428 XXXVI.-Ca?.boxyl-derivativRs of Benxoquiiaone. By J. U. NEF.* CHLORO- bromo- hydroxy- nitro- and amido-derivatives of benzo-quinone have long been known but attempts to obtain carboxyl-derivatives have hitherto been unsuccessful. The following experiments undertaken with the intention of filling up this gap were carried out at the suggestion of Prof. v. Baeyer, and it is an especial pleasure to the author t o have this opportunity of expressing his gratitude to Pi-of. v. Baeyer for the kindness and interest shown him during his stay in Muiiich. Durene was chosen as the startinq-point f o r the experiments on account of the especial interest attaching t o the production of quinono-tetracarboxylic acid C,oH,O, = C,O,( COOH), which has the same percentage composition as croconic acid C5H305 one of the products obtained by Gmelin (Qmelin Handbuch 4 Aufl.5 478) from the explosive bye-product formerly obtained in the preparation of potas-sium. The Russian chemist Basaroff (Dict. de Chemie par WZrtz, 2 1363). expressed the opinion that croconic acid and quinone-tetracarboxylic acid were identical but very recently Nietzki and Benckiser (Bey. 19 293 772) by a series of remarkable experiments, have shown that croconic acid has the formula C5H305 thus render-ing i t impossible that it can be identical with a derivative of benzene. The following synthesis of ethylic quinonetetracarboxylate confirms the conclusion that the two compounds are totally differem. A close investigation of etliylic quinonetetracarboxylate has led to the discovery of a series of compounds which stand in a remarkably close relation to ethylic succinosuccinate and its derivatives.A study of durylic-acid-qninone C,O,(CH,),COOH has shown that carboxylated quinones are very unstable substances and that the carboxyl-group which is usually so stable in benzene-derivatives. can easily be displaced by other groups ; the nitro-group for example. On the whole however they exhibit the general chemical reactions charactmising other quinone-derivatives On reduction they take up two atoms of hydrogen forming colourless quinols which on oxida-tion yield the original quinone. A very noteworthy fact is the existence of many carboxyl-pamdi-hydroxy-derivatives of benzene which cannot under any conditions be oxidised to the corresponding quinones.The oxidation to a * Compare Annalen 237 1 NEF CARBOXTL-DERIVATIVES OF BENZOQUINONE. 4 I 9 quinone can only be accomplished in the case of hexa- substituted benzene-derivat,ives which in this respect exhibit a much greater stability than other incompletely substituted derivatives of benzene. The durene used was obtained from the chemical factory of Lang-feld and Reuter Bramom near Rostock an2 was made from mono-bromopseudocumene C,H,(CH,),Br [I 2 4 51 (m. p. 71") by Fittig's method. It was nut pure as the boiling point ranged from 185" to 220", and contained moreover a liquid boiling a t about 170° and a solid compound melting a t loo" and boiling at 240". D u r j l i c acid was prepared from pseudocumidine CsH,( CH,).N.H2 [l 2 4 51 by Sandmeyer's method of displacing the amido- by the cyanogen-group (Hdler Ber.18 93). Although the yield of durotiitrile is poor (10-15 per cent.) arid the preparation of the acid therefore wearisome this is by far the best method of making durylic acid and the product obtained is pure (m. p. 149'). As pseudo-cumidine is now much used in the colour industry in making azo-dyes and therefore easily procurable I have recently tried to better the yield of duronitrile by modifjing Sandmeyer's method somewhat. Cnprous cyanide was used the solution in potassic cyanide being kept boiling over a flame while the diazo-pseudocumene solution was added. After extracting with ether washing the ethereal solution with dilute sodic hydrate and sulphuric acid and distilling off the ether the nitrile was distilled over with steam.It was thus obtained pure fusing a t 57.5". It can be saponified quantitatively by heating it with concentrated hydrochloric acid (1.18) in a sealed tube at 160-170". The yield thus obtained was about 50 per cent. which is far better than that afforded by the usual method. The course pursued in the iitvestigation was the fo€lowing :--The durene was converted by the usual methods into duroquinone, C60,( CH,),. As attempts to oxidise the methyl-groups to carboxyl-groups were unsuccessful the durene was used solely to make durylic acid which on nitration gives dinitrodurylic acid ; this acid was then converted into durylic-acid-quinone and pyromellithic-acid-quiiione. I. Duroquinone from Dureize.Dinitro-durene was made according to Jannasch and E'ittig's method (Zeit. f. Chem. 1870 161). The durene (5 grams) was added slowly to pure fuming nitric acid (1*49) 40 c.c. kept at a tempe-rature below 0". The substance rotates rapidly and then dissolyes with a passing deep-brown coloration. After allowing the mixture t o remain for two hours it is poured on to ice when the dinitro-durene separates out as a white flocculent precipitate. By crystal-lisation from alcohol i t can be separated from an oily bye-product 430 NEF CARBOSTL-DERIVA'I'IVES OF BESZOQUISONE. nnd obtained in colourless prisms melting a t 205". It has all the properties ascribed to it by Jann,zsch and Fittig; it dissolves easily in benzene and ether sparingly in hot and very little in cold alcohol.I t sublimes without decomposition in needles and volatilises with steam. It is unusually stable in the presence of oxidising agents; dilute nitric acid or a mixture of 1 vol. glacial acetic acid 1 vol. concent~ated nitric acid and 1 vol. of water (in which the substance dissolves) not having the slightest action on it. Dinitrodurol is best reduced as follows :-A hot solution of the substance in acetic acid is treated with zinc-dust until it no longer gives any turbidity on addition of water. The zinc is then precipitated from the diluted solution with sulphuretted hydrogen and the filtrate concentrated. On addition of sodic hjdrate diamidodurene is pre-cipitated in colourless pearly plates ; this was not analysed but con-verted directly into duroquinone.It is easily Roluble in chloroform and alcohol ; less so in ether. When exposed to the air in solution, or in a moist condition it turns green. Thc hydrochloride of the base in sparingly s o h ble in concentrated hydrochloric acid. An aqueous solution thereof treated i n the cold with ferric chloride, platinic chloride or sodic nitrite is oxidised to duroquinone. I n the case of platinic chloride the addition of nlcoliol precipitates am-monium platinochloride wliilst duroquinone remains in solution, (NH4),PtCl6 + 2HC1 + H,O. A similar easy elimination of the amido-groups with formation of a quinone was also observed in the case of diamidodurylic acid. The duroquinone was made by treating a solution of diamido-dcrene in hydrochloric acid with an excess of ferric chloride in the cold.A temporary deep-green coloration is noticed and then the solution becomes yellow and the quinone crystallises out in yellow needles. The yield is almost quantitative and the product obtained is pure. If the diarnidodurene is not isolated after the reduction with zinc-dust and the ferric chloride is added directly to the acetic acid solution (from which the zinc has been removed) the quinone is obtained in small quantity as a very dark-coloured impure product. nuroquinone was obtained by cryst:illisation from light petroleum ill beautiful long yellow needles melting a t 111". 0.1220 gram substance dried in a vacuum gave 0.3259 gram CO, and 0.0815 H,O. Cti(CH,),(NH,Cl)? + HzPtC1 + 2HZO + 0 = Ctj(CH,),O + CtLlculilted for c (GI3 3) 4 0 2 .Found. C . 73-18 72-86 H . 7-32 7.42 The quinone is volatile with steam and sublimes at 100" in needles NEB' CBRBOXYL-DERIVATIVES OF CENZOQUINONE. 43 I which possess a feeble but distinct odour characteristic of quinone. It is very easily soluble in alcohol ether chloroform and in hot light petyoleum ; insoluble in water or alkalis. It was not possible under any condition to oxidise the methylated side-chains to carboxyl ; oxidising agents in alkaline neutral or acid solution decomposed the benzene ring compietely. Worthy of note is the great stability of duroquinone in tlie presence of concentrated nitric acid (1*4) in which it dissolves with ease but even after long heating on a water-bath remains in great part unchanged. Treated with zinc-dust and acetic acid the quinone is reduced and on adding water a colourless substance melting at about 210" is pre-cipitated.Oxidising agents such as ferric chloride or nitric acid, oxidise it again to duroquinone. It is probable that the substance is dihydroxydurene but want of material has prevented me from obtain-ing the product in amount sufficient for an analysis. IT. Dinitrodurylic Acid out of Durol. DtwyZic A c i d C,H,(CH,),*COOH.-Jannasch (Zeit. f. Chem. 1870, 449 and 1871 33) has shown that durene heated for a loiig time wibh dilute nitric acid yields in about equal amount two acids one of which durylic acid C6H,( CH3),-COOH is volatile with steam ; the other cumidic acid,* C6HL(CH3)2(COOH)2 is not volatile with steam. I n my experiments however the great object was to obtain as much of the durylic acid as possible.When durene is heated with dilute nitric acid for three or four hours only and the acid formed is then separated froni unchanged durene I find that durylic acid alone is formed. Durylic acid is also the sole oxidation-product obtained on treating a solution of durene in acetic acid with the calculated amount of chromic acid also dissolved in acetlic acid. The yield in this case is however not so good as when the following method is employed. 20 grams durene is heated for three to four hours with 500 C.C. dilute nitric acid (1 V G ~ . concentrated nitric acid sp. gr. 1.4 to 3 vols. water) taking care t o add a few pieces of porous porcelain to pre-vent bumping. After filtering from the nitric acid the product is treated with dilute sodium carbonate solution aud the unchanged durene heated agaiu with nitric acid as above.After the second and third treatment with sodium carbonate an insoluble yellowish oil remains which can easily be separated from the soda solution by means of ether. On distilling off the ether an oil is left which cou-tains nitrogen and undergoes little change on further heating with nitric acid.? * This has recently been shown to consist of two isomeric acids (Schnapituff, Be? 19 2510). + The oil has a peculiar camphor-like odour and is very easily soluble in ether 432 NEF CARBOXYL-DERIVATIVES O F BENZOQUISOSE. On adding hydrochloric acid to the mixed sodium carbonate solu-tion impure durylic acid separates as a flocculent precipitate.It is first treated in acetic acid solution with zinc-dust to get rid of the nitro-products possibly formed and then distilled with steam. As the whole is volatile with steam there can be no bibasic cnmidic acid formed. The durylic acid thus driven over with steam is pure, melting a t 149" and has all the properties assigned to it by Jannasch (Z'eit. f. Chem. 1870 449). It is very sparingly soluble even in boiling water and volatilises with steam without melting. The yield obtained by the above method was 15-20 grams of impure durylic acid and 20-22 grams of oily bye-product from 40 grams of durene. I n the further experiments, the impure durylic acid obtained on adding hydrochloic acid to the soda solutions was used. (2.) Dinitrodurylic Acid C6(N0,),(CH,),GOOH.-The nitration of durylic acid according to Gissrnann's method was not found advan-tageous as the reaction is too violent.The following process gives a quantitative yield :-20 grams finely pulverised durylic acid (dried at lOO0) is dissolved in pure concentrated sulphuric acid and the solution cooled down to - 10" or - 15" by means of a freezing mixture. A solution of 28 grams potassic nitrate in concentrated sulphuric acid is now added as quickly as possible-taking care to stir well and to keep the temperature below 5". A deep brown colour appears at first which changes finally to light yellow while dinitrodurylic acid partly separates. After allowing the mixture to remain a t the ordinary temperature for four or five IIOUTS o r over night the solution is poured cwefully on to ice.The nitro-acid separates out as a flocculent yellowish precipitate which is collected and well washed. If pure durylic acid is used the product thus obtained is sufficiently pure; if however the impure acid is employed the nitro-derivative must be purified by means of its calcium salt. The crude product is heated with water and finely pulverised marble until the reaction is neutral ; on filtering and concentrating the calcium salt crystallises out in colourless radiating needles. When lieated on platinum-foil it explodes violently. An analysis of the oil as well as of a solid product obtained therefrom by cooling below Oo and pressing between filter-paper (melting point 35-40") gave figures agreeing with those required by mononitrodurene.The oil however does not yield dinitrodurene on treatment with fuming nitric acid and it is lherefore more proba-ble that the oil is a nitrate of a trirnetliylated bcnzyl alcohol. This i the more probikble as by oxidation with potassium permanganate in soda solution acids are obtained which contain no nitrogen. I n this way I succeeded in isolating an acid possessing all the prbperties of pyromellitic acid and the analp3is of the silver salt gave figures which are exactly those required by C,II,(COOAg),. (Gissmann Annalen 216 20s. NEF CARBOSYL-l)ERIVATIVES OF BENZOQUINONE. 433 0.3155 gram substance dried a t 180" gave 0.0760 gram CaS04 = 7.10 per cent. calcium ; calculated 7.32 per cent. The free acid obtained by the addition of hydrochloric acid t o the purified calcium salt is a yellowish amorphous powder melting a t about 205".It has all the properties mentioned by Gissmann The potassium salt is sparingly soluble in cold water and crystal-(loc. cit.). lises in colourless lustrous needles. 111. Bury lic Acid Quiraone f r o m Dinitrodurylic Acid. (1.) Diarnidodurylic Acid Cs(NH,)2(CH3),*COOH.-A solution of dinitrodurylic acid in acetic acid (50 per cent,.) is treated with zinc-dust in small portions; a violent reaction takes place at first and finally the soliltion becomes colourless. Sfter filtering hot from the excess of zinc-dust the solution is diluted wit'h about 4 volumes of water and allowed to cool when the diamidodurylic acid crystallises out almost completely i n colourless silky needles.The small portion remaining in solution is recovered by precipitating khe zinc concen-trating the a,mmoniacal solution and adding dilute acetic acid. The yield is almost quantitative. For analysis the substance was crystallised from water several times, and then dried carefully a t 110". 0.1642 gram anhydrous substance gave 0.1105 gram H,O and 0.0950 gram anhydrous substance gave 12 C.C. nitrogen a t 15' and 0.371 gram CO,. 723 mm. Calclrlated for CloH,,h',02. Found. C . 61.85 61.62 H . 7.22 7.48 N . 14.48 14.10 The figures obtained for carbon were a t first always about 1 per cent. too low because the substance sticks together on healing and it is difficult to remove the last trace of water of crystallisation. Diamidodurylic acid forms salts with bases and with mineral acids.The compounds with hydrochloric acid and with snlphuric acid are easily soluble in water. On adding platinic chloride to a solution of the hydrochloride in the cold ammonium platinochloride separates and dui-ylic-acid-quinone is formed. The oxidation to the quinone also takes place with equal ease on treatment with ferric chloride or sodic nitrite. Diamidodurylic acid is sparingly solable in cold alcohol or water ; much more soluble in hot; it is insoluble in ether. It melts at 221 434 NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE. with decomposition. When heated wit,h acetic anhydride for several hours at 140" an acetyl-derivative was formed. After getting rid of the acetic anhydride the substance was dissolved in ammonia and pre-cipitated therefrom by hydrochloric acid; in this way i t was pre-cipitated in colourless quadratic plates melting at about 275".When heated in a test-tube the substance loses carbonic anhydride and a compound sublimes in colourless needles ; it is insoluble in alkalis. 0 CH /\ COOH (2.) Durylic-acid-quiriowp .-This compound the CH3 (2 CH3 first carboxylated qninone-derivative known is formed quantitatively on adding an excess of ferric chloride to a solution of pure diamido-durylic acid in hydrochloric acid and allowing the mixture t o remain for about half an hour. No colour reactions are noticed but the yellow solution bas a faint but distinct quinone-like odour. If the solution is sufficiently concentrated part of the acid crjstallises out in yellow rhombohedrons.The solution is extracted three or four times with ether and the ether dried with calcium chloride ; after distilling off part of the ether and allowing the remainder to evaporate spontaneously the quinone acid remains as a dark-yellow crystalline pubstance having a distinct quinone odour. I f the above directions are carefully followed it is only necessary to wash away the traces of adhering ferric chloride with a little water in order to obtain a perfectly pure product. On heating with ferric chloride or on trying to convert dinitrodurylic acid directly into the quinone without isolating the diamido-compound a very impure yellow oily product is obtained. The c r j stals obtained from the ethereal solution are radiating flat needles of a deep yellow colour.Heated in a capillary tube they become red a t 127-129" and decompose at 130" with brisk evolution of gas. I. 0.1340 gram substance dried in a vacuum gave 0.3038 gram CO, 11. 0.1540 gram substance dried in a vacuum gave 0.346 gram GOz and 0.065 gram H,O. and 0.076 gram H,O. Calculated for Cs02(CH,),.COOH. C 61-85 H 5.15 Durylic-acid-quinone dissolves easily solvents except light petroleum. It Found. r---7 I. 11. 61.83 61-68 5.40 5.52 in alcohol and in other organic is sparingly soluble in col XEF CARBOSYL-DERIVATIVES OF BEKZOQUIXONE. 4315 water more soluble in hot water (the quinone odour is then especially noticeable) but does not crystallise out well on cooling. It dissolves in benzene in the cold unchanged but on heating t8he yellow solution changes to an intense dark red.Durylic-acid-quinone difl'ers from other quinones in being lion-volatile and easily decomposed. Being ail acid it decomposes soluble carbonates in the cold and dissolves in alkalis and in ammonia with a deep yellow colour. The alkaline solutions can be partly evaporated without change. In other respects, the quinone acid resembles the other paraquinone compounds of benzene; thus for example it can be reduced in alkaline or acid solution to the corresponding quinol or paradihydroxy-compound which on oxidation yields the original quinone compound again. An intermediate compound corresponding to quinhydrone was not noticed. Like quinone itself the above quiiione acid reacts with phenol and resorcinol forming dark-red compounds.On allowing a dilute alcoholic solution of the quinone acid to stand for two days with an excess of hydroxylamine hydrochloride a yellow unstable compound containing nitrogen is formed ; this is probably the quinoneoxime. Of the salta of durylic-acid-quinone the silver salt was prepared by adding silver nitrate to a concentrated solution of the neutral animoninm salt. It was thus obtained in yellow needles little soluble in cold water more so in hot. It is unstable towards heat and light, and explodes when heated quickly on platinuni foil. 0.146 gram substance dried in a vacuum gave 0.041 gram H,O, 0.2125 gram COz and 0.0525 gram Ag. Calculated for C6 02(CH,) - co Odg . Found. c 39.87 39-70 H 2.99 3-12 Ag 35.88 35.96 A concentrated solution of the ammonium salt of durylic-acid-quinone treated with lead acetate gives the lead salt as an insoluble yellow amorphous precipitate.Bark chloride gives the barium salt as a yellow crystalline precipitate dissolving on heatii1,g. Copper sulphate gives the copper salt as a pale-yellow sparingly soluble granular precipitate. ( 3 . ) Dihydrox y dury l ic Acid C (OH) *( CH,) 3*C 0 OH .-Dury lic-aci d-quinone can be redaced quantitatively in alkaline or acid solution to dihydroxydurylic acid. I n my first experiments I used zinc-dust and sodic hydrate. A dilute solution of the substance in sodic hydrate was treated in the cold with zinc-dust added in small portions a t 436 NEF CARBOXYL-DERIVATIVES O F BESZOQUINOSE. time. The solution turns green then violet and finally becomes colourless ; it is then acidified with concentrated hydrochloric acid, and extracted with ether.On distilling off the ether a wliite amor-phous mass remains having a characteristic sweet sniell. This method of reduction is uncertain on account of the instability of dihydroxydurylic acid in alkaline solution and it is much better to reduce with sulphurous acid which a t once gives a pure product. On adding a concentrated cold aqueous solution of sulphurous acid to finely pnlverised durylic-acid-quinone and shaking the whole mass is converted into dihydroxydurylic acid without dissolving ; in order t o obtain a colourless product the mixture is heated in a sealed tube a t 100" for two to three hours and then repeatedly crystallised out of water containing sulphurous acid.Dihydroxydurylic acid is thus obtained in colourless spheres made up of very fine radiating needles ; i t melts at 210" with decomposition. I. 0.1432 gram substnnce dried a t 110" gave 0.3208 gram CO, 11. 0-1556 gram substance dried a t 110" gave 0.3485 gram CO, and 0.084 gram H,O. and 0.089 gram H,O. Found. Cidcnlated for r-7 C (OH) 2 (C H3) 3.C 0 OR. I. 11. C 61.27 Gl.l.0 61.08 H . 6.13 6-52 6.36 This quinol dissolves easily in alcohol and ether sparingly in benzene and chloroform. Ferric chloride added to a dilute alcoholic solution oxidises it with passing green coloration t o the quinone. The colourless alkaline solution of the acid is unstable when exposed to the air changing quickly to a yellow red red-violet and filially to a deep violet.An ammoniacal solution of the acid reduces silver nitrate in the cold. Lead acetate added t o a solution of the ammo-nium salt gives a colourless amorphous precipitate. (4.) 13thyZ quinonedumjhte C6O,(CHr,),.CO0C,H5 was made by allowing the sili-er salt of durylic-acid-quinone to stand for 24 hours with a dry ethereal solution of ethyl iodide. The ethereal solution, filt)ered from the silver iodide was washed with sodic carbonate and dried by means of calcic chloride. After distilling off the ether cz yellow oil remained which on rubbing with a glass rod solidified readily to yellow radiating needles. On recrystallisation from hot light petroleum it was obt~~iried in long yellow needles melting a t 51".0.12% gram substance dried in a vacuum gave 0.0706 gram H,O and 0.2902 gram CO, NEF CARBOSYL-DERIVATIVES OF BENZOQUINONE. 43 7 Calculated for C6O9 (CH,) ,*COOC,H+ Found. c 64.86 64-61 H 6.31 6.40 The ethereal salt unlike the free acid is odourless. Like other quinones it sublimes with ease and without decomposition. It is easily soluble in alcohol and ether ; insoluble in water or alkalis. It is saponified when heated with sodic hydrate. (5.) Ethyl Dihyclroqdzwy Zate C6(OH),(CH,),*C00C?Hs.-on heating ethyl quinonedurglate with a concentrated aqueous solution of sulphurous acid it is reduced and on cooling ethyl dihydroxydnrylate crystallises out in needles. The substance may be purified by dis-solving it in alcohol and adding water containing sulphurous acid t o the hot alcoholic solution.It is thus obtained in colourless broad needles melting a t log" and having a characteristic sweet smell. 0.0935 gram substance dried i n a vacuum gave 0.2183 gram CO, Calculated for and 0.06 gram H,O. C,(OH),(CH,),.COOC,H~. Found. c 64.29 63.73 H . 7.14 7.13 The ethereal salt is soluble in alcohol and other organic solvents except light petroleum. It is somewhat soluble i n cold water ; much more so in hot. It dissolves in sodic hydrate forming a colourless solution ; on heating however hydrolysis occurs and the colour of the solution changes to deep violet. In alcoholic solution ferric chloride oxidises it to the quinone with passing green coloration. Nitric acid (sp. gr. 1.4) also oxidises it to the quinone.An intermediate product corresponding to quinhydrone was not noticed. (6.) Action o,f Concentrated Nitric Acid on Dury1ic-acid-quinone.-Durylic-acid-quinone dissolves without change in nitric acid (sp. gr. 1*4} in the cold. On heating on a water-bath however a strong evolution of gas is observed. The addition of water now pre-cipitates a beautiful Fellow crystalline substance which is insoluble in alkalis and therefore no longer an acid. Further investigation has shown that the carboxyl-group of durylic-acid-quinone has been displaced and in fact quantitatively by a nitro-group according to the reaction :-Cp,O,(CH,),*COOH + HNO = c6o,(cH,),*NO + H,O + CO,. The resulting conipound nitropseudocurnoquinono is one of the most stable of quinone compounds 438 NEF CARBOSPL-DEKIVATIVES OF BENZOQTJINONE.By heating durylic-acid-qixinone with alcoholic a,mmonia in a sealed tube a t loo" the elements of carbon dioxide are removed and a deep-red solution is obtained. Want of material has prevented me thus far from investigating the product more closely. These observations aye in harmony with those of Hermann (Annalen 211 343) in the case of paradihydroxyterephthalic acid. The latter when treated with concentrated nitric acid gives no tere-phthnlic-acid-quinone but 2 mols. CO are eliminated in the cold with formation of nitrnnilic acid. All these facts go to show that the carboxyl-group i n carboxylated quinoce-compounds is very loosely bound. 0 heating durylic-acid-quinone on a water-bath for about half an hour with nitric acid (sp.gr. 1.4) until the erolution of carbonic anhydride ceases. On adding water the quinone separates out completely and crystallises in yellow plates melting at 113". 0.1485 gram substance dried in a vacuum gave 0.3023 gram CO and 0.230 gram substance dried in a vacuum gave 15 C.C. nitrogen at 0.0655 gram H,O. 16" and 710 mm. Calculated for C692(CH,),.N02. Found. (3 55.38 55.52 H . 4.62 4.9 N . 7-18 7.09 This substance is vei-y stable. It sublimes easily and without decom-position. Heated on a water-bath for hours with concentrated nitric acid it remains in great part unchanged. On cry~tallisation from nitric acid light petroleum or especially from dilute alcohol it is obtained in yellow plates melting a t 113" and possessing a faint but distinct quinone-likc odour.(7.) Nitropseudocurno~lsilzol C6( OH),( CH,),*NO,.-This was ob-tained by treating nitropseudocumoquinone in a sealed tube at 100" with a concentrated solution of sulphurous acid in 25 per cent. alcohol. On cooling the tube is filled with long broad yellow needles which are quite pure and remain unchanged on repeated treatment with sulphurous acid solutions. For analysis the substance was crystallised from ether when it was obtained in flat yellow needles melting a t 106" NEF CARBOSYL-DERIVATIVES O F BESZOQUINONE. 439 0.1797 gram substance dried in a vacuum gave 0.3592 gram CO,, and 0.0908 gram H,O. Calculated for C,(OH)?(C'HJ,.NOp Found. C . . 54.82 54.52 H 5.58 5-58 Nitropseudocumoquinol is easily soluble in chloroform ether and alcohol.It is somewhat soluble in hot water crystallising therefrom in flat needles. Sodic hydrate colours i t bluish-violet and it then dissolves with a pale greenish-yellow colour. Ferric chloride or con-centrated nitric acid oxidiaes it to the corresponding quinone. On treating a solution of the quinone or of the quinol in acetic acid with zinc-dust reduction takes place and finally a colourless solution is obtained which on exposure to the air soon becomes deep red-violet. On neutralising with sodic carbonate after precipitation of the zinc as zinc sulphide ether extracts a deep red substance which with hydrochloric acid forms a colouriess hydrochloride crystallising in needles and sparingly soluble. IV. Corwersion of D i n i t r o d w y lic Acid i n t o E t h y l Quinonetetracar-b oxy 1 ate.(1.) Diniti.o~yro?nelZitic A c i d C,(NO,)?( COOH),*.-The oxidation of the three methyl side-chains in dinitrodurylic acid to carboxyl was carried out by means of potassium permanganate and gave an almost quantitative yield. 20 grams of dinitrodurylic acid (dried a t 100') and 50 gr:i.ms of anhydrous potassic carbonate were dissolved in 2 litres of water and heated on a water-bath. To this a solution of 72 grams of potassic permanganate was slowly added. A t first oxidation takes place rapidly, but towards the end very slowly and is finished after four or five days. The manganese dioxide formed was filtered off and the filtrate concentrated. After acidifying with concentrated hydrochloric acid and dilute sulphuric acid the solution was extracted four times with ether ; on distilling off the ether a white mass remained behind con-sisting of a mixture of the tribasic and the tetrabasic acid.After many trials the following method of separation was found to be exceedingly sharp. The mixture was dissolved in water and heated * The direct nitration of pyromellific acid seems to be impossible. This acid can be heated with a mixture of concentrated sulphuric acid and potassic nitrate for 10 hours in a sealed tube at 150" without the slightest change. The application of Burkhardt's method by which as is well known terephthalic acid has been nitrated (Beviehie 10 145) was in this case entirtlj fruitless 440 NEE' CARBOXTL-D ERIVATIVES O F BENZOQUISONE. on a water-bath with finely pulverised marble until the reaction WRS neutral and alcohol (about 2 volumes) was then added to the concentrated filtrate until no further precipitation took place.In this way the calcium salt of the tetrabasic acid is precipitated com-pletely in pale-yellow needles while the calcium salt of the tribasic acid remains in solution and in fact cannot be precipitated at all by addition of more alcohol. The tribasic acid was then converted into the tetrabasic acid by renewed treatment with potassic permanganate. The l a s t methyl-group in the tribasic acid is attacked by oxidising agents with great difficulty so that it takes from four to five days to oxidise 10 grams. The yield was excellent; 20 grams of dinitro-durylic acid gave from 20 to 23 grams of pure anhydrous dinitro-pyromellittic acid.The abovementioned calcium salt of dinitropyromellitic acid was purified by dissolving it in water and precipitating by addition of alcohol. It crystallises in pale-yellow needles. When heated in an air-bath a t 180" it loses its wat'er of crystallisation and becomes deep golden-yellow. I. 0-1260 gram salt dried a t 180" gave 0.0802 gram CaS04. 11. 0.1602 ,? 9 , 0.1022 , 9 , Found. Calculated for (--A- 7 c (NO,) 2 (CO 0) 4Caz. I. 11. Ca 19.05 18-73 18.i6 To obtain the acid the purified calcium salt was dissolved i n water, and the strongly acidified solution extracted three or four times with pure ether. On distilling off the ether and drying at loo" pure dinitropyromellitic acid is left. The acid is colourless has a very strong acid smell and decomposes marble in the cold with violent evolution of carbonic anhydride.It crystallises from ether or water in long silky needles. It is easily soluble in cold water still more so in hot water and deliquesces in ether vapour. It is also readily soluble in alcohol and acetic acid but insoluble in chloroform benzene, and light petroleum. With the exception of the silver lead and barium salts the salts of this acid are easily soluble in water and most of them are precipitated from their aqueous solutions by alcohol. The barium salt is easily soluble in hot water much less so in cold water. A separation of the tribasic and tetrabasic acids can also be accomplished by means of their barium salts as the salt of the tribasic acid is much more easily soluble.When heated in a capillary tube dinitropyromellitic acid loses water (100-160") and becomes slightly yellow. Above 160° i NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE. 441 becomes gradually deeper and deeper yellow and decomposes with violent evolution of gas between 208" and 225" according as the tem-perature rises slowly or quickly. If heated in an air-bath at 140" for several hours the acid decomposes and is converted into a deep-yellow substance. When dried at loo" the acid is colocrless and anhydrous as was deduced from the precipitation and analysis of the silver salt. I. 10.2 grams acid gave 22.5 grams silver salt. Calculated 22.88 , 9 7 11. 21.3 grams acid gave 47% 9 7 7 , Calculated 47.77 , 7 , The silver salt was prepared by adding a neutral solution of the ammonium salt of the acid to a hot solution of silver nitrate.It forms a stable amorphous yellow precipitate quite insoluble in water For analysis it was dried at loo" and as it explodes on heat-ing it was carefully mixed with the copper oxide. I. 0.3954 gram substance gave 0.2278 gram CO and 0.0210 H,O. II. 0.4115 gram substance gave 0.2340 gram GO2 and 0.0120 H20. 0.223 gram substance gave 0.1650 gram AgCl or 0.1242 Ag. 0.5668 gram substance gave 18.75 C.C. nitrogen at 7" and i14 mm. Found. Calculated for (----C6 (NO,),(COOAd 4 I. 11. Ag 55.96 55.70 -C 15.55 15-71 15-51 H - 0.59 0.52 N 3-63 3.73 -It is not possible to convert dinitropyromellitic acid into the noymal ethereal salt by the ordinary methods An alcoholic solution saturated with gaseous hydrogen chloride and allowed to stand for two days did not yield it and the acid heated with alcohol and a few drops of concentrated sulphuric acid for two days on a water-bath remained in great part unchanged.With reducing agents the acid shows a very characteristic reaction ; all reducing agents whether applied in neutral acid or alkaline solu-tion produce a deep red coloration. Ether extracts from acid soluiions a red substance ; the deep red ethereal solution has a very m-arked yellowish-red fluorescence. The red substance remains unchanged even when heated wit'h reducing agents for a long time; it is un-doubtedly as will be seen below from results obtained with the ethereal salt free diamidopyromellitic acid.The reduction takes a VOL. LIII. '2 I 442 NEF CARBOXPL-DERIVATIVES OF BENZOQUINONE. different course when granulated zinc and dilute sulphuric acid are used in the cold. In this case a yellow solution is produced from which ether extracts a yellow substance. The yellow ethereal solu-tion has a marked green fluorescence. The tribasic acid methylparadinitrotrimellitic acid, c,(No,),(cH~)(cooH)~, which is obtained along with dinilropyromellitic acid on oxidising dinitrodurylic acid forms a deliquescent calcium salt. With basic lead acetate it gives a fine yellow basic lead salt. The silver salt comes down after long heating with silver nitrate in sparingly soluble yellow plates. The acid itself is easily soluble in hot water ; much less so in cold.It dissolves with the greatest ease in alcohol and ether. It is obtained in long colourless needles by crystallisation from water containing hydrochloric acid. (2.) Ethy 1 DinitropyromeZZitafe C,(NO&( COOC2H5)4.-The ethyl salt of dini tropyromellitic acid was prepared in considerable quantity by treating the silver salt with ethyl iodide. The neutral silver salt can be precipitated quantitatively as follows :-A solution of 20 grams of dinitropyromellitic acid carefully neutralised with ammonium hydrate is added slowly to a warm (50') solution of 50 grams of silver nitrate. The silver salt was dried a t 100" and then allowed to stand over night with ethyl iodide (l$ times the calculated amount) and dry ether ; to complete the reaction t'he mixture was heated for five or six hours on a water-bath.The ethereal solution was now filtered from the silver iodide washed with dilute sodic carbonate and then dried with calcic chloride ; after distilling off the ether the solid ethyl salt, was left suffic.iently pure for farther work. For analysis it was crys-tallised twice from alcohol when it formed long colourless needles melting a t 130". 0.1155 gram substance gave 0.2006 gram GOz and 0.0465 gram Calculated for H20. c6 (N02)2 (C00C!2H5)4- Found. c 47.37 47-38 H . 4.39 4-47 It is easily soluble in benzene chloroform acetic acid and acetone ; less so in ether. By crystallisation from hot alcohol or light petroleum, i t is obtained in beautiful needles. It dissolves in alcoholic ammonia with a fugitive red colour.The yield was seldom over 60 per cent. of the theoretical. (3.) EthyZ Diamidopyromellitate c& NH&( COOC2H8,)4.-Ethyl di NEF CXRBOXYL-DERIVATIVES OF BENZOQUINONE. 443 nitropyromellitate on reduction with zinc-dust and acetic acid gives a beautiful red compound which I long regarded as ethyl azopyro-mellitate. It was found impossible in spite of repeated efforts to reduce this substance further with elimination of the nitrogen. I thought therefore that it might possibly be an inner azo-compound as it sublimes without decomposition like a quinone. I tried to ascertain its molecular weight by means of a vapour-density determination, using Victor Meyer's method but withoutJ success. Nevertheless the analytical figures obtained did not agree sharply ; the hydrogen especially was always too high.Later when I had larger quantities of the substance in my hands and analysed a product purified with especial care I obtained figures agreeing exactly with those required by ethyl diamidopyromellitate. Finally the formation and isolation of a colourless diacetyl-derivative by treating it with acetic anhydride is a convincing proof that the red substance is actually a diarnido-compound. The deep red colour of the diamido-compound is now no longer remarkable as v. Baeyer (Beq 19 430) has recently shown that ethyl paradiamidoterephthalate has the colour of potassic dichro-mate. For reducing ethyl dinitropyromellitate the following method was found most advantageous and gave an almost quantitative yield.The substance was Grst dissolved inacetic acid and to the hot solution a little water and then zinc-dust in small portions at a time were added. When the addition of more zinc-dust no longer produced a violent reaction and the red colour had changed to yellow the solution was filtered hot from the zinc-dust and poured slowly (stirring well) into a dish containing water. The diamido-compound then separated completely in yellow flakes which soon changed to a red crystalline precipitate. For purification the substance was dissolved in ether, washed with dilute sodic carbonate and the ethereal solution dried with calcic chloride was then allowed t o evaporate over sulphuric acid ; the diamido-compound crystallised out in beautiful deep red four-sided prisms.For analysis the large crystals were mechanically picked out and dried at 100". They melted at 134" and sublimed without decomposi-tion and without leaving any residue. When melted and allowed to solidify again the substance has exactly the appearance of sealing-wax. I. 0.1231 gram substance gave 0.2447 gram GO and 0.0675 gram 11. 0.1774 gram substance gave 0.3554 gram CO and 0.1044 gram HZO. HZO. 2 H 444 111. Dr. NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE. 0.1842 gram substance gave 0.3681 gram COz and 0.1036 gram 0.29'70 gram substance gave 20.3 C.C. nitrogen a t 19" and Found. H,O. 723 mm. Theory for v- 7 C,(NH2)2(COOC2H,) 4' I. 11. 111. c 54-55 54.20 54-64 54.50 H 6.06 6.09 6.54 6.25 N . . 7.47 - - '7.07 Ethyl Diamidopyl.on~ellitate.W. Muthmann examined the above cryst,als of ethyl diamido-pyromellithate in the mineralogical laboratory of P. Groth- in Munich, and reports as follows :-Monoclinic system /3 61" 17'. a b c = 0.66964 1 0.62686. Planes .m = (110) = mP. = (001) = OP; o = (111) = +P. = (010) = wywo; 12 = (110) = my;?. b The majority of the crystals showed only the ci-j-stal faces w c, and b. Found. Calc dated. ( n o ) (iio) = *60° 50' - -(110) (001) = *65 31 - -(010) (in) = *60 55 - -(iio) (111) = 53 39 53" 35' (110) (120) = 19 7 19 10 (010) (001) = 90 3 90 NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE. 445 Colour reddish-orange (Radde's International colour scale). The plane of the optic axes is parallel to the plane of symmetry. One bisectrix lies i n the acute angle of the crystallographic axes and forms with the vertical one an angle of 154".An optical axis comes out very nearly perpendicular to (001). From these data the axial angle is found to be approximately 85-90'. The crystals are very soft and possess a remarkably perfect cleavage in the direction (001). Ethyl diamidopyromellitate has feeble basic properties. On adding to it concentrated hydrochloric acid it becomes colourless ; and on diluting with an equal volume of water and heating the colourless salt dissolves and crystallises out again in needles on cooling. More water decomposes the salt and the free base separates out in beautiful rhombic plates. The diamido-compound dissolves in concentrated sulphuric acid with a pale-yellow colour ; water precipitates the free base again unchanged.The substance dissolves easily in alcohol ether and acetic acid. The ethereal solution is red and shows a very marked yellowish-red fluorescence. On saponification with alcoholic potash and acidifying, ether extracts from the solution a red substance identical with that obtained by the reduction of dinitropyromellitic acid. On treating a hot alcoholic solution of the ethereal salt of the diamido-compound with sulphuric acid (1 H,SOa to 1 of water) and zinc-dust it is easily reduced and on adding water a colourless substance crystallising in needles separates out. This was found to contain no nitrogen and it was therefore supposed to be ethyl pyro-mellitate. A close study however of the substance after v.Baeyer had observed that ethyl paradiamidoterephthalate is converted by reduction into ethyl succinosuccinate has shown that it is identical with ethyl paradiketohexamethylenetetracarboxylate described below. On converting this product into ethyl quinoltetracarboxylate and into ethyl quinonetetracarboxylate the last doubt about the complete identity of the two substances vanished. Ethyl diamidopyromellitate is therefore reduced hy means of zinc-dust and sulphuric acid according to the following reaction :-C I O C2H5OOC*i /\ Y.COOC2H2 C H 5 0 0 C - H 6 'CH.COOC2H5 + H2 + 2E20 = C2H5OOC.C C400C2H5 CZH5OOC.HC CH.COOC2Hh I / \/ c:o \c/ I + 2NH3. NH2 Ethyl paradiketohexmethylene-tetracarboxylate 446 NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE. Oxidising agents such as ferric chloride platinic chloride or sodic nitrite in acid solution do not convert the ethereal salt of the diamido-compound iato the corresponding quinone chromic acid in acetic acid solution as well as potassic dichromate and sulphuric acid decompose it entirely.Concentrated nitric acid however oxidises it to the quinone. (4.) Ethyl Diacetyldiamido~yromellitate.-This was obtained by heating the diamido-compound with excess of acetic anhydride at 140" for four or five hours ; on adding alcohol the new compound was in great part precipitated. After getting rid of the excess of acetic anhydride the substance was purified by crystallisation from alcohol ; it forms colourless lustrous rhombic plates melting at 149". 0.1161 gram substance gave 0.2342 gram CO and 0.0614 gram HZO.Theory for C6(COOC,H,)4(NHAc),. Found. C 55.00 55.01 H 5.83 5.88 The acetyl-derivative dissolves easily in acetone chloroform and in warm alcohol. It cannot be again resolved into its components by treating with sodium hydrate or hydrochloric acid. The same is true of many other acetyl-derivatives of aniido-acids. (5. ) Ethyl QuinonetetracarBoxybte or Ethyl Quinonepyrornel1itate.-On heating ethyl diamidopyromellitate with concentrated nitric acid, or with sulphuric acid and potassic nitrate it is oxidised to the corresponding quinone. The presence of nitrous acid is detrimental, giving rise to oily bye-products. The best yield (50-55 per cent.) was obtained by the following method 5 grams of the substance were dissolved in 40 C.C.pure concentrated nitric acid (sp. gr. 1.4) and allowed to stand in the cold for two hours. The subst,ance dissolves at first with yellow colour and without change ; soon however the solution becomes much darker and the quinone begins to crystallise out i n yellow needles. A slow and constant evolution of gas is noticed after the end of the first hour. It is therefore well to stir from time to time. On pouring into wafer the quinone separates out com-pletely in yellow needies. For analysis it was crystallised twice from alcohol when it was obtained in long yellow needles melting at The results of the analysis agreed with the formula C,O,( CO0CzH5),. 0.1375 gram substance gave 0.2756 gram CO? and 0.0662 gram 148-149". HZO. Theory. Found. c 54.55 54.66 H 5-15 5.3 NEF CARBOXYL-DERJTATIVES OF BENZOQUINONE.447 The substance is of quinone-yellow colonr ; it is odourless ; it sub-limes readily without decomposition and unquestionably belongs to the class of true (para) quinone compounds. It is very stable in the presence of concentrated nitric acid in which it easily dissolves. The solution even when heated a long time undergoes little change and the quinone separates again completely on adding water. It is but little soluble in cold alcohol or ether ; much more so in warm. Hydroxylamine hydrochloride in neutral or acid solution reduces it almost immediately to the quinol compound. It is insoluble in cold potassic hydrate ; on heating gently however it liquefies turns dark-red and then goes into solution.By using the calculated amountl of potassic hydrate it was saponified and strangely enough the product obtained was dihydroxypyromellitic acid. On saponifying in acid solution with acetic acid and concentrated sulphuric acid dihydroxy-pyromellitic acid also was obtained and not as was expected pyro-mellitic-acid-quinone. ( 6 . ) Ethyl Dihlldroxypyromellitate or Ethyl Quinoltetracarboxylate. -On treating ethyl quinonetetracarboxylate with zinc-dust and acetic acid it is reduced with great ease to the corresponding quinol-compound C,(OH)2(COOC2H5)~ [l 4 2 3 5 61. This as will become evident from what follows bas a most remarkable resem-blance in chemical as well as physical properties to Herrmann’s (Annalen 211 327) so-called ethyl quinonedihydrodicarboxylate ; the latter according to recent results obtained by v.Baeyer (Ber. 19, 4281 is to be regarded as ethyl quinolparadicarboxylat,e, C,H,(OH),(COOC,H,)2 [l 4 2 51. Of especial interest moreover is the action of redncing agents on ethyl quinoltetracarboxylate. It is converted in a manner analogous to the conversion of ethyl quinoldicarboxylate t o ethyl succino-succinate-to a hexnmethylene-derivative or hexahydrated derivative o€ benzene-that is from a tertiary t o a secondary bound ring (v. Baeyer ‘‘ Ueber Nomenclatur,” Ber. 19 160). H 0 C I O /\ + H = I I \ / COOC2H5-HC CH-COOCZHS COOC,H,-HC CH*COOCzH, c-; 0 Ethyl quinoltetracarboxylate. Ethyl paradike1;ohexametliylenetetra-The resulting compound ethyl paradiketohexamet hylenente tracar-boxylate (Ber.19 160) has a remarkably deceptive resemblance in all respects to ethyl succinosuccinate. By means of bromine in carbon carboxylate 448 NEF CARBOXYL-DERZVATIVES OF BENZOQUINONE. bisulphide solution it can be reconverted to ethyl quinonetetracar-boxylate. I n order to reduce ethyl quinonetetracarboxylate it is dissolved in acetic acid and zinc-dust added in small portions. The yellow solu-tion immediately become colourless with a beautiful blue fluorescence. On adding water the quinol compound separates out in almost colour-3ess needles having a marked bluish tinge. For purification it is dissolved in acetic acid and precipitated again by addition of water. I. 0.1280 gram substance gave 0.2564 gram CO and 11. 0.1475 gram substance gave 0.2934 gram GO and H,O.HZO. Found. Theory for F--T C6(0H)2(C00C2H5)4* I. 11. 0.0676 gram 0.0752 gram C 54.27 54-62 54-25 H 5-53 5-86 5.66 Ethyl quinoltetracarboxylate crystallises in pale-yellow needles having a bluish tinge. It has no sharp melting point 127-129'. When carefully heated between two watch-glasses it sublimes without decomposition. It dissolves easily in alcohol ether and acetic acid and tlhe solutions show a very marked beautiful blue fluor-esence. On adding a drop of ferric chloride to an alcoholic solution of the substance a bluish-green coloration is produced. I f heated with concentrated nitric acid it dissolves and is oxidised quantitatively to the quinone. It remains unchanged when exposed to an atmosphere of bromine for several days.It dissolves un-changed with yellow colour in dilute sodic hydrate or sodic carbonate. Concentrated sodic hydrate precipitates a deep cin-nabar-red sodium salt which is also obtained on adding sodic ethylate to a dry ethereal solution of the quinone. On heating the sodium hydrate solution hydrolysis takes place and the sparingly soluble sodium salt of dihydroxypyromellitic acid separates out in yellowish prisms, Ethyl quinoneparadicarboxylate also forms a similar coloured sodium salt and gives bluish-green coloration with ferric chloride. The solution of the substance in ether also shows the same marked blue fluorescence. An investigation of the physical properties of ethyl quinoltetra-carboxylate has shown still further the remarkable resemblance existing between these two compounds.The study of the physical properties OF this substance-which certainly is one of the mos NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE. 449 remarkable substances physically in chemistry-was begun during the last Easter vacation in the new chemical crystallographical laboratory of Professor P. Groth. Dr. W. Muthmann reports thereon as follows :-Et hy 1 &&no 1 tetracarboxy late. Modification I. Monoclinic system p = 64" 36'. a b c = 2.3875 1 3.0601. Planes c = (001) = OP. p = (101) = -I&. r = (ioi) = +Pm. nt = (110) = a?. Elongation in the direction of the axis b ; often in thin needles. The base never predominates but always appears as a small plane. Found. Calculated. (110) ( i i o ) %go 45' - -(001) (101) '36 46 - -(001) (TOl) *68 45 - -(110) (101) 68 10 68" 29' (iio) ( i o i ) 73 13 73 27 The plane of the optic axes lies in the plan of symmetry.An optical axis appears very nearly perpendicular to (101) and as a consequence no total reflection is noticeable on this face. In (101) pretty strong dichroism-dark-yellow and light-yellow-noticeable ; this was not perceptible in ( i O l ) , Colour greenish-yellow, Modification 11. Monoclinic system /3 = 81" 53'. a b c = 1.7896 1 3.3206. Planes c = (001) = OP. 0 = (111) = -P. a = (100) = mpm. Of= (111) = +P 450 NEF CARBOXTL-DERIVATIVES OF BENZOQUINONE. Thick plates in direction (001). In many of the crystals the corre-sponding opposite planes were not quite parallel owing to disturbances in their growth.The angles used in the calculation of the axial ratios &c. are the mean of 20-25 angle measurements on eight of the best crystals. (111) (111) (111) (111) (111) (111) (100) (001) (001) (111) (001) ( i n ) (100) (111) (loo) (111; Found. 56" 43' 111 44 117 53 81 57 71 25 78 45 59 40 63 21 2alculatecl. .- -- -81' 53' 71 28 78 55 59 49 63 28 Colour pale-yellow. Dichroism-coloudess and bright-yellow-noticeable in (001). In the summer transparent crystals of the first modification alone were obtained from carbon bisulphide solution. These on heating become turbid and seem to undergo a transformation. The change begins at 111" and is complete at 115". The product then melts on higher heating sharply at 133.2-133-6".In the colder months (April) at a temperature of about -5", crystals of the second modification were obtained side by side with the first modification from a solution in carbon bisulphide. Modifica-tion I1 becomes transformed at 63-5-64' into a different modification. On cooling this does not change back t o the original again. The modification thus obtained at 64" shows a peculiar fusing point. It begins to melt at 123-124" and thereupon a part of the substance becomes solid again a8nd the whole melts entirely at 128.5". On cooling and reheating the same phenomena again appear. Besides these two modifications a third was observed t o crystallise from the carbon bisulphide solution at very low temperatures. The The total reflection in (001) exactly diagonal NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE.451 amount of material on haud has up to the present time been in-sufficient for the closer study of this peculiar modification. The modifications I and I T described are undoubtedly totally different from one another and it is not possible by heating or fusing t o convert either one into the other. There is first of all a remarkable difference in colour I is green; 11 pale-yellow. On fusing and cooling I appears dark-yellow ; 11 bright-yellow. One modification crystallises in needles the other in plates so that it is an easy matter to separate them mechanically. When either modification thus sharply separated is dissolved in carbon bisulphide the mixture of the two modifications side by side like two substances totally dis-tinct from one another is again obtained.I f these are carefully separated mechanically and the operation repeated the same result is invariably obtained showing a mixture of the two modifications side by side. The substance therefore on being dissolved in carbon bisulphide (the same result is obtained on using pure dry ether) must undergo some transformation. The molecules of the substance must either possess a very peculiar tendency to arrange themselves always in two definite states of equilibrium corresponding to the two totally different crystalline forms or what is equally probable certain atoms of the molecule of the substance itself must be in a perpetual unstable equilibrium and the two modifications observed would then corre-spond each t o a different molecular structure.Certain it is the solutions of the two modifications have exactly the same colour. The absorption-spectrum of each modification dissolved in dilute sodium carbonate solution was quantitatively determined for all parts of the spectrum and found in both cases to be exactly the same. The data &c. will be given later in connection with a study of the absorp-tion-spectra of various other closely allied substances. On oxidation with concentrated nitric acid both modifications I and I1 give the same perfectly homogeneous substance the quinone melting sharply at 149" whilst on treatment with zinc-dust and hydrochloric acid in alcoholic solution both give the same colourless and perfectly homo-geneous substance-the diketone compound melting at 144".It is worthy of note that quinol and many other derivatives of quinol or paradihydroxybenzene as well as paradiicmido-deriratives of benzene are dimorphous. It may be noted now that Dr. Muthmann has discovered two entirely distinct modifications of ethyl para-diamidoterephthalate (Ber. 19 430) and measured the crystals bf both forms only one modification has as yet been found of tbe very closely allied ethyl paradiamidopyromellitate. The fact that so many of the paradihydroxy-derivatives of benzene are dimorphous naturally leads one to suspect that this dimorphis 4-52 NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE. may be due to a difference in the molecular constitution of these substances. It might be supposed that one modification corresponds to the real paradihydroxy-derivative of benzene e.g., H 0 C // \.Xc cx (X = H COOC2H5 or any I )I other atom o r group of xc cx atoms). 0 H and that the other modification corresponds t o the pseudo- or ketone form e.g., co I XHC CX atoms). The transference of the hydrogen-atoms to adjacent carbon-atoms which is assumed here is analogous for instance to what takes place in the case of carbostyril isatin pyridone and phloroglucin (Ber. 19, 159). In all these instances t,he chemical behaviour of these sub-stances necessitates the assumption of pseudo-forms or of the migra-tion of the hydrogen-atom to an adjacent carbon- or nitrogen-atom. These compounds however have never been obtained in two different physical modifications.If the dimorphism of the quinol compounds is due to a difference in chemical constitution i t must be possible to prove this by chemical means; the experiments thus far carried out however have been insufficient to settle this question. A short time ago Nietzki and Kehrmann (Ber. 20 613) found that quinol reacts with hydroxylamine giving a compound C6H6N,O2. If the above theory is correct this compound ought to have the formula C,H,N,O (that is contain two more hydrogen-atoms). These experiments were repeated but the analysis of the substance confirmed the formula given by Nietzki and Kehrmann which seems to show that quinol is first oxidised to quinone and then acted on by the hydroxylamine. The study of the action of hydroxylamine on ethyl quinoltetracar NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE.453 boxylate as well as on tetrachloroquinol has n o t as yet given any positive results. Dr. Muthmann and myself have joined to make a combined chemical and physical study of this subject. We have decided to publish here the results thus far obtained because Hantzsch and his pupils (Hantzsch and Zeckendorf Ber. 20 1308 2796; Hantzsch and Hermann Ber. 20 2801) in their study of ethyl quinolparadi-carboxylate and its derivatives have found the same peculiar di-morphism t o exist. Hantzsch has gone further and puts especial stress on the colour of the modifications. If the modification is colourless he says this corresponds to the true paradihydroxy-compound of benzene as all true compounds of benzene must be colourless (?).I f the modifica-tion is coloured it corresponds t o the ketone or quinone form because all ketones or quinones me coloured (?). The question of colour is it seems to us purely arbitrary and does not prove anything whatever. More has hitherto been considered necessary in settling the constitution of a chemical compound. It is very difficult to see for instance why ethyl quinolparadicarb-oxylate which according to Hantzsch is a ketone in its two most stable modifications does not under any circumstances act chemically like a ketone but invariably like a dihydroxy-compound of benzene. Hermann’s argument (Ber. 19 2229 2235) that ethyl quinolpara-dicarboxylate forms isomorphous mixtures with ethyl succinosuccinate seems a much more powerful argument for its ketone form.On the other hand it must be remembered that ethyl succinosuccinate acts like a dihydroxy-compound as well as like a diketone so that it is more logical to conclude that in the isomorphous mixtures the dihydroxy modification of ethyl succinosuccinate is present. (7.) Dihy droxypyyomellitic Acid OT Quinoltetracarboxylic Acid.-On heating ethyl quinoltetracarboxylate with potassic hydrate it is easily saponified ; the product dihydroxypyromellitic acid resembles very closely paradihydroxyterephthalic acid (Awnalen 211 335). 2 grams of the ethereal salt were heated with twice the calcu-lated amount of potassic hydrate (4 grams) on a water-bath ; a dark-yellow solution with a green fluorescence was obtained. After neutralising with acetic acid the lead salt was precipitated by means of lead acetate ; the precipitate was collected carefully washed and then decomposed with sulphuretted hydrogen.After filtering hot from the lead sulphide and concentrating the yellowish solution a pale-yellow substance crystallised out almost completely on cooling in four- and six-sided prisms. The substance purified by recrystal-lisation from water was found nofi to be free from ash. On adding concentrated hydrochloric acid to a hot concentrated solution pur 454 NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE. dihydroxypyromellitic acid separates out slmost completely in pale-yellow broad flat needles. The acid dried on an air-bath a t loo" mas found to contain 1$ mols. H,O. 0.1227 gram substance gave 0.1706 gram CO and 0.0339 gram H,O.Theory for C6(OH)2(COOH)4 -t- 1+H20. Found. C 38.34 37.92 H 2-87 3-07 Heated to a higher temperature (150") the substance not only lost its water of crystallisation but was further changed. Dihydroxypyroineilitic acid dried at 100" is of a pale-yellow colouT and dissolves sparingly in most solvents. The yellow solutions shorn a characteristic green fluorescence. It dissolves easily in hot water and is precipitated therefrom by mineral acids. The aqueous solution is coloured deep blue on the addition of a few drops of ferric chloride. The colour remaiiis unchanged for days even when a great excess of ferric chloride is added. When heated in a test-tube it melts at a very high temperature, losing water and then sublimes without leaving any residue the subli-mate which is probably the dianhydride of the acid crystallising in yellow flat needles.It dissolves in dilute alcohol with yellowish-red colour and reddish-yellow fluorescence. On shaking the solution with air a peculiar violet tinge is noticed on the sides of the test-tube. On standing or more quickly on heating the colonr of the solution changes to the yellow tint and green fluorescence of dihydroxypyro-mellitic acid again. The salts of this acid are yellow and in solution exhibit a characteristic green fluorescence. The silver salt was more closely studied and analysed. It was obtained by adding a solution of the neutral ammonium salt to a neutral solution of silver nitrate in the cold. It separates out as a lemon-yellow flocculent precipitate.When dry it is stable a t ordinary temperature. Heated quickly on platinum foil it explodes; if heated slowly a volatile yellow substance sub-limes. 0.2283 gram substance dried over sulphuric acid in a vacuum gave 0.0562 gram substance gave 0.0338 gram silver. 0-1394 gram CO and 0.0114 gram H,O. Theory for C6(0H)2(C00Ag) 4. Found. C 16.80 16.65 H . . 0.28 0.55 Ag 60.51 60.1 NEF CARBOXYL-DERIVATIVES OF BENZOQUINONE. 455 Lead acetate throws down the lead salt as a light-yellow floccu-lent precipitate ; baric chloride gives the barium salt as a granular, pale-yellow precipitate. The sodiuni salt is especially characteristic. On heating the free acid with sodium hydrate the sodium salt sepa-rates out in pale-yellow prisms which are very sparingly soluble even in boiling sodium hydrate.The oxidation of dihydroxypyromellitic acid to pyromellitic-acid-quinone C60z(COOH)4 presented unexpected difficulties. Con-centrated nitric acid has i n the cold no effect on dihydroxypyro-mellitic acid ; on heating gently however the substance dissolves with a violent evolution of carbonic anhydride and as it would seem is completely decomposed. When heated gently with a solution of chromic acid a violent evolution of gas is noticed. Potassium permanganate oxidises the acid in neutral or acid solution in the cold but although as much as three times the calculated amount of the oxidising agent was used much unchanged dihydroxypyromellitic acid was the only product obtained. A solution of the acid allowed to stand in the cold for days with a great excess of ferric chloride retained the beautiful blue coloration unchanged and much unchanged acid was found t o be present.It may he mentioned here again that ethyl quinonetetracarboxylate on hydrolysis in acid or in alkaline solution invariably gives dihydroxypyromellitic acid and not the quinone. A similar resistance towards the formation of the qninone from a paradioxy-compound was observed in the case of paradihydroxy-benzoic acid C,H,( OH),*COOH and its ethyl salt. Repeated attempts were made to oxidise these substances to quinones varying the con-ditions in every conceivable way. Paradihydroxybenzoic acid was found to evolve carbonic anhydride even on heating gently with such a mild oxidising agent as ferric chloride.Ethyl paradihydroxy-benzoate when treated with oxidising agents either remained un-changed or was entirely decomposed. Similar negative results were obtained by Brunner (Monats. Chenh., 1881 464) in the case of toluquinolcarboxylic acid, C,Hz( CH j (OH) 2C OOH, and its ethyl salt. Hantzsch and Loewy (Bey. 19 26) could not convert ethyl quinol-paradicarboxylate into the corresponding quinoneterephthalate, although they succeeded in obtaining mother oxidat;ion product. (8.) Etli y 1 Paradiketohexamethy lenetetra cai-boa y late.-The conversion of ethyl quinoltetetracarboxylate into a derivative of hexahydroben-zene or hexamethylene-that is from a tertiary t o a secondary chain of six carbon-atoms-can readily be accomplished by v.Baeyer’ 456 NEF CARBOXYL-DERIVATIVES OF BENZOQUINOXE. method (Bey. 19 432). To a very concentrated hot alcoholic solu-tion of the substance (2 grams) zinc-dust (10 grams) was added and then concentrated hydrochloric acid. The change takes place almost instantly and quantitatively. On adding water to the filtrate a colourless substance separates out in needles this when purified by dissolving in a little alcohol and precipitating with water was ob-tained in long colourless needles. The substance contains water of crystallisation and melts therein at 94". Heated in an air-bath at 110" it loses its water of crystallisa-tion and a hard granular mass remains fusing a t 144". 0.1312 gram substance dried a t 110" gare 0.2500 gram CO and 0.0720 gram H,O.Theory for C1sH24010. Found. C 54-09 53.84 H . . 6-00 6.10 The anhydrous substance is very sparingly soluble in alcohol ether, and carbon bisulphide. It separates from hot carbon bisulphide solu-tion in beautiful colourless needles. The solutions show a very faint but distinct blue fluorescence. The substance containing water of crystallisation on the other hand is easily soluble in alcohol and ether. On adding sodium ethylate to a dry ethereal solution of the di-ketone compound a colourless salt separates out at first but on adding more sodium ethylate this changes to a rose-coloured salt. Ethyl succinosuccinate gives exactly the same reaction. Potassic hydrate colours the substance yellow ; on diluting with water it dissolves forming a yellow solution from which the substance is precipitated unchanged on the addition of an acid.A drop of ferric chloride added to an alcoholic solution produces a cherry-red coloration. By treatment with bromine in carbon bisul-phide solution the substance is reconverted quantitatively into ethyl quinolt e tracarbox yl ate. That the compound has free ketone-groups is proved by the ease with which it reacts with phenylhydrazine hydroxylamine and am-monic acetate. Heated over a flame with ammonic acetate a yellow substance is obtained which dissolves in ether with a marked green fluorescence. It contains nitrogen and melts between 115" and 120". It is not however converted by treatment with bromine and concen-trated sulphuric acid into ethyl diamidopyromellitate as might be expected from a di-imido-derivative (Rer.19 429). On further treatment with ammonic acetate products melting a t higher tempera-tures (180-200") were obtained ; these also could not be converte NEF 0-AHBOXYL-DERIVATIVES O F BESZOQUISOSE. 457 irito ethyl diamidopyromellitate. It was therefore considered pro-bable that the ethylated carboxyl-groups as well as the ketone-groups, were attacked by the ammonic acetate. On heating an alcoholic solution of ethyl paradiketohexamethylene-tetracarboxylate with phenylhydrazine (twice the calculated amount) for 10 hours a t a temperature of 110-120" in small amount a sub-stance crystallising in red needles is formed ; as it is quite insoluble in alcohol it is readily obtained pure. The chief products of the reac-tion are soluble in alcohol and form a red granular mixture ; want, of material has thus far prevented me from obtaining them in quantity sufficient to purify them for analysis.The red product insoluble in alcohol shows all the properties of dip heiiy ldi pyrrazohexahy drobenzene (Bey. 17,55 1,205 5) a compound which Knorr obtained by the action of phenylhydrazine on ethjl succinosuccinate. It dissolves in alkalis with purple-red colour ; also in concentrated mineral acids but is precipitated in yellow flakes on the addition of waker. On treatment with sodium nitrite in acetic acid solution a blue insoluble dye is formed. The same blue compound is produced on adding sodium nitrite to an alkaline solution and then acidifying with dilute sulphuric acid. The above facts suffice to show the very great resembIaiice between ethyl paradiketohexamethylenetetracarboxylate and eChyl succinosucci-nnte.This is not surprising when one reflects that the former com-pound is simply the ethereal salt of a dicarboxylated derivative of succinosuccinic acid. They all contain nitrogen. co co Hzi /\ rH*COOC2H5 COOC,H,.Hd bH.COOC,H, COOC2Hb.HC CH COOC,H,.HC CH*COOC2H5 l l \co/ \/ co Ethyl succinosuccinate. Ethyl paradiketohexamethylene-tetracarboxylate. To show the great resemblance between the derivatives obtained from ethyl succinosuccinate and those obtained from ethyl diamido-pyromellitate the following tables (pp. 458 459) are added showing the chief reactions of tjhe difFerenh compounds. The experiments on durylic-acid-quinone and pyromellitic-acid-quinone will be continued.VOL. LIII. 2 Name. I---- I--I-I- --Properties. I----Ferric chloride. Solutions. J3Eth;l succinosucci-nate Etrhyl psrarliketohexa-methylenetetracarb-oxylate Sodium ethylate. colourless needles, m. p. 126-127" -_---coloudess needles, in. p. 144' faint blue fluores-cence cherry - red coloration colourless, then rose-coloured salt, - -colourless, t,hen rose-coloured salt fa.int bluc 5uores-cence cherry - red colorat ion Name. Ethyl pamdihydroxy-tvrephthalate Properties. -----______:_____-Par:diliytlrox;r tere- yellow plates plithalic acid - ~ - -marked blue fluor-escence -___.--yellowish - green, with green fluor-escence ---yellow with green fluorescence - - ~ _ _ _ _ _ _ _ _ _ _ _ _ _ _ Dihydroxypyromellitic light-yellow flat nee-acid 1 dles bluish-green coloration intense blue coloration intense blue co I oration Ferric chloIide. Solut,ions. rnellitate --_________ marked blue fluor- ~ bluisli-green escence ~ coloration dles m. p. 133.5'; light-yellow plates, m. p. 128'5' Sodium ethylate. deep red-coloured salt -deep red-coloured salt ---Name. Ferric Sodium I chloride. 1 ethylate. solutions. Properties. brown with gold-yellow fiuores-cence -red with yellowish-red fluorescence --------^__-____-Ethyl 'diamidopyro-mellitate fire-red prisms m. p. 134' weak bas