ACTION OF ALCOHOLS ON ETHEREAL SALTS. 391 X X I X - A c t i o n of Alcohols on Ethereal Salts in Presence of S m l l Quantities of Xodic Alkylate. By T. PURDIE Ph.D. B.Sc. Professor of Chemistry in the University of St. Andrews and W. MARSHALL B.Sc. CONSIDERINCI the close analogy existing between metallic and ethereal halts on the one hand and metallic hydroxides and alcohols on the other it might well he expected that the chemical change ordinarily called double decomposition so constantly exhibited by nietallic salts iind bases would also be of frequent occurrence among their organic analogues the ethereal salts and alcohols. Various isolated instances of interchange of alcoholic radicle occuri*ing between ethereal salts and alcohols have been recorded. The Frobability of a double decomposition of this kind taking place in the course of operations on ethereal salts in alcoholic solution is indeed commonly recognised and proTided f o r by employing the alcohol con-taining the same radicle as the ethereal salt operated on ; but so far as we are aware these reactions have not except in a few instances, been made the subject of special study.Priedel and Crafts (L4nt~aZen 130 198 ; 133 807) heated various mixtures of alcohols and ethers amongst others mixtures of ethylic acetate with amylic alcohol and arnplic acetate with ethylic alcohol, 2 u 398 PURDIE AND MARSHALL ACTION OF and found that in each case interchange of radicle occurred. They concluded that this interchange was not due to varying specific affinities of the alcohols but was to be attributed rather to the in-fluence of mass.Rose (Annalcn 205,240) has shown on the other hand that in the case of the ethereal carbonates the alcohol of higher molecular weight can displace almost entirely the alcohol of lower molecular weight, vhile the converse reaction takes place to only a small extent. A. Bachmann (Annnlen 218 50) heated methylic and ethylic acetates with various alcohols of the C,H2,+,0H series to the boiling point of the higher boiling substance and did not observe any action. G. Bertoni (Ber. X. 17 251) prepared several ethereal nitrites by the action of alcohols on amylic nitrite and found contrary to the experience of R6se with the carbonic ethers that the amount of decomposition effected diminished with the increasing molecular weight of the alcohol.It has been shown by one of us (Trans. 1885,862 ; 1887,627) that, in general when an ethereal salt and alcohol a,re mixed and a minute quantity of sodic alkylate is added an extensive sometimes complete interchange of alkyl radicle takes place between the salt and the alcohol at the ordinary temperature. Several instances of this peculiar ackion of sodic alkylate noted by other observers were quoted in the paper referred to ; and recently Conrad and Epstein (Ber. 20, 3057) and Peters (Ber. 20 3323) have observed that the action is so complete in the case of mixtures of acetoacetates with alcohols that it furnishes a convenient method of preparation from ethylic accto-acetates of otlier acetoacetates which are procured with difficulty by the usual methods.The results of the experiments previously communicated by one of us to this Joizrnal show as might be expected that the amount of interchange of radicle which takes place in the presence of sodic alkylate is largely influenced by the rclative proportions of ethereal salt and alcohol used but they also give some indication that the interchange is conditioned to a certain extent by the composition of the alkyl radicles concerned in the action. Thus when sodic amylic oxide was added to a mixture of ethylic oxalate and amylic alcohol the ethylic oxalate almost entirely disap-peared axid much amylic oxalate was formed whilst in the converse case of amjlic oxalate and ethylic alcohol the formation of ethylic oxalate could not be detected.Similarly the action between ethylic acetate and amylic alcohol taking into account the unavoidable loss of amylic acetate incurred during fractional distillation was apparently greater than that between arnylic acetate and ethylic alcohol. The following investigation was undertaken with the view of determinin ALCOHOLS ON ETHEREAL SALTS. 393 whether in general the more complex radicle may be regarded as possessing under the conditions of our experiments a greater affinity for the hydroxyl oxygen of the acid than the less complex radicle. Onr results lead us to the conclusion that excluding the methyl radicle such is the case with respect a t all events to acetic acid and the radicles of the CnH2Bn+l type. Our experiments were conducted as follows :-The materials were taken in the proportion of 1 mol.of ethereal salt to 1 mol. of alcohol and i$a atom of sodium. In each cast, 50 grams of the ethereal salt mere employed and this was mixed with three-fourths of the alcohol. The sodium was dissolved in the remaiii-ing fourth and the solution of the alkylate then added slowly to the mixture of alcohol and ethereal salt. After the mixture had stocci 24 hours it was subjected to fractional distillation and the distillation continued until it could be inferred from the smallness of the inter-mediate fractions that a fairly complete separation of the low and high boiling ethereal salts had been effected. The lowest or in some cases the two lowest boiling fractions containing a mixture of alcohol a i d ethereal salt were mixed with a measured quantity of standard potassic hydrate in excess ; the mixture was allowed to stand some time in a well-stoppered flask and then digested using a reflux con-denser until hydrolysis was complete.The residual alkali was determined by means of standard sulphcric acid and the weight of ethereal salt in the fraction found by calculation. In the case of the action of a low boiling alcohol on it high boiling ethereal salt the percentage of ethereal salt which had undergone exchange of radicle was calculated directly from the amount of ethereal salt found in the low boiling fractions. In the converse case, it was impossible to determine the amount of high boiling ethereal salt formed by saponifying the higher boiling fractions much sub-stance being unavoidably lost from liquid left in the flask after each distillation.I n these cases accordingly the percentage of ethereal salt acted on had to be calculated from the difference between the weight of ethereal salt used and the weight of the same found in the low boiling fractions. I n every instance a blank experiment in which no sodic alkylate was used was performed as far as possible under the same conditions as in the actual experiment. As the first two experiments may be regarded as typical of the others some details of working are given which are omitted in succeeding experiments. I t may be added that in mofit cases as iu our second experiment the weight of the low boiling fraction agreed to within about 1 gram with the weight calculated on the assumption that the numerical result obtained by hydrolysis representcd th 394 PURDIE ASL) MARSHALL ACT103 OF actual amount of interchange of alkyl radicle that had occurred and that the separation of the ethereal salt's was complete.Isoamylic Acetate and Methylic Alcohol.+ Employed 50 grams of isoamylic acetate; methylic alcohol 12.3 grams ; sodium 0.088 gram. Percentage of isoamylic acetate con-verted (1) 31.7; (2) 30.7; mean 31.2. Blank experiment 2*:3. These results were obtained by hydrolysis of the fractions boiling under 70" in experiments (1) and (2) and under 70" and 70-80" in the blank experiment after three distillations. The weights of the several fractions were as follows :-Experiment (2). Blank experiment. Under 703 .14.2 9.0 70- 80 . 0.6 2.0 80-100 . 0.5 0.8 100-120 . 1.0 0.5 Over 120 . 40.5 45.7 I n experiment (2) the total weight of the two lowest boiling frac-tions is about 2.5 grams less than it should be on the assumption that, 30.7 per cent. of the amylic acetate had been coiiverted into met,hylic acetate and it is probable therefore that the actual amount of inter-change that occurred was several per cent. higher than that found, the error being due to imperfect separation of the methylic acetate from the higher boiling fractions and to loss by evaporation. It was found that in distilling 50 grams of methylic acetate an average of 0.3 gram of substance was lost during each distillation. Methylic Acetate nnd Isoani ylic Alcohol. Employed methylic acetate 50 grams ; isoamylic alcohol 59.4 grams ; sodium 0.155 gram.Percentage of methylic acetate converted (I) 52.2 ; (2) 51.2 ; mean 51.7. The fractions saponified were those boiling under 70" and 70-80" a,fter three distillations weighing in (2) 32.5 and 2.5 grams respec-tively. Supposing in this experiment 51.2 per cent. of the methylic acetate to have been converted into amylic acetat,e the mixture of residual methylic acetate with the methylic alcohol formed should weigh 35.5 grams which agrees fairly well with the weight actually found namely 35 grams. The corresponding fractions in the blank experiment weighed 48.2 and 0.8 gram. Blank experiment 6.8. * The amjl compounds used are from ordinary fermentation amplic alcohol ALCOHOLS ON ETHEREAL SALTS. 395 Isoamylic Acetate and E t h y l i c Alcohol.Employed isoamylic acetate 50 grams ; ethylic alcohol 17.7 grams; Percentage of isoamylic acetate converted 25.6. These numbers were obtained by sodium 0.088 gram. Blank experiment 4.1 per cent. hydrolysis of the fractions boiling under 80" after six distillations. Etlzylic Acetate and Isoaw ylic Alcohol. Employed ethylic acetate 50 grams ; isoamylic alcohol 50 grams ; sodium 0.13 gram. Percentage of ethylic acetate converted (1) 62.0 ; (2) 61.4 ; mean 61.7. Blank experiment 8.8. These numbers were obtained by hydrolysis of the fractions boiling in (1) and (2) under 80" after five distillations and in the blank experiment under 85" and 85-90' after six distillations. Isobutylic Acetate and MethyEic Alcohol. Employed isobutylic acetate 50 grams; methylic alcohol 13.8 grams; sodium 0.099 gram.Percentage of isobutylic acetate converted, 42.1. Blank experiment 8.8. The fractions saponified were those boiling xnder 70" and 70-80" after the distillations had been repeated four times. Methylic Acetate and Isobutylic Alcohol. Employed methylic acetate 50 grams ; isobutylic alcohol 50 grams ; sodium 0.155 gram. Percentage of methylic acetate converted (1) 45.6; (2) 44.6; mean 45.1. Blank experiment, 6.2. The same fractions as in the last experiment were saponified. Isobwtylic Acetate and E t h y l i c Alcohol. Employed isobutylic acetate 50 grams ; ethylic alcohol 19.8 grams ; sodium 0.099 gram. Percentage of isobutylic acetate Converted 29.3 ; in blank experiment 9.1. The fractions that were saponified were those boiling under $OD 80-85" and 85-90' after six distillations.The separation of the ethereal salt was probably less complete than in the previous experiments. Ethylic Acetate and Isobutylic Alcohol. Employed ethylic acetate 50 grams ; isobutylic alcohol 42 grams ; sodium 0.13 gram. Percentage of ethylic acetate converted 59.4 ; in the blank experiment 16.1. The mixtures were distilled six times, and the fractions saponified were those boiling under 85" and 85-90' 396 PURDIE ASI) MARSHALTJ ACTION OF The separation of the residual ethylic acetate was not very satisfactory, and the numbers found are probably several per cent. too high. Experiments similar to those described were undertaken in order to determine the amount of interaction between amylic acetate and propylic alcohol and between propylic acetate and amylic alcohol, but from consideration of the weights of the lower boiling fractions and the results obtained by their hydrolysis it was evident in these cases that tbe separation of the ethereal salts was very imperfect.With amylic acetate and propylic alcohol the percentage numbers obtained were as follows .-Experiment with sodic alkylate 20.9 ; blank experiment 4.3 ; and with propylic acetate and amylic alcohol -experiment with sodic alkylate 37 ; blank experiment 32.3. An experiment was also performed with methylic acetate tertiary butylic alcohol and sodic butyl oxide. The formation of butylic acetate could not be detected. The form in which the sodic alkylate is added does not seem to have much influence on the extent of the reaction.Thus the amount of interchange of radicle occurring between methylic acetate and amylic alcohol is much the same whether the sodium be added in the form of amyl oxide or methoxide. Interchange of radicle between alcohol and ethereal salt is induced also by the addition of zinc chloride thoEgh this substance is by no means so active as sodic alky late. I n all the reactions we have examined except that of propylic acetate and isoamylic alcohol the result of which is not to be relied on the influence of the sodic alkylate in inducing the interchange of alkyl radicle between the ethereal salt and the alcohol is very marked. It appears further that the extent of the interchange which occurs is not determined solely by the proportions in which ethereal salt and alcohol are present but is influenced by the specific affinities of the alkyi radicles concerned.This is apparent from inspection of the following table in which the experiments are arranged in pairs so that the result of each experiment may be compared with that of its converse. I n column A are given the percentages of interchange occurring when sodic alkylate is used; in column B the numbers obtained by subtracting from these the percentages found in the corresponding blank experiments. The numbers in column B repre-sent probably more accurately at all events in some cases the rela-tive amounts of interchange of alkyl radicle that take place as by subtraction of the apparent amount of interchange occurring in tlie blank experiment a rough correction is applied for errors due t o loss by evaporation imperfect separation &c ALCOHOLS OX ETHEREAL SALTS.397 Isoamylic acetate and methylic alcohol . . Methylic acetate and isoamylic alcohol . . . Isoamylic acetate and ethylic alcohol . . . . Ethylic acetate and isoamylic alcohol . . . . Isobutylic acetate and methylic alcohol . . Methylic acetate and isobutylic alcohol. . . Isobutylic acetate and ethylic alcohol . . . . Ethylic acetate and isobutylic alcohol . . . . { { A. 31 '3 5 7 25.6 61.7 42.1 45.1 29.3 59.4 B. 28.9 44.9 21.5 59.9 33.3 38.9 20.2 43.3 It is evident that a much greater interchange of alkyl radicle occurs when the alcohol of coniplex radicle acts on the ethereal salt of simple radicle than in the converse case except in the isobutyl-methyl reactions in which the extent of the interchange is about the same and this holds true whether the numbers in column A or I3 are considered.Experiments involving the separation of the ingredients of a com-plex mixture of liquids by fractional distillatiou are not capable of exact quantitative interpretation and we are aware that the differ-ences of extent of reaction shown by our numerical results may in some cases be accounted for by experimental error. We think never-theless we are justified in drawing the conclusion that excluding the methyl radicle the affinity of the alkyl radicle for the hydroxyl oxygen of the acid as measured by the special reaction under con-sideration increases with increasing complexity of the alkyl radicle, bnt that with regard to the methyl radicle its affinity is greater than that of the ethyl and less than that of the amjl radicle.The grounds of this conclusion will be made more evident by arranging the numerical results already given so as to show in juxtaposition the amounts of interchange which takes place when (1st) an alcohol acts on two different acetates and (2nd) two different alcohols act on one and the same acetate. 1st Isoamylic acetate and metbylic alcohol . . . { Isobut,ylic acetate and methylic alcohol . . Isoamylic acetate and ethylic alcohol . . . . Isobutylic acetate and ethylic alcohol . . . . Ethylic acetate and isobutylic alcohol . . . . Methylic acetate and isobutylic alcohol .. Ethylic acetate and isoamylic alcohol . . . . Methylic acetate and isoamylic alcohol . . A. 31.2 42.1 25-43 29.5 59.4 45.1 61-7 51-7 B. 28.- 9 33.3 21.5 20.2 43.3 38.9 52.9 44.9 2nd Isoamylic alcohol and methylic acetate . . 51.7 44.9 33.9 { Isobutylic alcohol and methylic acetate . . . 45. 398 COULDRIDGE SOME INTERAUTIOXS OF A. B. 2nd Isoamylic alcohol and ethylic acetate . . . . . 61.7 52.9 43.3 Ethylic alcohol and isobutylic acetate . . . . 20.2 Methylic alcohol and isobutylic acetate. . . 42.1 33.3 Ethylic alcoliol and isoamylic acetate . . . . 21-5 28.9 { Isobutylic alcohol and ethylic acetate . . . . 59.4 29.3 25.6 { Methylic alcohol and isoamylic acetate . . . 31.2 { The only numbers which do not accord with the conclusion stated above are those given in column B for the action of ethylic alcohol on isoamylic and isobutylic acetates. The reaction under consideration is a complex one and its course is no doubt influenced by the relative stability of the different sodic alkylates the etherificatiori values of the alcohols and other factors. Menschutkin has shown that rnethylic alcohol holds a distinctive position among its homobgues in respect of power of etherification. A similar though not identical peculiarity is also emphasised in the reaction we have studied. The power possessed by alcohols to effect interchange of alkyl radicle in presence of sodic alkylate would seem to be associated however with the limits and not with the initial velocity of etherification as in correspondence with the former, excluding methylic alcohol this power increases with increasing molecular weight of the alcohol. It may be added that our con-clusions accord with the observations of Rose and of Peters already alluded to