Non-targeted screening for food chemical safety

Danielle Cawdron, Analytical Scientist

A new member-funded research project will give members access to non-targeted analysis for 'unknown hazards' that gives more confidence in the safety of their food and drink products.

Transcript

One of the largest issues facing the food industry today is the identification of new and emerging contaminants. Although analytical techniques have improved in selectivity and sensitivity in recent years, the task of identifying components and a sample matrix is still daunting. Traditionally the analytical methods surrounding food safety have been specifically developed for a certain sample type or for a class of compounds. This targeted approach has been used with success for several decades. However, a targeted approach does come with a significant drawback - it will miss all compounds that have not been pre-selected at the beginning of the analysis. So, by comparison a non-targeted screening technique would enable us to rapidly detect and identify unknown or unexpected compounds within a food and drink sample without the need to pre-select the chemicals of interest.

At the end of the three-year project we aim to have developed a robust non-targeted screening approach across a range of analytical instruments and we'd also hope to demonstrate that this is not only a practical, but useful tool for the detection and identification of unknown contaminants.

We see this project breaking down into three main areas, the first of which is sample preparation. The final sample extraction method needs to be unselective and be able to solubilise and recover a large range of chemical compound classes. A pre-separation technique such as gas chromatography or liquid chromatography would also be employed to reduce the complexity of the matrix, to help separate out any isomers and to concentrate the analytes of interest. Mass spectrometry is one analytical technique that has been identified as a powerful tool in non-targeted screening primarily because prior knowledge of chemical content isn't required and it has a relatively large dynamic range. Thirdly, how we process and interpret the data is going to be critical. Often the bottleneck in a non-targeted workflow is your data analysis and interpretation. The mass spec output data from these high-resolution instruments is very information rich and a powerful data mining tool needs to be employed to enable identification and interpretation.