Effect of Applied Potential on the Formation of Self-Organized TiO 2 Nanotube Arrays and Its Photoelectrochemical Response

Self-organized TiO2 nanotube arrays have been fabricated by anodization of Ti foil in an electrochemical bath consisting of 1 M of glycerol with 0.5 wt% of NH4F. The effects of applied potential on the resulting nanotubes were illustrated. Among all of the applied potentials, 30 V resulted in the highest uniformity and aspect ratio TiO2 nanotube arrays with the tube’s length approximately 1 μm and pore’s size of 85 nm. TiO2 nanotube arrays were amorphous in as-anodized condition. The anatase phase was observed after annealing at 400◦C in air atmosphere. The effect of crystallization and effective surface area of TiO2 nanotube arrays in connection with the photoelectrochemical response was reported. Photoelectrochemical response under illumination was enhanced by using the annealed TiO2 nanotube arrays which have larger effective surface area to promote more photoinduced electrons.


Introduction
Since titanium oxide (TiO 2 ) is a commercial product in the early twentieth century, it is believed to be the most promising photocatalyst, due to its great capacity for oxidation, wide band gap, nontoxicity, low-cost, widespread availability, and long term stability [1][2][3][4][5].Due to its wide range of functions, TiO 2 photocatalyst can be applied to the field of environmental cleanup including deodorization, antibacterial protection, antifouling protection, water treatment, emission gas treatment, dye-sensitized solar cells, hydrogen generation by water photoelectrolysis, gas sensors, and so on [4][5][6][7][8].These applications can be roughly divided into "environmental" and "energy" categories.In this context, simple anodization method has caught the attention of the scientific community because vertically oriented and highly ordered TiO 2 nanotube arrays can be produced via this method [2,[9][10][11][12].
Heterogeneous photocatalysis is a well-known process in which a combination of photochemistry and catalysis are operating together [5,13].It implies that both light and catalyst are necessary to bring out the chemical reaction.Upon absorption of photons with energy larger than the band gap of TiO 2 , electrons are excited from the valence band to the conduction band, creating electron-hole pairs [4,[13][14][15].The valence band holes are powerful oxidants (+1 to +3.5 V versus NHE depending on the semiconductor and pH), while the conduction band electrons are good reductants (+0.5 to −1.5 V versus NHE).During photocatalytic hydrogen production, an essential photogeneration of electron-hole pairs is required.When photocatalysis is applied to perform water splitting process for hydrogen production, the reducing conduction band electrons become important as their role is to reduce protons to hydrogen molecules [14,[16][17][18].
The photocatalytic hydrogen production of TiO 2 is largely controlled by (i) the light absorption properties, (ii) reduction and oxidation rates on the surface by the electron and hole, and (iii) the electron-hole recombination rate [14,19].However, the main barriers of these activities are the rapid recombination of photogenerated electron-hole pairs as well as the backward reaction and the poor activation of TiO 2 by visible light [15,20].Shifting the threshold of its photoresponse into the visible region would definitely enhance its potential for solar energy conversion [14,21].
In response to these problems, continuous efforts have been made to fabricate a highly efficient photocatalyst with a suitable architecture that could minimize the recombination of photogenerated electron-hole pairs.Therefore, in this paper, a detailed study has been performed to evaluate the morphology of the anodized Ti foil in different applied potential.The one-dimensional (1D) structure of nanotube arrays provides a high surface-to-volume ratio and excellent charge transfer properties if the precise design and control of the geometrical features are determined [21][22][23][24].Such a mechanistic understanding is very important for the controlled growth of ordered TiO 2 nanotube structure that finds potential use in the development of viable hydrogen fuel cell for sustainable energy system.

Experimental
TiO 2 nanotube arrays were grown by anodic oxidation of titanium foils (4 cm × 1 cm) with 99.6% purity and 0.127 mm in thickness.Prior to anodization, the Ti foils were degreased by sonication in ethanol for 30 min.The foils were then rinsed in deionized water and dried with nitrogen stream.The anodization was performed in a twoelectrode configuration bath with the Ti foil as the anode and the platinum rod as the counterelectrode.The electrolyte was homogenized via magnetic stirring.An experiment was done by adapting optimum condition for TiO 2 nanotubes formation, that is, anodization of Ti foil is performed at the feed rate of 0.1 V/sec, and the electrolyte is 0.5 wt% of NH 4 F in 1 M of glycerol (85% glycerol and 15% water) for 30 minutes.The applied potential was varied from 10 to 50 V.The anodized Ti foils were cleaned using acetone and dried in nitrogen stream.After the cleaning and drying treatments, the as-anodized samples were annealed at 400 • C for 2 hours in air atmosphere to convert to the TiO 2 anatase phase.
The morphology of the anodized Ti foils was viewed using a field emission scanning electron microscope (FESEM) Zeiss SUPRA 35 VP at a working distance of around 1 mm.In order to obtain the thickness of the anodic oxide formed, cross-sectional measurements were carried out on mechanically bent samples, where a partial liftoff of the anodic layer occurred.On top of that, the elemental analysis, that is, atomic percentage of the TiO 2 , can be determined with EDX (energy dispersion X-ray) which is equipped in the FESEM.Phase determination of the anodic layer was determined by an X-ray diffraction (XRD) using Philips, PW 1729, operated at 45 kV and 40 mV.
The photoelectrochemical response of the samples were characterized using a three-electrode photoelectrochemical cell with TiO 2 nanotube arrays as the working photoelectrode, platinum rod as the counterelectrode, and saturated calomel electrode (SCE) was used as the reference electrode.1 M KOH with 1 wt% of ethylene glycol solution was used as the electrolyte in this experiment.All of the three electrodes were connected to a potentiostat (μAutolab III).A Xenon lamp with an intensity of 800 W/m 2 (Zolix LSP-X150) was used to produce a largely continuous and uniform spectrum, and 100% transmittance of the light was permitted by the quartz glass as the xenon lamp shined on the TiO 2 nanotube arrays (photoanode).The Xenon lamp was switched on after the three electrodes were connected to the potentiostat.Then, a linear sweep potentiometry (LSP) was swept from −0.5 V to 1.0 V at a scan rate of 5 mV/s.During the voltage sweeping, the corresponding photocurrent was measured.

Results and Discussion
3.1.Surface Morphology of TiO 2 Nanotube Arrays.In this part of the experiment, the effect of applied potential on the morphology of the TiO 2 nanotube arrays is discussed.The formation of TiO 2 nanotube arrays in fluorinated electrolyte was the result of three simultaneously occurring processes: (1) field-assisted oxidation of Ti metal to form TiO 2 , (2) field-assisted dissolution of Ti metal ions into electrolyte, and (3) chemical dissolution of Ti and TiO 2 in the presence of H + and F − ions.In general, it can be concluded that the electrochemical condition is an important factor for the formation of well-aligned TiO 2 nanotube arrays [25,26].
Figure 1 shows the FESEM images of surface morphology of the anodized Ti foils in electrolyte composed of 1 M of glycerol with 0.5 wt% of dissolved NH 4 F for 30 minutes, and potential was varied from 10 V to 50 V.Insets are the cross-section morphology of the oxides.From these images, the appearance of the anodic oxides on the Ti foils was dependent on the applied potential in the fluorinated electrolyte.At low anodizing potential (10 V), pits started to form on the TiO 2 surface (Figure 1(a)).Upon increasing the potential to 20 V, 30 V, and 40 V, well-aligned TiO 2 nanotubes were observed (Figures 1(b)-1(d)).It is also apparent that the diameter of the pores increased from 50 nm to 110 nm when potential was raised from 20 V to 40 V.At 50 V, the intertubes' spacing was filled, and discrete nanotubes became more interconnected to form a nanoporous-like structure (Figure 1(e)).The average diameter, length, and surface area of the tube anodized at different applied potential are summarized in Table 1.
The diameter of the tubes was found to increase with the applied potential.The main reason may be attributed to the severe electric field dissolution which accelerates the formation of pits during the early stage of anodization.These pits were etched to form larger pore in higher potential.In addition, the increase in nanotube's length is attributed to the increased driving force for ionic species (H + , F − , and O 2− ) to transport through the barrier layer at the bottom of the nanotube under electric field, which results in faster movement of the Ti/TiO 2 interface into the Ti metal.This improves the pore deepening and results in longer nanotube length.With the increase of applied potential, the rate of the fluoride (F − ) ions and the chemical dissolution at the barrier layer inside the nanotubes were increased, hence, allowing more oxygen (O 2− ) to enter and oxidize the underlying Ti.This condition disrupts the self-ordering of nanotubular structure due to the imbalance reaction between the chemical dissolution, electric field dissolution, and oxidation process [7,27,28].Besides, the anodic layer induces polarization of the Ti-O bond under higher applied potential and, hence, destroying the nanotubular structure.As a result, shorter  length of irregular nanotubular structure was observed at higher applied potential.The dimensions and uniformity of the morphology of TiO 2 nanotube arrays were a function of applied potential in anodization process.Based on this study, 30 V was found to be the optimum potential for the formation of uniform circular nanotubes with lengths approaching 1 μm.minutes.Such plot is important to explain the mechanism of TiO 2 nanotube formation.Initially, there is an abrupt decrease of current as compact oxide layer is formed, which induces the potential drop between the Ti foil and the electrolytes.The existence of 15 wt% of water in the glycerol will contribute O 2− ions.The migration of O 2− ions towards the Ti/TiO 2 interface induces further growth of the barrier layer.This is the passivation process of Ti (region I), and a schematic illustration leading to the oxide layer formation is shown in Figure 2(b).The reaction occurred is depicted in the following equation:

Formation Mechanism of TiO
In the second stage (region II), a slight increase in current was observed.As mentioned previously, high electric field across the thin layer will induce electric field dissolution.This is accompanied by breakdown of the passivated layer.Random pits are formed on the surface of the oxide due to polarization of Ti-O bond assisted by electric field across the sample.These random pits will grow into pores of various sizes and depths as the pits react with F − ions to produce [TiF 6 ] 2− complex ions as shown in (2).The chemical etching would slightly increase the current density as marked in region II.A schematic illustration regarding the dissolution of the barrier layer is exhibited in Figure 2(c) as follows: (2) In the third stage (region III), constant equilibrium was maintained with increasing anodization time whilst current density slightly reduced due to the change in pore's depth of the pits.The growth of pores is due to the competition between electrochemical oxide formation and chemical dissolution by F − ions.Thus, the nanotube structure will grow inwards [27].

EDX Analysis of the Anodized TiO 2 Nanotube Arrays.
The energy dispersive X-ray analysis (EDX) was employed to investigate the chemical stoichiometry of the Ti anodized and subsequently annealed at 400 • C under the air atmosphere.From Figure 3, it can be concluded that the atomic ratio of Ti to O was approximately 1 : 2, indicating that only Ti and O elements were present without any impurities.

XRD Analysis of the Anodized TiO 2 Nanotube Arrays.
It is well known that surface morphology change is closely related to crystal growth and phase transition.Therefore, XRD was used to investigate the effect of crystal growth and phase transition on the change in surface morphology of the TiO 2 nanotube arrays.Figure 4 depicts the XRD patterns of asanodized and annealed TiO 2 nanotubes fabricated at 30 V in glycerol containing 0.5 wt% NH 4 F. The presence of Ti phase was detected in the as-anodized sample, which represents the amorphous phase of TiO 2 (Figure 4(a)).On the other hand, the presence of anatase phase was detected in the sample subjected to annealing at 400 • C in air atmosphere (Figure 4 112), ( 200), (105), and (211) crystal planes for the anatase phase, respectively.

Photoelectrochemical Response of the TiO 2 Nanotube
Arrays.In order to evaluate the effect of surface area and crystallization of TiO 2 nanotube arrays on its photoelectrochemical response, the as-anodized and annealed samples fabricated from different applied potentials were used as photoelectrodes in the photoelectrochemical process.The photocurrent density-voltage transient was recorded under darkness and illuminated conditions, with a light intensity of approximately 800 W/m 2 .Linear sweep potentiometry (LSP) was firstly applied to investigate the photoelectrochemical behaviors of TiO 2 nanotube samples.The corresponding experimental results are presented in Figure 5.
All of the samples exhibited insignificant photocurrents less than 10 −6 A/cm 2 under dark condition (without illumination).This indicates inactive photoreaction of TiO 2 without generating photoinduced electrons under dark condition.However, the photocurrent density increased under illumination.This situation implies that TiO 2 is a good photoresponse semiconductor for the transfer and decay of photoinduced electrons through the sample [19].Based on the photocurrent density-voltage (I-V ) characteristic, the photocurrent density increased when the voltage was increased from −0.5 V to 1 V under illumination.
A maximum photocurrent density of up to 0.85 mA/cm 2 was observed from the TiO 2 nanotubes fabricated at 30 V; this is relatively high among all of the samples (Figure 5(a)).On the contrary, the as-anodized TiO 2 nanotube (without annealing) had the lowest photocurrent density compared with the annealed TiO 2 nanotube arrays, having a minimum photocurrent density as low as 0.02 mA/cm 2 (Figure 5 In Figure 5, the TiO 2 nanotubes were amorphous in the as-anodized condition, which showed poor photocurrent density.However, the photocurrent density was found to increase for annealed TiO 2 nanotube arrays.Transformation of amorphous structure into crystalline phase was observed after annealing at 400 • C in air atmosphere.The increase in photocurrent density is likely attributed to the higher photogenerated electron/hole pairs caused by higher content of anatase phase, whereas the presence of amorphous phase reduces the content of electron/hole pairs due to the existence of more recombination centers in amorphous phase [29].The photocurrent density-voltage (I-V ) characteristic shows that the higher photocurrent densities of the higher aspect ratio TiO 2 nanotube arrays are strongly dependent upon the availability of larger active surface area for photoelectrochemical reaction.Larger active surface area have better photon absorption as more photoinduced electrons have been effectively transported from TiO 2 nanotube photoanode to counterelectrode through the external circuit under illumination.In addition, photocurrent density can be enhanced by improving the uniformity of the nanotube's size due to the effectively triggered interfacial electron shift, which eventually leads to a more significant promotion of photocurrent [19,30].The highly uniform morphology of TiO 2 nanotube arrays provides a high degree of electron mobility along the tube axis.This is because the grains of uniform TiO 2 nanotubes are stretched in the tube growth direction, thus, results in better charge transfer properties compared to nonuniform TiO 2 nanoporous structure.
To estimate the quantitative correlation of light absorption to the different surface area and crystallization of the TiO 2 nanotubes, the photoconversion efficiency (η) of lightto-hydrogen energy was measured (Figure 6) and calculated based on the following equation: total power output − electrical power output light power input where j p is the photocurrent density (mA/cm 2 ), j p E 0 rev is the total power output; j p |E app | is the electrical power input, I o is the power density of the incident light (mW/cm 2 ), E 0 rev is the standard reversible potential (1.23 V/SHE), E app = E mean − E aoc , E mean is the electrode potential (versus SCE) of the working electrode where the photocurrent was measured Based on the photoconversion efficiency curve shown in Figure 6, the highest visible spectrum efficiency (about 1.4%) was obtained from the nanotube arrays fabricated at 30 V. The decrease in photocurrent efficiency was as follows: 1.30%, 0.90%, 0.65%, 0.38%, and 0.05%, which correspond to the anodic samples fabricated at 20 V, 40 V, 50 V, 10 V, and as-anodized TiO 2 nanotube arrays, respectively.The results clearly indicate that the photoelectrochemical properties are dependent on the crystal structure, surface morphology, and surface area of the nanotubular structure.The increase in number of charge carrier plays an important role on improving the efficiency of the water-splitting process.

Conclusion
The effect of applied potential on the surface morphology of TiO 2 nanotube arrays formed in glycerol electrolyte containing 0.5 wt% of NH 4 F was investigated.TiO 2 nanotubes having a tube's length of approximately 1 μm and a uniform diameter of 85 nm have been successfully synthesized by anodization of Ti at 30 V. At lower applied potential, the Ti surface consisted of oxide layer with random pits whereas, at higher applied potential, irregular structure was formed as the balance between the chemical dissolution and electric field dissolution, and oxidation process was interrupted.The higher aspect ratio of TiO 2 nanotubes arrays produced at 30 V shows higher photocurrent response among all of the samples.It can be concluded that the effective surface area and crystallization of TiO 2 nanotube arrays are important factors influencing the efficiency of photoelectrochemical performance.

Figure 1 :
Figure 1: FESEM images of TiO 2 nanotube arrays fabricated at different applied potential in 1 M of glycerol electrolyte containing 0.5 wt% NH 4 F for 30 minutes at (a) 10 V, (b) 20 V, (c) 30 V, (d) 40 V, and (e) 50 V.Insets are the cross-section morphology of the TiO 2 nanotube arrays.

2
Nanotube Arrays.In order to obtain the electrochemical information during the formation of TiO 2 nanotube arrays, a curve of current density versus time transient was plotted.Figure 2(a) shows the current density versus time curve recorded during anodization of Ti in 1 M of glycerol electrolyte containing 0.5 wt% NH 4 F with a sweep rate of 0.1 V/s to 30 V for 30

Figure 2 :
Figure 2: (a) Current density against time plot for anodized Ti foils at 30 V for 30 minutes in 1 M glycerol electrolyte containing 0.5 wt% NH 4 F, (b) schematic illustration to explain the formation of oxide layer (region I), and (c) schematic illustration to explain the dissolution of the barrier layer (region II).

Table 1 :
Dimension of the TiO 2 nanotube arrays anodized at different applied potential.