The acid solution of magnesium chloride is made use of several times in succession as so much hydrochloric acid, together with a quantity of fresh acid sufficient for the reaction. Finally, when the magnesium chloride has inconveniently accumulated, it is worked up by itself into magnesia and hydrochloric acid.
2. In order to obtain the soda, the sodium binoxalate is brought together with an equivalent quantity of magnesium carbonate and water in a tight cask. As soon as the remaining air has been nearly expelled by the generated carbonic acid gas, the cask is closed, and a stirring mechanism set in motion.
A pressure gauge attached to the cask indicates a gradual rise of the pressure to two atmospheres, but, on continual stirring, this diminishes, until, finally, the gauge stands again at 0°. The cask now contains a concentrated solution of sodium bicarbonate, and a precipitate of magnesium oxalate, which latter, being coarsely granular, is easily separated from the liquid, and is used over again, after washing, for a new operation.
The solution of sodium bicarbonate is boiled for a short time with magnesia, obtained in distilling magnesium chloride, and both are thereby converted into simple carbonates. Both reactions are shown in the following scheme:
| 1. | NaHC2O4 | + | MgCO3 | = | NaHCO3 | + | MgC2O4 |
| Sodium binoxalate. | + | Magnesium carbonate. | = | Sodium bicarbonate. | + | Magnesium oxalate. |
| 2. | 2NaHCO3 | + | MgO | = | Na2CO3 | + | MgCO3 | + | H2O |
| Sodium bicarbonate. | + | Magnesia. | = | Sodium carbonate. | + | Magnesium carbonate. | + | Water |
As the solution of sodium carbonate, after concentration to 40°B., is incapable of dissolving or retaining in solution any sodium oxalate, it follows that the whole of the oxalic acid is recovered. The magnesia which is required for the purpose is obtained by distilling magnesium chloride, which thereby splits up into hydrochloric acid and magnesia. One half of the latter receives, as we have seen, its carbonic acid by boiling with sodium bicarbonate; the other half is placed, whilst still moist, upon trays in great wooden closets, through which the gases of the furnace pass, and is thereby carbonated. The process may also be so modified that the sodium binoxalate is first decomposed by caustic magnesia, and that magnesium carbonate is afterwards added.
The whole mixture is then transferred to a stirring cask, provided with openings for the passage of cooled furnace gases, whereby the caustic soda present is very soon carbonated.
3. As soon as a large quantity of magnesium chloride solution has accumulated, it is tested as follows:—A small sample is mixed, while boiling, with magnesium oxalate, as long as the latter is dissolved, and then allowed to cool. There should be no crystalline deposit of sodium binoxalate formed, a proof that the solution does not contain any sodium chloride in excess, and is fit for distillation. It is first neutralised by adding some more magnesia, and evaporated over a naked fire in large kettles to a doughy consistence, short of driving off the hydrochloric acid. It is then transferred into the ordinary soda furnace, where it is distilled with a moderate fire. The eliminated hydrochloric acid is condensed in the usual manner.
The residuary mass should not be heated red hot, so as not to impair its porosity or its ready affinity for carbonic acid. If, however, the first-mentioned test shows the magnesium chloride to contain sodium chloride the whole mass must be mixed with magnesium oxalate, and after removal of the precipitated sodium oxalate, saturated with magnesia and distilled. The same process, in all its details, may also be employed for the manufacture of potassa and its carbonate.