OP‘l’ICAL IYFLUENCE OF CONTIGUITY O F UNSATURATED GHOU 1’s. 115 X II.-The @pticul In$?i,ence of Co.iztiguity of Unsaturated Gmmps. By JULIUS WILHELM BRUHL. THE following remarks have reference to the striking physical pro- perties of certain terpenes and their derivatives to which attention has been called recently by Kay and W. H. Perkin, jun. (Trans., 1906, 89,839) and are made with the object of correlating their results with those already obtained by myself and other inquirers. The determinations of magnetic rotatory power made by Sir W. €3. Perkin shorn that d-limonene (and inactive dipentene) differs in an altogether remarkable manner from A”5(Y)-p-menthadiene, the valuo abtained for the latter being “ abnormally ” high, The difference is I 2116 BRUHL: THE OPTICAL INFLUENCE OF ascribed by Kay and Perkin to the presence in the p-menthadiene of two contiguous ethenoid linkings.Now it can be shown not only that this explanation is a correct one, but also that it could have been foreseen with csrtainty that hydrocarbons so constituted would exhibit properties such as they are found to possess. It is a well known fact that magnetic rotatory power and refractive and dispersive power are correlative properties ; relations which de- monstrably hold good between structure and refractive or dispersive power must, therefore, also obtain between structure and magnetic rotatory power. During the past sixteen years, indeed, I have shown by numerous researches,+ carried out with a large body of material, that contiguous unsaturated groups, not only *C:C*C:C* but also -C:C*C:O, &c., always give rise to a striking increase of molecular refractive and still more of molecular dispersive power. I have often pointed out that the phenomenon is general, being observed not only in acyclic but also in the benzene, hydrobenzene and other series; moreover, that its existence affords a means, particularly in the case of terpenes and oxyterpenes (camphors), of determining whether or no contiguous unsaturated groups are present in a compound.Many Continental investigators have confirmed my conclusions, especially Eykman (J. F. Eykman, Ber., 1889, 22, 2736; 1890, 23, 855; 1892, 25, 3069), some of whose results are included in the following table, comprising a number of simple examples illustrative of the property under discussion : CH,:CHTH,*CH;CH:CH, , .. ,) 28-90 ,, 29-57 ,) 0.67 Diallyl, b. p. 61". Theory for C,H,, ......... ), 28'89 CH*CH:C H i CH,.. ..................... ll I CH*CH:CH Tropilidene. HC CH HC<L>*CH, .................. ,, 30'79 ), 31.63 ,, 0.84 HC CH Toluene. Theory for C,H8 ......... ,, 30.89 - ~ * My first observations on this subject are to be found in Zeit. physiknl. Chcm., 1891, 7, 140 ; my later work has been published partly in that periodical and also in the J. pr. Chem. and in the Berichte.CONTIGUITY HC CH CH,’ o*c/(~\c =/ *CH:CH-CH, HC CH Anethole. HC CH Methylchavicole. Theory for CloHl,O’ ..... CH,:CH*CHO ................... Theory for C,H,O” I=1 ......... Acrolein. HC C-OH HC=CH HC~\)C.CHO ............... Salicylaldehyde.Theory for C,H,O’O’ ...... OF UNSATURATED GROUPS. HC OH EIC~%*CH:CH-CH:CH*CO,H \=/ ,) HC CH Cinnamylideneacetic acid, dis- solved in acetone. Theory for C,lH,,O’O‘’ iZ5 .... 9 9 45.95 ,, 45.89 16-01 15.67 34.03 32.52 60 ‘42 50.06 117 2.95 2 ‘04 1 *75 0.73 0.58 2 -68 1 ‘28 9.70 2.04 These few examples, which could easily be multiplied ten-fold, show that in every case the compound in which the unsaturated groups are contiguous has the greater molecular refractive and dispersive power, the experimental values in such cases always exceeding those calculated in the conventional manner ; the differences, however, are not at all equal; they vary according to the character and number of un- saturated groups, being greatest and enormously in excess in the case of cinnamylideneacetic acid, in which there are several contiguous unsaturated groups.It is remarkable that the benzene nucleus of toluene, which is a system of contiguous ethenoid linkings, does not behave optically as though it were thus constituted, the values it affords being normal, not in excess of those calculated (see table). Benzene and its homologues all behave alike ; so do also benzene derivatives obtained by substitut- ing univalent atoms for hydrogen : for example, bromobenzene, &c. What may be the cause of these seeming exceptions? In benzenoid compounds, the closed ring consists of six equal carbon atoms forming three equally situated ethenoid groups which apparently neutralise one another. On this account, such compounds cannot118 BMUHL: THE OPTICAL INFLUENCE OF exbibit the properties, chemical or physical, of ordinary contiguous (conjugated) ethenoid systems.As soon, however, as the equality of the six-carbon atoms is destroyed, the balance is disturbed; the pro- perties characteristic of contiguous ethenoid linkings then a t once make their appearance. The disturbance of balance may be effected in various ways. One method is to insert another atom or radicle between two of the six carbon atoms of the benzene nucleus, as in the case of tropilidene, which is formed by insertion of CH, into the benzene ring (see table). A secocd method of destroying the equality of the carbon atoms in the benzene ring consists in associating one or more of these atoms with an unsaturated group, such as C:C, C:O, NO,, NH,, &c.These groups apparently exercise a special attractive influence on the nucleus and by conferring stability on the carbon atom with which they are com- bined disturb the balance within the ring. I t is easy to show that such special attractive influences are actually a t work. Thus allyl- benzene, Ph*CH,*CH:CH,, and all its derivatives are labile compounds which are easily convertible into propenylbenzene or cinnnmene, Ph*CH:CH*CH,, and its derivatives; the latter are stable and not reconvertible into their isomerides. Owing to the special attractive influence exercised by the ethenoid group 0:C of the lateral chain upon the carbon atom of the benzene ring t o which the group is attached, the carbon atom in the benzene ring acquires a particular quality and properties are developed which are characteristic of com- pounds containing contiguous ethenoid linkings ; in fact, every known cinnamyl derivative (a large number have been examined) has a remat kably high refractive and dispersive power in comparison with that of the isomeric allylbenzene derivatives, none of which gives an abnormal value.Anethole, C6H4( OCH,)*CH: C EX* CH,, and methyl- chavicole, C,H,(OCH,)*CH;CH:CHa, are good instances (see table) of such differences. It is to be supposed that two or more C:C liukings are attached to a benzene nucleus in naphthalene, phenanthrene and other condensed benzenoid hydrocarbons ; these certainly are the cause of the excep- tionally high refractive and disper5ive power known to be characteristic of such hydrocarbons.The aldehydes, acids and ethereal salts in which a carbonyl group C:O is attached to a benzene ring show a similar behaviour; for example, salicylaldehyde (see table), amino- and nitro-benzene and their derivatives. In the case of phthalic compounds, in which two CO groups are attached to the benzene nucleus, the increased optical effect is likewise rernwknhle. I: have ~ I F O drawn attention to theCON T I C; U IT I‘ 0 F UN S A‘I’U It A TE D G; R OUPX . 119 increased optical effect manifest in derivatives of anthrnnilic acid, C,H,<c$oH):o, h H, in a special investigation (Ber., 1903, 36, 3640). A third method of destroying the equality in benzene consists in either removing one of the ethenoid linkings from the ring into the side chain or in dispensing with it altogether; by the former process ; toluene, for example, would be changed into a compound having an increased optical effect, /==\-CH, (I); whilst if the latter process were applied to benzene, one of the isomeric hydrobenzenes, ’ (11) or /-\ (HI,), the prototypes of the terpenes, mould be formed ; of these, I1 would exercise a ‘‘ normal ” and 111 an increased optical effect.I have examined I1 niysclf (J. p ~ . Clhem., 1894, 40, 248); I11 is unknown, Aiiwers (Bey., 1906, 39, 3748) has recently referred to the fact that not only compounds of types I and I11 but \==/- /=\ \=/ \=/ /-\ also those of the fourth type, ’’ \-C:C (IV) all exhibit a remark- \-/ ~~ ably increased refractive and dispersive power ; he, however, expressly recognises that it was to be expected from iny researches that com- pounds of these three types would be supra-refractive and supra- dispersive; 11 had indeed confidently expressed the opinion to hiin before hs had carried out any of his determinations that such would turn out to be the case.The discovery made by Sir W. H. Perkin that the A3s(g)-p-men- thadienes (type IV) are more refractive and dispersive than limonene, in which there are no contiguous ethenoid linkings, serves to confirm the views which I have set forth in these pages. And considering the general correlation which exists bet ween refractive and dispersive power and magnetic rotatory power, it was to be foreseen that limonene would exhibit normal and the A3 S‘”-p-menthadienes an increased magnetic rotatory power, although the extent of the difference could not have been predicted. There is one other point OF interest on which I may be allowed to add some remarks. Sir W.H. Perkin has found that the magnetic rotatory power is slightly higher (13.061) in the case of optically active d-A3ns(9’-p-men- thadiene than in that of the inactive dl-compound (12*939).* The values found for d- and I-limonene and dipentene were much lower, namely, 11.246, 11.162 and 11.315. * In the coinparisons iiiacie by Kay and Pcrkin (1). 554), tlie sninc value is erroneously giveii to both co~iipoui~ds.120 BROHL: THE OPTICAL INFLUENCE OF Comparing the optical properties generally of the various isomerides, it will be observed, on reference to Kay and Perkin's paper, that the values all follow the same order but that the differences between the refractive and dispersive powers, especially the latter, of d- and dl-A3.scg)-p-methadiene are remarkably high in comparison with those between the magnetic rotatory powers.These differences are probably real, as they exceed the ordinary experimental errors. Unfortunately the refractive and dispersive constants of d- and I-limonene were not measured by Sir W. H. Perkin. But judging from former determinations, there is no reason to suppose that appre- ciable differences exist in either refractive or dispersive power between d- and Elimonene and inactive dipentene. The last substance can be obtained by simply mixing d- with I-limonene and is probably merely an inactive mixture of optical antipodes. What then can be the reason that d-A3.*cg)-menthadiene and the dl- variety differ to such an extent in molecular refractive and still more in molecular dispersive power, whilst limonene and dipentene do not ? I think that it is not improbable that dZ-A3*s(g)-p-menthadiene is not a mere inactive mixture but a racemic compound of the d- and I-antipodes.It is probable from Kay and Perkin's observations that dLA1-tetrahydro-p-toluic acid is not an inactive mixture but a racemic compound of the d - and I-components; it may well be that the hydro- carbon prepared from it by simple chemical transformations i a also racemic." Of course, a conclusion of such consequence requires to be con- firmed by further experiments, which, however, thanks to the progress in synthesis due to the researches of W.H. Perkin, j u a , and his collaborators, will not offer great difficulties. ADDENDUM. I n the above paper I mentioned that it was desirable to re- determine the spectrochemical constants of d- and I-limonene and of the inactive mixture of them (dipentene). By the kindness of Messrs. Schimmel and Go., I am now enabled to give these constants. This firm sent me d- and I-limonene which had been specially prepared for me with great care. d-Limonene was obtained from cumin oil (Cccrum car&), I-limonene from the oil of the cones of the silver fir {&&a pectiryata). Both, terpenes were fractionated until the rotation * It must, however, be mentioned that the dl-hydrocarbon was prepared by eliminating water from the corresponding alcohol by boiling it with potassiuni hydrogen snlphate.It seems possible that, by this means, a partial inversion of the terpene generated might have taken place and that the special properties of this dl-terpene are perhaps due to such an alteration. This can only be decided by new physical determinations with a sample not treated with potassium hydrogen sulphatc.CONTIGUITY OF UNSATURATED GROUPS. 121 remained constant. It mas, however, not possible to obtain the two limonenes of quite equal rotatory power, that of I-limonene remaining a little lower. It seems, according t o the opinion of Messra. Schimmel and Co., that E-limonene is :wcompanied by some other constituent not separable by fractional distillation. This view was confirmed by the fact that a mixture of equal weights of d-limonene (aD + 104'15') and E-limonene (a, - 1Ol03O'), which have the same specific gravities (di' = 0.8402 and u!:".~ = 0*8407), does not display a rotatory power equal to the difference .t 104.015' - 101'30'= + 2O45', but that the rotation was actually found to be + l029', tha mixture having again a practically unchanged specific gravity (di"85 = 0,8402).As there is no other method of purifying the limonenes except by fractional dis- tillation, the two samples prepared by Schimmel and co. were used directly for my purpose. I am obliged to my colleague Prof. A. Klages and Mr. E. Sommer for making the measurements. Four series of determinations were made : (1) on d-limonene, (2) on Llimonene, (3) on a mixture of equal weights of both, and (4) on a mixture obtained by adding to cl-limonene so much of the I-compound that the deviation in the polarimeter became inappreciable.TABLE. a19,5' = I. d-Lirnoneiie, b. p. (corr.) 175?&-176"/763 mm. ............... +104"15' 11. I-Limonene, b. p. (corr.) 175*5-176-5"/763 mm. ............... a:9'5"= - 101"SO' 175*6-176~5"/763 mm. .......................................... ,19"= $- l"29' inactive, b. 1). (cow.) 1$5'6-176~6"/763 rnin. .............. agoo= 0"OO' 111. Mixture of equal weights 01' d- and I-lirnonene, b. p. (corr.) I V. I-Liniouene added to d-liinonene uiitil the mixture bocitiiie n. . to. d;. Ha- D. H6. HY. I. 21'0" 0'8402 1.47124 1 '47428 1'48223 1.48886 11. 20.6 0.8407 1.47157 1.47468 1'48256 1'48924 111. 20.7 0'8409 1.17143 1.47448 1 '48239 1.48904 IV.20.86 0'8402 1*%7134 1'47443 1'48231 1 *4889 8 nz - 1 _ _ _ ~ (?iJ+ 2)d Ii". I. 0-3328 11. 0.3328 111. 0.3327 I V . 0.3329 D. IIg. 0.3347 0.3395 0.3347 0-3394 0 *3345 0.3393 0-3348 0.3395 H". I. 45'26 11. 45'26 111. 45'24 I V . 45.27 Theory for C,,H,, 44.97 VOL. XCI. ~ ~ ~~~ HY. 46.71 45'51 46'-t7 45'52 46-16 46'71 45'49 46'14 46-68 45'53 46'17 46.72 45'24 - 46'40 D. HP- HY. 0.3434 0.3434 0.3433 0'3435122 DIXON AHD HAWTHORN^ : On reference to the table, it is seen that the agreement between d- and I-limonene is so excellent that the amount of hetero- geneous constituents in I-limonene must be but very small. There is further a very close accordance between the theoretical values and those observed for the molecular refraction and an almost absolute agreement as regards molecular dispersion.My numbers are also in satisfactory harmony with the figures given by Sir W. H. Perkin for dipentene; mine are a little higher for moledular refraction, and somewhat lower for molecular dispersion, The main result of these determinations is the fact that d- and Glimonene and their inactive mixture (dipentene) display, except in rotatory power, almost absolutely equd constants in every respect : in boiling point, specific gravity, refractive indices for every wave-length, specific and molecular refractions and dispersions within the whole visible spectrum. Since d- and dl-h”8(g)-p-menthndiene, prepared by Kay and Perkin,’ show remarkable differences in these physical constants, it is obvious that their relation cannot be the same as that of d- and I-limonene on the one hand and dipentene on the other. Therefore dZ-A3~*(’)-p- menthadione is probbbly a racemic compound-if the properties are tiot altered during its preparation (boiling with potassium hydrogen sulphate). At present it seems more likely that the difference i n the optical behaviour of this cll-compound is due to racemisation and not to chemical change (conversion into terpenes of another kind), as the magnetic rotation of d- and dZ-A“~8(9)-p-menthadiene dis- plays but slight differences. A decisive conclusion, however, in this very interesting question can only be arrived a t by preparing the dl-compound avoiding treatment with potassium hydrogen sulphate and by redetermination of its physical properties. HEIDELBEEG,