Monitoring moisture content in parchment coffee beans during drying using Fourier Transform near infrared (FT-NIR) spectroscopy: A dataset for calibrating chemometric-based models for moisture prediction

Abstract

Maintaining the quality of coffee across each stage of the coffee value chain is critical, with proper bean drying being essential for preserving product shelf life and moisture stability. This work compiles a dataset collected during the mechanical drying process of parchment coffee beans, monitoring moisture content alongside their corresponding near-infrared (NIR) spectra. The aim was to evaluate the application of NIR spectroscopy for predicting moisture content during drying, leveraging NIR as a reliable, rapid, and non-destructive technology for routine monitoring of the coffee drying process. Drying kinetics of parchment coffee beans were determined using a mechanical coffee dryer, with moisture content gravimetrically monitored at various drying times. At each drying point, NIR spectra were acquired using a Spectrum Two N FT-NIR Spectrometer equipped with a high-resolution Indium Gallium Arsenide (InGaAs) detector, operating in diffuse reflectance mode. NIR spectra were collected over a wavelength range of 4000–12000 cm⁻¹ (830–2500 nm), with a 4 cm⁻¹ interval, 8 cm⁻¹ resolution, and 64 scans. This work explored moisture content from fresh coffee (52 % wet basis; w.b.) to 8 % w.b., examining spectral changes throughout the entire drying process. The compiled dataset includes experimental drying kinetics and FT-NIR spectra in Excel format, organized according to experimental conditions. This dataset provides a valuable foundation for further analysis and for calibrating predictive models of moisture content during coffee drying, highlighting the high potential of NIR spectroscopy for industrial-scale drying control and monitoring in the coffee industry.