Imagine if you were asked to work into a haybarn and a friendly farmer asked you to help to locate a specific bale of hay. Unfortunately, the bale look very similar to the rest of the bales in the stack but it accidentally came from a neighbouring farmer’s field and the farmer would like it back! The only differentiator is that the grasses that the neighbouring farmer used to make their bale of hay was from differing species compared to the rest of the bales in the barn and so, as a result, it would be technically possible to tell which was the interloper but very difficult to employ in practice.
This unusual agricultural analogy is, in a sense, similar to the challenge people face when measuring mineral oils from food. How do we detect specific analytes, where the physicochemical properties of the analytes of interest, are similar to the physicochemical properties of oils that are found in foods? One possible way would be to use a very specific technique like gas chromatography interfaced with quadrupole and time of flight mass spectrometry (GC-QTOF MS). In theory, GC-QTOF MS could identify and differentiate 1000s of components in a given food sample. For example, this approach is successfully deployed in lipidomic research, where 1000s of individual lipidomic signatures are detected and also sometimes quantitated, for biochemical and clinical research.
However, not only are GC-QTOF MS instruments expensive but also, extensive sample preparation would also need to be conducted to prevent matrix contamination and competition in the mass spec source from the natural oil components. Also, as mentioned in part one of this blog series, for MOSH MOAH analysis, specific detection of individual mineral oil components is largely unnecessary. The reason is that any mineral oil contamination in foods is, at best, unwanted and at worst, is or feared to be toxic. And so, the mere presence of any contamination is sufficient to cause concern. As a result, MOSH MOAH detection relies on the detection and quantitation of ‘hump’s on chromatograms from GC interfaced with Flame Ionisation Detectors (GC-FID). These humps are regions on the GC chromatogram which are enriched in mineral contamination (MOSH) or MOSH) and only are observed when food is contaminated.
As a gross simplification, the natural fats and oils found in foods are largely found in two forms, they are either as free fatty acids or, far more commonly, as triglycerides (fatty acid esters attached to a glycerol backbone). They are linear, mostly aliphatic and often have one or more double bond(s) in as cis isomers. The shorter the chain length and the more double bonds present in the fat/oil, the lower the melting point. By contrast, mineral oils are often just hydrocarbons, saturated or unsaturated, of varying chain lengths, many show multiple branch chains. Moreover, mineral oils can be either aliphatic (MOSH) or aromatic (MOAH) in nature. These general differences between food oils and mineral oils can be exploited in different ways to deplete samples of food oils, to reveal the aforementioned mineral oil humps.
Food and food packing matrices are really complicated, the can be wet or dry, rich or depleted in natural oils, or indeed various mixtures these properties. Let’s take for example a prepackaged tuna sandwich, where the tinned tuna had been preserved in sunflower oil. Here you have a mixed matrix where you get oils from the sunflower oil, margarine and also from the tuna itself. Moreover, the tuna will be wet and the bread will also have both some water and some potential seed oil content. That’s a pretty complicated matrix! And without relevant food safety controls employed, there could have been several potential points in the manufacture / sourcing & transport of the ingredients, or even in the assembly of the sandwich, where there could be a risk of mineral oil contamination. One could argue that it would be easier to analyse the individual ingredients of the sandwich prior to assembly, and that certainly simplifies the problem. But even when we consider just the tuna its self, it still is a wet matrix with more than one source of natural oil to contend with (fish and sunflower oil).
Compare the above to just the sandwich packaging; where this could have been made from recycled cardboard which could have, come into contact with mineral in the past oil for example. The packaging would be by its very nature, pretty anhydrous and very low in oil content but the contaminated oils that they harbour could be potentially very dangerous for the consumer, when the packaging becomes in contact with the sandwich.
The heterogeneity of the food / packaging matrix therefore represents big challenges for laboratories hunting for mineral oil contamination. What works for a tuna sandwich will probably not work for its packaging! These are some of the upfront challenges labs need to consider prior to hunt for MOSH MOAH humps.