Model based investigation of water desorption and infrared spectral properties in cupuassu (Theobroma grandiflorum L.) pulp

• Insights into cupuassu pulp’s water desorption enhance drying efficiency in food processing. • The Oswin model provided the best fit to represent cupuassu pulp’s Type III desorption behavior. • Drying energy requirement was quantified as 320.33 kJ kg –1 for moisture reduction from 6.62 to 0.09 kg kg –1 dry basis. • Mid-infrared spectroscopy combined with Principal Component Analysis differentiated pulp samples from different cultivation zones. • Integrated desorption and spectral analyses support the design of energy-efficient drying and traceability tools. Cupuassu pulp, a major by-product of its processing industry, holds great potential as a sustainable upcycled food ingredient. However, its water desorption isotherms and mid-infrared spectral properties remain insufficiently characterized, hindering accurate estimation of drying energy demands and bioactive compound preservation. Therefore, in this research, the experimental assessment and mathematical model-based investigation of water desorption and infrared properties of cupuassu pulp were addressed. Water desorption isotherms were experimentally determined at 25 °C in the range of water activities of 0.1–0.99 using the dynamic dew point (DDI) method. Mid-infrared spectra were acquired in the wavenumber range of 4000–650 cm –1 using the Attenuated Total Reflectance Fourier Transform Infrared (ATR-FTIR) spectroscopy method. Desorption isotherms were modeled using ten conventional sorption equations, while drying requirements were assessed through Gibbs free energy analysis. Principal Component Analysis (PCA) was calibrated on mid-infrared data to elucidate the influence of growing zones on spectral properties of cupuassu pulp. The Oswin model provided the best fit to the water desorption isotherms (adjusted coefficient of determination; R 2 adj >99 %). Gibbs free energy analysis indicated that the drying process of cupuassu pulp required 320.33 kJ kg −1 to reduce moisture content and water activity from 6.62 kg kg –1 dry basis and 0.99 a w to 0.09 kg kg –1 dry basis and 0.1 a w . PCA modeling effectively characterized the chemical profile and differentiated pulp samples from various cultivars. This approach elucidates drying energy requirements and ensures cupuassu pulp traceability via its infrared spectral fingerprints, thereby fostering its valorization as a sustainable food ingredient.