Computational determination of the enthalpy of formation of alkylthial S-oxides and alkylthione S-oxides: a study of (Z)-propanethial-S-oxide, the lachrymatory factor of the onion (Allium cepa)

Abstract

A general computation procedure for the determination of unknown enthalpies of formation of alkylthial or alkylthione S-oxides, based on the recent accurate determination of the enthalpy of formation of methanethial S-oxide, is proposed. The method relies on the use of isodesmic reactions involving the sulfoxide and sulfide of the species of interest, with the individual energies calculated at the CBS-QB3 level. A discussion is given about the convenience of using density functional theory (DFT) instead of the CBS-QB3 model chemistry, especially for large compounds. To exemplify the method, the enthalpies of formation of ethanethial-S-oxide (−74.7 ± 8.4 kJ mol⁻¹), propanethione-S-oxide (−112.1 ± 8.4 kJ mol⁻¹) and propanethial-S-oxide (−94.1 ± 8.4 kJ mol⁻¹) are calculated, as well as those of the necessary sulfides propanethial (50.4 ± 8.4 kJ mol⁻¹) and propanethione (28.0 ± 8.4 kJ mol⁻¹). These values were derived on the basis of an independent theoretical estimation of the enthalpy of formation of thioformaldehyde (113.2 ± 4.2 kJ mol⁻¹, in agreement with the experimental value, 118 ± 8.4 kJ mol⁻¹). The use of the same method for evaluating the enthalpy of formation of ethanethial however, gave 68.1 ± 4.2 kJ mol⁻¹, higher than the experimental value, 50 ± 8 kJ mol⁻¹. Convincing evidence is given that the theoretical estimation should be preferred to the experimental one in this case. Moreover, it was determined that the most stable isomer of propanethial S-oxide is in fact the Z-conformer, where the terminal methyl group is about 120° out of the plane of the CSO group. However, the CS and CH–CH₂– bond lengths are very different from the recent experimental determination. A comparison with the experimental and theoretical results for the lower members of the series suggests that the model assumed for deriving the bond lengths from the microwave spectrum may be at fault. The whole process is validated by comparison of the enthalpies of formation of thioaldehydes and thioketones to the known enthalpies of formation of aldehydes and ketones.