acid to dryness and dehydrating same at any convenient temperature between 105° and 120° to convert it to SiO2 H2O which is soluble to a slight extent only. The dried residue is then digested with a few c.c. of 12 M HCl for a few minutes to dissolve the accompanying soluble salts, after which a little water is added and the dehydrated silicic acid filtered off. This dehydrated silicic acid will be contaminated more or less by salts of the divalent and trivalent elements of the iron group if these latter are present, in which case it must be further treated as described in § 195. The filtrate will contain about 2-3% of the silicic acid in soluble form, and if it is necessary to separate this also, the filtrate is evaporated to dryness and the residue dehydrated as before.

The forms in which the element silicon may occur are as follows: as free silicon or in combination as a silicide or a silicate. In the first two cases it must first be oxidized to silicic acid by means of nitric acid; in the case of silicates we encounter two kinds of compounds,-those soluble in mineral acid and those insoluble. The latter must be rendered soluble by fusion with sodium carbonate. Thus to take the case of the insoluble silicate rock, feldspar, we have as a typical illustration:

4 KAISI308 + 12 Na2CO3 →K2CO3 + 2 KAIO2 + 12 Na2SiO3 + 11 CO2

All soluble silicates upon treatment with mineral acid decompose according to the scheme

Na2SiO3 + HCl → 2 NaCl + H2SiO3

but, generally speaking, we do not get meta-silicic acid alone as represented in the reaction, but a mixture of ortho, meta, tri and di-silicic acids in varying proportions according to conditions of concentration, acidity, temperature, etc. These various forms of silicic acid have a descending order of solubilities in the succession given, namely,

ortho-silicic acid H4SiO4 or SiO2-2 H2O, soluble

Upon evaporation to dryness and dehydration at 105°-120° the ortho, meta, and tri-silicic acids are converted to the di-silicic acid which is soluble only to the extent of about 2%, that is to say, if the silicic acid being dehydrated contained the equivalent of 0.100 g. SiO2, only about 2 mg. (figured as SiO2) would go back into solution when the soluble salts accompanying it were dissolved. If it is necessary to get rid of the soluble portion which goes back into solution, then the dehydrated silicic acid must be filtered off, and the filtrate containing the soluble silicic acid again evaporated to dryness and dehydrated; the newly dehydrated portion, however, like the first portion, will be soluble to the extent of 2-3%. For precise work when dealing with amounts of silica much over 30 mg. two or three dehydrations, always with intervening filtration, will be necessary according to the quantity of silica present.

184. Determination of Sulphur. The method that is most generally used for the determination of sulphur is the gravimetric method, which is based upon the procedure of getting it in the form of sulphate ion and then precipitating the latter as barium sulphate by means of a solution of barium chloride, according to the reaction

SO+Ba++ → BaSO4 ↓

The reason for the wide use of this method is to be found in the fact that sulphur occurs so often in the form of sulphate, and that when it does occur in any of its other forms, namely, as free sulphur, sulphide, sulphite, etc., these are readily oxidizable to sulphate. In the precipitation of sulphate ion as barium sulphate it is to be noted that carbonate ion, phosphate ion or chromate ion must be absent or their concentration sufficiently diminished by regulation of the acidity of the solution so that they will not be precipitated (§ 160). It is the practice therefore when precipitating barium sulphate to have the solution slightly acid with hydrochloric acid; namely, the solution should be about 0.1 molar with respect to HCl. Barium sulphate is one of the most insoluble precipitates, its solubility being about 2 mg. per liter at 20°. It, however, exhibits to a remarkable degree the phenomenon of co-precipitation as already mentioned

in § 152, dragging down barium chloride, sodium and potassium nitrates, ferric sulphate and many other soluble salts so that it is really a difficult matter to obtain a pure precipitate of barium sulphate. Barium sulphate is fairly stable on ignition and can be heated to 1000° without decomposition, provided reducing substances are not present. In the presence of carbon, barium sulphate is reduced to the sulphide at temperatures around 500° to 600° according to the reaction

BaSO4 + 2 C→ BaS+ 2 CO2

For this reason the use of filter paper in the filtration of barium sulphate is objectionable because, upon the subsequent ignition, the carbon which results from the charring of the paper reduces some of the sulphate to sulphide, and this latter must be reconverted to the sulphate. To accomplish this reconversion it is often erroneously stated that treatment with a few drops of concentrated nitric acid will effect the oxidation of the sulphide to the sulphate according to the scheme

3 BaS+ 8 HNO3 →3 BaSO4 + 8 NO + 4 H2O

In point of fact, however, treatment with nitric acid transposes the sulphide to nitrate and this latter upon ignition at temperatures of 500°-800° goes over to a mixture of the oxide and peroxide, thus,

BaS+ 2 HNO3 → Ba(NO3)2 + H2S

4 Ba(NO3)2 + ▲ → 2 BaO2 + 2 BaO + 8 NO2 + O2

The remedy for taking care of any sulphide that may be formed is to treat the precipitate with one or two drops of 18 molar sulphuric acid and then get rid of the excess sulphuric acid by heating to 300° in an air bath (§ 40)

BaS+H2SO4 → BaSO4 + H2S

The following duplicate determinations, made in the author's laboratory, show the mistake of trying to use nitric acid to convert the sulphide to sulphate

Weight of BaSO4 precipitate after burning off filter paper and igniting to constant weight

B

A

0.3426 g.

0.3433 g.

Summary of Conditions. The amount of sample to be weighed out should not contain an amount of sulphate greater than corresponds to about 0.500 g. BaSO4; the volume of solution should be about 300-350 c.c. and should be acidulated with 2 or 3 c.c. of 3 M HCl per 100 c.c. solution; only salts of the alkali group should be present, and in particular salts of the iron group; nitrates and chlorates must be absent. The temperature of precipitation should be 90°-95°, and the barium chloride solution (0.05 M) should be added dropwise and with constant stirring. The precipitate should be allowed to digest overnight (for amounts of precipitate under 5 mg. this time should be extended to 48 hours) and then filtered through a Gooch crucible, dried at 110° or ignited with the Bunsen burner.

Precision. While concordant checks to a precision of 1 part per 1000 can be obtained for duplicate determinations of sulphates, it is almost certain that these will be attended with a constant error of at least 5 parts per 1000 (cf. § 53) and if salts which are absorbed are present this constant error may reach 30 parts per 1000.

185. Exercise No. 13. Determination of Sulphate in a Mixture of Soluble Chlorides and Sulphates.1. Weigh out an amount of sample that will furnish about 0.500 g. BaSO4 and transfer it to a beaker, dissolve in 300 c.c. water, add 6 c.c. 3 M HCI, and heat to near the boiling point. Add dropwise from a pipette a solution of 0.05 M BaCl2 until in slight excess. Cover the beaker with a watch-glass and set aside for overnight so that the precipitate can digest and in the meantime prepare a Gooch crucible. After digestion and before filtration of the precipitate add to the clear solution a few drops of the BaCl2 solution, and if more precipitate forms continue the addition of the barium chloride until barium ion is in excess. Without disturbing the

1 Instead of such a mixture, a sample of magnesium sulphate, MgSO4-7 H2O, or a standard-ed solution of this salt, may be used.

precipitate, decant the clear solution through the Gooch crucible. Add about 50 c.c. of hot water to the beaker containing the precipitate, stir, allow to settle and decant through the filter as before. Repeat once more. Now transfer the precipitate to the filter by means of a jet of hot water from a wash bottle, using a "policeman" to detach that portion of the precipitate which adheres to the walls of the beaker. Wash the precipitate on the filter with successive portions of hot water until the last washing shows no test for chloride. After the precipitate has been sucked dry at the filter pump for a few minutes, transfer the crucible to an oven and dry at 110° or ignite gently over a Bunsen flame to constant weight. From the weight of the BaSO4 thus obtained calculate the percentage of sulphate in the original sample.

186. Exercise No. 14. Determination of Sulphur in Pig Iron, Steel, Washed Metal or Muck Bar. In these products the percentages of sulphur that may be expected usually lie within the following limits:

As shown by A. Blair 2 the sulphur present in different kinds of iron and steel, particularly pig iron, may exist in four different forms:

1. That evolved as hydrogen sulphide on treatment with hydrochloric acid; the major portion is in this form.

2. That evolved as an organic sulphide (CH3)2S, etc., which is very stable and not oxidized by aqua regia, bromine, or ammoniacal hydrogen peroxide; the amount of this is very small and probably lost by any method in use.

3. That which is not attacked by hydrochloric acid but is oxidized by concentrated nitric acid or aqua regia.

4. That which is unacted on by nitric acid or aqua regia and is only obtainable in solution after fusion with sodium carbonate and potassium nitrate; the amount of this is very small.

There are several methods for obtaining the solution of the

2 J. A. C. S. 19, 114 (1897).