Surfactant Imprinting Hyperactivated Immobilized Lipase as E ﬃ cient Biocatalyst for Biodiesel Production from Waste Cooking Oil

: Enzymatic production of biodiesel from waste cooking oil (WCO) could contribute to resolving the problems of energy demand and environment pollutions.In the present work, Burkholderia cepacia lipase (BCL) was activated by surfactant imprinting, and subsequently immobilized in magnetic cross-linked enzyme aggregates (mCLEAs) with hydroxyapatite coated magnetic nanoparticles (HAP-coated MNPs). The maximum hyperactivation of BCL mCLEAs was observed in the pretreatment of BCL with 0.1 mM Triton X-100. The optimized Triton-activated BCL mCLEAs was used as a highly active and robust biocatalyst for biodiesel production from WCO, exhibiting signiﬁcant increase in biodiesel yield and tolerance to methanol. The results indicated that surfactant imprinting integrating mCLEAs could ﬁx BCL in their active (open) form, experiencing a boost in activity and allowing biodiesel production performed in solvent without further addition of water. A maximal biodiesel yield of 98% was achieved under optimized conditions with molar ratio of methanol-to-WCO 7:1 in one-time addition in hexane at 40 ◦ C. Therefore, the present study displays a versatile method for lipase immobilization and shows great practical latency in renewable biodiesel production.


Introduction
Over the past decades, biodiesel has attracted great interest as a sustainable alternative for fossil fuels in virtue of the depletion of fossilized fuel resources and their environmental impacts [1]. Biodiesel is a renewable and clean energy, and possess favorable advantages in combustion emission like low emissions of CO, sulfur free, low hydrocarbon aroma, high cetanenumber, and high flash point [2].
The conventional chemical technologies for biodiesel production involve the use of acid or basic catalysts (i.e., NaOH, KOH, and H 2 SO 4 ), thus numerous disadvantages are inescapable, for example, high corrosive procedure, high energy consumption, high quantities of waste pollution, and costly in efficient product separation processes [3]. Furthermore, in order to prevent the hydrolysis reaction and saponification, high quality oils are required, with low contents of water and free fatty acids [4].
Feedstocks used for biodiesel can be allocated five categories, including edible vegetable oils, non-edible plant oils, animal fats, microalgae oils, and waste oils [5]. The global application of first-generation biodiesel produced by using edible oils, was restricted due to food scarcity and high cost of the edible oils [6]. Biodiesel production from waste cooking oils (WCO) could be a promising and cost effective candidate in handling issues associated with energy crisis, environmental concerns, and total cost reduction of biodiesel production [7]. Moreover, 15 million tons of WCO are Bioimprinting is a commonly used method for achieving hyperactivation of lipases in organic media. The principle of bioimprinting is to "anchor" the enzyme in its active form, which could be achieved by binding with imprint molecules (such as surfactants, natural substrates, substrate analogs etc.). From an applied point of view, the dramatic hyperactivation of lipases by low concentrations of surfactants is an expeditious and facile method for lipase interfacial activation [20].
To develop an efficient and environmentally benign process for the biodiesel production from waste cooking oils, in the present study surfactant imprinting strategy on BCL was implemented in combination with mCLEAs immobilization using HAP-coated MNPs. Subsequent cross-linking could "lock" BCL in its favorable conformation, while HAP-coated MNPs could facilitate the recovery of immobilized BCL and simplify the biodiesel purification. To the best of our knowledge, this is the first report on BCL immobilization integrating surfactant imprinting and mCLEAs. The optimal conditions for mCLEAs preparation, along with the effect of different surfactants (anionic, cationic, and non-ionic) on the catalytic activity of BCL mCLEAs in transesterification were studied. The optimized surfactant-activated BCL mCLEAs was further used in transesterification of waste cooking oils to biodiesel. In addition, a detailed analysis of solvents, methanol-to-oil molar ratio, and temperatures on the yield of biodiesel production was presented. The results obtained in the research are expected to provide a reliable basis for further exploration of lipase immobilization and efficient biodiesel production in industry.

Preparation and Characterization of Immobilized Lipase
In this study, the prepared MNPs encapsulated by hydroxyapatite (HAP) were used as immobilization supports. The amino functionalization of HAP-coated MNPs was carried out using 3-aminopropyltrimethoxysilane (APTES) for efficient enzyme attachment. Typically, the preparation procedure of immobilization supports and surfactant-activated BCL mCLEAs were performed according to the scheme shown in Scheme 1. The prepared magnetic supports and immobilized BCL were characterized by fourier transform infrared spectroscopy (FT-IR), transmission electron microscope (SEM) and vibrating sample magnetometer (VSM).

Scheme 1.
Preparation procedure of immobilization supports and surfactant-activated Burkholderia cepacia lipase (BCL) magnetic cross-linked enzyme aggregates (mCLEAs). FTIR characterization was performed to investigate the chemical composition of functionalized MNPs and immobilized BCL. Spectra were recorded on over the region from 4000 to 400 cm −1 . As shown in Figure 1, the strong peak at 588 and 639 cm −1 corresponds to the stretching vibration of Fe-O bond. The characteristic absorption bands related to the HAP appease at 565 and 1044 cm −1 , which are assigned to phosphate groups [21]. In the IR spectrum of modified MNPs and BCL mCLEAs, the characteristic absorption bands related to the functional groups of HAP emerged clearly, which demonstrated the successful incorporation of MNPs with HAP. For all immobilized lipases, including BCL CLEAs, BCL mCLEAs and surfactant-activated BCL mCLEAs, the typical IR bands responsible for the lipase that were chemically covalent-bonded to the functionalized MNPs were observed at 1642 cm −1 for amide I (C=O stretching vibration) and at 1539 cm −1 for amide II (N-H bending vibration), respectively. Besides, compared with the results shown in Figure 1, aliphatic C-H stretch band at 2859 and 2927 cm −1 , corresponding to C-H stretching vibrations, are clearly observed in all immobilized lipases, which also indicated the successful loading of lipase. In order to assess morphology, size and composition of functionalized MNPs and immobilized BCL, SEM images were collected and illustrated in Figure 2. As seen in Figure 2, bare Fe 3 O 4 MNPs formed significantly dense agglomeration, because of their high surface energy and strong dipole-dipole interactions. It is obvious that the structure of Fe 3 O 4 MNPs becomes looser and more evenly distributed after being functionalized with HAP ( Figure 2B) and APTES ( Figure 2C), suggesting that surface modification is favorable for preventing aggregation of Fe 3 O 4 MNPs. At the same time, the rough surface of Fe 3 O 4 MNPs also increased the surface area for attachment of enzyme.
The crucial structure factors in aggregated-based enzyme immobilization, including morphological topographies, structural arrangement and size, play an important role in affecting substrate affinity and stability of biocatalyst [22]. Besides, the particle size of enzymes is an important property of any heterogeneous catalysis since it can directly affect the diffusion of substrates and catalytic efficiency, especially in the internal enzymes of highly compact aggregates [23]. SEM images ( Figure 2D) of standard BCL CLEAs revealed no defined morphologies and large size particles. Moreover, standard BCL CLEAs presented a uniform and compact surface with the presence of few tiny pores. On the contrary, after the incorporation of functionalized MNPs, BCL mCLEAs formed spherical structures and small particle sizes, which could reduce inner steric hindrance in closely packed CLEAs. It is noteworthy that the presence of functionalized MNPs displayed large active surface available for lipase immobilization, therefore were important for development of a stabilized enzyme-matrix. Furthermore, a loose and homodispersed structure of Triton-activated BCL mCLEAs was found in Figure 2F, suggesting that the formation of large aggregates were forbidden by the imprinting of surfactants. From the SEM outcomes, it can be discerned that, thanks to the coating of surfactants, lipase could be uniformly dispersed on functionalized MNPs, which could contribute to a wider surface area with more catalytic sites and decrease the diffusion limit. Consequently, compared with standard BCL CLEAs, Triton-activated BCL mCLEAs could perform superior catalytic efficiency.  With further functionalization of MNPs, the saturation magnetization value decreased and correlated with the increase of the core-shell layer. Interestingly, it is obviously observed that the saturation magnetization value of Triton-activated BCL mCLEAs increased visibly compared to BCL mCLEAs. It might be due to the uniform dispersion of lipase on MNPs and availability of large surface area which decreased the shielding-effect of the out layer substances. As seen in Figure 3 (inner), Triton-activated BCL mCLEAs showed fast response (6s) to the external magnetic field andcould be easily recovered from the reaction mixture. After removing the external magnetic field, the magnetic immobilized BCL redispersed rapidly by a slight shake, indicating good dispersion and efficient recyclability in industrial application.

Optimization of the Immobilization Conditions
In this study, the enzymes were precipitated by adding water-miscible organic solvents (acetone, ethanol and 2-propanol), PEG 800 and ammonium sulfate. The optimum precipitant was selected by measuring the transesterification activity of the corresponding BCL mCLEAs. Compared with free BCL, all BCL mCLEAs prepared using different precipitants performed higher transesterification activity in organic solvent. Among the protein precipitants evaluated, ammonium sulfate showed maximum recovery of activity (Figure 4a), therefore was further used in BCL immobilization. Traditionally, glutaraldehyde has been extensively used as the cross-linking agent to prepare CLEAs of various enzymes and exhibited a strong effect on activity and particle size of enzyme aggregates. The activity recovery of CLEAs greatly depends on the type of enzyme and the concentration of glutaraldehyde [24]. Lower glutaraldehyde concentration affects the cross-linking efficiency, which might result in enzyme leakage in immobilization, while excessive glutaraldehyde can induce the flexibility of enzymes and the active site availability, consequently, decreasing the activity recovery of CLEAs [25,26]. In this study, the influence of glutaraldehyde concentration on activity of BCL mCLEAs was investigated by using various concentrations of glutaraldehyde in cross-linking. As shown in Figure 4b, the optimum glutaraldehyde concentration of BCL mCLEAs was 2.0% (v/v).

Hyperactivation of BCL mCLEAs with Surfactants
A pivotal challenge in lipase immobilization is to open the lid of lipases and fix their open form for the exposure of active site. Surfactant imprinting is an efficient approach to activate lipases by facilitating lid-opening. Like other lipases, BCL also consists of a mobile element at the surface, which composed of two helical elements (a5-and a9-helices) and covers the active site [18]. To improve the catalytic performance of BCL mCLEAs, BCL was imprinted in the presence of surfactants prior to immobilization. Thus, four different surfactants with different properties (cationic, anionic and non-ionic) were investigated for modulating the activity of BCL mCLEAs in biodiesel production. As seen in Figure 5, Triton X-100 exhibited maximum effect on the enhancement of lipase activity in low surfactant concentration, while the addition of sodium bis-2-(ethylhexyl) sulfosuccinate (AOT) showed the least influence. The optimal surfactant in proper concentration acting as a bipolar agent, could simulate the amphiphilic environment to benefit the exposure of hydrophobic regions in the active site. Meanwhile, surfactants may also promote the dissociation of large aggregates formed, thus slightly increase the enzymatic activity of lipase ( Figure 2). However, the increase of surfactant concentration led to gradual decrease of biodiesel yield in all cases, indicating that surfactants showed positive and negative effect on the activity of lipase. Additional detergent molecules may bind to the active site region of lipase, blocking the substrate access, inducing inhibition [27]. Compared with ionic surfactants (AOT and cetrimonium bromide (CTAB)), nonionic surfactants (Triton X-100 and Tween 80) were preferred aiming at regulating the activity of BCL ( Figure 5). As the main interaction between the enzyme and nonionic surfactants is hydrophobic interaction while anionic or cationic surfactants perform electrostatic interactions [28], mild hydrophobic interaction between BCL and the surfactant might be important to trigger the interfacial activation mechanism. Therefore, nonionic Triton X-100 and Tween 80 were further studied to confirm the optimal amphiphile and surfactant concentration. As performed in Figure 6, the maximum hyperactivation of BCL mCLEAs was observed in the pretreatment of BCL with 0.1 mM Triton X-100, and the optimal Triton-activated BCL mCLEAs were used for further experiments.

Biodiesel Production
For the economic feasibility of biodiesel production, solvents, methanol-to-oil molar ratio, and reaction temperature are important variables to optimize for transesterification step. As a result of oxidative reactions occurring during cooking and long-term storage in air, WCO generally exhibits a dramatic increase in viscosity and saponification value [7]. Compared to the fresh oil, high viscose WCO is not favored in biodiesel production. Using solvents could reduce the viscosity of the reaction medium and decrease the diffusion limitations, while it might also directly affect the enzyme structure and activity. In general, hydrophobic solvents could promote the interface and stabilize lipases on their open assembly, causing the hyperactivation of these enzymes. To select the most suitable medium, five hydrophobic solvents commonly used in transesterification were tested in biodiesel production (Figure 7a). It can be clearly seen that biodiesel yield is remarkably dependent on the type of solvent. Overall, Triton-activated BCL mCLEAs exhibited higher activity than BCL mCLEAs and free BCL in all the solvents tested, and the changing trend of their activity in various solvents was accord with BCL mCLEAs and free BCL. In case of Triton-activated BCL mCLEAs, the best results were achieved using n-hexane with a yield of up to 94% biodiesel, which was 3.3-fold higher than that in free BCL catalyzed reaction. Interestingly, surfactant hyperactivation in combination with immobilization could fasten lipase in their active conformation, allowing biodiesel production performed in solvent without further addition of water, which was in accordance with earlier reports [29,30]. The methanol:oil molar ratio can have a significant effect on the reaction yield because excess methanol increases the reaction rate and drives high yield of biodiesel, while a high concentration of methanol leads to inactivation of lipases. In this study, experiments were performed at different molar ratios of methanol to WCO ranging from 3:1 (stoichiometric ratio) to 11:1 both in hexane and cyclohexane with methanol added only once. As shown in Figure 7b, Triton-activated BCL mCLEAs exhibited higher biodiesel yields in one-time addition of methanol under all the experimental conditions, especially when hexane was used as solvent. Meanwhile, owing to the significant inhibitory effect of excess methanol in one time addition, free BCL showed low activity in biodiesel production. It is worth noting that yields of Triton-activated BCL mCLEAs catalyzed reactions in hexane exceeded 90% in a wide range of methanol-to-WCO ratio, while the reaction yield of BCL mCLEAs catalyzed reaction decreased with methanol-to-WCO ratio exceeding 6:1. Consequently, it can be confirmed that surfactants pretreatment provided not only hyperactivation but also protection to lipases from denaturation in excess methanol. In addition, the maximum biodiesel yield was observed at a methanol-to-WCO ratio of 7:1 for Triton-activated BCL mCLEAs. Consequently, the minimal stoichiometric methanol-to-WCO ratio of 7:1 was chosen in further experiments.
Most of the enzymatic transesterification depends on temperature, which could enhance reaction rate and improve the dispersion of immobilized particles in reaction medium with better mass transfer between the reactants [31]. However, thermal denaturation of the enzyme might occur with elevation of temperature, typically according to the property of enzyme and immobilized methods. The effect of temperature on the yield of biodiesel during the transesterification of WCO has been investigated over a temperature range from 35 to 55 • C. According to Figure 7c, the optimum operational temperature for Triton-activated BCL mCLEAs and BCL mCLEAs was 40 • C, with biodiesel yields of 98% and 76% respectively after 24 h, and further increase of temperature will result in decrease of biodiesel yields simultaneously. Besides, Triton-activated BCL mCLEAs showed better activity below 40 • C, while BCL mCLEAs performed higher biodiesel yield over 40 • C. The suitable covalent cross-linking with functionalized MNPs provided extra structure stabilization in mCLEAs, requiring much more energy to the disruption of this stable structure than free enzyme [32]. Nevertheless, the accessible active site of lipases achieved by surfactants pretreatment might be more sensitive to high temperature denaturation [33].
In summary, the optimal reaction conditions for Triton-activated BCL mCLEAs catalyzed transesterification of WCO are as follows: hexane used as solvent, molar ratio of methanol-to-WCO 7:1 in one-time addition, reaction temperature 40 • C. To verify the feasibility of the whole process at a larger scale, transesterification of WCO were performed under optimal conditions adding proper amount of Triton-activated BCL mCLEAs (the initial content of BCL was 240 mg in immobilization) to a mixture of 1 g WCO in 20 mL hexane. The biodiesel yield reached 94% after shaking at 40 • C for 48 h. Triton-activated BCL mCLEAs showed good activity and stability under higher oil content, indicating the possibility of its scale-up application in bioreactor systems.

Reusability
Reusability of immobilized enzyme is a chief criterion for its cost-effective use for potential industry applications. The utilization of functionalized MNPs facilitates the consequent reuse of immobilized enzyme. To investigate the reusability of BCL mCLEAs and Triton-activated BCL mCLEAs, the immobilized lipases were recovered by magnetic separation, and applied in the consecutive batches of biodiesel reactions under optimized conditions. Assessments of the operational stability were analyzed for 6 cycles and presented in Figure 7d. As observed, Triton-activated BCL mCLEAs showed no significant loss in the catalytic activity after subsequent consecutive reuse for 4 cycles, and kept 82% relative activity after continuous running 5 cycles. Meanwhile, the relative activity of BCL mCLEAs was 55% after 5 cycles, implying that BCL could possess good long-term stability with surfactant pretreatment. The protein denaturation in one time addition of methanol and byproduct inhibition might be account for the decrease in biodiesel yield in long-term reuses [34].

Preparation of Magnetic Support
Preparation of HAP-coated MNPs was carried out according to the previously reported method [35]. Initially, MNPs cores were prepared by the conventional co-precipitation method. Typically, FeCl 2 ·4H 2 O (1.1 g) and (3.0 g) of FeCl 3 ·6H 2 O were dissolved in 90 mL deionized water under the protection of argon, with subsequent addition of 25% ammonia solution (30 mL) under vigorous stirring at room temperature. After stirring for 30 min, a 60 mL aqueous solution composed of Ca(NO 3 ) 2 ·4H 2 O (7.1 g) and (NH 4 ) 2 HPO 4 (2.3 g) adjusted to pH=11 was added drop wise to the above suspension under continuous stirring. Subsequently, the resultant mixture was heated to 90 • C and stirred for 2 h. After cooling to room temperature and aging in the mother solution overnight, the obtained precipitates were washed several times with deionized water until neutral and lyophilized for 12 h. The HAP-coated MNPs were obtained by calcining the materials in air at 300 • C for 3 h.
To obtain 3-aminopropyl trimetoxysilane functionalized HAP-coated MNPs (APTES-HAP-coated MNPs), HAP-coated MNPs (1.0 g) were suspended in a solution composed of 30 mL anhydrous toluene and 0.44 g of APTES. The mixture was refluxed under Ar atmosphere for 12 h, and then washed several times with ethanol, magnetically separated, and subsequently lyophilized prior to use.

Lipase Immobilization
BCL mCLEAs were produced according to the procedure described in Scheme 1. Firstly, 10 mg of APTES-HAP-MNPs were dispersed in 1 mL of BCL solution (10 mg/mL, 0.1 M phosphate buffer, pH 7.0) and shaken for 15 min at 30 • C. Then 5 mL of precipitant was added with stirring at 4 • C for 30 min. After precipitation, glutaraldyhyde was added drop wise into the suspension and stirred for 3 h at 30 • C. Afterwards, BCL mCLEAs were collected by centrifugation and washed thrice with phosphate buffer and deionized water, lyophilized and finally stored at 4 • C.
During optimization of the immobilization conditions, the effects of precipitants (acetone, ethanol, isopropanol, PEG 800 (1 g/mL), and saturated ammonium sulfate solution) and concentration of glutaraldehyde on the activity recovery of BCL mCLEAs were investigated.
The surfactant-activated BCL mCLEAs was prepared using cationic (CTAB), anionic (AOT) and nonionic (Tween 80 and Triton Triton X-100) surfactants at various concentrations. Then, 1 mL of BCL solution and appropriate amount of surfactant were mixed and stirred at 4 • C for 30 min. After incubated for 24 h at 4 • C, the suspended solution was sequentially used for BCL mCLEAs preparation under optimal conditions.

Characterization
The prepared support matrix and immobilized lipase described above were characterized using FTIR, SEM and VSM. The Fourier transform infrared (FTIR) spectra were acquired using a Perkin Elmer Frontier spectrometer (Spectrum Two, Waltham, MA, USA) equipped with an Attenuate Total Reflection (ATR) accessory. Samples were analyzed as KBr pellets in the range of 400 to 4000 cm −1 at a resolution of 0.5 cm −1 . The morphology of the particle surface was observed using a scanning electron microscope (SEM, Sigma HD, ZEISS, Germany), with deposition of a thin coating of gold onto the samples prior to analyses. The magnetic properties were detected by a vibrating sample magnetometer (VSM, MicroSense EZ9, Lowell, MA, USA) at room temperature.

Activity Assay
In studying the optimal conditions for BCL mCLEAs preparation, the enzymatic transesterification activities of free lipase and immobilized BCLs were assayed via transesterification reaction of 2-phenyl ethanol with vinyl acetate according to the method introduced previously [36]. The reaction mixture contained 10 mg of 2-phenylethanol, 1 mL of vinyl acetate and 10 mg of lipase (the initial content of BCL was 10 mg in preparing BCL CLEAs and BCL mCLEAs), and the reactions were carried out at 30 • C with continuous shaking at 220 rpm. After 24 h of reaction, samples were withdrawn and analyzed by high-performance liquid chromatography (HPLC). All experiments were repeated at least three times. The relative activity of BCL mCLEAs was calculated with the following equation: Relative activity (%) = Transesterification yield of immobilized BCL Trasesterification yield of free BCL × 100

Enzymatic Transesterification for Biodiesel Production
The transesterification of WCO were carried out at 40 • C in a 10 mL screw-capped vessel for 24 h with continuous shaking at 220 rpm. Unless otherwise stated, a typical reaction mixture consisted of 50 mg WCO, 2.0 mL hexane, 10 mg of lipase (the initial content of BCL was 10 mg in preparing BCL CLEAs and BCL mCLEAs) and methanol using methanol: oil molar ratio of 4:1. Single factor optimization was conducted to determine optimal reaction parameters for transesterification of WCO to biodiesel. Various conditions including kinds and concentration of surfactants, solvents, molar ratio of methanol to oil and temperature ( • C) were investigated. The transesterification reaction of large scale with 1 g WCO were carried out as described in Section 2.4. All biodiesel reactions were performed in dried solvents without any water added. The yield of biodiesel (20 µL) was analyzed in different time intervals using gas chromatography.
Fatty acid methyl esters (FAMEs) were analyzed by a Fuli9790 plus gas chromatography (Fuli, Zhejiang, China) fitted with a flame ionization detectorcity (FID, Zhejiang, China), and a KB-FFAP column (30 m × 0.32 mm × 0.25 µm). Nitrogen gas was a carrier at continues flow of 1.0 mL/min. The oven (Zhejiang, China) temperature was set and at 160 • C maintained for 2 min, then a heating ramp was applied up to 240 • C at a rate of 10 • C /min, and the temperature of the oven was maintained at 240 • C for 15 min. The temperatures of the injector (Zhejiang, China) and the detector (Zhejiang, China) were set at 270 and 280 • C, respectively. Methyl tridecanoate was used as internal standard, and the biodiesel yield (%) was calculated by peaks area of standard FAME peaks.

Reusability
The reusability of Triton-activated BCL mCLEAs and BCL mCLEAs for the transesterification of WCO were also investigated under optimal conditions. After each batch reaction, immobilized BCL was recovered by magnetic separation and washed with n-hexane. The washed biocatalyst was reused consecutively in repetitive cycles. The biodiesel yield of the first reaction was set as 100% and the FAMEs yield in the subsequent reactions was calculated accordingly.

Conclusions
A facile and effectual surfactant imprinting method to expose the lipase active site integrating amino functionalized HAP-coated MNPs was established to immobilize CLEAs of BCL attaining enhanced activity and stability. The as-prepared Triton-activated BCL mCLEAs was subsequent processed in enzymatic transesterification of waste cooking oil for biodiesel production, and showed 98% biodiesel yield under optimal conditions, which was 5.3-fold higher than the free lipase. This study proved that hyperactivation with surfactant could significantly improve the resistance of lipase to methanol in one-time addition, when compared to BCL mCLEAs and free BCL. In addition, surfactant imprinting in combination with immobilization could fasten lipase in their active conformation, allowing biodiesel production performed in solvent without further addition of water, and thus displayed priority in downstream purification of biodiesel over ordinary immobilization methods. Besides, the green immobilization with functionalized MNPs facilitates fast and easy recovery of lipase, and the corresponding immobilized BCL was reused for 4 cycles without significant loss in the catalytic activity. Furthermore, this work provides a promising approach for immobilization of other lipases, which can be used with success in green and clean production processes.