Strengthening Mechanism of Ti-Zr Sintered Alloys with Sc Addition

Abstract

In this study, Ti-Zr-Sc sintered alloys with excellent biocompatibility were fabricated from the elemental mixture of pure Ti, ZrH₂ and ScH₂ powders, and their microstructures and mechanical properties were investigated to clarify the strengthening mechanism. In particular, the effect of scandium (Sc) used as the third alloying element on the strengthening behavior by grain refinement, oxygen (O) and Sc solid solution, and Sc₂O₃ particle dispersion was quantitatively evaluated by using the theoretical strengthening models. 0.2% YS of Ti-Zr-Sc alloys decreased with Sc content in the range of 0∼1.0 at.% Sc, and increased in the range of 1.0∼2.5 at.% Sc. In the former, the added Sc elements reacted with O solutes to form Sc₂O₃ particles and resulted in a significant decrease of O solid solution strengthening effect. On the other hand, when Sc content was over 1.0 at.%, the strengthening effects by both Sc solid solution and Sc₂O₃ dispersion were effective, and cause a remarkable increment of 0.2% YS of Ti-Zr-Sc sintered alloys.

 

This Paper was Originally Published in Japanese in J. Jpn. Soc. Powder Powder Metallurgy 71 (2024) 482–491, https://doi.org/10.2497/jjspm.23-00067. The citation in reference 35 is corrected.

1. Introduction

Titanium (Ti) has superior corrosion resistance and biocompatibility compared to other light metals and is used as a biomedical material. For example, implants placed in the body require the function of bone to support large loads. If the stiffness of the implant material is significantly greater than that of the bone, stress shielding occurs, preventing the transfer of loading stresses to the surrounding bone. As a result, bone atrophy occurs, leading to a decrease in bone density and an increased risk of fracture. For this reason, β titanium alloys with relatively low stiffness [1, 2] are commonly used. On the other hand, since an implantable artificial heart acts as a pump to deliver blood to the body, it is required to down-sizing (ultra-miniaturization) of the device in addition to maximizing and stabilizing blood flow. Therefore, in addition to high strength and high ductility, high stiffness is required for the metallic material for them, and application of α or near-α titanium alloys is desirable. Our group has focused on the fact that zirconium (Zr), which has excellent biocompatibility like Ti, is a all-proportional solid solution element of Ti [3], and in the α-phase temperature range of about 800 to 850°C, its self-diffusion coefficient is about 10² lower than that of Ti [4]. We investigated the solid solution strengthening and grain boundary strengthening by solute drag effect [5] by preparing Ti-Zr PM alloys. Specifically, in order to solve the problem of segregation of the Zr component near the grain boundary due to its superior thermal stability, we attempted to optimize the sintering and heat treatment conditions that contribute to uniform solid solution of Zr. As a result, high strength has been achieved based on solid solution strengthening by Zr solute and α-Ti grain refinement [6, 7]. In addition, Ti-Zr alloys were prepared by a laser powder bed fusion method, and the relationship between their microstructure and mechanical properties, as well as the plateau behavior during compressive deformation in porous structures, were investigated in detail [8, 9].

Based on the previous results, this study investigates the addition of a third element to Ti-Zr-based PM alloys to increase their strength. Specifically, scandium (Sc) was selected as the additive element because it is non-toxic to living organisms, as no cytotoxicity was observed in Sc-added Mg alloys for biodegradable implants [10, 11]. According to the Ti-Sc equilibrium phase diagram [12], solid solution strengthening by Sc is expected, as the maximum solubility limit of Sc in α-Ti is about 7 at.%. Therefore, we quantitatively analyzed the effect of Sc on the strengthening mechanism of Ti-Zr-Sc PM alloys by analyzing the microstructure and evaluating the mechanical properties. It is believed that part of the Sc reacts with oxygen solute in the Ti matrix to form a chemically stable oxide (Sc₂O₃) [13]. Therefore, the solid solution strengthening of the α-Ti phase by Sc solute atoms and the dispersion strengthening of the above oxide particles are quantitatively analyzed. When the metal Sc [14], which is extremely active like Zr, is mixed with Ti powder under dry conditions, there is concern about ignition phenomena due to rapid oxidation reactions. Therefore, we performed hydrogenation heat treatment on a small piece of metallic Sc to produce a hydride (ScH₂) that can be safely handled and used as a raw powder.

2. Experimental Procedure

2.1 Preparation of ScH₂ particles by hydrogenation heat treatment

First, a small piece of metallic Sc of about 5 mm (99% purity, Kojundo Chemical Laboratory Co., Ltd.) was cut into chips of 1 mm or less in length, and about 50 g of the Sc chips was evenly distributed in an alumina container and placed in a tube furnace (ARF-2-500, Asahi Rika Manufacturing Co., Ltd.) for heat treatment in a hydrogen gas atmosphere. The Sc chips were heated to 600°C at a rate of 20°C per minute and held at the same temperature for 2.4 ks in H₂ flow (flow rate: 1 L/min). The samples were then cooled in an argon (Ar) gas atmosphere (flow rate: 1 L/min). It was difficult to completely synthesize ScH₂ to near the bottom of the container in a single hydrogenation heat treatment. Therefore, the target ScH₂ sample was prepared by stirring the sample after cooling, and then performing hydrogenation heat treatment again under the same conditions. Then, 60 g of the ScH₂ sample was placed into a plastic bottle with a 10-mm-diameter ZrO₂ media ball (YPZ ball, Nikkato Co., Ltd.) of the same weight, followed by dry milling using a rocking mill (RM-05, Seiwa Giken Co., Ltd.) for 3.6 ks. ScH₂ powder with a maximum particle diameter of 45 µm or less were then prepared using a stainless steel sieve (45 µm aperture).

2.2 Preparation of Ti-Zr-Sc PM alloy

In addition to the above ScH₂ powder, CP-Ti powder (99.7% purity, TILOP-150, Osaka Titanium Technologies Co., Ltd.) and ZrH₂ powder (Purity 97.8%, Mitsuwa chemicals Co., Ltd.) were used as starting material. Particle size distribution measurements using a laser diffraction/scattering particle size analyzer (LA-950, Horiba, Ltd.) showed that the median diameter D₅₀ of each raw material was 73.3 µm for pure Ti, 7.17 µm for ZrH₂, and 20.9 µm for ScH₂. Based on previous research results [7], the composition of the base metal was Ti-10 wt.% Zr, and the Sc content was 0.0, 0.5, 1.0, 1.5 and 2.5 at.% (hereinafter referred to as Ti-10%Zr+(0∼2.5 at.%)Sc). They were then mixed by dry mixing using a rocking mill. The weight ratio of the mixed powder to ZrO₂ media balls was 10:1, and the mixing time was set to 3.6 ks.

Next, sintered compacts with a relative density of 98% or higher were prepared by solidifying each powder mixture using an SPS (SPS-1030S, SPS SYNTEX, Inc.) under the following conditions: temperature 1000°C, heating time 3.6 ks, applied pressure 30 MPa, and vacuum <6 Pa. The sintered compacts were heat-treated at 1000°C for 3.6 ks (vacuum <100 Pa) in a vacuum furnace (FT-1200R-250, Furutech Co, Ltd.) to completely remove hydrogen derived from ZrH₂ and ScH₂ used as starting materials. As in the previous study [7], the sintered compacts were then heated at 1000°C for 1.8 ks in an Ar gas atmosphere using an infrared gold-imaging furnace (RHL-P610C, Advance Science and Engineering Co., Ltd.), followed by immediate water quenching to ensure uniform solid solution of the Zr. The sintered compacts were then subjected to hot extrusion to achieve complete densification. In order to prevent the Zr component from diffusing during the preheating process and re-segregating and concentrating near the grain boundary, a temperature range of single α phase was selected. Specifically, the sintered specimens were heated at 800°C for 300 s in an Ar gas atmosphere and immediately extruded (extrusion ratio: 8.2, ram speed: 6 mm/s) to produce 15 mm diameter Ti-Zr-Sc extrudate.

Prior to the preparation of the above samples, the decomposition behavior of ZrH₂ and ScH₂ particles dispersed in the Ti matrix was investigated. First, solidification of the mixed powder was carried out at 500°C by the SPS method, and a Ti sintered samples containing hydride particles was prepared. The decomposition temperature of each hydride was determined by thermogravimetric differential thermal analysis (TG-DTA, DTG-60, Shimadzu Corporation), and the vacuum heat treatment temperature required for complete removal of residual hydrogen was decided based on this results.

2.3 Evaluation of microstructure and tensile properties

For the evaluation of microstructure, compositional analysis by oxygen, nitrogen, and hydrogen content measurement (EMGA 830, HORIBA, Ltd.) and microstructural analysis by X-ray diffraction (XRD-6100, Shimadzu Corporation), scanning electron microscopy (SEM, JSM-7100F, JEOL Ltd.) and accompanying energy dispersive X-ray spectroscopy (EDS, JED-2300, JEOL Ltd.), and electron backscatter diffraction (EBSD, HIKARI Detector, EDSA-TSL) were used. After mechanical polishing, electropolishing with an electrolytic polishing solution for Ti (CH₃COOH:HClO₄ = 95:5) (voltage 20 V, energization time 120 s) was used to prepare the samples for EBSD analysis. To evaluate the mechanical properties of each sample, a plate-like tensile specimen (parallel section width, 2 mm; thickness, 1 mm; length, 10 mm) was taken along the extrusion direction and tested at room temperature under a strain rate of 5.0 × 10⁻⁴/s using a tensile testing machine (AUTOGRAPH AG-X 50 kN, Shimadzu Corporation).

3. Results and Discussions

3.1 Thermal decomposition behavior of hydrides

Figure 1 shows the results of SEM-EDS analysis on the Ti-ZrH₂-ScH₂ mixed powder containing 2.5 at.% Sc and the two hydride particles. The ZrH₂ and ScH₂ particles were uniformly distributed on the spherical CP-Ti powder, and the fine hydride particles, especially those smaller than about 5 µm, were attached to the surface of the CP-Ti powder. Figure 2 shows the TG-DTA results for (a) ScH₂ particles and (b) Ti-ZrH₂-ScH₂ sintered material. The endothermic peaks at 760°C and 850°C were observed in the DTA curve of the ScH₂, and the TGA curve showed a decreasing mass at 850°C. Therefore, the thermal decomposition of the ScH₂ particles is considered to have occurred at 850°C. The endothermic peak at 760°C is considered to be a phase transformation peak of Sc remaining in ScH₂. On the other hand, endothermic peaks are detected around 670°C, 780°C and 840°C in the Ti-ZrH₂-ScH₂ sintered material (b). According to previous studies on the dehydrogenation reaction of ZrH₂ [15], the decomposition reaction occurs in the temperature range of 520°C to 840°C with a crystal structure change from ZrH₂ → εZrH_1.801 → δZrH_1.66 → γZrH → Zr. Our previous study [7] also showed that the thermal decomposition of ZrH₂ particles dispersed in Ti material proceeds at about 630°C. Therefore, the first peak observed around 670°C is considered to be caused by the thermal decomposition of ZrH₂ particles. The second peak around 780°C is an endothermic reaction associated with the transformation from α-Ti to β-Ti. This temperature is lower than the phase transformation point of pure Ti (882°C), but the solid solution of Zr and H is known to lower the phase transformation temperature [16, 17]. According to the equilibrium phase diagram of the Ti-Zr system, the β-phase transformation temperature decreases to about 815°C with solid solution of 5.4 at.% Zr. In addition, the effect of H on lowering the phase transformation point is more remarkable than Zr. When approximately 5 at.% of H is solid solutioned with Zr in Ti, the transformation point is lowered to around 750°C. According to the TG-DTA result of ScH₂ particles (a), the third endothermic peak corresponds to the dehydrogenation reaction of the ScH₂. Based on the above results, the vacuum heat treatment temperature was set to 1000°C to completely remove the H component derived from the ZrH₂ and ScH₂ particles from the SPS sample.

Fig. 1

(a)∼(d) SEM-EDS analysis on elemental mixture of Ti-9.55 wt.% ZrH₂-4.46 wt.% ScH₂ powder (Ti-10 wt.% Zr+2.5 at.% Sc) and SEM observation images of (e) ZrH₂ and (f) ScH₂ particles. (online color)

Fig. 2

TG-DTA analysis results of (a) ScH₂ particles and (b) Ti-ZrH₂-ScH₂ sintered material. (online color)

3.2 Microstructural Analysis and Sc₂O₃ Particle Formation in Ti-Zr-Sc PM alloy

We investigated the oxygen, nitrogen and hydrogen contents in the Ti-10%Zr+(0∼2.5 at.%)Sc. First, Fig. 3(a) shows the results of oxygen and hydrogen analyses for the Ti-ZrH₂-ScH₂ mixed powder containing 2.5 at.% Sc, its SPS sintered compact and hot extruded material. The oxygen content in all samples was around 0.2 wt.%, which is close to the upper limit specified in ASTM Grade 2. On the other hand, due to the residual hydrogen content in the SPS material from the ZrH₂ and ScH₂ particles, the hydrogen content was as high as 0.227%, which is the same as that of the mixed powder. In contrast, the hydrogen content of the extrudate subjected to vacuum heat treatment and uniform Zr solution treatment was reduced to 0.014%. This value meets the ASTM Grade 2 hydrogen specification and confirms that the hydrogen content derived from the raw material can be completely removed. Figure 3(b) shows the analysis results of each impurity component in the extrudate with different Sc addition. Oxygen, nitrogen and hydrogen were 0.175–0.20%, 0.006–0.008% and 0.013–0.015%, respectively. These values satisfy the ASTM Grade 2 specifications, and showing no significant correlation between the Sc addition and impurity elements.

Fig. 3

(a) Oxygen and hydrogen contents of elemental mixture powder, as SPSed and extruded materials (Ti-10%ZrZr+2.5 at.% Sc) and (b) Oxygen, hydrogen and nitrogen contents of extruded Ti-10%Zr+(0∼2.5 at.%) Sc alloys.

Next, the elemental distribution of Zr and Sc in the extrudates with Sc contents of 0, 0.5 and 2.5% was analyzed by SEM-EDS, as shown in Fig. 4. Zr was uniformly distributed in all extrudates. This indicates that the homogenization of Zr by solution treatment at 1000°C was maintained in the extrudates. On the other hand, the black dots in the SEM image increase as the amount of Sc increases, and EDS shows that the Sc is concentrated in the black dots. Elemental analysis on the black dots indicated by the arrow in the SEM image of (c) Ti-Zr+2.5% Sc is shown in Fig. 4(c-1). In the matrix (Nos. 3 and 4), Zr in solid solution was detected in addition to Ti, and the amount of Sc was 2.4 at.%, which was less than the addition amount. On the other hand, in the black point (Nos. 1 and 2), almost no Ti and Zr were detected, and consisted of Sc and O. The atomic number ratio calculated from the quantitative analysis of both elements is Sc:O = 2:3. As mentioned above, Sc reacts with oxygen atoms in solid solution in the matrix to form the more stable Sc₂O₃ [13]. Therefore, the black particles dispersed in the matrix are Sc₂O₃.

Fig. 4

(a)∼(c) SEM-EDS analysis results of Zr, Sc and O element mapping of Ti-10%Zr+(0, 0.5, 2.5 at.%) Sc alloys, and (c-1) EDS point analysis on dispersion particle in Ti matrix of Ti-10%Zr+2.5 at.% Sc alloy. (online color)

Next, in order to quantitatively analyze the dispersion of the oxide particles, SEM images of each sample in Fig. 4 were prepared at 200x, and black particles with an area of 40 µm² or larger, which could be visually confirmed, were subjected to image analysis software using binarization processing. The number of particles (n), area fraction (s), and average particle diameter (d₀) were calculated using image analysis software (Image-Pro Plus, Media Cybernetics). The results showed that n = 33, s = 0.76%, and d₀ = 14.03 µm for the Ti-10%Zr+0.5 at.%Sc material, and n = 31, s = 0.80%, and d₀ = 8.95 µm for the Ti-10%Zr+2.5 at.%Sc material, indicating that there is no significant difference in the size and dispersion of Sc₂O₃ particles within the range of 0.5∼2.5 at.% Sc. On the other hand, the amount of Sc₂O₃ particles was investigated based on the amount of Sc solid solution in each sample (results of EDS analysis). It is assumed that the difference between the total amount of Sc addition and the amount of Sc in solid solution contributed to the formation of Sc₂O₃. As shown in Table 1, the difference between Sc solid solution and the amount of Sc added is 0.18∼0.35 at.%, which shows no strong correlation with the amount of ScH₂ addition. Therefore, there is no difference in the amount of Sc₂O₃ particles produced within the range of 0.5∼2.5 at.% Sc, which is consistent with the SEM observation result that the amount of dispersed Sc₂O₃ particles does not strongly depend on the Sc content.

Table 1 Designed total content and solid solution content of Sc element in Ti-10%Zr+(0∼2.5 at.%) Sc alloys.

The main strengthening mechanism of the Ti-Zr-Sc PM alloys discussed in this study is solid solution strengthening by Zr and Sc. The solute atoms of Zr and Sc in the α-Ti crystal induce lattice distortion and suppress the movement of dislocations to increase the strength. Using the results of XRD analysis of Ti-10%Zr+(0∼2.5 at.%)Sc extrudate, the lattice constants in the a- and c-axis directions of the α-Ti crystal were calculated from four diffraction peaks of each sample using Bragg’s law [18]. The relationship between lattice constant and amount of Sc addition are shown in Fig. 5. Since part of the amount of Sc added contributes to the formation of Sc₂O₃ particles as described above, the amount of Sc in the Ti matrix analyzed by EDS shown in Table 1 was used as the solid solution amount. As shown in (a), the diffraction peaks corresponding to all α-Ti crystal planes were shifted to the low-angle side, indicating that the solid solution of Sc atoms expanded the α-Ti crystal in both the a- and c-axis directions. As shown in (b), the calculated lattice constants increase linearly with the amount of Sc solid solution, with an increase rate of Δa = 5.68 × 10⁻⁴ nm/at.%[Sc] and Δc = 6.27 × 10⁻⁴ nm/at.%[Sc], indicating a slightly larger crystal expansion in the c-axis direction.

Fig. 5

(a) XRD profiles of extruded Ti-10%Zr+(0∼2.5 at.%) Sc alloys and (b) dependence of lattice constant in a-axis and c-axis of α-Ti crystal matrix on Sc content.

Next, we used SEM-EBSD to analyze the microstructural changes in the Ti-10%Zr+(0∼2.5 at.%)Sc extrudate to investigate the dependence of grain size on Sc solid solution. First, SEM-EBSD and EDS analyses were carried out on the same area of the Ti-10%Zr+1.0 at.% Sc. The results are shown in Fig. 6. As mentioned above, the Sc₂O₃ particles in this alloy are shown as black dots in SEM images. These particles are enriched in Sc. EBSD analysis was performed on the area indicated by the square in Fig. 6(a) and (b). The black area in the IPF map (c) corresponds to the Sc₂O₃ particles shown in (a) and (b), and no clear pattern could be obtained at Sc₂O₃ particles. Here, IPF maps for pure Ti and Ti-10%Zr+0∼2.5 at.% Sc extrudate and the relationship between average grain size of α-Ti (d₀) and Sc addition are shown in Fig. 7. The grain boundaries can be clearly observed without noise in (a) pure Ti and (b) Ti-10%Zr without Sc, while noise was introduced in (c∼f) Ti-10%Zr+Sc by the dispersed Sc₂O₃ particles. In (a) pure Ti material, d₀ = 18.68 µm, while in (b) Ti-10%Zr material, d₀ = 2.85 µm, indicating significant grain refinement by Zr solid solution as in the previous study [7]. In (c∼f) Ti-10%Zr+Sc, d₀ = 1.92∼3.79 µm, indicating that d₀ is almost constant for the amount of Sc solid solution. In a previous study [7], it was reported that the Ti-10 wt.% Zr material (d₀ = 2.71 µm) exhibited grain refinement compared to the pure Ti (d₀ = 16.3 µm), and that the effect of grain refinement became constant when the Zr content exceeded 10 wt.%, while the rate of decrease of d₀ was more pronounced in the low Zr solid solution content. This means that the solute drag effect is saturated at about 10% Zr solid solution in α-Ti with grain size of 2∼3 µm generated by dynamic recrystallization during hot extrusion. Therefore, further grain refinement is limited even when the additional Sc is solid-solution in the α-Ti crystals constituting Ti-10 wt.% Zr. As a result, as shown in Fig. 7(b), d₀ is considered to be almost constant with respect to the amount of Sc.

Fig. 6

(a) SEM observation image, (b) EDS mapping of Sc element and (c) IPF map of extruded Ti-10%Zr+1.0 at.% Sc alloy. (online color)

Fig. 7

EBSD-IPF maps of (a) pure Ti and (b)∼(f) Ti-10%Zr+(0, 0.5, 2.5 at.%) Sc alloys, and mean grain size of α-Ti calculated using IPF maps. (online color)

3.3 Tensile properties of Ti-Zr-Sc PM alloy

The stress-strain curves for the Ti-10%Zr+0∼2.5 at.% Sc extrudate obtained from tensile tests at room temperature are shown in Fig. 8, and the mean values of the maximum tensile strength (UTS), 0.2% proof stress (0.2%YS) and elongation at break are summarized in Table 2. The UTS and 0.2% YS of the Ti-Zr+0 at.% Sc were 610.3 MPa and 491.9 MPa, respectively, while the tensile strength decreased with the addition of Sc, reaching 520.1 MPa and 402.0 MPa for the Ti-Zr+1.0 at.% Sc. However, the UTS and YS values increased when the Sc content over 1.0 at.% Sc, reaching UTS: 588.2 MPa and 0.2% YS: 476.8 MPa for the Ti-Zr+2.5 at.% Sc. The elongation at break for the Ti-Zr+0 at.% Sc was 22.1%, while it increased to 27.2% for the Ti-Zr+1.0 at.% Sc. However, the elongation decreased with the addition of 1.5 at.% Sc and 2.5 at.% Sc and were 14.7% and 9.6%, respectively. In summary, in the Ti-10% Zr+0∼2.5 at.% Sc extrudate, the ductility increased with the decrease of UTS and 0.2%YS in the range of 0∼1.0 at.% Sc addition, while the tensile strength increased in the range of 1.0∼2.5 at.%, but the elongation decreased significantly. The results of SEM-EDS analysis of the fracture surface of the Ti-10%Zr+2.5 at.% Sc after tensile testing are shown in Fig. 9. Most of the fracture surfaces show fine dimples indicating ductile fracture, but there are dispersed particles of 10∼30 µm in size, which are considered to be Sc₂O₃ particles as shown in Fig. 4, based on the EDS analysis results. At Sc additions of 1.5 at.% or less, the higher ductility of the α-Ti matrix compensated for the reduction in ductility due to the dispersion of brittle Sc₂O₃ particles. However, at Ti-10%Zr+2.5 at.% Sc, the ductility of the matrix decreased, which could no longer compensate for the reduction in ductility due to Sc₂O₃ particles, resulting in a lower elongation at break.

Fig. 8

Stress-strain curves of pure Ti and Ti-10%Zr+(0∼2.5 at.%) Sc alloys in tensile test.

Table 2 Ultimate tensile strength (UTS), 0.2% yield stress (YS) and elongation to failure of extruded pure Ti and Ti-10%Zr+(0∼2.5 at.%) Sc alloys.

Fig. 9

SEM-EDS analysis results of fractured surface of Ti-10%Zr+2.5 at.% Sc alloy specimen after tensile test. (online color)

3.4 Strengthening mechanism of Ti-Zr-Sc PM alloy

Based on the results of the tensile test, a quantitative analysis of the strengthening mechanism of the Ti-Zr-Sc extrudate is performed. Based on the results of the microstructural analysis described in Section 3.2, the following microstructural factors are considered to change with the addition of Sc: α-Ti grain size, Sc solid solution amount, Sc₂O₃ particle formation amount, and oxygen solid solution amount. Each of these factors contributes to an increase or decrease in the strengthening amount. In particular, the effect of solid solution strengthening by oxygen solute atoms in α-Ti crystals is remarkable [19, 20]. Therefore, the formation of Sc₂O₃ particles with the addition of Sc leads to a decrease in oxygen solid solution, which is expected to result in a decrease of 0.2% YS. Therefore, we derive the strengthening amount for (1) grain boundary strengthening by grain refinement, solid solution strengthening by (2) O, (3) Zr, and (4) Sc solute atoms, and (5) dispersion strengthening by Sc₂O₃ particles using empirical formula and theoretical models. For (4) Sc solid solution, the material constant F_m, which is the maximum interaction force between the solute atoms and dislocations, required to calculate the amount of solid solution strengthening by the Labusch model [21, 22]. However, the value of F_m is not known, so the value obtained by subtracting the calculated strengthening amount (1), (2), (3) and (5) from the strengthening amount obtained from tensile tests is used as (4) the solid solution strengthening amount by Sc.

First, the increase in 0.2%YS due to grain boundary strengthening is derived from the Hall-Petch law [23, 24] shown below.

  

\begin{equation} \Delta \sigma = \sigma_{0} + \text{kd}^{ - 1/2} \end{equation} (1)

where σ₀: frictional force, k: material constant, d: average grain size of matrix, and 15.7 MPa/mm^1/2 was used as the material constant k [25, 26].

Next, the amount of solid solution strengthening by (2) O and (4) Zr is derived using the Labusch model [21, 22], which is a statistical theory that can be used when the solute atoms is relatively high (concentration exceeds about 0.1 at.%). The content of O, Zr and Sc in solid solution in the Ti-Zr-Sc extrudate is well above 0.1 at.%. Therefore, Labusch model is selected to derive solid solution amount in this study. The strengthening amount in Labusch model can be expressed by the following equation, and we have already demonstrated that the model can calculate the amount of solid solution strengthening for Ti-O alloys [19, 20, 27], Ti-N alloys [19, 28], Ti-Zr alloys [7, 29] and other alloys with high accuracy.

  

\begin{equation} \Delta \sigma = \left(\frac{F_{\text{m}}{}^{4}w}{4Gb^{9}} \right)^{\frac{1}{3}}\frac{c^{\frac{2}{3}}}{S_{F}} \end{equation} (2)

where w: the range of interaction between dislocations and solute atoms, G: shear stiffness (4.50 × 10¹⁰ N/m²), b: Burgers vector, S_F: Schmidt factor, c: solute atom concentration. The S_F was measured by EBSD analysis for each sample (0.475∼0.481). b was the a-axis length (2.97 × 10⁻¹⁰ m) calculated for the reference material, Ti-10%Zr+0 at.% Sc extrudate, and w ≈ 5b [30] was estimated to be 1.485 × 10⁻⁹ m. The F_m values for the solid solution of O and Zr atoms in α-Ti crystals were reported to be 6.22 × 10⁻¹⁰ N and 1.38 × 10⁻¹³ N in previous studies [7, 19]. As for the solute content of each element, the solute content of Sc was obtained from the SEM-EDS point analysis results shown in Table 1. Regarding O solute contest, in Ti-10%Zr+0.5∼2.5 at.% Sc, Sc atoms react with some of the oxygen solute atoms in Ti to form Sc₂O₃ particles, so the total O in the sample can be classified into O in solid solution and O in the oxide. The amount of Sc in Sc₂O₃ is the difference between the total amount of added Sc and the amount of Sc solute in the Ti matrix (0.18∼0.35 at.%), as shown in Table 1. The results of microstructural analysis show that the same amount of Sc₂O₃ particles was formed regardless of the amount of Sc added. Assuming that 0.245 at.% Sc, which is the average of the difference, reacted with the oxygen solute atoms in the Ti crystal, 0.38 at.% of the oxygen solute atoms contributed to the formation of Sc₂O₃. In other words, 0.38 at.% of the oxygen solute is considered to have decreased compared to the Ti-10% Zr + 0 at.% Sc. The amount of oxygen solute in each sample calculated based on this is significantly reduced to 0.20∼0.25 at.% compared to the amount of oxygen solute in the Ti-10%Zr (0.65 at.%), as shown in Fig. 10. Finally, the dispersion strengthening of Sc₂O₃ particles is calculated. Geometrically Necessary Dislocation (GND) [31] occurs in composite materials due to the mismatch between the elastic moduli of the matrix and the dispersed particles. The strengthening amount due to this is derived from the following equation [32].

  

\begin{equation} \Delta \sigma_{\text{GND}} = 1.25Gb\sqrt{3\rho_{\text{GND}}} \end{equation} (3)

Fig. 10

Oxygen contents as solid solutes in α-Ti matrix and as contained in Sc₂O₃ particles of each specimen.

The strengthening amount due to the Sc₂O₃ dispersion can be obtained from eq. (4) [33–35].

  

\begin{equation} \Delta \sigma_{\text{disp}} = \frac{8f\varepsilon_{\text{y}}}{bd} + \frac{12 \cdot \Delta C \cdot \Delta T \cdot f}{bd} \end{equation} (4)

where f and d are the volume fraction and size of the dispersed particles, respectively; ε_y is the yield strain; ΔC is the difference in the linear expansion coefficient between the matrix and the dispersed particles; and ΔT is the difference between the sintering temperature and room temperature (=980°C). As mentioned above, f and d were calculated from the area fraction and average particle size of the Sc₂O₃ particles observed in the SEM images shown in Fig. 4. ε_y was measured from the stress-strain curve obtained from the tensile test. Since no report was found for linear expansion coefficient of Sc₂O₃, the linear expansion coefficient (7.5 × 10⁻⁶/°C) of oxides with the same crystal structure (alumina single crystal Al₂O₃ and yttria Y₂O₃) was substituted, and the value obtained by subtracting it from the linear expansion coefficient of pure Ti (8.4 × 10⁻⁶/°C). The ΔC was 9.0 × 10⁻⁷°C.

Based on the above measurement results, the material properties required to derive the strengthening factors (1) to (5) are summarized in Table 3. In order to quantitatively analyze the strengthening mechanism in Ti-Zr-Sc extrudate, which is the objective of this study, the strengthening amount associated with Sc addition is calculated. In this study, a Ti-10%+0 at.% Zr alloy was used as the reference material, and the strengthening amount for each factor was calculated as the increase or decrease from the reference material. The amounts of strengthening of (1) Δσ_[GR], (2) Δσ_[O-SS], (3) Δσ_[Zr-SS], (4) Δσ_[Sc-SS], and (5) Δσ_[Disp] corresponding to strengthening factors (1) to (5) can be derived using eqs. (1) to (4) and Table 3, and the experimental value Δσ_[Exp] obtained in tensile tests is expressed as follows.

  

\begin{align} \varDelta \sigma_{[\textit{Exp}]} &= \varDelta \sigma_{[\textit{GR}]} + \varDelta\sigma_{[\textit{O-SS}]} + \varDelta \sigma_{[\textit{Zr-SS}]} \\ &\quad + \varDelta\sigma_{[\textit{Sc-SS}]} + \varDelta\sigma_{[\textit{Disp}]} \end{align} (5)

Table 3 Zr, Sc and O solution contents, characteristics of Sc₂O₃ dispersoids, microstructural and mechanical properties of extruded Ti-10%Zr+(0∼2.5 at.%) Sc alloys.

As described above, (4) Δσ_[Sc-SS] is the value obtained by subtracting the strengthening amounts (1) to (3) and (5) from Δσ_[Exp], and the calculated results are shown in Table 4 and Fig. 11. Negative values indicate a decrease from the reference material. For example, Δσ_[O-SS] is negative for all Ti-10% Zr+0.5∼2.5 at.% Sc, indicating that the amount of strengthening due to oxygen solid solution has decreased compared to the reference material as a result of the decrease in oxygen solid solution due to the formation of Sc₂O₃.

Table 4 0.2% YS increment by each strengthening factor of extruded Ti-10%Zr+(0∼2.5 at.%) Sc alloys. Ti-10%Zr+0%Sc material is used as standard.

Fig. 11

Calculated strengthening increment by each factor of extruded Ti-10%Zr+(0∼2.5 at.%) Sc alloys.

First, regarding (1) grain boundary strengthening, as shown in Fig. 7, Δσ_[GR] increased only in the Ti-10% Zr+0.5 at.% Sc due to the formation of finer grains than in the reference material. On the other hand, the amount of strengthening tended to decrease by 28 to 40 MPa due to slight grain coarsening in the Ti-10 wt.% Zr+1.0∼2.5 at.% Sc. The solid solution strengthening of O decreased in the range of 114 to 132 MPa due to the decrease in oxygen solute caused by the formation of Sc₂O₃ as described above. The increase or decrease in the solid solution strengthening amount of Zr was limited to −2.4∼2.2 MPa because the Zr solute did not change with the addition of Sc. (5) The inclement of dispersion strengthening by Sc₂O₃ particles is limited to 8∼12 MPa. This is due to the average particle size of the dispersed particles was relatively large (about 10 µm). Based on these results, the Sc solid solution strengthening (Remain in Table 4) was calculated by eq. (5). The solid solution strengthening of Sc showed a tendency to increase with increasing Sc in solid solution. This trend consistent with the solid solution strengthening.

Focusing on eq. (2) used to calculate the solid solution strengthening, since (F_m⁴w/4Gb⁹)^1/3 is the constant part, a linear relationship is established between the increase in 0.2% YS Δσ and the variable part c^2/3/S_F, which includes c related to the amount of solid solution atoms. In other words, for Ti-10%Zr+0.5∼2.5 at.% Sc, Δσ and c^2/3/S_F can be plotted to obtain a linear relationship between these values. This method has been employed in previous studies on quantitative analysis of solid solution strengthening of O, N, and Zr in Ti, and good agreement between calculated and experimental results has been reported [7, 19, 29]. Figure 12 shows the correlation between the amount of Sc solid solution strengthening and c^2/3/S_F in eq. (2). The coefficient of determination (R² value) of the regression equation is 0.981, which is statistically strongly correlated with c^2/3/S_F. Here, the F_m value can be calculated from the constant part (F_m⁴w/4Gb⁹)^1/3, which corresponds to the slope of the linear approximation equation. Using this value, the amount of Sc solid solution strengthening Δσ_[Sc-SS] can be derived from eq. (2). The proportionality constant in the linear approximation is 776.9, which is equal to the constant part (F_m⁴w/4Gb⁹)^1/3 in eq. (2), resulting in F_m = 1.73 × 10⁻¹⁰ N for the Sc solid solution strengthening.

Fig. 12

Relationship between material’s factor in Labusch model (cSc2/3/SF) and strength increment by Sc solid solution of Ti-10%Zr+(0.5∼2.5 at.%) Sc alloys.

Based on the results of these analyses, the contribution of each strengthening factor to the 0.2%YS value is summarized: The 0.2%YS value decreased with Sc content in the range of 0∼1.0 at.% Sc addition and increased in the range of 1.0∼2.5 at.% Sc addition. In the former range, the oxygen solid solution strength decreased significantly from 234.6 MPa to 112.0∼129.3 MPa as a result of the decrease in oxygen solid solution due to the formation of Sc₂O₃. At the same time, the addition of Sc is accompanied by an increase in Sc solid solution strengthening of 54.0∼56.3 MPa and Sc₂O₃ dispersion strengthening of 8.1∼11.5 MPa. However, these totals are small compared with the above decrease in oxygen solid solution strengthening, resulting in a decrease of 0.2%YS. In the later range, while no big difference was observed in the oxygen solid solution strength, the 0.2% YS value tends to increase again. This is caused by the increase in Sc solid solution strengthening of 82.8∼127.6 MPa, in addition to the Sc₂O₃ dispersion strengthening of 7.7∼12.2 MPa. The change in the solid solution of the Zr was extremely small, and the effect of the increase or decrease in the Zr solid solution strengthening was negligible.

From these results, it can be concluded that in the Ti-Zr-Sc PM extrudate, the 0.2% YS value decreased mainly due to the decrease in oxygen solid solution strengthening resulting from the formation of Sc₂O₃. However, the further addition of Sc contributed to the 0.2% YS increase mainly due to the effect of Sc solid solution strengthening.

4. Conclusion

To improve the mechanical properties of biocompatible high-strength Ti-Zr PM alloys, we prepared Ti-Zr-Sc PM extrudate by adding Sc, a non-toxic third element, and quantitatively clarified each strengthening mechanism by analyzing the microstructure and evaluating the tensile properties. The following results were obtained.

1. (1)    When the active Sc was added as particles, hydride (ScH₂) was prepared as raw powder from the viewpoint of safety during handling. After mixing with Ti powder and ZrH₂ particles, vacuum sintering, heat treatment, and hot extrusion processing, the residual hydrogen content was reduced to a level meeting the ASMT standard.
2. (2)    The solid solution of Sc atoms in α-Ti expanded the crystal lattice in both the a- and c-axis directions. Both showed a tendency to increase linearly with the amount of Sc solid solution. On the other hand, the effect of Sc solute atoms on the further refinement of α-Ti grain was found to be limited.
3. (3)    Some of the added Sc reacted with oxygen solute in the Ti matrix to form Sc₂O₃ particles. This resulted in a decrease in the solid solution of oxygen in the α-Ti crystals.
4. (4)    The increase or decrease in 0.2% YS for the main strengthening factors in Ti-Zr-Sc PM alloys, i.e., grain refinement strengthening, solid solution strengthening by each solute atom, and Sc₂O₃ particle dispersion strengthening, were calculated, and the amount of Sc solid solution strengthening was derived from these results. Quantitative analysis of each strengthening factor showed that the 0.2% YS decreased with the addition of Sc, mainly due to the decrease in O solid solution strengthening resulting from the formation of Sc₂O₃, but that the 0.2% YS increased again with an increase in the amount of Sc solid solution strengthening as the amount of Sc increased.

Acknowledgments

This research was supported by the Osaka University OU Master Plan Acceleration Project and the Light Metal Scholarship Foundation’s Education and Research Fund. The authors would like to express their sincere appreciation to Ryoji Minamitani, Mitsuo Horie, and Hiroko Fujii of the Joining and Welding Research Institute, Osaka University, for their generous cooperation in conducting the experiments and analyses.

REFERENCES

• 1)   M.  Kaur and  K.  Singh: Review on titanium and titanium based alloys as biomaterials for orthopaedic applications, Mater. Sci. Eng. C 102 (2019) 844–862. doi:10.1016/j.msec.2019.04.064
• 2)   P.  Pesode and  S.  Barve: A review—metastable β titanium alloy for biomedical applications, J. Eng. Appl. Sci. 70 (2023) 25–60. doi:10.1186/s44147-023-00196-7
• 3)  Binary Alloy Phase Diagrams, ed. by T.B. Massalski, (ASM International, Materials Park, OH, 1990).
• 4)   R.  Arroyave,  T.W.  Eagar and  L.  Kaufman: Thermodynamic assessment of the Cu–Ti–Zr system, J. Alloy. Compd. 351 (2003) 158–170. doi:10.1016/S0925-8388(02)01035-6
• 5)  Controlled Grain Growth, ed. by F.F.Y. Wang, (Academic Press, New York, NY, 1976).
• 6)   A.  Alhazaa,  A.  Assaifan,  M.  Hezam,  M.A.  Shar,  J.  Umeda and  K.  Kondoh: Effect of sintering temperature on the microstructure and mechanical properties of the Ti-2.5Zr alloy, Mater. Res. Express 8 (2021) 016522. doi:10.1088/2053-1591/abd91f
• 7)   K.  Kondoh,  M.  Fukuo,  S.  Kariya,  K.  Shitara,  S.  Li,  A.  Alhazaa and  J.  Junko: Quantitative strengthening evaluation of powder metallurgy Ti–Zr binary alloys with high strength and ductility, J. Alloy. Compd. 852 (2021) 156954. doi:10.1016/j.jallcom.2020.156954
• 8)   A.  Issariyapat,  J.  Huang,  T.  Teramae,  S.  Kariya,  A.  Bahador,  P.  Visuttipitukul,  J.  Umeda,  A.  Alhazaa and  K.  Kondoh: Microstructure refinement and strengthening mechanisms of additively manufactured Ti-Zr alloys prepared from pre-mixed feedstock, Addit. Manuf. 73 (2023) 103649. doi:10.1016/j.addma.2023.103649
• 9)   A.  Issariyapat,  S.  Kariya,  A.  Alhazaa,  J.  Umeda and  K.  Kondoh: Additive Manufacturing and Characterization of High Strength Ti-Zr Gyroid Scaffolds Using Pre-Mixed Ti-ZrH₂ Powders, JOM 73 (2021) 4166–4176. doi:10.1007/s11837-021-04923-4
• 10)  C.T. Horovitz: Biochemistry of Scandium and Yttrium, (Springer, 1999).
• 11)   K.  Munir,  J.  Lin,  C.  Wen,  P.F.A.  Wright and  Y.  Li: Mechanical, corrosion, and biocompatibility properties of Mg-Zr-Sr-Sc alloys for biodegradable implant applications, Acta Biomater. 102 (2020) 493–507. doi:10.1016/j.actbio.2019.12.001
• 12)   H.  Okamoto: Sc-Ti (Scandium-Titanium), J. Phase Equilibria Diffus. 34 (2013) 158–159. doi:10.1007/s11669-012-0157-2
• 13)   O.  Takeda and  T.H.  Okabe: Current Status on Resource and Production Technologies of Scandium, J. Mining Metall. Inst. Japan 137 (2021) 36–44. doi:10.2473/journalofmmij.137.36
• 14)  Rare Metals Handbook, ed. by C.A. Hampel, (Reinhold Publishing Co., London, 1961).
• 15)   F.D.  Manchester and  J.M.  Pitre: The h- sc (Hydrogen- Scandium) system, J. Phase Equilibria 18 (1997) 194–205. doi:10.1007/BF02665706
• 16)   M.  Ma,  L.  Liang,  B.  Tang,  W.  Xiang,  Y.  Wang,  Y.  Cheng and  X.  Tan: Decomposition kinetics study of zirconium hydride by interrupted thermal desorption spectroscopy, J. Alloy. Compd. 645 (2015) S217–S220. doi:10.1016/j.jallcom.2015.01.054
• 17)   J.L.  Murray: The Ti–Zr (Titanium-Zirconium) system, Bull. Alloy Phase Diagr. 2 (1981) 197–201. doi:10.1007/BF02881478
• 18)   A.  San-Martin and  F.D.  Manchester: The H–Ti (Hydrogen-Titanium) system, Bull. Alloy Phase Diagr. 8 (1987) 30–42. doi:10.1007/BF02868888
• 19)   W.H.  Bragg and  W.L.  Bragg: The reflection of X-rays by crystals, Proc. R. Soc. Lond. A 88 (1913) 428–438. doi:10.1098/rspa.1913.0040
• 20)   S.  Kariya,  M.  Fukuo,  J.  Umeda and  K.  Kondoh: Quantitative Analysis on Light Elements Solution Strengthening in Pure Titanium Sintered Materials by Labusch Model Using Experimental Data, Mater. Trans. 60 (2019) 263–268. doi:10.2320/matertrans.Y-M2018849
• 21)   J.  Shen,  B.  Chen,  J.  Umeda and  K.  Kondoh: Microstructure and mechanical properties of CP-Ti fabricated via powder metallurgy with non-uniformly dispersed impurity solutes, Mater. Sci. Eng. A 716 (2018) 1–10. doi:10.1016/j.msea.2018.01.031
• 22)   R.  Labusch: A Statistical Theory of Solid Solution Hardening, Phys. Status Solidi 41 (1970) 659–669. doi:10.1002/pssb.19700410221
• 23)   R.  Labusch: Statistische theorien der mischkristallhärtung, Acta Metall. 20 (1972) 917–927. doi:10.1016/0001-6160(72)90085-5
• 24)   E.O.  Hall: Variation of Hardness of Metals with Grain Size, Nature 173 (1954) 948–949. doi:10.1038/173948b0
• 25)   N.J.  Petch: J. Iron Steel Inst. 173 (1953) 25–28.
• 26)   V.A.  Moskalenko,  A.R.  Smirnov,  Y.M.  Plotnikova,  I.S.  Braude and  R.V.  Smolianets: Fundamentals of titanium nanocrystalline structure creation by cryomechanical grain fragmentation, Mater. Sci. Eng. A 700 (2017) 707–713. doi:10.1016/j.msea.2017.06.057
• 27)   P.  Luo,  D.T.  McDonald,  W.  Xu,  S.  Palanisamy,  M.S.  Dargusch and  K.  Xia: A modified Hall–Petch relationship in ultrafine-grained titanium recycled from chips by equal channel angular pressing, Scr. Mater. 66 (2012) 785–788. doi:10.1016/j.scriptamat.2012.02.008
• 28)   K.  Kondoh,  E.  Ichikawa,  A.  Issariyapat,  K.  Shitara,  J.  Umeda,  B.  Chen and  S.  Li: Tensile property enhancement by oxygen solutes in selectively laser melted titanium materials fabricated from pre-mixed pure Ti and TiO₂ powder, Mater. Sci. Eng. A 795 (2020) 139983. doi:10.1016/j.msea.2020.139983
• 29)   K.  Kondoh,  A.  Issariyapat,  J.  Umeda and  P.  Visuttipitukul: Selective laser-melted titanium materials with nitrogen solid solutions for balanced strength and ductility, Mater. Sci. Eng. A 790 (2020) 139641. doi:10.1016/j.msea.2020.139641
• 30)   K.  Kondoh,  S.  Kariya,  A.  Khantachawana,  A.  Alhazaa and  J.  Umeda: Quantitative Strengthening Evaluation of Powder Metallurgy Titanium Alloys with Substitutional Zr and Interstitial O Solutes via Homogenization Heat Treatment, Materials 14 (2021) 6561. doi:10.3390/ma14216561
• 31)  M. Kato: Introduction to the Theory of Dislocations, (Shokabo Co. Ltd., Tokyo, 1999).
• 32)   D.L.  Foley,  M.I.  Latypov,  X.  Zhao,  J.  Hestroffer,  I.J.  Beyerlein,  L.E.  Lamberson and  M.L.  Taheri: Geometrically necessary dislocation density evolution as a function of microstructure and strain rate, Mater. Sci. Eng. A 831 (2022) 142224. doi:10.1016/j.msea.2021.142224
• 33)   F.  Tang,  I.E.  Anderson,  T.  Gnaupel-Herold and  H.  Prask: Pure Al matrix composites produced by vacuum hot pressing: tensile properties and strengthening mechanisms, Mater. Sci. Eng. A 383 (2004) 362–373. doi:10.1016/j.msea.2004.05.081
• 34)   M.  Guo,  K.  Shen and  M.  Wang: Relationship between microstructure, properties and reaction conditions for Cu–TiB₂ alloys prepared by in situ reaction, Acta Mater. 57 (2009) 4568–4579. doi:10.1016/j.actamat.2009.06.030
• 35)   J.C.  Shao,  B.L.  Xiao,  Q.Z.  Wang,  Z.Y.  Ma and  K.  Yang: An enhanced FEM model for particle size dependent flow strengthening and interface damage in particle reinforced metal matrix composites, Compos. Sci. Technol. 71 (2011) 39–45. doi:10.1016/j.compscitech.2010.09.014

Related Articles

• This Paper was Originally Published in Japanese in J. Jpn. Soc. Powder Powder Metallurgy 71 (2024) 482–491, https://doi.org/10.2497/jjspm.23-00067. The citation in reference 35 is corrected.