by J. Boeseken et G. Elsen, Rec. Trav. Chim., 48, 363 (1929)
Translated by Jeffrey Jenkins (Eleusis), whose comments are [bracketed]
In a previous communication1, one of us demonstrated how Peracetic Acid oxidation almost always gives a mixture of acetate diols. This reaction differentiates itself clearly from oxidation with Perbenzoic Acid almost always gives ethylene oxides. The same author then supposed that Peracetic Acid could be considered like a “monoacetyl perhydroxy” and that oxidation was the result of simple addition of this substance to a double bond (an addition with inversion, of [after?] Walden). One could expect that the initial product is always a monoacetate [ester?] and that the formation of a diacetate or a free diol is respectively due to an ulterior acetylation or saponification.
It is also possible that the course of a reaction strongly depends on the groups which are directly linked to the unsaturated carbon atoms [the alpha carbon, only?]. M. Derx2, at first, and then M.J. Petrus, Blumberger & one of us3 had found that the speed of the oxidation by Perbenzoic Acid depends on the placement of the double bond in relation to the benzyl core, thus Phenyl-1-Propene was attacked much more quickly than Allylbenzene, and Phenyl-1-Butene much more so than its isomers, etc...
It appeared to us that the character of the products obtained when using Peracetic Acid as an oxidizer would be more or less linked to the speed of the oxidation as well. For example, a rapid oxidation would primarily produce a double bonded acetate, providing that a slow oxidation would cause either the formation of more diacetate or a complication. We have oxidized C6H5CH=CHCH2C6H5, Anethol, beta-Isosafrole, and Indene, all with the double bond closer to the phenyl side. They are all attacked quickly; the first three giving the monoacetates and Indene giving a mixture of mono- and di- acetates. In contrast, Allylbenzene gives a diacetate probably because the monoacetate is transformed by the long oxidation required into the diacetate [a competing reaction].
Safrole gives a complicated mixture: it seems that the molecule was partially attacked in another sense.
In the meantime, Eugenol is a derivative of the CH2CH=CH2 bond so it is rapidly oxidized giving the monoacetate with a small amount of the diacetate.
M.V. Broek4 found that the tetrabromo derivative of the acid d-Eleostearic [?] gives the monoacetate while tetrabromodihydroxystearic acid is very slow to react.
These observations indicate that the reaction process begins with a simple addition of Peracetic Acid to the double bond. If this addition is very slow, the monoacetate of the diol can bond to another part of the molecule (safrole), it can be transformed into the diacetate (allylbenzene) or it may remain unchanged [ie - unreactive].
It goes without saying that for the cyclohexene (of which the monoacetate, diacetate and free diol were to be obtained), the speed of saponification during the isolation process plays an important role [ie - the cyclohexyl analogues are more sensitive to the neutralization with .1N KOH than their aromatic counterparts].
The Experimental Part
[All ignored except for Isosafrole]
We have chosen this substance because it is an example of cis-trans isomerism. However, we have only been able to obtain one of the possible isomers of Isosafrole: beta-Isosafrole. the would-be alpha-Isosafrole [a benzylenic, C6H5=CHCH3CH3?] isn’t a pure substance, even after troublesome fractional separation in an open cathodic device [electrolytic separation???] of refraction. The fractions presented such deviations that this substance should be a mixture of at least 2 substances with differing indices of refraction but of near-identical boiling points. Moreover, within this mixture of isomers there should be one which is rapidly oxidized [by Peracetic Acid]. This agrees with the results of H.I. Waterman & Priester (7) who found that the ordinary transformation of Safrole to Isosafrole, when attempting to prepare alpha-Isosafrole, was incomplete and actually gave a mixture of beta-Isosafrole and Safrole [assuming that alpha-Isosafrole is a benylenic, this makes sense].
[The preparation of beta-isosafrole is omitted - it is merely the fractional distillation of crude isosafrole using a Vigreux column.]
16.21g of beta-Isosafrole was dissolved in 70mL of 1.52M Peracetic Acid (an excess of 6.4% acid, that is to say) and from there cooled in an ice bath. The evolution of heat is striking! All of the Peracetic Acid was consumed after 2 days; the product of the reaction was treated in the manner below [the acetate is saponified using dilute potassium hydroxide - this step is not performed if one wishes to make MDP-2-P]. The oil thus obtained was a clear-brown in color and had a characteristic ethereal odor.
1g of oil required 43.4mL of .1N KOH soln. to saponify.
[The process using alpha-Isosafrole is summarized below working on the assumption that alpha-Isosafrole is not the one of interest, but just in case...]
16.3g alpha-Isosafrole, having an index of refraction between 1.5668 and 1.5676 [Sodium D Line, I asssume] was added to 75mL of 14.1M [!!! CAUTION !!!], or approximately 5% excess acid [perhaps they meant 1.41M?]. The evolution of heat was much less than that of beta-Isosafrole [which would agree with a benzylic intermediate]. .45g of the clear-brown oil obtained required 18.0mL of .1N KOH soln. to saponify. The presence of a byproduct is indicated by the lower amount of alkali needed.
[Other examples deleted as well.]
In summary, we can conclude that the action of Peracetic Acid on an unsaturated molecule at the double bond is a simple addition. What happens ultimately depends on the character of the formed monoacetate and the facility of the starting material towards oxidation. To wit: