Formation of Graded TiO₂ Layer on Ti Wire by Direct Alternating Current Discharge Plasma at Atmospheric Pressure

Abstract

Although metallic materials have been used as load-bearing materials in dental and biomedical fields since they have good mechanical properties such as good ductility and strength, their aesthetic properties are inferior to those of ceramic or resin. To obtain aesthetically improved Ti dental devices, the formation of white titanium oxide on pure Ti dental devices was studied. Direct atmospheric pressure plasma (APP) treatment using alternating current was carried out on pure a Ti plate and wire. It was found that a titanium oxide layer with enough whiteness can be obtained on pure Ti wire using direct APP treatment. Although delamination of the titanium oxide layer was found after a bending test, the concept of functionally graded materials (FGMs) can overcome the shortcoming.

1. Introduction

Metallic biomaterials are currently utilized as structural materials in artificial hip joints, bone plates and screws, and dental implantology; they are used mainly in implants that replace hard tissue. Among metallic biomaterials, Ti alloys have high biocompatibility, specific strength, corrosion resistance, and exhibit the most suitable characteristics for biomedical applications [1]. The Ti-6Al-4V extra-low interstitial (ELI) alloy with α + β (hcp + bcc) phases has been the most widely used for orthopedic implant material due to its excellent combination of biocompatibility, corrosion resistance, and mechanical properties. Since the toxicity of V has been reported, V-free α + β-type Ti alloys have been also developed. In the case of β-type Ti alloys, enhanced properties such as lower modulus of elasticity, increased corrosion resistance, and improved tissue response are expected when compared with α + β-type Ti alloys. β-type Ti alloy composed of non-toxic elements with lower moduli of elasticity and higher strength was developed by Niinomi et al. based on the d-electron alloy design method [1,2,3,4,5]. The developed alloy has the chemical composition of Ti-29Nb-13Ta-4.6Zr and it is abbreviated as TNTZ. One of the applications of TNTZ is dental devices. However, TNTZ is grayish silver, which is not esthetically suitable for dental devices, as it is conspicuous among natural teeth.

Meanwhile, titanium oxide films present a range of colors. This allows them to be used in decorative applications, electronics, the automotive industry, and art—for new designs, advertising, and jewelry [6]. It is also well known that TiO₂ powder is white [7] and is used as a pigment, “titanium white” [8,9,10]. Therefore, it is expected that the white TNTZ dental devices can be fabricated by a titanium oxide coating on their surface.

In our previous studies [11,12,13,14], the formation of a titanium oxide layer coating was investigated to improve the esthetic properties of TNTZ. Various techniques can be used to obtain a titanium oxide coating; two different thermal oxidation treatments were adopted in our studies. One is high temperature oxidation at elevated temperature using a conventional furnace [11,12]. It was found that brightness and yellowness increased with increasing layer thickness. Therefore, high-temperature oxidized TNTZ will be appropriate for the orthodontic arch wire, which can reduce the psychological burden of patients during treatment. However, the issue that needs to be addressed to make the oxide coating utilizable for the fabrication of the arch wire is to suppress the embrittlement of the metallic matrix around the interface between the coating and the metallic substrate caused by high-temperature oxidation at elevated temperature [13]. This issue can be overcome by reducing the processing time and temperature, and enhancing the oxidation reaction only on the surface.

Alternatively, another thermal oxidation treatment is also performed using a remote atmospheric pressure plasma (APP) jet [13,14], where APP treatment was reported to effectively control/alter the structural properties of the surface [15]. Remote APP has many advantages such as relatively low equipment cost compared to conventional low-pressure plasma treatment due to the absence of a vacuum system and high throughput [16,17,18]. There is a wide range of applications such as surface fabrication, display technology, nanoscience, biology, medicine, and so on [16,17]. Remote APP treatment in a shorter time was found to cause whitey surface oxidation on disc- [13] and wire-shaped [14] TNTZ without causing matrix embrittlement. Moreover, enhancing surface oxidation in TNTZ could be achieved by combination of shot peening (SP) pretreatment [13,14], where SP can cause severe plastic deformation of the surface due to the transferred high collision energy from the injected hard powder, resulting in the introduction of defects [19,20,21]. Therefore, it can be said that the APP treatment is a useful method to obtain aesthetically improved TNTZ dental devices.

However, in order to use the remote APP treatment, specialized equipment is necessary such as an extensive power supply and a plasma torch [22]. If a similar white titanium oxide coating could be obtained by a simpler apparatus, this processing for white titanium oxide coating would be suitable for biomedical applications such as dental devices. Smaller arrangements with appropriate power supplies, in which one of the electrodes is a conductive material to be treated, are also able to generate non-thermal equilibrium plasma jets. Hereafter, this treatment will be denoted as direct APP treatment. The aim of this study is to examine the oxidization behavior of pure Ti using direct APP treatment. If the conditions for aesthetics can be obtained with pure Ti, good conditions can also be provided with TNTZ, since the aesthetic properties of the oxide-coated TNTZ alloy are better than those of pure Ti.

2. Materials and Methods

2.1. Sample Preparation

2.1.1. Polishing

Two different shapes of pure Ti samples were used. One was plate-shaped commercially pure Ti (CP-Ti) with 99.5% purity and a thickness of 1 mm. It was cut into about 10 × 30 × 1 mm³ in size, and polished with up to #2000 emery paper and 1 µm diamond suspension paste. The influence of surface condition on the anodization coloring behavior of Ti has been studied by Kitayama et al. [23]. It was found that uniform coloring was observed on polished surfaces, while multiple coloring was observed on pickled surfaces. After polishing, the samples were ultrasonically cleaned using acetone to remove contamination from the surface before oxidation.

Another sample shape was wire. The CP-Ti wires were used with 0.685 mm or 0.8 mm in diameter, cut to about 50 mm in length. The wire samples were polished with #200 emery paper to remove the oxide layer on the surface and then polished with #1500 emery paper. After that, the polished samples were wiped with ethanol to clean the surface.

2.1.2. Pretreatment by SP

Some of the wire samples with 0.8 mm in diameter were subjected to SP pretreatment before oxidation by direct APP treatment. This pretreatment was performed using a circulation-type sand blaster (Daiei Dental Products Co., Ltd., Osaka, Japan, Cycle blaster Jr.). The injection pressure, time, and distance between the nozzle and the peening surface were 0.5 MPa, 60 s, and approximately 20 mm, respectively. After SP pretreatment, the SPed samples were polished with #1500 emery paper to remove the rough surface caused by SP treatment and then wiped with ethanol. When SP pretreatment was performed, a deformation-induced layer with 10~20 μm in thickness was formed. Since the surface roughness induced by SP was about 3 μm, the SPed sample had the deformation-induced layer even after polishing.

2.1.3. Oxidation by Direct APP Treatment

The plate-shaped Ti samples and wire-shaped Ti samples having different diameters with and without the SP pretreatment were subjected to direct APP treatment using alternating current to obtain white titanium oxide on their surfaces. Figure 1 shows the direct APP treatment apparatus in schematics. The high voltage required for plasma generation was obtained by a neon transformer (LECIP, Gifu, Japan, 100-B-15HCS) adjusted by a slidac (Tokyo Rikosha, Saitama, Japan, RSA-10). The direct APP treatment uses only electrode and electrical power. Therefore, the treatment is simple, safe, and harmless compared to conventional techniques such as galvanic plating and electroless plating by toxic and hazardous solutions or PVD and CVD coatings which are expensive and complex because of the vacuum technique. Two different shapes of electrode were adopted. One was stainless steel pipe and the other was bar-shaped CP-Ti with a sharp tip, as shown in Figure 1. To study the influence of process parameters on the whitening of Ti, the distance between the sample and electrode, z [mm], the processing time, t [s], and the voltage, V [kV], were changed. The discharge plasma was generated by applying a high voltage between the sample and the electrode. It was expected that the oxidation of Ti samples will proceed by high temperature caused by the discharge plasma and the oxygen radicals in the plasma. The conditions of direct APP treatment for plate-shaped Ti samples and wire-samples are listed in

Table 1. Conditions of direct APP treatment for plate-shaped Ti samples.

Sample Name	Shape and Size [mm × mm × mm]	Shape of Electrode	Distance between Electrode and Sample, z [mm]	Time, t [s]	Voltage, V [kV]
Sample P1	Plate, 10 × 30 × 1	Pipe	5	60	7.5
Sample P2				120
Sample P3			3
Sample P4					9.0

and

Table 2. Conditions of direct APP treatment for wire-shaped samples.

Sample Name	Diameter of Wire [mm]	Shot Peening Pretreatment	Shape of Electrode	Distance between Electrode and Sample z [mm]	Time t [s]	Voltage V [kV]	Grayscale
Sample W1	0.685	N/A	Pipe	5	60	9.0	N/A
Sample W2	0.8
Sample W3				3		12.0	⑦
Sample W4			Bar				③
Sample W5		Yes					④–⑤

, respectively, where the treating numbers of wires are 8 and 7 for the sample W4 and sample W5, respectively.

2.2. Evaluation

2.2.1. Observation of the Samples

The surface color of the direct APP-treated samples was evaluated using lightness L* by comparing with grayscale (Sekonic, Tokyo, Japan, Exposure Profile Target II) where the highest number of L* indicates the brighter tone and the lowest number indicates the darker tone [14]. Surface and cross-sectional observation was performed using a scanning electron microscope (SEM) (JEOL, Tokyo, Japan, JSM-5900LV) attached with an electron probe micro-analyzer (EPMA) (JEOL, Tokyo, Japan, JXA-8230).

2.2.2. Three-Point Bending Test

To fabricate the orthodontic arch wire, the bending behavior of the treated samples with white titanium oxide coating is one of the most important issues to study. A three-point bending test of the wire samples was carried out on the universal testing machine (SHIMADZU, Kyoto, Japan, Autograph AZ-1/250 kN) at a crosshead speed of 2 mm/min using jigs shown in Figure 2, in which the jigs are equipped with retainers for use in an immersion test. The lower span length of the bending test, the upper jig curvature and the maximum stroke were 15 mm, 2 mm, and 5 mm, respectively. After the three-point test, the appearance of the oxide layer was observed using SEM.

3. Results and Discussion

3.1. Direct APP Treatment on Plate Sample Using Pipe-Shaped Electrode

Figure 3a–d show photographs of the plate-shaped Ti samples (sample P1 to sample P4), respectively) after direct APP treatment using the pipe-shaped electrode. As shown in Figure 3c,d, clear plasma exposure regions are observed in sample P3 and sample P4 fabricated under stronger conditions, such as shorter irradiation distance between electrode and sample, longer irradiation time and higher voltage, where the color of the plasma exposure regions was not white, but metallic blue. It is known that the interference color, which takes place between the oxide and the incident light, results in the appearance of colors on the metal surface and oxides with different thicknesses generate different colors [24]. The interference colors of the thin oxide layers can be well defined by the thickness of the film [25,26]. The color of the oxide layer with each thickness has been reported by Gaul [27]. Based on the literature, the bluish-colored oxide layer has a thickness between 0.046 μm and 0.07 μm [27]. In our previous studies, an oxide layer with the thickness of ten μm order was required to obtain the white color [12,13]. Therefore, only a very thin oxide film seems to have been formed on the Ti plate samples by direct APP treatment using a pipe-shaped electrode. It must be noted that the plasma exposure regions have neither a circle shape nor a ring shape, but a crescent-like shape, even though the shape of the electrode was a pipe. Although it is known that the formation of discharged plasma is affected by the distance between the electrode and sample, it is difficult to obtain the situation of an accurate constant distance between pipe-shaped surface and plate-shaped sample surface and to control precise irradiation distance between them. Therefore, it is concluded that the pipe-shaped electrode is not the best shape for use in dental practice, because of the uneven surfaces and morphologies of prosthetic tooth restorations. On the other hand, the above effect is smaller in the case of the wire-shaped samples. Next, direct APP treatment on wire samples using the pipe-shaped electrode were studied.

3.2. Direct APP Treatment on Wire Samples

The photographs of sample W1 and sample W2 after direct APP treatment using a pipe-shaped electrode are shown in Figure 4a,b, respectively. As already described previously, the white oxide layer did not form in the plate-shaped sample treated under the most severe conditions among the samples studied, as shown in Figure 3d. On the other hand, the wire-shaped sample W1 and sample W2 had the white oxide layer on their surface even through the treatment conditions were weaker (longer irradiation distance between electrode and sample, shorter irradiation time, and lower voltage). Sample W1 and sample W2 were fabricated under the same conditions except the diameter of the samples, as shown in Table 2. One can see that sample W2 with a larger diameter had a grayish color, while the sample W1 with a smaller diameter had a white color. This means that the thickness of the oxide layer on the sample with the larger diameter was not enough to conceal the dark color of CP-Ti. During thermal oxidation, the thickness of the oxide film varies proportionately with time and the temperature of the metal [27]. To increase the treatment temperature, direct APP treatment using a pipe-shaped electrode on sample W3 was carried out under more severe conditions, i.e., higher voltage and shorter distance between the electrode and sample. An example of sample W3 is shown in Figure 4c. It is seen that the sample W3 is white, almost as white as the sample W1. From these results, it can be said that samples with larger mass were more difficult to oxidize than the samples with smaller mass by direct APP treatment. However, despite applying the same condition, three out of six samples did not have sufficient whiteness. This may come as unstable plasma will form by the pipe-shaped electrode.

During direct APP treatment, discharge plasma occurs by the insulation air breakdown between the electrode and the sample continuously caused by alternating high voltage [28]. However, the oxide layer starts to form at the surface of samples and electrode, as soon as they are exposed to discharge plasma. As a result of the newly generated oxide layer, having worked off insulation and making the insulation breakdown difficult at the same point, discharge plasma may move around. Therefore, because the pipe-shaped electrode has many possible routes of insulation breakdown, the discharge plasma may not have concentrated at one point.

This phenomenon also explains the difference in the whitening behavior of plate samples and wire samples using pipe-shaped electrodes. The whitening of Ti samples by direct APP treatment is caused by two reasons; one is the increase in temperature by plasma irradiation and another one is active species such as oxygen radicals generated by plasma. Plate samples with a larger heat sink did not achieve a higher temperature by plasma irradiation compared to wire samples with a smaller heat sink. This is the effect of increasing temperature by plasma irradiation. As shown in Figure 4c, an oxide layer was formed on the surface of the wire sample. It is important to note here that an oxide layer was found at not only the plasma exposure position, but also the opposite position, namely the plasma nonexposed position. This is evidence of the effects of increased temperature by plasma irradiation. However, the thickness of the oxide layer at the opposite position is thinner than that at the plasma exposure position. This cannot be explained by the effects of an increase in temperature by plasma irradiation. In our previous study, the Ti alloy was coated with a white oxide layer by high-temperature oxidation [12]. The enhanced surface oxidation on the Ti alloy could be achieved by APP treatment in a shorter time without causing matrix embrittlement, as APP has the ability to produce high-density reactive species [13]. As mentioned before, it is difficult to control precise irradiation distance between a pipe-shaped surface and plate-shaped sample surface, which results in movement of discharge plasma. However, in the case of the wire sample, discharge plasma can occur at limited routes. Therefore, in the case of the wire sample, the larger amounts of active species per unit area was generated by the plasma irradiation. In this way, the better whitening behavior of wire samples is attributed to two reasons; temperature increase by plasma irradiation and active species generated by plasma.

Therefore, discharge plasma can occur at very limited routes discharge plasma in the case of bar-shaped electrode with a sharp tip. This seems to have made the difference of the success rate. Figure 4d,e show photographs of sample W4 and sample W5 after direct APP treatment using a rod-shaped electrode, where the conditions of the treatment are also listed in Table 2. The number of treated wires was 8 and 7 for sample W4 and sample W5, respectively. The white oxide layer was successfully obtained in all samples. Hasegawa et al. investigated the variation in the color tone of the Japanese tooth and reported their average values and distribution [29]. The average values of L* lightness of the central incisor, the lateral incisor, and the canine are shown in Figure 5 [29]. The lightness L* values of the Japanese tooth ranged between 73 and 63. The lightness L* values in the samples W3 to W5 were evaluated by comparing with the grayscale, and the results are also listed in Table 2. Sample W3 and sample W4 had different lightness L* values; the difference in treatment condition between them was only the shape of the electrode. In the case of the pipe-shaped electrode, the discharge plasma may not have concentrated at one point, while a concentrated route is expected in the case of the bar-shaped electrode with a sharp tip. Therefore, the thicker oxide layer can be formed using the bar-shaped electrode compared to that using the pipe-shaped electrode, which results in the difference in the L* value. From these results, the shape of the electrode seems to greatly affect the formation of the oxide layer. Since the lightness L* values of the grayscale of ③, ④, and ⑤ were 81.3, 77.5, and 74.4, respectively, it is confirmed that the sample W4 and sample W5 have enough whiteness. In the following sections, cross-sectional observation, hardness tests, and bending tests were carried out on these samples.

3.3. Cross-Sectional Observation of Wire Samples

Figure 6a,c show the cross-sectional SEM images of sample W4 and sample W5, respectively. High magnification images of these samples are shown in Figure 6b,d, respectively. The oxide layer was observed in both samples and the thickness of the oxide layer found in the sample with and without SP pretreatment was approximately 3–8 µm and 2–3 µm, respectively. The oxide layer of the sample without SP pretreatment had an uneven thickness and some parts had been peeled off. However, the oxide layer found in the sample with SP pretreatment had an almost constant thickness and no parts had been peeled off. Furthermore, the gap between the oxide layer and the matrix observed in the sample with SP pretreatment was smaller than that without SP pretreatment. From these observations, it is considered that the SP pretreatment can suppress the delamination of the oxide layer from the matrix.

The EPMA composition mapping of Ti and O in sample W5 was carried out, and the results are shown in Figure 7a and b, respectively. From this mapping, the thickness of the oxide layer formed on sample W5 was confirmed to be approximately 2.5 µm.

Line analysis was also carried out along radial direction, and the results are shown in Figure 8a. From this figure, the oxide layer with approximately 2.5 µm in thickness was again confirmed in sample W5. Additionally, an oxygen diffusion region with compositional gradient having a thickness of approximately 3 µm was also found. To study the Ti/O molecular ratio of the oxide layer, the ordinate of this figure was converted into molecular percent of O (Figure 8b), where only two dominant elements of Ti and O were taken into account, and other elements were neglected for this conversion. No compositional gradient was found in the oxide layer, and the oxide layer was identified as being be stoichiometric TiO₂.

3.4. Hardness Distribution of the Sample

Figure 9 shows the position dependence of the hardness obtained by the nanoindentation test on sample W5. The hardness of sample W5 is almost constant, approximately 200 HV. This value fits well with the reported Vickers hardness of pure Ti. Considering that the oxygen-diffused region of the samples becomes harder than the other region, the little amount of oxygen seems to have diffused inside the sample from surface by direct APP treatment. This does not contradict with the results obtained by SEM observation and EPMA analysis described in the previous section. Therefore, the embrittlement of the matrix caused by the diffusion of oxygen may be prevented by direct APP treatment.

3.5. Three-Point Bending Test

Since the fabricated white oxide layer-coated Ti wire was planned to be used as the orthodontic arch wire, the bending behavior is one of the most important issues to study. Figure 10 shows the results of the 3-point bending test of 0.8 mm Ti wire with and without the SP pretreatment and direct APP treatment. The load and stroke reported in this figure are the force measured by the load cell and the stroke of the cross head, respectively. In particular, the stroke does not include the unexpected deformation, where the samples are greatly deformed and slid along the jig during the test. From the load-stroke curves shown in Figure 10, some important information can be readable. First, compared to the bending load-stroke curves of Ti wire samples with and without SP pretreatment, it is seen that the SPed Ti wire has a higher value than expected because of hardening by SP treatment. Second, a similar comparison can be made between samples with and without direct APP treatment; i.e., the bending behavior of the direct APP-treated samples is superior to that of the untreated samples. This seems to be caused by the difference in hardness, where TiO₂ is higher than Ti.

The photographs of Ti wire, SPed Ti wire, sample W4, and sample W5 after the three- point bending tests are shown in Figure 11a–d, respectively, where the stroke was 5 mm. Enlarged images of sample W4 and sample W5 are also shown in this figure. As seen, different deflections appear for each sample, even if the given stroke was same. The samples without direct APP treatment were bent at a sharp angle around the bending point, while the direct APP-treated sample W4 and sample W5 had a gentle arc around the bending point. The measured radius of curvature in the samples without direct APP treatment was 4.5 mm, while it was 9.6 mm and 12.8 mm for the direct APP-treated sample with and without the SP pretreatment (sample W4 and sample W5), respectively. Un-APP-treated samples have homogeneous microstructure and mechanical properties along the wire longitudinal direction. In contrast, direct APP-treated samples have graded microstructure and mechanical properties along the wire longitudinal direction, since the direct APP treatment was applied only at central part. Since the oxidized part is the hardest, deformation at the central part may have prohibited. This seems to have caused the difference in the shape of samples after the three-point bending test.

Another important feature observed in direct APP-treated samples was that the oxide layer formed was partly delaminated by the bending test. This phenomenon is more significant at the inner region, namely the compression side. To study this phenomenon in more detail, SEM observations were carried out. The SEM images of the surface of sample W4 and sample W5 after the three-point bending test are shown in Figure 12a,c and b,d, respectively, where the low magnification images are shown in (a) and (b), and the high magnification images are shown in (c) and (d). From the low magnification images, a larger amount of oxide layer was found to remain in sample W4 rather than in sample W5. However, from the high magnification images, some of the remaining oxide layer in sample W4 looked like it was almost peeling off. Therefore, there is a possibly that it will be easily removed by small impulse or vibration. On the other hand, such an oxide layer was not observed in sample W5. In this way, delamination of the oxide layer was found at the curved region, and a trade-off relationship between aesthetic property and bending property was found for direct APP-treated Ti wire. In the next section, the idea of overcoming the above trade-off problem will be discussed.

3.6. Offset Direct APP Treatment

In this study, anomalous bending behavior was found for direct APP-treated samples. This may come from the graded microstructure and mechanical properties along the wire longitudinal direction. Moreover, a trade-off relationship between aesthetic property and bending property was found for direct APP-treated Ti wire. To overcome the trade-off problems, the concept of functionally graded materials (FGMs) is often applied, where the formation of gradients of chemical composition, phase distribution, or microstructure represents a now fervently pursued concept in the design of advanced engineering components [30,31]. For this purpose, offset direct APP treatment and bending tests were carried out, in which the center of the bending test was shifted from the center of plasma exposure.

The result of offset three-point bending test of direct APP-treated 0.8 mm Ti wire with the SP pretreatment is shown in Figure 13, where the treatment conditions were same as sample W5. One can find that the sample was bent at a sharp angle around the bending point, where the radius of curvature was 4.4 mm. This value is almost the same as that without direct APP treatment. It is seen that the delamination is again found for the white oxide layer-coated Ti wire. It must be noted that the delamination was not found at the arched region, but around the center of plasma exposure. In other words, arching for white oxide layer-coated Ti wire was possible without delamination when the proper direct APP treatment was applied. As seen in Figure 5, a significant difference for L* depending on the type of tooth was reported, and the central incisor had the highest L*, and L* of the canine was the lowest among the three types of tooth. Meanwhile, there were many reports on mathematical analysis of the curve of dental arch [32,33,34]. AlHarbia et al. concluded that the fourth-order polynomial function may be used as a guide to fabricate customized arch wires due to its advantage in providing a more naturally smooth curve [32]. Generally, the fourth-order polynomial equation is expressed as:

y = a₄ x⁴+ a₃ x³+ a₂ x²+ a₁ x¹+ a₀ x⁰

(1)

where the a₄, a₃, a₂, a_1, and a₀ are polynomial coefficients. The curve of the dental arch obtained by Equation (1) is shown in Figure 14a, where a₄ = 2.5 × 10⁻⁵, and an assumption of a₃ = a₂ = a₁ = a₀ = 0 is adapted for simplicity. The radius of curvature, R, is described by the following equation:

R = (1 + f′²)^3/2/|f″| = [1 + (4a₄ x³)²]^3/2/12a₄ x²

(2)

and the result is shown in Figure 14b. In Figure 14, the minimum radius for arch wire bending was 16 mm, which is larger than the 12.8 mm bending radius, at which delamination was present for sample W5. Therefore, the arch wire with an APP oxide surface finish can avoid the delamination.

Moreover, it is seen that the front part (incisor region), where the esthetic properties are required, had a large radius, while the positions with the smallest radius were near the canines or premolars, where a strong demand for whiteness does not exist. Therefore, graded plasma treatment along the longitudinal direction of the wire is the best for our purpose. Although the wire Ti samples were fixed the position against the electrode by using a simpler apparatus, plasma treatment with reciprocating motion of 30 mm was carried out in our previous study [14]. Presently, the authors studied the fabrication method of the white oxide layer-coated Ti wire with high aesthetic and mechanical properties for dental use by using the concept of FGMs. The results will be presented in a future report.

4. Conclusions

In this study, it was aimed to obtain a white titanium oxide coating on pure Ti wire using a simple direct APP treatment. The main results are summarized as follows.

(1)

Although pipe-shaped electrodes form unstable plasma, a stable discharge plasma will be formed by a bar-shaped electrode with a sharp tip;

(2)

A titanium oxide layer with enough whiteness can be obtained on pure Ti wire using a bar-shaped electrode;

(3)

The oxide layer was partly delaminated by the bending test. Delamination after an offset three-point bending test was not found at the arched region but around the center of plasma exposure;

(4)

Arching for a white oxide layer-coated Ti wire is possible without delamination when the proper direct APP treatment is performed.