McCRAE : ETHYL SEC.OCTYL TARTRATE. 1103 CXIX.-Ethyl see. Octyl Tarti-ate and its Di7,cnqd ccnd Diacetyl Derivutives. By JOHN MCCRAE. DURING the course of an investigation of the monoacyl derivatives of tartaric esters (Frankland and McCrae, Trans., 1898, 73, 307), it was found that, no matter how large an excess of methyl tartrate mas taken, the action of an acid chloride on i t always led to the formation of a very large proportion of the diacyl compound. With diethyl tartrate, however, it was possible by using an excess of the ester to prevent the formation of the diacyl derivative, and so obtain the monoacyl com- pound. In order to ascertain whether i t would be easily possible to 4 ~ 21104 MCCRAE: ETHYL SEC.OCTYL TARTRATE AND ITS prepare monoacyl derivatives of a tartaric ester containing an alkyl group high in the series, the following investigation was undertaken.It was originally intended to prepare dioctyl tartrate and investigate the action of acid chlorides on it, but early in the course of the work an easy method of obtaining ethyl octyl tartrate was discovered, and, so far, only this and its diacyl products have been examined. The work is being carried further, in order to realise the original intention. The results obtained do not indicate that the introduction of the octyl group in place of the ethyl exerts any retarding influence on the production of diacyl derivatives. For the purpose of obtaining a tartaric ester with an alkyl group of high molecular weight, the most suitable alcohol to use appeared to be octyl alcohol.sec.Octy1 alcohol (methyl-n-hexylcarbinol) is a commercial article, and was adopted in the experiments. When the work mas started, the results of Marckwald and McKenzie (Be?.., 1901, 34,469) had not been published, and it was believed that this alcohol was a homogeneous material. Marckwald and McKenzie have shown that the substance is not pure, but consists of a mixture of a l~vorotatory and an inactive compound. Until the active component present in the octyl aIcohol is more definitely characterised, it would be superfluous t o enter into a discussion of the possible isomerides which may be formed from a combination of the two active com- pounds. As the octyl alcohol used throughout has been OF the same compo- sition, the results obtained for the different compounds are strictly comparable with each other (Landolt, '' Optiortl Activity and Chemical Composition," p.121). EXPERIMENTAL, The octyl alcohol was obtained from Kahlbaum, and had the same rotation as that employed by Marckwald and McKenzie, namely, aD = - 17' ( I = 4) (compare also Patterson, this vol., p. 490). Fifteen grams of dried tartaric acid mere heated in a staled tube with 60 grams of distilled octyl alcohol for 3 days a t 140-160". On distilling the mixture under diminished pressure, the excess of alcohol and the water passed over a t a comparatively low temperature, and at 280-240", under 15 mm. pressure, a substance distilled which was probably dioctyl tartrate. Much decomposition took place, and the yield was very poor. On account of the insolubility of tartaric acid in octyl alcohol, the hydrochloric acid method could not be used directly.Purdie and Marshall (Trans., 1888, 53, 391) have shown that it is possible to transform one ester into another by the action of anDIBENZOYL AND DIACETYL DERIVATIVES. 1105 alcohol, and it was thought possible that such a transformation might be effected if a tartaric ester could be obtained which is soluble in octyl alcohol. Meanwhile, Patterson and Dickinson (this vol., p. 280) had proved that a reciprocal transformation of methyl and ethyl tartrates is possible with ethyl and methyl alcohols and hydrochloric acid. This method was tried, and it has been found that after a single treatment of ethyl tartrate and octyl alcohol in proper proportions mit,h hydrochloric acid, pure ethyl octyl tartrate can be produced.Etlql Octyl Tartrate. Fifty grams of diethyl tartrate are dissolved in 200 grams of octyl alcohol and the solution saturated with dry hydrochloric acid in the cold. After standing for 24 hours, the hydrochloric acid is extracted under diminished pressure and the water and ethyl alcohol, as well as the excess of octyl alcohol, are distilled off. The residue is then dis- tilled, still under diminished pressure, and the fraction passing over between 170' and 205' under 15 mm. pressure is collected. As the boiling points of diethyl and ethyl octyl tartrates do not lie far apart, it, is not easy to separate them by fractional distillation, and the dis- tillate thus obtained constantly contains some unchanged diethyl tartrate.This is best got rid of by shaking with water, which dis- solves (and hydrolyses) the ethyl compound, but does not dissolve or decompose the ethyl octyl derivative. After washing three or four times with water, t'he oil is dried as completely as possible over sul- phuric acid in a vacuum. It is then distilled under diminished pres- sure once o r twice (b. p. 200-202° under 15 mm. pressure), or until the rotation is constant. It was generally found that the rotation did not alter after two distillations. I n one preparation, some of the excess of octyl alcohol which had been distilled off was washed. with water, then redistilled, and the rotation of the fraction boiling at 178' was found to be aD= - 18' ( I = 4), as compared with a,, = - 17' found for the original alcohol.There is here no evidence of a variation in the rate of esterification of the constituents of the alcohol, but i t must be remembered that the process was carried out very differently from the method used by Marckwald and McKenzie. Various preparations of ethyl octyl tartrate gave the same rotation, namely, +5'22' in a 66.04 mm. tube a t 16'. It is a viscous, colourless oil with a decidedly rancid odour. 0.1940 gave 0.4108 CO, and 0.1548 H,O. C=57*75 ; H=9*07. 0.1195 ,, 0.253'7 CO, ,, 0.0950 H,O. C=57.90 ; H=8*83. U1,H,,06 requires C = 57.93 ; H = 8.97 per cent.1106 McCRAE : ETHYL SEC OCTYL TARTRATE AND ITS A determination of the ethoxy-group by Zeisel's method gave the 0.1490 gave 0.1039 AgI. C,H, = 8.61. C,,H260, requires C,H, = 10.00 per cent.The density and rotatory power were then determined at various following result : temperatures, and the following numbers were obtained : d 15'/4"= 1.0664. d 30°/4'= 1.0533. d 4 5 ' / 4 O = 1.0402. d 60°/40= 1.0277. d 79*8'/4'= 1*0100. Rotation of ethyl oct9Z tartrate. (Length of polarimeter tube= 66-04 mm.) Temperature. to. 16 21 -5 41.5 50 62 71 9s Observed rotation. .______ -t 5"22' 5 29 6 27 6 56 7 23 7 41 8 21 Density at to. 1.0655 1'0608 1'0433 1.0360 1.0259 1.0177 0.9937 -I- 7-63" 7 -83 9 *36 10.13 10.89 11 *43 12.72 [ M I D . + 22.13" 22.71 27.14 29'38 31-58 33-15 36 '89 Ethyl Octyl Dibenxoylturtrate. By heating together ethyl octyl tartrate and benzoyl chloride at looo, reaction took place-rather more sluggishly than with diethyl tartrate-and from the high negative rotation of the resulting pro- duct it was evident, in analogy with the high negative rotation OF diethyl dibenzoyltartrate (Frankland and Wharton, Trans., 1896, 69, 1586), that much diacyl ester was formed even when no excess of acid chloride was used. The monoacyl diethyl tartrates were obtained in the solid state, and could be separated from the diacyl derivatives by repeated crystallisation; as no acyl derivative of ethyl octyl tartrate has been obtained in the solid state, and, further, as they seem t o de- compose on heating, so that they cannot be distilled, the task of separating the mono- from the di-acyl compound seemed hopeless, and the preparation of the monoacyl derivatives was abandoned.As has been noticed, ethyl octyl dibenzoyltartrate is constantly formed when benzoyl chloride is treated with excess of ethyl octyl tartrate, and, on the other hand, the substance is not obtained pure even when a very large excess of acid chloride is used, on account of the incompleteeess of the reaction.DIBENZOYL AND DIACETPL DERIVATIVES.1107 The method which ultimately led to its isolation was as follows: One hundred grams of benzoyl chloride were heated under a reflux condenser in a water-bath, and 36 grams of ethyl octyl tartrate were slowly dropped in with continual shaking. The heating was continued for 4 days, when no more hydrochloric acid was evolved. The product was washed with water, and then with sodium car- bonate solution. Most of the benzoyl chloride was thus removed, but as this is not easily completely destroyed by water, it was found to be necessary to dissolve the oil in ether and shake the ethereal solution with sodium carbonate solution for 3 or 4 days, occasionally renewing the carbonate solution.The ethereal solution was separated, dried as oompletely as possible over ignited potassium carbonate, and, after filtering, the ether was distilled off. The last trace of ether was expelled by warming the ester in the steam-oven, then placing it in an exsiccator, and exhausting the air. A yellowish oil was thus obtained weighing 56 grams. The rotation of this at 16.5' was found t o be - 32'16' in a 66 mm. tube. The substance is easily soluble in the common organic solvents ; it is quite insoluble in water, which appears to have no appreciable hydrolysing action on it, Heated in a vacuum, it undergoes decomposition, and benzoic acid (recognised by its melting point) was found in the distillate.For its purification, precipitation by water from an alcoholic soh- tion was adopted, The oil was dissolved in hot absolute alcohol and water was added until there was just a permanent opalescence at the high temperature. It is advisable not to boil the alcoholic solution, as an interchange of alkyl groups (ethyl for octyl) might possibly then take place, although this has not been actually observed, On cooling, the solution deposited an oil, the colour of which mas not so intense as before, The oil was dissolved in ether, and the ethereal solution dried and treated as before. The rotation of the product had then risen t o -33'20' a t 15.5' in the same tube.After repeating this process of purification, the rotation had further increased to - 34'47' at 15.5', and the next purification led to the production of a substanae with rotation - 35'55' at 11' in the 66 mm. tube, after which the rotation was not sensibly altered by further treatment, The whole of the substance in the alcoholic solution was not precipitated with water, and the various residues were combined and the whole separated by the addition of more water. The oil was extracted with ether, and after drying with ignited potassium carbon- ate and evaporating off the ether, the residue had a rotation of - 29O56' in the 66 mm. tube at 14O. This oil probably consisted for the most part of ethyl octyl dibenzoyltartrate, but whether the admixture was unchanged ethyl octyl tartrate or the monobenzoyl derivative (which presumably has a higher dextrorotation than the1108 McCRAE: ETHYL SEC.OCTYL TARTRATE AND ITS simple ester-see McCrae and Patterson, Trans., 1900, 77, 1107) was not determined.Another preparation had the same rotation. The compound has only been obtained as n clear, almost colourless, viscous oil. It was suspended in boiling liquid air and a t once set to a hard, glassy mass which melted on regaining the ordinary tempera- ture. It has been kept for more than 6 months under water, but shows no sign of solidifying, 0.2542 gave 0,6245 CO, and 0,1486 H,O. The influence of temperature on the density and on the rotation has The density observations actually made were : C = 67.00 ; H= 6.50.C28H3408 requires C = 67.47 ; H = 6.83 per cent. been determined. d 19Oj4O = 1.0884. d 6 l o / 4 O = 1.0533. d 38*5'/4O = 1.0728. d 7So/4O = 1.0396. Rotation of ethyl octyl dibennxopltat-irate. (Length of polarimeter tube=66*04 mm.) Temperature. to. 0 bserved rotation. 10.8 24'2 38 47'5 54 67 79 88 99 - 35"45' - 3 5 19 - 34 50 - 3 4 20 - 3 4 2 - 3 3 31 - 32 19 - 3 1 27 - 3 0 17 Density a t 1". 1.0956 1.0845 1 '0729 1.0649 1 '0595 1'0488 1'0386 1.0313 1'0220 - 49'41" - 49-31 - 49'16 - 48.81 - 48.64 - 48'39 - 47'12 - 46 '18 - 4 4 - a c MID. - 246.0" - 245.6 - 244.8 - 243*1 - 242.2 - 234'7 - 230 '0 - 241'0 - 223 '5 Et?hyl Octyl DiacetyZtart~*ate. Acetyl chloride reacts energetically on ethyl octyl tartrate, just as i t does on diethyl tartrate. Diethyl monoacetyl tartrate is ex- tremely difficult to separate from the diacetyl derivative, in spite of the fact that the latter can easily be obtained in the solid form (McCrae and Patterson, Zoc.cit.). As ethyl octyl diacetyltartrate has not been obtained in the solid state, it appeared at the outset a hope- less task to try t o prepare the monoacetyl derivative, and consequently no attempts have been made in this direction. To obtain the diacetyl compound, 30 grams of ethyl octyl tartrate were slowly run into 50 grams of acetyl chloride heated in a water-DIBENZOYJ, AND DIACETPL DERIVATIVES. 1109 18.5 44 61 81 *5 100 bath under a refiux condenser. The heating was continued until there was no further evolution of hydrochloric acid-which may require so long as 24 hours.The resulting product was poured into water and treated as described for the dibenzoyl compound. The oil produced in this way in one preparation had a rotation of + 2'30' in a 66 mm. tube a t 14'. For the purification, the same method (separation from alcoholic solution by means of water) has been adopted as for the dibenzoyl derivative. After three separations i t was found that the rotation was constant (aD = 2'58', 1=0*66, t = 20.5'). Another preparation had n rotation of 2'55' at 20' in t.he same tube. Ethyl octyl diacetyltartrate is a clear, viscous oil with a very slight rancid odour. It cannot be distilled, as on heating much above 100' it undergoes decomposition. 0.1772 gave 0-372s CO, and 0.1 148 H,O. The density and rotation have been determined a t various tempera- C = 57.38 ; H = 7.20.C,,H,,,O, requires C = 57-75 ; H = 8.02 per cent. tures. d 18*1'/4O = 1,0554. d 49.2'14' = 1.0271. d 31*5'/4O = 1.0429. d 70*5'/4" = 1*0071. + 2'56' 1.0553 + 4 '20" 3 14 1.0317 4 *63 3 36 1.0160 5 '37 4 14 0.9968 6'43 4 41 0.9800 7 -23 Rotation of ethJ octyl dictcetyltartrctte. (Length of polnrimeter tube = 66-04 mm.) Temperature. 1 Observed 1 Density at i to. rotation. [all). r 11 I D . ____- - + 15.71" 17 -32 20'08 24.05 27.04 Conclusions. From the results obtained with ethyl octyl tartrate, it is evident that the rotation increases with rise of temperature a t about the same rate as that of diethyl tartrate does : [al:Oo Diethyl tartrate (Pictet). ,. ... ... , . . Ethyl octyl tartrate ..................7-78 12.75 I€ we compare the rotations of ethyl octyl tartrate and its deriva- tives with those of diethpl tartrate, we notice a striking similarity, in agreement with the rule suggested by Guye (Guye and Babel, Arch. 7.66" 1399O1110 McCRAE : ETHYL SEC.OCTY L TARTRATE, Xci. phys. nat. Geneva, 1899, iii, 7, 114 ; and Guye, this vol., p. 476), namely, that substitution effected sufficiently far removed from the asymmetric carbon atom scarcely modiEes the rotatory power. Using the diethyl compounds as starting point, the following table shows that the replacement of a methylene hydrogen atom of one of the ethyl groups by n-hexyl gives rise to only a comparatively small change in the molecular rotation. [ M 1:o [MI? Diethyl tartrate" ........................+ 1 5 ~ 7 8 ~ + 27.38O Ethyl octyl tartrate .................... +22-55 +37*17 Ethyl octyl diacetyltartrate.. .......... + 15-80 + 27-04 Diethyl diacetyltartrate? ............... + 9.9 + 18-26 Diethyl dibenzoyltartratet ............ - 247.1 - 251.6 Ethyl octyl dibenzoyltartrate ......... - 245.8 - 222-8 The influence of temperature on the rotatory power of all these compounds is the same, namely, with rise of temperature the dextro- rotation increases (or the lavorotation diminishes). Frankland and Wharton (Zoc. c i t , ) have found that the laevorota- tion of diethyl dibenzoyltartrate increases with rise of temperature up to a maximum reached at about 60°, and thereafter it progressively decreases. Allowing, therefore, for this anomaly, the influence of temperature on diethyl dibenzoyltartrate is the same as on the other tartrates mentioned, although the two numbers given would appear to contradict this. No corresponding anomalous behaviour has been observed with ethyl octyl dibenzoyltartrate, but attention may be directed to the observation that the rotation of both this and the diacetyl derivative is considerably more affected by change of tempera- ture near looo than it is at low temperature. The expense incurred in the prosecution of t h i s work has been defrayed by a grant from the Research Fund of the Chemical Society, I desire here to acknowledge my indebtedness for this, and also to express my thanks to Dr. Patterson for so kindly placing his Laurent polarimeter at my disposal. THE YORKSHIRE COLLEGE, LXEDS. * Pictct, Jahresber., 1882, 856. t McCrae and Patterson (Zoe. cit.). $ Franklaud and Wharton (Zoc. cit.).