoh, M. Nakao, T. Tamamura,
Appl. Phys. Lett. 71 (1997) 2770.
[7] J. Choi, K. Nielsch, M. Reiche, R.B. Wehrspohn, U. Gösele, J. Vac.
Sci. Technol. B 21 (2003) 763.
[8] J. Choi, G. Sauer, K. Nielsch, R.B. Wehrsphon, U. Gösele, Chem.
Mater. 15 (2003) 776.
[9] V. Zwilling, E. Darque-Ceretti, A. Boutry-Forveille, D. David, M.Y.
Perrin, M. Aucouturier, Surf. Interface Anal. 27 (1999) 629.
(Fig. 12).
[10] D. Gong, C.A. Grimes, O.K. Varghese, W.C. Hu, R.S. Singh, Z.
Chen, E.C. Dickey, J. Mater. Res. 16 (2001) 3331.
[11] R. Beranek, H. Hildebrand, P. Schmuki, Electrochem. Solid-State
Lett. 6 (2003) B12.
4
. Conclusions
[12] K. Nielsch, J. Choi, K. Schwirn, R.B. Wehrspohn, U. Gösele, Nano
Lett. 2 (2002) 677.
We have investigated in detail the electrochemical an-
[13] O. Jessensky, F. Müller, U. Gösele, J. Electrochem. Soc. 145 (1998)
3735.
odization of titanium in different electrolytes, and their po-
tential and temperature dependence. It turned out that due
to the semiconducting properties of titania, a mirror image
of the behavior of porous alumina cannot be accomplished.
So-called porous titania in literature corresponds to the pit-
ting regime of aluminum where pores are created due to
dielectric breakdown of titania or alumina, respectively. Be-
low the breakdown potential of titania, only thick barrier
layers can be formed. However, by nanoimprint of titanium
and successive anodization of titania below the breakdown
potential, monodomain porous titanium oxide with a pore
[
14] A.P. Li, F. Müller, A. Birner, K. Nielsch, U. Gösele, J. Appl. Phys.
4 (1998) 6023.
8
[15] J.-L. Delplancke, A. Garnier, Y. Massiani, R. Winand, Electrochim.
Acta 39 (1994) 1281.
[16] T. Shibata, Y.C. Zhu, Corros. Sci. 37 (1995) 133.
[17] J.-L. Delplancke, R. Winand, Electrochim. Acta 33 (1988) 1539.
[18] C.E.B. Marino, E.M. de Oliveira, R.C. Rocha-Filho, S.R. Biaggio,
Corros. Sci. 43 (2001) 1465.
[19] H. Tang, K. Prasad, R. Sanjinbs, P.E. Schmid, F. Lévy, J. Appl.
Phys. 75 (1994) 2042.
[20] L. Arsov, M. Froelicher, M. Froment, A.H.-L. Goff, J. de Chim.
Phys. 72 (1975) 275.
[21] G. Blondeau, M. Froelicher, M. Froment, A.H.-L. Goff, Thin Solid
Films 42 (1977) 147.
2
depth of 60 nm on a cm -scale can be prepared. We believe
that this work is a first milestone on the way to find new
ordered porous structures for different valve metals.
[
22] Y.T. Sul, C.B. Johansson, Y. Jeong, T. Albrektsson, Med. Eng. Phys.
3 (2001) 329.
23] J.C. Marchenoir, J.P. Loup, J. Masson, Thin Solid Films 66 (1980)
57.
2
[
3
[
[
[
[
24] J. Yahalom, J. Zahavi, Electrochim. Acta 15 (1970) 1429.
25] J.F. MaAleer, L.M. Peter, J. Electrochem. Soc. 129 (1982) 1252.
26] J. Choi, K.H. Lee, Y. Tak, Electrochemistry 69 (2001) 843.
27] J.W. Diggle, T.C. Downie, C.W. Coulding, Chem. Rev. 69 (1969)
Acknowledgements
Authors thank Mr. A. Sobbe for the mechanical polish-
ing of Ti, Dr. M. Reiche for fabricating the imprint master
stamp and G. Sauer of the University Erlangen-Nuremberg
for stimulating discussions.
365.
[28] P.A. Malachesky, in: A.J. Bard (Ed.), Encyclopedia of Electrochem-
istry of the Elements, vol. VI, Marcel Dekker, New York, 1973,
p. 63.
[
29] R.B. Wehrspohn, A.P. Li, K. Nielsch, F. Müller, W. Erfurth, U.
Gösele, in: K.R. Hebert, R.S. Lillard, B.R. Mac Dougall (Eds.), The
Electrochemical Society Proceeding Series, PV 2000-4, Pennington,
NJ, 2000, p. 271.
References
[
[
[
1] J.W. Schultz, M.M. Lohrengel, Electrochim. Acta 45 (2000) 2499.
2] A. Aladjem, J. Mater. Sci. 8 (1973) 688.
3] R. Asahi, T. Morikawa, T. Ohwaki, K. Aoki, Y. Taga, Science 293
[30] R.S. Alwitt, H. Uchi, T.R. Beck, R.C. Alkire, J. Electrochem. Soc.
131 (1984) 13.
[31] M. Kozlowski, W.H. Smyrl, L.J. Atanasoka, R. Atanasoki, Elec-
trochim. Acta 34 (1989) 1763.
(2001) 269.