Tribological and corrosion behaviors of warm- and hot-rolled Ti-13Nb-13Zr alloys in simulated body fluid conditions

Introduction

Titanium and its alloys are regarded as one of the most promising materials for orthopedic biomedical implants and are considered as an alternative to conventionally used stainless steel due to their low elastic modulus, low density, higher corrosion resistance, and superior biocompatibility.^1,2 The conventional Ti-6Al-4V alloy, which falls under the category of (α+β) alloy, is extensively used in the fabrication of hip, knee, heart, and spinal implants due to high specific strength and corrosion resistance. Nevertheless, this material has three serious problems.³ First, it possesses a higher modulus of elasticity (110 GPa) compared to bone tissue (20–30 GPa) leading to stress shielding effect. Second, alloying elements Al and V were reported to have toxicity associated with Alzheimer’s disease. Third, a poor wear resistance of titanium and its alloys is another drawback that limits their usage as load-bearing components in the biomedical industry. Moreover, the accumulation of wear debris between tissue and implant induces an adverse cellular response, inflammation, and loosening of implant due to the process called osteolysis.^4,5

A number of methods have been suggested to enhance biomaterial performances of titanium and its alloys, such as alloy design, powder metallurgy, and thermomechanical process.⁶ In particular, the development of β- and near-β-type titanium alloys, such as Ti-Mo, Ti-Nb-Zr, Ti-Mo-Zr-Fe, and Ti-Nb-Zr-Ta systems, are considered as a better alternative as they show a low elastic modulus (45–90 GPa), high corrosion resistance, and biocompatibility due to their alloying elements.⁷ Some of these alloys have already been approved by American Society for Testing and Materials (ASTM) and/or US Food and Drug Administration (FDA) for an orthopedic implant material.⁸ Amongst various β- and near-β-type titanium alloys, Ti-13Nb-13Zr alloy has been extensively studied in the aspect of thermomechanical processing for implant applications.^9–11 Conventionally Ti-13Nb-13Zr is hot-rolled (HR) above the β-transus temperature, but recent research suggests the potential of warm-rolling performed below the β-transus temperature with improved mechanical properties.^11,12

Although the mechanical properties were found to be improved by the warm-rolling, until now no investigations have been done on the effect of this process on tribological and corrosion behaviors despite the importance of the topic. High resistance to corrosion is another essential factor for an implant material as the corrosion deteriorates mechanical properties and may lead to periprosthetic bone loss.⁵ Thus, the present work aims at clarifying the tribological response and corrosion resistance of Ti-13Nb-13Zr alloy manufactured by the warm- and conventional hot-rolling processes.

Material and methods

ASTM F1713-08 Ti-13Nb-13Zr (in wt%) alloy was prepared in two ways for this study: warm-rolled (WR) or conventionally HR conditions. First, Ti-13Nb-13Zr rods, 28 mm in diameter, were purchased from Hyundai Titanium Co., Incheon, Republic of Korea. They were solution-treated at 800°C for an hour followed by water quenching. The material was then reheated at 650°C for an hour and WR at the same temperature using a multi-pass caliber-rolling process.¹¹ The reduction of area was calculated to be 90%. Second, Ti-13Nb-13Zr plates were manufactured using the vacuum-arc technique at Defense Metallurgical Research Laboratory, India. The material was melted in the disc shape of 110 mm in diameter and 12 mm thickness in non-consumable vacuum arc furnace using titanium sponge, niobium powder, and zirconium chips. The melting was carried out in copper crucible under the partial pressure of ultra-high purity argon gas. This process was repeated five to six times flipping the disc in order to ensure chemical homogeneity. Afterward, the disc-shaped sample was HR from 800°C to 650°C with a thickness reduction of 63% followed by air cooling.

The microstructural analysis was carried out using scanning electron microscopy (SEM; JSM-6330F, JEOL, Tokyo, Japan) with samples which were chemically etched in Kroll’s reagent consisting of 2 mL HF, 2 mL HNO₃, and 96 mL H₂O to reveal the microstructures and wear behaviors. Electron backscatter diffraction (EBSD; Quanta 3D FEG, FEI, Hillsboro, OR, USA) analysis with electropolished samples was conducted to reveal grain and phase structures. The data were analyzed with Orientation Imaging Microscopy Data Analysis (OIM; EDAX Inc., Mahwah, NJ, USA) (eg, the average grain size, d). X-ray diffraction (XRD; Bruker XRD D8-Advance, Bruker Corporation, Billerica, MA, USA) analysis was performed at room temperature to confirm phase constituents with 2.2 kW Cu anodes and a scan rate of 0.278°/min.

The hardness was evaluated using a standard Vickers hardness tester with a load of 500 gf and dwell time of 10 seconds with mirror-polished samples. Hank’s solution was prepared in laboratory for evaluating the wear behavior. The solution consists of (0.185 g CaCl₂ [1.258 mol], 0.4 g KCl [0.00536 mol], 0.06 g KH₂PO₄ [0.0004 mol], 0.1 g MgCl₂·6H₂O [0.00049 mol], 0.1 g MgSO₄·7H₂O [0.00041 mol], 8.0 g NaCl [0.137 mol], 0.35 g NaHCO₃ [0.00417 mol], 0.48 g Na₂HPO·12H₂O [0.00269 mol], and 1.00 g D C₆H₁₃NO₅ HCl [0.004637 mol]). This solution was made up to 1,000 mL with distilled water. The wear test was performed using automated ball-on-plate reciprocating wear tester (TR-285M1, Ducom Instruments, Karnataka, India) with sliding motion against alumina balls of 6 mm diameter. Specimens of 30 mm ×20 mm ×4.5 mm were mechanically polished and ultrasonically cleaned with acetone prior to the experiment. The weight loss was measured with an electronic weighing machine (Adventure Pro AVG 264, Ohaus Corporation, Parsippany, NJ, USA) of 10⁻⁴ mg of accuracy to evaluate the wear rate.

Corrosion behaviors were investigated in Hank’s solution with 10 mm ×10 mm ×4.5 mm specimens using a potentiostat (Gill AC, ACM Instruments, Cumbria, UK). The saturated calomel plate and platinum foil were used as the reference and counter electrodes, respectively. Open-circuit potential test was performed for an hour to achieve a steady potential, and then potentiodynamic polarization test was started from an initial potential of 250 mV below the corrosion potential. The scan rate was 0.166 mV/s as per ASTM F2129 standards. The polarization tests were repeated three times for each specimen for reproducibility of the data.

Results and discussion

Figure 1 presents the SEM and EBSD data of the WR and HR alloys. The micrographs clearly revealed the presence of α and β phases, while martensitic laths were not confirmed. The WR alloy exhibited a significantly fine microstructure oriented along the rolling direction (RD), in which ultrafine equiaxed grains (d=0.3 μm) were formed. This suggests that the submicrocrystalline structure has been formed by the warm-rolling process of Ti-13Nb-13Zr. In contrast, the HR alloy exhibited a typical coarse lamellar structure and elongated grains, of which the thickness was measured to be 0.5 μm. It is obvious that its actual grain size is significantly larger compared to that of the WR Ti-13Nb-13Zr. Figure 2 demonstrates the XRD results revealing the (α+β) structure for both alloys, which is in good accord with the EBSD results. The martensitic transformation would be suppressed due to the diffusion of niobium atoms into the β phase.¹¹ It is interesting from the XRD data that the alloys demonstrated the different texture behaviors. The WR plane normal to the RD exhibited the strong (110)β peak. Such a texture development is typically observed in caliber-rolled body-centered cubic alloys.¹³ The (0002)α and (10-12)α peaks were significantly weak in this plane in contrast to the planes parallel to the RD. This type of texture can be observed in pure titanium and interpreted in terms of slip/twin behavior in a hexagonal close-packed system.^14,15

Figure 1 Microstructure of the investigated alloys. Notes: SEM micrograph of (A) WR and (B) HR Ti-13Nb-13Zr, and EBSD grain/phase map of (C) WR and (D) HR Ti-13Nb-13Zr. The horizontal axis is the RD for all images. Alpha and beta indicate each phase, respectively. Abbreviations: SEM, scanning electron microscopy; WR, warm-rolled; HR, hot-rolled; EBSD, electron backscatter diffraction; RD, rolling direction.

Figure 2 XRD line profiles of the WR and HR Ti-13Nb-13Zr. Note: The inset is the magnified image ranging from 37° to 39°. Abbreviations: XRD, X-ray diffraction; WR, warm-rolled; HR, hot-rolled; RD, rolling direction.

It is noted that the WR Ti-13Nb-13Zr exhibited the superior tensile properties in comparison to other forms of Ti-13Nb-13Zr alloy as discussed in our previous work; the WR alloy exhibited a tensile stress of 1,050 MPa, while both HR alloy and ASTM Ti-13Nb-13Zr possessed the value of 850 MPa.^16,17 Moreover, the WR alloy exhibited the higher hardness (HV 335±5) than the HR alloy (HV 215±4) as well as the solution-treated and cold-rolled sample (HV 285±7).¹⁸ This is understood by the conclusion that Ti-13Nb-13Zr exhibits a higher hardness when it consists of the finer microstructural features.¹⁹ These results prove the great improvement in mechanical properties via the warm-rolling process.

The friction coefficient of WR and HR alloys was measured to be around 0.42–0.45, which was similar to the values reported in the literature.^18,20 The wear rate was determined by the expression, WR = (V × S × N × f)/L where V is the volume loss, S =15 mm is the stroke length, N =30,000 is the number of cycles, f =2 Hz is the frequency, and L =10 N is the applied load. The results presented no significant variation between the WR (1.05×10⁻⁵ mm³/Nm) and HR Ti-13Nb-13Zr (4.44×10⁻⁵ mm³/Nm). In other words, no straightforward relationship was confirmed between hardness and wear properties in the investigated materials, as also supported by previous works.^10,21 Figure 3 presents worn scars produced during the wear sliding, suggesting that the abrasive wear is dominant in both Ti-13Nb-13Zr alloys. Moreover, the tribocorrosion also occurred in the HR Ti-13Nb-13Zr as evidenced by the formation of dark patches observed in Figure 3B; the WR alloy did not show such a microstructural characteristic.

Figure 3 SEM micrograph of wear tracks. Notes: (A) WR and (B) HR Ti-13Nb-13Zr. Abbreviations: SEM, scanning electron microscopy; WR, warm-rolled; HR, hot-rolled.

Figure 4 provides polarization curves of the WR and HR Ti-13Nb-13Zr where the WR alloy was evaluated in two planes to investigate a texture effect on corrosion behavior. The planes parallel to the RD exhibited the stable passivation up to 1,200 mV, while the WR alloy plane normal to the RD presented the formation of thick oxide layer above 1,200 mV. The specific corrosion data are summarized in Table 1. It should be noted that both WR planes exhibited the lower corrosion current densities (1.11 μA/cm² for plane normal to the RD and 1.33 μA/cm² for plane parallel to the RD) compared to the HR alloy (1.71 μA/cm²). The corrosion current density is directly proportional to the corrosion rate,²² indicating the better corrosion resistance of the WR alloy. These results explain the fact that only the HR alloy exhibited the tribocorrosion after the wear test, because the alloy was more prone to the corrosion in the simulated body fluid.

Table 1 Results of potentiodynamic polarization tests Abbreviations: WR, warm-rolled; HR, hot-rolled; RD, rolling direction; Ecorr, corrosion potential; Icorr, corrosion current density; Rp, polarization resistance; Ipass, passivation current density.

Figure 4 Potentiodynamic curves measured in Hank’s solution for the investigated Ti-13Nb-13Zr alloys. Abbreviations: WR, warm-rolled; HR, hot-rolled; RD, rolling direction.

Corrosion resistance is influenced by various factors such as chemical composition, surface roughness, grain size, and texture.^5,10,23 It is reasonable that the grain size and texture mainly contributed to the corrosion behavior in this work as the first two factors were comparable among the investigated materials. The enhanced corrosion resistance of WR alloy is attributed to its submicrocrystalline structure (ie, ultrafine grains); the grain refinement is beneficial to corrosion resistance in case of Ti-13Nb-13Zr, although it is not applicable to other metals.¹⁰ Such an improvement would be achieved by the formation of stable oxide layers on the surface composed of a considerable amount of grain boundaries.²⁴ The different corrosion behaviors depending on the plane of WR alloy originate from the variation in developed texture. The (110) plane is one of the crystallographic groups having the highest atomic density in the body-centered cubic structure.²⁵ It is of particular interest that corrosion resistance increases with increasing atomic density of crystallographic plane,²³ which rationalizes the lowest corrosion current density of WR plane normal to the RD with the strong (110)β texture.

Conclusion

The WR Ti-13Nb-13Zr demonstrated a submicrocrystalline structure formed by the warm-rolling process, which provided a significant increase in hardness and tensile stress compared to the HR alloy composed of coarse elongated grains. Although there was no notable difference in wear resistance between the two alloys, the WR Ti-13Nb-13Zr exhibited a better resistance to corrosion in the simulated body fluid conditions. The improvement was attributed to the submicrocrystalline structure with atomically dense crystallographic plane developed by the warm-rolling. The present study confirmed the enormous potential of WR Ti-13Nb-13Zr with not only great mechanical properties but also high corrosion resistance in the simulated body fluid.

Acknowledgment

The authors would like to thank Dr Fathima Zivic for her technical support.

Disclosure

The authors have no conflicts of interest to disclose.

References