In perester decompositions

Ethyl peracetate was the first ester of a peroxy acid, and was characterized by Baeyer and Villiger in 1901. Kinetic studies of perester decomposition were reported by Blomquist and Ferris in 1951, and in 1958 Bartlett and Hiatt proposed that concerted multiple bond scission of peresters could occur when stabilized radicals were formed (equation 46). As noted below (equation 57), polar effects in perester decomposition are also significant. 

A wealth of kinetic data concerning the effect of substituents in perester decompositions is reported in the literature. The data are divided into types of R groups in RC03C4H9-t. The trend from Table 91 is apparent. The rate coefficients... 

In a study of sulfur participation in perester decomposition it was of interest to eliminate steric acceleration as a factor. Hence the rates of thermolysis of (15) and (16) were compared as evidence that anchimeric assistance by sulfur rather than steric acceleration was responsible for the high relative rates observed for the o-phenylthioperesters (see Chapter 5). 

In the case of reaction 3, entries 1 and 2, that is, iert-butyl peracetate and (ert-butyl perpropionate, almost certainly decompose by a stepwise mechanism, rather than the concerted mechanism assumed for reaction 3. Entry 3, tert-butyl perisobutyrate, probably forms the least stable R radical by the perester decomposition mechanism which is still mostly concerted in nature (36). 

It is somewhat contradictory and not yet fully understood why the back strain effect on the rate of perester decompositions is so large. We had reasoned before from the discussion of conformational effects that the Ca-CO-bond of 25 is only stretched to a small extent at transition state. From an analysis of bond energies5 18 it becomes questionable if the homolysis of C-N-bonds (as in 20 ) and C-C-bonds (as in 25) is likely to be directly comparable5,12a 18 In addition the extent of Ca-CO-cleavage at the transition state of fragmentation of 25 may well be itself dependent on the... 

Disulfide was found to be the main product (yield 52.5%) of this perester decomposition. Accelerating action of ortAo-substituents with p- or tt-electrons is due to the formation of an intermediate bond of the O S or O C=C type in the transition state ... 

This bond formation compensates (partially) the activation energy for dissociation of the O—O bond in perester. The empirical peculiarities of anchimeric assistance decomposition are the following [3,4] ... 

Three different mechanisms of perester homolytic decay are known [3,4] splitting of the weakest O—O bond with the formation of alkoxyl and acyloxyl radicals, concerted fragmentation with simultaneous splitting of O—O and C—C(O) bonds [3,4], and some ortho-substituted benzoyl peresters are decomposed by the mechanism of decomposition with anchimeric assistance [3,4]. The rate constants of perester decomposition and values of e = k l2kd are collected in the Handbook of Radical Initiators [4]. The yield of cage reaction products increases with increasing viscosity of the solvent. 

The higher the viscosity of the solvent, the higher the amount of the parent molecules formed due to the geminate recombination of radicals. The observed rate constant of decomposition of the initiator decreases with an increase in viscosity [3,90], This was observed in the decomposition of peresters and diacetyl peroxide in various solutions. Subsequently, the fraction fT of the radical pairs recombining to the parent molecule increases with an increase in the viscosity ... 

Winstein estimated the rate acceleration due to anchimeric assistance in ionic perester decompositions as follows. By using Equation 6.66, in which the first term on the right-hand side is the difference in homolytic dissociation... 

Azo compound decomposition is much less susceptible to polar substituent effects, and so probably has less charge separation in the transition state,75 but is more sensitive to geometrical restrictions. Bridgehead azo compounds decompose at rates lower than expected on the basis of their tertiary nature, whereas peresters are much less strongly affected.70 The difference can be rationalized by the proposal that the transition state comes farther along the reaction coordinate in azo decomposition, so that the nonplanarity forced on the incipient radical by the ring system is felt more strongly there. 

Comparison of the cyclic systems in Table 17 leads to the opposite conclusion, however the destabilization of the 1-norbornyl radical relative to the 1-adamantyl is less for the azo decompositions. Perhaps the mechanism of the azo decompositions of the more unreactive systems is different from that of, for example, the f-butyl azo compound (i.e. the rate determining step of the 1-norbomyl azo compound may be a one bond homolysis rather than the synchronous two bond fission of the f-butyl system312, 315)). Also, the smaller 1-norbornyl/1-adamantyl rate ratio for the f-butyl perester decompositions may be due to a greater influence of polar effects in these reactions 309a). This problem is under active investigation 309a). 

For further examples of dichotomous solvent-influenced radical/ionic perester decompositions, see the base-catalyzed perester fragmentation shown in Eq. (5-39) in Section 5.3.2 [110], as well as the decomposition of t-butyl heptafluoroperoxybutyrate, C3p7-C0-0-0-C(CH3)3 [691]. The relative extent of monomolecular and induced thermal decomposition of disubstituted dibenzyl peroxydicarbonate, ArCH2-0-C0-0-0-C0-0-CH2Ar, is also substantially influenced by the reaction medium [692]. The thermolysis of suitable dialkyl peroxides can also proceed by two solvent-dependent competitive reaction pathways (homolytic and electrocyclic reaction), as already shown by Eq. (5-59) in Section 5.3.4 [564]. 

Attention was given to the kinetics of perester decomposition in previous reviews (refs. 399, 400). Induced decomposition is observed with these compounds. For the decomposition of /-butyl perbenzoate in p-chlorotoluene, the rate law was suggested to be the same as in the decompositon of diacyl peroxides , viz. 

FUNCTIONAL GROUP PROXIMITY EFFECTS IN THE DECOMPOSITION OF /-BUTYL PERESTERS IN CHLOROBENZENE... 

Several investigations have been concerned with the effect of solvent upon the rate of perester decomposition. These data also shed some light on the importance of the ionic structure (II) in the transition state. Some data on the effect of solvent may be obtained from previous tables. For the decomposition of peresters where R is a primary alkyl group in RCO3C4H9-/ and one-bond homolysis is the mechanism of choice, changes in solvent polarity have little effect on the rate of decomposition. The rate coefficients for the decomposition of r-butyl percaprate at 110 °C in chlorobenzene, nitrobenzene and diphenyl ether are 8.30 x 10 , 6.58 x 10" and 6.39 X 10" sec"S respectively" . The rates are also independent of initial concentration of the peroxide this may indicate that induced decomposition is unimportant cf. sub-section 13.4.1). 

Martin and his co-workers have provided a firm experimental foundation for anchimeric assistance in the homolytic decomposition of per-benzoate derivatives (Table 6). A single o-phenylthio group was found to increase the rate of decomposition of f-butyl perbenzoate by a factor of ca. 45 thousand. The effect is not steric since an o-r-butyl group hardly affects the rate of perester decomposition [compare (95) versus (99), Table 6]. A noticeable but weak effect is observed with the homologous sulfide (96). In addition to sulfur, anchimeric assistance by iodine and vinyl groups in (97) and (98), respectively, was observed. The bridged canonical species... 

The absence of radical products in the reaction of Scheme IV, and the formation of only a limited amount of product in the perester decomposition (Scheme HI) which could have arrived by a non-radical path (via could be explained in at least two ways. The two pair species, one from perester and the other from the pairing of ions, could differ in that the ion sources contribute potassium and triflate ions which could be included in an ion aggregate, possibly more likely to give ionic products than is the unperturbed sulfonium—tert-butoxide pair. Further work will be required to confirm or deny this explanation. The associated uncertainty remains a defect in the experiment. 

---