Khodalevich G. N. and Zharikov I. I.

On the Problem of Obtaining Anhydrous Aluminum Chloride from Siberian Clays

The West Siberian region has entered a period of unprecedented rates of industrialization and socialist reconstruction of its entire economy. Powerful centers of industry are being created in the form of the Ural–Kuzbass combine, individual “links of which are already entering the ranks of operating enterprises.” The first and second five-year plans for the region provide for the powerful development of the coal-chemical industry, many branches of which will require the construction of special-purpose plants, such as, for example, a plant producing anhydrous aluminum chloride—an important substance in a whole series of chemical productions, connected in one way or another with the cracking process, with the preparation of dyes, pharmaceutical and other preparations.

The studies available in the Soviet periodical scientific literature on the question of obtaining anhydrous aluminum chloride from domestic raw materials concern chiefly the clays of the Donbass. With the development of a powerful Siberian chemical industry, the organization of research work connected with the question of obtaining anhydrous aluminum chloride from the raw materials of the West Siberian region is, undoubtedly, timely. The work carried out by us is the first attempt at solving this problem.

Since the quality of anhydrous aluminum chloride prepared by the factory method depends first and foremost on the purity of the raw material used, in the sense of “contamination” of the latter by large amounts of iron and titanium compounds, it is clear that our first task was the selection of a suitable object of investigation. We were able to obtain samples of clays from the Nekrasovskoye and Moyskoye deposits, which were subjected to analytical study.

The results of the analysis are summarized in Table No. 1. According to the available literature data, the most reactive aluminous earth (and this is important for obtaining anhydrous aluminum chloride) can be obtained from clays with a high kaolinite content. Therefore we carried out repeated determinations of kaolinite in the samples of Siberian clays available to us. The determinations were performed by the method indicated in the article by I. E. Adadurov, “Obtaining Aluminum Chloride from Donbass Clays.” The results of our determinations are summarized in Table No. 2.

Table No. 1

Table No. 2

On the basis of the data of the empirical, or summation, analysis (see Table No. 1) of the clays investigated, and also from the results of determining kaolinite in the same samples, we concluded that the best object for obtaining anhydrous aluminum chloride may be the Nekrasov clay, since, in comparison with the other samples, it contains, first, a smaller amount of iron and, second, is richer in aluminum bound in the kaolin nucleus.

Anhydrous aluminum chloride was first obtained in 1825 by Oersted by passing chlorine over a heated mixture of alumina and carbon¹.

This reaction may proceed according to the following equations:

\[ Al_2O_3 + 3C + 3Cl_2 = 2AlCl_3 + 3CO \]

or

\[ 2Al_2O_3 + 3C + 6Cl_2 = 4AlCl_3 + 3CO_2 \]

And to this day Oerstedt’s method should be regarded as the basic one, since, in the chemical sense, a considerable variety of different new proposals for obtaining anhydrous aluminum chloride from alumina and alumina-containing materials—such as clays, bauxites, etc.—is nothing other than a modification of Oerstedt’s process, chiefly with respect to the choice of reducing agents. Thus, for example, in order to obtain reducing agents with a larger surface area, it is proposed to chlorinate alumina-containing materials not in the form of a simple mixture of them with coal or coke, but in the form of briquettes prepared by mixing alumina-containing materials with various carbon-containing substances (petroleum, starch, etc.) and subsequently coking the resulting mass. Other proposals concern the use of reducing agents in the gaseous state (such as carbon monoxide, carbon disulfide, naphthalene, etc.). Still other proposals concern the replacement of pure chlorine by phosgene, hydrogen chloride, and other substances.

Aluminum chloride of high purity can be obtained by chlorinating metallic aluminum with chlorine or hydrogen chloride. Of interest is the method for obtaining anhydrous \(\mathrm{AlCl_3}\) by heating alkali or alkaline-earth chlorides with aluminum sulfide

\[ \mathrm{Al_2S_3 + 6KCl = 3K_2S + Al_2Cl_6} \]

(see U.S. patent 1716102, 1929). A review of patents on the preparation of anhydrous aluminum chloride may be found in the book by Oliver Ralston, Anhydrous Aluminum Chloride (publ. 1924, Sov. Neft. PR.).

At present, abroad, anhydrous \(\mathrm{AlCl_3}\) is most often obtained by the classical Oerstedt method, by the action of chlorine on a calcined mixture of bauxite with coal, the bauxite used having as low a content as possible of iron and silica, i.e., being of the highest grade.

Since the most widespread, and consequently the cheapest, alumina-containing material is clay, it is natural that many authors in their works\(^1\) have focused attention on obtaining anhydrous \(\mathrm{AlCl_3}\) from clays. From the works cited it is evident that in clays, along with chlorination of alumina, an equally vigorous chlorination of silicon dioxide also takes place. N. N. Voronin and I. S. Galinker indicate that, when clays are chlorinated with chlorine in the presence of coal, only about 45% of the chlorine consumed goes to the chlorination of alumina, and they come to the conclusion that chlorination of clay can become economically advantageous only if there is demand for silicon tetrachloride. V. I. Spitsyn and O. M. Gvozdeva found that the addition of SiCl\(_4\) vapor to chlorine significantly—

\(^1\) V. I. Spitsyn and O. M. Gvozdeva, “Chlorination of Oxides and Natural Compounds.” Published in 1931 by the Institute of Applied Mineralogy.—I. E. Aladzurov, Journal of Chemical Industry, Nos. 21–22, 1928, and No. 24, 1930.—N. N. Voronin and I. S. Galinker, Journal of Chemical Industry, Nos. 2–3, 1930.

significantly lowers the chlorination of silica from clay, without appreciably lowering the activity of alumina. I. V. Adadurov, in the special article “Is It Profitable to Obtain Anhydrous Aluminum Chloride from Donbass Clays” (Journal of Chemical Industry, No. 24, 1930), disputes the correctness of the conclusions of N. N. Voronin and I. S. Galinker.

In view of a certain lack of clarity in the question of the economic feasibility of using clays for obtaining anhydrous \(\mathrm{AlCl_3}\) by the method of chlorinating clays with chlorine in the presence of carbon, and of the complete absence in the literature of data on the chlorination of Siberian clays, we carried out a whole series of experiments on the chlorination of certain Siberian clays with chlorine in the presence of carbon²). The experimental arrangement of these experiments was as follows. A refractory tube was placed in a horizontal electric furnace; on one side it was connected with a gasometer containing chlorine, and on the other side with a receiver, which was a glass tube bent twice in the shape of the letter U. The refractory tube and the receiver tube were selected to have the same diameter and were joined end to end by means of a piece of rubber tubing. After the receiver there followed 3 Drechsel bottles, the first of which contained a 5% solution of \(\mathrm{H_2SO_4}\), and the next two a concentrated solution of \(\mathrm{NaOH}\), serving to absorb excess chlorine.

Fig. 1.

Chlorine was obtained from \(\mathrm{KMnO_4}\) and \(\mathrm{HCl}\) and was collected in a graduated gasometer over a saturated solution of common salt. Chlorination was carried out with dry chlorine, for which purpose 2 Drechsel bottles with concentrated \(\mathrm{H_2SO_4}\) were placed between the gasometer and the electric furnace. The charge of clay to be chlorinated was placed

²) Unfortunately, only after this work had been prepared for publication, in Journal of Chemical Industry for 1932, No. 7 (the journal arrived in Tomsk belatedly), we read the article by B. P. Bruns and E. I. Shatunovskaya, “Obtaining Silica Gel from Silicon Tetrachloride, Which Is a Waste Product in the Production of Anhydrous Aluminum Chloride from Kaolin.”

in a glazed porcelain boat, which during the experiment was placed in a refractory tube in the zone of greatest heating of the electric furnace. The temperature in the furnace was monitored with a pyrometer; the thermocouple was located in the furnace beside the chlorination tube. Chlorination was carried out in the range from \(500^\circ\) to \(750^\circ\). After the experiment was completed, the residue of unchlorinated clay, after preliminary burning off of the excess carbon, was analyzed by the usual method for silicates.

The results of the experiments are summarized in Table No. 3.

The first seven experiments were carried out with briquettes prepared from finely ground clay mixed with powdered charcoal, taken in an amount of about 20% of the weight of the clay. The mixture was kneaded with sugar syrup and subjected to charring in a porcelain crucible at \(tt^\circ\) 600 and \(700^\circ\) without access of air. Analysis of the briquettes for carbon content gave an average figure of 31%. The size of the briquettes was 3–4 mm.

From the data presented in Table No. 3 it is evident that at temperatures of \(500\)—\(525^\circ\) the chlorination of the clay proceeds insufficiently intensively. It is seen, however, that silica and aluminum oxide begin to be chlorinated simultaneously, while the chlorination of the iron oxides proceeds with considerably higher yields.

At temperatures of \(600\)—\(630^\circ\), vigorous chlorination of the constituent parts of the clay is observed. The yield of aluminum chloride in 1 hour, at chlorine flow rates of 2 liters and 1 liter per hour, already reaches 80%. Under the same conditions, a one-and-a-half-hour experiment gives about 86% utilization of aluminum oxide.

With an increase in temperature to \(700\)—\(725^\circ\), the yield of \(\mathrm{AlCl_3}\), at the same chlorine flow rates and duration of the experiment, already reaches 90–92%. However, at these temperatures the difference in the chlorination rates of the iron oxides and aluminum oxide is smoothed out and, in addition, the percentage of chlorinated silica increases significantly.

The remaining experiments, from 8 to 16, are characterized by the fact that here briquettes prepared by simple kneading in water of clay with powdered charcoal and with subsequent calcination without access of air at the same temperature as the first ones were subjected to chlorination. These briquettes were prepared with different contents of carbon in them. The purpose was to determine the effect of an excess and deficiency of carbon on the process of chlorination of silica and aluminum oxide.

From the data obtained it may be concluded that no sharp difference is observed in the chlorination of these two constituent parts of the clay in this case either. It is interesting to note that, with a carbon content in the briquettes of 35%, without the use of a binding carbon-containing substance, chlorination proceeded extremely vigorously, and the utilization of aluminum oxide, as is seen from Table No. 3, reached 97%. However, at the same time the chlorination of silica also increased.

Further, we set up experiments on the chlorination of clay

Table 3

Experiments on the chlorination of Krasnoyarsk clay with carbon in briquettes

also in the form of briquettes, but without a content of coal in them; moreover, chlorination was carried out not with chlorine alone, but with a mixture of it with carbon monoxide. Chlorine and carbon monoxide were passed either in equimolecular quantities, or with a certain excess of carbon monoxide. The briquettes placed in a porcelain boat were, before chlorination, subjected to preliminary calcination at a temperature of 600–700°.

The results of the experiments are given in Table No. 4.

In the first four experiments the gas mixture, before being admitted into the reaction tube placed in an electric furnace, was heated in a specially arranged electric heater to a temperature of 450–500°.

In the following three experiments the gas mixture entered the reaction tube without preliminary heating.

Finally, the last experiment in this table gives the results for chlorination of clay with a gas mixture passed over heated activated carbon.

From all these experiments it is evident that the rate of the chlorination reaction, as compared with chlorination of clay by pure chlorine in the presence of coal, decreases considerably. However, the difference between the chlorination of silica and of alumina becomes more sharply expressed, especially in those cases where chlorination was carried out with some excess of carbon monoxide.

On the basis of the results of the experiments carried out by us and on the basis of the available literature data, the following conclusions may be drawn:

1. Siberian clays, taking the Nekrasovskoye deposit as an example, may serve as a quite satisfactory raw material for obtaining anhydrous aluminum chloride.

2. The chlorination of clays in the presence of coal at a temperature of 600–700° proceeds fairly intensively, despite the not entirely good contact between the briquettes and the chlorine (since, during chlorination, the briquettes were placed by us in a boat and did not in all parts come uniformly into contact with the passing gases).

3. As in the Donbass clays, along with the chlorination of alumina, the chlorination of silica also proceeds very vigorously. To carry out the chlorination of the constituent parts of the clay by phases, chiefly with respect to aluminum, silicon, and iron, is very difficult, despite the fact that the individual components are characterized by definite and different temperature conditions of chlorination (see the article by R. Wasmuth, Zeitschrift für angewandte Chemie, 1930, 5).

4. The separation of aluminum chloride from the chlorides of silicon and partially of iron may be effected by fractional condensation of the chloride vapors, if they pass from the reaction chamber into a series of successively connected receivers, provided that a definite temperature regime is maintained in the latter, since the vapors of AlCl₃ must pass into the solid phase at a temperature below 183°, and the vapors of FeCl₃ at a temperature below 280°.

Table No. 4

Experiments on chlorination of a mixture of Krasovsky clay oxides and chlorine.

SiCl₄ vapors, however, at a temperature below 57°. Thus fractional condensation makes it possible to obtain anhydrous AlCl₃ with a substantially lower iron content and with the complete absence of SiCl₄. If it were possible, before the fractional condensation of the iron and aluminum chlorides, to convert ferric chloride into ferrous chloride (which turns into vapor at temperatures above 800°), then, unquestionably, technical aluminum chloride would have a high degree of purity with respect to iron content.

1. A reduction in the consumption of chlorine for the chlorination of silica can be achieved, as we have repeatedly had occasion to verify, by adding SiCl₄ vapors to the chlorine; the same observations were made earlier by V. I. Spitsyn and O. M. Gvozdeva.

2. The reaction of chlorinating clay with a mixture of carbon monoxide and chlorine requires very serious study.

From the experiments partially given in Table No. 4 it is evident, as was also indicated above, that the limits, in terms of the quantities of silicon dioxide and aluminum oxide chlorinated, become sharply defined.

The only drawback is the insignificant rate of the chlorination reaction under the conditions of our experiments.

The determination of favorable conditions and a deeper and more thorough study of this reaction will constitute the subject of our forthcoming work.

On the significance of this reaction, see the article by Dr. C. Wurster (Zeitschr. f. angew. Chemie 1930, 40).

1. Pogg. Ann. 5. (1825), 132. ↩