832
cyanide, sulphocyanide, ferrocyanide. 6. Formationof insoluble sodium compounds, partly as ferro-sodium sulphide, and partly as composite silicates incombination with lime, magnesia, alumina, &c. 7.
Incomplete lixiviation. 8. Mechanical losses byleakages of liquors and scattering of salt and ashduring the operations. The total loss from all thesecauses amounts to 14 to 20 per cent, of the total.It is considered very good work if a yield of 75parts of soda ash, calculated as 48 per cent., from100 parts of sulphate, is obtained; this correspondsto 69'23 per cent, of 52 per cent, ash, or to 60T5per cent, of pure Da 2 C0 3 on 96 per cent, sulphate =62'66 on pure Da 2 S0 3 . The yield in German worksis stated = 64 per cent, pure carbonate from 100 puresulphate, instead of the theoretical 74'6—probably amaximum figure. The loss under the head Do. 6(by the formation of insoluble compounds) hasbeen carefully investigated by Sciieoker-Kestner,who has shown that it increases with the quantity oflimestone or chalk used in the ball-mixture, and thisis a strong reason why the proportion of chalk shouldbe kept as low as consistent with a complete decom-position of the sulphate. The loss from this sourcealone may amount to 5 per cent, of the availablesoda.
Defined Alkali. —The ordinary soda ash (yellowcarbonate) is sufficient for most purposes to whichsodium carbonate is applied in the arts. In mostcases it has to be dissolved in water before use,by which process its insoluble impurities at any rateare removed; but this is not the case where theash is used without dissolving, as in glassmakingand in a few other cases. There is, therefore, aconstant demand for a purer article than the ordinaryyellow carbonate, and a certain portion of the latteris consequently further refined. The usual refiningprocess removes the insoluble impurities, as well asall sulphur compounds below sulphate, and all iron;but it does not diminish the proportion of chlorideand sulphate in the ash. Where these are injurious(as in flint-glass making), a different article must beused, such as fished ash, or calcined crystals, or ashmade by the ammonia process. The ordinary re-fining process for 50 or 52 per cent, alkali iscarried out as follows:—Ordinary soda ash (un-ground) is dissolved in water heated by steam, gener-ally in a cylindrical vessel 8 to 10 feet in diameter,and about 6 feet high, provided with a central revolv-ing shaft with arms for agitating the liquid (see Figs.29, 30. AVheu the ash has been very well carbonated,nothing further is needed; but in the contrary case asmall quantity of bleaching powder is put in thedissolver before running it off, and the boiling iscontinued for a quarter of an hour. The colour ofthe solution, which should be perfectly bright andas clear as water, generally decides upon that point.In no case ought more than 6 or 8 lbs. of bleachingpowder per ton of ash be used. The bleachingpowder (chloride of lime) thoroughly destroys anyremains of organic substance (ulmates); it furtheroxidises the sulphite and the sulphide (if any) tosulphate, and the iron protosalts to persalts, from
which the iron is at once precipitated as ferric oxide,and carried down with the calcium carbonate formedby the action of the chloride of lime on the sodiumcarbonate. The clear solution now contains nothingbut sodium carbonate, chloride, and sulphate andhydrate; the silicate and aluminate generally also gounchanged through the operation. As the chlorideof lime destroys a corresponding portion of sodiumcarbonate, it is evident that no more of it must beused than is absolutely necessary for the purpose inquestion. Whenever it is employed the resultingrefined alkali, in spite of its containing an insolubleresidue, shows a slightly lower alkalimetrical test
Fig. 29.
Fig. 30.
than the soda ash from which it has been made.The solution is now run off, along with all precipi-tates, into settling tanks; very much indeed dependsupon the settling being done properly and com-pletely, and it is best not to trust to one settlingtank, but to employ two or even three. It is a goodplan to make them very long, and to let the solutionpass in at one end and out at the other. At thesame time the liquor must not be allowed to remainso long in the settlers that it can crystallize; theoriginal solution should not be made stronger than56° Tw. (— 1-28 specific gravity), to provide againstthis contingency. Fortunately, as is well known,
J