64 MILLS AND BAIN: THE CONFtQURATION OF THE1X.-The Conjguration of the Doubly LinkedNitrogen Atom. Optically Active Salts of tlaeSernicarbuzone ayid Benzoylphen~lhydrazone o jcycloHexanone-4-carboxylic Acid.By WILLIAM HOBSON MILLS and ALICE MARY BAIN.IN spite of the facts that a certain number of phenylhydrazonesand semicarbazones have been found to occur in isomeric modifica-tions, and that these are exclusively of the asymmetrical typeR,R,C:N*NHX, for several reasons-of which the most importantare the greater irregularity in the occurrence of the isomerism andthe absence of significant chemical differences between theisomerides, like the differences in ease of dehydration of thealdoximes, or in the course of the Beckmann transformation in theketoximes-it is probably generally held that the evidence for theexistence in these compounds of a geometrical isomerism, condi-tioned by the presence of nitrogen doubly linked to carbon inaccordance with the hypothesis of Hantzsch and Werner, is by nomeans so strong as it is in the case of the oximes. It thereforeseemed desirable to extend to the semicarbazones and phenylhydr-azones the method which we had previously employed (T., 1910,97, 1866) f o r the examination of the configuration of doubly linkednitrogen in the case of the oximes, namely, that of investigatingthe possibility of the existence in an optically active form ofphenylhydrazones and semicarbazones of such constitution that theyare molecularly symmetric or asymmetric according as the threevalencies of the doubly linked nitrogen which they contain are,or are not, in one plane.Our experiments on the oximino-groupwere carried out on the oxime of cycZohexanone4-carboxylic acid(Perkin, T., 1904, 85, 416). I n the present communication anaccount is given of the investigation from a similar point of viewof derivatives of the hydrazone of this acid. I f in the hydrazoneof cycZohexanone-4-carboxylic acid (or its N-substitution deriv-atives) the valencies of the doubly linked nitrogen lie in one planethe molecule possesses a plane of symmetry (Fig. l), and thecompound must therefore be incapable of optical activity. If, onthe other hand, the three valencies of this nitrogen atom are notin one plane (Figs. Za, Zb), then the molecule possesses neither aplane nor a centre of symmetry, and must therefore exist in twoenantiomorphous forms exerting equal and opposite effects on planepolarised light DOUBLYFIG.l.*H C@,H'*.. %\ /C/\ 7% p 2\/CH, CH,LINKED NITROGEN ATOM."./c *-./c65' Therefore if an N-substitution derivative of the hydrazone ofcy clohexanone-4-carboxylic acid can be obtained in an opticallyactive form, the conclusion must be drawn that in that compoundthe three valencies of the nitrogen atom do not lie in one plane.Unfortunately, the plienylhydrazone of cycZohexanone-4-carb-oxylic acid could not be investigated on account of experimentaldifficulties. The action of phenylhydrazine on this acid wasexamined by Perkin (loc. cit., p.428), and the resulting phenyl-hydrazone was found to be an un6table oil, readily decomposing,under the influence of acids, into ammonia and tetrahydrocarbazole-3-carboxylic acid.The X-substitution derivatives of the phenylhydrazone are, how-ever, naturally much more stable. W e prepared the N-methyl andY-benzoyl compouuds, and found them t o be well defined, crystal-line substances. The former did not give salts of a satisfactorycharacter with the common alkaloids, but the benzoylphenylhydr-azone was readily obtained in a dextrorotatory form by means ofits quinine salt. The sodium salt, obtained from the quinine saltby decomposing it with excess of sodium hydroxide, was found,after the removal oE the quinine, to have in the aqueous alkalinesolution a molecular rotation of [MID 238.6O.The optically activesolutions of sodium or ammonium salt obtained in this mannershowed marked autoracemisation, which, in the ammonium salt,proceeded in accordance with the unimolecular formula. As hadbeen found to be the case with the corresponding oxime, the auto-racemisation was retarded by the presence of alkaline hydroxides.The strong alkali hydroxides, however, even in dilute solution,brought about a t the same time some chemical alteration (probablyheavy lines, those behind the plane of the paper by dotted lines.* In these diagrams bonds projecting in froiit of the paper are represented byVOL. cv. E66 MILLS AND BAIN: THE CONFIGURATION O F THEin the first place elimination of the benzoyl group), and in thepresence of these the observed rate of diminution of optical activitydid not follow the unimolecular law.On acidification of the activesolutions complete racemisation waa brought about, apparentlyinstantaneously.Since the quinine salt contains the dextrorotatory form of theacid combined with the more highly laevorotatory quinine, itssolutions must show initially a small lzevorotation, increaaing as theautoracemisation of the acid component of the salt proceeds, untilit reaches the value corresponding with that of the quinine saltof the inactive acid. I n the case of both the quinine and themorphine salts of the corresponding oxime it was possible to observesuch mutarotation (Zoc. cit., p. ISSO), but in the benzoylphenyl-hydrazone the change proceeds so rapidly that unfortunately itwas invariably found to have entirely completed itself within thebrief time necessary to prepare a solution and observe its rotation.On account.of the rapidity of this change it was evidently hope-less t o attempt to obtain more highly active preparations of thebenzoylphenylhydrazone by recrystallisatisn of its quinine salt.The molecular rotation of the optically pure compound, therefore,could not be determined.These facts explain why the I-acid cannot be obtained from themother liquor from which the morphine salt of the d-acid hascrystallised. On concentration this yields nothing but the 6-acidsalt. The isolation of the dextrorotatory modification depends,therefore, on a process of activation, rather than of resolution, bymorphine.A similw phenomenon was observed by Pope andPeachey in the case of the methylethylpropyl tin salts (P., 1900,16, 42, 116).The semicarbazone of cyclohexanone-4-carboxylic acid (Perkin,Zoc. cit., p. 427) was similarly examined, and with the aid ofmorphine we succeeded in obtaining the ammonium salt of thiscompound also in an optically active condition. After crystallisingthe morphine salt from dilute alcohol, decomposing it withammonia, and removing the morphine, strongly dextrorotatorysolutions were obtained. The values observed for the molecularrotation of the ammonium salt which they contained lay in mostcases between [MI, 30° and [MI, 40°. Even at temperatures in theneighbourhood of Oo the solutions showed rapid autoracemisation,proceeding strictly in accordance with the unimolecular formula.The autoracemisation was retarded by free alkali, but even in thepresence of Lv / 5-soclium hydroxide (together with a little ammonia)the period of half change at 1.lo was as short ;t8 3.6 hours.Onacidifying the solutions the optical activity immediately disapDOUBLY LINKED NITROGEN ATOM. 67peared. There is thue a very close correspondence between theoptically active forms of the oXime, the semicarbazone, and thebenzoylphenylhydrazone of cycZohexanone4-carboxylic acid withregard t a the conditions which affect their autoracemisation.The effect of temperature on the rate of racemisation of thesemicarbazone in alkaline solution was also examined.The velocity-constant was increased 3.49 times by a rise of temperature of 7O.Assuming that the constant is increased in a fixed ratio by equalincrements of temperature, this corresponds with a quotient forloo of 5.96, a. number greatly in excess of the usual value 2 to 3.Dimroth (Aizmlen, 1904, 335, 8) found similarly high values inthe case of the transformation, in various solvents, of methyl5-hydroxy-l-phenyl-l : 2 : 3-triazole-4-carboxylate into the correspond-ing neutral (diazo) compound, and, regarding the two compoundsa t that time as a pair of keto-enol desmotropes, he suggested thata high temperature-coefficient might be characteristic of desmo-tropic change. The fact that the effect of temperature is unusuallyIarge in the present case also indicates the possibility of a hightemperature-coefficient being of frequent occurrence in intra-molecular transformations.The above observations prove beyond doubt that, in the formof their salts, the benzoylphenylhydrazone and the semicarbazoneof cycEohexanone-4-carboxplic acid can exhibit very considerableoptical activity. There would appear to be no reason to supposethat the salts differ in constitution from the acids from which theyare derived, or that the acids themselves are constituted otherwisethan in accordance with the general view as to the structure of thesemicarbazones and phenylhydrazones of cyclic ketones, Sincecompounds of the type C0,LI.CH<c,2--C: ' H2°GH2> C*NH*NR,R2 wouldalmost certainly be less acidic than the corresponding hydrazones,i t is most highly improbable that the active salts should be derivedfrom tautomeric modifications of the hydrazones of this character,containing an asymmetric carbon atom of the ordinary type.The conclusion would therefore appear justified that the molecularasymmetry which arises when the carbonyl oxygen of the symmetri-cal cyclohexanonecarboxylic acid is replaced by the hydrazoneresidue is only to be explained satisfactorily by supposing that theconfiguration assumed by this residue in the resulting hydrazoneacid is such that its distal portion, the substituted amino-group, isdrawn into a position considerably to one side o r the other of theplane which was previously the plane of symmetry of the keto-acid.These experiments accordingly very strongly support the.hypothesisthat in the hydrazones, as in the oximes, the three valencies of theF 68 MILLS AND BAIN: THE CONFIGURATION OF THEdoubly-linked nitrogen atom do not lie in one plane, but aredirected along the three edges of a trihedral angle.EXPERIMENTAL.cycloHexanone-4-carboxylic Acid Bensoylphenylhydrasone.When as-benzoylphenylhydrazine, cyclohexanone - 4 - carboxylicacid, and acetic acid are dissolved in equimolecular proportionsin a small amount of alcohol and the mixture is allowed t o remainat the ordinary temperature, colourless crystals of the abovehydrazone are slowly deposited. By recrystallisation from a largevolume of benzene the compound is obtained in a state of purity,melting a t 169-171O.The yield of recrystallised hydrazoneamounts t o 60-70 per cent. of the theoretical:0-2137 gave 0.5620 CO, and 0.1170 H,O. C=71-74 ; H= 6.08.0.1922 ,, 14.4 C.C. N, (moist) at 18O and 7.38 mm. N=8-4.C,,H,,O,N, requires C = 71-43 ; H = 5-95 ; N = 8.3 per cent.0-3543 required for neutralisation 10.60 C.C. N / 10-NaOH.Equivalent = 334. C20Hl,0,N2H requires 336.Optically Active Salts of cycloHexamne-4-carboxylic AcidBemzo ylphen ylh ydrason e.The dextrorotatory metallic salts of cyclohexanone-4-carboxylicacid benzoylphenylhydrazone were obtained by the aid of quinine.The benzoylphenylhydrazone (3.86 grams) and anhydrous quinine(3.72 grams) were dissolved in 34 C.C. of methyl alcohol; waterwas then added in quantity just short of that required to producea permanent turbidity (about two volumes were necessary), andthe solution inoculated with a few crystals of the quinine salt froma previous preparation.The I-quinine-d-acid salt was slowly deposited in the form ofsmall clusters of silky needles.It was recrystallised from dilutemethyl alcohol. The air-dried salt melts a t 139-141°, and containsone molecule of water of crystallisation :0.3615 lost 0'0095 KO. H20=2.63.The, salt after .drying for three hours at 90-looo gave the follow-0.2511 gave 0.6688 CO, and 0,1547 H,O. C= 72-64 ; H = 6-84.0-3078 ,, 22-26 C.C. N, (moist) at 14.9O and 771 mm. N=8-57.C,H,,03N2,C,,H,02N2,H20 requires H20 = 2 * 65 per cent.ing results on analysis:C20H,03N,,C20H~0,N2 requires C= 72-72 ; H = 6-73 ;N = 8-5 7 per cent.I n order to obtain the dextrorotatory sodium salt a weigheDOUBLY LINKED NITROGEN ATOM.69quantity of the quinine salt was triturated with such a quantityof sodium hydroxide in aqueous solution that the final solutionwould be fifth-normal with respect to the excw of sodium hydr-oxide. The separated quinine was washed with a little water, thefiltrate and washings were extracted three times with chloroform,made up to the required volume and polarimetrically examined, alloperations being conducted at the temperature of melting ice; forexample, 1.2 grams of quinine salt having been treated in thismanner with 0.197 gram of sodium hydroxide, the following polari-metric observation was made, the volume of the solution being15.5 c.c.:1 = 2, C *=7.74, a, =5*44O, [MID = 238.6'.Thle Autoracemisation Phenomena.(a) Arn?ionium Sdt inPresence of N/ 10-A mmonium Hydroxide.The solution of dextrorotatory ammonium salt was prepared ina manner similar to that described above for the potassium salt,1.0 gram of quinine-d-acid salt being employed, together with a suffi-cient excess of ammonia t o leave the solution approximately deci-normal with respect to ammonium hydroxide after decompositionof the quinine salt. The volume of the solution wil.8 12.1 c.c., thetemperature maintained at 19*9O, and the polarimetric observationsmade in a 2-dcm. tube:t (mins.).0 .o5.0810.4722 4330'1740'4249-03Rotation. 1 Itlog.a/a - 2.2.71" -2'22 0.01i01.79 0.01711.15 0'01660'85 O * O l c i i0.55 0*01710.40 0.0169(b) Potassium Salt in Presence of N f 20-Potassium Hydroxide.One gram of quinine-d-acid salt was trea-ted with 10.58 C.C. of aA7/5-solution of potassium hydroxide. The volume of the solutionmas made up after removal of the quinine to 12 c.c., the tempera-ture was maintained a t 19.7O, and the observations were made ina 2-dcm. tube:1, (mins.). Rotation. l/tlog.a/a - x.0 .o 5.93" -4 *54 5.49 0.0073710.01 5-16 0*0060421 24 4'63 0.0089630.03 4 '27 0.0047549.03 3'49 0.00469* This refers t o the ammonium salt70 MILTAS AND BAIN: THE CONFIGURATION OF THEThese tables show first, that in the presence of N/lO-ammoniumhydroxide solution racemisation proceeds rapidly, and in accurd-ance with the unimolecular formula, and secondly, that the substi-tution of potaasium hydroxide for ammonium hydroxide greatlyretards the racemisation (the initial rate is about twenty-six timesless in the latter case).That in the presence of potassium hydr-oxide the rate of diminution of optical activity as the racemisationproceeds shows marked deviation from that. to be expected in asimple unimolecular reaction is not surprising, for towards the endof the forty-eight hours during which the solutions were underobservation they became turbid and discoloured, indicating thatsecondary changes were taking place. As i t was also found bytitration that the quantity of free potassium hydroxide had greatlydiminished, it is probable that elimination of the benzoyl groupwas taking place, followed by decomposition of the resulting unstablephen yl hydrazone.cycloNezanone-4-carboxylic A cia! Phenylmet h ylh ydrazone.A solution of as-phenylmethylhydrazine (2.3 grams) in 50 percent. acetic acid (20 c.c.) was added to a solution of cyclohexanone-4-carboxylic acid (2.6 grams) in water (60 c.c.).As the hydrazoneotherwise tends to separate as an oil, a few crystalline nuclei froma previous preparation were added. After half an hour the crystal-line deposit was collected, dried, and recrystallised from ethylacetate. The phenylmethylhydrazone is a colourless compound,melting and decomposing at 138-139O. Its solutions become redon keeping:0.2026 gave 0.5103 CO, and 0.1327 H,O.C=68*7; H=7.3.0.2551 ,, 25.6 C.C. N, at 21° and 759 mm. N=11*4.CI4Hl8O2N2 requires C'= 68.2 ; H = 7.3 ; N = 11.4 per cent.Optically Active Salts of the Semicarbasone of cycloHexanone-4-car b oxylic A cid.The semicarbazone of cyclohexanone-4-carboxylic acid, preparedas described by Perkin (loc. cit., p. 427), was obtained in the formof its alkaline salts in a dextrorotatory modification with the aidof morphine.The semicarbazone (2.7 grams), dissolved in a hot mixture ofmethyl alcohol (60 c.c.) and water (16 c.c.), was added to a hotsolution of morphine (4.13 grams) in methyl alcohol (30 c.c.). Themorphine salt separated on cooling, usually in well-formed prisms,which contained no water of crystallisation. I n order to obtainthe dextrorotatory ammonium salt, the very finely-powderedmorphine salt was triturated a t Oo with so much ice-cold aqueouDOUBTJY LINKED NITROGEN ATOM.71ammonia that the excess of ammonia, after dilution of the solutionto the required volume, would be of decinormal strength. Theprecipitated morphine was removed by filtration, washed with alittle ice-water, and the filtrate and washinge, polarimetricallyexamined in a tube round which ice-cooled water was circulated.The amount of morphine left in the solution was found to be sominute and was so readily allowed for by observing the slight,laevorotation which appeared when the autoracemisation was com-plete, that it seemed inadvisable to attempt to remove it. Theammonium salt obtained in tkis manner from 1 gram of morphinesalt was examined polarimetrically with the following result :I = 2, C = 3.258, U, = 1*17O, [MI, = 38.8'.The solutions of ammonium salt obtained from different prepara-tions of morphine salt showed the following molecular rotation :[nil,, 37 80 37.1" 32.5" 27-9" 32.5"The Autoracemisation of the Optically Active Salts of th.eSemicarbasone.(I) Ammonium Salt in Presence of N 10-Ammonium Hydr-ozide.-Morphine salt (1 gram) was triturated with an aqueoussolution of ammonia (0.0584 gram), and, after removal of theprecipitated morphine, the filtrate and washings were made up to13.7 C.C.The observations were made in a 2-dcm. tube maintaineda t 2.3O:t (mins.). 1: o tation. l/tlog.a/(a - x).0.0 1'15" -4'45 1 '07 0.007113-4 0.89 0.008328 -4 0-69 0.007835 '8 0 60 0.007946 -3 0'48 0 -008255 '1 0.41 0.008265.7 0 -35 0.007980-2 0.25 0'0082(2) A mmonium Salt and Amntonium Hydroxide (N/lO) withthe Addition of Sodium Hydroxide (N/ lo).-The solution (11 c.c.)was prepared from 2-43 grams of morphine salt.The observationswere made in a 2-dcm. tube maintained at 1.9O:t (mins. ) Kotation. l/tlog.a/(a -a).0 '0 2-43" -9.5 2-16 0.005413 '1 1-95 0-005327'2 1.75 0 005240 '1 1 *52 0.005163 -0 1-15 0'0052105.7 0.70 0-0051152.1 0'41 0-0051(3) Ammonium Salt and A mmonium Bydroxidc (N/ 10) withthe ddition of Sodiim Eydroxide (N/5) at 1*l0.-The solutio72 CONFIGURATION OF THE DOTJBLY L I S I ~ D NITROGEN ATOM.(12 c.c.) was prepared from 2.5 grams of morphine salt.observations were made in a 2-dcm.tube:Thet (m ins.)0.015.82,s-254.6115'8234'0337.3453.6582-7Rotation.2-80'2-672.582 3 51'941'320 '940.650.445l/llog.Q(n - 2). -0.002310 001410.001390-001380.001390 *001410'00:390~00137(4) Amrn,mium Salt and Amrnonizcm Hydro-xide (N/10) with theaddition of Sodium Hydroxide (N/5) nt 8*l0.--The sulution wasprepared as in experiment (3), but the observation tube was main-tained a t a higher temperature:f (inins.)0'09 -624 *O48.1F4-7118 4155.611 o tation. 1 /tl og. n(n - x).2.42" -2 16 0 00511 *85 0-00481'44 0.00470.92 0'00490.66 0.00480'43 0'0048The first three tables show clearly the effect of an increase inthe hgdroxyl-ion concentration (or diminution in the hydrogen-ionconcentration) on the stability of the optical activity, the diminish-ing values of the constant in the three cases, 0*0081, 0*0052, and0.00139, corresponding with increasing periods of half racemisationof 37 mins., 58 mins., and 217 mins.Comparison of the third and fourth tables shows the effect oftemperatures on the rate of autoracemisation. A rise of tempera-ture from 1-lo to 8-1° was associated with an increase in thevelocity constant from 0-00139 to 0.00485, that is t o say, theconstant was increased 3.49 times by a rise of 7 O .Acidification produces an instantaneous disappearance of theoptical activity. I n illustration of this the following experimentmay be quoted. An alkaline ice-cold dextrorotatory solution wasacidified by addition of 1 C.C. of A'-hydrochloric acid; 1 C.C. ofN-sodium hydroxide solution was then added as rapidly as possible.On polarimetric examination the resulting alkaline solution wasfound to be quite inactive.The authors desire to express their .thanks to the ChemicalSociety for a grant in aid of this work.UNIVERSITY c 11 E M ICAL L A B 0 KAToRY, NOR'T'HF,I~N 1'0LY TPCHN I C 1 XSTI'I'LJTIC,Ca M B K I n Q E. HOLLOWAY, LONDON, N