Reduction of Titanium Dioxide to Metallic Titanium by Nitridization and Thermal Decomposition
2017 年 58 巻 3 号 p. 361-366
詳細
2017 年 58 巻 3 号 p. 361-366
We report a novel process to produce metallic titanium from titanium dioxide—the raw material typically used in the conventional production of titanium metal—via a titanium nitride (TiN) intermediate. TiN is more easily reacted than titanium oxides such as titanium monoxide and titanium dioxide, and shows thermodynamic reactivity equivalent to titanium tetrachloride (TiCl4), which is used industrially as an intermediate in the conventional metallic titanium manufacturing process. The thermal decomposition temperature of TiN (～3500 K), determined from a thermodynamic database, is also significantly lower than those of the oxides (～5300 K) and TiCl4 (～6200 K); thus, it is suitable for use in the available temperature range of an electric arc furnace (～4000 K). Here, we demonstrate the use of TiN as an intermediate for the manufacture of metallic titanium by thermolysis using the arc melting method.
Titanium and titanium alloys have been applied in the field of industrially manufactured articles, including medical implants and aircraft structures, and are used in the construction of marine equipment and power plants. The global market production and demand for titanium have been rapidly increasing on an annual basis1). The Kroll method2) is used industrially to produce high purity sponge titanium; however, it has a number of disadvantages in terms of industrial mass production when compared to other high performance processes, such as ironmaking. For example, in this process, raw TiO2 ore is converted to TiCl4 in the presence of coke and chlorine gas. This necessitates the use of metallic magnesium to reduce the TiCl4 intermediate, which is electrochemically recycled along with the chlorine from magnesium chloride (MgCl2) by electrolysis (and thus requires a large amount of electricity). Then, a slow vacuum distillation process to remove the Mg and MgCl2 residues resulted from the reduction is necessary. Further, the reaction chamber requires prolonged cooling during the titanium reduction process due to the exothermic nature of the reaction. In the typical titanium melting, the electrodes to be melted in an arc melting furnace is usually shaped using the sponge titanium, and require a special size and form suitable for melting operation; this form control process prior to melting necessitates difficult technologies.
Chen et al.3) recently reported a novel process for titanium production based on an electrochemical reaction in molten calcium chloride, resulting in the formation of small metallic titanium particles. Suzuki et al.4,5) reported a similar electrochemical reduction process in a molten salt medium, and carefully investigated the thermodynamic relationships underlying the reduction process. In addition, other groups6,7) have developed novel production processes based on the magnesiothermic and calciothermic reductions of TiCl4 and titanium concentrates using the salts MgCl2 and CaCl2, respectively. Because these reduction processes appear to produce metallic titanium particles, they are not fully or immediately suitable for replacing the conventional industrial process. Moreover, the chloride employed in these processes results in cracking of the titanium metal8) depending on increasing of its concentration causing by an unsatisfied distillation process and an usage environment, and decreases a glass-forming ability of metallic glasses widely amorphous materials9). Consequently, chloride is not suitable for use in the manufacture process of metallic material including titanium, especially titanium based-metallic glasses, and by extension, its use in titanium production should also be avoided, if it is possible.
Here we show an experimental investigation of a novel process to produce titanium from titanium dioxide. This method is based on the thermodynamics and avoids the use of chloride, and the electricity-consuming electrochemical process for the recycling of process byproducts is unnecessary. It is our anticipation that this processing technique, based on the chemical conversion of TiO2 to TiN and subsequent thermal decomposition of TiN, will be cost-effective and environmentally friendly.
The titanium ore consists mostly of ilmenite, an oxide of iron and titanium, and is usually upgraded to highly purified titanium oxide. The bonding strength of titanium to oxygen and other elements can be commonly compared based on the differences in their Gibbs free energies of formation as a function of temperature. Gibbs free energy changes can generally be found in the thermodynamic data10), and are depicted in Fig. 1 for the systems employed herein. Ilmenite (FeO·TiO2) is a main component of titanium ore, and is commonly employed in the production of metallic titanium, along with TiCl4. As proposed herein, titanium nitride (TiN) can be also used as an intermediate in this process; this material is also compared in thermodynamic terms with other intermetallic compounds in Fig. 1. TiO and TiO2 are typically highly stable, as demonstrated by their large negative Gibbs free energies of formation. TiCl4, however, has a less negative Gibbs free energy of formation, and so it should be more easily reduced to metallic titanium than either TiO or TiO2. Similarly, the free energy of formation of TiN is comparable to that of TiCl4, and so a similar reducibility would be expected. In addition, the sign of the Gibbs free energy of formation of TiN changes at ～3500 K, indicating its instability at higher temperatures, at which it is thermolytically separated into metallic titanium and nitrogen gas. Thus, when TiN is employed as the intermediate material for the production of titanium, metallic titanium can be produced thermodynamically by heating TiN to its decomposition temperature without requirement for a reductant such as metallic magnesium, which is employed in the conventional process. Moreover, because the thermal decomposition does not depend on a rate-determining surface reaction using a reductant, such as the reduction of TiCl4 by Mg, the a high reduction rate would also be expected.
Variation in the Gibbs free energies of formation of the titanium-based intermetallic compounds TiO, TiO2, TiCl4, and TiN.
Figure 2 compares the stabilities of TiO (as a highly stable titanium oxide) and TiN based on their Gibbs free energies of formation in both direct and indirect reactions (i.e., indirect reaction means combustion with carbon). Although the major intermetallic component of the initial ore is TiO2, the degree of oxidation of the titanium changes in turn to Ti4O7, Ti3O5 and TiO as the reduction proceeds. Among these species, TiO is the most stable oxide based on its Gibbs free energy of formation value, and therefore, for comparison purposes, TiO is conveniently considered as a model in Fig. 2. For the direct reaction, TiO is stable between ～1000 and 2200 K. For the indirect reaction, the Gibbs free energy of formation varies between positive and negative values depending on the temperature, with TiN becoming stable >1000 K. This indirect reaction was studied by White et al.11), who experimentally confirmed the formation of TiN at 1423 K. Furthermore, the directions of both reactions are thermodynamically determined by the N2 partial pressure, and are dependent on the equilibrium constants of the reactions10). For example, upon heating the direct reaction to 1800 K, the partial pressure of N2 is ～1 atm and that of O2 is 2.6 × 10−12 atm; thus, TiN stabilization by control of the atmosphere alone is difficult. In contrast, for the indirect reaction at 1800 K, as the partial pressures of N2 and CO are 0.9998 atm and 1.8 × 10−4 atm, respectively, the stabilization of TiN is easily achieved using commercial grade nitrogen gas.
Variation in the Gibbs free energies of formation of TiN from TiO by both direct and indirect reactions (where the latter also involves the combustion of carbon).
TiO2 powder (reagent grade, Wako Pure Chemical Industries, Ltd.) was lightly inserted into a carbon crucible and heated to 1910 K in a vertical electric furnace with an alumina tube at a rate of 300 Kh−1 (0.08 Ks−1). The final temperature was maintained for 1 h (3.6 ks) under a flow of nitrogen gas. The crucible was rapidly cooled to 300 K (room temperature), and the product formed in the carbon crucible was analyzed by X-ray diffraction.
Next, we carried out the conversion of TiN to metallic Ti by arc melting. Although the TiN powder produced in the above process was suitable for the following stages, the production of large quantities of TiN on laboratory scale was problematic. This was mainly due to the size limitations of the furnace and carbon crucible on laboratory scale, such that preparation of sufficient TiN quantities to use in the next step was difficult. Thus, we employed commercial TiN powder in the present study. The TiN powder (reagent grade, Kojundo Chemical Laboratory Co., Ltd.) was subjected to arc melting after its insertion into a die with an inner diameter of 20 mm and compression using an oil-hydraulic cylinder. The compressed TiN powder was heated to 3000–4000 K using a common arc melting furnace (DIAVAC Ltd.) under an Ar atmosphere. The product was subjected to arc melting several times, with clean-up of the melting chamber after each step due to staining from the generation of the gas from the product. The resulting product was analyzed by X-ray diffraction (for the metallic titanium, Ultima III, Rigaku Corporation; for TiN, D8 Discover with GADDS, Bruker AXS).
As shown in Fig. 3, a yellow product was formed from the white TiO2 inside the carbon crucible. This may suggest the formation of the yellow TiN.. The X-ray diffraction analysis of the product showed a pattern that was well matched both to the typical pattern for TiN (PDF #11-038-1420, International Centre for Diffraction Data), as shown in Fig. 4, and to that of the TiN powder employed in subsequent experiments.
Photograph of the yellow product formed by the indirect reaction of TiO2, N2, and C to give TiN.
X-ray diffraction pattern of the product shown in Fig. 3, formed from the reaction of TiO2 (as initial maternal, TiO2 is reacted via TiO to TiN as the inserted reaction) in a carbon crucible under a flow of N2 at 1910 K for 1 h.
Figures 5 (a)–(d) outline the TiN decomposition process. Figure 5(a) shows the initial state of the compressed TiN powder. Figure 5(b) shows a partially reduced product exhibiting a lustrous, crystalline, gold-colored surface, which is comparable in appearance to a typical TiN thin film. As the decomposition process progresses, a fully lustrous surface is formed, as shown in Figs. 5 (c), (d). The diffraction pattern of the product shown in Fig. 5 (d) is presented in Fig. 6. It indicates the presence of several different phases, with metallic titanium as the major product (hexagonal structure, triangles). A small amount of the titanium oxide Ti3O5 is also observed (squares). It is reasonable to consider that this titanium oxide forms from the slight amount of oxygen (10−5 to 10−7 atm) contaminating the commercial inert gas (Ar) during the arc melting process, which was resolved by purifying the Ar gas or by altering the melting method. In addition to these two phases, signals corresponding to a third phase were also observed (circles). The diffraction pattern for this phase correlates with that of nitrogen-dissolved titanium, which was experimentally manufactured as a standard under dilute nitrogen gas atmosphere. The nitrogen concentration of this standard is >1 mass%, which is the analytical upper limit of the infrared absorption method. The nitrogen-dissolved titanium and the standard are determined crystallographically to be either metallic titanium or titanium nitride, which have the Fm-3m crystallographic phase group based on the cubic structure.
Photographic images of the thermal decomposition process of TiN using the arc melting furnace. (a) The compressed TiN, prior to the reaction. (b) Partially decomposed TiN, having a lustrous surface that is partially covered by a gold-colored thin film of TiN. (c), (d) Metallic titanium product with a completely lustrous surface after decomposition of the TiN.
X-ray diffraction patterns of the product shown in panel (d) of Fig. 5, and matched diffraction patterns of a metallic titanium (hexagonal structure, triangles) as the major product, high-nitrogen-dissolved metallic titanium (circle) prepared as an original standard which was annealed under nitrogen gas atmosphere and referred standard materials.
From the previously reported phase diagram of the titanium nitride system12), metallic titanium exhibiting a cubic structure can dissolve nitrogen up to 10 at% (3 mass%) at 2320 K. Consequently, the saturated concentration of nitrogen is larger than the analytical concentration of the standard. Thus, if the third diffraction pattern in Fig. 5 (d) can be interpreted as a quenching structure containing significant quantities of nitrogen, the nitrogen concentration of the produced titanium would be expected to be <3 mass%. In addition, TiN is a non-stoichiometric compound with a nitrogen content ranging from 30–50 at% (11–23 mass%) over a wide temperature range (1380–2620 K). For the diffraction pattern containing both TiN and the (alpha-phase-stabilized) hexagonal multiphase metallic titanium, titanium with nitrogen dissolved at 20–30 at %(7–11 mass%) could be observed in the phase diagram of the titanium nitride system over a wide temperature range. Despite the fact that the issue of the nitrogen concentration remains to be unsolved, these results experimentally confirm that the conventional metallic titanium manufacturing process can be replaced by two simple steps, namely synthesis of TiN from TiO2 and thermal decomposition of TiN.
Figure 7 shows a flowchart illustrating the conventional and proposed processes for the manufacture of metallic titanium. As previously mentioned, the conventional process begins with TiO2, and produces sponge-like metallic titanium via the TiCl4 intermediate. Subsequently, the titanium sponge is subjected to a range of treatments: (1) the removal of magnesium and MgCl2 by distillation, (2) mechanical compression, and (3) arc melting. In addition, as both the magnesium and chlorine gas employed in the conventional process are electrochemically recycled, large amounts of electricity are required for this process, thus increasing the titanium fabrication costs.
Schematic flowchart comparing the current conventional titanium manufacturing process with the novel process presented herein.
As in the conventional process, TiO2 is used as the starting material in our new approach; however, in this case, metallic titanium is produced by the thermolysis of TiN using the arc melting process. TiN can be easily produced by the indirect reaction between carbon, nitrogen, and TiO2 in a process similar to that used for the production of TiCl4. Our novel route therefore allows to eliminate a number of high cost steps/processes to be eliminated, including the reduction of TiCl4 using metallic magnesium, the electrochemical reduction of MgCl2 to regenerate metallic magnesium, the vacuum distillation process, and the compaction of the titanium sponge.
Based on the diffraction patterns examined, the concentration of dissolved nitrogen in the produced titanium is expected to be 3 or 11 mass% as upper limits of nitrogen in these cases. The dissolved nitrogen concentration can be determined theoretically as expressed below:
| \[1/2 \mathrm{N}_{2 {\rm (gas \ in \ atmosphere)}} = \underline{\rm N} _{\rm in \ metal}\] |
| \[K = a_{\rm N} / \left( p_{\rm N2} \right)^{1/2} = {a_{\rm N2}}^* / \left( {p_{\rm N2}}^* \right)^{1/2}\] |
Through the experiments conducted for preparation of titanium via thermal decomposition of TiN using a laboratory-scale arc melting furnace, it was anticipated such that a commercial scale arc melting furnace would be advantageous to the reduction in heat losses because of heating a local spot on the liquid metal surface. The alternative way of reducing nitrogen concentration in the produced titanium is to use a highly purified inert gas to lower the nitrogen partial pressure. The electron beam melting furnace is also able to provide temperature higher than 4000 K and high vacuum to meet the operational conditions for thermal decomposition of TiN, thus may be of practical use, but only if a drastic hardware reconstruction is possible to withdraw considerably large and sound ingots.
We experimentally investigated a process for the reduction of titanium dioxide (TiO2) to metallic titanium based on thermodynamics. Titanium nitride (TiN) was employed as a suitable intermediate for this process, as opposed to the titanium tetrachloride (TiCl4) used in the conventional process. Comparison of the Gibbs free energies of formation of the two intermediates showed that TiN was easily decomposed at around 3000 K, compared to the decomposition temperature of 6200 K for TiCl4. TiN was easily synthesized in a carbon crucible under a nitrogen atmosphere at a reasonable temperature above 1000 K (carbothermal synthesis), which can be easily achieved in a conventional electric resistance furnace. The subsequent decomposition of TiN at 3500 K using an electric arc furnace produced metallic titanium that exhibited a lustrous surface. X-ray diffraction studies confirmed the successful synthesis of metallic titanium, as the obtained diffraction pattern was comparable to that of commercial grade material.
If a process can create increasing production quantity, then the possible range of titanium applications should be expanded with the reliable and economic supply of this material, similarly to steel. Further work will be oriented to the influence of dissolved nitrogen on the quality of the obtained metallic titanium. The present approach is only preliminary in nature. It suffices to say at this time that from the results presented there is positive evidence of extraction of ductile titanium via the proposed method. More extensive instrumentation and many more experiments are necessary for complete clarification of the details on this approach.
This work was supported by the Japan Titanium Society (Titanium Research Aid Program of 2014) and the Ministry of Education, Culture, Sports, Science and Technology (MEXT) of Japan (“Research and Development Project on Advanced Materials Development and Integration of Novel Structured Metallic and Inorganic Materials (2010–2016)”).
