251 POPE : FLUORONE DERIVATIVES. PART Ir.XXVI I I .-Flu or one Derivatives. Part ]I. Resorcinol-benxein.By FRANK GEORGE POPE.RESORCINOL-BENZEIN is the name which was originally given byDoebner t o the compound he obtained by the condensation ofresorcinol with benzotrichloride, when the reacting components wereheated together for some hours a t 180-190° (Annalen, 1883, 217252 POPE : FLUORONE DERIVATlVES. PART IT.234). The empirical formula C,,H,,OQ was assigned t o the benzein,and the reaction was considered t o take place in the followingmanner :C7H5C13 + 2C6H602 = 2HCl+ C,9H1504Cl,the intermediate chloro-derivative so formed being expected t oyield, by exchange of its chlorine atom f o r a hydroxyl group, atetrahydroxytriphenylcarbinol :C19H1504C1 + HOH = HCl + CI,H,,05.Apparently, however, anhydride f orination takes glace, andresorcinol-benzein is produced :2ClQH,504C1 + H 2 0 = 2HC1+ C3SH3009,and Doebner considers that structurally i t may be represented bythe expanded formula :[CI$%(OH)~’C(OH) (G3Hd *C&%(OH)]2O*The benzein was described as crystallising from a mixture ofalcohol and glacial acetic acid in large crystals, exhibiting a violet-red reflex.They remained unaffected when heated to looo, but onraising the temperature to 130°, two molecules of water were lost,and analysis of the residual compound pointed to the compositionC3,H,,07. No definite results were obtained either on acetylationor bromination.Somewhat later, Cohn ( B e y . , 1891, 24, 2064; J . p r .Chem., 1893,[ii], 48, 387) obtained the same compound by the condensation ofbenzoic acid with resorcinol in the presence of zinc chloride, andstated that it formed an unstable compound with hydrochloricacid, and yielded with bromine a series of different bromo-derivatives, according to the amount of halogen employed andthe temperature a t which the reaction was carried out.Here the matter rested for some time, until Kehrmann (Ber.,1908, 41, 3442; 1909, 42, 873) pointed out that resorcinol-benzeinwas probably identical with the 3-hydroxy-9-phenylfluorone that hehad obtained from the product of the action of benzotrichloride onacetyl-m ~aminophenol in nitroberizene solution. I n confirmationof this idea he prepared the benzein by Doebner’s method, andworking with a highly purified product showed that its molecularweight agreed with that of 3-hydroxy-9-phenylfluorone (C,gH1203),and that on methylation with methyl sulphate, the methyl etherof the Auorone was obtained.Pope and Howard (T., 1910, 97,1025) also drew attention t o the fact that the benzein might beconsidered as ideghical with Kehrmann’s 3-hydroxy-9-phenyl-fluorone.H. v. Liebig (J. pr. Chem., 1908, [ii], 78, 534) has, however,objected to Kehrmann’s views, and supports the older DoebnePOPE : FLUORONE DERIVATIVES. PART 11. 253view, considering that in the preparation of the benzein frombenzoic acid and resorcinol, benzoic anhydride may be used inplace of the acid, the reaction then proceeding:the compound thus obtained being Doebner's C3,H2,07.Numerousfresh derivatives of the benzein were described, amongst othersbeing an anhydro-compound, C,6H520,3,C2H,*OH, formed by theaction of ammonia on the benzein, a second anhydro-compound,C,6H,,0,4,CH3*C02H, obtained by heating the benzein with glacialacetic acid, a barium salt, C7,H4,0,,Ba,, a hydrochloride,C7,H,,0,,,4HC1, a compound, C3,H,,0g (m. p. 147*), formed byfusion of the benzein with an alkali hydroxide, an acetyl derivative,C,gH,20(O~CO*CH3)2,C2H5*OH, and a second acetyl compound,C3sH2,03(O*CO*CH3)4, prepared when the acetylation was carriedout in the presence of zinc dust.In 1912 v. Liebig (J. pr. Chem., 1912, [ii], 85, 97, 241) publishedtwo further communications dealing with the benzein, in whichno reference is made to the two different anhydro-compoundsobtained previously.Using the Doebner method of synthesis, heconcludes that the reaction mixture contains a-resorcinol-benzein, C19H1203, @-resorcinol-benzein, (C19H1203)3,C2H5*OH, andy-resorcinol-benzeh, (ClgHl,03)4,H20,C2H,*OH, and the dihydroxy-benzophenone compound of the y-benzein, of the compositionThe a-compound is considered as identical with Kehrmann's3-hydroxy-9-phenylfluorone, whilst the P- and y-benzeins are formu-lated respectively as :~ C ~ 9 ~ 1 2 0 ~ ~ 2 ~ H 2 0 ~ C 1 ~ H 1 ~ 0 ~ ~OH0/\/\/\OH/\B.Ph OEtOH OH0 0 0Using the Cohn method of synthesis, a 8-benzein, ( C,9E1203,R20)z,is also described, whilst the compound C38H3oOg (m. p. 147O),mentioned in the earlier paper, is now recognised as 2 : 4-dihydroxy-benzophenone254 POPE : FLUORONE DERIVATIVES. PART IT.I n the portion of v.Liebig’s paper dealing with his practicaldetails, the term resorcinol-benzein is used very indiscriminatelywithout reference to whether it applies to the a-, p-, y-, o r $-forms,and it is described as crystallising from nitrobenzene in red needles,having the composition (C,gH1203)4,3H20,C,H,*N02, and fromaniline in dark red needles with a blue reflex, having the com-position (C,9H,20,)4,C6H,*NH2. These two compounds containrespectively 1-05 and 1.12 per cent. of nitrogen, and the followinganalytical data are given:0*1128 gave 1.1 C.C. N, a t 21° and 722 mm.0.1092 gave 1.5 C.C. N, a t 15O and 724 mm.N=1*07.N=1*55.N (calc.) = 1-05.N (calc.) = 1.12.results which, seeing the small quantity of substance taken, and theconsequent amount of nitrogen obtainable, may be considered asopen to a considerable amount of doubt, whilst the origin of thewater of crystallisation ( !) is not apparent.Various hydrochlorides are also described, the a-benzein yieldingone of the composition C,gH,203,HCI; the p-, one of the composition(C,SH,20,)3,3HCl,C,H5*OH ; and the y-, one of the composition(C,g13[,203)4,4HC1,H20. The following observations may perhapsbe offered with regard to the /3- and y-benzeins.The 8-form havingthe composition (C,gH,203)3,3C2H,*OH, when heated t o 140°, losestwo molecules of aTcohol, and since the third molecule is difficult t oremove, the product obtained is considered as an ether-like com-bination of the benzein and alcohol, namely :OH 0 EX0 10 0 0 1 0OH/\/\/\CPh C CPh/\P h OEtbut it is difficult to see why the ethoxy-group should be attachedin this way, and no reason is given for affixing it in the positionshown.When heated with alcoholic ammonia and the excess ofammonia is subsequently removed, it is said to yield a compoundof the composition (ClgH,203)3,H,0,2C2H5*OH, which a t 140° loseswater and alcohol, giving (C19H1203)3,C2H5*OH, finally passing at240° into (c1gH&3)3.With regard to the y-benzein, obtained by the action of alcoholicammonia on dihydroxybenzophenone-y-resorcinol-benzein, this iPOPE : FLUORONE DERIVATIVES. PART 11. 255represented as (ClgHlZO3),,2HzO,C,H5~OH.A t water-bath tem-perature, by loss of water, i t passes into the compoundwhich loses alcohol when heated to 140°, yielding (C,,H120,),,H,0or0 0 0 0 0 0(C,,H,20,)4,H,0,CZH5'0H,OH OHI I I I I I I l l v\<y V>Q/ PhC--0---and finally, a t 240°, the last molecule of water is removed, and theanhydrous product, C19H1203, results. It seems curious that, onheating, the water molecule should be removed before the alcoholmolecule, and that v. Liebig in this instance does not make use ofan ether formulation t o explain the tenacity with which the alcoholmolecule is held, whereas such a method was adopted with the&compound.With the object of trying to decide between the conflictingopinions of Kehrmann and v.Liebig, resorcinol-benzein was pre-pared by the Doebner method, and the following results wereobtained. The crude product, when boiled with water, yielded anaqueous extract containing benzoic acid, some 2 : 4-dihydroxybenzo-phenone, unchanged resorcinol, and a small quantity of a redcolouring matter which was not further examined. The residualsolid was digested with dilute ammonia and the mixture filtered,when a reddish-coloured residue was left. This is apparently thebenzoyl derivative of the benzein, since it is insoluble in coldalkali, and on hydrolysis by means of mineral acid is decomposedvery slowly with formation of benzoic acid.The ammoniacal filtrate was acidified with acetic acid, and theprecipitated solid purified by conversion into its hydrochloride.The hydrochloride was dissolved in ammonia, the base then pre-cipitated by the addition of acetic acid, and finally recrystallisedfrom a suitable solvent.It was a t first thought that by fractional crystallisation of thehydrochloride, two isomerides were obtained, but on liberation ofthe base from the two specimens, compounds of identical meltingpoints were produced, and when a mixed melting point of the twospecimens was taken, no depression was observed.Moreover, onfusion of the two specimens wifh an alkali hydroxide, both yielded2 : 4-dihydroxybenzophenone, and thus apparently only one productof reaction is produced.The resorcinol-benzein so prepared is identical with Kehrmann's3-hydroxy-9-phenylfluorone, as is shown by the analytical dat256 POPE : FLTJOROKE DERIVATIVES. PART 11.obtained from the base, its hydrochloride, the sodium and bariumsalts, the acetyl derivative, and the acetyl derivative obtained onacetylation in presence of zinc dust.No trace of v. Liebig's varioushydrochlorides was observed, and the same analytical results wereproduced whatever solvents were used for recrystallising thebenzein. Liebig's anhydro-compounds could not be obtained eitherby the action of alcoholic ammonia or acetic acid; in each case theproduct formed gave tEa melting point of the unchanged benzeinand furnished correspmding data on analysis, whi1,st no depressionof melting point was observed on mixing the so-called anhydro-compounds with a specimen of the benzein.I n order to obtain confirmatory evidence of the fluorone structureof the benzeiD, i t was subjected to the action of phosphorus penta-chloride,, when the 3 : 6-dichloro-9-phenylxanthonium chlorideobt,ained by Howard and Pope (T., 1911, 99, 550) from the actionof the pentachloride on 3-hydroxy-9-phenylfluorone was produced,the identity being further established by the conversion of thexanthonium chloride into 3 : 6-dianilino-9-phenylxanthenyl chlorideand the corresponding 3 : 6-di-~-naphthylamino-compound on con-densation with aniline and P-naphthylamine respectively.These facts would seem to justify Kehrmann's view as to thestructure of resorcinol-benzein, and in the annexed table theformulze of the variousauthor are contrasted.Resorcinol- henzein ............a- Kesorcinol-benzein ............B-Elesorcinol- benzein .............y-Kehol.ciilol-l-lenzAin ............8-Resoi ci ri ol-be1 t zein ............Anhydroresorcinol-bcnzein (I)9 9 9' (11)Hydrochlorides :a ..............................8 ..............................y ..............................Barium salt ........................Acetyl derivative ..............Reduced acetate ...............Alkali fusion .....................products obtained by v.Liebig and theH. v. Liebig. F. G . Pope. I 'OH jIEXPERIMENTAL.For the preparation of resorcinol-benzein, the method adopted byKehrmann and Dengler (Ber., 1909, 42, 873) was used, and theproduct obtained was purified by means of its hydrochloride.It was thought that possibly the hydrochloric acid solution fromwhich the hydrochloride of 3-hydroxy-9-phenylfluorone waFOPE : FLUORONE DERIVATIVES.PART 11. 257originally obtained might contain an isomeride, conceivably of thestructure :0and with the object of ascertaining whether such were the case orno, the acid solution wap concentrated, and the crystalline residueobtained subjected t o purification by Kehrmann's method. Thebase obtained was, however, identical with 3-hyclroxy-9-plienyl-fluorone, and when the two were mixed, no depression of maitingpoint could be noticed, so that apparently only one product of thefluorone type is formed.Resorcinol-benzein is readily soluble in hot nitrobenzene, aniline,or pyridine, but sparingly so in alcohol. From the three formersolvents it crystallises in small, red needles, which melt at 330-331O(Found, C=79*29; H=4.25.Calc., C=79.17; H=4*17 per cent.).The hydrochloride, C19H120,,HC1, is obtained by dissolving thebase in alcohol and adding concentrated hydrochloric acid to thesolution (see above). It crystallises froin alcoholic hydrochloricacid in golden-yellow needles :0.2240 gave 0.5748 CO, and 0.0826 H20.0.1747 ,, 0.077 AgC1. C=70*0; H=4*09; C1=10*90.C19H,,03,HC1 requires C = 70.25 ; H = 4.01 ; C1= 10-94 per cent.The sodium salt is prepared by suspending the base in s;ater,and adding to the suspension one equivalent of sodium hydroxidealso dissolved in water. A deep red solution, showing a yellowfluorescence, is produced, and on concentration the sodium saltseparates in very small, red needles:0.2838 gave 0.0652 Na,SO,.Cl,Hl103Na requires Na = 7-42 per cent.v.Liebig states that a soluble barium salt, C,,H,,O1,Ba,(Ba = 32.35 per cent.), is formed when resorcinol-benzein is heatedwith baryta solution. On repeating this experiment it was foundthat the benzein is almost insoluble in cold baryta solution, butthat on boiling it appears t o dissolve, and a barium salt, in animpure condition, or possibly a basic salt, is formed. If, however,the sodium salt is first prepared and barium chloride added t o itsaqueous solution, a bright red precipitate of the barium salt of thefluorone is immediately produced. This was collected, well washedwith hot water, and dried:Na = 7.44.VOL. cv.258 POPE : FLUORONE DERIVATIVES. PART 11.0.1253 gave 0.0408 BaS04.(C19H,,03),Ba requires Ba = 19.27 per cent.The acetyl derivative, C,gH,,02-O*CO*CH,, was prepared byheating 5 grams of the benzein with 20 grams of acetic anhydrideand 5 grams of anhydrous sodium acetate for four hours. Themixture was then poured into cold dilute alcohol, and warmedon the water-bath until the excess of the anhydride was destroyed.The residue was collected, washed, dried, and crystallised fromalcohol or acetic acid, separating from the solvent in red needles:Ba = 19.14.0.1449 gave 0.4046 CO, and 0.057 H,O.By carrying out the acetylation in the presence of zinc dust, theC=76*15; H=4*37.C,,R,,O, requires C = 76.31 ; H = 4.27' per cent.colourless 3 : 6-dihydrozy-9-phenylzanthen diacetate is obtained :0and on crystallisation from alcohol, separates in colourless needles,which melt a t 175O:0.1048 gave 0.2840 CO, and 0.0436 H20.C,,H,,O, requires C = 73-80 ; H = 4-81 per cent.According to v.Liebig, a compound which he designates asanhydroresorcinol-benzein, C,,H,,O,,,C,H,*OH, is obtained whenthe parent substance is heated with alcoholic ammonia. In orderto test this statement, 4 gram? of the benzein were heated on thewater-bath for two days with successive quantities of alcoholicammonia, the product thoroughly dried, and compared with theoriginal benzein, with the following results :C = 73.91 ; H =4*62.Besominol- benzcin.Red needles, m. p. 330-3.31".Hydrochloride, yellow.A nh ydroreso rciii o 1 - benze in.Red needles, m.p. 330-331".No depression of m. p. on inixiiigwith original substancc.Hydrochloride, yellow.Analysis of the so-called anhyclro-compound gave the following0.1758 gave 0.5084 CO, and 0.0652 H,O.v. Liebig's anhydro-compound would require C = 74.88 ; H = 4*64per cent. It would thus seem apparent that alcoholic ammoniais without action on resorcinol-benzein a t water-bath temperature.Again, a second anhydro-compound, C,,H,O,,,CH,*CO,R, is saidto be formed on heating the benzein with glacial acetic acid. Thisexperiment was repeated by heating the benzein for twelve hoursdata :C=78.88; H=4*12.C1gHI2O, requires C = 79.17 ; H = 4-17 per centPOPE : FLUORONE DERIVATIVES.PART 11. 259with the acid. Itmelted a t 330-331°, and was in all respects identical with theunchanged benzein :The product obtained was collected and dried.0.2667 gave 0‘7691 CO, and 0*1006 H,O. C=78.65; H=4.19.C1,H,,O3 requires C = 79.17 ; H = 4.17 per cent.C,,H,Ol,,CH,*CO,H requires C = 74-88 ; H =4.64 per cent.In order further to establish the identity of the benzein with3-hydroxy-9-phenylfluorone, the reaction between the benzein andphosphorus pentachloride was examined. For this purpose 5 gramsof the benzein were dissolved in 20 grams of phosphoryl chlorideand warmed on the water-bath, with the gradual addition of 10grams of phosphorus pentachloride. The blood-red liquid obtainedwas warmed until all the pentachloride dissolved, allowed to remainfor five minutes, and then poured into 300 C.C.of light petroleum.The yellow precipitate obtained was collected, well washed withlight petroleum, and dried in a vacuum. It is insoluble in non-hydroxylic solvents, but readily decomposes in the presence ofhydroxylic solvents. A chlorine estimation served to establish itsidentity with the 3 : 6-dichloro-9-phenylxanthonium chloridedescribed by Howard and Pope (T., 1911, 99, 550) (Found,C1= 29.06.Confirmation of this result is seen in the reaction of the abovechloro-derivative when brought into contact with aniline and with8-naphthylamine. I n the first case 5 grams of the dichlorophenyl-xanthonium chloride were heated for five minutes with 25 C.C. ofaniline, and the resulting paste was added t o 200 C.C.of water, thewhole rendered alkaline, and distilled in a current of steam t oremove the excess of aniline. The residue was acidified, the pre-cipitate collected, well washed with water, and dried. For purifi-cation, the bluish-purple solid was recrystallised from alcohol,separating from the solvent in very small needles, which dissolvein alcohol to a purple solution. Analysis showed that the substanceobtained wa3 identical in composition with the dianilinophenyl-xanthenyl chloride obtained fsom 3-hydroxy-9-plienylfluorone(Found, C = 78.60 ; H = 4.6 ; C1= 7.28. Calc., C = 78-39 ; H = 4-84 ;Cl=7.48 per cent.).Similar results were obtained by the condensation of the dichloro-xanthonium chloride with B-naphthylamine, the xanthenyl chlorideformed recrystallising from nitrobenzene in small crystals showinga fine bronze reflex, and giving analytical data corresponding withthe xanthenyl chloride obtained similarly from 3-hydroxy-9-phenyl-fluorone (Found, C1= 5.97.From the above results it thus appears that resorcinol-benzeinas obtained by Doebner method is identical with Kehrmann’sCalc., C1= 29.49 per cent.).Calc., Cl = 6-18 per cent.).s 260 WORLEY: THE SURFACE TENSION OF MIXrURES.PART I.3-hydroxy-9-pl1enylfluorone, and is not a mixture of various forms,since the derivatives obtained by working with the two compoundsare the same in both cases.On fusion with alkali hydroxides, or on boiling with a moderatelyconcentrated solution of potassium hydroxide, the benzein yieldsresorcinol and 2 : 4-dihydroxybenzophenone. To recover the latcer,the fusion was acidified with hydrochloric acid and filtered. Theprecipitated matter was extracted with cold alcohol, the alcoholicsolution evaporated to dryness, and the residue repeatedly re-crystallised from hot water. I n this manner, pale yellow needles,melting a t 143-144O, are obtained. This is the melting point of2 : 4-dihydroxybenzophenone, and analysis furnished data confirm-ing this conclusion (Found, C = 72.98 ; H = 4.7. Calc., C = 72.90 ;H=4.7 per cent.)."I n conclusion, I desire to express my thanks to Mr. E. Turner,B.Sc., for his help in the practical work, and to the Research FundCommittee of the East London College for a grant to defrayexpenses.EAST LONDON COLLEQE