Theoretical calculations involving singlet molecular oxygen (O2(1g)) are challeng- ing due to their inherent multi-reference character. We have tested the quality of re- stricted and unrestricted DFT geometries obtained for the reaction between singlet oxy- gen and a series of alkenes (propene, 2-methylpropene, trans-butene, 2-methylbutene and 2,3-dimethylbutene) which are able to follow the ene-reaction. The electronic en- ergy of the obtained geometries are rened using 3 dierent methods which account for the multi-reference character of singlet oxygen. The results show that the mechanism for the ene-reaction is qualitatively dierent when either one or two allylic-hydrogen groups are available for the reaction. When one allylic-hydrogen group is available the UDFT calculations predict a stepwise addition forming a biradical intermediate, while, the RDFT calculations predict a concerted reaction where both hydrogen abstrac- tion and oxygen addition occur simultaneously. When two allylic-hydrogen groups are available for the reaction then UDFT and RDFT predict the same reaction mechanism, namely that the reaction occurs as a stepwise addition without a stable intermediate between the two transition states. The calculated rate constants are in reasonable agreement with experimental data, except for trans-butene where the calculated rate constant is three orders of magnitude lower than the experimental one. In conclusion we nd that the simple bypassing scheme tested in this paper is a robust approach for calculations of reaction involving singlet oxygen in the limit that the transition state processes low multi-reference character. 2

Ab initio calculations on systems involving singlet molecular oxygen (O2 (1g)) are challenging due to signicant multi-reference character arising from the degeneracy of the HOMO and LUMO orbitals in singlet oxygen. Here we investigate the stragegy of bypassing singlet oxygen’s multi-reference character by simply adding the experimen- tally determined singlet/triplet splitting (22.5 kcal/mol) to the triplet ground state of molecular oxygen. This method is tested by calculating rate constants for the reac- tions of singlet molecular oxygen with furan, 2-methylfuran, 2,5-dimethylfuran, pyrrole, 2-methylpyrrole, 2,5-dimethylpyrrole, and cyclopentadiene. The calculated rate con- stants are within a factor of 15 compared to experimentally determined rate constants. The results show that energy renement at the CCSD(T)-F12 level of theory is cru- cial to achieving accurate results. The reasonable agreement with experimental values validates the bypassing approach which can be used for other systems involving the 1,4-cyclo-addition of singlet oxygen. 2