ACTION OF CHLOROFORM, ETC., ON BlEThMIDOBENZOIC ACID. 151 3 XCVI 11. -A c t ion of Chloroform ai 1 d Pot ussium .Hydroxide on, l%?etarnidobenxoic acid. By WALTER JOHN ELLIOTT, M.A. AS the action of chloroform and potash on amido-acids of the aro- matic series had not been investigated, it seemed of interest to ascertain whether the presence of the carboxyl group would cause any vayiation in their ordinary action on compounds containing the amido-group, especially as Dr. Ruhemann and the author have shown that, in the case of phenylhydrazine, the reaction is not normal, the substances formed not being isonitriles (Ruhemann and Elliott, Trans., 1888, 53, 850; Ruhemann, Trans., 1889, 55, 242). In the early experiments, a solution of potash in alcohol was used, but a solution in water was afterwards found to give a better yield.5 grams of metamidobeneoic acid dissolved i n spirit were added to 8 solution of 15 grams of potash also dissolved in spirit., and the mix- ture was heated on a water bath in a flask furnished with a reflus condenser; 8 grams of chloroform were then carefully added, and the whole boiled for some time. On cooling, diluting considerably with water, and acidifying with dilute hydrochloric acid, a yellow, amorphous precipitate was thrown down after a time; in some of1514 ELLIOTT: ACTION OF CHLOROFORM AND the experiments, a slight isonitrile odour was detected before acidi- fying. After collecting and washing the product, attempts were made to purify it by crystallisation, but no solvent for i t has yet been found ; it is insoluhle in water, alcohol, ether, chloroform, benzene, petroleum, acetic acid, and acetone ; consequentl;p, its purification has been a matter of great difficulty, increased by the fact that it does not melt below 300°, arid darkens and decomposes above that tern- peratwe.It has, however, acid properties, dissolving in alkalis and in sodium carbonate solution, conseqnently an attempt was made to purify i t by dissolving i t in sodium carbonate Bolution, boiling with animal charcoal, and acidifying, but no precipitate was thrown down, and, on shaking with ether, nothing was removed from the solution. The substance must, therefore, be decomposed by hot alkaline solu- tions ; it was found, however, t o be reprecipitated by acids from cold alkaline solutions, and t h i s method of piirificntion was tried.After washing the reprecipitated substance, it was dried a t ZOOo and analysed. Since the substance does iiot melb, the relative purity of different specimens could only be determined by analysis. The numbers I, 11, and III were obtained by analysis of the repro- cipitated substance, the crude product (after washing with hot water and hot alcohol) giving the results I V and V. I. 11. 111. IV. V. Sldehyde. Carbon .. .. .. .. 57.4 56.66 - 55.4 55-72 58.18 Hydrogen . . .. . . 4.3 4.48 - 5-02 5.13 4.24 - 8.16 - - 8.48 Nitrogen . . . . . . . - As the results of analysis given above show that the percentage of carbon is less than in metamidobenzoic acid itself, the substance cannot be an isonitrile; but it led me to examine the reactions of the product with Fehling’s solution and with phenylhydrazine, with a view of ascertainirig whetber it was an aldehyde or allied compound.A compound having the composition NH,*C6H3( COOH)*COH would have the percentage composition given above in the last column, headed “ Aldehyde ” ; these numbers are in fairly close agreement with the results of analysis. B’ehliug’s solution is readily reduced by it on heating, and almost decolorised even at the ordinary temperature. When suspended in dcohol and treated with a quantity of phenylhydrazine calculated on the basis of the aldehyde formula, the Substance dissolves slowly a t ordinary temperatures, and readily on heating, yielding a deep red solution from which red crystals are precipitated on the addition of water. These reactions indicate that the product is an aldehyde or a compound allied to an aldehyde.In order to obtain further evidence of the aldehyde nature of thePOTASSIUM EYDROXIDE ON METAMIDOBENZOIC ACID. 151 5 product, an aqueous solution was substituted for the alcoholic solu- t,ion of potash, in fact the ordinary method for the preparation of hydroxy-aldehydes from phenols was used ; this was found to give a far better yield. 10 grams of the acid were dissolved i t 1 a solution of 20 grams of potash in about PO0 C.C. of water, 14 grams of chloroform gradually added, and the whole boiled for about three quarters of an hour; the solution was then boiled with animal charcoal for a short time, filtered, cooled, acidi6ed with acetic acid, and left for five or six hours, when about 6 grams of the crude product were deposited.As all other methods of purification had failed, fractional pre- cipitation from the potash solution was tried. The solution acidified with acetic acid was allowed to remain for two hours, filtered from the precipitate which had formed, and again allowed to stand for several hours. The precipitate then formed was well washed with warm water, and dried in il vacuum. Theory for r--J-- 7 CSHjNOP I. 11. 111. Carbon ........ 58.18 57.80 58.11 - Hydrogen. ..... 4.24 5.02 4.7’4 - Nitrogen ....... 8.48 - - 8-36 Found. The substance, therefore, has the empirical formula, CeH,NO,. With the ammoniacal solution of the acid, silver nitrate gives a jellowish precipitate, which readily decomposes.The barium salt was then prepared by adding barium chloride to a neutral solution of the acid in ammonia, the yellow precipitate was collected, washed with water, and dried in a vacuum. On analysis i t was found to contain ‘29.97 per cent. of barium, theory requiring 29.46 per cent. for a salt of thc formula (C8H,N03),Bn. Since the acid was decomposed when boiled with alkalis, it seemed possible that it might also decompose when boiIed with water. 5 grams of the pure acid, when boiled for some time with water gradually dissolved, yielding a clear yellow solution, from which, on cooling, clusters of yellow crystals were deposited ; more of the sub- stance being obtained on shaking the mother liquors with ether, and crystallising the residue left on distilling off the ether.The crystals resembled metamidobenzoic acid in appearance, and melted at the same temperature (174”) ; this was confirmed hy the result of an analysis. Theory for NH2.C6H,.COOH. Fouud. Nitrogen.. ........ 10.21 10.4 Hence on boiling with water metamidobonzoic acid is formed. The fact that a solution of the acid in an alkali, after being boiled and1516 ELLIOTT: ACTION OF CHLOROFORM AND acidified, gave nothing when shaken with ether, is due to the met- amidobenzoic acid formed having combined with the mineral acid to give a salt insoluble in ether. The Pherbylhydrazine Compourtd. Owing to the small yield of this compound, and the fact that it is not very stable, the examination is incomplete, but enough infor- mation has been obtained to wawant a suggestion as to its constitu- tion.Five grams of the pure acid were suspended in alcohol, 3.5 grams of phenylhydrazine added, and the mixture gently heated until solu- tion was complete. After filtering, water was added, and the red precipitate collected and purified by several recrystallisations from dilute alcohol or benzene. It crystallises from alcohol in red needles, which soften and decompose slightly when heated for any length of time at looo, hence the substance should be dried i n a vacuum, The pure dry product melts at 116O, and decomposes with some vigour a t a higher temperature. It is soluble in alcohol, ether, and benzene, and slightly so in water. Found. 7 Theory for --J--- C14H13N4* I. Ir. III. IT. Carbon . . . . , . 71.18 70.75 70.72 - - Hydrogen .. . . 5.08 5.41 5.28 - A Nitrogen . . . . . 23.73 - - 24-11 24.07 This substance, therefore, has the empirical formula C,H,N2, and contains no oxygen. When treated with hydrochloric acid, a dark coloured product is obtained which is easily decomposed by washing with water, yielding the original compound ; it is, however, slightly soluble in the acid, forming a crimson solution. The red compound appears to have, therefore, feeble basic properties ; with dilute sulphuric acid also, it forms a dark coloured product, decomposed by water; strong sulphuric acid dissolves i t with a deep bluish-green coloration. The alcoholic solution dyes a deep yellow, which is turned crimson by acids, but the latter colour is destroyed by water. These facts are accounted for if the substance is considered to be a disazo-compound, with the following formula : and another reason for this formula is given later. In order to ascertain whether any other product was obtained by the action of phenylhydrazine, the original filtrate from the red com- pound was diluted with more water and shaken with ether, and the ether distilled off, The residue, which conkained phenylhydrazine, was crystallised from alcohol and the product purified by several CsK5*N:N*CH:CH*N:N*CsH, ;POTASSIUM HYDROXIDE ON METAMIDOBENZOIC ACID.15 17 recrystallisations from alcohol and from water; it then melted at 173", and, on analysis, was found to contain 10.20 per cent. of nitrogen ; it is, therefore, metamiilobenxoic acid. No other com- pound could be isolated.An attempt was also made to obtain a hydrazone in the usual way, but without result. When the barium salt is suspended in alcohol and heated with phenylhydmzine, the solution becomes deep red, and, on filtering, and adciing water to the filtrate, a red substance is precipitated which has all the properties of the red crystals mentioned above ; the quantity obtained, however, was insufficient for analysis, the yield being small. Conclusion. The results of analysis of the acid formed by the action of chloro- form and potash on metamidobenzoic acid point to the formula CeNH703, but, unlike the aldehydes, it forms no condensation product with strong ammonia solution, and no compound with sodium hydro- gen sulphite solution; moreover, it is insoluble in a,ll solvents. With phenylhydrazine, it yields no hydrazone, but only metamido- benzoic acid and an azo-compound of the formula Cl4Hl2N4, and, when boiled with water and alkalis, it decomposes, yielding meta- midobenzoic acid.These reactions are explained if the following constitution is assigned to this compound :-c7NH602*CH( oH)*co*c7NH60z, that is, diamidodicarboxybenzoh. This constitution will explain the ease with which the compound decomposes. When boiled with water, i t is oxidised, giving metamidobenzoic acid and carbon dioxide. A similar reaction takes place when the compound is boiled with alkalis, With phenyl hydrazine, an osazone is probably first formed according t o the following equation : 2H20 + az, but the hydrogen instead of reducing the phenylhydrazine, as is usually the case, reacts on the osazone and sets free metamidobenzoic acid ; at the same time, a molecular transformation takes place with the formation of a disazo-compound of the formula CeH,*N:N*CH:CH*N:N*C~H~. A similar molecular transformation takes place when phenylhydr- azine acts on the naphthaquinones, azo-compounds being formed instead of hydraxon es.1518 JOKETT ON ATISINE, The react,ions between chloroform, aqueous potash, and various other aromatic compounds, such as phenylhjdrazine and metanitro- benzoic acid, are now being investigated, and I hope to be able to communicate the results t o the Society in a short time. My thanks are due to Mr. F. Beames, R.Sc., for help rendered during the courss of this research. !The G?*amnnr ,School, BristoZ.