Classification of roasted coffee bean products using laser-induced breakdown spectroscopy: a novel variable selection...

Area of Science:

• Food Science
• Analytical Chemistry
• Spectroscopy

Background:

• Coffee bean authentication is crucial for industry authenticity and consistency.
• Existing methods like spectroscopy and chromatography have limitations, including complex sample prep or molecular data instability during roasting.
• Laser-induced breakdown spectroscopy (LIBS) offers fast, minimally invasive elemental analysis suitable for food authentication.

Purpose of the Study:

• To evaluate the feasibility of combining LIBS with the k-nearest neighbors (k-NN) algorithm for classifying roasted coffee bean products.
• To identify key elemental discriminatory variables for an accurate coffee classification model.
• To assess the stability of the chosen analytical markers during the coffee roasting process.

Main Methods:

• Collected LIBS spectra from 12 roasted coffee bean products from South Korean markets.
• Analyzed spectra for elemental emission peaks (Li, Na, K, Rb, Mg, Ca, C, H, O) and molecular bands (CN, C2).
• Employed the k-NN algorithm with a novel 'inter-to-intraclass variation ratio' to classify beans based on elemental intensities, particularly Li, Na, and Rb.

Main Results:

• LIBS successfully detected key elements and molecular bands in roasted coffee beans.
• The emission intensities of Li, Na, and Rb were identified as the most significant discriminatory variables.
• The k-NN model achieved high classification accuracies: 96.0% with k=1 and 98.5% with k=3 after scaling.
• The elemental composition (Li, Na, Rb) proved stable and unaffected by the roasting process.

Conclusions:

• LIBS combined with k-NN is a practical and efficient tool for authenticating roasted coffee products.
• The method leverages high sensitivity to alkali metal elements for rapid and accurate classification.
• This approach offers a stable alternative to methods relying on labile molecular information affected by roasting.