Partial least squares with discriminant analysis and UV– visible spectroscopy for qualitative evaluation of Arabica and Robusta coffee in Lampung

Abstract

Arabica coffee is considered to be of better quality than Robusta coffee. It has superior taste and aroma, better than Robusta coffee. To develop an authentication system for Arabica coffee, it is highly necessary to discriminate between pure Arabica coffee and Arabica adulterated with Robusta coffee. Ground roasted coffee samples are most difficult to discriminate from each other: visual inspection by the naked eye or even machine vision methods becomes very problematic. For this reason, we here propose a relatively new analytical method based on UV-visible spectroscopy for discrimination the pure and adulterated Arabica ground roasted coffee. In this study, 100 samples were used as samples with different degrees of adulteration (0%-60% of Robusta concentration in an Arabica-Robusta coffee blend). Spectral data of samples were acquired using a UV-visible spectrometer in the range of 190-1100 nm (Genesys 10s, Thermo Scientific, USA). Partial least square discriminant analysis (PLS-DA) was applied to discriminate between the pure and adulterated Arabica coffee based on UV-visible spectra data. Several pre-processing spectra were also tested to determine which one provides an appropriate discrimination model. The PLS-DA model has coefficient of correlation 0.89 (R 2 = 0.79) with low Root mean square error of calibration (RMSEC) 0.226. The full-cross validation resulted in Q 2 = 0.74 and low Root mean squared error of cross-validation (RMSECV) 0.254. Using this PLS-DA model, a total rate of correct classification of 97.5% was obtained in the prediction set. In conclusion, UV-visible spectroscopy in tandem with PLS-DA is a promising analytical method for differentiating between pure and adulterated Arabica ground roasted coffee.

Quality Control and Authentication of Black Betel Leaf Extract (Piper acre Blume) from East Kalimantan as an Antimicrobial Agent Using a Combination of High-Performance Liquid Chromatography and Chemometric Fourier Transform Infrared

“…The characteristics between the groups were almost the same and the response values were not much different. These results also show that there are groups of samples with response values close together and, even if the buildup occurs, they have chemical characteristics that are very similar to one another [32][33][34][35].…”

“…In addition, the combination modeling was also able to identify piperenamide A compounds based on the number of levels in different samples. Therefore, piperenamide A can be considered as a characterization compound for black betel [32][33][34][35].…”

Discrimination between arabica and robusta coffees using NIR-integrating sphere spectroscopy coupled with hierarchical clustering analysis

“…However, most of the reported studies are laborious, time-consuming, and expensive. On the other hand, discrimination between the arabica and robusta coffee using spectral data as discriminator have been well documented by using UV-visible to NIR spectroscopy [1,[11][12][13]. UV-visible spectroscopy is cheaper in instrumentation but laborious in sample preparation [11].…”

“…On the other hand, discrimination between the arabica and robusta coffee using spectral data as discriminator have been well documented by using UV-visible to NIR spectroscopy [1,[11][12][13]. UV-visible spectroscopy is cheaper in instrumentation but laborious in sample preparation [11]. NMR [1], mid-infrared [12], and NIR spectroscopy [13] are available, however, with expensive spectrometers.…”