Optimisation and evaluation of performic acid epoxidation for the determination of mineral oil aromatic hydrocarbons (MOAH) in edible oils and fats - a promising global approach.

Removing biogenic interferences is a necessity for the unambiguous analysis of mineral oil aromatic hydrocarbons (MOAH) in many matrices, and routinely performed by epoxidation with meta-chloroperbenzoic acid in ethanol. With new reagents like performic acid (PFA) emerging, currently available protocols were assessed, optimised and subsequently compared to determine the most suitable method for the future. For a comprehensive evaluation, pitfalls and challenges of PFA epoxidation were identified for the first time, highlighting the need to remove polar solvents prior to analysis and the risk of losing internal standards during solvent evaporation. Precipitating epoxidation side-products could be removed by a novel miniaturised silica clean-up column, which also eliminated residual peroxides, thus ensuring sample stability. Reduced reagent volumes also proved equally suitable, minimising unwanted MOAH losses. In addition, 1-chlorobutane was found to be an appropriate substitute for chloroform, with enhanced occupational safety. Furthermore, polar solvents allowed PFA epoxidation with 30% instead of 50% hydrogen peroxide, hence enabling a global reach of the method. A combination of these innovative developments made it possible to design optimised routine suitable workflows. Overall, PFA in n-hexane was identified as the best possible method for manual use, demonstrating the least discrimination of MOAH and an effective removal of interferences, paired with the most user-friendly handling. PFA in 1-chlorobutane was summarized as the best compromise between performance and global availability. Moreover, two-dimensional gas chromatography (GC×GC) analysis revealed persistent interferences to likely be of aromatic structure, limiting further removal by epoxidation without additional losses of MOAH.