ALUM.—Pbecipitation.
of about 1-40, which is the most suitable state ofconcentration for producing the flour, or meal, forthe next operation.
To economize fuel the furnace is very long, andthe fire communicates with an evaporating pan, sothat the whole heating power may be made availableas far as practicable. The draught of the furnaceinsures,- in the simplest manner, a constant currentof air over the liquid, which is favourable to evapora-tion by sweeping the vapours generated onwards tothe chimney. Luring the evaporation of the liquorin the furnace a copious sediment is thrown down,consisting of basic ferric sulphate, sulphate of lime,and other salts, of which it is necessary to clear theliquor in the cooling vats ; but however impure theliquor, or however rich in aluminium sulphate thesolution may be, its evaporation should not be carriedon so far as to cause a deposition of salt upon cool-ing; the ingredients must be entirely held in theclear solution, even after cooling the liquor in thevats set apart for the production of the alum flour.
The furnaces are fired incessantly for eight days,and the boiler is replenished with liquor from thesupply cistern till the eve of the eighth day, when acharge of mother liquor is run in which deposits nosedimentary matter on boiling, like the fresh liquor;this is boiled down till it has the proper degree ofstrength, and then allowed to settle for twelve hours.After adding the mother liquor it is customary to findthe strength of the solution. This is done by takingout a known quantity of it, and concentrating it to theproper strength with potassium chloride and allowingthe solution to cool for twenty-four hours; the crys-tals produced, when washed with a little cold water,dried and weighed, furnish an approximate test ofthe value of the liquor as to its real percentage ofalumina. The alum-maker also employs this processto test the quality of his potassium chloride; but inthe latter case he has a standard solution of alumin-ium sulphate. These experiments point out thequantity of alkaline chloride which should be addedper gallon of liquor.
At the Whitby alum-works leaden pans are em-ployed, in consequence of the liquors from the cal-cined shale, of this locality containing sulphate ofmagnesia, which renders surface evaporation inappli-cable, as already explained; the pans are 10 feetlong, 4 feet 9 inches wide, 2 feet 2 inches deep atone end, and 2 feet 8 inches at the other. In theselead pans the liquor is rapidly concentrated by a fireat one end, the flue of which runs along in parallelchannels under a plate or course of brickwork onwhich the pans rest. Every morning the pans areemptied into a settling cistern of stone or lead. Thespecific gravity of the liquor should be about T40 to1-50; this density varies, however, according to theviews of the manufacturer. For a liquor which con-sists of two parts of aluminium sulphate and one offerrous sulphate, a specific gravity of 1-25 may besufficient; but for a solution of one part of alumi-nium sulphate and two parts of sulphate of iron,which latter is to be abstracted by crystallization, adensity of T40 is requisite.
165
Preparation of the Powder .—From the evapor-ating basins the liquor is run into the flour orpowder cisterns, in which the aluminium sulphate iscombined with potassa or ammonia (the “flux” or“precipitating substance”), in such a manner as toeffect the first purification of the alum. The alumis only very slightly soluble in the cold mixture ofthe liquor with the solution of the precipitating sub-stance, and for this reason the greater portion of itseparates in the form of minute crystals, or “flour,”from the solution ; an agitating motion given to theliquor in the cisterns favours the formation of thecrystals. To obtain the largest yield of flour it isnecessary for the precipitating solution to be asstrong as possible, which is secured by dissolving theprecipitant in the least possible quantity of boilingwater. The proper amount of the precipitant shouldbe determined, as near as possible, by a preliminaryexperiment on the small scale; but still these data,though they serve to give an idea of the quantity,do not correspond always with the extensive opera-tions of the manufacturer; so that even when pre-vious experiments are made to ascertain the truequantity to be added, much caution is necessarywhen operating on the whole bulk of mother liquor,as an excess of the precipitant may often cause muchfuture inconvenience and labour. The best preci-pitating substances are potassium and ammoniumsalts, as these form alums -which are very insoluble incold water, only 9*25 parts of the potassium alum,and 9*16 parts of the ammonium alum being dis-solved by 100 parts of water at 50° Fahr. (10° Cl.).Sodium salts are very disadvantageous precipitatingagents in the formation of alum, on account of thegreat solubility of the resulting compound, 4G partsof which are dissolved by 100 of water at 50° Fahr.;they are therefore seldom used.
In the event of potassium salts being employed,preference is given to potassium chloride or sulphate;of the latter 50*9 parts are required to precipitate100 parts of aluminium sulphate, from which results334 parts of crystallized alum. 'The preceding enu-merated tests, however, will not answer in thefactory, because the mother liquor is indefinite asto the amount of impurities which accompany thealuminium sulphate. To show this more clearly, letit be supposed that the mother liquor is saturatedwith ferrous sulphate and aluminium sulphate, andthat its temperature is 50° Fahr. In this case it willcontain, for every
100 - parts of water,
33-2 parts of anhydrous forrous tulpliato, and
33*5 parts of anhydrous aluminium sulphate.
To precipitate these products 17*1 parts of neutralpotassium sulphate are required, for the solution ofwhich 06 parts of boiling water will be necessary.
On mixing this solution with the lie, 96 parts ofalum will be produced, together with the accumulatedwater of the lie and precipitant, which, collectively,will amount to 166 parts. Of the 96 parts of alumformed, only 16 can remain in solution after theliquor has cooled to the primary temperature of 50°Fahr. (10° C.) ; therefore 80 parts will separate in