Phase behaviour of the pseudo-ternary system carbon dioxide + ethanol + fish oil at high pressures

1 Phase behaviour of the pseudo-ternary system carbon dioxide + ethanol + fish oil at high pressures Rodrigo Melgosa, Mª Teresa Sanz*, Ángela G. Solaesa, Sagrario Beltrán Department of Biotechnology and Food Science (Chemical Engineering Section) University of Burgos, Pza. Misael Bañuelos s/n 09001 Burgos, Spain *Corresponding author. Tel.: +34-947258810; Fax: +34-947258831; e-mail: tersa[email protected] ABSTRACT This work provides experimental fluid phase equilibrium data of the pseudo-ternary mixture CO2 + ethanol + fish oil, a system of interest in pharmaceutical and food-industry applications such as the production of omega-3-enriched lipid derivatives at mild, non-oxidative conditions. Experimental tie-lines were obtained by means of an analytical isothermal method with recirculation of the vapour phase. Measurements were carried out in the temperature range 323.15 K-343.15 K and at pressures from 10 MPa to 30 MPa. The Peng-Robinson equation of state coupled with the conventional van der Waals mixing rules with two adjustable parameters was used for experimental data correlation. Keywords Phase equilibrium, Supercritical carbon dioxide, Ethanol, Fish oil, Thermodynamic modelling, Peng-Robinson equation of state. 2 1. INTRODUCTION Dense carbon dioxide is expected to play an important role as a reaction medium in ecologically friendly processing. Enzyme-catalysed ethanolysis of lipid sources in supercritical carbon dioxide (SC-CO2) or CO2-expanded media can be used to improve the production of lipid derivatives, including concentrates of omega 3 polyunsaturated fatty acids (n-3 PUFAs) from fish oil [1,2]. However, the relatively low solubility of the reactants in SCCO2 limits the reaction performance in this medium; thus, a CO2-expanded media is preferred to minimise mass transfer limitations [3]. Understanding the phase behaviour of the ethanol + fish oil substrate mixture with CO2 would help in the selection of adequate ethanolysis conditions (pressure, temperature, ethanol-to-oil ratio, and amount of dissolved CO2 in the reaction mixture). This knowledge can be also extended to other potential applications and may be of interest in the fish oil industry, since SC-CO2 or CO2-expanded media can be used throughout all the n-3 PUFA concentration process, including the supercritical extraction of the fish oil [4], the refining step [5], the separation and fractionation of the reaction products [6,7], or the formulation of the final product by means of particle formation techniques [8]. Experimental data related with high-pressure phase equilibrium of the binary system CO2 + ethanol are extensively reported in the literature [9–14]. Besides, ternary and higher systems comprising CO2 and pure triglycerides, other lipid derivatives and their mixtures have been also investigated [15]. However, only a few data regarding the phase equilibrium of pseudoternary mixtures of CO2, ethanol, and edible oils can be found in the literature, and these data are usually related to the solubility of the lipid compounds in CO2 with ethanol as a co-solvent [16]. Several publications have previously reported fluid phase equilibrium of pseudo-ternary mixtures of CO2 + ethanol + vegetable oils. However, to our knowledge, this is the first work 3 dealing with oils rich in n-3 PUFAs from animal sources, such as fish oil. Geana and Steiner [17] reported fluid phase equilibrium data for the pseudo-ternary system CO2 + ethanol + rapeseed oil in the temperature range 313 K-353 K and at pressures from 6 MPa to 12 MPa, satisfactorily correlating the phase behaviour with the Peng-Robinson equation of state (PR EoS) [18] coupled with the conventional van der Waals mixing rules with two adjustable parameters (vdW2). Ndiaye et al. [19] studied the fluid phase equilibria of binary and ternary mixtures involving CO2, ethanol, soybean oil, castor oil, and their fatty acid ethyl esters. The pseudo-ternary system CO2 + ethanol + castor oil was studied at fixed ethanol-to-oil ratios, temperatures ranging from 313.15 K to 343.15 K and pressures from 2.13 MPa to 27.13 MPa. Experimental data were correlated both with PR EoS vdW2 and the Statistical Associating Fluid Theory (SAFT) [20] with one binary interaction parameter. Among these two models, the authors considered that SAFT EoS described better the phase behaviour of the pseudoternary system, yet they pointed out at some deviations from their experimental results, such as the over-prediction of the cloud point pressure at high ethanol ratios [20]. Hernández et al. [21] investigated the fluid phase equilibrium behaviour of the pseudo-ternary mixture CO2 + ethanol + sunflower oil at two different conditions of temperature and pressure (313.15 K and 13 MPa; 333.15 K and 20 MPa). A group contribution equation of state (GC EoS) [22] was used to correlate the experimental data. Two different sets of parameters were adopted for the interaction between the triglyceride and the alcohol groups, one of them corresponding to the Liquid + Liquid (L1+L2) 2-phase region, and the other to the Vapour + Liquid (V+L2) 2phase region. More recently, Dalmolin et al. [23] studied the phase transitions in the system CO2 + ethanol + rapeseed oil, at temperatures in the range 313.15 K-343.15 K and pressures up to 22.53 MPa. They found a 3-phase region with a Vapour + Liquid + Liquid (V+L1+L2) phase transition that occurred at higher pressures when increasing temperature, and satisfactorily explained their experimental results with the PR EoS vdW2 model. 4 In this work, the phase behaviour of the pseudo-ternary mixture CO2 + ethanol + fish oil in the temperature range from 323.15 K to 343.15 K and pressures from 10 MPa to 30 MPa has been determined by means of an analytical isothermal method with recirculation of the vapour phase (AnTVcir, as described by Dohrn and Brunner [24]). The main goal of the study involves a contribution towards understanding the phase behaviour of systems containing CO2, ethanol, and oils rich in n-3 PUFAs. The knowledge obtained will be useful in the development of applications involving these pseudo-ternary mixtures, such as the previously mentioned enzymatic reactions, supercritical extraction and fractionation, and particle formation techniques. 2. EXPERIMENTAL 2.1. Materials Fish oil was provided by AFAMSA S.A. (Pontevedra, Spain) being a mixture of tuna (Thunnus sp.) and sardine (Sardina pilchardus) refined oils. The fatty acid profile and free fatty acid content of the fish oil have been determined according to AOCS methods [25]. The fatty acid profile has been previously reported [26] and is also provided in Table 1. Free fatty acid content was found to be 0.29 ± 0.04 % oleic acid. Density of fish oil was also measured in an Anton Paar DMA 5000 instrument, finding values of ρ323.15 K = 906.53 kg·m-3 and ρ343.15 K = 895.83 kg·m-3 (u(ρ) = ± 0.05 kg·m-3). Absolute ethanol (0.999 mass fraction) was purchased from Merck KGaA. Carbon dioxide (0.999 mass fraction in the liquid phase) was supplied by Air Liquide S.A. (Spain). Compounds have been used as provided by the manufacturers without further purification. 5 The potential reactivity of fish oil in contact with ethanol could lead to some extent of transesterification and reaction products could be formed during phase equilibria measurements, mainly fatty acid ethyl esters (FAEE). Although reaction in the absence of catalysts is very slow, the presence of FAEE and other intermediate components was analysed after phase equilibria measurements by NP-HPLC. Chromatographic method is reported elsewhere [27]. FAEE content was found less than 0.001 mole fraction and it was considered not to affect the phase equilibria. Hidrolysis of the fish oil was also evaluated by means of the free fatty acid content [25], finding a maximum increase up to 0.81 ± 0.05 % oleic acid at the highest studied temperature (343.15 K). Fatty acid profile was also determined after phase equilibria measurements, finding no significant changes compared to the initial profile (Table 1). Additionally, since fish oil is rich in polyunsaturated fatty acids (Table 1) it is very prone to oxidation. Therefore primary oxidation was evaluated by means of peroxide value analysis [28] before and after phase equilibria measurements. Although PV slightly increased from 2.0 ± 0.2 meq O2/kg to 3.5 ± 0.1 meq O2/kg, oxidation products are minor components present in small amounts that would not affect the phase equilibria of the system. 2.2. Apparatus and procedure A schematic diagram of the high-pressure apparatus used for fluid phase equilibrium measurements is shown in Figure 1. It was built by Eurotechnica GmbH (Germany) and consists of an equilibrium cell made of stainless steel (SS-316) and equipped with a sapphire window for observing the content of the cell during measurements. Internal volume of the cell ranges from 40 to 70 mL, adjustable through a manual screw piston. The cell includes a pressure transducer and an immersed thermocouple. Both of them calibrated and connected to a Data Acquisition System (DAS). The equipment was placed inside an oven that allowed temperature control of the system. Mixing of the components of the system was achieved by 6 continuously taking the vapour phase and passing it back into the equilibrium cell through the liquid phase by means of a gear pump (Micropump IDEX). A 750 μL loop that could be isolated by means of a 6 way valve (VICI) was placed in the recirculation path for sampling the vapour phase with minimal equilibrium disturbance. Besides, a micro-metering valve was connected to the bottom of the equilibrium cell through a 1/16” capillary for sampling the heavy phase. Pressure drop occurring when sampling the liquid phase was compensated by reducing the volume of the cell through the manual screw piston. Maximum specifications of the apparatus are p = 32 MPa and T = 393 K. A typical experiment with the high-pressure variable-volume view cell began with the preheating of the system up to the desired temperature. When the temperature was achieved, the equilibrium cell was gently purged with low pressure CO2 to sweep the residual air inside the cell. Immediately afterwards, known volumes of fish oil and ethanol were introduced into the cell by means of a binary HPLC pump (Agilent 1200 Series). A certain amount of CO2 was then charged into the cell by using a high-pressure syringe pump (ISCO 260D). The exact amounts of fish oil and ethanol were calculated using their respective densities at room temperature, whereas the mass of CO2 charged into the cell was measured by a Coriolis mass flow meter (Rheonik RHE015). Once the cell was charged and the desired pressure was adjusted by actuating the manual screw piston, the gear pump was connected and recirculation of the vapour phase was performed for at least 2 h to facilitate the mixing of the components and its distribution in the different phases of the system. The system was then let to stand for another 2 h at constant temperature and pressure. Phase separation was visually verified through the sapphire window and samples from the vapour and liquid phases were taken by the 6-way and the micro-metering valve, respectively. Pressure variations up to ±0.1 MPa were observed during sampling, while temperature change was not detected. Overall standard 7 uncertainties in the equilibrium measurements were u(p) = ± 0.15 MPa, u(T) = ± 0.1 K, and u(wi) = ± 0.005; being wi the mass fraction of component i. Samples were decompressed to atmospheric pressure and released CO2 was measured by means of a thermal mass flow meter (Bronkhorst F-110C). Ethanol and fish oil were separated from CO2 and collected in an ice-cooled glass trap. The amount of each component was determined by weighing the vials in a precision analytical balance (accurate to ± 0.0001 g) before and after evaporation of ethanol at T = 373.15 K. 3. RESULTS AND DISCUSSION 3.1. Experimental data In order to check the reliability of the apparatus and experimental procedure, preliminary measurements of the binary system CO2 + ethanol were carried out. Experimental highpressure phase equilibria data of the CO2 + ethanol binary system are summarised in Table 2. The results obtained were compared with data taken from literature [9,12], finding a good agreement (Figure 2). Experimental high-pressure phase equilibrium data of the pseudo-ternary system CO2 + ethanol + fish oil are listed in Tables 3-5. Due to the large differences in molecular weight of the components of the system, compositions are expressed in terms of mass fraction instead of mole fraction. Experimental results at T = 323.15 K showed two different 2-phase regions for the two pressures investigated (10 MPa and 30 MPa). In one of them, two liquid phases could be distinguished, being the light and heavy phases rich with ethanol and oil, respectively 8 (L1+L2). In the other 2-phase region, a light vapour phase rich with CO2 and a heavy liquid phase rich with fish oil (V+L2) were observed. Homogeneous monophasic mixtures were visually and analytically verified, samples taken from the top and bottom of the equilibrium cell were similar with differences smaller than the experimental uncertainty. In the case of the phase equilibrium at T = 343.15 K and p = 10 MPa, three 2-phase regions (L1+L2, V+L1, and V+L2) and a 3-phase region (V+L1+L2) were observed. The appearance of the V+L1 region is consistent with the published phase equilibrium data of the binary system CO2 + ethanol at T = 343.15 K [9,12]. Results obtained for the binary ethanol + fish oil tie-lines at T = 323.15 K, and pressures of 10 MPa and 30 MPa (Figure 3) are similar to those obtained by Bucio et al. in previous works at the same temperature and atmospheric pressure [26], indicating that, in the range investigated, pressure does not significantly affect phase equilibrium of this binary mixture. The composition of the two liquid phases in the L1+L2 region became more similar as more CO2 is dissolved. The same was true for the V+L2 region, where the compositions of the vapour and liquid phases in equilibrium tend to merge with increasing amounts of dissolved ethanol. A similar trend has been found for other pseudo-ternary mixtures of CO2 + ethanol + natural lipids, such as castor oil [19] and sunflower oil [21]. The homogeneous monophasic region at p = 30 MPa appears to be slightly larger, although no strong effect of pressure on the phase behaviour of the mixture at T = 323.15 K can be observed. From the phase diagram at T = 343.15 K and p = 10 MPa (Figure 4), it can be observed that, starting from the binary sides of the phase diagram, increasing amounts of the third component made the tie-lines of the 2-phase regions approach the sides of the 3-phase region triangle. Inside this 3-phase region, theoretical mixtures split in a V+L1+L2 system, each phase with a composition defined by the vertices of the triangle. 9 Comparing the phase behaviour at T = 343.15 K, p = 10 MPa (Figure 4) with the 323.15 K isotherm at the same pressure (Figure 3a), it is noticeable that the former presents a lower amount of CO2 dissolved in the liquid phase (V+L2 region), probably because of the temperature-driven increase in the CO2 vapour pressure. On the contrary, it can be observed that raising temperature from 323.15 to 343.15 K increases the solubility of fish oil in ethanol (L1+L2 region) from near 0.075 mass fraction up to 0.35 mass fraction, yet adding CO2 has a slight de-entraining effect at 343.15 K since the more CO2 is dissolved, the wider the tie-lines become. 3.2. Data correlation In this work, fish oil has been treated as a pseudo-component, assuming that the different triacylglycerols present in the fish oil behave in a similar way, which has been previously verified [26]. Experimental phase equilibrium data of the pseudo-ternary mixture were correlated with the Peng-Robinson equation of state (PR EoS) in combination with the conventional van der Waals mixing rules with two adjustable parameters (vdW2). The PR EoS, was used as originally defined by Peng and Robinson [18]: b)b(vb)v(v a(T) bv T p−++ − − =R (1) where the parameter a is related to the intermolecular attractive forces and b to the size of the molecules. For a pure fluid, constant b takes the value: C C p RT 0.07780=b (2) 16 ACKNOWLEDGEMENTS The authors thank the Junta de Castilla y León for financial support, BU055U16, co-financed by the European Regional Development Fund (ERDF). RM acknowledges MINECO [BES2013-063937] and AGS University of Burgos for their pre-doctoral contracts. REFERENCES [1] N. Rubio-Rodríguez, S. Beltrán, I. Jaime, S.M. de Diego, M.T. Sanz, J.R. Carballido, Production of omega-3 polyunsaturated fatty acid concentrates: A review, Innov. Food Sci. Emerg. Technol. 11 (2010) 1–12. [2] R. Melgosa, M.T. Sanz, Á.G. Solaesa, E. de Paz, S. Beltrán, D.L. Lamas, Supercritical carbon dioxide as solvent in the lipase-catalyzed ethanolysis of fish oil: Kinetic study, J. CO2 Util. 17 (2017) 170–179. [3] G.R. Akien, M. Poliakoff, A critical look at reactions in class I and II gas-expanded liquids using CO2 and other gases, Green Chem. 11 (2009) 1083–1100. [4] N. Rubio-Rodríguez, S.M. De Diego, S. Beltrán, I. Jaime, M.T. Sanz, J. Rovira, Supercritical fluid extraction of fish oil from fish by-products: A comparison with other extraction methods, J. Food Eng. 109 (2012) 238–248. [5] B.M. Bhosle, R. Subramanian, New approaches in deacidification of edible oils - A review, J. Food Eng. 69 (2005) 481–494. [6] O.J. Catchpole, J.B. Grey, K.A. Noermark, Fractionation of fish oils using supercritical CO2 and CO2 + ethanol mixtures, J. Supercrit. Fluids. 19 (2000) 25–37. 17 [7] A.P.A. Corrêa, C.A. Peixoto, L.A.G. Gonçalves, F.A. Cabral, Fractionation of fish oil with supercritical carbon dioxide, J. Food Eng. 88 (2008) 381–387. [8] Á. Martín, S. Varona, A. Navarrete, M.J. Cocero, Encapsulation and co-precipitation processes with supercritical fluids: Applications with essential oils, Open Chem. Eng. J. 5 (2010) 31–41. [9] J.S. Lim, Y.Y. Lee, H.S. Chun, Phase equilibria for carbon dioxide–ethanol–water system at elevated pressures, J. Supercrit. Fluids. 7 (1994) 219–230. [10] C.J. Chang, C.-Y. Day, C.-M. Ko, K.-L. Chiu, Densities and p-x-y diagrams for carbon dioxide dissolution in methanol, ethanol, and acetone mixtures, Fluid Phase Equilib. 131 (1997) 243–258. [11] H. Chen, H. Chang, E.T.S. Huang, T. Huang, A new phase behavior apparatus for supercritical fluid extraction study, Ind. Eng. Chem. Res. 39 (2000) 4849–4852. [12] S.N. Joung, C.W. Yoo, H.Y. Shin, S.Y. Kim, K.P. Yoo, C.S. Lee, et al., Measurements and correlation of high-pressure VLE of binary CO2 – alcohol systems (methanol, ethanol, 2-methoxyethanol and 2-ethoxyethanol), Fluid Phase Equilib. 185 (2001) 219–230. [13] Ž. Knez, M. Škerget, L. Ilič, C. Lütge, Vapor–liquid equilibrium of binary CO2– organic solvent systems (ethanol, tetrahydrofuran, ortho-xylene, meta-xylene, para-xylene), J. Supercrit. Fluids. 43 (2008) 383–389. [14] C. Secuianu, V. Feroiu, D. Gean, D. Geană, Phase behavior for carbon dioxide + ethanol system: Experimental measurements and modeling with a cubic equation of state, J. Supercrit. Fluids. 47 (2008) 109–116. 18 [15] Ö. Güçlü-Üstündağ, F. Temelli, Solubility behavior of ternary systems of lipids in supercritical carbon dioxide, J. Supercrit. Fluids. 38 (2006) 275–288. [16] Ö. Güçlü-Üstündağ, F. Temelli, Solubility behavior of ternary systems of lipids, cosolvents and supercritical carbon dioxide and processing aspects, J. Supercrit. Fluids. 36 (2005) 1–15. [17] D. Geana, R. Steiner, Calculation of phase equilibrium in supercritical extraction of C54 triglyceride (rapeseed oil), J. Supercrit. Fluids. 8 (1995) 107–118. [18] D. Peng, D.B. Robinson, A new two-constant equation of state, Ind. Eng. Chem. Fundam. 15 (1976) 59–64. [19] P.M. Ndiaye, E. Franceschi, D. Oliveira, C. Dariva, F.W. Tavares, J.V. Oliveira, Phase behavior of soybean oil, castor oil and their fatty acid ethyl esters in carbon dioxide at high pressures, J. Supercrit. Fluids. 37 (2006) 29–37. [20] S.H. Huang, M. Radosz, Equation of state for small, large, polydisperse, and associating molecules: Extension to fluid mixtures, Ind. Eng. Chem. Res. 30 (1991) 1994– 2005. [21] E.J. Hernández, G.D. Mabe, F.J. Señoráns, G. Reglero, T. Fornari, High-pressure phase equilibria of the pseudoternary mixture sunflower oil + ethanol + carbon dioxide, J. Chem. Eng. Data. 53 (2008) 2632–2636. [22] S. Skjold-Jørgensen, Group contribution equation of state (GC-EOS): A predictive method for phase equilibrium computations over wide ranges of temperature and pressures up to 30 MPa, Ind. Eng. Chem. Res. 27 (1988) 110–118. 19 [23] I. Dalmolin, A.A. Rigo, M.L. Corazza, P.M. Ndiaye, M.A.A. Meireles, E.A.C. Batista, J.V. Oliveira, Phase behaviour and thermodynamic modelling for the system (grape seed oil + carbon dioxide + ethanol) at high pressures, J. Chem. Thermodyn. 68 (2014) 71–74. [24] R. Dohrn, G. Brunner, High-pressure fluid-phase equilibria: Experimental methods and systems investigated (1988-1993), Fluid Phase Equilib. 106 (1995) 213–282. [25] Official Methods and Recommended Practices of the American Oil Chemists’ Society, American Oil Chemists’ Society, Champaign, IL, 1995. [26] S.L. Bucio, A.G. Solaesa, M.T. Sanz, S. Beltrán, R. Melgosa, Liquid-liquid equilibrium for ethanolysis systems of fish oil, J. Chem. Eng. Data. 58 (2013) 3118–3124. [27] Á.G. Solaesa, M.T. Sanz, M. Falkeborg, S. Beltrán, Z. Guo, Production and concentration of monoacylglycerols rich in omega-3 polyunsaturated fatty acids by enzymatic glycerolysis and molecular distillation, Food Chem. 190 (2016) 960–967. [28] Official Methods of Analysis of the Association of Official Analytical Chemists. AOAC Official Method 965.33. Peroxide Value of Oils and Fats, AOAC International, Washington, DC, 2000. [29] Y. Medina-Gonzalez, T. Tassaing, S. Camy, J.-S. Condoret, Phase equilibrium of the CO2/glycerol system: Experimental data by in situ FT-IR spectroscopy and thermodynamic modeling, J. Supercrit. Fluids. 73 (2013) 97–107. [30] B.E. Poling, M. Prausnitz, J.P. O’Connell, The Properties of Gases and Liquids, fifth ed., McGraw-Hill, New York, 2001. 20 [31] L. Constantinou, R. Gani, J.P. O’Connell, Estimation of the acentric factor and the liquid molar volume at 298 K using a new group contribution method, Fluid Phase Equilib. 103 (1995) 11–22. [32] L.P. Cunico, A.S. Hukkerikar, R. Ceriani, B. Sarup, R. Gani, Molecular structurebased methods of property prediction in application to lipids: A review and refinement, Fluid Phase Equilib. 357 (2013) 2–18. [33] D.S.H. Wong, S.I. Sandler, A theoretically correct mixing rule for cubic equations of state, AIChE J. 38 (1992) 671–680. [34] J.O. Valderrama, C.A. Faúndez, Thermodynamic consistency test of high pressure gas-liquid equilibrium data including both phases, Thermochim. Acta. 499 (2010) 85–90. [35] H.C. Van Ness, Exact forms of the unrestricted Gibss-Duhem equation, Chem. Eng. Sci. 10 (1959) 225–228. [36] J. Wisniak, J. Ortega, L. Fernández, A fresh look at the thermodynamic consistency of vapour-liquid equilibria data, J. Chem. Thermodyn. 105 (2017) 385–395. 21 Table 1. Fatty acid composition of the fish oil [26]. Fatty acid % wt. myristic C14:0 3.8 palmitic C16:0 21.0 palmitoleic C16:1n-7 6.1 stearic C18:0 6.0 oleic C18:1n-9 18.4 vaccenic C18:1n-7 3.0 linoleic cis (LA) C18:2n-6 2.4 α-linolenic (ALA) C18:3n-3 0.7 stearidonic C18:4n-3 0.9 eicosenoic C20:1n-9 2.1 eicosatrienoic C20:3n-3 2.0 eicosapentaenoic (EPA) C20:5n-3 6.9 docosapentaenoic (DPA) C22:5n-3 1.8 docosahexaenoic (DHA) C22:6n-3 24.9 Standard uncertainties are u(percentage) = ± 0.5 22 Table 2. Vapour + Liquid phase compositions (mole fraction) of the binary system CO2 (1) + ethanol (2). p / MPa x 1 y 1 p / MPa x 1 y 1 T = 323.15 K T = 343.15 K 4.67 0.2647 0.9840 7.01 0.2930 0.9679 5.12 0.2771 0.9859 8.05 0.3785 0.9656 5.34 0.3058 0.9819 9.04 0.4260 0.9614 5.78 0.3128 0.9837 9.98 0.4968 0.9484 7.07 0.4128 0.9814 10.98 0.6130 0.9432 7.69 0.4795 0.9739 11.50 0.6910 0.9060 7.96 0.5069 0.9764 8.16 0.5455 0.9754 8.44 0.6348 0.9740 8.46 0.6279 0.9671 8.51 0.6535 0.9679 8.69 0.6875 0.9630 Standard uncertainties are u(p) = ± 0.15 MPa, u(T) = ± 0.1 K, and u(w i ) = ± 0.005 23 Table 3. Liquid + Liquid (L1+L2) and Vapour + Liquid (V+L2) phase compositions (weight fraction) of the pseudo-ternary system CO2 (1) + ethanol (2) + fish oil (3) at T = 323.15 K and p = 10 MPa. heavy phase light phase type of phase eq. w1 w2 w3 w1 w2 w3 0.0000 0.2652 0.7348 0.0000 0.9256 0.0744 L1+L2 0.0423 0.2836 0.6741 0.0420 0.8692 0.0888 L1+L2 0.0761 0.2966 0.6273 0.0780 0.7957 0.1263 L1+L2 0.1012 0.3087 0.5901 0.1016 0.7638 0.1346 L1+L2 0.1696 0.3898 0.4406 0.1710 0.6176 0.2114 L1+L2 0.2247 0.0000 0.7754 0.9868 0.0000 0.0132 V+L2 0.2440 0.0272 0.7288 0.9580 0.0280 0.0140 V+L2 0.2776 0.0747 0.6478 0.9003 0.0751 0.0246 V+L2 0.3330 0.1181 0.5490 0.7924 0.1657 0.0419 V+L2 0.3622 0.1541 0.4837 0.7332 0.2193 0.0476 V+L2 0.3730 0.1710 0.4560 0.6949 0.2525 0.0526 V+L2 0.3804 0.1841 0.4355 0.6207 0.3010 0.0783 V+L2 0.2026 0.2224 0.5751 0.2024 0.2222 0.5754 homogeneous 0.2745 0.4691 0.2564 0.2747 0.4689 0.2564 homogeneous 0.3039 0.2838 0.4023 0.3030 0.2844 0.4026 homogeneous 0.4590 0.3640 0.1770 0.4593 0.3638 0.1769 homogeneous Standard uncertainties are u(p) = ± 0.15 MPa, u(T) = ± 0.1 K, and u(wi) = ± 0.005 24 Table 4. Liquid + Liquid (L1+L2) and Vapour + Liquid (V+L2) phase compositions (weight fraction) of the pseudo-ternary system CO2 (1) + ethanol (2) + fish oil (3) at T = 323.15 K and p = 30 MPa. heavy phase light phase type of phase eq. w1 w2 w3 w1 w2 w3 0.0000 0.2219 0.7781 0.0000 0.9066 0.0934 L1+L2 0.0649 0.3010 0.6341 0.0771 0.8185 0.1044 L1+L2 0.1550 0.4288 0.4162 0.1520 0.6654 0.1826 L1+L2 0.2559 0.0000 0.7441 0.9832 0.0000 0.0168 V+L2 0.2843 0.0674 0.6484 0.9089 0.0639 0.0272 V+L2 0.3073 0.1174 0.5753 0.8588 0.1131 0.0281 V+L2 0.3404 0.1401 0.5194 0.8252 0.1417 0.0331 V+L2 0.3844 0.1666 0.4490 0.7079 0.2367 0.0555 V+L2 0.4922 0.2338 0.2740 0.6089 0.3036 0.0875 V+L2 0.2926 0.2224 0.4851 0.2928 0.2219 0.4853 homogeneous 0.2919 0.4499 0.2582 0.2902 0.4488 0.2610 homogeneous 0.4528 0.416 0.1312 0.4481 0.4163 0.1356 homogeneous 0.5219 0.3261 0.1520 0.5198 0.328 0.1522 homogeneous Standard uncertainties are u(p) = ± 0.15 MPa, u(T) = ± 0.1 K, and u(wi) = ± 0.005 25 Table 5. Liquid + Liquid (L1+L2) and Vapour + Liquid (V+L1, V+L2) phase compositions (weight fraction) of the pseudo-ternary system CO2 (1) + ethanol (2) + fish oil (3) at T = 343.15 K and p = 10 MPa. heavy phase light phase type of phase eq. w1 w2 w3 w1 w2 w3 0.0000 0.2829 0.7171 0.0000 0.6497 0.3503 L1+L2 0.0363 0.2327 0.7310 0.0728 0.6418 0.2854 L1+L2 0.0618 0.2102 0.7280 0.1725 0.6291 0.1984 L1+L2 0.0890 0.1851 0.7259 0.3132 0.5677 0.1191 L1+L2 0.1157 0.1607 0.7236 0.4183 0.5088 0.0729 L1+L2 0.1624 0.1162 0.7214 0.4981 0.4199 0.0820 L1+L2 0.4854 0.5146 0.0000 0.9461 0.0539 0.0000 V+L1 0.4850 0.5014 0.0136 0.9239 0.0589 0.0172 V+L1 0.4870 0.4752 0.0378 0.9002 0.0680 0.0318 V+L1 0.1820 0.0000 0.8180 0.9796 0.0000 0.0204 V+L2 0.1843 0.0396 0.7761 0.9455 0.0325 0.0220 V+L2 0.1794 0.0717 0.7489 0.9143 0.0542 0.0315 V+L2 0.1814 0.0957 0.7229 0.9055 0.0607 0.0338 V+L2 Standard uncertainties are u(p) = ± 0.15 MPa, u(T) = ± 0.1 K, and u(wi) = ± 0.005