CHAPTER VIII

THE CHEMISTRY OF AROMATIC COMPOUNDS

The suggestions of Couper and Kekulé that an explanation of the properties of chemical compounds should be sought in the nature and mutual affinities of their constituent elements rather than of their radicals were not wholly accepted at the time they were first made. Speculative ideas have to justify themselves by facts. The value of an hypothesis depends upon its usefulness and expediency, and on its power of indicating the lines of future inquiry. How far it is inductively sound and deductively useful is a matter for individual judgment. Consequently the tendency to pass from purely rational and constitutional formulæ to formulæ which sought to symbolise the inner structure—the very skeleton, as it were—of a molecule, was resisted for a time, and by no one more strongly than by Kolbe.

Kolbe’s attitude to the new doctrine may be said to have had its justification in his own work. His remarkable prediction, based on considerations which had nothing in common with Kekulé’s doctrine, of the existence of the secondary and tertiary alcohols, so soon to be confirmed by Friedel’s discovery of secondary propyl alcohol, and by Butlerow’s isolation of tertiary butyl alcohol, served to retard the general adoption of Kekulé’s views by showing that apparently they were no more fruitful in suggestiveness than those they were intended to supplant. But it was exactly in their suggestiveness with regard to the development of isomerism that structural formulæ based upon valency were gradually found to be most useful. It was perceived that it was now possible not only to foretell the existence of isomers, but to determine their number, and to some extent to forecast their properties and modes of decomposition. Cayley, for example, calculated the number of possible isomers of the hydrocarbons of the generic formula CnH2n+2 up to C6H14 all those that theory predicted have been discovered. In no single case have more been obtained than the number indicated by theory. The accumulated weight of this and similar testimony ultimately established the doctrine of chemical structure on a firm basis.

This conception received a great extension as the result of Kekulé’s application of his ideas to the explanation of the chemical constitution of the group of substances of vegetable origin—consisting of essential oils, balsams, resins, and their products, which, on account of their characteristic odours, were collectively known to the older chemists as the _aromatic compounds_. Some of these, like oil of bitter almonds, gum benzoin, coumarin, oil of wintergreen, oil of anise, oil of cinnamon, oil of cumin, balsam of tolu, phenol, and certain of their derivatives, such as benzene, aniline, salicylic acid, cinnamic acid, toluene, cymene, had already been investigated with important theoretical results; but as a class they had received far less attention than the derivatives of the great group of homologous radicals of which methyl is the initial member. Of course it was part of the doctrine of Liebig—the discoverer of benzoyl—that the aromatic compounds also contained specific radicals; but the relation of these compounds to those we now call aliphatic (fatty) compounds was not understood, although certain analogies were recognised.

In 1866 Kekulé drew attention to the following significant peculiarities of the aromatic compounds: (1) All aromatic compounds, even the simplest, are comparatively richer in carbon than the corresponding class of fatty (aliphatic) compounds; (2) among the aromatic substances, as among fatty compounds, numerous homologous compounds exist; (3) the simplest aromatic substances contain at least six atoms of carbon; (4) all decomposition products of aromatic substances show a certain family resemblance; the main product of the decomposition contains at least six atoms of carbon—_e.g._, benzene C6H6, phenol C6H6O, etc., which would seem to indicate that all aromatic substances contain a nucleus or atomic grouping containing six carbon atoms. Within this nucleus the carbon atoms are in closer connection or denser combination, from which it follows that all aromatic compounds are comparatively rich in carbon. More carbon atoms can then be added to this nucleus according to the same laws that govern the fatty compounds. In this way the existence of homologous compounds may be explained.

On the assumption that carbon is tetravalent and that its valency is constant, Kekulé showed how, by linking together six carbon atoms by alternate single and double bonds, six affinity units may be left free. If we assume that six carbon atoms are attached to one another according to this law of symmetry, we obtain a group which, regarded as an _open chain_, contains _eight_ unsaturated units of affinity:

——C══C——C══C——C══C—— | | | | | |

By making the further assumption that the two carbon atoms at the ends of the chain are linked together by one unit of affinity each, a _closed chain_ (a symmetrical ring) is obtained which now contains _six_ unsaturated units of affinity:

+--------------+ | | C══C——C══C——C══C | | | | | |

From this _closed chain_ all the substances usually designated as “aromatic compounds” are derived. In these a common nucleus may be assumed: it is the closed chain C6A6, where A denotes an unsaturated affinity. The six affinities of this nucleus may be satisfied by six monovalent elements. They may also, wholly or in part, be satisfied by one affinity of polyvalent elements, the latter necessarily bringing with them other atoms into the compound, thus producing one or more side chains, which in their turn may be lengthened by the addition of other atoms.

If each of the free units is satisfied by an atom of hydrogen, we obtain benzene, which, as Kekulé demonstrated, becomes the centre round which the great group of aromatic compounds might be said to revolve. Benzene was discovered by Faraday in 1825 among the volatile liquids condensed from the oil-gas made by the Portable Gas Company. It had already played a notable part in the development of chemical theory in connection with the discovery of isomerism. It was now to play a far more important rôle: to become, in fact, the progenitor of a great family of substances, not only of theoretical value, but of great economic importance.

The limits of this work preclude any attempt to trace in detail the development of the conception with which Kekulé enriched science, or to dilate upon the great extension of benzenoid or cyclic chemistry which has resulted from it during the past forty years. It has been said that Kekulé’s memoir on the benzene theory is the most brilliant piece of scientific prediction to be found in the whole range of organic chemistry. Of course, on its promulgation it had to run the gauntlet of criticism; and an army of eager, active workers was soon engaged in testing its sufficiency and in developing the rich province which it first made known. As the facts multiplied, other statical formulæ were suggested by Dewar, Ladenburg, and Claus, but they have not proved more adequate to explain the facts as these have become better understood. Observations which seemed to contradict Kekulé’s theory, or which seemed to be imperfectly explained by it, have, in the light of fuller knowledge, been shown to be in harmony with it; and such additional proofs of agreement have thereby served to strengthen its position. Its capacity for development is, indeed, as in the case of every other hypothesis of the first rank, one of its cardinal qualities. It adequately explains the constitution of great numbers of derivatives whose analogies and relations, apart from it, would have remained obscure and in many cases unintelligible. The symmetrical distribution of the carbon and hydrogen atoms in the benzene molecule, assumed by Kekulé on indirect grounds, has been established by the independent investigations of Ladenburg and others, and its ring structure has been demonstrated by Baeyer and Perkin. Purely physical evidence, based upon its thermo-chemical and optical characters can be adduced in its support. Determinations of the molecular volume and magnetic rotation of its compounds further serve to substantiate it.

│Friedrich August Kekulé│ was born at Darmstadt on September 7, 1829. After passing through the gymnasium of his native town, he entered, in 1847, the University of Giessen, with the intention of becoming an architect. Attracted by Liebig’s teaching, he turned to chemistry, and worked with Will on _amyl sulphuric acid_ and its salts. In 1851 he went to Paris, heard Dumas’s lectures, and formed a friendship with Gerhardt, whose _Traité de Chimie Organique_ largely moulded his views. He became an assistant to Von Planta, occupying himself with the chemistry of the alkaloids. Subsequently he came to London, worked under Stenhouse, and made the acquaintance of Williamson, then in the full vigour of his scientific activity. Here he discovered _thioacetic acid_. It was at this time, also, that his ideas with regard to structural chemistry began to take shape. Returning to Germany, he attached himself to the University of Heidelberg as a _privat-docent_, and had for a pupil Baeyer, who took up the study of the organo-arsenic compounds. In 1858 he published his memorable paper “On the Constitution and Metamorphoses of Chemical Compounds and on the Chemical Nature of Carbon,” in which he developed his views on the linking of atoms, out of which has grown our system of constitutional formulæ. The immediate result of this celebrated memoir was a call to the chemical chair of the University of Ghent, where Kekulé had among other students Baeyer, Hübner, Körner, Ladenburg, Linnemann, and Dewar. Here he remained nine years, and here he published his classical _Lehrbuch der Organischen Chemie_. The years he spent in Ghent were the most productive time of his career, and it was there that he developed his benzene theory—a conception as fruitful as that of his doctrine of atom-linking. In 1867 Kekulé was called to Bonn to take charge of the newly erected laboratory which Hofmann had designed. Although he continued to work, mostly in collaboration with his pupils, among whom may be named Anschütz, Bernthsen, Thorpe, Carnelley, Claisen, Dittmar, Franchimont, Van ’t Hoff, Japp, Schultz, Wallach, and Zincke, his health after 1876 began to fail. He died on July 13, 1896.

* * * * *

Of course no statical formula can be the ultimate representation of the constitution of benzene. However convenient and suggestive such a formula may be, it can be only a transitional phase in its complete chemical and physical history. Kekulé was early conscious of this fact, and suggested a dynamical hypothesis based upon a mechanical conception of valency. This he imagined might represent the number of contacts with other atoms which a vibrating atom experienced in the unit of time. Two atoms of tetravalent carbon, each linked by one combining unit, will experience four oscillations, striking each other and three other atoms in the unit of time, while the monovalent hydrogen atom makes only one oscillation. The doubly linked carbon will collide with its neighbouring atom twice, and also with two other atoms within the same period. The assumption that the carbon atom has a more rapid motion than the hydrogen atom is, however, not warranted by the kinetic theory. Other dynamic formulæ have been proposed by Knorr and by Collie. Collie and Baly have further suggested that the absorption bands of benzene observed in the ultra violet of its spectrum point to synchronous oscillations of its molecule, due to dynamic changes in the making and breaking of the links between the several pairs of the carbon atoms, setting up vibrations in the benzene ring comparable with those of an elastic ring in the act of expanding and contracting.

* * * * *

The large group of the _essential oils_, containing hydrocarbons similar to oil of turpentine, and classed under the generic term of _terpenes_, might, from their origin and mode of occurrence, be expected to be allied in constitution to the aromatic compounds; and such is found to be the case. The terpenes are isomeric hydrocarbons of the formula C10H16. They are found sometimes singly, at other times mixed, in a great variety of plants, associated with _sesquiterpenes_ C15H24, and oxygenated substances, such as camphor, borneol, menthol, etc., some of which have long been known and valued for their medicinal properties and technical applications. The elucidation of their constitution has taxed the skill of many workers during the past thirty years; but, thanks to the labours of Wallach, Baeyer, Perkin, Tiemann, Bredt, Komppa, and others, an insight has been gained into their nature and analogies. They are apparently all cyclic compounds with certain attributes which connect them with hydrocarbons of the aliphatic series. _Pinene_, the characteristic constituent of oil of turpentine, obtained by distilling the resinous exudations of many species of pines, exists in two modifications, distinguished by differences in optical activity, known respectively as _australene_, found in American, Russian, and Swedish turpentine, and _terebenthene_, found mainly in French turpentine. It would seem from their empirical formulæ, as well as from their association in nature, that the terpenes and _camphor_, which Dumas first showed to have the composition C10H16O, should be closely allied in constitution, and that it ought to be readily possible to effect their mutual transformation. The constitution of camphor was long one of the standing problems of organic chemistry, and dozens of formulæ have been suggested at various times during the last twenty years in explanation of its structure. That it contained a benzene nucleus seemed to be proved by the ease with which it yielded toluene, cymene, and other benzene homologues. The first real insight into its structure was gained when Bredt ascertained the constitution of _camphoronic acid_, C6H11 (CO2H)4—a product, together with _camphoric acid_, of the oxidation of camphor—which he found broke up into trimethylsuccinic acid and _iso_butyric acid, and the structure of which was established by Perkin and J. F. Thorpe.

The result of the Japanese monopoly has been to greatly enhance the price of natural camphor; during the last ten years it has practically trebled. This has naturally stimulated endeavours to prepare this substance by synthetical means. _Artificial camphor_ is now made from _pinene_ by transforming the hydrocarbon into bornyl chloride by the action of hydrochloric acid. From this _camphene_ is prepared; by treatment with glacial acetic acid it forms _isobornyl acetate_. On hydrolysis this is transformed into _isoborneol_, which by oxidation yields _camphor_, differing from the naturally occurring variety only in the fact that it is optically inactive. All so-called aromatic compounds are not necessarily cyclic systems, for it has been recognised within the past few years that some of the most valuable natural perfumes, such as that of the rose, lavender, and orange blossom, lemon-grass, geranium, ylang-ylang, neroli, etc., owe their characteristic aroma to the presence of terpenes and camphors, which are not strictly benzenoid or cyclic compounds, but “ruptured rings” behaving like open-chain or aliphatic substances. To judge from past experience, it may confidently be stated that, now the constitution of these substances is understood, their synthetical preparation on an industrial scale is practicable. The discovery by Cahours in 1844 that _oil of wintergreen_ is substantially methyl salicylate led to its artificial production from synthetically prepared salicylic acid. Sir William Perkin in 1868 effected the synthesis of _coumarin_, the aromatic principle in woodruff and hay. Fittig and Mielck in 1869 synthesised _heliotropin_, and in 1871 Tiemann and Haarmann obtained _vanillin_, the characteristic aromatic body in the vanilla pod, by synthetic means, and established its manufacture on a commercial scale. The chemical nature of the characteristic odoriferous substances in oil of cumin, anise, rue, cinnamon, heliotrope, jasmine, violet, parsley, etc., has now been established and some of them are made industrially. The artificial essence of violets known as _ionone_, prepared by Tiemann in 1893, and now made commercially, is similar but not identical in structure with the true perfume—_irone_. What is known as _artificial musk_ is a trinitro-butyl toluol. _Artificial orange-flower oil_ is a methyl ester of anthranilic acid.

In Vol. I. a short account has been given of the early history of the large and important group of vegetable products known as the _alkaloids_. Many of these have long been valued on account of their powerful physiological action. As they all contain nitrogen and are generally basic, they were regarded by Berzelius, and subsequently by Liebig and Hofmann, as akin to ammonia in constitution, and were classed as amines. The first experimental evidence of their nature was obtained by Gerhardt, who found that, when strychnine and certain of the alkaloids belonging to the quinine group are treated with potash, an oil was obtained which he termed _quinoline_, and which was recognised by Hofmann as identical with a substance obtained in 1834 from coal-tar by Runge, and at that time known as _leucol_. By other modes of treatment certain alkaloids—_e.g._, nicotine and conine—are found to yield pyridine, a basic substance found by Anderson, in 1846, in the fœtid liquor obtained by distilling bones, and since found in coal-tar. Others of them—_e.g._, papaverine, narcotine, etc.—yield _isoquinoline_, an oil also discovered in coal-tar, by Hoogewerff, and Van Dorp, in 1885. These three substances—quinoline, _iso_quinoline, and pyridine—constitute so many nuclei in the constitution of a large number of alkaloids. Pyridine resembles benzene in being a cyclic compound, consisting of five carbon atoms and one nitrogen atom. Quinoline stands to pyridine in much the same relation that naphthalene stands to benzene. It can be obtained synthetically, as first shown by Koenigs and Skraup, and subsequently by Doebner and Von Miller, from benzene derivatives.

_Iso_quinoline, isomeric with quinoline, differs from that substance in the position of the nitrogen atom. It, too, has been synthetically prepared from benzene derivatives in a number of ways.

Among the naturally occurring pyridine alkaloids may be named _piperine_, found in black pepper, and _conine_, the poisonous principle of hemlock (_conium maculatum_). The latter alkaloid was prepared synthetically by Ladenburg in 1886; as first obtained it differed from the naturally occurring product, which is dextro-rotatory, in being optically inactive. Ladenburg surmised that the synthetic preparation stood to the naturally occurring compound in the same relation that racemic acid stood to tartaric acid, and that, by treatment in the manner employed by Pasteur, the racemic modification of conine might be separated into its dextro- and lævo-constituents. This was found to be the case; but the separated dextro component was now found to be distinctly more optically active than the pure, natural variety. It was, in fact, an isomeric modification—_iso_-conine. By heating this to 300° it was transformed into ordinary conine, identical in all respects with the natural alkaloid. Ladenburg has also effected the synthesis of piperine by condensing piperidine and piperinic acid.

_Nicotine_, the alkaloid of tobacco, was discovered by Posselt and Reimann in 1828. Its constitution was first ascertained by Pinner, and it was synthetically obtained by Amé Pictet, in 1904, as an inactive substance, capable of being resolved by the crystallisation of its tartrates into a dextro- and lævo-modification, the latter of which was identical with that found in the tobacco leaf.

_Atropine_ and _hyoscyamine_—the poisonous principles of belladonna and henbane—are isomeric alkaloids, the former of which is optically inactive, and the latter is lævo-rotatory. Atropine is, in fact, the racemic modification. The constitution of both alkaloids is known, and their synthesis is now possible.

The successive steps may be thus indicated:

1. _Synthesis of glycerin_ (Faraday, Kolbe, Melsens, Boerhave, Friedel, and Silva).

2. _Glycerin to glutaric acid_ (Berthelot and De Luca, Cahours and Hofmann, Erlenmeyer, Lermantoff, and Markownikoff).

3. _Glutaric acid to suberone_ (Brown and Walker, Boussingault).

4. _Suberone to tropidine_ (Willstätter).

5. _Tropidine to tropine_ (Willstätter, Ladenburg).

6. _Synthesis of tropic acid_ (Berthelot, Fittig and Tollens, Friedel, Ladenburg, and Rügheimer).

7. _Tropine and tropic acid: atropine_ (Ladenburg).

The alkaloid _cocaïne_, contained in the leaves of _erythroxylon coca_ and now employed as a local anæsthetic, was discovered by Niemann in 1860. It has been shown to be closely related to atropine in constitution, and has now been synthetically prepared in the dextro-modification.

The alkaloids _papaverine_, _narcotine_, _narceïne_, contained in opium, are derivatives of _iso_quinoline, as also is _berberine_, found in the common barberry (_berberis vulgaris_). Papaverine, which occurs in opium to the extent of about one per cent., was first isolated by Merck in 1848. Its constitution has been established by Goldschmidt. _Narcotine_ is, next to morphine, the most abundant constituent of opium. The study of the products of its hydrolysis and oxidation—viz., _opianic acid_ and _cotarnine_, both of which substances have long been known—has indicated its probable structure. _Narceïne_ is closely allied to narcotine, and can, indeed, be obtained from it by combining the latter alkaloid with methyl iodide and treating the compound with caustic potash. The constitution of _berberine_, which is one of the few coloured vegeto-alkaloids known, has been worked out by Perkin. As yet nothing definite is known concerning the structure of the most important and largest constituent of opium—viz., _morphine_; or of its congeners _codeïne_ and _thebaine_. Grimaux, however, in 1881 converted morphine into codeïne by treatment with methyl iodide and potash; hence the two alkaloids stand in a relation somewhat similar to that in which narceïne stands to narcotine. There is very little doubt that the three alkaloids are very closely related, and that a knowledge of the constitution of one of them would immediately elucidate the structure of the others. They are probably phenanthrene derivatives.

_Quinine_ and _cinchonine_, the most important of the cinchona alkaloids, are quinoline compounds, and are closely related in constitution. But the many attempts to unravel their structure have yielded no definite results up to the present.