CHLORINATION OF A COLLECTIVE TITANIUM–ZIRCONIUM CONCENTRATE

R. A. Beskrovnaya, S. I. Smolyaninov

(Presented at the interuniversity conference dedicated to M. V. Lomonosov)

Chlorination as a metallurgical method has undergone broad development; chlorine has become one of the cheapest chemical reagents, and materials resistant to chlorine at high temperatures have appeared. Thus, two obstacles that hindered the introduction of the chlorine method into practice have been eliminated.

Titanium tetrachloride is the basic raw material for obtaining this metal, the rapid development of whose production is connected with its important economic and strategic significance. The use of zirconium is most promising in reactor construction for nuclear technology. Nuclear fuel containing zirconium is proposed to be processed by the chlorine method [1].

The literature contains a large number of papers and patents on the chlorination of oxides and ores of rare metals. In view of the fact that the chlorination conditions were varied arbitrarily by each author, the data obtained are in most cases qualitative in nature. The kinetics of chlorination of briquettes made from titanium or zirconium oxides mixed with carbon are presented in works [2, 3]. The thermodynamics of the chlorination reactions of titanium and zirconium oxides in the presence of carbon with gaseous chlorine have been studied [2, 4]. It is considered that the regularities revealed in the chlorination of pure oxides are fully applicable to concentrates as well. Titanium tetrachloride is obtained by chlorinating ore concentrates (ilmenite, rutile) or titanium dioxide obtained from them, titaniferous slags, carbides, and carbonitrides [6, 5, 8].

Zirconium tetrachloride is obtained by chlorinating zirconium concentrate, carbide, or carbonitride [7].

However, in nature there occur complex titanium–zirconium deposits, during the development of which it is easy to obtain a collective titanium–zirconium concentrate, whereas selective beneficiation requires a complex flowsheet, is accompanied by large losses, and the concentrates are contaminated with accompanying minerals. Hence the idea arises: in preparing raw material by beneficiation methods, is it not possible to limit oneself to the separation of a collective concentrate, and to carry out the subsequent separation and isolation of the individual components by metals during the metallurgical conversion?

From this point of view, the use of the chlorine method for processing a titanium–zirconium concentrate is extremely interesting, since in a single operation of chlorination and condensation of chlorides, ...

it is possible to open up the concentrate and separate the chlorination products into individual chlorides or groups of chlorides with identical physicochemical properties.

The importance of this question is evident if only from the fact that, in the cost of rare metals, up to 90% is accounted for by the cost of the raw material.

Experimental Part

Starting materials and their preparation. For the study a titanium–zirconium concentrate was taken, with a content of \(TiO_2 = 38\%\), \(ZrO_2 = 13.82\%\), \(Fe_2O_3 = 17.99\%\), \(SiO_2 = 23.66\). The principal mineralogical components were ilmenite and zircon. Rutile, leucoxene, and certain other minerals were present in small amounts.

Peat from the Tagan deposit was taken as the reducing agent:

\[ W^p = 35\%,\quad A^c = 7.3\%,\quad V^r = 70.0\%. \]

Peat surpasses the reducing agents currently used (soot, charcoal, petroleum coke, anthracite) in high reactivity, availability, low cost, simplification of the briquette technology, and elimination of the necessary binder additive [9].

Briquettes were prepared with the theoretical amount of reducing agent and with a 2- and 3-fold excess of it. Testing of the mechanical strength of air-dried and thermally treated briquettes (at temperatures of 400 and 800°C) showed that their strength is quite sufficient for use in chlorinators with a fixed bed [9].

For the study, briquettes were taken with a 2-fold content of reducing agent relative to the theoretical amount, thermally treated at 800°C. The briquettes were crushed to pieces 3–5 mm in size, and the samples were stored in jars with ground-glass stoppers; before chlorination they were kept in a drying cabinet at 105°C for 2 hours. In doing so, their high hygroscopicity was established, and special attention was paid to drying the charge before chlorination.

Method of Investigation

Studies of the process of chlorinating the titanium–zirconium concentrate were carried out in a laboratory apparatus, the scheme of which is shown in Fig. 1. A porcelain bottom with holes, which supported the weighed portion of the substance under investigation, was inserted into a quartz tube 7 (\(d = 12.8\) mm) heated by a vertical tube furnace. Gaseous chlorine came from cylinder 1. To check for the absence of moisture, it was passed through a column filled with blue silica gel. The chlorine rate was measured by means of capillary rheometer 6 and regulated by needle valve 2 and manostat 5. The chlorides obtained were collected in the lower part of the reaction tube and in receiver 9 with a trap filled with glass wool. For control, a U-shaped tube filled with glass wool was installed. The gas released during chlorination was sampled at definite intervals into pipettes 12 and analyzed on a VTI gas analyzer. A definite weighed portion of briquette was charged into the reaction tube, and air was displaced from the system by passing nitrogen from cylinder 3 through it. Nitrogen was passed into the system until the temperature required by the experimental conditions was established, and then chlorine was admitted into the system at the specified rate. At the end of the experiment the system was purged with nitrogen; in a nitrogen atmosphere the residue after chlorination was cooled to room temperature. The reaction tube, receiver, and U-shaped tube were weighed before and after the experiment; thus

in this way the weight of the residue after chlorination and the total weight of the chlorides obtained were determined.

The degree of chlorination was judged from analysis of the residue, in which the content of the oxides of iron, titanium, and zirconium was determined. The amount of metal removed from the initial charge in the form of chloride was then calculated from the formula, in %,

\[ M=\left(1-\frac{b}{a}\frac{M_b}{M_a}\right)100\%, \]

where \(M\) is the amount of metal removed from the charge;
\(a\) is the weight of the charge for chlorination;
\(b\) is the weight of the residue after chlorination;
\(M_a\) is the metal content in the charge before chlorination;
\(M_b\) is the metal content in the charge after chlorination.

Fig. 1. Schematic of the apparatus.

Discussion of Results

Three series of experiments were carried out to determine:

1. The effect of temperature on the degree of extraction of the components into the sublimate (iron, titanium, and zirconium).
2. The effect of experiment duration at different temperatures on the degree of chlorination.
3. The effect of the chlorine feed rate on the chlorination process. The effect of temperature, at different experiment durations, on the chlorination process of the titanium–zirconium concentrate.

The experiments were carried out at temperatures of 300–800°C. The duration of the experiment was 15, 30, 60, and 120 min. The chlorine rate was 150 ml/min. The results of the experiments are presented in Fig. 2.

The investigations showed that the iron and titanium contained in the titanium–zirconium concentrate begin to chlorinate at 250°C. Moreover, iron is not chlorinated selectively. Up to a temperature of 500°C, iron is chlorinated predominantly; above 500°C, titanium. At a temperature

At 800°C in 60 min all titanium is driven off in the form of TiCl₄. At 800°C iron is converted into chloride only by 90%. The onset of the chlorination reaction of zirconium is about 300°C; at 800°C in one hour, at a chlorine rate of 150 ml/min, zirconium is chlorinated by 65%. In a number of studies the optimum chlorination temperature of iron oxides pressed with carbon was determined as 280°C, and of titanium oxides with carbon as 400–420°C [11, 10].

In the chlorination of natural compounds, which are less reactive than oxides, the optimum chlorination temperatures are higher.

Thus, ferrititanate raw material with carbonaceous material was chlorinated at 500–600°C [12]. Ilmenites and slags are chlorinated at 600–1000°C [13]. The optimum chlorination temperature for the raw material studied lies in the temperature interval 700–800°C for iron and titanium, and for zirconium, apparently, considerably above 1000°C (Fig. 3).

In this same series of experiments the composition of the gas phase was checked for the content of CO, CO₂, and Cl₂ (Table 1).

Table 1

It is shown that the content sharply falls by 800°C. The CO content increases with increasing temperature; the CO₂ content is maximal at a temperature of 500°C.

According to literature data, during the chlorination of titanium dioxide in a mixture with carbon the following reactions occur:

$$ \mathrm{TiO_2 + 2Cl_2 + C \rightleftarrows TiCl_4 + CO_2 + 49\ kcal,} \tag{1} $$

$$ \mathrm{TiO_2 + 2Cl_2 + 2C \rightleftarrows TiCl_4 + CO + 78\ kcal,} \tag{2} $$

$$ \mathrm{TiO_2 + 4Cl_2 + 2C \rightleftarrows TiCl_4 + 2COCl_2 + 68\ kcal.} \tag{3} $$

As I. N. Godnev and A. V. Pamfilov [3] showed, the equilibrium pressure of phosgene at 500–800°C is vanishingly small, and reactions 1 and 2 practically predominate. At temperatures above 600°C the process proceeds mainly according to reaction 2, and at 900–1000°C only according to this reaction.

During the chlorination of zirconium dioxide in a mixture with carbon, according to Stephens and Gilbert [14], the following principal reactions take place:

$$ \tfrac{1}{2}\mathrm{ZrO_2 + C + Cl_2 = \tfrac{1}{2}ZrCl_4 + CO,} \tag{1} $$

$$ \tfrac{1}{2}\mathrm{ZrO_2 + CO + Cl_2 = \tfrac{1}{2}ZrCl_4 + CO_2,} \tag{2} $$

$$ \tfrac{1}{2}\mathrm{ZrO_2 + \tfrac{1}{2}C + Cl_2 = \tfrac{1}{2}ZrCl_4 + \tfrac{1}{2}CO_2.} \tag{3} $$

At temperatures above 700°C chlorination proceeds predominantly according to reaction 1. Reaction 3 predominates at low temperatures. Analysis of our data shows that, in the chlorination of titanium–zirconium raw material, in general the same regularities are preserved as in the chlorination of pure oxides.

Fig. 2. Dependence of chlorination on experiment duration at different temperatures: A — amount of chlorinated iron, titanium, %; B — experiment duration (min). For iron, curves at 300° (1), 400° (2), 500° (3), 800° (4). For titanium — 300° (5), 400° (6), 500° (7), 800° (8).

Fig. 3. Dependence of chlorination on temperature. Experiment duration — 1 h. A — amount of chlorinated iron (2), titanium (1), zirconium (3), in %; B — experiment temperature (°C).

Effect of the Chlorine Flow Rate on the Degree of Chlorination of Titanium–Zirconium Raw Material

Experiments were carried out at temperatures of 300–800°C and an experiment duration of 1 hour. The charge was 15 g; the chlorine flow rate was 150, 300, 500, and 1000 ml/min.

The results of the experiments are presented in Fig. 4.

The data obtained show that, at low chlorination temperatures, the chlorine feed rate does not affect the yield of titanium iron chlorides. Beginning at a temperature of 400°C, the influence of the chlorine flow rate on the chlorination process becomes noticeable. The amount of zirconium chloride formed increases only slightly with an increase in the chlorine flow rate at the temperatures selected for study, 300–800°C (Fig. 4).

Fig. 4. Dependence of chlorination on temperature at different chlorine feed rates: A—the amount of chlorinated component, %; B—the chlorine feed rate, ml/min. For iron, curves 150 ml/min (1), 300 (2), 500 (3), 1000 (4); for titanium—150 (5), 300 (6), 500 (7), 1000 (8); and for zirconium—150 (9), 300 (10), 500 (11), 1000 (12).

Conclusions

1. The optimum chlorination temperatures for iron and titanium, which are constituents of this titanium–zirconium concentrate, lie in the temperature range 600–800°C, i.e., higher than for the pure oxides.

2. From the kinetic chlorination curves for a mixture of collective titanium–zirconium concentrate and peat, it may be assumed that the chlorination process up to temperatures of 350–400°C belongs to the kinetic region. In this interval the chlorination rate increases rapidly with temperature and is almost independent of the chlorine rate; from 400 to 600°C there is an intermediate region; above 600°C the process is diffusion-controlled; the reaction rate, like the diffusion rate of chlorine, depends only weakly on temperature and increases with increasing chlorine flow rate.

Literature

1. F. H. Culler, K. E. Planko. Ind. End. Chem., 5, 1, 32, 1959.
2. B. D. Vasilenko, A. N. Vol’skii. Collection of Scientific Works of Mints vetzoloto, 27, 119 (1957).

1. I. S. Morozov, S. L. Stefanyuk. Zh. n. kh., 3, 10, 2366, 1958.
2. I. N. Godnev, A. V. Pamfilov. Zh. o. kh., 7, 1264, 1937.
3. Chlorination of slags rich in titanium dioxide. J. Applied Chemistry, vol. XIII, 12, 1760—1769, 1940.
4. I. P. Bardin, V. A. Reznichenko. Metallurgy of the USSR, 1917—1957 and 583, 1958.
5. I. S. Morozov, B. G. Korshunov. Chemistry of Rare Elements, 2, 102, 1955.
6. Gorski C. H. J. Metals, 191, No. 2, 131—133, 1951.
7. S. I. Smolyaninov and R. A. Beskrovnaya. The use of peat as a reducing agent and binder in the chlorination of certain natural compounds. Proceedings of TPI, vol. 112, 1963.
8. A. V. Pamfilov, A. S. Khudyakov and E. G. Shtandel. Zh. o. kh., 5, 605, (1935).
9. A. V. Pamfilov and E. G. Shtandel. Journal of Applied Chemistry, vol. 9, issue 10, 1936.
10. Austrian patent 166812, 23, 02, 56.
11. USA, 2793903, 15, 11, 55.
12. Stephens. Gilbert Journ. of Metals 4, No. 7, (1953), 733.