Separation of sodium lactate from Span 80 and SDS surfactants by ultrafiltration

1 Sepa a ion o sodium lac a e om Span 80 and SDS su ac an s by ul a il a ion La a Roque, Isabel Escude o*, José M. Beni o Depa men o Bio echnology and Food Science, Uni e si y o Bu gos, Plaza Misael Bañuelos s/n, 09001 Bu gos, Spain E-mail add esses: l 0010@uni e sidaddebu gos.es (L. Roque), [email p o ec ed] (J.M. Beni o) *Co esponding au ho . Tel.: +34-947258809; ax: +34-947258831. E-mail add ess: [email p o ec ed] Abs ac The ul a il a ion p ocess o sepa a ion o sodium lac a e om so bi an monoolea e (Span 80) and sodium dodecyl sul a e (SDS) su ac an s using Z O2 la -disc ul a il a ion memb anes was s udied in his wo k. The s udy is ocused on he in luence o he nominal molecula weigh limi o he memb ane (NMWL), he ansmemb ane p essu e (TMP), and ini ial lac ic acid concen a ion (CA) on he pe mea e lux (Jp) and ejec ions obse ed o ion lac a e (RA) and SDS (RS) using a ull cen al composi e expe imen al design and esponse su ace me hodology. Expe imen s we e conduc ed in ou s ages: a i s s age o lac ic acid ex ac ion wi h niosomes o mula ed wi h Span 80 (20 mol/m3) and SDS (4 mol/m3), a second back- ex ac ion s age conduc ed by NaOH addi ion un il pH > 12 o niosomes b eaking and sodium lac a e eleasing, and a hi d and ou h ul a il a ion s ages a 25 oC o sepa a e he lac a e ions om he mixed su ac an s. Memb ane NMWL, TMP and hei in e ac ions p esen ed s a is ically signi ican in luence on he pe mea e lux. Rejec ions o lac a e ion and SDS we e lowe han 4.5% and highe han 86%, espec i ely, whe eas Span 80 ejec ion was 100% in all ange o expe imen al condi ions es ed. The op imal condi ions we e es ablished o maximum alues o pe mea e lux, and hey we e ob ained o a 2 ba TMP and 15 kDa NMWL memb ane. Unde hese condi ions, he ejec ions o SDS su ac an and lac a e ion we e 87.3% and 4.31%, espec i ely, wi h a pe mea e lux o 42.63 L/m2h. The an agonis ic e ec be ween pe mea e lux and SDS ejec ion is also p o ed. Keywo ds Lac ic acid; Span 80; SDS; Ul a il a ion; Expe imen al design me hodology 1. In oduc ion Lac ic acid is o pa amoun impo ance in pha maceu ical and ood indus ies due o i s p ope ies as a p ese a i e, acidulan , pH egula o , and la o ing. I s use has conside ably inc eased in he las yea s because o he inc eased p oduc ion o polylac ic acid (PLA) biodeg adable he moplas ic [1–4]. In lac ic acid biop oduc ion, unsus ainable and high ene gy consump ion con en ional sepa a ion echniques, such as p ecipi a ion wi h calcium hyd oxide o 2 sol en ex ac ion, a e usually used o he lac ic acid sepa a ion om e men a ion b o hs [5,6]. Memb ane-based sepa a ion echniques [7] ha e p o en o be e ec i e because hey can a oid accumula ion o lac ic acid in he e men a ion b o hs, p e en ing p oduc inhibi ion and inc easing p oduc i i y o he e men a ion p ocess. In his way, hollow- ibe con ac o s using o ganic sol en s [8–10] and micella enhanced ul a il a ion (MEUF) using su ac an s ha e been s udied [11–14]. Mo e ecen ly, he use o niosomes as lac ic acid ex ac ion agen s has also been s udied [15]. Niosomes a e esicles o med by one o mo e bilaye s o non-ionic su ac an s enclosing an aqueous inside ca i y. Niosomes a e widely used in medical and pha macological applica ions o hei abili y o mic oencapsula e compounds o di e en na u e [16–20]. Howe e , he use o niosomes as ex ac ion agen s o solu es om e y low concen a ion aqueous solu ions is a new applica ion in he ield o sus ainable p ocesses ha has ba ely been explo ed. F aile e al. [21] obse ed ha he addi ion o sui able amoun s o he anionic su ac an sodium dodecyl sul a e (SDS) o non-ionic su ac an Span 80 (so bi an monoolea e) o mula ions yields a s abilizing e ec on he niosome bilaye , imp o ing lac ic acid en apmen e iciency. Howe e , he addi ion o ionic su ac an s o he niosome dispe sions can lead o he comple e solubiliza ion o esicles. The solubiliza ion p ocess o Span 80 niosomes by addi ion o SDS has been ecen ly s udied [22]. I was iden i ied as a h ee-s age micelliza ion p ocess: SDS adso p ion un il sa u a ion, in ensi ica ion o he bilaye solubiliza ion by mixed micelles o ma ion, and comple e bilaye solubiliza ion by micelliza ion. The c i ical poin s co esponding o SDS concen a ion o niosome sa u a ion and o al solubiliza ion we e iden i ied o se e al Span 80 niosome concen a ions, being 12 and 16 mol/m3 o SDS, espec i ely, o he 20 mol/m3 Span 80 o mula ion. The memb ane hyb id p ocess o lac ic acid ex ac ion by niosomes o mula ed wi h Span 80 and SDS in p e-sa u a ion concen a ions, using a 0.20 µm po e size la - disc TiO2 mic o il a ion memb ane and 0.3 ba o ansmemb ane p essu e, has been s udied in a p e ious wo k [15]. Bes esul s showed a 33% lac ic acid ex ac ion deg ee a e 30 min equilib ium ime, using niosomes o Span 80 (20 mol/m3) and SDS (4 mol/m3) as ex ac ion agen s, pH < pKa o lac ic acid (pKa = 3.4), and a SDS/lac ic acid mola a io o 0.01. Back-ex ac ion o lac a e ion was conduc ed by addi ion o NaOH un il pH > 12 whe e b eaking o niosomes was obse ed. Howe e , a signi ican pe mea e lux decline wi h espec o wa e lux (Jp/Jw = 0.38) was ob ained du ing he sepa a ion o componen s due o ouling by mixed micelles and SDS monome s in he pola iza ion laye and wi hin he la ge po es o he mic o il a ion memb ane. These esul s ha e led o he p esen wo k ocused on he use o ul a il a ion (UF) memb anes in o de o educe ouling and o imp o e he pe mea e lux du ing he back-ex ac ion s age a pH > 12. This wo k is a con inua ion o he p e ious one [15] and aims o model and op imize he emo al o lac a e ion om back-ex ac ion aqueous solu ions a pH > 12 con aining Span 80 and SDS su ac an s in he s a ed concen a ions, using ul a il a ion memb anes. A Cen al Composi e Design (CCD) and Response Su ace Me hodology (RSM) we e used o s udy he e ec o he ac o s (lac a e ion concen a ion, ansmemb ane p essu e, and memb ane nominal molecula weigh limi ), on he pe mea e lux and ejec ion o componen s. RSM app oach was also used o gain an unde s anding o he concen a ion pola iza ion phenomenon. The op imiza ion o he p ocess condi ions was conduc ed in o de o achie e maximum pe mea e lux and su ac an s ejec ion, and minimum lac a e ion ejec ion. 3 2. Ma e ials and me hods 2.1. Chemicals DL-Lac ic acid (>90% pu i y, Fluka) was used as solu e. The non-ionic su ac an so bi an monoolea e (Span 80, >95% pu i y, Sigma-Ald ich) and he anionic su ac an sodium dodecyl sul a e (SDS, 99%, Sigma-Ald ich) we e used in he o mula ion o niosomes. O he chemicals such as me hanol (HPLC g ade, HiPe Sol Ch omano m), maleic acid (>99%, Fluka), phospho ic acid (>85%, Sigma-Ald ich), disodium hyd ogen phospha e dodecahyd a e (>98%, Pan eac), po assium dihyd ogen phospha e (>99.5%, Me ck), sodium hyd oxide (analysis g ade, Scha lau), and phenolph halein (99%, Pan eac) we e used h oughou he expe imen s. Fo he de e mina ion o SDS concen a ion he ollowing chemicals we e used: e hyl iole (99%, Sigma-Ald ich), glacial ace ic acid o analysis quali y (Pan eac), sodium ace a e o analysis (Me ck), anhyd ous sodium sul a e o analysis (Scha lau), oluene (>99.5%, Anala No mapu VWR Chemicals) and e hylenediamine e aace ic acid (EDTA, >99%, Sigma-Ald ich). Ul apu e deionized Milli-Q wa e (Millipo e, USA), wi h a conduc i i y o 0.1 μS/cm, was used o he p epa a ion o all solu ions. 2.2. Niosome o ma ion Niosomes we e p epa ed by ul asonica ion o 10 cm3 aqueous solu ions o Span 80 (20 mol/m3) and SDS (4 mol/m3). These concen a ions we e chosen on he basis o p e ious wo ks [15,21]. The applica ion o ul asounds was ca ied ou o e a 5-min e ec i e ime, by pulses e e y 5 s (5 s on and 5 s o , 60 cycles; 30% ampli ude, 500 W), o a oid o e hea ing o he sample, using a high-in ensi y ul asonic p ocesso (Vib a-Cell VCX 500, Sonics & Ma e ials Inc., USA) equipped wi h a 3 mm-diame e i anium alloy bicylind ical p obe. Following, samples we e cen i uged (Eppendo 5804 cen i uge) o 15 min a 9000 pm, in o de o emo e any ace o me al de ached om he p obe. 2.3. Expe imen al p ocedu e UF expe imen s we e ca ied ou using a Spi lab il a ion cell (TAMI Indus ies, F ance) wi h 90 mm diame e la -disc ce amic memb anes (INSIDE DisRAM, TAMI Indus ies, F ance), made o a Z O2 ac i e laye suppo ed on TiO2, wi h 56.3 cm2 o e ec i e a ea. The nominal molecula weigh limi s (NMWL) o he memb anes we e 3, 8 and 15 kDa. All expe imen s we e conduc ed using he ollowing ou -s age p o ocol: 1. Ex ac ion s age: i was ca ied ou by mixing 400 cm3 o aqueous solu ion con aining lac ic acid (CA = 5, 10 and 15 mol/m3), named as Fo, and 10 cm3 o dispe sed phase con aining niosomes, named as Fd. The mix u e was con inuously s i ed a 375 pm and 20 oC o 30 min o each equilib ium. 2. Back-ex ac ion s age: his s age was pe o med by addi ion o a equi ed olume o NaOH (1 N) aqueous solu ion o he abo e men ioned dispe sion un il pH abou 12.2 ± 0.2. I was allowed 45-50 min o each equilib ium and hen a 60 cm3 sample was wi hd awn o analysis. The sample and emaining dispe sion we e iden i ied as Fbis. 3. UF s age in cons an concen a ion mode. The eed solu ion (Fbis) was ed o he ul a il a ion cell by a pe is al ic pump (Mas e lex l/s economy d i e Cole Pa me , CRS o o EW-07518-00) a a p e ixed low a e and p essu e. Pe mea e and e en a e s eams we e eci cula ed o he 1 L jacke ed eed ank, 4 whe e he eed solu ion was kep a cons an empe a u e (20 ºC) and s i ed a 375 pm. Adjus men o ansmemb ane p essu e (TMP) was achie ed by a needle al e loca ed in he e en a e s eam. The sys em is also equipped wi h a lowme e and a p essu e gauge, bo h placed a he inle o he il a ion cell. Expe imen s we e un o 30 min unde speci ic TMP (1, 1.5 o 2 ba ) in o de o achie e s able condi ions in he pola iza ion laye and memb ane. Subsequen ly, a 60 cm3 sample was wi hd awn o analysis and he sample and emaining dispe sion we e iden i ied as F. 4. UF s age in concen a ion mode. Once he equilib ium wi h he memb ane was eached, he eed solu ion F was ul a il a ed in concen a ion mode, emo ing con inuously he pe mea e s eam and eci cula ing he e en a e o he eed ank up o a olume concen a ion a io (VCR, he quo ien be ween ini ial eed olume and e en a e olume) a ound 2. The pe mea e lux was calcula ed by measu ing he ime needed o collec ing 10 cm3 pe mea e samples. Finally, pe mea e and e en a e we e sepa a ed o analysis and named as P and R, espec i ely. Fig. 1 shows a scheme o he ou -s age p ocedu e and he se -up o he UF expe imen al equipmen . Table 1 summa izes he analy ical measu emen s made o di e en samples h ough he expe imen al p ocess. EXTRACTION STAGE BACK-EXTRACTION STAGE F 0 F d F bis (pH > 12) NaOH (1) (2) (3) (4) F (5) (2) (3) (4) (5) R P (1) CONSTANT CONCENTRATION MODE UF STAGECONCENTRATION MODE UF STAGE Figu e 1. Schema ic diag am o he ou -s age expe imen al p ocedu e. Fd: dispe sed phase, Fo: con inuous phase, Fbis: dispe sion a pH > 12 wi hou memb ane con ac , F: eed dispe sion a pH > 12 in con ac wi h he memb ane unde UF condi ions a cons an concen a ion, P and R: pe mea e and e en a e a e UF in concen a ion mode, 1: eed ank, 2: pe is al ic pump, 3: p essu e gauge, 4: memb ane module, 5: needle al e. Memb ane cleaning was accomplished a e wa ds by insing wi h deionized wa e o emo e he oam, ollowed by washing wi h 0.1 N sodium hyd oxide solu ion o 30 min, and hen wi h 0.17 w .% phospho ic acid solu ion o 30 min. A inal insing s ep wi h deionized wa e un il neu ali y was su icien o es o e he ini ial wa e lux o he memb ane. 5 Table 1. Summa y o analy ical measu emen s made o samples h ough he expe imen al p ocedu e. Sample Desc ip ion Analy ical measu emen s Fo Lac ic acid aqueous solu ion (ini ial con inuous phase) Lac ic acid concen a ion and pH Fd Aqueous dispe sion o niosomes (ini ial dispe sed phase) Size, PDI, ze a po en ial, and pH Fbis Equilib ium dispe sion a pH > 12 (wi hou memb ane) Lac a e ion concen a ion, SDS monome s concen a ion, size, PDI, ze a po en ial, and pH F Bulk dispe sion a pH > 12 unde s eady- s a e UF condi ions in cons an concen a ion mode Lac a e ion concen a ion, SDS monome s concen a ion, size, PDI, ze a po en ial, and pH P Final pe mea e a e UF in concen a ion mode (VCR = 2) Lac a e ion concen a ion, SDS monome s concen a ion, size, PDI, ze a po en ial, and pH R Final e en a e a e UF in concen a ion mode (VCR = 2) Lac a e ion concen a ion, SDS monome s concen a ion, size, PDI, ze a po en ial, and pH 2.4. Analy ical me hods Lac a e ion concen a ion was de e mined by high pe o mance liquid ch oma og aphy using a HPLC Shimadzu. A e e se phase column ACE 5C18 (ACE HPLC columns) and a UV- is de ec o a 216 nm we e used. De ailed condi ions o he analy ical me hod can be ound elsewhe e [15]. Samples we e measu ed in iplica e and he analy ical e o was lowe han ± 0.001 mol/m3. SDS monome concen a ion was de e mined by spec opho ome y a 615 nm wi h a Hi achi U-2000 equipmen , using he e hyl iole me hod [23]. Samples we e measu ed in iplica e and he analy ical e o was lowe han ± 0.002 mol/m3. The pa icle size dis ibu ion, he mean hyd odynamic diame e and he polydispe si y index (PDI) o he samples we e ca ied ou by dynamic ligh sca e ing (DLS) using a Ze asize Nano ZS appa a us (Mal e n Ins umen s L d., UK). The PDI is a dimensionless measu e o he wid h o he size dis ibu ion anging om 0 o 1, a highe alue being indica i e o a b oade dis ibu ion o pa icle size. The a e age alue and he ela i e e o o he 3 eplica es, each o 5 measu emen s a 20 oC, was conside ed o each sample. Ze a po en ial measu emen s we e conduc ed wi h he a o emen ioned Ze asize Nano ZS appa a us, using he Lase Dopple Velocime y echnique. They we e pe o med on he same sample p e iously p epa ed o measu e he pa icle size, bu using he app op ia e DTS1061 disposable olded capilla y cell equipped wi h elec odes o allow he passage o elec ic cu en and he mo emen o he pa icles acco ding o hei cha ge [24]. Six eplica es o 11 measu emen s we e pe o med o each sample a 20 oC. The pH measu emen was pe o med a 20 oC using a C ison GLP 22 pH-me e i ed wi h a C ison 52-02 glass pH elec ode (C ison, Spain), wi h an e o o ± 0.01 pH uni s. 6 Mo phological analysis o niosomes was pe o med by nega i e s aining ansmission elec on mic oscopy (NS-TEM), using a JEOL-2000 EX-II TEM ope a ing a 160–180 kV, wi h an image esolu ion o 1 nm, loca ed a he Uni e si y o O iedo (Spain). A d ople o he selec ed sample was placed on a ca bon-coa ed coppe g id, and he sample excess was emo ed using a piece o il e pape . Then, a d op o phospho ungs ic acid solu ion (2% w/ ) was applied o he ca bon g id and le o 2 min. Once he excess o s aining agen was emo ed by abso bing wi h he il e pape , he sample was ai -d ied and he hin ilm o s ained niosomes was obse ed by TEM. 2.5. Expe imen al design and s a is ical analysis Response Su ace Me hodology (RSM) and Cen al Composi e Design (CCD) wi h h ee le els o each independen a iable we e used o s udy he e ec o NMWL (X1: 3–15 kDa), TMP (X2: 1–2 ba ) and lac ic acid ini ial concen a ion (X3: 5–15 mol/m3) on he pe mea e lux (Jp), lac a e ion obse able ejec ion (RA), and SDS obse able ejec ion (Rs). The ac o s and le els s udied a e summa ized in Table 2. Based on he selec ed high and low le els, he NMWL ideal cen al poin should be 9 kDa. Howe e , a 8 kDa memb ane was used a he cen al poin s in his s udy, assuming ha his change does no signi ican ly in luence he expe imen al design. Table 2. Fac o s and le els s udied. Fac o s Le els Low (-1) Cen e (0) High (+1) X1: NMWL (kDa) 3 8 15 X2: TMP (ba ) 1 1.5 2 X3: CA (mol/m 3 ) 5 10 15 The esponse a iables we e calcula ed using he ollowing equa ions: A V J p × = (1) ( ) ( ) Fbis i pi iC C 1R − = (2) whe e V is he olume o he pe mea e sample collec ed, is he ime needed o collec ing he pe mea e sample, A is he memb ane e ec i e a ea (56.3 cm2), and Ci(p) and Ci(Fbis) a e he o al concen a ion o lac a e ion o SDS in he inal pe mea e and dispe sion a pH > 12 (Fbis dispe sion), espec i ely. The CCD model gene a ed 17 expe imen al uns wi h h ee eplica es a he cen al poin which highligh he ep oducibili y o he expe imen s. A second-o de deg ee polynomial equa ion was used o exp ess each p edic ed esponse (Y) as a unc ion o he independen a iables unde s udy (X1, X2 and X3). The model equa ion is as ollows: 322331132112 2 333 2 222 2 1113322110 XXaXXaXXaXaXaXaXaXaXaaY +++++++++= (3) 7 whe e Y ep esen s he esponse a iable (Jp, RA, and RS, in his case), a0 is a cons an , and ai, aii, aij a e he linea , quad a ic, and in e ac i e coe icien s, espec i ely. Analysis o a iance (ANOVA) and leas signi ican di e ence (LSD) es we e applied o de ec he e ec o he ac o s and s a is ically signi ican di e ences among alues, espec i ely. The model was i ed by mul iple linea eg essions (MLR). The alidi y o he empi ical model was es ed wi h ANOVA. The signi icance o each es ima ed eg ession coe icien was assessed h ough alues o he s a is ic pa ame e s F and p (p obabili y) wi h a 95% con idence le el. The expe imen al design and da a analysis we e pe o med using STATGRAPHICS Cen u ion XVI (S a poin Technologies, Inc., Wa en on, VA, USA). Op imal condi ions we e de e mined wi h he help o he STATGRAPHICS Cen u ion XVI so wa e, in o de o each he maximum pe mea e lux and SDS ejec ion, and he minimum lac a e ion ejec ion, acco ding wi h he wo k objec i es. 3. Resul s and discussion 3.1. E ec o NaOH addi ion on he b eakup o niosomes Pa icle size measu emen in he dispe sed phase (Fd) e eals niosomes o 200 nm a e age diame e and a PDI alue o 0.27, which indica es a homogeneous popula ion (Fig. 2a). O he wise, de ailed analysis o DLS in ensi y da a o dispe sions a pH > 12 e ealed ha he main peak obse ed in Fd, which is a ibu ed o niosomes, disappea s in hese samples indica ing niosome des uc ion by he addi ion o NaOH un il pH > 12. Howe e , peaks associa ed wi h mixed micelles o 78–80 nm in size and Span 80 agg ega es la ge han 1000 nm we e obse ed in any o he Fbis, F and R dispe sions. Resul s co esponding o a R dispe sion a e also depic ed in Fig. 2a. They a e acco ding wi h p e ious wo ks [15,21]. As expec ed, ze a po en ial alues be ween –45 and –38.5 mV we e ob ained in he Fd samples used in di e en expe imen s (– 40.8 mV o Fd sample shown in Fig. 2b), which indica e he p esence o nega i ely cha ged niosomes due o he SDS adso bed in hei su ace. Besides, as shown in Fig. 2b, wo pa icle popula ions a e obse ed in he R dispe sion, wi h ze a po en ial alues o –8 and –20 mV, indica ing weakly nega i ely cha ged pa icles. Simila esul s we e ob ained o any o Fbis and F dispe sions (no shown), which co obo a e he b eakup o he niosomes a pH > 12. No pa icles we e ound in pe mea es, ega dless o he memb ane NMWL. The p esence and mo phology o niosomes in he dispe sed phase (Fd) ha e been con i med by TEM measu emen s. Fig. 3 shows wo pho omic og aphs o o mula ions o 20 mol/m3 o Span 80 and 4 mol/m3 o SDS, whe e he whi e a eas co espond o he g id. Fig. 3a shows he p esence o sphe ical niosomes o abou 200 nm in Fd, whose sizes ag ee wi h hose measu ed by DLS. Fig. 3b shows absence o niosomes in he dispe sion a pH > 12 (Fbis). Fig. 3b could co espond o la ge s uc u es o Span 80, as i s concen a ion (20 mol/m3) is well abo e i s CMC (≈ 0.1 mol/m3 in wa e [25]), and hey would be in acco dance wi h he la ge pa icles shown in Fig. 2a. 8 (a) (b) Figu e 2. (a) Pa icle size dis ibu ions co esponding o dispe sed phase, Fd, and e en a e, R, and (b) Ze a po en ial o Fd and R dispe sions, as desc ibed in Table 1. (a) (b) Figu e 3. TEM mic og aphs. (a) Niosomes o Span 80 (20 mol/m3) and SDS (4 mol/m3) in he dispe sed phase (Fd). (b) Fo mula ion o Span 80 (20 mol/m3) and SDS (4 mol/m3) in aqueous solu ion a pH > 12 (Fbis). Scale ba s: 0.2 µm. I is well documen ed ha addi ion o low concen a ion o ca ions o anionic su ac an (SDS) solu ions dec eases he epulsi e o ces be ween head g oups o SDS monome s due o he elec os a ic shielding e ec , esul ing in he o ma ion o micelles a lowe concen a ion han i s CMC (8.1 mol/m3 in wa e [11,26–32]). Howe e , beyond a c i ical concen a ion, he sodium ions s a dis up ing he micella packing, esul ing in less s able micelles [29,30]. In ligh o he esul s, i can be assumed ha he p esence o Span 80 monome s is highly imp obable in dispe sions a pH > 12 due o i s hyd ophobic cha ac e (HLB = 4.3 [20]), whe eas he coexis ence o la ge Span 80 agg ega es wi h mixed micelles and SDS su ac an monome s is highly p obably. 9 3.2. Expe imen al design The ma ix o he CCD and expe imen al alues o he esponse a iables a e gi en in Table 3. Table 3. Ma ix o he cen al composi e design (CCD) and expe imen al alues o he esponse a iables: pe mea e lux (Jp), lac a e ion obse able ejec ion (RA), and SDS obse able ejec ion (RS). Expe imen Fac o s Responses NMWL (kDa) TMP (ba ) C A (mol/m 3 ) J p (L/m 2 h) RA RS 1 15 1 5 15,99 abc 0.0191 a 0,897 a 2 8 1,5 10 24,51 cd 0.0182 a 0,864 a 3 3 1 15 8,53a 0.0115a 0,888a 4 8 1,5 5 26,64 cd 0.0160 a 0,865 a 5 3 1,5 10 12,79 ab 0.0056 a 0,881 a 6 15 1,5 10 34,10 de 0.0315 a 0,880 a 7 8 1,5 15 19,18cd 0.0171a 0,891a 8 8 1 10 14,92 abc 0.0117 a 0,881 a 9 8 1,5 10 22,38 cd 0.0208 a 0,866 a 10 15 1 15 17,05 abcd 0.0193 a 0,887 a 11 3 1 5 9,06a 0.0025a 0,902a 12 3 2 5 20,25 bcd 0.0313 a 0,889 a 13 15 2 5 42,63 0.0431 a 0,873 a 14 15 2 15 40,50 0.0448 a 0,870 a 15 8 1,5 10 23,45cd 0.0077a 0,871a 16 8 2 10 27,71 de 0.0055 a 0,869 a 17 3 2 15 17,05 abcd 0.0313 a 0,890 a Values wi h di e en le e s in each column a e signi ican ly di e en (LSD es , p < 0.05) Table 3 shows ha Jp alues p esen la ge a ia ion, be ween 8 and 43 L/m2h. Howe e , RA and RS alues we e lowe han 4.5% and highe han 86%, espec i ely, wi h e y simila alues among hem o all expe imen s. LSD es was applied o each esponse a iable, e ealing ha RA and RS alues we e no signi ican ly di e en (p > 0.05). Howe e , Jp alues iden i ied in Table 3 ha do no sha e a same le e we e conside ed s a is ically di e en among hem (p < 0.05). ANOVA o he i ed model o he Jp esponse shows ha he model was s a is ically signi ican (p- alue = 0.001). Table 4 shows ha NMWL, TMP and hei in e ac ion a e s a is ically signi ican (p < 0.05) on Jp. F alues indica e ha , o he ange o ac o s s udied, NMWL and TMP ac o s ha e he s onge in luence on Jp, and also ha in e ac ion be ween NMWL and TMP has syne gis ic e ec on Jp. O he wise, CA ac o was no signi ican on Jp, indica ing ha i does no con ibu e on he Jp esponse, in he ange o concen a ions es ed. 16 Figu e 10. Rela ionships be ween he pe mea e lux and TMP o di e en medium composi ions. Symbols: expe imen al da a. Lines: beha io ends. 4. Conclusions Z O2 ce amic ul a il a ion memb anes can be success ully used o he sepa a ion o ion lac a e om he su ac an s Span 80 and SDS in aqueous solu ions a pH > 12. P ocess op imiza ion by RSM showed ha , in he ange o condi ions s udied (TMP: 1– 2 ba , NMWL: 3–15 kDa, and CA: 5–15 mol/m3), bes esul s we e ob ained o he 15 kDa memb ane and a ansmemb ane p essu e o 2 ba . Unde hese condi ions he pe mea e lux (Jp) was 42.63 L/m2h and SDS ejec ion (Rs) was 87.3%. Ion lac a e concen a ion e ec was no s a is ically signi ican on Jp and i s ejec ion was lowe han 4.5%. Span 80 ejec ion was 100% in all ange o expe imen al condi ions es ed, as i o ms la ge agg ega es ha a e e ained by memb anes. Al hough he di e ences be ween Rs alues we e no s a is ically signi ican wi h 95% signi icance le el, an an agonis ic beha io be ween Rs and Jp has been expe imen ally es ed. I was obse ed ha UF p ocess was mainly con olled by con ec ion and Jp inc eased as TMP inc eases, being his e ec mo e app eciable o memb anes wi h highe NMWL. Besides, as Jp inc eases pa o SDS molecules adso bed on he memb ane o accumula ed in he pola iza ion laye pass h ough he memb ane dec easing he SDS monome s ejec ion. The e en ion o su ac an s a pH > 12 a e in luenced by h ee p edominan e ec s: he memb ane sie ing ha yields e en ion o mixed micelles and la ge agg ega es o Span 80, he de-compac ion o he pola iza ion laye due o he micelles des abiliza ion caused by he excess o Na+ ions, and he shielding o he nega i ely cha ged memb ane su ace by Na+ ions which imp o es he pe mea ion o SDS monome s as Jp inc eases. This s udy complemen s a p e iously one pe o med wi h a 0.20 µm TiO2 mic o il a ion memb ane and 0.3 ba TMP [15] whe e a signi ican pe mea e lux decline was ob ained du ing he sepa a ion o componen s due o SDS monome s 0 10 20 30 40 50 60 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 2.1 Jp (L/m2 h) TMP (ba ) Span 80, pH = 12.4 Span 80 + SDS, pH = 7 Span 80 + SDS, pH = 12.5 SDS, pH = 6.9 SDS, pH = 12.6 17 accumula ed in he pola iza ion laye and adso bed wi hin he la ge po es o he mic o il a ion memb ane. A compa ison be ween bo h s udies shows ha Jp ob ained wi h he 15 kDa ul a il a ion memb ane (Jp = 42.63 L/m2h a TMP = 2 ba ) was highe han he ob ained wi h he 0.20 µm mic o il a ion memb ane (Jp = 19.19 L/m2h a TMP = 0.3 ba ). Besides, smalle dec ease o Jp wi h ela ion o pu e wa e lux (Jw) was ob ained wi h he UF memb ane (Jp/Jw = 0.61) han wi h he mic o il a ion one (Jp/Jw = 0.38). These esul s indica e an imp o emen in he ex ac ion-backex ac ion p ocess o lac ic acid wi h Span 80 and SDS niosomes using ul a il a ion memb anes. Acknowledgmen s Financial suppo om he Minis e io de Economía y Compe i i idad (MINECO, Spain) h ough p ojec CTQ2011-25239, and om Jun a de Cas illa y León h ough p ojec BU055U16 co inanced by he Eu opean Regional De elopmen Fund (ERDF- FEDER) is g a e ully acknowledged. The au ho s would like o hank D . Ca los Ál a ez (Scien i ic Technical Se ices, Uni e si y o O iedo, Spain) o his aluable help and assis ance wi h TEM measu emen s. Re e ences [1] M.A. Abdel-Rahman, Y. Tashi o, K. Sonomo o, Lac ic acid p oduc ion om lignocellulose-de i ed suga s using lac ic acid bac e ia: o e iew and limi s, J. Bio echnol. 156 (2011) 286–301. [2] R. Da a, S.-P. Tsai, P. Bonsigno e, S.-H. Moon, J.R. F ank, Technological and economic po en ial o poly(lac ic acid) and lac ic acid de i a i es, FEMS Mic obiol. Re . 16 (1995) 221–231. [3] K.M. Nampoo hi i, N.R. Nai , R.P. John, An o e iew o he ecen de elopmen s in polylac ide (PLA) esea ch, Bio esou . Technol. 101 (2010) 8493–8501. [4] S. Taskila, H. Ojamo, The cu en s a us and u u e expec a ions in indus ial p oduc ion o lac ic acid by lac ic acid bac e ia, in: J. Ma celino Kongo (Ed.), Lac ic Acid Bac e ia – R& D o Food, Heal h and Li es ock Pu poses. InTech, 2013. doi: h p://dx.doi.o g/10.5772/51282. [5] K.L. Wasewa , A.A. Yawalka , J.A. Moulijn, V.G. Panga ka , Fe men a ion o glucose o lac ic acid coupled wi h eac i e ex ac ion: a e iew, Ind. Eng. Chem. Res. 43 (2004) 5969–5982. [6] D. Yanko , J. Molinie , J. Albe , G. Malma y, G. Kyuchouko , Lac ic acid ex ac ion om aqueous solu ions wi h i-n-oc ylamine dissol ed in decanol and dodecane, Biochem. Eng. J. 21 (2004) 63–71. [7] P. Pal, J. Sikde , S. Roy, L. Gio no, P ocess in ensi ica ion in lac ic acid p oduc ion: a e iew o memb ane based p ocesses, Chem. Eng. P ocess. 48 (2009) 1549–1559. [8] H. Huang, S.T. Yang, D.E. Ramey, A hollow- ibe memb ane ex ac ion p ocess o eco e y and sepa a ion o lac ic acid om aqueous solu ion, Appl. Biochem. Bio echnol. 114 (2004) 671–688. [9] R.-S. Juang, J.-D. Chen, H.-C. Huan, Dispe sion- ee memb ane ex ac ion: case s udies o me al ion and o ganic acid ex ac ion, J. Memb . Sci. 165 (2000) 59– 73. [10] M. Rod íguez, M.J. González-Muñoz, S. Luque, J.R. Ál a ez, J. Coca, Ex ac i e ul a il a ion o he emo al o ca boxylic acids, J. Memb . Sci. 274 (2006) 209– 218. [11] R.M. Gean a, M.O. Ruiz, I. Escude o, Micella -enhanced ul a il a ion o he eco e y o lac ic acid and ci ic acid om bee molasses wi h sodium dodecyl sulpha e, J. Memb . Sci. 430 (2013) 11–23. 18 [12] J. Landabu u-Agui e, E. Pong ácz, R.L. Keiski, Sepa a ion o cadmium and coppe om phospho ous ich syn he ic wa e s by micella -enhanced ul a il a ion, Sep. Pu i . Technol. 81 (2011) 41–48. [13] J. Landabu u-Agui e, E. Pong ácz, A. Sa pola, R.L. Keiski, Simul aneous emo al o hea y me als om phospho ous ich eal was ewa e s by micella - enhanced ul a il a ion, Sep. Pu i . Technol. 88 (2012) 130–137. [14] M.O. Ruiz, J.M. Beni o, B. Ba iuso, J.L. Cabezas, I. Escude o, Equilib ium dis ibu ion model o be aine be ween su ac an micelles and wa e : applica ion o a micella -enhanced ul a il a ion p ocess, Ind. Eng. Chem. Res. 49 (2010) 6578–6586. [15] L. Roque, I. Escude o, J.M. Beni o, Lac ic acid eco e y by mic o il a ion using niosomes as ex ac ion agen s, Sep. Pu i . Technol. 151 (2015) 1–13. [16] E. Acos a, Bioa ailabili y o nanopa icles in nu ien and nu aceu ical deli e y, Cu . Op. Colloid In e ace Sci. 14 (2009) 3–15. [17] Y.-M. Hao, K. Li, En apmen and elease di e ence esul ing om hyd ogen bonding in e ac ions in niosome, In . J. Pha m. 403 (2011) 245–253. [18] C. Ma ianecci, L. Di Ma zio, F. Rinaldi, C. Celia, D. Paolino, F. Alhaique, S. Esposi o, M. Ca a a, Niosomes om 80s o p esen : he s a e o he a , Ad . Colloid In e ace Sci. 205 (2014) 187–206. [19] D.J. McClemen s, E.A. Decke , Y. Pa k, J. Weiss, S uc u al design p inciples o deli e y o bioac i e componen s in nu aceu icals and unc ional oods, C i . Re . Food Sci. Nu . 49 (2009) 577–606. [20] A.Y. Waddad, S. Abbad, F. Yu, W.L.L. Munyendo, J. Wang, H. L , J. Zhou, Fo mula ion, cha ac e iza ion and pha macokine ics o Mo in hyd a e niosomes p epa ed wi h a ious non-ionic su ac an s, In . J. Pha m. 456 (2013) 446–458. [21] R. F aile, R.M. Gean a, I. Escude o, J.M. Beni o, M.O. Ruiz, Fo mula ion o Span 80 niosomes modi ied wi h SDS o lac ic acid en apmen , Desalin. Wa e T ea . 56 (2015) 3463–3475. [22] L. Alonso, L. Roque, I. Escude o, J.M. Beni o, M.T. Sanz, S. Bel án, Solubiliza ion o Span 80 niosomes by sodium dodecyl sul a e, ACS Sus ainable Chem. Eng. 4 (2016), 1862–1869. [23] B.A. Uzoukwu, L.M.L. Nolle , Analysis o su ac an s, in: L.M.L. Nolle (Ed.), Handbook o Wa e Analysis, Ma cel Dekke , New Yo k., 2000, pp. 767–784. [24] M. Kaszuba, J. Co be , F.M. Wa son, A. Jones, High-concen a ion ze a po en ial measu emen s using ligh -sca e ing echniques, Philos. T ans. R. Soc. A-Ma h. Phys. Eng. Sci. 368 (2010) 4439–4451. [25] J. Wei, Q. Xue, E ec s o su ac an s on he ibological p ope ies o a C 2O3 coa ing, Wea 162–164 (1993) 229–233. [26] E. Sampe , M. Rod íguez, M.A. De la Rubia, D. P a s, Remo al o me al ions a low concen a ion by micella -enhanced ul a il a ion (MEUF) using sodium dodecyl sul a e (SDS) and linea alkylbenzene sul ona e (LAS), Sep. Pu i . Technol. 65 (2009) 337–342. [27] K. Xu, G.-M. Zeng, J.-H. Huang, J.-Y. Wu, Y.-Y. Fang, G. Huang, J. Li, B. Xi, H. Liu, Remo al o Cd2+ om syn he ic was ewa e using micella -enhanced ul a il a ion wi h hollow ibe memb ane, Colloid Su . A-Physicochem, Eng. Aspec s 294 (2007) 140–146. [28] J.-S. Yang, K. Baek, J.-W. Yang, C oss low ul a il a ion o su ac an solu ions, Desalina ion 184 (2005) 385–394. [29] I. Kowalska, M. Kabsch-Ko bu owicz, K. Majewska-Nowak, T. Winnicki, Sepa a ion o anionic su ac an s on ul a il a ion memb anes, Desalina ion 162 (2004) 33–40. [30] A. Pa is , P.D.T. Huibe s, B. Deneka, D.O. Shah, E ec o e aalkylammonium chlo ides on oaming p ope ies o sodium dodecyl sul a e solu ions, Langmui 14 (1998) 4471–4474. 19 [31] K. Majewska-Nowak, I. Kowalska, M. Kabsch-Ko bu owicz, Ul a il a ion o SDS solu ions using polyme ic memb anes, Desalina ion 184 (2005) 415–422. [32] E. Fe nández, J.M. Beni o, C. Pazos, J. Coca, Ce amic memb ane ul a il a ion o anionic and nonionic su ac an solu ions, J. Memb . Sci. 246 (2005) 1–6. [33] L. Suá ez, M.A. Diez, F.A. Rie a, T anspo mechanisms o de e gen ing edien s h ough ul a il a ion memb anes, Sep. Pu i . Technol. 136 (2014) 115–122. [34] F.H. Quina, P.M. Nassa , J.B.S. Bonilha, B.L. Bales, G ow h o sodium dodecyl sul a e micelles wi h de e gen concen a ion, J. Phys. Chem. 99 (1995) 17028– 17031. [35] M. Sammalko pi, M. Ka unen, M. Haa aja, Ionic su ac an agg ega es in saline solu ions: sodium dodecyl sul a e (SDS) in he p esence o excess sodium chlo ide (NaCl) o calcium chlo ide (CaCl2), J. Phys. Chem. B 113 (2009) 5863– 5870. [36] C.S. Gangabadage, A. Najda, D. Bogdan, S.S. Wijmenga, M. Tessa i, Dependence o he size o a p o ein-SDS complex on de e gen and Na+ concen a ions, J. Phys. Chem. B. 112 (2008) 4242–4245. [37] A.-S. Jönsson, B. Jönsson, The in luence o nonionic and ionic su ac an s on hyd ophobic and hyd ophilic ul a il a ion memb anes, J. Memb . Sci. 56 (1991) 49–76. [38] L. Villa aña-López, M. Á ila-Rod íguez, M.P. González-Muñoz, S udy o he ze a po en ial and s eaming cu en o ul a il a ion memb anes in con ac wi h an anionic su ac an , Desalin. Wa e T ea . 56 (2015), 3456–3462. [39] I. Kowalska, Su ac an emo al om wa e solu ions by means o ul a il a ion and ion-exchange, Desalina ion 221 (2008) 351–357. [40] I. Kowalska, K. Majewska-Nowak, M. Kabsch-Ko bu owicz, In luence o empe a u e on anionic su ace ac i e agen emo al om wa e solu ion by ul a il a ion, Desalina ion 198 (2006) 124–131.