Bioresource Technology 334 (2021) 125233 Contents lists available at ScienceDirect Bioresource Technology journal homepage: www.elsevier.com/locate/biortech Second- and third-generation biodiesel production with immobilised recombinant Rhizopus oryzae lipase: Influence of the support, substrate acidity and bioprocess scale-up Josu López-Fernández , Maria Dolors Benaiges , Francisco Valero * Department of Chemical, Biological and Environmental Engineering, School of Engineering, Universitat Autònoma de Barcelona, Bellaterra, 08193 Barcelona, Spain H I G H L I G H T S G R A P H I C A L A B S T R A C T • Rhizopus oryzae lipase proved an excel­ lent biocatalyst for biodiesel production. • Hydrocarbon chains on the support increased the biocatalyst operational stability. • Low oil acidity negatively influenced biocatalyst operational stability. • Acyl-acceptor concentration and oil acidity had a synergistic effect. • Biodiesel production from waste cook­ ing oil was successfully scaled up. A R T I C L E I N F O A B S T R A C T Keywords: Rhizopus oryzae lipase immobilised onto differently functionalised polymethacrylate (Purolite®) and magnetite Biodiesel superparamagnetic supports was assessed as a catalyst for biodiesel production with pomace oil. The presence of Biocatalysis surface hydrocarbon chains increased the operational stability of the biocatalysts supported on Purolite® and Rhizopus oryzae lipase superparamagnetic particles up to 9 and 2 times, respectively. By contrast, the presence of functional groups had Immobilisation no effect on the initial transesterification rate, which was twice higher with the lipase immobilised onto Pur­ Substrate acidity olite®. Also, functionalising Purolite® with epoxide and octadecyl groups led to the highest biodiesel and volumetric productivity. This biocatalyst with other substrates including makauba, jatropha, waste cooking oil, and microbial oil, led to similar initial reaction rates. However, simply raising substrate acidity from 0.5 to 2% increased the operational stability of the biocatalysts 15 times. A synergistic effect between acyl-acceptor con­ centration and substrate acidity was observed. The transesterification reaction was successfully scaled up to 50 mL. 1. Introduction paradigm to enable a climate-neutral society and avoid environmental collapse. In this scenario, a wide variety of clean energy sources have the Global warming is an unavoidable process that requires a change of potential to jointly replace polluting fossil fuels (Connolly et al., 2016). * Corresponding author. E-mail address: francisco.valero@uab.cat (F. Valero). https://doi.org/10.1016/j.biortech.2021.125233 Received 19 March 2021; Received in revised form 23 April 2021; Accepted 24 April 2021 Available online 2 May 2021 0960-8524/© 2021 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). J. López-Fernández et al. Bioresource Technology 334 (2021) 125233 One such source is biodiesel, which is likely to play a major role in the esterases and its ability to grow at high cell densities (ca. 100 g L− 1 dry process by virtue of its substantial advantages; thus, this clean fuel is cell weight) in defined media (López-Fernández et al., 2020). rROL is a biodegradable, renewable, non-toxic, usable by existing engines, locally 1,3-regioespecific lipase which was used for enzymatic second- and produced and carbon–neutral —carbon in biodiesel exhaust is recently third-generation biodiesel production. Its transesterification gives the fixed from the atmosphere (Ranganathan et al., 2008; Rodrigues et al., corresponding mono-alkyl esters and, provided acyl-migration is 2016). Biodiesel is a mixture of mono-alkyl esters of long-chain fatty controlled (Canet et al., 2017), 2-monoacylglyceride as well, thereby acids obtained by transesterification of a wide range of oily substrates avoiding glycerol formation and giving a product with a high added and classified as first-, second- or third-generation biodiesel —the latter value for the cosmetics and food industries (Luo et al., 2016). The lipase two are designated “advanced biodiesel” in European Directive 2015/ used here was covalently immobilised onto three glutaraldehyde-treated 1513. First-generation biodiesel is produced from edible oils, con­ polymethacrylate-based supports containing both epoxide and hydro­ sumption of which poses severe ethical problems. On the other hand, phobic functional groups. The effects of functional groups on lipase second- and third-generation biodiesel are obtained from non-edible oils immobilisation and diffusional restrictions, and also on the initial re­ (e.g., waste cooking oil, WCO), oils from plants growing in agriculturally action rate and operational stability of the catalyst, were assessed. unsuitable land (e.g., jatropha, makauba) (Hama and Kondo, 2013) or Glutaraldehyde-treated superparamagnetic nanoparticles containing microbial oils respectively (Ma et al., 2018; Navarro López et al., 2016). hydrophobic functional groups at variable densities were also used for Global biodiesel production is primarily focused on first-generation covalent immobilisation of lipase and evaluation of biodiesel produc­ biodiesel, the demand for which is expected to grow to 110.6 million tion. Tests were conducted by using alperujo oil (non-edible olive barrels per day by 2035 (Ogunkunle and Ahmed, 2019) and hence to pomace oil with a high FFA content obtained in the olive extraction strengthen the ethical dilemma on this type of biofuel. In response, process; Bonet-Ragel et al., 2018), as a model substrate for trans­ public institutions are taking steps to shift production from first- to esterification. Subsequently, the best biocatalyst was used for produc­ second- and third-generation biodiesel. For instance, the European tion of second- and third-generation biodiesel from alternative Union (EU/2015/1513) has amended its previous policies in this regard substrates of industrial use, including jatropha, makauba, WCO, and by aiming at decreased first-generation biofuel consumption and microbial oils. Finally, transesterification with WCO was scaled-up to a increased production of second- and third-generation biodiesel. Even so, 50 mL mini-laboratory reactor. most existing biodiesel production plants have been designed for first- generation biodiesel production, which is based on chemical trans­ 2. Materials and methods esterification in the presence of homogeneous basic catalysts. Under these conditions, the high content in free fatty acids (FFA) of second- and 2.1. Materials third-generation substrates facilitates soap formation unless they are previously neutralised. Alternative solutions involving trans­ Olive pomace oil was a gift from Professor Eulogio Castro (University esterification with acids and heterogeneous solid catalysts have been of Jaen, Spain); microbial oil from a modified strain of Rhodosporidium proposed; most, however, are uneconomical, require more severe reac­ toruloides was supplied by Neol Biosolutions (Granada, Spain); jatropha tion conditions (e.g., higher alcohol-to-oil mole ratios) or result in lower and makauba oils were kindly donated by Professor Denise Freire of the reaction rates (Bonet-Ragel et al., 2018; Hama and Kondo, 2013; Ran­ Universidade Federal do Rio de Janeiro (Brazil); and WCO was obtained ganathan et al., 2008). Enzymatic transesterification with lipases from a local public waste management company. All oils were centri­ (glycerol ester hydrolases, E.C.3.1.1.3) has emerged as an effective fuged prior to use. Polymethacrylate matrix supports D6307, D6308 and choice for biodiesel production without the previous shortcomings. In D6309 were kindly supplied by Purolite® (King of Prussia, PA, USA). fact, lipases enable transesterification reactions with substrates con­ The colorimetric kit for enzymatic assay 11821729 was obtained from taining high FFA concentrations under milder conditions, might avoid Roche (Mannheim, Deutschland) and bovine serum albumin standards glycerol formation —an unwanted outcome for the biodiesel industry (Ref. 11811345) were supplied by Thermo Fisher Scientific (Waltham, (Rodrigues et al., 2017) — and the need for solvents. Enzyme-based MA, USA). Heptane, ethanol and methanol were purchased from Pan­ transesterification uses less energy and involves fewer downstream reac (Barcelona, Spain). n-Butylamine solution, 3-aminopropyltriethox­ steps (He et al., 2018; Xu et al., 2017); however, lipases for use as bio­ ysilane (APTS), ammonium sulphate, NaBH4, FeCl2, FeCl3, standards of catalysts by the biodiesel industry are expensive and unstable (especially methyl/ethyl palmitate, methyl/ethyl stearate, methyl/ethyl oleate, in the presence of alcohols such as methanol, which diminishes enzyme methyl/ethyl linoleate, methyl linoleate and all unstated reagents were reusability and compromises economic feasibility) (López-Fernández obtained from Sigma–Aldrich (St. Louis, MO, USA). et al., 2019). Immobilising enzymes provides an effective way of circumventing 2.2. Lipase heterologous production the shortcomings of biocatalysts by improving their stability against deactivation and enabling their reuse (Ariaeenejad et al., 2021; Mad­ Recombinant Rhizopus oryzae lipase (rROL) formed by its mature havan et al., 2017). A great variety of lipase immobilisation methods sequence was produced by the Bioprocess Engineering and Applied have been reported including a number for biodiesel production. For Biocatalysis Group of the Universitat Autònoma de Barcelona (Barce­ instance, hydrophobic adsorption is broadly used on the grounds of the lona, Spain) as described elsewhere (Arnau et al., 2010). Production hydrophobicity of the active sites of lipases and their widely docu­ runs were followed by centrifugation, microfiltration, ultrafiltration and mented catalytic hyperactivation (Canet et al., 2016). Other strategies lyophilisation of the culture broth to remove biomass and concentrate such as covalent immobilisation (Bonet-Ragel et al., 2015; Da Rós et al., the enzyme (Guillén et al., 2012). 2010; Rodrigues et al., 2016), entrapment (Noureddini et al., 2005) and the use of cross-linked enzyme aggregates (CLEA) have also been used in search for highly efficient biocatalysts (Khanahmadi et al., 2016). Also, 2.3. Lipase activity measurement and protein determination the previous approaches have been adapted for use with supports based on superparamagnetic nanoparticles (Cruz-Izquierdo et al., 2014) or Lipase activity was determined on a Cary 300 spectrophotometer with novel lipase-displaying whole-cell biocatalysts (Jin et al., 2013). from Varian (Mulgrave, VIC, Australia), using the Roche lipase colori­ In this work, mature sequence of Rhizopus oryzae lipase (rROL) was metric kit in 200 mM Tris-HCl buffer at pH 7.25 at 30 ◦ C. Measurements expressed in the methylotrophic yeast Komagataella phaffii, which has were made in triplicate at 580 nm (Resina et al., 2004). been claimed as the most suitable cell factory for rROL production due to Protein concentrations were determined by using the Bradford its secretion capacity, its lack of endogenous extracellular lipases or method with bovine serum albumin as standard (Bradford, 1976). 2 J. López-Fernández et al. Bioresource Technology 334 (2021) 125233 2.4. Synthesis of magnetic nanoparticles (MNPs) and functionalisation Germany) and under orbital stirring at 350 rpm. Ethanol or methanol with amino groups was added in 1, 5 or 10 pulses by splitting the stoichiometric volume of alcohol (2:1 alcohol/oil) into 1, 5 or 10 portions, respectively, that were Nanoparticles of magnetite (Fe3O4) were synthesised and function­ added during the reaction to reach the theoretical maximum yield alised with amino groups (MNPs-NH2) as described elsewhere (Cruz- (Bonet-Ragel et al., 2018). The reaction was scaled up to a 50 mL Izquierdo et al., 2014). All solutions used for MNPs synthesis and Scharlau HME-R mini-reactor from Scharlab (Sentmenat, Barcelona, functionalisation were prepared in nitrogen-bubbled ultrapure water. Spain) with mechanical stirring. The above-described procedures were The dry weight and concentration of MNPs were determined by using a performed with a final WCO mass of 40 g. VR-1/120/240 vacuum concentrator from Heto Lab Equipment (Zea­ Olive pomace oil was used as model to investigate diffusional re­ land, Denmark). strictions, the initial reaction rate and the operational stability of the biocatalysts. The best biocatalyst was used for further evaluation of the 2.5. Functionalisation of supports with aldehyde groups initial reaction rate and operational stability with alternative substrates including jatropha oil, WCO, makauba oil and microbial oil. All re­ Two types of immobilisation supports based on polymethacrylate actions components were pre-equilibrated for water activity, using and magnetite were used. The polymethacrylate-based supports saturated KOH (aw = 0.093) overnight (a minimum of 16 h) (Bonet- included Purolite® D6307 with epoxide and butyl functional groups Ragel et al., 2018). (EB); Purolite® D6308 with epoxide and octadecyl groups (EO) and The initial reaction rate of all evaluated biocatalysts was calculated Purolite® D6309 with epoxide and divinylbenzene groups (EDVB). All by adding 0.16 mL of methanol to 8 g of pomace oil (viz., the stoi­ were conditioned as described elsewhere (Bonet-Ragel et al., 2015). chiometric amount needed to obtain a yield of ca. 14%) and a total Epoxide groups were converted into aldehyde groups with ethylenedi­ amount of biocatalyst of 32 000 AU (Bonet-Ragel et al., 2015). An amine first and glutaraldehyde then. identical procedure was followed to assess diffusional restrictions but Previously functionalised MNPs-NH2 were simply pretreated with using an identical mass of biocatalyst (200 mg) with variable immobi­ glutaraldehyde following the protocol described elsewhere (Bonet-Ragel lised lipase activity. et al., 2015). The resulting aldehyde-functionalised MNPs (MNP-CHO) Operational stability was assessed in duplicate tests by causing the were further modified by incubation in a 0–1 M butylamine solution in biocatalyst to deposit in the vial bottom by decantation (Purolite®) or 100 mM phosphate buffer pH 8 at room temperature for 2 h to obtain application of a magnetic field (magnetite) and removing the medium MNP-But-CHO. MNPs were washed three times with PBS after each step after each reaction run (Bonet-Ragel et al., 2018). and stored at 4 ◦ C until use. The optimum butylamine concentration for MNP-CHO functionali­ sation was established by assessing the operational stability of the 2.6. Lipase immobilisation onto functionalised supports modified biocatalysts (MNP-But-CHO). Test were conducted as described above but in triplicate, using a scaled-down reaction volume Lipase was immobilised onto the above-described polymethacrylate- (1.5 mL vials) and only 1 pulse of ethanol. and magnetite-based supports by using a modified version of a previ­ ously reported method (Bonet-Ragel et al., 2015). Unless otherwise 2.9. Diffusional restrictions stated, 1 g of glutaraldehyde-treated support (dry weight for MNPs) was mixed with a 3500 AU lipase mL− 1 solution at 4 ◦ C for 42 h. The poly­ The Weisz–Prater criterion, which is a dimensionless number used to methacrylate biocatalysts thus obtained (EB-rROL, EO-rROL, EDVB- assess internal diffusional restrictions (Harvey W. Blanch, 1997; Weisz rROL) were vacuum-filtered and dried on silica gel, whereas the and Prater, 1954) was calculated by substituting experimental data from magnetite-based biocatalysts were recovered by application of a mag­ the biodiesel reactions into Eq. (2): netic field, washed three times with PBS and concentrated to 2 mg mL− 1. robs ρp ( )2 Vp Schiff bases and unreacted aldehyde groups were reduced by incu­ Φ= (2) Deff Cm,0 Ap bating the supports with a 1 mg mL− 1 NaBH4 solution in 100 mM phosphate buffer at pH 8 at room temperature for 2 h (Cruz-Izquierdo Where robs denotes transesterification rate (mol g–1 − 1 particle s ), ρp particle et al., 2014). density (g cm ), VP particle volume (cm ), Ap particle area (cm2), Cm,0 − 3 3 The specific activity of the biocatalysts (immobilised AU mg bulk concentration of methanol (mol cm− 3) and Deff (cm2 s− 1) the support− 1) was calculated as the difference between those of the blank effective diffusivity coefficient as calculated from Eq. (3): and supernatant solutions divided by the final weight (MNP dry weight) of biocatalyst. The immobilisation yield (IY) for the polymethacrylate Deff = Dm,a εp σ (3) supports was estimated by exposing them to lipase solutions at con­ τ centrations from 55 to 1000 AU mg support− 1: Dm,a (cm2 s− 1) being the molecular diffusivity of methanol in the Biocatalyst specific activity reaction medium (pomace oil) as estimated from the Nakanishi corre­ IY (%) = x100 (1) lation (Poling et al., 2001), εp particle porosity, σ the constriction factor offered activity units and τ particle tortuosity. Φ values under 0.3 ensure the absence of in­ ternal diffusional restrictions since the resulting effectiveness factor is 2.7. Quantification of fatty acid methyl and ethyl esters close to unity (Harvey W. Blanch, 1997). Fatty acid methyl esters (FAMEs) and fatty acid ethyl esters (FAEEs) 2.10. Substrate acidity were quantified on a model 7890A gas chromatograph from Agilent (Santa Clara, CA, USA) equipped with a 19095 N-123 capillary column Total acidity was determined in accordance with the protocols in and an autosampler (Canet et al., 2014). The relative standard de­ Commission Regulation (EEC) No 2568/91 Annex II, amended by viations (RSD) for the FAMEs and FAEEs never exceeded 3%. Commission Regulation (EC) No 702/2007. 2.8. Transesterification reactions Reaction runs were conducted in 10 mL closed vials containing 8 g of oil at 30 ◦ C that were placed in a KS 400 incubator from IKA (Staufen, 3 J. López-Fernández et al. Bioresource Technology 334 (2021) 125233 Fig. 1. (A) Biocatalyst activity (black symbols, ) and immobilisation yield (white symbols, ) at a variable offered activity. (B) Initial transesterification rate for each biocatalyst: EB-rROL (circles, ), EO-rROL (squares, ) and EDVB-rROL (triangles, ). Table 1 suggested the absence of internal diffusional restrictions under the Weisz–Prater criterion for pomace oil transesterification with methanol in the conditions studied. However, because mass transfer limitations may be presence of various biocatalysts immobilised onto a polymethacrylate matrix. Φ an artefact on enzyme stability assessment (Klibanov, 1983; Ollis, and Deff were calculated from Eq. (2) and (3), respectively, using σ = 1 and τ = 1972), the absence of restrictions was confirmed by calculating the 1.41 in the latter. Porosity (εp) and specific volume of the supports were 0.6 and Weisz–Prater Criterion (Φ). Table 1 shows the results for the three 1.4 cm3 g− 1 support. Purolite® biocatalysts with the highest immobilised activity. Φ was less Parameters Biocatalyst than 0.3 in all cases, which supports the previous conclusion. EB-rROL EO-rROL EDVB-rROL robs (mol s− 1 cm− 3) × 107 6.44 ± 0.05 6.57 ± 0.02 5.20 ± 0.1 3.2. Polymethacrylate supports. Initial transesterification rate and Biocatalyst weight (g) 0.2 ± 0.05 0.2 ± 0.03 0.2 ± 0.01 operational stability ρp (g cm− 3)* 0.72 0.72 0.72 Specific area (m2 g− 1) * 152 139 59 robs (mol g–1 − 1 particle s ) × 10 5 2.89 ± 0.02 2.95 ± 0.01 2.33 ± 0.07 The catalytic performance of immobilised enzymes and the phys­ Cm,0 (mol cm− 3) × 104 4.41 4.41 4.41 ico–chemical conditions of their microenvironment are strongly influ­ Dm,a (cm2 s− 1) × 106 1.18 1.18 1.18 enced by functional groups in the supports (Bolivar and Nidetzky, 2019). Deff (cm2 s− 1) × 107 5.04 5.04 5.04 Φ × 107 0.78 ± 0.05 0.95 ± 0.03 4.19 ± 0.08 This led us to compare the initial reaction rate with methanol and *Data kindly provided by Purolite operational stability, with both ethanol and methanol as acyl-acceptors, ®. of the biocatalysts obtained by immobilising rROL onto Purolite® sup­ ports containing both epoxide and hydrophobic functional groups (viz., 3. Results and discussion EB-rROL, EO-rROL and EDVB-rROL) with those of rROL immobilised onto a polymethacrylate-based support containing epoxide groups only 3.1. Lipase immobilisation onto polymethacrylate-based supports and (EX-rROL) (Bonet-Ragel et al., 2018). diffusional restrictions The initial transesterification rate was similar with all polymethacrylate-based biocatalysts irrespective of the particular Recombinant Rhizopus oryzae lipase was covalently immobilised functional group —EB-rROL 24.9 ± 0.9, EO-rROL 25.1 ± 0.51, EDVB- onto three glutaraldehyde-treated polymethacrylate-based Purolite® rROL 24.4 ± 0.33 and EX-rROL 24.6 ± 0.78 µmol FAME min− 1 cm− 3. supports, namely: EB-rROL, EO-rROL and EDVB-rROL. Lipase solutions Contrary to the expectations, these results showed that chemical dif­ with an activity ranging from 55 to 1000 AU per mg support were used ferences in surface composition among supports had no effect on the to evaluate the immobilisation yield and maximum lipase activity initial reaction rate under the conditions studied. loading. All supports studied were identical in these parameters irre­ On the other hand, the operational stability of the biocatalysts was spective of their characteristics (Fig. 1A). Considering lipase immobili­ dramatically influenced by the functional groups of the supports. sation trend, offered activity values lower than 55 AU mg− 1 support Fig. 2A–2C shows the relative yield obtained by exposing each biocat­ resulted in immobilisation yields of 100% and less than 90 AU mg− 1 alyst to a variable number of pulses of the different alcohols in consec­ support in greater than 80%. As expected, increasing offered lipase ac­ utive reaction cycles. Operational stability with 1 pulse of methanol was tivity decreased the immobilisation yield. Also, biocatalysts activity not assessed owing to the low yield of the first reaction cycle by effect of increased in proportion to a maximum lipase activity loading of 200 deactivation of the enzyme (Fig. 2D) (Bonet-Ragel et al., 2018). Because immobilised AU mg− 1 support with all supports. Immobilised activity rROL is a 1,3-regiospecific enzyme, the maximum expected yield was was maximal with lipase solutions of 400 AU mg− 1 support, above 66%. Using a stoichiometric amount of alcohol (viz., a 2:1 alcohol-to-oil which no further improvement was observed. mole ratio) avoided the presence of too high alcohol concentration in the Mass transfer limitations in immobilised enzymes usually reflect in reaction medium —and hence potentially adverse effects on the opera­ diffusional restrictions, whether external or internal. The latter are tional stability of the biocatalysts. usually more severe with enzymes embedded in a solid porous matrix The joint presence of epoxide/butyl and epoxide/octadecyl groups in (Illanes, 2008) such as Purolite® supports. Diffusional restrictions were supports EB and EO, respectively, substantially increased the opera­ experimentally evaluated by examining the initial transesterification tional stability of the biocatalysts relative to EX-rROL upon exposure to rate with methanol of biocatalysts containing variable amounts of 1 pulse of ethanol and 5 of methanol addition strategies. By way of immobilised enzyme (Fig. 1B). The proportional relationship observed example, Fig. 2C shows the most extreme case, in which EB-rROL and 4 J. López-Fernández et al. Bioresource Technology 334 (2021) 125233 Fig. 2. Relative yield (%) in consecutive transesterification runs with biocatalysts immobilised on polymethacrylate-based supports upon exposure to (A) 1 pulse of ethanol, (B) 5 pulses of ethanol and (C) 5 pulses of methanol. The respective first reactions yields were taken to represent 100% yield. EB-rROL (black, ), EO- rROL (up-striped grey, ), EDVB-rROL (dark grey, ) and EX-rROL (down-striped white, ) (Bonet-Ragel et al., 2018). (D) First transesterification run with 1 pulse of methanol. The solid line represents the average FAME yield for the biocatalysts as a whole. EB-rROL (circles, ), EO-rROL (squares, ) and EDVB- rROL (triangles, ). EO-rROL were roughly 9 times more stable than EX-rROL after 5 reac­ reduced, formation of these bases is reversible in presence of water. tion cycles with 5 pulses of methanol. In contrast to EB-rROL and EO- However, EB-rROL performed identically in terms of initial reaction rate rROL, EDVB-rROL exhibited no increase in stability —which was and operational stability with ethanol with reduced and unreduced similar to that of EX-rROL. These results suggested that the increased Schiff bases. Consequently, there was no enzyme leakage (i.e., formation stability of EB-rROL and EO-rROL was not a consequence of microen­ of Schiff bases was not reversed owing to the low water concentration in vironmental changes due the presence of hydrophobic functional groups the transesterification medium and the well-known high complexity of —in fact, EDVB-rROL did not follow this trend—, but rather to that of the interaction between aldehyde groups in glutaraldehyde-treated surface hydrocarbon chains in the support possibly enhancing catalytic supports and immobilised enzymes) (Sheldon and van Pelt, 2013). performance of the enzyme (Urrutia et al., 2018). In addition, EO-rROL was 20% more stable than EB-rROL against 1 pulse of ethanol (Fig. 2A), which testified to the positive effect of the hydrocarbon chain length on 3.3. Magnetite-based supports the operational stability of the biocatalysts. The increased operational stability of EO-rROL and EB-rROL addi­ Superparamagnetic nanoparticles (MNPs) have been widely used as tionally made them resistant to 5 pulses of methanol. In fact, unlike EX- supports for immobilised enzymes by virtue of their easy recovery by rROL and EDVB-rROL, both biocatalysts exhibited an identical relative application of a magnetic field. The MNP matrix is non-porous, spher­ yield close to 100% after 5 reaction cycles irrespective of the alcohol ical-like solid magnetite (Fe3O4) that can be functionalised with a broad used (Fig. 2B and 2C) —being methanol a more powerful lipase inacti­ variety of chemicals for immobilisation purposes (Del Arco et al., 2021). vator than ethanol (Bonet-Ragel et al., 2018). However, neither EO- In this work, MNPs containing aldehyde groups (MNP-CHO) were rROL nor EB-rROL resisted deactivation by 1 pulse of methanol further functionalised with butylamine to obtain particles with both (Fig. 2D), the final FAME yield being around 15% rather than the ex­ aldehyde and butyl groups. The particles were treated with butylamine pected theoretical maximum: 66% (Fig. 2D). solutions of eight different concentrations to obtain functionalised de­ Covalent lipase immobilisation causes the formation of Schiff bases rivatives from MNP-CHO (the blank control, with no butyl groups) to through condensation of amino groups in lysines with aldehyde groups MNP-1MBut-CHO (functionalised with the most concentrated butyl­ in the supports (Sheldon and van Pelt, 2013). Unless they have been amine solution). Functionalised nanoparticles were then used to immobilise rROL and the biocatalysts thus obtained were assessed for 5 J. López-Fernández et al. Bioresource Technology 334 (2021) 125233 Fig. 3. (A) Relative yield (%) of consecutive transesterification runs (1 pulse of ethanol) in the presence of rROL immobilised on MNP-based supports functionalised with solutions ranging from 0 (blank control) to 1 M butylamine in 1.5 mL vials. The bars correspond to the functionalised MNPs obtained by using a butylamine concentration of 1 M (MNP-1MBut-rROL, black, ), 100 mM (MNP-100mMBut-rROL, white, ), 10 mM (MNP-10mMBut-rROL, up-striped white, ), 1 mM (MNP-1mMBut-rROL, horizontally striped grey, ), 100 µM (MNP-100µMBut-rROL, up-striped grey, ), 10 µM (MNP-10µMBut-rROL, horizontally striped white, ), 1 µM (MNP-1µMBut-rROL, down-striped grey, ) and 0 M (MNP-rROL, blank, down-striped white, ). (B) Relative yield (%) of consecutive transesterification runs (1 pulse of ethanol) with MNP-1mMBut-rROL (horizontally striped grey, ) and MNP-rROL (blank, down-striped white, ) in 10 mL vials. The respective first reactions yields were taken to represent 100% yield. the influence of surface butyl groups in the support on biocatalyst Table 2 performance. Productivity and volumetric productivity of polymethacrylate- and magnetite- Fig. 3A shows the results of the screening with ethanol as acyl- based biocatalysts exposed to a variable number of pulses of methanol or acceptor for the optimum butylamine concentration for use in 1.5 mL ethanol. vials. As can be seen, MNPs functionalised with solutions containing a Biocatalyst Productivity Volumetric productivity butylamine concentration above 10 µM provided biocatalysts more than (µmol min− 1) (µmol min− 1 cm− 3) twice as stable as those immobilised on MNPs treated with lower EtOH EtOH MetOH EtOH EtOH MetOH butylamine concentrations or no amine (blank). Also, no differences in 1 pulse 5 pulses 5 pulses 1 pulse 5 5 pulses operational stability were observed with butylamine concentrations pulses above 10 µM. Therefore, a concentration of 1 mM was chosen to assess EB-rROL 38.27 74.89 48.76 ± 0.86 1.68 1.10 ± the influence on the initial transesterification rate and operational sta­ ± ± 1.96 ± 0.36 0.30 ± 0.04 0.02 0.01 bility in 10 mL vials. EO-rROL 41.56 73.43 48.04 ± 0.94 1.65 ± 1.08 ± MNP-rROL and MNP-1mMBut-rROL exhibited a similar initial ± 0.91 ± 0.42 0.35 ± 0.02 0.01 0.02 transesterification rate in 10 mL vials. As with the Purolite® bio­ EDVB-rROL 34.83 67.07 34.25 ± 0.78 1.51 ± 0.77 ± ± 0.28 ± 0.39 0.38 ± 0.01 0.01 0.01 catalysts, the presence of surface hydrocarbon chains on the support had EX-rROL 23.00 60.36 19.62 ± 0.52 1.36 ± 0.44 ± no effect on the initial rate. However, the initial reaction rate with the ± 0.33 ± 0.40 0.33 0.02 0.02 MNP-based biocatalysts was roughly one-half that obtained with the MNP-rROL 22.16 0.50 Purolite®-based counterparts —MNP-rROL 10.3 ± 0.25 and MNP- ± 0.66 ± 0.02 1mMBut-rROL 12.2 ± 0.85 µmol FAME min− 1 cm− 3. Operational sta­ MNP- 35.36 0.79 1mMBut- ± 1.50 ± 0.03 bility (Fig. 3B) followed the same trend as in the screening tests rROL (Fig. 3A), MNP-1mMBut-rROL being twice more stable than MNP-rROL after 5 consecutive reaction cycles. Whereas MNP-rROL and EX-rROL (i. e., two catalysts with homologous supports as regards functional groups) exhibited better productivity and volumetric productivity than MNP- were similar in operational stability, MNP-1mMBut-rROL was 20% less based biocatalysts also containing surface chains (e.g., 20% higher stable than EO-rROL and 15% less so than EB-rROL against 1 pulse of with EO-rROL than with MNP-1mMBut-rROL). Also, productivity and ethanol addition strategy. Therefore, no further research was performed volumetric productivity among Purolite®-based biocatalysts were 1.5–2 with superparamagnetic biocatalysts. times greater in EO and EB-rROL than they were in EX-rROL with 1 pulse of EtOH and 5 of MetOH, respectively —differences were smaller with 5 pulses of EtOH as a result of the weaker adverse effect of this acyl- 3.4. Biocatalyst performance: Productivity and volumetric productivity acceptor and of its application being split over several pulses (Bonet- Ragel et al., 2018). In addition, adding ethanol in 5 pulses resulted in In addition to initial reaction rate and operational stability, the roughly twice better productivity and volumetric productivity than biocatalysts were assessed in terms of productivity (µmol FAEE/FAME using a single pulse with any biocatalyst. Further research splitting the min− 1) and volumetric productivity (µmol FAEE/FAME min− 1 cm− 3) in amount of alcohol in more pulses was not performed as it would reduce 5 consecutive reaction cycles for use at the industrial scale. biodiesel productivity (Bonet-Ragel et al., 2018). As can clearly be seen from Table 2, the biocatalysts immobilised The productivity, volumetric productivity and operational stability onto supports functionalised with hydrocarbon chains performed better results led us to choose EO-rROL and 5 pulses of either alcohol for than those onto supports containing none irrespective of matrix type further testing. (polymethacrylate or magnetite nanoparticles). Therefore, the presence of hydrocarbon chains had a favourable effect on biocatalyst perfor­ mance. Purolite®-based biocatalysts with surface hydrocarbon chains 6 J. López-Fernández et al. Bioresource Technology 334 (2021) 125233 Table 3 R. toruloides is also similar to previously reported values (Singh et al., Percent acidity of alternative oily substrates and initial transesterification rate 2018). Although the acidity of WCO falls within the typical ranges (less with EO-rROL. than 0.5% for refined oils and 0.5–15% for used oils; Lam et al., 2010), Substrate Acidity (%) Initial reaction rate its low value suggests that it was not extensively cooked —the presence (µmol FAME min− 1 cm− 3) of FFAs in waste oil is mainly a result of oxidative, hydrolytic and Makauba oil 12.16 ± 0.85 21.4 ± 0.15 thermolytic reactions during frying (Hama et al., 2013). Jatropha oil 7.85 ± 0.22 24.5 ± 0.22 The presence of FFAs has been reported to have a favourable effect Microbial oil 2.47 ± 0.19 22.06 ± 0.34 on the initial reaction rate and the operational stability of enzymes WCO 0.77 ± 0.13 22.8 ± 0.68 during transesterification (Bonet-Ragel et al., 2015). Although EO-rROL Olive pomace oil 18.93 ± 0.8 24.4 ± 0.33 exhibited a similar initial rate irrespective of substrate acidity (Table 3) —what might be explained because not only the initial FFA content is 3.5. Alternative substrates: second- and third-generation biodiesel crucial, but also the immediately generated FFAs (Lin et al., 2011)— this property had a marked effect on its operational stability. With ethanol as Although oil pomace was used as a model substrate to assess bio­ acyl-acceptor in the transesterification reaction, oil acidity had no catalysts performance, the potential usefulness of this substrate for the appreciable effect (Fig. 4A) because this alcohol is known to scarcely food industry could still stir the ethical debate between food and fuels. inactivate lipase. With methanol, a more powerful lipase inactivator We therefore chose to investigate alternative oily substrates such as non- (Bonet-Ragel et al., 2018), FFAs did influence the results (Fig. 4B). Thus, edible vegetable oils from makauba and jatropha, microbial oil from with the least acidic substrate (WCO), the operational stability of the Rhodosporidium toruloides and WCO. biocatalyst was rather poor —more than 95% of the initial relative yield Table 3 shows the acidity of each alternative substrate. The values for was lost after only 5 reaction cycles. Meanwhile, substrates with acidity vegetable oils from makauba (Aguieiras et al., 2014) and jatropha (Jain >2% maintained at least 60% of the initial value —roughly 30 times and Sharma, 2010) are consistent within previous reports. However, the more than the relative yield obtained with WCO. Therefore, substrate acidity of these oils depends markedly on factors such as the harvest and acidity strongly influenced the operational stability of the biocatalyst. storage conditions (Souza et al., 2016). The acidity of microbial oil from Also, the results are suggestive of a synergistic effect of the alcohol Fig. 4. Relative yield (%) of consecutive transesterification cycles of various types of oil with 5 pulses of ethanol (A) or 5 of methanol (B) in the presence of biocatalyst EO-rROL. The bars correspond to pomace (black, ), makauba (up-striped grey, ), jatropha (striped white, ), microbial (white, ) and WCO (grey, ). (C) Comparison of the relative yield (%) obtained with 5 (grey, ) and 10 pulses of methanol (squared white, ) and WCO as substrate. The initial reactions yields were taken to be 100%. 7 J. López-Fernández et al. Bioresource Technology 334 (2021) 125233 outperformed by Purolite®-based biocatalysts. Oil acidity and acyl- acceptor concentration proved crucial and synergistically influential on biocatalyst performance. The fact that the enzymatic trans­ esterification of WCO with EO-rROL was successfully scaled up vali­ dated its potential for industrial implementation to enable biodiesel production from an inexpensive substrate and management of the waste under the circular economy principles. CRediT authorship contribution statement Josu López-Fernández: Conceptualization, Formal analysis, Inves­ tigation, Methodology, Writing - original draft. Maria Dolors Benaiges: Data curation, Methodology, Supervision, Validation, Writing - review & editing. Francisco Valero: Data curation, Funding acquisition, Methodology, Project administration, Supervision, Validation, Writing - review & editing. Fig. 5. Relative yield (%) of consecutive transesterification cycles of WCO with Declaration of Competing Interest 5 pulses of ethanol (black circles, ) and 10 of methanol (white circles, ) in the presence of biocatalyst EO-rROL. The times on the x-axis are those at which The authors declare that they have no known competing financial each cycle was started. The initial reactions yields were taken to be 100%. interests or personal relationships that could have appeared to influence the work reported in this paper. concentration and substrate acidity on deactivation of the biocatalysts. This hypothesis was checked by using a reduced concentration of Acknowledgment methanol in the reaction medium (specifically, by splitting the amount used in 10 pulses) with WCO as substrate. As can be seen from Fig. 4C, This work was funded by the Spanish Ministry of Science and Inno­ the operational stability of the biocatalyst was 20 times greater under vation (Project PID2019-104666GB-100). JLF was additionally funded these conditions. Therefore, the origin of the synergistic effect was the by a Basque Government Scholarship for the training of pre-doctoral increased polarity of the reaction medium in the presence of FFAs, which researchers (PRE_2017_1_0110). acted as a buffering agent for the high polarity of methanol —and hence for its also high deactivation capacity (Canet et al., 2016). Appendix A. Supplementary data 3.6. Transesterification scale-up Supplementary data to this article can be found online at https://doi. org/10.1016/j.biortech.2021.125233. Once WCO was found to be the most suitable substrate on the grounds of its low cost, high production and potential for use in accor­ References dance with the principles of circular economy (Hama et al., 2013; Aguieiras, E.C.G., Cavalcanti-Oliveira, E.D., De Castro, A.M., Langone, M.A.P., Freire, D. Olkiewicz et al., 2016), it was used to scale up the enzymatic production M.G., 2014. Biodiesel production from Acrocomia aculeata acid oil by (enzyme/ of biodiesel to a 50 mL mini-reactor for industrial proof of concept. enzyme) hydroesterification process: use of vegetable lipase and fermented solid as Transesterification runs in 10 mL vials under orbital stirring were low-cost biocatalysts. Fuel 135, 315–321. https://doi.org/10.1016/j. fuel.2014.06.069. successfully scaled up to a stirred tank reactor under mechanical stirring Ariaeenejad, S., Motamedi, E., Hosseini Salekdeh, G., 2021. Application of the (Bonet-Ragel, 2018). As can be seen from Fig. 5, the relative yields ob­ immobilized enzyme on magnetic graphene oxide nano-carrier as a versatile bi- tained in consecutive cycles using 5 pulses of EtOH and 10 of MetOH functional tool for efficient removal of dye from water. Bioresour. Technol. 319, with the laboratory-scale mini-reactor were similar to those obtained in 124228 https://doi.org/10.1016/j.biortech.2020.124228. Arnau, C., Ramon, R., Casas, C., Valero, F., 2010. Optimization of the heterologous 10 mL vials (Fig. 4). Therefore, the proposed enzymatic biodiesel pro­ production of a Rhizopus oryzae lipase in Pichia pastoris system using mixed duction method can be easily implemented on an industrial proof of substrates on controlled fed-batch bioprocess. Enzyme Microb. Technol. 46, concept scale. Also, around 80% of the initial relative yield was main­ 494–500. https://doi.org/10.1016/j.enzmictec.2010.01.005. Bolivar, J.M., Nidetzky, B., 2019. The microenvironment in immobilized enzymes: tained after 10 consecutive reaction cycles with either acyl-acceptor. methods of characterization and its role in determining enzyme performance. Although few studies on the use of WCO in combination with Rhizopus Molecules 24, 3460–3484. https://doi.org/10.3390/molecules24193460. oryzae lipase have been reported (López-Fernández et al., 2020), similar Bonet-Ragel, K., 2018. Enzymatic synthesis of biodiesel from high free fatty acid feedstock using a recombinant Rhizopus oryzae lipase. Universitat Autònoma de operational stabilities have been obtained with other substrates (Duarte Barcelona, Cerdanyola del Vallés. et al., 2015; Su et al., 2014). Also, there was virtually no difference Bonet-Ragel, K., Canet, A., Benaiges, M.D., Valero, F., 2018. Effect of acyl-acceptor between using 5 pulses of EtOH and 10 pulses of MetOH, in order to stepwise addition strategy using alperujo oil as a substrate in enzymatic biodiesel synthesis. J. Chem. Technol. Biotechnol. 93, 541–547. https://doi.org/10.1002/ minimise the negative synergistic effect of the alcohol concentration and jctb.5399. substrate acidity on biocatalyst activity at the expense of longer reaction Bonet-Ragel, K., Canet, A., Benaiges, M.D., Valero, F., 2015. Synthesis of biodiesel from times, reduced productivity and increased operational costs for a po­ high FFA alperujo oil catalysed by immobilised lipase. Fuel 161, 12–17. https://doi. org/10.1016/j.fuel.2015.08.032. tential biodiesel plant. Bradford, M.M., 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 4. Conclusions 72, 248–254. https://doi.org/10.1016/0003-2697(76)90527-3. Canet, A., Benaiges, M.D., Valero, F., Adlercreutz, P., 2017. Exploring substrate specificities of a recombinant Rhizopus oryzae lipase in biodiesel synthesis. Rhizopus oryzae lipase proved a suitable biocatalyst for biodiesel N. Biotechnol. 39, 59–67. https://doi.org/10.1016/j.nbt.2017.07.003. production from various oily substrates. The hydrocarbon chain length Canet, A., Bonet-Ragel, K., Benaiges, M.D., Valero, F., 2016. Lipase-catalysed was found to play a key role in increasing the operational stability of transesterification: viewpoint of the mechanism and influence of free fatty acids. Biomass Bioenergy 85, 94–99. https://doi.org/10.1016/j.biombioe.2015.11.021. rROL immobilised onto polymethacrylate- and magnetite-based sup­ Canet, A., Dolors Benaiges, M., Valero, F., 2014. Biodiesel synthesis in a solvent-free ports. Despite their easy recovery, magnetite-based biocatalysts were system by recombinant Rhizopus oryzae lipase. Study of the catalytic reaction 8 J. López-Fernández et al. Bioresource Technology 334 (2021) 125233 progress. JAOCS. J. Am. Oil Chem. Soc. 91, 1499–1506. https://doi.org/10.1007/ Luo, X., Ge, X., Cui, S., Li, Y., 2016. Value-added processing of crude glycerol into s11746-014-2498-y. chemicals and polymers. Bioresour. Technol. 215, 144–154. https://doi.org/ Connolly, D., Lund, H., Mathiesen, B.V., 2016. Smart energy Europe: the technical and 10.1016/j.biortech.2016.03.042. economic impact of one potential 100% renewable energy scenario for the European Ma, Y., Gao, Z., Wang, Q., Liu, Y., 2018. Biodiesels from microbial oils: opportunity and Union. Renew. Sustain. Energy Rev. 60, 1634–1653. https://doi.org/10.1016/j. challenges. Bioresour. Technol. 263, 631–641. https://doi.org/10.1016/j. rser.2016.02.025. biortech.2018.05.028. Cruz-Izquierdo, Á., Picó, E.A., López, C., Serra, J.L., Llama, M.J., 2014. Magnetic cross- Madhavan, A., Sindhu, R., Binod, P., Sukumaran, R.K., Pandey, A., 2017. Strategies for linked enzyme aggregates (mCLEAs) of Candida antarctica lipase: an efficient and design of improved biocatalysts for industrial applications. Bioresour. Technol. 245 stable biocatalyst for biodiesel synthesis. PLoS One 9, 1–22. https://doi.org/ Part B, 1304–1313. https://doi.org/10.1016/j.biortech.2017.05.031. 10.1371/journal.pone.0115202. Navarro López, E., Robles Medina, A., González Moreno, P.A., Esteban Cerdán, L., Martín Da Rós, P.C.M., Silva, G.A.M., Mendes, A.A., Santos, J.C., de Castro, H.F., 2010. Valverde, L., Molina Grima, E., 2016. Biodiesel production from Nannochloropsis Evaluation of the catalytic properties of Burkholderia cepacia lipase immobilized on gaditana lipids through transesterification catalyzed by Rhizopus oryzae lipase. non-commercial matrices to be used in biodiesel synthesis from different feedstocks. Bioresour. Technol. 203, 233–244. https://doi.org/10.1016/j.biortech.2015.12.036. Bioresour. Technol. 101, 5508–5516. https://doi.org/10.1016/j. Noureddini, H., Gao, X., Philkana, R.S., 2005. Immobilized Pseudomonas cepacia lipase biortech.2010.02.061. for biodiesel fuel production from soybean oil. Bioresour. Technol. 96, 769–777. Del Arco, J., Alcántara, A.R., Fernández-Lafuente, R., Fernández-Lucas, J., 2021. https://doi.org/10.1016/j.biortech.2004.05.029. Magnetic micro-macro biocatalysts applied to industrial bioprocesses. Bioresour. Ogunkunle, O., Ahmed, N.A., 2019. A review of global current scenario of biodiesel Technol. 322, 124547 https://doi.org/10.1016/j.biortech.2020.124547. adoption and combustion in vehicular diesel engines. Energy Reports 5, 1560–1579. Duarte, S.H., del Peso Hernández, G.L., Canet, A., Benaiges, M.D., Maugeri, F., Valero, F., https://doi.org/10.1016/j.egyr.2019.10.028. 2015. Enzymatic biodiesel synthesis from yeast oil using immobilized recombinant Olkiewicz, M., Torres, C.M., Jiménez, L., Font, J., Bengoa, C., 2016. Scale-up and Rhizopus oryzae lipase. Bioresour. Technol. 183, 175–180. https://doi.org/10.1016/ economic analysis of biodiesel production from municipal primary sewage sludge. j.biortech.2015.01.133. Bioresour. Technol. 214, 122–131. https://doi.org/10.1016/j.biortech.2016.04.098. Guillén, M., Benaiges, M.D., Valero, F., 2012. Biosynthesis of ethyl butyrate by Ollis, D.F., 1972. Diffusion influences in denaturable insolubilized enzyme catalysis. immobilized recombinant Rhizopus oryzae lipase expressed in Pichia pastoris. Biotechnol. Bioeng. 14, 871–884. https://doi.org/10.1002/bit.260140603. Biochem. Eng. J. 65, 1–9. https://doi.org/10.1016/j.bej.2012.03.009. Poling, B.E., Prausnitz, J.M., O’Connell, J.P., 2001. The properties of gases & liquids - 5th Hama, S., Kondo, A., 2013. Enzymatic biodiesel production: an overview of potential edition. McGraw Hill Education. New York, USA. https://doi.org/10.1016/0894- feedstocks and process development. Bioresour. Technol. 135, 386–395. https://doi. 1777(88)90021-0. org/10.1016/j.biortech.2012.08.014. Ranganathan, S.V., Narasimhan, S.L., Muthukumar, K., 2008. An overview of enzymatic Hama, S., Yoshida, A., Tamadani, N., Noda, H., Kondo, A., 2013. Enzymatic production production of biodiesel. Bioresour. Technol. 99, 3975–3981. https://doi.org/ of biodiesel from waste cooking oil in a packed-bed reactor: an engineering approach 10.1016/j.biortech.2007.04.060. to separation of hydrophilic impurities. Bioresour. Technol. 135, 417–421. https:// Resina, D., Serrano, A., Valero, F., Ferrer, P., 2004. Expression of a Rhizopus oryzae lipase doi.org/10.1016/j.biortech.2012.06.059. in Pichia pastoris under control of the nitrogen source-regulated formaldehyde Harvey W. Blanch, D.S.C., 1997. Biochemical engineering, second edition - CRC Press. dehydrogenase promoter. J. Biotechnol. 109, 103–113. https://doi.org/10.1016/j. Boca Raton, USA. jbiotec.2003.10.029. He, Y., Wu, T., Wang, X., Chen, B., Chen, F., 2018. Cost-effective biodiesel production Rodrigues, A., Bordado, J.C., Dos Santos, R.G., 2017. Upgrading the glycerol from from wet microalgal biomass by a novel two-step enzymatic process. Bioresour. biodiesel production as a source of energy carriers and chemicals - a technological Technol. 268, 583–591. https://doi.org/10.1016/j.biortech.2018.08.038. review for three chemical pathways. Energies 10, 1817–1853. https://doi.org/ Illanes, A., 2008. Enzyme biocatalysis: principles and applications, in: enzyme 10.3390/en10111817. biocatalysis: principles and applications. 2008th edition. Springer. https://doi.org/ Rodrigues, J., Perrier, V., Lecomte, J., Dubreucq, E., Ferreira-Dias, S., 2016. Biodiesel 10.1007/978-1-4020-8361-7. production from crude jatropha oil catalyzed by immobilized lipase/acyltransferase Jain, S., Sharma, M.P., 2010. Kinetics of acid base catalyzed transesterification of from Candida parapsilosis in aqueous medium. Bioresour. Technol. 218, 1224–1229. Jatropha curcas oil. Bioresour. Technol. 101, 7701–7706. https://doi.org/10.1016/j. https://doi.org/10.1016/j.biortech.2016.07.090. biortech.2010.05.034. Sheldon, R.A., van Pelt, S., 2013. Enzyme immobilisation in biocatalysis: why, what and Jin, Z., Han, S.Y., Zhang, L., Zheng, S.P., Wang, Y., Lin, Y., 2013. Combined utilization of how. Chem. Soc. Rev. 42, 6223–6235. https://doi.org/10.1039/C3CS60075K. lipase-displaying Pichia pastoris whole-cell biocatalysts to improve biodiesel Singh, G., Jeyaseelan, C., Bandyopadhyay, K.K., Paul, D., 2018. Comparative analysis of production in co-solvent media. Bioresour. Technol. 130, 102–109. https://doi.org/ biodiesel produced by acidic transesterification of lipid extracted from oleaginous 10.1016/j.biortech.2012.12.020. yeast Rhodosporidium toruloides. 3. Biotech 8, 434. https://doi.org/10.1007/s13205- Khanahmadi, S., Yusof, F., Chyuan Ong, H., Amid, A., Shah, H., 2016. Cocoa pod husk: a 018-1467-9. new source of CLEA-lipase for preparation of low-cost biodiesel: an optimized Souza, G.K., Scheufele, F.B., Pasa, T.L.B., Arroyo, P.A., Pereira, N.C., 2016. Synthesis of process. J. Biotechnol. 231, 95–105. https://doi.org/10.1016/j.jbiotec.2016.05.015. ethyl esters from crude macauba oil (Acrocomia aculeata) for biodiesel production. Klibanov, A.M., 1983. Stabilization of enzymes against thermal inactivation. Adv. Appl. Fuel 165, 360–366. https://doi.org/10.1016/j.fuel.2015.10.068. Microbiol. 29, 1–28. https://doi.org/10.1016/S0065-2164(08)70352-6. Su, F., Li, G., Zhang, H., Yan, Y., 2014. Enhanced performance of Rhizopus oryzae lipase Lam, M.K., Lee, K.T., Mohamed, A.R., 2010. Homogeneous, heterogeneous and immobilized on hydrophobic carriers and its application in biorefinery of rapeseed enzymatic catalysis for transesterification of high free fatty acid oil (waste cooking oil deodorizer distillate. Bioenergy Res. 7, 935–945. https://doi.org/10.1007/ oil) to biodiesel: a review. Biotechnol. Adv. 28, 500–518. https://doi.org/10.1016/j. s12155-014-9415-y. biotechadv.2010.03.002. Urrutia, P., Arrieta, R., Alvarez, L., Cardenas, C., Mesa, M., Wilson, L., 2018. Lin, Y.H., Luo, J.J., John Hwang, S.C., Liau, P.R., Lu, W.J., Lee, H.T., 2011. The influence Immobilization of lipases in hydrophobic chitosan for selective hydrolysis of fish oil: of free fatty acid intermediate on biodiesel production from soybean oil by whole the impact of support functionalization on lipase activity, selectivity and stability. cell biocatalyst. Biomass and Bioenergy 35, 2217–2223. https://doi.org/10.1016/j. Int. J. Biol. Macromol. 108, 674–686. https://doi.org/10.1016/j. biombioe.2011.02.039. ijbiomac.2017.12.062. López-Fernández, J., Barrero, J.J., Benaiges, M.D., Valero, F., 2019. Truncated Weisz, P.B., Prater, C.D., 1954. Interpretation of measurements in experimental catalysis. prosequence of Rhizopus oryzae lipase: key factor for production improvement and Adv. Catal. 6, 143–196. https://doi.org/10.1016/S0360-0564(08)60390-9. biocatalyst stability. Catalysts 9, 961. https://doi.org/10.3390/catal9110961. Xu, J., Liu, C., Wang, M., Shao, L., Deng, L., Nie, K., Wang, F., 2017. Rotating packed bed López-Fernández, J., Benaiges, M.D., Valero, F., 2020. Rhizopus oryzae lipase, a reactor for enzymatic synthesis of biodiesel. Bioresour. Technol. 224, 292–297. promising industrial enzyme: biochemical characteristics, production and https://doi.org/10.1016/j.biortech.2016.10.045. biocatalytic applications. Catalysts 10, 1277. https://doi.org/10.3390/ catal10111277. 9