Coffee varietal differentiation based on near infrared spectroscopy

Area of Science:

• Analytical Chemistry
• Food Science
• Spectroscopy

Background:

• Accurate coffee varietal identification is crucial for quality control and authenticity.
• Distinguishing between Arabica and Robusta coffee, and their blends, presents analytical challenges.
• Near-infrared spectroscopy (NIRS) offers a rapid, non-destructive method for chemical analysis.

Purpose of the Study:

• To evaluate the efficacy of NIRS for discriminating pure Arabica and Robusta coffee varieties.
• To assess NIRS capabilities in classifying coffee blends with varying Robusta content.
• To investigate the impact of the Direct Orthogonal Signal Correction (DOSC) pre-processing method on classification accuracy.

Main Methods:

• Acquisition of 191 roasted coffee NIR spectra from pure varieties and blends (0-60% Robusta).
• Application of the Direct Orthogonal Signal Correction (DOSC) pre-processing technique to remove irrelevant spectral information.
• Development of classification models using raw and DOSC-corrected NIR spectra with the potential functions method.

Main Results:

• DOSC pre-processing significantly improved the accuracy of coffee varietal classification models.
• NIRS, especially after DOSC correction, demonstrated high suitability for identifying coffee varietal composition.
• Classification models effectively discriminated between pure varieties and blends across a range of Robusta percentages.

Conclusions:

• The DOSC method is a valuable pre-processing step for enhancing NIRS-based coffee varietal identification.
• NIRS is a powerful tool for the rapid and accurate authentication of coffee varieties and blends.
• This approach facilitates improved quality control and traceability in the coffee industry.