Prof. V. I. Minaev.

Associate Professor of Chemical Technology.

Optimum Temperature and Time of Reduction Smelting of Sodium Sulfate with Coal

1. Topic. The Leblanc soda-smelting process is expressed by the following scheme:

\[ 5\,\mathrm{Na_2SO_4}+10\,\mathrm{C}=5\,\mathrm{Na_2S}+10\,\mathrm{CO_2}\ldots\ldots\ldots(1) \]

\[ 5\,\mathrm{Na_2S}+5\,\mathrm{CaCO_3}=5\,\mathrm{Na_2CO_3}+5\,\mathrm{CaS}\ldots\ldots(2) \]

\[ 2\,\mathrm{CaCO_3}+2\,\mathrm{C}=2\,\mathrm{CaO}+4\,\mathrm{CO}\ldots\ldots\ldots(3) \]

The theoretical side of this classical process has been elucidated by the investigations of Gossage, Kolb, and especially by the works of Scherer—Kestner. Reading these equations, we see that first of all the reduction of sodium sulfate ($\mathrm{Na_2SO_4}$) to sodium sulfide ($\mathrm{Na_2S}$) takes place, with the simultaneous liberation of carbonic anhydride ($\mathrm{CO_2}$); the $\mathrm{Na_2S}$ formed at the moment of smelting impregnates the carbonic lime and, entering with it into double decomposition, gives calcium sulfide and soda. In addition, from the action of heat, part of the $\mathrm{CaCO_3}$ decomposes into lime and carbonic-acid gas, and the latter with an excess of coal gives carbon monoxide, which at the end of the decomposition becomes evident owing to the appearance of a blue flame. D. I. Mendeleev expresses the whole sum of the reactions of soda smelting by a single equation:

\[ 2\,\mathrm{Na_2SO_4}+3\,\mathrm{CaCO_3}+9\,\mathrm{C} =2\,\mathrm{Na_2CO_3}+\mathrm{CaO}\cdot 2\,\mathrm{CaS}+10\,\mathrm{CO}\ldots\ldots(4), \]

indicating that the quantity of substances mixed in factory production does in fact approach the ratio required by this equation, and noting the special importance of the excess of lime in the soda melt in the form of the almost water-unchangeable system $\mathrm{CaO}\cdot 2\,\mathrm{CaS}$. This compound has a particularly important significance in the subsequent lixiviation of the soda melt.¹)

For our investigation, from the entire totality of chemical transformations taking place in the Leblanc soda melt, we took only the first reaction in an isolated form, i.e., as it is represented by equation (1). It is, of course, beyond doubt that the more completely this reaction proceeds, i.e., the greater the yield of $\mathrm{Na_2S}$ by which it is accompanied under suitable conditions, the more advantageous will also be the further process of soda formation; on the other hand, this reaction and the clarification of the conditions of its course are of interest irrespective of soda smelting, all the more because the Russian and German reference chemical literature available to us here did not give us sufficiently clear and exhaustive materials on the question that interests us. Therefore we think that the small investigation proposed here is of some general interest.

The motivating cause for the present investigation was, on the one hand, the quite natural aspiration of our time to arouse state and private-public industrial initiative for raising the productive forces of vast Siberia by utilizing local raw-material resources in the form of enormous reserves of natural self-deposited Glauber salt (mirabilite) and sulfate (thenardite), called in the local dialect “gudzhir,”²) and, on the other hand, professional duty and academic

¹) See Principles of Chemistry by D. I. Mendeleev, 8th ed., 1906, p. 207 and add. [340]

²) To this question was devoted my report “On the Soda Industry for Siberia” in the Society of Siberian Engineers (November 1917), printed as a separate “Note on the organization of the soda business on the reserves of natural self-deposited Glauber salt,” 1918.

A second report on the same subject was read at the Sibpromburo of the Supreme Council of National Economy (November 1921), and a memorandum in revised form, “On the Basic Chemical Industry for Siberia” (December 1921), was submitted to the Siberian State Planning Commission.

obligation to involve in this work, in a living way, their pupils—future workers in the native chemical industry¹.

2. Literature data concerning the reaction of formation of \( \mathrm{Na_2S} \) are found in Lyubavin (Technical Chemistry), Mendeleev (Principles of Chemistry), and others. The most complete information is given by Dammer (Anorganische Chemie). Thus, according to Kircher, \( \mathrm{Na_2S} \), in the form of a meat-red substance, with evolution of heat, is formed under laboratory conditions by passing \( \mathrm{H_2S} \) through powdered \( \mathrm{NaOH} \). From a solution of \( \mathrm{NaOH} \), half saturated with hydrogen sulfide, \( \mathrm{Na_2S \cdot H_2O} \) precipitates on evaporation and cooling.

According to Berthelot, \( \mathrm{Na_2S} \) is formed by fusing 100 parts \( \mathrm{Na_2SO_4} \) with 20 parts coal \((5:1)\); in this case a meat-red or black mass (from excess coal) is obtained, soluble in hot water. In order to prevent the melt from adhering to the walls of the vessel in which the fusion is carried out, Jean adds to the mixture of sulfate and coal \( \mathrm{BaSO_4} \) in considerable quantity, which makes the mass porous. Magnesia should also satisfy this requirement. Belgrup succeeded in finding, in a strongly compressed mixture of coke powder with graphite, a material, highly resistant, for vessels for the melt.

Weldon first brings the sodium sulfate and coal together to a red-hot state, after which the reaction proceeds smoothly and rapidly and almost only \( \mathrm{Na_2S} \) is obtained.

The method given by Delbig for obtaining \( \mathrm{Na_2S} \) from soda residues deserves attention; namely, a mixture of the latter with \( \mathrm{Na_2SO_4} \) is extracted with water, whereby calcium hydroxide hydrate and calcium sulfhydrate are first formed:

\[ 2\mathrm{CaS} + 2\mathrm{H_2O} = \mathrm{Ca(OH)_2} + \mathrm{Ca(SH)_2} \ldots \ldots (5); \]

then calcium hydroxide hydrate, reacting with soda, gives \( \mathrm{NaOH} \), which with calcium sulfhydrate forms \( \mathrm{Na_2S} \).

In Dammer, besides the above-mentioned proportion \( \mathrm{Na_2SO_4} : \) coal \(= 5:1\) (according to Berthelot), there is an indication of the proportion \(4:1\), and also of the physical properties of the melt—its color and porosity.

In Lyubavin we find an indication that Leblanc, in his recipe for preparing soda, used wood charcoal, which, owing to its porosity and low ash content, is very suitable for this purpose, but now it is too expensive and therefore coal is usually taken. Coke is also used for the fusion; the latter, however, should be taken in a somewhat larger amount than coal, and the fusion should be conducted longer (1½ times as long as usual), because coke enters into the reaction with greater difficulty. The materials must be ground and mixed.

As regards the proportions of sodium sulfate and coal (not counting chalk or limestone), in the Leblanc process, according to Lyubavin, the following quantities were used:

a) according to Leblanc: for 100 parts of calcined \( \mathrm{Na_2SO_4} \)—50 parts of wood charcoal,

b) according to Kolb: for 106 parts of commercial \( \mathrm{Na_2SO_4} \)—52.5 parts of fat bituminous coal,

c) according to MacTear: for 100 parts \( \mathrm{Na_2SO_4} \)—40 parts coal,

d) according to Lunge: for 100 parts \( \mathrm{Na_2SO_4} \)—46–52 parts bituminous coal,

e) factory proportions (according to Lunge): for 100 parts \( \mathrm{Na_2SO_4} \)—from 30 to 75 parts bituminous coal, while on average the ratio of the named materials is taken b. or m. close to \(2:1\), which differs from the data cited above, taken from Dammer.

The literature data concerning the temperature conditions of the Leblanc process are likewise far from complete. In addition to what was indicated above about bringing the mixture of sulfate with coal to the incandescent state (according to Weldon), F. Fischer gives, for the Leblanc process, more precise indications that the temperature at the first hearth of the soda furnace, where the heat reaches light-red incandescence, attains \(713^\circ\), and by the end of the process rises to \(932^\circ\).

Mendeleev, in describing the process of converting Glauber’s salt into soda, speaks only of the “strong heating” of a mixture of \(Na_2SO_4\) with coal and limestone, etc. We consider it especially important to note here that D. I. Mendeleev, among the many diverse methods of obtaining soda, dwells\(^{1}\) on Gossage’s method, which prepares \(Na_2S\) from \(Na_2SO_4\) by calcination with coal, dissolves it in water, and subjects the solution, in coke towers, to the action of an excess of \(CO_2\), whereby \(H_2S\) is formed (a gas which, on complete combustion, can yield \(SO_2\), and on incomplete combustion, sulfur) and sodium bicarbonate:

\[ Na_2S + 2CO_2 + 2H_2O = H_2S + 2HNaCO_3 \ldots\ldots (6), \]

from which, on heating, soda and \(CO_2\) are obtained. This method, in the opinion of D. I. Mendeleev, is suitable for the treatment of natural \(Na_2SO_4\), all the more so because \(H_2S\) gives sulfur as a by-product.

Having at our disposal the above-cited literature material, one can only state that such important questions as those of the most advantageous smelting temperature, the duration of the process, the influence of time and temperature on the course of the given reaction and on its final effect, i.e., on the physical properties of the melt and on the yield of \(Na_2S\), are elucidated far from fully; at the same time, the solution of these questions is not only of great theoretical interest, but also of practical interest. To elucidate the noted influences of temperature and time on the course of the reaction and on its final effect, our work was undertaken under laboratory conditions; at the same time we succeeded in observing certain characteristic features of the course of the reaction in question, which may have application in practice in conducting and controlling the process.

3. Starting materials for our work were: natural Siberian sulfate, bituminous coal, and later also birch charcoal. The sulfate was provided to us from samples available in one of the laboratories of the Institute, from one of the lakes (unknown to us) of the Kulunda steppe lake chain. This sulfate had the following composition:

\(99.80\% + 0.20\) insoluble residue.

Let us note here, incidentally, that subsequently we had mirabilite from the dried-up Lake Tuskol, with a content of up to 98% Glauber’s salt, calculated as the dehydrated product, with an insignificant content of \(NaCl\) (0.14–0.18% in aqueous compound) and \(CaSO_4\) (0.34–0.45%), and also in the absence of magnesium salts, which is very important for the Leblanc soda process.

As for the bituminous coal, for its characterization we give here its analysis:

\(^{1}\) Principles of Chemistry, 8th ed. (1906), p. 578, footnote [340].

According to the above composition of our sulfate and coal, a correction for purity was introduced into the quantitative ratios of $\mathrm{Na_2SO_4:C}$ in their mixture and, correspondingly, into the calculation of the yield of $\mathrm{Na_2S}$.

4. Determinations of $\mathrm{Na_2S}$ in the melts obtained were carried out by the Lestelle method. The basis of this method is the insolubility of $\mathrm{Ag_2S}$ in ammonia. The preparation of the ammoniacal-silver solution required for titration is prescribed by the following recipe: dissolve 13.345 g of pure Ag in pure $\mathrm{HNO_3}$, add 250 cc of ammonia solution, and dilute with water to 1 liter. Each cc of such solution corresponds to 0.005 g of $\mathrm{Na_2S}$.

For preparing the ammoniacal-silver solution we used pure $\mathrm{AgNO_3}$ in the corresponding calculated amount (21.006 g). The solution was kept in a closed bottle wrapped in black glazed paper. Our entire usual melt (from 4 g sulfate and 1 g coal) was dissolved in 500 cc of water.

For titration, 25 cc portions of the solution of the melt under examination were taken, placed in a conical flask, heated to boiling, ammonia was added, and the ammoniacal-silver solution was run in dropwise from a burette until the formation of a black precipitate of $\mathrm{Ag_2S}$ ceased. The reaction scheme of the titration may be represented as follows:

$$ 2\,\mathrm{AgNH_4(NO_3)_2}+\mathrm{Na_2S}=\mathrm{Ag_2S}+2\,\mathrm{NaNO_3}+2\,\mathrm{NH_4NO_3}\ldots \tag{7} $$

In order to observe the end of the reaction more precisely, the solution being titrated should be shaken vigorously at short intervals; the fine particles of the $\mathrm{Ag_2S}$ formed collect into a black mass, whereupon the solution again becomes transparent, though colored yellowish. The end of the titration is noticed distinctly and, with some practice, very easily: from one excess drop the solution at once acquires (always after shaking) a violet-brown tint, which indicates the end of the reaction forming $\mathrm{Ag_2S}$, since one or two more cautiously added drops no longer change the color. If the solution proves to have been over-titrated, this can readily be verified by back-titration, after which a yellow coloration must necessarily appear again.

5. Preliminary experiments. Having no indications as to the duration and temperature conditions of melting, we conducted the first experiments for orientation. For the experiments we took a mixture of our sulfate and coal in the proportion $4:1^{1)}$. The weighed portions of the components were thoroughly mixed and placed in porcelain crucibles with lids. All crucibles of the same capacity with identical charges (5 g of mixture) were subjected to calcination and melting on Bunsen burners, carefully selected for equal flame strength; each crucible remained on the fire for a different time, from $2^{1}/_{2}$ to $6^{1}/_{2}$ hours. This first experiment cannot be considered successful, but it gave valuable indications that prolonged holding of the melt on the fire not only does not increase the yield of $\mathrm{Na_2S}$, but directly reduces its amount to zero. The cause of this phenomenon, apparently, is the process of oxidation of $\mathrm{Na_2S}$ and its conversion into more stable compounds—$\mathrm{Na_2SO_3}$, $\mathrm{Na_2S_2O_3}$, etc. The melting temperature in this first experiment was not determined, but it was assumed to have been the same in all cases. The results of the determinations and calculations of the yield of $\mathrm{Na_2S}$ are summarized in Table I and, correspondingly, are shown in the diagram by curve I—I.

$^{1)}$ Theoretically, according to equation (1), for 4 parts of $\mathrm{Na_2SO_4}$, 0.7 parts of carbon are required for smelting, or, by calculation, 0.78 parts of the coal of the above composition. In accordance with the purity of our sulfate (95.63% $\mathrm{Na_2SO_4}$), in the indicated ratio instead of 4 parts it would have been necessary to take 4.2 parts. Some excess of coal (e.g., 1.00 part instead of 0.78) is practically necessary, since, undoubtedly, part of it burns before it has had time to reduce $\mathrm{Na_2SO_4}$ to $\mathrm{Na_2S}$, and part simply does not enter into the reaction.

First experiment.
Table I.

Proportion of the mixture: 4 g of sulfate (= 3.8 g Na₂SO₄) : 1 g of coal.

Theoretically, from 4 g of our sulfate (= 3.8 g Na₂SO₄), 2.69 g of Na₂S (100%) should be obtained; the highest yield attained here is only 0.265 g (12.68%).

Since the heat of the Bunsen burner was evidently insufficient to heat through and melt the entire charge of the crucible in a short interval of time and more or less simultaneously and uniformly (in the crucibles of the first experiment there remained spots of the sulfate-and-coal mixture that had not reacted at all), the second experiment with 10 crucibles and with a charge of the same composition was modified in the sense that each crucible, after standing the prescribed time on the Bunsen burner, and thus having been heated through, was subjected to an additional 15-minute ignition on a Teclu burner. The additional crucibles were intended to determine the behavior of the melt over shorter intervals of time, beginning with one hour and up to four hours. The results of this experiment are given in Table II and are represented in the diagram by curve II—II.

Second experiment.
Table II.

Proportion of the mixture: 4 g of sulfate (= 3.8 g Na₂SO₄) : 1 g of coal.

All melts were carried out twice, and melts 5, 6, and 7 three times. The appearance of the melts was different; their solutions were also differently colored. The greatest yield of Na₂S was in melts with a reddish tint, which on leaching gave yellowish solutions. In general, the following sequence could be noted here: yellow solutions (from reddish melts) gave the highest yield of Na₂S; next, in decreasing order, came solutions that were pale yellow, dark green, green, and pale green. All the melts of the second experiment were only slightly porous, although in this respect the best were melts Nos. 1, 2, and 3, i.e., the melts with the highest content of Na₂S in them.

The second experiment already gave a result satisfactory to a considerable degree, since the yield of Na₂S in crucible No. 2 was 62.2% as against 12.68% for the first experiment. From a comparison of the results of these experiments and the conditions under which they were reproduced, it is evident that in the crucibles of the first experiment the mixture of sulfate with coal, adjacent to the walls of the crucible, had time in 1–1½ hours to heat to melting, so that the reduction reaction in this part of the melt proceeded completely; but the heat of the Bunsen burner was insufficient to melt the whole mass, and therefore the middle of the charge only barely sintered. On prolonged standing in the heat, the melt at the walls gradually thickened and began to oxidize; thus the only part that had reacted, and that contained Na₂S in the melts kept over the flame for 5½ and 6 hours, no longer revealed the sought product.

The crucibles of the second experiment were in better conditions in this respect, especially the one which stood over the flame for 1½ hours; 15 minutes’ heating on the Techlu burner caused the mass to react more completely, while oxidation of the outer part of the melt did not have time to occur fully, in view of the short interval of time.

The third experiment was set up under the same heating conditions as the second, but the mixture of sulfate with coal was taken in the ratio of 5 parts of the first to 2 parts of the second. This experiment was intended to determine the effect of an excess quantity of coal on the yield of Na₂S. The results are summarized in Table III and represented by curve III—III.

Third experiment.
Table III.

The conclusion that increasing the coal content in the mixture leads to a decrease in the yield of Na₂S would be incorrect, i.e., the introduction of a larger quantity of coal requires an increase in the furnace temperature; whereas, by increasing the amount of coal in the charge, we did not change the thermal conditions of the melting, remaining with the same burners as in the second experiment, and therefore our yield of Na₂S decreased. The third experiment again confirms that prolonged heating leads to a decrease in the yield of Na₂S. On leaching the melts of the third experiment, a large residue of coal was found that had not entered into the reaction; moreover, it is characteristic that we did not obtain the melt proper here as such, but had a black seized mass with clearly visible free coal. In view of the obvious disadvantage and uselessness of the taken ...

in the third experiment, the ratio of sulfate to coal being (5:2), we did not repeat this experiment, but undertook, with the former mixture (4:1), a new experiment—the fourth—with melting in an electric furnace, in which, first, it was possible to measure accurately the temperature of the furnace and of the melt and, secondly, by placing the porcelain crucible with its charge in the fireclay crucible of the furnace, to prevent access of air to the molten reacting mixture. Unfortunately, our furnace allowed the loading of only one crucible.

When melt No. 1 was heated, the pyrometer placed in the furnace space showed a temperature of 900°; the temperature of the melting charge itself was determined by the pyrometer as 790–800°; thus, the drop between the temperature of the furnace space and that of the charge was 100–110° C, evidently owing to the thick walls of the fireclay crucible. The melting in crucible No. 1 proceeded very violently; the melt obtained was extremely dense and was difficult to leach even with hot water; the crucible was impregnated with melt to a certain depth, so that it was necessary to digest it for a prolonged time; this in turn leads to a decrease in the amount of Na₂S, etc., and therefore the determination itself, and consequently the figure for the Na₂S content, lost its meaning.

Melts No. 2 and No. 3, at a furnace-space temperature of 800° C, showed within themselves a temperature of 700–720°, while the melting proceeded very calmly and uniformly. This observation and determination of the temperature in the melt may be regarded as the chief valuable result of the fourth experiment. In view of the impossibility of using current at any time of day, and because of the slowness of the work with one crucible, it was necessary to abandon the electric furnace and subsequently to carry on the work along the path indicated by the preliminary experiments, in a muffle furnace, giving a temperature up to 1000°, though more difficult to regulate.

6. Main experiments. The fifth experiment, with six crucibles and the adopted charge of 5 g of mixture (4 parts sulfate : 1 part hard coal), was carried out in a muffle furnace at temperatures fluctuating from 950 to 890° C. The crucibles were placed in the muffle when the temperature in the latter had reached the above-stated limits. The results are collected in Table IV and represented on the diagram of curve IV–IV′.

Fifth experiment:
Table IV.

Mixture proportion: 4 g sulfate (= 3.8 g Na₂SO₄) : 1 g coal.

As is seen from the table presented, the difference in the duration of sintering of the melts in the furnace was slight; this was done in order to determine precisely the inflection of the curve, and then to ascertain whether overburning, even by several minutes, noticeably lowers the yield of Na₂S. The features of the melts obtained were as follows:

Melts No. 5 and No. 6 had an almost black color, were well fused, and were difficult to leach with hot water; melts No. 3 and No. 4 were ocher-yellow in color and leached more readily; finally, melts No. 2 and No. 1 were reddish-green in color and leached very well. It is evident from the table that the greatest yield corresponds to a 40-minute stay of the melt in the muffle; a difference in the duration of the melts’ stay in the furnace at a temperature of 950—890° over 9 minutes entails a loss in the yield of Na₂S of 11% (compare melts No. 2 and No. 6).

The yield attained in this experiment, 78.96% Na₂S, already appears sufficiently high, and this apparently is explained by the uniformity of the heat and by the reducing or, at least, neutral atmosphere of the furnace (coals lay in the muffle).

Since in the experiment with the electric furnace it was established that melting proceeds more calmly at a melt temperature of 720° C or, what is the same, at a furnace temperature of 780°, we carried out a sixth experiment:

Experiment six.

Table V.
(See curve V—V).

Mixture proportion: 4 g sulfate: 1 g coal.

From this experiment it follows that lowering the temperature in the furnace to 780° does not entail a lowering of the yield of Na₂S. From a comparison of the results of the fifth and sixth experiments it is evident that at a higher temperature the course of the reaction proceeds faster, and overholding the melt at the higher heat leads to a rapid decrease in the yield of Na₂S; consequently, in practice it will be very difficult to conduct the melting and regulate the melting time at high furnace temperatures. Moreover, melt No. 2, which gave a yield of Na₂S of 79.9%, is readily leached and again has a reddish color with greenish patches.

Since in practice, in the interest of saving fuel, it is more advantageous to work with the lowest possible furnace temperatures, a seventh experiment was set up with the furnace heated to 700° C.

Experiment seven.

Table VI.
(See curve VI—VI).

Mixture proportion: 4 grams sulfate: 1 g stone coal.

The seventh experiment, which gave very good results, confirms that the yield of \(Na_2S\) reaches a maximum when the melt is kept in the furnace for an average of about 40–45 min. (cf. experiments five, six, and seven); before this interval of time has elapsed, the high temperature of the furnace does not have a strongly negative effect on the yield, but after this period the high temperature \((950—890^\circ)\) is reflected more noticeably in a lowering of the yield of \(Na_2S\) than is the temperature of \(700^\circ\), which is necessary and sufficient for our smelting (cf. in the seventh experiment melts Nos. 2, 3, and 4).

The smelting of the mixture in the crucibles of the seventh experiment was very calm; the melts were leached very easily, but melt No. 4 was leached somewhat more difficultly and, quite evidently, only because, as compared with melt No. 3, it spent an extra 15 min. in the furnace at \(700^\circ\), although it did not thereby lower its \(Na_2S\) content. Melts No. 2 and No. 3 again had a reddish color with greenish patches, while melt No. 4 was brown, and melt No. 1 was greenish-gray. The residue of unreacted coal was negligible.

7.—These experiments clarified to a sufficient degree the picture of the reducing smelting of sulfate with hard coal. Further experiments were undertaken on the smelting of sulfate with birch charcoal. Here an interesting surprise was noted at once: a mixture of sulfate with charcoal (4 : 1), introduced in crucibles into a muffle furnace at \(700^\circ C\), after 20 min. had not begun to melt, whereas in all the preceding experiments smelting usually began 8—10 minutes after the crucibles were placed in the furnace. Since it could not be assumed that the elapsed 20 min. had been insufficient for the mixture to become heated through, it was necessary to suppose that a temperature of \(700^\circ\), with charcoal in the mixture, was insufficient for smelting (probably because of the absence of fluxing agents in charcoal as compared with hard coal, which contains more mineral substances, Fe, S, etc.). Therefore the crucibles were removed from the furnace; the latter was set to \(890—900^\circ C\), after which the crucibles were again introduced into the muffle. The results of this eighth experiment are summarized in Table VII; in the diagram—the curve VII—VII.

Eighth experiment.  Table VII.

Melt No. 1—very porous, red-green in color. Melt No. 2—inside flesh-red in color, very porous; outside yellow in color, had fallen away from the walls of the crucible. Melt No. 3—brown in color, not porous, fused completely.

Ninth experiment (check experiment)  Table VIII.
(The mixture is the same as in the eighth experiment.)  (see curve VIII—VIII).

This experiment fully confirms the result of the eighth experiment, if they are related to a temperature of 900° C, i.e., in other words, heating for 20 min. of the melts in the eighth experiment at 700° C had no influence whatever on the final result, since it is obvious that at 700° the fusion of sulfate with charcoal does not occur and does not even begin partially.

The tenth experiment, carried out at a temperature of 880° C, gave results entirely coinciding with the results of the two preceding ones: the highest yield (= 83.22% Na₂S) was obtained from the melt that had remained in the furnace for 50 minutes.

Having changed the ratio of sulfate to charcoal to the proportion 4:0.8, we set up the eleventh experiment at 800° C. We place the results in Table IX (see also curve IX—IX in the diagram):

Eleventh experiment.                    Table IX.

All the melts of this experiment are porous. Melts No. 3 and No. 4 are greenish-yellow on the outside, and flesh-red inside. Comparing the results of the ninth and eleventh smeltings, we see that the decrease in the charcoal content had a very slight effect on lowering the yield of Na₂S (by 0.95%), but for that the temperature of the furnace was lowered by 80–100°.

8. Conclusion. Since in all experiments with a sulfate-to-coal ratio of 4:1 a portion of unreacted coal is observed in the melt, it is clear that an excess of coal in the mixture is useless and always disadvantageous, since, besides the unproductive expenditure of it in the mixture, it entails increased fuel consumption and a loss of time for burning it out; the latter, being connected with a lowering of the Na₂S yield, will make an excess of coal in the mixture harmful. Under laboratory working conditions the rational ratio of sulfate to coal should evidently be close to 4:0.8 and in any case no higher than 4:0.9. Under technical working conditions, owing to inevitable losses of coal, such a ratio will probably be close to 4:1 and perhaps not lower than 4:0.9.

The greatest yield of Na₂S from the melt, determined in the experiments described above as 83.25%, is always accompanied by the porosity of the melt and by a characteristic flesh-red color with greenish oases; these signs are indicators of the good quality of the melt and, consequently, can be used in the technical control of production.

The experiments described above have fully clarified the rate of progress of the reaction of the reduction smelting considered; namely, we see that a 25–30-minute residence of the mixture in the proportion 4:0.8 in a heated furnace, or 45–50 minutes for the mixture 4:1, is sufficient for completion of the reaction, for beyond this not only is no increase in yield observed, but a gradual decrease in the Na₂S content in the melt takes place. Under technical working conditions the time of the most advantageous smelting found here will probably change somewhat, depen-

... depending on the time required to heat the charge until melting begins—and, consequently, will depend on the amount of mixture charged into the furnace, the degree to which it has cooled (the season), the dimensions of the furnace (its heat capacity), etc.

As for the optimum temperature under the conditions of the reducing smelting under consideration, the most advantageous furnace temperature was found to be 700–780° when smelting sulfate with hard coal, and 800–880° when smelting with charcoal. Increasing the coal content in the mixture with sulfate entails the need to raise the heat in the furnace; the lower temperature (by 100°) when working with hard coal in the mixture probably depends on the presence in it of melts, which facilitates melting.

To confirm all the conclusions that had suggested themselves, an experiment was carried out with a larger quantity of mixture than the crucibles allowed. A mixture of sulfate with charcoal in the proportion 4 : 0.8 and in an amount of 190 g (160 g sulfate + 32 g charcoal) was placed at once, in a porcelain cup with a lid, into a furnace heated to 810° C. The mixture was kept in the furnace for 20 minutes and then removed. The melt obtained was porous and, in color, similar to melt No. 4 in experiment eleven. The same experiment with a 4 : 1 mixture in an amount of 200 g (160 g sulfate + 40 g charcoal) gave excellent results when kept in the furnace for 50 minutes at a temperature of 850° (according to the data of experiment nine).

In these experiments, as in all the preceding ones, one further characteristic phenomenon is observed, deserving mention: 5–6 minutes after the crucibles and cups are placed in the furnace, tongues of flame invariably appeared from under their lids, which went out after a certain interval of time; from observations it was established that this moment corresponds to the highest content of \(\mathrm{Na_2S}\) in the melt and, consequently, to the completion of the reduction reactions. In all probability it may be assumed that what can burn here is CO and the finest particles of coal carried along by the stream of carbon dioxide being evolved (the color of the flames is not bluish, but yellow). Thus, it is quite probable that the reaction proceeds somewhat more complexly than is expressed by equation 1). At the same time it is obvious that the observed phenomenon of combustion, and the moment of its cessation, may also serve as signs for monitoring the course of the reaction, which is important in practice, since overholding the melt in the furnace invariably leads to a decrease in the yield of \(\mathrm{Na_2S}\).

As regards the yield of \(\mathrm{Na_2S}\) in the reducing smelting described, under our conditions we reached a maximum of 79.9% in the melt with hard coal and 83.25% with charcoal; as to what the practical situation is under factory conditions, we have no exact information, but it seems that the yield is usually not higher than 60–65% ¹).

9. Supplement. Subsequently the work was reduced to obtaining crystalline \(\mathrm{Na_2S \cdot 9H_2O}\) and to studying the conditions for using the mother liquors (after crystallization of the indicated nonahydrate salt) to obtain \(\mathrm{Na_2S_2O_3 \cdot 5H_2O}\).

In the cycle of a Leblanc-type soda industry, with the utilization of sulfur from soda waste, sulfur compounds—\(\mathrm{Na_2S \cdot 9H_2O}\), \(\mathrm{Na_2S_2O_3 \cdot 5H_2O}\), and others—occupy a prominent place. Finally, when soda production according to the Gossage scheme (equation 6) is established, the utilization of S will find its place.

¹) In Prof. N. N. Lyubavin’s Technical Chemistry, distinguished by the general completeness of its data, in the section on hyposulfite there is only a brief indication of the preparation of \(\mathrm{Na_2S}\): “\(\mathrm{Na_2SO_4}\) is calcined with coal in a furnace with a reducing flame for 6 hours and \(\mathrm{Na_2S}\) is extracted with water. In this, however, some sodium polysulfide, sodium sulfhydrate, and sodium sulfite are obtained.” There are no data on the yield of \(\mathrm{Na_2S}\); the duration of the smelting (6 hours) is hardly expedient...

The preparation of sodium thiosulfate proceeds as indicated in the literature. It is only necessary to note that, after obtaining a neutral reaction, an excess of sulfurous acid in the reaction should be avoided:

\[ 2\,\mathrm{Na_2S} + 3\,\mathrm{SO_2} = 2\,\mathrm{Na_2S_2O_3} + \mathrm{S}\ldots\ (8), \]

and, after removing the precipitated sulfur, the clarified solution should be concentrated at a temperature not exceeding \(70^\circ\mathrm{C}\), since a higher temperature entails decomposition of the hyposulfite with separation of sulfur. The formation and growth of small crystals proceeds rapidly without vacuum. The hyposulfite obtained met all purity requirements: it contained no \(\mathrm{Na_2SO_4}\) or \(\mathrm{Na_2S}\), gave a slightly alkaline solution, and had a melting point of \(48^\circ\) (according to Dammer—from \(45^\circ\) to \(50^\circ\)).

e. Minas, V. T. Optimum temperature and time for reduction smelting of Na₂SO₄ with coal.

Vertical axis: Yield of Na₂S as percent

Horizontal axis: Time in hours

Preliminary experiments

Maxim. 62.2%

Maxim. 45.9%

Maxim. 12.68%

I

II

III

Main Experiments

1. In the 1916–17 academic year, by commission of the Chemical Faculty of the Tomsk Technological Institute, I gave a special course on the chemical technology of mineral substances. Among the students specializing in this field, A. P. Selivanov and M. N. Mikhailov took the liveliest part in the development of our topic. ↩