JAPP AND LANDER: SYNTHESIS O F PENTACARBON RINGS. 139 IX.-Synthesis o f Pentucurbon Rings. Part II. Con- densation of Benxil with Acetone Dicurboxylic Acid. By FRANCIS ROBERT JAPP, F.R.S., and GEORGE DRUCE LANDER, B.Sc. BENZIL and acetonedicarboxylic acid, when gently warmed with dilute alcoholic potash, condense according to the equation yielding anhydracetoneBenxi1cu~boxylic acid (melting at 167-1 68"). I n the preliminary note on this subject which we published (Proc., 1896, log), this acid was formulated as an analogue of isophenanthroxyl- C H *C=C COOH ene-acetoacetic acid l 6 I >CO (Japp and Klinge- C6H4' C(OH)* UH, mnnn, Trans., 1891, 59, 2). Further investigation, however, has shown140 JAPP AND LANDER : SYKTHESIS OF PENTACARBON RINGS. that this analogy does not hold.Thus, whilst isophenanthroxyleneaceto acetic acid, dissolved in sodium carbonate, is stable towards permangan- ate in the cold, anhydracetonebenzilcarboxylic acid is rapidly oxidised. This difference in behaviour is in keeping with the formulze here as- cribed to these acids, the oxidisable acid containing,and the stable acid not containing, hydrogen directly united to the ethylene group. The position of the carboxyl group in anhydracetonebenzilcarboxylic acid is thus identical with that of the alkyl group in the mon-alkyl derivatives of anhydracetonebenzil (see preceding paper), When anhydracetonebenzilcarboxylic acid is boiled for a few minutes with fuming hydriodic acid, it is reduced and at the same time deprived of carbon dioxide, yielding a di~~~enylcyclopent~none, C H *C*CH, II >CO (m.p. l l O o ) , identical with that obtained from C,H,* C-CH, anhydracetonebenzil itself (see preceding paper). By oxidation with sodium hypobromite, anhydracetonebenzilcarboxylic acid yields a mixture of diphenylmuleic and diphenylficrnaric acids, these two substances being produced in approximately equal quantity. A change in the position of the double bonds ta,kes place during this process. The corresponding change which occurs in the formation of diphenylcyclopentenme has been discussed in the preceding paper. When oxidised with chromium trioxide in acetic acid solution, it parts with 2 atoms of hydrogen, yielding an acid of the formula C118H220, (melting a t 205 -207", with decomposition). The primary product is a, yellow compound, apparently the hydrazone ; but this readily changes, especially on recrystallisation, into dark red needles of a substance melting indefinitely about 200") formed by elimination of 1 mol.of water from 2 mols. of the hydrazone, The action of phenylhydrazine is complex. EXPERIMENTAL Pyepumtion of Anhydrucetonebenxilcccrboxylic Acid .-Twent y-one grams of finely-powdered benzil and 15 grams of acetonedicarboxylic a,cid were introduced into a flask along with sufficient alcohol to dissolve the benzil in the subsequent process of warming. To this were added 17 grams of potassium hydroxide dissolved in 20 C.C. of wat.er and 100 C.C. of alcohol, after which the mixture was very gently warmed on the water bath until everything had dissolved. The flask was then re- moved from the water bath, and allowed t o stand.Too long heating, or too high a temperature, decomposes the acetonedicarboxylic acid, andJAPP AND LAXDER: SYNTHESIS OF PENTACARBON RINGS. 191 must therefore be avoided; and for the same reason, the benzil should be previously powdered, in order that it may the more readily dissolve. On standing, the liquid deposited a potassium salt, which was separated by filtration, washed with cold alcohol to remove adhering caustic potash, and then boiled with alcohol, in which it is only sparingly soluble ; it was thus freed from benzil and dark-coloured impurities. From the solution of this salt in water, the new acid was precipitated by dilute sulphuric acid. It was purified by dissolving it in hot glacial acetic acid and di- luting with hot water ; on standing, the liquid deposited crystals of the new acid. It may also be recrystallised from a large bulk of boiling water, From either of these solvents, it crystallises in two forms : (I) small, thickish, oblong plates, frequently grouped into rosettes ; this form is ailhydrous and melts at 167-168"; and (2) thin, very lustrous plates, or flat needles, containing 1H,O; these melt on the water bath, but the fused substance speedily re-solidifies in crystals of the first fonn, and then melts a t 167-168".With concentrated sulphuric acid, it gives an intense red coloration, resemblirig that produced by benzilic acid under the same circumstances. I t s sodium salt is rapidly oxidised by permanganate in the cold (von Baeyer's test for non-saturation). Analysis* of a specimen recrpstallised from a mixture of ethylic acetate and light petroleum gave figures agreeing with the formula Cl8H1404.0.2972 gave 0.7996 CO, and 0.1304 H,O. C = 73.37 ; H = 4.S7. 0.2084 ,, 0.5605 CO, ,, 0.0922 H,O. C = 73.35 ; H = 4.91. C18H1404 requires C = 73.47; H = 4.76 per cent. I n order to indicate its relationship to anhydracetonebenzil, we have 0-4585 of the lustrous plates, deposited from water, lost, on heating named it unhyd~.ucetonebenxilcurboxylic acid. a t loo", 0.0254. H,O = 5-54 per cent. Cl8H1,O4,H,O requires H,O = 5-76 per cent. The d v e r salt was obtained as a white precipitate by adding silver nitrate to a solution of the ammonium salt. 0.3756, dried at 80°, gave, on ignition, 0.1006 Ag = 26.78 per cent.C18H1,04Ag requires Ag = 26.93 per cent. Action of Concentrated Ifgdriodic Acid on An~gdracetonebenxiur- boxylic Acid. Formation of Diphenylcyclopentenone.-Ten grams of an- hydracetonebenzilcarboxylic acid were boiled with excess of concen- * The acid was prepared and analysed by Dr. J. Bishop Tingle, who began this research, jointly with one of us, in the Chemical Laboratory of tlie Royal College of Science, London, in 1890. The work was interrupted at the time by Dr. Tingle's departure from London.142 JAPP AND LANDER : SYNTHESIS O F PENTACARBON RINGS. trated hydriodic acid (sp. gr. 1.96) for 5 minutes. The solid mass left on cooling was washed with water, extracted with ether, and the ethereal solution freed from iodine in the usual way with sulphurous acid.It was difficult to purify, but as we could perceive that it was diphenylcyclopentenone, we recrystallised it, as recommended by Japp and Burton, from a large volume of boiling water and afterwards from alcohol. It was thus obtained in thin, yellowish prisms melting a t 1 lo", which is the melting point of diphenylcyclopentenone. Analysis gave figures agreeing with the expected formula C17H,,0. Found : C = 87.02 H = 6-24. Calcu- lated : C = 87-18 ; H= 5.98 per cent. The product of reduction was not an acid. It is formed according to the equation Cl8HI4O4 + H2 = C17H140 + CO, + H2O. Oxidation of Anhydraceto.r.Le6enxiEcar6oxylic Acid with #odium Hypo- twomite. ETormation of Dipheny Zmaleic and DiphenyZjunzaric Acids. - Twenty grams of anhydracetonebenzilcarboxylic acid were dissolved in caustic soda, a solution of 60 grams of bromine in excess of caustic soda was added, and the mixture allowed to stand at the ordinary tempera- ture for a fortnight, after which it was saturated with sulphur dioxide, precipitated with dilute sulphuric acid, and the precipitate taken up with ether.The ethereal solution was extracted, first with aqueous sodium carbonate, which removed an organic acid, and then with aqueous caustic soda, which took up a substance insoluble in the car- bonate. On acidifying the caustic soda solution, a yellowish precipitate was obtained; this was recrystallised from benzene, from which it was deposited in the characteristic, greenish-yellow, fluorescent crystals of diphenylmaleic anhgdvide, melting at 156".Found : C = 76.71 ; H = 4.05. Calculated for C16Hlo0,: C=76*80; H=4*00 per cent. The yield was 3 grams. The sodium carbonate extract gave, on acidification, a precipitate of an acid. This was deposited from ethylic acetate, on the addition of benzene, in colourless needles, melting, when rapidly heated, at 276" with evolution of gas (aqueous vapour). The yield was 3 grams. Analysis showed that the substance had the composition of diphernyl- fumaric acid. Calculated for C16H1204 : C = 71.64 ; H = 4.48 per cent. Reimer (Ber., 1882, 15, 1627) gives 260" as the melting point of this compound. It does melt a t that, temperature, if kept there long enough ; but, when rapidly heated, it melts as above at 276". The lower melting point is doubtless due to the slow conversion of the substance into diphenylmaleic anhydride and water.Owing to this discrepancy in the melting point, we thought it necessary Found : C = 71 -42 ; H = 4.50.JAPP AND LANDER: SYNTHESIS OF PENTACARBON RINGS 143 to identify the diphenylfumaric acid still further by transforming it into diphenylmaleic anhydride. For this purpose, a portion of it was heated in a Sprengel vacuum at 260O. Only water was given off, while a yellowish substance sublimed and collected in crystals in the upper part of the flask. The product was deposited from benzene in greenish- yellow, fluorescent prisms, melting a t 156", and was diphenylmuleic ccnhy d d e . Oxidation of An~?/drucetonebenxilcarbox~Zic Acid with Ch~omizcm B*ioxide.-Ten grams of anhydracetonebenzilcarboxylic acid were dis- solved in a little glacial acetic acid, and the solution was diluted with water, but not so as to cause precipitation. Eight grams of chromium trioxide dissolved in a little water were then added, and the liquid was warmed on the water bath until the oxidation appeared to becomplete, after which the dark green solution was poured into excess of water.The precipitate was collected, extracted with sodium carbonate, and the organic acid reprecipitated and purified by recrystallisation, first from benzene, and afterwards from ethylic acetate. It crystallised in minute needles, melting with evolution of gas at 205-2079 (The melting point 201O was erroneously given in our first note.) Analysis gave figures agreeing with the foi inula C,,H,,O,, showing that the new acid had been formed from anhydracetonebenzilcarboxylic acid by the withdrawal of two atoms of hydrogen.0.1194 gave 0.3241 CO, and 0.0463 H,O. C = 74.02 ; H=4.31. 0.1850 ,, 0.4994 CO, ,, 0.0682 H,O. C = 73.63 ; H = 4.09. C,,H,,O, requires C = 73.97 ; H = 4.1 1 per cent. The siEvev scclt was obtained as a white precipitate by adding silver nitrate to a solution of the ammonium salt. It was dried at 80'. 0,1691 gave 0.0457 Ag. Ag = 27-02. C,,H,,O,Ag requires Ag = 27.07 per cent. Action of Phenylhydmxine o n Anhydrcccetonebenxilcarboxylic Acid.- Five grams of anhydracetonebenzilcarboxylic acid were boiled with phenylhydrazine in alcoholic solution. Almost immediately, a crystal- line substance of a clear, yellow colour was deposited.This was separated by filtration and extracted with boiling benzene, in which it is only sparingly soluble. The yellow residue was then dissolved in boiling alcohol, which deposited it in dark red needles, the substance having undergone a change during recrystallisation. The benzene extract also yielded red needles. The melting point was very indefinite, as the substance decomposed in melting : the point of decomposition lay somewhat over 200°. Some specimens which had been purified by144 JAPP -4ND MURRAY : SYNTHESIS OF PENTACARBON RINGS. recrystallisation from a mixture of ethylic acetate and light petroleum, or acetone and light petroleum, melted lower than this (180-182'). I f acetic acid is used as a solvent in the preparation of the compound, the red substance is formed at once ; but too high a temperature must be avoided, otherwise the product is resinified. The substance separates from the acetic acid solution on dilution with water. Analyses of specimens prepared by both methods gave identical figures. 0.1756 gave 0.4947 CO, and 0.0832 H,O. C = 76.83 ; H = 5.26. 0.1838 ,, 12.20 C.C. moist nitrogen a t 15Oand 753 mm. N = 7.69. 0.1123 ,, 7.40 ,, ,? ,, ,, 16' ,, 750-5 mm.N = '7.58. C,,H,,N,O, requires C = 76.80 ; H = 5-07 ; N -- 7.47. 0.1134 ,, 0.3181 CO, ,? 0.0542 H,O. C=76.50; H=5*31. We have already pointed out that the formation of such a compound might be accounted for by supposing 2 mols. of the normal hydra- zone, C,,H,,N,O, (possibly represented by the yellow compound), to unite with elimination of 1 mol. of water. The unstable yellow compound was not analysed. CHEMICAL DEPARTMENT, UNIVERSITY OF ABERDEEN.