A number of methodologies for the oxidation of alkenes to allylic alcohols have been developed (Fig. 1).1) For example, selenium dioxide and chromate reagents generally oxidize the allylic position without rearrangement of the carbon-carbon double bond. In contrast, singlet oxygen affords the allylic alcohol with the rearrangement of the carbon-carbon double bond. Epoxidation of alkenes followed by epoxide opening with metal amide is also well-known. [2][3][4][5][6] During the course of the synthetic investigation of diversifolin, we found that treatment of g,d-unsaturated carboxylic acid 1 with 2.0 eq of phenylselenenyl chloride (PhSeCl) in pyridine at 50°C, followed by oxidation with hydrogen peroxide, provided g-butyrolactone 2 in 44% yield (Chart 1).
7)This result means that allylic oxidation of the trisubstituted alkene proceeded in addition to the planned selenolactonization of the 4-alkenoic acid moiety. Herein we wish to report the scope and limitations of the phenylselenenyl chloridemediated allylic oxidation of alkenes.At first, we investigated the optimization of this oxidation using citronellyl acetate 3 as a model substrate. As shown in Table 1, the reaction of 3 proceeded in a similar manner (PhSeCl-pyridine, 50°C) as 1 to give the corresponding allyl alcohol 4 in 66% yield (entry 1).8) When other readily available phenylselenium reagents were applied in the reaction, neither phenylselenenyl bromide (PhSeBr) nor N-PSP (Nphenylselenenylphthalimide) were effective for this transformation (entries 2, 3). Although the reactions in benzene, triethylamine, and 2,6-lutidine yielded poor results, the reaction in acetonitrile gave 4 in 31% yield (entries 4-7). As shown in entries 8 and 9, higher reaction temperatures resulted in a slight decrease in the yield probably due to decomposition of 4. These results indicate that treatment with PhSeCl in pyridine at 50°C gave the best result.With optimum conditions established, we next carried out the oxidation of a variety of other substrates 5a-j as shown in Table 2.8) The reaction of 2-methyl-2-undecene (5a) gave the corresponding allyl alcohol 6a in 55% yield (entry 1), while 1,1-diethyl substituted derivative 5b gave 6b in a very low yield due to steric hindrance around the trisubstituted olefin (entry 2). In contrast, cyclohexylidene derivative 5c produced 6c in higher yield than in the case of 5b (entry 3). This result might be attributed to inflexibility of the cyclohexane ring in 5c. Although the allylic oxidations of nerol 5d and citronerol 5f afforded the corresponding diols 6d and 6f A phenylselenenyl chloride (PhSeCl)-mediated allylic oxidation to give allylically rearranged alcohol has been developed. A possible mechanism for the present reaction is generation of allylic selenide from prenyl moiety via [1,3]-sigmatropic rearrangement, followed by oxidation and [2,3]-sigmatropic rearrangement to afford 3-isopenten-2-ol.