94 LOWRY, PICKARD, AND KENYON.: THE ROTATORY DISPERSIVEXII. - T h e JZotutory Dispersive Power of OrganicConapounds. Payt V. A Comparison o j theOptical and Magnetic Rotatoi-y Dispersions inSome Optically Active Liquids.By THOMAS MARTIN LOWRY, ROBERT HOWSON PICHARD, andJOSEPH KENYON.ONE of the first objects aimed a t in the present experiments on therotatory dispersive power of organic compounds was to compare(1) the natural rotatory disperson and (2) the magnetic rotatorydispersion in a number of optically active liquids. The substanceswhich have been available hitherto for making such a comparisonhave usually been of complicated structure, as in the case of pineneand nicotine; or they have been difficult to prepare in a state ofoptical purity, as in the case of turpentine and optically activeamyl alcohol; they may even have shown anoinalous rotatorydispersion, as in the case of ethyl tartrate. The preparation of alarge number of active alcohols in a state of optical purity (sePOWER OF ONGANIC COMPOUNDS.PART V. 95Pickard and Kenyon, T., 1911, 99, 45; 1912, 101, 620; 1913, 103,1923; and P., 1912, 28, 42) provided a unique supply of materialfor the comparison referred to above, and led directly to theco-operation in experimental work which forms the basis of thepresent paper. The magnetic rotations and dispersions in mostof the compounds used in this investigation have *been given inPart IV of this series of papers; the corresponding optical rotationsand dispersions are set out in the following pages.Wiede.mann’s Law.The comparison of the optical and magnetic rotatory dispersionsin an optically active liquid was first made by G.Wiedemann in1851, five years after the ‘(magnetisation of light” had been ais-covered by Faraday. Wiedernann made a series of comparativemeasurements in the case of turpentine oil with solar light of fivedifferent wave-lengths. After working out the ratio of the tworotations for each wave-length, he concluded that “ These numbersagree so well together that one may assume that the law of pro-portionality of the rotation of the plane of polarisation producedby the current in light of different wave-lengths, with the rotationalready existing in turpentine-oil may be regarded as correct ”(Ann.Phys. Chem., 1851, [ii], 82, 231).After an interval of fifty years Wiedemann’s law was tested byDisch (ktnn. Physik, 1903, [iv], 12, 1153), who discovered markeddeviations, especially in substances showing anomalous rotatorydispersion, but concluded that these were due to lack of homogeneityin the material. Darmois (Ann. Chim. Pltys., 1911, [vii], 22, 247,495), from very similar data, concluded that “the law of pro-portionality was quite inexact, and that Wiedemann’s result was theresult of a pure chance.”A series of experiments on quartz (Lowry, Phil. Tram., 1912, A,212, 295) showed that in this case the proportionality between thenatural and artificial rotatory powers was exact within the limitsof experimental error over a wide range of the visible spectrum.The experiments now described show that this proportionality,which is perhaps a general property of optically active crystals,does not exist in the case of optically active liquids.Occasionally,as in the case of phenylmethylcarbinol, the optical and magneticdispersion-ratios come very close together, but the accidentalcharacter of this agreement is revealed by the wide disagreementin the dispersion-ratios a,35s/a5asl of the next homologue, thus :{ $ 9 (mag.) ... 1.789) { 5 ,, (mag.). 1.731)CtjH,’CH(OH)’CH, (opt.) ...... 1.736 C H *CH(OH)*C,H, (opt.) . 1’6796 LOWRY, PICKARD, AND KENYON: THE ROTATORY DISPERSIVEEqually emphatic evidence that Wiedemann’s law cannot beapplied to organic liquids is afforded by the behaviour of the seriesof fatty alcohols.Thus we find that the characteristic dispersion-ratios for the following series of carbinols are :“ Methyl ” series CH,‘CH(OH)’R (opt.) ..................J “ Ethyl ” series C,H,‘CH(OH)*R (opt.) ..................1 *6511.6391.663{ ’ 9 9’ ,, (mag.) .................. 1.637................ 1 Y 9 8’ 1 9 mag.) 1 633f “hoPropy1” series (CH,),CH*CH(OH)*R (opt.) .......1 $ 9 9 ) 1 ) (mag.) ......... 1 *635The lowest members of the series to show optical activity containa t least four carbon atoms; their optical rotatory dispersions showmarked anomalies (p. 84), but their magnetic dispersions areperfectly normal, thus proving again that Wiedemann’s law cannotbe applied to them.In the case of more complex compounds the discrepancy is stilllarger. Thus, in the case of a complex ester, there is the followingremarkable contrast between ;he two dispersion-ratios :2 -048 Methyl camphorcarboxylate (opt.). ............. { ’9 9 9 (mag ) ............ 1.632Ethyl tartrate, which shows anomalous optical rotatory dispersion,agrees with this substance in giving a magnetic dispersion-ratio1.630, which does not differ from the ratio characteristic of suchsimple esters as ethyl acetate.It is not easy t o explain this inequality in dispersive power.In seeking for an explanation the chief clue is to be found in thefact that the difference between f i e optical and magnetic dispersionsvanishes in the case of quartz, where the optical activity is due t oasymmetry in the crystal structure instead of in the molecule.Inthis case it may be suggested that every part of the moleculecontributes its quota both to the optical and to the magneticrotatory power. In optically active liquids, on the other hand, itis probable that the natural rotatory power is influenced to a muchgreater degree by the atoms or groups of atoms which are nearestto the centres of asymmetry, whilst the magnetic rotatory power isinfluenced to an equal extent by all atoms of a given kind. Theexistence of a rough “additive law” for magnetic rotations andthe highly constitutive character of optical rotatory power may thusperhaps account for the inequality of the two dispersions.The deviations from Wiedemann’s law in the case of opticallyactive liquids suggest a further question, to which a t the presenttime no decisive answer can be given.The inequality in the twodispersions might be attributed : (a) to a change in the value of thPOWER OF ORGANIC COMPOUNDS. PART V. 97dispersion-constant,” ~ ~ 2 , in the equation a= -!k...- (see part 11),A2 - A o ior ( 6 ) to the introduction (as in the case of ethyl tartrate) of asecond term into the dispersion-equation, which then becomes :I n the case of ethyl tartrate, the second term in the equationfor the optical rotation is negative in sign; it is therefore easy t odetect its influence in the anomalous rotatory dispersion of thesubstance (Lowry and Dickson, P., 1913, 29, 185). No anomalyexists, however, in the magnetic rotatory dispersion of this ester,and no deviation from the normal form of the dispersion-curve canbe detected, in spite of the great probability that the liquid containstwo dynamic isomerides with independent dispersion-constants.Itis therefore evidently very difficult to detect the presence of a secondterm in the dispersion-equation unless the two terms differ in signor contain dispersion-constants differing very widely in magnitude.From the theoretical point oi view much might be said in favourof adopting the second of the explanations set out above, but inactual practice no such option exists. There is, in fact, no alterna-tive t o the employment of the simple formula which assumes thatthe deviations in Wiedemann’e law are due to changes in the valueof the dispersion-constant A:.This formula, as has been shownin Part I1 of this series of papers, expresses the form of thedispersion-curves, both for optical and for magnetic rotations, withan accuracy which exceeds that which can be attained in any oneindividual series of observations ; it would therefore be quiteimpossible to determine the magnitude of the four constants in thetwo-term equation, even if it were known that this equation wasthe correct one to apply.Optical Rotatory Dispersion in Rornotogous Series.There is a marked tendency for the optical (like the magnetic)rotatory dispersions in homologous series of compounds 50 settledown t o a steady value after the first members have been passed;but the steady value does not appear until the ccmpound containsfive or six carbon atoms, the abnormalities usually persisting untilthe growing chain has established itself as the largest of the fourgroups attached to the asymmetric carbon atom. Thus we have:VOL.cv. 98 LOWRY, PICKARD, AND KENYON : THE ROTATORY DISPERSIVEMethglcarbinols,CH,'CH(OH)'R. Opt. Mag.R=Ethyl R = Propyl t o decyl ..... i:::;} 1-637 .................Ethylcarbinols, *C,H ,*CH( OH)*R.R = Methyl ............... 1.662R= Rutyl .................. 1 *650R = Amyl, hesyl ......... 1 $391R=isoPropyl ............ 1,6611z'soPropy lcarbi 11 ols,(CH,),CH(OH)*R. Opt. Mag.1'6631 1'635R = Met hy 1 ............... 1 '6 97 1R=Ethyl to octyl ......isoBn t vlcarbinols,R=Methyl ............... 1'631 1'644R=$Ethyl ...............1'633f 1.635R=fPropyl ............... 1.657 1'639(CH,),CH*CH;CH(OH)*R.* The diethylcarbinol is, of coursc, inactive ; the ethylpropylcarbinol has toosmall a rotatory power for an exact measurement of the disllersion ratio.t The agreement with the magnetic dispersion ratio 1'685 is exceptionally close,but does not reappear in the two adjacent homologues and must therefore beregarded as fortuitous.$ These samples were probably not quite pure, but this would not be likely to affectthe dispersion-ratios.The dispersive power of methylisopropylcarbinol is remarkable,especidly in contrast with the steady value 1.663 of the dispersion-ratio in six higher homologues. It ie. however, fully justified by aconsideration of the dispersion-ratios in the series set out below :C H,'CH( OH)*C)H, ........................inactiveCH,'C H (0 H )'C H2 Me ..................... 1 '6621 *697CH,*CH(OH~*CAIC,J ........................ 1'707CH;C H (OH) 'C H Ale, .....................Attention may also be directed to the high dispersive power o f :CHOptically active amyl alcohol, s'CH'CI-I,'OH ........... 1 *700c,H,/CH,Optically active valcric acid, \CH.CO'OH ............... 1 *710C2H5'These two compounds are very similar in structure, and differfrom the active secondary alcohols in that the asymmetric carbonatoms are linked entirely to carbon and hydrogen, instead of t ocarbon, hydrogen, and oxygen. The fact that the oxygen has beenshifted away from the asymmetric atom, instead of diminishing therotatory dispersion, actually itxreases it. It appears, in fact, thatoxygen contributes relatively little to the rotatory dispersion ofthese compounds, which seems to be influenced more by carbon thanit is by oxygen.A bsolufe Molecular Rotation (Optical).In dealing with the magnetic rotatory power of liquids, Perkinselected water as the standard both of specific and of molecularrotation.This choice was justifid by the fact that it is mucheasier t o measure the rotation produced in a deciinetre length oPOWER OF ORGANIC: COMPOUNDS. PART V. 99water than to determine the strength of the magnetic field whichproduces this effect. I n dealing with the natural rotatory powerof optically active liquids, no such standard is needed, as thisproperty can be measured directly in angular degrees per decimetrelength of the liquid; a correction for density or concentration givesthe (‘ specific rotalion,” from which the “ molecular rotation ” mayeasily be calculated without introducing any other substance as abasis for comparison.This fundamental difference in the method of dealing withnatural and magnetic rotations leads to a further contrast when theattempt is made to eliminate the influence of dispersion by workingout the “absolute” rotatory powers of a substance.I n the caseof magnetic rotatory powers the comparative method was extendedt o this final stage in the working out of the experimentalobservations, the magnetic rotation in water being reduced to“ absolute ” wave-length as well as that in the substance.In con-sequence of this extension ot the comparative method the finalreduction produces only slight alterations in the values of themolecular rotatory powers; in the case of substances, such astert.-butyl alcohol, which have the same dispersive-power as water,the “ absolute ” molecular rotation is actually identical with thatfor sodium or for mercury light; 5 table showing the small changesproduced in the case of typical aliphatic compounds is given inPart IV (this vol., p. 88). In dealing with optical rotatorypowers, it is necessary, in order to eliminate the influence ofdispersion, t o reduce the readings f o r the substance to the standardwave-length, a t which h2= 1 + ~ 2 .This produces large alterations,which cannot be diminished o r removed by the use of relative valuesfor two substances. The actual alterations which are produced are$own in the following table fx a series of typical dispersion-ratios :ap5t)/a546, = 1.636 1.644 1’651 1.673 1 735 1.764a,ba,/as6, = 0.2784 0.2763 0.2745 G.2691 0.2555 0.2499a,bs./a5993 = 0.3275 0.3254 0,3237 0.3181 0.3045 0.2989andIn khis table it is seen that the absolute rotations range from28 to 25 per cent. of the rotations for green mercury light, and from33 to 30 per cent. of the rotations f o r sodium light.Tabulated Measureme.rLts.I n table I two series of values are given for the absoluteThe values given in the molecular rotation of each compound.I3100 LOWRY, PICKARD, AND KENYON : THE ROTATORY DISPERSIVEfifth column are calculated from the readings of column 4, whichwere made in London with peen mercury light; the values in thesixth column are calculated from readings which were made inBlackburn with sodium light. I n nearly every case the two valuesagree together closely, but in the few cases in which markeddifferences are observed, the second value is t o be regarded as themore trustworthy.Thus in the case of methylethylcarbinol, thelower value obtained with mercury light is almost certainly duet o the absorption of water, as it was found to diminish steadily insuccessive series of experiments; so also in the case of methyl-IL-butylcarbinol, the sample used for measuring the dispersion wasknown to be of slightly lower rotatory power than the sample usedwhen the sodium readings were taken.The dispersion-ratios and dispersion-constants are set out incolumns 2 and 3.The rotatory dispersion in these compounds isso steady that no new facts are disclosed when the effects of dis-persion are eliminated ; all the essential characteristics of thedrift of rotatory power in honologous series can be seen equallywell in the readings for sodium or for mercury light. Attentionmay, however, be directed to the fact that the small rotations inthe " ethyl "-carbiiiols are associated with exceptionally small dis-persion, whilst the high optical activity of the '' isopropyl "-carbinolsis associated with a high dispersive power ; the '' methyl "-carbinolsoccupy an intermediate position both in rotatory power and indispersive power.This parallelism between rotation and dispersiondoes not apply to the abnormality of the initial members of eachseries, where a low rotatory power may be associated with highdispersive power.Optical Rotatory Dispersion and Absolute Molecular Rotation inSome Optically Active Liquids.ObservedDispersive rotation, ALs. mol. rot.Dispersion- power, a5461 &ratio. 1 0 0 ~ 2 . (100 nim.). L. F'. & I<.( a ) " Methyl 'j Carbinols (normal).Methyl ethyl carbinol .........,, propyl , , .........,, butyl ,, .........,, amyl ,, .......,, hexyl ,, .........,, heptjl ,, .........,. octyl ,, .........,, nonyl ,, .........,, dccyl ,, .......1 '6611-6521-6531.6511'6531.6511 -6491'6511 -6532.622'402 '422.372 '422-372.322.372-4212-57' 1.3.131 3.3013.25 3.95 3-9110.84 [3*72] 3-819.88 3.85 3-889.54 4'14 4'108'68 4.17 4.178'44 4'45 4.437-87 4.50 4-527-52 4'61 4-6POWER OF ORGANIC COMPOUNDS.PART V. 101Optical Rotatory Dispersion and A bsolute Moleculnr Rotation insome Optically Active Liquids (continued).0 bservedDispersive rotation, Abs. mol. rot.Dispersion- power, a5461 &ratio. ~ O O A , ~ . (100 mm.). L.( 6 ) ' I Ethyl " Carbinols.Ethyl prop91 carbinol ......... [1-615~ * [1.39] * 1-60" ,, biityl ,, ......... 1.650 2-34 7.94 .. amyl . . . . . . . . . . 1.639 2.05 8-07 ,, hexyl ), ......... 1.639 2.05 7.80,) octyl ,, ........1.634 1 92 6'1 4(c) " isopropyl " Carbinols.isoPropy1 incthyl carbinol ... 1.697 3-46 4-74"), ethyl . . . . . 1.661 2-62 14-71 ,) propyl ,, ... 1-665 2-74 20'62 ,) butyl ), ... 1.665 2.74 24-97 ), nmyl . . . . . 1.663 2.67 22'46 ,, h x y l ,, ,., 1-661 2.62 51-16,, octyl ,, ... 1.661 2-62 18.38,, decyl . . . . . 1-669 2.86 15'59(d) " Butyl" Carbinols.isoRutyl methyl carbinol.. .... 1 *631 1-92 19'44"tert. -Kntylmethylcarbinol ... I 707 3-68 7-87( e ) Aromatic Secondary Alcohols.Phenylmethylcarbinol ...... 1.736 4 -29 52'49"Phenylethylcarbiiiol ......... 1'674 2 93 32-37Benzylmethylcarbiuol ......... 1.833 6-13 32'47B-Phenylethylmethylcarbinol 1,679 3.03 16 *55d f ) Amy1 alcohol (active) ...... 1.700 3.54 5.40"isoValeric acid .................. 1.710 3-74 20'50f ,, ethjl .. . . . . . . 1.633 1.90 9 *39f ,, propyl ........ 1.651 2-37 4 '990 573-092-543-783-561 *344 -957-8910.6910-6311 *0111.1510.826-852-57--16.1411.9110-576.795.80-1'. & K.0.673'133-503.i83 -561.344.947 8910.6410.5310.8611.0611 -066 -84---15.9311'98 --- -* The rotatory power was too smd1 t o give trustworthy figures for the dispersivet These saniples were of doubtful purity.power of this compound.Summary and Conclusions.1. Optical rotatory dispersion varies more widely than magneticrotatory dispersion, but remains constant in homoIogous series ofsecondary aliphatic alcohols when the " growing chain ') of carbonatoms has established itself as the heaviest radicle in the asym-metric molecule.2. Wiedemann's law, which applies exactly in the ca.se of quartz,does not hold good in the case of optically active liquids. Theoptical is usually greater than the magnetic rotatory dispersion,but the converse is sometimes observed in aromatic compounds102 PERKINS: THE POROSITY OF IRON.3. Remarkable variations of optical rotatory dispersion areobserved in the series:C6H,'CH(OH)'CH, ................ a&a,4G1 = 1.736C,HB'CHiOH~*CII,*CH, ........... ,, = 1'674C, H B*C H,-CH( 0 tI )'C H ............ > IC,Hj.CH2'CH;C1H(O11)'C.H:: ..... > )and in the series:C H ;C H (0 H ) 'CH2 Me ...............CH,'CH(OH)'CHMe, .............. 9 9CH;CH(OH)'CMe, ................. 9 99 ,High dispersion,s are also observed in : '= 1.833= l $ i 9= 1'662= 1.697= 1.707= 1.700= 15'10The expenses incurred in the researches described in this and inthe preceding paper have been defrayed in part by generous grantsfrom the Government Grant Committee of the Royal Society, whichare hereby gratefully acknowledged. Our thanks are also due toMr. W. P. Paddison, t o Mr. H. W. Southgate, and t o Mr. H. R.Courtman for assistance in carrying out the long and tedious seriesof observations recorded in these papers.GUY'S HOSPITAL.LONDON, S.E.MUNICIPAL TECFINICAL SCHOOL.BLACKBURN