RUHEMANN : TRIPHENY L-2-PYRONE. 457XL IX .- Trip hen y 1- 2 -p yrone.By SIEGFRIED RUHEMANN.AFTER having found that, in the presence of sodium ethoxide, ethylphenylpropiolate condenses with P-ketonic esters or P-diketones to2-pyrone derivatives (Trans., 1899, 75, 245, 411), and that theester reacts with acetone or acetophenone to yield 4-pyrone com-pounds (Trans., 1908, 93, 431, lZSl), I have, for various reasons,thought fit to extend this research. I n the first place, it wasnecessary to examine whether the formation of 4-pyrone derivativesgenerally takes place by the action of the acetylenic ester on ketoneswith the grouping -CH,*CO. This inquiry led to the result that,so far as my observations go, this reaction is limited to those twocases, for p-tolyl methyl ketone, propiophenone, or methyl ethylketone do not condense with ethyl phenylpropiolate; the only changewhich occurs, is the transformation of the ester into the correspond-ing acid.As previously stated (Zoc. &.), this hydrolysis takesplace to some extent, also, on using acetone or acetophenone insteadof the above-named ketones, and accompanies the formation of the4-pyrone derivatives.The behaviour of ethyl phenylpropiolate towards deoxybenzoinis different. I have undertaken this investigation in connexionwith the further study on the diketopyrrolines, the results of whichwill be published at a later date. In the light of the work ondeoxybenzoin carried out by Victor Meyer and other chemists, itwas to be expected that ethyl phenylpropiolate would react withthe ketone, not'as it does with acetone or acetophenone, but inthe manner similar to the behaviour of the acetylenic ester towards&ketonic esters or P-diketones, thus :CPhiC*CO,Et + CH,Ph*COPh = O<ggc:g>CPh + C,H,O,and yield 4 : 5 : 6-triphenyl-2-pyrone.This reaction does, indeed,take place. I have subjected this compound to a, closer study withthe view of ascertaining whether, under the influence of potassiumhydroxide, it decomposes thus :C23HI609 + 2KH0 = C6H,*C(CH2*CcH,) :CH*CO,K + C,H,*@O,K.I find that only part of the pyrone compound breaks up in thisway to form B-benzylcinnamic acid, melting a t 168-169O. Theformation of this acid is of some interest for the following reason.Whereas the homologues of cinnamic acid with the formulaC,H,-CH:CR=CO,Et are readily produced according to Perkin'458 RUHEMANN : TRIPHENYL-2-PYRONE,reaction, it is only recently that Schroeter (Ber., 1904, 37, 1090;1907, 40, 1589; 1908, 41, 5) was able to obtain P-alkylcinnamicacids from the products of the action of zinc or magnesium on themixture of aromatic ketones and ethyl iodoacetate (see, also, Rupeand Busolt, Ber., 1907, 40, 4537).Like Michael and Palmer's(Amer. Chem. J., 1885, 7, 69) a-benzylcinnamic acid, which on re-duction yields dibenzylacetic acid, (C6H5-CH2)2*CH*C02H, p-benzyl-cinnamic acid, on treatment with sodium amalgam, is transformedinto p-phenyl-P-benzylpropionic acid,C6H,*CH(CH2*C6H,) *CH2-C02H.Again, just as a-benzylcinnamic acid, under the influence of coldconcentrated sulphuric acid, condenses to form benzylidene-a-hydrindone, C6H,<!2>C:CH*C6H5 (see Schmid, J.pr.Chem., 1900, [ii], 62, 550), p-benzylcinnamic acid is transformedinto 3-phenyl-1-napht hol :C( 0 H): $lHThe formation of P-benzylcinnamic acid (m. p. 168-169O) isaccompanied by the production of an isomeride melting between76" and 90°, which, as yet, I have failed to obtain with afixed melting point. But the action of potassium hydroxide ontriphenyl-2-pyrone is still more complicated, because part of thepyrone compound undergoes the following change :C,H4<CH =rC*C,H,'The fact that the product C,,Hl,O, which is formed, does notmelt sharply, indicates that it is a mixture, and leads to the viewthat the compound CH2:C(C6H,)*CH(CGH,)*COoCsH5, which maybe expected to be first produced, under the influence of the alkali,is partly or wholly transformed into a mixture of stereoisomeridesof phenylethylidenedeoxybenzoin, CH3=C(C,H5):C(C6H,).CO*C6H,.All attempts to separate this mixture by crystallisation have beenunsuccessful. Compounds which belong t o the same type are knownalready. These are the alkylidenedeoxybenzoins which Klages andTetzner (Ber., 1902, 35, 3965) prepared from deoxybenzoin.Isomeric with the product C2,H180, which is formed from triphenyl-2-pyrone, is p-methylbenzylidenedeoxybenzoin,This substance was obtained by those chemists on condensingdeoxybenzoin with p-tolusldehyde, and was found to exist in twoforms, which could be separated.CH,*C6H,'cH : c( C6H,)*CO*C6H5HUHEMANN : TRIPHENYL-2-PYRONE.459EXPERIMENTAL,Formation of 4 : 5 : 6-Tr~phenyl-fL-pyl.onc,O<Co-cH>CPh. CPh:CPhThis substance is formed on adding deoxybenzoin (9-8 grams) todry sodium ethoxide (3.4 grams) suspended in absolute ether, andthen ethyl phenylpropiolate (8.7 grams) to the solution which isshortly produced. The mixture becomes red, and a yellow solidseparates. The reaction is complete in the course of a day; wateris then added, when part of the solid (8 grams) remains undissolved.This is insoluble in water or alcohol; it dissolves sparingly in cold,moderately in boiling, glacial acetic acid, and, on cooling, crystallisesin faintly yellow prisms, which melt at 245-246O:6.2030 gave 0.6342 CO, and 0.0930 H,O.Cz3Hl6O2 requires C = 85.18 ; H = 4.94 per cent.The yield of this substance is only 50 per cent.of the theoretical;this is due to the fact that part of the ethyl phenylpropiolate istransformed into the acid, which, together with the unattackeddeoxybenzoin, is contained in the ethereal layer from the product ofthe reaction.C = 85.20 ; H = 5-09.Action, of Potassium Hydroxide o n Triphenyl-2-pyrone.On boiling the pyrone with an excess of alcoholic potassiumhydroxide on the water-bath, it dissolves, yielding a deep redsolution, and finally a yellow solid. After four to five hours'heating, the alcohol is distilled off, water added to the residue, andthe whole extracted with ether. On evaporation of the ether, ayellowish-red oil is left behind, which gradually sets almost com-pletely, t o a solid.This is washed with a little dilute alcohol,when it becomes quite white. This product is a mixture of thestereoisomerides of phenylethylidened eoxy b eneoin,CH3-C( C6H5) :c( C6H,)-C0 C6H,.It readily dissolves in hot alcohol, and, on cooling, crystallises incolourless prisms, which begin to soften at 76O, and are completelymelted a t 90°:0.1882 gave 0.6110 CO, and 0*1090 H,O. C = 88.54 ; H = 6-13.0.1970 ,, 0.6400 CO, ,, 0*1080 H20. C=88*60; H=6*09.C,,H,,O requires C = 88.59 ; H = 6.04 per cent.The substance is very soluble in carbon disulphide, and moderatelyso in light petroleum; it dissolves slowly in cold concentrated sul-phuric acid, forming a red solution.I have repeatedly crystallise460 RUHEMANN : TRIPHENYL-2-PYRONE.the product from alcohol as well as from light petroleum, but allthe fractions, on melting, showed practically the same behaviour,and, on analysis, gave the same results.The alkaline solution from the product of the action of potassiumhydroxide on the pyrone, which contains a mixture of benzoic acidand the isomeric 8-benzylcinnamic acids, is mixed with an excess ofdilute sulphuric acid and repeatedly extracted with ether. Onevaporation of the ether, a yellow oil is left behind, which sets toa solid after a few hours. When this is washed with cold dilutealcohol, @-benzyIcinnamic acid remains undissolved.The acid is only sparingly soluble in cold alcohol, but readilyso in boiling alcohol, and, on cooling, separates in long, colourlessneedles, which melt at 168-1 69O :0-2028 gave 0-5992 CO, and 0-1080 H;O.Cl6Hl,O, requires C = 80.64 ; H = 5.88 per cent.8-Benzylcinnamic acid is readily soluble in ammonia, and thissolution, on the addition of silver nitrate, yields a white silver salt,which is sparingly soluble in boiling water :C =80*58; H=5*92.0.2595 gave 0.0810 Ag.Ag=31.21.C,,H&,Ag requires Ag = 31-30 per cent.8-PheqZ-8-6 enzylpropionic A cid, C,H,-CH (CH,*C,H,)*CH,*CO,H.The reduction of P-benzylcinnamic acid readily takes place ondissolving it in dilute sodium hydroxide and shaking the solutionwith an excess of sodium amalgam (Zi per cent.) for about half anhour. The alkaline liquor is poured off and mixed with dilutehydrochloric acid, when an oil separates which solidifies in thecourse of a day.The solid is sparingly soluble in light petroleum,readily so in chloroform or alcohol; on adding water to the hotalcoholic solution until it becomes turbid, colourless prisms graduallyseparate, which melt. at 95-96O :0.1978 gave 0.5810 CO, and 0.1190 H,O.C16H1602 requires C = 80.0 ; H = 6.67 per cent.The silver salt, which is formed on the addition of silver nitrateto the ammoniacal solution of the acid, is white, and does notchange on drying in the water-oven :Ag= 31-09.C1,Hl,0,Ag requires Ag = 31.12 per cent.C=80*10; H=6*71.0.2438 gave 0.0758 AgRUHEMANN : TRIPHENY L-8-PYRONE. 461CH=F*C,H,C(0H):CH '3-Phenyl- 1-napht hol, C,H,<On adding cold concentrated sulphuric acid to P-benzylcinnamicacid, the crystals gradually dissolve, forming a red solution which,when kept overnight and then slowly poured into cold water, yieldsan oil.This gradually sets to a solid, which is sparingly soluble inboiling water, readily so in ether, chloroform, or alcohol; it ispurified by adding water to the alcoholic solution until an emulsionis produced, from which light brown needles, melting at 100-lO1°,separate :0.1883 gave 0.6023 CO, and 0.0950 H,O.C16H,,0 requires C = 87.27 ; H=5*45 per cent.This compound is very soluble in alkalis, and does not give acolour reaction on the addition of ferric chloride to its alcoholicsolution.The alkaline solution, which is formed by the action of potassiumhydroxide on triphenyl-2-pyroneY contains, besides P-benzylcinnamicacid, melting a t 168--169O, its isomeride.This, being very solublein alcohol, remains in the original alcoholic filtrate from the formeracid, together with benzoic acid, which, also, is produced in thereaction. In order to remove the benzoic acid, water is added tothe filtrate, and the whole extracted with ether. The productwhich is left behind on evaporation of the ether is distilled in acurrent of steam, the benzoic acid contained in the distillate isextracted with ether, and, after evaporation of the ether, re-crystallised from water. The isomeride of P-benzylcinnamic acidis not volatile with steam, and remains in the distilling flask as anoil, which slowly solidifies. The solid is very soluble in alcohol,ether, warm carbon disulphide, or chloroform, sparingly so in lightpetroleum. The solution in the latter solvent gradually depositsstout prisms, which soften a t 104O, and are completely melted at1 1 6 O . That this substance is isomeric with fl-benzylcinnamic acid,melting a t 168O, is proved by the following analysis :C = 87-23 ; H = 5.60.0.1900 gave 0.5620 CO, and 0-1023 H,O.C,,H1,O, requires C = 80.64 ; €1 = 5.88 per cent.The fact that this substance does not fuse sharply seems toindicate that it is not pure, but after repeated crystallisations fromdilute alcohol or light petroleum, the behaviour on melting ispractically unaltered.C = 80.67 ; H= 5-98.UNIVERSITY CHEMICAL LABORATORY,CAMBRIDGE