SEPARATION FROM TITANIUM.

Titanium is precipitated entirely with the uranyl phosphate in a form very closely approximating to TiO2. P2O5. It occurs so sparingly in ores that a colorimetric estimation satisfies most requirements. In larger amounts it can be separated by adding 30 to 50 c.c. acetic acid to the hot solution, and then sufficient ammonium acetate to destroy the free mineral acid. A large excess of ammonium acetate is to be avoided. (See "Separation of Iron.")

Titanium is completely precipitated by boiling its acid solution with an excess of thiosulphate in the presence of much free acetic acid; uranium is not at all precipitated under these conditions, providing phosphoric acid is absent, but quantitative separations on this principle from synthetic mixtures have not been attempted. When phosphates are present uranium and titanium must be precipitated together, ignited, fused with sodium carbonate, extracted with water, and the residue (consisting of all the titanium and most of the uranium) dissolved in acid, and separated as above.

D

SEPARATION FROM ALUMINIUM.

Since the introduction of alumino-thermic reductions, experimental metallurgy often requires rapid assays to be made by such means as are not interfered with by aluminium, and it is therefore to be regretted that one's choice of means of separating this element from uranium are so limited. Of the methods proposed for the separation of iron and uranium only two are applicable to the separation of uranium and aluminium, and these with no great measure of success.

The separation of aluminium as basic acetate is never very satisfactory unless preponderating amounts of iron are also present. Its separation as hydroxide with ammonium carbonate is less successful than usual when applied to a mixture containing uranium. Uranium passes entirely into solution when a hot and slightly acid mixture of the two metals is poured in a thin stream into an agitated solution of ammonium carbonate, but much of the aluminium also goes into solution, and it cannot with certainty be assumed that all this will again separate even if allowed to stand twelve or fourteen hours in a warm place, and a small amount of aluminium left unprecipitated has an important influence on the estimation of the

uranium on account of its comparatively low combining weight. Boiling, of course, cannot be resorted to, for fear of precipitating uranium.

All the aluminium passes into solution when a mixture of the two metals is poured into aqueous caustic soda, but much uranium also goes into solution, and no amount of boiling causes it to separate again. Uranium is precipitated from acetic acid solutions with nitroso-ß-naphthol, and aluminium is not; but persistently reliable results were not obtained with this reagent, which is to be regretted, because the uranyl naphtholate filters and washes much better than most other precipitated compounds of uranium.

Generally in the course of analysis aluminium. and uranium are obtained together as phosphates. In this form the separation with ammonium carbonate is much more successful; in fact, it is entirely successful if the phosphates are dissolved in nitric acid, most of the acid neutralized,' and the clear heated solution poured into the agitated alkaline liquid. If filtered at once a few milligrams of

1 A pale yellow crystalline precipitate forms on adding ammonia, and not an amorphous cloud such as when solutions of either aluminium or uranium phosphate are neutralized. The former precipitate does not redissolve so quickly in hydrochloric acid as the latter, and more care must therefore be taken over the neutralization. I do not know the composition of the lemon-coloured crystalline precipitate.

aluminium phosphate remain in solution, and possibly also traces of uranyl phosphate remain in the precipitate. Standing for an hour or two in a warm place minimizes both these errors.

A modification of this separation, which is sometimes to be preferred, is to pour the dissolved metallic phosphates into an excess of sodium carbonate. The clear or merely opalescent liquid is then heated to boiling, but not boiled, removed from the flame, and ammonium chloride (in amount more than equivalent to the excess of sodium carbonate) added. This solution also, if filtered at once, contains a few milligrams of aluminium phosphate; but it is an easy matter to dissolve the uranyl pyrophosphate subsequently obtained (and weighed) in nitric acid, pour the solution into ammonium carbonate, collect the precipitate, and weigh it. The following results were obtained after separating from a gram of aluminium :

The behaviour of the ignited pyrophosphate

when fused with sodium carbonate is referred to

on page 12. It might be imagined that a separation of practical utility could be based on this reaction if supplemented by a colorimetric estimation of the small fraction of uranium extracted by digesting with water. Experiment, however, shows that the amount of uranium passing into solution is very much greater when the fused pyrophosphate contains aluminium. But this opportunity may be taken to say that small amounts of uranium, either alone or associated with other metals, may often be conveniently estimated by adding an excess of potassium carbonate and half a gram or so of sodium peroxide, and then matching the colour formed by running a standard uranyl solution into an equal volume of distilled water in which are dissolved similar amounts of potassium carbonate and sodium peroxide. A milligram of uranium can be detected in 60 to 100 c.c. of solution by these means, and the indication is more characteristic than that given with potassium ferrocyanide.