Extractability, selectivity, and comprehensiveness in supercritical fluid extraction of seaweed using ternary mixtures of carbon dioxide, ethanol, and water.

It is well-known that an ideal extraction method enabling quantitative analysis should give complete extraction of the target analytes as well as minimal co-extraction of unwanted matrix substances. If the extraction method is part of a nontarget screening protocol, the desired analytes can differ widely in terms of chemical properties. In chromatography, terminologies such as recovery, selectivity, and comprehensiveness are well-established and can easily be determined. However, in extraction, these concepts are much less developed. Hence, the aim of our research is to develop and scrutinize theory in extraction with respect to numerical descriptors for extractability, selectivity, and comprehensiveness. Our approach is based on experiments determining the extractability of target analytes and selected interferences. As a case study, we use a pooled sample of three species of seaweed (Alaria esculenta, Laminaria digitata and Ascophyllum nodosum). Target analytes are β-carotene, fucoxanthin, δ-tocopherol, and phloroglucinol; and selected interferences are carbohydrates, proteins, ash, arsenic, and chlorophyll a. As a "green and clean" extraction technique, supercritical fluid extraction (SFE) using mixtures of CO, ethanol and water were explored using a design of experiment. The temperature was varied between 40-80°C, and the pressure was held constant at 300 bar. Obtained results clearly demonstrate that highest relative selectivity was achieved with CO containing only 5 vol% of ethanol and no water, which primarily enabled high extractability of β-carotene, and yielding an extract free of carbohydrates, proteins, and toxic metals such as arsenic. Best methods for highest extractability of the other target analytes varied quite widely. Analytes requiring the highest water content (fucoxanthin and phloroglucinol), also resulted in the lowest relative selectivity. Maximum relative comprehensiveness was achieved using CO/ethanol/water (40/55/5, v/v/v) at 70°C and 300 bar. Our study demonstrates the feasibility of using relative quantitative descriptors for extractability, selectivity, and comprehensiveness, in optimization strategies for analytical extractions.