Synthetic Tannins, Their Synthesis, Industrial Production and Application Part 2

The successful resolving of racemic leucodigallic acid into both of its optically active components can only be brought about through the _d_- or _l_-hexacarbethoxyleucodigallic acid on introducing the latter into a 1 per cent. pyridine solution and heating to 45-50 C., whereby the _d_- or _l_-acid is formed accompanied by a strong evolution of carbon dioxide.

Hydrolysis of leucogallic acid yields gallic acid and gallic aldehyde; oxidation by means of hydrogen peroxide yields ellagic acid and luteic acid, and oxidation with pota.s.sium persulphate and sulphuric acid, in acetic acid solution, yields purpurotannin (see below) [Footnote: Liebig's _Ann_., 1912, 386, 318.].

Another distinct difference between digallic acid and leucodigallic acid is the fact that the formaldehyde condensation product of the former resembles gallic acid, whereas that of the latter resembles tannin; it is therefore probable that the leucodigallic acid part of the tannin molecule imparts this characteristic property to tannin.

---CO.O--- ^ ^ | | | | HO V OH COOH V OH OH OH [Ill.u.s.tration: Digallic Acid becomes...]

---CO.O--- ^ ^ OH | | | | HO V OH COOH V OH OH OH [Ill.u.s.tration: Luteic Acid becomes...]

---CO.O--- ^ ^ OH | | | | HO V --O.CO-- V OH OH OH [Ill.u.s.tration: Ellagic Acid becomes...]

COOH COOH ^ _______ ^ | | | | HO V ---O--- V OH OH OH [Ill.u.s.tration: Purpuro Tannin.]

3. Ellagic Acid

Ellagic acid was discovered in 1831 by Braconnot, who named it "acide ellagique." Its presence in the vegetable kingdom was not quite comprehended for some time, and Nierenstein [Footnote: _Chem. Ztg._, 1909, 87.] was the first to prepare this substance from algarobilla, dividivi, oak bark, pomegranate, myrabolarms, and valonea. The acid is obtained by precipitating it with water from a hot alcoholic extraction of the plants referred to, and recrystallising the precipitate from hot alcohol. Another method of preparation consists in boiling the disintegrated plants with dilute hydrochloric acid, washing the residue, and extracting with hot alcohol, from which the acid will then crystallise. According to Lowe, [Footnote: _Zeits. f. a.n.a.lyt. Chem._, 1875, 35.] it may be obtained from dividivi, an aqueous extract of which is heated to 110 C. in a tube closed at both ends, when crystalline ellagic acid is deposited. Heinemann [Footnote: Ger. Pat., 137,033 and 137,934.] obtained ellagic acid by simply boiling repeatedly aqueous tannin solutions.

Lowe [Footnote: _Jour. f. prakt. Chem._, 1868, 103, 464.] first synthesised ellagic acid by heating gallic acid with a.r.s.enic acid or silver oxide. Herzig [Footnote: _Monatshefte fur Chemie_, 1908, 29, 263.] states that ellagic acid is deposited when air is conducted through a mixture of the ethyl or methyl ester of gallic acid and ammonia. Perkin [Footnote: _Proc. Chem. Soc._, 1905, 21, 212.] obtained a substance very similar to ellagic acid by electrolysis of gallic acid in sulphuric acid solution; on oxidising gallic acid in concentrated sulphuric acid solution, Perkin and Nierenstein [Footnote: _Ibid._, 1905, 21, 185.] obtained flavellagic acid. Ellagic acid is also obtained by heating luteic acid in a 10 per cent. soda solution.

Ellagic acid thus prepared crystallises with 2 molecules of water as yellow micro-crystalline rhombic prisms or prismatic needles. The crystals lose this water when heated to 100 C., and it is possible that it is water of const.i.tution, in which case the substance would be hexoxydiphenylcarboxylic acid, and the substance left after drying at 100 C., the dilactone.[Footnote: _Arch. d. Pharm_., 1907, 244, 575.]

Ellagic acid is slightly soluble in water, alcohol, and ether, but is easily soluble in caustic potash. With concentrated nitric acid the product a.s.sumes a red colour, which appears to be due to the presence of impurities; ellagic acid is commercially known as "alizarin yellow."

The const.i.tution of ellagic acid was uncertain for a long time, and different structural formulae were proposed which more or less corresponded to its properties. The most satisfactory structural formula was proposed by Graebe--[Footnote: _Chem. Ztg_., 1903, 129.]

---CO.O--- ^ -------- ^ OH | | | | HO V --O.CO-- V OH OH

This would represent a tetroxydiphenylmethylolide.

The probability of the correctness of this formula is supported by the possibility of the following derivatives: monomethylellagic acid, C'14H'6O'7(O.CH'3); dimethylellagic acid, C'14H'4O'6(O.CH'3)'2; tetramethylellagic acid, C'14H'2O'4(O.CH'3)'4; phenylhydrazinellagic acid, C'14H'6O'8.N'2H'3C'6H'5.

By the electrolytic reduction of ellagic acid, hexoxydiphenyl, (OH)'3C'6H'2-C'6H'2(OH)'3, is obtained; the ordinary methods of reduction yield leucoellagic acid, C'14H'10O'8, which crystallises in small sharp needles, melting with decomposition at 294-295 C. Leucoellagic acid is soluble in ethyl and methyl alcohols, and in glacial acetic acid, insoluble in chloroform, benzene, toluene, carbon tetrachloride, and petrol ether; it gives a bluish-green colour with ferric chloride which quickly turns black. Leucoellagic acid is soluble in alkalies, the solution a.s.suming a deep-red coloration; it reduces silver nitrate in the cold, but is not adsorbed by mordanted cotton cloth, in which respect it differs from ellagic acid.[Footnote: Liebig's _Ann_., 1912, 394, 249.

ELLAGITANNIC ACID, C'26H'28'O'10-3H'2O, is closely related to ellagic acid; the former consists of faintly yellow needles, M.P. 329-336C.

It is soluble in water, precipitates gelatine, and is adsorbed by hide powder. It occurs with gallic acid, tannin, and ellagic acid in dividivi, myrabolams, algarobilla, and chestnut wood extracts.

Other bodies of this cla.s.s include:--

METELLAGIC ACID, Cl_4H_6O_5, derived from methoxybenzoic acid, and recrystallised from acetic acid, forms small crystalline needles, M.P. 273-276 C., and yields fluorene on distillation with zinc dust.

----CO.O---- ^ ---------- ^ | | | | V ---O.CO--- V OH

FLAVELLAGIC ACID, C_14H_6O_9, is obtained by the oxidation of gallic acid with concentrated sulphuric acid and pota.s.sium persulphate. It crystallises from pyridine in prismatic needles melting above 360 C. Distillation with zinc dust yields fluorene (see above)--

----CO.O---- ^ ---------- ^ OH | | | | HO V ---O.CO--- V OH OH OH

By heating ellagic acid for three-quarters of an hour at 185 C. with concentrated sulphuric acid, ceruleo-ellagic acid (dioxyellagic acid), C_14H_6O_10, is formed as yellowish needles, M.P. 360 C., which are but little soluble in the usual solvents. The acid is slightly soluble in strong caustic soda solution, the colour of the solution, on diluting, changing to green and blue.

LUTEIC ACID (Luteo Saure, pentoxybiphenylmethylolide carboxylic acid),C_14H_8O_9, occurs, in addition to ellagic acid, in myrabolams-- [Footnote: _Ber_., 1909, 42, 353.]

----CO.O---- ^ ---------- ^ OH | | | | HO V OH HOOC V OH OH OH

It is obtained by extracting myrabolams for one hour and a half, under reflux condenser, with pyridine, filtering and adding twice the volume of water to the filtrate and boiling till complete solution is obtained. After about thirty hours a reddish powder deposits, from which ellagic acid may be extracted with pyridine; the mother-liquor on being concentrated yields luteic acid. It is also obtained by oxidising tannin with hydrogen peroxide, the other oxidation product being ellagic acid, and the two may then be separated as indicated above. Luteic acid forms reddish needles which are decomposed, with evolution of gas, at 338-341 C. Heated with 10 per cent. caustic soda solution it yields ellagic acid. In pyridine solution the carboxyl group maybe eliminated by hydrogen iodide, whereby pentoxybiphenylmethylolide is formed as long silky needles, which do not melt below 300 C. The same substance may also be obtained when ellagic acid is boiled with concentrated caustic potash solution. When luteic acid is treated with diazomethane, it yields the methyl ester of pentamethoxybiphenylmethylolidcarboxylic acid.

4. DEPSIDES

The most common decomposition products of the natural tannoids are hydroxybenzoic acids, notably gallic and proto-catechuic acids; furthermore, other aromatic and aliphatic hydroxy compounds frequently occur. So far, however, attempts at explaining the const.i.tution of the complex decomposition products obtained by hydrolysing high molecular tannoids have not been successful. On the other hand, the const.i.tution of the simpler natural tannoids is known to a greater or less extent; of these, lecanoric acid (Lecanorsaure) is the best known, being an ester anhydride of orsellic acid (a dihydroxytoluylic acid). It combines with erythrite, forming another tannoid, erythrine. The fact that hydroxybenzoic acids are constantly encountered together with the products obtained on hydrolysis of the tannins, seems to point toward the conclusion that anhydrides of hydroxybenzoic acids are frequent const.i.tuents of the natural tannoid molecules.

The a.s.sumption that, for instance, in tannin at least part of the gallic acid radicals are combined with one another is highly probable, and is supported by the formation of tri- and dimethylgallic acid from methylotannin, [Footnote: Herzig, _Monatshefte f. Chemie_, 1909, 30, 343.] and by the formation of ellagic acid when tannin is oxidised.

[Footnote: Nierenstein, _Ber_., 1908, 41, 3015.] Further proof is brought forward by the existence of the pentacetyl-tannin, [Footnote: Schiff, _Ann. d. Chem_., 1873, 170, 73.] and by the results of hydrolysis which has yielded up to 104 per cent. anhydrous gallic acid fiom tannin [Footnote: Sisley, _Bull. Soc. Chim_. 1909, 5, 727.]

Of the three cla.s.ses of isomeric anhydrides which can be formed from hydroxybenzoic acids, the chemistry of the natural tannins is only concerned with the cla.s.s comprising the ester anhydrides. If the carboxyl of the first molecule combines with a hydroxyl of the second molecule (ester formation), then a substance possessing character similar to that of a hydroxybenzoic acid is formed, which is capable of combining up with a further molecule in the same way. It is natural to a.s.sume that this ester form is much more prevalent in Nature than a combination of two carboxyls by the elimination of water. From the point of view of the chemistry of the tannins, therefore, the starting-point would naturally be that of synthesising the ester anhydrides of hydroxybenzoic acids. Amongst the small number of synthetically prepared ester anhydrides of hydroxybenzoic acids, a few occur exhibiting the properties of the natural tannoids.

In order to simplify the terminology of these substances, Fischer [Footnote: Liebig's _Ann_., 1910, 372, 35.] proposed the name "Depsides" from [Greek: depheiv] = to tan. In a.n.a.logy with peptides and saccharides, the names di-, tri-, and polydepsides of hydroxybenzoic acids would be suitable for these substances.

The principles underlying the synthesis of depsides are the following:--If the chlorides of carbomethoxy (or carbethoxy) hydroxybenzoic acids are coupled with the sodium salts of hydroxybenzoic acids, esters are formed, _e.g._,

CH_3CO O.O.C_6H_4.CO.Cl + NaO.C_6H_4.COO.Na = NaCl + CH_3.COO.O.C_6H_4.CO.O.C_6H_4.COO.Na

On gently saponifying the esters, these are converted into the corresponding hydroxy derivatives--

OH.C_6H_4.CO.O.C_6H_4.COOH

According to Fischer and Freudenberg, [Footnote: Liebig's _Ann._, 1909, 372, 32.] this method possesses the following advantages:--

1. The synthesis takes place at low temperatures, so that any intramolecular rearrangements are improbable.

2. The composition of the substances is controlled by the intermediary compounds, the carboalkyloxy derivatives.

3. The synthesis permits of more definite evidence as regards the structure of the resulting compounds.

4. The substances obtained are easily purified.

Depsides produced in this manner are by no means new, and were obtained by Klepl by simply heating _p_-hydroxy-benzoic acid (_cf._ Introduction, p. 4). This simple procedure, however, is not applicable to most other hydroxybenzoic acids which are decomposed at the high temperature necessary to induce reaction. Lowe and Schiff (_loc. cit._) have obtained products similar to tannins, the latter investigator by removing the elements of water from gallic acid, protocatechuic acid, salicylic acid, _m_-hydroxybenzoic acid, cresotinic acid, phloretinic acid, and pyrogallolcarboxylic acid. These depsides, however, are amorphous substances, and it is hence difficult to substantiate their h.o.m.ogeneity.

Carbomethoxylation of Hydroxybenzoic Acids

Amongst other compounds chlorphydroxybenzoic acid is used in the preparation of the materials employed in the synthesis of depsides; the free phenolic group, however, exerts a disturbing influence when aromatic acids are acted upon by phosphorus chloride, and another group, which can subsequently be easily removed, must therefore be introduced to cover the disturbing influence referred to. For this purpose, Fischer [Footnote: _Ber_., 1908, 41, 2860.] chose the carbomethoxy group, and this investigator succeeded, by the action of chlorocarbonic alkyl ester and alkali upon hydroxybenzoic acid in cold aqueous solution, in obtaining substances with the properties required. [Footnote: _Ber._, 1908, 41, 2875.] In such substances (_e.g._, salicylic acid) where the hydroxyl occupies the ortho-position to the carboxyl, complete carbomethoxylation does not take place, whereas the _m_- or _p_- positions offer no hindrance. In the case of the _o_-position, however, the action of chlorocarbonic alkyl ester is successfully a.s.sisted by the presence of dimethylaniline in an inert solvent, _e.g._, benzene.[Footnote: U.S. Pat, 1,639,174, 12, xii., 1899.] The difficulty encountered by the _o_-position is eliminated when the carboxyl is not directly linked to the benzene nucleus, _e.g._, _o_-c.u.maric acid. Many hydroxybenzoic acids require an excess of chlorocarbonic methyl ester, which then also, to some extent, attacks the carboxyl group; but on dissolving the product in acetone and treating it with bicarbonate the carboxyl group as such is again restored without splitting off the carbomethoxy group.[Footnote: _Ber._, 1913, 46, 2400.] In this way all hydroxybenzoic acids may be carbomethoxylated. [Footnote: _Ibid._, 1908, 41, 2877, 2881, 2882; 1909, 42, 226, 218, 223, 225; Liebig's _Ann._, 1912, 391, 357, 366; _Ber._, 1913, 46, 1145, 2390, 2400.] The carbomethoxy group is easily removed by excess of aqueous alkali in the cold, and is also partially removed when insufficient alkali is present; the latter fact is of importance in the synthesis of didepsides.

Chlorides of Carbomethoxyhydroxybenzoic Acids

The chlorides of these compounds are obtained when phosphorus pentachloride is allowed to act upon the acids, and are as a rule crystalline. For the purpose of synthesis they may be employed as follows:

1. They readily form esters with alcohols, which on subsequent saponification with alkali are converted into the esters of the free hydroxybenzoic acids.

2. The chlorides interact energetically with esters of amino-acids, and may be coupled with amino-acids in aqueous alkaline solution. On subsequently removing the carbo-methoxy group derivatives of hydroxybenzoic acids are obtained, _e.g._,