Abstract/Summary

Vegetable by-products are abundant in interesting phytochemicals, such as phenolics and carotenoids, which exhibit high antioxidant activity and are valuable in the production of foods, chemicals and cosmetics. Different techniques have been used for the extraction of these molecules, with Supercritical Fluid Extraction (SFE) being one of the most sustainable, efficient and reliable approaches to this end. A variety of fruit and vegetable matrices have already been successfully extracted by SFE, including the targeting of such bioactives. However, there is a lack of studies dealing with further purification of the extracts for obtaining more valuable fractions that could potentially find finer applications as natural additives in the food and nutraceutical sectors. The aim of this work was to assess and optimise an integrated protocol for carotenoid extraction and purification from carrot peels and study the application of the optimised conditions to other carotenoid-rich vegetable matrices. The extraction process was evaluated through a Central Composite Design of Experiments at different temperatures, pressures and co-solvent concentrations, as well as by kinetics experiments and modelling, scalability potential and extract characterisation. The statistical and kinetic extraction models were validated successfully and the optimised conditions were: temperature 59.0 °C, pressure 349 bar, 15.5% ethanol, 15 g/min of CO2 flow rate, and total extraction time of 30 min. These resulted in a carotenoid recovery of 86.1% and the process was shown to be potentially scalable, since recoveries as high as 96.7% were observed in runs performed with 10-fold the initial sample mass. Furthermore, the supercritical-fluid carotenoid-rich extracts were purified by Hydrophobic Interaction Chromatography, through the optimisation of batch and in-column adsorption parameters such as resin capacity, kinetics and elution. The adsorbent employed in the purification stage showed an adsorption capacity of 10.4 μg of carotenoid per mg of resin and the global yield of the process was of 88.4% for total carotenoids and 92.1% for carotenes. Analyses of the final eluate confirmed the evolution of the purification, leading to a 2.1 and 4.7-fold increase in antioxidant activity and carotenoid concentration, respectively. Finally, the SFE-optimised extraction conditions were extrapolated to other vegetables samples, inherently rich in carotenoids. The model was deemed applicable to other vegetables, such as sweet potato, apricot, pumpkin, green and yellow pepper. It can be concluded that the protocol set up in this work can be employed with confidence to efficiently extract and purify carotenoids from vegetable matrices and represents a tangible alternative for waste valorisation.