Chapter VIII: Part V: Lime-Roasting of Galena (1)

THE HUNTINGTON-HEBERLEIN PROCESS

(July 6, 1905)

It is a fact, not generally known, that the American Smelting and Refining Company is now preparing to introduce the Huntington-Heberlein process in all its plants, this action being the outcome of extensive experimentation with the process. It is contemplated to employ the process not only for the desulphurization of all classes of lead ore, but also of mattes. This is a tardy recognition of the value of a process which has been before the metallurgical profession for nine years, the British patent having been issued under date of April 16, 1896, and has already attained important use in several foreign countries; but it will be the grandest application in point of magnitude.

The Huntington-Heberlein is the first of a new series of processes which effect the desulphurization of galena on an entirely new principle and at great advantage over the old method of roasting. They act at a comparatively low temperature, so that the loss of lead and silver is reduced to insignificant proportion; they eliminate the sulphur to a greater degree; and they deliver the ore in the form of a cinder, which greatly increases the smelting speed of the blast furnace. They constitute one of the most important advances in the metallurgy of lead. The roasting process has been the one in which least progress has been made, and it has remained a costly and wasteful step in the treatment of sulphide ores. In reducing upward of 2,500,000 tons of ore per annum, the American Smelting and Refining Company is obliged to roast upward of 1,000,000 tons of ore and matte.

The Huntington-Heberlein process was invented and first applied at Pertusola, Italy. It has since been introduced in Germany, Spain, Great Britain, Mexico, British Columbia, Tasmania, and Australia, in the last at the Port Pirie works of the Broken Hill Proprietary Company. Efforts were made to introduce it in the United States at least five years ago, without success and with little encouragement. The only share in this metallurgical improvement that this country can claim is that Thomas Huntington, one of the inventors, is an American citizen, Ferdinand Heberlein, the other, being a German.

LIME-ROASTING OF GALENA

(September 22, 1905)

The article of Professor Borchers (see p. 116) is, we believe, the first critical discussion of the reactions involved in the new methods of desulphurizing galena, as exemplified in the processes of Huntington and Heberlein, Savelsberg, and Carmichael and Bradford, although the subject has been touched upon by Donald Clark, writing in the _Engineering and Mining Journal_. It is perfectly obvious from a study of the metallurgy of these processes that they introduce an entirely new principle in the oxidation of galena, as Professor Borchers points out. Inasmuch as there are already three of these processes and are likely to be more, it will be necessary to have a type-name for this new branch of lead metallurgy. We venture to suggest that it may be referred to as the “lime-roasting of galena,” inasmuch as lime is evidently a requisite in the process; or, at all events, it is the agent which will be commonly employed.

When the Huntington-Heberlein process was first described, it did not even appear a simplification of the ordinary roasting process, but rather a complication of it. The process attracted comparatively little attention, and was indeed regarded somewhat with suspicion. This was largely due to the policy of the company which acquired the patent rights in refusing to publish the technical information concerning it that the metallurgical profession expected and needed. The history of this exploitation is another example of the disadvantage of secrecy in such matters. The Huntington-Heberlein process has only become thoroughly established as a new and valuable departure in metallurgy, a departure which is indeed revolutionary, nine years after the date of the original patent. In proprietary processes time is a particularly valuable element, inasmuch as the life of a patent is limited.

From the outset the explanation of Huntington and Heberlein as to the reactions involved in their process was unsatisfactory. Professor Borchers points out clearly that their conception of the formation of calcium peroxide was erroneous, and indicates strongly the probability that the active agent is calcium plumbate. It is very much to be regretted that he did not go further with his experiments on this subject, and it is to be hoped that they will be taken up by the professors of metallurgy in other metallurgical schools. The formation of calcium plumbate in the process was clearly forecasted, however, by Carmichael and Bradford in their first patent specification; indeed, they considered that the sintered product consisted largely of calcium plumbate.

Even yet, we have only a vague idea of the reactions that occur in these processes. There is undoubtedly a formation of calcium sulphate, as pointed out by Borchers and Savelsberg; but that compound is eventually decomposed, since it is one of the advantages of the lime-roasting that the sintered product is comparatively low in sulphur. Is it true, however, that the calcium eventually becomes silicate? If so, under what conditions is calcium silicate formed? The temperature maintained throughout the process is low, considerably lower than that required for the formation of any calcium silicate by fusion.

Moreover, it is not only galena which is decomposed by the new method, but also blende, pyrite and copper sulphides. The process is employed very successfully in the treatment of Broken Hill ore that is rather high in zinc sulphide, and it is also to be employed for the desulphurization of mattes. What are the reactions that affect the desulphurization of the sulphides other than lead?

There is a wide field for experimental metallurgy in connection with these new processes. The important practical development is that they do actually effect a great economy in the reduction of lead sulphide ores.

THE NEW METHODS OF DESULPHURIZING GALENA[18]

BY W. BORCHERS

(September 2, 1905)

An important revolution in the methods of smelting lead ore, which had to a large extent remained for centuries unchanged in their essentials, was wrought by the invention of Huntington and Heberlein in 1896. More especially is this true of the roast-reduction method of treating galena, which consists of oxidizing roasting in a reverberatory furnace and subsequent smelting of the roasted product in a shaft furnace.

The first stage of the roast-reduction process, as carried out according to the old method, viz., the oxidizing roast of the galena, serves to convert the lead sulphide into lead oxide:

PbS + 3O = PbO + SO₂.

Owing to the basic character of the lead oxide, the production of a considerable quantity of lead sulphate was of course unavoidable:

PbO + SO₂ + O = PbSO₄.

As this lead sulphate is converted back into sulphide in the blast-furnace operation, and so adds to the formation of matte, it has always been the aim (in working up ores containing little or no copper to be concentrated in the matte) to eliminate the sulphate as completely as possible, by bringing the charge, especially toward the end of the roasting operation, into a zone of the furnace wherein the temperature is sufficiently high to effect decomposition of the sulphate by silica:

PbSO₄ + SiO₂ = PbSiO₃ + SO₃.

But in the usual mode of carrying out the roast in reverberatory furnaces, the roasting itself on the one hand, and the decomposition of the sulphates on the other, were effected only incompletely and with widely varying results.

Little attention has been paid in connection with the roast-reduction process to the reaction between sulphates and undecomposed sulphides, which plays so important a part in the roast-reaction method of lead smelting. As is well known, lead sulphate reacts with lead sulphide in varying quantities, forming either metallic lead or lead oxide, or a mixture of both. A small quantity of lead sulphate reacting with lead sulphide yields under certain conditions only lead:

PbSO₄ + PbS = Pb₂ + 2SO₂.

Within certain temperature limits this reaction even proceeds with liberation of heat. In order to encourage it, it is necessary to create favorable conditions for the formation of considerable quantities of sulphate right at the beginning of the operation. This was first achieved by Huntington and Heberlein, but not in the simplest nor in the most efficient manner. And, indeed, the inventors were not by any means on the right track as to the character of their process, so far as the chemical reactions involved are concerned.

At first sight the Huntington-Heberlein process does not even appear as a simplification, but rather as a complication, of the roasting operation. For in place of the roast carried out in one apparatus and continuously, there are two roasts which have to be carried out separately and in two different forms of apparatus; nevertheless, the ultimate results were so favorable that the whole process is presumably acknowledged, without reservation, by all smelters as one of the most important advances in lead smelting.

It is useful to examine in the light of the German patent specification (No. 95,601 of Feb. 28, 1897) what were the ideas of its originators regarding the operation of this process and the reactions leading to such remarkable results. They stated:

“We have made the observation that when powdered lead sulphide (PbS), mixed with the powdered oxide of an alkaline earth metal, _e.g._, calcium oxide, is exposed to the action of air at bright red heat (about 700 deg. C.), and is then allowed to cool without interrupting the supply of air, an oxidizing decomposition takes place when dark-red heat (about 500 deg. C.) is reached, sulphurous acid being expelled, and a considerable amount of heat evolved; if sufficient air is then continuously passed through the charge, dense vapors of sulphurous acid escape, and the mixture gradually sinters together to a mass, in which the lead of the ore is present in the form of lead oxide, provided the air blast is continued long enough; there is no need to supply heat in this process—the heat liberated in the reaction is quite sufficient to keep it up.”

The inventors explained the process as follows:

“At a bright-red heat the calcium oxide (CaO) takes up oxygen from the air supplied, forming calcium peroxide (CaO₂), which latter afterward, in consequence of cooling down to dark-red heat, again decomposes into monoxide and oxygen; this nascent oxygen oxidizes a part of the lead sulphide to lead sulphate, which then reacts with a further quantity of lead sulphide, with evolution of sulphur dioxide and formation of lead oxide.”

Assuming the formation of calcium peroxide (CaO₂), the process leading to the desulphurization would therefore be represented as follows:

1. at 700° C. CaO + O = CaO₂
2. at 500° C. 4CaO₂ + PbS = 4CaO + PbSO₄
3. at the melting point PbS + PbSO₄ = 2PbO + 2SO₂ (?)

Reactions 1 and 2 combined, assuming the presence of sufficient oxygen, give:

PbS + 4CaO + 4O = PbSO₄ + 4CaO.

Now the invention consists in applying the observation described above to the working up of galena, and other ores containing lead sulphide, for metallic lead; and the essential novelty of the process therefore consists in passing air through the mass cooled to a dark-red heat (500 deg. C.).

This feature sharply distinguishes it from other known processes. It is true that in previous processes (compare the Tarnowitz reverberatory-furnace process, the roasting process used at Munsterbusch near Stolberg, and others) the lead ore was mixed with limestone or dolomite (which are converted into oxides in the early stage of the roast) and the heat was alternately raised and lowered; but in all cases only a surface action of the air was produced, the air supply being provided simply by the furnace draft. Passing air through the mass cooled down, as indicated above, leads to the important economic advantages of reducing the fuel consumption, the losses of lead, the manual labor (raking) and the dimensions of the roasting apparatus.

In order to carry out the process of this invention, the powdered ore is intimately mixed with a quantity of alkaline earth oxide, _e.g._, calcium oxide, corresponding to its sulphur content; if the ore already contains alkaline earth, the quantity to be added is reduced in accordance. The mixture is heated to bright-red heat (700 deg. C.) in the reverberatory furnace, in a strongly oxidizing atmosphere, is then allowed to cool down to dark-red heat (500 deg. C.), also in strongly oxidizing atmosphere, is transferred to a vessel called the “converter,” and atmospheric air is passed through at a slight pressure (the inventors have found a blast corresponding to 35 to 40 cm. head of water suitable).[19] The heat liberated is quite sufficient to keep the charge at the reaction temperature, but, if desired, hot blast may also be used. The mixture sinters together, and (while sulphurous acid gas escapes) it is gradually converted into a mass consisting of lead oxide, gangue and calcium sulphate, from which the lead is extracted in the metallic form, by any of the known methods, in the shaft furnace. The operation is concluded as soon as the mass, by continued sintering, has become impermeable to the blast. If the operation is properly conducted, the gas escaping contains only small quantities of volatile lead compounds, but on the other hand up to 8 per cent. by volume of sulphur dioxide. This latter can be collected and further worked up.

“In place of the oxide of an alkaline earth, ferrous oxide (FeO) or manganous oxide (MnO) may also be used.”

According to the reports on the practice of this process which have been published,[20] conical converters of about 1700 mm. (5 ft. 6 in.) upper diameter and 1500 mm. (5 ft.) depth are used in Australian works. At a new plant at Port Pirie (Broken Hill Proprietary Company) converters 2400 mm. (7 ft. 10 in.) in diameter and 1800 mm. (5 ft. 11 in.) deep have been installed. These latter will hold a charge of about eight tons. In the lower part of these converters, at a distance of about 600 mm. (2 ft.) from the bottom, there is placed an annular perforated plate, and upon this a short perforated tube, closed above by a plate having only a limited number of holes.

No details have been published with regard to the European installations. The general information which the Metallurgische Gesellschaft[21] placed at my disposal upon request some years ago, for use in my lecture courses, was restricted to data regarding the consumption of fuel and labor in roasting and smelting the ores, which was figured at about one-third or one-half of the consumption in the former processes, to the demonstration of the large output of the comparatively small converters, and to the reduced size of the roasting plant as the result. But the European establishments which introduced this process were bound by the owners of the patents, notwithstanding the protection afforded by the patents, to give no information whatever regarding the process to outsiders, and not to allow any inspection of the works.

On the other hand, a great deal appeared in technical literature which was calculated to excite curiosity. Moreover, as professor of metallurgy, it was my duty to instruct my pupils concerning this process among others, and it was therefore very gratifying to me that one of the students in my laboratory took a special interest in the treatment of lead ore. I gave him opportunity to install a small converter, in order to carry out the process on a small scale, and in spite of the slender dimensions of the apparatus the very first experiments gave a complete success.

However, I could not harmonize the explanation of the process given by the inventors with the knowledge which I had acquired in my many years’ practical experience in the manufacture of peroxides. It is clear from the patent specification that in the roasting operation at 700 deg. C. a compound must be formed which functions as an excellent oxygen carrier, for on cooling to 500 deg. C. the further oxidation then proceeds to the end not only without any external application of heat, but even with vigorous evolution of heat. No more striking instance than this could be desired by the theorists who have of recent years again become so enthusiastic over the idea of catalysis. Huntington and Heberlein regarded calcium peroxide as the oxygen carrier, but that is a compound which cannot exist at all under the conditions which obtain in their process. The peroxides of the alkaline earths are so very sensitive that in preparing them the small quantities of carbon dioxide and water must be extracted carefully from the air, and yet in the process, in an atmosphere pregnant with carbon dioxide, water, sulphurous acid, etc., calcium peroxide, the most sensitive of the whole group, is supposed to form! This could not be.

The only compounds known as oxygen carriers, and capable of existing under the conditions of the process, are calcium plumbate and plumbite. I have emphasized this point from the first in my lectures on metallurgy, when dealing with the Huntington-Heberlein process, and, in point of fact, this assumption has since been proved to be correct by the work of L. Huppertz, one of my students.

During my practical activity (1879-1891) I had prepared barium peroxide and lead peroxide in large quantities on a manufacturing scale, the last-mentioned through the intermediate formation of plumbites and plumbates:

2NaOH + PbO + O = Na₂PbO₃ + H₂O

or:

4NaOH + PbO + O = Na₄PbO₄ + 2H₂O.

An experiment made in this connection showed that calcium plumbate is formed just as readily from slaked lime and litharge as the sodium plumbates above. Litharge is an intermediate product, produced in large quantities in lead works, and must in any case be brought back into the process. If, then, the litharge is roasted at a low temperature with slaked lime, the roasting of the galena could perhaps be entirely avoided by introducing that ore together with calcium plumbate into the converter, after the latter had once been heated up. Mr. Huppertz undertook the further development of this process, but I have no information on the later experimental results, as he placed himself in communication with neighboring lead works for the purpose of continuing his investigation, and has not since then given me any precise data. I will therefore confine myself to the statement that the fundamental idea for the experiments, which Mr. Huppertz undertook at my suggestion, was the following:

To dispense with the roasting of the galena, which is necessary according to Huntington and Heberlein; in other words, to convert the galena by direct blast, with the addition of calcium plumbate, the latter being produced from the litharge which is an unavoidable intermediate product in the metallurgy of lead and silver. (Borchers, “Elektrometallurgie,” 3d edition, 1902-1903, p. 467.)

This alone would, of course, have meant a considerable simplification of the roast, but the problem of the roasting of galena has been solved in a better way by A. Savelsberg, of Ramsbeck, Westphalia, who has determined the conditions for directly converting the galena with the addition of limestone and water and without previous roasting. He has communicated the following information regarding these conditions:

In order that, in blowing the air through the mixture of ore and limestone, an alteration of the mixture may not take place owing to the lighter particles of the limestone being carried away, it is necessary (quite at variance with the processes in use hitherto, in which for the sake of economy stress is laid on the precaution of charging the ore as dry as possible into the apparatus) to add a considerable quantity of water to the charge before introducing it into the converter. The water serves this purpose perfectly, also preventing any change in the mixture of ore and limestone, which invariably occurs if the ore is used dry. The water, moreover, exerts a very beneficial action in the process, inasmuch as it aids materially in the formation and temporary retention of sulphuric acid, which latter then, by its oxidizing action, greatly enhances the reaction and consequently the desulphurization of the ore. Furthermore, the water tends to moderate the temperature in the charge by absorbing heat in its volatilization.

In carrying out the process the converter must not be filled entirely all at once, but first only in part, additional layers being charged in gradually in the course of the operation. In this way a uniform progress of the reaction in the mass is secured.

The following mode of procedure is advantageously adopted: A small quantity of glowing fuel (coal, coke, etc.) is introduced into the converter, which is provided at the bottom with a grate (perforated sheet iron), the grate being first covered with a thin layer of crushed limestone in order to protect it from the action of the red-hot coals and ore. Upon this red-hot fuel a uniform layer of the wetted mixture of crude ore and limestone is placed. When the surface of the first layer has acquired a uniform red heat, a fresh layer is charged on, and this is continued, layer by layer, until the converter is quite full. While the layers are still being put on, the blast is passed in at quite a low pressure, and only when the converter is entirely filled is the whole force of the blast, at a rather greater pressure, turned on. There then sets in a kind of slag formation, which, however, is preceded by a very vigorous desulphurization. After the termination of the process, which can be recognized by the fact that vapors cease to be evolved, and that the surface of the ore becomes hard, the converter is tipped over, and the desulphurized mass drops out as a solid cone of slag, which is then suitably broken up for the subsequent smelting in the shaft furnace.

Savelsberg explains the reaction of this process as follows:

“1. The particles of limestone act mechanically, gliding in between the particles of lead ore and separating them from one another. In this way a premature sintering is prevented, and the whole mass is rendered loose and porous.

“2. The limestone moderates the reaction temperature produced in the combustion of the sulphur, so that the fusion of the galena, the formation of dust and the separation of metallic lead are avoided, or at least kept within the limits permissible. The lowering of the temperature of reaction is due partly to the decomposition of the limestone into caustic lime and carbon dioxide, in which heat is absorbed, and partly to the consumption of the quantity of heat which is necessary in the further progress of the operation for the formation of a slag from the gangue of the ore and the lead oxide produced.

“3. The limestone gives rise to chemical reactions. By its decomposition it produces lime, which, at the moment of its formation, is converted into calcium sulphate at the expense of the sulphur in the ore. The calcium sulphate at the time of slag formation is converted into silicate by the silica present, sulphuric acid being evolved. The limestone therefore assists directly and forcibly in the desulphurization of the ore, causing the formation of sulphuric acid at the expense of the sulphur in the ore, the sulphuric acid then acting as a strong oxidizing agent toward the sulphur in the ore.”

The most conclusive proof for the correctness of the opinion which I expressed above, that it is very important to create at the beginning of the operation the conditions for the formation of as much sulphate as possible, has been furnished by Carmichael and Bradford. They recommend that gypsum be added to the charge in place of limestone. At one of the works of the Broken Hill Proprietary Company (where their process has been carried on successfully, and where lead ores very rich in zinc had to be worked up) the dehydrated gypsum was mixed with an equal quantity of concentrate and three times the quantity of slime from the lead ore-dressing plant, as in the table given herewith:

─────────────────┬────────┬─────────────┬──────────┬────────
│ OF THE │ OF THE │ OF THE │ OF THE
CONTENTS │ SLIME │ CONCENTRATE │ CALCIUM │ WHOLE
│ │ │ SULPHATE │ CHARGE
─────────────────┼────────┼─────────────┼──────────┼────────
Galena │ 24 │ 70 │ │ 29
Zinc blende │ 30 │ 15 │ │ 21
Pyrites │ 3 │ │ │ 2
Ferric oxide │ 4 │ │ │ 2.5
Ferrous oxide │ 1 │ │ │ 1
Manganous oxide │ 6.5 │ │ │ 5
Alumina │ 5.5 │ │ │ 3
Lime │ 3.5 │ │ 4.1 │ 10
Silica │ 23 │ │ │ 14
Sulphur trioxide │ │ │ 59 │ 12
─────────────────┴────────┴─────────────┴──────────┴────────

The charge is mixed, with addition of water, in a suitable pug-mill. The mass is then, while still wet, broken up into pieces 50 mm. (2 in.) in diameter, which are then allowed to dry on a floor in contact with air; in doing so they set hard, owing to the rehydration of the gypsum.

As in the case of the Savelsberg process, the converters are heated with a small quantity of coal, are filled with the material prepared in the manner above described, and the charge is blown, regulating the blast in such manner that, after the moisture present has been dissipated, a gas of about 10 per cent. SO₂ content is produced, which is worked up for sulphuric acid in a system of lead chambers.

The reactions are in this case the same as in the Savelsberg process, for here also calcium sulphate is formed transitorily, which, like other sulphates, reacts partly with sulphides, partly with silica.

Where gypsum is available and cheap, the Carmichael-Bradford process must be given preference; in all other cases unquestionably the Savelsberg process is superior, owing to its great simplicity.

LIME-ROASTING OF GALENA

BY W. MAYNARD HUTCHINGS

(_October 21, 1905_)

Much interest attaches to the paper by Professor Borchers, recently presented in the _Engineering and Mining Journal_ (Sept. 2, 1905) on “New Methods of Desulphurizing Galena,” together with an editorial on “Lime-Roasting of Galena”; it is a curious coincidence that the same issue contained also an article on the “Newer Treatment of Broken Hill Sulphides,” in which is shown the importance of the new methods as a contribution to actual practice.

For some years it had been a source of surprise to me that a new process, so interesting and so successful as the Huntington-Heberlein treatment of sulphide ores, should have received scarcely any notice or discussion. This lack, however, now appears to be remedied. The suggestion that the subject should be discussed in the _Journal_ is good, as is also that of the designation “Lime-Roasting” for a type-name. Such observations and experiments on the subject as I have had occasion to record have, for many years, figured in my note-books under that heading.

Whatever may be the final results of the later processes, now before the metallurgical world or still to come, there can be no doubt whatever that full and exclusive credit must be given to Huntington and Heberlein, not only for first drawing attention to the use of lime, but also for working out and introducing practically the process. It has been a success from the first; and so far as part of it is concerned, it seems to be an absolute and fundamental necessity which later inventors can neither better nor set aside. The other processes, since patented, however good they may be, are simply grafts on this parent stem.

It is, however, quite certain that Huntington and Heberlein, in the theoretical explanation of the process, failed to understand the most important reactions. Their attributing the effect to the formation and action of calcium peroxide affords a sad case of _a priori_ assumption devoid of any shred of evidence. As Professor Borchers points out, calcium peroxide, so difficult to produce and so unstable when formed, is an absolute and absurd impossibility under the conditions in question. Probably many rubbed their eyes with astonishment on reading that part of the patent on its first appearance, and hastened to look up the chemical authorities to refresh their minds, lest something as to the nature of calcium peroxide might have escaped them.

Fortunately the patent law is such that there was no danger of a really good and sound invention being invalidated by a wrong theoretical explanation by its originators. But, nevertheless, it was a misfortune that the inventors did not understand their own process. Had they known, they could have added a few more words to their patent-claims and rendered the Carmichael patent an impossibility.

Professor Borchers appears to consider that the active agent in the new process is calcium plumbate. That this compound may play a part at some stage of the process may be true; this long ago suggested itself to some others. We may yet expect to hear that the experiments undertaken by Professor Borchers himself, and by others at his instigation (in which calcium plumbate is separately prepared and then brought into action with lead sulphide), have given good results. But it does not appear so far that there is any real proof that calcium plumbate is formed in the Huntington-Heberlein or other similar processes; and it is difficult to see at what stage or how it would be produced under the conditions in question. This is a point which research may clear up, but it should not be taken for granted at this stage. Indeed, it seems to me that the results obtained may be fairly well explained without calling calcium plumbate into play at all.

Of course the action of lime in contact with lead sulphide excited interest many years before the new process came into existence. My own attention to it dates back more than a dozen years before that time (I was in charge of works where I found the old “Flintshire process” still in use).

Percy pointed out, in his work on lead smelting, that on the addition of slaked lime to the charge, at certain stages, to “stiffen it up,” the mixture could be seen to “glow” for a time. When I myself saw this phenomenon, I commenced to make some observations and experiments. Also (as others probably had done), I had observed that charges of lead with calcareous gangue are roasted more rapidly and better than others, and to an extent which could not be wholly explained by simple physical action of the lime present.

Simple experiments made in assay-scorifiers in a muffle, on lime roasting, are very striking, and I think quite explain a good part of what takes place up to a certain stage in the processes now under consideration. I tried them a number of years ago, on many sorts of ore, and again more recently, when studying the working of the new patents. For illustration, I will take one class of ore (Broken Hill concentrate), using a sample assaying; Pb, 58 per cent.; Fe, 3.6 per cent.; S, 14.6 per cent.; SiO₂, 3 per cent. The ore contained some pyrite. If two scorifiers are charged, one with the finely powdered ore alone, and one with the ore intimately mixed with, say, 10 per cent. of pure lime, and placed side by side just within a muffle at low redness, the limed charge will soon be seen to “glow.” Before the simple ore charge shows any sign of action, the limed charge rapidly ignites all over, like so much tinder, and heats up considerably above the surrounding temperature, at the same time increasing noticeably in bulk. This lasts for some time, during which hardly any SO₂ passes off. After the violent glowing is over, the charge continues to calcine quietly, giving off SO₂, but is still far more active than its neighbor. If, finally, the fully roasted charge is taken out, cooled and rubbed down, it proves to contain no free lime at all, but large quantities of calcium sulphate can be dissolved out by boiling in distilled water. For instance, in one example where weighed quantities were taken of lime and the ore mentioned, the final roasted material was shown to contain nearly 23 per cent. of CaSO₄; the quantity actually extracted by water was 20.2 per cent. Further tests show that the insoluble portion still contains calcium sulphate intimately combined with lead sulphate, but not extractable by water.

There is no doubt that when lead sulphide (or other sulphide) is heated with lime, with free access of air, the lime is rapidly and completely converted into sulphate. The strong base, lime, apparently plays the part of “catalyzer” in the most vigorous manner, the first SO₂ evolved being instantly oxidized and combined with the lime to sulphate, with so strong an evolution of heat that the operation spreads rapidly and still goes on energetically, even if the scorifier is taken out of the muffle. Also, the “catalytic” action starts the oxidation of the sulphides at a far lower temperature than is required when they are roasted alone.

If, in place of lime, we take an equivalent weight of pure calcium carbonate and intimately mix it with ore, we obtain just the same action, only it takes a little longer to start it. Once started, it is almost as vigorous and rapid, and with the same results. It does not seem correct to assume (as is usually done) that the carbonate has first to be decomposed by heat, the lime then coming into action. The reaction commences in so short a time and while the charge is still so cool, that no appreciable driving off of CO₂ by heat only can have taken place. The main liberation of the CO₂ occurs during the vigorous exothermic oxidation of the mixture, and is coincident with the conversion of the CaO into CaSO₄.

If, in place of lime or its carbonate, we use a corresponding quantity of pure calcium sulphate and mix it with the ore, we see very energetic roasting in this case also, with copious evolution of sulphur dioxide, only it is much more energetic and rapid and occurs at a lower temperature than in the case of a companion charge of ore alone.

It is very easily demonstrated that the CaSO₄ in contact with the still unoxidized ore (whether it has been introduced ready made or has been formed from lime or limestone added) greatly assists the further roasting, in acting as a “carrier” and enabling calcination to take place more rapidly and easily and at a lower temperature than would otherwise be the case.

The result of these experiments (whether we mix the ore with CaO, CaCO₃, or CaSO₄) is that we arrive with great ease and rapidity at a nearly dead-sweet roast. The lime is converted into sulphate, and the lead partly to sulphate and partly to oxide. Two examples out of several, both from the above ore, gave results as follows:

No. 1—Roasted with 20 per cent. CaCO₃ (= 11.2 per cent. CaO); sulphide sulphur, 0.02 per cent.; sulphate sulphur, 9.30 per cent.; total sulphur, 9.32 per cent.

No. 8—Roasted with 27.2 per cent. CaSO₄ (= 11 per cent. CaO); sulphide sulphur, 0.05 per cent.; sulphate sulphur, 11.28 per cent.; total sulphur, 11.33 per cent.

If these calcined products are now intimately mixed with additional silica (in about the proportions used in the Huntington-Heberlein process) and strongly heated, fritting is brought about and the sulphur content is reduced by the decomposition of the sulphates by the silica. Thus, the resultant material of experiment No. 1, above, when treated in this manner with strong heat for three hours, was sintered to a mass which was quite hard and stony when cold, and which contained 6.75 per cent. of total sulphur. Longer heating drives out more sulphur, but a very long time is required; in furnaces, and on a large scale, it is with great difficulty and cost that a product can be obtained comparable with that which is rapidly and cheaply turned out from the “converters” of the new process.

To return to the Huntington-Heberlein process, working, for example, on an ore more or less like the one given above, we may assume that, during the comparatively short preliminary roast, the lime is all rapidly converted into CaSO₄ and that some PbSO₄ is also formed (but not much, as the mixture to be transferred from the furnace to the converter requires not less than 6 to 8 per cent. of sulphur to be still present as sulphide, in order that the following operation may work at its best). As the blast permeates the mass, oxidation is energetic; no doubt that CaSO₄ here plays a very important part as a carrier of oxygen, in the same manner as we can see it act on a scorifier or on the hearth of a furnace.

What the later reactions are does not seem so clear. They are quite different from those on the scorifier or on the open hearth of a furnace, and result in the rapid formation (in successive layers of the mixture, from the bottom upward) of large amounts of lead oxide, fluxing the silica and other constituents to a more or less slaggy mass, which decomposes the sulphates and takes up the CaO into a complex and easily fused silicate. It is true that, as a whole, the contents of a well-worked converter are never very hot, but locally (in the regions where the progressive reaction and decomposition from below upward is going on) the temperature reached is considerable. This formation of lead oxide is so pronounced at times that one may see in the final product considerable quantities of pure uncombined litharge.

When the work is successful, the mass discharged from the converters is a basic silicate of PbO, CaO, and oxides of other metals present, and nearly all the sulphates have disappeared. A large piece of yellow product (which was taken from a well-worked converter) contained only 1.1 per cent. of total sulphur.

It may be that calcium plumbate is formed and plays a part in these reactions; but its presence would be difficult to prove, and its formation and existence during these stages would not be easy to explain. Neither does it seem necessary, as the whole thing appears to be capable of explanation without it.

While the mixture in the converter is still dry and loose, energetic oxidation of the sulphides goes on, with the intervention of the CaSO₄ as a carrier. As soon as the heat rises sufficiently, fluxing commences in a given layer and sulphates are decomposed. The liberated sulphuric anhydride, at the locally high temperature and under the existing conditions, will act with the greatest possible vigor on the sulphides in the adjacent layers; these layers will then in their turn flux and act on those above them, till the whole charge is worked out. The column of ore is of considerable hight, requiring a blast of 1½ lb., or perhaps more, in the larger converters now used. This pressure of the oxidizing blast (and of the far more powerfully oxidizing sulphuric anhydride, continuously being liberated within the mass of ore, locally very hot) constitutes a totally different set of conditions from those obtained on the hearth of a furnace with the ore in thin layers, where it is neither so hot nor under any pressure. It is to these conditions, in which we have the continued intense action of red-hot sulphuric anhydride under a considerable pressure (together with the earlier action of the CaSO₄), that the remarkable efficiency of the process seems to me to be due.

In the Carmichael process, the preliminary roast is done away with, CaSO₄ being added directly instead of having to be formed during the operation from CaO and the oxidized sulphur of the ore. The charge in the converter has to be started by heat supplied to it, and the work then goes forward on the same lines as in the Huntington-Heberlein process, so that we may assume that the reactions are the same and come under the same explanation.

Carmichael was quick to see what was really an important part and a correct explanation of the original process. He was not misled by wrong theory about any mythical calcium peroxide, and so he obtained his patent for the use of CaSO₄ and the dispensing of the roast in a furnace.

This process would always be limited in its application by the comparative rarity of cheap supplies of gypsum, but it appears to be a great success at Broken Hill; there it is not only of importance in working the leady ores, but also for making sulphuric acid for the new treatment of mixed sulphides by the Delprat and Potter methods. For this purpose, the use of CaSO₄ will have the additional advantage that the mixture to be worked in the converter will contain not only the sulphur of the ore, but also that of the added gypsum; on decomposition, it will yield stronger gases for the lead chambers of the acid plant.

Finally comes the Savelsberg patent, which is the simplest of all; not only (like the Carmichael process) avoiding the preliminary roast with its extra plant, but also not requiring the use of ready-made CaSO₄, as it uses raw ore and limestone directly in the converter. I have no knowledge as to actual results of this process; and, so far as I am aware, nothing on the subject has been published. But Professor Borchers evidently has some information about it, and regards it as the most successful of the methods of carrying out the new ideas. On the face of it, there seems no reason why it should not attain all the results desired, as the chemical and physical actions of the CaO, and of the CaSO₄ formed from it, should come into play in the same manner and in the same order as in the original process; as it is carried out in the identical converter used by Huntington and Heberlein, the final reactions (as suggested above) will take place under the same conditions as to continuous decomposition _under considerable heat and pressure_, which I regard as the most vital part of the whole matter.

It is well to emphasize again the fact that the idea, and the means of obtaining these vital conditions, owe their origination to Huntington and Heberlein.

THEORETICAL ASPECTS OF LEAD-ORE ROASTING[22]

BY C. GUILLEMAIN

(March 10, 1906)

It is well known that the process of roasting lead ores in reverberatory furnaces proceeds in various ways according to the composition of the ore in question. Thus in roasting a sulphide lead ore rich in silica, one of the reactions is:

PbS + 3O = PbO + SO₂.

But this reaction is incomplete, for the gases which pass on in the furnace are rich in SO₂ and in SO₃. And so it is found that whatever lead oxide is formed passes over almost immediately into lead sulphate, according to the reaction:

PbO + SO₂ + O = PbSO₄.

This reaction is the chief one which takes place. Whether the silicious gangue serves as a catalyzer for the sulphur dioxide, or whether it serves merely to keep the galena open to the action of the gases, the end result of the roast is usually the formation of lead sulphate according to the above reaction.

In the case of an ore rich in galena, a slow roast is essential, for it is desired to have the following reaction take place during the latter part of the roast:

PbS + 3PbSO₄ = 4PbO + 4SO₂.

Now, if the heating were too rapid, not enough lead sulphate would be found to react with the unaltered galena. The quick roasting of a rich ore would result in the early sintering of the charge, and sintering prevents the further formation of lead sulphate. Whether this sintering (which takes place so easily and which is so harmful in the latter part of the process) is due to the low melting point of the lead sulphide, whether the heat evolved by the reaction

PbS + 3O = PbO + SO₂

is sufficient to melt the lead sulphide, or whether other thermochemical effects (notably the preliminary sulphatizing of the lead sulphide) come into play, must for the present be undecided. Suffice it to say that the sintering of the charge works against a good roast.

In the Tarnowitz process a definite amount of lead sulphide is converted into lead sulphate by a preliminary roast. The sulphate then reacts with the unaltered lead sulphide, and metallic lead is set free, thus:

PbS + PbSO₄ = 2Pb + 2SO₂.

But when a very little of the sulphide has been transformed into sulphate, and when there is so little of the latter present that only a small amount of lead sulphide can be reduced to metallic lead, the mass of ore begins to sinter and grow pasty. Very little lead could be formed were it not for the addition of crushed lime to the charge just before the sintering begins. This lime breaks up the charge and cools it, prevents any sintering, and allows the continued formation of lead sulphate.

It scarcely can be held that the lime has any chemical effect in forming lead sulphate, or in forming a hypothetical compound of lead and calcium. Even if such theories were tenable from a physico-chemical point of view, they would be lessened in importance by the fact that other substances, such as purple ore or puddle cinder, act just as well as the lime.

There are now to be mentioned several new processes of lead-ore roasting whose operations fall so far outside the common ideas on the subject that their investigation is full of interest. For a long time the attempt had been made to produce lead directly by blowing air through lead sulphide in a manner analogous to the production of bessemer steel or the converting of copper matte. In the case of the lead sulphide, the oxidation of the sulphur was to furnish the heat necessary to carry on the process.

After many attempts along this line, Antonin Germot has perfected a method wherein, by blowing air through molten galena, metallic lead is obtained.[23] About 60 per cent. of a previously melted charge of galena is sublimed as lead sulphide, and the rest remains behind as metallic lead. The disadvantages of the process are the difficulties of collecting all of the sublimate and of working it up. Moreover, it is impossible as yet to secure two products of which one is silver-free and the other silver-bearing. The silver values are in both the metallic lead and in the sublimed lead sulphide.

While the process just described answers for pure galena, it fails with ores which contain about 10 per cent. of gangue. In the case of such ores, they form a non-homogeneous mass when melted, and the blast penetrates the charge with difficulty. If the pressure is increased the air forces itself out through tubes and canals which it makes for itself, and the charge freezes around these passages.

Messrs. Huntington and Heberlein have gone a little farther. Although they are unable to obtain metallic lead directly, they prepare the ore satisfactorily for smelting in the blast furnace, after their roasting is completed. The inventors found that if lead sulphide is mixed with crushed lime, heated with access of air, and then charged into a converter and blown, the sulphur is completely removed in the form of sulphur dioxide. The charge, being divided by the lime, remains open uniformly to the passage of air, and sinters only when the sulphur is eliminated.

The inventors announce, as the theory of their process, that at 700 deg. C. the lime forms a dioxide of calcium (CaO₂) which at 500 deg. C. breaks down into lime (CaO) and nascent oxygen. This nascent oxygen oxidizes the lead sulphide to lead sulphate according to the reaction:

PbS + 4O = PbSO₄.

Furthermore it is claimed that the heat evolved by this last reaction is large enough to start and keep in operation a second reaction, namely

PbS + PbSO₄ = 2PbO + 2SO₂.

The theory, as just mentioned, cannot be accepted, and some of the reasons leading to its rejection will be given.

It is well established that the simple heating of lime with access of air will not result in further oxidation of the calcium. The dioxide of calcium cannot be formed even by heating lime to incandescence in an atmosphere of oxygen, nor by fusing lime with potassium chlorate. Moreover, calcium stands very near barium in the periodic system. And as the dioxide of barium is formed at a low temperature and breaks up on continued heating, it seems absurd to suppose that the dioxide of calcium would act in exactly the opposite manner. Moreover, a consideration of the thermo-chemical effects will disclose more inconsistencies in the ideas of the inventors. The breaking up of CaO₂ into CaO and O is accompanied by the evolution of 12 cal. The reaction of the oxygen (thus supposed to be liberated) upon the lead sulphide is strongly exothermic, giving up 195.4 cal. So much heat is produced by these two reactions that, if the ideas of the inventors were true, the further breaking up of the calcium dioxide would stop, as the whole charge would be above 500 deg. C. It appears, then, that the explanations suggested by Messrs. Huntington and Heberlein are untrue.

In the usual roasting process, as carried out in reverberatory furnaces, it is well established that the gangue, and whatever other substances are added to the ore, prevent mechanical locking up of charge particles, since they stop sintering. It is not at all improbable that in the new roasting process the chief, if not the only, part played by the lime is the same as that played by the gangue in reverberatory-furnace roasting. A few observations leading to this belief will be given.

It is known that other substances will answer just as well as lime in this new roasting process. Such substances are manganese and iron oxides. Not only these two substances, but in fact any substance which answers the purpose of diminishing the local strong evolution of heat, due to the reaction:

PbS + 3O = PbO + SO₂,

serves just as well as the lime. This fact is proved by exhaustive experiments in which mixtures of lead sulphide on the one hand, and quartz, crushed lead slags, iron slags, crushed iron ores, crushed copper slags, etc., on the other hand, were used for blowing. All these substances are such that any chemical action, analogous to the splitting up of CaO₂, or the formation of plumbates as suggested by Dr. Borchers, cannot be imagined. The time is not yet ripe, without more experiments on the subject, to assert conclusively that there is no acceleration of the process due to the formation of plumbates through the agency of lime. But the facts thus far secured point out that such reactions are, at least, not of much importance.

Theoretical considerations point out that it ought to be possible to avoid the injurious local increase of temperature during the progress of this new roasting process, without having to add any substance whatever. To explain: The first reaction taking place in the roasting is

PbS + 3O = PbO + SO₂ + 99.8 cal.

Now the heat thus liberated may be successfully dispersed if there is, in simultaneous progress, the endothermic reaction:

PbS + 3PbSO₄ = 4PbO + 4SO₂ - 187 cal.

Hence if there could be obtained a mixture of lead sulphide and of lead sulphate in the proportions demanded by the above reaction, then such a mixture ought to be blown successfully to lead oxide without the addition of any other substance. Such a process has, in fact, been carried out. The original galena is heated until the required amount of lead sulphate has been formed. Then the mixture of lead sulphide and of lead sulphate is transferred to a converter and blown successfully without the addition of any other substance.

The adaptability of an ore to the process just mentioned depends on the cost of the preliminary roast and the thoroughness with which it must be done. As is known, when lead sulphide is heated with access of air, it is very easy to form sintered incrustations of lead sulphate. If these incrustations are not broken up, or if their formation is not prevented by diligent rabbling, the further access of air to the mass is prevented and the oxidation of the charge stops. If ores with such incrustations are placed in the converter without being crushed, they remain unaltered by the blowing. If the incrustations are too numerous the converting becomes a failure.

It has been found that the adoption of mechanical roasting furnaces prevents this. Such furnaces appear to stop the frequent failures of the blowing which are due to the lack of care on the part of the workmen during the preliminary roasting. Moreover, in such mechanical furnaces a more intimate mixture of the sulphide with the sulphate is obtained, and the degree of the sulphatizing roast is more easily controlled.