RUHEMANN : TRIKETOHYDRINDENE HYDRATE. 2025CCXII. -Triketohydrz'nderie Hydmite.By SIEGFRIED RUHEJIANN.IN a recent paper (this vol., p. 1438) it was shown that the condensa-tion product of a-hydrindone with p-nitrosodimethylaniline, on treatmentwith dilute sulphuric acid, decomposes with the formation of trike'to-hydrindene hydrate. Its formula was represented thus :according to which the elements of water are united with the %ketonicgroup of triketohydrindene. There cannot be any doubt as to thecorrectness of this formula, because the union of water with anyother ketonic group of the triketone would produce a colouredVOL. XCVII. 6 2026 RUHEMANN : TRIKETOHYDRINDENE HYDRATE.mmpound owing to the proximity of two ketonic groups, whereastriketohydrindene hydrate is colourless, as is also 1 : 3-cliketo-CO hydrindene, C6H,<CO>CH,. Further evidence in support of theabove formula is the fact that the hydrate, on treatment withphosphorus pentachloride, is transformed into the colourless 2 : 2-di-chloro-1 : 3-diketohydrindene, C6H,<CO>CC12.CO Triketohydrindenealso forms additive products with other substances, such as guanidineor benzamidine, which are colourless, and therefore are to be repre-sented by formuls similar to that of triketohydrindene hydrate.The hydrate further reacts with hydrogen cyanide t3 yield theunstable cyanohydrin, C,H,<CO>C(OH)*CN. COA closer study of the remarkable behaviour of potassium hydroxidetowards the hydrate which was described previously (loc.cit., p. 1448)led t o the following result. The reaction proceeds in three distinctphases, which are indicated by colour changes. On the addition ofthe alkali to the crystals OF the hydrate, they turn yellow and thendissolve to form a yellow solution; this subsequently becomes blue,even at the ordinary temperature, if the alkali is concentrated. Theblue colour, bowever, is very fugitive, and disappears on dilution withwater t o yield a colourless solution. On using dilute potassiumhydroxide (about 15 per cent.), the blue colour does not appear unlessthe temperature is raised immediately after the addition of the alkalit o the hydrate. The colourless alkaline solution, which represents thelast phase of the reaction, contains the potassium salt of o-carboxy-mandelic acid, CO,H*C,H,*CH(OH)*CO,H, because, on treatment withdilute sulphuric acid, it yields phthalidecarboxylic acid,The formation of this acid leads t o the conclusion that, under theinfluence of the alkali, the five-carbon ring of triketohydrindeneruptures with the formation of phenylglyoxal-o-carboxylic acid,CO,H*CBH,*CO*CHO, which finally undergoes the change to phthalide-carboxylic acid.This behaviour resembles in every respect thechange which, by the action of alkalis, phenylglyoxal undergoes t omandelic acid :C6H5*CO*CH0 -+ C6H5*CH(OH)*C0,H.The result arrived at in examining the behaviour of triketohydrin-dene hydrate towards potassium hydroxide supports the view whichwas expressed before (Zoc. tit.) concerning the product of the action ofammonia on the triketone.The formation of phenylglyoxal-o-carboxylicacid is to be regarded as the first change of the triketonic compounRUHEMANN : TKIKETOHYDRINDENE HYDRATE. 2027which is produced by the alkali, and this view follows, also, from thefact that the yellow alkaline liquor reduces Fehling’s solution.The explanation of the intermediate phase of the reaction which ischaracterised by the blue colour of the alkaline solution is difficult,because the solution readily loses its colour, and passes into the finalphase of the reaction, It is, therefore, only from analogy to thechanges which the diketopyrrolines and the compounds with similarstructure undergo, on treatment with alkalis, that a view can beexpressed concerning the nature of the blue product.The bluesolution which, for example, diketodiphenylpyrroline yields withpotassium hydroxide was explained (Trans., 1909, 95, 984) by thechange into its tautomeric form :$!Ph---CO CPh = ?*OHCPh-NH.60 -+ bPh:N*CO ’which contains a phenolic group and has an o-quinonoid structure,A similar arrangement may be assumed to exist in the potassiumcompound which is formed in the second phase of the action of thea1 kali on t r iketo hydrindene hydrate. Accordingly, pheny lglyoxal-o-carboxylic acid, which is first produced, undergoes ring formation,thus :in which, also, phenolic groups are associated with an o-quinonoidstructure. A substance with this constitution may be supposed t oyield blue salts with alkalis.These, like the corresponding saltsof the diketopyrrolines, are unstable, and, on dilution with water,are transformed into the salts of o-carboxymandelic acid :The further study of triketohydrindene hydrate led to results whichappear to be of great interest. It was found that a deep blue colouris produced on warming a mixture of aqueous solutions of this com-pound and an aliphatic or an aliphatic-aromatic amino-acid whichcontains the amino-group in the side-chain. As shown below, thisreaction has been successfully applied to a number of a-amino-acids,but a8 yet only two P-amino-acids have been tested, and they werefound to differ markedly from the a-amino-acids in their behaviourtowards the triketone, because with them the colour reaction takesplace less readily, and in the case of p-amino-P-phenylpropionic acid isfar less intense than with the corresponding a-amino-acids.No6 ~ 2028 RUEEMANN : TRIKETOHYDRINDENE HYDRATE.coloration, however, is produced by the triketone in solutions ofaromatic amino-acids which contrain the amino-group in the nucleus,nor does it occur with substituted amino-acids, such as phenylglycineor hippuric acid. On the other hand, triketohydrindene hydrate givesa blue reaction with peptone, and this fact in the. light of the resultsindicated above leads to the conclusion that in the peptones, compoundsoccur which contain the free amino-group of amino-acids.The same coloration is produced i n normal human urine on warmingit with a n aqueous solution of the reagent.This behaviour agrees withthe observations of Abderhalden and Pregl (Zeitsch. pAysio2. Chem., 1905,46, 19 ; see also Abderhalden and Schittenhelm, ibiu?., 1906, 47, 1396),according to which the urine contains a protein-like substance. Theauthors showed that it does not contain free amino-acids, but that theseare formed from it on hydrolysis.EXPERIMENTAL.Triketohydrindene hydrate was prepared in the manner describedbefore (Zoc. cit.) by warming slightly the product of the action ofp-nitrosodimethylaniline and a-hydrindone with dilute sulphuric acid ;its extraction from the dark solution which is formed, is more con-veniently effected by ethyl acetata than by ether on account of itsgreater solubility in that boloent.Additive Product of ~r&tohydr~ndene with Guunidinc,c,B~~~~>c(oH).NH.c(:NH).NH,.This is formed by adding triketohydrindene hydrate (1 gram), dis-solved in water, to an aqueous solution of guanidine, obtained from itschloride (0.6 gram) and the calculated quantity of potassium hydroxide.The mixture turns red and soon deposits a colourless solid, whichis insoluble in water, alcohol or benzene; it does not melt, butbegins to darken at about 190" and finally becomes black :0,2015 gave 0,4045 CO, and 0.0740 H20.0.1945 ,, 32.6 C.C.N, at 30' and 754.5 mm. N= 19 02.C1,H90,N, requires C = 54.79 ; H = 4.1 1 ; N = 19.18 per cent.On heating hhe compound with water for some time, slight decom-position takes place, and a small quantity dissolves to yield a redsolution.C-54.75; H=4.08RUHEMASN : TRIKETOHYDRINDENE HYDRATE.2029Additive Product of Tviketohydrindene with Benxamidine,This is prepared in a similar way to the former substance, namely,by adding sodium carbonate to the mixture of triketohydrindenehydrate (1.2 grams) and benzaxnidine hydrochloride (1 gram) dissolvedin hot water. A white solid is soon precipitated, which is insolublein water; it is sparingly soluble in boiling alcohol, but does notseparate from the solution without the addition of water, whencolourless prisms are formed, which darken at about 200" and decomposea t 229-230' with evolution of gas :0.2011 gave 0.5050 CO, and 0.0805 H,O.0.2051 ,, 17.8 C.C. N, a t 17" and 762 mm. N = 10.10.This snbstance is insoluble in sodium carbonate ; it is decomposed byC = 68.49 ; H = 4.44.C,,H,,0,N2 requires C = 68.58 ; H = 4-29 ; N = 10.0 per cent.hot potassium hydroxide with the formation of benzaldehyde.Additive Product of Friketohydrindene with Hydrogen Cyanide,Whilst the additive compounds mentioned above are fairly stable, thecyanohydrin, which triketohydrindene hydrate forms with hydrogencyanide, is readily decomposed.X t is obtained by mixing the triketonehydrate (0.5 gram), dissolved in water, with potassium cyanide(0.5 gram), and then adding dilute hydrochloric acid to the deep redeolution which is produced, The colour disappears and pale brownneedles separate which sinter and darken at about 120°, and at 148'melt with evolution of gas.The compound dissolves in boiling water,but at the same time decomposes. For analysis, the crystals formedi n the reaction were mashed with cold water and dried in a vacuumdesiccator over sulphuric acid :0.2053 gave 13.8 C.C. N, at 1 8 O and 749 mm. N = 7.64.C1,1J50,N requires N = 7.49 per cent,Aciiolz of Pho~phorus Pentach loride on TTiketoh ydrindene Bydrate.On mixing the triketone hydrate (1 mol.) with phosphorus penta-chloride (2 mols.), no reaction ta.kes place at the ordinary temperature;at loo", the mixture evolves hydrogen chloride and assumes a redcolour, which is probably due to the removal of water and the formationof triketohydrindene. The action proceeds more readily if phosphorylchloride is used together with phosphorus pentachloride ; on warming2030 RUHEMANN : TRIKETOHYDRINDENE HYDRATE.the whole dissolves t o yield a yellow solution, which is cooled andthen poured on ice, when an oil is precipitated which soon solidifies.The solid is washed with light petroleum, which removes a yollowproduct, whilst a colourless substiance remains undissolved.Thisis readily soluble in hot dilute alcohol, and, on cooling, crystallises i ncolourless plates. The yield is poor owing to the formation of theyellow by-product, The compound mas identified as 2 : 2-dichloro-l : 3-di-ketohydrindeno, C,H,<CO>CC12, co by the melting point, 124-1 25O,and its chemical bebaviour (compare Zincke and Gerland, Ber., 1858,21, 2390).Fmmation of Phthalidmarboxylic Acid from Triketohydrindene Eydrate.The action of potassium hydroxide on the triketone hydrate wasdescribed in a previous paper (this vol., p.1448) and more fullyon p. 2026. It was stated that the final phase of the reaction yieldsa colourless solution. This, when treated with an excess of dilutesulphuric acid and digested on the water-bath for an hour, containsphthalidecarboxylic acid, yo' '>CH*CO,H. It is isolated from theC,H,acid solution by extraction with ether, and, on evaporation of theether, is left as a white solid. This readily dissolves in hot water, and,on cooling, crystallises in colourless plates melting at 150-151°, Theyield is almost theoretical. (Found : C = 60.70 ; H = 3-37, Calc.,C = 60.67 ; H = 3.37 per cent.)The compound is identical with the acid which Zincke (Bey., 1894,27, 743), in the course of his researches on the action of bleachingpowder on quinones, obtained from monochloro-P-naphthaquinoneas well as from ~eocoumar~ncarboxylic acid.Action.of 'I.iketoh~dl.i~enehydrate on Amino-acids.On mixing a slightly warmed aqueous solution of triketobydrindenehydrate and glycine, an intense blue colour develops immediately, and,after a short time, a dark solid separates. The same behaviour isshown by all the a-amino-acids which 1 have been able to obtain;several of them I owe to the kindness of Dr. Hopkins, whomI have also to thank for the interest he took in the progress ofthe work. The reaction has been applied to alanine, valine,leucine, tyrosine, to a-amino-/I-phenylpropionic, aspartic, and glutamicacids, to tryptophan and cystine, Owing to the fact that thelatter compound is sparingly soluble in water, it must be boiledwith a solution of tbe triketone hydrate; in all the other caseRUHEMANN : TRIKETOHYDRINDENE HYDRATE. 2031the reaction takes place almost with the same rapidity as with glycine,and requires mere traces of the reagents, The sensitiveness of thereagent is indicated by the fact that the blue colour is still per-ceptible on slightly warming glycine with a solution which contains1 part of triketohydrindene hydrate dissolved in 15,000 parts ofwater.In the same dilution, the reagent yields with ammonia B yellowcoloration, whereaq the colour is red, or reddish-violet, on using moreconcentrated solut'ions (compare this vol., p. 1447).The shade of theblue coloration that is produced shows but slight variations with thedifferent a-amino-acids. Of the P-amino-acids only two have beentested, namely, P-aminopropionic acid and /3-amino-P-phenylpropionicacid. The latter, which was prepared according to Posner's directions(Bey., 1903, 36, 4313) by the action of hydroxylaruine on cinnamicacid, differs remarkably from its isomeride, a-amino-/3-phenylpropionicacid, in its behaviour towards triketohydrindene hydrate ; no colouris produced on mixing the reagents dissolved in hot water, A faintviolet coloration, however, develops if the solution is boiled for sometime ; on the other hand, the corresponding a-amino-acid readily reactswith the triketonic compound to yield a deep blue colour. Thedifference between a- and P-aminopropionic acids in their behaviourtowards the reagent, although not so striking as with the formerisomeric acids, is yet apparent, especially as regards the rapidity ofthe reaction, which in this case, also, is greater with the a-amino-acid.It is, however, necessary t o examine a larger number of the @amino-acids before, by means of the reagent, a general distinction betweenthe two groups of amino-acids can be established, Neither phenyl-aminoncetic acid (phenylglycine) nor hippuric acid yield a colorationwith the triketone hydrate, even on boiling the aqueous solutions ofthe mixtures, and this result indicates that the reaction depends onthe amino-group of the amino-acid being intact. Aromatic amino-acids which contain the amino-group in the nucleus, for example,o-aminobenzoic acid, also do not respond to the test.The chemical nature of the colour reaction has not yet beenascertained, but experiments in this direction are already in progress,and will be recorded shortly.UNIVERSITY CHEMICAL LABORATORY,CAMBRIDGE