The efficacy of ultrasound-assisted extraction (UAE) of fucoidan from four frozen thalli of Arctic brown algae (Fucales, Phaeophyceae) Fucus vesiculosus, Fucus distichus, Fucus serratus, and Ascophyllum nodosum was studied for the first time and compared with the heat-assisted extraction (HAE). The chemical composition and molecular weight distribution of fucoidans were characterized using HPLC-RID, HPSEC, ferric reducing antioxidant power (FRAP) assay, total antioxidant capacity (TAC). The composition of fucoidan was heterogeneous and depended on the extraction method and the species: F. vesiculosus, F. distichus, F. serratus and A. nodosum. UAE was more efficient for fucoidan extraction when compared with HAE. Two-level quarter fractional factorial design (FFD) was applied to optimize the extraction UAE process of fucoidan from frozen F. vesiculosus. Five parameters of extraction were studied: extraction time (20-60 min), ultrasonic amplitude (60-100 %), pH (2-6), liquid-to-raw-material ratio (10-30 mL g-1), and particle size (1-3 mm). The highest yield was achieved at the extraction time of 55 min, the ultrasonic amplitude of 80%, pH 3.4, liquid-to-raw-material ratio of 25 mL g-1 and particle size of 1.4 mm. It was found that the power law and the Peleg's model describe the kinetics of fucoidan extraction with the highest accuracy. The environmental friendliness of the proposed extraction technique is shown. Overall, a novel biorefinery process for fresh frozen algal biomass was established to produce a lipid fraction, mannitol, and two distinctive polysaccharide products such as fucoidan and alginic acid from a single seaweed feedstock and optimized fucoidan UAE extraction. Ultrasound was effective to enhance the extraction of fucoidan at a relatively low temperature from frozen algae thalli.
