EVANS: THE DISTILLATION OF MIXTURES, ETC. 2233CCXXXII1.-The Distillation of Mixtures o f Enantio-morphously Related Substances.By WILLIAM CHARLES EVBNS.ALTHOUGH most of the possible types of behaviour which can ariseduring the distillation of mixtures have been experimentally studied,one of the simplest appears hitherto to have escaped investigation.This type is the one in which the two components of the liquidmixture have the same boiling point under all the pressures dealtwith, and in which the boiling points of mixtures of all com-positions are identical with those of the components under thesame pressure. Under these conditions it would be expectedthat all mixtures of the two components should behave on dis-tillation like a single substance; the composition of the vapourshould be the same as that of the liquid even when the pressureis varied, and no separation by fractional distillation should bepossible .Probably the only instances in which conditions of this highlyspecialised character can be experimentally realised are to be foundamongst mixtures of enantiomorphously related substances, and inthe study of such cases the delicate nature of the method availablefor determining the composition of the mixtures greatly facilitatesthe practical examination.It is, of course, well known that noseparation of the optically active components of an externally com-pensated substance can be effected by fractional distillation; but ithaa not previously been shown and is not immediately evidentVOL. XCVII.7 2234 EVANS: THE DISTILLATION OF MIXTURES OFthat mixtures of two enantiomorphously related isomerides in anycasual proportion would resist separation by fractional distillation.The following observations of the behaviour of optically activemixtures of d- and I-camphor and also of d- and I-tetrahydro-quinaldine were made at the suggestion of Professor Pope for thepurpose of obtaining the lacking experimental data.A. Distillation of Mixtures of d- and 1-Camphor.The mixtures were prepared from natural d-camphor and arti-ficially prepared externally compensated camphor in suitableproportions. In the following series of experiments, the camphorwas distilled from a retort heated by a naked flame, care beingtaken to prevent the distillate from solidifying in the neck of theretort; the distillate was collected in a number of fractions, and thespecific rotatory power of each determined in benzene solution,(1) d-Camphor alone was distilled, and the specific rotatorypowers of the first and last fractions of the distillate, in 10 per cent.solutions, found to be [a], + 40'85O and + 40-95O respectively; thesevalues are ident,ical within the limits of experimental error.(2) On distilling externally compensated camphor in the samemanner, the first and last fractions were found to be opticallyinactive.(3) A mixture consisting approximately of one part of Z-camphorand two parts of d-camphor was separated into five fractions bydistillation as above described under atmospheric pressure ; thespecific rotatory powers of the five fractions in 8 per cent.benzenesolution were [a], + 27*15O, + 27*02O, + 27.37O, + 26*66O, and+ 26'14O respectively. A similar distillation was performed withanother mixture, and yielded five fractions with the specific rotatorypowers [a], + 21.40°, + 20'09O, + 19*75O, + 20*21°, and + 20s050respectively. It will be noted that the rotatory powers of the fivefractions composing either series are not identical within the limitsof experimental error, so that some slight degree of separation isindicated. I n these determinations, however, no special precautionswere taken to ensure thorough admixture of the two componentsbefore distillation; it therefore seemed possible that, owing to thereadiness with which camphor sublimes, the two components, presentin the solid state in different quantities, might have sublimed atdifferent rates determined by the surface exposed and the tem-perature attained by the solid.(4) The contingency just indicated waa obviated by meltingtogether the mixturea of d- and I-camphor before introducing theminto the retort; after taking this precaution, one mixture, similarto those examined in (3), gave five fractions, of which the specifiENANTIOMORPHOUSLY RELATED SUBSTANCES.2235rotatory powers in 8 per cent. solutions were [a], + 20*24O, + 20*66O,+ 20*55O, + 20.60°, and + 20-56O respectively. These five values areidentical within the limits of experimental error, and it is thusindicated that intimate mixtures of d- and Z-camphor cannot bealtered in composition by distillation under atmospheric pressure.B.Distillation of Mixtwres of d- and l-Camphor in Steam.(5) A mixture prepared from d-camphor and externally com-pensated camphor, and consisting of about two parts of d-camphorand one of Z-camphor, was distilled in a current of steam, threefractions being collected; these, and the residue left in the distillingflask, were collected and dried in the air. The three fractions andthe residue gave the specific rotatory powers, in 4 per cent. benzenesolution, of [a],, + 16'44O, + 15*24O, + 14'55O, and + 27*90° respec-tively ; repetition of the operation with another similar mixtureyielded four fractions and a residue, which gave the specific rotatorypowers of [u],, + 1 5 * 9 5 O , + 14*28O, + 1 4 * 1 5 O , + 14*49O, and + 27015~respectively.No precautions were taken to ensure intimate admix-ture of the CE- and dl-camphor previous to the steam distillation, andthe variations in the specific rotatory power of the distillate andthe large difference between these values and those referring to theresidues left in the distilling flask might, it was anticipated, bedue to differences in the rate of sublimation in steam of the solidactive and externally compensated substances.(6) That the cause just suggested of the variable rotatory powersof the distillate and residue is the true one was demonstrated by thefollowing trials. A mixture of approximately one part of d-camphorand two parts of externally compensated camphor WBS melted andallowed to solidify, then roughly ground, and subjected to steamdistillation as described under (5).The three fractions of camphorwhich distilled and the residue which remained in the still, examinedin 5 per cent. benzene solution, gave the specific rotatory powers[ u ] ~ + 21*54O, + 21*26O, + 21'62O, and + 21'55O respectively.The close approximation to constancy of these numbers shows thatmixtures of d- and Z-camphor, if care is take to ensure intimateadmixture, behave like a single substance on distillation in acurrent of steam.C. Dietillatioolc of Mixtures of d- and l-Tetra~ydroqwircaldine.As indicated in the previous pages, the investigation of theproblem under consideration with the aid of mixtures of d- andLcamphor is complicated by the fact that these substances are solidat the ordinary temperatures, and that their melting and boilingpoints do not differ greatly.Further information ww therefore7 ~ 2236 EVANS: TEE DISTILLATION OF MIXTURES7 ETC.sought from the study of mixtures of d- and I-tetrahydro-quinaldine.Externally compensated tetrahydroquinaldine wa8 treated withd-a-bromocamphor-wsulphonic acid, and the laevo-component of thebase in large measure separated by Pope and Peachey's method(Trans., 1899, 75, 1068); the mixture of the d-base with a smallproportion of the I-isomeride remaining after the separation of theI-tetrahydroquinaldine d-a-bromocamphorsulphonate was isolated bydistilling the mother liquors in a current of steam after additionof lime.Such mixtures were used in the following series ofexperiments.(7) Two specimens of the mixed d- and I-tetrahydr~uinaldine,containing (a) about 75 per cent. of d- and 25 per cent. of I-tetr*hydroquinaldine, and ( b ) about 56 per cent, of d- and 44 per cent.of I-tetrahydroquinaldine, were distilled from Wurtz flasks undera pressure of 300 mm. of mercury. The rotatory powers, a,, ofthe several fractions collected were determined in 100 mm. tubesat 16O:(a. ) (a. 1Fractions. QD. an.1 + 28-39' + 6 *85'2 28'47 6.873 28-48 6-874 28.46 6-895 28-41 6.876 28'44 6 '887 28'43 6.878 28'479 28'49(8) Two specimens of the mixed bases containing ( c ) about 65 percent. of d- and 35 per cent. of I-tetrahydroquinaldine, and (d) and(e) about 78 per cent.of d- and 22 per cent. of I-tetrahydroquinaldine,were distilled from a Wurtz flask over a naked flame, ( c ) underpressures increased gradually from 55 to 415 mm. of mercury, (d)under atmospheric pressure, and (e) under pressures which werecaused to rise and fall during distillation between 140 and 600 mm.of mercury. The rotatory powers observed in 100 mm. tubes ofthe various fractions collected are stated in the appended table:Fractions.12345678910(c. 1a,.3- 11.34"11'3911 '4011.3811.4011.3811'3911-3611-3911 *39(d. 1an.+ 33-42'33.4333'4633-4533 '4233'46(e. 1a,. + 33-40"33-3833'3933.4533.3933'4TERTIARY ACIDIC, ETC., DERIVATIVES OF D-CAMPHORIMIDE. 2237From the above results, it is to be concluded that on distillingmixtures of enantiomorphously related compounds under varyingconditions of pressure, no difference in composition is to be observedbetween the various fractions of the distillate by determination ofthe rotatory powers, and that< therefore no separation of suchmixtures into their optically active components is possible bydistillation. The fact that no change in rotatory power can bedetected in partly compensated mixtures of either d- and E-camphoror d- and I-tetrahydroquinaldine on fractional distillation maydoubtless be regarded as a demonstration that under the conditionsprevailing during distillation no combination of a racemic characterexists between the enantiomorphously related isomerides.The above are the first recorded instances of the theoreticallysimplest case arising in the distillatmion of mixtures, but no doubtsuch instances could be easily multiplied amongst other mixtures ofenantiomorphously related compounds.THE CHEMICAL LABORATORY,UNIVERSITY OF CAMBRIDGE