Mechanically Activated Rutile and Ilmenite as the Starting Materials for Process of Titanium Alloys Production

The consumptive conventional process of titanium alloys production needs new inno‐ vative processes. As starting materials for aluminothermic reduction, natural TiO 2 and FeTiO 3 concentrates can be used. The keynote of the present chapter is mechanical acti‐ vation as a pre‐treatment step for these concentrates, which is realized by the milling in a vibratory industrial mill. Mechanically activated rutile ore used in aluminothermic reduction saved 30% booster expenses and decreased Cl 2 emissions. Mechanical activa‐ tion of ilmenite and ilmenite/aluminum mixtures was performed, and the kinetics of subsequent hydrometallurgical production of synthetic TiO 2 by pressure and normal leaching were studied. New processes with the coupling of aluminothermic production of titanium alloys were proposed.


Introduction
The production of titanium alloys by using the conventional process route is tremendously cost-intensive. Therefore, there is a huge endeavor for an alternative process route. In an aluminothermic reduction process, titanium oxide can be used to achieve a titanium alloy. The energy which is needed for the autothermic reduction is released by the reduction of titanium oxides as well as by metal oxides which are needed for the titanium alloy composition. As the energy released by these oxides is not sufficient, there are boosters like KClO 4 or CaO 2 that are necessary. To reduce the amount of these boosters, the process of mechanical activation of titanium oxide (rutile, anatase, and ilmenite) is helpful.

Fundamentals of mechanical activation
According to Butyagin, the mechanical activation of solids is defined as an increase in reaction ability due to stable changes in solids structure [1]. Mechanical activation represents a multi-step process with changes in the energetic parameters and the amount of accumulated energy of solids in each step, and it is followed by the process of the defects accumulation, amorphization, formation of metastable polymorphous forms, and chemical reactions [2]. Depending on the number of solid phases involved, the solid-state reactions by mechanical activation are classified into single (homogeneous or inhomogeneous reactions) and multi-phase systems (heterogeneous reactions) (see Table 1). The homogeneous single-phase solid-state reactions are the defect reactions and inhomogeneous reactions are inter-solid diffusions with the concentration gradients. During the heterogeneous multi-phase solidstate reactions, the mass transport diffusion occurs across the phase boundaries, and one or more product phases are created [3][4][5][6].

Type of milling devices
The main problem for the technical application of the mechanical activation of minerals or ores is a suitable type of the mill. For initiation of mechanical activation, the impact load mechanism is required. This is performed in the vibratory mills that are manufactured in industrial scale. For mechanical activation examination only the lab-scale mills are commonly used. They are based on the pressure load mechanism due to centrifugal force of motion. An industrial eccentric vibratory mill (type: ESM 656-0.5 ks, Siebtechnik GmbH, Germany) for the mechanical activation of TiO 2 (rutile) and FeTiO 3 (ilmenite) concentrates was used (see Figure 1).

Mechanical activation
In order to identify the reaction ability of the mechanically activated rutile at the subsequent metallurgical reaction, the mechanical activation degrees, I/I 0 are assigned. The ratios I/I 0 represent the X-ray diffraction intensities at the lattice plane (110) of rutile, whereas I 0 is the measured value for the untreated rutile (defined as 100% < 10 μm), and I is value for mechanically activated rutile. This ratio I/I 0 is a parameter that gives information about grain size, lattice defects, and solid-state reactions. The results in Figure 2 showed the activation degree of the rutile concentrate milled for 1, 2, and 3 h was dependent on the specific energy consumption of milling. The degree of crystallinity, I/I 0 of rutile crystal structure, decreased from 0.7 to 0.25, with an increase in milling time [6].

Aluminothermic reduction process for titanium alloys production
The materials with a decreased crystallinity due to mechanical activation can be very useful for the aluminothermic reduction process to reduce the amount of boosters. An aluminothermic reaction presents the reduction of a metal oxide with aluminum as a reductant. The reduction is feasible when Al shows a greater chemical affinity for the non-metal element of the compound than the desired metal which should be reduced. Regarding the redution of TiO 2 with  Al, the intrinsic heat of the reaction is not sufficient to maintain the self-sustained reaction [7]. Therefore, boosters (e.g., KClO 4 , CaO 2 ) are added to reach the needed energy density. The modeling of the aluminothermic reaction mixture needs input parameters such as the required energy density, targeted adiabatic temperature, estimated heat losses, as well as aimed slag composition in order to improve the metal/slag separation. In order to reach the targeted temperature, changes in the mixture are not allowed to effect or vary the product alloy composition [7]. Because of little difference in the energy density from 50 to 100 J.g −1 of aluminothermic reaction, the process window has to be exactly defined to supply slow propagation of reaction and early solidification. Rutile ore concentrate contains additional by-components which make the determination of the process conditions more complex. Table 2 gives an overview of the reduction reactions of the by-components included in rutile concentrate [6].
The heats of these reactions differ significantly. Although the amount of these by-components are not very high, the released heats of each reaction have to be taken into account due to the above explained narrow process window. The reduction of several by-components is more favorable than the reduction of TiO 2 thus remaining within the metal phase. Therefore, the composition of the master alloy Ti-6Al-4V was carefully chosen, which corresponds to 60 wt.% of Ti, 24 wt.% of Al, and 16 wt.% of V. The investigations were focused on the definition of the required energy charging for the aluminothermic reaction with mechanically activated rutile to obtain a stable product. Important is the composition of the product, especially the contents of Ti, Al, and V and the by-components as well as the oxygen content in order proceed in the upcoming refining steps to achieve a valuable titanium alloy that can be produced cost-efficiently and can be used therefore in the car industry and other light-weight applications. It has to be balanced out on how much booster can be saved by charging a reasonable mechanically activated rutile, minimizing the introduced energy for milling and maximizing the saved amount of booster [6]. Based on the promising results of the preliminary trials, further trials were conducted in small (8.4 kg of mixture) and mid-scale (18.2 kg of mixture) levels, with a variation in the reaction parameters such as time of mechanical activation of rutile concentrate (0, 1, 2, and 3 h), grain size of Al (90-300, 500-800, and 700-1200 μm), and KClO 4 addition to achieve energy density of 2400 and 2500 J.g −1 for the aluminothermic reaction. The selected range for variation of each reaction parameter is shown in Figure 3.  The non-activated rutile (0 h) was used to compare the efficiency of the activation. The activation duration varies between 1 and 3 h. There are three types of Al grain sizes. Due to pre-trials, the two energy densities of 2400 and 2500 J.g −1 are used by KClO 4 addition [6]. This experimental setup indicates clearly that the trials with 2 h activation time and fine Al particles showed the best results according to the reaction time of the aluminothermic reduction and a good metal-slag separation. The metal composition was also the best in these trials. Small amounts of Fe and Si could be detected which would not interfere the final target alloy, Ti-Al6-V4 [6].

Equations of the by-component reactions
Because of avoiding the production of TiO 2 via chlorination, the savings were up to 50%, and because of mechanical activation, the use of KClO 4 decreased by 30% remarkably [6]. Due to the fact that lime is used as slag component to decrease the liquid temperature of the final slag, there are investigations to avoid the KClO 4 completely by using CaO 2 which will be reduced by Al to CaO and remains as slag component without producing any gas emissions. As the released heat by KClO 4 is nearly four times as high as the one by CaO 2, there is much more CaO 2 required to reach the needed energy density. Due to thermodynamic calculations, the amount of CaO should not exceed 50% because the liquid temperature of the slag will increase again and because of economic reasons, it should be used as sparsely as possible. As the energy density is decreased by mechanical activation, it is theoretical possible to reach a good amount of CaO in the slag and the needed energy density.
First trials were conducted with a 2 h mechanically activated rutile ore with CaO 2 as booster, varying the energy density. Besides the four trials which were conducted with an energy density from 2250 up to 2350 J.g −1 , the results were comparable to the experimental trials before. Due to the lower energy density in these trials, the metal-slag separation was not sufficient which resulted in high oxygen content in the alloy. The results for the oxygen content for the various trials varying the energy density can be seen in Figure 4.
A closer look to the trials with an energy density of 2450 J.g −1 shows that low oxygen contents can be reached. But there is also a large deviation for the oxygen content. This leads to the assumption that the oxygen may not be completely diluted. Instead, there could be small oxide particles in the metal fraction which is a result of the slag-metal separation. Further investigations have to be performed. Therefore, a bigger-scale experiment to improve the separation needs to be done as well. Remelting this material in a vacuum induction furnace with a special ceramic crucible will be investigated to homogenize the material.
Mechanically activated rutile ore can be used for aluminothermic reduction to produce a titanium master alloy. KClO 4 can be used, as well as CaO 2 as booster. In this regard, a remarkable amount of KCl gas could be avoided. Using just CaO 2 as a booster could avoid the production of KCl completely.

FeTiO 3 concentrate, ilmenite
The beneficiation of ilmenites requires pyrometallurgical or chemical process steps to separate the iron content of approximately 30%. There are some methods of iron separation from FeTiO 3 , which vary in their technical and energy demands. The chemical sulfate process, with H 2 SO 4 at < 220°C, yielding TiO 2 in pigment quality, is difficult and complex because of the low solubility of ilmenite in H 2 SO 4 [3]. Nevertheless, for the production of synthetic TiO 2 concentrates except for the pyrometallurgical processes [8,9], the connected pyro-and hydrometallurgical processes [10][11][12][13] were developed. One possibility of the direct hydrometallurgical processing of FeTiO 3 is its pre-treatment by mechanical activation [14]. The studies on the solubility of FeTiO 3 after mechanical activation using different mills for ultrafine grinding such as vertical ball mills (attritors), planetary ball mills, and drum mills have been published, in which the structural changes of FeTiO 3 without technical applicability was focused [15][16][17]. In general, due to mechanical activation, the solids are exposed to high mechanical stress, which is responsible for their specific surface area increase, the crystalline structure defects formation and leads to the enthalpy increase. Hence, mechanically activated solids and minerals with low solubility are more leachable in subsequent hydrometallurgical process [18]. The application of mechanical activation of FeTiO 3 by energy-efficient milling as a pre-treatment step in the hydrometallurgical process of synthetic TiO 2 concentrates production might be a new realization to utilize such TiO 2 materials (~95%) in aluminothermic Ti alloys generation.

Mechanical activation and the kinetics of subsequent hydrometallurgical production of synthetic TiO 2 by pressure leaching
The investigations were carried out with FeTiO 3 concentrates (>95% FeTiO 3 , <5% SiO 2 ) of Russian origin (GMD, Mineral Trade Company For determination of the optimal milling conditions, a parameter study was done: mill feed quantities varied from 100 to 300 g/charge, activation times ranged from 15 to 60 min, the amplitude of inhomogeneous vibrations was 20 mm, and the revolutions of the motor of the mill, 960 min −1 , were constant. The steel balls of 30 mm diameter were applied. Activation degrees and the ratios I/I 0 at lattice plane (104) of FeTiO 3 after and before mechanical activation as a function of mechanical activation time are shown in Figure 5. It is obvious that the ilmenite structure is strongly strained by the mechanical activation.
The influence of the activation time of FeTiO 3 on the TiO 2 extraction to the precipitation product is shown in Figure 6. The curve for the total Fe extraction in the product reflects the dissolution of FeTiO 3 . The leaching tests confirm the dependence on the activation degree of FeTiO 3 . A critical point is that, after 15 min of mechanical activation, most of FeTiO 3 is dissolved. This finding is extremely interesting from a technical point of view. The hydrolytic precipitation of TiO 2 is influenced by an initial dissolution accelerated with increasing activation.
At temperatures >50°C and a pH value >1.5, the hydrolysis of titanyl sulfate solution, TiOSO 4 to TiO 2, is triggered. In the investigated temperature range of 100-150°C, the dissolution of FeTiO 3 and the simultaneous precipitation of TiO 2 take place in parallel.
The leaching time was varied in a range from 15 to 90 min. A general fact is that the dissolution rate of FeTiO 3 is fast. Figure 7 shows the dissolution of ~64% FeTiO 3 (corresponding to the dissolution of Fe) during 30 min of leaching at 120°C. By increasing the reaction temperature to 150°C, ~86% of FeTiO 3 dissolves during 30 min of leaching. It results from Figure 7 that the leaching time had a relatively low effect on the hydrolysis. After 60 min of leaching, the recovery   of precipitated TiO 2 achieved ~93% and the content of Fe in TiO 2 decreased to <6%. The synthetic TiO 2 product assigned a relatively high purity. By decreasing the leaching temperature from 150 to 120°C, the obtaining of the high-quality synthetic TiO 2 concentrate is impossible [19].
It was detected that the addition of Fe powder, used as a reduction agent, influences the initial rate of FeTiO 3 dissolution (86% with an addition of 6% Fe and 89-92% with an addition of 12% Fe) and after a leaching time of 60 min, the influence of the Fe additive is no longer identifiable [19].
Sulfate process requires ~2 tons of concentrated H 2 SO 4 per 1 ton of FeTiO 3 for digestion. Subsequently, the generated digestion cake is leached with H 2 O by pH < 1.5. Therefore, the direct leaching of mechanically activated FeTiO 3 carried out by a described procedure requires a solid/acid ratio of 1:2 at least, which corresponds to 10% H 2 SO 4 and 77% dissolution of FeTiO 3 (mechanically activated for 15 min) after 60 min of leaching, at 150°C, with addition of 10% Fe. By increasing the initial acid concentration to 20%, which corresponds to a ratio of 1:4, FeTiO 3 dissolves to 89%. For technical dimensioning, the ratio of ilmenite to acid should be <1:4. An excessively high amount of acid would complicate the hydrolytic conditions.
The solid contents from 50 to 200 g.l −1 were investigated. In order to show the influence of mechanical activation on the leaching of FeTiO 3 with different s/l ratios, two test series with activation times of 15 and 30 min were carried out. As expected, the residual Fe content in the product increases to 7% with increasing s/l ratio to 200 g.l −1 at a higher activation time. That means the activation time, 15 min, is sufficient to achieve 92% of dissolution of FeTiO 3 with s/l ratio of 200 g.l −1 and with 30% H 2 SO 4 at 150°C.
According to literature, the research on the leaching of the mechanically activated ilmenites demonstrated no technical applicability. For various types of the mills, the milling times of up to 200 h were used, and low s/l ratios (only 10 g.l −1 ) for the leaching process were employed [15-17, 20, 21]. The operating conditions determined in this investigation fulfill the requirements for a technical implementation of the process for the production of a synthetic TiO 2 product (95.23% TiO 2 , 3.32% Fe 2 O 3 , 1.7% SiO 2 , 1.2% CaO, 0.34% Al 2 O 3 , 0.007% P 2 O 5 , and 0.004% ZrO 2 ), which is suitable for use in aluminothermic alloys production. Figure 8 shows the process flowsheet on the coupling of hydrometallurgical processing of FeTiO 3 into synthetic TiO 2 concentrate (anatase) with the aluminothermic production of TiAl alloys.
The energy required per ton of synthetic anatase is expected to be 506 kWh. Based on current costs for energy, this corresponds to approximately 212 €/t of synthetic anatase. This investigated pre-treatment of ilmenite with mechanical activation and leaching enables new cost-effective production methods for titanium-based alloys. So far, only high-purity rutile pigments have been used for the aluminothermic production of TiAl alloys [7,22]. At 95% TiO 2 , the synthetic TiO 2 concentrate (anatase) that we produced meets the requirements for aluminothermy.

Mechanical activation with metal addition and the kinetics of subsequent hydrometallurgical production of synthetic TiO 2 by normal leaching
Mechanical activation of FeTiO 3 with Al powder in the stoichiometric ratio 1:2 caused its mechanochemical reduction already after 120 min. The product phases found on the thermodynamic calculations with 1100-1700°C (Al 2 O 3 , TiO, Fe 2 Ti, and FeTi) have already appeared after 360 min of mechanochemical reduction. Such mechanochemical processing of FeTiO 3 concentrate decreases the temperature and subsequently the time of thermal reduction of FeTiO 3 , which could be also used in the titanium alloys production [23]. For subsequent leaching in H 2 SO 4 (40%) at 50°C, mechanically activated FeTiO 3 with Al in ratio 1:0.3 was tested. In this case, the necessary agglomerates of FeTiO 3 /Al are created even after 15 min of mechanical activation. Figure 9 shows the measured grain size distribution of the unmilled FeTiO 3 and the activated FeTiO 3 /Al mixtures.
The SEM image in Figure 10 shows the formed agglomerates. The contact pressure caused by the impact stress on the agglomerates during the second stage of ultrafine milling (agglomeration stage) leads to structures similar to briquettes with the highest bulk density [24,25].  Figure 11 shows that the maximum Ti recovery of activated FeTiO 3 reaches 23%. Activated FeTiO 3 /Al mixture shows a 53% Ti dissolution after only 15 min of activation.    With a further increase to 300 g/charge, the Ti recovery decreases to 69%. For 600 g/charge, the Ti-recovery reaches to 56%, which is still a high value. Ti recovery for the unmilled FeTiO 3 was 0.3% only. A check of the s/l ratio per liter showed that the optimal ratio was 150 g.l −1 of FeTiO 3 /Al mixture. For technical processes, at least 300 g.l −1 must be feasible. Figure 12 evidences that the leaching rate of FeTiO 3 /Al mixture at temperature 50°C is very fast. After 5 min of leaching, the Ti recovery was 69% and after 60 min of leaching, almost 80% of Ti was leached out.
In Figure 13, the flowsheet of the described process for production of synthetic TiO 2 was proposed [26]. In an open agitator vessel, FeTiO 3 /Al mixture is dissolved at a temperature of 50°C since the strong bond in the briquetted mixture follows the shrinking core model [26]. Undesirable metals, Fe and Al, can be crystallized as a mixture of Fe, Al sulfates at temperature < 15°C. Ti(OH) 3 precipitates during hydrolysis and by calcination oxidizes to TiO 2 (anatase) with the following composition: 99% TiO 2 , 0.59% Fe 2 O 3 , <0.1% Al 2 O 3 , <0.2% SiO 2 , <0.03 ZrO 2 , and 0.01% Cr 2 O 3 .

Conclusion
New technological processes of using titanium oxide concentrates (rutile, ilmenite) as starting materials for titanium alloys production were proposed. The first one is based on mechanical activation of natural rutile concentrate with milling, which caused the reducing of booster (KClO 4 ) amount around 30% and saving expenses for previous chlorination in subsequent aluminothermic reductions. Promising steps to replace the KClO 4 could be observed but are still under further investigations. The second process produces synthetic TiO 2 (anatase) from mechanically activated ilmenite concentrates treated by pressure leaching in dilute H 2 SO 4 at 150°C. The third process presents the improvement of the previous one because ilmenite concentrate can be hydrometallurgically processed to synthetic TiO 2 by normal leaching in dilute H 2 SO 4 at 50°C. Ilmenite concentrate has to be pre-treated by mechanical activation with aluminum as a reducing agent which leads to formation of FeTiO 3 /Al agglomerates and a new effective reaction mechanism of leaching. After the separation of iron and aluminum by means of crystallization, it is possible to extract synthetic TiO 2 as a raw material for aluminothermic production of titanium alloys.