Classification of commercial charcoal for domestic use by near infrared spectroscopy

Abstract

Charcoal for domestic use presents great variation in quality because in most furnaces the carbonization process is difficult to and often inadequate raw material is used, generating a heterogeneous product. Near infrared (NIR) spectroscopy is a fast and reliable method to classify bio-based materials. Thus, the aim of this study was to apply NIR spectroscopy coupled with multivariate statistics in order to classify commercial vegetable charcoal for domestic use into categories and estimate its quality. Seventy-six charcoal specimens from nine producers were selected for NIR spectra acquisition on the transverse (TV) and rolling surface (RS) using the integrating sphere and fiber optic probe on raw (untreated) and sanded charcoal. Principal Component Analysis (PCA) of the spectra was performed to verify possible clusters among producers or quality classes (in terms of FCC levels). Discriminant Analyses based on Partial Least Squares (PLS-DA) were performed to classify the charcoals according to their producers and to predict their quality class. The PCA of spectra was not able to distinguish groups indicating high heterogeneity between treatments. However, PLS-DA models correctly classified up to 95% of the charcoal specimens both by producers and quality classes using spectra obtained by the integrating sphere or fiber optic probe. NIR spectroscopy coupled with multivariate analyses presented potential to be an efficient and rapid technique to classify charcoal. PLS-DA models can be applied in unknown charcoal specimens for reliable classification.