THE CONSTITUENTS OF AN ESSENTIAL OIL OF RUE. 1585 CLVIIL-The Coiistituents of un Essential Oil of Rice. By FREDERICK B. POWER and FREDERIC H. LEES. THE incentive to this investigation arose from the requirement of preparing some methyl nonyl ketone, the most convenient source of which is the essential oilof rue. A portion of the oil was therefore obtained from an English firm of distillers, and although there was no reason to doubt its purity, i t v a s found to possess some decidedly abnormal characters. As i t bore the designation 01. Rutae Ang., we inferred that not only had the oil been distilled in Britain, but possibly native-grown herb had been used, inasmuch fts the ordinary commer- cial oil appears to be always distilled from fresh material. This afforded a possible explanation of its deviation from the accepted standards of the ordinary oil, and, under these circumstances, i t seemed to merit a more complete investigation. Soon after beginning the work, we observed a brief preliminary notice by Soden and Henle (Phnrm.Zeit., 1901, 277) of an Algerian oil of rue, which was sub- sequently further examined by them (ibid., 1901, 1026). As the oil supplied t o us, on further examination, was found to resemble, in some of its physical and chemical characters, that described by Soden and Henle, we sought information as t o its origin. It was thus ascertained that it had not been distilled from native-grown herb, and as it appears that the rue has never been cultivated in this country for the purpose of distillation, it is highly probable that thesrticle supplied t o us was an Algerian oil.VOL. LXXXI 5 01586 POWER AND LEES: Thoms (Ber. deut. pharm. Ces., 1901, 10, 3) has shown that, besides the chief constituent and the only one which had hitherto been posi- tively identified, namely, methyl nonyl ketone, ordinary oil of rue contains about 5 per cent. of methyl It-heptyl ketone, together with some free fatty acids and a very small amount of a phenol-like sub- stance, from which crystals separated which melted a t 155--156O, but were not further characterised. Soden and Henle (Zoc. cit), in their investigation of the Algerian oil, found that while, like the ordinary European oil, it contains about 90 per cent. of ketones, there is about twice as much methyl heptyl ketone as methyl nonyl ketone. They have also indicated that, ‘‘besides these ketones, the Algerian oil contains a not inconsiderable amount of esters which boil chiefly at 200-210°, and appear to consist, for the most part, of the acetic esters of the secondary alcohols correspond- ing to the two ketones, or of similar aliphatic alcohols, but that they were not yet able to determine the true nature of these esters on account of insufficient material.” I n view of the fact that our investigation of what was assumed to be an English oil had progressed to a considerable extent before the probability of its Algerian origin became apparent, and that we have not only identified the alcohols but also several minor con- stituents, it seemed desirable that the results of the investigation should be recorded. EXP E RIME NT A L.The oil was light yellow in colour ; it had a density d 15*5O/16*= =0-8405, and a,, -3’48‘ in a 100 mm. tube. It was completely soluble in 2 parts of 70 per cent. alcohol. The total amount of oil employed was 1800 grams. Treatment with 30 per cent. Sulphuric Acid.-As Schimmel & Co. (Semi-annwd Report, Oct., 1901, 46) have obtained from ordinary oil of rue a small quantity of a basic oil, possessing a blue fluorescence, which they regarded as methyl methylanthranilate, it seemed desir- able to ascertain whether our oil contained a similar substance. It was therefore extracted with 20 per cent. sulphuric acid, and this liquid made alkaline with sodium carbonate and shaken out with ether. From the ethereal liquid about 0.5 gram of a light brown syrup was obtained, which had an odour suggestive of quinoline or its homologues; it was alkaline to litmus, and caused white fumes in contact with hydrochloric acid.On standing for some time, a few crystals were formed, but were too small in amount to admit of separ- ation. The substance, when dissolved in hydrochloric acid, gave with auric chloride a brownish precipitate, which almost immediatelyTHE CONSTITUENTS OF AN ESSENTIAL OIL OF RUE. 1587 became resinous. It was evident that the basic substance in this oil was not methyl methylanthranilate. Treatment with Sodium Carbonate.-In order t o separate any un- combined fatty acids, the oil was subsequently extracted with a 5 per cent. solution of sodium carbonate. The alkaline liquid was concentrated? acidified with sulphuric acid, and distilled with steam.The distillate, which contained oily drops, was shaken out with ether, and from the ethereal liquid a small amount of a light brown oil was obtained. This, on distillation, gave two fractions : (I) below 200°, and (2) 220-270O. These were converted into silver salts of which one salt (a) was pre- pared from fraction (I), and two salts ( b ) and (c), from fraction (2). They were respectively well washed, dried, and analysed : (a) 0.0601 gave 0.0262 Ag. Ag = 43.5 per cent. (b) 0.1004 ,, 0.0390 Ag. Ag=38.8 ,, (c) 0 0948 ,, 0.0400 Ag. Ag=42.2 ,, It was evident from these results that the free fatty acids con- tained in the oil consist of a mixture, the constituents of which were too small in amount t o admit of identification. Idemtzycation of Hethyl Salicylute.Treatment with Potassium Hydroxide.-After the treatment with sodium carbonate, the oil was extracted with a 5 per cent. solution of potassium hydroxide. The alkaline liquid was acidified with sulphuric acid and extracted with ether. The ethereal liquid afforded a small amount of a dark brown oil, which had the characteristic odour of methyl salicylate, and on standing a short time formed a crystalline paste. This was spread on a porous tile, when the oily mother liquor was removed and a small quantity of white crystals obtained, which, after recrystallisation from hot water, melted at 155-156', gave a deep violet colour with ferric chloride, and were identified as salicylic acid.* On analysis : 0,1304 gave 0.2899 CO, and 0.0528 H,O.C = 60.6 ; H = 4.5. C,H608 requires C = 60.9 ; H = 4.4 per cent. From the porous tile, by extraction with warm ether, a small amount of an oily liquid was obtained which distilled between 190' and 250' This, on hydrolysis, yielded a further amount of salicylic acid. As the original oil had been previously extracted with sodium car- bonate, the salicylic acid could not have existed in it as such, but must * The substance obtained by Thoms (Zoc. cit.), which melted a t 155-156", wonld also appear to have been salicylic acid. 5 0 2I588 POWER AND LEES: have been present as an ester which had become partially hydrolysed in the process of isolation. It was evident that this ester was methyl salicylate. Ketonic Constituents of the Oil.Treatment with h'odium Bisulphite.-The ;whole of the oil, after the treatment just described, was diluted with ether and then repeatedly shaken with a saturated solution of sodium bisulphite. The crystal- line compound was well washed with ether, drained, and pressed. The combined ethereal filtrate and washings were first shaken wibh a solu- tion of sodium carbonate t o remove sulphurous acid, then with water, finalIy dried with calcium chloride, and the ether removed. The residual oil, thus deprived of the larger portion of its ketonic conetitu- ents, 'was set aside for subsequent examination. The bisulphite compound was suspended in ether and decomposed with a 10 per cent. solution of sodium hydroxide. The ethereal solu- tion was washed with water, dried with calcium chloride, and the ether removed.The residual oil was then fractionally distilled. After pro- longed fractionation, it was resolved chiefly into two large fractions, one distilling at 190-200' and the other at 225-235'. Identification of Methyl n-Heptyl Ketone. Fraction 1 90-200°.-On redistilling this fraction, the greater part boiled a t 194-197'; it was most constant at 1945-195.5' under 763 mm. pressure and this portion had a density d 14°/160==0*8296. On analysis : 0,0934 gave 0:2617 CO, and 0-1080 H,O. The semicarbazone melted at 119--120*. 0.1272 gave 0.2790 CO, and 0.1214 H,O. C = 76.4 ; H= 12.8. C9H,,0 requires C = 76.1 ; H = 12.7 per cent. On analysis : C = 59.8 ; H = 10.6. C,,,H,,ON, requires C = 60.3 ; H = 10.6 per cent. Identification of Methyl n-Nonyl Ketone.Traction 225--235'.-0n redistilling this fraction, i t boiled for the The semicarbazone melted at 122O. 0.1376 gave 0.3195 00, and 0.1358 H,O. C,,H,,ON, requires C = 63.4 ; H = 11.0 per cent. The boiling point of methyl nonyl ketone is generally stated t o be 224' ; Thoms (Zoc. cit.) likewise records it as 223-224' under 774 mm. most part at 229-233' under 759'mm. pressure. On analysis : C = 63.3 ; H = 11.0.THE CONSTITUENTS OF AN ESSENTIAL OIL OF RUE. 1589 pressure, whereas Soden and Henle (Zoc. cit.) have indicated it to be 230-231' under 740 mm. pressure. We have confirmed this observa- tion by regenerating the ketone from the repeatedly crystallised semi- carbazone ; it then distilled completely at 231.5-232*5' under 761 mm.pressure, and had a density d 20-5°/160 = 0.8263. On analysis : 0.0845 gave 0.2398 GO, and 0.0994 H,O. C = 77.4 ; H= 13.1. C,,H,,O requires C = 77.6 ; H = 12.9 per cent. Non-htonic Constituents of the Oil. The portion of the original oil which did not combine with sodium bisulphite was subjected to fractional distillation. Under the ordinary pressure, a portion distilled over below 207' ; this on further distillation afforded the following fractions : 120-145', 145-165', 165-170', 170-180', and 180-185'. The portion distilling above 185' was finally mixed with that which boiled above 207" in the first instance. Tractions 120-145', 145-165', and 165-170°.-These fractions were examined as follows : Rraction 120-145°.-This had a density d 16'/16' = 0.8416 and was optically inactive.On analysis : 0,0932 gave 0.2516 CO, and 0.0987 H20. C= 73.6 ; H = 11.8. Rmction 145-1 65'.-This had a density d 16O/ 16' = 0.852 and On analysis : aD -2'45' in a 25 mm. tube. 0,098 gave 0.282 CO, and 0.106 H,O. C = 78.5 ; H = 12.0. These two fractions had a combined weight of not more than They possessed an odour of terpene and also one recalling 5 grams. that of ethyl valerate. Ident$cation of cb Vderic Estey. The above two fractions were mixed and boiled for some hours with a dilute aqueous solution of potassium hydroxide. There still floated on the surface of the liquid a considerable proportion of the oil, which was taken up with ether and examined in connection with the fraction boiling at 165-170'. The alkaline liquid was acidified, and again extracted with ether, which dissolved a small amount of a fatty acid from which a silver salt was prepared.This by recrystallisation from hot water was separated into three fractions, which were analysed : (1) 0.0638 gave 0.0324 Ag. Ag = 50.8. (2) 0.0558 ,, 0.0287 Ag. Ag=51*4. It was thus evident that the oil contained an exceedingly small C,H:,O,Ag requires Ag = 51.7 per cent. amount of a valeric ester, which was probably ethyl valerate.1590 POWER AND LEES: Iclentijcation of Pinene. The oil separated by the hydrolysis of the two preceding fractions, 120-145' and 145--165O, was mixed with the small fraction 165-1'70O. The total liquid, which amounted to 3 grams, evidently contained a terpene, and from it a small amount of a nitrosochloride was prepared, The latter was converted into the nitrolepiperidide which, after recrystallisation from methyl alcohol, melted at Identijcation of Limonene and Cineol.11 9-1 20'. Practions 170--1 80' and 180--185°.--These two fractions weighed Traction 170-18O0.-This had a density d 15*5'/16O = 0.8606 and (1) 0,1306 gave 0.3987 CO, and 0.1405 H,O. Fraction 180-185°.-This had a density d 15*5O/16'= 0.8588 and uD - 10'30' in a 25 mm. tube. Both these fractions were insoluble in 70 per cent. alcohol, and were otherwise practically identical. Their general characters indi- cated that they contained a considerable proportion of a terpene, but they also had an odour suggesting the presence of cineol. The presence of limonene in the two preceding Fractions was deter- mined by means of the tetrabromide, which melted at 103'; the high laevorotation of the fractions made it evident that it mas I-limonene." The fraction 170-180" was specially tested for phellandrene, but with a negative result.The identification of cineol in these two fractions was effected by means of its compound, C,,H,,O C,I,NH, with t et raiodopyrrole (Hirschsohn, Pharm. Zeit. RUSS., 1893, 32, 49, 67), and also by the additive product with hydrogen bromide. The former melted with decomposition at 114-115', and the latter a t 55-56'. On warming the hydrobromide with dilute alkali, it developed the unmistakable odour of cineol. respectively 11 and 5 grams, and had the following characters : uD -20'16' in a 50 mm. tube. On analysis : C = 83.3 ; H = 12.0. (2) 0,1238 ,, 0.3780 CO, ,, 0.1340 H,O.C=83*3; H=12*0. Portion of Oil boiling above 185O.-This was hydrolysed with alcoholic potassium hydroxide.' The alcohol was then removed, water added, * It is stated in " Die aetherischen Oele," by Gildemeister and Hoffmann, p. 596, that ( ' the terpene found in oil of rue by different investigators is in all cases to be referred to an adulteration with oil of turpentine," and Thorns (Zoc. cit.) also states that it contains no terpene. No constituent of the genuine oil has, however, hitherto been isolated which would account for its optical activity, and the possi- bility of the presence of a small amount of terpene would therefore not seem to be excluded.THE CONSTITUENTS OF AN ESSENTIAL OIL O F RUE. 1591 and the separated oil taken up with ether.The ethereal solution was washed, dried quickly with calcium chloride, and the ether removed. The alkaline liquid was acidified, distilled with steam, and the dis- tillate neutralised with sodium carbonate, concentrated, acidified, and extracted with ether. From the ethereal solution a n appreciable quantity of a light yellow, pungent liquid was obtained. This distilled almost completely below 120°, leaving a very small amount of an acid of higher boiling point in the flask. Ident$cation of Acetic Acid. The portion of acid which distilled below 120' was converted into its barium salt, and from a portion of this a silver salt was prepared. On analysis : 0.2208 gave 0,1426 Ag. Ag= 64.6. C,H,O,Ag requires Ag = 64.7 per cent. The higher boiling portion of volatile acid was dissolved in a little dilute aqueous ammonia, decolorised by animal charcoal, and converted into a silver salt, which was for the most part insoluble in hot water.On analysis : 0.0704 gave 0.0304 Ag. The small amount of this acid did not permit of its identification. Ag= 43.2 per cent. 5% Alcohols. The oil obtained by the above hydrolysis, when tested with semicarb- azide, was found still to contain a considerable amount of ketone, not- withstanding the previous exhaustive treatment with bisulphite. A preliminary distillation showed that the greater part passed over be- tween 190' and 250'; that boiling above 250', which will be subse- quently referred to, possessed a beautiful indigo-blue colour. The oil was therefore fractionally distilled until that passing over below 250' was quite free from any blue colour.I n order to separate from this oil the last portion of ketones, recourse was had to the following method. 20.7 Grams of sodium were dissolved in 250 C.C. of absolute alcohol, and 100 grams of semicarbazide hydrochloride, made into a paste with alcohol, were introduced. An alcoholic solution of semicarbazide was thus obtained, sufficient in amount t o combine with the 128 grams of oil employed, had the latter consisted entirely of methyl heptyl ketone. This mixture was boiled for half an hour, and then, while still hot, filtered from the sodium chloride. On concentrating the filtrate, a considerable amount of the crystalline semicarbazones separated, which1.592 POWER AND LEES: was removed by means of a pump, and washed well with cold alcohol. On the addition of water to the filtrate, there separated the substances which had not combined with the semicarbazide, together with a small amount of the semicarbazones.These were taken up with ether, which was subsequently removed, and the oily residue rapidly dis- tilled in steam, The non-ketonic substances quickly came over, and i t was only after these had distilled that the semicarbazones began slowly to decompose, The oil was extracted with ether, dried with potassium carbonate, and, after the removal of the ether, fractionally distilled. The principal fractions were collected at 1 90-200' and 220-235', and of these the latter was much smaller than the former. Smaller fractions were also collected a t 200-205', 205-210', and 2 10-220'.IdentiJication of Nethyl -n-hept ylcadinol. Fraction 1 90-200°, on redistillation, yielded a fraction boiling con- (1) 0.0931 gave 0.2605 GO, and 0.1 130 H,O. Fvaction 200-205°.--'L'his was also analysed : 0.1007 gave 0.2780 CO, and 0.1258 H,O. As the principal fraction was evidently not quite pure, the separa- tion of any trace of terpene it might still contain was effected by converting the alcohol into its acetic ester. Por this purpose, the two fractions were mixed and boiled with an equal weight of acetic anhydride and a little anhydrous sodium acetate. A product was thus obtained which distilled chiefly at 212-217', and most constantly at 213-215'. This was analysed : stantly at 195-196'. On analysis : C = 76.3 ; H = 13.5.(2) 0.1048 ,, 0.2931 CO, ,, 0.1285 H20. C=76.3; H=13*6. C = 75.3 ; H= 13.9. C,H,,O requires C = 75.0 ; H = 13.9 per cent. 0.1210 gave 0.3174 GO, and 0.1300 .H20. C,,H,,O, requires C = 71*0 ; H = 11.8 per cent. Methyl-n-heptylcarbinyl ucetate, CH,*CH(O*C,H,0)*C7H15, was a colourless oil of pleasant odour, and had a density d 20m5*/16*= 0.8605 and aD - 3'3' in a 50 mm. tube. Hydrolysis afforded methyl-n-heptyl carbinol, which boiled at 198-200' under 765 mm. pressure, and had a density d 19'/16' = 0.8273 and aD - 3'44' in a 50 mm. tube. 0.1223 gave 0.3348 GO, and 0,1529 H,O. On oxidation with chromic acid, this optically active alcohol afforded C=71*5; H=11*9. On analysis : C= 74.7 ; H= 13.9. C,H,,O requires C = 75.0 ; H = 13.9 per cent.THE CONSTITUENTS OF AN ESSENTIAL OIL OF RUE.1593 methyl n-heptgl ketone, which had the characteristic odour, distilled a t 195-199O, and yielded a semicarbazone melting at 118-1 1 9'. The proof was thus afforded that this alcohol is methyl-a-heptylcarbinol. IdentiJccction of Methy I-n-non ylcarbinol. The fraction boiling a t 220-235' was relatively very small in amount. It was acetylated, as in the case of the alcohol of lower boiling point. The product, on fractionation, distilled principally between 240° and 248', and was most constant at 245'. Hydrolysis afforded the alcohol, which boiled a t 230-235', and was most constant at 231-233't; i t had ccD - 1'18' in a 25 mm. tube, This fraction was analysed : 0.0991 gave 0,2793 GO, and 9.1220 H20. This optically acbive alcohol afforded, on oxidation with chromic It was C = 76.9 ; H= 13.7.C,,H,,O requires C = 76.7 ; H = 14.0 per cent. acid, a ketone, which formed an oxime melting a t 46-47'. thus identified as methyl-n-nonylcar binol. A Blue Oil. It has already been noted that from that portion of the oil used for the identification of the alcohols a fraction was separated, boiling above 250C; which possessed a deep blue colour. I t was relatively small in amount, and had a rank, weed-like odour. On redistillation, it passed over between 250' and 320°, but was at no point constant. Theliquid was dichroic, the colour, in r e d ~ t e d ~ l i g h t , being a deep indigo-blue? and in transmitted light, red. It was sparingly soluble in 70 per cent. alcohol. When dissolved in glacial acetic acid, it gave a reddish- brown colour on the addition of a drop of hydrochloric acid.The presence of a blue oil in a n oil of rue has not hitherto been observed. The small amount, and the complex character of this liquid, rendered it impossible t o determine its composition. sumrnury . From the preceding details it will be seen that this essential oil of rue, which mas apparently of Algerian origin, contained the following substances : 1. Methyl-n-hegtylketone. 2. Methyl-n-nonylketone. 3. Methyl-n-heptylcarbinol . 4. Methyl-m-nonylcar binol. 5. A blue oil of high and inconstant boiling point.1594 LEES : METEYL P-METHYLHEXYL KETONE. 6. Acetic acid, in combination with the alcohols. 7. A basic substance, having the odour of quinoline. 8. A mixture of free fatty acids. 9. Methyl salicylate. 10. Ester of valeric acid, apparently ethyl valerate. 11. Pinene. 12. I-Limonene. 13. Cineol. The relative proportion of the above constituents is approximately as follows : The two ketones represented 80 per cent. of the oil, and were present in about equal amounts. The two alcohols represented about 10 per cent., and were present partly in the uncombined state and partly as acetic esters. The methyl-wheptylcarbinol preponderated. The two terpenes, together with cineol, represented about 1 per cent. of the oil. There was very little pinene, and the amounts of limonene and cineol were about equal. The amount of blue oil was about 0.5 per cent., and finally there was separated from the non-ketonic portion of the oil a small amount of undistillable, viscous substance, which was probably a decomposition product. The fact that the substances associated with the ketones and alcohols are present in such small amount, and that the limonene is the more rarely occurring I-form, enables us to conclude that all these substances are natural constituents of the oil. THE WXLLCOME CHEMICAL RESEARCH LABORATORIES.