SEPARATION OF ISOMERIDES CONTAINED IN XYLIDINE. 65 VL-A Method of Separation of the various Isomerides contained in ordinary co mmercia 1 Xy lid h e . * By W. R. HODGKINSON and LEONHARD LIMPACH. ORDINARY technical, or commercial, xylidine, obtained by the nitration and subsequent reduction of the coal tar xylene distilling between * This is a continuation of some work on xylidines and xylenols (compare Trans., 1892, 61, 420 ; 1893, 63, 104). VOL. LXXVLI. F66 HODGRINSON AND LIMPACH : A METEOD OF SEPARATION OF 136' and 142', contains, with the exception of the symmetrical 5-arnino-nz-xylene, all the possible isomerides from the three xylenes. The main constituents are the 4-amino-na-xylene and the 2-amino-p- xylene, which together make up about 70 per cent, of the total bases, In the present paper, an account is given of a process for effecting a complete separation of all the isomerides present in commercial xylidine, making use of the method aIready devised by one of the authors for the isolation of 4-amino-m-xylene as acetate, and of 2-amino-p-xylene as hydrochloride (Germ, pat., 39947, of Septembeic 19th, 1886; Eng.pat., 11822, 1886), and of 2-amino-rn-xylene by means of its formyl derivative (Ber., 1899, 32, 1008). 4- Amino-m-xylens. When acetic acid is added t o ordinary xylidine, the acetate of the 4-amino-rn-xylene is the only one to crystalliae out, The method we have worked out, and employed t o obtain the different xylenols, de- pends in the first instance on this fact. As an example, 300 gramv of commercially pure xylidine were treated with slightly more than the necessary amount of glacial" acetic acid to convert the 4-amino-rn-xylene into its acetate,? The crystallisa- tion of the acetate commenced a t once, and was practically completed after 24 hours standing.The crystals were separated in a centrifugal separator (a filter pump would be equally effective), and, after press- ing, were found to be almost pure 4-amino-m-xylene acetate. As a rule, the yield of the base by this method amounts to from 40 to 42 per cent. of the original xylidine. The acetate was hydrolysed with sodium hydroxide, and the base converted into the formyl derivative, which, after the first crgatal- lisation from dilute alcohol, melted a t 113.5'. 2 -A mino- p-x y Zene . The filtrate and mother liquors contain a considerable amount of the 2-amino-p-xylene 2 and smaller quantities of 3-amino-o-xyIene, 4-amino-o-xylene, and 2-amino-m-xylene.To this mixture, hydrogen chloride (ordinary strong solution) was added until the whole dissolved. On cooling, the 2-amino-p-xylene hydrochloride crystallised out, and after a couple of recrystallisations was obtained quite pure. * Glacial acid is not absolutely necessary ; 80 per cent. acid acts quite well, but the product takes a longer time to cryatallise. t This can be ascertained sufficiently nearly on a very sinall scale, '$ This is valuable for the production of pseudocumidine.THE VARIOUS ISOMERIDES CONTAINED IN XYLIDINE. 67 It may be remarked that the product obtained in the first crystal- lisation is sufficiently pure for conversion into pseudocumidine.Some of the product from the first crystallisation was made into the formyl derivative. This melted a t 116-117', aEter the first crystallisation from weak alcohol. As indicated by this method, about 30 per cent, of this isomeride is, as a rule, contained in ordinary xylidine. 2-Amino -m-xy Zene. The residue (filtrate) from the para-base now contains the two ortho- and the meta-isomerides. We find that the 2 -amino-m-xylene can be very completely separated in the following manner. The mother liquors are evaporated in a retort. After the excess of acid and water has distilled off, the contents of the retort become semi- solid. On now gently heating, the hydrochloride of 2-amino-m-xylenc sublimes in very beautiful, silky, needle-shaped crystals. One re- sublimation gave a perfectly pure product.From 300 grams of xylidine we have obtained 1OA6 grams of this salt, indicating the presence of 2.7 per cent. of this isomeride. Another method, adapted for larger operations, is to distil off about 20 per cent. of the contents of the retort by heating rapidly; the distillate then solidifies very quickly. It can then be converted into the formyl derivative, and easily be obtained perfectly pure by one cry stallisation. The formyl derivative of 2-amino-m-xylene crystallises in very beautiful, long needles which, when slowly heated in the usual way, melt at 164O, but when rapidly heated, melt only at 176-171'. It is completely hydrolysed after boiling for 6 hours with alcoholic potassium hydroxide, The acetyl derivative melts a t 177", and also crystallises in long needles.The base distils completely a t 216' under 735 mm. pressure (therm. in vapour), and has a sp. gr. 0.980 at 15'. From this isomeride, we have obtained, by diazotisation, the m-2- xylenol." It crystallises from water, in which it is somewhat soluble, in long, hard crystals, and from dilute alcohol in long prisms. It melts at 47' and boils at 199.6' (therm. in vapour) under 738 mm. pressure. * We have prepared a number of derivatives from this xylenol, which we hope to describe, with others, later.68 SEPARATION OF ISOMERIDES CONTAINED IN XYLIDINE. 3-Amino-o-xylena. The black residue in the retort, from which the 2-amino-m-xylehe hydrochloride had been sublimed, consisted of the hydrochlorides * of the two ortho-bases.? It was dissolved in water, and the bases, set free by the addition of excess of sodium hydroxide, were separated from the alkaline liquid, distilled in steam,$ and then converted into their formyl derivatives.This is particularly easily accomplished by the simple addition of the necessary amount of formic acid, of 28 per cent. strength, and heating on a water-bath. The two formyl compounds were now separated by taking ad- vantage of a slight difference in their rate or facility of crystallising. After standing for a few days, the mixture became semi-solid; the liquid portion was then pressed out, and the solid portion recrystal- lised. They were hydrolysed by alcoholic sodium hydroxide, and the base converted into the acetyl derivative, which melted a t 132'. This base, therefore, is 3-amino-o-xylene, and the amount contained in xylidine amounts roughly t o 9-11 per cent.The first batch of crystals melted a t 100-102°. 4-Amino-o-xylene. The mother liquor pressed from the formyl derivative of 3-amino-o- xylene consists for the most part of that of 4-amino-o-xylene, with small quantities of the other bases which have escaped separation. It is very difficult to induce this formyl derivative to crystallise. Cooling with a freezing mixture to -20' for some time was without result. It was therefore hydrolysed, and the base separated. This dis- tilled over completely at 222' (therm. in vapour), and after standing for 3 or 4 months during the winter, formed large, monoclinic crystals, which melted at 48'. A formyl derivative prepared from them was induced to crystallise by considerable cooling and agitation, and the base obtained from this crystallised material by hydrolysis melted at 5 2 O . There is therefore no doubt it is the 4-amino-o-xylene. We have prepared considerable quantities of the xylenols corre- sponding to these xylidines, and will shortly give an account of some of their derivatives. WOOLWICH AND ERLANGEN. * For some purposes these may be used without further purification. -t About 20 to 25 per cent. of these isomerides is contained in ordinary xylidine. $ These bases also distil alone without decomposition,