Optimized saponification for MOSH and MOAH analysis in food matrices: addressing internal standard variability

Mineral oil saturated hydrocarbons (MOSH) and mineral oil aromatic hydrocarbons (MOAH) are a complex mixture of food contaminants with high lipophilicity. Due to the complexity of MOH and food matrices, an efficient sample preparation method must be performed, which depends on the type of sample and the possible interferences. However, one step commonly included in the sample preparation for MOH analysis is the saponification, which allows the complete extraction of MOH from the lipids present in the matrix, improving the analytical sensitivity [1].
The official ISO 20122:2024 method for the determination of MOSH and MOAH determination in vegetable oils included a saponification step [2]. However, two major issues arise with the ISO method. First, it introduces substantial variability (15-25%) in MOAH quantification due to inconsistent partitioning of internal standards during saponification, affecting the ratio between 1,3,5-tri-tert-butylbenzene (TBB) and 2-methylnaphthalene (2-MN) [3]. Although the expected ratio is 1.00, values exceeding 1.15 have been reported, causing discrepancies in quantification. Second, the ISO saponification conditions (60°C, 30 min) was validated primarily on low-melting vegetable oils, which means that an adaptation is necessary for other type of food matrices.
This work presents an optimization of the ISO method using a microwave-assisted extraction (MASE) system for high-melting fats and different types of meat, as a more complex food sample. For the meat, unprocessed meat samples differing on the percentage of fat were studied: pig rib, beef rib and bacon (with 30%, 23% and 16%, respectively).
The performance of the saponification was evaluated in terms of internal standards distribution and recovery. The methods proposed in this work allows the complete saponification of both matrices, obtaining TBB/2-MN ratios closer to the expected value (1.00):
i) For high melting fats: the TBB/2-MN ratio improved (from approximately 1.15 to 1.05), while maintaining high recoveries across diverse edible oils.
ii) For meat: the TBB/2MN ratio was consistent among the three different types of meat tested and was on average 1.08 ± 0.02 and recoveries of 98% for MOSH and 96% for MOAH.
In conclusion, the MASE proposed in this work has shown to be adequate not only for MOH analysis in high-melting fats, but also for other types of matrices with a more complex composition.