Efficient and Stable Magnetic Chitosan-Lipase B from Candida Antarctica Bioconjugates in the Enzymatic Kinetic Resolution of Racemic Heteroarylethanols

Lipase B from Candida antarctica immobilized by covalent binding on sebacoyl-activated chitosan-coated magnetic nanoparticles proved to be an efficient biocatalyst (49.2–50% conversion in 3–16 h and >96% enantiomeric excess) for the enzymatic kinetic resolution of some racemic heteroarylethanols through transesterification with vinyl acetate. Under optimal conditions (vinyl acetate, n-hexane, 45 °C), the biocatalyst remains active after 10 cycles.


Introduction
In the last decades enzymes have become popular catalysts, not only in academic research, but also in fine chemical production, lipases being by far the most popular enzymes for industrial applications [1]. Alongside their high chemo-, region-and stereoselectivity, lipases are able to transform efficient water-insoluble substrates, since their natural substrates-the triglycerides are also insoluble in water and are transformed at the oil-water interface. The high enantioselectivity shown in the kinetic and dynamic kinetic resolution of racemates turned lipases into powerful tools for the preparation of the enantiomers of pharmaceuticals and fine chemicals [2,3].
In industrial applications not only the catalytic activity, but also the catalyst operational stability is important, thus besides the modulation of enzyme activity and selectivity, the need to increase the stability of the enzymes and to transform them into robust, recyclable catalyst, operating during several catalytic cycles, has become crucial. Accordingly, many effective immobilization methods have been developed, using a large palette of materials as support [4].
Lipase B from Candida antarctica (CaL-B), first isolated from the sediments of Lake Vanda in Antarctica [5] possessing a high lipolytic activity [6], was intensively studied, important biomedical application and structural properties [7][8][9] being reported. Based on its high selectivity, commercial availability, high tolerance to organic solvents, temperature and large substrate tolerance, CaL-B is currently used in many industrial applications [1,2,10] and presents a high potential as catalyst in the kinetic resolution processes of secondary chiral alcohols [11], allowing the preparation of both stereoisomers in their enantiopure form.
Molecules 2019, 24, x FOR PEER REVIEW  3 of 15 heretoarylmethylketones 1a-d,g,i,j, while the heretoarylethanols rac-2e,f,h were obtained by Grignard reaction from the corresponding aldehydes 1e,f,h. Further, the racemic ethanols were used as substrates in the enzymatic transesterification reactions and also as starting materials for the synthesis of the corresponding racemic acetates rac-3a-j, used to set-up the chromatographic separation methods (Scheme 1) [16][17][18]54].

Covalent Immobilization of CaL-B on MNP-CS Supports
With the aim to provide an efficient lipase preparate, four approaches were used for the covalent immobilization of CaL-B on chitosan-coated MNP (MNP-CS), involving a variety of interactions between enzyme and the nano-support (Scheme 2): (A) the direct covalent binding of activated enzyme to the chitosan amino groups [32,[44][45][46] allows the preparation of the most simple nanobioconjugate, but the newly formed amidic bond can represent an important barrier for lipase mobility since its direct binding to the support generates protein conformational changes, negatively affecting the activity/selectivity; (B) the covalent binding through the residual epoxy groups of glycerol-diglycidyl ether (GDGE) cross-linker, covalently attached to amino groups of chitosan, that allows enzyme binding through a more flexible arm in comparison to the classical glutaraldehyde [48][49][50]; (C) the covalent binding through an alkyl-diamine linker previously bonded to the GDGE activated nanoparticle [51] that provides longer, even more flexible spacer-arm as compared with (B) and (D) covalent binding on the sebacoyl (SC)-derivatized MNP-CS via the activated ester method [52,53], the dicarboxylic acid halides acting as activating group in enzyme immobilization, while the hydrophobic alkyl chain was shown to be beneficial for maintaining the active conformation of the immobilized lipase [53].
Since the lipases' catalytic activity in organic solvents is not rigorously reflected by their hydrolytic activities determined in aqueous solution, Wang et al. [55] have been developed a colorimetric method based on the amount of released acetaldehyde when activated vinyl esters are used in transesterification reactions, named by them synthetic activity. Using this parameter, a first evaluation was performed. The obtained results for all new biocatalysts, for the lyophilized lipase and Novozym 435 as references are presented in Table 1.

Covalent Immobilization of CaL-B on MNP-CS Supports
With the aim to provide an efficient lipase preparate, four approaches were used for the covalent immobilization of CaL-B on chitosan-coated MNP (MNP-CS), involving a variety of interactions between enzyme and the nano-support (Scheme 2): (A) the direct covalent binding of activated enzyme to the chitosan amino groups [32,[44][45][46] allows the preparation of the most simple nanobioconjugate, but the newly formed amidic bond can represent an important barrier for lipase mobility since its direct binding to the support generates protein conformational changes, negatively affecting the activity/selectivity; (B) the covalent binding through the residual epoxy groups of glycerol-diglycidyl ether (GDGE) cross-linker, covalently attached to amino groups of chitosan, that allows enzyme binding through a more flexible arm in comparison to the classical glutaraldehyde [48][49][50]; (C) the covalent binding through an alkyl-diamine linker previously bonded to the GDGE activated nanoparticle [51] that provides longer, even more flexible spacer-arm as compared with (B) and (D) covalent binding on the sebacoyl (SC)-derivatized MNP-CS via the activated ester method [52,53], the dicarboxylic acid halides acting as activating group in enzyme immobilization, while the hydrophobic alkyl chain was shown to be beneficial for maintaining the active conformation of the immobilized lipase [53].
Since the lipases' catalytic activity in organic solvents is not rigorously reflected by their hydrolytic activities determined in aqueous solution, Wang et al. [55] have been developed a colorimetric method based on the amount of released acetaldehyde when activated vinyl esters are used in transesterification reactions, named by them synthetic activity. Using this parameter, a first evaluation was performed. The obtained results for all new biocatalysts, for the lyophilized lipase and Novozym 435 as references are presented in Table 1.  The synthetic activity of the new enzyme preparation bearing the sebacoyl group was much higher as compared with the lyophilized enzyme (Entries 8 and 1 in Table 1), probably due to the hydrophobic residue reducing the enzyme-enzyme interactions at the support surface. The most active biocatalyst, MNP-CS-SC-CaL-B (Entry 8 in Table 1), has a synthetic activity similar to that of the commercially available Novozym 435 (Entry 2 in Table) when comparing the synthetic activity reported on enzyme preparation weight (column 3 in Table 1). It is notable that despite its widespread use, Novozym 435 is unstable, and even even fragile under stirring, and through breakage of the biocatalyst particles, the enzyme can be released, an important loss of enzyme activity being often observed and reported [56,57], issues which can be alleviated through the use of robust covalent enzyme immobilizates.  The synthetic activity of the new enzyme preparation bearing the sebacoyl group was much higher as compared with the lyophilized enzyme (Entries 8 and 1 in Table 1), probably due to the hydrophobic residue reducing the enzyme-enzyme interactions at the support surface. The most active biocatalyst, MNP-CS-SC-CaL-B (Entry 8 in Table 1), has a synthetic activity similar to that of the commercially available Novozym 435 (Entry 2 in Table) when comparing the synthetic activity reported on enzyme preparation weight (column 3 in Table 1). It is notable that despite its widespread use, Novozym 435 is unstable, and even even fragile under stirring, and through breakage of the biocatalyst particles, the enzyme can be released, an important loss of enzyme activity being often observed and reported [56,57], issues which can be alleviated through the use of robust covalent enzyme immobilizates.

Enzyme Catalyzed Transesterification Studies
As known, in the lipase catalyzed transformation of secondary alcohols, the individual enantiomers are transformed with different reaction rates and in accordance to the Kazlauskas-rule the R-esters are formed faster, while the slow reacting S-ethanols remain untransformed [58].
First, in order to confirm the results obtained for the efficiency of the new CaL-B preparations based on the synthetic activity, the new biocatalysts (containing the same quantity of CaL-B) were tested in the EKR of the racemic 1-phenyl-1-ethanol rac-2a chosen as model substrate, using vinyl acetate as irreversible acyl donor, since the formed vinyl alcohol spontaneous tautomerized to acetaldehyde, favoring the product formation (Scheme 1b).
The conversion c and enantiomeric excesses ee P and ee S (see Section 3) were used as measure of biocatalyst activity and selectivity, in order to quantify the biocatalyst efficiency.
The obtained results ( Figure 1) are in accordance with the activity study; as consequence MNP-CS-SC-CaL-B was chosen as optimal biocatalyst, allowing the maximum conversion (≈50%) in shorter reaction time (only 13 h) and high enantiomeric excesses (ee S , ee P > 99%).
Molecules 2019, 24, x FOR PEER REVIEW 5 of 15 As known, in the lipase catalyzed transformation of secondary alcohols, the individual enantiomers are transformed with different reaction rates and in accordance to the Kazlauskas-rule the R-esters are formed faster, while the slow reacting S-ethanols remain untransformed [58].
First, in order to confirm the results obtained for the efficiency of the new CaL-B preparations based on the synthetic activity, the new biocatalysts (containing the same quantity of CaL-B) were tested in the EKR of the racemic 1-phenyl-1-ethanol rac-2a chosen as model substrate, using vinyl acetate as irreversible acyl donor, since the formed vinyl alcohol spontaneous tautomerized to acetaldehyde, favoring the product formation (Scheme 1b).
The conversion c and enantiomeric excesses eeP and eeS (see Section 3) were used as measure of biocatalyst activity and selectivity, in order to quantify the biocatalyst efficiency.
The obtained results ( Figure 1) are in accordance with the activity study; as consequence MNP-CS-SC-CaL-B was chosen as optimal biocatalyst, allowing the maximum conversion (≈50%) in shorter reaction time (only 13 h) and high enantiomeric excesses (eeS, eeP > 99%). Since the temperature has an important effect on the reaction rate and based on the known high termostability of CaL-B, the EKR of racemic 1-phenyl-1-ethanol rac-2a with vinyl acetate in n-hexane was performed at temperature gradually increased with 5 °C between 30 and 60 °C. Samples were taken after 2 h and analyzed on HPLC. A continuous increased was observed until 45 °C; afterwards the conversion decreased ( Figure 2). The obtained results are in concordance with the literature data, reporting an increased termostability of CaL-B bioconjugates as compared with the free enzyme [59]. As consequence 45 °C was chosen as optimal temperature, all further experiments were performed at this temperature. Since the temperature has an important effect on the reaction rate and based on the known high termostability of CaL-B, the EKR of racemic 1-phenyl-1-ethanol rac-2a with vinyl acetate in n-hexane was performed at temperature gradually increased with 5 • C between 30 and 60 • C. Samples were taken after 2 h and analyzed on HPLC. A continuous increased was observed until 45 • C; afterwards the conversion decreased ( Figure 2). The obtained results are in concordance with the literature data, reporting an increased termostability of CaL-B bioconjugates as compared with the free enzyme [59]. As consequence 45 • C was chosen as optimal temperature, all further experiments were performed at this temperature. As known, in the lipase catalyzed transformation of secondary alcohols, the individual enantiomers are transformed with different reaction rates and in accordance to the Kazlauskas-rule the R-esters are formed faster, while the slow reacting S-ethanols remain untransformed [58].
First, in order to confirm the results obtained for the efficiency of the new CaL-B preparations based on the synthetic activity, the new biocatalysts (containing the same quantity of CaL-B) were tested in the EKR of the racemic 1-phenyl-1-ethanol rac-2a chosen as model substrate, using vinyl acetate as irreversible acyl donor, since the formed vinyl alcohol spontaneous tautomerized to acetaldehyde, favoring the product formation (Scheme 1b).
The conversion c and enantiomeric excesses eeP and eeS (see Section 3) were used as measure of biocatalyst activity and selectivity, in order to quantify the biocatalyst efficiency.
The obtained results ( Figure 1) are in accordance with the activity study; as consequence MNP-CS-SC-CaL-B was chosen as optimal biocatalyst, allowing the maximum conversion (≈50%) in shorter reaction time (only 13 h) and high enantiomeric excesses (eeS, eeP > 99%). Since the temperature has an important effect on the reaction rate and based on the known high termostability of CaL-B, the EKR of racemic 1-phenyl-1-ethanol rac-2a with vinyl acetate in n-hexane was performed at temperature gradually increased with 5 °C between 30 and 60 °C. Samples were taken after 2 h and analyzed on HPLC. A continuous increased was observed until 45 °C; afterwards the conversion decreased ( Figure 2). The obtained results are in concordance with the literature data, reporting an increased termostability of CaL-B bioconjugates as compared with the free enzyme [59]. As consequence 45 °C was chosen as optimal temperature, all further experiments were performed at this temperature. The nature of the solvent can play an important role in the enzyme efficiency, since the enzyme conformation strongly depends on the reaction media. Five nonpolar and polar aprotic solvents were used in a similar manner for the model substrate. The obtained experimental data confirm as optimal solvent the n-hexane (42% conversion after 6 h, as compared with approx. 12% conversion obtained in toluene and under 2% in the more polar tetrahydrofuran, methyl tert-butyl ether and dichloromethane).
Next, aiming to enhance the productivity of EKR process, the effect of substrate-enzyme ratio was studied using similar conditions (5 mg of substrate and 2 equiv. of vinyl acetate in n-hexane-1 mL, at 45 • C and 1000 rpm), modifying the quantity of enzyme. Samples were taken after 6 h and analyzed on HPLC. The conversion increases from 12% at 15:1 ratio (w rac-2a /w biocatalyst ) as compared to approx. 45% at 5:1 ratio, decreasing to 30% when a 2.5:1 ratio was used. The optimal substrate-enzyme ratio established in conclusion to 5:1 was used in further experiments.
Since the nature of the acyl donor could also significantly influence the selectivity of the enantiomer selective acylation of secondary alcohols catalyzed by CaL-B [60], the acylation of rac-2a was next performed in the already established optimal conditions (n-hexane, at 45 • C, 5:1 ratio and 1000 rpm) with three activated vinyl esters: vinyl acetate, butyrate and decanoate (2 equiv.). Vinyl acetate proved to be the best acylation agent, since the maximum conversion (50%) was reached in the shortest reaction time (13 h, Table 2, Entry 1). Further the effect of the substrate-vinyl acetate molar ratio was investigated using the same amount of rac-2a (5 mg) and biocatalyst CaL-B (1 mg), in n-hexane (1 mL), at 45 • C. Samples were taken after 13 h and analyzed on HLPC. The highest activity (50% conversion after 13 h) and selectivity (ee S , ee P > 99.9%) resulted at 1:2 and 1:4 molar ratio with vinyl acetate (Figure 3). The nature of the solvent can play an important role in the enzyme efficiency, since the enzyme conformation strongly depends on the reaction media. Five nonpolar and polar aprotic solvents were used in a similar manner for the model substrate. The obtained experimental data confirm as optimal solvent the n-hexane (42% conversion after 6 h, as compared with approx. 12% conversion obtained in toluene and under 2% in the more polar tetrahydrofuran, methyl tert-butyl ether and dichloromethane).
Next, aiming to enhance the productivity of EKR process, the effect of substrate-enzyme ratio was studied using similar conditions (5 mg of substrate and 2 equiv. of vinyl acetate in n-hexane-1 mL, at 45 °C and 1000 rpm), modifying the quantity of enzyme. Samples were taken after 6 h and analyzed on HPLC. The conversion increases from 12% at 15:1 ratio (wrac-2a/wbiocatalyst) as compared to approx. 45% at 5:1 ratio, decreasing to 30% when a 2.5:1 ratio was used. The optimal substrateenzyme ratio established in conclusion to 5:1 was used in further experiments.
Since the nature of the acyl donor could also significantly influence the selectivity of the enantiomer selective acylation of secondary alcohols catalyzed by CaL-B [60], the acylation of rac-2a was next performed in the already established optimal conditions (n-hexane, at 45 °C, 5:1 ratio and 1000 rpm) with three activated vinyl esters: vinyl acetate, butyrate and decanoate (2 equiv.). Vinyl acetate proved to be the best acylation agent, since the maximum conversion (50%) was reached in the shortest reaction time (13 h, Table 2, Entry 1). Further the effect of the substrate-vinyl acetate molar ratio was investigated using the same amount of rac-2a (5 mg) and biocatalyst CaL-B (1 mg), in n-hexane (1 mL), at 45 °C. Samples were taken after 13 h and analyzed on HLPC. The highest activity (50% conversion after 13 h) and selectivity (eeS, eeP > 99.9%) resulted at 1:2 and 1:4 molar ratio with vinyl acetate (Figure 3).

Analytical-Scale Lipase-Mediated O-Acylation of Racemic Heteroarylethanols rac-2a-j
The substrate domain for the new biocatalyst was investigated next using the determined optimal conditions (Scheme 1b, Table 3). Samples were taken periodically from the reaction mixtures, diluted with n-hexane and analyzed on appropriate HPLC chiral columns (Table 3). Very good results (conversions > 49% and high optical purities, eeP > 99%, eeS > 96%, E > 200) [61], were obtained for all tested substrates, demonstrating the excellent catalytic properties and a large substrate domain of MNP-CS-SC-CaL-B.

Analytical-Scale Lipase-Mediated O-Acylation of Racemic Heteroarylethanols rac-2a-j
The substrate domain for the new biocatalyst was investigated next using the determined optimal conditions (Scheme 1b, Table 3). Samples were taken periodically from the reaction mixtures, diluted with n-hexane and analyzed on appropriate HPLC chiral columns (Table 3). Very good results (conversions > 49% and high optical purities, ee P > 99%, ee S > 96%, E > 200) [61], were obtained for all tested substrates, demonstrating the excellent catalytic properties and a large substrate domain of MNP-CS-SC-CaL-B.

Reuse Experiments.
The reusability of the best performing biocatalyst, MNP-CS-SC-CaL-B, was finally studied in the enantioselective acylation of rac-2e ( Figure 4A) since it represents an important requirement for any application, allowing the development of sustainable processes.

Reuse Experiments.
The reusability of the best performing biocatalyst, MNP-CS-SC-CaL-B, was finally studied in the enantioselective acylation of rac-2e ( Figure 4A) since it represents an important requirement for any application, allowing the development of sustainable processes.  The reaction was performed 10 times with the same biocatalyst as previously described. After each cycle, the enzyme preparate was washed with n-hexane (3 × 0.5 mL) and immediately reuse in the next cycle. The reaction was performed 10 times with the same biocatalyst as previously described. After each cycle, the enzyme preparate was washed with n-hexane (3 × 0.5 mL) and immediately reuse in the next cycle.
The recycling capacity, but also the long term stability of the enzyme preparate was studied similar to the EKR of 1-(10-methyl-10H-phenothiazin-3-yl)ethan-1-ol rac-2i ( Figure 4B), taking samples after 10 h in order to rich higher conversions, closed to 50%.
As can be observed in Figure 4A, B the activity remains high in both cases even after 10 cycles, decreasing by less than 5%. As consequence, the operational and long term stability of the new prepared immobilized lipase is promising for further experiments in a continuous-flow system, allowing higher productivity.

Materials and Methods
Lipase B from Candida antarctica (CaL-B) was purchased as solution from Novozymes (Copenhagen, Denmark). Medium molecular weight chitosan   Solvents used in the chemical synthesis were used as purchased (methanol, diethyl ether) or dried over molecular sieve (dichloromethane).
Chiral HPLC (Agilent 1200 Series instrument, Agilent, Waldbronn, Germany) was used for the quantitative analysis of the enzymatic kinetic resolution products, samples being taken periodically and investigated by chiral chromatography.
The enantiomeric excesses of the substrate and of the product (ee S , ee P ) were calculated from peak areas of HPLC chromatograms. The enantiomeric ratio (E) was calculated with the known Equation (1) using the conversion c calculated with the equation (2) as well-known [61]: For the quantitative spectrophotometric determinations of enzyme load through BCA method, an Agilent 8453 UV-Vis spectrophotometer equipped with thermostat was used. An Alpha 1-2 LO plus lyophilizer (Gore, Felkirchen-Westerham, Germany) was used to dry the intermediate materials and newly obtained enzyme preparations.

Covalent
(D) The fourth type was obtained after MNP-CS derivatization with sebacoyl chloride (SC) (Scheme 2D). The remained free carboxylic groups generated after hydrolysis were finally conjugated with CaL-B through the active ester method using EDAC and NHS. MNP-CS (50 mg) were suspended in n-hexane (2 mL) containing SC (320 µL) and stirred for 10 min. After magnetic separation, the preparation (MNP-CS-SC) was washed with n-hexane (3 × 1 mL) for removing the unreacted SC and dried at room temperature. The complete hydrolysis of the unreacted acyl chloride groups was performed immersing the MNP-CS-SC in deionized water at 37 • C for 30 min. The MNP-CS-SC were submerged in saline phosphate buffered (PBS, 50 mM, 100 mM NaCl, pH 6.4, 2 mL) containing EDAC (0.1 mol, 15.5 mg) and NHS (0.1 mol, 11.5 mg) with gentle shaking at room temperature for 1 h. The resulted activated nanoparticles were separated and washed with deionized water (3 × 1 mL). Next, the activated magnetic nanoparticles were immersed in PBS (50 mM, pH 6.4, 2 mL) containing CaL-B (3 mL, 7.47 mg/mL) and left under stirring overnight. In order to completely remove the physically adsorbed enzyme, the biocatalyst was separated and washed with distilled water (5 × 1 mL) and finally vacuum freeze dried for 24 h (Scheme 2D) [52,53].

CaL-B Load on MNP-CS Based Supports
The immobilization yields and the enzyme load were calculated for each case using the protein content of the initial protein solution, of the separated filtrate and of all unified washing solutions, determined spectrophotometrically with the BCA assay (see Table 4).

The Protocol for the Synthetic Activity
A fast and sensible colorimetric method was used for the synthetic activity of the new biocatalyst evaluation. The method is based on n-butanol transesterification with vinyl acetate (2 equiv.) in n-hexane at 30 • C and 1000 rpm. The released acetaldehyde was determined after derivatization with 3-methyl-2-benzothalininone hydrazone (MBTH) as hydrochloride salt, when the formed aldazine is further converted by oxidative coupling with another MBTH molecule in the presence of ammonium iron(III) sulfate (NH 4 Fe(SO 4 ) 2 × 12H 2 O) into a blue tetraazapentamethylene-cyanine (TAPMC) with a maximum absorption at 598 nm (Scheme 3) [55].

CaL-B Load on MNP-CS Based Supports
The immobilization yields and the enzyme load were calculated for each case using the protein content of the initial protein solution, of the separated filtrate and of all unified washing solutions, determined spectrophotometrically with the BCA assay (see Table 4). A fast and sensible colorimetric method was used for the synthetic activity of the new biocatalyst evaluation. The method is based on n-butanol transesterification with vinyl acetate (2 equiv.) in n-hexane at 30 °C and 1000 rpm. The released acetaldehyde was determined after derivatization with 3-methyl-2-benzothalininone hydrazone (MBTH) as hydrochloride salt, when the formed aldazine is further converted by oxidative coupling with another MBTH molecule in the presence of ammonium iron(III) sulfate (NH4Fe(SO4)2 × 12H2O) into a blue tetraazapentamethylenecyanine (TAPMC) with a maximum absorption at 598 nm (Scheme 3) [55]. The test transesterification reaction using as substrate n-butanol (9.14 µL), vinyl acetate (2 Eq.) as acylation agent, 1 mg of enzymatic preparation and n-hexane (1 mL) as solvent, was performed at 30 °C for 5 min. A sample of 5 µL was taken from the reaction mixture, mixed with MBTH solution (1% in dH2O, 400 µL) and double distilled water (395 µL). The mixture was left for 10 min at 30 °C and H4FeNO4S2 × 12H2O solution (1% in HCl 10%, 200 µL) was added; the resulted mixture was stirred for another 30 min at the same temperature and the absorbance at 598 nm was finally measured [55].

General Protocol for Enzymatic O-Transesterification
In order to find the best performing biocatalyst, all new enzyme preparates were tested on racemic 1-phenyl-1-ethanol rac-2a as model substrate. The EKR process was further optimized regarding the solvent, acylation agent, quantity of enzyme, quantity of acylation agent and temperature. The substrate domain in the optimal conditions was finally established. The test transesterification reaction using as substrate n-butanol (9.14 µL), vinyl acetate (2 Eq.) as acylation agent, 1 mg of enzymatic preparation and n-hexane (1 mL) as solvent, was performed at 30 • C for 5 min. A sample of 5 µL was taken from the reaction mixture, mixed with MBTH solution (1% in dH 2 O, 400 µL) and double distilled water (395 µL). The mixture was left for 10 min at 30 • C and H 4 FeNO 4 S 2 × 12H 2 O solution (1% in HCl 10%, 200 µL) was added; the resulted mixture was stirred for another 30 min at the same temperature and the absorbance at 598 nm was finally measured [55].

General Protocol for Enzymatic O-Transesterification
In order to find the best performing biocatalyst, all new enzyme preparates were tested on racemic 1-phenyl-1-ethanol rac-2a as model substrate. The EKR process was further optimized regarding the solvent, acylation agent, quantity of enzyme, quantity of acylation agent and temperature. The substrate domain in the optimal conditions was finally established.

Enzyme Catalyzed Transesterification Studies
Determination of the Optimal Biocatalyst The new biocatalysts were tested in the EKR of the racemic 1-phenyl-1-ethanol rac-2a chosen as model substrate (5 mg) with vinyl acetate (2 equiv.) as acylation agent, using n-hexane (1 mL) as solvent and the quantity of each biocatalysts containing 1 mg of CaL-B, at 50 • C and 1000 rpm. Samples were taken every 2 h and analyzed on HPLC (Figure 1).

Establishing of the Optimal Temperature
The EKR of 1-phenyl-1-ethanol rac-2a (5 mg) with vinyl acetate (2 equiv.) in n-hexane (1 mL) in presence of the biocatalyst (3.63 mg enzyme preparate containing 1 mg of lipase) was next performed at temperature gradually increased with 5 • C between 30 • C and 60 • C at 1000 rpm. Samples were taken after 2 h and analyzed on HPLC ( Figure 2).

Determination of the Optimal Solvent
The mixtures of the substrate (rac-2a, 5 mg), vinyl acetate (2 equiv), biocatalyst (3.63 mg enzyme preparation containing 1 mg of lipase) and the tested solvent (1 mL) were shaken (1000 rpm) in reaction vials at the optimal temperature (45 • C). Samples were withdrawn after 6 h, diluted, filtered and analyzed through HPLC.

Optimal Substrate-Enzyme Ratio
The effect of substrate-enzyme ratio was studied using 5 mg of rac-2a as substrate, 2 equiv. of vinyl acetate in n-hexane (1 mL), at 45 • C and 1000 rpm), modifying the quantity of enzyme. Samples were taken after 6 h and analyzed on HPLC.

Optimal Acyl Donor
The acylation of rac-2a (5 mg) was next performed in the optimal established conditions with three vinyl esters: vinyl acetate, vinyl butyrate and vinyl decanoate (2 equiv.) in n-hexane (1 mL) with the optimal biocatalyst (3.63 mg enzyme preparation containing 1 mg of lipase) at 45 • C and 1000 rpm ( Table 2).

General Protocol for Recycling Experiments
The reusability and the stability of the optimal enzyme preparation was further tested in reuse experiments in the EKR of rac-2e and rac-2i, using 5 mg of substrates, 3.63 mg of MNP-CS-SC-CaL-B (containing 1 mg enzyme), 1 mL of n-hexane as solvent, 2 equiv. of vinyl acetate at 45 • C and 1000 rpm. Samples were taken after 1 h for rac-2e and after 10 h for rac-2i. After each cycle the enzyme preparation was washed with n-hexane (3 × 0.5 mL) and immediately reused in the next cycle ( Figure 4A,B).

Conclusions
In this study, the covalent immobilization of CaL-B on chitosan-coated magnetic nanoparticles through covalent binding was achieved. The best performing enzyme preparation, with a sebacoyl group for support activation and as spacer arm, proved to be active and selective in the EKR of various 1-heteroarylethanols. Moreover, based on its high operational stability the new biocatalyst represents a promising candidate for further continuous flow experiments, in order to achieve higher productivities.