ABSORPTION SPECTRA OF DlKETOPYRROLINES. 2535CCLIX.--The Absorption Spectra of Various Diketo-pyrroline Compounds.By JOHN EDWARJI PURVIS.A SERIES of coloured diketopyrrolines have been described byItuhemann (Trans., 1909, 95, 984, 1603; this vol., pp. 462, 1438),and it appeared to be of some interest to study these compounds inrelation to their absorption and constitution; the aim of thiscommunication is to show how far the absorption is connected with(1) the ketonic groups, (2) the replacement of the oxygen of theketonic groups, and (3) the replacement of hydrogen of thearomatic sidechains by various groups.N / 1000-alcoholic solutions of the following substances wereexamined. The method of examination has been described before.?o--co>NH ; dark red.CPh :CPh 2 : 3-Diketo-4 : 5-diphenylpyrroline,>NH;co- coCPh:C( C,H,) 2 : 3-Diketo-4-phenyl-5-tolylpyrrolines (0-, rn-, p ) , 1..-dark red.>NH ; dark 2 : 3-Diketo-l-phenyl-5-p-cumylpyrroline,red.From the absorption curves (Fig. l), it is evident that theabsorptions are all of the same type. The differences correspondwith differences in the shades of red of the different compounds.For example, the ortho-tolyl compound (I) is not quite so darkred as the met* (111) and para- (V) compounds, and the curvesshow these differences, in that the rays are transmitted throughgreater thicknesses of the solution. It is noticeable, also, that thediphenyl compound (11) is a shade darker red than the ortho-tolylcompound, and not so deep red as the meta- and para-compounds.This is also clear from the curves, for the rays are transmittedthrough greater thicknesses than either the met* or para-tolylcompounds. Or, to put it another way, the absorption band ofthe ortho-tolyl compound is not so strong as that of either themeta- or para-tolyl compound; and the absorption band of thediphenyl compound is a little stronger than that of the ortho-tolylcompound, but not so strong as the bands of the metit- or para,-methyl derivatives.N / 10,000-solutions' of the substances were also examined, but nobands were observed in the ultra-violet regions. The positionsGO---GObPh: C(C9Hl,)8 c 2536 PURVlS: THE ABSORPTION SPECTRA OFwhen general absorption begin, expressed in oscillation frequencies,are :Diketodiphenylpyrroline 30 mm.thick .......................... 32469 , 9 9 ,, 10 ,, ,, ........................... 3997Diketo-o-tolylpyrroline 30 ,, ,, ........................... 3603), 9 9 ), 10 ), ,, ........................ 4199Diketo-m-tolylpyrroline 30 ,, ,, ........................ 33209 , Y 2 ,, 10 ), ,, ............ : .............. 3981Diketo-p-tolylpyrroline 33 ,, ,) ........................... 3148,J $ 9 ,, 20 ), ,, ........................... 3965Diketo-p-cumylpyrroline 30 ,, ,, ........................... 31539 , 9 9 ,, 10 ,, ,, ........................... 3593I. N/1000-alcohoEic solution of 2 : 3-dikato-4-phen~l-5-o-tolylpyrroli~.I I I. N/lOOO-alcoholic , , 2 : 3 -diketo-4-phenyl-5-rn-tolyl~jrroline.IV.N/lOOO-alcoholic ,, 2 : 3-diketo-4-phenyl-5-p-cumylpyrroline.V. NJ1000-alcoholic ,, 2 : 3-diketo-4-phenyl-5-p-tolylpyrroli~.11. N/lOQO-alcoholic ,, diketodiphenylpyrrolineVARIOUS DIKETOPYRROLINE COMPOUNDS. 2537Therefore the result of the substitution of hydrogen by analiphatic group in an aromatic side-chain does not fundamentallyalter either the colour or the absorption. The observed bandcorresponds in each case with the colour, and there is no bandproduced in the ultr%violet region.When the two diketo-groups are still left intact, and the sub-stitution in an aromatic sidechain is by the methoxy- or methylene-dioxy-groups, the dark red colour remains unchanged, but anotherFIG. 2.Oscillation frequencies.28 20 22 24 26 28 30 32 34 36 38I.N/ 10,000-alcoholic solution of 2 : 3-dik~to-4-phenyl-5-p-anisyllpyrro~~~e.11. N/10,00Q-alcoholic ,, 2 : 3-diketo-4-phenyl-5-pipermy@yrrolinc.band in the ultr%violet region is produced.two derived substances are:The formulae of these>NH ;TO- co 2 : 3-Diketo-4-phenyl-5-p-anisylpyrroline,2 : 3-Dike to-4-phenyl-5-piperonylpyrroline,CPh:C(C,H,*OMe)dark red.>NH;co---GOCPh : C( C6H,: O,:CH,)Idark red.N / 10,000-solutions were examined, and from the curves (Fig. 2)it will be seen that each substance has two bands. The lessrefrangible band of each corresponds with the bands of the previoussubstances. On the other hand, the more refrangible bands haveno corresponding bands in the original diphenyl compounds, andthey also differ from each other both in position and persistency.I n other words, neither the original colour nor the correspondingabsorption is fundamentally altered by the introduction of 2538 PURVIS: THE ABSORPTION SPECTRA OFmethoxy- or a methylenedioxy-group in an aromatic side-chain, butanother band is produced in the ultracviolet region correspondingwith the new type of side-chain.Further, the phenylhydrazones of the diphenyl compound andof the p-tolyl compound were examined in N / 10,000-solutions.Theconstitutional formulae of these subst,ances are :?(: N *N HPh)*CO>N .CPh====CPh Diketodipheny 1 pyrrolinephen y 1 hydrazone, Ycarmine-red.CP h- 7 : NCPh*NH.CN Diphenylpyrrolinophenazine, I I >CsH4 ; lemon-yellow.Diketopheny l-p-to1 y lpyrrolinepheng lh y drazone,?( :N:NHPh)*COCPh==C( C,H7) >NH ;carmine-red,GPh-- y:N >C,H, ; lemon- Phen yl-p-tolylpyrrolinophenazine,C(C,H,)*NH*C:Nyellow.Considering the carmine-red phenylhydrazones, it will be seenfrom the curves (Fig.3 and 4) that they show two bands. Theless refrangible band corresponds with the single band of theoriginal unsubstituted diketopyrrolines slightly shifted towards themore refrangible side; and the smaller, more refrangible one resultsfrom the introduction of the hydrazine radicle in place of oxygenof one of the ketonic groups. Considering the yellow-colouredphenazine compounds, the curves also show two bands, a smaller,less refrangible one, and a stronger one in the more refrangibleside.That is to say, the successive elimination of the oxygen ofboth the ketonic groups produces a change in colour from darkred, through carmine-red to yellow; a decrease in the intensity ofthe band of the original diketonic substances; and the productionof another band on the more refrangible side, the intensity of whichincreases by the successive elimination of the ketonic groups. It isimportant to notice that the increased weight of the molecule doesnot shift the band or the general absorption towards the red end,as is usually the case. On the contrary, the shift is towards themore refrangible end of the spectrum.General Results and Discussion.To sum up these observations, it is clear (1) that the diketonicstructure means the production of a dark red colour, anda corresponding well-marked absorption band, the positionof which differs slightly, corresponding with differences inthe dark red shade; (2) the introduction of a methyl oVARIOUS DIKETOPYRROLINE COMPOUNDS.2539FIG. 3.Oscillation frequenciPs.18 20 22 24 26 28 30 32 34 36 38\I. N/1OOO -alcoholic solution of diketodiphenylp yrroline.111. NJ10,QOO-alcoholic ,, diphenylpyrrolinophenazine.11. N/10,0OO-alcoholic , , diketodiphenylpyrrolinephenylhydrazonc.FIG. 4.Oscillation frepzcencies.18 20 22 24 26 28 30 32 34 36 38I. N/lOOQ-alcoholic solution of 2 : 3-diketo-4-phenyl-B-p-tolylpyrrolinc.11. N/10,000-alcoholic ,, 2 9 9 , phen ylhydrazone.111. NJ10,OOO-alcoholic ,, phcnyl-p-tolylpyrrolinophenazine2540 ABSORPTION SPECTRA OF DIKETOPYRROLINES.a propyl group in an aromatic sidechain produces nofundamental change either in the colour or in the absorptionband; (3) if either a methoxy- or a methylenedioxy-group isintroduced in an aromatic sidechain, the deep red colour andcorresponding band are still retained, but another more refrangibleband is produced characteristic of the type of the introducedradicle; (4) when one of the oxygen atoms of the diketonic groupsis replaced by the :N*NHPh group, the colour is changed fromm darkred to carmine-red ; the corresponding less refrangible absorptionband is reduced in intensity and its position is shifted towards themore refrangible side; and another weaker band is produced beyondthis on the more refrangible side; (5) the replacement of both oxygenatoms of the diketonic groups, and the production of the phenazine:N ring :N>C,H,, changes the colour from carmine-red t o lemon-yellow,and the corresponding band is also shifted towards the morerefrangible end.This band is also less intense, and the secondmore refrangible band becomes much stronger ; and (6) correspond-ing with these changes in colour and selective absorption, thepwitions of general absorption are shifted towards the morerefrangible region of the spectrum.The observations, then, indicate that the absorption is intimatelyconnected with the presence of a diketonic grouping. Baly andStewart (Trans., 1906, 89, 502) have suggested that the residualaffinities of the dicarbonyl compounds studied by them are oscillatingbetween two extreme phases.But in these compounds it mag beequally valid to say that the maximum valencies of the oxygenatoms come into action, resulting in the production of it closedring produced by the two oxygen atoms, and the consequent pro-duction of absorption. On this suggestion the ring is destroyedby the elimination of the oxygen in the hydrazone linking; theintensity of the original band is lessened, and another band isproduced, accompanied by changes in the colour and absorption inthe visible spectrum from the less to the more refrangible regions.The. further production of the phenazine ring does not whollydestroy the original absorption; it simply decreases it, and anotherband is produced characteristic of the new type of ring.From these considerations, it seems to be difficult to resist theconclusion that the origind band, which corresponds with the deepred colour, is caused by the oscillation or vibration of the originaldiketopyrroline ring, and that it is modified by the eliminationof the ketonic groups. In connexion herewith, it should bementioned that the author found no bands in phenylhydrazine insolutions of N / lo-, N / loo-, N / 1000-, and N / 10,000-strengthTAYLOR : RESEARCHES ON BLEACHING POWDER. 2541through varying thicknesses of 2 mm. to 30 mm. Also, Hartleyand Dobbie (Trans., 1898, 73, 598) found no bands in alcoholicsolutions of pyrrole, and this has been confirmed by the author(this vol., p. 1648).I have again to thank the Government Grant Committee of theRoyal Society, by whose assistance the spectroscope used in thisresearch was obtained, and also Dr. Ruhemann for specimens of thepure substances.UNIVERSITY CHEMICAL LABORATORY,CAMBRIDGE