Multivariate data visualisation methods based on multi-elemental analysis of wines and coffees using total reflection X-ray fluorescence analysis
Stephen J. Haswell, A. D. Walmsley · Journal of Analytical Atomic Spectrometry · 1998
The potential of applying TXRF for multi-elemental analysis and chemometric techniques for the classification of wines and coffees has been investigated. The wine samples were all of the same grape variety and coffee samples were sourced from various manufacturers. Elemental concentrations were determined using a vanadium standard. Principal components analysis (PCA) of the wine data showed that each wine type could be clearly distinguished, and cluster analysis showed the separation between the sample types. Star plots were produced for the wines and they showed that the variability in the replicate samples was much less than the differences between plots for each wine type. The coffee samples were not so clearly separated by PCA and cluster analysis, but most of the samples were found to be grouped by manufacturer, rather than country of origin. Again, it was demonstrated that star plots could be used to show the differences between samples. Clearly the application of simple chemometric techniques to multi-elemental analysis based on TXRF data is useful for the rapid screening and classification of samples such as wine and coffee and this type of methodology is enhanced through the use of descriptive visualisation techniques such as star plots.
Cited by 64 works. doi:10.1039/a705317g
Authors
References
- prediction of wine quality and geographic origin from chemical measurements by parital least-squares regression modeling
- Chemometric Characterization and Classification of Five Venetian White Wines
- Methodology for the selection of suitable sensors for incorporation into a gas sensor array
- Research on white and red wine blending in the production of rosé wines by means of the partial least squares method
- Chemometrics—the key to sensor array development
- Prediction of some physico-chemical parameters in red wines from ultraviolet–visible spectra using a partial least-squares model in latent variables