Authentication of Six Indonesian Ground Roasted Specialty Coffees According to Variety and Geographical Origin using NIR Spectroscopy with Integrating Sphere

Abstract

Several factors such as type of bean (Arabica or Robusta), geographical origin (elevation, soil type, temperature, and solar radiation), and agronomic/postharvest practices (fertilization, bean processing, fermentation, and type of roasting) have a great influence on the quality of final taste of the coffee. In this research, an authentication of six Indonesian ground roasted specialty coffees according to different in variety and geographical origin using NIR spectroscopy with integrating sphere was evaluated. A total of 70 coffee samples of Arabica and Robusta coffee from different geographical origins were used as samples. NIR spectral data in the range of 1175.79 nm to 1651.222 nm were obtained using a portable NIR spectrometer equipped with an integrating sphere. The unsupervised classification was performed using PCA and supervised classification was conducted using the SIMCA method. The result of PCA shows that the samples were well clustered according to variety and geographical origin along the PC1 and PC2 axis both using original and preprocessed spectra. The SIMCA performed good results both for original and preprocessed spectral data, most of the testing sample set were properly classified to their corresponding classes.