Electrochemical Surface Treatment of a β-titanium Alloy to Realize an Antibacterial Property and Bioactivity

In this study, micro-arc oxidation (MAO) was performed on a β-type titanium alloy, namely, Ti-29Nb-13Ta-4.6Zr alloy (TNTZ), to improve not only its antibacterial property but also bioactivity in body fluids. The surface oxide layer formed on TNTZ by MAO treatment in a mixture of calcium glycerophosphate, calcium acetate, and silver nitrate was characterized using surface analyses. The resulting porous oxide layer was mainly composed of titanium oxide, and it also contained calcium, phosphorus, and a small amount of silver, all of which were incorporated from the electrolyte during the treatment. The MAO-treated TNTZ showed a strong inhibition effect on anaerobic Gram-negative bacteria when the electrolyte contained more than 0.5 mM silver ions. The formation of calcium phosphate on the surface of the specimens after immersion in Hanks’ solution was evaluated to determine the bioactivity of TNTZ with sufficient antibacterial property. As a result, thick calcium phosphate layers formed on the TNTZ specimen that underwent MAO treatment, whereas no precipitate was observed on TNTZ without treatment. Thus, the MAO treatment of titanium-based alloys is confirmed to be effective in realizing both antibacterial and bioactive properties.


Introduction
Titanium (Ti) and Ti alloys are widely used in both the orthopedic and dental fields because of their good mechanical properties, high corrosion resistance, and biocompatibility.In particular, β-type Ti alloys show a relatively low Young's modulus, and they have attracted attention for the development of novel metallic biomaterials that can reduce the stress-shielding effect of bone-fixating devices and increase the shock-absorbing property of dental implants.A β-type Ti alloy, Ti-29Nb-13Ta-4.6Zr(TNTZ), which is composed of non-toxic and non-allergenic elements, has been developed.It has a high potential for use in biomedical applications because of its excellent mechanical properties and low cytotoxicity along with a low Young's modulus [1,2].In a previous study, the hard-tissue compatibility of TNTZ was evaluated in vivo [3]: The contact area of TNTZ with rabbit femur bone

MAO Treatment
The TNTZ specimen was fixed in a polytetrafluoroethylene holder with an o-ring.The exposed area in contact with the electrolyte was 39 mm 2 (7.0 mm in diameter) or 398 mm 2 (22.5 mm in diameter).Details of the working electrode are described elsewhere [25].A Type-304 stainless steel plate was used as a counter electrode.The base composition of the electrolyte for MAO treatment was 100 mM calcium glycerophosphate and 150 mM calcium acetate.In this study, 0 to 5 mM silver nitrate (AgNO 3 ) was also added to the electrolyte for incorporation of Ag into the oxide layer formed by MAO treatment.After pouring the electrolyte into the electrochemical cell, both electrodes were connected to a DC power supply (PL-650-0.1,Matsusada Precision Inc., Shiga, Japan), and then, a positive voltage, with the adjustment of the constant current density condition of 251 A¨m ´2 (9.8 mA and 100.0 mA for 39 mm 2 and 398 mm 2 specimens, respectively), was applied for 10 min.Thus, the major part of the specimen was MAO-treated with an annular untreated area, 0.5 mm from the edge.All the surface characterizations described below were performed within the MAO-treated area.

Surface Characterization and Bioactivity Evaluation
The surface morphologies of the specimens were analyzed using a scanning electron microscope (SEM, S-3400NX, Hitachi High-Technologies Corp., Tokyo, Japan).
The chemical compositions of the specimens were analyzed using X-ray fluorescence spectrometer (XRF, XGT-1000WR, HORIBA Ltd., Kyoto, Japan).Quantitative analyses of the XRF results were performed using the fundamental parameter method with standard samples.
Bioactivity was evaluated by immersion in simulated body fluids in general.A SBF solution [26] is recognized as the most common fluid for bioactive evaluation.In this study, Hanks' solution without glucose was used because of more adequate judgement on intricately-structured surface.The composition of the solution is shown in Table 1.Perfluoroalkoxy alkane bottles were washed with nitric acid and rinsed with ultra-pure water in advance.Specimens were immersed in bottles with 40 mL of Hanks' solution at 37 ˝C for 7 days.The solution was replaced with a fresh one 3.5 days after the first immersion.The amount of calcium phosphate precipitated on the specimens was determined by the following method.Reagent-grade nitric acid (60%-61%) and ultra-pure water were mixed with a volume ratio of 1:99.Specimens were immersed and stirred in the nitric acid solution to completely dissolve the precipitated calcium phosphate.Inductively coupled plasma atomic spectrometer (ICP-AES, ICPS-7000 ver.2, Shimadzu Corp., Kyoto, Japan) was used to determine the concentrations of both Ca and P ions in the solution.Prior to the measurement, standard solutions of Ca and P ions with concentrations of 0.01, 0.1, and 1 ppm were also prepared.

Bacterial Adhesion Test
To evaluate the antibacterial property of the MAO-treated specimen, the bacterial adhesion test was performed with anaerobic Gram-negative bacterium (Escherichia coli, NBRC3972, NITE, Tokyo, Japan) in accordance with the standard test method (ISO 22196:2011).The suspension medium was prepared by a five-hundred fold dilution of the nutrient broth containing 3 gL ´1 meat extract, 10 gL ´1 peptone, and 5 gL ´1 sodium chloride.The pH of the suspension medium was adjusted to be between 6.8 and 7.2 with sodium hydroxide or hydrochloric acid.The bacteria were added to the suspension medium to obtain 5 ˆ10 6 colony forming units (CFU) mL ´1.The bacterial suspension of volume 0.1 mL was dropped onto a specimen and a cover film was placed immediately.The specimens and the cover films were incubated at 35 ˝C for 24 h; then, they were washed using 9.9 mL of sterile physiological saline.The CFU of the living bacteria dispersed into the saline was determined using the culture medium sheet for E. coli (JNC Corp., Tokyo, Japan).
Three specimens were used in each treatment conditions.Statistical analyses were performed using Student's t test (two-tailed) with p < 0.05 considered statistically significant.

Surface Properties after MAO Treatment
Figure 1 shows the SEM images of TNTZ before and after MAO treatment.The untreated TNTZ showed a relatively smooth surface and the grinding grooves could be observed in the higher magnification image.On the other hand, the MAO-treated specimens showed a typical structure of MAO-treated valve metals [14,27] reported in previous studies.The morphology of the porous oxide layer with a number of small holes less than 10 µm in diameter was similar to MAO-treated TNTZ without Ag addition in the electrolyte, reported in our previous study [13].Thus, it is supposed that the presence of Ag ions in the electrolyte does not influence the formation of the porous oxide layer during MAO treatment when the concentration of AgNO 3 does not exceed 2.5 mM.However, when MAO treatment on TNTZ was tried with 5.0 mM AgNO 3 , the voltage was not increased completely and the micro-arc did not appear.This phenomenon might be derived from the change in the electrical resistance against the positive applied voltage of the oxide layer formed initially.Incorporation of excess Ag into the oxide layer altered its semiconducting property.As a result, the applied positive current was consumed only by the oxide evolution reaction that took place on the outermost surface of the oxide layer.Therefore, we set the concentration of AgNO 3 added to the electrolyte for MAO treatment as a maximum of 2.5 mM and performed the following investigation.physiological saline.The CFU of the living bacteria dispersed into the saline was determined using the culture medium sheet for E. coli (JNC Corp., Tokyo, Japan).Three specimens were used in each treatment conditions.Statistical analyses were performed using Student's t test (two-tailed) with p < 0.05 considered statistically significant.

Surface Properties after MAO Treatment
Figure 1 shows the SEM images of TNTZ before and after MAO treatment.The untreated TNTZ showed a relatively smooth surface and the grinding grooves could be observed in the higher magnification image.On the other hand, the MAO-treated specimens showed a typical structure of MAO-treated valve metals [14,27] reported in previous studies.The morphology of the porous oxide layer with a number of small holes less than 10 μm in diameter was similar to MAO-treated TNTZ without Ag addition in the electrolyte, reported in our previous study [13].Thus, it is supposed that the presence of Ag ions in the electrolyte does not influence the formation of the porous oxide layer during MAO treatment when the concentration of AgNO3 does not exceed 2.5 mM.However, when MAO treatment on TNTZ was tried with 5.0 mM AgNO3, the voltage was not increased completely and the micro-arc did not appear.This phenomenon might be derived from the change in the electrical resistance against the positive applied voltage of the oxide layer formed initially.Incorporation of excess Ag into the oxide layer altered its semiconducting property.As a result, the applied positive current was consumed only by the oxide evolution reaction that took place on the outermost surface of the oxide layer.Therefore, we set the concentration of AgNO3 added to the electrolyte for MAO treatment as a maximum of 2.5 mM and performed the following investigation.The results of the XRF analyses are summarized in Table 2.In addition to the metal elements (Ti, Nb, Ta, and Zr), relatively high concentrations of Ca and P were detected from the porous oxide layer formed by MAO treatment.Thus, these elements were incorporated into the oxide layer during the rapid growth and breakdown of the oxide during MAO treatment.This incorporation of the electrolyte elements has been reported in our previous study [13,14], as described in the introduction chapter.The concentrations of Ag in the MAO-treated samples were 0.01% and triple standard deviation, 3σ, were also 0.01%.This implies that the concentration of Ag was almost equivalent to the detection limit of the measurement and the presence of Ag was still questionable.Figure 2 shows the XRF spectra of the fluorescence X-ray energy region of Ag Kα.Small peaks could be observed in the spectrum around 22.1 keV for MAO-treated TNTZ with 0.5 and 2.5 mM Ag, whereas no peaks were observed for untreated TNTZ.Thus, we confirmed the incorporation of Ag as well as Ca and P The results of the XRF analyses are summarized in Table 2.In addition to the metal elements (Ti, Nb, Ta, and Zr), relatively high concentrations of Ca and P were detected from the porous oxide layer formed by MAO treatment.Thus, these elements were incorporated into the oxide layer during the rapid growth and breakdown of the oxide during MAO treatment.This incorporation of the electrolyte elements has been reported in our previous study [13,14], as described in the introduction chapter.The concentrations of Ag in the MAO-treated samples were 0.01% and triple standard deviation, 3σ, were also 0.01%.This implies that the concentration of Ag was almost equivalent to the detection limit of the measurement and the presence of Ag was still questionable.Figure 2 shows the XRF spectra of the fluorescence X-ray energy region of Ag Kα.Small peaks could be observed in the spectrum around 22.1 keV for MAO-treated TNTZ with 0.5 and 2.5 mM Ag, whereas no peaks were observed for untreated TNTZ.Thus, we confirmed the incorporation of Ag as well as Ca and P from the electrolyte in the oxide layer during MAO treatment.However, the amount of incorporated Ag was much lower than those of the incorporated Ca and P because of the difference in their original electrolyte concentrations: Ca was 250 mM, P was 150 mM, and Ag was 2.5 mM maximum.
Ag was much lower than those of the incorporated Ca and P because of the difference in their original electrolyte concentrations: Ca was 250 mM, P was 150 mM, and Ag was 2.5 mM maximum.Oxygen is exempted element for the equipment used in this study.

Antibacterial Property
To precisely investigate the effect of Ag concentration on the antibacterial property of TNTZ, MAO-treated specimens with lower AgNO3 concentrations and Ag-free (treated with base electrolyte composition) specimens were also prepared.Figure 3 shows the result of the bacterial adhesion test.If the number of the bacteria is unchanged, CFU mL −1 in Figure 3 counted from the mixture of the suspension and the saline can be calculated to be 5.0 x 10 4 .Hence, the number of viable bacteria on untreated TNTZ did not change during 24-h incubation.In other words, mechanical grinding does not impart antibacterial property to TNTZ, as is the case of most metallic materials.On the other hand, the specimens treated with 0.5 and 2.5 mM AgNO3 showed no viable bacterium on their surface.Thus, the incorporated Ag in the porous oxide layer was quite effective against bacterial adhesion, and it could completely kill the bacteria within 24 h even though the concentration of Ag inside the porous oxide layer was minimal.The efficacy of Ag deteriorated when the concentration of AgNO3 was lowered.It was found that the threshold AgNO3 concentration for obtaining antibacterial property was around 0.05 to 0.5 mM for TNTZ.Oxygen is exempted element for the equipment used in this study.

Antibacterial Property
To precisely investigate the effect of Ag concentration on the antibacterial property of TNTZ, MAO-treated specimens with lower AgNO 3 concentrations and Ag-free (treated with base electrolyte composition) specimens were also prepared.Figure 3 shows the result of the bacterial adhesion test.If the number of the bacteria is unchanged, CFU mL ´1 in Figure 3 counted from the mixture of the suspension and the saline can be calculated to be 5.0 ˆ10 4 .Hence, the number of viable bacteria on untreated TNTZ did not change during 24-h incubation.In other words, mechanical grinding does not impart antibacterial property to TNTZ, as is the case of most metallic materials.On the other hand, the specimens treated with 0.5 and 2.5 mM AgNO 3 showed no viable bacterium on their surface.Thus, the incorporated Ag in the porous oxide layer was quite effective against bacterial adhesion, and it could completely kill the bacteria within 24 h even though the concentration of Ag inside the porous oxide layer was minimal.The efficacy of Ag deteriorated when the concentration of AgNO 3 was lowered.It was found that the threshold AgNO 3 concentration for obtaining antibacterial property was around 0.05 to 0.5 mM for TNTZ.

Calcium Phosphate Formation in Hanks' Solution
SEM images of the specimens after immersion in Hanks' solution for seven days are shown in Figure 4.There are almost no precipitates on the surface of untreated TNTZ specimens, and bare metal surface was still exposed even after immersion for seven days in simulated body fluid.On the other hand, MAO-treated TNTZ was completely covered by the newly formed precipitates or layers with apparent thickness.The cracks in the newly formed layer were observed on the specimen treated with 2.5 mM Ag.It was confirmed that the cracks were introduced during the SEM observation process under vacuum because they were not observed just after electron beam irradiation.
Quantitative analysis was performed to determine the amount of Ca and P species formed on the specimen surface per unit area with ICP-AES.The results are summarized in Figure 5a.It is obvious that the amounts of Ca and P formed on the MAO-treated specimens were much higher than those formed on the untreated one.As shown in Figure 4, both the MAO-treated specimens were covered by newly formed precipitates.Thus, the precipitate on the MAO-treated specimens after immersion in Hanks' solution was confirmed to be spontaneously formed calcium phosphate.

Calcium Phosphate Formation in Hanks' Solution
SEM images of the specimens after immersion in Hanks' solution for seven days are shown in Figure 4.There are almost no precipitates on the surface of untreated TNTZ specimens, and bare metal surface was still exposed even after immersion for seven days in simulated body fluid.On the other hand, MAO-treated TNTZ was completely covered by the newly formed precipitates or layers with apparent thickness.The cracks in the newly formed layer were observed on the specimen treated with 2.5 mM Ag.It was confirmed that the cracks were introduced during the SEM observation process under vacuum because they were not observed just after electron beam irradiation.

Calcium Phosphate Formation in Hanks' Solution
SEM images of the specimens after immersion in Hanks' solution for seven days are shown in Figure 4.There are almost no precipitates on the surface of untreated TNTZ specimens, and bare metal surface was still exposed even after immersion for seven days in simulated body fluid.On the other hand, MAO-treated TNTZ was completely covered by the newly formed precipitates or layers with apparent thickness.The cracks in the newly formed layer were observed on the specimen treated with 2.5 mM Ag.It was confirmed that the cracks were introduced during the SEM observation process under vacuum because they were not observed just after electron beam irradiation.
Quantitative analysis was performed to determine the amount of Ca and P species formed on the specimen surface per unit area with ICP-AES.The results are summarized in Figure 5a.It is obvious that the amounts of Ca and P formed on the MAO-treated specimens were much higher than those formed on the untreated one.As shown in Figure 4, both the MAO-treated specimens were covered by newly formed precipitates.Thus, the precipitate on the MAO-treated specimens after immersion in Hanks' solution was confirmed to be spontaneously formed calcium phosphate.Quantitative analysis was performed to determine the amount of Ca and P species formed on the specimen surface per unit area with ICP-AES.The results are summarized in Figure 5a.It is obvious that the amounts of Ca and P formed on the MAO-treated specimens were much higher than those formed on the untreated one.As shown in Figure 4, both the MAO-treated specimens were covered by newly formed precipitates.Thus, the precipitate on the MAO-treated specimens after immersion in Hanks' solution was confirmed to be spontaneously formed calcium phosphate.
Ca and P were also detected from the MAO-treated specimen even before immersion in Hanks' solution (Table 2).To distinguish the Ca and P originating from the oxide layer as incorporated elements, the MAO-treated specimens before immersion in Hanks' solution were also tested in the same way.
The results are shown in Figure 5b.Certain amounts of Ca and P were detected from the MAO-treated specimens before immersion in Hanks' solution; however, the amounts were about one-fifth of those detected from the specimens after immersion.In other words, the Ca and P detected by ICP-AES from the MAO-treated specimens containing higher amount of Ag were mainly derived from the newly formed calcium phosphate on their surface.Thus, calcium phosphate formation on TNTZ was considerably accelerated on the oxide layer by MAO treatment even in the presence of Ag with sufficient concentration for eliminating bacteria.

Efficacy of Ag Incorporated on TNTZ Surface
In this study, we confirmed the following two aspects of the Ag incorporated in the resulting surface oxide layer by MAO treatment: (1) It showed excellent antibacterial property against E. coli when sufficient AgNO3 was present in the electrolyte for the treatment; and (2) it did not affect the calcium phosphate formation during immersion in simulated body fluid and maintained the superior bioactivity of the MAO-treated surface.Our previous study demonstrated the availability of MAO treatment for improvement of hard-tissue compatibility of TNTZ [13].From the results obtained in this study, we concluded that antibacterial property could be achieved in addition to hard-tissue compatibility by a simple modification of the treatment condition of the MAO process.This technique may be utilized as a novel surface treatment for development of next-generation implant devices.Nevertheless, the effect of Ag ions dissolved slightly from the oxide layer on living organs or tissues still remains unclear.It is easily supposed that an amount in excess of the appropriate metabolism level of Ag may result in unanticipated adverse effects on a living body.Pratten et al. [28] reported that a low level of Ag ions below 35 ppb was effective as an antimicrobial agent against staphylococcus bacterium.On the other hand, Hardes et al. [29] reported that Ag ions in the blood below 56.4 ppb could be considered as non-toxic in humans.The most important issue to establish a novel process for implants utilizing Ag as an antimicrobial agent may be how to control the release and accumulation rates of Ag inside the living body.From the viewpoint of these incompatible demands, the behavior of Ag ion released from the oxide layer will be investigated as the next step in this study.In addition, biological testing in vitro and in vivo are also necessary to ensure the biosafety of Ag-containing material surfaces.Ca and P were also detected from the MAO-treated specimen even before immersion in Hanks' solution (Table 2).To distinguish the Ca and P originating from the oxide layer as incorporated elements, the MAO-treated specimens before immersion in Hanks' solution were also tested in the same way.
The results are shown in Figure 5b.Certain amounts of Ca and P were detected from the MAO-treated specimens before immersion in Hanks' solution; however, the amounts were about one-fifth of those detected from the specimens after immersion.In other words, the Ca and P detected by ICP-AES from the MAO-treated specimens containing higher amount of Ag were mainly derived from the newly formed calcium phosphate on their surface.Thus, calcium phosphate formation on TNTZ was considerably accelerated on the oxide layer by MAO treatment even in the presence of Ag with sufficient concentration for eliminating bacteria.

Efficacy of Ag Incorporated on TNTZ Surface
In this study, we confirmed the following two aspects of the Ag incorporated in the resulting surface oxide layer by MAO treatment: (1) It showed excellent antibacterial property against E. coli when sufficient AgNO 3 was present in the electrolyte for the treatment; and (2) it did not affect the calcium phosphate formation during immersion in simulated body fluid and maintained the superior bioactivity of the MAO-treated surface.Our previous study demonstrated the availability of MAO treatment for improvement of hard-tissue compatibility of TNTZ [13].From the results obtained in this study, we concluded that antibacterial property could be achieved in addition to hard-tissue compatibility by a simple modification of the treatment condition of the MAO process.This technique may be utilized as a novel surface treatment for development of next-generation implant devices.Nevertheless, the effect of Ag ions dissolved slightly from the oxide layer on living organs or tissues still remains unclear.It is easily supposed that an amount in excess of the appropriate metabolism level of Ag may result in unanticipated adverse effects on a living body.Pratten et al. [28] reported that a low level of Ag ions below 35 ppb was effective as an antimicrobial agent against staphylococcus bacterium.On the other hand, Hardes et al. [29] reported that Ag ions in the blood below 56.4 ppb could be considered as non-toxic in humans.The most important issue to establish a novel process for implants utilizing Ag as an antimicrobial agent may be how to control the release and accumulation rates of Ag inside the living body.From the viewpoint of these incompatible demands, the behavior of Ag ion released from the oxide layer will be investigated as the next step in this study.In addition, biological testing in vitro and in vivo are also necessary to ensure the biosafety of Ag-containing material surfaces.

Conclusions
To realize the antibacterial property for TNTZ implants, the formation of a bioactive porous oxide layer containing Ag via MAO treatment was attempted.Incorporation of Ag into the oxide layer as well as Ca and P from the electrolyte components during MAO treatment was confirmed.When the electrolyte contained over 0.5 mM of AgNO 3 , the MAO-treated specimens showed excellent antibacterial property against E. coli.In addition, these specimens were completely covered by a thick calcium phosphate layer during immersion in Hanks' solution for 7 days, whereas only a small amount of calcium phosphate formed on untreated TNTZ.Thus, multiple biofunctions such as antibacterial property and bioactivity could be achieved, and this technique was considered to be a promising method to enhance the biocompatibility of TNTZ and other metallic biomaterials made of valve metals.

Figure 1 .
Figure 1.Scanning electron micrographs of TNTZ before and after MAO treatment.

Figure 1 .
Figure 1.Scanning electron micrographs of TNTZ before and after MAO treatment.

Figure 2 .
Figure 2. XRF spectra of (a) untreated, (b) MAO-treated with 2.5 mM Ag, and (c) MAO-treated with 0.5 mM Ag around the Ag Kα energy region.

Figure 2 .
Figure 2. XRF spectra of (a) untreated, (b) MAO-treated with 2.5 mM Ag, and (c) MAO-treated with 0.5 mM Ag around the Ag Kα energy region.

Figure 3 .
Figure 3. Result of bacterial adhesion test using E. coli on TNTZ specimens.

Figure 4 .
Figure 4. Scanning electron micrographs of specimens after immersion in Hanksʹ solution for 7 days.

Figure 3 .
Figure 3. Result of bacterial adhesion test using E. coli on TNTZ specimens.

Figure 3 .
Figure 3. Result of bacterial adhesion test using E. coli on TNTZ specimens.

Figure 4 .
Figure 4. Scanning electron micrographs of specimens after immersion in Hanksʹ solution for 7 days.Figure 4. Scanning electron micrographs of specimens after immersion in Hanks' solution for 7 days.

Figure 4 .
Figure 4. Scanning electron micrographs of specimens after immersion in Hanksʹ solution for 7 days.Figure 4. Scanning electron micrographs of specimens after immersion in Hanks' solution for 7 days.

Figure 5 .
Figure 5. Amount of calcium and phosphate ions formed on TNTZ specimen surfaces in Hanksʹ solution determined by ICP-AES.

Figure 5 .
Figure 5. Amount of calcium and phosphate ions formed on TNTZ specimen surfaces in Hanks' solution determined by ICP-AES.

Table 1 .
Composition of Hanks' solution used in this study (mol¨L ´1).

Table 2 .
Concentrations of the detected elements by XRF on TNTZ specimen surfaces with/without MAO treatments.

Table 2 .
Concentrations of the detected elements by XRF on TNTZ specimen surfaces with/without MAO treatments.