Optimization of annealing treatment and comprehensive properties of Cu-containing Ti6Al4V-xCu alloys

1. Introduction

At present, nosocomial infections caused by implant materials still maintain an important problem to be solved [1,2]. The implants related infections may lead to possible complications, such as prolonged hospitalization, complex revision procedures, implant failure, complete removal, patient suffering, and even death [3,4]. Once infection occurs, it will cause great pain and heavy economic and psychological burdens to the patients. Much has been done in terms of preventing infections, for example, control of asepsis, sterile procedures, adequate protocols of perioperative antibiotic prophylaxis and appropriate management of antibiotics for medical treatment [5]. But the abuse of antibiotics may give rise to the threat of drug-resistant bacteria [6]. Another important strategy that has been progressively gaining ground over the years is the use of new antibacterial biomaterials with self-antibacterial function [[7], [8], [9]]. It is generally believed that copper (Cu) is a necessary trace element in the human body and Cu²⁺ ions possess inherent antibacterial ability [10]. From the view of composition design, Cu has been added into the implant materials by many researchers to endow the materials with antibacterial properties. For example, it was reported [11,12] that by addition of Cu, stainless steel can possess excellent antibacterial property.

In orthopedic and dental areas, traditional Ti6Al4V alloy has been widely used due to its characteristics of low modulus, high strength, excellent corrosion resistance, good biocompatibility and so on [13]. However, Ti6Al4V alloy is bio-inert and hence does not show any antibacterial activity. In recent years, previous studies have tried to add Cu into Ti6Al4V alloy in order to obtain a new kind of antibacterial titanium alloy, targeting for applications in orthopedic and dental implants, such as artificial joints (hip, knee, shoulder, ankle, elbow, wrist, knuckle, etc.), spine systems, bone fixations (intramedullary nail, plate, screw, etc.), and even dental implants [14]. Ren et al. [14] have found that the as-cast Ti-6Al-4V-xCu (x = 1, 3, 5 wt%) alloys possess good corrosion resistance and biocompatibility. Meanwhile, the Ti6Al4V-5Cu alloy revealed better antibacterial property than the Ti6Al4V-3Cu and Ti6Al4V-1Cu alloys, indicating that the antibacterial property was enhanced with the increase of Cu content. The research from Wan et al. [15] and Shirai et al. [16] also showed that the antibacterial activity of Cu-containing material is strongly dependent on the Cu content. Aoki et al. [17] reported that, for the as-cast Ti6Al4V-xCu (x = 1, 4, 10 wt%) alloys, the ductility was significantly deteriorated with increase of the Cu content, and the elongation was in the range of 0.2%-2.8%, far lower than the requirement of ISO 5832-3:2016 Standard for the medical Ti6Al4V alloy bars (not less than 10%). Other researchers [[18], [19], [20], [21]] also reported that the unsatisfied ductility was a common problem for the Cu-containing alloys. In our previous study [24], we optimized the ductility of Ti6Al4V-5Cu alloy by hot-processing and annealing treatment, in which the alloy annealed at 740 °C showed 6% higher ductility than the standard required. The Cu content in titanium alloys not only affects the mechanical properties, but is also related to the biocompatibility. It was reported that excessive Cu intake could cause stomach upset, nausea and diarrhea, and could also lead to tissue injury and disease [22]. The inevitable wear or corrosion in the use of Cu-bearing titanium alloys may also accelerate the excessive release of metal ions and increase the risk of toxicity [23]. According to the above studies, we can conclude that the more Cu addition into the titanium alloy, the better antibacterial property it possesses, but higher Cu content will deteriorate the ductility, corrosion resistance and biocompatibility of the alloy. Therefore, it’s necessary to find the most appropriate range of Cu content for the Ti6Al4V-xCu alloys. Although the ductility of Ti6Al4V-5Cu alloy was optimized by hot-processing and annealing treatment, how to guarantee the mechanical properties of other Cu-containing Ti6Al4V-xCu alloys is still unclear.

In this study, relatively higher Cu contents (4.5, 6 and 7.5 wt%) were added into Ti6Al4V alloy for the reason of increasing the antibacterial properties. Furthermore, the annealing temperatures for each Cu-bearing alloy were optimized in a narrow range (720-760 °C) in order to obtain both high strength and enough ductility. Then, each alloy with optimized annealing treatment was studied to optimize the Cu content that endows the alloy with the best comprehensive properties, including tribological property, corrosion resistance, antibacterial property and biocompatibility. The novelty of our study is that we have optimized the comprehensive properties of such high Cu-containing Ti6Al4V-xCu (x = 4.5, 6, 7.5 wt%) alloys by optimizing annealing treatment and controlling Cu contents. It is of importance in medical transformation and clinical application for this novel antibacterial Cu-containing titanium.

2. Materials and methods

2.1. Materials

Three Cu-containing titanium alloys, Ti6Al4V-xCu (x = 4.5, 6, 7.5 wt%), were prepared by a 25 kg vacuum consumable melting furnace. The chemical compositions are listed in Table 1. The alloy ingots were hot-forged into round bars with the dimension of Φ12 mm. The microstructures of the hot-processed alloys were observed by an optical microscope (Axiovert 200 MAT), as shown in Fig. 1. It can be found that primary β grain boundaries are completely broken in the Ti6Al4V-4.5Cu alloy. But in the other two alloys, Ti6Al4V-6Cu and Ti6Al4V-7.5Cu, primary β grain boundaries can still be observed. The α+Ti₂Cu→β and α→β transformation temperatures of different alloys were obtained by differential thermal analysis (DTA), and the results are inserted in Fig. 1, respectively. It can be observed that the α+Ti₂Cu→β transus temperatures of Ti6Al4V-xCu alloys are increased and the α→β transus temperatures are decreased with increase of the Cu content. After hot-processing, all the annealing treatments were carried out in the α+β+Ti₂Cu ternary phase region, with temperatures of 720, 740 and 760 °C, respectively, duration time of 1 h, and cooling in air. The annealed samples are named as 4.5Cu-720, 4.5Cu-740, 4.5Cu-760, 6Cu-720, 6Cu-740, 6Cu-760, 7.5Cu-720, 7.5Cu-740 and 7.5Cu-760, respectively. Commercial medical grade Ti6Al4V bars were used as control.

View original graphic| Download| PPT slide

Fig. 1.   Optical microstructures of hot-processed Ti6Al4V-xCu alloys and corresponding differential thermal analysis (DTA) graphs: (a) Ti6Al4V-4.5Cu alloy; (b) Ti6Al4V-6Cu alloy; (c) Ti6Al4V-7.5Cu alloy.

2.2. Methods

2.2.1. Microstructure observation

Scanning electron microscope (SEM, SHIMADZU SSX-550) and transmission electron microscope (TEM, Tecnai G² 20) with energy dispersive spectroscopy (EDS) were used to observe the microstructures of alloys. For SEM observation, samples were mechanically ground and polished, and then etched by a Kroll reagent. For TEM observation, samples were mechanically ground to 40 μm in thickness and ion-milled at a low angle of 4°. A Rigaku D/max2500 pc type X-ray diffraction (XRD) was used to identify different phases, using a CuK_α irradiation, with tube voltage of 50 kV, tube current of 300 mA, and scan velocity of 4°/min.

2.2.2. Tensile properties measurement

Tensile properties measurements were carried out on an Instron-8872 universal testing machine at ambient temperature. The stretching rate was 0.5 mm/min. Standard dimension specimens with diameter of 5 mm and working length of 25 mm was used. Three parallel samples were used for each test.

2.2.3. Tribological performance

Tribological tests were carried out by a universal multifunctional tester (CETR-UMT-2 M) in a reciprocating sliding style. The frequency was 2 Hz and displacement amplitude was 10 mm. The loading force was 10 N and loading time was 30 min. Si₃N₄ sphere balls with diameter of 5 mm were used as the counter body. Samples were prepared by mechanical grinding and polishing. The variation of friction coefficient with sliding time was recorded and average values were calculated. Weight loss values of the alloys were analyzed by an analytical balance. The wear areas of scars on surface, the maximum wear depths caused by fretting wear, and the cross-section profiles of wear track were measured by a surface profilometer (Alpha-Step IQ). The morphologies of the worn surfaces were observed by SEM.

2.2.4. Corrosion behavior

Corrosion resistance of alloys was investigated by both potentiodynamic polarization and electrochemical impedance spectroscopy (EIS) tests using a potentiostat/galvanostat (Reference 600 T M, Gamry Instruments, Inc., USA). The electrolyte was 0.9 wt% NaCl solution. The temperature was 37 ± 1 °C. Samples were served as the working electrode with exposure area of 0.785 cm². A platinum foil was used as the counter electrode and a saturated calomel electrode was served as the reference electrode. Firstly, the open circuit potential test was applied for 1 h until the potential value became stable. Then the EIS test was conducted from the frequency of 10⁵ Hz down to 10^-2 Hz at the open-circuit potential, and alternative current amplitude of ±10 mV was applied. The impedances were analyzed by the Gamry Echem Analyst software. After the EIS test, potentiodynamic measurement was performed at a scanning rate of 0.5 mV/s, starting from -0.5 to 2 V_OCP (OCP: open circuit potential). The corrosion potential (E_corr) and the corrosion current density (I_corr) were obtained by Tafel fitting of the experimental curves. The pitting potential (E_p) was obtained where the stable pitting was formed. The passive current density (I_p) was taken at 1 V_SCE (SCE: saturated calomel electrode). Each test was repeated for three times.

2.2.5. Antibacterial test

The antibacterial test was carried out according to the China Standard GB/T 2591 (equivalent to JIS Z 2801-2000, ASTM G21-96, and NEQ). Before the test, samples were mechanically ground up to 1000# sandpaper, ultrasonically cleaned in ethyl alcohol for 15 min, and sterilized at 121 °C for 20 min. A typical kind of Gram-positive bacteria, Staphylococcus aureus ATCC25923 (S. aureus), was chosen for the test. 50 μL bacterial suspension with a concentration of 10⁵ cfu/mL was inoculated on the sample surface for 24 h at 37 °C. Then each sample was put into a centrifuge tube and 3 ml of phosphate buffered saline (PBS) was added into the tube. After vortex shaking for 1 min, the bacteria were washed down from the surface of samples. Afterwards, 100 μL of the bacterial solution was cultured on a nutrient agar petri dish at 37 °C for 24 h. The petri dishes were prepared by adding 5 g beef extract, 5 g NaCl, 10 g peptone and 20 g agar powder into 1000 ml distilled water, and the pH value was adjusted to 7.2-7.4. Then the bacterial colonies appeared on the petri dishes were counted and the antibacterial rates were calculated from the following formula:

Antibacterial rate (%) = (A_control ‐ A_tes) / A_control ×100% (1)

where A_control represents for the average amount of bacteria colonies on the petri dish of Ti6Al4V alloy, and A_test represents for that of the Ti6Al4V-xCu alloys. The test was repeated for at least 5 times.

2.2.6. Cytotoxicity assay

Human umbilical vein endothelial cells (HUVECs) were grown in a RPMI.1640 medium that composed of 10% fetal bovine serum, 80 U/mL penicillin and 0.08 mg/mL streptomycin. After 80% of the cells were confluent, warm phosphate buffer solution (PBS) was used to wash the cells. Then a 0.25 wt% trypsin-EDTA solution was added to detach the cells. After the HUVECs became round, some amount of complete medium was added into the flask, and the cells were removed into a centrifuge tube. Then the cells were centrifuged and resuspended for passage. Then the HUVECs were cultured in a standard culture incubator at 37 °C, with humidified air in it, containing 5% CO₂.

The effect of different alloys on cell proliferation was evaluated by the CCK-8 assay. Samples were put in the bottom of 96-well plates, and 100 μL medium that contained 5 × 10⁴ cells were added into each well. After culturing for 1 d, 3 d, and 7 d, respectively, the medium was replaced by 110 μL solution that was prepared by mixing the medium and CCK-8 solution at the ratio of 10:1. After incubation for 1 h at 37 °C, 100 μL solution was transferred to a fresh 96-well plate. Then a microplate reader was used to measure the absorbance at 450 nm.

2.2.7. Statistical analysis

The statistical difference of CCK-8 results was analyzed by independent-sample t-test in SPSS 13.0 software. Statistical significance was defined as p < 0.05.

3. Results

3.1. Microstructures

The Ti6Al4V-xCu (x = 4.5, 6, 7.5 wt%) alloys were annealed at 720, 740 and 760 °C, respectively, and the microstructures are shown in Fig. 2. The samples of 4.5Cu-720, 6Cu-740 and 7.5Cu-760 were chosen to identify the phases by XRD (Fig. 3(a)). The results reveal that three kinds of phases can be identified, which are α, β and Ti₂Cu. The SEM images were used to calculate the volume fractions of the three phases in different alloys (Fig. 3(b)). It can be observed that for each alloy, with increase of the annealing temperature, volume fractions of both α and Ti₂Cu phases were decreased, but that of the β phase was increased. Comparing the samples at the same annealing temperature, it can be seen that with increase of the Cu content, the volume fraction of α phase was decreased, but those of Ti₂Cu and β phases were increased.

View original graphic| Download| PPT slide

Fig. 2.   SEM microstructures of Ti6Al4V-xCu alloys annealed at different temperatures: (a) 4.5Cu-720; (b) 4.5Cu-740; (c) 4.5Cu-760; (d) 6Cu-720; (e) 6Cu-740; (f) 6Cu-760; (g) 7.5Cu-720; (h) 7.5Cu-740; (i) 7.5Cu-760.

View original graphic| Download| PPT slide

Fig. 3.   (a) XRD patterns of 4.5Cu-720, 6Cu-740 and 7.5Cu-760 samples, (b) calculated volume fractions of different phases in Ti6Al4V-xCu alloys annealed at 720 °C, 740 °C and 760 °C, (c) TEM bright images and (d) selected diffraction patterns of Ti2Cu phase in 6Cu-740 sample.

Furthermore, the sample of 6Cu-740 was used to examine the microstructure of α+β+Ti₂Cu by TEM (Fig. 3(c)). The chemical compositions of point 1-3 in Fig. 3(c) were determined by EDS, and the results are summarized in Table 2. It can be found that the size of the intermetallic Ti₂Cu phase (point 1) is about a few hundred nanometers, with compositions of 40 wt% Cu, 59.1 wt% Ti, and very little amounts of Al and V. The selected diffraction patterns in Fig. 3(d) confirms the structure of Ti₂Cu phase. In the β phase (point 2), the contents of β-stabilizer elements, V (13.1 wt%) and Cu (8.6 wt%), are higher than the nominal composition, but the content of α-stabilizer element, Al (2.8 wt%), is lower. While in the α phase (point 3), it contains higher content of Al (6.8 wt%), but less V (1.7 wt%) and Cu (3.2 wt%) than the nominal composition.

3.2. Tensile properties

The engineering tensile stress-strain curves of Ti6Al4V-xCu (x = 4.5, 6, 7.5 wt%) alloys are plotted in Fig. 4(a)-(c), respectively. The mechanical properties (yield strength, tensile strength and elongation) of different samples are summarized in Fig. 4(d). According to the requirement of ISO 5832-3:2016 Standard for the medical Ti6Al4V alloy bars, the yield strength, tensile strength and elongation should be higher than 860 MPa, 930 MPa and 10%, respectively. Thus, for the Ti6Al4V-xCu alloys in the present study, all the yield strength and tensile strength are higher than the requirement of the Standard, but some of the elongations (4.5Cu-760 and 6Cu-760) are lower than the requirement. For the Ti6Al4V-4.5Cu alloy, the elongation was significantly decreased from 15.0% to 9.7% with the increase of annealing temperature from 720 to 760 °C. For the Ti6Al4V-6Cu alloy, the elongation was firstly increased a little from 12.7% to 13.4%, then was dramatically decreased to 8.2% when the annealing temperature was elevated to 760 °C. For the Ti6Al4V-7.5Cu alloy, the elongation was slowly increased from 10.5% to 11.8% as the annealing temperature increased from 720 to 760 °C. Therefore, samples of 4.5Cu-720, 6Cu-740 and 7.5Cu-760 possessed the highest elongation values for each kind of alloy. It can also be seen that, in these three optimized samples of 4.5Cu-720, 6Cu-740 and 7.5Cu-760, the ductility was deteriorated with increase of the Cu content.

View original graphic| Download| PPT slide

Fig. 4.   Engineering tensile stress-strain curves of annealed (a) Ti6Al4V-4.5Cu, (b) Ti6Al4V-6Cu and (c) Ti6Al4V-7.5Cu alloys and (d) mechanical properties of different samples.

3.3. Tribological characteristics

The samples of 4.5Cu-720, 6Cu-740 and 7.5Cu-760, with the optimized mechanical properties, were subsequently used in the later experiments to study the effect of Cu content on the wear resistance of alloys, in order to further optimize the Cu content.

Variations of friction coefficients vs. sliding time and the average friction coefficients of the optimized samples are shown in Fig. 5(a) and (b), respectively. It can be seen from Fig. 5(a) that all the Cu-containing samples show lower friction coefficients than that of Ti6Al4V alloy. Fig. 5(b) also shows that the average friction coefficients of Cu-containing alloys are significantly lower than that of Ti6Al4V alloy. Meanwhile, the average friction coefficients of 6Cu-740 and 7.5Cu-760 samples are a little bit lower than that of 4.5Cu-720 sample. The weight loss (Fig. 5(c)), wear area and maximum wear depth of scars (Fig. 5(d)) for all the Cu-containing samples are significantly lower than those of Ti6Al4V alloy. All the above results indicate that the wear resistance is increased by the Cu addition.

View original graphic| Download| PPT slide

Fig. 5.   Variations of friction coefficients vs. sliding time (a), average friction coefficients (b), weight loss (c), and wear area and the maximum wear depth of scars (d) for samples of Ti6Al4V, 4.5Cu-720, 6Cu-740 and 7.5Cu-760.

The cross-section profiles of wear tracks and the SEM morphologies of wear scars on sample surfaces are shown in Fig. 6, Fig. 7, respectively. In Fig. 6(a), Ti6Al4V alloy reveals much deeper wear depth than all the Cu-containing alloys (Fig. 6(b)-(d)), indicating that the Cu-containing alloys possessed better wear resistance than the Ti6Al4V alloy, which is in accordance with the results of friction coefficient measurement. But the profile of 7.5Cu-760 (Fig. 6(d)) alloy is much rougher than those of 4.5Cu-720 and 6Cu-740 alloys, indicating a more damaged surface for 7.5Cu-760 alloy. Similarly, in Fig. 7(a)-(d), surface morphologies of the wear scars show that 4.5Cu-720 and 6Cu-740 alloys experienced less wear damages compared to the Ti6Al4V alloy. But the 7.5Cu-760 alloy shows a seriously damaged surface, which confirms the characteristics observed in profiles of wear scars (Fig. 6). Therefore, the 4.5Cu-720 and 6Cu-740 alloys showed better wear resistance than Ti6Al4V alloy, but high addition of Cu in the 7.5Cu-760 alloy deteriorated the wear resistance. What’s more, EDS analysis of the delamination regions on all the samples revealed that the compositions were from both of the substrate alloy and the Si₃N₄ counter-face material, indicating a typical adhesive wear process on the alloys. Meanwhile, numerous ploughing grooves and many wear debris were observed on all the alloys, as shown in Fig. 7(a1)-(d1), which indicates a typical abrasive wear process. On the surface of 7.5Cu-760 alloy, some micro-cracks are also observed.

View original graphic| Download| PPT slide

Fig. 6.   Cross-section profiles of wear tracks on different samples: (a) Ti6Al4V; (b) 4.5Cu-720; (c) 6Cu-740; (d) 7.5Cu-760.

View original graphic| Download| PPT slide

Fig. 7.   SEM images of wear scars on (a) Ti6Al4V, (b) 4.5Cu-720, (c) 6Cu-740 and (d) 7.5Cu-760 and their corresponding magnified graphs in (a1), (b1), (c1) and (d1), respectively.

3.4. Corrosion resistance

The results of potentiodynamic polarization are illustrated in Fig. 8(a) and the electrochemical parameters are summarized in Table 3. It can be found that the 4.5Cu-720 alloy shows higher corrosion potential, E_corr, and lower corrosion current density, I_corr, than those of the Ti6Al4V alloy, indicating that the addition of 4.5 wt% Cu enhanced the corrosion resistance of the Ti6Al4V alloy. As the Cu content increased to 6%, the E_corr is decreased and I_corr is increased to a small extent, indicating that for the 6Cu-740 alloy, the resistance to uniform corrosion is a little bit deteriorated, comparing with the 4.5Cu-720 alloy. However, when the Cu content is increased to 7.5%, the E_corr is significantly decreased and the I_corr is significantly increased compared to the 6Cu-740 alloy, indicating that the Cu content should not be higher than 6%, otherwise the uniform corrosion resistance of the Ti6Al4V-xCu alloy will be significantly deteriorated. Meanwhile, as the Cu content increases from 4.5% to 7.5%, the pitting potential, E_p, is decreased, and the passivation current density, I_p, is increased to a small extent, indicating that the addition of Cu had some negative effect on protectiveness of the passive film.

View original graphic| Download| PPT slide

Fig. 8.   (a) Typical potentiodynamic polarization curves, (b) Bode plots, (c) Nyquist diagrams and (d) equivalent electrical circuit used for fitting EIS data, while R_s, R₁, Q₁, R₂ and Q₂ are solution resistance, charge-transfer resistance, CPE of interface of electrolyte/passive film/substrate, passive layer resistance, CPE of the passive layer, respectively.

The results of electrochemical impedance spectroscopy test and the fitted equivalent electrical circuit (EEC) model are presented in Fig. 8(b)-(d). As seen from the Bode diagram (Fig. 8(b)), at the lowest frequency of 10^-2 Hz, the absolute impedances of the Cu-containing alloys are all higher than that of Ti6Al4V alloy, which indicates that the Cu addition enhanced the corrosion resistance of alloy. The absolute impedance between the high frequencies of 10³-10⁵ Hz represents the solution resistance of R_s [25], showing little difference among the samples. For the Nyquist diagram (Fig. 8(c)), it is well known that larger diameter of a semi-circle means a higher corrosion resistance [26,27]. The diameters of the curves for different samples are in the following order: 4.5Cu-720 > 6Cu-740 > 7.5Cu-760 > Ti6Al4V, implying that the Ti6Al4V alloy showed the worst corrosion resistance, and with increase of the Cu content, the corrosion resistance is deteriorated. An equivalent circuit model (Fig. 8(d)) was well fitted to the EIS curves and the fitted results are summarized in Table 4. The fitting errors are at a small order of magnitude of 10^-3. In this circuit, a pair of elements of Q₁ and R₁ in parallel represent the charge transfer process at the surface. Another pair of elements of Q₂ and R₂ in parallel is used to describe the dielectric properties of the surface passive layer of the alloys. R_s is used to present the solution resistance. Here, two constant phase elements (CPE) of Q₁ and Q₂ are used to replace the capacitance, because the ideal capacitance hardly exists in a real electrochemical process due to the surface induced deviations. The CPE is given as:

ZCPE=$\frac{1}{(jω)^n}$ (2)

where Q represents the magnitude of the CPE, ω is the angular frequency, and n is the deviation parameter. This is a dispersion formula, when n = 1, it represents the capacitance, when n = 0, it represents the resistance [25]. In this study, all the n values are close to 1, indicating a capacitive characteristic. According to the above EIS fitted parameters, the impedance of electrode (Z_ω) and the polarization resistance (R_p), can be calculated with the following relation:

Zω=$\frac{1}{\frac{1}{R1} Q1(jω)^{n}}+ \frac{1}{\frac{1}{R_{2}} Q2(jω)^{n}}$ (3)

Based on Eq. (3), R_p can be calculated from the following formula:

Rp= ${lim \atop ω→0}$(Zω) =R1+R₂ (4)

Then the R_p values can be used to identify the corrosion resistance of different alloys. As presented in Table 4, the calculated R_p values of samples show an order of 4.5Cu-720 > 6Cu-720 > 7.5Cu-720 > Ti6Al4V, in which the R_p values of Cu-containing alloys are significantly higher than that of Ti6Al4V alloy. The maximum R_p value (1500 kΩ) is obtained from the 4.5Cu-720 sample, indicating the best corrosion resistance among the alloys.

3.5. Antibacterial property

Typical morphologies of S. aureus colonization after culturing with different samples for 24 h are shown in Fig. 9(a)-(d). The bacterial colonies on the petri dish of Ti6Al4V alloy are much more than those of the Cu-containing titanium alloys, indicating that the Cu-containing alloys possessed strong antibacterial abilities. As shown in Fig. 9(e), the antibacterial rate of 4.5Cu-720 sample is about 80%, and with increase of the Cu content, the antibacterial property is enhanced.

View original graphic| Download| PPT slide

Fig. 9.   Typical photographs of S. aureus colonization after culturing with samples of (a) Ti6Al4V, (b) 4.5-720, (c) 6-740, (d) 8-760, (e) calculated antibacterial rate of Cu-containing alloys and (f) absorbance of cells that cultured on different alloys for 1, 3 and 7 days.

3.6. Cytotoxicity evluation

CCK-8 essay was carried out to evaluate the cytotoxicity of all the samples. The results are shown in Fig. 9(f), indicating that at day 1 and day 3, there is no difference among the Ti6Al4V-xCu alloys and the Ti6Al4V alloy. The Cu-containing alloys show a good cytocompatibility. At day 7, the Ti6Al4V-7.5Cu alloy shows significant lower absorbance than that of Ti6Al4V alloy, meaning that the addition of 7.5% Cu into the Ti6Al4V alloy will deteriorate the cytocompatibility.

4. Discussion

In order to develop a new antibacterial implant material, Cu, a strong antibacterial metal element, has been added into the traditional Ti6Al4V alloy. In this study, relatively high contents of Cu, 4.5 wt%, 6 wt% and 7.5 wt%, respectively, were used for ensuring the antibacterial properties. However, high Cu addition into the Ti6Al4V alloy might lead to poor ductility, risk of toxicity and influence other crucial properties, such as wear resistance and corrosion resistance [17,23]. This study was aimed at optimizing the comprehensive properties, such as tensile mechanical properties, tribological property, corrosion resistance, antibacterial property and biocompatibility of the high Cu-containing Ti6Al4V-xCu alloys by choosing the optimal annealing heat treatment and Cu content.

The mechanical properties of Ti6Al4V-xCu alloys were optimized, especially the ductility, by optimizing the annealing temperature. Three different annealing temperatures (720, 740 and 760 °C) in the Ti₂Cu+α+β ternary phase region were chosen for this study, because after annealing in the ternary phase region, an equiaxed microstructure could be obtained, which possessed better ductility than the lamellar microstructure [28,29]. It is observed from Fig. 4(f) that all the Ti6Al4V-xCu alloys show significantly higher strength than the requirement of ISO Standard, but not all the elongations are high enough. The microstructures are crucial for the tensile properties of alloys [29]. It can be seen from Fig. 4(b) that samples with different annealing temperatures and different Cu contents show different proportions of α, β and Ti₂Cu phases in the microstructures, which can affect tensile properties of the alloy. It is generally believed [30] that in titanium alloys the α phase is softer than the β phase, and the precipitated intermetallic compound, Ti₂Cu, is much harder than both α and β phases. Slip of dislocations is easy to initiate in the α phase but is pinned by the harder phase. Therefore, more proportions of β and Ti₂Cu phases mean stronger resistance to the dislocations movement, which contributes to higher strength, but lower ductility. For the Cu-containing Ti6Al4V-xCu alloys, with increase of the annealing temperature from 720 to 760 °C, the ductility is influenced by the transformation of α+Ti₂Cu→β, because the volume fraction of β phase is increased and that of the Ti₂Cu phase is decreased at the same time, which show a contradictory effect on the ductility. For the Ti6Al4V-4.5Cu alloy, when the annealing temperature is increased from 720 to 760 °C, the Ti₂Cu phases are almost dissolved, and the proportion of β phase is significantly increased, so the dominating impact factor that resulted in the decreasing trend of the ductility should be the increase of β phase. For the Ti6Al4V-6Cu alloy, when the annealing temperature is increased from 720 to 740 °C, the ductility is slightly increased, which should be mainly influenced by the dissolution of Ti₂Cu phases. When the annealing temperature is furtherly increased to 760 °C, the ductility is sharply decreased, which should be mainly caused by the significantly increased proportion of β phase. For the Ti6Al4V-7.5Cu alloy, the proportion of β phase is not increased too much when the annealing temperature is increased from 720 to 760 °C, so the ductility should be influenced by the dissolution of Ti₂Cu phase, which shows a small increasing trend. From the above analysis, it can be concluded that samples of 4.5Cu-720, 6Cu-740 and 7.5Cu-760 possess the best ductility among samples of each alloy, which meet the requirements of mechanical properties for the medical Ti6Al4V alloy bars in the ISO 5832-3:2016 Standard. Meanwhile, with increase of the Cu content, the ductility values of these three samples are decreased because the proportions of both Ti₂Cu and β phases are positive correlation with the Cu content. In the study of Sun et al. [31], the elongation values of forged Ti6Al4V alloys with an equiaxed microstructure were about 12.5%-18%, and in the study of Shi et al. [32], the elongation values of Ti6Al4V alloys with various lamellar microstructure features were range from 9.3% to 14.5%. In this study, the elongation values of the 4.5Cu-720, 6Cu-740 and 7.5Cu-760 samples are 15%, 13.4%, and 11.8%, respectively, which are comparable to the ductility of Ti6Al4V alloy.

Then the optimized samples (4.5Cu-720, 6Cu-740 and 7.5Cu-760 alloys) with the best mechanical properties were further optimized for comprehensive properties including tribological property, corrosion resistance, antibacterial property and cytotoxicity. According to the SEM observations of samples after friction tests in Fig. 7, all the alloys show dominant abrasive wear mechanism with secondary adhesive wear mechanism. The representative abrasive wear morphology is characterized by numerous ploughing grooves and many wear debris [33]. The EDS analysis on the delamination regions confirm that the detected elements came from both substrate and counter-face after the adhesive wear process. The friction tests (Fig. 5) indicate that all the Cu-containing samples possess better friction resistance than that of the Ti6Al4V alloy. For the samples of Ti6Al4V, 4.5Cu-720 and 6Cu-740, the friction resistance of samples is in the following order: 6Cu-720 > 4.5Cu-740 > Ti6Al4V, indicating that the friction resistance is enhanced with increase of the Cu content. It was reported that the wear process of ductile materials was often accompanied with extensive plastic deformation, so materials with higher yield strength could have stronger resistance to plastic deformation, which contributed to the better tribological characteristics [34]. Therefore, for the samples of Ti6Al4V, 4.5Cu-720 and 6Cu-740, higher Cu content in the alloy results in higher strength and stronger resistance to plastic deformation, which enhances the friction resistance. However, when the Cu content is increased to 7.5%, although the friction coefficient of the alloy is lower than that of the Ti6Al4V alloy, it shows more severely damaged wear scars (Fig. 6, Fig. 7) than another two Cu-containing alloys. The wear scar of 7.5Cu-760 alloy shows more obvious abrasive wear behavior, with more extensively damaged regions than both 4.5Cu-720 and 6Cu-740 samples, indicating that the Cu content should not be too high for the alloy. From Fig. 7(d1), it can be observed that there are many micro-cracks next to the damaged regions. Cracking and wear particles formation are two main kinds of damage mechanisms for the materials in fretting wear [35]. As mentioned above, in the 7.5Cu-760 alloy, the proportion of the hard phase, Ti₂Cu, is the highest among the three Cu-containing alloys. Hard phases can enhance the strength of the alloy, but cracks are also easy to initiate from the hard phase, and propagate along the voids that were formed due to the interfacial decohesion [36]. Therefore, the excessive amount of Ti₂Cu phases in the 7.5Cu-760 alloy has a deleterious effect on the wear resistance. In conclusion, the 6Cu-740 alloy, with the Cu content of 6%, shows the best wear resistance.

The samples of 4.5Cu-720, 6Cu-740 and 7.5Cu-760 were used to study the effect of higher Cu content on the corrosion resistance of the alloy. According to the potentiodynamic polarization and EIS tests, 4.5Cu-720 and 6Cu-740 samples show more positive E_corr, decreased I_corr, and higher polarization resistance R_p than the Ti6Al4V alloy. This result is also in agreement with other previous studies [24]. Zhang et al. [30], which confirmed that the Ti-Cu alloy showed better corrosion resistance than the pure titanium because Cu could suppress the dissolution rate of the alloy [37,38]. For the 7.5Cu-760 sample, although its polarization resistance R_p is higher than that of the Ti6Al4V alloy, its uniform corrosion resistance is deteriorated, with a significantly higher I_corr value than those of 4.5Cu-720 and 6Cu-740 samples. Meanwhile, for the 7.5Cu-760 sample, the protectiveness of its passive film is the weakest among all the alloys, with the lowest E_p and the highest I_p. This is because that excessive addition of Cu (7.5%) results in more proportion of β and Ti₂Cu phases, which act as the cathode compared to α phase, and therefore the effect of galvanic corrosion among the three phases of α, β and Ti₂Cu is accelerated [39]. As a result, only when the Cu content is lower than 6 wt.%, the corrosion resistance of alloy could be better than the commercial Ti6Al4V alloy.

The results of antibacterial test indicate that with increase of the Cu content, the antibacterial ability of the alloy is enhanced. Many researchers have reported [22,40] that the Ti₂Cu phase played an important role in killing the bacteria and inhibiting the formation of bacterial biofilm. Therefore, a higher volume fraction of Ti₂Cu phase contribute to a better antibacterial property in this study. But when the Cu content is increased to 7.5%, the cytotoxicity of the alloy is also presented. It is probably because that according to the polarization curve, the corrosion rate of 7.5Cu-760 sample is significantly higher than those of the other alloys that contain lower Cu contents, and then more metal ions are release from the alloy, thereby causing cytotoxicity.

5. Conclusion

Annealing treatments and comprehensive properties of Cu-containing Ti6Al4V-xCu alloys were optimized in this study. According to the results, it can be concluded that the tensile properties of 6Cu-740 sample meet the requirements of the ISO 5832-3:2016 Standard for the implant material of Ti6Al4V alloy. Meanwhile, 6Cu-740 sample possesses better wear resistance and corrosion resistance than the commercial Ti6Al4V alloy, as well as strong antibacterial effect and low cytotoxicity, which all show great potential for medical applications.

Acknowledgments

This work was financially supported by the National Key Research and Development Program of China (Nos. 2018YFC1106600 and 2016YFC1100600), the Innovation Fund Project of Institute of Metal Research, Chinese Academy of Sciences (No. 2017-ZD01), the National Natural Science Foundation (Nos. 51631009 and 51811530320) and Key Projects for Foreign Cooperation of Bureau of International Cooperation Chinese Academy of Sciences (No. 174321KYSB2018000).