THE BEACTION BBTWEEN BENZYLAMINE, ETC. 2879CCLXI X. - The Reaction Betweeyx Benzylarnine and theDibromosuccinic Acids.By EDWARD PERCY FRANKLAND.IT has been shown by the author (T., 1911, 99, 1775) that mzpso-dibromosuccinic acid -Ic and benzylamine react to form a dibenzyl-aminosuccinic acid and, simultaneously, the dibenzylamidel ofi-tartaric acid. Analogous results have now been obtained withr-dibromosuccinic acid,* which yields an isomeric dibenzylamino-succinic acid, together with a small quantity of the dibenzylamideof racemic acid. A comparison of the properhies of the twodibenzylamino-acids suggests that the one obtained from 7neso-dibromosuccinic acid has a structure corresponding with that ofi-tartaric acid, and resembles the diaminosuccinic acid preparedby Lehrfeld from meso-dibro~nosuccinic acid, which Tafel has shownto yield i-tartaric acid when treated with nitrous acid.On theassumption that the configurakions of the groups about the twoasymmetric carbon atoms are similarly affected by the reagent(nitrous acid), the configuration of the diamino-acid will be tliesame as that of the tartaric acid derived from it. Consequently,we may be justified in calling Lehrfeld's diamino-acid meso-diaminosuccinic acid (I), and the more soluble diamino-acidobtained by Tafel, which yields racemic acid with nitrous acid, isr-diaminclsuccinic acid (11) :CO,H*$JH*NH R CO,H*FH*NHR NHR*CO*yH*OHCO,H*CH*NHR R N H* CH*CO,H OH*CH*CO*NHR(1.) (IT.) (IIr.)The new dibenzylaminosuccinic acid from r-dibromosuccinic acid,on account of its greater solubility, has most probably a structureanalogous to that of the r-diarninosuccinic acid [(11), whereR=C,H7]. It has not been possible to prove this directly, sincethe benzylamino-acids cannot be converted into the amino-acids orinto tartaric acids by any simple means.The simultaneous appear-ance of the benzylamide of racemic acid (111) cannot be taken asevidence of a racemic configuration for the dibenzylamineacid,since the former substance is pro'duced by an entirely differentreaction (first discovered by Lutz in the case of monobromosuccinicacid), and is not generated from bromofumaric acid, t.he initialstage in the formation of the dibenzylamino-acid.The smallness of the yield of racemobenzylamide (one-twentieth* It is coilsidered that these designations are more suitable than those usedIceviously, namely, s-dibromosucciuic acid and isodibromosnccini~ acid I espectively,9 B 2880 FRANKLAND : THE REACTION BETWEENt o one-thirtieth of tlie weight of the isomeric dibenzylamino-acid)is probably due to the extreme readiness with which r-dibromo-succinic acid parts with hydrogen bromide to form bromofumaricacid.The action of benzylamine on r-dibromosuccinic acid in wateror chloroform solution appears to, lead, in the first instance, t o tlieformation of a mixture of mono- and di-benzylamine salts ofbromofumaric acid ; the mono-salt being sparingly soluble in water,i t may separate out when the mixture is acidified with hydrochloricacid.This salt yields broniofumaric acid and benzylamine hydro-chloride when treated with ethereal hydrochloric acid, and whenheated with aqueous benzylamine passes into the (r)-dibmzylamino-acid. * This reaction may proceed along two different lines: eitherthe benzylamine may combine a t the double bond t o form it bromo-benzylamino-acid, and the bromine is then displaced by the benzyl-amino-group, or the bromine is displaced first, yielding a benzyl-C0,H-S*NH*C7 liH*C*CO,Haminofumaric acid, ’, which then attaches benzyl-aniine t o the double bond.I n the previous communication on meso-dibromosuccinic acid i twas suggested t h a t the corresponding formation of bromomaleicacid was followed by the addition of benzylamine a t the doublebond t o yield a bromobenzylamino-acid (benzylamine salt) (IV).Further investigations on a product isolated from the reactionmixture, and having this compoaition, C,8H,,0,N2Br, render i t mostprobable t h a t the substance in question is the isomeric dibeiizyl-amine salt of bromomaleic acid (V):$!H Er* C0,H,C7H7*NH,CH(NH*C,FI 7)*C0,H(IV.)Si*E* C0,H,C7H, *NH,HC* C0,H,C,H7*N H,(V.)Thus, when treated with ethereal hydrochloric acid, this sub-stance yielded bromomaleic acid and benzylamine hydrochloride(2 mols.), and i t appeared to be identical with a product obtainedby allowing an alcoholic solution of bromomaleic acid and benzyl-amine (2 mols.) t o crystallise in the cold.These facts render itimprobable that we are here dealing with a bromobenzylamino-acid.Investigations, as yet unpublished, by the author in conjunc-tion with H.Webb and C. Mowinclcel on the action of ammonia* The tcrnis ( T - ) and (meso-) applied to distinguish the two isomeric dibeuzyl-amino-acids are bascd on the apparent analogy between these acids and the r- andmeso-diaminosuccinic acids respectively. Incidentally it may be noted that thedibenzylamino-acids (r-) and (meso-) arc derived from the corresponding r- and meso-dibroniosiiccinic acids, although,lof course, this ielationship affords no direct indicationof the assumed configurations of the dibe?izylamiiio-acidsBENZYLAMlNE AND THE DIBRUMOSUCCINIC ACIDS. 2881on the two dibromosuccinic acids have not resulted in the isolationof bromoaminosuccinic acids,* only ammonium salts of bromo-maleic and broniofumaric acids being obtained.It has been found that when the dibenzylamine salt of bromo-maleic acid is heated in aqueous solution i t undergoes a reaction,yielding some (meso-)dibenzylaminosuccinic 'f- acid, together with amore soluble substance.When heated in alcoholic solution, thedibenzylmino-acid is accompanied by the monobenzylamine salt ofbromomaleic acid. Circumstances having compelled the author tointerrupt these and other investigations proceeding at the Uni-versity of Birmingham, Edgbmton, i t is hoped that in time afurther communication may be made in which some light will bethrown on the mechanism of the reaction whereby the bromo-unsaturated acid passes into the saturated dibenzylamino-acid.Reaction products other than the above have been noted in caseswhere r-dibromosuccinic acid has been treated in chloroform solu-tion ; these also await further investigation.EXPERI M E N T A L .Eight grams of r-dibromosuccinic acid (prepared by McKenzie'smethold) were treated with 12.4 grams of benzylamine (4 mols.) in25 C.C.of water, and the mixture was heated to looo for threehours. An orange-yellow precipitate appeared, which was collectedfrom the cooled mixture and washed with water. The filtrate, onbeing heated and kept for some time, yielded a further smallquantity of precipitate' (colourless).These precipitates contain (r-)dibenzylaminosuccinic acid, prob-ably together with some benzylamine salts of the same, also alittle dibenzylamide of racemic acid and brown, tarry matter.(r-) Dib enz ylaminosuccinic Acid.The precipitates were heated on the steam-bath with diluteaqueous ammonia; a tarry, semi-crystalline residue was separated,and the solu€ion concentrated to a small bulk.The r-dibenzyl-aminosuccinic acid crystallised in flakes and crusts on the surfaceof the liquid, the yield amounting t o 3.9 grams, including a smallamount extracted subsequently from the tarry residue. This pro-duct was purified by several recrystallisations from dilute ammonia,from which solvent it separated in small, acicular prisms, meltingand decomposing, on quick heating, a t 25OO.z Pure (meso-)dibenzyI-* Claus (Ber., 1832, 15, 1849) claims to havc isolated the silver salt of a bromo-aminosucciuic acid derived from the action of ammonia on meso-dibromosuccinic acid.t See foregoing note on the (r-)dibenzylamino-acid.$ All melting points uncorrected2882 FRANKLAND : THE REACTION BETWEENaminosuccinic acid, when similarly heated, decomposed a t 260°,and mixtures of the two acids a t 244-247O.(r-)Dibenzylamino-succinic acid is only sparingly soluble in hot water (although con-siderably more readily soluble than the meso-acid), and practicallyinsoluble in alcohol. It dissolves in ammonia o'n warming and inmoderately concentrahd hydrochloric acid. Unlike the meso-acid, it is not readily precipitated by diluting the acid 60111-tion, but separates out on the addition of sodium hydroxideor sodium carbonate solution.The ?--acid undergoes considerabledecomposition on long heating with potassium hydroxide solution,but is little affecbed by heating in sealed tubes with concentratedaqueous ammonia :0.2094 gave 0.0799 gave 97.66 C.C. (20, and 5.51 C.C. Nz.*14.32 C.C. N,. C=65*53; N=8*63, 8.56.CI8H2,O4N, requires C = 65-86 ; N = 8-54 per cent.Bibenzylamide of Racemic Acid.The tarry residue separated from the ammoniacal solution of thereaction product (see above) was extracted again with ammonia,and then warmed with a little methyl alcohol. The tar dissolved,and a crystalline precipitate remained. This was extracted againwith hot ammonia, and the residue weighed 0.18 gram and meltedat 207-210°. A further quantity of the same substancej0.05 gram) was obtained from alcoholic washings.Theaminoniacal extracts yielded some (r-)dibenzylaminosuccinic acid(see above).The dibenzylamide of racemic acid was purified by recrystallisa-tion from alcohol, from which solvent it separates in glistening,oblong, rectangular plates, so>me of which have truncated angles,forming pentagons and hexagons. It melts at 208-210° :0.0756 gave 92.73 C.C. CO, and 5-28 C.C. N,. C=65*70; N=8*74.C,,H,,04N2 requires C = 65-86 ; N = 8-54 per cent.The dibenzylamide of racemic acid was synthesised by warmingdimethyl racemate with benzylamine in methyl-alcoholic solution.The substance crystallised out on the addition of a few crystals ofthe product obtained from r-dibro,rnosuccinic acid. After recrystal-lisation from abso1ut.e ethyl alcohol, the substance presented thesame appearance under the microscope as the above-described pro-duct, and melted over the same range of temperature (208-210°),as did also a mixture of the two substances.(Found: N=8*63.Calc., N=8.54 per cent.)The yield of the dibenzylamide from ?-dibromosuccinic acid is++ Dry GO, a i d N, st N.T.P. ; C and N combustion in a vacuuInBENZYLAMINE ANI) THE DIBROMOSUCCINIC ACIDS. 2883smaller than that of the corresponding substance from ?neso-dibromosuccinic acid; 5 grams of the latter subst.ance, when heatedwith aqueous benzylamine (2 mols.), gave 0.6 gram, and with4 mo1s., 0.7 gram of the dibenzylamide of i-tartaric acid, a purifiedspecimen of which crystallised in rhomboidal plates, accompaniedby a few hexagons, and melted a t 203O.A mixture of raceniobenzylamide from r-dibromosuccinic acidwith synthetic i-tartarobenzylamide melted at 194O.,4 ctioit of Bemylamitie oi& r-Dib?*omosuccitbic d c i d in Noti-uqueous Solvents.The principal product of the reaction was apparently thedibenzylamine salt of bromofumaric acid.The amount ofdi benzylamino-acid was again very small.( a ) ChZoroform.-Five grams of freshly prepared r-dibromo-succinic acid (Fo'und, Br =57*11, 57-22. Calc., Br = 57-97 percent.) were h a t e d with a solution of 8 grams of benzylamine in60 C.C. of chloroform, and heated to boiling under reflux for twohours.The precipitate (7.1 grams) contained bemylamine hydro-bromide, together with.less soluble substances. These wereseparated by treating with water, and eventually by extractingany residual benzylamine hydrobromide with absolute alcohol.The less soluble products (2.8 grams) consisted of benzylamine saltsof bromof umaric and (r-)dibenzylaminosuccinic acids, together witha little dibenzylamide' of racemic acid.The chloroform mother liquor and washings were evaporated,and the rmidual tarry residue heated with water. A small amountof (r-)dibenzylaminosuccinic acid (0.66 gram, decomposing a t236-238O) was separated, and the solution contained a littlebenzylamine hydrobromide, together with about 0.3 gram of aproduct melting and decomposing a t 205-225O.( b ) Alcohol.-Five grams of r-dibromosuccinic acid were dis-solved in 20 c .~ . of warm absolute ethyl alcohol, and treated with8 grams of benzylamine. After keeping for about three hours, acrystalline precipitate began to form in the liquid. About sixteenhours later this precipitate was collected, washed with alcohol andwith ether, and dried. It weighed 2.6 grams, and melted anddecomposed a t 172O. An aqueous solution of this substance de-colorised cold alkaline perinanganate almost instantly, showing thepresence of bromofumaric acid. The aqueous solution was slightlycloudy, even on heating, perhaps owing to the presence of a littledibenzylamino-acid.On acidifying with dilute hydrochloric acid, a crystalline pre2884 FRANKLAND : THE REACTION BETWEENcipitate appeared, which was collected and recrystallised fromwater.The substance melted and decomposed a t 219O, and provedto be the monobenzylamine salt of bromofumaric acid (see below).The alcoholic mother liquor and washings on concentrationyielded about 1.3 grams of a substance melting at 205-211°,followed by a little (r-)dibenzylaminosuccinic acid (decomposinga t 248O).I n another experiment, 8 grams of the r-dibromo-acid weretreated with 16 grams of benzylamine (more than 5 mols.) in120 C.C. of alcohol, and the solution was heated t o boiling underreflux for four hours. On concentrating the solution, a crystallineprecipitate separated, which was collected and extracted with hotwater. The residue, which was soluble in boiling water, decom-posed a t 2 3 1 O . This substance is probably the nio?zobenzyZami?iesalt of (r-)dibenaylaminoszLccinic acid, but cannot readily be purifiedfrom possible admixtures of the free acid and its dibenzylaminesalt :0.1429 gave 10.98 C.C. N,.N:=9.61.C,8El&04N,,@,H,N req=xires N = 9.65 per cent.Action of Bemylamine o n Brornofurnnric Acid.The following experiments were carried out with a view toinvestigate more fully the intermediate products of the reactionbetween benzylamine and r-dibromosuccinic acid. The reaction isin this case simplified by the' non-appearance of the dibenzylamideof racemic acid, and by taking a smaller proportion of amine t othe acid, the' amount of dibenzylamino-acid produced becamealmost negligible.A chloroform solution (50 c.c.) of 4.4 grams of bromofumaricacid was treated with 6 grams of benzylamine (2g mols.) andheated to boiling under reflux for three hours.The resultingprecipitate weighed 9.3 grams, and a portion dissolved in watercontained only a trace of ionic bromine. Theory requires 9-2 gramsof the dibenzylamine salt of bromofumaric acid. When the pro-duct was dissolved in water, only a very small precipitate re-mained, but when a large proportion of amine was used (3 mols.)this precipitate appeared in greater quantity, and proved to be(r-)dibenzylaminosuccinic acid (decomposing a t 242O).The aqueous solution was treated with 3.7 grams of concentratedhydrochloric acid, whereupon a crystalline precipitate was de-posited, which was collected, washed with water, alcohol, and ether.The mother liquor was shown to contain benzylamine hydro-chloride. The substance, after recrystallisation from water, de-composed at 221°, and appeared on analysis to be the monobenzylBENZYLAMINE AND THE DIBROMOSUCCINIC ACIDS.2885amiue salt of bromofumayic acid. The yield was 5.75 grams, whilsttheory requires 6.8 grams. The substance is rather sparinglysoluble in hot water, and almost insoluble in cold water:0.1238 gave 96.84 C.C. CO, and 4.69 C.C. N,. C=43*93; N=4*74.C4H30,Br,C,H,N requires C =43-71; N=4-64 per cent.On the results of this analysis, the substance might equally wellbe the isomeric bromobenzylaminosuccinic acid, C,,H,,04NBr ;however, the aqueous solution rapidly decolorises cold alkaline per-manganate, and when treated with ethereal hydrochloric acid thesubstance yields benzylamine hydrochloride and bromof umaricacid.Two grams of the substance were shaken with dry ethereal hydro-chloric acid.The precipitate weighed 1 gram, and melted a t 258*,whilst theory requires 0.95 gram of benzylamine hydrochloride.The ethereal mother liquor yielded some bromofumaric acid. Bylong-continued heating with aqueous benzylamine, the monobe,nzyl-amine salt of bromofumaric acid passes into (r-)dibenzylamino-succinic acid.Thus 1.3 grams of the above-described product were heated a tlooo with 1.3 grams of benzylamine in 10 C.C. of water for fourhours. The substance gradually dissolved, and, on cooling, a pre-cipitate of small crystals appeared, which weighed 0.25 gram anddecomposed at about 233O (compare second experiment withalcoholic benzylamine on r-dibromosuccinic acid).After recrystal-lisation from dilute ammonia, the substance decomposed a t 242O[ (r-)dibenzylaminosuccinic acid].Experiments with meso-Bibromosuccinic Acid.These experiments were undertaken with the object of investi-gating the structure of the compound melting a t 156O described ina previous communication (T., 1911, 99, 1779). It was also thoughtpossible that an intermediate compoand might be found betweenthis substance and the final product, (meso-)d'ibenzylaminosuccinicacid.Seven grams of meso-dibromosuccinic acid were treated with11 grams of benzyiamine (4 mols.) in 100 C.C. of absolute ethylalcohol, and the mixture was heated to boiling for five minutesand then kept overnight'.The resulting precipitate weighed8.6 grams. After extraction with boiling alcohol 0.6 gram remained(crude benzylamine salt of (meso-)dibenzylaminosuccinio acid). Thealcoholic solution on cooling deposited 4.1 grams of the substancemelting a t 156O, which was investigated as follows: After shakingwith dry ethereal hydrochloric acid the resulting precipitat2886 THE REACTION BETWEEN gENZYLAMINE, ETC.weighed 2.85 grams, and proved t o be benzylamine hydrochloride,a specimen melting a t 257O after recrystallisation from a mixtureof alcohol and chloroform. Assuming the substance t o be the di-benzylamine salt of bromomaleic acid, theory requires 2.87 gramsof benzylamine hydrochloride, whereas if a bromobenzylamino-acid(benzylamine salt) had been present only half this quantity ofbenzylamine hydrochloride should be obtained.The ethereal mother liquor deposited colourless crystals of bromo-nialeic acid, melting a t 138-140° (Found, Br =4P15.Calc.,Br = 41.03 per cent.).A substance melting at 1 5 6 O and apparently identical with theabove, except for its more rapid reducing action on permanganate,was obtained by allowing an alcoholic solution of bromomaleic acidand benzylamine (2 mols.) to crystallise in the cold.The mother liquor from the 4.1 grams of substance mentionedabove deposited, on concentration, 0.25 gram of a substance decom-posing a t 239--242O, presumably crude (mesa-)dibenzylaminosuccinicacid, and then, on the addition of chloroform and further concen-tration, a precipitate of needle-shaped crystals, which melted at142--148O, and proved on analysis to be the monobenzylamine saltof bromomaleic acid (Found, Br = 26.06.Calc., Br= 26.01 percent.) (compare T., 1911, 99, 1779).It is suggested that, contrary to what was sta<ted in the previouscommunication, the substance melting a t 15 6 O is the dibenzylaminesalt of bromomaleic acid, and that on heating in alcoholic solutioni t is converted partly into the dibenzylamino-acid with simultane-ous formation of the monobenzylamine salt of bromomaleic acid :I3r.E C0,H,C,H7.X 11 Il C;( LCT H*C,H ,)*CO,H + -H*C*C02H,C7H7*NH2 HC(NH*C,H7)*CO,H 2+ C7H7*N H,, H B I-.BIS*CO,HHC* CO,H, C,H,*NH,I n a second experiment 7 grams of meso-dibromosuccinic acidwere treated with 8.2 grams of benzylamine (more than 3 mols.) in60 C.C. of absolute ethyl alcohol. The mixture was heated t o boilingfor five minutes, and then kept for about twp and a-half hours.The precipitate weighed 8.57 grams, and melted and decomposeda t 1 4 2 O . The mother liquor yielded 0.9 gram of coarse needlesmelting a t 156O (ciibenzylamine sa.lt of bromomaleic acid) and alittle nearly pure (meso-)dibenzylaminosuccinic acid (decomposingat 250-254O).The 8-57 grams of substance were dissolvetd in water, separatedfrom some undissolved dibenzylamino-acid, the solution acidifiedwith hydrochloric acid, and extracted with ether. The etherealextract yielded some meso-dibromosuccinic acid (decomposing a FIRTH AND MYERS: SOME I’ROYERl’IES OF SOLUTIONS. ETC. 2887284-285O; found, Br = 58.13. Thisshows t h a t the reaction product obtained under these conditionsstill contains unchanged nzeso-dibromosuccinic acid (as benzylaminesalt). A similar result was obtained by hesating the acid andamine together f o r two hours in boiling chloroform solution.Calc., Br = 57.97 per cent,.).CIIJCNTCAL I)RIBAI:TMEIW,BIRMINGHAM.THE UKIVERSIIT~, E D G I ~ \ S I O S