156

XXII.-Some Notes on Hydrated Ferric Oxide and its Behaviour with Hydrogen Sulphide.

By LEWIS T. WRIGHT.

FOR the purpose of studying some points in the reaction between hydrogen sulphide and hydrated ferric oxide, I had occasion to prepare some of the latter material by precipitation of ferric hydrate from a ferric chloride solution by means of liquid ammonia in the usual way; but I did not find it possible in this manner to prepare a material perfectly free from basic chloride. I used both the method of pouring excess of liquid ammonia into ferric chloride solution, and that of pouring ferric chloride solution into excess of liquid ammonia, and always experienced that the resulting precipitates of ferric hydrate, when washed most exhaustively with boiling water with the aid of a filter-pump, never failed to yield minute traces of chlorine even in the last washings.

After many washings, the filtrates cease to yield any decided precipitate with silver nitrate, but continue to give a persistent opalescence indicative of minute traces of chlorine. On drying the washed ferric hydrate at about 100° in a water-bath for a few hours, and again washing with boiling water, the first portions of the wash-water give a strong chlorine reaction, showing that in the process of drying some basic chloride has been dissociated. I have also noticed that the first washings are acid, as might be supposed if basic chloride had been dissociated. On repeating the drying and washing of the ferric oxide in the same manner as before, I found that the wash-water gave both acid and chlorine reactions.

In order to avoid the great inconvenience of handling a bulky gelatinous precipitate, I thought to obtain a material in a more compact form by the following method of preparation.

Ferric chloride solution was added slowly to excess of liquid ammonia, with constant stirring. This operation was effected in a porcelain evaporating dish, which was then placed in a water-bath, and its contents evaporated to dryness. By this means, a dirty redbrown mass was obtained, showing here and there patches of ammonium chloride. This mass, on being treated with water, fell mostly into an impalpable powder, a large portion of which, on attempted filtration, passed through the filter-paper and even through many successive filters. The filtrates were of bright red colour, and when dilute, had in certain lights a purple appearance, suggestive of the purple colour of ignited ferric oxide. Each filtrate, on being boiled

and refiltered, left on the paper a small quantity of red hydrated ferric oxide, but the filtrate was still red. Portions of these red filtrates, which all had an acid reaction, on treatment with a little ammonia to alkaline reaction and boiling, partially coagulated, and left decided quantities of bright red powder on the filter, but the filtrates were still red, not having the appearance of solutions, being muddy and opaque, with what appeared to be an unfilterable precipitate. This is probably similar to the material called "colloïdal ferric hydrate,” "dialysed iron," or "fer Bravais." Magnier de la Source (Compt. rend., 90, 1352-1354) in describing a similar condition of ferric hydrate, or compound of ferric hydrate and ferric chloride, supposes that under certain conditions ferric hydrate is soluble in water; but I cannot think that this is a case of true solution, but rather of pseudo" solution; for the filtrates I obtained had all the appearance of holding in suspension an intensely impalpable powder. They deposited small quantities of red powder on standing for some weeks, and in cases where they had been treated with ammonia and boiled to neutrality, they exhibited after long standing an acid reaction. Now, however, a drop of liquid ammonia and the application of a little heat, caused the pseudo solution to coagulate in light red flocks, which speedily settling, left the supernatant liquor perfectly bright and clear. I have made some further observations which may explain the presence of basic salt in the ferric hydrate prepared in presence of ammonium chloride, and on the difficulty of preparing ferric hydrate free from basic salt.

Some ferric hydrate washed, dried at 100°, rewashed and redried at 100°, lost on ignition

1

gram of this hydrated ferric oxide boiled with about 100 c.c. of pure water in a retort, gave a distillate quite neutral to litmus.

5 grams NH.Cl in 100 c.c. pure water on boiling gave a weak acid distillate. On pouring the ammonium chloride solution into the retort containing the hydrated ferric oxide, and distilling to dryness, free ammonia equal to 0·005 gram was obtained. On repeating this experiment many times, I always obtained an alkaline distillate containing an appreciable quantity of ammonia, the action

FeCl + 6NH, + 6H,O = Fe,H.O. + 6NH CH

being apparently reversed—

Fe2HO+6NH,Cl = FeCl + 6NH3 + 6H2O.

Having prepared numerous samples of hydrated ferric oxide by means of Fe2Cl, and NH3, I have noticed that they all possess different

shades of colour, no two samples having exactly the same shade of red. They range in colour from dirty brown to bright brick-red. D. Tommasi (Bull. Soc. Chim. [2], 38, 152-153) divides ferric hydrate into two classes: red, obtained by precipitating a ferric salt with alkalis, and yellow, by oxidation of ferrous hydrate, ferrosoferric hydrate, or ferrous carbonate; but (Bull. Soc. Chim. [2], 37, 196— 197) he has found that ferric hydrate kept under water for a year is converted to the extent of 0.3 per cent. into a soluble modification identical with Graham's "colloïdal hydrate.”

I am inclined to think that his ferric hydrate was slightly impure, and that the colour differences and the formation of a small quantity of "colloïdal hydrate" are due to contained basic salt.

The object I had in view, viz., the study of the reaction between ferric hydrate and hydrogen sulphide was much hampered by the impurity of the ferric hydrate, and also by want of uniformity between the various samples I had prepared.

Recently precipitated ferric hydrate in the gelatinous state suspended in water, on being saturated with hydrogen sulphide, becomes black, and is completely soluble in excess of potassium cyanide, with formation of potassium ferrocyanide and sulphide: FeS + 6KCN = K2S + K,Fe(CCN), but ferric hydrate washed and dried at 100° does not behave in quite the same manner, being more or less converted into a form of ferric oxide inactive with H.S. Tommasi found that ferric hydrate kept under water for a year became converted to the extent of 30 per cent., into a modification insoluble in dilute acids. Some portions of the filtrates before spoken of containing "colloïdal hydrate" remained quite unchanged in colour when saturated with hydrogen sulphide. Others retained a bright red colour for ten minutes after the solution was saturated with hydrogen sulphide, and then suddenly flashed off black. On standing a few moments, black flocks collected, and settling left the solution bright and colourless, or slightly yellow. On boiling with potassium cyanide, the black flocks were dissolved; but particles of purple-red ferric oxide were left undissolved. This was a form inactive to hydrogen sulphides. Some portions of the filtrates containing "colloïdal hydrate," which deposited black flocks on treatment with hydrogen sulphide were filtered. Sometimes the filtrate was quite colourless, and after standing became opalescent from decomposing hydrogen sulphide. Others passed through of a slight yellow colour, and on boiling deposited black flocks of ferrous sulphide, and on filtration gave colourless filtrates, becoming opalescent on standing. On boiling, this opalescent water became quite clear, the free sulphur being removed, probably thus:—

H2O + S = H2S + 0.

Cross (Chem. Soc. J., 1879, 1, 250) has noticed an action similar to this, to which I shall have again to refer.

These previously opalescent filtrates now contained iron in solution as ferrous sulphate,

which was estimated with potassium dichromate after expulsion of all hydrogen sulphide.

As a general rule, on boiling the iron sulphide suspended in water (produced by treating ferric hydrate suspended in water with excess of hydrogen sulphide), it is found impossible to finally expel all hydrogen sulphide from solution whilst any iron sulphide (ferrous sulphide ?) remains, for hydrogen sulphide is continually being generated and ferrous sulphate formed.

Supposing the ferric hydrate to react with hydrogen sulphide in the following manner :—

Fe2HO + 3H2S = 2FeS + S + 6H2O,

some of the free sulphur decomposes water in the manner noticed by Cross (loc. cit.),

4S + 4H2O = 3H2S + H2SO..

the sulphuric acid reacting with the ferrous sulphide, thus:

H2SO. + FeS FeSO, + H2S.

There are two equations in use in text-books for the purpose of explaining the reaction between hydrated ferric oxide and hydrogen sulphide, a reaction largely made use of in gasworks for removing the latter from crude coal-gas. They are:

(A.) Fe2Oз,H2O + 3H2S = Fe2S3 + 4H2O.

3,

(B.) Fe,O,H2O + 3H2S= 2FeS +S+ 4H2O.

If either one of these two reactions took place alone, it would be easy to decide which of the two did actually occur: for if the reaction proceeded according to equation (A) no free sulphur would be found; if, according to the equation (B), then free sulphur equal to one-third of the total sulphur entering into action would be found in the free state.

In my experiments I have always found free sulphur as one of the products of the reaction. By treating FeO3, H2O, suspended in bisulphide of carbon with gaseous hydrogen sulphide, and filtering the clear bisulphide after the black iron sulphide has settled, I have

VOL. XLIII.

M

obtained very appreciable quantities of sulphur. I mixed Fe20,3H,0 with powdered quartz for the purpose of moderating the reaction, and placed it in tubes. From the tubes the air was expelled by a current of hydrogen, and then hydrogen sulphide was passed through. When the ferric oxide appeared saturated, hydrogen was again passed through the tubes to expel the excess of hydrogen sulphide; and then without contact with air the material was washed with pure carbon bisulphide. When the washing was complete, the material was turned out of the tubes and exposed to the air until the oxidation was complete. The material was again washed with carbon bisulphide, the washings were evaporated, and the free sulphur was weighed. In this manner I obtained the following numbers :

The sulphur in the free state before oxidation does not appear to bear any relation to the sulphur produced by the oxidation of the iron sulphide or sulphides.

I have noticed also, in the case of the hydrated ferric oxide used in coal-gas purification, that the free sulphur produced in the reaction between Fe2O, and H2S is about one-fifteenth to one-thirtieth of that formed on the oxidation of the "fouled oxide."

It might be supposed that all the free sulphur formed in the reaction between Fe2O, and H2S was not soluble in carbon bisulphide. This I do not think probable, for then it would be necessary to assume that the insoluble sulphur became converted into the soluble variety during the oxidation of the sulphide; otherwise the oxide material used in gasworks would accumulate large quantities of insoluble sulphur, which is not the case in my experience.

For the purpose of further studying the nature of the reaction, I carefully estimated the water formed. This, of course, does not affect the question as to which reaction, (A) or (B), takes place, or whether both occur, for the quantity of water formed in each case is the same. The ferric oxide used lost on ignition 10:15 per cent., 10.09 per cent., 10-13 per cent., 10:10 per cent., and therefore closely approximated to the hydrate Fe2O3, H2O. It had been finely powdered, dried at 100°, and was kept in a desiccator over concentrated sulphuric acid. A U-tube fitted with glass stopcocks was weighed full of dry hydrogen,