SCISSION OF SUBSTITUTED CYCLIC CATECI-TOL ETHERS. 929 LXX VIK-The Scission o j Certain Substituted Cyclic Ccl tech01 Ethers. By GERTRUDE MAUD ~EOBINSON and ROBERT ROBINSON. IT was observed (G. M. Robinson this vol. p. 113) that on attempting t o produce an azoxy-derivative from 4-nitrocatechol methylene ether (4-nitromethylenedioxybenzene) (I) by the action of sodium methoxide in methyl-alcoholic solution the sole product was a nitrophenol although the corresponding reaction with nitro-veratrole proceeds smoothly in the normal manner. On further investigation the nitrophenol was readily identified as 5-nitro-guaiacol (11) and the process appeared a t first sight to be one of reduction. However an altlernative view suggested itself namely, that t.he methoxy-group was derived from the methyl alcohol used as solvent and this was proved to be the case since when ethy 930 ROBINSON AND ROBINSON THE SCISSION OF alcohol was used a quantitative yield of 5-nitro-2ethoxyphenol (111) was obtained.(1.) (1I.j (111.) The mechanism of the reaction is therefore to be represented in the following manner : Blanksma (Chem. Weekblad 1909 6 313) has demonstrated the reversible interchange of alkyloxyl groups in nitro- and especially dinitro- phenol ethers and for example 2 4-dinitro-anisole is changed in ethyl-alcoholic solution in the presence of traces of alkali into 2 4-dinitrophenetole; but in the majority of instances mononitrophenol ethers are unchanged under these con-ditions. I n the case of nitrocatechol methylene ether the greater reactivity may be due to the influence of the oxygen atom in the meta-position with respect t o the nitro-group on the distribution of affinity and this is confirmed by the observation that we find i t impossible to realise a similar reaction in the case of nitroethylene-dioxybenzene (IV) in which substance the oxygen atom is one atom further removed from the point of attack.Cardwell and Robinson (T. 1915 107 255) a showed that 5-nitroguaiacol is obtained by the hydrolysis of nitroveratrole with hydrobromic acid and that its acetyl derivative results when acetylguaiacol is nitrated. Similarly the hydrolysis of nitrocatechol diethyl ether with hydrobromic acid is now found t o yield 5-nitro-2-ethoxy-phenol and the benzayl derivative (V) of this substance is obtained by the nitration of 2-benzoyloxyphenetole.NO2 YH,*Of\NO BzO “‘NO FH,*O()NO, CH”O\/ E d \/ I CH,*O\,NO, (IV). (V.) (VI.) In order to examine the scissioii of the ethylenedioxy-ring we submitted 4 5 6-trinitroethylenedioxybenzene (VI) to the action of ammonia. It was quickly changed to a mixture of 3:5-dinitro-2 4-diamino-@-hydroxyethoxybenzene (VII) and 4 5-dinitro-6-aminoethylenedioxybenzene (VIII). The benzoyl derivative (IX) of the former substance exists in two highly characteristic chromo CERTAIN SURST1TUTE:D CYCLIC CATECHOL ETHERS. 931 isometric modifications which are described in the experimental portion on-p. 938. NO2 NH2 NH,/\NH2 7H2*O/\N02 HO*CH,~CH,.O~ \/ /NO CH,*O!,)NO, (VII.) The substance VII is rapidly and quantitatively hydrolysed by hot aqueous alkali hydroxides with the production of two mole-cules of ammonia and the phenol X which was transformed into the well-known dinitrodiaminoanisole (XI) as shown below.NO2 EO, \/ NO2 NO2 llO/\OH HBr HO/\OH M e 2 S O ) --+ HOI /NO G o $ H O ~ H~*CH,=OI ]NO \/ (X.1 MeO/'O.\le N H ~ NH,/\NH, MeC()NO -f MeO\)NO, I n view of the close analogy which exists between the hydroxyl of the carboxyl group and the hydroxyl of opdinitrophenols it is interesting to note that the substance X shows no tendency to form an internal anhydride. It would be fair t o conclude from this that lactone-formation in some way involves the whole carb-oxyl group and not merely the acidic hydroxyl of that group. I n the nitrophenols it is the whole conjugated system extending from the nitro-group to the hydroxyl that corresponds with the carbmyl in the acids and the steric conditions favourable to lactone-formation are accordingly not necessarily present in the dinitro-phenol (X).The constitution of the dinitroamine VIII was proved by eliminating the amino-group when 4 5-dinitroethylenedioxy-benzene was obtained. The nitro-groups in positions 4 and 5 in trinitroveratrole have previously been shown to be displaceable under certain conditions but this is the first example of the dis-placement of the 6-nitro-group in a trinitrocatechol ether. We are inclined to assign the result to the unexplained influence which a fused ring has on the a-position in rendering atoms and groups (XI.) VOT. 0x1. 0 932 EOBINSON AND ROBINSON THE SCISSION 03’ attached a t this point more liable to substitution and displace-ment.Naphthalene for example yields a-nitronaphthalene on nitration and the reactivity of groups in the 1- 4- 5- or 8-posi-tions in substituted anthraquinones may also be cited. E X P E R I M E N T A L. Prepration of 4-Nitrocatechol Methylene Ether. Salway has shown (T. 1908 95 1163) that the nitration of piperonal yields nitrocatechol methylene ether in addition to nitropiperonal and when the former substance is the object’ of preparation the nitroaldehyde may be converted as described in the foregoing communication into 5-nitro-4-aminocatechol methylene ether (compare p. 915) from which the amino-group may be eliminated in the usual manner. 5-Nitro-4-aminocatechol methylene ether (9 grams) was mixed with alcohol (150 c.c.), sulphuric acid (15 c.c.) and powdered sodium nitrite (5 grams), and the solution boiled during five minutes after which a further quantity of sodium nitrite (4 grams) was added and the heating continued for five minutes longer.The crystalline precipitate obtained after the addition of water was collected and dried and weighed 8 grams. After crystallisation from alcohol the substance melted a t 147O and at the same temperature when mixed with a specimen of 4-nitrocat echo methylene ether obtained by the nitra-tion of piperonal. Formcr f i o i ~ of 5-Nitrogunincol b y the Internction of Nitrocrrtechol Methylene Ether and Sodium Methoxide. Nitrocatechol methylene ether (1.1 grams) was heated on the steam-bath with a solution of sodium methoxide (from 3 grams of sodium and 40 grams of methyl alcohol) until the red sodium salt of the nitrophenol crystallised from the solution.This requires from one to three minutes. After cooling the salt was separated, washed with ether dissolved in water and the solution acidified with hydrochloric acid. The yellow precipitate was collected (1.0 gram when dry) and recrystallised from water when it was obtained in needles melting a t 105O and at the same temperature when mixed with 5-nitroguaiacol. 5-Nitro-2-ethozyphenol (111). An almost quantitative yield of this substance was produced when nitrocatechol methylene ether was treated in ethyl-alcoholic solution with sodium ethoxide under precisely the same condition CERTAIN SUBSTITUTED CYCLIC CATECHOL ETHERS.933 as are described above for the corresponding reaction with sodium methoxide. I n this case no salt separated from the solution but the reaction was as rapid as in the former example. The sub-stance is more sparingly soluble in water and alcohols than is 5-nitroguaiacol and separates from methyl alcohol containing a little water in pale yellow well-defined prisms melting a t 113-114': 0.1478 gave 0.2854 CO and 0.0671 B20. C,H90,N requires C=52*5; H=4*9 per cent. The substance dissolves in sodium carbonate solution and the orange-red colour produced can scarcely be distinguished from that of a similar>solution of 5-nitroguaiacol. I n view of the fact that 4-nitroguaiacol dissolves in sodium carbonate t o a yellow solution this behaviour indicates the constitution of the sub-stance and the matter is placed beyond doubt by the preparation of 5-nitro-2-ethoxyphenol by the two followiiig methods which in the corresponding methoxy-series lead to 5-nitroguaiacol.( A ) 2-Ethoxyphenol was benzoylated by an application of the Schotten-Baumann reaction and the dry benzoyl derivative dis-solved in an excess of cold nitric acid (D 1.42). Undue rise of temperature was checked and after ten minutes the clear solu-tion was added t o water and the precipitated oil induced to solidify by the usual methods. The solid was collected and crystallised from methyl alcohol when i t was obtained in felted masses of slender colourless needles melting a t 101-102° : 0.1108 gave 4.7 C.C.N a t 18O and 762 mm. C,,H,,O,N requires N = 4.9 per cent. 5-Nitro-2-ethoxyphenyl beiizoute was dissolved in boiling ethyl alcohol and hydrolysed by the addition of a solution of three times its weight of potassium hydroxide dissolved in water. After heat-ing on the steam-bath during three minutes i t was found that a sample was completely soluble in water. Excess of alcohol was removed the residue dissolved in water just acidified with hydro-chloric acid and then treated with sufficient aqueous sodium carbonate to restore a faint orange colour to the liquid. Under these conditions the benzoic acid remained in solution and the nitrophenol was precipitated. It was collected crystallised from aqueous methyl alcohol and melted a t 112-114O and at the same temperature when mixed with some of the substance prepared from nitrocatechol methylene ether.( B ) 4-Nitrocatechol diethyl ether * (20 grams) was heated on * 4 5-Dinitrocatechol diethyl ether C,,H,,O,N, does not appear to have been previously prepared. It was obtained in quantitative yield by dissolving the mononitro-derivative in nitric acid (D 1.42). In about half an C=52*7; H=5*0. N=5.0. 0 0 934 ROBINSON AND ROBINSON THE SCISSION OF the steam-bath during three hours with a saturated solution (80 grams) of hydrobromic acid in acetic acid. After the addition of water the solution was rendered alkaline by the addition of sodium hydroxide and the unchanged substance removed by filtra-tion (3 grams). The filtrate was acidified and the precipitated mixture of nitrophenols collected (14 grams).After several crystallisations from aqueous methyl alcohol 8 grams of pure 5-nitrcl2-ethoxyphenol melting a t 112-114O were obtained. The melting point was not depressed by admixture with a specimen of the substance obtained by the method ( A ) described above. The alcoholic mother liquors were added to the aqueous solution from which the crude phenol was separated and the whole was extracted with ether. The recovered mixed nitrophenols were converted into benzoyl derivatives by treatment with benzoyl chloride and sodium hydroxide in aqueous solution and the mixture of the benzoates was then fractionally crystallised from methyl alcohol. The most sparingly soluble substance crystallised in slender needles, and was recrystallised from acetic acid.It melted at 156O and was identified as the dibenzoyl derivative of nitrocatechol. The substance was also obtained in the following manner o-Phenylene benzoate (13 grams) was added to cold nitric acid (50 c.c. D 1-42), when the benzoate became an oil which was well mixed with the acid. After remaining overnight the nitration was completed and hour the dinitro-derivative began t o separate from the solution which was gently warmed to ensure the completion of the reaction. Water was added and the solid collected and crystallised from alcohol from which it separated in characteristic citron-yellow micaceous flakes melting at 11 3" and rather sparingly soluble in alcohol. It is quantitatively converted by nitric acid in sulphuric acid solution to the 4 5 6.trinitrocatechol diethyl ether which Blanksma (Rec.trav. chim. 1905 24 40) has obtained lcly the nitration of 4 6-dinitrocatechol diethyl ether. The substance melted a t 122" and was converted by alcoholic ammonia into 3 5-dinitro-2 4-diarizinophenetole, crystallising from nitrobenzene in hematite-like plates melting a t 257" (Nietzki; Annalen 1882,215 153 and Blanksma Zoc. cit. give 245"). 0.4951 distilled with 10 per cent. aqueous potassium hydroxide evolved NH, which neutralised 40.8 C.C. N/lO-HCl whereas this amount of a substance, C,H,,0,N4 yielding 2NH requires 40.9 C.C. 4 5-Dinitrocatechol diethyl ether dissolved in hot aqueous alcoholic hydrochloric acid was reduced by the addition of zinc dust. The colourless solution was mixed with sodium acetate and filtered.The filtrate was then heated with a solution of phenanthraquinone in hot sodium hydrogen sulphite solution and a voluminous pale yellow precipitate of the phenazine derivative was quickly formed. The substance was collected and crystallised from xylene. 2 3-Diethoxyphenanthraphenazine C,,H,,O,N, crystallises in pale flesh-coloured needles which change on keeping in contact with the solvent into rectangular plates melting a t 230'. It dissolves in sulphuric acid to a bright magenta solution yellow on dilution with water and in benzene to a pale yellow solution which exhibits intense violet fluorescence CERTAIN SUBSTITUTED CYCLIC CATECHOL ETHERS. 935 the derivative had solidified. Water was added and the substance collected and crystallised from acetic acid and acetone.The suh-stance melted a t 156O and a t the same temperature when mixed with the product obtained as above: 0.2091 gave 6.4 C.C. N a t 1 7 O and 756 mm. C,,H,,O,N requires N = 3.9 per cent. After the separation of the relatively small quantity of the dibenzoate a substance crystallised in colourless prisms which is undoubtedly the b enzoyl derivative of 4-nitro-2-ethoxyphenol. It could not' be obtained in a pure condition and always melted over a considerable range of temperature. A few crystals were mechanically separated and it was noted that on hydrolysis a nitrophenol was obtained which gave a pure yellow solution in aqueous alkali whereas nitrocatechol gives a blood-red and 5-nitro-2-ethoxyphenol an orange-red solution.From the mother liquors considerable quantities of the benzoyl derivative of 5-nitro. 2-ethoxyphenol melting a t 1 0 1 O were obtained. N=3*6. 4 5 6-Trinitroethylenedioxybenzene (VI). The ethylene ether of nitrocatechol is less readily nitrated than the corresponding dimethyl or diethyl ethers and it was found desirable t o operate under the following conditions which are more convenient than those employed by Ghosh (T. 1915 107 1591) for the same purpoee: 4-Nitroethylenedioxybenzene (10 grams) was dissolved in nitric acid (100 grams D 1-5) and after remaining overnight at the ordinary temperature water was added and the precipitate collected and crystallised from alcohol. The substance was obtained in prismatic needles melting a t 132-133O and was occasionally also obtained in leaflets melting a t the same tempera-ture.The substance is readily reduced to a diamine which gives a red coloration with ferric chloride and condenses with phenanthraquinone to the sparingly soluble 2 3-ethylenedioxyphem~thraphenazine, which crystallises from acetic acid in clusters of slender yellow needles melting a t 239-240°. The colour reaction in sulphuric acid and the fluorescence in benzene are indistinguishable from those exhibited by 2 3-dimethoxyphenanthraphenazine. 4 5-Dinitroethylenedioxybenzene (22 grams) was finely powdered and dissolved as far as possible in sulphuric acid (100 c.c.) and gradually nitrated by the addition with cooling of nitric acid (20 c.c. D 1.5) in sulphuric acid (20 c.c.).The dinitro-compound It separated from acetic acid in plates. C22HI*O2NZ 936 ROBINSON AND ROBINSON THE SCISSION OF passed into solution and the trinitro-derivative crystallised out. The mixture was poured into water and the colourless solid col-lected and crystallised from alcohol. The substance was sparingly soluble and separated in laminae melting a t 155-156O as state'd by Ghosh (Zoc. cit.). Frcm concentrated solutions i t was obtained in prismatic needles : 0 1169 gave 0.1509 CO and 0'0141 H,O. Each stage in the nitration of ethylenedioxybenzene proceeds in C=35*2; H=1*4. C,H,O,N requires C = 35.4 ; H = 1.8 per cent,. a quantitative manner. 3 5-Dini t7-0-2 4-dicr m in 0-p- h y drox y p t hoxy b e m 4 ne (VII) . This substance is obtained by the action of ammonia on trinitro-ethylenedioxybenzene but when the reaction was carried out in ethyl-alcoholic solution it was found that the percentage of carbon obtained on analysis was consistently about 1 per cent.too high, and the amount of ammonia obtained on hydrolysis with aqueous potassium hydroxide was also greater than the theoretical. This indicated that a diamino-derivative of similar constitution and properties but of smaller molecular weight contaminated the pro.-duct and eventually proof was obtained that the impurity was dinitrodiaminophenetole. The production of this substance clearly showed that the trinitroethylenedioxybenzene under the influence of the alcoholic ammonia was converted in part into trinitro-catechol diethyl ether and accordingly the use of alcohol as a solvent was avoided.The following conditions were ultimately adopted Trinitroethylenedioxybenzene (12 grams) was dissolved in pyridine (50 grams) and heated on the steam-bath under reflux with aqueous ammonia (50 c.c. D 0.880). I n a short time the product of the reaction separated in red needles and a further quantity of ammonia solution (50 c.c.) was then added and the heating continued f o r half an hour. After cooling and adding water the orange-red solid was collected and when dry weighed 9.5 grams. This material is a mixture containing about 75 per cent. of dinitrodiaminohydroxyethoxybenzene and about 25 per cent. of a dinitroaminoethylendioxybenzene. The two sub-stances may be approximately separated by crystallisation from xylene in which the former is the more sparingly soluble.The further purification of the second crop of crystals is described below but the intense red crystals which separate first can be obtained in a pure condition by two recrystallisations from nitro-benzene. It is perhaps better to extract the crude product with hot xylene insufficient to dissolve the whole and then to crystal CERTAIN SUBSTITUTED CYCLIC CATECIIOL ETHERS. 937 lise the residue from nitrobenzene. red needles melting a t 240O: The substance occurs in deep 0-1543 gave 0.2127 CO and 0.0583 H,O. 0.0978 , 18.2 C.C. N a t 2 4 O and 750 mm. N=21*1. C,H,,O,N requires C == 37.2 ; H = 3.9 ; N = 21.7 per cent. 0.3977 distilled with 50 C.C. of 10 per cent. aqueous potassium hydroxide evolved NH, which neutralised 31 0 C.C.NIlO-HCl, whereas this amount of a substance CPH1006N4 yielding 2NH3, requires 30.8 C.C. This substance resembles in its behaviour the similarly consti-tuted dinitrodiaminophenetole. It is quite devoid of basic character and is very sparingly soluble in organic solvents. When its solution in sulphuric acid is treated with sodium nitrite a reaction of unexplained character occurs and on the addition of water there is produced a transient intense purple coloration. This quickly disappears and the yellow solution contains traces of some diazo-salt as indicated by coupling with “R-salt,” but the major part of the substance has undergone decomposition. C=37*6; H=4*2. 3 5-Diiiitro-2 4diamino-P-b enzoylozyethoxyb en z e u e (IX).It was noticed that the amino-groups of dinitrodiaminoanisole and the corresponding phenetole derivative cannot be benzoylated by the action of pyridine and benzoyl chloride on these substances. Recourse was therefore had to this reaction in order to prove the presence of a hydroxyl group in the substance described in the last section. The dinitrodiaminohydroxyethoxybenzene (4 grams) was dissolved in pyridine (30 c.c.) and benzoyl chloride (15 c.c.). A certain amount of heat was developed and when the reaction had subsided ether was added and the red precipitate collected and washed with water. I t was dried in the air and then crystal-lised from ethyl acetate in which i t is sparingly soluble: 0.1526 gave 0.3800 CO and 0.0560 H,O. C1,H,,0,N4 requires C = 49.7 ; H = 3.9 per cent.0.3827 distilled with aqueous potassium hydroxide evolved NH,, which neutralised 21.5 C.C. N/lO-HCl whereas this amount of a substmce Cl5HI4O7N4 yielding ZNH, requires 21 *2 C.C. On acidifying the alkaline solution from this experiment, benzoic acid and 3 5-dinitro-2 4-dihydroxy-P-hydroxyethoxy-benzene (X) were obtained as a crystalline precipitate. The latter substance was identified after removal of the benzoic acid by repeated evaporations of the solution in water. This benzoyl derivative is dimorphous and occurs in two distinct chromoisomeric modifications. Both forms melt at 180-181O C=50*0; H=4*1 938 ROBINSON AND ROBINSON THE SCISSION OF alone or mixed. The crimson form crystallises readily from most solvents and may be easily obtained by crystallisation from xyleiie or by the addition of ether alcohol or light petroleum to a solu-tion of the substance in pyridine.It occurs in crimson plates with parallel edges. The orange-yellow modification is character-ised by its sparing &solubility in ethyl acetate and is the first to crystallise from this solvent. It is obtained by heating the red form or a mixture of the two to near the melting point and then extracting the material with ethyl acetate. It crystallises from this solvent in needles. The change from red to yellow on heating also occurs slowly a t temperatures above loo@ but nothing in the nature of a transition point could be determined. Recrystallisa-tion of either form from nitrobenzene resulted in crimson leaflets containing solvent of crystallisat.ion.3 5-Dinitro-2 4-dih ydroxy-P-h ydroxy e t hox y b en z ene (X) . This substance was readily obtained by boiling dinitrodiamino-P-hydroxyethoxybenzene with a 10 per cent. solution of potassium hydroxide until all the ammonia was evolved. The reaction was complete in a few minutes and the orange solution was acidified with hydrochloric acid. Bronze-yellow needles separated on cool-ing and the substance was purified by recrystallisation from dilute hydrochloric acid. It is readily soluble in water or alcohol less readily so in dilute hydrochloric acid and it has powerful dyeing properties. The substance forms well-defined brit,tle needles which darken at 160° and melt and decompose at 170O: 0.1379 gave 0.1884 CO and 0.0395 H,O.This compound shows no tdndency to change by loss of water to a lactone-like ethylenedioxy-derivative. Its constitution was proved in the following manner The glycol residue was removed by boil-ing several hours with an excess of concentrated hydrobromic acid, and the resulting phenol converted into its trimethyl ether by means of methyl sulphate and potassium carbonate in boiling nitro-benzene solution. This substance was then heated a t looo in a sealed tube with concentrated aqueous ammonia and the resulting red precipitate crystallised from nitrobenzene. The red plates were identified as consisting of 3 5-dinitro-2 4-diaminoanisole and melted a t 255O and a t the same temperature when mixed with a specimen of the substance obtained from trinitroveratrole.C=37.3; H=3*2. C,H,O,N requires c = 36.9 ; H = 3.1 per cent CERTAIN SUBSTITUTED CYCLIC CATECHOL ETHERS. 939 4 Ei-Dini tro-6amino e t h y len edio xy b enz en e (VIII). This by-product of the preparation of dinitrodiaminohydroxy-ethoxybenzene from trinitroethylenedioxybenzene is separated in an impure form by extracting the crude product with xylene as already described. It may be freed from its congener by taking advantage of the ready attack of the latter by warm aqueous potassium hydroxide. The crude crystals obtained from xylane were carefully warmed with a 5 per cent. solution of potassium hydroxide until the precipitate had a pure yellow colour. The substance was then collected washed with water and several times crystallised from ethyl acetate from which i t separated in orange-yellow rhombic prisms melting a t 202O: 0.1133 gave 0.1676 CO and 0.0340 H,O.The substance is sparingly soluble in moat solvents and has weak basic character. On boiling with aqueous potassium hydroxide it is slowly attacked yielding a cherry-red solution. The base is diazotisable in dilute hydrochloric acid suspension and the diazonium salt couples with P-naph tho1 to a crimson azo-compound' When trinitroethylenedioxybenzene is reduced in alcoholic hydro-chloric acid with zinc dust a colourless solution is obtained which gives an intense green coloration with ferric chloride. The sub-stance now under consideration shows the same behaviour. On boiling with alcoholic sulphuric acid and powdered sodium nitrite, the amino-group was easily removed and after the addition of water pearly leaflets separated and were collected.The substance was recrystallised from alcohol and melted a t 132-133O alone or mixed with a specimen of 4 5-dinitroethylenedioxybenzene. Since the conEtitution of the latter substance is known by inference only, the material obtained from the above experiment was hydrolysed by heating with concentrated aqueous hydrobromic acid and the dinitrocatechol so obtained methylated in the usual manner and the product identified with 4 5-dinitroveratrole. A cetyl Deriva tiwe.-The dinitroamine could Be recrystallised from acetic anhydride but on the addition of a trace of sulphuric acid the aminegroup was acetylated and the derivative was col-lected after decomposition of the excess of acetic anhydride by means of dilute hydrochloric acid. The substance was recrystal-lised from alcohol in which i t is sparingly soluble and obtained in hexagonal prisms melting at 257O. The substance appears colourless when first prepared but the compact crystals have a pde yellow colour. C=40.2; H=3.3. C,H,O,N requires C= 39.8 ; H = 2.9 per cent. 0 o 940 HINDMARSH KNIGHT AND ROBINSON : The similarly constit*uted 4 5-dinitro-3-acetylaminoveratrole has recently been prepared by Gibson Simonsen and Rau (this vol., p. 78) who find that it dissolves in aqueous potassium hydroxide t o a yellow solution which on acidification gives a precipitate of the unchanged substance. This ethylene ether shows the same characteristic behaviour. UNIVERSITY OF LIVERPOOL. [Received September 4th 1917.