1076 MELDOLA AND EYRE: CXII1.-Additional Notes on Dinitloo-o-anisidine. A Chemical Reaction in which one of the Products continues the same Reaction. By RAPHAEL MELDOLA, F.R.S., and JOHN VARGAS EYRE. THE dinitroanisidine described last year by one of the authors and Elkan Wechsler (Trans,, 1900,77,1172) has been since shown to have the constitution 4 : 5-dinitro-2-aminoanisole (Proc., 1901, 1’7, 131 ; also Freyss, Bull. SOC. had. Mulhouse, 1901,70, 375). Further evidence of this constitution is given in the following paper as well as some experiments which throw light on the remarkable action of nitrous acid on the compound. The Diphenyhxines from the Triaminomisole and its Acstyl Dsrivatives. Dinitroacetnnisidide (m. p. 1 62-163O) was reduced in glacial acetic acid solution with zinc dust and a few drops of hydrochloric acid to start the reaction. The solution containing tbe acetyltriaminoanisole was mixed with an acetic acid solution containing the calculated quantity of benzil, and the mixed solutions heated for some hours on a water- bath.The azine is thrown out on dilution with water, and neutralisa- tion with ammonia as an ochreous powder which dissolves in boiling alcohol with a brownish colour. The solution is slightly fluorescentADDITIONAL NOTES ON DINITRO-0-ANISIDINE. 1077 when cold, awl the azine separates in the form of ochreous needles melting at 823-224". 0.1 163 gave 11.8 C.C. moist nitrogen a t 20.5"and 764 mm. N = 11.62. 0.1186 ,, 11.75 ,, ,, 19.5O ,, 752.4 mm. N = 11-33. C,,H,,O,N, requires N = 11.39 per cent. The dinitroanisidine on siiiiilar ti eatment gave an azine cryst allising from alcohol in brown needles with serrated edges, and from benzene, in which i t is very soluble, in minute, p,de, ochreous needles.The melting point is 194-195". 0.1092 gave 12.2 C.C. moist nitrogen at 18.5Oand 755.S mm. N = 12-78. 0.0903 ,, 10 $ 9 9 , 20.5" ,, 770.5 mm. N = 12.80. C2,HI7ON, requires N = 12*S4 per cent. The dilute alcoholic solution of the azine has a distinct green A UoreFcence which disappears on heating, and reappears on cooling. The compound is basic and dissolves in strong hydrochloric acid with a claret red colour, which disappears on dilution with water, owing to the dissociation of the salt. The constitution of these compounds is shown by the formulae : NH A~()N : ~ c , H , NH,<)N : ~ c , H , C€I,*O\/N :C*C,H;* C! Li3*O\/N:C*C,H,* The mine ring may of course be represented in the usual way with a cross linking between the nitrogen atoms : or the left- hand benzene ring may be written on the quadrivalent (quinonoid) type : :N*E*C,H, :N*C*C,H; *r *g*C,H, *N.C*C,€I,' Quctnlhtive Diaxotiscdion of L)initi*o-o-anisidine.It has been proved in former p p e r s that this compound loses a nitro-group on diazotisation, and we have shown in our last note (Proc., 1901, 17, 131) that the nitro-group thus eliminated i s the one occupying the para-position with respect t o the amino-group, the resulting compound being a diazoxide. I n the note referred to, we suggested that the nitro-group might be eliminated in the form of nitrous acid according to the scheme : N02*C,H,((>CH,)< N;O NO, H -+ N02-C,H2(OCH,)<f? - .If this view were correct, me should have the somewhat remarkable case of a chemical reaction in which one of the products (nitrous acid) V O I . L X X J X . 4 D1078 MELDOLA AND EYRE. is the same a s the reagent added, and is thus capable of carrying on the diaaotisation. ID order to submit this to the test of experiment, a method had to be devised for measuring quantitatively the amount of diazotising work done by a known quantity of nitrous acid. Pre- liminary experiments with a standardised solution of sodium nitrite, and fiolutions of known strength of dinitroanisidine in glacial acetic acid, showed that the ordinary method of ascertaining the end of diazotisation by the liberation of free nitrous acid was quite inapplic- able in the present case.Free nitrous acid, as shown by the potassium iodide and starch test, shows itself from the very first addition of the nitrite solution, and does not cease to be present until after some days. I n other words, the diazotising process has no time limit sharp enough to be fixed by any colour test. 'Under these circumstances, we were led to use the gravimetric method formerly applied with success in the case of the quantitative resolution of the diazoamido-coinpounds (Meldola and Streatfeild, Trans., 1887, 51, 438 ; 1888, 53, 675). The method as applied to the present reaction depends upon the following conditions : I f sodium nitrite solution is added to an acetic acid solution of dinitroanisidine, keeping the latter in excess, there are present after a certain interval (1) diazoxide, (2) unaltered dinitroanisidine, and (3) nitrous acid.After a sufficiently long interval (about 3 days), free nitrous acid is no longer detectable, but even if free nitrous acid is present this does not interfere with the results. On adding such a solution to an alkaline solution of P-naphthol, the diazoxide combines at once to form the azo-compound, N02*C6H2( OCH,) (OH) *N2* ~loH6*oH.* On making strongly acid with hydrochloric acid, the free azo-compound is precipitated, whilst the unaltered dinitroanisidine remains in solu- tion if a sufficient volume of water is present. The azo-compound, being practically insoluble in water (even when hot), can be collected on a tared filter, washed with dilute acid, and finally with hot water, until free from all soluble compounds, and then dried and weighed.The weight of azo-compound gives the weight of diazoxide formed. Two sets of experiments mere made, using the following quantities : A.-First Set.-Twenty-five C.C. of a 1 per cent. solution of dinitru- anisidine in acetic acid. Sodium nitrite solution 1 C.C. = 0.0029 gram NaNO,. Nitrite solu- tion 1 C.C. = 0*00906 gram NaNO,. B.--Xecond 8et.-Same quantities of dinitroanisidine. * The formula of this compound given in the preliminary note (Trans,, 1900, 77, 1173) does not represent it as containing a hydroxy- as well as B methoxy-group, because the nature of the reaction had not been a t that time fully made out. The percentage of nitrogen does not differ considerably in the two cases.ADDITIONAL NOTES ON DINITKO-O-ANISIDINE.1079 The quantity of dinitroanisidine was the same throughout, 0.25 gram, and the same quantity of nitrite, 0.0202 gram, was added in each case, namely, 6.98 C.C. in set A, and 2-23 C.C. in set B. This quantity of nitrite is onefourth the calculated quantity required on the assump- tion that one molecule of dinitroanisidiiie requires o m molecule of nitrite. The results are given below : Dinitro- anisidine. Weight of 1 Weight of 1 Nitrite' 1 &naphthol. azo-compound. * f I ... 0.25 I 4 d R V S I 0'0202 ' 0.3498 028C7 1 0.3195 9 2 1 0'3bO6 v ... I 9 1: ~ 0.3694 9 , I I 9 2 ... I r: I The P-naphthol is slightly in excess of that required by theory, namely, 0.2 instead of 0.17 gram, but this excess is completely re- moved during the treatment.The quactity of azo-compound thcoreti- cnlly producible by the weight of nitrite taken (0.0202 gram) is 0.0992 gram, The weight of azo-compound theoretically producible on the assumption that all the dinitroanisidine is converted into diazoxide is 0.3968 gram. Thus che results show t h a t whilst the theo- retical limit of diazotisation is not reached in five days, the actual amount of diazoxide formed is greatly in excess of that capable of being produced by the weight of nitrite added. The nitro-gyoup eliminated t Jt,us continues the diaxotisation. The method, although not claiming to give results within very close limits of accuracy, is sufficiently exact to bring out this main conclu- sim in a very striking manner. The only sources of error that can be foreseen are the solubility of the azo-compound in water leading to a deficiency and the possible retention of dinitroanisidine leading to an excess in weight. The first of these errors we believe to be quite negligible. The second was provided against by washing the azo- compound (after being collected) off the filter, redissolving in hot dilute sodium hydroxide, reprecipita ting by acid, and collecting and washing again before drying and weighing. The azo-compound in each experi- ment was thus twice precipitated from a large volume of dilute acid, and although a slight loss may have been incurred by this treatment, the final product was pure azo-compound as shown by analysis. 0.1174 of No. I1 gave 12.85 C.C. moist nitrogen a t 21° and 757.4 mm. N = 12.40. C,7H1,0,N3 requires N = 12.39 per cent. FINSBURY TECHNICAL COLLEGE. 4 ~ 2