QUANTITATIVE ANALYSIS OF INORGANIC SUBSTANCES.
Quantitative analysis is far more difficult than mere qualitative analysis, and cannot be treated adequately in a general encyclopaedia. As has been said,, unless the nature of the sub stance is known in some other manner a pre liminary qualitative analysis must be made. Several usual methods of making quantitative analyses will be described below.
Electrolytic Method.— When the substance to be analyzed is an alloy, or a simple mixture of metallic salts, its metallic components may often be readily separated by the electrolysis of its solution, the separation of the metals being based upon the known fact that in electrolysis the nature of the deposit depends largely upon the nature and degree of concentration of the solution, the sizes of the electrodes and the strength of the electric current that is employed. By systematic study of the effect of these Con ditions in the electrolysis of mixtures of metal lic salts, it is found to be possible to deposit one metal upon the cathode, while the others remain in solution. The electrolytic method has been developed to a considerable extent, and promised to be of great value. Thus far, how ever, it is not in extensive general use. For details concerning it, consult Classen, 'Quanti tative Analysis by Electrolysis,) and various papers published by Prof. Edgar F. Smith, a few years ago, in the Journal of the Franklin Institute. Scott's 'Standard Methods of Chemical Analysis) contains the latest details concerning approved electrolytic methods. See also the article ELECTROLYSIS.
Gravimetric Methods.--Strictly, any method of analysis in which the quantity of each con stituent is determined by weighing is a gravi metric method; but the term is usually under stood to exclude the electrolytic method just mentioned. In gravimetric work the components that are to be weighed may be separated by fire-methods, or by selective precipitation from colution_ as in the scheme of nualitative 2nalv ids outlined above. The fire-methods are com monly used in the estimation of gold and silver and are described in the article ASSAYING. The wet methods do not differ in general theory from the method give.' above for qualitative an alysis; for it is evident that any precipitate which contains only one base may be isolated and weighed, and that the quantity of the base presentbe calculated from the observed weight of e precipitate, and its known chemi cal formula. But the practical case is by no means as simple as this statement would indi cate, because certain matters of detail, that are not of the slightest importance in qualitative work, must be attended to with ,great care in quantitative analysis. For example the sub stance that is to be determined must be isolated by a method that will ensure its perfect separa tion from every other substance that may be present; that precipitate that is to be weighed must be granular enough• to be filtered easily and without loss; the precipitate must not be liable to oxidation nor to other change upon exposure to air for such time as its manipula tion may demand; it must be of such nature that it can be thoroughly dried; and it must not be hygroscopic enough to absorb sensible quantities of water from the air, from the time it is dried until the weighing has been com pleted. Thus in qualitative analysis aluminum may be recognized by the precipitation of the hydrate; but in quantitative analysis it is neces sary to reduce the metal to the form of the oxide. The hydrate is usually gelatinous when freshly precipitated, and it retains traces of the acid with which the metal was previously com bined, and also traces of the alkali that was used in the precipitation of the hydrate. These facts are of no consequence in qualitative work, but their importance in quantitative investiga tion is evident.
Volumetric Methoda.—In volumetric analy sis the quantities that are to be measured are determined by the measurement of volumes, and weighings are resorted to only in the prepara tion of the standard reagents that are to be used. The reagents are made up in certain standard strengths, according to the purposes for which they are wanted; but the usual strengths are those designated as cnormall and "decinormalo solutions. A "'normal* solution is one having such a strength that one litre of the solution contains as many grams of the reagent as there are units in the reagent's chemical equivalent. Thus the chemical equivalent of sodium hydrate, NaOH, is (in round numbers) 23 + 16 ? 1==.40; and hence a anormalp solu tion of sodium hydrate is one which contains 40 grams of that substance to the litre. A odecitionmal* solution of this reagent contains 4 grams of it per litre, and a kentinormal* solution contains 0.4 gram per litre. If the reagent is a bivalent acid, or a salt of a bivalent base, the number of grams of it present in each litre of solution must be equal to half the molecular weight. Thus sulphuric acid, contains two atoms of replaceable hydrogen, and its molecular weight is 2 + 32 + 64=98. Hence a normal solution of sulphuric acid con tains 49 grams of the anhydrous acid, per litre of solution. The general idea is to have all of
the reagents of such strength that if one cubic centimeter of any (normal) acid solution be added to one cubic centimeter of any (normal) alkaline solution, the mixture will be precisely neutral. As an illustration of the method that is followed, let it be assumed that a manufac turer buys a carboy of oil of vitriol, and wishes to know what proportion of pure sulphuric acid it contains. If the acid were quite free from water, 49 grams of it would be exactly neu tralized by 1,000 cubic centimeters (that is, one litre) of any normal alkali solution. It is more convenient to work with one-tenth of this quantity of acid and reagent; so that the ex periment will consist in
out 4.9 grams of the oil of vitriol, diluting it with water, adding a piece of litmus paper, and letting a normal alkali solution pass into it (pref erably from a graduated burette) until the acid is precisely neutralized. If 100 cubic centi meters of the alkali were required to effect the neutralization, the given sample of oil of vitriol would be known to contain 100 per cent of its weight of sulphuric acid,— or to be en tirely free from water. On the other hand, if only 53.9 cubic centimeters of the normal alkali solution were required to effect neutraliza tion, the sample would be known to contain 53.9 per cent of its own weight of sulphuric acid. As a further example, suppose it is de sired to ascertain the percentage of sodium oxide present in a given sample of crude soda ash, without raising the question as to whether the sodium actually occurs as oxide, hydrate or carbonate. The formula of sodium oxide is Na,O, and its molecular weight is 46 -I-
As it contains two atoms of sodium to the molecule, we weigh out 3.1 grams (not 6.2 grams) of it, dissolve in water and dilute and add litmus paper as before. Then into the solution we pass a normal acid solution until neutralization is effected. If 46.7 cubic centi meters of the normal acid solution are re quired, the alkali present in the sample, when computed as sodium oxide, constitutes 46.7 per cent of the weight of the whole. This process is called 'Taciditnetry° when it is used for esti mating the strength of acids, and aallcalimetryp when it is used in estimating the strengths of alkalis. As a further illustration of volumetric methods the estimation of chlorine (known as uchlorimetryD) may be considered. If the sub stance to be examined for chlorine is bleaching powder, 10 grams of the powder are dissolved i by rubbing with water in a mortar, and the solution is diluted till it occupies a litre. It is then well shaken, and 100 cubic centimeters are drawn off into a beaker, by means of a pipette, and treated with a decinormal solution of arsenious acid (As,O,) until a drop of the mixture, when withdrawn by a glass rod, gives no blue stain upon filter paper that has been soaked in starch liquor and iodide of potassium. The number of cubic centimeters of decinormal arsenious acid solution required is to be multi plied by the constant multiplier 0.00355, and the product is the weight of available chlorine in grams, contained in each gram of the original powder. (For explanation of the multiplier 0.00355, and for full details of this process and of volumetric analysis generally, consult Fran cis Sutton,
The analysis of gases is of so special a character that it is treated under a separate heading. See GASOMETRIC ANALYSIS.
In the analysis of organic compounds, no general scheme can be given, corresponding to that which is used in the systematic examina tion of inorganic substances. The number of possible organic compounds is so great, that practically nothing can be done in the way of effecting a analysis of a compound concerning whose general nature we have no preliminary information. For the more or less general methods that have been developed for the examination of special classes of organic substances, advanced books on organic analysis must be consulted. The ultimate analysis of an organic substance consisting of oxygen, hy drogen and carbon may be effected by burning the substance in a glass tube in a current of oxygen gas. The carbon is converted into car bon dioxide, which is absorbed by potash and estimated quantitatively by observing the gain in weight of the potash; and the hydrogen is converted into water, which is similarly esti mated by absorption of calcium chloride. The oxygen of the original compound is then esti mated by difference. When nitrogen is also present, the process is somewhat more com plicated. In this case the gases of combustion may be passed over red-hot metallic copper to absorb the oxygen, and the nitrogen may be measured in the free state, the oxygen being finally concluded by difference, as before.
Consult (in addition to the works mentioned above), Fresenius, 'Manual of Qualitative Chemical Analysis) and 'System of Instruction in Quantitative Chemical Analysis); Lunge, 'Technical Methods of Chemical Analysis); Thorpe,