The scope of the authenticity and adulteration problem ranges from an accidental substitution of one product for another (e.g., two crops physically resemble one another and are mislabeled during or after harvest) to fraud through intentional substitution of one lower-cost commodity for another to downright threats to public health involving addition of components to boost certain attributes of the product (e.g., melamine in milk to increase the measured results of a non-specific protein determination). The natural products industry is required to confirm authenticity of ingredients and needs guidance on the best practices for detecting known adulterants as well as for identifying inconsistencies in a product profile that may indicate potential adulteration.
In 2018, Exercise O of the DSQAP included a study to evaluate the ability of various authenticity methods to identify adulteration of botanical samples. Participants were provided 16 samples in Set A containing Ginkgo biloba plant materials and Set B containing Ginkgo biloba extract materials. Each sample contained between 0 % and 15 % (by weight) of Sophora japonica extract. The results from participants indicated that no single method was able to correctly identify the presence of Ginkgo biloba, the plant part, and the level of adulteration in every sample. The laboratories that were most successful in this study utilized multiple fit-for-purpose methods. The full study report is available on the NIST website (NIST IR 8266).
Building on experience from the food metrology program, NIST is exploring the use of multi-technique chemical fingerprinting for food authenticity using honey to pilot the effort. Initial chemical analysis of a wide variety of honey samples will involve 1H-NMR spectroscopy. Additional analytical techniques that may be explored in this effort include vibrational spectroscopy (IR, near-IR, Raman), UV-Vis spectroscopy, elemental profiling via ICP-MS, isotope ratio mass spectrometry, and chromatography with various detection modes. The resulting data will be investigated using both traditional data analysis and chemometric modeling approaches aimed at understanding the range of chemical signatures associated with natural variability, origin and potential adulteration with exogenous sugar sources, additives and colorings. From these data, sampling models and recommended statistical approaches to building an appropriate model will be investigated.