Estimation of Cobalt

Cobalt admits of estimation in a variety of ways, the most important of which, however, are:

The main difficulty is to obtain the cobalt in the form of one of its compounds entirely free from nickel or other metal. This may be accomplished by the cyanide method described under "Wet Tests". The solution, after separation of the nickel, is evaporated with dilute sulphuric acid until white fumes are evolved. On addition of water a solution of cobalt sulphate results, and the cobalt may be precipitated in a variety of ways, for example: as oxide with hypobromites or persulphates, as sulphide, as cobalti-nitroso β-naphthol, or as basic carbonate. In each case the precipitate is ignited and reduced to metallic cobalt in a current of pure hydrogen.

The procedure with α-nitroso β-naphthol is as follows: When it is added to a solution of a cobalt salt acidulated with acetic acid it yields a voluminous precipitate of cobalti-nitroso p-naphthol. Any nickel originally present remains in solution, a very complete separation of the metals being thereby effected. Indeed, the reaction is so sensitive that a distinct turbidity is produced with such traces of cobalt as fail to be precipitated by the nitrite process described above. The naphthol solution is somewhat unstable, and should be made up fresh once a month by dissolving 4 grams of the solid in 150 c.c. of cold glacial acetic acid, and subsequently diluting with an equal quantity of water.

The requisite amount of this solution is added hot to one containing the nickel and cobalt salts acidulated with acetic acid, and the whole allowed to stand overnight. On filtering, the precipitate is washed with 13 per cent, hydrochloric acid to remove any nickel carried down mechanically with the cobalt, and after washing with water and drying, the whole is incinerated with a little pure oxalic acid, and the cobalt reduced to the metallic condition in a current of hydrogen. Copper, chromium, and iron are also precipitated by the naphthol, and should not, therefore, be present in the original solution.

α-dimethyl glyoxime may be used for separating small quantities of nickel from cobalt. Excess of the glyoxime is added to a boiling ammoniacal solution of the cobalt salt, when the scarlet nickel α-dimethyl glyoxime is precipitated. α-diphenyl glyoxime (α-benzil dioxime) may be used in a similar manner with corresponding results.

A convenient method of estimating cobalt and nickel in an ore consists in decomposing about one gram of the last named with nitric acid and potassium chlorate and removing the insoluble residue by filtration. The solution is acidified with hydrochloric acid, a little bromine water added, and any iron precipitated as ferric hydroxide by excess of ammonia. The solution is again acidified, and any copper, etc., precipitated with hydrogen sulphide. Ammonium chloride is now added and excess of a 1 per cent, alcoholic solution of dimethylgly oxime to the boiling solution. It is now rendered alkaline with ammonia, allowed to stand for thirty minutes, and the nickel dimethyl glyoxime filtered off in a Gooch crucible, dried at 110° C. for one hour, and weighed. The compound has the formula NiC₈H₁₄N₄O₄, and contains 20.33 per cent, of nickel. The filtrate containing the cobalt in solution is made strongly acid by addition of concentrated sulphuric acid and evaporated till fumes appear. This destroys the organic matter present. On cooling, dilute hydrochloric acid is added and the whole warmed till all solid matter has dissolved. Its temperature is reduced to 50° C., and a warm solution of nitroso β-naphthol in 50 per cent, acetic acid added until precipitation ceases. After standing an hour at 50° C. the precipitate is washed with dilute hydrochloric acid, then with hot water, and finally ignited, being weighed as Co₃O₄.

Volumetric Methods for the estimation of cobalt have been suggested at various times, but owing to the special conditions that have to be observed, their utility is seriously curtailed.

Colorimetric Methods have frequently been suggested, but of these, that originated by Skey and studied by several others appears to be the most useful. It hinges on the fact that potassium thiocyanate yields a blue colour with solutions of cobalt salts, due to the formation of cobalt thiocyanate. On adding alcohol and ether to the liquid, a blue layer is produced. This is destroyed by mercuric chloride, sodium acetate, or sodium thiosulphate, and is masked by the presence of iron salts in consequence of the intense red colour of ferric thiocyanate; consequently these substances should not be present when the colorimetric test is applied.

In carrying out the test a small quantity of standard cobalt sulphate solution is taken, and acidulated with dilute sulphuric acid. Ammonium thiocyanate is then added and the volume of the liquid made up to 10 c.c. A mixture of alcohol and ether is now added until the total bulk is 20 c.c. and the blue colour of the supernatant ethereal layer is compared with that from a solution of the liquid to be tested, and which has been treated in a precisely similar manner. Electrolytic Methods have also been employed.