202 XXIII --Ox Aroma& Nidrosamifies. B ~ O T T O N. WITT, Ph.D. IK 1864 Griess published in the Philosophical Transactions his great paper on the diazo-compounds, in which he so thoroughly described the action of nitrous .acid on primary aromatic amines that subsequent research has brough% to light no new facts of importance. In 1874 I made a few experiments on the action of nitrous acid, and especially its ethers, on secondmy and tertiary amines, and was aston- ished to find that very beautiful substances were foimed in these reactions. A few weeks after I had communicated my results to the Chemical Society of Zurich, B ae y e r and Car o published their paper on nitrosodimethyianiline, the final product of the action of amylic nitrite on dimethylaniline; I have therefore in my research con- fined myself to the action of nitrous acid and its salts on secondary %mines.The facts observed are shortly recorded in the Bem'ciite of the Berlin Chemical Society ; and a detailed account of the formation and properties of diphenylnitrosamine has been given in my inaugural dissertation. Since the publication of the latter, in 187.5, I have studied more carefully some complicated reactions of diphenylnitro- samine, and the object of this paper is to bring before the Society a description of the phenomena observed, of the new substances pre- pared, and a short statement of my theoretical views on the subject. Whenever a secondary amine is acted upon by nitrous acid, water and a nitrosamine are formed. The term 4' nitrosamine " I apply to any substituted ammonia which contains, instead of at least one atom of hydrogen, the univalent nitrosyl group, -NO, in immediate con- nection with the ammoniacal nitrogen.Thus the action of nitrous acid on diethylamine forms diethylnitrosamine- C*& C2H5 I I NH I I CZH5 + NO.OH = HZ0 + N.NO; c2H5 whilst diphenylnitrosamine is obtained by bringing diphenylamine into contact with ethylic or amylic nitrite- C6H5 C6H5 I I NH + NO.OCzHs = N.NO + C2H,.OH, I c ' a 5 I ca5WITT ON AROMATIC NITROSAMINES. 203 or by the action of nitrous vapours on diphenylamine, or of metallic nitrites on the hydrochloride of diphenylamine. Diphenylnitrosamine is a substance remarkable for its great power of crystallisation. It can be easily obtained in large honey-coloured monoclinic crystals, melting at 665", easily soluble in alcohol, ben- zene, and similar solvents, and giving a striking reaction with concen- trated snlphuric acid, in which it rapidly dissolves, imparting to it a beautiful blue coloration, large quantities of pure nitric oxide gas being given off.Powerful reducing agents cause the formation of ammonia and reproduction of diphenylamine ; but by careful reduc- tion it may be converted into diphenylhydrazine, one of a beautiful series of new bases recently discovered by Emil Fischer. The action of aniline, or any other primary monamine, on diphenylnitros- amine iE1 violent; if carefully controlled, however, the result is simple enough, being represented by the following equation :- c6H5 I ca5 CsH, N I I II N.NO + 2(NH,.CsHa) = NH + N + HzO.I I I CSH, c& NH I A corresponding amount of diphenylamine is regenerated, whilst the nitroso-group is employed in the formation of diazobenzene, which immediately combines with another molecule of aniline, forming diazo- amidobenzene. The latter then undergoes the usual ti-ansformation into amidoazobenzene. Under certain circumstances a secondary re- action takes place, giving rise to a beautiful substance of the formula C,H& which is remarkable for a peculiar reaction. If heated with concentrated sulphuric acid to the boiling point of the latter, a, fine blue colonring-matter is obtained, possessing a beautiful crimson fluoremence and a characteristic absorption spectrum with three dark bands. If, instead of pure ethylic or amylic nitrite, the ordinary prodnct obtained by leading nitrous vapours into ethylic alcohol be employed in the preparation of diphenylnitrosamine, or if nitrous vapours be introduced into a solution of diphenylamine, the products obtained differ greatly from diphenylnitrosamine, and a few experiments suffice to prove that they are mixtures of different substances in varying proportions.Their separation is, however, a matter of extreme diffi- culty : for gradual decomposition sets in if repeated recrystallisation is resorted to, and it was not until I found a means of preparing each of the different substances, independently and separately, that I could204 WITT ON AROMATIC MTROSAMINES. attempt a successful inquiry into their constitution. Here I encoun- tered another serious difficulty : all these substances, when subjected to organic analysis, suddenly give off large quantities of nitrous vapours; and this takes place so ra;pidly that even the use of a number of spirals of metallic copper does not always completely de- compose the gas.On the other hand, the introduction of a large amount of metallic copper is detrimental to the accumy of the analysis, as recently pointed out by Lietzenmayer (Ber., xi, 306). When the nitrous vapours are given off a, peculiar kind of very denRe charcoal remains behind, the final combustion of which can be aceom- plished only by the prolonged passage of oxygen through the tube. Owing to all these difficulties my analytical data are marcely within such close limits as I could have wished, although I have made quite an unusual number of combustions with the greatest care and by methods which, from long experience, I know to be perfectly tmst- worthy.On the other hand, I have in almost all cases determined not only carbon and hydrogen, but also nitrogen by the volumetric method. It will be easily understood that analyses attended with such serious difficulties could not suffice to clear up the nature of my new sub- stances, and I therefore had recourse to a closer study of their mode of formation. I soon found that ordinary ethylic nitrite contains invariably large quantities of nitric acid, the presence of which was easily accounted for, when I convinced myself that none of the methods employed for the production of what is commonly called nitrous anhydride, N203, produce this compound; and that by far the hrger portion of the product obtained by the action of nitric acid of different strengths on either arsenious anhydride, or starch, or treacle, is nitric tetroxide, N204, which may be regarded as a mixed anhydride of nitric and nitrous acid.In contact with water or alcoliol it furnishes both nitrite and nitrate, and this accounts for the invariable presence of nitric acid in all nitrous ethers prepared by the introduction of nitrons vaponrs into the corresponding alcohol. The nitric acid present proved also to be the cause of the peculiar difference between the action of pure and ordinary ethylic nitrite on diphenylamine. This conclusion arrived at, I completely abandoned the use of ordi- nary nitrite, and with the purpose of obtaining a thorough insight into my experiments only employed mixtures of known quantities of pure nmylic nitrite and nitric acid ; acting with this reagent upon weighed quantities of diphenylamine and moderating or accelerating the reac- tion by the use of various solvents.I f 20 grams of finely powdered diphenylamine be placed in a dry flask and a mixture of 15 C.C. of nitric acid, sp. gr. 1.424, 35 gmms of pure amylic nitrite, 100 C.C. of methylated alcohol,WITT ON AROMATIC NITROSAMINES. 205 be poured upon it, the diphenylamine dissolves instantaneously with a dark coloration, and a rise of temperature is perceptible; the flask is kept in constant motion, and the reaction, if necessary, accelerated by gentle heating. Suddenly crystals begin to settle out ; and if the flask be now placed in iced water, a beautifully crystalline deposit i R obtained.When crystals cease to form, the precipitate is filtered off and carefully washed with spirit, dried and recrystallised from chloroform. It forms beautiful plates of light yellow colour fusible a t 133.5", easily soluble in chloroform, benzene, hot alcohol, and glacial acetic acid. The analytical data obtained with this substance are the follow- ing :- I. 11. 111. IV. V. VI. VII. C . . . 39-90 59.70 59.84 60.03 59.60 - - H .. 420 4.30 430 440 440 - - N... - c - - - 18.60 18.10 This substance may easily be recognised as a nitrosamine ; for it imparts a fine violet coloration to pure concentrated sulphuric acid, at the same time giving off nitric oxide gas.Now as the three available ahmicities of the nitrogen in diphenylamine are thus proved to be engaged in a manner identical with diphenylnitrosamine, the only possible conclusion was, that some substitution had taken place in one or both of the two phenyl groups of the diphenylnitrosamine, and that the new substance was the nitrosamine of a substituted diphenyl- amine. We possess, as I have shown before, a means of regenerating the secondary amine from a nitrosamine by the action of aniline, which was resorted to and proved to be effectual ; and equally good results were obtained by splitting off the nitroso-group by means of alcoholic potash. Whichever process was applied, it resulted in the formation of a new substance crystallising from dilute alcohol in small glistening scales of a pale yellow colour, melting at 132". This substance does not colour sulphuric acid, but it dissolves with a fine red colour in alcoholic potash ; subjected to organic analysis i t proved to be mono- ni trodip hen y lamine : - Theory for Experiment.CSHpNH.CBEd.NOZ. I. 11. C ........ 67.20 67.20 67.32 H ........ 4.66 5-06 5.35 N ........ 13.09 0 ........ 14.95 99.90 - - - - -206 WITT ON AROMATIC NITROSAMINES. The product, from which this mononitrodiphenylamine wm ob- tained, is therefore mononitrodiphenylnitrosamine ; the following num- bers- C .................... 39-30 H .................... 3.70 N .................... 17.30 0 .................... 19.70 CsH5. N . NO. C sH4. N 0 2 = IOO*OO indicated for this substance by theory agree as closely with the results actually obtained rn can be expected under the dificulties which the analysis of these substances offers.Mononitrodiphenylnitrosaine, when dissolved in glacial acetic acid and subjected to the action of bromine, yields a precipitate of a canary-yellow colour, which is easily soluble in benzene and crystal- lises from this solvent in long silky needles fusible at 208-5-209". If this substance be again treated with bromine, and heated with it for some time on the water-bath, another product is formed crystallising from benzene, in which it is not so easily soluble as the first product, in heavy, well-shaped small prisms, melting at 214-5-215". Mononitrodiphenylamine is a colonring matter, though a very weak one. The colonring properties of this substance are due to the entrance of a nitro-group into diphenylamine, which at the same time renders the imido- hydrogen replaceable by metals- Silk may be dyed yellow in its solution.Mononitrodiphenylamine. Potaeeium mononitro- dipheny lamine. But as soon as the imido-hydrogen is replaced by the nitroso-group, the colonring properties of the compound are destroyed, the influence of the nitro-group being paralysed. This may be easily demonstrated by introducing mononitrodiphenylnitrosamine into a very cold solution of potash in alcohol. In the first instance no coloration is observed. But on heating, or even on standing for a short time, a dark scarlet coloration appears, due to the elimination of the nitroso-group and the formation of mononitrodiphenylamine, which, as already observed, is a colonring matter.If the proportions of nitric acid and amylic nitrite be altered, and the spirit be exchanged partly or entirely for another solvent, whichWITT ON AROMATIC NITROSAMINES. 207 expedites the reaction, more nitro-groups may be introduced into the molecule of diphenylnitrosamine. The following method has been found to give satisfactory results. 17 grams of diphenylamine are introduced into a dry flask sur- rounded by iced water. Then a mixture of- 50 C.C. of glacial acetic acid, 40 C.C. of nitric mid, 1.424, 50 C.C. methylated spirit, 48 grams nitrite of amyl, is added all at once. The diphenylamine dissolves with a dark colour, and very soon small granular crystalline aggregates begin to settle out.When the precipitate no longer increases, it is filtered off on the filter- pump, the mother-liquor is carefully removed, and the precipitate washed with spirit and finally with ether. It is then dried at the ordinary temperature, and appears, when dry, as a sandy powder of light yellow colour. Recrystallisation may be effected from chloro- form, in which the new compound is but sparingly soluble, but I have not been able to detect, any difference in the analytical results ob- tained with either the recrystallised or the crude product. Their behaviour with reagents and their characteristic appearance are also very much the same. The reaction of this body with sulphuric acid, in which it dissolves with a dirty violet coloration, giving off nitric oxide at the same time, proves it to be a nitrosamine.The analytical results, however, do not agree with any simple formula, but the num- bers for at least the carbon and hydrogen are intermediate between those required for di- and for tri-nitrodiphenylnitrosamine- Theory for Experiment. C12HSN405* CI,H;N,O,I. Crude. Recrystalbed. I. 11. 111. 0.. .... 50-00 43.25 46.95 - 47.01 H...... 2.75 2.10 3.99 - 3.18 19-04 - N.. .... 19.45 21.05 - 0.. .... 27.80 33.60 - - - - 100*00 100*00 I therefore concluded that the new substance, although apparently uniform and not separable by repeated recrystallisation, was contami- nated with some impurity, and hoped to effect a more successful puri- fication on removing the nitroso-group by means of alcoholic potash or aniline. For this purpose 20 grams of the product were placed in a flask, together with 100 C.C.of spirit, and 50 C.C. of a 20 per cent. solution of potassic hydrate in alcohol. On heating, a splendid purple coloration208 WITT ON AROMATIC NITROSAMINES. appeared, which gradually increased in intensity until all was dis- solved. Dilute hydrochloric acid was now added until the colour had changed to a bright orange, and, finally, a large quantity of water was introduced; the precipitate was collected on a calico filter, washed carefully, pressed, and treated with about 400 C.C. of boiling methylic alcohol. The solution was filtered from the part which remained undis- solved (A), and on cooling deposited crystals of an obviously impure product (B). If aniline be employed for the removal of the nitroso-group, the following method may be adopted :- 20 grams of the new product, 30 grams of pure hydrochloride of aniline, 25 grams of aniline, 50 C.C.of spirit, are hated for about half an hour on the water-bath. Plenty of dilute hydrochloric mid is then added, and the precipitate boiled out three successive times with extremely dilute nitric acid. All the amidoazo- benzene is thm removed, and may be found in the extract. What remains undissolved is treated with methylic alcohol as before, and the part insoluble in the boiling alcohol (A) treated separately from that which deposits on cooling (B). The latter, B, is successively extracted with small quantities of boiling methylated spirit and glacial acetic acid, and both solutions removed by hot filtration.After this treatment the new substance requires one crystallisation from ethylic, or, better, isobutylic alcohol to be quite pure. It then forms long, thick, pointed needles of great brilliancy and lustre, of a dark yellow colour, with a slight bluish dichroism, fusible at 214". They dissolve in alcoholic potash with a bright purple colour, which is destroyed by the addition of cold wafer, but reappears on boiling. Analysis proved them to be dinitrodiphenyl- amine :- Theory Experiment. For C12H9N304. I. 11. 111. IV. C ........ 55-60 56.50 54.90 56.39 - H ........ 3.50 4.80 409 4.07 - N.. ...... 16-20 - - - 16.M 0 ........ 24.70 - - - - The substance A, insoluble in methylic aloohol, was boiled out two or three times with this solvent, and then recryshllised from pure sylene.From this hydrocarbon it is deposited elowly in small, indistinctWITT ON AROMATIC NITROSAMINES. 209 granular crystals of a vermillion colour, fusible at 211.5'. It dissolves in alcoholic potash with scarlet colour, and the solution behaves in a similar manner to the purple solution of the body first described. On analysing it I found that if was isomeric with the yellow substance, and corresponds also to the formula of dinitrodiphenylamine :- For ClJ&,Na04. I. 11. In. IV. C 36.80 56-32 - - ........ 55-60 H.. ...... 3.50 4.25 4.17 - - Nb... ..... 16.20 - - 9790 16.29 0. ........ 24.70 Theory Experiment. - - - 100~00 This fact ie by no means surprising. If we suppose that, in the dinitrodiphenylamines obtainable by direct nitration, the two nitro-groups enter two different phenyl-group, there are still three dif€erent anbstanm likely to arise :- NO* 0 NH Dipamnitro- Diortbnitro- Orthapmnitm- diphenylamine.diphenylamine. diphenylamine. The metaposition is out of the question, ~ E J a nitro-group entering into an a;l'omatic molecule never occupies a meta-position to a, basic group already present. The two dinitrodiphenylamines here described most likely correspond to the firat and third formulae, for they both arise from further nitration of the same mononitrodiphenylamine, which, in all probability contains its one nitro-group in the p m - position. All the substances hitherto described give, when treated with nitric acid of the highest concentration, hexanitrodiphenylamine, a sub- stance originally discovered by E.Kopp, more closely investigated by Gnehm, and repeatedly introduced as an orange dye into commerce. On careful nitration diphenylnitrosamine gives, as I observed some time ago, and Gn e hm recently published, tetranitrodiphenylamine. Thus diphenylamine, through the agency of its nitrosamine, aliows of210 WITT ON AROMATIC NITROSAMINES. the introduction of one nitro-group after the other up to six, with exception of three and five, into its molecule, and we now know, irrespectively of isomerides obtained with picric chloride and dinitro- chlorobenzene by Clemm and P. T. Austen, the following nitro- derivatives of diphenylamine :- Mononitrodiphenylamine, C12H,,,N202. Dinitrodiphenylamine, CI2H9N,Oa (two iaomers). Tetranitrodiphenylamine, Cl2H7N5O8.Hexanitrodiphenylamine, Cl2K5N7Ol2. I do not doubt that my new method of nitrating substances in dilute solutions with the aid of alkylic nitrites, will be found applicable in a, great many cases, where nitration has hitherto offered unusual diffi- culties. It would also appear that this method furnishes some sup+ port to the theory already advanced, that the formation of nitroso- compounds always precedes the introduction of the nitro-group into the molecule of aromatic substances. Thus far I have described as much of my research on the action of nitrous acid on secondary amines as has become fit for publication since my last communication to the Berichte. I should have preferred to wait until I could have given some account of the reduction-pro- ducts of my new nitro-compounds, and investigated more closely their brominated derivatives, had it not been for an incident which made it desirable to publish without delay.In February, Mr. Meldola communicated to the Chemical News a note on “A New Colouring-matter,” and, at the same time, was good enough to send me two specimens of his preparations, one represent- ing the substance arising from the action of nitric tetroxide on diphenylamine in an acetic acid solution, the other the product obtained from the first by the action of aldoholic potash. On closely examin- ing and comparing our specimens, we found the first product to repre- sent a mixture of the nitrosamines described in this paper, whilst the second was a mixture of the nitrated diphenylamines, which corre- spond to the nitroso-compounds. On ascertaining that the subjects of our investigations were identical, and that the action of nitrous acid on diphenylamine and other secondary amines had occupied me for several years, Mr. Meldola was kind enough to relinquish his claim to the continuation of this research, for which I have much pleasure in thanking him : at the same time I wish to’ point out that the prac- tical application of mono- and di-nitrophenylamine as a dye-stuff, called citronine, is solely and entirely due to Mr. Meldola. The shades obtained on silk with citronine are truly magnificent, and no other artificial colour can compete with it as regards this particular tint ; unfortunately, owing to the inefficiency of tho nitro-group as a chro- mophor, and to the extreme weakness of the salt-forming groups inPERKIN ON SOME NEW DERIVATIVES, ETC. 211 citronine, it is not a strong colonring-matter, and this, combined with its insolubility in water, renders it of less general applicability than it otherwise deserves. Still, it is a very interesting fact, that in citro- nine an artificial colonring-mtter has for the first time been found, which, if it were not for the little difficulties just mentioned, might with advantage replace turmeric and qnercitron. I have the intention of bringing before the Society, in a second paper, the results which 'I hope to obtain in the continuation of this research.