Due to the nature of economically motivated adulteration (EMA) and mislabeling, it is difficult to predict the exact nature of potential threats, so many in the food industry are looking to detection techniques that help detect “unknown-unknowns.” The analytical testing strategy identified to provide this type of detection is known as “food fingerprinting.” Unlike conventional approaches, which rely on detection of a known number of analytes as an indicator of authenticity, food fingerprinting measures a large number of variables and applies mathematics to generate a fingerprint specific to authentic samples of the commodity or ingredient of adulteration concern. A wide number of analytical techniques have been identified as useful in this approach, including nuclear magnetic resonance, molecular spectroscopy, stable isotope analysis, and mass spectrometry.
For a new approach to be successful and adopted widely, several characteristics are desirable. Namely, it should provide a rapid answer and be deployable in a manner that allows a large number of samples to be screened. Of course, it is highly desirable that it incurs minimal additional testing expense.
Fingerprinting of high-risk food types such as milk powder is valuable and NIR spectroscopy clearly has a role to play given its ubiquity in raw materials testing.
Near infrared spectroscopy (NIR) is an ideal choice as it is extensively used today in the food industry, and as a result, capital investments in new detection instruments are minimized. In addition, NIR does not demand laboratory-type sample preparation protocols, lab-based environmental conditions or specific gases, and it generally provides an answer in less than a minute. This enables NIR to be deployed in manufacturing facilities and operated by non-laboratory trained personal, resulting in cost-effective, fast screening for adulteration issues.
Example: Fingerprinting of Milk Powder
Milk powder is one of the most widely traded food commodities, with over 2.5 million metric tons exported annually, and is used in a huge array of food products, from infant formula to baked goods and confectionary. NIR is already widely applied to measure concentrations of key quality parameters such as protein, moisture, lactose, ash, and fat. Protein is a key quality parameter in milk linked to its value, and standard methods for protein analysis rely on a simple nitrogen assay with the protein concentration inferred from the nitrogen content. Addition of chemicals rich in nitrogen can artificially increase the apparent protein and the price demanded. Whilst regulators have responded and enforced tight regulations around some high nitrogen containing chemicals such as melamine, the “chemical space” is vast, and there are many more high-nitrogen compounds that could potentially be used in the same way. To stay ahead of criminals, it’s important to look beyond currently known adulterants and consider other possibilities.
NIR’s capability can be easily extended to screen samples of these potential unknown threats. NIR spectra contain information about the whole sample—including any adulterants present. There is no physical separation process at work, so the spectra must be processed with appropriate chemometric and mathematical tools to separate the contributions of the milk powder matrix and any adulterants.
A principal components analysis (PCA)-based method such as Soft Independent Modelling Class Analogy or SIMCA, in which a “fingerprint” is built for the unadulterated milk powder, and the degree of fit of the sample spectrum to this model is used to determine whether the result is a pass or a fail, can be used. While this approach is truly non-targeted and potentially sensitive to any adulterant, there is no indication of why a failing sample has failed (no identification of the adulterant) and, because the method makes no use of the adulterant spectrum, the sensitivity cannot be expected to be as high as a quantitative method.