Home Search Collections Journals About Contact us My IOPscience
The influence of hydroxide on the initial stages of anodic growth of TiO2 nanotubular arrays
This article has been downloaded from IOPscience. Please scroll down to see the full text article.
2010 Nanotechnology 21 505601
(https://www.doczj.com/doc/bc2144288.html,/0957-4484/21/50/505601)
View the table of contents for this issue, or go to the journal homepage for more
Download details:
IP Address: 202.115.35.14
The article was downloaded on 22/02/2011 at 06:58
Please note that terms and conditions apply.
IOP P UBLISHING N ANOTECHNOLOGY Nanotechnology21(2010)505601(8pp)doi:10.1088/0957-4484/21/50/505601
The in?uence of hydroxide on the initial stages of anodic growth of TiO2 nanotubular arrays
Zainab T Y Al-Abdullah1,Yuyoung Shin1,Rantej Kler1,
Christopher C Perry2,Wuzong Zhou3and Qiao Chen1,4
1Department of Chemistry and Biochemistry,University of Sussex,Brighton BN19QJ,UK
2Division of Biochemistry,School of Medicine,Loma Linda University,Loma Linda,
CA92350,USA
3School of Chemistry,University of St Andrews,North Haugh,St Andrews,Fife KY169ST,
UK
E-mail:qiao.chen@https://www.doczj.com/doc/bc2144288.html,
Received10June2010,in?nal form30September2010
Published22November2010
Online at https://www.doczj.com/doc/bc2144288.html,/Nano/21/505601
Abstract
Understanding the mechanism for growing TiO2nanotubes is important for controlling the
nanostructures.The hydroxide nano-islands on the Ti surface play a signi?cant role at the initial
stage of anodization by forming the very?rst nano-pores at the interface between hydroxide
islands and substrate and eliminating the H2O electrolysis.A quantitative time dependent SEM
study has revealed a nanotube growth process with an initial linear increase of pore diameter,
?lm thickness and number of pores.During the anodization of titanium,different current
transient curves are observed for Ti samples with or without hydroxide on the surface.The
transient current pro?le has been quantitatively analyzed by?tting several distinctive stages
based on a growth mechanism supported by SEM observations.It is found that a saturated cubic
dependent equation is appropriate to?t a short current upturn due to the increase of the surface
area.
(Some?gures in this article are in colour only in the electronic version)
1.Introduction
Titanium dioxide is one of the most widely studied oxide semiconductors due to its superior photocatalytic activity[1–3],biocompatibility[4]and stability in water[5]. Moreover,modifying the microstructures and controlling the electronic structure of TiO2can potentially improve the ef?ciency of its application in photovoltaic devices[3,6–9]. TiO2nanotubes are of great interest among the various known titania nanostructures.To date,the most effective way of creating vertically aligned TiO2nanotubular arrays is by anodizing a Ti plate,in a similar manner to the process for fabricating anodic aluminum oxide(AAO)[10],but using different electrolytes containing?uoride[2,3,11–16]. Successful attempts have been reported for controlling the geometry,such as the diameter[15,17],length[9,18],wall 4Author to whom any correspondence should be addressed.thickness[19]and even external structures of the aligned nanotubes[8],by manipulating the anodizing conditions. More recent efforts have been focused on establishing the growth mechanism with the aim of?ne tuning the nanotube morphology.Microstructural studies of AAO and anodic titanium oxide(ATO)have been performed in order to understand this mechanism[20].Qualitative description[15] and quantitative modeling of the electric?eld distribution[21] and growth process[22]have also been established based on experimental observations.However,knowledge about early stage pore growth in AAO and ATO is still very limited[23].
Herein,we report the catalytic behavior of hydroxide species on the initial porous growth.We discuss the possible mechanism of a localized anodization process.We present the result of a time dependent scanning electron microscopy (SEM)study and its correlation with the current transient behavior under potentiostatic conditions.This quantitative
1μ
m
d
Figure 1.SEM images showing the top view (a),bottom view (b)and pro?le view (c)of TiO 2nanotubular arrays.The layer was grown on Ti substrate pretreated with an acid solution containing 1wt%H 3PO 4and 0.1wt%HF in water for 5min followed by 45min anodizing in ethylene glycol electrolyte containing 0.6wt%NH 4F and 2wt%water.The TEM image (d)shows the internal structural details of the same nanotubes.The inset shows the magni?ed bottom of the nanotubes.
analysis allows us to improve our understanding of the early stage growth in ATO.
2.Experimental details
High purity titanium plate (0.5mm,95%)was ultrasonicated in acetone and rinsed in deionized water.The plate was mechanically polished by using different grades of diamond paste down to 100nm grain size followed by sequential ultrasonic cleaning in isopropanol and water.The hydroxide islands were created by dipping the polished Ti plate in an acidic solution containing 1wt%H 3PO 4and 0.1wt%HF for 5min,followed by sonicating in deionized water for 4min.H 3PO 4is used to adjust the solution to pH 1.5,while F ?is used to initialize the formation of an hydroxide layer.Longer time dipping will form a much rougher surface and 4min in contact with the solution is optimized for an appropriate density of isolated hydroxide islands.
The anodization was performed in a home-made electrochemical cell using a clean Ti plate as a cathode.The electrolyte contains 0.6wt%NH 4F (analytical grade)and 2wt%water in ethylene glycol.The anodizing voltage was kept constant at 60V DC with a ?xed electrode separation of 65mm.The surface area exposed to the electrolyte is about 5.0cm 2(2.5cm 2each side).The experiments were performed at room temperature under aerated non-stirred conditions.
The time dependent transient current was recorded using a USB data logger (U12,Labjack).The creation of the hydroxide layer and the evolution of TiO 2nanotubes were
monitored using an SEM (JSM 820M,Jeol)operating at 30kV and a transmission electron microscope (TEM)(Hitachi-7100)operating at 100kV .X-ray photoelectron spectroscopy (XPS)data were recorded with 12keV bias and 10mA emission current.The 2p signals from the clean Ti sample were used to calibrate the spectroscopy.
3.Results and discussion
Figure 1shows the typical morphology of anodized Ti in the electrolyte containing ?uoride.The SEM images show the circular cross section of the nanotubular arrays from the (a)top,(b)bottom and (c)pro?le view directions.The nanotubes are open at the top (?gure 1(a))and closed at the bottom (?gure 1(b))with an average outer diameter of 120nm.In the bottom view SEM image,a dark region is visible in the middle of each nanotube which is probably due to the presence of the hollow tubular structure.Such image contrast was also observed by Chen et al ,recently [24].The pro?le view (?gure 1(c))shows that straight nanotubes have equal length of about 2.5μm.The TEM image,(?gure 1(d))gives detailed morphology of the individual nanotubes with a smooth wall structure.It is clear that the top of the anodized tubes have a consistent inner diameter (about 80nm)and wall thickness (about 20nm)for individual nanotubes.However,at the bottom of the nanotube,the inner diameter is reduced to 30nm with an increased wall thickness of 45nm and the thickness of the barrier oxide layer is about 84nm,which is much thicker than the nanotube wall.
Figure2.SEM images of Ti plates after(a)polishing,(b)acid treatment for30s,(c)acid treatment for90s and(d)acid treatment for90s followed by ultrasonic cleaning in water.
In the following sections,we focus on the mechanism of
the early stage growth of ATO.
3.1.Formation of the hydroxide layer
After the?ne mechanical polishing of the Ti plate with100nm
diamond paste,the surface has a mirror?nish.Ultrasonic
cleaning of the polished sample with isopropanol and water has
no effects on the surface.At this stage,the sample becomes
hydrophobic.The SEM image,shown in?gure2(a),gives a
typical example of a polished surface,which is very smooth
with a low density of defects.The defects are dominated by
polishing grooves with a width of100nm.
In order to create hydroxide nano-islands,the polished
sample is dipped in the acidic solution(1wt%H3PO4and
0.1wt%HF in water).SEM images in?gures2(b)and(c)
show the evolution of the surface morphology,recorded at30
and90s in contact with the acidic solution.A breakdown
of the smooth surface can be clearly identi?ed,which could
be related to the acid-induced corrosion.After90s,the
surface becomes rougher and more hydrophilic.The color of
the sample changes to a gray matt?nish.For removing the
loose fragments,the sample is sonicated in deionized water
for5min.The surface morphology at this stage is shown in
?gure2(d).Many smooth islands are formed with relatively
regular size and shape,possibly in the form of TiO x(OH)4?2x.
The formation of hydroxide species on the acid treated
sample is clearly evidenced by the XPS study,shown in
?gure3,with two components in the O1s peak with binding
energy at531.5and530.1eV,contributed from hydroxide
and oxide respectively.No signi?cant?uorine1s XPS signal
Binding Energy (eV)
Figure3.O1s XPS spectrum of the acid treated Ti sample.
is found at584.8eV.A linear background and a linear
combination of Gaussian and Lorentzian peak shape have been
chosen to?t the spectrum.Similar assignments have been
identi?ed for the TiO2nanotubes[25].However,we have also
noticed that the relative intensity of the hydroxide against the
oxide is much higher on our acid treated surface in comparison
with that found on the TiO2nanotubes[25].
b
Figure4.SEM images recorded from an acid treated Ti plate anodized at different times.
3.2.Time dependent SEM study at different stages of anodization
In order to understand the effects of preformed hydroxide islands on the growth of nanotubes,we monitored the evolution of the sample morphology at each stage of the anodization. Figure4demonstrates a sequence of SEM images taken from acid treated plates anodized for different lengths of time.Each sample was prepared individually with identical conditions. As shown in?gure4(a),?ne etch pits were formed as early as45s into the anodization.What is more signi?cant is that the etching pits are exclusively located at the boundaries between the preformed hydroxide nano-islands and the thin metal oxide?lm formed in the anodization.Therefore,both the oxidation of Ti and dissolution of oxide processes underneath the hydroxide particles are faster than on the clean metal surface.This suggests that the activation energy for creating the etching pits is relatively low on such sites.It can also be observed that the shapes of some etching pits follow the pro?le of the islands rather than being perfectly circular.
One of the possible mechanisms for this preference of initial location for etching is that relatively high concentration, lower valency cations,such as Ti3+,are available at the interfaces between metal and metal hydroxide caused by oxygen vacancies.The oxidation of Ti3+to Ti4+requires much less polarization potential therefore it is kinetically favorable. The oxidation of Ti3+to Ti4+requires oxygen species from the dissociation of water,following the equation(1):
Ti2O3+H2O?2e
?→2TiO2+2H+.(1)
Such oxidation will also increase the local concentration of H+which will promote the dissolution of formed TiO2 species.Alternatively,the presence of hydroxide islands can cause an inhomogeneous distribution of the electric?eld.In particular,the?eld at such interfaces is much stronger than on the?at surface which generates a locally focused positive electrical?eld(sample as anode)and drives the initiation of the etching pits.Such a focused?eld could attract the F?anion moving towards the hydroxide island on the anode.The
ability of F?to form soluble species,such as TiF2?
6,leads to
a permanent chemical attack(dissolution)of formed TiO2and prevents TiO x(OH)4?2x precipitation.
Figure4(b)shows the SEM image of the surface after 90s anodization.The density and diameter of the etching pits increase while some etching pits start to develop in the areas between the hydroxide islands.At this stage,the average pore diameter is about30±8nm(averaged over65 pores).Nevertheless,the shape of the pore is still irregular. Further anodizing creates more etching pits with slightly increased diameter,40±8nm,as shown in?gure4(c).More signi?cantly,as the dimension of the nanopore increases,the number of such pores decreases in between200and500s. This suggests that there is a merging of nearest pores as they
become bigger.At1400s,almost perfect nanotubular feature can be observed on some parts of the image(middle bottom).
For further understanding the growth behavior of the TiO2 nanotubes,the diameter and number of pores,together with the oxide?lm thickness as a function of anodization time are also measured from the SEM images,shown in?gure5.It is clear that,at the early stage of anodization,both the?lm thickness and the pore diameter linearly increase with time, while the number of pores actually reaches its maximum value of180μm?2at250s.This gives quantitative evidence of the merging of the porous structures.Until1200s,nanotubes with the maximum diameter are achieved,which indicates the transition from nano-porous to https://www.doczj.com/doc/bc2144288.html,bining this pro?le for the number of pores together with the diameter of the pores,one can quantitatively predict that the maximum surface area is achieved around250s.
The oxide?lm thickness was measured from the cross sectional view of the SEM image.Here,we are only able to identify the overall thickness rather than detailed information, such as oxide barrier layer thickness and nanotube length. Nevertheless,it is worth mentioning that the thickness of the oxide?lm increases almost linearly from the beginning of the anodization until1200s,at which time the growth rate gradually decreases.This could be due to the slowing of the oxidation rate on the bottom of the nanotube or the increasing of dissolution at the top of the nanotubes.Since the anodization current does not show signi?cant decrease,we can conclude that the decrease of the growth rate is likely to be the result of the dissolution.
3.3.Current transient of the anodization process
In order to further understand the growth mechanism, anodization current is recorded as a function of time.Curve A in?gure6(a)shows the typical current behavior of anodizing a freshly polished sample(sample A,without acidic treatment), while curve B corresponds to the current transient for the sample covered with hydroxide islands(sample B).For sample A,at the beginning of the anodization,large current accompanied with evolution of gas bubbles(O2)on the anode is observed.The current decreases sharply within9s from 37.5mA,followed by a steady high current at35.0mA,which is then gradually reduced to20.0mA at450s.Meanwhile,the release of O2bubbles is gradually diminished at this
stage.Figure5.(a)Measured number of pores(green,crossed line)and diameter of pores(red,dotted line);(b)measured oxide?lm thickness as a function of anodization time.
There are several competitive electrochemical reactions on the anode,which include the oxidation of Ti and dissociation of H2O.The initial large evolution of O2gas indicates that direct oxidation of the metallic Ti is less effective because the bias is focused at the interfacial region between the metal/electrolyte which offers excessive over-potential for H2O hydrolysis. Similar?eld enhanced water dissociation during the AAO formation has been studied by using a computational chemical method recently[26].The process is more important in ATO, because the produced OH?can stay in the nanotubes for longer time and even form a double-layered wall[20,21].The involvement of multiple contributions of competitive reactions in the initial anodization process makes it impossible to build a quantitative model for the transit behavior.During the?rst 450s,an oxide layer is gradually developed and the anodic voltage is dropped within this oxide layer.This leaves much less anodic bias at the oxide/electrolyte interface for H2O electrolysis.Meanwhile,the oxide layer is suitable for the migration of oxide anions and therefore the ionic conductivity increases.The overall current becomes stable at this stage. In this region,the nature of conductivity changes from being electronic to ionic dominated.
Surprisingly,the O2gas evolution is completely eliminated on the acid treated sample.Thus,the contribution of water hydrolysis in the anodization current is minimized.
Figure 6.Anodization current as a function of time.(a)Curve A (blue)was from a polished Ti,and curve B (red)from an acid treated Ti.Anodizing regions are also indicated (D2and D3).The current behavior from the acid treated sample is analyzed in (b)with the exponential decay (curve C,brown)?tted at the leading edge of the current drop (curve B,red)and the residual current (curve D,black).(c)The increased current (black dot)is ?tted with a saturated cubic function (D2,solid,green).(d)The residual current decay (red)is ?tted with an exponential decay function (D3,black).(e)A summary of the three best ?tting independent functions.(f)Overall curve ?tting (blue)overlapped on the experimental current curve (red).
Such a sample (B)gives a very different I –t plot.The typical transient current curve can be separated into three regions (curve B in ?gure 6(a)).In the ?rst,the current shows a leading exponential decay within 68s down to 11.2mA,during which a compact oxide layer is formed.In the second,the decay is followed by a short up turn of the current up to 15.5mA,which can be generally described as the result of growth of pores with increasing surface area.Finally,the current undergoes a steady decrease to 11.6mA at 2800s,indicating the growth of nanotubes.Although similar current behavior has been previously reported by several groups [27,28],to our knowledge,this is the ?rst time that the catalytic role of hydroxide islands has been identi?ed.More importantly,without the interference of water hydrolysis,the anodization current is dominated by the migration of oxide anions associated with the formation of nanostructures.Only under such controlled conditions does it becomes possible to model the anodization process.
For a quantitative understanding of the three regions of the current behavior recorded from sample B,we ?t the current curve with three independent functions,each re?ecting their physical properties with de?ned parameters and functionality.The three regions have been speci?ed as the initial formation of an oxide barrier layer (0s The initial current transient behavior can be described by an exponential decay function,equation (2): I 1=a exp (?bt )(2) where the parameter a corresponds to the initial anodization current and b de?nes the curvature of the decay process,both being related to the electrolyte concentration,surface area,working temperature,anodizing voltage,electrode separation and,more importantly,the growth rate of the oxide layer.Speci?cally,the initial anodization current,a ,is determined by the ion conductance of the electrolyte solution.Our experimental measurement has identi?ed that parameter a is linearly proportional to the electrode surface area,electrode separation and anodization voltage.The rapid decay of anodization current is due to the increase of the barrier thickness.Parameter b represents the rate constant for the growth of the oxide barrier layer.The quantity a /b de?nes the integration area of the current exponential decay curve,which corresponds to the total charge (Coulomb),Q ,used during the formation of the initial oxide layer.With the de?ned surface area,S ,of the Ti electrode,the correlation between the oxide layer thickness,T ,and the measured a ,b parameters of the current curve can be de?ned as: Q = I 1d t =a /b =S ×T ×4e /V c (3)where e is the electron charge and V c is the volume of each TiO 2unit,about 31.2?A 3.The quantity S ×T ×4e /V c represents the charge used for creating an oxide layer with a volume of S ×T .The best matching of the exponential decay is shown in curve C in ?gure 6(b)with parameters a =35.15mA and b =0.0196s ?1.The surface area (S )of this sample is about 5cm 2(both sides).Using equation (3), we can estimate the maximum thickness of the formed metal oxide of about175nm,which gives a growth factor of 2.9,higher than previously reported by Schmuki[29].The difference could be attributed to the fact that the maximum oxide thickness could only be achieved at the equilibrium condition at which the anodization current is reduced to zero. With previous experimental data and observation methods[29] such an equilibrium could not be physically achieved.Instead, our extrapolating method by quantitatively analyzing the exponential decay curvature of the initial anodization process has enabled us to extract the true growth factor without the limitation of achieving an equilibrium condition. The process of creating the porous structure at a very early stage increases the anodization current by increasing the surface area.By assuming that the pores are hemispherical, the surface area,S(t),as a function of time is described in equation(4): S(t)=S(0)?Nπr2+4/2Nπr2(4) where S(0)is the initial?at surface area,N and r represent the number and the radius of the pore structures at time t,Nπr2 corresponds to the loss of the?at circular area and4/2Nπr2 corresponds to the addition of hemispherical pore surface area. It is statistically reasonable to assume that both the number of pores(N)and the radius of the pores(r)increase linearly with time(t).In other words,N=nt and r=mt,where n and m are proportionality constants dependent on the growth behavior.The assumption satis?es the initial condition at t=0s,N=0and r=0nm and allows equation(4)to be transferred as a function of time S(t)=S(0)?ntπ(mt)2+4/2ntπ(mt)2 =S(0)+ntπ(mt)2=S(0)+Lt3(5) where L=nπm2.Most importantly,the surface area increases following a cubic dependence on the anodization time.It should be emphasized that the increase of the surface area will be limited by a maximum value when the porous cross sectional area Nπr2is approaching the initial?at surface area, S(0).This leads to a simple conclusion that the maximum anodization surface area equals2S(0),double that of the initial surface area. To re?ect the cubic dependence at the initial stage,as well as the limitation of the maximum surface area,we propose a three parameter saturated cubic formula,equation(6),to describe the increasing of anodization current,based on an assumption that the anodization current is proportional to the effective surface area at the electrolyte/oxide interface: I2=ct3/(t3+d)+e.(6) The factor d is introduced in the above equation so that when t3 d,the curve follows the cubic behavior,signifying the growth of porous structure,while for t3 d,the current becomes?at with a saturated value of(c+e)suggesting the maximum current is achieved,which is related to the maximum surface area.Here,e represents the initial current before the formation of the porous structure and c represents the maximum anodic current contributed by the increasing of surface areas.At the initial stage of increasing porous structure,I2≈ct3/d+https://www.doczj.com/doc/bc2144288.html,pared with equation(5),it is clear that the combined parameter c/d is proportional to nπm2, corresponding to the increasing rate of the surface area.The 3 √ d represents a critical tim e after which the rate o f increasin g of surface area decreases. In?gure6(b),the residual current,curve D,is calculated by subtracting the curve C from the measured curve B.The best?tting of this part of the process(curve D2,region2) is shown in?gure6(c),with parameters c=19.72mA, d=915624s3and e=?3.07mA.With a similar saturated formula,instead of t3,we have also tested the best curve?tting as a function of a variety powers of t.The standard deviations are0.38,0.25,0.09and0.18mA for the least squares?tting as a function of t,t2,t3and t4respectively.Therefore,the cubic dependent equation gives the best overall result.It is worth mentioning that if the increasing of surface area is dominated by the increase in the number of pores,the current will follow a linear relationship.Alternatively,if the process is dominated by the increase of pore diameter,it will then follow a square relationship.Under our particular experimental conditions,the current curve follows a cubic relationship,which suggests that both the number of pores and the size of the pores increase in that period of time,as proved with SEM observations(see section4).The curve D2(region2)passes0.0mA at about 55±10s,indicating the true starting point of pore formation (?gure6(c)). While the effective anodization area is increasing,the length of the nanotubes will also start to grow.However,its in?uence on anodization current is too small to be noticed at this early stage of the growing process.Once the surface area is maximized,a slow decay of anodization current is observed,as shown in?gure6(d)(red).Carefully examining the curvature reveals the subtle change of the decay rate during the rest of the anodization.We?t this current behavior with an exponential decay curve as I3=f0+f1exp(?gt),with f0=9.5mA,f1=6.2mA and g=3.6×10?4s?1. The parameter f0represents the steady(minimum)anodization current.This gradual decay of the anodization current was previously attributed to the effect of limited ion diffusion and concentration gradient within the nanotubes[30].Here,the major error comes from the dif?culty in identifying the leading edge of the linear decay of current,which might contain residual contributions from the initial exponential decay and continuous increasing of the surface area.The details of the growth mechanism that contributes to this exponential decay of current have been discussed with a quantitative SEM analysis in section3.2. Compared with the SEM study,it is clear that although the diameter of the nanotube is gradually increasing,the number of nanotubes has increased to its maximum value at about200s. Therefore,the total surface area is maximized at this time.This corresponds to the end of region2with a peaked anodization current.Also,the length of the nanotube is almost linearly increasing until about1500s.This will cause the slow decay of anodization current in region3which will then reach a steady current towards the end of the anodization. Figure6(e)summarizes three functions used to analyze the three distinctive regions of the anodization process.The overall ?tting of the experimental result(thick blue curve)is shown in ?gure6(f)with a standard deviation of0.14mA.For individual regions,the leading edge exponential decay(C),D2and D3, the standard deviations are0.96,0.09,and4.0mA respectively. Therefore,the analytical error is mainly originating from the D3region,which is due to some extended contribution from region2. Comprehensive analysis of the transient current pro?le has allowed us to establish quantitative models based on individual growth mechanisms.Furthermore,the understanding of such processes would enable us to control and optimize the anodization process for obtaining the desired morphology of the nanotubes.For instance,larger diameter nanotubes may be achieved by current ramping in region2.The analysis is also helpful for us to understand how the parameters,such as pH, [F?]and temperature,are correlated with current density and the nanotube dimensions. 4.Conclusion The anodization current pro?les of the polished and acid pretreated Ti plates are different.In the latter case,the acid treatment results in hydroxide nanoparticles on the metal surface.These hydroxide particles can enhance the formation of pits to initiate the pore formation,while they are slowly dissolved under the anodization conditions.The current transient pro?le has been?tted into four distinctive stages.It was found that a cubic dependent equation is appropriate to?t the current increases after the exponential decay. Acknowledgments The authors thank Dr J Thorpe at the Sussex Centre for advanced Microscopy for TEM investigation and Dr J F C Turner at Sussex University for helpful discussions. The authors also thank Professor B E Hayden at Southampton University for taking XPS data.QC thanks Dr S Firth at University College London for donating the Jeol SEM.ZTY A gratefully acknowledges the Iraqi Ministry of Education for the scholarship. References [1]Liu Y B,Gan X J,Zhou B X,Xiong B T,Li J H,Dong C P, Bai J and Cai W M2009J.Hazard.Mater.171678 [2]Allam N K and Grimes C A2009Langmuir257234 [3]Grimes C A2007J.Mater.Chem.171451 [4]Park J,Bauer S,Schmuki P and von der Mark K2009Nano Lett.93157 [5]Sordo C,Van Grieken R,Marugan J and Fernandez-Ibanez P2010Water Sci.Technol.61507 [6]Gao X F,Sun W T,Hu Z D,Ai G,Zhang Y L,Feng S,Li F and Peng L M2009J.Phys.Chem.C11320481 [7]Kang T S,Smith A P,Taylor B E and Durstock M F2009Nano Lett.9601 [8]Kim D,Ghicov A,Albu S P and Schmuki P2008J.Am.Chem. Soc.13016454 [9]Kim D,Lee K,Roy P,Birajdar B I,Spiecker E and Schmuki P2009Angew.Chem.Int.Ed.489326 [10]Su Z X,Hahner G and Zhou W Z2008J.Mater.Chem. 185787 [11]Bauer S,Park J,Faltenbacher J,Berger S,von der Mark K and Schmuki P2009Integr.Biol.1525 [12]Berger S,Kunze J,Schmuki P,Valota A T,LeClere D J, Skeldon P and Thompson G E2010J.Electrochem.Soc. 157C18 [13]Chen P,Brillet J,Bala H,Wang P,Zakeeruddin S M and Gratzel M2009J.Mater.Chem.195325 [14]Macak J M,Albu S P and Schmuki P2007Phys.Status Solidi RRL1181 [15]Macak J M,Hildebrand H,Marten-Jahns U and Schmuki P2008J.Electroanal.Chem.621254 [16]Ghicov A and Schmuki https://www.doczj.com/doc/bc2144288.html,mun.2791 [17]Yoriya S and Grimes C A2010Langmuir26417 [18]Yoriya S,Prakasam H E,Varghese O K,Shankar K,Paulose M, Mor G K,Latempa T J and Grimes C A2006Sensor Lett. 4334 [19]Mohamed A E,Kasemphaibulsuk N,Rohani S and Barghi S2010J.Nanosci.Nanotechnol.101998 [20]Su Z X and Zhou W Z2008Adv.Mater.203663 [21]Su Z X and Zhou W Z2009J.Mater.Chem.192301 [22]Thebault F,Vuillemin B,Oltra R,Kunze J,Seyeux A and Schmuki P2009Electrochem.Solid State Lett.12C5 [23]Lu S J,Su Z X,Sha J and Zhou W https://www.doczj.com/doc/bc2144288.html,mun. 5639 [24]Chen Y,Wang X M,Lu S S and Zhang X2010Electrochem. Commun.1286 [25]Lee B G,Choi J W,Lee S E,Jeong Y S,Oh H J and Chi C S 2009Trans.Nonferr.Met.Soc.China19842 [26]Su Z X,B¨u hl M and Zhou W Z2009J.Am.Chem.Soc. 1318697 [27]Valota A,LeClere D J,Skeldon P,Curioni M,Hashimoto T, Berger S,Kunze J,Schmuki P and Thompson G E2009 Electrochim.Acta544321 [28]Mor G K,Varghese O K,Paulose M and Grimes C A2005Adv. Funct.Mater.151291 [29]Yasuda K,Macak J M,Berger S,Ghicov A and Schmuki P2007J.Electrochem.Soc.154C472 [30]Macak J M,Tsuchiya H,Ghicov A,Yasuda K,Hahn R, Bauer S and Schmuki P2007Curr.Opin.Solid State Mater. Sci.113
相关主题
文本预览