Modiﬁcation of Silica Xerogels with Polydopamine for Lipase B from Candida antarctica Immobilization

: Silica xerogels have been proposed as a potential support to immobilize enzymes. Improving xerogels’ interactions with such enzymes and their mechanical strengths is critical to their practical applications. Herein, based on the mussel-inspired chemistry, we demonstrated a simple and highly effective strategy for stabilizing enzymes embedded inside silica xerogels by a polydopamine (PDA) coating through in-situ polymerization. The modiﬁed silica xerogels were characterized by scanning and transmission electron microscopy, Fourier tranform infrared spectroscopy, X-ray diffraction, X-ray photoelectron spectroscopy and pore structure analyses. When the PDA-modiﬁed silica xerogels were used to immobilize enzymes of Candida antarctica lipase B (CALB), they exhibited a high loading ability of 45.6 mg/g support , which was higher than that of immobilized CALB in silica xerogels (28.5 mg/g support ). The immobilized CALB of the PDA-modiﬁed silica xerogels retained 71.4% of their initial activities after 90 days of storage, whereas the free CALB retained only 30.2%. Moreover, compared with the immobilization of enzymes in silica xerogels, the mechanical properties, thermal stability and reusability of enzymes immobilized in PDA-modiﬁed silica xerogels were also improved signiﬁcantly. These advantages indicate that the new hybrid material can be used as a low-cost and effective immobilized-enzyme support. and SiO 2 –CH 3 were investigated in detail. The results show that SiO 2 –CH 3 a signiﬁcant CALB-embedding ability. Compared with the SiO 2 –CH 3 the results showed that PDA-modiﬁed SiO 2 –CH 3 had better mechanical properties, thermal stability, storage stability and reusability. This indicated that the new hybrid silica xerogel could be used as a low-cost and relatively effective immobilized-enzyme support.


Introduction
Biocatalysts play a vital role in various scientific fields due to their unique advantages, such as high substrate specificity, outstanding catalytic ability and mild reaction conditions. Biocatalysis, applied in ester synthesis, is useful and its synthetic products can be identical to natural products. Recently, a transesterification reaction catalyzed by lipase (triacylglycerol ester hydrolase, EC 3.1.1.3) has been performed to produce esters [1]. Lipase-catalyzed reactions have been applied to the synthesis of chiral drugs [2], wax esters [3], structural lipids [4] and biodiesel [5]. However, the main bottlenecks of enzyme application are its low thermal stability, poor operational stability and the difficulty of reusing enzymes. Therefore, significant effort has been devoted to exploiting immobilization strategies to stabilize enzymes and endow them with greater stability and reusability [6,7].
In immobilizing enzymes, it is necessary to select an appropriate support material, which can improve the properties of enzymes. Lipases are widely recognized to have a hydrophobic domain [8]. The hydrophobic immobilization of a lipase can act upon its domains, to increase its activity and stability, by interfacial activation [9,10]. Thus, materials comprised with ordered mesoporous organosilica, in which organic hydrophobic groups are homogeneously distributed within their frameworks, may be ideal supports for lipase immobilization. Silica xerogel, thanks to a high specific surface area, good mechanical strength, inertness and stability at high temperature, has attracted much attention in had better mechanical properties, thermal stability, storage stability and reusability. This indicated that the new hybrid silica xerogel could be used as a low-cost and relatively effective immobilized-enzyme support. Figure 1a shows a FTIR spectral comparison of SiO2-CH3-CALB and SiO2-CH3-CALB@PDA. The adsorption peaks at 777 cm −1 and 445 cm −1 corresponded to the Si−O−Si group [29]. The band at 1277 cm −1 was assigned to the characteristic peak of Si-CH3, which proved that MTMS successfully deposited a methyl polymer layer on the silica surface. Other bands, at 3388 cm −1 and 1643 cm −1 , belonged to the stretching and bending vibrations of −OH [30]. After modification by polydopamine, the vibration absorption peak at 3388 cm −1 was significantly enhanced and widened, which was related to the catechol composition of polydopamine [31]. In addition, SiO2-CH3-CALB@PDA showed a peak at 1508 cm −1 , which was ascribed to the bending vibrations of indolequinone groups [32]. The XRD patterns of SiO2-CH3-CALB and SiO2-CH3-CALB@PDA are illustrated in Figure  1b. There was a relatively wide peak at 2θ = 22°, which is characteristic of amorphous silica [33]. The peak at 2θ < 10° was due to the siloxane network and the xerogel's structure, composed of ordered organic layers [34]. After modification by polydopamine, the intensity of the characteristic peak at 10° became weak, implying the microstructure of SiO2-CH3-CALB had changed due to the uniformly distributed deposition of polydopamine within the structure of the xerogel [35]. These results indicate that a polydopamine coating formed on the silica xerogel through self-polymerization. Microstructural images of SiO2-CH3-CALB@PDA are shown in Figure 2. Figure 2a,b shows the SEM images of the SiO2-CH3-CALB and SiO2-CH3-CALB@PDA prepared in this work, respectively by panel. They were constituted by the agglomeration of many silica clusters in uniform shape. Compared with SiO2-CH3-CALB, the surface of SiO2-CH3-CALB@PDA was rougher and looser between clusters, indicating that a polydopamine layer had formed on the Si−O−Si surface. In this structure, the formation of a protective enzyme barrier can absorb and disperse most of the energy from external forces, preventing the xerogel from breaking [36]. TEM images of monodispersed SiO2-CH3-CALB@PDA showed that its particles have a relatively uniform, nano-scale size ( Figure  2c); a more intuitive expression is shown in Figure 2d. The shape of the SiO2-CH3- Microstructural images of SiO 2 -CH 3 -CALB@PDA are shown in Figure 2. Figure 2a,b shows the SEM images of the SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA prepared in this work, respectively by panel. They were constituted by the agglomeration of many silica clusters in uniform shape. Compared with SiO 2 -CH 3 -CALB, the surface of SiO 2 -CH 3 -CALB@PDA was rougher and looser between clusters, indicating that a polydopamine layer had formed on the Si−O−Si surface. In this structure, the formation of a protective enzyme barrier can absorb and disperse most of the energy from external forces, preventing the xerogel from breaking [36]. TEM images of monodispersed SiO 2 -CH 3 -CALB@PDA showed that its particles have a relatively uniform, nano-scale size ( Figure 2c); a more intuitive expression is shown in Figure 2d. The shape of the SiO 2 -CH 3 -CALB@PDA particles irregularly spherical. Additionally, the SiO 2 -CH 3 -CALB@PDA surface had openframework channels (Figure 2e) that facilitated the diffusion of the substrate and product molecules [37]. Elemental mapping analysis (Figure 2g−j) demonstrated that PDA was uniformly distributed within the xerogel (as these contained nitrogen), and oxygen, carbon and silicon were also found in the SiO 2 -CH 3 -CALB@PDA surface. Notably, the oxygen content was high. CALB@PDA particles irregularly spherical. Additionally, the SiO2-CH3-CALB@PDA surface had open-framework channels (Figure 2e) that facilitated the diffusion of the substrate and product molecules [37]. Elemental mapping analysis (Figure 2g−j) demonstrated that PDA was uniformly distributed within the xerogel (as these contained nitrogen), and oxygen, carbon and silicon were also found in the SiO2-CH3-CALB@PDA surface. Notably, the oxygen content was high.  (Figure 3c), two peaks were observed at 531.9 eV and 532.7 eV, respectively, which belonged to O atoms of polydopamine in the form of quinone and catechol [38]. The high-resolution spectra of N 1's peak are shown in Figure 3d. The main peak, at 399.4 eV, indicated the existence of R2NH and RNH2, while the peak at 401.2 eV was attributed to R3N [39]. This result indicates that an adhesive polydopamine coating formed on the surface of the Si−O−Si network structure of the xerogel by self-polymerization.  (Figure 3c), two peaks were observed at 531.9 eV and 532.7 eV, respectively, which belonged to O atoms of polydopamine in the form of quinone and catechol [38]. The high-resolution spectra of N 1's peak are shown in Figure 3d. The main peak, at 399.4 eV, indicated the existence of R 2 NH and RNH 2 , while the peak at 401.2 eV was attributed to R 3 N [39]. This result indicates that an adhesive polydopamine coating formed on the surface of the Si−O−Si network structure of the xerogel by self-polymerization.   (Figure 4a,d). Then, they entered the gel phase, the SiO2-CH3-CALB showed a milky white gel block, while the SiO2-CH3-CALB@PDA showed a black transparent gel block (Figure 4b,e). This may be explained by dopamine's having begun to self-polymerize into polydopamine on the gel network. After the final drying stage, The final samples of SiO2-CH3-CALB and SiO2-CH3-CALB@PDA were obtained by grinding (Figure 4c,f).

Characterization
With the aim of further prove the PDA can delay shrinkage of xerogel, analysis of BET of SiO2-CH3-CALB and SiO2-CH3-CALB@PDA were taken into account. It can be seen from their adsorption-desorption curves, in Figure 4g, that they had strong interaction with N2 at low pressure and presented typical Langmuir type IV curves. The H2 hysteresis loops were also observed, indicating the mesoporous structure and the characteristics of 'ink bottle' pores [40]. From the pore size distribution curve in Figure 4h, it can be seen that the pore size (15.02 nm) and BET surface area (165.84 m 2 ·g −1 ) of SiO2-CH3-CALB@PDA were larger than those of SiO2-CH3-CALB (13.52 nm and 121.67 m 2 ·g −1 ), which we believe to be due to the PDA coatings and deposits on the surface of the Si−O−Si network structure during the sol-gel process, delaying the gel shrinkage [41].
The immobilization capacity of SiO2-CH3-CALB@PDA was evaluated by investigating the CALB loading. As shown in Figure 4i, the amount of CALB immobilized on SiO2-CH3-CALB@PDA increased with increasing CALB concentration. When the CALB concentration was 14.5 mg/mL, the CALB loading increased to 45.6 mg/g. However, when the With the aim of further prove the PDA can delay shrinkage of xerogel, analysis of BET of SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA were taken into account. It can be seen from their adsorption-desorption curves, in Figure 4g, that they had strong interaction with N 2 at low pressure and presented typical Langmuir type IV curves. The H2 hysteresis loops were also observed, indicating the mesoporous structure and the characteristics of 'ink bottle' pores [40]. From the pore size distribution curve in Figure 4h, it can be seen that the pore size (15.02 nm) and BET surface area (165.84 m 2 ·g −1 ) of SiO 2 -CH 3 -CALB@PDA were larger than those of SiO 2 -CH 3 -CALB (13.52 nm and 121.67 m 2 ·g −1 ), which we believe to be due to the PDA coatings and deposits on the surface of the Si−O−Si network structure during the sol-gel process, delaying the gel shrinkage [41].
The immobilization capacity of SiO 2 -CH 3 -CALB@PDA was evaluated by investigating the CALB loading. As shown in Figure 4i, the amount of CALB immobilized on SiO 2 -CH 3 -CALB@PDA increased with increasing CALB concentration. When the CALB concentration was 14.5 mg/mL, the CALB loading increased to 45.6 mg/g. However, when the CALB concentration was more than 14 mg/mL, a decline in the activity recovery of the immobilized CALB was observed. The loading reached a maximum at a high enzyme concentration (~16 mg/mL), and there is a slightly continuous decrease in the enzyme activity when the enzyme concentration exceeds 14.5 mg/mL. This can be explained by the fact that excess CALB loading will easily lead to the congestion of the enzyme molecules. Therefore, the resulting spatial constraint can increase the mass transfer resistance of the substrate and product, which is expressed as reducing activity [42]. Therefore, the optimum CALB concentration was chosen as 14.5 mg/mL. In this case, the CALB loading is efficient (activity recovery higher than 93%) without sacrificing excess enzyme to unnecessary use. Meanwhile, compared with the enzyme loading of 28.5 mg/g on pristine SiO 2 -CH 3 -CALB at an initial CALB concentration of 14.5 mg/mL, the enzyme loading on SiO 2 -CH 3 -CALB@PDA reached as high as 45.6 mg/g, nearly twice as high as that on SiO 2 -CH 3 -CALB. As mentioned above, the modification of PDA provided a barrier for the enzyme, and covalent linking enhanced the interaction between the enzyme and the support, effectively preventing enzyme leakage.
Catalysts 2021, 11, x FOR PEER REVIEW 6 of 15 concentration (~16 mg/mL), and there is a slightly continuous decrease in the enzyme activity when the enzyme concentration exceeds 14.5 mg/mL. This can be explained by the fact that excess CALB loading will easily lead to the congestion of the enzyme molecules. Therefore, the resulting spatial constraint can increase the mass transfer resistance of the substrate and product, which is expressed as reducing activity [42]. Therefore, the optimum CALB concentration was chosen as 14.5 mg/mL. In this case, the CALB loading is efficient (activity recovery higher than 93%) without sacrificing excess enzyme to unnecessary use. Meanwhile, compared with the enzyme loading of 28.5 mg/g on pristine SiO2-CH3-CALB at an initial CALB concentration of 14.5 mg/mL, the enzyme loading on SiO2-CH3-CALB@PDA reached as high as 45.6 mg/g, nearly twice as high as that on SiO2-CH3-CALB. As mentioned above, the modification of PDA provided a barrier for the enzyme, and covalent linking enhanced the interaction between the enzyme and the support, effectively preventing enzyme leakage.  Figure 5 shows the internal microstructure of the SiO2-CH3-CALB-and SiO2-CH3-CALB@PDA-immobilized enzyme and the mechanism of the polydopamine-modified immobilized enzyme. The Si−O−Si polymer network skeleton was obtained by hydrolysis and a condensation reaction with TMOS and MTMS as co-precursors, and the enzyme molecules were embedded in the Si−O−Si network. In hydrolysis reaction, the methyl group of MTMS was not involved in hydrolysis, replacing and cross-linking with hy-  Figure 5 shows the internal microstructure of the SiO 2 -CH 3 -CALB-and SiO 2 -CH 3 -CALB@PDA-immobilized enzyme and the mechanism of the polydopamine-modified immobilized enzyme. The Si−O−Si polymer network skeleton was obtained by hydrolysis and a condensation reaction with TMOS and MTMS as co-precursors, and the enzyme molecules were embedded in the Si−O−Si network. In hydrolysis reaction, the methyl group of MTMS was not involved in hydrolysis, replacing and cross-linking with hydroxyl groups on Si−O−Si network, which provided a necessary condition for the development of hydrophobic properties [43]. The polydopamine-modified immobilized enzyme was based on the synergistic sol-gel mechanism [44]. In short, dopamine nanoparticles were uniformly mixed into the sol. In this system, dopamine hydrochloride was self-polymerized into PDA under alkaline conditions and deposited on the surface of Si−O−Si network. Moreover, the residual quinone functional groups presented in the polydopamine coat-ing were reactive toward nucleophilic groups, and CALB could couple covalently with polydopamine through Michael-type addition or Shiff-based formation [45,46]. We expected that the resulting SiO 2 -CH 3 -CALB@PDA xerogels would have excellent mechanical strengths and enzyme activity stabilities.

Strategy for Immobilizing CALB and Possible Mechanism
droxyl groups on Si−O−Si network, which provided a necessary condition for the development of hydrophobic properties [43]. The polydopamine-modified immobilized enzyme was based on the synergistic sol-gel mechanism [44]. In short, dopamine nanoparticles were uniformly mixed into the sol. In this system, dopamine hydrochloride was selfpolymerized into PDA under alkaline conditions and deposited on the surface of Si−O−Si network. Moreover, the residual quinone functional groups presented in the polydopamine coating were reactive toward nucleophilic groups, and CALB could couple covalently with polydopamine through Michael-type addition or Shiff-based formation [45,46]. We expected that the resulting SiO2-CH3-CALB@PDA xerogels would have excellent mechanical strengths and enzyme activity stabilities.

Mechanical Properties
In practical applications, xerogel is prone to deformation under external force, resulting in enzyme leakage or inactivation. Therefore, strength is crucial for the application of xerogel in organic catalysis. We experimentally compared the strengths of SiO2-CH3-CALB and SiO2-CH3-CALB@PDA. The compressive stress-strain curves for SiO2-CH3-CALB and SiO2-CH3-CALB@PDA are presented in Figure 6. A macroscopic compression experiment showed that the SiO2-CH3-CALB@PDA xerogel model could withstand higher pressures (12.55 Mpa) than that of SiO2-CH3-CALB (9.00 Mpa), and the strain of SiO2-CH3-CALB@PDA (9.64%) was greater than that of SiO2-CH3-CALB (9.07%). In addition, the fracture modes of the two materials were also significantly different. The fracture mode of SiO2-CH3-CALB was similar to that of brittle materials, while the fracture mode of SiO2-CH3-CALB@PDA was similar to that of viscoelastic materials [47,48]. This can be ascribed to two factors. On the one hand, PDA was deposited on the surface of the Si−O−Si network, which reduced the capillary force generated by the shrinkage of the xerogel during drying [29]. On the other hand, the surface of the PDA contained a large number of functional groups that could interact with Si−O−Si chains, serving as crosslinking sites to increase the mechanical strength of the SiO2-CH3-CALB@PDA xerogel, preventing it from breaking under pressure [32]. Overall, the internal network structure of the xerogel and polydopamine coatings played a key role in the whole compression process, confirming the formation of a stable xerogel. After modification by PDA nanoparti-

Mechanical Properties
In practical applications, xerogel is prone to deformation under external force, resulting in enzyme leakage or inactivation. Therefore, strength is crucial for the application of xerogel in organic catalysis. We experimentally compared the strengths of SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA. The compressive stress-strain curves for SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA are presented in Figure 6. A macroscopic compression experiment showed that the SiO 2 -CH 3 -CALB@PDA xerogel model could withstand higher pressures (12.55 Mpa) than that of SiO 2 -CH 3 -CALB (9.00 Mpa), and the strain of SiO 2 -CH 3 -CALB@PDA (9.64%) was greater than that of SiO 2 -CH 3 -CALB (9.07%). In addition, the fracture modes of the two materials were also significantly different. The fracture mode of SiO 2 -CH 3 -CALB was similar to that of brittle materials, while the fracture mode of SiO 2 -CH 3 -CALB@PDA was similar to that of viscoelastic materials [47,48]. This can be ascribed to two factors. On the one hand, PDA was deposited on the surface of the Si−O−Si network, which reduced the capillary force generated by the shrinkage of the xerogel during drying [29]. On the other hand, the surface of the PDA contained a large number of functional groups that could interact with Si−O−Si chains, serving as crosslinking sites to increase the mechanical strength of the SiO 2 -CH 3 -CALB@PDA xerogel, preventing it from breaking under pressure [32]. Overall, the internal network structure of the xerogel and polydopamine coatings played a key role in the whole compression process, confirming the formation of a stable xerogel. After modification by PDA nanoparticles, the mechanical properties of the xerogels were improved. This occurred because polydopamine can interact with the xerogel matrix, increasing xerogel hardness and improving brittleness. cles, the mechanical properties of the xerogels were improved. This occurred because polydopamine can interact with the xerogel matrix, increasing xerogel hardness and improving brittleness.

Stability of Immobilized CALB
The free and immobilized CALB was incubated at 60 °C for a certain time to investigate their thermal stability. The influence of temperature towards the stability of CALB is illustrated in Figure 7a. With the increasing of incubation time, the hydrolytic activity of free CALB and SiO2-CH3-CALB decreased, and free CALB was entirely deactivation after 3 h treatment. However, the SiO2-CH3-CALB@PDA exhibited better stability, which maintained 36.5% of its activity after 6 h of incubation. These results revealed that the better thermal stability of SiO2-CH3-CALB@PDA among free CALB and SiO2-CH3-CALB was attributed to the strong covalent bonds that formed through the reaction between the amine in the enzyme and the electrophilic groups in the PDA [49]. In addition, the PDA layer provides a stiffer external backbone to protect the CALB molecule from high temperatures [50]. Improvements in thermal stability will expand the range of applications for immobilized enzymes.
In order to investigate the storage stability of SiO2-CH3-CALB and SiO2-CH3-CALB@PDA, the examination was carried out at room temperature for 90 days. As shown in Figure 7b, SiO2-CH3-CALB exhibited 66.3% of its initial activity after 90 days, while SiO2-CH3-CALB@PDA exhibited approximately 71.4% under the same conditions. The high storage stability exhibited by CALB can be explained by the protective effect of the Si−O−Si network in the silica structure, which protects the enzyme activity inside, and further enhances its structural stability. The interactions of different geometries of the enzyme and support may have a significant influence on the enzyme activity. Generally, it is accepted that the highly curved surface reduces the possibility of enzyme denaturation and inhibits lateral interactions between adjacent enzymes, further leading to the structural stability and persistent activity of the adsorbed enzyme [51,52]. Additionally, multiple points of binding were observed between the PDA support and CALB in SiO2-CH3-CALB@PDA, which formed the PDA coating on the surface of the polymer network inside the xerogels, acting in a protective role [53,54]; this could explain their greater activity in external environments.

Stability of Immobilized CALB
The free and immobilized CALB was incubated at 60 • C for a certain time to investigate their thermal stability. The influence of temperature towards the stability of CALB is illustrated in Figure 7a. With the increasing of incubation time, the hydrolytic activity of free CALB and SiO 2 -CH 3 -CALB decreased, and free CALB was entirely deactivation after 3 h treatment. However, the SiO 2 -CH 3 -CALB@PDA exhibited better stability, which maintained 36.5% of its activity after 6 h of incubation. These results revealed that the better thermal stability of SiO 2 -CH 3 -CALB@PDA among free CALB and SiO 2 -CH 3 -CALB was attributed to the strong covalent bonds that formed through the reaction between the amine in the enzyme and the electrophilic groups in the PDA [49]. In addition, the PDA layer provides a stiffer external backbone to protect the CALB molecule from high temperatures [50]. Improvements in thermal stability will expand the range of applications for immobilized enzymes.
In order to investigate the storage stability of SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA, the examination was carried out at room temperature for 90 days. As shown in Figure 7b, SiO 2 -CH 3 -CALB exhibited 66.3% of its initial activity after 90 days, while SiO 2 -CH 3 -CALB@PDA exhibited approximately 71.4% under the same conditions. The high storage stability exhibited by CALB can be explained by the protective effect of the Si−O−Si network in the silica structure, which protects the enzyme activity inside, and further enhances its structural stability. The interactions of different geometries of the enzyme and support may have a significant influence on the enzyme activity. Generally, it is accepted that the highly curved surface reduces the possibility of enzyme denaturation and inhibits lateral interactions between adjacent enzymes, further leading to the structural stability and persistent activity of the adsorbed enzyme [51,52]. Additionally, multiple points of binding were observed between the PDA support and CALB in SiO 2 -CH 3 -CALB@PDA, which formed the PDA coating on the surface of the polymer network inside the xerogels, acting in a protective role [53,54]; this could explain their greater activity in external environments.

Transesterification and Reusability
Some enzymes have been used as biocatalysts to synthetize ester compounds, among which CALB can form high value-added ester products by transesterification reactions. As one of the major biocatalysts for ester synthesis, CALB can catalyze the transesterification of n-butanol with ethyl acetate to produce butyl acetate, which is an excellent organic solvent. Figure 8 shows the CALB-catalyzed synthesis of butyl acetate by transesterification of n-butanol with ethyl acetate. The reaction is a solvent-free system, and was carried out in a batch reactor at 70 °C. In a solvent-free system, the enzyme directly acts on the substrate, increases the substrate concentration, improving the reaction rate and selectivity and reducing the damage of organic solvents to the enzyme [55]. Therefore, we chose ethyl acetate as a reactant, as it also acts as a solvent in the CALB-catalyzed synthesis of butyl acetate.
The transesterification of n-butanol with ethyl acetate was selected as a target reaction to evaluate the conversion efficiency and reusability of immobilized CALB in the present work. The conversion of n-butanol and the reusability of SiO2-CH3-CALB and SiO2-CH3-CALB@PDA were compared under optimal active conditions. As shown in Figure 9, in the first cycle, the conversion of n-butanol of SiO2-CH3-CALB and SiO2-CH3-CALB@PDA retained 52.42% and 57.67%, respectively. For SiO2-CH3-CALB, CALB molecules were encapsulated in the xerogel polymer network by physical adsorption, Virgen-Ortíz et al. have reported some substrates/product may produce the enzyme's release from physically absorbed enzymes, so the leakage of CALB was prone to denaturation or inactivation during the reaction [56]. The decrease in conversion was observed in the first five cycles. After the fifth cycle, the activity began a slow deceleration state, lasting for the next three cycles. After eight cycles, SiO2-CH3-CALB@PDA retained more than a 30.84% conversion of n-butanol. SiO2-CH3-CALB retained a 25.04% conversion of n-butanol. The conversion of n-butanol loss could be due to enzyme leakage during washing and enzyme deactivation during repeated uses [57]. As the reaction produces a by-product of ethanol in the batch reactor system, resulting in enzyme inhibition, inhibition will reduce lipase activity. High concentrations of n-butanol inhibit the synthesis of butyl acetate catalyzed by immobilized CALB. This inhibitory effect has been found in the reaction among butyric acid and lauric acid with ethanol [58,59]. Therefore, operational stability of the enzyme is not too high. Considering SiO2-CH3-CALB@PDA had better reusability, storage stability

Transesterification and Reusability
Some enzymes have been used as biocatalysts to synthetize ester compounds, among which CALB can form high value-added ester products by transesterification reactions. As one of the major biocatalysts for ester synthesis, CALB can catalyze the transesterification of n-butanol with ethyl acetate to produce butyl acetate, which is an excellent organic solvent. Figure 8 shows the CALB-catalyzed synthesis of butyl acetate by transesterification of n-butanol with ethyl acetate. The reaction is a solvent-free system, and was carried out in a batch reactor at 70 • C. In a solvent-free system, the enzyme directly acts on the substrate, increases the substrate concentration, improving the reaction rate and selectivity and reducing the damage of organic solvents to the enzyme [55]. Therefore, we chose ethyl acetate as a reactant, as it also acts as a solvent in the CALB-catalyzed synthesis of butyl acetate.   The transesterification of n-butanol with ethyl acetate was selected as a target reaction to evaluate the conversion efficiency and reusability of immobilized CALB in the present work. The conversion of n-butanol and the reusability of SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA were compared under optimal active conditions. As shown in Figure 9, in the first cycle, the conversion of n-butanol of SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA retained 52.42% and 57.67%, respectively. For SiO 2 -CH 3 -CALB, CALB molecules were encapsulated in the xerogel polymer network by physical adsorption, Virgen-Ortíz et al. have reported some substrates/product may produce the enzyme's release from physically absorbed enzymes, so the leakage of CALB was prone to denaturation or inactivation during the reaction [56]. The decrease in conversion was observed in the first five cycles. After the fifth cycle, the activity began a slow deceleration state, lasting for the next three cycles. After eight cycles, SiO 2 -CH 3 -CALB@PDA retained more than a 30.84% conversion of n-butanol. SiO 2 -CH 3 -CALB retained a 25.04% conversion of n-butanol. The conversion of n-butanol loss could be due to enzyme leakage during washing and enzyme deactivation during repeated uses [57]. As the reaction produces a by-product of ethanol in the batch reactor system, resulting in enzyme inhibition, inhibition will reduce lipase activity. High concentrations of n-butanol inhibit the synthesis of butyl acetate catalyzed by immobilized CALB. This inhibitory effect has been found in the reaction among butyric acid and lauric acid with ethanol [58,59]. Therefore, operational stability of the enzyme is not too high. Considering SiO 2 -CH 3 -CALB@PDA had better reusability, storage stability and mechanical strength, SiO 2 -CH 3 -CALB@PDA is more applicable for practical applications.

Comparison of Butyl Acetate Production Using Previous Lipase Biocatalysts
The prepared catalyst of SiO2-CH3-CALB@PDA possessed the advantages of biocompatibility, environmental friendliness, operating convenience and safety. Compared with previous lipase catalysts, such as SiO2-CH3-CALB and Novozyme 435, the catalytic efficiency of SiO2-CH3-CALB@PDA (57.67%) in the transesterification reaction system was slightly higher than those of SiO2-CH3-CALB (52.42%) and Novozyme 435 (55.30%) [60]. Although the improvement in operational stability and catalytic performance is not obvious, the polydopamine modification strategy is worth adopting; it can improve the immobilized enzyme loading and balance the mechanical properties of the supports, which expands the application range of immobilized enzymes in some special cases.

Comparison of Butyl Acetate Production Using Previous Lipase Biocatalysts
The prepared catalyst of SiO 2 -CH 3 -CALB@PDA possessed the advantages of biocompatibility, environmental friendliness, operating convenience and safety. Compared with previous lipase catalysts, such as SiO 2 -CH 3 -CALB and Novozyme 435, the catalytic efficiency of SiO 2 -CH 3 -CALB@PDA (57.67%) in the transesterification reaction system was slightly higher than those of SiO 2 -CH 3 -CALB (52.42%) and Novozyme 435 (55.30%) [60]. Although the improvement in operational stability and catalytic performance is not obvious, the polydopamine modification strategy is worth adopting; it can improve the immobilized enzyme loading and balance the mechanical properties of the supports, which expands the application range of immobilized enzymes in some special cases.

Preparation of SiO 2 -CH 3 -CALB
SiO 2 -CH 3 -CALB was prepared by sol-gel method. Firstly, TOMS (0.54 g), MTMS (1.934 g), methanol (3.39 g), PEG (0.14 g), NaF solution (0.49 g, 1 M), water (1.26 g) and CALB enzyme solution (3.39 g) were mixed and stirred at 0 • C, and the mixture was transferred to a clean petri dish. Then, the petri dish was sealed and placed at room temperature for 2 h to form a gel network. Finally, the petri dish was opened to evaporate the water and solvent in the gel completely, and then dried at room temperature for 48 h.

Preparation of SiO 2 -CH 3 -CALB@PDA
Dopamine hydrochloride (0.02 g) was dispersed in methanol (3.39 g), then NaF solution (0.98 g, 1 M) was added and mixed for 10 min. The use of NaF solution rather than Tris buffer was due to the fact that primary amine group in Tris can covalently interact with PDA, which could affect the deposition of PDA and the continuous coupling of CALB with PDA.
To the obtained mixture we added TMOS (0.54 g) and 1.934 g of MTMS (1.934 g), PEG (0.14 g), water (1.26 g) and CALB enzyme solution (3.39 g), which was then mixed and stirred at 0 • C, and the mixture was transferred to a clean petri dish. Then, the petri dish was sealed and placed at room temperature for 4 h to form a gel network. Finally, the petri dish was opened to evaporate the water and solvent in the gel completely and then dried at room temperature for 48 h.

Characterization
The microstructures of the samples were observed using a transmission electron microscope (TEM, Talos F200S, Hillsboro, FL, USA) and scanning electron microscopy (SEM, Nova Nano SEM 450, Hillsboro, FL, USA). Fourier transform infrared (FT-IR) spectra of the samples were collected from 4000 to 400 cm −1 on a Bruker Tensor 27 analyzer (Bremen, Germany) using KBr pellets method. X-ray diffraction (XRD) patterns were measured by a Bruker D8 Discover (Bremen, Germany) with scanning rate of 6 • min −1 under Cu Kα radiation (λ = 0.154056 nm). Samples were mounted on a low background silicon substrate and diffraction scans covered a 2θ range of 5 • to 80 • . X-ray photoelectron spectra (XPS, Al-Kα) were recorded on an X-ray photoelectron spectrometer (ESCALAB 250Xi, Hillsboro, FL, USA), and the C 1 s of 284.8 eV was referred to for calibrating the binding energy. The N 2 adsorption-desorption isotherms were measured by a pore sizespecific surface area analyzer (SSA-6000, Beijing, China) at 77 K. The pore size distribution and surface area were determined through calculating N 2 adsorption-desorption according to the Brunauer-Emmett-Teller (BET) method. A spectrophotometer (UV-2600, Shimadzu, Kyoto, Japan) was used to analyze the concentration and activity of the enzyme.

Determination of Enzyme Loading
The Bradford method was used to determine enzyme embedding in the silica xerogels by measuring of the protein concentrations in the initial enzyme solutions and immobilized enzyme phosphate detergents. A calibration curve was plotted, using Coomassie Brilliant Blue G-250 solutions as standards. The enzyme concentration in the solution was able to be determined with UV-vis spectrophotometry, by measuring the absorbance at 595 nm. The amount of enzyme embedded in silica xerogels was calculated by the following equation: where C 0 is the initial enzyme concentration (mg/g), C 1 is the enzyme concentration in phosphate detergent (mg/g).
3.6. Properties of Free CALB and the Immobilized CALB 3.6.1. Assay of the CALB Activity The free CALB and samples of immobilized CALB activities were determined by using p-NPP (5 mg/mL in ethanol) as the substrate. Typically, 200 µL of p-NPP solution was added to the solution consisting of the samples (2 mg) and PBS (0.1 M, pH 7.5, 3 mL). After reaction for 3 min, the filtrate of the reaction that contained 4-nitrophenol (p-NP), and the concentration of p-NP was quantified via absorbance at 410 nm on a spectrophotometer. One unit (U) of lipase hydrolytic activity was regard as the lipase mass that liberates 1 nmol of p-NP under these test conditions per minute. The relative enzymatic activity was related to a percentage of this highest activity (100% means the highest enzymatic activity). The activity recovery was calculated from the value of the activity of the initial CALB solution divided by the activity value of immobilized CALB obtained immediately after the immobilization procedure.
3.6.2. Thermal and Storage Stability of the Free CALB and Immobilized CALB Free CALB, SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA were incubated in PBS (50 mM, pH 7.5) at 70 • C for 6 h to examine their thermal stabilities. The p-NPP assay was employed to measure residual activity as described in Section 3.6.1. To evaluate the storage stability, the residual activity of SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA was tested after a given treatment duration at 25 • C, respectively. The residual activity of each sample under treatment was measured at given time intervals and used for comparison with the original activity.

Mechanical Performance Tests
The mechanical performances of SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA were tested using a microcomputer control electron universal testing machines (CMT6104, Shenzhen, China) with a 5000-N load cell. To facilitate testing, samples were made into rectangular specimens. Compression strain tests of the samples (lengths, 23 mm; widths, 13.28 mm; thicknesses, 6 mm) were performed at a compression rate of 2 mm/min.

Transesterification and Reusability
The reaction for the transesterification of n-butanol with ethyl acetate was performed in a glass three-necked reactor with a volume of 250 mL at 343 K and 101.3 kPa. The electric stirring was controlled up to 3000 rpm to achieve uniform mixing of the reactive mixture. In the experiment, the mixture of reactants ethyl acetate and n-butanol (molar ratio of ethyl acetate to n-butanol was 1:1) were heated to 343 K in a water bath, then the catalysts (the catalyst dosage was 10% of the mass of n-butanol, and the catalysts were SiO 2 -CH 3 -CALB and SiO 2 -CH 3 -CALB@PDA) were set in the reactor to start the reaction. Samples were withdrawn from the reactor every 30 min with a syringe during the reaction for composition analysis until the 5 h. Finally, the catalysts were washed with PBS (0.1 M, pH 7.5) buffer and dried for 12 h before next cycle.
The composition of the product was analyzed by gas chromatography (GC-2010 Pro, Shimadzu, Kyoto, Japan) equipped with a flame ionization detector (FID) and an InertCap FFAP capillary column (30 m × 0.25 mm × 0.25 mm). Typically, n-propanol was used as the internal standard substance. N 2 with purity of 99.99 wt% was used as carrier gas at 1 mL/min. The temperature of the injection port and the detector were controlled at 473 K and 493 K, respectively. 0.4 µL sample was injected each time.

Conclusions
In this work, the immobilization of CALB in PDA-modified silica xerogels was successfully prepared by the self-polymerization of dopamine on the Si−O−Si network surfaces of silica xerogels. The modified silica xerogels showed an excellent embedding ability for CALB compared with conventional silica xerogels. They exhibited a high capacity of 45.6 mg/g support for CALB encapsulation. The mechanical strength and thermal and storage stability of the immobilized CALB were greatly elevated. Moreover, the immobilization of an enzyme in PDA-modified silica xerogels was utilized in the transesterification between n-butanol with ethyl acetate, which retained 30.84% conversion of n-butanol after eight cycles. In short, the SiO 2 -CH 3 -CALB@PDA catalyst was prepared by a simple and practical method, which is expected to overcome the related problems of shrinkage and