Reaction of Organic Sulfides with Singlet Oxygen. A Revised Mechanism Frank Jensen,* ,† Alexander Greer, ‡ and Edward L. Clennan ‡ Contribution from the Departments of Chemistry, Odense UniVersity, DK-5230 Odense M., Denmark, and UniVersity of Wyoming, Laramie, Wyoming ReceiVed NoVember 3, 1997. ReVised Manuscript ReceiVed February 19, 1998 Abstract: On the basis of ab initio calculations we propose a revised mechanism for the reaction of organic sulfides with singlet oxygen, which is more consistent with experimental evidence than previous schemes. In aprotic solvents the reagents initially form a weakly bound peroxysulfoxide, with a small barrier due to entropy. The peroxysulfoxide may decay back to ground state (triplet) oxygen, be trapped by sulfoxides, or rearrange to a S-hydroperoxysulfonium ylide with a barrier of ∼6 kcal/mol. The latter is ∼6 kcal/mol more stable than the peroxysulfoxide, and can be trapped by sulfides or rearrange to a sulfone. In some cases, like five- membered rings or benzylic sulfides, the S-hydroperoxysulfonium ylide may undergo a 1,2-OOH shift to an R-hydroperoxysulfide, which eventually leads to cleavage products. In protic solvents the peroxysulfoxide is rapidly converted to a sulfurane by solvent. Introduction The photooxidation of organic sulfides (R 2 S) was originally reported by Schenck and Krauch in 1962. 1 It is now generally accepted that the reaction proceeds via the lowest excited singlet state of molecular oxygen, commonly referred to as singlet oxygen. The major reaction product is the sulfoxide (R 2 SO), with varying amounts of sulfone (R 2 SO 2 ) depending on the substrate and reaction conditions. Some substrates, like benzylic sulfides, also give products corresponding to oxidation of the carbon framework. The reaction mechanism has proven to be very complex. 2-6 Both the reaction efficiency and kinetic behavior depend on substrate, temperature, and solvent. Kinetic and trapping experiments suggest that there are at least two distinct inter- mediates present in aprotic solvent, while a single intermediate is sufficient for explaining the experimental data in protic solvents. 2 It has not been possible to detect any of these intermediates spectroscopically under normal reaction condi- tions. Low-temperature matrix isolation studies have obtained IR bands of an unstable species, 7 but the nature of this species is not clear. 8 We have previously performed ab initio calculations with the aim of evaluating the viability of various proposed intermediates, and to describe the detailed pathways for their formations. 9,10 These calculations suggested that the energetics of the proposed mechanism were inconsistent with experimental facts. None of the previous mechanisms have been able to explain all available experimental facts. We here wish to report new calculations which suggest a revised reaction scheme. The central feature is the presence of a new intermediate, a S-hydroperoxysulfonium ylide. 11 After a brief review of key experimental observations, the results of our calculations are presented, and it is shown how the experimental facts fit the new reaction scheme. During the course of this work Ishiguro et al. reported semiempirical calculations, which also suggest a S-hydroper- oxysulfonium ylide intermediate. 12 Just prior to completion of this work we learned that Prof. McKee had performed similar type calculations for dimethyl sulfide using DFT methods for structural features and QCISD(T) calculations for energetics. 13 With a single exception (discussed below), we find very good agreement with these results. Computational Details All calculations have been performed with the Gaussian-94 14 and ACES II 15 program packages, using standard basis sets. 16,17 Singlet oxygen is a delta state, which requires a complex wave function within a single determinant framework. At the MP2 level the experimental singlet-triplet energy difference of 22.5 kcal/mol 18 is reproduced to † Odense University. ‡ University of Wyoming. (1) Schenck, G. O.; Krauch, C. H. Angew. Chem. 1962, 74, 510-510. (2) (a) Gu, C.-L.; Foote, C. S.; Kacher, M. L. J. Am. Chem. Soc. 1981, 103, 3, 5949-5951. (b) Liang, J.-J.; Gu, C.-L.; Kacher, M. L.; Foote, C. S. J. Am. Chem. Soc. 1983, 105, 4717-4721. (3) Gu, C.-L.; Foote, C. S. J. Am. Chem. Soc. 1982, 104, 6060-6063. (4) Jensen, F. In AdVances in Oxygenated Processes; Baumstark, A. L., Ed.; JAI Press: Greenwich, CT, 1995; Vol. 4, pp 1-48. (5) Clennan, E. L. In AdVances in Oxygenated Processes; Baumstark, A. L., Ed.; JAI Press: Greenwich, CT, 1995; Vol. 4, pp 49-80. (6) Sawaki, Y.; Watanabe, Y.; Ishikawa, S.; Ishiguro, K.; Hirano, Y. In The role of oxygen in chemistry and biochemistry; Ando, W., Moro-oka, Y., Eds.; Elsevier: Amsterdam, 1988; pp 95-98. (7) Akasake, T.; Yabe, A.; Ando, W. J. Am. Chem. Soc. 1987, 109, 8085-8087. (8) Previous calculations at the MP2/6-31G(d) level did not provide a good match for the observed frequencies (ref 9). We have improved the theoretical level to MP2/6-311+G(2df), with only minor changes in calculated frequencies and isotopic shifts. Frequencies for the S-hydroper- oxysulfonium ylide 4 have also been calculated, but these do not agree with experimental values either. (9) Jensen, F. J. Org. Chem. 1992, 57, 6478-6487. (10) Jensen, F.; Foote, C. S. J. Am. Chem. Soc. 1988, 110, 2368-2375. (11) (a) Greer, A.; Chen, M.-F.; Jensen, F.; Clennan, E. L. J. Am. Chem. Soc. 1997, 119, 4380-4387. (b) Clennan, E. L.; Chen, M.-F.; Greer, A.; Jensen, F. J. Org. Chem. Submitted for publication. (12) Ishiguro, K.; Hayashi, M.; Sawaki, Y. J. Am. Chem. Soc. 1996, 118, 7265-7271. (13) McKee, M. J. Am. Chem. Soc. 1998, 120, 3963-3969. 4439 J. Am. Chem. Soc. 1998, 120, 4439-4449 S0002-7863(97)03782-7 CCC: $15.00 © 1998 American Chemical Society Published on Web 04/28/1998