Evaluation of a new microwave assisted saponification/extraction method for mineral oil analysis in food

Mineral oil hydrocarbons are currently being investigated regarding their toxicity to humans. The analysts working on MOHs are now focusing to reduce the variability associated to this kind of analysis. The variability associated with this analysis mainly originates from the interpretation of the chromatograms and their integration. This aspect is being addressed through the use of GC×GC and automatic software integration. However, important variability is also associated with sample preparation, an aspect still underevaluated.

Recently, two interlaboratory trials has been organised: one by the DGF for oil and fat, and another by the JRC for infant formula. The sensitivity reached in oil samples was at 2 mg/kg for the MOAH while the JRC reached a LOQ of 1 mg/kg. One of the main issues reported was the distribution of the internal standards for the MOAH fraction. In all the cases reported the ratio between TBB and the MN standards were at least at 1.15 instead of 1. This discrepancy in the ratio it is associated to a different partition behaviour of the two internal standards, which obviously can be reflected in a different partition of the contamination itself, depending on its profile.
In order to mitigate this variability, it was decided to compare the different saponification method and to develop a method using microwave to lower this discrepancy of extraction between the compounds. At the end, the ratio between the standard was reduced from between 1.11 and 1.18 to between 1.02 and 1.08 with acceptable recoveries in different matrices. The recoveries were calculated spiking matrices with gravex and QC-10 standard to cover all the range of volatility.

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.