Exploration of the saponification conditions for mineral oil analysis through the utilization of microwave-assisted saponification/extraction.

Mineral oil hydrocarbons are a class of pollutants that potentially possess harmful properties, particularly the aromatic hydrocarbons (MOAH) subset. The process of analytical determination is found to be associated with a considerable degree of variability, mainly attributable to the procedures involved in the sample preparation phase that requires enrichment and purification, alongside the interpretation of chromatograms. In recent years, efforts have been undertaken to automate the chromatogram interpretation, resulting in a decrease in variability.
Regarding the sample preparation, two interlaboratory trials conducted in the previous year – one by the JRC for infant formula, and another by the DGF for oil and fat – have highlighted the challenge of achieving low sensitivity in detecting the MOAH fraction in fat samples, resulting in a limit of quantification (LOQ) of 2 mg/ml. An issue noted was the observed discrepancy in the recovery of the internal standards. It can be observed that TBB exhibits a superior recovery compared to the MN standards. Indeed, the mean ratio of TBB compared to the MN standards stands at 1.15. One issue encountered was that, in one trial, TBB yielded superior recovery outcomes while 2MN resulted in better recoveries in another, indicating that the recovery of contamination could vary according to its specific composition. The objective of minimizing variability was addressed through the use of an innovative saponification technique, which leveraged the benefits associated with microwave-assisted saponification/extraction. Subsequently, evaluations were conducted on the partitioning behaviour of internal standards, across a diverse range of matrices and saponification conditions. In summary, the observed ratio disparity among all matrices scrutinized was diminished from a range of 1.11 to 1.18 to a range of 1.02 to 1.08.
Acknowledgement
This work is supported by Fonds de la Recherche Scientifique Belgique (FNRS) (PDR projects-ToxAnaMOH, T.0187.23). The authors thank Milestone, LECO and Restek for their support.