Supercritical carbon dioxide as solvent in the lipase-catalyzed ethanolysis of fish oil: kinetic study

Supe c i ical ca bon dioxide as sol en in he lipase- ca alyzed e hanolysis o ish oil: kine ic s udy Rod igo Melgosaa, M. Te esa Sanz*a, Ángela G. Solaesaa, Es he de Paza,Sag a io Bel ána, Daniela L. Lamasb aDepa men o Bio echnology and Food Science (Chemical Enginee ing Sec ion), Uni e si y o Bu gos, Bu gos, Spain bIns i u o Nacional de In es igación y Desa ollo Pesque o (INIDEP), Buenos Ai es, A gen ina *[email p o ec ed] Fax: 0034947258831 Abs ac Supe c i ical ca bon dioxide (SC-CO2) has been used as g een sol en in he lipase-ca alyzed e hanolysis o ish oil by Lipozyme RM IM a mild, non-oxida i e condi ions and wi h no sol en esidues. The e ec o expe imen al condi ions, ini ial subs a e e hanol/oil mola a io (2-38), p essu e (7.5-30 MPa), and empe a u e (323.15-353.15 K) on equilib ium con e sion, eac ion a e and oxida i e s a us o he p oduc s has been s udied. No e hanol inhibi ion has been obse ed a high concen a ions o e hanol, when pu ing in con ac i s he ish oil wi h he enzyme a oiding di ec con ac be ween he bioca alys and e hanol. Ope a ing p essu e a ec ed posi i ely he eac ion pe o mance in he ange in es iga ed. Visual obse a ion o he phase beha iou o he ini ial eac ion mix u e showed an “expanded liquid phase” ha helped enhancing eac ion a e, and a gas phase. Raising empe a u e accele a ed he eac ion up o a limi (343.15 K), obse ing highe enzyme he mal s abili y han in o he eac ion media (313.15 K). Howe e , lipid oxida ion inc eases wi h empe a u e. Up o 86 ± 1 % FAEE yield has been ound a MR = 6:1, 30 MPa and 323.15 K. Kine ic da a ha e been co ela ed by using a ma hema ical model based on he elemen a y eac ions o he 3-s ep anses e i ica ion. Kine ic a e cons an s, appa en ac i a ion olumes and ene gies a e epo ed o he i s ime o a lipase-ca alyzed e hanolysis eac ion in SC-CO2. Keywo ds Omega 3, lipase, e hanolysis, supe c i ical ca bon dioxide. 1. In oduc ion Fish oil is a na u al sou ce o omega-3 polyunsa u a ed a y acids (n-3 PUFAs) such as eicosapen aenoic acid (EPA, 20:5 n-3) and docosahexaenoic acid (DHA, 22:6 n-3). Heal h bene i s o hese compounds ha e been well es ablished in he li e a u e [1]. As a consequence, unc ional oods en iched wi h n-3 PUFAs ha e been he ype o unc ional ood p oduc s whose p oduc ion in Eu ope and USA has inc eased he mos in he las yea s [2]. Ne e heless, in a ecen ly published e iew i has been ound ha he excess o oxida ion in comme cial n-3 PUFA supplemen s a ec s be ween 11-62% o he analysed supplemen s [3]. T adi ional me hods o p oduc ion o n-3 PUFA concen a es om hei na u al sou ces ha e been ecen ly e iewed, and a numbe o no el echniques ha e been p oposed [2]. Among he la es , enzyma ic modi ica ion o oils ich in n-3 PUFAs in supe c i ical luids (SCFs) ises as an al e na i e o ob aining less oxidized ish oil de i a i es, compa ed o con en ional me hods. Se e al s udies ha e been ca ied ou on enzyma ic eac ions in di e en SCF media. A comp ehensi e e iew on his subjec was ca ied ou by Knez [4], wi h e e ences on di e en enzyma ic eac ions in dense gases, such as oxida ion, hyd olysis, es e i ica ion and anses e i ica ion. Supe c i ical ca bon dioxide (SC-CO2) is p obably he mos used SCF due o i s bene i s (non- oxic, non- lammable, eadily a ailable a high pu i ies and low cos s, and ela i ely mild c i ical condi ions, easily sepa a ed om he eac ion p oduc s by simple dep essu iza ion) ha a e appealing when choosing en i onmen al eplacemen o o ganic sol en s. Besides, by a ying he empe a u e and p essu e i allows he ac iona ion o he p oduc s. Some p e ious s udies o enzyma ic e hanolysis o na u al lipid sou ces in SC-CO2 ha e been epo ed in he li e a u e. Di e en immobilized lipases ha e been used as bioca alys , such as he non-speci ic lipase No ozyme 435 om Candida An a c ica [5-8] and he sn-1,3- egiospeci ic lipase Lipozyme TL-IM om The momuces lanuginosa [9, 10]. In his wo k, Lipozyme RM IM om Rhizomuco miehei, a sn-1,3 speci ic lipase, has been used as bioca alys . E hanolysis o palm ke nel oil in SC-CO2 by he homologous Lipozyme IM was s udied by Oli ei a and Oli ei a [7]. Al hough he bioca alys was epo ed o be sn-1,3- speci ic, i did no beha e as a egiospeci ic lipase and conside able amoun s o glyce ol we e ound in he eac ion p oduc s. Fu he mo e, he eac ion con e sion was ollowed in e ms o glyce ol p oduc ion, no aking in o accoun he eac ion in e media es (di- and monoacylglyce ides). Kondo e al. [11] ca ied ou he syn hesis o a y acid e hyl es e s (FAEEs) om SC-CO2-ex ac ed canola oil in a con inuous supe c i ical ex ac ion- eac ion (SFE-SFR) sys em by using Lipozyme RM IM as bioca alys , obse ing a dec ease in he FAEE p oduc ion a high e hanol concen a ion. Enzyma ic eac ions in SC-CO2 can be a ec ed by ope a ing p essu e in di e en ways. Acco ding o ansi ion s a e heo y and s anda d he modynamics, ope a ing p essu e can a ec he eac ion a e cons an s. Besides, densi y- ela ed changes in he physical pa ame e s o SC-CO2 may indi ec ly a ec he enzyme ca aly ic ac i i y, and hus he eac ion pe o mance [12]. Loss e al. [13] ha e ecen ly e iewed di e en applica ions o supe c i ical luids as al e na i e sol en o bioca alysis p ocesses, concluding ha he e seems o be no “ ule o humb” o p edic ing he e ec o p essu e on enzyme ac i i y in SC- CO2. Di ec e ec s o p essu e on enzyme esidual ac i i y and s abili y o Lipozyme RM IM ha e been p e iously in es iga ed, inding ha almos no changes occu ed in he ange be ween 10 and 25 MPa a 323.15 K [14]. Di e en esul s ha e been ound in he li e a u e ega ding he e ec o p essu e on Lipozyme RM IM-ca alyzed eac ions in SC-CO2. Fo ins ance, in he s udy o es e i ica ion o s ea ic acid wi h e hanol ca alyzed by Lipozyme IM in SC-CO2 in he ange om 6 o 20 MPa a 323.15 K, Nakaya e al. [15] ound an inc ease in es e i ica ion a e wi h an inc ease in p essu e, bu a maximum was ound in he hyd olysis a e o he co esponding e hyl s ea a e. Laudani e al. [16] pe o med a de ailed kine ic and he modynamic s udy o he es e i ica ion o oleic acid wi h 1-oc anol ca alyzed by Lipozyme RM IM in dense ca bon dioxide. These au ho s epo ed a posi i e e ec o inc easing p essu e om 8 o 10 MPa a 323.15 K. Fu he inc ease in p essu e up o 30 MPa led o a dec ease in eac ion con e sion om 84 % o 77 %. The e o e, he e ec o p essu e on enzyme ac i i y in CO2 is e y dependen no only on he speci ic enzyme, bu also on he eac ion s udied and he phase beha iou o he sys em a di e en p essu e and empe a u e condi ions. In his wo k, he e ec o he ini ial mola a io o subs a es, p essu e, and empe a u e on equilib ium yield and eac ion a e has been s udied o he e hanolysis o ish oil in SC-CO2, co e ing a wide ange han p e ious wo ks epo ed in he li e a u e. Addi ionally, oxida ion pa ame e s o he e ined ish oil and he eac ion p oduc s ob ained om he eac ions in SC- CO2 ha e been de e mined and compa ed wi h hose ob ained om eac ions pe o med in con en ional o ganic sol en s and in sol en - ee media a a mosphe ic p essu e. This way, op imal eac ion condi ions conside ing kine ic aspec s and quali y o he p oduc s can be de e mined. Expe imen al da a ha e been sa is ac o ily co ela ed by a simple semi-empi ical kine ic model based on he elemen a y eac ions ha may occu in his sys em and aking in o accoun eac ion in e media es. 2. Ma e ial and Me hods 2.1. Ma e ials Lipozyme RM IM, a lipase om Rhizomuco miehei immobilized on a mac opo ous esin, was pu chased om No ozymes A/S (Denma k). Re ined ish oil was kindly p o ided by AFAMSA S.A. (Spain) being a mix u e o una (Thunnus sp.) and sa dine (Sa dina pilcha dus) oil. Fa y acid p o ile o he ish oil has been p e iously epo ed wi h a 24 % mol o docosahexaenoic acid (DHA) and 7 % mol o eicosapen anoic acid (EPA) [17]. Absolu e e hanol (99.9 %) was pu chased om Me ck KGaA. Ca bon dioxide (99.9%) was supplied by Ai Liquide S.A. (Spain). All o he chemicals used in di e en analyses we e o analy ical o HPLC g ade. 2.2. E hanolysis o ish oil in SC-CO2 The e hanolysis eac ion has been pe o med in a high p essu e ba ch s i ed ank eac o (HP- BSTR) made o s ainless s eel (SS-316) and ha ing an in e nal olume o 100 mL. A schema ic diag am o he expe imen al appa a us is shown in Figu e 1. In a ypical expe imen , a weighed amoun o enzyme (5.0 % w . o subs a es) was added in o he eac o oge he wi h a known amoun o ish oil. Subsequen ly, e hanol was added acco ding o he es ablished ini ial subs a e mola a io. This p ocedu e was adop ed in o de o a oid di ec con ac o e hanol wi h he enzyme, which may cause inac i a ion o he ca alys . The eac o was hen closed, connec ed o he p essu e ci cui and placed in a he mos a ic wa e ba h a he desi ed ope a ing empe a u e. Subsequen ly, SC-CO2 was ed in o he eac o by means o a high p essu e pump (ISCO 260 D) up o he desi ed p essu e, which was main ained by a digi al p essu e con olle . A Bou don p essu e gauge also p o ided a seconda y lec u e. Once he es ablished condi ions ha e been eached, magne ic s i ing was connec ed and he eac ion was ini ia ed. Figu e 1. Schema ic diag am o he high p essu e appa a us used o he e hanolysis eac ions in SC-CO2. 1: CO2 ese oi ; 2: sy inge pump; 3: c yos a ; 4: up u e disk; 5: high p essu e ba ch s i ed ank eac o ; 6: he mos a ic wa e ba h; 7: magne ic s i e ; 8: sampling de ice. Ope a ing empe a u e and p essu e ha e been a ied in he ange be ween 323.15-353.15 K and 7.5-30 MPa, espec i ely. The e ec o he ini ial subs a e mola a io has been s udied in he ange om 2:1 o 38:1 (e hanol: ish oil). Samples we e aken pe iodically du ing 24 h h ough a siphoned capilla y equipped wi h a mic o il e made o sin e ed s eel, which p e en ed he wi hd awal o he enzyme om he eac ion mix u e. Samples we e collec ed in glass sc ew- op ials imme sed in a cold ap and s o ed a -18 ºC up o analysis. P essu e d ops up o 0.5 MPa we e obse ed du ing he wi hd awal o he samples, which we e compensa ed by eeding esh SC-CO2 a he desi ed p essu e in o he eac o . Acco ding o he low mass o he samples (ca. 0.1 g) compa ed wi h he ini ial loading o he HP-BSTR, dis u bances o he ba ch p ocess we e conside ed negligible. 2.3. De e mina ion o he composi ion Neu al lipid p o ile o he samples ( a y acid e hyl es e s, FAEEs; Monoacylgyce ides, MAGs; diacylglyce ides, DAGs; and un eac ed iacylglyce ides, TAGs) has been de e mined by no mal phase HPLC. Ch oma og aphic equipmen , me hod and calib a ion p ocedu e ha e been p e iously desc ibed in de ail [18]. Ch oma og aphic analysis o glyce ol (GLY) con en in he eac ion samples was pe o med using High-Tempe a u e Gas Ch oma og aphy (HTGC). Me hod and calib a ion p ocedu e ha e been p e iously desc ibed [17]. GLY con en in he eac ion samples was also heo e ically calcula ed by a balance o he glyce ol backbone, as p oposed by So o á e al. o he enzyma ic hyd olysis o blackcu an oil in SC-CO2 [19]. A modi ied exp ession o an e hanolysis eac ion gi es: nGLY = (nFAEE – nDAG – 2·nMAG)/3 (1) whe e nFAEE , nDAG , and nMAG a e he FAEE, DAG and MAG mole con en in he eac ion samples and nGLY is he heo e ical GLY mole con en . Theo e ical calcula ion de ia ed less han 10% om expe imen al da a, hus HT-GC de e mina ion o GLY and neu al lipid p o ile analysis we e sa is ac o ily ela ed. Un eac ed E OH was heo e ically calcula ed conside ing he eac ion s oichiome y, in which he p oduc ion o 1 mol o FAEE consumes 1 mol o E OH, gi ing: nE OH = nE OH,o – nFAEE (2) whe e nE OH,o is he ini ial mole con en o e hanol. 2.4. Measu emen o lipid oxida ion De e mina ion o he pe oxide alue (PV), p-anisidine alue (p-AnV), and acid alue (AV) o he samples be o e and a e he kine ic expe imen s ha e been pe o med in o de o e alua e po en ial lipid oxida ion p ocesses du ing he e hanolysis eac ions. The pe oxide alue, PV, measu es he concen a ion o pe oxides and hyd ope oxides o med in he ini ial s ages o lipid oxida ion (p ima y oxida ion). The p-anisidine alue (p-AnV) is an es ima ion o he concen a ion o seconda y oxida ion p oduc s. De e mina ion o he acid alue (AV) has been also pe o med as an es ima ion o he hyd oly ic ancidi y o he ish oil and he eac ion samples. All de e mina ions we e pe o med acco ding o s anda d me hods [20-22]. In he case o eac ion samples, lipid ac ions we e ob ained by means o e apo a ion o un eac ed e hanol in a acuum o a y e apo a o (Heidolph). 2.5. Kine ic model Lipase-ca alyzed e hanolysis o iacylglyce ydes (TAG) o ish oil can be conside ed as a 3- s ep anses e i ica ion. A each s ep, one molecule o FAEE and a glyce ide con aining one ewe es e bond a e ob ained. Glyce ides in ol ed in he eac ion a e di- and monoacylglyce ides (DAG and MAG), and glyce ol (GLY) as he las p oduc . Following he p oposed model, he eac ion akes place h ough he ollowing s eps: 1. Con e sion o i- o diacylglyce ides: TAG +E OH k1 ⇄ k−1 DAG +FAEE (3) 2. Con e sion o di- o monoacylglyce ides: DAG +E OH k2 ⇄ k−2 MAG + FAEE (4) 3. Con e sion o monoacylglyce ides o glyce ol: MAG + E OH k3 ⇄ k−3 GLY +FAEE (5) To co ela e he expe imen al kine ic da a, a semi-empi ical model based on he mass balance equa ions o all he species in he eac ion sys em has been employed. Al hough he sn-1,3- speci ic ca alys canno deacyla e he sn-2 posi ion o he acylglyce ide, s ep 3 (Eq. 5) should be conside ed because isome iza ion o 2-MAG o 1(3)-MAG (acyl-mig a ion) may occu . As he egioisome s 1,2- and 1,3-DAG; and 1(3)- and 2-MAG could no be dis inguished wi h he applied analy ical p ocedu e, no di e ence was made be ween hem in he model. Hyd olysis eac ion has no been aken in o accoun since no ee a y acids we e de ec ed (<0.1%). The kine ic equa ions in ol ed in he e hanolysis sys em a e he ollowing: d(nTAG/n o al)/d = – k’1·xTAG·xE OH + k’–1·xDAG·xFAEE (6.1) d(nDAG/n o al)/d = k’1·xTAG·xE OH – k’–1·xDAG·xFAEE – k’2·xDAG·xE OH + k’–2·xMAG·xFAEE (6.2) d(nMAG/n o al)/d = k’2·xDAG·xE OH – k’–2·xMAG·xFAEE – k’3·xMAG·xE OH + k’–3·xGLY·xFAEE (6.3) d(nGLY/n o al)/d = k’3·xMAG·xE OH – k’–3·xGLY·xFAEE (6.4) d(nFAEE/n o al)/d = k’1·xTAG·xE OH – k’–1·xDAG·xFAEE + k’2·xDAG·xE OH – k’–2·xMAG·xFAEE + k’3·xMAG·xE OH – k’–3·xGLY·xFAEE (6.5) d(nE OH/n o al)/d = – k’1·xTAG·xE OH + k’–1·xDAG·xFAEE – k’2·xDAG·xE OH + k’–2·xMAG·xFAEE – k’3·xMAG·xE OH + k’–3·xGLY·xFAEE (6.6) Equilib ium cons an s o each eac ion s ep, e alua ed as Ki = k’i/k’–i, a e also epo ed. To ou knowledge, his is he i s ime ha kine ic a e cons an s a e epo ed o he h ee s eps o a lipase-ca alyzed e hanolysis in SC-CO2. Figu e 4. a) Equilib ium FAEE yield (%) s. MR (e hanol :oil) in he e hanolysis o ish oil ca alyzed by Lipozyme RM IM in SC-CO2 medium. Solid line is om he non-linea eg ession o he expe imen al da a (Eq. 9). b) Loga i hmic ela ionship be ween ini ial eac ion a e and MR. Dashed line is om he linea eg ession o he expe imen al da a. Expe imen al condi ions: p = 10 MPa, T = 323.15 K and enzyme loading 5 % w . o subs a es As i can be obse ed om Table 2, he o wa d and e e se a e cons an s ollow he o de k’3 > k’2 > k’1 and k’–2 > k’–3 > k’–1. An inc ease in MR leads o an inc ease in bo h he o wa d and he e e se a e cons an s. In he MR ange s udied, he o wa d a e cons an o he hi d s ep (MAG o p oduce FAEE, k’3) is la ge han he o he wo o wa d a e cons an s, being he ini ial b eakdown o TAG he slowes s ep and he e o e he a e-limi ing s ep o he e hanolysis eac ion. This beha iou has been also desc ibed in he li e a u e o ei he acid o base-ca alyzed anses e i ica ion [30]. I can be also no iced ha equilib ium cons an s mono onously dec eased wi h inc easing MR due o he excess o e hanol employed. Besides, he equilib ium cons an o he hi d eac ion s ep, K3, is one o de o magni ude highe han hose o he o he wo s eps. In he li e a u e, di e en mechanisms ha e been p oposed o lipase-ca alyzed anses e i ica ion sys ems, mos o hem based on Ping-Pong Bi-Bi models [31]. Compa ison wi h hese s udies is di icul since in mos cases he modynamic pa ame e s a e no p o ided. In his wo k, a simple model was adop ed and su p isingly, leads o kine ic and equilib ium pa ame e s o he same o de as hose epo ed by chemical- ca alysis [30]. Table 2. E ec i e o wa d (k’i) and e e se (k’–i) eac ion a e cons an s, equilib ium cons an s (Ki) and objec i e unc ion (O.F.) alues o he p oposed kine ic model o he e hanolysis o ish oil by Lipozyme RM IM in SC-CO2 a di e en ini ial mola a io o subs a es (MR). Reac ions we e pe o med a p = 10 MPa, T = 323.15 K, enzyme loading 5 % w . o subs a es. MR (e hanol:oil) k’ 1 (min -1 ) k’ –1 (min -1 ) K1 k’ 2 (min -1 ) k’ –2 (min -1 ) K2 k’ 3 (min -1 ) k’ –3 (min -1 ) K3 O.F. 2:1 0.0971 0.5998 0.1543 1.3972 2.4197 0.5511 3.7926 0.9194 4.1315 0.1704 4:1 0.1703 0.8898 0.1786 1.5682 4.0452 0.3684 5.3823 1.3874 4.0460 0.0592 6:1 0.4764 1.3465 0.4867 3.2605 5.9020 0.5290 7.4221 2.5333 2.6911 0.0309 10:1 0.5177 6.7735 0.0946 22.2968 164.8420 0.1426 11.5229 7.5545 2.1538 0.0137 38:1 3.5370 34.7671 0.0152 16.3786 158.3258 0.1178 27.8656 56.5390 0.2191 0.0047 3.2.P essu e e ec P essu e has been a ied in he ange om 7.5 o 30 MPa a ixed MR = 6:1 (e hanol: ish oil), 323.15 K and an enzyme loading o 5 % w . o subs a es. Table 1 shows ha ini ial eac ion a e s eadily inc eases wi h p essu e om 7.5 o 30 MPa, and FAEE yield a equilib ium inc eases om 51.4 % nea he c i ical p essu e (7.5 MPa) o 80 % a 9 MPa, showing a pla eau a ound 81-85 % a highe p essu es up o 30 MPa (Figu e 5). In he ange in es iga ed (7.5-30 MPa), no decay in he eac ion pe o mance was obse ed when inc easing p essu e. Figu e 5. E ec o ope a ing p essu e (p) on he e hanolysis o ish oil ca alyzed by Lipozyme RM IM in SC- CO2 medium. Expe imen al condi ions: MR = 6:1, T = 323.15 K and enzyme loading 5 % w . o subs a es. Lines ep esen he i ing o he p oposed kine ic model o he expe imen al da a Figu e 6 shows ha a semi-loga i hmic ela ionship can be es ablished be ween ini ial eac ion a e and ope a ing p essu e (dashed line, ln o = 0.4220·ln p + 3.6157; R2 = 0.9774). Besides, a simila exp ession o he one p oposed in Eq. 9 can be es ablished o desc ibe he e ec o p essu e on he equilib ium FAEE yield. Ope a ing p essu e was exp essed in e ms o educed p essu e (p = p/pC). Non‐linea eg ession was pe o med by using he Ma qua d algo i hm (S a g aphics) gi ing a limi ing equilib ium FAEE yield o 83.47 %, b = 0.079 and p 0 = 0.982 (p0 = 7.22 MPa) wi h R2 = 0.973. The con inuous line in Figu e 6 co esponds o his adjus men . Figu e 6. Open iangles: Equilib ium FAEE yield (%) s. educed ope a ing p essu e (p = p/pC) in he e hanolysis o ish oil ca alyzed by Lipozyme RM IM in SC-CO2 medium. Solid line is om he non-linea eg ession o he expe imen al da a (Eq. 9). Filled ci cles: loga i hmic ela ionship be ween ini ial eac ion a e and p . Dashed line is om he linea eg ession o he expe imen al da a. Expe imen al condi ions: MR = 6:1, T = 323.15 K and enzyme loading 5 % w . o subs a es The e ec o ope a ing p essu e on he e hanolysis o ish oil by Lipozyme RM IM in SC-CO2 could be explained conside ing he phase beha iou o he ish oil/e hanol eac ion mix u e in SC-CO2. Visual obse a ions o he ini ial eac ion mix u e a he di e en p essu es assayed in his wo k showed ha he eac ion sys em consis ed o an “expanded” liquid and a gas phase. A he condi ions s udied in his wo k, solubili y o CO2 in e hanol is high [32], whe eas Bo ch-Jensen and Molle up [33] epo ed mode a e solubili y o CO2 in ish oil a p essu es om 6 o 65 MPa and empe a u es om 293.15 o 393.2 K. A none o he empe a u es, ca bon dioxide and ish oil we e comple ely miscible, bu a each empe a u e, CO2 solubili y in ish oil was ound o inc ease wi h inc easing p essu e [33]. This inc ease in CO2 solubili y wi h p essu e may co espond o he inc ease in he eac ion a e wi h p essu e as shown in Figu e 6, since CO2 sol a ion esul s in a be e mass ans e eac ion medium due o an inc ease in di usi i y and a educ ion o medium iscosi y. The di e en e ec s o p essu e depending on he phase beha iou o he sys em can be no ed when compa ing he esul s ob ained in his wo k wi h hose om he es e i ica ion o oleic acid wi h n-oc anol in SC-CO2 [16]. Phase beha iou o he sys em oleic acid + n-oc anol + CO2 shows ha CO2 is highly soluble in he eac ion mix u e (oleic acid and e hanol) and he liquid phase can con ain up o 70 % mole o CO2 nea he c i ical poin o CO2 [16]. Fu he inc ease in ope a ing p essu e p omo ed a dec ease in con e sion due o mo e CO2 sol a ing in he eac ion bulk and leading o dilu ion o he subs a es. On he con a y, he mix u e ish oil + e hanol can dissol e a much lowe amoun o CO2 (30 % mole a MR = 6:1, 10 MPa and 323.15 K, acco ding o analy ical de e mina ion o he phase beha iou in a high p essu e iew cell). The e o e, dilu ion o he subs a es a high p essu e is no supposed o s ongly a ec he eac ion pe o mance in he ish oil + e hanol + CO2 eac ion sys em. Oli ei a and Oli ei a [7] adop ed a Taguchi expe imen al design o assess he in luence o he p ocess a iables on he e hanolysis o palm ke nel oil in SC-CO2 ca alyzed by Lipozyme IM. Acco ding o hei esul s, Lipozyme IM was posi i ely a ec ed by p essu e, al hough MR was he a iable ha mo e s ongly a ec ed he con e sion. Fo his enzyme hey ound an op imum a 14.6 MPa. On he con a y, in he p esen wo k, no maximum in he p essu e was ound in he ange om 7.5 o 30 MPa, al hough simila phase beha iou could be expec ed o bo h eac ion mix u es ( ish oil + e hanol + CO2 o palm ke nel oil + e hanol + CO2). The e ec o p essu e on he kine ic a e cons an s has been aken in o accoun by using he ansi ion-s a e heo y and classical he modynamics. Following his app oach, he a ia ion o he eac ion a e cons an k’i wi h p essu e o a bimolecula eac ion “A + B = M* = p oduc s” can be exp essed as ollows [12]: (∂ln (k)/∂p)T = – ΔV*/RT (10.1) whe e k is he a e cons an o he eac ion exp essed in p essu e-independen concen a ion uni s, ΔV* is he appa en ac i a ion olume, T is he ope a ing empe a u e and R is he gas cons an . The di ec in eg a ion o Eq 10.1 is no s aigh o wa d since ac i a ion olume changes wi h p essu e [12]. Howe e , wi hin a small ange o p essu e i can be assumed ha ΔV* does no change wi h p essu e and Eq. 10.1 can be easily in eg a ed, gi ing: k’i = k0,i·exp(– p·ΔVi*/RT) (10.2) whe e he subsc ip i e e s o he di e en s eps in he e hanolysis eac ion. Following his exp ession, p e-exponen ial kine ic cons an s, k0i, and appa en ac i a ion olumes, ΔVi*, ha e been simul aneously es ima ed o he expe imen s pe o med in he ange 9-30 MPa. Resul s ob ained a e lis ed in Table 3. Kine ic pa ame e s o he expe imen pe o med a p = 7.5 MPa we e es ima ed sepa a ely because o he ma ked changes in he physical p ope ies o he sol en nea he c i ical egion (p = 1.02). In he p essu e ange 9-30 MPa, appa en ac i a ion olumes we e nega i e (ΔVi* < 0) o all he eac ion s eps, which indica es ha eac ion a e cons an s will inc ease wi h inc easing p essu e. I can be obse ed ha ΔV* o he o wa d i s and hi d eac ion s eps a e lowe (o highe in absolu e alue) han he co esponding ΔV* o he e e se eac ion, being his s eps mo e sensi i e o an inc ease in ope a ing p essu e. On he con a y, he DAG o MAG con e sion shows simila ΔV* alues o he o wa d and e e se eac ions, which sugges li le in luence o p essu e in his s ep. O e all conside a ion o ΔV* alues indica es ha FAEE p oduc ion may be a ou ed by inc easing p essu e, as i can be obse ed om he expe imen al esul s. To ou knowledge, no ΔV* alues ha e been p e iously epo ed o lipase-ca alyzed e hanolysis in SC-CO2. Howe e , a simila ΔV* alue o ca. -206 cm3·mol-1 was epo ed by He e al. [34] o he anses e i ica ion o soybean oil wi hou ca alys in sub- and supe c i ical me hanol be ween 8.7 MPa and 36 MPa (553 K and MR me hanol:soy bean oil = 42:1). In any case, Kama e al. [12] s a ed ha he use o ΔV* in enzyme-ca alyzed eac ions mus be ea ed wi h cau ion, since changes in p essu e will esul in mul iple a iables being changed ha also in luence he abili y o he enzyme o ca alyse a gi en eac ion. The e o e ΔV* should no be used o compa e he e ec s o p essu e o ca alyzed and unca alyzed eac ions. Fo an unca alyzed eac ion, da a a e only dependen o di ec p essu e-e ec s and no indi ec e ec s o p essu e a e ansmi ed ia he enzyme. Table 3. Values o he o wa d (k0,i) and e e se (k0,–i) p e-exponen ial cons an s, appa en ac i a ion olume o each o wa d (ΔV*i) and e e se (ΔV*–i) eac ion s ep and alues o he objec i e unc ion (O.F.) o he p oposed kine ic model o he e hanolysis o ish oil by Lipozyme RM IM in SC-CO2 a 7.5 MPa and in he ange 9- 30 MPa. Reac ions we e pe o med a MR = 6:1, T = 323.15 K, enzyme loading 5 % w . o subs a es. p (MPa) k 0,1 (min -1 ) k 0,–1 (min -1 ) ΔV* 1 (cm 3 mol -1 ) ΔV* –1 (cm 3 mol -1 ) k 0,2 (min -1 ) k 0,–2 (min -1 ) ΔV* 2 (cm 3 mol -1 ) ΔV* –2 (cm 3 mol -1 ) k 0,3 (min -1 ) k 0,–3 (min -1 ) ΔV* 3 (cm 3 mol -1 ) ΔV* –3 (cm 3 mol -1 ) O.F. 9 – 30 0.2366 1.1770 -158.0394 -47.6628 1.5824 2.6906 -126.3835 -127.7673 3.0135 1.7217 -145.7125 -88.7143 0.0889 7.5 k’1= 0.3099 k’-1= 43.1938 k’2= 2.3257 k’-2= 12.1366 k’3= 43.8870 k’-3= 45.9396 0.0087 3.3. Tempe a u e e ec To assess he e ec o empe a u e on he kine ics o he e hanolysis o ish oil by Lipozyme RM IM in SC-CO2, ope a ing empe a u e has been a ied be ween 323.15 and 353.15 K. Ini ial subs a e mola a io (6:1 e hanol: ish oil), p essu e (10 MPa) and enzyme loading (5% w . o subs a es) emained unchanged. Figu e 7 shows ha FAEE yield a equilib ium was simila in he empe a u e ange om 323.15 o 343.15 K. The hea o eac ion is gene ally small o many anses e i ica ion sys ems; he e o e he equilib ium con e sion obse ed o he e hyl es e s is essen ially empe a u e independen [35]. Raising empe a u e om 323.15 o 343.15 K esul ed in an inc ease o he ini ial eac ion a e (Table 1 and Figu e 8), p obably because o a highe kine ic ene gy o he molecules. Besides, lowe iscosi y and highe di usi i y o he sol en a highe empe a u es may lead o lowe mass ans e limi a ions [4]. The highes empe a u e assayed in his wo k (353.15 K) led o lowe equilib ium FAEE yield (Table 1 and Figu e 7), which may be due o he mal deac i a ion o he ca alys . The e ec o empe a u e on ini ial eac ion a e is shown in he A henius plo (Figu e 8). F om his igu e, i can be seen ha he A henius dependence is no longe alid a empe a u es highe han 343.15 K. As i has been p e iously men ioned, his beha iou may co espond o he mal deac i a ion o he ca alys . Figu e 7. E ec o ope a ing empe a u e (T) on he e hanolysis o ish oil ca alyzed by Lipozyme RM IM in SC-CO2 medium. Expe imen al condi ions: MR = 6:1, p = 10 MPa and enzyme loading 5 % w . o subs a es. Lines ep esen he i ing o he p oposed kine ic model o he expe imen al da a. Simila esul s o he mal beha iou o Lipozyme RM IM ha e been epo ed in he li e a u e o he syn hesis o n-oc yl olea e in SC-CO2 [16]. Con e sion a ound 80 % was obse ed in he ange 308.15-333.15 K a 10 MPa, whe eas highe empe a u es (343.15 and 353.15 K) led o lowe con e sion (a ound 65 %) ye sligh ly highe ini ial eac ion a es, which may indica e ha he mal deac i a ion is no immedia e. Oli ei a and Oli ei a [7] epo ed T = 324 K as he op imum empe a u e o he e hanolysis o palm ke nel oil by Lipozyme IM in SC-CO2, whe eas 313.15 K was ound o be he op imum o he same eac ion in n-hexane [8]. Recen ly, Cale o e al. [28] ha e ound he same ope a ing empe a u e (T = 313.15 K) as he op imum o he e hanolysis o sun lowe oil by Lipozyme RM IM in sol en - ee media Figu e 11. E ec o he e hanolysis o ish oil ca alyzed by Lipozyme RM IM on he pe oxide alue (PV) in di e en eac ion media: SC-CO2 a di e en ope a ing p essu e, SF (sol en - ee media) and 20 w % o TP. Dashed line ep esen s he ecommended limi se by he FDA [38]. To ou knowledge, no o he s udies in he li e a u e assessed lipid oxida ion p ocesses du ing enzyma ic e hanolysis o ish oil in SC-CO2. Pa k e al. [39] de e mined PV and conjuga ed diene (CD) con en o comme cial salmon oil be o e and a e enzyma ically (Lipozyme IM) and chemically ca alyzed e hanolysis a a mosphe ic p essu e and n-hexane as he eac ion media. They ound ha bo h me hods inc eased PV and CD, ye li le oxida ion and isome iza ion o PUFAs we e ound when Lipozyme IM was used as he ca alys [39]. Resul s ob ained in his wo k show ha enzyma ic e hanolysis o ish oil in SC-CO2 can be conside ed a sui able me hod o ob ain less oxidized eac ion p oduc s compa ed o hose ob ained by enzyma ic e hanolysis in con en ional o ganic sol en s o in sol en - ee media a a mosphe ic p essu e. 4. Conclusions SC-CO2 has been used as a g een sol en in he anses e i ica ion o ish oil by Lipozyme RM IM, p o iding an en i onmen ally benign eac ion medium. Ad an ages o using SC-CO2 include eplacing o ganic sol en s, enhancing eac ion kine ics by educing mass ans e limi a ions and p e en ing oxida ion due o displacemen o oxygen. The la e is especially impo an when wo king wi h easily oxidizable compounds such as n-3 PUFAs. Enzyme and phase beha iou a e key pa ame e s o unde s and biocon e sion in SC-CO2. The lipase showed highe he mal s abili y in SC-CO2 eac ion medium han in o he con en ional eac ion media. Besides, no e hanol inhibi ion has been obse ed when a oiding high concen a ions o e hanol in he enzyme en i onmen . Ope a ing p essu e a ec ed posi i ely he eac ion pe o mance due o sol a ion o CO2 in he eac ion mix u e, which educes iscosi y and imp o es di usion coe icien s. Lipase-ca alyzed e hanolysis in SC-CO2 has been shown as sui able me hod o ob ain less oxidized n-3 PUFA FAEE compa ed o o he eac ion media. Co ela ion o he kine ic da a o a semi-empi ical model showed ha he a e-limi ing s ep is he b eakdown o iacylglyce ides. Simila ends o kine ic and equilib ium pa ame e s ha e been obse ed as hose epo ed by chemical-ca alysis. Acknowledgemen s To he Spanish Go e nmen (MINECO) and he Eu opean Regional De elopmen Fund (ERDF) o inancial suppo o he p ojec CTQ2012-39131-C02-01. Financial suppo om he Jun a de Cas illa y León and Eu opean Regional De elopmen Fund (ERDF) h ough p ojec BU055U16 is also g a e ully acknowledged. RM acknowledges MINECO o a g an (BES-2013-063937). AGS acknowledges Uni e si y o Bu gos o a ellowship. EP acknowledges MINECO o a ellowship (FJCI-2014-19850). Re e ences [1] C. H. S. Rux on, S. C. Reed, M. J. A. Simpson and K. J. Milling on, Jou nal o Human Nu i ion and Die e ics, 2004, 17, 449–459. [2] N. Rubio-Rod íguez, S. Bel án, I. Jaime, S. M. de Diego, M. T. Sanz and J. R. Ca ballido, Inno a i e Food Science and Eme ging Technologies, 2010, 11, 1–12. [3] B. B. Albe , D. Came on-Smi h, P. L. Ho man and W. S. Cu ield, BioMed Resea ch In e na ional, 2013, 2013, 464921. [4] Ž. Knez, Jou nal o Supe c i ical Fluids, 2009, 47, 357–372. [5] H. Gunnlaugsdo i and B. Si ik, Jou nal o Ame ican Oil Chemis s’ Socie y, 1995, 72, 399–405. [6] H. Gunnlaugsdo i , J. Ma ias and B. Si ik, Jou nal o Supe c i ical Fluids, 1998, 12, 85–93. [7] J. V. Oli ei a and D. Oli ei a, Indus ial and Enginee ing Chemis y Resea ch, 2000, 39, 4450–4454. [8] D. Oli ei a and J. V. Oli ei a, Jou nal o Supe c i ical Fluids, 2001, 19, 141–148. [9] S.-K. Shin, J.-E. Sim, H. Kishimu a, B.-S. Chun, Jou nal o Indus ial and Enginee ing Chemis y, 2012, 18, 546-550. [10] H.-Y. Lee, Tanbi ul Haque A.S.M., S.-B. Kim, Y.-B. lee, B.-S. Chun, Jou nal o Indus ial and Enginee ing Chemis y, 2014, 20, 1097-1102. [11] M. Kondo, K. Rezaei, F. Temelli and M. Go o, Indus ial and Enginee ing Chemical Resea ch, 2002, 41, 5770–5774. [12] S. V Kama , E. J. Beckman and A. J. Russell, C i ical Re iews in Bio echnology, 1995, 15, 41–71. [13] R. Loss, L. Le in, J. V. Oli ei a and D. Oli ei a, eds. M. A. Z. Coelho and B. D. Ribei o, Whi e Bio echnology o Sus ainable Chemis y, The Royal Socie y o Chemis y, 2015, 104–135. [14] R. Melgosa, M. T. Sanz, Á. G. Solaesa, S. L. Bucio and S. Bel án, Jou nal o Supe c i ical Fluids, 2015, 97, 51–62. [15] H. Nakaya, K. Nakamu a and O. Miyawaki, Jou nal o he Ame ican Oil Chemis s’ Socie y, 2002, 79, 23–27. [16] C. G. Laudani, M. Habulin, Ž. Knez, G. Della Po a and E. Re e chon, Jou nal o Supe c i ical Fluids, 2007, 41, 92–101. [17] S. L. Bucio, A. G. Solaesa, M. T. Sanz, S. Bel án and R. Melgosa, Jou nal o Chemical and Enginee ing Da a, 2013, 58, 3118–3124. [18] A.G. Solaesa, S.L. Bucio, M. T. Sanz, S. Bel án, and S. Rebolleda, Jou nal o Oleo Science, 2014, 63, 449–460. [19] H. So o á, M. Za e úcka, P. Be nášek and M. S amenić, Chemical Enginee ing Resea ch and Design, 2008, 86, 673–681. [20] Ame ican Oil Chemis ’ Socie y, AOCS O icial Me hod Cd 3d-63, 2009. [21] AOAC In e na ional, AOAC O icial Me hod 965.33, 2002. [22] Ame ican Oil Chemis ’ Socie y, AOCS O icial Me hod Cd 18-90, 2011. [23] A. Millq is Fu eby, C. Vi o, P. Adle c eu z and B. Ma iasson, Bioca alysis and Bio ans o ma ion, 1996, 14, 89–111. [24] X. Xu, A. R. H. Skands, C. E. Høy, H. Mu, S. Balchen and J. Adle -Nissen, Jou nal o he Ame ican Oil Chemis s’ Socie y, 1998, 75, 1179–1186. [25] O. N. Ci ci and F. Temelli, Jou nal o Supe c i ical Fluids, 2011, 58, 79–87. [26] W. Piya hee awong, Y. Iwasaki, X. Xu and T. Yamane, Jou nal o Molecula Ca alysis B: Enzyma ic, 2004, 28, 19–24. [27] V. Kuma i, S. Shah and M. N. Gup a, Ene gy & Fuels, 2007, 21, 368–372. [28] J. Cale o, C. Ve dugo, D. Luna, E. D. Sancho, C. Luna, A. Posadillo, F. M. Bau is a and A. A Rome o, New Bio echnology, 2014, 31, 596–601. [29] D. M. Ches e ield, P. L. Roge s, E. O. Al-Zaini and A. A. Adesina, Chemical Enginee ing Jou nal, 2012, 207-208, 701–710. [30] A. Si asamy, K. Y. Cheah, P. Fo nasie o, F. Kemausuo , S. Zino ie and S. Mie us, ChemSusChem, 2009, 2, 278–300. [31] B. Chei silp, A. H-Ki ikun and S. Limka anyu, Biochemical Enginee ing Jou nal, 2008, 42, 261–269. [32] J. S. Lim, Y. Y. Lee and H. S. Chun, Jou nal o Supe c i ical Fluids, 1994, 7, 219–230. [33] C. Bo ch-Jensen and J. Molle up, Fluid Phase Equilib ia, 1997, 138, 179–211. [34] H. He, S. Sun, T. Wang and S. Zhu, Jou nal o he Ame ican Oil Chemis s’ Socie y, 2007, 84, 399–404. [35] S. L. Bucio, Á. G. Solaesa, M. T. Sanz, R. Melgosa, S. Bel án and H. So o á, Jou nal o Oleo Science, 2015, 64, 431–441. [36] D. Oli ei a, A. C. Feih mann, A. F. Rubi a, M. H. Kuni a, C. Da i a and J. V. Oli ei a, Jou nal o Supe c i ical Fluids, 2006, 38, 373–382. [37] R. C. Rod igues and R. Fe nandez-La uen e, Jou nal o Molecula Ca alysis B: Enzyma ic, 2010, 66, 15–32. [38] U.S. Depa men o Heal h and Human Se ices, Food and D ug Adminis a ion. GRAS No ice 000200, 2006. [39] S.-B. Pa k, Y. Endo, K. Ma uyama and K. Fujimo o, Food Science and Technology Resea ch, 2000, 6, 192–195.