STUDIES OF DYNAMIC ISODIXBISM. PART XX. 2883 CCCXCIX.-c?tudies of Dynamic Isomerh. Part X X . Amphoteric Solvents as Catalysts for the Muturota-tion of the Sugars. By THOU h m L O ~ Y and I~VINE JOHN FAULKNEB. 1. h the preceding paper of this series (this vol. p. 1385) evidence waa adduced to show that the mutarohtion of the sugars is not due to a mere spontaneous tautomeric change involving only the molecules of the sugar itself but to a molecular rearrangement in FIG. 1. FIG. 2. Mutorotation of tetrmnedhylglucoee (a) in @ine and water @ (b) in @ne and creaol k. Muturotatwn of glucose in pydine andwrrter. 100 80 60 40 20 0100 80 60 40 20 0 "/b P y r idine . which other constituents of the solution also play an essential part. Of the constituents which interact catalytically with the sugar to bring about this rearrangement water is the most im-portant since it waa already obvious in 1904 from Irvine's work on the methylated sugars that the mutarotation is retarded enorm-ously and in some cases may be almost stopped by working in non-aqueous solvents.The opinion was therefore expressed (J., 85 1566) that all the data then available could be interpreted by attributing the mutarotation to a process of " hydrocatalysis " of the sugar. At that time only one assertion could be made in reference ta 5 E 2884 LOWRY AND FAULKNER : the mechanism of this hydrocatalysis namely that since glucose hydrate exhibits just the same mutarotation as glucose itself the change of structure which causes the changes of rotatory power could not take place during the initial hydration of the sugar, but must occur a t some subsequent stage e.g.by a molecular rearrangement or isomeric change in the hydrated sugar. A further consideration of the mechanism of mutarotation (this VO~., p. 1371) especially in comparison with the mechanism of hydrolysb, led however to a much more definite suggestion namely that water probably acted as a catalyst for mutarotation in virtue of the fact that it possesses both acidic and basic properties and that in general the essential condition for the mutarotation of a sugar is that the solvent should possess amphoteric properties. This conclusion has the advantage of providing an explanation of the well-known fact that acids and bases produce even higher velocities of mutarotation than water alone and of the further fact (which was discovered in the come of our own experiments) that pyridine which is a powerful catalyst in the presence of water, is not a catalyst at all for the mutarotation of tetramethyl glucose when used in the absence of water.2. The experiments described in the present paper had their origin in the conception of the mechanism of mutarotation which is outlined above. The most important results of the experiments are to show : (i) That pyridine which is inactive when dry gives when mixed with about twice its weight of water a maximum velocity of mutarotation which is about twenty times as great as the velocity for solutions of the same sugars (glucose and tetramethyl glucose) in pure water.(ii) That cresol like pyridine has no appreciable catalytic properties when water is not present showing that acids and bases may alike be rendered ineffective if used in the absence of water. (iii) That mixtures of cresol and pyridine each sufficiently dry to give only a negligible velocity of mutarotation when mixed in the proportion of about 2 parts of cresol to 1 part of pyridine are again about twenty times as active as water in promoting the mutarotation of tetramethyl glucose. It is of course possible to argue that perfectly dry pyridine and perfectly dry cresol would be incapable of producing muta-rotation; but this criticism even if it could be vindicated by experiments on the effect of extreme drymg would not affect the present discussion since it is clear that n mixture of pyridin STUDIES OF DYNAMIC ISOMERISM.PART XX. 2885 and crew1 is able to do what neither solvent can do by itself namely, to give a velocity of mutarotation which is far greater than in a solvent containing 100% of water. In other words if a trace of water is e-ssential (as it may be even in order to enable the acid and base to interact to form an unstable salt),* the intense activity of the mixed solvent is due to its pronounced amphoteric character, and not to a direct catalytic action upon the sugar of any minute trace of water which it may contain. 3. It is interesting to notice the theoretical conclusions which follow from the experimental facts thus established namely (i) that (just as in the case of nitrocamphor) it is impossible for a profon to wander directly from one position to another within the molecule of the sugar and (ii) that in order to effect this transfer it is neces-sary to provide a medium into which a proton can escape from the sugar and from which a proton can be provided to replace the proton thus lost by the sugar.These conclusions are supported by the fact that the only sub-stances which are now known to act as catalysts for the muta-rotation of the sugars are those which possess either acidic or basic properties or both. Oxygenated solvents such as acetone and ethyl acetate even when used in presence of water retard rather than accelerate the mutarotation; and even their " polar " character and relatively high dielectric constants do not enable them to develop any catalytic activity.The mechaniam of muta-rotation suggested above which involves an interchange of protons between the sugar and the medium (just as in the interaction of an acid and a base) appears therefore to be the only one which is in accord with the experimental facts and no alternative mechanism is yet known which can replace it. E X P E R I M E N T A L . 4. The sugars were the same samples as those used in Part ,YILi. The solvents used were as follows (i) Air-free distilled water. (ii) Pyridine (B.D.H. "Extra Pure") shaken twice with quick-lime for 24 hours and distilled twice shaken with barium oxide and twice fractionated giving about 800 C.C. boiling from 114-8" to 115.2" (corr.) from about 1 litre. The sample gave a residual velocity of mutarotation of tetramethyl glucose not greater than O-OOO4 in minute-units.(iii) o-Cresol (B.D.H. " for cineol-deter-minations ") from R sealed flask was distilled immediately before use; when liquefied by the addition of a little benzene and used * The melting-point curve for pyridine (m. p. -41') and o-cresol (m. p. 30") rises to a maximum at -+-lo in an equimolecular mixture of the acid and base (Bramlcy J. 1916 109 476). 5e* 2886 STODlES OF DYXAMIO ISOMERISM. P U T XX. as a solvent for tetramethyl glucose it gave an arrest of mutra-rotation during a period of 6 hours followed by a slow change with a velocity coefficient of about 0.0005. m-Cresol (B.D.H., specially prepared) was redistilled and gave a velocity coefficient of 0-0003 only. 5.Muturotatkm-curves were plotted as follows One gram of sugar waa weighed out in a 20 C.C. flask and the mixed solvent, previously made up by weight in another vessel was added at zero time. The solution was transferred quickly to an air-dried polarimeter tube (2 dcm.) and readings were taken as soon as possible namely about 2 minutes from zero time in the case of water and pyridine but about 5 minutes in the case of solutions rich in cresol which were much more viscous. 6. The experimental results are summarised in Tables I 11, TABLE I. TABLE 11. TABLE 111. Mutarotation of Tetra-methyl Glucose in Mutarotstion of Tetra-Mutarotation of Glucose in mixtures of Pyrid- methyl Glucose in mixtures of Pyridine and ine and Water at mixtures of Cresol Water at 20".20". and Pyridine at 20". k. % Creaol. k. % H,O- k.* % H*O. 100 95 90 85 80 70 65 60 60 40 30 20 10 0 0.015 0.140 0.258 t 0.274 0.282 0.272 0.269, 0.225 0.277 0.288 0.210 0.190 0-221 0.236 0.166 0-168 0,106 0.100 0-046 0.012 0-0008 0-258 100 97.5 95 92-5 90 85 80 76 70 60 50 40 30 20 10 0 0-0128 0.099 0.177 0.219 0.264 0.261 0.272 0.254 0.267 0-273 0.300 0-317 0-238 0.318 0.291 0.251 0.203 0.179 0.108 0.040 0.042 0.0088 0.0003 100 0.0003 99 0-027 97.5 0.060 95 0-082 92-6 Mutarotation to too rapid for 55.0 measurement. 52.5 0.180 49.9 0.148 40.4 0.067 21.1 0-0168 1-62 0.0013 0 0-0003 k = l / t . (log (ao - a ) - log (q - a )).In calculating this coefficient Values shown in italics correspond with a half-change period of less than $ m-Cresol was used for these four solutions as o-cresol is solid at these the time was always expressed in minutea. 3 minutes and are not regarded as trustworthy. concentrations. and 111 and are represented graphically in Figs. 1 and 2. The individual readings have not been reproduced. A complete muta-rotation-curve is therefore represented by a single monomolecular coefficient $ in the tables and by a single point on the curves. 9 The evidence for the monomolecular character of the curves will be dis-cussed later in another psper BEAUI'IONS OF ORGAXIU TEIWULPHATES. 2887 Even under the moat favourable conditions the fastest muta-rotations for which fairly trustworthy data can be given correspond with a velocity.coefficient of less than 0.2 since this velocity coefficient already corresponds with a half-change period of only 2.3 minutes. Values greater than this are therefore shown in italics in the tables and are indicated only by a dotted line in Figs. 1 and 2. In the case of glucose the mixtures which gave the highest velocities also gave cloudy solutions which were difficult to read; but the general come of the curve can be inferred fairly accurately from the velocity coefficients for stronger and weaker solutions for which much more concordant values were obtained. The mixtures rich in cresol suffered from the two-fold disadvantage that they were very viscous and showed only a relatively small change of rotatory power ; the velocity-concentration curve for pyridine-cresol mixtures therefore includes a longer dotted section than the two curves for mixtures of pyridine and water. We desire to thank the Depa,rtment of Scientific and Industrial Research for a maintenance grant to one of us (I. J. F.) during the period in which this research was carried out. UNIVERSITY CHEM~CAL LABORATORY, CAlKBRIDGE. [Received November Znd 1926.