28 EASTERFIELD AND SELL ON CITRAZINIC ACID. V.-Studies on Citraxinic acid. Part 11. By T. H. EASTERFIELD, MA., and W. J. SELL, M.A., F.I.C. Conrersion of Diammonic Citrate into Citrazinic acid. SO many methods have been described for the conversion of " an- hydro "-citric into citrazinic acid derivatives, that we have been led to investigate the action of heat on the ammoniuui salts of citric acid, with the object of directly preparing citrazinic acid. Behrmann and A. W. Hofmann (Bey., 17, 2688) have, indeed, attacked this problem, but were unable to obtain definite results, probably because their experiments were carried out a t too high a temperature. Sabanin and Lasknomky (Zeit. anal. Chem., 17, 74) have shown that when ammoniacal solutions of citric acid are heated, in sealed tubes, at 110-120" during seveid hours, and subsequently exposed to the air, a blue-green colour is gradually developed, and we find that, if a solution of diammonic citrate is heated on the water bath for several days in an open dish, and then allowed to cool, it behaves in a similarEASTERFIELD AND SELL ON CITRAZNIC ACID.29 manner. This development of colonr, on exposure, is, however, shared by cold dilute ammoniacal solutions of citrazinic acid, and led us to suppose that a slow condensation of citric t o citrazinic acid had occurred, even under this gentle treatment. When the action was slightly forced, by heating the citrate of ammonia at 130" for three hours in an open vessel placed in an air bath, a considerable quantity of citrazinic acid (about 6 per cent.of the citric acid taken) was formed, but, at temperatures above 160: no citrazinic acid could be isolated. The acid produced in this way always showed a rather high per- centage of nitrogen ; after treatment with potash and animal charcoal, it was identified as citrazinic acid by the fluorescence of its alkaline solutions, its behaviour with a warm solution of potassic nitrite, and by an analysis. 0.1287 gave 0.0395 H,O and 0.2202 CO,. 0.1571 ,, 12.7 C.C. dry nitrogen at 11" and 760 mm. N = 9.62. This is the simplest condensation of citric acid into a pyridine derivative which has been hitherto described. Preparation of Citrazinamide f r o m Ethylic Citrazinate.-Though it could scarcely be doubted that citrazinamide is, in reality, the amide of citrazinic acid, it seemed desirable to check this conclusion by preparing the amide from an etihereal salt of the acid.Ethylic citrazinate was heated with an excess of strong ammonia at 120-130", f o r two hours in a sealed tube ; a crystalline substance was formed which, after recrystallisation from dilute ammonia, was recognised by its appearance and a nitrogen determination, as the anhydrous ammonium salt of citrazinamide described by us in a former paper. 0.1347 gave 28.4 C.C. dry nitrogeu a t l2"and 764 mm. N = 25.01. CsHJVz0,,NH3 requires N = 24.55 per cent. C = 46-63; H = 3.40. Cke[,NO, requires C = 46-45 ; H = 3.23 ; N = 9.03 per cent. Reduction of Citratinamide by Sodium Amalgam-Emil Fischer (Ber., 23, 933) has pointed out that, whereas amides of the benzenoid series are readily reduced to alcohols by the action of sodium amalgam, those of the aliphatic series undergo no such reduction, and the recent work of Max Marx (Annakm, 263, 249-259), and more par- ticularly that of Arthur Hutchinson (Trans., 1890, 57, 957), has em- phasised the correctness of Fischer's statement.As far as we are aware, no experiments have yet been made with the object of finding out whether amides of the pyridine series would behave as fatty, or as benzenoid compounds, on reduction. In the case of citrazinamide, reduction readily takes place in alkaline solution, citrazinyl alcohol being one of the chief products ; and as in the case of some benzeno'id30 EASTERFIELD AND SELL ON CITRAZINIC ACID. amides, a small quantity of a hydrobenzoin is simultaneously formed.These experiments may, therefore, be regarded as yielding additional evidence of the benzenoid habitns of the derivatives of pyridine. 10 grams of citrazinamide were dissolved in a small quantity of caustic soda solution and diluted to 200 C.C. ; the solution was then reduced with 29 per cent. sodium amalgam, of which 215 p m (rather more than 4 atoms of sodium) were required. During the early stages of reductmion, the liquid darkened considerably, but, towards the end of the process, it became of a light amber colonr. As soon as an escape of hydrogen showed that the reduction was complete, the liquid was acidified with dilute sulphuric acid, and the small quantity of resinous matter which separated was filtered off and neglected.On concentrating the solution to about one-quarter of its bulk, and allowing it to cool, a brown substance separated in minute spherules, subsequentIy identified as a hydrobenzoin (see below) ; the solution was then neutralised with ammonia, poured into 10 times its volume of methylated spirit, and, after filtration from the moist sulphates, the spirit was distilled off. The solution, thus left, was acidified with hydrochloric acid and evaporated to a small bulk in a vacuum over sulphuric acid and potash; the crystalline sub- stance which was deposited was then dissolved in a small quantity of hot water, from which it separated, on cooling, in clusters of almost colourless prisms. Analysis agrees with the supposition that it is citrazinyl alcohol, formed according to the equation CsH4302*C:ONHz + 2H2 = CsH4NO2.CHz.OH + NH,.The crystals contain 1H20, which they do not lose in a vacuum a t the ordinary temperature, but do so readily at 100". After two recrystallisations, the substance melted at 158" without under- going decomposition. 0.1635 gave 0.0830 H,O and 0.2695 CO,. 0.1539 lost 0.0174 H,O and gave 11.4 C.C. dry nitrogen at 770 mm. CsH7N03,H20 requires C = 45.28; H = 5.66; N = 8.80; H,O = Another specimen gave C = 45.07 ; H = 5.44 ; N = 9.12 per cent. C = 44.95 ; H = 5.64. and 12". 11.32 per cent. H,O = 11.30; N = 8.88.HEYCOCK AND NEVILLE: FREEZNQ POINTS OF ALLOYS. 31 The aqueous solution of the alcohol is very acid to litmus paper, and gives a precipitate with ammoniacal nitrate of silver ; if this pre- cipitate is dissolved in a drop of ammonia, and the solution warmed, a silver mirror is formed.Citrazinyl alcohol dissolves easily in hot water, comparatively sparingly in cold water, but readily in a i d or alkaline solutions, ethyl and methyl alcohol, acetone, and ethylic acetate. I n ether, benzene, chloroform, and light petroleum, it is either very sparingly soluble, or insoluble. The yield was about 16 per cent. of the amide employed. vH(OH)-yH(OHj C C /\ /\ Citrazinyl Hydrobenzoiiz, Hv QHz Hv HZ H0.C CO HO*C CO It has been mentioned that, during the preparation of citrazinyf alcohol, a substance consisting of brown sphernles was obtained ; this dissolves readily in hot water, much more sparingly in cold water, and its aqueous solution reddens blue litmus ; it is very soluble in alkalis, and is reprecipitated on the addition of a mineral acid. When analysed, it gave numbers which agreed with those required for a hydrobenzoh of the above formula ; the yield of the compound was very small, and we have been unable to make further experiments with it. Different specimens ga-re the following analytical resnlh. 0.1374 gave 0.0523 HzO and02615 (30,. C = 51.9; H = 4-22. 0.1135 ,, 0.0425 ,, ,, 0.2114 ,? C = 50.8; H = 4-14. 0.1192 ,, 9.7 C.C. dry nitrogen a t 11" and 758 mm. N = 9.67. 0.1135 ,, 9-6 C.C. ,, ,, 10" and 7T0 mm. N = 1023. Uniz ersity Lab m a f ory , (C,HaNO,*CHOH)~ requires C = 51.42; = 4.20; N = 10.00 per cent: Cambridge.