THE PRO(11IESSlVE BROMINATION OF TOLUENE. 501LI I. - The Progress h e Bromination 0 f To 1 u enc.By JULIUS BEREND COHEN and PAVITRA KUMAB DGTT.THE general rules of substitution in the benzene nucleus have beenarrived a t mainly from the study of the di-substitution products.The formation of tri-derivatives has been examined only in a fewcases. Cohen and Dakin published a series of papers during1901-1904 on the progressive chlorination of toluene. With theobject of discovering some general rule they studied the orienta-tion of the entrant chlorine atom where one or more were alreadypresent in the nucleus.Their results in the case of chlorination of the isomeric mono-chlorotoluenes may be briefly stated as follows (T., 1901, 79, 1111) :The ortho- and para-compounds yield all the possible isomerides,whereas the meta-compound gives only two out of the four, neitherthe 2 : 3- nor the 3 : 5-compound being formed.The only investigation which we have been able to find on thenature and amount of the products of bromination, is one byA.K. Miller (T., 1892, 61, 1023). Ortho- and para-bromotolueneswere brominated by keeping the substances in contact with thecalculated quantity of bromine for several days with the additionof a trace of iodine. Miller found that the ortho-compound givestwo out of four possible dibromotoluenes, and the para-compoundgives both of the possible derivatives. He also obtained a tribromo-toluene melting at 112-113O from the products of bromination ofpbromotoluene, the substance being identical with the 2 : 4 : 5-tri-bromotoluene (m.p. 11 1-2-1 12.8O) obtained indirectly by Nevileand Winther (Bey., 1880, 13, 970).I n comparing these results with’ those of the chlorination oftoluene considerable discrepancies appear, and in considering thisdifference in the orienting effect of the two halogens we thought itdesirable to repeat Miller’s work and extend it along the samelines as previously followed in studying the chlorination of toluene.It was therefore necessary to prepare a t the outset the six iso-meric dibromotoluenes and such of their derivatives as would affordthe best means of identification. The following table gives a listof these derivatives and their melting points. They have beenprepared for the most part in the same way as the derivatives ofthe dichlorotoluenes (T., 1901, 79, 1111), and hence many of thedetails of the processes have been omitted in this paper.F o r com-parison the melting points of the corresponding dichloro-compoundsare subjoined.L L 502 COHEN AND DUTT:M e l t k g Yoiiits of Berivntives of the Dibromotoluenes.Mono-Substance. nitro-.2 : 3 (m. p. 30-31") 58-59"2 : 4 liquid ............. 79-802 : 5 liquid. ............ 87-882 : 6 (m. p. 5.5") .... 49-50*3 : 4 liquid. ............ 86-873 : 5 (m. p. 39") ...... 124Sulphonyl Sulphon- BenzoicDinitro-. chloride. amide. acids.107-109" 93" 214' 149-150"129-130 86-88 212-213 168-169142-143 73-74 210-211 153161-1 62 101 204 146-147129-130 104-105 212 229-230-\ (156-157 104.5-105j \ liquid 195 213-214Jlelting Points of Derivatives of the Dichlorotoluenes.2: 3 ..................50-5-51.5" 71-72" liquid 222' 163"2 : 4 .................. 54-55 104 71 176 159-1602: 5 .................. 50-51 100--101 45-46 191-192 1532 : 6 .................. 53 121-122 liquid 207 139-1403 : 4 .................. 63-64 91.5-92.5 81 190-191 200-2013 : 5 .................. 61-62 99-100 44-45 168-169 182-183* Nevile and Winther (Bey., 1880, 13, 970 ; 1881, 14, 419), from a preparation ofthe 2 : 6-compo1u1d obtained indirectly from nz-toluidine, give 56.8- 57", to whichthey assign the constitution 2 : 6-dibromo-p-nitrotolueiie. We have recently dis-covered a paper by J. J. Elanksma (Chesn. IVcekbZnd., 1912, 9, 968), in which theauthor gives the melting point of the mononitro-compound as 50" and that of thedinitro-derivative 161" (their constitution being 2 : 6-dtbromo-3-niiro- and 2 : 6-dibromo-3 : 5-dinitro-toluenes respectively), rcsults which agree substantially with ours.It will be seen from the above table that in the case of the di-bromotoluenes the melting points of the dinitro-derivatives and thedibromobenzoic acids present the widest range of temperature,whereas those of the sulphonamides are too close together t o be satis-factorily used as a means of identification.The crystalline appear-ance and solubilities of the barium sulphonates have afforded a use-ful method of distinguishing between the 2 : 5-compound, whichforms highly refractive rhombohedra1 prisms, the 3 : 4-compound,which crystallises in thin plates, and the sulphonates of the 2 : 3-and 2 : 4-compounds, which are much more soluble and crystallisein needles. Advantage was also taken of the insolubility of thesulphonic acid of the 3 : 5-compound.On the same principle we have carried our investigation as faras the tribromotoluenes in order to see in what respects the resultsof bromination of the dibromotoluenes differ from those of chlor-ination of the dichlorotoluenes (Cohen and Dakin, T., 1902, 81,1324).Each of the six isomeric dibromotoluenss was brominatedseparately, and the products in each case were examined by com-paring them with the derivatives of the tribromotoluenes preparedfor this purpose.The tribromotoluenes have been obtained syn-thetically by various methods, and their derivatives, such as nitro-THE PROGRESSIVE RROMINATION OF TOLUENE.503dinitro-tribromotoluenes and tribromobenzoic acids, have also beenobtained. The preparation and constitution of all the six isomerictribromotoluenes have been described by Nevile and Winther (Ber.,1880, 13, 974; 1881, 14, 419), but not of their derivatives.The following table gives a list of the melting points of thetribromotoluenes and their derivatives as determined by ourselves :Substance. Mononitro-. Dinitro-. Benzoic acids.2 : 3 : 4 (m. p. 44-45") . . . 107-108"2 : 3 : 5 (m. p. 53-54"] ... { 88.5-990}2 : 3 : 6 (m. p. 5&-59") . . . 91-92 202-203 not oxidisable2 : 4 : 5 (m. p. 112-113") 130-131.5 278-279 195-1962 : 4 : 6 (m.p. 65-66'). . . 74-75.5 21 7-21 8 1%-1873 : 4 : 5 (m. p. 88-S9") . . . 104-105 211-5 2351 97- 1 99" 197-198"209-210 193-194 67-68The methods adopted for separating and identifying the productswill be referred to individually. I n some cases the melting pointsand solubilities of the dinitro-compounds furnish the best means,whereas in others those of the benzoic acids are most suitable,especially in dealing with the diortho-substituted acids, which arenot esterified by Fischer and Speier's method.While engaged in this investigation we also studied the formationof monobromotoluenes from toluene in order to see whether anytrace of the meta-compound was formed in the reaction. It iswell-known that the first action of bromine on toluene gives theortho- and para-compounds in the proportion of about 40 per cent.of the former to 60 per cent.of the latter. No evidence existsas to the presence'of the meta-compound in the final product, whichmay be due to the small quantity present in the mixture.This was found to be the case, a small quantity of m-bromo-toluene being formed with the ortho- and para-compounds.I n continuation of this investigation the chlorination of the mono-bromotoluenes and the bromination of the inonochlorotoluenes arebeing examined.EXPERIMENTAL.For purposes of reference the experimental part has been sub-divided as follows :I. Bromination of toluene.11. (a) Preparation of the six isomeric dibromotoluenes.( b ) Bromination of the isomeric monobromotoluenes.111.( a ) Preparation of the isomeric tribromotoluenes and their( b ) Bromination of the isomeric dibromotoluenes.derivatives504 COHEN AKD DUTT:PART I.-The Bromanation of Toluene.The bromination was effected by dropping 26.5 C.C. of bromineinto 46 grams of freshly distilled and chemically pure toluenecontaining a few pieces of freshly prepared aluminium-mercurycouple (about 0.5 gram), the operation being conducted in thesame way as in the preparation of bromobenzene (Cohen and Dakin,T., 1899, 76, 894). The product, after being washed with sodiumhydroxide solution, was dried over calcium chloride and distilled.The fraction distilling a t 180-190° was collected separately(40 grams). Twenty-five grams of this fraction were oxidised tobromobenzoic acids by heating in sealed tubes with dilute nitricacid (lHNO, : 3H,O) a t 130-135O for six hours.After cooling,the contents of the tubes were treated with sodium hydroxide solu-tion, and the unchanged bromotoluenes removed by shaking withchloroform. The free acids were then liberated by the additionof hydrochlbric acid to the clear alkaline solution, collected, andcrystallised fractionally from. dilute alcohol.After repeated crystallisation, a fraction was obtained meltinga t 135-145O. This was then converted into the acid chloride bymeans of phosphorus pentachloride in the usual way, and themethyl ester prepared by treating the acid chloride with methylalcohol. The mixed methyl esters were then distilled fractionallyunder diminished pressure.From this a portion was obtainedwhich solidified in the condenser tube, and also in the receiver aftercooling. The solid portions on crystallising twice from dilutealcohol gave 0.3 gram of ester melting a t 29'31O. The meltingpoint of the methyl ester of the metabromobenzoic acid is 31-32O.The only other ester which melts very near t o this is the 2:4-di-bromo-derivative melting a t 33O. I n order to confirm the presenceof m-bromobenzoic acid, 0.2 gram of the ester was hydrolysed withalcoholic potassium hydroxide, and the acid liberated by means ofhydrochloric acid. After crystallising t'wice from hot water, i tmelted a t 155-156O, and thus the presence of m-bromotoluene 'inthe original product was confirmed.The 2 : 4-dibromobenzoic acidmelts at 168-169O.PART 11.-(a). Preparation of the Six Isomeric Dibrortiotoluenes.The 2 : 3-, 2 : 4-, and 2 : 5-dibromotoluenes were prepared fromthe corresponding nitrotoluidines (Cohen and Zortman, T., 1906,89, 49). The 2 : 3-compound was also obtained by Wyiine's methodfrom o-toluidinesulphonic acid, which gave a better yield than thatfrom the nitrotoluidine (T., 1892, 61, 1040), and from 6-bromo-o-toluidins by the method described or1 p. 514. The 2 : 4-compounTHE PROGRESSIVE BROMINATION OF TOLUENE. 505was also obtained by brorninating pnitrotoluene in the presenceof iron (Scheufelen, AnnaZen, 1903, 231, 171), and then reducingthe 2-bromo-4-nitrotoluene (m. p 76-77O) formed in the reaction,and finally replacing the amino-group by bromine.The 2 : 5-corn-pound was also obtained from both m-acetotoluidide and o-aceto-toluidide by bromination, and then replacing the amino-group bybromine.The 3 : 4-compound was obtained from the corresponding nitro-toluidine and also from pacetotoluidide by bromination.The 3 : 5-compound was prepared by removing the amino-groupfrom 3 : 5-dibromo-ptoluidia e by diazotising in alcoholic solution,the latter substance being obtained by brominating pacetotoluidideor ptoluidine itself,The 2 : 6-compound was prepared from 2 : 6-dinitrotoluene byalternate reduction and diazotisation. The conversion of 6-bromo-o-toluidine into dibromotoluene cannot be effected satisfactorilyin the ordinary way; if diazotisation is carried out in hydrobromicacid, the diazoamino-compound is produced, and a very little of thedibromotoluene ; if in presence of hydrochloric acid and subsequenttreatment with cuprous bromide in hydrobromic acid, a mixtureof chlorobromo- and dibromo-toluene is produced.The method finally adopted was as follows: Seventeen gramsof 6-bromo-o-toluidine were dissolved in 100 C.C.of absolute alcohol,to which were added 60 C.C. of hydrobromic acid. Twelve C.C. ofamyl nitrite were then slowly run in keeping the temperature below20°. The mass became almost solid, and was filtered and washedwith alcohol. It was then dissolved in hydrobromic acid and cooledto Oo, and the mixture left in ice for an hour. It was then distilledin steam, the distillate extracted with chloroform, dehydrated overanhydrous sodium sulphate, the chloroform distilled off, and theresidue distilled under diminished pressure from a paraffin-bath.The major portion (12 grams) was collected a t 122O/23 mm.as acolourless liquid (m. p. 5-6O).Before this method was finally adopted, two others given belowwere tried, but these were found to be tedious and inconvenient,and, in addition, gave very poor yields:1. From 2 : 6-dibromo-4-nitro-m-toluidine (Nevile and Winther,Ber., 1880, 13, 974); which again was prepared from m-aceto-toluidide, as described under 2 : 3 : 6-tribromotoluene (see later).2. From 2-bromo-4-nitrotoluene, by bromination in presence ofiron (Scheufelen, An.naZen, 1903, 231, 178).By this method aportion only of the bromonitrotoluene is converted into the 2 : 6-di-bromo-compound when heated in sealed tubes, and the separation oft h e two constituents is very difficult506 COHEN AND DUTT:PART 11.-(b). Z'he Bromiization of the Isomeric Mono bromo-toluenes : Bromination of o-Brontotoluene.The process was carried out in the following way, and was appliedwithout modification t o the other bromotoluenes.Fifty-two grams of o-bromotoluene were dissolved in an equalweight of carbon tetrachloride in a flask provided with a refluxcondenser, and 0.5 gram of aluminium-mercury couple was added.The liquid was cooled in ice, and 12 C.C. of bromine in an equalvolume of carbon tetrachloride were slowly dropped in from atap-funnel attached to the upper end of the reflux condenser.Substitution takes place rapidly; after some time the product waspoured into water, shaken with sodium hydroxide solution, de-hydrated and distilled under diminished pressure.Thirtysevengrams, boiling a t 130--133O/15 mm., were obtained, about 10 gramsof a fraction of lower boiling point, consisting chiefly of unchangedproduct, and about 1 gram boiling above 133O. The process wasrepeated several times with about the same quantity of materialand with similar results.Sulphonation of the Mixed Dibromotoluenes.Thirty-three grams of the above product were mixed with 64grams of fuming sulphuric acid, shaken, and then heated on thewater-bath, until a test sample dissolved completely in water; Aftercooling, the liquid was poured into water and boiled with excessof barium carbonate, filtered, and the residue extracted repeatedlywith boiling water, until no more dissolved.On cooling 37 gramsof the barium sulphonate crystallised, having the appearance oflarge, highly refractive, rhombohedra1 plates, characteristic of the2 : 5-compound. The mother liquors were concentrated to a smallbulk, and deposited 1.5 grams of needles. Both fractions weretreated se,parately with phosphorus pentachloride and subsequentlywith ammonium carbonate, to convert the sulphonyl chloride intothe sulphonamide, and the product crystallised from alcohol. Theless soluble salt crystallising .in plates gave a sulphonamide meltinga t 207O (2 : 5 =210--311°), and the sulphonamide obtained from themore soluble salt, crystallising in needles after three recrystallisa-tions, melted a t 209O (2 :4 =212O).Although, as stated above, themelting points of the sulphonamides afford no definite indicationof the isomerides present in a mixture, the characteristic appar-ance of the barium salt of the 2 : 5-compound distinguishes it quitesharply from the other three isomerides. The results of nitrationand oxidation leave no doubt that the other barium salt is thaTHE PROGRESSIVE BROMINATlON OF TOLUENE. 507of the 2 : 4-compound. The result confirms Miller’s observationthat the main product of bromination of the ortho-compound isthe 2 : 5-derivative.Nitration of the Mixed Dibrornotolaenes.The mixed dibromotoluenes were converted into the dinitro-compounds by heating with fuming nitric acid and concentratedsulphuric acid on the water-bath, pouring into water, and fraction-ally crystallising the product.After repeated crystallisation from alcohol, the substance wasseparated into two fractions, a large fraction crystallising in fineneedles and melting a t 142-143O (2:5=142--143O), and a smallfraction crystallising in prisms with pointed ends characteristic ofthe 2 : 4-compound and melting a t 124-126O (2 : 4=129-130O).The result is in entire agreement with that derived from sulphon-ation. The 2 : 4-compound formed a small proportion of the mixtureand was not quite pure.Very careful examination was made forthe 2 : 6-dinitro-compound, which has a much higher melting pointthan any of the other isomerides, but wibhout result.Oxidation of the Mixed Dibromotoluenes.An attempt was made to repeat Miller’s method of oxidationby boiling the dibromotoluenes with dilute nitric acid a t theordinary pressure, but although the process was repeated withdifferent preparations and different strengths of acids and con-tinued f o r a week a t a time, scarcely any acid was formed, and weare unable to explain Miller’s result.Ullmann’s permanganatemethod was also tried ineffectually. The oxidation was then carriedout in sealed tubes a t 130° with dilute nitric acid (lHNO, : 3H20).Fifteen grams were taken, and a t the end of three hours thematerial was completely oxidised. The solid product was dissolvedin sodium hydroxide solution, shaken with ether to remove impuri-ties, and the acid precipitated with hydrochloric acid.It was thenesterified by Fischer and Speier’s method in the hope of separatingany 2 : 6-compound which would remain unesterified. The amountleft unchanged was, however, so small as to be scarcely weighable,and the absence of the 2 :6-compound confirmed.I n a second oxidation, the mixed acids were separated by frac-tional crystallisation of the barium salts. The acid obtained fromthe main fraction crystallised from alcohol in silky needles meltinga t 150°, which is very nearly the melting point of the 2 : 5-acid. Avery small quantity of the 2:4-acid was also separated, melting,although not sharply, a t 169O508 COHEK AND DUTT:Bromination of m-Brornotol~uene.The meta-compound was brominated in the same manner as theortho-compound.Ten grams of m-bromotoluene gave 11.5 grams,boiling at 139-142O/ 36 mm., which was practically the whole ofthe product. I n a second preparation a small quantity of residuesolidified, and was identified as 2 : 4 : 5-tribromotoluene.Sulphonntion of the Mixed Dibromotoluenes.The sulphonation was carried out as described in the foregoingsection, and the barium salts were extracted with water. Thebarium salt was separated by crystallisation into four fractions.From the first fraction after repeated crystallisation a smallportion was obtained, which, on treatment with phosphorus penta-chloride, gave a sulphonyl chloride melting a t 98-101° (3 : 4, m.p.104-105°), which on being converted into the amide melted a t 2 1 2 O .It was' evidently the 3:4-denvative, and formed only a smallproportion of the whole mixture. The last fraction was treated inthe same way, and gave a sulphonyl chloride melting a t 73-74Oand amide melting at 210-211O. This was evidently the 2 : 5-com-pound. The intermediate fractions were less pure, and were notfully examined. The crystalline appearance of the barium salts inthe last three fractions was also characteristic of the 2 : 5-compound.No indication was obtained of the presence of the 2 : 3-compound.I n order to test for the presence of any trace of the 3 : 5-compound,about 3 grams of the mixed dibromotoluenes were heated on thewater-bath for fifteen minutes with about an equal weight of weaklyfuming sulphuric acid, and then poured into water.A little whiteprecipitate settled on keeping, which was found to be insolublein boiling water, characteristic of the 3 : 5-compound. The residuewas then heated on platinum foil to test for any mineral matter,which was found to be absent, but it burned with a smoky flameand disappeared. From this we infer the presence of a trace of the3 : 5-compound in the bromination product.Nitration of the Mixed Dibromotoluenes.From10 grams of the mixed dibromotoluenes 12 grams of crude nitro-compounds were obtained, which, after several recrystallisationsfrom benzene and light petroleum, gave 3 grams of nearly puredinitro-compound of 2 : 5-dibromotoluene melting at 140-1 42O.The mother liquor on evaporation gave a solid residue of indefinitemelting point, and was therefore re-nitrated and treated in theThe nitration was effected in the same way as beforeTHE PROGRESSIVE BROMINATION OF TOLUENE. 509same way a,s above.I n this way 1 gram of the 2 : 5-compound wasobtained, and the rest of the material consisted of an oily mass,from which a solid separated on keeping, melting completely, butnot sharply, at 123O, and no definite substance could be isolatedfrom it.Oxidation. of the Mixed Dibromotoluen.es.The material was oxidised in sealed tubes a t 13O0, as previouslydescribed in the cme of the products from o-bromotoluene, and gavea product melting indefinitely from 125O to 150O. After repeatedcrystanisation, a pure acid crystallising in needles melting a t 231Owas obtained, which is that of the 3:4-compound.The motherliquors yielded an acid melting at 146-148O, which is the impure2:5-acid, melting at 153O, and formed the main bulk of theproduct. The 3:5-acid was present in too small a quantity to bedetected by this method.These results furnish evidence of the presence of 2:5- as themain product, the 3 : 4-, and- a trace of the 3 : 5-derivatives out of thefour possible isomerides, a result which agrees substantially withthat obtained on chlorination.Brominntion of p-Bromotoluene.I n brominating p-bromotoluene, 20 grams of the substance weretaken, and yielded 13 grams of dibromotoluenes boiling a t135--145O/36 mm.A small quantity of residue solidified, and aftercrystallisation melted a t 1 1 2 O , and was 2 : 4 : 5-tribromotoluene.Only two isomerides can be formed from pbromotoluene, namely,the 2:4- and 3:4-compounds, but as the melting points of t$edinitro-compounds and sulphonamides lie too near together to offera satisfwtory means of identification, recourse was had to oxidationand the formation of the dibromobenzoic acids.Oxidation of t h e Mixed Dibromotolzcenes.The oxidation was performed as previously described, and theproduct was dissolved in a solution of sodium carbonate, and shakenwith ether to remove a small quantity of an oily substance. Theacids were liberated, and converted into the barium salts and frac-tionated. I n this way two fractions were separated.From the lesssoluble portion hydrochloric acid precipitated an acid, whichcrystallised from dilute alcohol in needles, and melted a t 228O(3 : 4-acid). The mother liquor after evaporation to a small bulkwas fractionally precipitated by hydrochloric acid. The lastfraction gave an acid melting a t about 158O, which, after crystal-lisation, melted a t 166-168O (the 2 :4-acid melfs a t 169O), andformed the main bulk of the product510 COHEN AND DUTT:PART 111.-(a) Preparation of the Zsomeric TribromotoluenesI n describing the preparations in this section, the main outlinesof the processes have been indicated, but no details of experimentshave been givqn.2 : 3 : 4-Tribrornotoluei~e,and their Derivatives.The starting point f o r the preparation of this substance was3-bromo-pacetotoluidide, which by nitrating in a, mixture offuming nitric acid and glacial acetic acid a t 5O gave 3-bromo-5-nitro-pacetotoluidide (m. p.2 loo). This substance on boiling withdilute sulphuric acid (equal parts of acid and water) on a sand-bath was hydrolysed to the free base (m. p, 64-65O). The amino-group was then replaced by bromine in the following way:The base was dissolved in glacial acetic acid, the calculatedquantity of concentrated hydrochloric acid added, and diazotisedwith sodium nitrite. The, whole of the substance passed intosolution, which was added to the solution of cuprous bromide inhydrobromic acid. The mixture was then warmed on the water-bath until the evolution of nitrogen ceslsed, and steam passed infor a few minutes.The brown liquid which separated solidified,on cooling, to a crystalline mass, which was then collected, andafter one crystallisation from alcohol was pure. The yield wasalmost theoretical (m. p. 63-65O).The nitro-group was then reduced with iron and acetic acid, andthe baae separated by steam distillation (m. p. 58-59O). I n someexperiments tin and hydrochloric acid were used to reduce thenitro-group, but the yield was very poor. The base was thenconverted into the acetyl compound (m. p. 163*5--164O), which on%rominating in acetic acid gave 2 : 3 : 4-tribromo-rn-acetotoluidide.The bromination takes place slowly, and the mixture must beallowed t o stand before the bromine is completely absorbed.An attempt was made to brominate the free base, either inacetic or hydrochloric acid, but the product obtained was in eachcase 2 : 3 : 4 : 6-tetrabromo-m-toluidine (m.p. 223-224O).2 : 3 : 4-Tribromo-m-toluidine was obtained by hydrolysing theacetyl compound with boiling dilute sulphuric acid (equal volumesof concentrated acid and water). The product was poured intowater, and the solution made alkaline and extracted with ether,which leaves behind any unchanged acetyl compound. Theamino-group was then replaced by hydrogen in the followingway, as the ordinary method gives a very poor yield. The sub-stance was dissolved in ten times its weight of absolute alcohol,and a quantity of concentrated sulphuric acid (about twice thTHE PROGRESSIVE BROMINATION OF TOLUENE.511weight of the base) was added. The sulphate of the baseseparated in finely divided condition. The mixture waa then cooledto about 40°, and a quantity of amyl nitrite slightly in excma ofthe calculated amount was then added. On heating, nitrogen wasevolved, and a clear solution obtained. The product was pouredinto water, when the tribromotoluene solidified, and was collected,dried, and crystallised from glacial acetic acid (m. p. 45-46O). Theyield was theoretical.2 : 3 : 4-Tribrornouvltrotolue?ze.Ten C.C. of fuming nitric acid were placed in a small flask cooledin ice-water. About 5 grams of the tribromotoluene were thengradually dropped in, and the flask was occasionally shaken. Afterhalf-an-hour the product was poured into cold water.The solidnitro-compound was collected, washed with water, and crystallisedfrom alcohol (m. p. 107-108°) :0*1210 gave 0.1808 AgBr. Br = 63.58.CE7H,0,NBr, requires Br= 64.14 per cent.2 : 3 : 4-Tribromodinitrotoluelze was prepared by heating the sub-stance with an equal amount of fuming nitric acid and concentratedsulphuric acid on the water-bath for about an hour, after which theproduct was poured into water, collected and crystallised fromalcohol, from which it separated in prisms melting a t 197-199O.2 : 3 : 4Tribromobenaoic 4 cid.-This was prepared by oxidisingthe tribromotoluene with dilute nitric acid (1 to 4), by heating insealed tubes a t 150-160° for about six hours. The product wasmade alkaline with sodium hydroxide solution, and the unchangedtribromotoluene extracted with chloroform.The free acid wasthen liberated by adding hydrochloric acid to the clear alkalinesolution. The acid was collected, purified by boiling with animalcharcoal in alcoholic solution, and finally crystallised from benzene.It forms colourless needles melting a t 197-198O :0.1246 gave 0.1962 AgBr. Br = 67-00.C7H302Br3 requires Br = 66.85 per cent,2 : 3 : 5-Tm'bromotoluene.1. o-Acetotoluidide, on bromination in acetic acid solution, gave5-bromo-o-acetotoluidide (m. p. 156-157O). An attempt was madeto brominate it further by keeping the substances together over-night, but no action took place. This is probably a case of sterichindrance.The bromotoluidine (m. p.58-59O) was then made the startingpoint, and was prepared by hydrolysing the above acetyl compound512 CoHEN AND DUTl:It was also obtained by direct bromination of o-toluidine, but theyield was poor. The base, on bromination in acetic acid solution,gave 3 : 5-dibromo-o-toluidine (m. p. 45-46O). The amino-groupwas then replaced by bromine by Sandmeyer's reaction, and the2 : 3 : 5-dibromotoluene, thus obtained, melted a t 53-54O.2. I n a second method, the starting point was 5-nitro-o-toluidine,which was formed along with the 2:3-compound in the nitrationof o-acetotoluidide (Reverdin and CrBpieux, Ber., 1900, 33, 2498).It was brominated in acetic acid (Cohen and Raper, T., 1904, 85,1269), and the 3-bromo-5-nitro-o-toluidine, when crystallised fromglacial acetic acid, melted a t 180-181°.The amino-group wasthen replaced by bromine, and the 2 : 3-dibromo-5-nitrotoluene(m. p. 104-105°) was reduced t o 2:3-dibromo-m-toluidine (m. p.83-85O), by means of tin and hydrochloric acid. On replacingthe amino-group by bromine, 2 : 3 : 5-tribromotoluene was obtained,melting a t 53-54O.2 : 3 : 5-Tr'ribromonitrotoluene.Five grams of the tribromotoluene were treated with aboutdouble the amount of fuming nitric acid in the cold. The substancedissolved in a few minutes, and was then poured out into ice-coldwater. After repeated crystallisation from alcohol, both thepossible isomerides were separated, one melting a t 67-68O, and theother a t 88.5-90°, the former being obtained in greater quantity.The constitution of these has not yet been determined:0.0866 (m.p. 67-68O) gave 0.1302 AgBr.0-1390 (m. p. 88*5-90°) gave 0.2090 AgBr.C,H40,NBr3 requires Br = 64.14 per cent.2 : 3 : 5-Tribromodinitrotoluene.-The tribromotoluene was heatedon the water-bath with fuming nitric acid and concentrated sul-phuric acid for an hour, and afterwards poured into water. Oncrystallising several times from glacial acetic acid, the nitro-com-pound was obtained quite pure, and melted a t 209-210°:Br=63*96.Br=63*98.0-2156 gave 0'2904 AgBr. Br=57*31.2 : 3 : 5-Tribromobenzoic acid was prepared by two methods :1. By oxidising the tribromotoluene with dilute nitric acid (1 to3) in sealed tubes a t 135-140° for six hours. The contents of thetube were then dissolved in sodium hydroxide solution, the un-changed tribromotoluene was removed by steam distillation, andthe free acid liberated by means of hydrochloric acid. The acidobtained in this way was coloured slightly yellow, and was de-C,H304N,Br3 requires Br = 57-28 per centTHE PROGRESSIVE BROM LNATION OF TOLUENE. 513colorised by digesting the alcoholic solution with animal charcoal.It crystallised in colourless needles melting a t 193-194O.2.The starting point was anthranilic acid, which, on brominatingin acetic acid solution, gave 3 : 5-dibromo-2-sminobenzoic acid(m. p. 225O). The amino-group was then replaced by bromine(Rosanoff and Prager, J. Amer. Chem. SOC., 1908, 30, 1902).These authors obtained it as yellow needles melting a t 193*5O, buton purifying it in the same way as described above, we obtained itin colourless needles melting a t 193-194O.2 : 3 : 6-Tribromotoluene.1.2-Bromo-5-acetotoluidide was nitrated in the cold ( 5 O ) withfuming nitric acid and glacial acetic acid. Two nitro-compoundswere formed in this way as pointed out by Nevile and Winther(Ber., 1880, 13, 971), who used concentrated sulphuric acid insteadof acetic acid. The crude nitration product, after crystallising onc0from alcohol, melted a t 115-124O, and after repeated crystallisationfrom alcohol, melted sharply a t 125-125-5O. It is 2-bromo-$-nitro-5-acetotoluidide, and forms the main product in the reaction.Another nitro-compound remaining in the mother liquor was pre-cipitated on adding water, and melted indefiniteIy from 90° to l l O o ,and after removing the acetyl group and crystallising several timesfrom alcohol, the product melted at 102-103O.The constitutionof this base was determined in the following way : the amino-groupwas replaced by bromine, and the dibromonitro-compound thusobtained (yellow crystals, m. p. 78-80°) was reduced by meansof tin and hydrochloric acid. The base separated from alcohol incolourless crystals melting a t 5 6-5 7O. Finally the amino-groupwas replaced by bromine, and the tribromotoluene thus obtainedmelted a t 55-57O, and is, therefore, the 2 : 3 : 6-compound, whichmelts at 58-59O. The original substance is, therefore, 6-bromo-2-nitr o-m-toluidine.The 2 : 4 : 5-compound mentioned above was then hydrolysed witiidilute sulphuric acid (equal vols.) on the water-bath, and the baseprecipitated by adding water.On crystallisation from alcohol, thelatter was obtained in reddish-brown needles melting a t 179-181O.It was then brominated in acetic acid solution on the water-bath.The 2 : 6-dibromo-4-nitro-m-toluidine formed in the reaction wascrystallised from glacial acetic acid and alcohol (yellow needles,m. p. 130-133O). The yield was theoretical. Nevile and Win-tlier's method of brominating in sulphuric acid solution with anaqueous solution of bromine was also tried, but it was not sosatisfactory as the one described above514 COHEN AND DUTT:It was observed in this connexion that the crystals first separatingfrom the hot solution were bulky needles, which filled the liquid,but on keeping overnight the mass was transformed into prisms,which deposited a t the botom of the vessel.ThO amino-group wasthen replaced by bromine, and the 2 : 3 : 6-tribromo-4-nitrotoluenecrystallised from acetic acid (m. p. 106.5-107°). The nitregroupwas reduced by means of iron and acetic acid, and the base crystal-lised from alcohol (m. p. 118-119°). The amino-group was thenremoved by diazotising in alcoholic solution, and the tribromo-toluene crystallised from alcohol in colourless needles melting a t2. I n the following method the starting point was 6-bromo-o-toluidine (see above), and gave a better yield. It w a converted intothe acetyl compound, which crystallised from alcohol in colourlessneedles melting at 159*5-161°.One atom of bromine was thenintroduced into it, and the product crystallised from alcohol incolourless needles melting a t 165-5-166.5O. The substance was5 : 6-dibromoaceto-o-toluidide. I& constitution was determined byremoving the amino-group and distilling the product underdiminished pressure. It crystallised on cooling, melted a t 29-30°,and gave a dinitro-compound melting at 106-107°, and is therefore2 : 3-dibromotoluene. As the yields throughout are quite satisfac-tory, this furnishes a useful method for preparing 2 : 3-dibromo-toluene.The acetyl group was removed by boiling with concentratedhydrochloric acid, and the base melted a t 58O. On replacing theamino-group by bromine the tribromotoluene crystallised fromalcohol in needles melting at 58-59O.2 : 3 : 6-Tm'bromonitrotoluene was obtained in the same way asthe .2 : 3 : 5-com~ound, and crystallised from alcohol in light, straw-coloured needles, melting at 91-92O.2 : 3 : 6-Tribromodinitrotoluene was prepared in the same way asthe 2 : 3 : 5-compound, and crystallised from alcohol and acetic acidin light yellow prisms, melting a t 202-203O.2 : 3 : 6-Tribromobenzoic acid could not be prepared by directoxidation.The tribromotoluene was heated in sealed tubes withdilute nitric acid for several days together, first a t 140--145O, andthen at 150-160°, but not a trace of acid was formed. On raisingtho temperature above 170° the substance ww completely destroyed.Different strengths of acid were also used, and the heating wascontinued for about two weeks a t a time, but without any result.Other oxidising agents, for example, chromic acid and potassiumpermanganate, were also tried ineffectually. It is curious that the2 : 4 : 6-tribromo-compound, which is also a diortho-substituted deriv-5 8-5 goTHE PROGRESSIVE BROMINATlON OF TOLUENE.515ative, is easily attacked by nitric acid, whilst the 2 : 3 : 6-compoundis not.2 : 4 : 5-Tribromotoluene.1. From 3-Bromo-p-toZuidine.-The substance was nitrated withconcentrated nitric acid in the presence of concentrated sulphuricacid in the cold, that is, by adding the dry nitrate of the base to-cold concentrated sulphuric acid (Cohen and Dakin, T., 1902, 81,1334).The 5-bromc~2-nitro-ptoluidine crystallised from alcohol inbrown needles, and melted sharply a t 118y1190. The amino-groupwas replaced by bromine, and the product crystallised from alcohol(m. p. 83.5-84.5O). The nitro-group was then reduced by meansof tin and hydrochloric acid, and the base crystallised from alcoholin colourless needles melting at 96-97O. On replacing the amino-group by bromine, 2 : 4 : 5-tribromotoluene (m. p. 112-113O) wasobtained in colourless needles from alcohol.2. Prom 2-Bromo-5-acetotoluididJe, m. p. 101-102°.-Thesubstance was brominated in acetic acid solution, and the 2 : 4di-bromo-5-acetotoluidide (m. p. 168-169O) that was formed wasprecipitated by the addition of water. The acetyl group wasremoved by boiling with concentrated hydrochloric acid, and theamino-group replaced by bromine.The tribromotoluene crystallisedfrom alcohol in colourless needles melting at 112-113O.2 : 4 : 5-Tribromonitrotoluene was prepared in the same way asthe 2 : 3 : 5-derivative. Only one mononitro-compound could beisolated, and this separated from alcohol and acetic acid in colour-less crystals melting at 130-131*5°.2 : 4 : 5-Tribromo&nitrotoZuelze.-This separates from alcohol andglacial acetic acid in colourless crystals melting at 278-279O.2 : 4 : 5-Tribromobenzoic acid was obtained by oxidising thetribromotduene with dilute nitric acid in a sealed tube; it crystal-lises from benzene in colourless needles melting a t 195-196O:0-1278 gave 0.2020 AgBr. Br = 67-24.C,H,O,Br, requires Br = 66.85 per cent.2 : 4 : 6-TribromotoZuene.m-Toluidine was brominated in acetic acid solution with thequantity of bromine calculated to fom tribromotoluidine. Theproduct when crystallised from alcohol melted at 1OO-l0lo.Theamino-group was then removed by diazotising in alcoholic solution.The tribromotduene crystallised from alcohol in colourless needlesmelting at 65-66O.2 : 4 : 6-TribromonitrotoZuene forms colourless crystals melting at74-75-50 :VOL. cv. M 516 COHEN AND DUTT:0.1360 gave 0.2044 AgBr.2 : 4 : 6-Tribromodinitrotoluene separatw from alcohol in colour-less crystals melting a t 217-218O.2 :4 : 6-T?ibromobemzoic acid formed colourless crystals fromalcohol, melting a t 186-187O. This is identical with the acid(m.p. 186G) obtained by Rosanoff and Prager (Loc. cit.) fromtribromoaniline.3 : 4 : 5-Tribromotoluene.From 3-Bromo-p-toZuidine.-This substance on bromination gave3 : 5-dibromo-ptoluidine (m. p. 74-75O), which, on replacing theamino-group by bromine, gave 3 : 4 : 5-tribromotoluene (colourlessneedles from alcohol, m. p. 88-89O). It was subsequently foundthat ptduidine itself could be directly brominated to form thedibromo-compound with equally satisfactory results.3 : 4 : 5-Tribromonitrotoluene separates from alcohol in colourlesscrystals melting a t 104--105O.3 : 4 : 5-Tribromodinitrotoluene forms colourless priems, meltinga t 211-5O.3 : 4 : 5-Tm'bromobenzoic acid crystallises from benzene in colour-less needles melting at 235O.Br = 63.94.C,H,0,NBr3 requires Br = 64.14 per cent.PART 11.-(b).The Brominatiolt of the Six IsomericDi b ro m o t oluelt e s.Bromdnation of 2 : 3-Dibromotoluene.-Ten grams of the dibromo-toluene were dissolved in an equal volume of carbon tetrachloride,*about 0.5 gram of f reshly-prepared aluminium-mercury couple wasadded, and 2.2 C.C. of bromine dissolved in 5 C.C. of the same solventwere then gradually dropped in. Reaction took place very readily,and was prevented from becoming too violent by carefully coolingthe vessel. I n the end a voluminous dark mass was obtained.Water was added to the product, and the mass extracted withcarbon tetrachloride, shaken with sodium hydroxide solution, andthen separated. The carbon tetrachloride was then distilled offon the water-bath, the last traces being removed by aspirating acurrent of air through the hot residue, which was afterwards dis-tilled under diminished pressure.The yield was 9.2 grams boilinga t 190-200°/60 mm.After one crystallisation from alcohol, a major fraction (6 grams)was obtained, melting a t 56-58O (A). The mother liquor, onevaporation, gave a product melting indefinitely from 40° to* In one exporiment no solvent was used and tho product consisted mostly oftetrabromotoluenes together with a little of the tribromo-compoundsTHE PROQRESSIVE BROMINATION OF TOLUENE. 5174 6 O ( B ) . The product ( A ) was convered into the dinitro-com-pound which, after crystallising twice from alcohol, melted sharplya t 202-203O, and is, therefore, the 2 : 3 : 6-compound.The fraction (B) was also converted into the dinitro-compound,from which a product was obtained which melted a t 206-208';the 2 : 3 : 5-tribromodinitrotoluene melts a t 209-210°.The pres-ence of the 2 : 3 : 5-compound was further confirmed by the mys-talline appearance of the dinitro-compound ; the dinitro-derivativeof the 2 : 3 : 5-compound crystallises in plates, whilst that of the2 : 3 : 6-compound crystallises in prisms.From the mother liquors of the nitro-compounds of both ( A ) and(B), me have not yet been able to isolate the dinitro-compound of2 : 3 : 4-tribromotoluene (m. p. 197-199O), which, if formed a t all,seems very difficult to detect in this way.The product from a second bromination was heated in a sealedtube with dilute nitric acid in order to prepare the benzoic acids,but without any results.This is a further confirmation of thepresence of the 2 : 3 : 6-compound as the main product, for, as pre-viously stated, this substance is not attacked by nitric acid.It will thus be seen that the result of bromination of 2:3-di-bromotoluene is quite different from that of chlorination of di-chlorotoluene. In the latter case the 2 : 3 :4-compound is the soleproduct, whereas in the former, it is the 2 : 3 : 6-derivative thatforms the main bulk of the proETuct, together with a little of the2 : 3 : 5-derivative, whilst the presence of the 2 : 3 : 4-compound isdoubtful.Bromination of 2 : 4-Dibromotoluene.I n this case, the bromination was effected without the use of asolvent.* The product was dissolved in chloroform, shaken withsodium hydroxide solution, and separated.The chloroform wasthen distilled off, and the residue crystallised from alcohol. I nthis way, 8 grams of a substance were obtained which melted a t109--ill*; the weight of material originally taken being 10 grams.The mother liquor, on evaporation, gave 3.5 grams of a voluminous,pasty mass, which consisted of the unchanged dibromotoluene,mixed with some of the tribromo-compounds.The fraction melting a t 109-11lo was converted into the dinitrGcompound, which, after crystallising from alcohol and acetic acid,melted a t 278-279O ; the 2 : 4 : 5-tribromodinitrotoluene melts atThe second fraction wm similarly converted into the dinitro-compound, from which one fraction melting a t 276-278O (2 : 4 : 5)278-279'.* I n a second experiment, carbon tetrachloride was nsed as solvent and gave thesame result.M M 518 COHEN AND DU":was separated; another fraction melted at 212-215O (probably the2 : 4 : 6-tribro8modinitrotoluene, which melts at 217-218O) ; and twoother portions melted a t 189-193O and 195-205O, but the quanti-ties of these last two fractions were so small that it was hopelessto detect in this way any trace of the 2:3:4-derivative, which isthe third possible isomeride.The product of bromination of the dibromotoluene was thenoxidised by heating in a sealed tube with dilute nitric acid (1 to 3)a t 160° for two days, and the benzoic acids were separated in theusual way.The mixed acids were aterified by Fischer and Speier'smethod, when a small quantity was left unesterified, and was identi-fied as the 2 : 4 : 6-acid. The esterified portion was then hydrolysedwith alcoholic potassium hydroxide. On liberating the free acidand crystallising it from alcohol, it melted a t 192-194O. The2 : 4 : 5-acid melts at 195-196O.Thus the results of bromination of 2 : 4-dibromotoluene agreesubstantially with those obtained in the case of chlorination.Bromination of 2 : 5-Dibromotoluene.Ten grams of material were brominated, using carbon tetmchloride as solvent. In a second experiment the solvent wasomitted, but the product consisted of a mixture of tri-, tetra-, andpent a-b r omot oluenes .The bromination product was treated in the same way as thatfrom 2:3-dibromckoluene.The yield was 11 grams, the substancemelting indefinitely a t 70-looo and boiling at 155--165O/20 mm.A portion was converted into the dinitrecompound, and wascrystallised several times from alcohol and acetic acid. The mainbulk of the product was a substance melting a t 276-278O, identicalwith the dinitro-compound of 2 : 4 : 5-tribromotoluene. Afterrepeated crystallisations another product was obtained, whichmelted a t 200-202° and crystallised in prisms characteristic ofthe 2 : 3 : 6-derivative. No evidence was obtained of the presenceof the third possible isomeride, the 2 : 3 : 5-compound.The oxidation of the bromination product has also been carriedout, but no definite evidence was obtained of the presence of the2:3:5-acid, as it has about the same melting point as the2 : 4 : 5-acid.It will thus be seen that these results agree with those obtainedon chlorination in so far as the same two isomerides are formed,but the proportion is reversed.I n the case of chlorination the2:3:6-compound is the main product, whilst in the present caseit is the 2 : 4 : 5-compoundTHE PKOGRESSIVE BROMINATION OF TOLUENE. 519Bromination of 2 : 6-Dibromotolzterce.The bromination wm effected in the same way as in the case of2 : 4-dibromotoluene. The product was crystallised from alcohol,and melted sharply at 59O. It was then converted into the dinitro-derivative in the usual way. After several crystallisations fromalcohol it melted at 202-203O, which corresponds with the2 : 3 : 6-compound.The mother liquor left after crystallising the brominationproduct was evaporated, and from it a product was obtained whichwas a mixture of the unchanged dibromotoluene and a little ofthe 2 : 3 : 6-compound.2 : 6-Dibromotoluene can only give two tribromo-derivatives, the2:3:6- and 2:4:6-compounds, and of these only one, the former,is produced in ths above reaction.No indication of the presence ofthe 2 : 4 : 6-tribromodinitro-compound (m. p. 217-218O) could bedetected on nitration. The remit is thus in entire agreement withthat of the chlorination of 2 : 6-dichlorotoluene.Brominatiorc of 3 : 4Dibromotoluene.This was brominated exactly in the same way as the previoussubstance. The product, after one crystallisation from alcohol,gave 8 grams of a substance (m.p. 110-112°) from 10 grams ofmaterial originally taken. It was converted into the dinitro-com-pound, which after one crystallisation melted sharply at 278-279O,and was identical with the 2 : 4 : 5-tribromodinitrotoluene. No otherdinitrecompound could be detected in this fraction.The mother liquor on evaporation yielded 3 grams of a viscid oil,which was probably a mixture of the original unchanged substanceand a little of the tribromo-compound. This was also convertedinto the dinitrecompound, from which only one substance (m. p.207-209O) could be isolated after repeated crystallisation. It isthe dinitro-compound of 3 : 4 : 5-tribromotoluene (m.p. 211.5O).This was further confirmed by its crystalline appearance, whichis that of truncated prisms. No trace of the 2:3:4-tribromo-dinitrotoluene (m. p. 197-199O) could be detected in the aboveproduct.This result agrees substantially with that of chlorination, exceptfor the presence of a minute amount of the 3 : 4 : 5-derivative520 THE PROGRESSIVE BROMl NATION OF TOLUENE.Bromination of 3 : 5-Dibromotoluene.This was brominated in the manner already described. Theproduct, on crystallisation from alcohol, melted at 51-53O. Alittle of it was converted into the dinitro-compound, which afterseveral crystallisations from alcohol and acetic acid melted at209--210°, and is identical with the 2 : 3 : 5-tribromodinitro-com-pound. The mother liquor on evaporation gave a mixture of theunchanged dibromotoluene and 2 : 3 : 5-tribromotoluene. The othertribromobluene which is theoretically possible, namely, the 3 : 4 : 5(m. p. 88-89O>, could not be detected by repeated crystallisationof either the bromination product or its dinitro-derivative. Theresult thus agrees entirely with that of chlorination.I n one experiment carbon tetrachloride was used as solvent, but,curiously enough, the product consisted of a mixture of tri- andtetra-bromotoluenes, a material which melted a t a much highertemperature than any of the tribromotoluenes.I n the following table the results of bromination and chlorina-tion are given side by side, the principal product being placed firstin each case, and a trace indicated by brackets.Brominationproducts.Ortho-, para-,Toluene. (meta-).Ortho- 2 : 5 ; 2:4Para- 2 : 4 ; 3:4Meta- 2 : 5 ; 3 : 4 ; (3:5)2: 3 2 : 3 : 6 : 2:3:5Mono-halogencompounds(2:4 2:4:53 (2:4:6)2: 5 3 : 4 : 5 ; (2:3:6)2: 6 2 : 3 : 63: 4 2 : 4 : 5 ; (3:4:5)2 : 3: 5halogencompoundsDi- I 3 : 5Chlorinationproducts.Ortho-, para-.2 : 4 ; 2 : 3 ; 2 : 6 ; (2:5)2:4; 3 : 42:5.; 3:42: 3: 42 : 4 : 5 ; 2 : 3 : 4 ; (2:4:6)2:3:6; 2:4:52:3:62 : 4 : 52: 3: 5Attention is directed to the difference in the orienting effects ofchlorine and bromine produced by chlorination and brominationof ortho-halogen derivatives, an effect which is most marked in thecase of the mono-halogen compound and the 2 : 3- and 2 : 5-dihalogenderivatives.It would appear from this that the orienting effect of the methylgroup is more emphasised in the case of chlorine substitution andof the halogen in the case of bromine substitution. In the highersubstituted derivatives the tendency is to form compounds havingthe symmetrical structure,HARTLEY U- AND B-TRIMETHYL COBALTICYANIDE. 521as was pointed out by Cohen and Dakin (T., 1904, 85, 1274) andCohen and Hartley (T., 1905, 87, 1360).In conclusion, we wish to express our thanks to the ResearchFund Committee of the Chemical Society for a grant towards theexpenses of this investigation.THE UNIVERSITY, LEEDS