As the tannins are uncrystallisable, and incapable of being distilled without decomposition, they are exceedingly difficult to obtain in a state of purity, and, owing to the considerable differences in their character, no one method is equally applicable to the whole group. As their successful separation requires considerable chemical training, and experience, detailed description is outside the scope of the present work, but some particulars of the more important methods employed are given on p. 43, L.I.L.B.
Their chemical constitution is complex and in most cases imperfectly understood, but all the natural tannins which have been investigated prove to be derivatives of the trihydric phenol, pyrogallol, or of the dihydric phenol, catechol, the latter of which is often accompanied by a trihydric phenol, phloroglucol, which is isomeric with pyrogallol. The phenols are, themselves, a class of derivatives of benzene, C6H6, in which one or more of the hydrogen atoms are replaced by OH groups. Common phenol or “carbolic acid” is their simplest representative. Many of them, including pyrogallol and catechol, are used as photographic “developers.” The phenols on replacing another hydrogen by carboxyl (CO.OH) form true acids, of which salicylic corresponds to common phenol, protocatechuic to catechol, and gallic to pyrogallol; and the tannins are apparently complicated acids, in which one of the two latter acids is linked to a second molecule of the same or another acid as an anhydride, in some cases possibly with the addition of phenols or other organic groups. For more detailed information, see L.I.L.B., p. 45. Gallotannic acid is apparently digallic acid, in which two molecules of gallic acid are linked together after giving up the elements of a molecule of water. Natural gallotannic acid and many other tannins are glucosides, or at least contain glucose, which in many cases can be removed by purification.
From what has just been said, it is obvious that a classification of the tannins according to constitution, is at present impracticable, not only from our imperfect knowledge, but from the difficulty of separating and determining the products of their decomposition. It is not, however, difficult to distinguish the catechol- from the pyrogallol-tannins by their chemical characteristics, apart from actual separation of the phenols, and the division is important as it is marked by certain broad differences in their properties which affect their use in tanning.
The catechol-tannins, dissolved in water, yield a precipitate when bromine-water is added till the solution smells strongly of it. The precipitate is occasionally crystalline, but generally amorphous, and of yellowish or brownish colour. When the infusion of tannin is very weak, the precipitate is sometimes only slight, or forms slowly. Pyrogallol-tannins give no precipitate with bromine-water. Another reaction, which is generally characteristic of catechol-tannins, is that if concentrated sulphuric acid is added to a single drop of the infusion in a test-tube, a dark red, or crimson ring is formed at the junction of the two liquids, and on dilution with water, the solution is generally pink. Pyrogallol-tannins on the other hand give a yellow, or at most a dark brown ring, which dilutes to a yellowish solution. This reaction is of great delicacy, which may be further increased by the use of an alcoholic instead of an aqueous extract. It is often given also by the non-tannin residue of catechol-tannins which is left after treatment with hide-powder, in which case it is probably due to the presence of catechins allied to the tannins. With ferric salts (preferably a solution of iron-alum), pyrogallol-tannins give blue-blacks, while catechol-tannins generally give greenish blacks, though the reaction is apt to be rendered uncertain by the presence of colouring matters, or perhaps in some cases by the constitution of the tannin. Thus aqueous infusions of common oak-bark (Quercus robur) give a decidedly blue black with iron, though the tannin is a catechol one, and the purified tannin gives a green-black. Most of the barks of American oaks, such as Q. prinus, give green-blacks without purification. The Australian mimosas generally give dull purple-blacks with iron-salts, though they all contain catechol-tannins. The iron test was first proposed by Stenhouse as a means of classification. Trimble has shown that while the purified pyrogallol-tannins only contain about 52 per cent. of carbon, the catechol-tannins have about 60 per cent.[150]
[150] ‘The Tannins,’ ii. p. 131.
Only two tannins of the pyrogallol group have been definitely distinguished, though it is very possible that more exist. These are ordinary tannic acid of gall-nuts (probably digallic acid), which yields gallic acid when heated with dilute acids, or by the action of certain unorganised ferments or zymases ([p. 16]) which are generally present in tanning materials; and ellagitannic acid (usually present in greater or less proportion in mixture with the gallotannic acid), which, under the same conditions yields “bloom” (an insoluble deposit of ellagic acid), as one of its products. Hence it happens that most pyrogallol tannins deposit “bloom” on leather, though in very different proportions, gall-nuts and sumach giving very little, and myrobalans, valonia and divi-divi a great deal. English oak-bark deposits a good deal of “bloom” on leather, though it is certain that its principal tannin is a catechol one, but it is possible that the blue-black which it gives with iron salts may be due to the presence of ellagitannic acid, though gallotannic acid is known to be absent. The tannins of oakwood, chestnut and valonia are principally if not entirely pyrogallol derivatives, closely allied to, if not identical with the two just named, but, if so, very difficult to obtain in a pure condition. It is noteworthy that so wide a difference exists between the various products of the oak; galls, bark, fruit and wood yielding tannins of very varied properties. The tannin of other galls, such as those of the sumach and pistacio, generally contain gallotannic acid, even when, as in the last case, the remainder of the plant yields catechol-tannins.
The tannins of the catechol group appear to present much more variety than the pyrogallol-tannins, though it is possible that many apparent differences may be due to the presence of impurities. It is, however, at least certain that the tannins of gambier and cutch contain phloroglucol as one of their constituents, while it is absent from most other tanning materials. Its presence is easily detected by moistening pine-wood (a deal shaving) with an infusion of the tannin in question, and applying a little concentrated hydrochloric acid, when after a few minutes, a bright red or purple stain is produced. The catechol-tannins, on boiling with acids, yield no gallic acid, or bloom, but generally a deposit of “reds,” insoluble in water but soluble in alkaline liquids, and in alcohol, and which are closely allied to resins, and especially to the red resin known as “dragon’s blood.” These reds are anhydrides of the tannins, that is, are produced from them by the abstraction of water; and are consequently formed by any agency which tends to remove water, such as long boiling or high temperature. The lower anhydrides (that is, those from which least water has been abstracted) are not wholly insoluble, but form the “difficultly soluble” tannins which are naturally present in many materials. They are much more readily soluble in hot than in cold water, which is one of the causes why liquors made by the aid of heat generally give darker colour to leather than those extracted cold. They exist in large quantity in hemlock extract and quebracho. Attempts have been made to utilise their alkaline solutions for tanning, but without much success; though alkalies or alkaline sulphites are frequently used to obtain “soluble” quebracho extracts ([p. 338]).
Many catechol tanning materials, and especially gambier, cutch and quebracho, contain in addition to the tannin, considerable portions of colourless bodies called catechins, which are only slightly soluble in cold water, but readily in hot, and which crystallise out on cooling. These bodies do not tan, but are in a sense the source of the tannins, which appear to be their first anhydrides, the reds being formed by the successive loss of further molecules of water. These bodies very probably ultimately become converted into tannins by changes in the tanyard. The change may be brought about very rapidly by heating to a temperature of 100 to 120° C.[151] The catechin of gambier, by crystallising on and in the leather, is the cause of a trouble known as “whites,” which is common where gambier is largely used.
[151] Some doubt exists as to the exact temperature at which catechins become converted into anhydrides, and Perkin puts it higher than that stated.
An unfortunate peculiarity, apparently common to all catechol-tannins, is that, however light-coloured the leather produced by them, it darkens and reddens rapidly by exposure to strong light, and ultimately becomes quite friable and rotten.[152] Cp. [pp. 234], [272].