Mechanochemical biomimetic mineralization of UiO-66-NH₂-immobilized cellulase for enhanced catalytic stability and efficiency

Abstract

Biomimetic mineralization is a crucial biotechnology for encapsulating enzymes within metal–organic frameworks (MOFs). While this technique is extensively employed in solvent-based systems, its applicability in mechanochemical systems remains to be explored. In addition, the structural units of MOF composition may significantly influence this process. In this study, we reported that dodecanuclear zirconium clusters act as precursors to facilitate the mechanochemical process of biomimetic mineralization, leading to the construction of cellulase@UiO-66-NH₂ (with cellulase abbreviated as Cel). The results demonstrate that dodecanuclear zirconium clusters promote the biomimetic mineralization of Cel@UiO-66-NH₂, preventing enzyme degradation by organic solvents. This approach leads to a 20% increase in enzyme activity per unit mass and a 78% improvement in the encapsulation rate. It also enhances catalytic efficiency and substrate affinity compared with Cel@UiO-66-NH₂ synthesized with hexanuclear zirconium clusters. The biomimetic mineralization was attributed to the increased local concentration of structural units of MOFs surrounding Cel, as well as transformations in chemical bonding and alterations in enzyme structure. We demonstrated the stability of Cel@UiO-66-NH₂ compared to traditional physical adsorption methods and explored its applications in the saccharification of carboxymethylcellulose and microcrystalline cellulose and the high-temperature sequential extraction of polysaccharides from Naematelia aurantialba. Our results revealed that Cel@UiO-66-NH₂ retained over 50% of its catalytic activity after eight cycles of carboxymethylcellulose saccharification and maintained 50.9% enzyme activity after five cycles of treatment with microcrystalline cellulose. In addition, we achieved a polysaccharide extraction yield of 10.35% at 70 °C.

---

Green foundation

1. Traditional enzyme immobilization methods lead to significant carbon emissions. This study introduces a mechanochemical biomimetic mineralization strategy, reducing the need for organic solvents and energy-intensive processes, while improving enzyme recyclability for more sustainable and cost-effective biocatalysis.

2. By optimizing the UiO-66-NH₂ metal precursors, biomimetic mineralization in the mechanochemical process enhanced enzyme encapsulation efficiency by 78% and activity by 20%. Compared to traditional methods, this approach reduces carbon emissions and improves catalytic performance. After eight cycles of carboxymethyl/microcrystalline cellulose hydrolysis, over 50% activity is retained, and it can also be used for high-temperature extraction of plant polysaccharides.

3. Future research should focus on: (1) Expanding the use of degradable and renewable MOFs precursors; (2) Optimizing mechanochemical conditions and exploring alternative energy sources, such as solar-assisted milling; (3) Investigating the molecular mechanisms of enzyme-MOF interactions.

1. Introduction

The demand for enzymes as catalysts is increasing in several industrial fields. Enzymes are often immobilized on supports to achieve the recyclability of enzymes.^1–3 However, the synthesis of supports usually requires large amounts of chemicals. Chemicals and waste are a primary source of global greenhouse gas emissions, with a projected increase to 1.2 billion tons of carbon dioxide equivalent by 2030, causing economic losses of up to $1.7 trillion.⁴ In addition, enzymes are highly sensitive to the conditions of the reaction medium. Factors such as high temperatures and extreme pH can cause a decrease in enzyme activity.⁵ Current immobilization technologies still do not fully protect enzymes in a fast, simple, stable, and efficient manner. Therefore, there is an urgent need to find a more advanced and environmentally friendly platform for enzyme immobilization.

Biomineralization is a pervasive bio-induced self-assembly process that creates biocomposites with superior mechanical properties by controlling their crystal morphology and composition under physiological conditions.^6,7 Inspired by biomineralization in nature, the strategy of biomimetic synthesis of new materials has been pioneered. In this strategy, biomolecules such as deoxyribonucleic acid, enzymes, and peptides are used as templates, and crystals of organic/inorganic materials are induced to proliferate around their surfaces. This process protects the biomolecules and allows for their efficient use.^8,9 Biomimetic mineralization is an advanced method of enzyme immobilization. This process is regulated by various external factors, such as the forces between the biomolecules and the material, the nature of the assembled interfaces, and the additives. These factors can affect the final morphology, structure, and function of the composites.¹⁰ As organic–inorganic hybrid materials with customizable structural topology, pore function, and crystal morphology, MOFs possess a natural affinity for biomolecules.^11–14 Biomimetic mineralization technology currently relies on solvent-based processes. Kim et al. prepared enzyme@MIL-88A composites by injecting fumaric acid (organic phase) into Fe³⁺ and proteins (aqueous phase).¹⁵ Liang et al. prepared enzyme@zeolitic imidazolate framework-8 (ZIF-8) composites using biomacromolecules, such as urease and horseradish peroxidase.¹⁴ However, the types of MOFs that can be synthesized at room temperature have limitations. ZIF-based MOFs are unstable and prone to disintegration under alkaline or elevated temperature conditions.¹⁶ Moreover, solvent-based methods for synthesizing MOFs are lengthy and costly, often requiring high temperatures and hazardous organic solvents that do not align with the current green concept.

As people gradually focus on using clean and sustainable technologies, many studies have started to develop green and environmentally friendly methods for preparing metal–organic materials. Recently, mechanochemical synthesis has emerged as a new process for MOF synthesis. Mechanochemical reactions rely on the ball milling process, which generates chemical energy through the impact between the balls and the reactants. This process triggers a chemical reaction and the formation of structural stresses, new bonds, and free radicals.¹⁷ Mechanochemical reactions can use insoluble metal sources, such as metal oxides and salts, as metal precursors.¹⁸ In addition, adding small amounts of solvents can facilitate the chemical synthesis process in a more environmentally friendly and safe manner and provide a faster and more biofriendly option for synthesizing many MOFs that previously required harsh conditions for their synthesis.¹⁹ The discovery of this method offers the possibility for different types of MOFs encapsulating enzymes. In 2016, He et al. proposed a simple mechanochemical encapsulation method to synthesize enzyme@ZIF-8 by introducing lipase into a mortar and grinding it during the MOF synthesis process.²⁰ In 2019, Wei et al. successfully encapsulated β-glucosidase in MOFs such as ZIF-8 and UiO-66-NH₂.²¹ While biomimetic mineralization of solvent systems has been extensively studied, this concept remains to be applied to mechanochemical systems. The two reaction systems are very different, and it remains unknown whether and how biomimetic mineralization can occur in mechanochemical systems.¹⁷ Moreover, irrational structural units of MOFs hinder mechanochemical reactions and affect the activity of biomolecules.²² Among the MOFs synthesized through mechanochemical methods, UiO-66-NH₂ is the most promising for industrial conversion owing to its high chemical stability and large-scale production.²³ An essential step in the mechanochemical approach is the synthesis of insoluble metals, which serve as precursors and are essential for determining the topology of MOFs. Currently, the insoluble metal precursors used for the mechanochemical synthesis of enzyme@UiO-66-NH₂ are mainly hexanuclear zirconium methacrylate clusters [Zr₆O₄(OH)₄(CH₂C(CH₃)COO)₁₂].²¹ Mechanochemical reactions based on these hexanuclear zirconium methacrylate clusters may be hampered by the formation of inert intermediates and the polymerization of methacrylic acid released during the milling process.^23–26 This may negatively affect the final morphology, structure, and functionality of the enzyme@MOF composite.

In this study, we reported that the metal precursor, dodecanuclear zirconium acetate clusters [Zr₁₂O₈(OH)₈(CH₃COO)₂₄], promoted biomolecule (Cel)-induced mechanochemical biomimetic mineralization. Compared with the hexanuclear zirconium methacrylate cluster, the dodecanuclear zirconium acetate clusters helped the enzyme avoid degradation by organic solvents during synthesis and improved the catalytic efficiency of the enzyme and the encapsulation rate of UiO-66-NH₂. Furthermore, biomimetic mineralization was confirmed using small-angle X-ray scattering (SAXS). Biomimetic mineralization during mechanochemical processes was attributed to the prenucleation of UiO-66-NH₂ clusters around biomolecules. This process is facilitated by increased local concentrations of metal cations and organic ligands, chemical bonding transformations, and changes in the Cel structure. Moreover, biomimetic mineralization was verified to mediate the enzyme-protective effect of the UiO-66-NH₂ shell. Compared to traditional physical adsorption methods, this shell attenuated the biological, thermal, and chemical degradation of Cel and improved its stability and viability. Finally, the synthesized Cel@UiO-66-NH₂ was used for hydrolysis (saccharification) of carboxymethylcellulose/microcrystalline cellulose and the high-temperature sequential production of Naematelia aurantialba polysaccharides to validate the utility of mechanochemistry-based biomimetic mineralization. In conclusion, this work provides insights into the research and application of biomimetic mineralization techniques for immobilized enzymes and mechanochemistry. It provides a direction for more in-depth exploration of this method in the future.

2. Experimental section

2.1 Materials

All chemicals were obtained commercially and used without additional purification. 2-Aminoterephthalic acid (BDC-NH₂, 98%, Macklin), Methacrylic acid (>99%, Aladdin), Zr(OPr)₄ solution (70% solution in n-propanol, Macklin), Acetic acid (≥99.5%, Aladdin), Cellulase from Trichoderma reesei (Cel, Shanghai yuanye Bio-Technology Co., Ltd), Carboxymethyl cellulose (CMC, Shanghai yuanye Bio-Technology Co., Ltd), Citric acid monohydrate (≥99.5%, Macklin), Citric acid trisodium salt dehydrate, (≥99.5%, Macklin), 3-amino-1,2,4-triazole (3-AT, 96.0%, Macklin), Urea (99.0%, Macklin), Fluorescein-5-Isothiocyanate (FITC, >95.0%, Macklin), D-Modified Eagle Medium (MEM, Cell Resource Center of the Institute of Basic Medical Sciences, Chinese Academy of Medical Sciences), Penicillin/streptomycin (Gibco), Trypsin-EDTA (0.25%, Gibco), Penicillin (Gibco), Streptomycin (Gibco), All test kits were purchased from Beijing Solarbio Science & Technology Co., Ltd (China).

2.2 Synthesis and characterization

2.2.1 Synthesis of hexanuclear zirconium methacrylate cluster [Zr₆O₄(OH)₄(CH₂C(CH₃)COO)₁₂]. The precursor (methacrylate metal cluster) synthesized by this method is uniformly denoted as cluster 2. The synthesis of cluster 2 was carried out following the literature method with slight modifications.²⁵ Cluster 2 was synthesized by reacting methacrylic acid with Zr(OPr)₄ solution (70% solution in n-propanol). There was a mixture of 1.4 mL methacrylic acid, one drop of water, and 2 mL of Zr(OPr)₄ solution (70% solution in n-propanol). Then, the mixture was heated in an oven at 60 °C for 3 days, and the formed colorless solid was obtained by filtration under vacuum and washed three times with 2-propanol.

2.2.2 Synthesis of dodecanuclear zirconium acetate clusters [Zr₁₂O₈(OH)₈(CH₃COO)₂₄]. The precursor (dodecanuclear zirconium acetate clusters) synthesized by this method is uniformly denoted as cluster 1. The synthesis of cluster 1 was carried out following the literature method with slight modifications.²⁴ Cluster 1 was synthesized by reacting acetic acid with Zr(Pro)₄ in a 70% solution in n-propanol. 3.5 mL of acetic acid and 0.95 mL of Zr(OPr)₄ solution (70% solution in n-propanol) were mixed, and then the mixture was left at room temperature overnight, and the colorless solid was obtained by filtration under vacuum and washed three times with acetic acid.

2.2.3 Milling synthesis of Cel@UiO-66-NH₂ from the cluster 1 (Cel@UiO-66-NH₂ (1)). The Cel@UiO-66-NH₂ composites synthesized by this method are uniformly denoted as Cel@UiO-66-NH₂ (1). The LAG-assisted milling reaction of Cel@UiO-66-NH₂ (1) was performed on a JXFSTPRP-CL-24 high-speed low-temperature grinding instrument at 50 Hz, −30 °C, and 30 min. The reactants (acetate cluster 1 : 2-Aminoterephthalic acid, molar ratio of 1 : 6), Cel (10 mg) and 50 μL of anhydrous ethanol were added to a grinding tube (2 mL) with three stainless steel balls (3 mm diameter) was used. The collected powder was centrifuged, washed with water, and vacuum-dried at 40 °C.

2.2.4 Milling synthesis of UiO-66-NH₂ from the cluster 2 (UiO-66-NH₂ (2)). The UiO-66-NH₂ samples synthesized by this method are uniformly denoted as UiO-66-NH₂ (2). The LAG-assisted milling reaction of UiO-66-NH₂ (2) was performed on a JXFSTPRP-CL-24 high-speed low-temperature grinding instrument at 50 Hz, −30 °C, and 30 min. The reactants (methacrylate cluster 2 : 2-Aminoterephthalic acid, molar ratio of 1 : 6) and 50 μL of anhydrous ethanol were added to a grinding tube (2 mL) with three stainless steel balls (3 mm diameter). The collected powder was centrifuged, washed with water, and vacuum-dried at 40 °C.

2.3 Characterization

An X-ray diffractometer (Smartlab, Rigaku, Japan) was used to collect powder X-ray diffraction (PXRD) patterns. The adsorption–desorption isotherms of N₂ were measured at 77 K using an automatic chemical adsorption instrument (AutoChemII 2920, Micromeritics, USA) for analysis of the Brunauer–Emmett–Teller (BET) surface area and pore structure of the samples. All samples were degassed at 200 °C for 6 h prior to measurement. TEM images were obtained from transmission electron microscopy (TEM, Hitachi 7700, Tokyo, Japan) with an accelerating voltage of 200 kV and equipped with an energy dispersive X-ray spectrometer (EDX, XFlash 6130, Bruker, Germany). Field emission scanning electron microscope (FE-SEM, Sigma 500, ZEISS, Germany) with an accelerating voltage of 5 kV was used to obtain SEM images of the samples. The surface morphology and elemental composition of the samples were characterized by energy dispersive X-ray spectroscopy (XFlash 6130, Bruker, Germany). Information on the structural changes of the proteins was analyzed by using an ultraviolet–visible spectrophotometer (Shimadzu-2600, Kyoto, Japan) and a fluorescence spectrometer (F-7000, Hitachi, Tokyo, Japan). The Fourier transform infrared spectra (FTIR) were obtained on a Fourier transform infrared spectroscopy (Nicolet iS50, TMO, USA) at 400–3600 cm⁻¹. Thermogravimetric analysis (TGA) was performed on a thermogravimetric analyser (TGA-50, Shimadzu, Japan). Zeta-potentials were performed by a nanoparticle size analyzer (Litesizer™ 500, Anton Paar, Austria). X-ray photoelectron spectroscopy (XPS) spectra were collected from an XPS spectrometer (Kalpha, Thermo Scientific, USA) equipped with a monochromatic Al Kα X-ray source (hv = 1486.6 eV) operating at 100 W. All peaks were calibrated with adventitious carbon with a C 1s peak binding energy of 284.8 eV. Enzyme-material interactions were measured using a quartz crystal microbalance with dissipation monitoring (Q-Sense Explorer, Biolin Scientific, Gothenburg, Sweden). Differential scanning calorimetry (DSC) was performed using a differential scanning calorimeter (DSC204F1, Netzsch, German). Determine the presence and spatial location of fluorescently labeled biomolecules in (or on) MOF composites using the confocal laser scanning microscopy (CLSM) CLSM technique (LSM 880, Zeiss, Germany).

2.4 The kinetic parameters of Cel in Cel@UiO-66-NH₂

The K_m, V_max, and K_cat/K_m of the hydrolysis reaction were calculated using the Michaelis–Menten equation for the same time (30 min) and different concentrations of carboxymethyl cellulose (0.2%, 0.6%, 1%, 1.4%, 2% w/v).²⁷where V₀ is the initial catalytic rate; V_max is the maximum rate conversion, which is obtained when the catalytic sites on the enzyme are saturated with substrate; K_cat is the catalytic constant (units: s⁻¹), representing the number of substrate molecules converted per enzyme molecule per second; K_m is the Michaelis–Menten constant; [S] is the initial substrate concentration.

2.5 Quartz crystal microbalance with dissipation monitoring

The interaction of cellulase with UiO-66-NH₂ (synthesized using different metal precursors) was measured using a quartz crystal microbalance with dissipation monitoring (Q-Sense Explorer, Biolin Scientific, Gothenburg, Sweden). Briefly, UiO-66-NH₂ nanoparticles were spin-coated onto the surface of the gold sensors. Before the adsorption step, ultrapure water was introduced into the flow cell until a stable baseline was established. Subsequently, cellulase solution was injected into the flow cell and the adsorption process continued until adsorption saturation was reached. Finally, the sensors were flushed with solvent to remove any cellulase loosely bound to the UiO-66-NH₂ nanoparticles until a stable signal was reached. The mass of the enzyme adsorbed on the UiO-66-NH₂ nanoparticles was estimated using the viscoelastic model in the instrument software (Dfind, Q-Sense, Biolin Scientific AB, Gothenburg, Sweden).

2.6 Synchrotron small angle X-ray scattering analysis

Small angle X-ray scattering (SAXS) analyses were performed on Xeuss 2.0 (Xenocs, France) with a Cu X-ray source of 30 W (wave length of 0.1542 nm) and a detector of Pilatus 300 K (Dectris). The sample-detector distance were set to 538 mm to cover a broad Q range using the virtual detector mode (0.074–4 nm⁻¹). The 1-D scattering profiles were reduced from the 2-D data using the XSACT package.

2.7 Guinier analysis

The dataset was imported into Sasfit (version 0.94.11), an open-source analytical program for generating Guinier fits. Beaucage describes how Guinier's law and structurally finite power laws can be derived from mutually exclusive scattering events. In the simplest case, the observed scattering is the sum of two components.²⁸where G is the Guinier pre-factor of the larger structure; B is a pre-factor specific to the type of power-law scattering: B is defined according to the regime in which the exponent P falls; G_s: G_s is the Guinier pre-factor of the smaller structure; B_s: B_s is a pre-factor specific to the type of power-law scattering: B_s is defined according to the regime in which the exponent P_s falls; R_g: large scale structure; R_sub: surface-fractal cut-off radius of gyration, R_sub defines the high-Q cutoff for the intermediate power law; R_s: size R_s of small subunits; P: scaling exponent of the power law assigned to the larger structure R_g; P_s: scaling exponent of the power law assigned to the smaller structure R_s.

2.8 Statistics

Data in this work were presented as mean ± standard deviations. One-way ANOVA test and Student's t-test were used to assess the statistical significance of variance. The error bar represents the standard deviation (n = 3). Statistical significance was considered when p < 0.05 (*), p < 0.01(**), and p < 0.001(***).

3. Results and discussion

3.1 Effects of metal precursors on biomimetic mineralization

3.1.1. Synthesis and characterization of materials. During the mechanochemical synthesis of MOFs, the choice of metal precursors significantly influences their structural formation.^21,24,25,29 Additionally, in the process of biomimetic mineralization, metal precursors can bind to the active sites on the surface of biomolecules. This interaction promotes nucleation around the biomolecule, leading to the growth of the mineralized material.¹⁰ Based on previous studies, we investigated the effects of different metal precursors on the structure of synthesized MOFs, the mineralization process, and their enzyme encapsulation. We used two primary metal precursors for synthesizing UiO-type MOFs: dodecanuclear zirconium acetate and hexanuclear zirconium methacrylate clusters. First, Cel was encapsulated by ball milling into MOFs synthesized with dodecanuclear zirconium acetate and hexanuclear zirconium methacrylate clusters as metal precursors, referred to as Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2), respectively (refer to the “Methods” section for detailed grinding and cleaning procedures). The crystal structures were analyzed using powder X-ray diffraction. Specific peaks at 2θ = 7.4°, 8.6°, and 25.8° for UiO-66-NH₂ (1) and 2θ = 7.4°, 8.3°, and 27.0° for UiO-66-NH₂ (2) correspond to the (1111), (200), and (600) crystal planes, confirming their successful synthesis (Fig. 1a). However, the presence of peaks at 2θ = 15.0° and the increased peak intensity upon enzyme addition suggest possible defects in the crystal structure of UiO-66-NH₂ (2) (Wei et al., 2019).²¹ The active conformation of Cel, before and after immobilization, was further analyzed using Fourier transform infrared spectroscopy (Fig. 1b and S1†).^30,31 Compared with UiO-66-NH₂ (1), characteristic peaks at 1078, 1020, and 1640 cm⁻¹ were observed in both Cel and Cel@UiO-66-NH₂ (1) These peaks correspond to the C–N stretching vibrations of the Cel polypeptide's aliphatic amine (1078 and 1020 cm⁻¹) and the C=O double bond (2931 cm⁻¹).^32–34 Unlike UiO-66-NH₂ (2), a shift in the C=O peak around 1640 cm⁻¹ in Cel@UiO-66-NH₂ (1) suggests the formation of hydrogen bonds or coordination interactions between Cel and the MOFs. The enhanced C–N peaks at 1078 cm⁻¹ and 1020 cm⁻¹ may result from interactions between the MOFs’ amino groups (–NH₂) and the enzyme's carboxyl or amino groups, altering the local environment. The more prominent peak at 1371 cm⁻¹ in Cel@UiO-66-NH₂ (1) indicates the involvement of the MOF structure in enzyme immobilization.³⁵ These results demonstrate that biomimetic mineralization via UiO-66-NH₂ (1) not only stabilizes the enzyme but also enhances its structural integrity through specific chemical bonding.

Fig. 1 Characterization of Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2) biocomposites. (a) X-ray diffraction, (b) Fourier-transform infrared spectroscopy, (c) confocal laser scanning microscopy images, (d) zeta potential, (e) cumulative pore volume, (f) incremental pore volume, (g) enzyme encapsulation rate at different enzyme additions, and (h) relative activity at different enzyme additions for Cel@UiO-66-NH₂ (1), UiO-66-NH₂ (1), Cel@UiO-66-NH₂ (2), and UiO-66-NH₂ (2). RS (mg) represents the amount of reducing sugar determined by DNS assay and serves as an indirect measure of catalytic performance. Higher RS values indicate greater hydrolytic activity under the same reaction conditions.

In addition, the spatial distribution of fluorescein-tagged Cel (FCel) in the samples was assessed using confocal laser scanning microscopy (Fig. 1c and S3†).³⁶ Equal amounts of FCel were added to all samples during synthesis. The images revealed that the FCel molecules were more homogeneously distributed in Cel@UiO-66-NH₂ (1) than in Cel@UiO-66-NH₂ (2), where a substantial portion of FCel was localized in the outer region. This suggests that MOFs synthesized using different metal precursors result in distinct enzymatic spatial distributions owing to their structural differences.³⁷ Surface zeta potential measurements (Fig. 1d) revealed that UiO-66-NH₂ (1) had a potential of 2.26 mV, while UiO-66-NH₂ (2) exhibited a potential of −1.54 mV. After enzyme encapsulation, these potentials increased to 7.58 and 9.32 mV, respectively. This suggests that there is differential enzyme binding to the two MOFs, possibly facilitated by electrostatic interactions.^38,39 Additional tests were conducted to assess the nitrogen adsorption–desorption of multiple samples (Fig. 1e and f). The results showed that the specific surface areas of Cel@UiO-66-NH₂ (1) and UiO-66-NH₂ (1) (331.20 and 343.82 m² g⁻¹) were more extensive than those of Cel@UiO-66-NH₂ (2) and UiO-66-NH₂ (2) (143.84 and 199.86 m² g⁻¹). In addition, the specific surface area of Cel@UiO-66-NH₂ (1) gradually decreased with increasing enzyme loading (Fig. S4†). This suggests that enzyme molecules occupy the pores of the MOFs, reducing the available surface area, particularly in the micropores and mesopores, while also indicating a higher degree of enzyme immobilization. Additionally, the formation of mesopores may result from the assembly of MOFs from nanoparticles, with interparticle interactions leading to the creation of larger pores.⁴⁰ Moreover, the results regarding the incremental pore volume of the samples showed that Cel@UiO-66-NH₂ (1) and UiO-66-NH₂ (1) had larger pore volumes than Cel@UiO-66-NH₂ (2) and UiO-66-NH₂ (2). This difference explains why UiO-66-NH₂ (1) could immobilize more enzymes than UiO-66-NH₂ (2). The larger pore volume provides the enzymes more space to move, which can improve their activity.¹⁴ To compare the effects of different metal precursors on enzyme immobilization, varying amounts of enzymes were added. The results showed that the encapsulation rate of Cel@UiO-66-NH₂ (1) consistently surpassed that of Cel@UiO-66-NH₂ (2) (Fig. 1g), with a maximum increase of 78%. As the amount of enzyme added increased, the enzyme activity gradually increased. The enzyme activity per unit mass in Cel@UiO-66-NH₂ (1) was always higher than that in Cel@UiO-66-NH₂ (2), with a maximum enhancement of 20% (Fig. 1h).

To further analyze the encapsulation of enzymes in MOFs, they were disintegrated by successively adding 3% hydrofluoric acid and ethylenediamine tetraacetic acid buffer (pH 10.0, 1.0 M). The supernatant obtained after sample elution and the solution following MOF disintegration were analyzed using sodium dodecyl sulfate–polyacrylamide gel electrophoresis (Fig. S5†). This analysis showed that for the digested Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2), bands corresponding to the molecular weights of 68, 48, and 35 kDa for Cel were observed, akin to free Cel. This affirms that the enzyme was encapsulated during grinding. In the supernatants of Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2), more unbound enzymes were observed in Cel@UiO-66-NH₂ (2), confirming that UiO-66-NH₂ (1) has a stronger affinity for enzymes than UiO-66-NH₂ (2).

3.1.2. Kinetic parameters and activity of immobilized enzymes. To fully assess the differences in catalytic performance between Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2), the enzyme kinetic parameters (Michaelis constant) and curves were analyzed. The values of V_max, K_m, K_cat, and K_cat/K_m for Cel@UiO-66-NH₂ (1), Cel@UiO-66-NH₂ (2), and Cel in the hydrolysis reaction are shown in Fig. 2a. In the hydrolysis reaction, Cel@UiO-66-NH₂ (1) showed a higher substrate affinity, as its K_m value (0.058 mM) was lower than that of Cel (0.072 mM) and Cel@UiO-66-NH₂ (2) (0.077 mM). Moreover, the K_cat/K_m of Cel@UiO-66-NH₂ (1) was 14.645 L s⁻¹ mmol⁻¹, higher than that of Cel@UiO-66-NH₂ (2), which was 9.588 L s⁻¹ mmol⁻¹. The enzyme kinetic profiles of Cel@UiO-66-NH₂ (1), Cel@UiO-66-NH₂ (2), and free Cel (Fig. 2b) demonstrated that the catalytic efficiency of Cel@UiO-66-NH₂ (1) was lower than that of free Cel but higher than that of Cel@UiO-66-NH₂ (2). Cel@UiO-66-NH₂ (1) retained approximately 68% of its enzyme activity. Details on the calculation of enzymatic catalytic efficiency can be found in the ESI.†

Fig. 2 Enzymatic activity and kinetics of enzyme@MOFs. (a) Calculated catalytic kinetic parameters of Cel@UiO-66-NH₂ (1), Cel@UiO-66-NH₂ (2), and Cel. (b) Plot of reaction velocity against the substrate for Cel@UiO-66-NH₂ (1), Cel@UiO-66-NH₂ (2), and Cel.

Furthermore, unlike the findings reported by Wei et al.,²¹ the organic solvent did not compromise enzyme activity during the direct synthesis of Cel@UiO-66-NH₂ (1) using the “one-step” method. In this process, enzyme activity was better preserved, which can be attributed to the biomimetic mineralization that occurred during the synthesis (Fig. 2a). Structural units formed around the nucleation of biomolecules, thereby preventing enzyme degradation by organic solvents.^41,42 In conclusion, these results indicate that dodecanuclear zirconium acetate clusters are used as metal precursors for synthesizing UiO-type MOFs. This synthesis helps regulate the spatial distribution of the enzyme and improve the immobilization efficiency and viability of the enzyme.

3.2. Validation of mechanochemical biomimetic mineralization

Biomimetic mineralization is the unique ability of amino acids, peptide fragments, and more intricate biological components to aggregate inorganic cations for seed biomineral formation.⁴³ The biomimetic mineralization of MOFs highlights the affinity of biological macromolecules for the structural units of the MOFs, which is generated through interactions such as intermolecular hydrogen bonding and hydrophobic interactions.⁴⁴ Further data analysis revealed that the power-law slope α of the SAXS fitted curve in the low Q region decreased from −2.82 to −3.3 following the addition of the enzyme. The power-law slope α in the high Q region increased from −3.32 to −3.16 (Fig. 3a and b). This change suggests that after the enzyme was added, many small pores (small size structure) in UiO-66-NH₂ (1) were transformed into large pores (big size structure); that is, the effect of biomimetic mineralization occurred.¹⁴ Therefore, it is hypothesized that Cel acts as a biomimetic mineralization agent, and the dodecanuclear clusters promote the biomimetic mineralization of Cel@UiO-66-NH₂ (1) as a synthetic feedstock.⁴⁵ However, this effect was not observed in Cel@UiO-66-NH₂ (2) (Fig. 3c and d). To further investigate the impact of grinding time on Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2), we measured the SAXS patterns at different grinding times. The results showed significant changes in the SAXS patterns of Cel@UiO-66-NH2 (1) with varying grinding time. In the q(A⁻¹) range of 0.005 to 0.05 and 0.05 to 0.15, the scattering intensity first increased and then decreased, while in the q range of 0.15 to 0.2, the scattering intensity continuously decreased. The changes in the slope were consistent with the trend of scattering intensity. The slope gradually increased in the range of 0.005 to 0.05, while in the 0.05 to 0.2 range, the slope increased initially and then gradually slowed down. This trend indicates that, with increasing grinding time, the pores in Cel@UiO-66-NH₂ (1) gradually formed and transformed, indicating the occurrence of biomimetic mineralization, and the crystal structure transitioned from disorder to order. In contrast, the SAXS patterns of Cel@UiO-66-NH₂ (2) did not show significant time-dependent regular changes was observed (Fig. 3e and f).⁴⁶

Fig. 3 Characterization of and tracking of the mechanochemical biomimetic mineralization process. (a–d) Representative Guinier approximations (red) for raw small-angle X-ray scattering spectra (black) at room temperature, (e and f) SAXS patterns of Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2) samples at 1, 10, and 20 min during mechanochemical biomimetic mineralization, showing time-dependent nanostructural evolution. (g and h) XRD patterns of Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2) at different time points (1–20 min, 2θ = 5–40°) during mechanochemical biomimetic mineralization, showing the evolution of crystallinity, (i) The XRD peaks corresponding to Miller indices (111), and (200) of Cel@UiO-66-NH₂ (1) all display correlated time-dependent intensity changes.

To investigate the biomimetic mineralization behavior during mechanochemical synthesis, time-resolved XRD patterns of Cel@UiO-66-NH₂ (1) and (2) were collected at various grinding times (1–20 min). As shown in Fig. 3g and i, sample (1) exhibited a progressive increase in the intensity of characteristic diffraction peaks, such as (111) and (200), reflecting a typical crystallization trajectory involving nucleation, growth, and stabilization. The sigmoidal trend in peak development over time suggests a biomimetic mineralization process, likely modulated by Cel through the formation of coordination intermediates and facilitation of nucleation.^46,47

In contrast, Cel@UiO-66-NH₂ (2) exhibited diffraction peaks from the initial time point (1 min), and the diffraction peaks were broad and low in intensity, accompanied by minimal changes in peak intensity throughout the subsequent time course (Fig. 3h). This behavior indicates a spontaneous, non-regulated crystallization process lacking enzymatic involvement, and without the sequential structural evolution characteristic of biomimetic mineralization.

Together, these results highlight the critical role of Cel in directing the stepwise formation of MOF structures, supporting the occurrence of enzyme-mediated biomimetic mineralization in sample (1), but not in sample (2).

The scanning electron microscopy images in Fig. 4c and d indicate that both composites have unique three-dimensional flower-like structures derived from MOFs with two-dimensional nanosheets. Specifically, Cel@UiO-66-NH₂ (1) is composed of regular rectangular nanosheets (Fig. 4c and S6†), while Cel@UiO-66-NH₂ (2) consists of needle-like nanosheets (Fig. 4d and S7†). The difference in morphology was attributed to the biomimetic mineralization of Cel@UiO-66-NH₂ (1).⁴⁸ Furthermore, the transmission electron microscopy images (Fig. S8†) showed more shadows in the image of Cel@UiO-66-NH₂ (2), which could be attributed to the poor separation of the lamellae of UiO-66-NH₂ (2). The dispersion of Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2) in water was observed separately (Fig. S8†), with Cel@UiO-66-NH₂ (1) exhibiting better water dispersion. The energy-dispersive X-ray spectrum revealed a homogeneous distribution of various elements, including C, O, N, Zr, S, and P, consistent with the previous conclusions obtained using confocal laser scanning microscopy evaluation (Fig. S9 and S10†). The biocompatibility of the two materials was compared by incubating with HaCaT cells (Fig. S11†). The results showed that UiO-66-NH₂ (2) was more cytotoxic than UiO-66-NH₂ (1). This increased cytotoxicity could be attributed to the corrosive and reactive nature of methacrylic acid, suggesting that UiO-66-NH₂ (1) is more suitable for biocatalysis.

Fig. 4 Characterization of the hierarchical pore structure. (a) Schematic illustration of different pore sizes present in Cel@UiO-66-NH₂ (1). (b) Two-dimensional small-angle X-ray scattering patterns of Cel@UiO-66-NH₂ (1) and scanning electron microscopy images of (c) Cel@UiO-66-NH₂ (1) and (d) Cel@UiO-66-NH₂ (2).

Furthermore, the N₂ adsorption-resolution isotherms of several materials were analyzed (Fig. S12†). The isotherms of UiO-66-NH₂ (1), Cel@UiO-66-NH₂ (1), UiO-66-NH₂ (2), and Cel@UiO-66-NH₂ (2) showed type IV curves. The hysteresis loops observed in the N₂ isotherms suggest a mesoporous structure, as verified by the pore size distribution (Fig. S13†). SAXS experiments are sensitive to structural changes within the mesostructural range.^49,50 Three regions were characterized using Guinier fits (computational model): Q = 0.01–0.03, 0.03–0.1, and 0.1–0.2 Å⁻¹. This analysis provided a detailed view of the hierarchical pore structure (Fig. 4b, and S14†). The results showed three types of structural spaces in Cel@UiO-66-NH₂ (1): radii of 3.7, 8.5, and 33.4 nm. The spaces with radii of 3.7 and 8.5 nm were used for enzyme storage. In contrast, those with a radius of 33.4 nm facilitated the entry and exit of substrates and products during the reaction between the enzyme and the substrate (Fig. 4a). Moreover, the results revealed that Cel in Cel@UiO-66-NH₂ (1) has a broader space for exchange with the external environment, compared with Cel@UiO-66-NH₂ (2).

These results indicate that dodecanuclear zirconium clusters significantly enhance the biomimetic mineralization of Cel@UiO-66-NH₂ (1) compared with hexanuclear zirconium methacrylate clusters, providing a larger internal space for enzyme immobilization.

3.3. Mechanism studies of the biomimetic mineralization in mechanochemistry

3.3.1. Adsorption kinetics of MOFs with enzymes. We investigated the mechanism through which dodecanuclear zirconium acetate clusters promote biomimetic mineralization by analyzing the differential binding kinetics between Cel and UiO-66-NH₂ (1) and UiO-66-NH₂ (2). Quartz crystal microbalance with dissipation monitoring (QCM-D) was used to assess the interactions between Cel and UiO-66-NH₂ (1) and UiO-66-NH₂ (2). QCM-D allows for simultaneous monitoring of changes in frequency (Δf) and energy dissipation (ΔD). These changes are linked to mass adsorption on the sensor surface and variations in the viscoelastic properties of the adsorbed layer, which can be observed as variations in Δf and ΔD for the different samples.^51,52 The results showed that dodecanuclear zirconium clusters had a higher affinity for enzymes than hexanuclear zirconium clusters over the same period. The binding rate of dodecanuclear zirconium clusters with enzyme molecules remained consistently higher than that of hexanuclear zirconium clusters during the 0–1000 s observation period, particularly in the early stages. At 200 s, the binding rate of dodecanuclear zirconium clusters with enzyme molecules was 1.93 ng cm⁻² s⁻¹, while hexanuclear zirconium clusters had a rate of 1.39 ng cm⁻² s⁻¹ (Fig. 5a and b). This indicates that the higher affinity and rapid adsorption capability of dodecanuclear zirconium clusters are key factors in promoting biomimetic mineralization.⁵³ Additionally, the same test was conducted on both MOFs. The results showed that both materials could bind a certain amount of enzyme. At adsorption saturation, UiO-66-NH₂ (2) exhibited a Δf decrease of −32 Hz, while UiO-66-NH₂ (1) showed a decrease of −40 Hz (Fig. S15†).^54,55 Subsequently, the Voigt-Voinova model was used to calculate the masses of different protein layers at saturation(Fig. 5c and d), confirming the same phenomenon. At saturation, the total enzyme adsorption was 575.91 ng cm⁻² for UiO-66-NH₂ (2) and 622.37 ng cm⁻² for UiO-66-NH₂ (1). To further validate our findings, two concentrations of cellulase solution were applied, and the flowing medium was switched from water to citrate buffer containing 10% ethanol. As shown in Fig. 5e–h, compared with UiO-66-NH₂ (2), UiO-66-NH₂ (1) exhibited a faster adsorption rate and a higher enzyme-binding capacity within a shorter time frame. Subsequently, the Voigt–Voinova model was used to estimate the mass of the adsorbed protein layer at saturation. The results revealed that the total enzyme adsorption capacities of UiO-66-NH₂ (2) were 648.94 and 660.04 ng cm⁻², while those of UiO-66-NH₂ (1) were 642.85 and 631.61 ng cm⁻², respectively (Fig. 5i–l).

Fig. 5 Multiple interactions between Cel and the structural units of metal–organic frameworks were investigated using QCM-D. (a) Frequency shift (Δf) over time as Cel flows over substrates capped with dodecanuclear or hexanuclear zirconium clusters. (b) Mass change profiles over time corresponding to (a). (c) Mass change profiles of Cel interacting with UiO-66-NH₂ (1). (d) Mass change profiles of Cel interacting with UiO-66-NH₂ (2). (e and f) Δf profiles showing adsorption and desorption behavior of Cel with UiO-66-NH₂ (1). (g and h) Δf profiles showing adsorption and desorption behavior of Cel with UiO-66-NH₂ (2). (c, i, and j) Mass change profiles of Cel interacting with UiO-66-NH₂ (1) under aqueous conditions with pH = 4.8 buffer as the flowing medium. (d, k, and l) Mass change profiles of Cel interacting with UiO-66-NH₂ (2) under aqueous conditions with pH = 4.8 buffer as the flowing medium.

These results suggest that dodecanuclear zirconium clusters enhance the affinity between biomolecules and structural units, increasing the local concentration of structural units. This promotes the formation of prenucleation clusters of MOF structural units around biomolecules, facilitating biomimetic mineralization.

3.3.2. Chemical bonding between MOFs and enzymes. X-ray photoelectron spectroscopy was used to investigate the chemical states on the surface and within a depth of 10 nm of Cel, UiO-66-NH₂ (1), UiO-66-NH₂ (2), Cel@UiO-66-NH₂ (1), and Cel@UiO-66-NH₂ (2) biocomposites. This analysis aimed to explore biomimetic mineralization through the lens of chemical bonding. The elements C, N, O, and Zr are present in Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2), as well as small amounts of S and P elements (Fig. S16†). The deconvoluted C 1s spectrum of UiO-66-NH₂ (1), illustrated in Fig. 6a, showed four different peaks at 288.8, 285.43, 284.61, and 283.82 eV. These peaks correspond to the O–C=O bond in the MOF backbone, the typical C–N bond of aromatic amines, the C–C and C=C bonds of aromatic rings, coordination carboxylic acids, and dodecanuclear zirconium cluster, respectively.⁵⁶ The analysis of the chemical bond ratios revealed an increase in the C=C bond ratio in Cel@UiO-66-NH₂ (1) compared with Cel@UiO-66-NH₂ (2) (Fig. 6i, e, and S14†). This increase is likely due to C=C bonds in the enzyme's R groups or the chemical bonds formed during the synthesis of the enzyme and UiO-66-NH₂ (1).³⁰ The deconvoluted O 1s spectrum of UiO-66-NH₂ (1) in Fig. 6b shows four different peaks at 533.58, 531.87, and 530.12 eV corresponding to the C–O bond in the MOF backbone, the C=O bond, and the Zr–O bond in the metal cluster.^57,58 As shown in the analysis of the chemical bond ratios (Fig. 6i, f, and S14†), Cel@UiO-66-NH₂ (1) showed an increased proportion of Zr–O bonds and a decreased proportion of C=O bonds compared with Cel@UiO-66-NH₂ (2). This suggests that chemical bonds were formed and transformed during the binding process between the enzyme and UiO-66-NH₂ (1), forming more Zr–O bonds. The deconvoluted Zr three-dimensional spectra of Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2), shown in Fig. 6d, h, and S16,† also confirmed this idea, and the ratio of the two spin–orbit splitting peaks changed with the addition of the enzyme. As shown in Fig. 6c, the deconvoluted N 1s spectrum of Cel@UiO-66-NH₂ (1) revealed two distinct peaks at 398.57 and 398.05 eV, corresponding to C–N bonds and –NH bonds in the enzyme and the MOF backbone, respectively.^31,59 Compared with Cel@UiO-66-NH₂ (2), the proportion of C–N bonds in Cel@UiO-66-NH₂ (1) increased, as shown in Fig. 6g, i, and S16.† XPS results confirmed the enhanced covalent binding between UiO-66-NH₂ (1) and the enzyme, including C=C, Zr–O, and C–N bonds. These interactions improve the stability of the biocomposites and facilitate biomimetic mineralization.

Fig. 6 X-ray photoelectron spectroscopy analysis reveals chemical bonding changes. (a and e) C 1s high-resolution spectra, (b and f) O 1s high-resolution spectra, (c and g) N 1s high-resolution spectra, and (d and h) Zr three-dimensional high-resolution spectra of UiO-66-NH₂ (1), Cel@UiO-66-NH₂ (1), UiO-66-NH₂ (2), and Cel@UiO-66-NH₂ (2). (i) Atomic percentages of Cel, UiO-66-NH₂ (1), and Cel@UiO-66-NH₂ (1) based on the X-ray photoelectron spectroscopy spectra.

3.3.3. Structural changes in enzymes. Furthermore, the structural changes in the enzyme during bionic mineralization were elucidated. Solid-state UV–Vis analysis showed two absorption peaks between 200 and 310 nm for Cel, Cel@UiO-66-NH₂ (1), and Cel@UiO-66-NH₂ (2), corresponding to different aromatic amino acids in the enzyme (Fig. 7a).⁶⁰ Furthermore, the second derivative spectrum was further explored; the enzyme showed two peaks at 307 and 278 nm, which were blue-shifted after being encapsulated into MOFs. The two peaks of Cel@UiO-66-NH₂ (1) were blue-shifted to 300 and 263 nm (Fig. 7b). This indicates that the side chain groups of the aromatic amino acids in the enzyme were surrounded by other non-polar amino acid residues, which diminished their interaction with polar molecules in aqueous solutions.⁶¹ To obtain more information about the changes in the Cel tertiary structure, the Cel/UiO biocomposite was examined using fluorescence spectroscopy. This technique reflects the degree of oxidation of aromatic amino acid residues and changes in the microenvironment.⁶² The λ_max values for Cel@UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (2) were blue-shifted to 331.6 and 334.8 nm, respectively, compared with 340.8 nm for free Cel. This indicates that the aromatic amino acids moved from a hydrophilic to a hydrophobic environment, which may lead to further folding of the enzyme's spatial structure (Fig. 7c).⁶³

Fig. 7 Computer-bound spectroscopy analysis to reveal the Cel conformational changes. (a and b) Solid-state UV–Vis spectra and the corresponding second derivative UV–vis spectra of Cel, Cel@UiO-66-NH₂ (1), and Cel@UiO-66-NH₂ (2). (c) Fluorescence spectra of Cel, Cel@UiO-66-NH₂ (1), and Cel@UiO-66-NH₂ (2). (d) Active site pocket and the amino acid residues of Cel interacting with the substrate.

Cel mixture from Trichoderma reesei, a multienzyme complex (they complete the hydrolysis process of cellulose through a progressive reaction), contains three components: exoglucanase (EC:3.2.1.91), β-glucosidase (EC:3.2.1.21), and endoglucanase (EC:3.2.1.4).^57,64 Molecular operating environment software shows (Fig. S17†) endoglucanase (EC:3.2.1.4) with a size of 4.55 nm × 4.98 nm × 4.93 nm, β-glucosidase (EC:3.2.1.21) with a size of 5.13 nm × 5.93 nm × 6.61 nm, and exoglucanase (EC:3.2.1.91) with a size of 5.98 nm × 7.41 nm × 7.38 nm. Furthermore, the active site pocket of the Cel mixture and the amino acid residues interacting with the substrate (Fig. 7c and S18†) show many aromatic amino acids distributed on the enzyme surface; for example, the aromatic amino acid residues interacting with the substrate in the active pocket of endoglucanase (EC:3.2.1.4) are Trp-231, Trp-41, Trp-162, etc. Combined with the above results (blue shift in the fluorescence and UV spectra following enzyme encapsulation), it is speculated that Cel has undergone further folding during the biomimetic mineralization process. As a result, the internal voids of the proteins appear to have decreased in volume, which creates smaller spaces.^65,66 This reduction allows the proteins to move more freely in the cavities of the MOFs, which is more conducive to binding to the MOF structural units and facilitates the onset of biomimetic mineralization.

In conclusion, our experimental results demonstrate that the enhanced affinity between the enzyme and the structural units of MOFs, the increase in C=C, Zr–O, and C–N bonds, and the change in enzyme structure promote mechanochemical mineralization.

3.4. Validation and application of biomimetic mineralization–mediated Cel protection

Finally, we investigated the stability of Cel@UiO-66-NH₂ (1) and compared it with Cel-on-UiO-66-NH₂ (1), exploring its ability to hydrolyze carboxymethylcellulose/microcrystalline cellulose and extract plant polysaccharides. To compare the conventional physical adsorption method with the mechanochemical approach, the Cel-on-UiO-66-NH₂ (1) composite was synthesized by co-incubating Cel with UiO-66-NH₂ (1) at room temperature. In the hydrolysis reaction, Cel-on-UiO-66-NH₂ (1) (19.25 mg) demonstrated higher catalytic activity than Cel@UiO-66-NH₂ (1) (16.53 mg). Furthermore, the K_cat/K_m ratio of Cel@UiO-66-NH₂ (1) was higher than that of Cel-on-UiO-66-NH₂ (1), indicating superior catalytic efficiency of Cel@UiO-66-NH₂ (1) (Fig. 2 and S19†). Thermogravimetric analysis data showed a decrease in the weight loss percentage of Cel@UiO-66-NH₂ (1) compared with Cel and Cel@UiO-66-NH₂ between 250 °C–400 °C, indicating improved thermal stability of Cel (Fig. 8a and S19†). Differential scanning calorimetry analysis (Fig. 8b)⁶⁷ showed that the denaturation temperature of the enzyme increased from 76 °C to 91 °C for Cel@UiO-66-NH₂ (1), whereas Cel-on-UiO-66-NH₂ (1) only increased to 86 °C. The results of the hydrolysis of carboxymethyl cellulose using Cel and Cel@UiO-66-NH₂ (1) at different temperatures (Fig. 8c) showed that Cel@UiO-66-NH₂ (1) provided better high-temperature protection than Cel and Cel-on-UiO-66-NH₂ (1). Cel and Cel-on-UiO-66-NH₂ (1) retained only 19.7% and 58% of their initial activity, respectively, and Cel@UiO-66-NH₂ (1) retained 91% of its original activity when exposed to 80 °C.³⁷ Transmission electron microscopy images taken after incubation at different temperatures showed no significant disintegration of Cel@UiO-66-NH₂ (1) (Fig. S20†). The fluorescence spectra of Cel@UiO-66-NH₂ (1) showed no significant change in the λ_max of Cel in Cel@UiO-66-NH₂ (1) after incubation (Fig. S21†). Additionally, compared with Cel and Cel-on-UiO-66-NH₂ (1), Cel@UiO-66-NH₂ (1) provided better protection against enzymatic activity in both strong acid and base environments (Fig. 8d).

Fig. 8 Reusability and stability of the enzyme@MOFs. (a) Thermogravimetric analysis curves, (b) differential scanning calorimetry curves, (c) thermal stability, (d) pH stability, (e) chemical stability, and (f and g) recyclability of Cel, Cel-on-UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (1); and (h) recyclability and (i) extraction/glucose yield of N. aurantialba polysaccharides extracted at 70 °C and using Cel and Cel@UiO-66-NH₂ (1).

In addition, we compared and explored the biological and cyclic stability of Cel and Cel@UiO-66-NH₂ (1) when hydrolyzing carboxymethyl cellulose. Urea can disrupt hydrophobic interactions by solventizing nonpolar groups inside the protein with water molecules, leading to protein unfolding and loss of its biological function.⁶⁸ Next, the enzymatic activities of free Cel, Cel-on-UiO-66-NH₂ (1) and Cel@UiO-66-NH₂ (1) were measured after 30 min of incubation in a 0.5 M urea solution (Fig. 8e). The results showed that the activities of free Cel and Cel-on-UiO-66-NH₂ (1) reduced to 46.5% and 81.3% of their initial activity (without urea), respectively, whereas the activity of Cel@UiO-66-NH₂ (1) decreased only to 96.0%. This finding suggested that MOFs restricted the enzyme, thereby reducing its capacity to undergo structural changes.⁶⁹ To further test the hypothesis, negative control experiments were performed by exposing the catalyst to the inhibitor 3-amino-1,2,4-triazole (3-AT). 3-AT does not denature Cel by unfolding but covalently binds to the enzyme and inhibits Cel without changing the structural conformation of the enzyme.⁷⁰ The results showed that the activities of free Cel and Cel-on-UiO-66-NH₂ (1) decreased to 52.9% and 91.7% of their initial levels (without 3-AT), respectively. The activity of Cel@UiO-66-NH₂ (1) only decreased to 94.7% (Fig. 8e). To evaluate the general applicability of this strategy, catalase (CAT) was also encapsulated into UiO-66-NH₂ (1). Upon the addition of 3-AT, the activity of free CAT decreased to 62.2% of its original level (without 3-AT), whereas CAT@UiO-66-NH₂ (1) retained 95.8% of its initial activity (Fig. S22†). In addition, the proteinase K assay further confirmed the successful encapsulation of Cel within UiO-66-NH₂ (1) (Fig. S23†). Moreover, the recoverability of Cel@UiO-66-NH₂ (1)-hydrolyzed carboxymethyl cellulose was evaluated. This showed that the catalytic activity of Cel@UiO-66-NH₂ (1) decreased continuously over 10 consecutive cycles, and Cel@UiO-66-NH₂ (1) successfully retained more than 50% of its catalytic activity after eight consecutive reuse cycles, whereas Cel-on-UiO-66-NH₂ (1) retained only 28% (Fig. 8f). Subsequently, the stability of Cel@UiO-66-NH₂ (1) was further investigated using QCM-D (Fig. 8g). QCM-D was used to detect whether the Cel encapsulated inside UiO-66-NH₂ (1) was stably present inside the material during the rinsing process. The results showed no significant mass drop, and a rise in the Δf signal was observed for the material over time. This indicates that the internally encapsulated enzyme is stable.

Finally, the recoverability and extraction efficiency of Cel@UiO-66-NH₂ (1)-hydrolyzed microcrystalline cellulose and extracted phytopolysaccharides were investigated. Developing efficient plant cell wall degrading enzymes is important for converting plant biomass and acquiring active ingredients in the industry.^71,72 Taking N. aurantialba as an example, it is a type of edible and medicinal mushroom in China, and its main active substance is a polysaccharide. N. aurantialba polysaccharides are usually obtained through high-temperature water extraction, and it is usually necessary to add Cel to hydrolyze its cell wall to accelerate polysaccharide dissolution. However, high temperatures can severely damage the activity of Cel, and its cost increases because it cannot be recycled.^73,74 Therefore, the recyclability of N. aurantialba polysaccharides extracted using Cel@UiO-66-NH₂ (1) was investigated. As shown in Fig. 8h, Cel@UiO-66-NH₂ (1) retained 54.5% enzyme activity after five cycles. N. aurantialba polysaccharides tend to be extracted using higher temperatures, which may impair the viability of the incorporated enzymes. Therefore, we explored the ability of composites to promote the solubilization of N. aurantialba polysaccharides at 70 °C. The results showed that Cel@UiO-66-NH₂ (1) exhibited a better catalytic ability at 70 °C than Cel, and the extraction of N. aurantialba polysaccharides reached up to 10.35% (Fig. 8i). Carboxymethyl cellulose and microcrystalline cellulose are modified or degraded cellulose products with different physicochemical properties and applications. The cellulase-catalyzed hydrolysis of microcrystalline cellulose is one of the most efficient methods for industrial biomass utilization and value addition. Therefore, we investigated the recoverability of hydrolyzed microcrystalline cellulose by Cel@UiO-66-NH₂ (1). The results showed that Cel@UiO-66-NH₂ (1) retained 50.9% enzyme activity after five cycles (Fig. 8h). Additionally, the hydrolysis efficiency of Cel@UiO-66-NH₂ (1) was higher than that of Cel at 5 h of hydrolysis time (Fig. 8i).

In conclusion, the results indicated that, compared to traditional physical adsorption methods, biomimetic mineralization facilitated the protective effect on biomacromolecules, preventing their biological, thermal, and chemical degradation and improving the stability and vitality of biocomposites. It can be better applied to hydrolysis of carboxymethyl cellulose and microcrystalline cellulose, extraction of plant polysaccharides, and other application scenarios.

4. Conclusions

This study reported the mechanochemical biomimetic mineralization process induced by dodecanuclear zirconium clusters as metal precursor–promoted biomolecules (Cel). The results are as follows: (I) UiO-66-NH₂ synthesized with dodecanuclear zirconium clusters as precursors improves the encapsulation rate and catalytic efficiency of Cel. (II) SAXS confirmed the mechanochemical biomimetic mineralization of Cel@UiO-66-NH₂ (1). (III) The occurrence of mechanochemical biomimetic mineralization can be attributed to the increase in the local concentration of MOF structural units around the enzyme, the conversion of chemical bonds, and the structural changes of the enzyme. (IV) Compared to traditional physical adsorption methods, the biological, thermal, and chemical stability of Cel is improved and can be applied in scenarios such as saccharification of carboxymethylcellulose/microcrystalline cellulose and high-temperature continuous production of N. aurantialba polysaccharides.

Author contributions

Xiaoyang Sun carried out the synthesis and characterization. Xiaoyang Sun and Linyu Nian analyzed the results. Xiaoyang Sun and Chongjiang Cao conceived and planned the study. All authors contributed to the preparation of manuscript. All authors have given approval to the final version of the manuscript.

Data availability

Data available on request from the authors. The data that support the findings of this study are available from the corresponding author, [Chong Jiang Cao, ccj33@163.com], upon reasonable request.

Conflicts of interest

The authors declare no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

The authors are gratefully thankful for the support from the National Key Research and Development Program of China (2022YFF1100600), Key Research and Development Program of Shandong Province (2021SFGC1205), and Youth Program of Natural Science Foundation of Jiangsu Province (BK20231008).

References

1. H. E. Schoemaker, D. Mink and M. G. Wubbolts, Science, 2003, 299, 1694–1697.
2. P. V. Iyer and L. Ananthanarayan, Process Biochem., 2008, 43, 1019–1032.
3. J.-M. Choi, S.-S. Han and H.-S. Kim, Biotechnol. Adv., 2015, 33, 1443–1454.
4. M. Imran, K. Zaman, A. A. Nassani, G. Dincă, H. u. R. Khan and M. Haffar, Geosci. Front., 2023, 101608.
5. P. Cao, H. Liu, D. Wu and X. Wang, Chem. Eng. J., 2021, 405, 126695.
6. M. Grzelczak, J. Vermant, E. M. Furst and L. M. Liz-Marzán, ACS Nano, 2010, 4, 3591–3605.
7. A. Sánchez-Iglesias, N. Winckelmans, T. Altantzis, S. Bals, M. Grzelczak and L. M. Liz-Marzán, J. Am. Chem. Soc., 2017, 139, 107–110.
8. G. González-Rubio, P. Díaz-Núñez, A. Rivera, A. Prada, G. Tardajos, J. González-Izquierdo, L. Bañares, P. Llombart, L. G. Macdowell, M. Alcolea Palafox, L. M. Liz-Marzán, O. Peña-Rodríguez and A. Guerrero-Martínez, Science, 2017, 358, 640–644.
9. T. Udayabhaskararao, T. Altantzis, L. Houben, M. Coronado-Puchau, J. Langer, R. Popovitz-Biro, L. M. Liz-Marzán, L. Vuković, P. Král, S. Bals and R. Klajn, Science, 2017, 358, 514–518.
10. Q. Li, Y. Wang, G. Zhang, R. Su and W. Qi, Chem. Soc. Rev., 2023, 52, 1549–1590.
11. H. Furukawa, K. E. Cordova, M. O'Keeffe and O. M. Yaghi, Science, 2013, 341, 1230444.
12. W. Liang, P. Wied, F. Carraro, C. J. Sumby, B. Nidetzky, C.-K. Tsung, P. Falcaro and C. J. Doonan, Chem. Rev., 2021, 121, 1077–1129.
13. X. Wang, P. C. Lan and S. Ma, ACS Cent. Sci., 2020, 6, 1497–1506.
14. K. Liang, R. Ricco, C. M. Doherty, M. J. Styles, S. Bell, N. Kirby, S. Mudie, D. Haylock, A. J. Hill, C. J. Doonan and P. Falcaro, Nat. Commun., 2015, 6, 7240.
15. G.-Y. Jeong, R. Ricco, K. Liang, J. Ludwig, J.-O. Kim, P. Falcaro and D.-P. Kim, Chem. Mater., 2015, 27, 7903–7909.
16. J.-D. Xiao, Q.-N. Wang, Z.-D. Feng, S. Tang, Y. Liu and C. Li, J. Catal., 2022, 406, 165–173.
17. S. Głowniak, B. Szczęśniak, J. Choma and M. Jaroniec, Mater. Today, 2021, 46, 109–124.
18. D. Tan and F. García, Chem. Soc. Rev., 2019, 48, 2274–2292.
19. K. K. Sarmah, T. Rajbongshi, A. Bhuyan and R. Thakuria, Chem. Commun., 2019, 55, 10900–10903.
20. H. He, H. Han, H. Shi, Y. Tian, F. Sun, Y. Song, Q. Li and G. Zhu, ACS Appl. Mater. Interfaces, 2016, 8, 24517–24524.
21. T.-H. Wei, S.-H. Wu, Y.-D. Huang, W.-S. Lo, B. P. Williams, S.-Y. Chen, H.-C. Yang, Y.-S. Hsu, Z.-Y. Lin, X.-H. Chen, P.-E. Kuo, L.-Y. Chou, C.-K. Tsung and F.-K. Shieh, Nat. Commun., 2019, 10, 5002.
22. B. Szczęśniak, S. Borysiuk, J. Choma and M. Jaroniec, Mater. Horiz., 2020, 7, 1457–1473.
23. B. Karadeniz, A. J. Howarth, T. Stolar, T. Islamoglu, I. Dejanovic, M. Tireli, M. C. Wasson, S.-Y. Moon, O. K. Farha and T. Friscic, ACS Sustainable Chem. Eng., 2018, 6, 15841–15849.
24. A. M. Fidelli, B. Karadeniz, A. J. Howarth, I. Huskić, L. S. Germann, I. Halasz, M. Etter, S.-Y. Moon, R. E. Dinnebier, V. Stilinović, O. K. Farha, T. Friščić and K. Užarević, Chem. Commun., 2018, 54, 6999–7002.
25. K. Užarević, T. C. Wang, S.-Y. Moon, A. M. Fidelli, J. T. Hupp, O. K. Farha and T. Friščić, Chem. Commun., 2016, 52, 2133–2136.
26. H. Ali-Moussa, R. N. Amador, J. Martinez, F. Lamaty, M. Carboni and X. Bantreil, Mater. Lett., 2017, 197, 171–174.
27. M. Huo, L. Wang, Y. Chen and J. Shi, Nat. Commun., 2017, 8, 357.
28. G. Beaucage, J. Appl. Crystallogr., 1996, 29, 134–146.
29. J. Andersen and J. Mack, Green Chem., 2018, 20, 1435–1443.
30. Y. Cheng, O.-M. Lai, C.-P. Tan, W. Panpipat, L.-Z. Cheong and C. Shen, ACS Appl. Mater. Interfaces, 2021, 13, 4146–4155.
31. N. S. Mohammadi, M. S. Khiabani, B. Ghanbarzadeh and R. R. Mokarram, J. Biotechnol., 2020, 323, 189–202.
32. I. N. Ahmed, X.-L. Yang, A. A. Dubale, R.-F. Li, Y.-M. Ma, L.-M. Wang, G.-H. Hou, R.-F. Guan and M.-H. Xie, Bioresour. Technol., 2018, 270, 377–382.
33. A. Barth, Biochim. Biophys. Acta, Bioenerg., 2007, 1767, 1073–1101.
34. M. Jackson and H. H. Mantsch, Crit. Rev. Biochem. Mol. Biol., 1995, 30, 95–120.
35. J. B. Costa, M. J. Lima, M. J. Sampaio, M. C. Neves, J. L. Faria, S. Morales-Torres, A. P. Tavares and C. G. Silva, Chem. Eng. J., 2019, 355, 974–985.
36. B. Jiang, Q. Zhao, H. Shan, Y. Guo, X. Xu, D. J. McClements, C. Cao and B. Yuan, J. Agric. Food Chem., 2022, 70, 6540–6551.
37. W. Liang, H. Xu, F. Carraro, N. K. Maddigan, Q. Li, S. G. Bell, D. M. Huang, A. Tarzia, M. B. Solomon, H. Amenitsch, L. Vaccari, C. J. Sumby, P. Falcaro and C. J. Doonan, J. Am. Chem. Soc., 2019, 141, 2348–2355.
38. M. Moradi, R. Razavi, A. K. Omer, A. Farhangfar and D. J. McClements, Trends Food Sci. Technol., 2022, 120, 75–87.
39. J. Liang, M. Y. Bin Zulkifli, J. Yong, Z. Du, Z. Ao, A. Rawal, J. A. Scott, J. R. Harmer, J. Wang and K. Liang, J. Am. Chem. Soc., 2022, 144, 17865–17875.
40. H. Xu, J. Han, B. Zhao, R. Sun, G. Zhong, G. Chen, Y. Yamauchi and B. Guan, Nat. Commun., 2023, 14, 8062.
41. Y. Pan, H. Li, M. Lenertz, Y. Han, A. Ugrinov, D. Kilin, B. Chen and Z. Yang, Green Chem., 2021, 23, 4466–4476.
42. P. Deria, J. Yu, T. Smith and R. P. Balaraman, J. Am. Chem. Soc., 2017, 139, 5973–5983.
43. G. Wang, R.-Y. Cao, R. Chen, L. Mo, J.-F. Han, X. Wang, X. Xu, T. Jiang, Y.-Q. Deng, K. Lyu, S.-Y. Zhu, E. D. Qin, R. Tang and C.-F. Qin, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 7619–7624.
44. A. Lukton, Nature, 1961, 192, 422–424.
45. R. Murty, M. K. Bera, I. M. Walton, C. Whetzel, M. R. Prausnitz and K. S. Walton, J. Am. Chem. Soc., 2023, 145, 7323–7330.
46. S.-W. Lin, P. K. Lam, C.-T. Wu, K.-H. Su, C.-F. Sung, S.-R. Huang, J.-W. Chang, O. Shih, Y.-Q. Yeh, T. H. Vo, H.-K. Tsao, H.-T. Hsieh, U. S. Jeng, F.-K. Shieh and H.-C. Yang, ACS Nano, 2024, 18, 25170–25182.
47. H. Jiang, Q. Wang, H. Wang, Y. Chen and M. Zhang, Catal. Commun., 2016, 80, 24–27.
48. W. Jiang, Z.-b. Qu, P. Kumar, D. Vecchio, Y. Wang, Y. Ma, J. H. Bahng, K. Bernardino, W. R. Gomes, F. M. Colombari, A. Lozada-Blanco, M. Veksler, E. Marino, A. Simon, C. Murray, S. R. Muniz, A. F. de Moura and N. A. Kotov, Science, 2020, 368, 642–648.
49. S.-Y. Chen, W.-S. Lo, Y.-D. Huang, X. Si, F.-S. Liao, S.-W. Lin, B. P. Williams, T.-Q. Sun, H.-W. Lin, Y. An, T. Sun, Y. Ma, H.-C. Yang, L.-Y. Chou, F.-K. Shieh and C.-K. Tsung, Nano Lett., 2020, 20, 6630–6635.
50. X. Lian, Y.-P. Chen, T.-F. Liu and H.-C. Zhou, Chem. Sci., 2016, 7, 6969–6973.
51. J. Y. Chen, L. S. Penn and J. Xi, Biosens. Bioelectron., 2018, 99, 593–602.
52. S. Gentil, S. M. Che Mansor, H. Jamet, S. Cosnier, C. Cavazza and A. Le Goff, ACS Catal., 2018, 8, 3957–3964.
53. H.-E. Lee, J. Lee, M. Ju, H.-Y. Ahn, Y. Y. Lee, H.-S. Jang and K. T. Nam, Mol. Syst. Des. Eng., 2018, 3, 581–590.
54. P. Del Pino, B. Pelaz, Q. Zhang, P. Maffre, G. U. Nienhaus and W. J. Parak, Mater. Horiz., 2014, 1, 301–313.
55. T. Kopac, Int. J. Biol. Macromol., 2021, 169, 290–301.
56. S. He, T. Zeng, S. Wang, H. Niu and Y. Cai, ACS Appl. Mater. Interfaces, 2017, 9, 2959–2965.
57. J. M. Yarbrough, R. Zhang, A. Mittal, T. Vander Wall, Y. J. Bomble, S. R. Decker, M. E. Himmel and P. N. Ciesielski, ACS Nano, 2017, 11, 3101–3109.
58. B. Qi, J. Luo and Y. Wan, Bioresour. Technol., 2018, 268, 577–582.
59. Z. Liu, H. Wang, J. Ou, L. Chen and M. Ye, J. Hazard. Mater., 2018, 355, 145–153.
60. M. Cao, A. Cao, J. Wang, L. Cai, J. Regenstein, Y. Ruan and X. Li, Food Chem., 2018, 266, 498–507.
61. M. Zhang, F. Li, X. Diao, B. Kong and X. Xia, Meat Sci., 2017, 133, 10–18.
62. X. Sui, H. Sun, B. Qi, M. Zhang, Y. Li and L. Jiang, Food Chem., 2018, 245, 871–878.
63. F. Hillger, D. Hänni, D. Nettels, S. Geister, M. Grandin, M. Textor and B. Schuler, Angew. Chem., Int. Ed., 2008, 47, 6184–6188.
64. S. Paul, M. Gupta, K. Dey, A. K. Mahato, S. Bag, A. Torris, E. B. Gowd, H. Sajid, M. A. Addicoat, S. Datta and R. Banerjee, Chem. Sci., 2023, 14, 6643–6653.
65. J.-B. Rouget, T. Aksel, J. Roche, J.-L. Saldana, A. E. Garcia, D. Barrick and C. A. Royer, J. Am. Chem. Soc., 2011, 133, 6020–6027.
66. J. Navarro-Sánchez, N. Almora-Barrios, B. Lerma-Berlanga, J. J. Ruiz-Pernía, V. A. Lorenz-Fonfria, I. Tuñón and C. Martí-Gastaldo, Chem. Sci., 2019, 10, 4082–4088.
67. A. Rutenberg, V. V. Vinogradov and D. Avnir, Chem. Commun., 2013, 49, 5636–5638.
68. P. Li, S.-Y. Moon, M. A. Guelta, L. Lin, D. A. Gómez-Gualdrón, R. Q. Snurr, S. P. Harvey, J. T. Hupp and O. K. Farha, ACS Nano, 2016, 10, 9174–9182.
69. F.-S. Liao, W.-S. Lo, Y.-S. Hsu, C.-C. Wu, S.-C. Wang, F.-K. Shieh, J. V. Morabito, L.-Y. Chou, K. C.-W. Wu and C.-K. Tsung, J. Am. Chem. Soc., 2017, 139, 6530–6533.
70. M. Ueda, H. Kinoshita, T. Yoshida, N. Kamasawa, M. Osumi and A. Tanaka, FEMS Microbiol. Lett., 2003, 219, 93–98.
71. W. Ren, D. Zhang, Y. Zhou, H. Wang, L. Xia, C. Tan, F. Guo, X. Zhang, R. Yang and Y. Yu, Chem. Eng. J., 2024, 490, 151422.
72. Y. Wang, A. Wang, S. Liu, L. Li, Y. Liu, Y. Liu, G. Nie, S. Nie, S. Yao and H. Yu, Chem. Eng. J., 2024, 489, 151439.
73. T. Sun, R. Wang, D. Sun, S. Li, H. Xu, Y. Qiu, P. Lei, L. Sun, X. Xu and Y. Zhu, Bioresour. Technol., 2020, 318, 124268.
74. T. Sun, X. Xu, Y. Ma, H. Jiang, K. Yang, R. Wang, Y. Gu, S. Li, Y. Qiu, D. Sun, H. Xu and P. Lei, Food Hydrocolloids, 2023, 142, 108848.

---

Footnotes

† Electronic supplementary information (ESI) available: Fig. S1–S18: FTIR spectra, CLSM images, particle size distribution, SDS-PAGE images, N₂ adsorption–desorption isotherms, pore size distributions, SAXS spectra, SEM images, TEM images, EDX images, cytotoxicity, spatial structure of Cel, active site pocket, TGA curves, and fluorescence spectra. See DOI: https://doi.org/10.1039/d5gc00628g

‡ These authors contributed equally to this work.

---