In Figure 1, samples from regions within China are closer to each other compared to the two Egyptian regions (Faiyum and Sikkim). Clear separation of each group is also achieved within the different Chinese regions. These results indicate that the elemental composition of teas is influenced by where the tea leaves are grown. Along the first dimension, explaining 70 percent of the total variance ratio, macro elements like Na, Mg, and Ca, micro elements like Sr, Zr, and Mo, as well as trace elements, including the rare earth elements La and Ce, were found to be higher in the Faiyum teas. Meanwhile, Mn, Ni, Zn, Rb, and Re were higher in the other teas, leading to the clear separation between the different regions.
A similar clear separation is shown in Figure 2, where teas were classified by type. The Moroccan mint teas clearly separate from the other teas, along the first dimension, explaining 68 percent of the total variance ratio. Further, green teas (i.e., simply green, green, and gunpowder green) were located closer to each other, with the gunpowder green teas being closer towards the black and fermented teas (i.e., Chinese breakfast, oolong, and Pu-erh).
Discriminating elements along the first dimension include the macro elements Mg, Na, Ca, and Fe, micro elements like Sr, Mo, and Hf, and rare earth elements like Ce, Pr, Nd, Sm, Eu, and Gd. All of these elements were higher in the Moroccan mint teas, while Mn, Ni, Cd, and Re showed higher levels in the green, black, and fermented teas.
Along the second dimension, where green teas separate from black and fermented teas, levels of Rb, Mn, W, Re, and Tl were higher in the black and fermented teas.
Once a model has been established based on a full range of samples, elemental profiling can be used to identify the geographical origin of specialty foods and beverages.
In this study, multiple elements were determined in digests of 29 tea samples using ICP-MS to demonstrate the capability of multi-elemental profiling for authentication purposes. The teas, which were sourced from a reliable importer, included different varieties of tea, teas grown in different growing-regions of southern China, and teas from Egypt.
The methodology could be run on a single quadrupole ICP-MS. However, the use of a triple quadrupole ICP-MS provides better detection limits and accuracy for some analytes that are prone to complex interferences in certain matrices (e.g., rare earth elements).
Easy-to-use MPP chemometric techniques were applied to process the large dataset and to visualize the differences between samples. An exploratory analysis with discriminant analysis as a supervised multivariate analysis showed a clear separation of teas by region and by type.
Future studies will aim to strengthen the model with additional data from the analysis of more teas from different geographical origins, as well as tea mixtures. A market basket study on all commercially available Pu-erh teas can then be carried out. The method has the potential to quickly check the authentication of tea on a routine basis.
Dr. Nelson is a market development spectroscopy scientist for Agilent CrossLab Group. Reach her at firstname.lastname@example.org. Dr. Hopfer is an assistant professor of food science at the Department of Food Science for Pennsylvania State University. Reach her at email@example.com.