Beryllium Production

Vauquelin originally obtained beryllia by fusing beryl with potassium hydroxide, dissolving the melt in water and boiling the solution to reprecipitate the beryllia. Since the adequate isolation of beryllia from beryl involves the troublesome separation of the two very similar oxides, beryllia and alumina, a large number of methods have been devised and used to effect it. The best methods can be classified under the two main headings of (A) alkali fusion and (B) fluoride fusion. Beryllia is almost invariably prepared from beryl, though other minerals containing beryllium have been occasionally employed.

(A) In the alkali-fusion method the finely powdered beryl is fused with about twice its weight of potassium carbonate or potassium hydroxide. The melt is ground, covered with water, and treated with excess of sulphuric acid. The whole is then heated, with stirring, until white fumes of sulphuric acid are evolved and the residue is reduced to a fine white powder. This decomposes the silico-aluminate of beryllium and potassium and makes the silica insoluble. After dissolving out the sulphates of beryllium, aluminium, and potassium with water, it has been usual to crystallise out the aluminium as potash alum, after suitable concentration. The mother-liquor from the separated alum contains the beryllium as sulphate, and also potassium sulphate, sulphuric acid, some aluminium sulphate, and small amounts of impurities, including slight amounts of iron. Britton advises, after filtering off the silica, adjusting the concentration of the sulphuric acid to about 5N by adding potassium hydroxide, saturating the boiling solution with potassium sulphate, and crystallising at a low temperature, preferably at 0° C. If alcohol is used to procure a complete separation of the aluminium as alum, some beryllium sulphate also crystallises out. Whether a preliminary separation of aluminium through potash alum is made or not, a final isolation of the beryllia is necessary.

(B) Lebeau fused powdered emerald with twice its weight of fluorspar, poured the molten mass into water and heated the resulting friable mass with sulphuric acid. After volatilisation of the silicon fluoride the bulk of the aluminium was removed as potash alum. He also heated emerald in the electric furnace, which volatilised some of the silica and left a mass easily attacked by hydrofluoric and sulphuric acids.

A solution of beryllium fluoride is obtained by heating beryl to dull redness with 6 parts of ammonium hydrogen fluoride and extracting with water. The solution also contains traces of aluminium and other impurities - including iron.

Copaux heats 1 part of finely powdered beryl with 2 parts of sodium silicofluoride to 850° C. The beryllia is converted into a double fluoride, thus

2BeO+SiF₄ = SiO₂+2BeF₂;

2NaF+BeF₂ = Na₂BeF₄.

The alumina is analogously converted into artificial cryolite, Na₃AlF₆. Boiling water extracts the soluble double fluoride of beryllium from the residual mass, and leaves the bulk of the insoluble fluoride of aluminium.

The final isolation of pure beryllia from the crude products of these and other methods of extraction from beryl is difficult, and many modern improvements have been made on the older methods, which were often very elaborate.

In a much used method the solution containing the beryllium salt and associated impurities is poured into a large quantity of ammonium carbonate solution. Aluminium hydroxide is precipitated. Most of the iron separates on warming, and the rest is thrown down as sulphide by adding ammonium sulphide. On boiling the filtered solution the beryllium is precipitated as basic carbonate. This precipitate is easily washed, and very suitable for conversion into other salts of the metal. In this method appreciable quantities of beryllium hydroxide are carried down by the aluminium hydroxide, and the beryllium precipitate is contaminated with aluminium hydroxide.

The original method employed by Vauquelin, precipitation of beryllium hydroxide from its solution in potassium hydroxide (or sodium hydroxide) by boiling, has been a constant subject of controversy. Several workers have obtained good results by this method, but others concluded that it was quite unreliable. Britton decided that this method affords a satisfactory separation of beryllium hydroxide if the following conditions are observed -

1. The solution, which should not contain more than about 0.3 grm. BeO and 0.4 grm. Al₂O₃, should be evaporated to 25 c.c., brought to room temperature, and the precipitated hydroxides just redissolved by adding 6N NaOH. The beryllium hydroxide subsequently deposits more slowly and incompletely if excess of the alkali is added.
2. The solution should then be diluted to 500 c.c. and boiled for forty minutes. Longer boiling may precipitate some aluminium hydroxide.
3. The precipitate should then be filtered off at once to avoid resolution of any beryllium hydroxide and decomposition of the sodium aluminate.

Other methods are -

1. The hydroxides of beryllium and aluminium, precipitated by ammonia, are suspended in a concentrated solution of ammonium chloride. If the solution is boiled till ammonia is evolved, the beryllium hydroxide is completely dissolved, while the aluminium hydroxide is not attacked. This method is tedious, and even with careful attention to conditions is probably not quantitative.
2. The hydroxides of beryllium and aluminium are dissolved in sulphurous acid or ammonium sulphite and the solution boiled till the sulphur dioxide has been expelled. Aluminium hydroxide is precipitated and the beryllium remains in solution. All the alumina is precipitated, but it adsorbs some beryllia.
3. Sodium carbonate precipitates nearly all the alumina from a solution containing aluminium and beryllium salts, and the beryllium remains in solution.
4. Pollok, after fusing beryl with sodium hydroxide and separating the silica, precipitated the aluminium as chloride by saturating the solution with hydrogen chloride. Havens originally dissolved the co- precipitated hydroxides of aluminium and beryllium in hydrochloric acid and concentrated to about 10 c.c. The solution, which did not contain more than 0.1 grm. of the respective oxides, was mixed with an equal volume of ether and saturated with hydrogen chloride. After filtering off the aluminium chloride the beryllium can be precipitated from the filtrate by ammonia. The separation from aluminium is complete.
5. The methylamines and ethylamines precipitate beryllium hydroxide and retain aluminium hydroxide in solution. Precipitation can be effected from either nitric acid or hydrochloric acid solution of the two earths, but the beryllium hydroxide precipitated is very gelatinous, and probably adsorbs much alumina.
6. Beryllium can also be separated through its basic acetate. The basic acetate is dissolved away from other elements, including iron and aluminium, by hot glacial acetic acid or chloroform. According to one method, beryllium-containing minerals are heated with acetic acid, the beryllium converted, by a second treatment with acetic acid, into a form soluble in chloroform and extracted with that solvent.
7. Parsons and Barnes treated the hydrochloric acid solution of the beryllium hydroxide with sodium bicarbonate. This precipitates aluminium hydroxide and retains beryllium in solution, as described under the estimation of beryllium.

The two preliminary treatments, (A) and (B), have been variously combined with the different methods of final isolation. Britton obtains from beryl a solution of beryllium sulphate, as described in (A), throws down "crystalline " beryllium hydroxide by treating the solution with sodium hydroxide and boiling, as previously described, and dissolves it in nitric acid. The solution is boiled to oxidise any iron, cooled, neutralised, diluted, and saturated with sodium bicarbonate. On boiling for half a minute, with vigorous stirring, any ferric or aluminium hydroxide precipitates completely, and pure beryllium hydroxide can be separated from the filtrate by ammonia. The sodium hydroxide process may be omitted, but care is then needed to prevent the aluminium hydroxide from carrying down some hydroxide of beryllium.

Bussy and Wohler, working independently, first isolated the metal in an impure state by decomposing the anhydrous chloride with potassium.

Sodium may be used instead of potassium, and the simple chloride may be replaced by the double fluoride of potassium and beryllium, 2KF.BeF₂. About 1883 it was discovered that metallic beryllium can be obtained by electrolysing this double fluoride, though the halides of beryllium do not conduct electricity in the pure state.

Subsequently, a fused mixture of beryllium chloride and sodium or ammonium chloride was electrolysed, and Lebeau electrolysed the double fluoride of sodium and beryllium in a nickel crucible. The anode was a rod of graphite, and the crucible itself acted as cathode. Beryllium minerals can be electrolysed in the presence of fluorine or one of its compounds halogen compounds of the alkalies or alkaline earths may be added.

The electrolytic method is now generally used; it consists essentially of electrolysing beryllium compounds in the presence of fluorides, and its development is largely due to Lebeau.

Attempts to reduce beryllium oxide with magnesium have not been very successful. The converse reaction

Be+MgO = BeO+Mg

occurs more readily.

The metal can be purified by volatilisation in hydrogen. Oesterheld, in 1916, claimed to have secured metallic beryllium of 99.5 per cent, purity.