Available online at www.sciencedirect.com J. of Supercritical Fluids 43 (2008) 447–459 Study of the phase equilibrium formed inside the flash tank used at the separation step of a supercritical fluid extraction unit Thais M. Takeuchi a , Patr´ ıcia F. Leal a , Rog´ erio Favareto b ,L´ ucio Cardozo-Filho b , Marcos L. Corazza c , Paulo T.V. Rosa a,1 , M. Angela A. Meireles a,∗ a LASEFI-DEA/FEA (College of Food Engineering), UNICAMP (State University of Campinas), Cx. Postal 6121, 13083-862 Campinas, S˜ ao Paulo, Brazil b Universidade Estadual de Maring´ a (UEM), Department of Chemical Engineering, Av. Colombo, 5790 Bloco D-90, 87020-900 Maring´ a, Paran´ a, Brazil c Universidade Integrada do Alto Uruguai e das Miss˜ oes-Campus de Erechim (URI), Department of Food Engineering, Av. Sete de Setembro, 1621 Bloco 09, CEP 99700-000 Erechim, Rio Grande do Sul, Brazil Received 20 February 2007; received in revised form 7 August 2007; accepted 10 August 2007 Abstract In the present work the influence of a non-ideal separation step of a supercritical fluid extraction (SFE) unit was studied; the solvent used was carbon dioxide. The behavior of clove bud (Eugenia caryophyllus), vetiver grass (Vetiveria zizanioides), and fennel (Foeniculum vulgare) was analyzed. The starting point was a previous study on the same subject, which considered that no solute is lost in the vapor phase and a fixed fraction of 2% of the CO 2 is lost within the heavy phase. The flash separation step was simulated using the SuperPro Designs 6.0 ® Software, which calculates the phase equilibrium that occurs during the separation step using the Peng–Robinson equation of state. Experimental data for extraction kinetics from vetiver grass were obtained at 313 K and 20 MPa. Phase equilibria for CO 2 /vetiver extract were measured at pressures from 7.5 to 30 MPa and temperatures of 303.2, 318.2 and 333.2 K; CO 2 weight fraction varied from 0.1 to 0.99. The flash tank experimental data was obtained for fennel at 4 MPa and 293 K; the resulting anethole loss was 1.45%. The costs of manufacturing (COM) of the SFE extracts were determined according to Turton et al. using Tecanalysis v 1.0; the influence of the non-ideal flash separation step on COM was studied for clove buds. © 2007 Elsevier B.V. All rights reserved. Keywords: Clove; Fennel; Phase equilibrium; Separation step; Supercritical fluid extraction; Vetiver 1. Introduction Questions related to the use of techniques that avoid or min- imize damages to the environment have been fully debated in the past few years. Supercritical fluid extraction (SFE) embodies processes that use fluids at high pressures as extraction solvents (carbon dioxide being the most important example) and is known as a clean technology. The resulting products are free of toxic residues and generally present high quality when compared to products obtained by conventional techniques. Quality is a characteristic that will be more and more requested by the consumption market in the near future. This fact associated with today’s pursuit for healthy products makes the ∗ Corresponding author. Tel.: +55 19 3521 4033; fax: +55 19 3521 4027. E-mail address: meireles@fea.unicamp.br (M.A.A. Meireles). 1 Current address: Institute of Chemistry, UNICAMP, Cx. Postal 6154, CEP 13084-862, Campinas, SP, Brazil. production of vegetable extracts by supercritical fluid extraction a very interesting option for the industrial sector. These tendencies can be thoroughly explored by countries such as Brazil, which has appropriate climate and soil for agri- culture, and has the world’s largest biodiversity, what could provide high quality and low-cost raw material. However, to take advantage of its potential, the country needs to develop and/or to adapt technology that is economically viable and ecologically correct. In that context, several studies have been fulfilled in order to investigate the physicochemical phenomena that occur dur- ing the extraction process. Two major groups of experiments have been used: the global yield isotherms (GYI), and the over- all extraction curves (OECs). It becomes clear that using GYI to determine process temperature and pressure is convenient if you consider the following two points: (i) global yield (max- imum yield of extract that can be obtained after exhaustive extraction) is an intensive property [1] that depends only on process pressure and temperature, which are difficult to select 0896-8446/$ – see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.supflu.2007.08.002