Efecto de la oxidación térmica en nanoestructuras de TiO2 sobre la nanodureza y resistencia a la corrosión

This article aimed to analyze the effect of the thermal oxidation in the corrosion resistance and the hardness properties of TiO2 nanostructures obtained by the anodizing process in the HF/H3PO4 solution. TiO2 nanostructures on Ti6Al4V obtained by anodizing processes were subjected to thermal oxidation (TO) treatments over a temperature range from 500 oC to 620 oC for 2 hours. Surface morphology was evaluated by using scanning electron microscopy; the hardness properties of TiO2 nanostructures were obtained by Nanoindentation measurements using a Berkovich probe with a tip radius of 150 mm. The corrosion behavior of the samples was studied using potentiodynamic polarization and electrochemical impedance spectroscopy (EIS). The results showed that TiO2 nanostructures, modified by thermal oxidation, increased the surface properties of hardness and corrosion resistance, compared to the substrate, maintaining its mixed or tubular structure. On the other hand, a transformation of nanotubes to nanopores after 600oC was evidenced, generating significant changes in the mechanical properties of these structures.


INTRODUCTION
The use of titanium alloys in the medical field has responded very well due to its properties of biocompatibility, corrosion resistance, low elastic modulus compared to stainless steels or cobaltbased alloys [1][2][3][4]. However, it has been found that chloride ions present in the different extracellular of body fluids such as blood and interstitial fluids generate significant corrosion in this type of material [5][6][7][8]. Processes of superficial modification, such as anodizing and thermal oxidation, have provided a solution to tribological and corrosion problems present in this type of alloys [9][10][11]. Anodizing is a method of superficial modification that produces different types of oxide films in the metal, either barrier-type films or nanostructured films [12][13]. Some the electrolytes used for the formation of these structures are the different diluted solutions of sulfuric acid, phosphoric acid, acetic acid, among others [14]. For the formation of nanotubes, it is essential that the electrolyte used in the anodizing process of titanium contains fluoride ions [15] which allow generating the pits of the oxide layer in formation, thus, facilitating the kinetic growth process of TiO 2 nanotubes [16]. In addition, it is necessary to apply a constant potential using a DC power supply to ensure the homogeneity of nanotubes [17][18].
Self-organized TiO 2 nanotubes films, generated from anodizing processes, have allowed a greater interaction with cell tissue due to their nature, compared with barrier-type films. However, it is known that the mechanical properties of these structures should be analyzed and studied depending on the type of application used, making necessary the development of treatments that allow improving their mechanical properties at a nanometric level [19][20][21][22]. According to some researches, the development of anodized films allows a better bioactive surface behavior with increased cellular adhesion; however, the mechanical resistance or the tribological behavior of these films is limited [23][24][25][26]. TiO 2 exists in three phases: anatase, brookite and rutile; however, anatase and rutile are the phases of major study in most of the biomedical applications [27], particularly for orthopedic and osteosynthesis applications due to the high surface reactivity, which in turn depends on the crystallinity [10,21,28].
Thermal oxidation of titanium and its alloys has already been investigated for different biomedical applications with the aim of improving its hardness and wear resistance [10]. The passive behavior of the oxide film generated by thermal oxidation plays an important rol for its suitability as an implant material. Likewise, it reduces the kinetics of the corrosion process by blocking the diffusion of the aggressor ions present in the body fluids of the surface of the metal [29]. Corrosion behavior is one of the most important characteristics for the application of implants; therefore, the corrosion resistance of nanotubes should be investigated before being used in clinics [30].
The manufacture of self-organized layers of oxide nanotubes as a suitable surface treatment can improve the mechanical characteristics and corrosion resistance of titanium alloys. In terms of corrosion resistance, the layers of TiO 2 nanotubes on titanium have a better resistance in the simulated biofluids compared to Ti without being modified [31][32][33]. The growth of nanostructures in titanium alloys Ti6Al4V, as well as the development of modification treatments, has been a subject of great interest in terms of their adhesion properties, corrosion resistance, biocompatibility, and wear resistance [14]. However, the study of mechanical properties of these nanostructures remains being a subject of study due to the different variables that might be modified to improve their development (type of electrolyte, time, type of surface modification, voltage, current density, and structure obtained). Some authors have reported hardness values that vary from 0.7GPa to 12GPa depending on the load and assay condition [21][22][34][35].
Therefore, the objective of this article is to analyze the effect of thermal oxidation on corrosion resistance and hardness properties of TiO 2 nanostructures obtained by anodizing process in the HF/H 3 PO 4 solution.

MATERIALS AND METHODS
In this investigation, Ti6Al4V ELI alloy (ASTM F136) discs with 12.7 mm diameter and 3 mm of thickness were used. Before subjecting to anodizing and thermally oxidation process, the samples were ground with different grades of SiC papers (grit size 240 up to 1500) and then polished with diamond pastes of 3 and 1 µm average particle size. Finally, the samples were ultrasonically cleaned in ethanol for 10 minutes and air-dried at room temperature. Anodization (AN) was carried out in an electrochemical cell with a two-electrode configuration [36]. Ti6Al4V ELI discs were used as anode and stainless-steel plate as cathode. The anodization was conducted at room temperature. A constant potential of 14V was applied using a DC power supply. The electrolytic solution of 1 M H 3 PO 4 + 0.2% v/v HF was used to obtain nanotube morphology. The distance between anode and cathode was maintained at a constant distance of 2 cm. All of samples were anodized for about 1 hour and then rinsed with distilled water and air-dried at room temperature.
Based on a literature review and taking into account the studies of Sarraf, Zalnezhad, Bushroa and Hamouda [37], Deen, Farooq, Raza and Haider [32] and Bessauer et al., the relationship between variables for the anodization on the Ti6Al4V alloy was determined. After anodizing, thermal oxidation (TO) treatments were carried out in a muffle furnace over range temperatures from 500 ºC to 620 ºC for 2 hours, and subsequently furnace cooled. The heating rate was approximately 5 ºC/min. The morphology of the samples was observed by scanning electron microscopy (SEM), using a scanning electron microscope model Quanta FEG 650 equipped with X-ray dispersive spectroscopy EDX (Apollo X). Data of nanotube diameters for different samples, were obtained using image J software and expressed as mean ± standard deviation (SD)and were statistically analyzed by Statgraphics software; values of p < 0.05 were considered significant.
Morphological exams of the corresponding surfaces were performed as part of the microstructural characterization, through optical profilometry and using an Ultra Precision 3D Profilometer Talysurf CCI. Additionally, nanoindentation test were carried out using the triboindenter TI 950 (Hysitron INC.) An array of 10X10 indentations was performed by applying a force of 500 mN using a Berkovich type indenter. Electrochemical studies were conducted using a Potentiostat/Galvanostat Autolab PSGTAT 302 with impedance module FRA2 (frequency response analyzer). A microcell was used to evaluate the different regions of each sample, with an area of 0.031 cm 2 . Three tests were performed for each condition. The conventional three electrode cell was used to carry out the electrochemical studies. The tests were conducted in Ringer's solution (8.4 g/L NaCl, 0.302 g/L KCl, 0.298 g/L CaCl 2 .2H 2 O, 0.15 g/L NaHCO 3 ) which simulates human body fluid at 25 ± 2 ºC, with pH of 7.3 ± 1.0, Ti6Al4V was kept as the working electrode; Pt wire and Ag/ AgCl (KCl sat) were used as counter and reference electrodes, respectively. Impedance spectroscopy (EIS) measurement was carried out in the frequency range of 10 5 Hz to 10 -2 Hz with an excitation voltage of 10 mV (peak-to-peak). The measurements were initiated after 2 hours of open circuit potential. After EIS measurements, potentiodynamic polarization studies were carried out from -250 mV to +1200 mV (versus open circuit potential) at a scan rate of 1 mVs -1 .

Morphology of nanostructures
The Figure 1 shows the morphology of the TiO 2 nanostructures obtained by combined processes of anodizing and thermal oxidation at temperatures from 500 ºC to 620 ºC. Figure 1a shows the formation of TiO 2 nanostructures to 14V obtained by electrolytic anodizing; in the surface a uniform and orderly distribution of nanotubes is observed with average diameters of 66.9 ± 7.6 nm, similar data were found by Mohan, Anandan and Rajendran [39].
Figures 1b, 1c, 1d, 1e and 1f show TiO 2 nanostructures of the anodized and thermally oxidized samples at 500, 540, 580, 600 and 620 ºC respectively. Figures 1b, 1c and 1d exhibit a self-aligned and randomly disordered TiO 2 nanotubes with a resulting average diameter of 62.59 nm and 61.52 nm, respectively. A change of morphology with a deformation and enclosure of the nanotubes was seen, with a densification effect of the nanotube film; effect that is generated by submitting the samples to processes of thermal oxidation, thus, improving the mechanical properties as expressed by Sarraf, Zalnezhad, Bushroa and Hamouda [37]. In addition, it was observed that small crystalline structures were formed from 600 ºC, covering the nanostructures, which was more evident in the thermally oxidized sample at 620 ºC, as shown in previous studies [25]. Table 1 shows the diameters of the nanostructures obtained. It shows a diameter growth from 500 ºC to 560 ºC, however after 580 ºC a slight decrease in the nanotube diameter was observed after thermal oxidation, as indicated by Munirathinam and Neelakantan [21], which can be attributed to the sintering occurred by the crystallization effect. A transformation of nanotubes to nanopores occurred after 600 ºC, the transformation that is also evidenced by other authors such as Sarraf, Zalnezhad, Bushroa and Hamouda [37] and Mohan, Anandan and Rajendran [39], where the increase in temperature caused significant changes in microstructures, making the nanotubes to collapse completely and adopt compact film morphology of TiO 2 (as shown in Figure 1e and 1f). The structure on the nanopores is recognized because there is no intertubular space between them, as evidenced by nanotubes.
Bauer, Park, Von Der Mark and Schmuki [40] using 1M concentrations of H 3 PO 4 with different concentrations of HF, found self-organized nanostructures as presented in this research, however with a smaller nanotube diameter due to the absence of heat treatment.  and Rsm (mean spacing of the roughness profile).

Roughness measurements of nanostructures
(See Table 2). Figure 2a presents an orderly and uniform surface of the substrate; this also shows the change of surface as the temperature increased as demonstrated in Figures 2b, 2c, 2d.
The increase in the temperature allowed observing an increment of the roughness, represented in Ra values from 0.14 µm to 0.22 µm for 500 ºC and 620 ºC respectively, compared to Ra values of 0.03 µm of the substrate. According to Biswas, Manna, Chatterjee, Bhattacharyya and Majumdar [41], this increase in roughness could favor cellular adhesion in relation to very smooth surfaces, possibly due to the increase of water in the surface, as was reported in previous studies [36]. Guéhennec, Soueidan, Layrolle and Amouriq [42] found that the roughness increment in the surface favors the biomechanical stability and the osseous anchorage of the implants subjected to superficial treatments. Rp values remained constant between 500 ºC and 580 ºC, but with a lower temperature the values increased from 0.4 ± 0.02 µm to 0.7 µm by the transformations present and the formation of titanium oxides.
In general, the values of Rku (kurtosis) were lower than 3, showing that the distribution of the peaks had a Gaussian tendency, similar data were found by Gadelmawla, Koura, Maksoud, Elewa and Soliman [43]. The average spacing between the peaks increased from 560 ºC, Rsm, compared to the one observed in low temperature samples; meanwhile, Rv value  of anodized and thermally oxidized samples was greater than the Rv value of substrate. Figure 3a shows the cFFomparison of the elastic module with the hardness for the substrate and thermally modified samples (nanostructures TiO 2 film) between 500 ºC and 620 ºC, and figure 3b corresponds to the contact depth and maximum penetration depth depending on the temperature.

NANOMECHANIC PROPERTIES
It was found that, with the increase in temperature, hardness and elastic modulus values of the nanostructures TiO 2 film, were increasing from 1.8 ± 0.7 to 3.2 ± 0.8 for 500º and 620 ºC respectively. Similarly, elastic modulus values increased from 94GPa to 120GPa for the same temperatures. Hardness and elastic modulus values for nanostructures TiO 2 film were lower than those of the substrate (5,2 Gpa and 122Gpa), similar behavior that was found by other authors [22,31].
The lowest contact depth and maximum depth values for the substrate and the samples treated at 600 and 620 ºC correspond to the maximum hardness values 5.2, 3.9 and 3.1 GPa respectively. On the other hand, it was found that the elastic modulus values of the modified surfaces did not exceed the value of the elastic modulus of the substrate, which reduced the efforts between the substrate and the coating and improved the behavior of the surfaces.

Potentiodynamic tests
Potentiodynamic curves for thermally oxidized samples at different temperatures and substrate  of Ti6Al4V are shown in Figure 4. Polarization curves, in all cases, went from an active behavior to a passive behavior, which is typical of this type of materials [30,38,42].
The highest current density was presented in the substrate showing a higher dissolution rate; On the other hand, the nanostructures of TiO 2 exhibited a passive behavior with low current densities. It was found that the samples subjected to 540 ºC, 560 ºC and 600 ºC presented the lowest corrosion rate (Vcorr), however all the TiO 2 nanostructures showed a lower corrosion rate compared to the substrate, due to the applied heat treatment. In addition, the nanostructures showed a good performance in media containing Clions, because no pits were found on the surface, during and after the potentiodynamic tests performed. The corrosion potentials (Ecorr) of the TiO 2 nanostructures presented more positive potentials compared to the substrate and anodized material (Table 3).
Electrochemical Impedance Spectroscopy Figure 5 shows the Bode diagrams a) phase angle and b) impedance against frequency, which had variations in relation to the substrate and the thermally oxidized and anodized samples. The phase angle diagram in Figure 5a presents in low frequency areas (0.01 Hz-100 Hz) a phase angle that went from 20º to 70º for the substrate, compared to the anodized material that showed a reduction in the rank from 55º to 45º. At high frequencies, the substrate presented a lower angle compared to the anodized and thermally oxidized material. Nanostructured TiO 2 sample at 620 ºC was the one that presented the highest angles at high frequencies. On the other hand, the maximum phase angle was presented in the substrate, typical aspect of passive films formed on the surface [39,45]. EIS response of the nanostructured TiO 2 sample to 620 ºC, showed two time constants: one at low frequency (0.01 Hz) and another at high frequency (1000Hz) with possible double layer formation; the first generated by the initial anodizing layer and the second by the thermal oxidation process [46]. Figure 5b presents the impedance graph for the different studied temperatures, indicating that the nanostructured samples showed high impedance at high frequencies compared to the substrate. In the case of the substrate, a section of horizontal straight line was found at high frequency, consequence of the resistance to the electrolyte, which is a very similar behavior for thermally oxidized samples above 580 ºC. In addition, samples anodized and thermally oxidized at 540, 560 and 620 ºC show  higher values of impedance at high frequencies, caused by a difference in the values of resistance to the solution, this can be due to the absorption of species by the nanostructures, which manages to generate an increase in the electrical resistance of the Ringer's solution. This species absorption indicates that the nanostructured TiO 2 surfaces are highly reactive compared to the untreated alloy [36].
In relation to nanostructured and thermally oxidized samples, it was found that the best treated sample condition was the one developed at 620 ºC, where high impedance was shown at low frequencies compared with other thermally oxidized samples. Y. Liu and J.Chen et al., reported that, in some cases, modified samples showed inflection points that demonstrated variants of time, making possible a mass transfer and generating a different behavior both for the base as for the walls of the nanotube, presenting a diffusion process in the film [47][48]. Figure 6 shows the equivalent circuits used to model the behavior of the substrate, anodized samples and the anodized and thermally oxidized samples.
In the model applied to anodized and thermally oxidized samples, R1 was the resistance to the solution, R2 and R3 were the resistance of the internal barrier and the outer layer; W was the Warburg impedance and CPE1 and CPE2 were the constant phase elements. The resistance of the internal barrier layer corresponded to the polarization resistance of the film, while the resistance of the outer layer corresponded to the load transfer resistance of the film. CPE1-P and CPE2-P elements indicated the TiO 2 solution-nanotubes interface and TiO 2 -substrate nanotubes, respectively. The values obtained for CPE1-P indicated a highly capacitive behavior associated with the opposition to the load transfer generated by the TiO 2 nanostructures and the titanium oxide formed naturally. On the other hand, the values for CPE2-P, demonstrated a resistivecapacitive behavior, which can be associated with the morphology of that interphase. Table 4 shows the values of the equivalent circuit elements for the different samples. Chi squared (χ) parameter was used for the curve parameter, where values of 10 -3 and 10 -4 were found.
Thermal oxidation processes in untreated titanium alloys (Ti6Al4V) generated crystalline structures of anatase and rutile, especially after 600 ºC, where a transformation of amorphous to crystalline structures was evidenced, which increased the surface hardness of the coating, as it was shown in previous works [49]. However, it has been found that nanotubular structures, subjected to thermal oxidation processes, not only favor cell growth, according to previous studies [50][51], but also improve the mechanical properties (hardness) in this type of nanostructures, generating alternatives for possible tribological applications.  On the other hand, in this investigation, thermal oxidation processes below 620 ºC and applied to TiO 2 nanostructures, increased surface roughness, reduced the corrosion rate and maintained a tubular and organized structure, which favors the use for different applications in biomedical implants.

CONCLUSIONS
SEM images showed that nanotubular arrays were completely transformed into nanopores after 600 ºC and that the coarsening of the TiO 2 nanotube arrays occurred at temperatures higher than 580 ºC The thermally oxidized and anodized samples showed a passive behavior with low current densities compared to the substrate, being the samples AN+TO600 ºC and 620 ºC the ones with the best electrochemical behavior.
The oxidized samples at 620 ºC had, in the equivalent circuits, two-time constants with possible double-layer formation, an internal barrier and an external poroustype of the oxide film, indicating a resistive-capacitive behavior, both in the base as in the walls, presenting a diffusion process in the film formed. Additionally, there was an increment in the roughness (Ra) that was favored by the increase in temperature, making structures friendlier with the cell tissue.

ACKNOWLEGMENT
This project was financed by Universidad Autónoma del Caribe under the Project # 003-VI Call for Projects. Likewise, the authors are thankful by the support provided by the laboratory of Universidad de São Paulo.