J. of Supercritical Fluids 35 (2005) 119–127
Mass transfer modeling of apricot kernel oil extraction with supercritical carbon dioxide ¨ S.G. Ozkal, M.E. Yener ∗ , L. Bayındırlı Food Engineering Department, Middle East Technical University, 06531 Ankara, Turkey Received 15 April 2004; received in revised form 6 December 2004; accepted 14 December 2004
Abstract Effects of process parameters on extraction of apricot (Prunus armeniaca L.) kernel oil with supercritical carbon dioxide (SC-CO2 ) were investigated. The parameters included particle size (mean particle diameter < 0.425–1.5 mm), solvent flow rate (1–5 g/min), pressure (300–600 bar), temperature (40–70 ◦ C) and co-solvent concentration (up to 3.0 wt.% ethanol). The model of broken and intact cells represented the apricot kernel oil extraction well. Grinding was necessary to release the oil from intact oil cells of kernel structure. About 99% apricot kernel oil recovery was possible if particle diameter decreased below 0.425 mm. Two extraction periods were distinguished. The released oil on the surface of particles was extracted in the fast extraction period while the unreleased oil in the intact cells was extracted in the slow extraction period. The amount oil recovered in the slow extraction period was negligible compared to the amount recovered in the fast extraction period. Extraction rate in the fast extraction period increased with increase in solvent flow rate, pressure, temperature, and ethanol addition. Mass transfer coefficients in the fluid phase and in the solid phase changed between 0.7 and 3.7 min−1 , and between 0.00009 and 0.0005 min−1 , respectively. Mass transfer coefficient in the fluid phase increased with decrease in particle size and pressure, and with increase in solvent flow rate, temperature and ethanol concentration. © 2004 Elsevier B.V. All rights reserved. Keywords: Supercritical carbon dioxide; Apricot kernel oil; Extraction; Mass transfer; Solubility
1. Introduction Supercritical fluid extraction (SFE) eliminates the disadvantages of conventional solvent extraction which leads to degradation of heat sensitive compounds and leaves traces of toxic solvents in the solute, as well. This is a concern for food and medicinal extracts, because of the increasing regulations on solvent use. Besides achieving high yield and quality, SFE can be operated under a wide range of conditions to selectively extract specific end products with improved functional and/or nutritional characteristics for use in creating new formulated foods [1,2]. Extraction of oils from seed matrices with supercritical carbon dioxide (SC-CO2 ) is very common. Various applications include extraction of rapeseed [3], peanut [4], canola [5,6], almond [7], sesame [8], pistachio nut [9] and walnut [10] oils. ∗
Corresponding author. Tel.: +90 312 210 5630; fax: +90 312 210 1270. E-mail address:
[email protected] (M.E. Yener).
0896-8446/$ – see front matter © 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.supflu.2004.12.011
Mathematical models used for extraction of solutes from natural matrices are classified as (1) empirical models [11–13], (2) models based on heat transfer analogy [14,15], (3) shrinking core model [16–19] and (4) models based on differential mass balance [7,20–22]. The use of differential mass balance equations is common for extraction of essential oils and especially seed oils. Most of these models include the resistances in both or in one of the bulk phases. They take into account particle and bed characteristics via porosity and diameter. Although the models imply many assumptions and/or determination of several coefficients involved in the equations, they reflect the various mechanisms that contribute to overall behavior of an extraction process [12]. Some authors modeled the extraction of oil from oilseeds taking into account only the mass transfer resistance in the fluid phase [5,23,24]. On the other hand, mass transfer resistance in the solid phase was found to be important in the extraction of sage leaves essential oil [25]. The physical representation of the solid matrix is first included in the model of broken and
¨ S.G. Ozkal et al. / J. of Supercritical Fluids 35 (2005) 119–127
120
Nomenclature a E G h J k m N Q t x w y yr Y Z
specific interfacial area (area/volume of fixed bed) mass of extract (mass) grinding efficiency, fraction of the total oil released (0 ≤ G ≤1) dimensionless axial coordinate (1 ≤ h ≤1) mass transfer rate per volume of fixed bed (mass/volume time) mass transfer coefficient (length/time) oil recovery (mass of oil extracted/ mass of initial oil in kernel feed) mass of the oil free solid phase (mass) mass flow rate of solvent (mass/time) time (time) solid phase concentration (mass/mass) oil yield (mass of oil extracted/mass of kernel feed) solvent phase concentration (mass/mass) solubility (mass/mass of solvent) parameter of the slow extraction period, Eq. (11) parameter of the fast extraction period, Eq. (10)
Greek letters ε void fraction (volume/volume of bed) ρ density (mass/volume) ψ dimensionless time Superscript + at interfacial boundary Subscripts f solvent phase k boundary between the fast and slow extraction periods s solid phase 0 initial condition 90 time is 90 min
intact cells where mass transfer resistances both in the fluid and in the solid phases are considered [20,26]. The model was successfully used for the extraction of essential oil from black pepper [27], chamomile [28], and oil from almond [7], various seeds such as grape [29], sunflower, coriander, tomato, and peanut [22]. Apricot kernel contains about 48% oil of which 68% is oleic and 25% is linoleic acid [30]. This work aimed at the recovery of apricot kernel oil by SC-CO2 extraction and the analysis of the mass transfer. The effects of particle size, solvent flow rate, extraction pressure and temperature, and co-solvent concentration on oil yield were investigated. The
mass transfer coefficients were determined using the model of broken and intact cells. 2. Experimental 2.1. Materials Unshelled and dried apricot kernel samples were obtained from local market and stored at +4 ◦ C in sealed glass jars. The moisture and oil contents were 3.9 and 48.1%, respectively. The samples were ground into small sizes by using kitchen type grinder (Arc¸elik, Turkey), sieved and fractionated according to particle size by certified test sieves (Endecotts Ltd., London, England). Sieving was performed by a shaker (Octagon 200, Endecotts Ltd., London, England). The fractions between two successive sieves were assigned a size number as shown in Table 1. CO2 was purchased from Habas¸ (Turkey). 3. Methods 3.1. Supercritical fluid extraction system The solubility measurements and SFE experiments were performed by using Supercritical Fluid Extraction System (SFX System 2120, Isco Inc., Lincoln, NE). The system consists of an extractor (SFX 220) and two syringe pumps (Model 100DX). The pumps could pump up to 690 bar with flow rates ranging between 0.1 l/min and 50 ml/min. The volumetric flow rate is measured as liquid CO2 at the extractor pressure and at 5 ◦ C, which is the cooling temperature of the pump. The temperature in the extractor chamber could be controlled up to 150 ◦ C and the system enables addition of co-solvent if required. The extractor is a 10 ml steel cartridge where SC-CO2 flows downward. The extract is passed through a coaxially heated adjustable restrictor and the solute is precipitated in test tubes. 3.1.1. Extraction Effects of particle size (sizes 1, 2, 3 and 4 (Table 1)), solvent flow rate (1, 2, 3, 4 and 5 g/min), pressure (300, 375, 450, 525 and 600 bar), temperature (40, 50, 60 and 70 ◦ C) and co-solvent (ethanol) concentration (0, 0.5, 1.0, 1.5 and 3.0 wt.%) on oil yield were investigated. The standard extraction conditions were selected as, particle size 2, 3 g/min solvent flow rate, 450 bar, 50 ◦ C and 0% ethanol concentration. One parameter was changed at a time while the other parameters were kept constant at these standard conditions. In the extractions 5 g of apricot kernel sample was used and the extract was passed through a coaxially heated adjustable restrictor where temperature was set above 110 ◦ C. The oil was precipitated in test tubes containing glass wool. The yield was determined gravimetrically at definite time intervals. Each extraction was continued until no significant
¨ S.G. Ozkal et al. / J. of Supercritical Fluids 35 (2005) 119–127
121
Table 1 Parameters of the mass transfer model at different particle sizes (extraction conditions: P = 450 bar, T = 50 ◦ C, Q = 3 g/min, ρf = 951 kg/m3 ) Particle size
Dpm (mm)
yr (g/g)
G
tk (min)
wk (g/g kernel)
m90 (g/g initial oil)
kf a (min−1 )
ks a (min−1 )
AAD (%)
1 2 3 4
