Fabrication and Performance of ZnO Doped Tantalum Oxide Multilayer Composite Coatings on Ti6Al4V for Orthopedic Application

Ti6Al4V titanium alloy has been widely used as medical implant material in orthopedic surgery, and one of the obstacles preventing it from wide use is toxic metal ions release and bacterial implant infection. In this paper, in order to improve corrosion resistance and antibacterial performance of Ti6Al4V titanium alloy, ZnO doped tantalum oxide (TaxOy) multilayer composite coating ZnO-TaxOy/TaxOy/TaxOy-TiO2/TiO2/Ti (ZnO-TaxOy) was deposited by magnetron sputtering at room temperature. As a comparison, monolayer TaxOy coating was prepared on the surface of Ti6Al4V alloy. The morphology and phase composition of the coatings were investigated by field emission scanning electron microscopy (FE-SEM) and X-ray diffraction (XRD), the elemental chemical states of coating surfaces were investigated by X-ray photoelectron spectroscope (XPS). The adhesion strength and corrosion resistance of the coatings were examined by micro-scratch tester and electrochemical workstations, respectively. The results show that the adhesion strength of multilayer ZnO-TaxOy coating is 16.37 times higher than that of single-layer TaxOy coating. The ZnO-TaxOy composite coating has higher corrosion potential and lower corrosion current density than that of TaxOy coating, showing better corrosion inhibition. Furthermore, antibacterial test revealed that multilayer ZnO-TaxOy coating has a much better antibacterial performance by contrast.


Introduction
Ti6A14V titanium alloy, as a kind of ideal biomedical material, has been widely used in tooth implant, bone trauma products, artificial joints and other hard tissue substitutes or prostheses due to its excellent properties, such as biocompatibility, acceptable corrosion resistance, and comprehensive mechanical properties [1][2][3]. However, with the increase of service time, the release of Al and V ions owing to the interaction between Ti6Al4V and human body fluid may cause some toxic-relevant side effects (such as poisoning, allergies, and carcinogenesis), which eventually lead to failure of the implant [4][5][6][7]. Meanwhile, frequent bacterial infections during and after implantation can also lead to implant failure. Practically, 20% of implant failure are reported to be attributed to implant-related bacterial infections [8][9][10]. Therefore, improving the corrosion resistance and antibacterial properties  Figure 1 shows the structures of coated samples. As can be seen from the figure, ZnO-Ta x O y multilayer composite coatings consists of five layers, of which the 1st layer to 3rd layer are Ti metal layer, TiO 2 ceramic layer and TiO 2 -Ta x O y ceramic mixed layer, respectively. Those three layers were used as the intermediate transition layers and elements penetrated at two adjacent layers interface where many elements exist together, which would be helpful to alleviate the interfacial stress and improve coating adhesion [46][47][48][49]. The 4th layer of Ta x O y is mainly provided for improving the corrosion resistance of the Ti6Al4V titanium alloy, and the 5th layer (top layer) is Ta x O y doped with ZnO which acts as an antibacterial agent. Ti6Al4V (BAOTI Group Co., Ltd., Baoji, Shaanxi, China) substrate of a thickness 0.6mm was cut to the size of 10 mm × 10 mm, and its nominal composition in wt% is: Al, 6.8; V, 4.5; Fe, 0.3; O, 0.2; C, 0.1; N, 0.05; H, 0.015, and the balance, Ti. The substrates were polished using SiC emery paper with a grit size of 300, 600, 1000, and 2000, respectively. Thereafter, the substrates were ultrasonically cleaned in acetone and alcohol, each for 15 min, then dried in a pre-vacuum dryer (ZKT-6050, Shanghai Hasuc Instrument Manufacture Co., Ltd., Shanghai, China). A high vacuum magnetron sputtering system (JCP-450, Beijing Technol Science Co., Ltd., Beijing, China) was used for plasma cleaning and depositing coating. Prior to deposition, the substrate and targets were cleaned in turn by plasma cleaning with the following process parameters: background pressure of 1.0 × 10 −3 Pa, argon flow of 20 sccm, cleaning power of 200 W and the cleaning time of 20 min. Afterward, ZnO-Ta x O y multilayer composite coating was deposited on Ti6Al4V surfaces by sequentially depositing Ti film, TiO 2 film, Ta x O y -TiO 2 film, Ta x O y , and ZnO-Ta x O y film. Ta and Ti targets (Zhongnuo New Material Technology Co., Ltd. Beijing, China) had a purity of 99.99% and a size of ϕ 75 × 5 mm. During deposition, argon and oxygen were used as sputtering gas and reaction gas, respectively, both with the purity of 99.99%. The distance between the substrate and target was 75 mm, and the base pressure was 1 × 10 −3 Pa. The sputtering mode, deposition parameters of coatings are shown in Table 1.

Characterization of Coatings
The phase composition of the coatings was analyzed by X-ray diffraction (XRD, Ultima IV, Rigaku Corporation, Tokyo, Japan) with Cu Kα radiation. The surface and cross-section morphologies of coated samples was investigated by a field emission scanning electron microscopy

Characterization of Coatings
The phase composition of the coatings was analyzed by X-ray diffraction (XRD, Ultima IV, Rigaku Corporation, Tokyo, Japan) with Cu K α radiation. The surface and cross-section morphologies of coated samples was investigated by a field emission scanning electron microscopy (FESEM, SU8000, Hitachi Group, Tokyo, Japan) equipped with energy dispersive X-ray spectroscopy (EDS). For taking a cross-section measurement, the coating samples were suspended in the epoxy resin, cured at 25 • C for 24 h, and then cut into 4 mm × 4 mm × 4 mm. Next, the cross-section of coating sample was ground with 600 to 2000 grit SiC paper. The elemental compositions and chemical states of coating surfaces were investigated by X-ray photoelectron spectroscope (XPS, EscaLab 250Xi, Thermo Fisher Scientific Inc., Waltham, MA, US), equipped with monochromatic Al K α radiation (6 mA, 12 kV and 1486.68 eV). To remove surface contaminants, it is indispensable to sputter the surface using 2 kV Ar + with raster area of 4 mm 2 for 20 s. The scanning range was from 5 • to 90 • , with scan speed of 2 • /min.

Scratch Test
Scratch tests was carried to estimate the adhesion strength of coated samples with MFT-4000 scratch tester (MFT-4000, Lanzhou Institute of Chemical Physics of the Chinese Academy of Sciences, Lanzhou, China). During scratch test, a conical diamond indenter (angle 120 • and radius 200 µm) subjected to a progressive normal load from 0.1 N to 50 N moved across the surface of the coated samples with loading rate of 50 N/min. simultaneously, the variation of friction force, normal force and acoustical signal in terms of scratch distance were recorded continuously. Two critical loads, i.e., L cl and L c2 in the scratch test were defined for the failure of the coatings. L cl was the first critical load, which belongs to cohesive failure characterized by local coating puncture. With the gradual increase of normal load, the diamond indenter is passed through the coating into the substrate, cracks appear in the bottom and sides of the scratch, and the coating near the scratch appears to flake off, eventually the coatings had completely fallen off from the substrate along the scratch path. At this time, the friction force is abruptly changed, and an inflection point can be observed on the friction curve. This load force was recorded as the second critical load, L c2 , which can be used to indicate the adhesion strength of the coatings. In order to determine the failure modes of the coating and associate them with the load at which they occurred, the scratch images were observed using an optical microscope (Seika Machinery, Inc. a subsidiary of Seika Corporation, KH-7700, Tokyo, Japan).

Contact Angle Measurement
The contact angles (CAs) measurements were performed to estimate the surface wettability at 20 • C and the ambient humidity (50%), using a contact angle goniometer (JC20001, Shanghai Zhongchen Digital Technology Co., Ltd., Shanghai, China) by the sessile drop method. The droplets were laid onto the sample surfaces by a standard micro-syringe, and the droplet images were captured using a camera. To obtain accurate water contact angle data, the measurements were repeated at five different locations of the samples.

Electrochemical Measurements
The electrochemical characteristics of the samples were detected with electrochemical workstation (SP-15/20A, Bio-Logic Science Instruments, Seyssinet-Pariset, France) with a conventional three-electrode system in which platinum sheet was used as the counter electrode (CE), a saturated Ag/AgCl electrode as the reference electrode (RE) and the samples as the working electrode (WE). The electrolyte was simulated body fluid (SBF) [50], at pH 7.4. One cm 2 of the sample surface was exposed to the SBF solution. Potentiodynamic polarization curves were conducted in the range of −0.3~2.0 V with a scanning rate of 1 mV/s [51,52]. The corrosion potential (E corr ) and corrosion current density (I corr ) were calculated by the Tafel extrapolation method. All the experiments were repeated three times.

Antibacterial Experiment
The plate counting method is commonly used to quantitatively evaluate the antimicrobial properties of materials [53,54]. In this study, the bacterial strain of S. aureus (ATCC6538, Guangzhou Institute of Microbiololgy, Guangzhou, China), which is a classic strains of implant associated infections [53], was employed to analyze the antibacterial properties of coated samples in vitro by the plate counting method. Prior to the antibacterial test, all samples were placed in sealed test tubes and sterilized at 121 • C, 0.1 MPa for 30 min using a fully automatic autoclave. A series of bacterial suspension with concentrations of 10 5 -10 8 CFU/ml using 0.9% NaCl solution were prepared for experiments. Then 60 µL of the bacterial suspension with a concentration of 10 7 CFU/mL was put onto the sample surface, and cultured in a shaking incubator at 37 • C for 24 h. Thereafter, the sample with the suspension was placed into a sterile glass tube with 4 mL 0.9% NaCl solution and uniformly mixed with a vortex mixer. The suspension was then diluted by 100 times with 0.9% NaCl solution and mixed thoroughly again. The 200 µL of the diluted 100-fold bacterial solution was evenly spread on the counting agar plates using a screw inoculator, and then incubated in a shaking incubator at 37 • C for 24 h. Then these plates were photographed and the bacterial colonies were counted by an automatic colony imaging analysis system (Sphere Flash, Barcelona, Spain). Antimicrobial ratio (K) of the specimens was calculated using the formula [54]: where A and B was the average number of bacterial colonies (CFU/mL) for uncoated Ti6Al4V as the control group and for the coating sample, respectively. The obtained value represented an average of three test data

Microstructural Characterization of the Coatings
The surface morphology of the uncoated Ti6Al4V and coated Ti6Al4V are revealed in Figure 2.   Figure 2f. In these figures, some directional grooves with different depths and widths are clearly visible, which was formed by the polishing grit. Although the groove was gradually covered by the coating with the increase of the thickness of the deposited coatings, it can still be found that there are grain boundaries or gaps with the same direction. Compared with Ta x O y coatings, Ta x O y -TiO 2 coating has less directional grooves, better surface smoothness, indicating a better coating quality. It is suggested that a small size abrasive should be used to polish the surface of the substrate to obtain high quality coating surface with low roughness. Figure 3 displays the cross-section images of coated samples. The coating thicknesses of the Ta x O y and ZnO-Ta x O y samples were 3.97 µm ( Figure 3a) and 5.2 µm (Figure 3b), respectively. As can be seen, no obvious discontinuity or crack was detected between the coating and the Ti6Al4V substrate in ZnO-Ta x O y coatings samples, indicating that the multilayer ZnO-Ta x O y coatings are well bonded to the substrate. However, some brittle cracks occur in the monolayer Ta x O y coating samples, which suggests its bonding strength to Ti6Al4V substrate is lower than that of ZnO-Ta x O y coating sample. should be used to polish the surface of the substrate to obtain high quality coating surface with low roughness.   (Figure 3b), respectively. As can be seen, no obvious discontinuity or crack was detected between the coating and the Ti6Al4V substrate in ZnO-TaxOy coatings samples, indicating that the multilayer ZnO-TaxOy coatings are well bonded to the substrate. However, some brittle cracks occur in the monolayer TaxOy coating samples, which suggests its bonding strength to Ti6Al4V substrate is lower than that of ZnO-TaxOy coating sample. The XRD spectra of uncoated and coated Ti6Al4V samples are shown in Figure 4. Compared with the XRD pattern of bare Ti-6Al-4V substrate, no characteristic peaks can be found both for Ta x O y and ZnO-Ta x O y films in Figure 4a, indicating an amorphous structure. The results are consistent with previous studies [20,46,47]. The reason for this amorphous nature of the deposited film may be deposition temperature [55], sputtering power [56], flow ratio of oxygen to argon [57], element doping [57], etc. The local magnification pattern shows that the diffraction pattern of Ta x O y films is composed of diffuse-scattering curves and two humps appears in the 2θ range of 20 • -65 • in Figure 4b. The positions of humps coincide with the peak position of several possible tantalum oxides, such as TaO, TaO 2 and Ta 2 O 5 based on data in JCPDS card No.78-0724, 19-1296 and 25-0922, which signifying that the deposited Ta x O y films may be one and more phases [47]. Besides, compared with Ta x O y films, the diffraction curve of ZnO-Ta x O y films becomes flat with broader hump, implying the lower crystallinity of ZnO-Ta x O y films. The degradation of crystallinity may be caused by the incorporation of ZnO [58]. In addition, according to the data in JCPDS card NO.16-1451, the diffraction peak position of ZnO is around 31 • of 2θ, and it is also in the hump area. The specific stoichiometric composition of ZnO-Ta x O y films needs further analysis by XPS technique. The XRD spectra of uncoated and coated Ti6Al4V samples are shown in Figure 4. Compared with the XRD pattern of bare Ti-6Al-4V substrate, no characteristic peaks can be found both for TaxOy and ZnO-TaxOy films in Figure 4a, indicating an amorphous structure. The results are consistent with previous studies [20,[46][47]. The reason for this amorphous nature of the deposited film may be deposition temperature [55], sputtering power [56], flow ratio of oxygen to argon [57], element doping [57], etc. The local magnification pattern shows that the diffraction pattern of TaxOy films is composed of diffuse-scattering curves and two humps appears in the 2θ range of 20°-65° in Figure 4b. The positions of humps coincide with the peak position of several possible tantalum oxides, such as TaO, TaO2 and Ta2O5 based on data in JCPDS card No.78-0724, 19-1296 and 25-0922, which signifying that the deposited TaxOy films may be one and more phases [47]. Besides, compared with TaxOy films, the diffraction curve of ZnO-TaxOy films becomes flat with broader hump, implying the lower crystallinity of ZnO-TaxOy films. The degradation of crystallinity may be caused by the incorporation of ZnO [58]. In addition, according to the data in JCPDS card NO.16-1451, the diffraction peak position of ZnO is around 31° of 2θ, and it is also in the hump area. The specific stoichiometric composition of ZnO-TaxOy films needs further analysis by XPS technique.    The XRD spectra of uncoated and coated Ti6Al4V samples are shown in Figure 4. Compared with the XRD pattern of bare Ti-6Al-4V substrate, no characteristic peaks can be found both for TaxOy and ZnO-TaxOy films in Figure 4a, indicating an amorphous structure. The results are consistent with previous studies [20,[46][47]. The reason for this amorphous nature of the deposited film may be deposition temperature [55], sputtering power [56], flow ratio of oxygen to argon [57], element doping [57], etc. The local magnification pattern shows that the diffraction pattern of TaxOy films is composed of diffuse-scattering curves and two humps appears in the 2θ range of 20°-65° in Figure 4b. The positions of humps coincide with the peak position of several possible tantalum oxides, such as TaO, TaO2 and Ta2O5 based on data in JCPDS card No.78-0724, 19-1296 and 25-0922, which signifying that the deposited TaxOy films may be one and more phases [47]. Besides, compared with TaxOy films, the diffraction curve of ZnO-TaxOy films becomes flat with broader hump, implying the lower crystallinity of ZnO-TaxOy films. The degradation of crystallinity may be caused by the incorporation of ZnO [58]. In addition, according to the data in JCPDS card NO.16-1451, the diffraction peak position of ZnO is around 31° of 2θ, and it is also in the hump area. The specific stoichiometric composition of ZnO-TaxOy films needs further analysis by XPS technique.    Figure 5 shows the elements mapping images on the surface of ZnO-Ta x O y coating samples. The constituent elements of the coating are Ta, Zn and O. It can be observed from the figure that each element is uniformly distributed throughout the coating. Combined with XRD analysis, the Ta element exists in its oxides, while a small amount of Zn comes from ZnO. These results show that ZnO has been incorporated into the Ta x O y film. Ta x O y has a significant effect on improving the corrosion resistance and biocompatibility of Ti6Al4V [1,[9][10][11][12][13][14][15][16][17][18][19][20], while the presences of ZnO have a positive effect on improving the antibacterial properties of the ZnO-Ta x O y coatings [32]. corrosion resistance and biocompatibility of Ti6Al4V [1,[9][10][11][12][13][14][15][16][17][18][19][20], while the presences of ZnO have a positive effect on improving the antibacterial properties of the ZnO-TaxOy coatings [32].     corrosion resistance and biocompatibility of Ti6Al4V [1,[9][10][11][12][13][14][15][16][17][18][19][20], while the presences of ZnO have a positive effect on improving the antibacterial properties of the ZnO-TaxOy coatings [32].      Figure 7b. The Ta 4f 7/2 peaks associated with Ta 4+ (TaO 2 ), Ta 3+ (Ta 2 O 3 ), and Ta 2+ (TaO) are located at 24.4 eV, 24 eV, 23.6 eV respectively, and the spin-orbit splitting of Ta 4f 7/2 to Ta 4f 5/2 is 1.8ev, agrees with reported values [59][60][61]. The value state of Ta is related to the concentration of oxygen and the temperature during or after oxidation. The results indicate that three kinds of tantalum suboxides exist in ZnO-Ta x O y coating surface and Ta 2 O 5 does not appear. The binding energies of Zn 2p peak at 1021.8 eV and 1044.8 eV corresponding to Zn 2p3/2 and Zn 2p1/3 indicated that Zn element is present in the form of ZnO in the coatings [62]. As shown in Figure 7c Figure 7 presents the XPS survey spectrum and elemental high-resolution spectra of ZnO-TaxOy coated samples. It is clear that Ta, Zn and O are detected from the outermost coating of ZnO-TaxOy coated samples. The Ta 4f high-resolution spectrum contains three pairs of Ta 4f double peaks belonging to three chemical states of tantalum, shown in Figure 7b. The Ta 4f7/2 peaks associated with Ta 4+ (TaO2), Ta 3+ (Ta2O3), and Ta 2+ (TaO) are located at 24.4 eV, 24 eV, 23.6 eV respectively, and the spin-orbit splitting of Ta 4f7/2 to Ta 4f 5/2 is 1.8ev, agrees with reported values [59][60][61]. The value state of Ta is related to the concentration of oxygen and the temperature during or after oxidation. The results indicate that three kinds of tantalum suboxides exist in ZnO-TaxOy coating surface and Ta2O5 does not appear. The binding energies of Zn 2p peak at 1021.8 eV and 1044.8 eV corresponding to Zn 2p3/2 and Zn 2p1/3 indicated that Zn element is present in the form of ZnO in the coatings [62]. As shown in Figure 7c

Adhesion Strength
The adhesion strength between the deposited coating and substrate was investigated by scratch test. The critical loads of the scratch test were assessed by optical microscopy observations, with the help of scratch curves. As shown in Figure 8 and Figure 9, with the increase of the scratch length, the loading force increases proportionally, while the friction force shows the characteristic of oscillatory

Adhesion Strength
The adhesion strength between the deposited coating and substrate was investigated by scratch test. The critical loads of the scratch test were assessed by optical microscopy observations, with the help of scratch curves. As shown in Figures 8 and 9, with the increase of the scratch length, the loading force increases proportionally, while the friction force shows the characteristic of oscillatory rise. The fluctuation of the friction force against scratch length may be the occurrence of film cracking and shifting due to weak interface bonding. For Ta x O y coating samples, when the scratch length is 0.65 mm, continuous perforation and peeling of the Ta x O y coating occurs (Figure 8b), and the critical load L c2 is 5.42 N (Figure 8a), which means that the binding strength of Ta x O y is 5.42 N. Some of Ta x O y films were accumulated at the end of the scratch, exposing the Ti6Al4V substrate along the scratch track, as shown in Figure 8c. Figure 9 shows the scratch curve of the ZnO-Ta x O y coating and the magnified image of scratch track. The scratch direction is from left to right. As local coating perforation was observed at the scratch distance of 1.14 mm in Figure 9b, the first critical load is 9.46 N (Figure 9a). When the scratch length is 4.99 mm, the continuous perforation appeared, with a normal load of 41.58 N, indicating that the binding strength of ZnO-Ta x O y coating was 41.58 N. After several intermediate layers introduced into the ZnO-Ta x O y multilayer coating, the composition changed gradually from the bulk Ti6Al4V substrate to the ZnO-Ta x O y coating, preventing from evident interphase/interface (possible interface reaction or diffusion mechanism) [63,64] and cracks in coating (As shown in Figure 3). As a result of it, the multilayer coating demonstrated a fine binding with the substrate. What's more, it should be noted that Ta x O y coating and the Ti6Al4V substrate differ remarkably in thermal expansion coefficient (CTE); an environmental temperature fluctuation may cause the peel-off or dysfunction/defect of the whole coatings. The intermediate layer plays a buffering effect to reduce the CTE mismatch between the Ti6Al4V (α Ti6Al4V = 8.9 × 10 −6 K −1 ) [65] substrate and Ta x O y coating (α Ta-O = 2.9 × 10 −6 K −1 ) [66]. Under some temperature variations, it maintains the binding ability and post heating treatment could facilitate the interfacial reactions and diffusions, in turn enhancing the adhesion strength between the Ti6Al4V substrate and the multilayer coating.
track. The scratch direction is from left to right. As local coating perforation was observed at the scratch distance of 1.14 mm in Figure 9b, the first critical load is 9.46 N (Figure 9a). When the scratch length is 4.99 mm, the continuous perforation appeared, with a normal load of 41.58 N, indicating that the binding strength of ZnO-TaxOy coating was 41.58 N. After several intermediate layers introduced into the ZnO-TaxOy multilayer coating, the composition changed gradually from the bulk Ti6Al4V substrate to the ZnO-TaxOy coating, preventing from evident interphase/ interface (possible interface reaction or diffusion mechanism) [63,64] and cracks in coating (As shown in Figure 3). As a result of it, the multilayer coating demonstrated a fine binding with the substrate. What's more, it should be noted that TaxOy coating and the Ti6Al4V substrate differ remarkably in thermal expansion coefficient (CTE); an environmental temperature fluctuation may cause the peel-off or dysfunction/defect of the whole coatings. The intermediate layer plays a buffering effect to reduce the CTE mismatch between the Ti6Al4V (αTi6Al4V = 8.9 × 10 −6 K −1 ) [65] substrate and TaxOy coating (αTa-O = 2.9 × 10 −6 K −1 ) [66]. Under some temperature variations, it maintains the binding ability and post heating treatment could facilitate the interfacial reactions and diffusions, in turn enhancing the adhesion strength between the Ti6Al4V substrate and the multilayer coating.
In addition, Figure 8 and Figure 9 obviously reveal that under the same experimental condition, the scratch width of ZnO-TaxOy sample is lower than that of TaO sample. As the coating thickness increases, the scratch resistance of coating surface enhances, and the scratch width and depth are decreased [67]. The coating thickness of ZnO-TaxOy sample is greater than that of sample (Figure 3), so it has higher surface resistance to scratch, and shows smaller scratch width compared with TaxOy sample. Further observation can be found that the first spallation occurred on the edges of scratch track near the LC1 position (Figure 8b and Figure 9b), and then more spallation dominated on both side of the scratch track, which revealed that the coating's adhesive strength was higher than its cohesive strength, since spallation normally appear in brittle coatings and is induced by cracks on the coating surface or within the coating [68]. For TaO sample, the spallation appeared at a lower critical load, compared with ZnO-TaxOy sample, indicating that the adhesive strength of the TaxOy coating was lower than that of ZnO-TaxOy coating.  In addition, Figures 8 and 9 obviously reveal that under the same experimental condition, the scratch width of ZnO-Ta x O y sample is lower than that of TaO sample. As the coating thickness increases, the scratch resistance of coating surface enhances, and the scratch width and depth are decreased [67]. The coating thickness of ZnO-Ta x O y sample is greater than that of sample (Figure 3), so it has higher surface resistance to scratch, and shows smaller scratch width compared with Ta x O y sample. Further observation can be found that the first spallation occurred on the edges of scratch track near the L C1 position (Figures 8b and 9b), and then more spallation dominated on both side of the scratch track, which revealed that the coating's adhesive strength was higher than its cohesive strength, since spallation normally appear in brittle coatings and is induced by cracks on the coating surface or within the coating [68]. For TaO sample, the spallation appeared at a lower critical load, compared with ZnO-Ta x O y sample, indicating that the adhesive strength of the Ta x O y coating was lower than that of ZnO-Ta x O y coating.

Corrosion Behavior
The representative potentiodynamic polarization curves of the un-coated and coated Ti6Al4V samples in the SBF are shown in Figure 10. The corrosion potential (Ecorr) and corrosion current density (Icorr) derived by Tafel extrapolation method and were shown in Table 2. The corrosion potential describes the substrates' tendency to corrode and the corrosion current density indicates the corrosion rate [18]. The corrosion potential of uncoated Ti6Al4V substrate was approximately −0.19 V, while the corrosion potential of TaxOy and ZnO-TaxOy coated Ti6Al4V shift towards the positive potentials, the corrosion potential of TaxOy and ZnO-TaxOy coated samples was −0.11 V and 0.02 V, respectively. The corrosion current densities measured from Ti6Al4V, TaxOy and ZnO-TaxOy coated Ti6Al4V was about 7.07 μA/cm 2 , 3.85 μA/cm 2 , and 1.12 μA/cm 2 , respectively, exhibiting a descending trend in the SBF. These indicated that the TaxOy and ZnO-TaxOy coatings diminish appreciably the corrosion rate of Ti6Al4V substrate. The excellent corrosion resistance of the coated samples may be concerned with the superior stability of TaxOy ceramic coatings. Besides, the ZnO-TaxOy coatings showed better corrosion resistance than TaxOy coating, this can be mainly explained by the fact that ZnO-TaxOy coatings has higher surface quality and better bond with Ti6Al4V substrate than TaxOy coatings does. The above analyses reveal that the ZnO-TaxOy coatings might ensure favorable anti-corrosion property in implant application.

Corrosion Behavior
The representative potentiodynamic polarization curves of the un-coated and coated Ti6Al4V samples in the SBF are shown in Figure 10. The corrosion potential (E corr ) and corrosion current density (I corr ) derived by Tafel extrapolation method and were shown in Table 2. The corrosion potential describes the substrates' tendency to corrode and the corrosion current density indicates the corrosion rate [18]. The corrosion potential of uncoated Ti6Al4V substrate was approximately −0.19 V, while the corrosion potential of Ta x O y and ZnO-Ta x O y coated Ti6Al4V shift towards the positive potentials, the corrosion potential of Ta x O y and ZnO-Ta x O y coated samples was −0.11 V and 0.02 V, respectively. The corrosion current densities measured from Ti6Al4V, Ta x O y and ZnO-Ta x O y coated Ti6Al4V was about 7.07 µA/cm 2 , 3.85 µA/cm 2 , and 1.12 µA/cm 2 , respectively, exhibiting a descending trend in the SBF. These indicated that the Ta x O y and ZnO-Ta x O y coatings diminish appreciably the corrosion rate of Ti6Al4V substrate. The excellent corrosion resistance of the coated samples may be concerned with the superior stability of Ta x O y ceramic coatings. Besides, the ZnO-Ta x O y coatings showed better corrosion resistance than Ta x O y coating, this can be mainly explained by the fact that ZnO-Ta x O y coatings has higher surface quality and better bond with Ti6Al4V substrate than Ta x O y coatings does. The above analyses reveal that the ZnO-Ta x O y coatings might ensure favorable anti-corrosion property in implant application. Table 2. Corrosion parameters derived from polarization curves of Figure 10.

Sample
Ti6Al4V Ta  samples may be concerned with the superior stability of TaxOy ceramic coatings. Besides, the ZnO-TaxOy coatings showed better corrosion resistance than TaxOy coating, this can be mainly explained by the fact that ZnO-TaxOy coatings has higher surface quality and better bond with Ti6Al4V substrate than TaxOy coatings does. The above analyses reveal that the ZnO-TaxOy coatings might ensure favorable anti-corrosion property in implant application.

Wettability
Wettability refers to the ability or propensity of a liquid to spread on a solid surface. The wettability of the solid is usually evaluated by the contact angle. The smaller the contact angle, the better the wettability of the solid. Figure 11 shows the measured 4 contact angle data and the images of water drops on the surface of un-coated and coated Ti6Al4V samples. The water contact angle of Ti6Al4V alloy substrate is 84.27 • ± 2.6 • , indicating the hydrophilic surface. While Ta x O y and ZnO-Ta x O y coated on Ti6Al4V alloy exhibits the water contact angle greater than 90 • with hydrophobic properties. The hydrophobicity of the surface helps prevent the initial adhesion of bacteria and the formation of biofilms on the surface [66,69], and is beneficial for improving the corrosion resistance of the surface.

Wettability
Wettability refers to the ability or propensity of a liquid to spread on a solid surface. The wettability of the solid is usually evaluated by the contact angle. The smaller the contact angle, the better the wettability of the solid. Figure 11 shows the measured 4 contact angle data and the images of water drops on the surface of un-coated and coated Ti6Al4V samples. The water contact angle of Ti6Al4V alloy substrate is 84.27° ± 2.6°, indicating the hydrophilic surface. While TaxOy and ZnO-TaxOy coated on Ti6Al4V alloy exhibits the water contact angle greater than 90° with hydrophobic properties. The hydrophobicity of the surface helps prevent the initial adhesion of bacteria and the formation of biofilms on the surface [66,69], and is beneficial for improving the corrosion resistance of the surface.
While the hydrophobic surface can prevent cells from directly adhering on the coating surface, the amount of protein previously adsorbed on the hydrophobic surface is higher than that on the hydrophilic surface [70,71]. Protein can induce indirect cell adhesion, so the hydrophobic surface is beneficial to indirect cell adhesion. In addition, the hydrophilic surface, which was good for direct cell adhesion, actually causes side effects on protein adsorption followed by indirect cell adhesion. Therefore, the initial cell adsorption surface is neither too hydrophobic nor too hydrophilic [72]. In addition, the wettability of the surface can be controlled by doping the coating, decreasing surface roughness of the substrate, and adopting appropriate sputtering parameters [73].

Antibacterial Property
The antibacterial properties of the ZnO-TaxOy coating were compared with those of bare Ti6Al4V alloy and TaxOy coating. S. aureus was selected as the test bacteria. Figure 12a-c shows the image of bacteriological tests of S. aureus on solid agar plates incubated at 37 °C for 24 h on Ti6Al4V surface, TaxOy and ZnO-TaxOy coatings. It is clear that there are a large number of S. aureus colonies incubated on Ti6Al4V surface, while several S. aureus colonies are observed on ZnO-TaxOy coating. These results revealed that ZnO incorporated TaxOy films strongly inhibited S. aureus growth, which had an antibacterial rate of 90.65%, while the rate for the TaxOy coating was 19.78% (Figure 12d). The antimicrobial activity of ZnO-TaxOy coatings are due to the generation of reactive oxygen species (ROS), and release of Zn 2+ ions [74]. The production of ROS including superoxide anion (O2 2− ), hydrogen peroxide (H2O2), and hydroxide (OH − ). The O 2− and OH − cannot penetrate into the cell membrane due to their negative charges, while H2O2 molecules are able to pass through the bacterial While the hydrophobic surface can prevent cells from directly adhering on the coating surface, the amount of protein previously adsorbed on the hydrophobic surface is higher than that on the hydrophilic surface [70,71]. Protein can induce indirect cell adhesion, so the hydrophobic surface is beneficial to indirect cell adhesion. In addition, the hydrophilic surface, which was good for direct cell adhesion, actually causes side effects on protein adsorption followed by indirect cell adhesion. Therefore, the initial cell adsorption surface is neither too hydrophobic nor too hydrophilic [72]. In addition, the wettability of the surface can be controlled by doping the coating, decreasing surface roughness of the substrate, and adopting appropriate sputtering parameters [73].

Antibacterial Property
The antibacterial properties of the ZnO-Ta x O y coating were compared with those of bare Ti6Al4V alloy and Ta x O y coating. S. aureus was selected as the test bacteria. Figure 12a-c shows the image of bacteriological tests of S. aureus on solid agar plates incubated at 37 • C for 24 h on Ti6Al4V surface, Ta x O y and ZnO-Ta x O y coatings. It is clear that there are a large number of S. aureus colonies incubated on Ti6Al4V surface, while several S. aureus colonies are observed on ZnO-Ta x O y coating. These results revealed that ZnO incorporated Ta x O y films strongly inhibited S. aureus growth, which had an antibacterial rate of 90.65%, while the rate for the TaxOy coating was 19.78% (Figure 12d). The antimicrobial activity of ZnO-Ta x O y coatings are due to the generation of reactive oxygen species (ROS), and release of Zn 2+ ions [74]. The production of ROS including superoxide anion (O 2 2− ), hydrogen peroxide (H 2 O 2 ), and hydroxide (OH − ). The O 2− and OH − cannot penetrate into the cell membrane due to their negative charges, while H 2 O 2 molecules are able to pass through the bacterial cell wall, subsequently leading to injuries and destruction, and finally triggering cell death [75]. The ROS produced by ZnO can kill bacteria as well as affect the growth of osteoblasts. ROS arrests osteoblasts proliferation, decrease osteoblasts growth and/or differentiation, and promotes osteoblasts death by activating various signaling [76]. The key problem of ZnO used in implant is to determine its ideal content, because of that Zn has been shown to promote osseointegration or have cytotoxic effects at low and high concentrations, respectively [77][78][79]. The released Zn 2+ ions penetrate into the interior of the cell through the cell membrane, then interact with the genome and plasmid DNA, which interferes with the growth of the bacteria and destroys the amino acid metabolism, eventually, leading to cell death [80,81]. ROS produced by ZnO can kill bacteria as well as affect the growth of osteoblasts. ROS arrests osteoblasts proliferation, decrease osteoblasts growth and/or differentiation, and promotes osteoblasts death by activating various signaling [76]. The key problem of ZnO used in implant is to determine its ideal content, because of that Zn has been shown to promote osseointegration or have cytotoxic effects at low and high concentrations, respectively [77][78][79]. The released Zn 2+ ions penetrate into the interior of the cell through the cell membrane, then interact with the genome and plasmid DNA, which interferes with the growth of the bacteria and destroys the amino acid metabolism, eventually, leading to cell death [80,81].

Conclusions
In this paper, ZnO doped tantalum oxide (Ta x O y ) multilayer composite coatings (ZnO-Ta x O y /Ta x O y /Ta x O y -TiO 2 /TiO 2 /Ti, coating code ZnO-Ta x O y ) were successfully fabricated on the Ti6Al4V substrate surface with magnetron sputtering technique. The results suggest that the ZnO-Ta x O y coating had great potential for improving the corrosion resistance and enhancing antibacterial property against S. aureus for Ti6Al4V implants. This study provides an alternative modified coating on Ti6Al4V for orthopedic application. However, the cytocompatibility and more optimization of the ZnO-Ta x O y coatings (e.g., preparation parameters, coating thickness and composition) remain to be completed in the future.