Effect of Ultrasonic Waves on the Formation of TiO2 Nanotubes by Electrochemical Anodization of Titanium in Glycerol and NH4F

The present work describes the anodic growth of self-organized TiO2 nanotubes (TiNT) on titanium samples in an electrolyte containing 3% NH4F and 5% H2O in glycerol. Two stages anodization were proceeded. Initially, the potential was ramped from 0V to 30, 40 and 60 V. In the second stage the potential was held at the given potential up to 5 hrs. The application of ultrasonic waves during anodization was found to increase the reaction rates as it decreased the concentration gradient of the anodization products around the electric double layer. This resulted in high current and faster growth rates of the nanotubes. The diameters of the TiNT created under ultrasonic wave condition were found to be smaller than the corresponding TiNT created under magnetic stirring. However, the morphology of the TiNT was badly affected as a result of surface perturbation resulting from heterogeneity in surface energy density. A better route for good looking TiNT is to grow them under magnetically stirred solution followed by cleaning by sonication in deionized water for 30 seconds. Increasing the anodization potential from 30 to 60 V increased the dimensions of the TiNT from around 90 nm to around 150 nm, respectively. [DOI: 10.1380/ejssnt.2009.84]


I. INTRODUCTION
Nanotubes are of great interest due to their high surface-to-volume ratios and both size and composition dependant properties.In addition, their high degree of organization can provide direct charge transport along the length of the nanotubes, avoiding random charge hopping across and trapping at nanoparticles grain boundaries [1].Several recent studies have indicated that titania nanotubes have improved properties compared to any other form of titania for application in photocatalysis [2,3], sensing [4][5][6][7], photoelectrolysis [8][9][10][11] and photovoltaics [11][12][13][14].For biocombatibility, hydroxyapatite was deposited onto the titania nanotubes [15].
The two basic criteria for growth of the nanotubes array are sustained oxidation of the titanium and pore growth by chemical/field-assisted dissolution of the formed oxide.The dimensions of the nanotubes are determined by the dynamic equilibrium between growth and dissolution processes.The electrochemical anodization of titanium in fluorinated electrolytes is a relatively simple method to synthesize porous or tubular structures.TiO 2 nanotubes fabricated by this method are highly ordered, have high-aspect ratios and are oriented perpendicular to the substrate, e.g.Ti foil.The nanotube arrays have a well-defined pore size, wall thickness and tube length.Thus far, various electrolytes, such as HF electrolytes [17], HF/H 2 SO 4 mixtures [18], KF/NaF electrolytes [19], sulfate electrolytes containing small amounts of fluoride ions and citric acid [19], chromic acid/HF mixtures [20], NaF/Na 2 SO 4 solutions [21], NH 4 F/(NH 4 ) 2 SO 4 electrolytes [22], (NH 4 )H 2 PO 4 /NH 4 F solutions [23], H 3 PO 4 /HF mixtures [23], HF/acetic acid mixtures [24] and 2-propanol-water-NH 4 F [25] have been used to form the TiO 2 nanotube arrays.Organic solvents provide alternative for the electrolytes based on mineral acids or salts.The mineral acids are more corrosive and less controllable while the former are easier to control the TiO 2 dissolution.
As most of the applications utilising the TiNT are based in properties like dimensions (surface area), cleanness, and high degree of anodization, this necessitates improving several parameters in the preparation procedures.These include as e.g.mode of stirring, nature of electrolyte, potential, temperature, current density distribution (geometry of the working electrode and the cell), and the nature and surface finish of the anodizing target material.This paper compares the electrochemical anodization of Ti in glycerol-3% NH 4 F-5% H 2 O as electrolyte under ultrasonic wave and magnetic stirring.The ultimate goal is to improve understanding the interfacial reactions and their relation with some surrounding factors during the anodization process.The effect on the integrity of the TiNT will be discussed.In addition, the effect of potential on controlling the dimensions of the TiNT will be also discussed.

II. EXPERIMENTAL
Pure titanium sheet (99.5 % and thickness 0.5 mm) was purchased from Nilaco, Japan.Pieces of 1.5 cm × 1.5 cm were used in all experiments.Prior to anodization, the samples were finely polished using 9 µm diamond paste.The samples were then sonicated consecutively with acetone, 2-propanol and methanol, then rinsed with deionized (DI) water and dried in air stream.Electrochemical anodization was conducted in a self fabricated Teflon cell.The samples were pressed against a hole with o-ring.The electrolyte is a mixture of 3% NH 4 F-5% H 2 O-glycerol.All chemicals were purchased from Wako, Japan.They were used without further purification.The anodization was conducted using binary electrode setup with a platinum counter electrode.The electrochemical device used was a Potentiostat/Galvanostat HA-3001A equipped with arbi- trary function generator HB-105, both are from Hokuto-Denko Corp. Japan.The potential was swept from 0 V to 30-60 V at a rate of 200 mV/s followed by holding at 30-60 V for almost 5 hrs.
The morphology of the obtained nanostructure was inspected by the Field Emission Scanning Electron Microscope (FESEM), Hitachi S-4800.

A. Anodization process
The electrochemical anodization was conducted in two stages.In the first stage the potential ramped from 0 to 30 V with a scan rate of 200 mV/s.This was found to help building a barrier TiO 2 layer.The corresponding current was recorded as a function of potential.Figure 1(a) shows the time-variation of potential during the two stages of the anodization.In region I, the current increases more slowly with potential.This is believed to correspond to slower formation of TiO 2 while the reaction is under activation control.In region II, the potential is high enough to generate electric field that overcomes the activation barrier and a faster oxidation occurs.
The second stage of the anodization involves a poten- tial hold of titanium in 3%NH 4 F -5% H 2 O-glycerol at 30 V for almost 5 hrs. Figure 2 shows the variation of current with time during this potentiostatic anodization stage.The effect of ultrasonic waves in comparison with magnetic stirring is studied.In general, the current decreased exponentially with time.Initially, the fast decay in current indicates the formation of the high resistant TiO 2 film, according to the following equation, Under sufficient high potential the electric field will be strong enough to take the Ti 4+ ions off the surface leaving behind some voids in the interpore areas [26].Fluoride ions are able to dissolve the TiO 2 forming complex compound, as follows; The driving force for reaction (2) is high as judged from a comparison of the standard free energies of formation of -820 and -2180 kJ/mol for TiO 2 and [TiF 6 ] 2− aq , respectively [27].It is noteworthy that, under the same conditions, the anodization reactions are faster under ultrasonic waves compared with that under magnetic stirring.This was judged by the higher current values observed under ultrasonic waves conditions, see Fig. 2. Interestingly, the current under magnetic stirring decays rapidly reaching almost steady state constant values.In contrast, under ultrasonic waves one can observe oscillation in current.This may suggest the instability of the formed film or the nanotubes under ultrasonic waves.
The key reactions responsible for the formation of anodic nanoporous Al 2 O 3 [28][29][30][31][32][33][34] and TiO 2 [36][37][38][39] apparently are similar.Initially, oxide grows at metal surface due to interaction of metal with O 2− or OH − ions [29].Then these anions, as the oxide thicken, need to diffuse through to reach the metal surface.At the same time Ti 4+ ions diffuse out to the oxide electrolyte interface and this is facilitated by strong electric field.Simultaneously, http://www.sssj.org/ejssnt(J-Stage: http://www.jstage.jst.go.jp/browse/ejssnt/) field assisted dissolution of oxide at oxide/electrolyte interface takes place [29].Due to high electric field the Ti-O bond undergoes polarization and is weakened promoting dissolution of the metal ions.Chemical dissolution of TiO 2 by fluoride (Eq.( 2)) takes place.The rate of oxide growth at the metal/oxide interface and the rate of oxide dissolution at the pore-bottom/electrolyte interface reach equilibrium.Thereafter, the thickness of the barrier layer remains unchanged although it moves further into the metal making the pore/tube deeper.

B. Morphology of the anodized surface
Figure 3 shows FESEM images of the Ti samples anodized, as in Fig. 2, under different conditions.Figure 3(a) shows the surface of the samples anodized under ultrasonic waves condition.This sample shows formation of corrupted nanotubes where many of them are broken.Figure 3(b) shows the surface after anodization with no ultrasonic but the electrolyte was stirred by magnetic stirrer.The image shows homogeneous nanotubes that are overlayed by some broken nanotubes.When this sample was ultrasonically cleaned for 30 seconds in deionized water the broken nanotubes removed and the surface became clean showing self organized homogeneous titania nanotubes (TiNT).It is noteworthy to observe that the dimensions of the nanotubes created under ultrasonic conditions are 50-60 nm, Fig. 3(a).Under magnetic stirring condition, the sizes of the TiNT are 70-90 nm, Fig. 3(c).It is believed that the higher homogeneity in the electrolyte under ultrasonic conditions has generated more nucleation sites for the initiation of the nanotubes.This results in formation of TiNT with smaller diameters.
To explain the effect of ultrasonic waves on the integrity of the nanotubes we have to consider the different forces (stresses) and strains that may affect the nanotubes.The mechanical stability or instability of the TiNT may be controlled by interaction of two competing processes, i.e. , (1) surface energy acting as a stabilizing force, and (2) increase in strain energy due to electrostriction, electrostatic and recrystallization stresses trying to destabilize the surface.The electrostriction stress (σ er ), can be expressed as in Ref. [40] where γ is the coefficient of σ er in the direction of the electric field E, and Y is Young's modulus.Van Sterkenburg [41] stated that electrostriction strain is too small and is dilatational for the non-ferroelectric materials.So in case of TiO 2 this component can be neglected as thermal expansion could be longer than electrostriction strain [40].The electrostatic (Maxwell) stress, σ es , can be expressed as: where ε 0 and ε are permittivity of space and relative dielectric constant of the TiNT layer.The volume expansion due to formation of oxide layer induces compressive stress can be given by: where (∂v/v) is the volumetric strain and ν is the Poisson's ratio.
The total energy is given by These equations ( 3)-( 6) hold good for a stable surface with no perturbation.In case of a surface subjected to sinusoidal wave perturbation e.g.ultrasonic waves, the energy density varies across the top (crests) and the bottom (valleys) of the sinusoidal wave on the surface.The valleys will show high stress concentration effect consequently have higher strain energy.This heterogeneity in strain energy density along the surface acts as a driving force for the instability and breakdown of the nanotubes.http://www.sssj.org/ejssnt(J-Stage: http://www.jstage.jst.go.jp/browse/ejssnt/) e-Journal of Surface Science and Nanotechnology

C. Effect of potential
The anodization process was conducted at higher potentials (40 and 60 V) in glycerol-5%H 2 O-3%NH 4 F electrolyte under magnetic stirring condition.Figures 4(a) and (b) show the anodization at 40 and 60 V, respectively.Increasing the potential of anodization to 40 V has improved the degree of order of the nanotubes.The diameters of the nanotubes have increased relatively, from 60-100 nm at 30 V increased to 90-130 nm at 40 V and up to 150 nm at 60 V. Higher anodization voltages increase the oxidation and field-assisted dissolution hence increase in the dimensions of the nanotubes can be achieved before equilibrating with the chemical dissolution.
The total amount of electric charge that has passed through the cell is given by Eq. ( 7).This amount can be estimated by calculating the area under each curves of Fig. 2. Consequently the mass of Ti correspond to Eq. ( 2) is calculated using Faraday's law, Eq. ( 9) FIG.5: Equivalent electric charge during anodization, calculated from areas under the curves in Fig. 2, and the corresponding mass of TiO2 that forms, under ultrasonic or magnetic condition.
where Q is the amount of charge in coulomb, i is the current in A , t is time in seconds, m is the mass in grams, for titanium n = 4 and the atomic mass = 47.9 and f is the Faraday's constant.
The mass of Ti is converted to the equivalent TiO 2 , that would form during the anodization process.In these calculations we assume that reaction (2) is the only faradic reaction occurring on the anode surface.Note that the mass given here is the integral mass of TiO 2 that would form.This includes the proportion used for the base layer, for the TiNT layer and that breaks or dissolves and falls into the electrolyte.Figure 5 shows the amounts of Q and the mass of TiO 2 that would form ultrasonic and magnetic stirring conditions.It can be seen clearly that the ultrasonic waves have an observable effect on increasing the amount of TiO 2 that form by increasing the rate of oxidation of titanium.As the ultrasonic waves are responsible for repeatedly oxide break down (which was shown by the oscillation in current Fig. 2) this means that they expose more surface and hence speed up the rate of oxidation.Another possible explanation is when the tube breaks away from selective sites this causes pores or cracks or channels which speed up the ions diffusion.This in turn increases the rate of oxidation of the surface.

IV. CONCLUSIONS
We observed that self-organized titania nanotubes have been grown on the surface of Ti in 3% NH 4 F -5% H 2 O http://www.sssj.org/ejssnt(J-Stage: http://www.jstage.jst.go.jp/browse/ejssnt/) -glycerol.The ultrasonic waves were found to perturb the surface causing surface corruption of the TiNT, due to formation of heterogeneous surface energy density distribution.Although mostly broken, the TiNT created under ultrasonic conditions appears to have smaller diameters.Magnetic stirring although caused some nanotubes to break, the surface titania nanotubes were more homogeneous and well organized.In the latter case post anodization cleaning in deionized water for 30 seconds was found to be necessary to remove the overlayed broken nanotubes.Increasing the anodization potential from 30 to 60 V increased the dimensions of the TiNT from around 90 nm to around 150 nm, respectively.
FIG. 1: (a) Variation of potential with time during anodization of Ti, and (b) potentiodynamic polarization with a scan rate of 200 mV s −1 during the initial stage of the anodization in glycerol-5%H2O-3% NH4F solution under ultrasonic condition at 25 • C.

Figure 1 (
b) shows the current-potential relationship for the anodization at the first stage.Figure 1(b) is clearly classified into two regions.

FIG. 3 :
FIG. 3: FESEM images of the surfaces of Ti samples anodized in glycerol -5% water -3% NH4F solution at 30V, (a) under ultrasonic waves, (b) magnetic stirring and (c) is sample (b) after cleaning by sonication in deionized water for 30 seconds.

FIG. 4 :
FIG. 4: FESEM images of the surfaces of Ti samples anodized in glycerol -5% water -3% NH4F solution at (a) 40V, and (b) and (c) 60 V.Both samples anodized under magnetic stirring condition and cleaned by sonication in deionized water for 30 seconds.