Materials Chemistry
Production of Titanium Hydride Powder from Titanium Tetrachloride Using Magnesium Metal in Hydrogen Gas Atmosphere
2023 Volume 64 Issue 4 Pages 904-913
Details
2023 Volume 64 Issue 4 Pages 904-913
The development of titanium (Ti) powder production process is important because Ti powder metallurgy (PM) is a promising method that can result in large cost savings by reducing material loss and the number of steps in the conventional manufacturing process of Ti metal and its alloys. This study investigated the process of producing Ti hydride powder directly from titanium tetrachloride (TiCl4) using magnesium (Mg) metal and hydrogen gas (H2). The experiments were conducted at 1073–1173 K when the TiCl4 feeding rate was in the range of 19.16–57.42 g·min−1 with the use of cooling gas as Ar or H2 gas. A mixture of Ti and Ti hydride (TiH1.5) was obtained in an iron (Fe) crucible by the dehydrogenation of Ti hydride as the TiCl4 reduction proceeded. The concentration of oxygen (O) decreased with a decrease in the specific surface area and/or an increase in the proportion of TiH1.5 in the powder. As a result, the concentration of O of the mixture of Ti and TiH1.5 decreased to 0.116 mass% under a certain condition.
Titanium (Ti) is the ninth most abundant element and the fourth most abundant structural metal after aluminum (Al), iron (Fe), and magnesium (Mg) in the Earth’s crust.1) In addition, Ti metal exhibits high specific strength, high-temperature performance, excellent corrosion resistance, and biocompatibility. Therefore, it is applied in various industrial fields, such as aerospace, marine, chemical, and biomedical industries. However, Ti metal is mainly consumed in the aerospace industry because of the high cost of the manufacturing process.
In current industrial practice, the conventional manufacturing of the final product of Ti metal and its alloys is based on ingot metallurgy (IM) processing using Ti sponge as a raw material, produced by the Kroll process. The Ti sponge is sequentially converted to Ti or Ti alloy ingots, mill products, and final components by melting, mill processing, and machining, respectively. However, Ti has a high affinity with oxygen (O), nitrogen (N), and carbon (C) which negatively affect the mechanical properties of the final product. These make the manufacturing process more costly. In addition, a large amount of scrap is generated during mill processing and machining. For example, when 90–120 tons of Ti alloy are utilized to manufacture a Boeing 787, approximately 85% of the alloy feed ends up as a swarf.2) Therefore, decreasing the number of steps during manufacturing as well as the amount of scrap generated is required to reduce the cost of the production process.
In order to decrease the steps during manufacturing Ti metal and its alloys, many studies have been conducted in the past.3) One method is the utilization of powder metallurgy (PM), which can produce near-net-shape (NNS) products. This method can achieve large cost savings by minimizing material loss and eliminating several steps during manufacturing, such as melting, mill processing, and machining which account for 86% of the total manufacturing cost in the conventional production of aircraft parts.4) However, the PM process has two major drawbacks. One is unsatisfactory mechanical properties and the other is the significantly high cost of Ti products using PM.5)
Currently, Ti powder for PM is produced via gas atomization (GA), plasma atomization (PA), or the plasma rotating electrode process (PREP).6) However, these processes are expensive. In addition, to produce Ti products using PM with mechanical properties equivalent to those of conventional wrought Ti products, pressure-assisted consolidation processing, such as hot isostatic pressing (HIP), is required. Unfortunately, the use of pressure-assisted consolidation processing increases the cost of Ti products using PM. Thus, Ti products can be cost-effectively produced even using PM when the high costs of Ti powder and processing are solved.
To resolve these two major disadvantages of Ti products using PM, many studies have been conducted in the past. One solution for reducing the expensive price of Ti powder is to develop the direct production process for Ti powder from raw materials, such as titanium tetrachloride (TiCl4), without the production of a Ti sponge, as shown in Table 1.2,3,5,7–36) Among several processes, the production of Ti hydride powder from TiCl4 using Mg as a reducing agent in a hydrogen gas (H2) atmosphere was suggested by ADMA.35,36) This method is promising because the use of Ti hydride powder as a feedstock for PM can improve the mechanical properties of Ti products. Owing to the brittleness of Ti hydride, bulk Ti hydride can be easily pulverized into a powder form. In addition, the Ti hydride powder can be easily fractured during cold compaction, resulting in a higher density. Moreover, dehydrogenation of Ti hydride powder during the sintering of the compacted powder in vacuum results in the improvement of densification, which contributes to a higher sintered density of compact compared with that obtained using Ti powder.37,38) Furthermore, Ti hydride has a higher resistance to oxidation than Ti powder.
In the ADMA process, Ti hydride in bulk is produced by the reduction of TiCl4 using Mg in an H2 gas atmosphere at 1023–1123 K. Afterward, vacuum distillation to remove the MgCl2 and remaining Mg in the Ti hydride and hydrogenation of Ti hydride in bulk, in which Mg and MgCl2 are removed, are sequentially and repeatedly conducted at 1273–1323 K and 873–913 K, respectively. In this process, the remaining Mg metal is recovered as the metal itself. In addition, because a wet method is not utilized to remove MgCl2 and remaining Mg, O contamination of Ti is restricted. However, vacuum distillation followed by hydrogenation is conducted at different temperatures, twice. In addition, a long processing time is required.
In this study, the production of Ti hydride powder from TiCl4 using Mg in an H2 gas atmosphere followed by leaching was investigated, as shown in Fig. 1. Furthermore, a systematic investigation of the magnesiothermic reduction of TiCl4 in an H2 gas atmosphere was carried out which was not reported for the ADMA process. In addition, a leaching method for removing MgCl2 and the remaining Mg was conducted in order to decrease the number of processing steps, shorten the processing time, and simultaneously maintain the O concentration in the residues obtained at smaller than 1800 ppm, meeting ASTM Gr. 1 for Ti. However, it should be addressed that a detailed evaluation regarding energy-efficiency or economic feasibility is required with regard to the utilization of a leaching method in the process of Ti powder production by reacting TiCl4 and Mg metal in an H2 atmosphere.
Flowchart of the experimental procedure for the Ti and TiH1.5 powder production in this study.
The assessed Ti–H phase diagram was redrawn and presented in Fig. 2(a). This diagram shows the isothermal variations of equilibrium hydrogen partial pressure ($p_{\text{H}_{2}}$) at 576–1273 K as a function of the atomic ratio of H to Ti.39) In addition, this phase diagram provides the thermodynamic information of the equilibrium composition of Ti hydride formed at a certain $p_{\text{H}_{2}}$ and temperature.
In this study, experiments were conducted at 1073–1173 K in an H2 gas atmosphere. Owing to the lack of thermodynamic information under the experimental conditions, the equilibrium $p_{\text{H}_{2}}$ as a function of the atomic ratio of H to Ti at 1073–1173 K was drawn by extrapolating eq. (1), which was developed by Wang,40) as shown in Fig. 2(b). In eq. (1), pβ is the hydrogen partial pressure in the β-Ti region (atm), T is the absolute temperature (K), and X is the atomic ratio of H to Ti. In addition, the Ti–H phase diagram for the δ-Ti hydride region is shown in Fig. 2(c).41)
| \begin{align} \log p_{\beta}\ (\text{atm}) &= 6.879 + 0.304X \\ &\quad + 2\log\ (X/(2 - X)) - 6948.712/T \end{align} | (1) |
As shown in Fig. 2(b), when $p_{\text{H}_{2}}$ is 1 atm, the atomic ratios of H to Ti are approximately 0.66, 0.54, and 0.44 at 1073 K, 1123 K, and 1173 K, respectively, under equilibrium. This indicates that the H concentration in β-Ti increased with decreasing temperature at identical $p_{\text{H}_{2}}$, which is likewise in Fig. 2(c). These results indicate that the dissolution of H into Ti is strongly affected by the temperature when $p_{\text{H}_{2}}$ in the reaction system is identical.
Figure 2(c) shows that the temperature required for the stability of TiH2, which exhibits the H concentration of approximately 4.04 mass%, is lower than 677 K under 1 atm of $p_{\text{H}_{2}}$ when the results of Wang40) are used. However, as shown in Fig. 2(c), the equilibrium $p_{\text{H}_{2}}$ for TiH2 formation is 1 atm at 1047 K when the thermodynamic data of Barin is used.41) Therefore, in this study, both thermodynamic data were carefully utilized to consider the equilibrium composition of Ti hydride formed in an H2 gas atmosphere, despite the lack of data on the equilibrium $p_{\text{H}_{2}}$ for TiHx formation in the region where the atomic ratio of H to Ti is lower than 2.41)
Figure 3 shows the schematic of the experimental apparatus used in this study. TiCl4 (purity > 99.0%, Daejung Chemicals & Metals Co., Ltd.), used as a feedstock, was prepared inside a disposable glove box. Mg metal (ingot, purity: 99.9%, Silica Trading Co., Ltd.) was used as a reducing agent and placed inside an Fe crucible (ϕtop = 93 mm, O.D. of top, ϕbottom = 60 mm, O.D. of bottom, t = 1 mm, thickness, h = 176 mm, height). The Fe crucible was placed inside a stainless steel (SUS) reactor (ϕ = 142 mm, O.D., t = 6 mm, thickness, h = 493 mm, height). The lower end of an alumina (Al2O3) tube for the injection of TiCl4 was positioned at a distance of 3 mm from the upper end of the Fe crucible and the thermocouple was positioned at a distance of 60 mm from the bottom of the Fe crucible. After the TiCl4 injection tube and thermocouple were assembled with the flange and set up in the reactor, the reactor was placed in an electric furnace.
Schematic diagram of the experimental apparatus.
Before heating, the inside of the reactor was purged thrice by conducting a series of evacuations for 15 min and filling argon (Ar, purity: 99.999%) gas to an internal pressure of 1 atm. Subsequently, the reactor was first heated to 473 K in an Ar gas atmosphere, and the inside of the reactor was purged again thrice at 473 K by conducting an identical series. After the reactor was purged at 473 K, it was heated to 1073 K, 1123 K, or 1173 K. When the temperature reached 1073 K, 1123 K, or 1173 K, the flow of Ar gas was stopped, and H2 (purity: 99.999%) gas flowed into the reactor while maintaining an internal pressure of 1 atm. After 15 min, the flow rate of the H2 gas was maintained at 1500 sccm and the injection of TiCl4 was started. TiCl4 was carried into the crucible by supplying Ar gas into the TiCl4 bottle and the TiCl4 feeding rate was varied by regulating the flow rate of Ar gas to 20, 30, or 40 sccm. After the TiCl4 injection was completed, the reactor was cooled to room temperature in either Ar gas or H2 gas atmosphere. The flow rates of Ar and H2 gases used in this study were controlled using a mass flow controller (MFC).
After cooling, the Fe crucible was cut in half, and water leaching followed by acid leaching using hydrochloric acid (HCl) was conducted to remove magnesium chloride (MgCl2) and residual Mg in the Fe crucible. First, water leaching was conducted using distilled water at 298 K for 24 h without stirring, and acid leaching was conducted using 1 M HCl solution at 298 K for 3 h with stirring at 300 rpm. Subsequently, the residues were rinsed with distilled water, followed by ethyl alcohol, and crushed using a mortar and pestle. The crushed residues were leached again using 1 M HCl solution at 298 K for 2 h with stirring at 300 rpm. The residues were filtrated and rinsed with distilled water followed by ethyl alcohol. Subsequently, the residues were dried at 298 K for 1 h in air.
The chemical compositions of the samples were analyzed using X-ray fluorescence spectroscopy (XRF: Oxford instruments, X-MET 5100), and their crystalline phases were identified using an X-ray diffractometer (XRD: Rigaku, D/Max-2500, Cu-Kα radiation). The morphological characterization of the samples was performed using field-emission scanning electron microscopy (FE-SEM: Hitachi, SU8220). The concentrations of O and hydrogen (H) in the samples were analyzed using the O/N determinator (Eltra, ON900) and O/N/H determinator (Eltra, ONH2000), respectively.
Table 2 lists the experimental conditions used in this study. In addition, Fig. 4 shows the photographs of the residues obtained in various positions in the reactor after the reduction of TiCl4 using Mg in H2 gas atmosphere at 1123 K (Exp. no. 200826). As shown in Fig. 4(a), grey and dark grey deposits were found on the upper and lower part of the TiCl4 injection tube, respectively. The deposits were loosely attached to the tube. The boiling temperature of TiCl4 is 409 K, and the vapor pressure of pure Mg at 1123 K is 0.089 atm which is sufficient for evaporation. As a result, the deposits on the tube surface were produced by the reaction between gaseous TiCl4 and Mg.
Representative photographs of residues obtained in various positions after the experiment (Exp. no. 200826): (a) The mixture of Ti and TiHx deposited on the TiCl4 injection tube, (b) MgCl2 produced before water leaching, (c) The mixture of Ti and TiH1.5 after water leaching, and (d) The mixture of Ti and TiH1.5 after acid leaching.
The white-colored residue in bulk residues in the Fe crucible was found before leaching, as shown in Fig. 4(b). It was removed from the Fe crucible by water leaching, as shown in Fig. 4(c). These results imply that the residues were a mixture of the residual Mg and MgCl2 produced, according to the overall chemical reactions shown in eqs. (2) and (3), respectively. After removing the white-colored residues, the dark grey residues were found to be firmly attached to the Fe crucible at the center, wall, and bottom of the crucible. Figure 4(d) shows the photograph of the residues after HCl leaching. Subsequently, crushing of the residues in Fig. 4(d) followed by HCl leaching was carried out.
| \begin{equation} \text{TiCl$_{4}$ ($g$)} + \text{2 Mg ($l$)} = \text{Ti ($s$)} + \text{2 MgCl$_{2}$ ($l$)} \end{equation} | (2) |
| \begin{align*} \Delta G^{\circ}{}_{\text{r}} & = -309.3\,\text{kJ}\ \text{at}\ 1123\,\text{K}^{\text{41)}}\\ \Delta H^{\circ}{}_{\text{r}} & = -426.2\,\text{kJ}\ \text{at}\ 1123\,\text{K}^{\text{41)}} \end{align*} |
| \begin{align} &\text{TiCl$_{4}$ ($g$)} + \text{2 Mg ($l$)} + \text{3/4 H$_{2}$ ($g$)} \\ &\qquad\qquad\qquad\ \ = \text{TiH$_{1.5}$ ($s$)} + \text{2 MgCl$_{2}$ ($l$)} \end{align} | (3) |
As shown in Table 3, the concentration of O in the residue in the center of the Fe crucible after the first HCl leaching was 0.441 mass%. However, the concentration of O in the residue after the second HCl leaching was 0.169 mass%. These results indicate that a certain amount of a compound containing O, captured in the residues obtained after the first HCl leaching, was further removed after the second HCl leaching. When magnesiothermic reduction and leaching are considered, magnesium hydroxide (Mg(OH)2) is expected to be produced by the reaction between residual Mg and H2O during water leaching, as shown in eq. (4), and it remained in the residues despite the first HCl leaching. However, the crushing of the residues led to the exposure of Mg(OH)2, and that was dissolved during the second HCl leaching, as shown in eq. (5). Therefore, the concentration of O in the residue decreased to 0.169 mass%.
| \begin{equation} \text{Mg ($s$)} + \text{2H$_{2}$O ($l$)} = \text{Mg(OH)$_{2}$ ($s$)} + \text{H$_{2}$ ($g$)} \end{equation} | (4) |
| \begin{equation*} \Delta G^{\circ}{}_{\text{r}} = -359.4\,\text{kJ}\ \text{at}\ 298\,\text{K}^{\text{42)}} \end{equation*} |
| \begin{equation} \text{Mg(OH)$_{2}$ ($s$)} + \text{2HCl ($aq.$)} = \text{MgCl$_{2}$ ($aq.$)} + \text{2H$_{2}$O ($l$)} \end{equation} | (5) |
| \begin{equation*} \Delta G^{\circ}{}_{\text{r}} = -103.7\,\text{kJ}\ \text{at}\ 298\,\text{K}^{\text{42)}} \end{equation*} |
Figures 5(a) and (b) show a schematic of the reactor after the experiment and the temperature profile of the reactor before the reactions, respectively. As shown in Fig. 5(b), the temperatures at the center and bottom of the crucible, where most of the TiCl4 reduction reaction took place, were 1123 K before the reactions. The reaction between TiCl4 and Mg was highly exothermic, as shown in eq. (2). Figure 6 shows the actual temperature measured at the center of the Fe crucible when the reduction of TiCl4 by Mg was performed at the different reaction temperatures or TiCl4 feeding rates. When the reaction proceeded at 1123 K, as shown in Fig. 6(a), the temperature increased from 1123 K to 1313 K. The eutectic temperature of the binary phase diagram of the Fe and Ti system is 1358 K. Consequently, the Fe concentration of the residues obtained in the bottom of the Fe crucible was 2.43 mass%, as shown in Table 4. However, Fe was not detected in other residues. Thus, the contamination of Fe from the Fe crucible is necessary to be prevented by controlling the supplying rate of TiCl4 to increase the production for the practical application.
(a) Schematic of the reactor after the experiment and (b) temperature profile of the reactor which was measured along the vertical distance from the bottom of the reactor when the temperature was 1123 K before reactions.
The influence of the (a) reaction temperature and (b) TiCl4 feeding rate on the temperature inside the center of the Fe crucible during the experiments.
Furthermore, the increase in the temperature owing to the highly exothermic reaction contributed to the O and H concentrations as well as phases in the residues obtained. The results of the XRD analysis and the analytical results of the concentrations of O and H of the residues obtained in various positions after the magnesiothermic reduction of TiCl4 using Mg in an H2 gas atmosphere at 1123 K (Exp. no. 200826) are shown in Figs. 7 and 8, respectively. TiH1.924 was only found in the upper part of the TiCl4 injection tube, as shown in Fig. 7(a), whereas a mixture of Ti and TiH1.5 was produced on the lower part of the TiCl4 injection tube and in all positions of the Fe crucible, as shown in Figs. 7(b)–(e).
Results of XRD analysis of the residues obtained in various positions after the experiment (Exp. no. 200826): (a) the upper and (b) lower part of TiCl4 injection tube, and (c) the center, (d) wall, and (e) bottom of the crucible.
The concentrations of oxygen and hydrogen in the residues obtained in various positions after the experiment (Exp. no. 200826).
The production of TiH0.54 is expected at 1123 K in 1 atm of $p_{\text{H}_{2}}$ under equilibrium conditions, as shown in Fig. 2(b). However, a mixture of Ti and TiH1.5 was produced because the reaction was not conducted under equilibrium. Equilibrium could not be realized because the actual temperature fluctuated during the reaction time owing to the exothermic reaction, and also the reaction time was insufficient to reach equilibrium. The average H concentration in the residues excluding the residue obtained at the upper part of the tube was 0.99 mass%, as shown in Fig. 8.
In contrast, the temperature in the region of the upper part of the TiCl4 injection tube was less affected by the exothermic reaction. This was because that region was located at a distance of 28 cm from the center of the crucible, as shown in Fig. 5(b). In addition, as shown in Fig. 2(c), the equilibrium $p_{\text{H}_{2}}$ for TiH2 at 908–974 K was 0.32–0.58 atm.41) Therefore, the production of TiH1.924 was reasonable because the temperature range of the upper part of the TiCl4 injection tube was 908–974 K and $p_{\text{H}_{2}}$ was 1 atm in the experiments. As a result, the production of TiH1.924 at the upper part of the TiCl4 injection tube was in line with the thermodynamic data on the equilibrium $p_{\text{H}_{2}}$ for TiH2 formation.41)
Figures 7(b)–(e) show that although the experiments were carried out at 1 atm of $p_{\text{H}_{2}}$, the production of metallic Ti was inevitable and TiH1.5, instead of TiH1.924, was produced. This is because dehydrogenation of Ti hydride occurred. Ti hydride was expected to be produced at the early stage of TiCl4 reduction because $p_{\text{H}_{2}}$ was 1 atm. However, as the reduction proceeded, the dehydrogenation of the Ti hydride occurred, producing a mixture of metallic Ti and TiH1.5. One reason for the dehydrogenation of Ti hydride is the restriction of the H2 gas supply. As shown in eq. (2), the amount of MgCl2 produced increased as TiCl4 reduction proceeded. As a result, the diffusion of H2 gas to the residues was impeded because the produced MgCl2 covered the surface of the residues, as shown in Fig. 4(b).
Another reason for the production of a mixture of metallic Ti and TiH1.5 is the increase in actual temperature owing to the highly exothermic reaction. As shown in Fig. 6, the actual temperature at the center of the Fe crucible increased as the reduction proceeded, owing to the highly exothermic reaction. Equation (6) shows that when $p_{\text{H}_{2}}$ is maintained at 1 atm, the production of TiH2 by the reaction of Ti with H2 gas would proceed forward at a temperature below 1047 K. However, the atomic ratio of H to Ti of Ti hydride will decrease as the temperature increase when a temperature is larger than 1047 K because the tendency of the dehydrogenation of TiH2 will increase as the temperature increase. Therefore, the dehydrogenation of Ti hydride occurred as the TiCl4 reduction proceeded, owing to the temperature increase.
| \begin{equation} \text{Ti ($s$)} +\text{H$_{2}$ ($g$)} = \text{TiH$_{2}$ ($s$)} \end{equation} | (6) |
| \begin{equation*} \Delta G^{\circ}{}_{\text{r}} < 0\ \text{when}\ T < 1047\,\text{K}\ \text{under}\ p_{\text{H${_{2}}$}} = 1\,\text{atm}^{\text{41)}} \end{equation*} |
In addition, although the lower part of the TiCl4 injection tube was located in the region 974–1089 K, the lower part of the TiCl4 injection tube was also affected by the exothermic reaction because it was close to the center of the crucible. Therefore, the extent of dehydrogenation of Ti hydride that occurred in the lower part of the TiCl4 injection tube was identical to that in the center and bottom of the crucible. In summary, the formation of the mixture of metallic Ti and TiH1.5 resulted from the dehydrogenation of Ti hydride because the temperature increased owing to the highly exothermic reaction, and the supply of the H2 gas into the residues produced was impeded owing to the covering of the surface of the residues by MgCl2.
The results of the H analysis of the residues obtained were in good agreement with the results of the XRD analysis, as shown in Fig. 8. The highest H concentration was 3.56 mass% for the residue obtained in the upper part of the TiCl4 injection tube, identified as TiH1.924. Additionally, for the residues obtained in the lower part of the TiCl4 injection tube and in the center and wall of the Fe crucible, identified as a mixture of Ti and TiH1.5, the concentration of H was 1.02 mass%, 1.11 mass%, and 0.85 mass%, respectively.
However, the concentration of O in the residues obtained in the TiCl4 injection tube was higher than that in the Fe crucible, as shown in Fig. 8. The concentration of O in the residues obtained in the upper and lower parts of the TiCl4 injection tube and in the center and wall of the Fe crucible was 0.555 mass%, 0.489 mass%, 0.169 mass%, and 0.151 mass%, respectively. Figure 9 shows FE-SEM images of the residues obtained in various positions after the experiment (Exp. no. 200826). Aggregates of fine primary particles were found in the residues obtained from the upper part of the TiCl4 injection tube, as shown in Fig. 9(a). The size of the primary particles increased as the residues obtained were located closer to the center of the crucible, as shown in Figs. 9(b)–(d). Figure 9(e) shows TiH2 powders produced by the ADMA process35) whose primary particle size is significantly larger than the residues obtained in this study. This is because TiH2 was severely sintered during vacuum distillation at 1273–1323 K for a long processing time.
FE-SEM images of the residues obtained in various positions after the experiment (Exp. no. 200826): (a) the upper and (b) lower part of TiCl4 injection tube, and (c) the center and (d) wall of the crucible, and (e) microstructure of TiH2 powder produced by the ADMA process.35)
The concentration of O in a powder is determined by the O dissolved in the powder as a solid solution and the oxide layer on the surface of the powder. As the particle size decreased, the influence of the oxide layer on the concentration of O in the powder increased. As a result, when the sintering effect was enhanced with an increase in the temperature, the primary particles were further coarsened and densified, and thus, the size of the primary particles increased with temperature. Hence, a decrease in the specific surface area of the residues resulted in a decrease in the concentration of O in the Ti powder. In addition, it is worth noting that the influence of contamination during leaching on the concentration of O will decrease when the surface area of the residues decreased by increasing the primary particle size. Therefore, the concentration of O in the residues obtained from the TiCl4 injection tube was higher than those from the Fe crucible owing to the high specific surface area of the residues, as shown in Fig. 9.
4.3 Influence of cooling gas on the production of Ti hydrideFigures 10, 11, and 12 show the results of the XRD analysis, the concentrations of O and H, and the FE-SEM images of the residues obtained in the center of the crucible after the experiments, respectively, according to various experimental variables, such as reaction temperature and TiCl4 feeding rate.
Results of XRD analysis of the residues obtained in the center of the crucible after the experiments according to various experimental variables; (a) Cooling in H2 atmosphere; TiCl4 feeding rate: (b) 39.20 g·min−1 and (c) 57.42 g·min−1; Temperature: (d) 1073 K and (e) 1173 K.
The influence of the (a) cooling gas used, (b) TiCl4 feeding rate, and (c) reaction temperature on the concentrations of oxygen and hydrogen in the residues obtained in the center of the Fe crucible after the experiments.
FE-SEM images of the residues obtained in the center of the crucible after the experiments according to various experimental variables; TiCl4 feeding rate: (a) 39.20 g·min−1 and (b) 57.42 g·min−1; Temperature: (c) 1073 K and (d) 1173 K.
To demonstrate the feasibility of the production of TiH1.924 by further hydrogenation of the mixture of Ti and TiH1.5 produced in the Fe crucible, H2 gas was used for cooling instead of Ar gas after the experiment at 1123 K at an H2 gas flow rate of 1500 sccm and TiCl4 feeding rate of 55.83 g·min−1. Although the supply of H2 gas was impeded by the MgCl2 that covered the surface of the particles of the residues, further hydrogenation of the mixture of Ti and TiH1.5 may occur when the diffusion of H2 gas to the Ti mixture through MgCl2 was not highly limited.
Unfortunately, Fig. 10(a) shows that TiH1.924 was not produced. However, Fig. 11(a) shows that the hydrogenation further proceeded because the concentration of H increased to 1.98 mass%. These results indicate that the proportion of TiH1.5 increased whereas that of Ti decreased. Because only TiH1.5 was produced as Ti hydride, as shown in Fig. 10(a), the proportions of Ti and TiH1.5 in the residues were calculated. Consequently, the proportion of TiH1.5 increased from approximately 35% to 65% when the cooling gas changed from Ar gas to H2 gas. Additionally, with an increase in the proportion of TiH1.5 in the mixture of Ti and TiH1.5, which demonstrated high resistance to oxidation and dilute acid solutions, the amount of contamination during leaching decreased.43,44) As a result, the concentration of O decreased from 0.212 mass% to 0.182 mass% when H2 gas was used instead of Ar gas for cooling, as shown in Fig. 11(a).
4.4 Influence of TiCl4 feeding rate on the production of Ti hydrideThe mixtures of Ti and TiH1.5 were also obtained after the magnesiothermic reductions of TiCl4 using Mg metal at 1123 K when the H2 gas flow rate was 1500 sccm and TiCl4 feeding rates were 39.20 g·min−1 and 57.42 g·min−1, as shown in Figs. 10(b) and (c), respectively. As shown in Fig. 6(b), the actual temperatures in the center of the Fe crucible increased from 1123 K to 1336 K and 1388 K when the TiCl4 feeding rates were 39.20 g·min−1 and 57.42 g·min−1, respectively. This was because the amount of heat of the reaction increased with an increase in the TiCl4 feeding rate. Therefore, the concentration of H in the residues decreased from 1.11 mass% to 0.91 mass% with an increase in the TiCl4 feeding rates from 23.43 g·min−1 to 57.42 g·min−1, as shown in Fig. 11(b), because dehydrogenation further proceeded as the temperature increased. The proportion of TiH1.5 in the mixture of Ti and TiH1.5 decreased from approximately 37% to 30% when the TiCl4 feeding rate increased from 23.43 g·min−1 to 57.42 g·min−1.
Figure 11(b) also shows the change in the concentration of O in the residues depending on the TiCl4 feeding rate. When the TiCl4 feeding rates were 23.43 g·min−1, 39.20 g·min−1, and 57.42 g·min−1, the concentrations of O in the residues were 0.163 mass%, 0.150 mass%, and 0.212 mass%, respectively. Notably, although the actual temperature in the center of the Fe crucible increased with increasing TiCl4 feeding rate, the sintering effect was not sufficiently strengthened, as shown in Figs. 12(a) and (b). The aggregates of the primary particles shown in Fig. 12(b) appeared smaller than those shown in Fig. 12(a). An increase in the TiCl4 feeding rate increased the temperature inside the reactor. However, the coarsening was not sufficiently advanced because the reaction time was shortened with the increase in the TiCl4 feeding rate, resulting in a shorter dwelling time of residues at high temperatures. Therefore, despite the increased temperature due to increasing the TiCl4 feeding rate, the concentration of O in the residues was higher when the TiCl4 feeding rate was 57.42 g·min−1 than when it was 23.43 g·min−1 and 39.20 g·min−1, owing to the decrease in the primary particle size and the proportion of TiH1.5 in the mixture of Ti and TiH1.5.
4.5 Influence of reaction temperature on the production of Ti hydrideFigures 10(d) and (e) show that the mixture of Ti and TiH1.5 was obtained when the magnesiothermic reductions of TiCl4 using Mg metal were performed at 1073 K and 1173 K, respectively, and the TiCl4 feeding rates were in the range of 23.81–23.47 g·min−1 and the H2 gas flow rate was 1500 sccm. As shown in Fig. 6(a), the actual temperatures at the center of the Fe crucible increased from 1073 K to 1255 K and from 1173 K to 1324 K as the reduction of TiCl4 proceeded in an H2 atmosphere because the reaction was highly exothermic. The concentration of H in the residues decreased from 1.36 mass% to 1.29 mass% with an increase in reaction temperature from 1073 K to 1173 K, as shown in Fig. 11(c). In addition, the proportion of TiH1.5 in the mixture of Ti and TiH1.5 slightly decreased from approximately 45% to 43% when the reaction temperature increased from 1073 K to 1173 K. These results indicate that the dehydrogenation proceeded further as the temperature increased.
As shown in Fig. 11(c), the concentration of O decreased from 0.172 mass% to 0.116 mass% in the residues as the temperature increased from 1073 K to 1173 K. The coarsening and densification of the particles proceeded further when the temperature increased, as shown in Figs. 12(c) and (d). As a result, the specific surface area decreased when the temperature increased from 1073 K to 1173 K. Furthermore, these results show that the concentration of O in the residues was mainly affected by the specific surface area of the particles of the residues rather than the proportion of TiH1.5 in the mixture of Ti and TiH1.5 for these experiments. This was because the concentration of O decreased from 0.172 mass% to 0.116 mass% even though the concentration of H in the residues decreased from 1.36 mass% to 1.29 mass% when the reaction temperature increased from 1073 K to 1173 K.
The direct production of Ti hydride from TiCl4 using Mg metal and H2 gas was investigated. A mixture of metallic Ti and TiH1.5 was obtained in an Fe crucible after the experiments. The dehydrogenation of Ti hydride occurred with the TiCl4 reduction because of the restriction in the H2 gas supply owing to increased MgCl2 amount as well as increased temperature owing to the heat of reaction. Consequently, the concentration of hydrogen in the residues increased with a decrease in the TiCl4 feeding rate and reaction temperature. In addition, the concentration of O in the residue decreased with a decrease in specific surface area and/or an increase in the proportion of TiH1.5 in the mixture of Ti and TiH1.5. As a result, the concentration of O in the residue decreased to 0.116 mass% when the experiment was carried out at 1173 K with a TiCl4 feeding rate of 23.47 g·min−1 under an H2 gas flow rate of 1500 sccm. Thus, the process investigated in this study demonstrated the feasibility of the production of the Ti hydride powder that satisfied the criteria for the O concentration of ASTM Gr. 1 for Ti by combining a magnesiothermic reduction of TiCl4 with leaching.
This research was funded by the Ministry of Trade, Industry, and Energy (MOTIE, Korea) (grant number: 10063143).
