THE FORMATION OF 2 3 6-TRINITROTOLUENE ETC. 1615 CLXXXI.-The Formation of 2 3 6-Trinitrotohene in the Nitration of Toluene. By ROYSTON BARRY DREW. IN the mixture of trinitrotoluenes obtained by the direct nitration of toluene previous workers have identified only 2 4 6- 2 3 4-, and 3 4 6-trinitrottoluenes of the six possible isomerides. Molinari and Giua (Zeitsch. ges. Schiess. ‘16. Sprengstoffw. 1914 9 239) claimed to have isolated the 2 3 6-compound previously unknown, to which they assigned a nielting point of 79*5O but this com-pound has been synthesised by Korner and Contardi ( A t t i R. Accad. Lincei 1916 [v] 25 ii 339) and by Brady and Taylor (this vol. p. 876) and found to melt at 1 1 1 O ; the latter authors have suggested that Molinari and Giua’s compound was probably a molecular compound of 2 3 4- and 3 4 6-trinitrotoluenes.Will (Ber. 1914 47 704) has studied the nitration of 2:3- and 3 6-dinitrotoluenes and states that the sole products of their nitra-tion are 2 3 4- and 3 4 6-trinitroto1uene.s. More recently how-ever Marqueyrol Koehler and Jovinet (Bull. SOC. chim. 1920, [iv] 27 420) have obtained a third trinitrotoluene in addition to the 2 3 4- and 3 4 6-compounds from the nitration of m-nitro-toluene. Their results are incomplete and they have not identified the compound but as they found it was formed to a small amount in the nitration of 2 3- and 2 5-(or 3 6-)dinitrotoluenes they conclude it must be either the 2 3 6- or the 2 3 5-trinitro-deriv-ative. The present author previous to the publication of Marqueyrol Koehler and Jovinet’s results had investigated the nitration of 2 3- and 3 6-dinitrotoluenes and found that in eaoh case the product contains about 15 per cent.of 2:3:6-trinitro toluene. Other workers have shown that about 4 per cent of m-nitro-tduene is formed in the mono-nitration of toluene under the usual conditions and that on further nitration of this compound about half the product is a mixture of 2 3- and 2 5-dinitrotoluenee. In crude trinitrotoluene obtained by the direct nitration of toluene, there must be therefore about 0.3 per cent. of 2 3 6-trinitro-toluene. Opportunity however has not arisen to isolate this small amount from commercial samples of trinitrotoluene. Further a new method of preparing 2 3 6-trinitrotoluene has been devised which is easier than that described by Korner and Contardi (Zoc.c i t . ) and starts from trinitro-m-crewl ; this is mor 1616 DREW THE FORMATION OF 2 3 6-TRtNITROTOLUENE readily obtainable than the 2 6-dinitrotoluene required in Brady and Taylor's synthesis (Zoc. cit.). The reactions involved are aa follows : Methylpicramic acid obtained by the reduction of trinitro-m-cresol has been oriented only indirectly by Borschc and Heyde (Ber. 1906 39 4092) and the above synthesis fully confirms the view of these authors. E X P E R I Y E N T A L . Nitration of 3 6-Dinitrotoluene.-Forty grams of 3 6-dinitro-toluene were added to 250 C.C. of a mixture of sulphuric and nitric acids (HNO = 17.5 ; E2S0 = 78 ; H,O = 4.5 per cent.).The mix-ture was mechanically stirred and the temperature raised to 80° and kept at that point for an hour when the temperature was raised to looo for a further two hours and finally to 120° for fifteen minutes. The mixture was then cooled and poured into water the solid collected and washed repeatedly under water with the injection of steam until free from acid. The freezing point of the product after drying was found to be 95-4O as against 103O for pure 3:4:6-trinitrotoluene. The mixture was recrystallised from acetic acid to remove most of the 3 4 6-trinitrotoluene and by diluting the filtrate a mixture richer in 2 3 6-trinitrotoluene was obtained. Attempts were made to separate this mixture by crystallisation from various solvents but without success a mixture melting constantly at 7 8 O being obtained.A portion was treated with alcoholic ammonia in the hope that i t would be possible to separate the corresponding dinitrotoluidines but an oily intractable product was obtained. A successful separation was ultimately brought about by the action of hydrazine hydrate. It had been observed by Brady (ppivate rommum'cation) that hydrazine hydrate reacts readily with 2 3 4-and 3 4 6-trinitrdtoluenes giving red crystalline compounds sparingly soluble in alcohol. It has been found that 2:3:6-tri-nitrotoluene behaves in a different manner towards this reagent IN THE NITRATION OF TOLUENE. 1617 which can therefore be used to effect a separation. The mixture of 3 4 6- and 2 3 6-trinitrotduenes was dissolved in methyl alcohol a solution of hydrazine hydrate added and the mixture warmed on the water-bath for some time.On cooling the red precipitate was collected and fractionally crystallised from alcohol, and from the more readily soluble portion a colourless compound was isolated which was identified as 2 3 6-trinitrobluene by the method of mixed melting points with it sample of this compound prepared by Brady and Taylor’s method. Nitration of 2 3-BinitrotoZuene.-This was effected in an exactly similar manner to the nitration of 2 6-dinitrotoluene. The product had a freezing point of looo as compared with 112O in the case of pure 2:3:4-trinitrotoluene. The separation was brought about by means of hydrazine hydrate and the 2 3 6-tri-nitrotoluene isolated and identified.The approximate amount of 2 3 6-trinitrotoluene present in the mixtures obtained by the nitration of 3 6- and 2 3-dinitrotoluenes was determined by making up mixtures of pure 3 4 6- and 2 3 4-trinitrobluenes with 2 3 6-trinitrotoluene and determining their freezing points. It was found that these corresponded with the freezing points of the products of nitration when about 15 per cent. of the 2 3 6-compound was present. 2 6-Dinitro-m-cresoZ.-This was prepared from 2 6-dinitrol-4-amino-mcrewl. The method given for the preparation of this compound by Rellner and Beilstein (Annulen 1863 128 166) did not give good yields so a modified method was adopted. Twenty grams of trinitro-m-cresol were dissolved in 50 C.C. of boiling methyl alcohol and 200 C.C.of 15 per cent. ammonium sulphide slowly added. Vigorous action took place and when it subsided, the mixture was cooled diluted and acidified with hydrochloric acid. The sulphur was filtered off and the filtrate evaporated to small bulk. On keeping the 2 6-dinitro-4-amino-m-cresol separated and was crystallised from alcohol with the addition of animal charcoal. It was found that the usual method for the removal of the amino-group by the action of sodium nitrite on a boiling alcoholic solution of the amine containing sulphuric acid did not give good results with this cornpound owing to the stability of the diazo-compound formed ; accordingly the following method was adopted. Ten grams of 2 6-dinitro-4-amino-mcresol were dissolved in 70 C.C.of hot alcohol 20 C.C. bf ciinceiii;rated hydrochloric acid added and then gradually 10 grams of sodium nitrite dissolved in a minimum amount of water. As soon as the reaction ceased, the mixture was cooled and the precipitate collected. There wa 1618 THE FORMATION OF 2 3 6-TRINITROTOLUENE ETC. thus obtained a stable greenish-yellow crystalline material which deflagrated on heating and was probably the diazo-oxide corre-sponding with that obtained from picramic acid. This compound was treated with 30 C.C. of concentrated formic acid and a small quantity of copper polwder added when a violent reaction took place nitrogen and carbon dioxide being evolved. After filtering from the copper and diluting tlie filtrate an oil separated which slowly crystallised; the crystals were pressed on it porous tile and recrystallised from benzene when 2 6-dinitro-m-cresol separated as a whiter crystalline compound melting a t 1 3 3 O (Found N=14.1.C,H,O,N requires N = 14.1 per cent.). 2 6-Binitro-m-tolyl Methyl Ether.-The ammolnium salt of 2 6-dinitro-m-cresol was prepared by dissolving the cresol in alcoholic ammonia and evaporating the solution on the water-bath until crystallisation took place on cooling. The filtered crystals were dissolved in the minimum amount of alcohol and the calculated quantity of alcoholic silver nitrate was added when the silver salt separated in lustrous grey crystals which deflagrate violently on heating. Five grams of the silver salt were added to a solution of 5 C.C. of methyl iodide in alcohol when immediate action took place and after removal of the silver iodide the solu-tion was diluted and the precipitated 2 6-dinitro-m-tolpl methyl ether crystallised from alcohol when it was obtained in white needles melting a t 1 1 5 O (Found N= 13.4.C8H,0,N2 requires N = 13.2 per cent.). 2 6-Dinitro-m-toluidine and 2 3 6-Trinitrotoluene.-Five grams of 2 6-dinitro-m-tolyl methyl ether were heated to 130° with 5 c.a. of ammonia and 20 C.C. of alcohol. On cooling and diluting, 2 6-dinitro-m-toluidine was precipitated and identified by the method of mixed meflting points with a sample of this compound prepared by Cook and Brady (this vol. p. 750). 2 6-Dinitro-m-toluidine can be converted into 2 3 6-trinitrotoluene by Korner and Contardi’s method (Zoc. cit.). The author wishes t a express his thanks to the Director of RESEARCH DEPARTMENT, Artillery for permission to publish this work. ROYAL ARSENAL WOOLWICH. [Received October 26th 1920.