2025 Volume 66 Issue 5 Pages 616-621
Titanium has the highest strength-to-weight ratio of any metal. Titanium and its alloys are strong and lightweight, and are therefore used in many industrial applications. The ω-phase is known to form metastably during quenching and aging processes of titanium alloys, and the presence of this phase causes increases in strength and decreases in ductility. However, the mechanical properties of the ω-phase are poorly measured because this phase precipitates as nanograins in titanium alloys. Here we report the fabrication of bulk polycrystalline ω-phase samples in pure titanium under high pressure and temperature conditions of 12 GPa and 400°C. They are single-phase and randomly oriented polycrystalline materials with an average grain size of 21 ± 8 µm. The 0.2% offset yield strength was determined from compressive stress-strain curves to be 913 ± 3 MPa. The ω-phase is more than twice as strong as the α-phase in pure titanium and could be used as a non-toxic structural material for biomedical applications.
Bulk polycrystalline ω-phase samples in pure titanium were fabricated under high pressure and temperature conditions. The ω-phase is more than twice as strong as the α-phase.
Titanium crystallizes into a hexagonal closed-packed structure at ambient conditions, which is referred to as the α-phase. The α-phase transforms into a body-centered cubic structure, known as the β-phase, under high temperature. The α-phase is stronger than the β-phase, which exhibits better deformability and a lower Young’s modulus [1]. It is possible to control the mechanical properties of titanium alloys by annealing and aging treatments, which allow for the coexistence of these two phases. In certain instances, the metastable precipitation of the ω-phase with a simple hexagonal structure as nanograins occurs after undergoing the heat treatment processes [2]. The presence of the ω-phase is undesired as it reduces the ductility of titanium alloys, which is referred to as ω-phase embrittlement [3].
It is necessary to measure the physical properties of a single phase of the ω-titanium in order to comprehend those of titanium alloys containing the ω-phase. Since the ω-phase exhibits thermodynamic stability under high pressure [4, 5], some high-pressure experimental studies have been conducted to produce pure ω-phase titanium. Todaka et al. [6] utilized high-pressure torsion (HPT) processing under 5 GPa to severely deform a pure α-phase titanium. They obtained 90 mass% polycrystalline ω-phase, which was reported to be harder than the α-phase. Tane et al. [7] used the same technique to fabricate polycrystalline pure titanium specimens exclusively consisting of the ω-phase. A complete set of elastic stiffness of the ω-phase was successfully measured. Recently, Li et al. [8] employed a static high-pressure technique to synthesize a bulk specimen consisting of single-phase ω-titanium at a pressure of 12.2 GPa and room temperature. A specimen with dimensions of 3 mm × 3 mm × 4 mm was obtained and subjected to Vickers hardness and elastic modulus measurements. While the hardness values and elastic properties of the ω-phase have already been reported in the previous studies, the mechanical strength has not yet been reported. The potential of polycrystalline ω-titanium as a structural material has not been demonstrated.
In the present study, we fabricated bulk polycrystalline ω-phase samples consisting of randomly-orientated grains using a Kawai-type multianvil high-pressure apparatus [9] at simultaneous high pressure and high temperature. Compression tests were conducted and the stress-strain curves were obtained. Analysis of the data allowed for determination of the 0.2% offset yield strength and comparison with that of the α-phase.
High pressure and high temperature synthesis runs were conducted using a Kawai type apparatus [9] with a Walker module (mavo press LPR 1000 400/50, Max Voggenreiter GmbH, Mainleus, Germany), which is installed in Laboratory for Materials and Structures, Tokyo Institute of Technology, Japan. The maximum force of this instrument is 1000 tonf. The second stage anvils were tungsten carbide cubes with a truncated edge length of 11 mm. An 18 mm edge-length octahedron of magnesia doped with Cr2O3 was used as the pressure transmitting medium. A cylindrical LaCrO3 furnace (Nikkato Corp., Osaka, Japan) was employed, and a sample container made of MgO was embedded into the furnace. The starting samples were commercially pure (CP) titanium rods (99.5 mass% in titanium, CP Grade 2) with dimensions of 4 mm in diameter and 3 mm in height, which were purchased from the Nilaco Corporation, Tokyo, Japan. No heat treatment was applied before the high pressure process. Frist, the starting sample was compressed up to 12 GPa at room temperature and then the temperature was increased to 400°C. High-pressure and temperature conditions were maintained for 3 hours. After cooling to room temperature, the run product was recovered through a 2-hour decompression process.
2.2 X-ray diffraction measurementsX-ray diffraction (XRD) measurements were performed with two different geometries, reflection and transmission. The reflection mode measurements were conducted using a benchtop X-ray diffractometer (Miniflex600, Rigaku, Tokyo, Japan) with a Cu anode (X-ray energy, 8.048 keV). The flat surfaces of the rod-shaped samples were analyzed for phase identifications. The transmission measurements were performed at BL08W, SPring-8, Japan [10]. An energy of the incident X-rays was 115 keV (λ = 0.1077 Å) and the diffracted X-rays were recorded using a 16 inch flat-panel detector (XRD 1621 CN3, PerkinElmer Inc., CA, USA). The sample to detector distance (795 mm) and the orientation parameters of the detector were calibrated using a NIST CeO2 standard (NITS 674b). An XRD pattern was collected by averaging 15 frames with an exposure time of 1 s per frame.
2.3 Density and sound velocity measurementsThe densities of recovered samples were determined by measuring their volumes and masses. The volume was measured using a He gas pycnometer (AccuPyc II 1340, Micromeritics Inst. Corp., GA, USA), while the mass was measured using an analytical balance. Travel time of ultrasonic waves through the samples was measured by a pulse echo overlap method [11]. A LiNbO3 transducer was bonded to one side of an alumina buffer rod to generate both longitudinal and transverse acoustic waves. On the opposite end of the buffer rod, the starting material or the recovered sample after chemical polishing was attached. Prior to measuring sound velocities, the length of the sample was measured using a micrometer.
2.4 Mechanical property measurementsVickers hardness (HV) was measured using a microhardness indenter (HM 221, Mitsutoyo, Kawasaki, Japan). One of the flat surfaces of the rod-shaped sample was indented for the measurement. A hard steel standard with 500 HV was used. The holding time under the indentation load was 15 s. HV was calculated using the following equation: HV = 1.8544 P/d2, where P is the applied load (N) and d is the arithmetic mean of the two diagonals (µm) of a Vickers indentation trace.
A rod-shaped specimen (5 mm long and 2 mm in diameter) was used for uniaxial compression testing. The testing apparatus used in the present study was a universal testing machine (INSTRON 8516, Instron, MA, USA). A load cell was used to measure the applied load. A laser displacement sensor was used to measure the sample length during deformation.
X-ray diffraction (XRD) measurements were performed on the starting materials and samples recovered from high-pressure (hereafter called “recovered samples”). The results of reflection mode measurements are shown in Fig. 1. The starting material exhibited the α-phase (Fig. 1(a)), which transformed into the ω-phase by the high-pressure and temperature treatment. After recovery, we observed the presence of thermodynamically metastable ω-phase with magnesia that was used as the sample capsule (Fig. 1(b)). The magnesia materials were removed by grinding the sample surfaces with P800 silicon carbide abrasive paper. Subsequently, an XRD pattern was collected, revealing the coexistence of α- and ω-phase (Fig. 1(c)). This result suggests that a shear stress induced the back transformation from ω to α-phase on the sample surfaces during gliding. To remove the transformed surface layer, we conducted a chemical polishing procedure using a commercially available liquid containing a nitric hydrofluoric acid mixture (S-CLEAN S-22, Sasaki Chemical Co., Ltd., Kyoto, Japan). Subsequent to the chemical polishing, we observed the existence of only the ω-phase on the surfaces of the recovered samples (Fig. 1(d)).
X-ray diffraction profiles of bulk polycrytalline ω-phase in pure titanium fabricated at 12 GPa and 400°C. The data were obtained by the reflection mode (see text). (a) starting material; (b) as recovered; (c) after mechanical polishing to remove magnesia sample capsule; (d) after chemical polishing. A photograph of bulk polycrystalline ω-Ti after chemical polishing is shown as an inset of (d). (online color)
Results of high-energy XRD measurements with transmission geometry were are shown in Fig. 2. The measurements were made on both the starting material (α-Ti shown in Fig. 1(a)) and the recovered sample obtained after chemical polishing (ω-Ti shown in Fig. 1(d)). The 2D-XRD pattern in Fig. 2(a) indicates that the starting α-Ti material displays lattice preferred orientation, thought to have developed during the manufacturing process of CP Grade 2 titanium. The integrated 1D-XRD profile is also shown in Fig. 2(b). Furthermore, Fig. 2(c) shows a 2D-XRD pattern of the ω-phase. The diffracted peak intensities were almost constant, regardless of the azimuth angle, which suggests that the microstructure of the starting material (α-phase) vanished through the nucleation and growth process of the ω-phase in its stability filed under high pressure and temperature. The 1D-XRD pattern (Fig. 2(d)) confirms that only the ω-phase remains inside the recovered sample.
X-ray diffraction data obtained by the transmission mode (see text). (a) a 2D profile of the starting material (polycrystalline α-Ti); (b) the integrated profile of the 2D profile shown in (a); (c) a 2D profile of bulk polycrystalline ω-Ti (a recovered sample after chemical polishing); (d) the integrated profile of the 2D profile shown in (c). A schematic illustration of the transmission XRD geometry is shown as an inset of (b). (online color)
Figures 3(a) and (b) show optical micrographs of the starting material and the recovered sample after the chemical polishing, respectively. One of the flat surfaces of the rod-shaped sample was observed. The average grain sizes were measured by a liner intercept method. The measured values of the starting (α-phase) and the recovered (ω-phase) samples are 31 ± 11 and 21 ± 8 µm, respectively. These observations confirm the successful production of bulk polycrystalline materials in which the ω-phase has micrometer-sized grains with no preferred orientation (Fig. 2(c)). This was achieved through high pressure and temperature synthesis, mechanical grinding, and chemical polishing.
Optical micrographs of polished surfaces of polycrystalline Ti samples. (a) α-Ti as the starting material; (b) ω-Ti after chemical polishing. The average grain sizes of the starting (α-phase) and the recovered (ω-phase) samples are 31 ± 11 and 21 ± 8 µm, respectively. (online color)
We measured the density (ρ), longitudinal (VP), and transverse (VS) velocities of both the α-phase and ω-phase materials (Table 1). Elastic moduli (bulk, shear, Young’s moduli, and Poisson’s ratio) at ambient conditions were then calculated using the obtained values, and the results are summarized in Table 2. The bulk modulus of the ω-phase was found to be similar to that of the α-phase [12], while the shear and Young’s moduli were about 1.5 times greater than those of the α-phase [12]. A decrease in Poisson’s ratio is observed with the α-ω transition. The present measurements are in agreement with earlier experimental findings [7, 8] and first-principles calculations [13].
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/MT-MC2024015tab01.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
;%0A%09%09%09newWindow.document.open();%0A%09%09%09newWindow.document.write('<img src=%22./Graphics/MT-MC2024015tab02.jpg%22>');%0A%09%09%09newWindow.document.close();%0A%09%09)
Vickers hardness of both the α-phase and ω-phase materials were evaluated as a function of indentation load. Results show that the hardness values decrease slightly with indentation load. At a load of 4.9 N, the α-phase and ω-phase materials possess hardness values of 1.51 ± 0.05 and 2.41 ± 0.07 GPa, respectively. The present value obtained for the ω-phase is consistent with that reported in the Li et al. study [8], but is lower than that reported by Todaka et al. [6] (Table 2). This difference can be partially attributed to variations in grain size. The specimens examined in the present study and by Li et al. [8] were fabricated using static high-pressure technique, resulting in microcrystalline grains measuring 21 ± 8 µm and 5–10 µm, respectively. In contrast, the grain size of the specimens subjected to severe plastic deformation by HPT processing was a few hundred nanometers [6] or smaller [14]. The smaller grain size may have resulted in Hall-Petch strengthening [15, 16].
To investigate the mechanical strength of the ω-phase in pure titanium, uniaxial compression tests were performed on both the α-phase and ω-phase materials at a constant strain rate of 6.7 × 10−4 s−1. The stress-strain curves are depicted in Fig. 4. The 0.2% offset yield strength of the ω-phase was found to be 913 ± 3 MPa. After yielding, the ω-phase sample exhibits a steady state flow in which the stress remains almost constant at around 950 MPa with strain, while the α-phase exhibits work-hardening behavior. The bulk polycrystalline ω-phase sample is more than two times stronger than the α-phase in pure titanium. We terminated the deformation of the ω-phase when the nominal strain exceeded 5.5%, and we released the applied load. No fracture was detected when the deformed ω-phase specimen was recovered.
Stress-strain curves of α- and ω-phases in pure titanium under compression. The 0.2% offset yield strength of the α- and ω-phases were determined to be 330 and 913 ± 3 MPa, respectively. The latter value is the average of three measured values obtained from independent tests with one standard deviation, whereas the former was obtained from a single test. (online color)
Commercially pure (CP) titanium is a commonly used biomaterial [17] due to its excellent corrosion resistance and non-cytotoxicity, which make it compatible with both bone and soft tissue. However, its mechanical strength is relatively low, approximately 300 MPa for CP-Ti Grade 2 [18]. Consequently, titanium alloys with higher strength such as Ti-6Al-4V have been employed [19] in cases where higher loads must be supported, such as in artificial joints and bone fixators. Nevertheless, this titanium alloy also has limitations, because vanadium present in this alloy causes cytotoxicity [20]. Accordingly, since the 1960s, CP titanium has been the preferred material for dental implant devices [21]. The present study used CP titanium as the starting material and fabricated bulk polycrystalline ω-phase materials with strength exceeding twice that of the α-phase. Hence, the bulk polycrystalline ω-phase material has the potential to serve as a nontoxic biomaterial with high strength. Additional studies are needed to assess the biocompatibility of the ω-phase material.
Ductility is a critical property of CP titanium as a biomaterial. CP titanium shows an elongation to failure of about 30%, while the Ti-6Al-4V alloy exhibits limited ductility of only around 10% [22]. The ductility of the bulk polycrystalline ω-phase material remains unknown. However, in the present study, we observed that back transformation from the ω-phase to the α-phase occurs after the mechanical grinding (Fig. 1(c)), which could be connected to this issue. This result suggests that stress triggers the reverse transformation (from high-density ω phase to low-density α phase) (see Table 1). It is widely known that stress-induced transformation from a high- to a low-density phase results in improved ductility in steel [23] and toughening in zirconia ceramics [24]. Garvie et al. [24] observed that the transformation from a tetragonal to a less-dense monoclinic phase on ground and polished surfaces of zirconia leads to a significant increase in transverse rupture strength. Additionally, previous studies [25–27] have reported that the transformation from stishovite, which is a high-pressure phase in silica [28] and the hardest oxide [29], to a less-dense amorphous phase causes toughening. Our observation of the ground surfaces of the bulk ω-phase material is similar to these phenomena observed in zirconia and stishovite. Recently, Zahiri et al. [30] theoretically predicted that stress-induced ω → α martensitic transformation could lead to substantial plasticity of the ω-phase in titanium. They referred to this phenomenon as transformation-induced plasticity in ω-titanium. It is crucial to perform tensile testing on bulk polycrystalline ω-phase titanium to investigate its elongation property.
Bulk polycrystalline ω-phase samples in pure titanium were fabricated at a pressure of 12 GPa and a temperature of 400°C. The starting samples were commercially pure (CP) titanium rods (99.5 mass% in titanium, CP Grade 2) with dimensions of 4 mm in diameter and 3 mm in height. The ω-phase samples are single-phase and randomly oriented polycrystalline materials with an average grain size of 21 ± 8 µm. The 0.2% offset yield strength was determined from compressive stress-strain curves to be 913 ± 3 MPa. The ω-phase is more than twice as strong as the α-phase in pure titanium. Additional studies are needed to measure tensile strength and elongation to failure of the ω-phase material. Biocompatibility of the ω-phase should be assessed and this material could be used as a non-toxic structural material for biomedical applications.
We thank T. Hanawa, H. Hosoda, and A. Holtzheid for discussion. We also thank E. Oshinoya, E. Yoshida, A. Ushijima, and Y. Nakamura for technical assistance. This research was supported by a Grant-in-Aid for Scientific Research on the innovation area “Science of New-Class of Materials Based on Elemental Multiplicity and Heterogeneity (Grant No. 18H05452)” partially to N.N.
