ABSTRACT: Chilean hazelnut (Gevuina avellana) oil is highly appreciated in the cosmetic and pharmaceutical industries. Hazelnut oil (oil content calculated on 49% dry basis) is tradi- tionally obtained by pressing, a low-efficiency process that re- sults in a low-quality product. In this work, the conventional process was compared with two enzymatic alternatives in which commercial enzymes were used to increase the oil ex- traction yield: (i) extraction in aqueous medium and (ii) extrac- tion by pressing after an enzymatic treatment. The effect of vari- ous parameters on the extraction yield was studied to define the most satisfactory processing conditions. These included reac- tion time, temperature, enzyme concentration, and, in the aque- ous medium extraction process, the water/seed ratio, particle size, and pH. Although pressing is the better alternative, in both processes enzyme treatment improved extraction yields (94 and 98% for aqueous medium extraction and pressing after enzyme treatment, respectively, compared to 52% obtained in the con- ventional process). Moreover, the quality of the oil obtained is the same as or better than that of oil obtained by the conven- tional process. Paper no. J10261 in JAOCS 80, 33–36 (January 2003). KEY WORDS: Enzymatic extraction, enzyme-based oil extrac- tion, Gevuina avellana, hazelnut oil. Gevuina avellana, the hazelnut tree, is native to Austral Chile. The seed, which is the edible portion of the fruit, contains high protein and lipid contents, 12.4 and 49.3% on a dry basis, respectively (1). Owing to its high cost, hazelnut oil cannot compete with soybean or cotton oil for human con- sumption. However, because of its physicochemical proper- ties, it is best suited for applications in cosmetology, as it ab- sorbs short-wavelength UV radiation, allowing the passage of only that radiation producing a suntan without damaging the skin (2,3). Hazelnut oil contains a high percentage of un- saturated FA, so the skin readily absorbs it. For this reason, it is one of the major ingredients in the formulation of products such as suntan protective lotion and other protective creams (2,3); it is also used as a vehicle in the design of transport sys- tems for skin nutrients and protective and regenerating sub- stances in, e.g., creams, soaps, and shampoos (4,5). Chilean hazelnut oil is traditionally obtained by pressing. Before pressing, hazelnuts are peeled, to eliminate the skin and the cuticle, ground, and heat-treated. Oil and the residual hazelnut paste are the products of pressing. The oil extraction yield of this process is low, and the low-quality paste so pro- duced has a high residual oil content (6). Enzymes have been applied to the extraction of oil from sev- eral fruits and vegetables: in aqueous processes (7–13), in a sys- tem of low humidity for a later solvent extraction (14,15), or by pressing (16). In the present work, the oil extraction yield from Chilean hazelnuts obtained in two processes involving enzymes is compared with the conventional extraction process. In the first process an aqueous treatment is proposed, whereas in the sec- ond the enzymes are added to the hazelnuts prior to pressing. MATERIALS AND METHODS Hazelnut seeds were obtained from the south of Chile, sepa- rated from their rinds and husks, ground in a coffee mill or a mixer, and sieved to a particle size ≤1.4 mm. Total oil in the seed was measured by the Soxhlet extraction method (17). Three processes for oil extraction were performed, and the re- sults were compared: conventional (cold pressing), pressing with enzymatic pretreatment, and enzymatic aqueous pro- cessing. In a preliminary assay, the hazelnut mass was pre- treated with enzymes to evaluate their effects on oil extrac- tion, using three commercial enzymes: Olivex, Ultrazym, and Celluclast, all from Novozymes (Bagsvaerd, Denmark). The first two are cocktails designed to digest vegetable cell walls; they combine several activities including cellulase, pectinase, and hemicellulase; the last one contains a cellulolytic com- plex. Enzyme concentration is reported as vol/wt, expressed as % (where 1% = 1 mL/100 g, for example) as they are pro- vided in liquid form (vol) and are applied to a solid (wt) or referred to the mass of substrate. The overall enzymatic ac- tivity was measured by following the release of reducing sug- ars from the tissue with the dinitrosalicylic acid (DNS) method (19). All experiments were carried out in at least trip- licate. The SD for our results never exceeded more than 5%. The flow diagram for these processes is shown in Scheme 1. Conventional extraction. Samples of hazelnut seeds were used in a classical oil extraction. The seeds were dried at 60°C to 3–5% humidity and were then pressed in a hydraulic press (Carver Laboratory Press; Fred S. Carver Inc., Wabash, IN) at 39.2 MPa for 15 min at room temperature. The oil extrac- tion yield was reported as the difference between total and residual oil determined by the Soxhlet extraction method. Copyright © 2003 by AOCS Press 33 JAOCS, Vol. 80, no. 1 (2003) *To whom correspondence should be addressed at Instituto de Biotecnología UNAM, México, Apartado Postal 510-3, Cuernavaca, Morelos, 62271, México. E-mail: agustin@ibt.unam.mx Enzymatic Extraction of Oil from Gevuina avellana, the Chilean Hazelnut R.I. Santamaría a , C. Soto b , M.E. Zúñiga b , R. Chamy b , and A. López-Munguía a, * a Instituto de Biotecnología UNAM, Cuernavaca, Morelos, 62271, Mexico, and b Universidad Católica de Valparaíso, Valparaíso, Chile