The objective of this study was to obtain and characterize a freeze-dried powder of taioba leaves (Xanthosoma t aioba) as a source of bioactive substances and potential food applications. The fresh leaves were cooked and lyophilized for further analysis. Its bioactive contents expressed in flavonoids, ascorbic acid and total polyphenols were quantified. Total antioxidant activity (TAA) was determined by ABTS assays. The chemical composition via infrared spectroscopy (FTIR) and its microstructure were visualized by scanning electron microscopy (SEM). The bioactive contents of flavonoids, ascorbic acid and polyphenols were 17.15 mg/100 g, 58.3 mg/100 g and 24.15 mg Eq. of gallic acid/100 g, respectively. A high content of TAA was found. ABTS 37.35 (µg TE/g). The FTIR spectrum revealed high-intensity bands at 3350 cm -1 , 2928 cm -1 , 1637 cm -1 , and 1055 cm -1 related to vibrations associated with typical bands of −OH groups present in cellulose membranes, hemicellulose, carbohydrates, lignin and water. The micrographs showed irregular structures of the ground leaves with a fibrous structure. These results indicate a high potential of this raw material in food formulations as a source of bioactives suitable for applications in various industrial segments.

Annatto seed oil (ASO) and cupuassu seed fat (CSF) were combined at the ratios: 30:70, 50:50, and 70:30 (% w/w). Their fatty acid profile, nutritional quality, FTIR (Fourier Transform Infrared) spectra, and rheological behavior were evaluated. ASO increased the content of polyunsaturated fatty acids in the blends; whereas CSF conferred higher contents of monounsaturated fatty acids. The blends exhibited low atherogenicity and thrombogenicity indices, suggesting nutritional advantages. The Newtonian fluid behavior and FTIR results suggested that mixing ASO and CSF at different proportions did not affect the functional groups. ASO showed an activation energy value which indicated that this fat viscosity was more sensitive to temperature changes. The Newtonian model proved to be suitable to describe the behavior of samples, according to statistical fit parameters R2, χ2, and RSS. The resulting blends presented improved physicochemical properties and nutritional attributes, indicating their feasibility for the development of new products.