Зображення сторінки
PDF
ePub

Calcu- II. Product of the double decomposition, I. II. III. lated at 356° F., of protochloride of iron and composition. persulphuret of potassium. Oxide of zinc.. 0·639 0·636 0.635 0.648

Carbonic acid

(by the dif

ference).... 0·361 0·364 0·365

0.352 1.000 1.000 1.000 1.000

SULPHURETS OF MANGANESE.

Protosulphuret (MnS).

Amorphous black dust, sometimes adhering as a steel-gray covering to the sides of the vessel; rather alterable in humid air; it appears unalterable in dry air. This The products I., II. and III. were pre-sulphuret is prepared with an alkaline propared in the same circumstances as the corresponding products of manganese.

The neutral carbonate of copper (mysorine) and the blue hydrocarbonate appear to require conditions of temperature which I have not yet succeeded in realising. I have regularly obtained only the green carbonate, whose composition corresponds to that of malachite.

ARTIFICIAL FORMATION OF SOME
SULPHURETS.

A great number of metallic sulphurets are readily precipitated by the humid way in laboratories; those only which have not yet been produced in this manner will be treated of here. I obtained them by the double decomposition, at a high temperature, of a soluble metallic salt, and of a more or less sulphuretted alkaline sulphuret.

When the double decomposition precipitates sulphur in excess, the latter fuses into globules and separates.

I enclosed the alkaline sulphuret in a tube empty of air, with a bulb containing the metallic salt, and a bubble of air to break it.

BISULPHURET OF IRON (FeS2). An amorphous black dust, sometimes adhering as a yellow metalloid covering on the sides of the glass tube; scarcely alterable in the air when it is moistened, unalterble when it is dry; combustible; unattackable by hydrochloric acid. Bisulphuret of iron thus prepared, and exposed to the air for more than a year, was not converted into sulphate, notwithstanding its fine state of division. It is therefore very probable that this product corresponds to yellow pyrites. Bisulphuret of iron formed, at the ordinary temperature, by the slow reaction of the sulphurets on decomposing organic matters, effloresced, on the contrary, very rapidly, and appeared to be a white pyrites.

Calculated

I. II. composition. Iron ... 0454 0.458 0.467 Sulphur. 0.541 0.545 0.533

0.995 1.003 1.000

I. Product of the double decomposition, at 329° F., of sulphate of iron and persulphuret of potassium.

tosulphuret. When the latter contains an excess of sulphur, the product contains bisulphuret of manganese; when it contains an excess of alkali, the product contains some oxide. An alkaline protosulphuret mixed with carbonate would give a sulphuret of manganese itself, mixed with carbonate like the native sulphuret.

Calculated

composition. 0.625 0.633

Manganese...
Sulphur (by the difference) 0-375 0-367
1.000 1.000

The product was obtained at 365° F.

Bisulphuret (MnS2).

Amorphous brick-red dust, sometimes adhering as a transparent red covering to the sides of the tubes; scarcely alterable in the air when it is moistened, and quite unalterable when it is dry. This compound corresponds to hauerite; it is obtained, like the preceding, with an alkaline persulphuret.

I.

II. Calculated composition. 0.460 0.468 0.535 0.532

1.000

Manganese... 0.462
Sulphur...... 0.533

0.995 0.995

[blocks in formation]

Calculated drogen very soon gave a solution more or composition. less charged with hydrosulphuric acid, at Nickel (by the difference) 0-637 0.649 the same time as a deposit of sulphur, Sulphur. 0.360 0.351 which however presented no inconvenience. The tube was exposed to heat, and the sulphuret, at least partially dissolved, was found to have a crystalline appearance.

...

0.997 1.000 Product obtained at 320° F., with chloride of nickel and a protosulphuret of potassium.

Sulphuret (Ni3S1).

Heat appears to disengage the hydrosulphuric acid from its solution, for violent explosions occur. This gas appears to differ

The same characters; likewise forms on from carbonic acid, because it dissolves the glass a yellowish metalloïd covering.

Calculated composition.

[blocks in formation]

Zinc....
Sulphur.

Calculated

composition. 0.662 0.670 0.443 0.330

1.005

1.000

Product prepared at 347° F. All the sulphurets thus prepared by sudden double decomposition are almost amorphous; they take the metalloïd state and color only in being applied to glass. However, they are found crystallised in nature; they must, therefore, have been formed slowly, or have undergone the action of a solvent.

Hydrosulphuric acid appears to be, in certain conditions of temperature and pressure, a solvent of the metallic sulphurets. Indeed, I have been enabled to dissolve some of them by means of this gas; but the experiment presents some difficulties, and we can operate on only very small quantities of matters.

Hydrosulphuric acid, indeed, is sparingly soluble in water, and, besides, of itself, has only a very feeble solvent power. This faculty appears, it is true, to increase with the temperature; but a great increase of pressure is required to retain the gas united to the liquid.

sulphurets less easily than the latter does the carbonates, and because it is less soluble in water when the temperature increases; so that at equal heat, its solutions should possess more powerful tension.

This difference would explain very well the concentration of the sulphurets at a depth in all metalliferous beds, and the accumulation of carbonates in the upper part of them, since the solvent faculty of hot waters saturated with the two gases diminishing in proportion as the latter would be disengaged unequally under a decreasing hydrostatic pressure, the dissolved matters should be precipitated gradually, the sulphurets before the car

bonates.

I will add to the foregoing experiments some results which have no immediate application to the formation of concreted veins, but which, perhaps, are not uninteresting in a geological point of view.

Very finely divided amorphous anhydrous oxide of iron is diffused in many geological formations, and its beds presents a double character.

Sometimes it impregnates a whole collection of rocks of mechanical formation, evidently stratified, and which have not undergone any important derangement since their deposit; so that it is impossible to help regarding the oxide of iron as precipitated in the anhydrous state in the water which held in suspension the muddy and sandy matters. Sometimes the presence of the anhydrous oxide is local, and evidently related in neighborhood and origin to certain eruptive mineral masses.

In these two essentially different metallic veins the oxide of iron frequently accompanies dolomites, rock salt, anhydrites, and gypsum; and although these matters show themselves in different manners, the narrow connection of age and metallic vein existing between them, their enduring association, Into a thick glass tube of small bore, even in general conditions of formation almost filled with boiled water, I introduced absolutely opposed, prove that so constant a few parcels of metallic sulphuret and bi-a concurrence of effects could not be fortusulphuret of hydrogen; the tube exhausted of air, was then hermetically sealed. The decomposition of the bisulphuret of hy

itous, and permit us to sec in it only various products of the same natural operation. It must therefore be concluded that the same

phenomena or at least analogous phenomena ¦ of anhydrites, red oxide of iron and rock have occurred in the liquid masses which salt. deposited certain stratified rocks, and have accompanied the appearance of certain eruptive rocks.

It is already known from Haïdinger's beautiful experiment, that dolomite and anhydrite may be a double epigenic product of the reaction by the humid way of sulphate of magnesia on carbonate of lime; from the foregoing experiments it results that calcareous magnesian salts and even pure carbonate of magnesia will be deposited in certain cases in superheated waters; perhaps a step may have been made towards the solution of the problem if we show that the dishydratation of the sesquioxide of iron is operated in the same conditions, and does not necessarily suppose reactions comparable to those in the dry way.

I decomposed a solution of perchloride of iron by carbonate of lime or carbonato of soda at a temperature of at least 392° F. maintained for forty-eight hours, and I obtained a precipitate of very finely divided red, anhydrous oxide of iron, sparingly soluble in nitric acid. I repeated this experiment with the same results at temperatures between 320° F. and 356° F., maintained for eight days; finally, I dishydrated, in the same conditions, the hydrated peroxide of iron in suspension in a saturated solution of common salt or chloride of calcium, and in pure water. Nothing, however, indicates that these temperatures were the lowest at which the same effects could be obtained.

If it now be remarked that the sulphate of lime is precipitated anhydrous in a superheated liquid, and that the solubility of carbonate of soda, superior to that of chloride of sodium, increases much more rapidly with heat; we may take as a basis the reactions which have just been separately studied, and which are comparatively simple, in order to arrive at a conception of the complex phenomena which have given rise to the above-mentioned multiple association of mineral matters.

These phenomena may, indeed, be explained, if not in their details, at least collectively, by various combinations which would likewise satisfy the general data of the problem; it would be sufficient, for example, that waters charged with chlorides of calcium, magnesium and iron become mixed, under certain conditions of temperature, and, consequently, of pressure, with waters saturated with carbonate and containing more or less sulphate of soda, in order to give rise at the same time to deposits of calcareous magnesian salts,

ON THE EMPLOYMENT OF HYDROGEN IN THE ANALYSES OF MINERAL SUBSTANCES.*

BY M. L. E. RIVOT, MINING ENGINEER,

DIRECTOR OF THE LABORATORY OF THE SCHOOL OF MINES.

THE reductive action exerted by dry hydrogen gas, at a more or less elevated temperature, on certain metallic oxides, may be employed with great advantage for separating these oxides from other fixed bases, on which hydrogen gas has no action at any temperature. To the investigations already made with the view of turning this action to account, I think I may add the results of several separations, and of several analyses, difficult of performance by the processes hitherto employed. By employing hydrogen, I have been enabled to separate oxide of iron from the earths, alumina, glucina and zirconia; oxide of iron from oxide of chromium; and oxide of tin from silica.

I will first explain the method of separation and the results which I have obtained, and will afterwards point out the process of analysis on which I decided for chrome iron ore and native oxide of tin. SEPARATION OF OXIDE OF IRON AND

ALUMINA.

The process of separation described in the various treatises on chemical analysis consists in treating the weighed mixture of the two oxides by potassa, either by the dry way in a silver crucible, or by the humid way, after solution in an acid, in separating the oxide of iron and washing it thoroughly in boiling water. This process is very long, and often inaccurate, when the potassa employed is not perfectly pure. The separation of the two oxides of lead being very often required in the analysis of mineral substances and the products of manufactories, I sought for a quicker and more accurate process. I have obtained very accurate results by reducing, by dry hydrogen, at a bright red heat, the weighed mixture of oxide of iron and alumina, precipitated by ammonia from an acid solution, allowing it to cool in the hydrogen, and afterwards treating with very dilute cold nitric acid the mixture of alumina and metallic iron; the iron dissolves very easily, whilst the alumina is not attacked.

* Annales de Chimie et de Physique, October, 1850.

Very few details are necessary for explaining the mode of operation.

The oxide of iron and alumina are precipitated together by ammonia; the precipitate is dried and separated from the filter; the filter is burned, and the ashes are united to the two hydrates; the whole is calcined in a platinum crucible at a full red heat. The mixture of the two oxides is porphyrised and accurately weighed, then placed in a porcelain boat, likewise weighed, in a porcelain tube placed horizontally in a reverberatory furnace. To one of the extremities is adapted a glass tube drawn out at one

[blocks in formation]

gr. gr. gr.

end; at the other a gentle stream of ling Loss of weight in the 20-156 0-132 0-1616

drogen is passed in, being dried by passing through a tube containing chloride of calcium, and a flask containing monohydrated sulphuric acid.

When the air of the apparatus is completely expelled, the porcelain tube is gradually heated to bright redness, and this temperature is maintained so long as water is deposited on the sides of the glass tube at the extremity of the apparatus. In none of my experiments has it taken more than an hours' heating to accomplish the reduction.

The tube is then allowed to cool, the current of hydrogen being continued. When the tube is quite cold, the porcelain boat is removed and weighed. The loss of weight indicates the oxygen of the peroxide of iron, and permits its proportion in the mixture to be calculated. However, it is not advisable to be satisfied with this determination when the alumina is in large proportion, and when care has not been taken throughout to make the gas pass in very slowly, a little alumina may be carried away by the hydrogen, and, consequently, the loss of weight would lead to rather too large a proportion of peroxide of iron.

The mixture of metallic iron and alumina is digested for twenty-four hours without heat, in very weak nitric acid. The ordinary pure acid must be employed diluted with at least thirty times its bulk of water, or, better still, acid much more diluted, or pure water, adding at different times small quantities of nitric acid, so as to maintain a very slow effervescence of hydrogen gas; by operating thus, we are certain to completely dissolve the iron without attacking the alumina. The solution of the iron is complete when the alumina has become nearly white.

The alumina is separated by filtration. In the nitric solution, previously heated to peroxidise the iron, the peroxide of iron is precipitated by ammonia. In this manner, the two oxides separated may be weighed.

hydrogen...

Corresponding to

Peroxide of iron.... 0.510

1

1

0.431 0.527

[blocks in formation]

From these numbers it results that the separation of the two oxides by the process pointed out is very accurate: that the two oxides may be accurately estimated by weighing directly the peroxide of iron, and calculating the alumina by the difference; and, finally, that the loss of weight in the hydrogen leads to rather too high a proportion of peroxide of iron when the quantity of alumina is somewhat considerable, owing to the latter being carried away by the gaseous current, and on the contrary, to a very accurate result when the quantity of alumina is small.

I must remark that, if the hydrogen be passed into a mechanical mixture of alumina and oxide of iron instead of operating on the mixture of the two oxides precipitated by ammonia, the alumina is much more easily carried away by the current of gas.

To calculate the quantity of peroxide of iron from the loss of weight in the hydrogen, I took for the equivalent of iron, the number formerly admitted, 339. It is that to which two experiments, made with the greatest care, by the reduction, in hydrogen gas, of 1 gramme (15.434 grs.) of very pure peroxide of iron led me. I obtained :

[blocks in formation]

1

SEPARATION OF OXIDE OF IRON AND

ZIRCONIA.

The separation of these two bases, by the reductive action of hydrogen on the peroxide of iron, is effected with the same facility and the same accuracy as that of oxide of iron and alumina. The result obtained by weighing zirconia is more accurate, because this earth, being heavier than alumina, is not perceptibly carried away by the gaseous current. We may also use nitric acid less diluted to dissolve the iron reduced to the metallic state, because calcined zirconia does not dissolve at all in the acids.

I submitted to experiment the mixture:-
Peroxide of iron....
gr. 0·660
Zirconia ....
gr. 0.377

......

.....

I obtained the following results :Loss of weight in hydrogen gr. 0.205 corresponding to

.....

Peroxide of iron .....
gr. 0.667
Afterwards by weighing the zirconia and

oxide of iron, I obtained:

[blocks in formation]

Zirconia... gr. 0.375 Peroxide of iron gr. 0.668 These numbers prove that the separation of the two oxides by the means proposed is quite accurate, and that the composition of the mixture may be deduced from the loss of weight resulting from the action of the hydrogen, considered as the oxygen of the peroxide of iron.

This process of separation is more rapid and more accurate than those hitherto proposed.

SEPARATION OF OXIDE OF IRON AND

GLUCINA.

This separation, made with the same precautions as that of alumina and oxide of iron, leads to quite as accurate results. Glucina, like alumina, may be carried away by the current of gas, when its quantity is considerable; and, consequently, when the glucina is in large proportion, the composition of the mixture should not be determined solely by the loss of weight resulting from the action of the hydrogen. The metallic iron must be dissolved in dilute nitric acid without heat. The acid to be used should not contain less than thirty parts of water to one part of ordinary pure acid.

I submitted to experiment the mixture:-
Peroxide of iron
gr. 0.815
Glucina
gr. 0.399

.....

I obtained the following numbers :Loss of weight in hydrogen corresponding to

.....

gr. 0.249

.... 0-812

Peroxide of iron On afterwards treating the mixture of metallic iron and glucina by dilute nitric acid, I found:

Peroxide of iron
Glucina

gr. 0.816 gr. 0.397 The process indicated succeeds therefore very well for glucina and oxide of iron.

I have not made any experiments as to the separation of oxide of iron from yttria and thoria.

The reductive action of hydrogen on the oxides of cobalt and nickel, may usefully be applied to their separation from alumina.

When, in an acid solution containing the oxides of cobalt and nickel with alumina, the alumina is precipitated by ammonia, the precipitate retains a certain proportion of oxides of cobalt and nickel. Quite pure alumina may be obtained by calcining the precipitate given by ammonia, and treating it by dry hydrogen, and afterwards, like the mixture of iron and alumina, by very dilute cold nitric acid, which dissolves only the cobalt and nickel. SEPARATION OF OXIDE

SILICA.

OF TIN AND

The reductive action of hydrogen on oxide of tin may be employed with advantage for the accurate separation of oxide of tin and silica.

I operate in the following manner;

The

The mixture of oxide of tin and silica, strongly calcined and weighed, is placed in a porcelain boat likewise weighed,-in a porcelain tube as in the experiment already described. The hydrogen is made to pass in very slowly, in order that the silica may not be carried away by the gaseous current, and heated only to dull redness. reduction of the oxide of tin takes place very rapidly. After cooling in hydrogen, the mixture of metallic tin and silica is presented under the form of a grey powder, without metallic globules, at least if the temperature be not carried too high, or if the oxide of tin be not nearly pure. It is weighed the loss of weight gives the oxygen of the oxide of tin and enables us to calculate approximatively the composition of the mixture. However, as the silica may have been carried away in small quantity by the gaseous current, it is better to dissolve the metallic tin in nitro-hydrochloric acid, to weigh the undissolved silica, and, when the weight does not exactly correspond with the quantity of silica calculated from the loss sustained by the mixture in hydrogen, to estimate the oxide of tin in the nitro-hydrochloric solution. The process which I found most successful in this estimation is the following:- I saturate the acids with ammonia; I add sufficient hydrosulphate of ammonia to completely dissolve the tin; I decompose the sulphosalt of tin with hydrochloric acid,

« НазадПродовжити »