Substrates emulsification process to improve lipase-catalyzed sardine oil glycerolysis in different systems. Evaluation of lipid oxidation of the reaction products

Accepted Manuscript Substrates emulsification process to improve lipase-catalyzed sardine oil glycerolysis in different systems. Evaluation of lipid oxidation of the reaction products Ángela García Solaesa, María Teresa Sanz, Rodrigo Melgosa, Sagrario Beltrán PII: S0963-9969(17)30370-8 DOI: doi: 10.1016/j.foodres.2017.07.048 Reference: FRIN 6845 To appear in: Food Research International Received date: 18 May 2017 Revised date: 18 July 2017 Accepted date: 19 July 2017 Please cite this article as: Ángela García Solaesa, María Teresa Sanz, Rodrigo Melgosa, Sagrario Beltrán , Substrates emulsification process to improve lipase-catalyzed sardine oil glycerolysis in different systems. Evaluation of lipid oxidation of the reaction products, Food Research International (2017), doi: 10.1016/j.foodres.2017.07.048 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. ACCEPTED MANUSCRIPT 1 Substrates emulsification process to improve lipase-catalyzed sardine oil glycerolysis in different systems. Evaluation of lipid oxidation of the reaction products Ángela García Solaesa, María Teresa Sanz  , Rodrigo Melgosa, Sagrario Beltrán Department of Biotechnology and Food Science (Chemical Engineering Section), University of Burgos, 09001 Burgos. Spain Abstract Monoand diacylglycerols rich in omega-3 have a great interest due to their good bioavailability and oxidation stability compared with other kind of omega-3 concentrates. The main drawback in monoand diacylglycerols production by glycerolysis is the immiscibility of the substrates, oil and glycerol. To improve mass transfer rates, avoiding the use of organic solvents, emulsification of both reactants as reverse micelles (glycerol-in-oil) was carried out previous to lipase-catalyzed sardine oil glycerolysis. Substrate emulsification yielded higher reaction rates compared to kinetics with no previous emulsification, but still lower than in organic solvents. To avoid the use of organic solvent, SC-CO2 was used as reaction medium but no kinetic advantages were demonstrated in the pressure range from 15 to 25 MPa. By increasing temperature, from 40 to 90 ºC, reaction rates increased both in a solvent-free system and in SC-CO2 medium. It was also found that an increase in temperature does not lead to an increase in the final oxidation status of the reaction products. This behavior was due to the sorption capacity of the Lipozyme 435 support, giving lower oxidation status at the highest temperature, 80-90 ºC. Keywords: fish oil, glycerolysis, microemulsion, SC-CO2, peroxides adsorption. Chemical compounds studied in this article: Glycerol (PubChem CID: 753); Eicosapentaenoic acid (PubChem CID: 446284); Docosahexaenoic acid (PubChem CID: 445580); Aerosol OT (PubChem CID: 23673837); Tween 80 (PubChem CID: 5281955).  Corresponding author. Tel.: +34 947 258810. Fax: ++34947258831. E-mail address [email protected] ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 2 1. Introduction The importance of omega-3 polyunsaturated fatty acids (n-3 PUFA), especially eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA), in human nutrition and disease prevention is fully recognized scientifically (Kris-Etherton, Harris, & Appel, 2002; Riediger, Othman, Suh, & Moghadasian, 2009). n-3 PUFA supplements are available in different chemical forms. Among the different types of lipid derivatives containing n-3 PUFA concentrates, monoacylglycerols (MAG) and diacylglycerols (DAG) have good bioavailability and oxidation stability (Hernandez, 2014; Lawson & Hughes, 1988). Additionally, it must be also considered that dietary TAG are hydrolyzed in the small intestine to sn-2-MAG being the most favorable structure for n3 PUFA to be adsorbed by intestinal mucosa (Bandarra et al. 2012). In addition, MAG or its mixtures with DAG account for 75% of the worldwide emulsifier production (Zhong et al., 2009). The well-known drawbacks of the conventional chemical glycerolysis technique (energy intensive, low yields (30–40%), oxidized products) have prompted a growing interest in the development of alternative processes for the production of MAG and DAG rich in n-3 PUFA. Enzyme-catalyzed reaction is an attractive alternative since the reaction can be carried out under mild conditions (Bornscheuer, 1995; Feltes, de Oliveira, Block, & Ninow, 2013). To overcome the problem of the immiscibility of glycerol and oil, different approaches have been used in the literature to improve the contact between the reactants and hence reduce mass transfer limitation. Lipase-catalyzed glycerolysis has been carried out in different reaction media such as organic solvents (Damstrup et al., 2006), compressed fluids (Moquin, Temelli, King, & Palcic, 2005) and ionic liquids (Guo & Xu, 2006), in order to improve the mass transfer. The cost, toxicity and energy required for solvent removal from the product mixture, are important aspects to be considered when dealing with conventional solvent systems (Prat, Hayler, & Wells, 2014). Recently, the uses of different surfactants to increase the interfacial area, and ultrasound irradiation have been also proposed to reduce mass transfer limitation (Fiametti et al., 2012; Valério, Rovani, Treichel, De Oliveira, & Oliveira, 2010). Biocatalytic processing in microemulsion system has received attention in order to increase contact between substrates. The formation of a microemulsion of the reactants (glycerol-in-oil) as reverse micelles can help to improve mass transfer rates. Furthermore, lipases demonstrate high interfacial activity in micelle systems because the formation of the active site during the reaction ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 3 occurs at the interface between the substrates and the enzyme. Several food grade surfactants are able to stabilize the micellar system improving system homogeneity (Carvalho & Cabral, 2000; Stamatis, Xenakis, & Kolisis, 1999). Nevertheless, it must be taken into account that some food grade surfactants have chemical functions that could be modified by lipases. For instance, the lipase Novozym 435 presented activity at particular conditions towards some surfactants as soy lecithin and Tween in glycerolysis reactions (Camino Feltes, Villeneuve, Baréa, de Oliveira, & Ninow, 2012). To avoid this problem, other synthetic surfactants, such as sodium (bis-2-ethyl-hexyl) sulfosuccinate (aerosol-OT or AOT), have been used. AOT has been reported to form micelles in a great number of nonpolar substances and several other polar solvents such as glycerol (Fiametti et al., 2009). In this case, good results have been obtained in glycerolysis systems when adding more that 7.5% of AOT (Fiametti et al., 2009). However, the high amount of this surfactant may generate problems during removal processes (Stamatis, Xenakis, & Kolisis, 1994). Another alternative to organic solvents is the use of the supercritical fluids (SCFs) as reaction medium. Supercritical carbon dioxide (SC-CO2) is probably the most used SCF due to its additional benefits (non-toxic, non-flammable, readily available at high purities and low costs, and relatively mild critical conditions) that are appealing when choosing environmental replacement for organic solvents (Matsuda, 2013; Rezaei, Temelli, & Jenab, 2007). SC-CO2 has liquid-like density but gas-like viscosity resulting in high mass transfer being a clean alternative to replace organic solvents. Enzymatic concentration of n-3 PUFA in supercritical fluids (SCFs) is an interesting option for the prevention of oxidation during processing of fish oil (Lin, Chen, & Chang, 2006; Roh, Kim, & Choi, 2015). Besides, SC-CO2 can be easily separated from the reaction products by simple depressurization and allows fractionation of the reaction products. Some previous studies of enzymatic reactions of different lipid sources in SC-CO2 have been reported in the literature. However, in case of enzymatic glycerolysis, other compressed fluids such as propane, n-butane, and acetone, have been used (Esmelindro et al., 2008; Tai & Brunner, 2011; Valério et al., 2010). Some studies of glycerolysis of vegetable oils in SC-CO2 at high temperatures can be found but with no enzymatic catalyst (Moquin et al., 2005; Temelli, King, & List, 1996). In a previous work, a detail kinetic study of glycerolysis of sardine oil using Lipozyme 435 form Candida antarctica B as biocatalyst in an optimized amount of tert-butanol was performed (Solaesa, Sanz, Beltrán, et al., 2016). Tert.butanol helped to create a ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 4 homogeneous phase and to reduce mass transfer limitations. However, organic solvents present different environmental concerns. In this work, to improve contact between substrates, avoiding the use of organic solvents, emulsification of glycerol and oil before glycerolysis reaction was considered. Glycerolysis reaction has been performed in a solvent free system at atmospheric pressure and in SC-CO2 as reaction medium with previous susbstrates emulsification. The effect of adding a surfactant, AOT or Tween 80, to stabilize the emulsion, on glycerolysis performance has been also studied. Glycerolysis has been determined at different operating temperaturesat atmosphere pressure, 0.1 MPa, and in SC-CO2 medium in the pressure range from 15 to 25 MPa. Since n-3 PUFA are highly susceptible to oxidation; the oxidative status of the final reaction products was evaluated through the peroxide and anisidine values. Reaction yields and the oxidation values of the reaction products were compared for both systems. 2. Materials and methods 2.1 Materials Refined sardine oil was provided by Industrias Afines S.L. (Spain) with 18.3% of EPA and 7% of DHA and a water content of 0.2% (Solaesa, Bucio, Sanz, Beltrán, & Rebolleda, 2014). Glycerol was purchased from Sigma Aldrich with a purity of ≥ 99.5% and a water content of 0.18%. The food grade lipase Lipozyme 435 from Candida antarctica B (immobilized on a macroporous hydrophobic acrylic resin), was donated by Novozymes A/S (Bagsvaerd, Denmark). Carbon dioxide (99.9%) was supplied by Air Liquide S.A. (Spain). Polyoxyethylene sorbitan monooleate (Tween 80) and sodium bis (2-ethylhexyl) sulfosuccinate (Aerosol AOT or AOT), used as food grade surfactants, were purchased by Sigma Aldrich. All other chemicals used in different analyses were of analytical or HPLC grade. 2.2 Emulsification process Microemulsions of the glycerolysis system of sardine oil were prepared at a fixed mole ratio of 3:1 (glycerol:oil) since this mole ratio was found as the optimum in a previous kinetic study (Solaesa, Sanz, Beltrán, et al., 2016). A high-speed blender (Miccra D9 equipped with a DS-20/PF EMR rotor–stator) at different speeds, from 16000 to 35000 rpm, was used by pulses during 3 minutes. To prepare the surfactant-free emulsion as reverse micelles, the appropriate amount of glycerol (10 g) was added drop by drop to ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 5 the suitable amount of oil (30 g) while being completely mixed at high speed. Dispersed (glycerol) and continuous (sardine oil) phases were identified by the dilution test (Mize et al., 2013). Furthermore, different concentrations (0.5, 1 and 1.5% in glycerol or oil as indicated in Table 2) of two food grade surfactants, AOT and Tween 80, were tested in order to improve the stability of the emulsion. A defined quantity of each surfactant was dissolved in oil or in glycerol, depending on its solubility. The characterization of the emulsions was performed 10 min after emulsification to avoid any creaming or coalescence effect. Particle size distribution (PSD), mean droplet diameter and polydispersity index (PDI) of samples were measured by dynamic light scattering (DLS), using a Zetasizer Nano ZS apparatus (Malvern Instruments Ltd., UK) to evaluate the best conditions to produce a stable emulsion with small (or the smallest) droplet size. 2.3 Lipase-catalyzed glycerolysis of sardine oil in different systems A comparative study of lipase-catalyzed glycerolysis in different systems was carried out. All the experiments were conducted in a batch mode keeping constant the enzyme concentration at 5 wt% (by weight of substrates) and the substrate mole ratio (3:1, glycerol to oil) according to previous work (Solaesa, Sanz, Beltrán, et al., 2016). Table 1 summarizes all glycerolysis reactions that have been done in this work. Experiments 1 - 6 have been carried out at atmospheric pressure in a solvent free system in a 100 mL jacketed batch reactor. First of all, experiments 1 and 2 were carried out to evaluate the effect of previous substrates emulsification on reaction rate. Experiments 3 and 4 were performed with emulsified substrates stabilized by adding a food grade surfactant, AOT and Tween 80 respectively, at the optimum concentration previously determine in section 2.2. Experiments 2, 5 and 6 were performed to evaluate the effect of reaction temperature, 50, 80 and 90ºC respectively. Glycerolysis reaction was carried out as follows. Once emulsion was prepared, it was charged into the reactor. Later, the lipase was added and a nitrogen stream was applied. The reactor was then closed and the stirring system by impellers was connected. A thermostatic water bath allows working at the desired temperature. The reactor was covered with foil paper to avoid the light exposure. On the other hand, experiments 7-13 have been carried out in SC-CO2 as reaction medium. They were performed in a high pressure batch stirred tank reactor made of stainless steel, having an internal volume of 100 mL (Melgosa et al., 2017). A freshly ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 6 prepared emulsion and the lipase were charged into the reactor provided with magnetic agitation and then it was closed, placed in a thermostatic water bath and connected to the pressure circuit. Subsequently, SC-CO2 was fed into the reactor by means of a high pressure pump (ISCO 260 D) up to the desired pressure, which was maintained by a digital pressure controller. Operating pressure and temperature have been varied in the range between 15-25 MPa (Exp.7-9) and 40-90ºC (Exp. 7 and 10-13). In both systems, samples were taken periodically during 8 h, filtered and stored at -18ºC up to analysis. 2.4 Analysis of the reaction products The neutral lipid profile (TAG, DAG, MAG and FFA) was analyzed by a normal phase high performance liquid chromatography (NP-HPLC). The chromatographic apparatus consisted of a HPLC system (Agilent 1200) formed by a quaternary pump and an autoinjector. The chromatographic separation of the compounds was carried out at room temperature with a Lichrospher Diol column (5 μm, 4 mm × 250 mm) and detection was performed by an evaporative light scattering detector (Agilent 1200 series) at 35ºC and 0.35 MPa. Gradient elution was achieved by mobile phases A (isooctane) and B (methyltert-butyl ether:acetic acid = 99.9:0.1, v/v). The method and calibration procedure have been previously reported (Solaesa, Sanz, Falkeborg, et al., 2016). The regioisomers of DAG and MAG could not be distinguished by the applied analytical procedure, so the total amount of MAG and DAG was reported for the kinetic experiments. The lipid profile results were expressed in glycerol free basis. 2.5 Lipid oxidation analysis The oxidation status has been determined using two assays: peroxide value (PV) and anisidine value (AV). The PV measures the concentration of hydroperoxides formed in the initial stages of lipid oxidation (primary oxidation). PV was determined following the AOAC Official Method 965.33 by an automatic titrator Methrom 905 Titrando (AOAC Official Method 965.33, 2000). The AV is an estimation of the concentration of non-volatile secondary oxidation products (mainly 2-alkenals and 2,4-dienals). The AV was measured according to AOCS official method (Cd 18–90), using a UV-Visible spectrophotometer (AOCS Official Method Cd 18-90, 2017). PV and AV allow calculating total oxidation (TOTOX) by the formula: TOTOX = 2PV + AV [1] ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 7 PV and AV have been determined for the supplied refined sardine oil and the final reaction mixtures obtained after 8 h at the different temperatures The lipid phase was separated for analysis from the lipase and the remained glycerol by centrifugation at 5000 rpm and 35ºC during 10 minutes. The upper phase, free of glycerol, formed by the lipid fraction (TAG, DAG, MAG and FFA) was collected under N2 atmosphere and stored at -18ºC up to analysis. 2.6 Statistical analysis All analyses were conducted using software Statgraphics X64. The results are presented as a mean  standard deviation of at least three replicates. The significance of the differences was determined based on an analysis of the variance with the Tukey’s honestly significant difference (HSF) method at p-value ≤ 0.05. 3. Results and discussion 3.1 Optimization of the emulsification process and characterization of the emulsion 3.1.1 Surfactant-free emulsions The effect of emulsification speed on emulsion stability without the addition of surfactants has been evaluated by measuring the polydispersity index (PDI) and the droplet diameter of the emulsion obtained in the range from 16000 to 35000 rpm. At any of the emulsification speeds essayed, droplet diameter was lower than 2 µm but 29000 rpm were needed to obtain a PDI below 1. The lowest polydispersity index was obtained at the highest speed assayed in this work, 35000 rpm; however, foaming was observed. Therefore 29000 rpm was selected for further substrate emulsifications. At this speed the mean droplet diameter of the emulsion was 301 ± 34 nm and the PDI around 0.4. The surfactant-free emulsion presented a PDI lower than 1 only up to 20 minutes, although at longer times, still a translucent and homogeneous system was visually observed. In any case, the emulsion was prepared and immediately used as reaction media. 3.1.2 Surfactant stabilized emulsions The use of a surfactant was also tested in this work to improve the emulsion stability and reaction rates. Two food grade surfactants, AOT and Tween 80, with hydrophilic ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 8 lipophilic balance (HLB) values of 10 and 15 respectively, were used at different amounts (0.5, 1 and 1.5%). PDI of the emulsions prepared adding these surfactants were measured at specific times to evaluate their stability (Table 2). Although a surfactant was added to stabilize the emulsions, PDI in the different emulsions increased with time in all cases (Table 2). The higher stability was observed when 0.5% of Tween 80 and 1.5% of AOT were previously dissolved in glycerol. In these cases the emulsion was found to be stable for at least 1 h. PSD was evaluated for emulsions with the highest stability formed by adding 0.5 % of Tween 80 and 1.5 % of AOT in glycerol and compared with those obtained in surfactant-freeemulsion. Smaller micelles were obtained when a surfactant was added to the system with medium particle sizes values of 67 ± 5 nm, 94 ± 4 nm and 301 ± 34 nm for AOT 1.5 % and 0.5 % of Tween 80 dissolved in glycerol and surfactant-free emulsion, respectively. 3.2 Glycerolysis reaction of sardine oil by Lipozyme 435 3.2.1 Effect of substrates emulsification on the reaction rate Fig. 1 compares the kinetics of the glycerolysis reaction in a solvent free medium at atmospheric pressure with and without previous emulsification of the substrates (Exp 1 and 2 respectively). As it can be observed, when no previous emulsification of the reactants was carried out, mass transfer limitations lead to lower initial reaction rate. These limitations are reflected in the values of the initial slope of TAG composition as function of time being 0.15  0.01 (mol TAG %·min-1) without substrates emulsification and 0.279  0.008 (mol TAG %·min-1) for substrate emulsification. For a reverse micelle system, higher interfacial area is provided, which favors lipasecatalyzed reactions. At longer reaction times, reaction rates become similar due to the MAG and DAG formation as emulsifiers. The low HLB values of MAG and DAG mean that they tend to stabilize reverse micelles systems (O’Brien, 2004). Similar results were observed by Awadallak et al. (Awadallak, Voll, Ribas, Cardozo, & Edson, 2013) in the enzymatic palm oil hydrolysis under ultrasound irradiation to produce DAG. They also performed a control reaction (without ultrasound influence) to compare the degree of hydrolysis in both systems, being around 20% after 12 h in the control reaction and almost 40% when ultrasound was used before the reaction. But at longer reaction times (24 h) the degree of hydrolysis becomes similar. Therefore, they also ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 15 Oil Chemists’ Society. http://doi.org/10.1007/s11746-006-1171-5 Esmelindro, Â. F. A., Fiametti, K. G., Ceni, G., Corazza, M. L., Treichel, H., de Oliveira, D., & Oliveira, J. V. (2008). Lipase-catalyzed production of monoglycerides in compressed propane and AOT surfactant. Journal of Supercritical Fluids. http://doi.org/10.1016/j.supflu.2008.05.004 European Pharmacopoeia 5.0 (2005). Omega-3-acid Triglycerides. Feltes, M. M. C., de Oliveira, D., Block, J. M., & Ninow, J. L. (2013). The Production, Benefits, and Applications of Monoacylglycerols and Diacylglycerols of Nutritional Interest. Food and Bioprocess Technology. http://doi.org/10.1007/s11947-012-0836-3 Fiametti, K. G., Ustra, M. K., De Oliveira, D., Corazza, M. L., Furigo, A., & Vladimir Oliveira, J. (2012). Kinetics of ultrasound-assisted lipase-catalyzed glycerolysis of olive oil in solvent-free system. Ultrasonics Sonochemistry, 19, 440–451. http://doi.org/10.1016/j.ultsonch.2011.09.005 Guo, Z., & Xu, X. (2006). Lipase-catalyzed glycerolysis of fats and oils in ionic liquids: a further study on the reaction system. Green Chemistry, 8, 54–62. http://doi.org/10.1039/b511117j Hernandez, E. M. (2014). Issues in fortification and analysis of omega-3 fatty acids in foods. Lipid Technology, 26(5), 103–105. http://doi.org/10.1002/lite.201400004 Kris-Etherton, P. M., Harris, W. S., & Appel, L. J. (2002). Fish consumption, fish oil, omega-3 fatty acids, and cardiovascular disease. Circulation, 19, 2747–2757. http://doi.org/10.1161/01.CIR.0000038493.65177.94 Lawson, L. D., & Hughes, B. G. (1988). Human absorption of fish oil fatty acids as triacylglycerols, free acids, or ethyl esters. Biochemical and Biophysical Research Communications, 152(1), 328–335. Lin, T. J., Chen, S. W., & Chang, A. C. (2006). Enrichment of n-3 PUFA contents on triglycerides of fish oil by lipase-catalyzed trans-esterification under supercritical conditions. Biochemical Engineering Journal, 29, 27–34. http://doi.org/10.1016/j.bej.2005.02.035 Matsuda, T. (2013). Recent progress in biocatalysis using supercritical carbon dioxide. Journal of Bioscience and Bioengineering, 115(3), 233–241. http://doi.org/10.1016/j.jbiosc.2012.10.002 ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 16 Medina-Gonzalez, Y., Tassaing, T., Camy, S., & Condoret, J.-S. (2013). Phase equilibrium of the CO2/glycerol system: Experimental data by in situ FT-IR spectroscopy and thermodynamic modeling. The Journal of Supercritical Fluids, 73, 97–107. http://doi.org/10.1016/j.supflu.2012.11.012> Melgosa, R., Sanz, M. T., Solaesa, Á. G., Paz, E. De, Beltrán, S., & Lamas, D. L. (2017). Supercritical carbon dioxide as solvent in the lipase-catalyzed ethanolysis of fi sh oil : Kinetic study. Journal of CO2 Utilization, 17, 170–179. http://doi.org/10.1016/j.jcou.2016.11.011 Mize, H. E., Lucio, A. J., Fhaner, C. J., Pratama, F. S., Robbins, L. A., & Karpovich, D. S. (2013). Emulsions of Crude Glycerin from Biodiesel Processing with Fuel Oil for Industrial Heating. Journal of Agricultural and Food Chemistry, 61, 1319–1327. http://doi.org/dx.doi.org/10.1021/jf304883t Moquin, P. H. L., Temelli, F., King, J. W., & Palcic, M. M. (2005). Kinetic modeling of the glycerolysis reaction for soybean oils in supercritical carbon dioxide media. JAOCS, Journal of the American Oil Chemists’ Society. http://doi.org/10.1007/s11746-005-1118-x Nunes, A. V. M., Carrera, G. V. S. M., Najdanovic-Visak, V., & Nunes Da Ponte, M. (2013). Solubility of CO2 in glycerol at high pressures. Fluid Phase Equilibria, 358, 105–107. http://doi.org/10.1016/j.fluid.2013.07.051 O’Brien, R. D. (2004). Fats and Oils. Formulating and Processing for Applications (Second Ed.). CRC Press LLC. Prat, D., Hayler, J., & Wells, A. (2014). A survey of solvent selection guides. Green Chemistry, 16. http://doi.org/10.1039/c4gc01149j Rezaei, K., Temelli, F., & Jenab, E. (2007). Effects of pressure and temperature on enzymatic reactions in supercritical fluids. Biotechnology Advances, 25, 272–280. http://doi.org/10.1016/j.biotechadv.2006.12.002 Riediger, N. D., Othman, R. A., Suh, M., & Moghadasian, M. H. (2009). Acids in Health and Disease. Journal of the American Dietetic Association, 109, 668–679. http://doi.org/10.1016/j.jada.2008.12.022 Roh, M. K., Kim, Y. D., & Choi, J. S. (2015). Transesterification of fish oil by lipase immobilized in supercritical carbon dioxide. Fisheries Science, 81, 1113–1125. http://doi.org/10.1007/s12562-015-0926-z ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 17 Solaesa, Á. G., Bucio, S. L., Sanz, M. T., Beltrán, S., & Rebolleda, S. (2014). Characterization of Triacylglycerol Composition of Fish Oils by Using Chromatographic Techniques. Journal of Oleo Science, 63(5), 449–460. http://doi.org/10.5650/jos.ess13202 Solaesa, Á. G., Sanz, M. T., Beltrán, S., & Melgosa, R. (2016). Kinetic study and kinetic parameters of lipase-catalyzed glycerolysis of sardine oil in a homogeneous medium. Chinese Journal of Catalysis, 37, 596–606. http://doi.org/10.1016/S1872-2067(15)610403 Solaesa, Á. G., Sanz, M. T., Falkeborg, M., Beltrán, S., & Guo, Z. (2016). Production and concentration of monoacylglycerols rich in omega-3 polyunsaturated fatty acids by enzymatic glycerolysis and molecular distillation. Food Chemistry, 190, 960–967. http://doi.org/10.1016/j.foodchem.2015.06.061 Stamatis, H., Xenakis, A., & Kolisis, F. N. (1994). Studies on Enzyme Reuse and Product in Microemulsions ". Annals of the New York Academy of Sciences, 1, 237–241. Stamatis, H., Xenakis, A., & Kolisis, F. N. (1999). Bioorganic reactions in microemulsions : the case of lipases. Biotechnology Advances, 17, 293–318. http://doi.org/S07349750(99)00007-5 Tai, H. P., & Brunner, G. (2011). Mono-and di-acylglycerol synthesis in CO 2 -expanded acetone. The Journal of Supercritical Fluids, 59, 87–91. http://doi.org/10.1016/j.supflu.2011.07.021 Tao, M., Li, Q., Qu, J., & Zhang, M. (2013). Enzymatic Synthesis of Dipalmitin in Supercritical Carbon Dioxide and Mechanism Study. Industrial and Engineering Chemistry Research, 52, 13528–13535. http://doi.org/10.1021/ie4015364 Temelli, F., King, J. W., & List, G. R. (1996). Conversion of Oils to Monoglycerides by Glycerolysis in Supercritical Carbon Dioxide Media. JAOCS, 73(6), 699–706. Valério, A., Rovani, S., Treichel, H., De Oliveira, D., & Oliveira, J. V. (2010). Optimization of mono and diacylglycerols production from enzymatic glycerolysis in solvent-free systems. Bioprocess and Biosystems Engineering, 33, 805–812. http://doi.org/10.1007/s00449-0090402-1 Zhong, N., Li, L., Xu, X., Cheong, L., Li, B., Hu, S., & Zhao, X. (2009). An Efficient Binary Solvent Mixture for Monoacylglycerol Synthesis by Enzymatic Glycerolysis. JAOCS, Journal of the American Oil Chemists’ Society, 86, 783–789. http://doi.org/10.1007/s11746-009-1402-7 ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 18 Fig. 1 .TAG composition as a function of time in the glycerolysis of sardine oil in solvent free system with () and without (◊) substrates emulsification at atmospheric pressure (0.1 MPa). Reactions were performed at MR = 3:1 (glycerol:oil), T = 50ºC, enzyme loading 5 % wt. of substrates. 0 20 40 60 80 100 0 100 200 300 400 500 TAG (mol %) Time (min) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 19 Fig. 2. TAG composition as a function of time in the glycerolysis of sardine oil previous substrates emulsification with 1.5% of AOT in glycerol (□), 0.5% of Tween 80 in glycerol (○) and surfactant free () at atmospheric pressure (0.1 MPa).. Reactions were performed at MR = 3:1 (glycerol:oil), T = 50ºC, enzyme loading 5 % wt. of substrates. 0 20 40 60 80 100 0 100 200 300 400 500 TAG (mol %) Time (min) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 20 Fig. 3. Time course of lipase-catalyzed glycerolysis reaction of sardine oil in tert-butanol medium (hollow symbols) and in solvent free (solid symbols) with previous emulsification of the substrates at atmospheric pressure (0.1 MPa). Legend: TAG (, ▲), DAG (□, ■), MAG (◇, ◆) and FFA (○, ●). Reactions were performed at MR = 3:1 (glycerol:oil), T = 50ºC, enzyme loading 5 % wt. of substrates. 0 20 40 60 80 100 0 100 200 300 400 500 Lipid composition (mol %) Time (min) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 21 Fig. 4. Effect of pressure in SC-CO2 as reaction medium on glycerolysis conversion of sardine oil and products yield with previous substrates emulsification . (∆) conversion of TAG, (□, ◊, and ○) yields of DAG, MAG and FFA, respectively after 7 h of reaction time. Reactions were performed at MR = 3:1 (glycerol:oil), T = 50ºC, enzyme loading 5 % wt. of substrates. 0 20 40 60 80 100 0 5 10 15 20 25 Conversion or Yield (mol %) Pressure (MPa) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 22 Fig. 5. Effect of temperature on MAG + DAG composition as function of time in the glycerolysis of sardine oil with previous substrates emulsification in SC-CO2 at15 MPa: 40ºC (◊), 50ºC (□), 65ºC (∆), 80ºC (○) and 90ºC (x). Reactions were performed at MR = 3:1 (glycerol:oil) and enzyme loading 5 % wt. of substrates. 0 20 40 60 80 100 0 100 200 300 400 500 600 MAG + DAG (mol %) Time (min) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 23 Fig. 6. Influence of reaction temperature on PV (white bars) and AV (grey bars) in the final reaction mixture after 7 h at 15 MPa in SC-CO2 as reaction medium with previous substrates emulsficiation. Measurements given are mean values based on four determinations. Limit allowed is the maximum of each axis. Values with different letters in each type of analysis (PV or AV) are significantly different when applying the Tukey's honestly significant difference (HSD) method at p-value ≤ 0.05. ab c b b a ab ab a ab b b b 10 15 20 25 30 0 2 4 6 8 10 Initial 40 50 65 80 90 AV PV (mEq peroxide kg-1 oil) Reaction temperature (ºC) ACCEPTED MANUSCRIPT ACCEPTED MANUSCRIPT 24 Table 1. Summary of the reaction conditions for lipase-catalyzed sardine oil glycerolysis reactions carried out in this work. Exp. Reaction medium Pressure (MPa) Temperature (ºC) Emulsification Surfactant 1 Solvent free 0.1 50 No - 2 Yes - 3 Yes AOT 4 Yes Tween 80 5 80 Yes - 6 90 Yes - 7 SC-CO2 as solvent 15 50 Yes - 8 20 Yes - 9 25 Yes - 10 15 40 Yes - 11 65 Yes - 12 80 Yes - 13 90 Yes - ACCEPTED MANUSCRIPT