Statistical mixture design investigation of CO2–Ethanol–H2O pressurized solvent extractions from tara seed coat

Tara (Caesalpinia spinosa) fruit seed coat is an abundant Peruvian agro-industrial residue that could be explored as a raw material for the production of valuable phenolic- and antioxidant-rich extracts. In this work, a mixture design was applied to optimize solvent mixtures for selective extraction of phenolic compounds from tara fruit seed coat. Pressurized solvent extractions (PSE) were performed at 313 K and 20 MPa, using homogeneous CO2, ethanol (EtOH) and H2O solvent mixtures. Extraction kinetics was obtained and phenolic contents, antioxidant activities (measured by the coupled oxidation of β-carotene and linoleic acid) and anti-inflammatory activities (measured by the lipoxygenase inhibition activity) of extracts were evaluated. A linear mixture model predicted total yields and it was verified that H2O was the most effective solvent to obtain high extraction yields. A quadratic model predicted extracts’ phenolic contents and it was verified that H2O and CO2 had an antagonistic effect and that maximum phenolic contents can be expected for EtOH-rich mixtures. A special cubic model predicted anti-inflammatory activities and the maximum activities were also predicted for EtOH-rich mixtures. For the antioxidant activity, no model revealed to be significant (at a 95% confidence level). Nevertheless, high antioxidant activities were observed after 3 h of assay (above 80% of oxidation inhibition for nine out of eleven extracts) and even after 6 h of assay (above 70%). The obtained results strongly suggest that tara seed coat is a residue that has great potential to be valorized for the production of phenolic-rich extracts presenting high anti-inflammatory and antioxidant activities.