Magnetic Nanozyme Based on Loading Nitrogen-Doped Carbon Dots on Mesoporous Fe3O4 Nanoparticles for the Colorimetric Detection of Glucose

The simple and accurate monitoring of blood glucose level is of great significance for the prevention and control of diabetes. In this work, a magnetic nanozyme was fabricated based on loading nitrogen-doped carbon dots (N-CDs) on mesoporous Fe3O4 nanoparticles for the colorimetric detection of glucose in human serum. Mesoporous Fe3O4 nanoparticles were easily synthesized using a solvothermal method, and N-CDs were then prepared in situ and loaded on the Fe3O4 nanoparticles, leading to a magnetic N-CDs/Fe3O4 nanocomposite. The N-CDs/Fe3O4 nanocomposite exhibited good peroxidase-like activity and could catalyze the oxidation of the colorless enzyme substrate 3,3′,5,5′-tetramethylbenzidine (TMB) to blue TMB oxide (ox-TMB) in the presence of hydrogen peroxide (H2O2). When the N-CDs/Fe3O4 nanozyme was combined with glucose oxidase (Gox), Gox catalyzed the oxidization of glucose, producing H2O2 and leading to the oxidation of TMB under the catalysis of the N-CDs/Fe3O4 nanozyme. Based on this mechanism, a colorimetric sensor was constructed for the sensitive detection of glucose. The linear range for glucose detection was from 1 to 180 μM, and the limit of detection (LOD) was 0.56 μM. The recovered nanozyme through magnetic separation showed good reusability. The visual detection of glucose was also realized by preparing an integrated agarose hydrogel containing the N-CDs/Fe3O4 nanozyme, glucose oxidase, and TMB. The colorimetric detection platform has an enormous potential for the convenient detection of metabolites.


Introduction
Diabetes is a metabolic disease characterized by an elevated glucose (Glu) concentration in the blood, which affects millions of people around the world [1]. A long-term high blood glucose may lead to organ failure and tissue damage, causing various complications such as blindness, cardiovascular disease, and kidney failure [2]. Therefore, the accurate and timely monitoring of blood glucose level is of great significance for the prevention and control of diabetes. With the development of clinical testing, daily home-based health monitoring, and personalized treatment, the convenient glucose detection has attracted widespread attention. On the one hand, glucose detection using body fluid samples such as sweat, skin interstitial fluid, tears, saliva, and urine as substitutes for blood is widely researched. On the other hand, methods such as visual, continuous, non-invasive, or wearable device-based glucose monitoring is constantly investigated. Until now, various technologies have been developed for the detection of glucose, such as electrochemical sensors, colorimetric methods, fluorescence methods, and chemiluminescence methods [3][4][5][6]. simple synthesis, low cost, high catalytic ability, and good reusability, indicating its great potential for the convenient detection of glucose.

Strategy for the Colorimetric Detection of Glucose Based on the N-CDs/Fe 3 O 4 Magnetic Nanozyme
Mesoporous materials have high specific surface area, regular and ordered pore structure, narrow pore size distribution, and continuously adjustable pore size, which play an important role in adsorption, separation, catalytic reactions, and other applications [35][36][37][38]. As illustrated in Figure 1, mesoporous Fe 3 O 4 nanomaterial was firstly synthesized using a solvothermal method [39]. Subsequently, nitrogen-doped carbon dots (N-CDs) with peroxidase-like activity were synthesized in situ using a hydrothermal method in the presence of the Fe 3 O 4 nanomaterial. A common nitrogen-containing amino acid, histidine, was used as the carbon source for the synthesis of N-CDs. The as-prepared N-CDs/Fe 3 O 4 nanocomposite has the advantages of a high peroxidase-like activity, magnetic separation ability, simple preparation, and low cost. By combining the production of H 2 O 2 by the glucose oxidase (Gox)-catalyzed oxidation of glucose, the colorimetric detection of glucose could be achieved. preparing an integrated agarose hydrogel containing the N-CDs/Fe3O4 nanozyme, glucose oxidase, and TMB. The developed magnetic nanozyme has the advantages of simple synthesis, low cost, high catalytic ability, and good reusability, indicating its great potential for the convenient detection of glucose.

Strategy for the Colorimetric Detection of Glucose Based on the N-CDs/Fe3O4 Magnetic Nanozyme
Mesoporous materials have high specific surface area, regular and ordered pore structure, narrow pore size distribution, and continuously adjustable pore size, which play an important role in adsorption, separation, catalytic reactions, and other applications [35][36][37][38]. As illustrated in Figure 1, mesoporous Fe3O4 nanomaterial was firstly synthesized using a solvothermal method [39]. Subsequently, nitrogen-doped carbon dots (N-CDs) with peroxidase-like activity were synthesized in situ using a hydrothermal method in the presence of the Fe3O4 nanomaterial. A common nitrogen-containing amino acid, histidine, was used as the carbon source for the synthesis of N-CDs. The as-prepared N-CDs/Fe3O4 nanocomposite has the advantages of a high peroxidase-like activity, magnetic separation ability, simple preparation, and low cost. By combining the production of H2O2 by the glucose oxidase (Gox)-catalyzed oxidation of glucose, the colorimetric detection of glucose could be achieved.

Characterization of N-CDs/Fe3O4 Magnetic Nanozyme
The N2 adsorption-desorption technology was used to explore the pore structure of the produced Fe3O4 nanomaterials. Figure 2a shows the N2 adsorption-desorption isotherm of the Fe3O4 nanomaterial. As seen, the isotherm was a typical type IV isotherm with an H3 hysteresis loop, indicating the mesoporous nature of the Fe3O4 nanomaterial [40]. Based on the BJH method, two types of mesoporous structures were revealed. One, with an average diameter of 6.5 nm, corresponded to pores formed between the stacked Fe3O4 nanoparticles. The other, with an average diameter of 7 nm, indicated the mesopores on the surface of the nanoparticles. Thus, the as-synthesized Fe3O4 is a mesoporous nanomaterial.

Characterization of N-CDs/Fe 3 O 4 Magnetic Nanozyme
The N 2 adsorption-desorption technology was used to explore the pore structure of the produced Fe 3 O 4 nanomaterials. Figure 2a shows the N 2 adsorption-desorption isotherm of the Fe 3 O 4 nanomaterial. As seen, the isotherm was a typical type IV isotherm with an H3 hysteresis loop, indicating the mesoporous nature of the Fe 3 O 4 nanomaterial [40]. Based on the BJH method, two types of mesoporous structures were revealed. One, with an average diameter of 6.5 nm, corresponded to pores formed between the stacked Fe 3 O 4 nanoparticles. The other, with an average diameter of 7 nm, indicated the mesopores on the surface of the nanoparticles. Thus, the as-synthesized Fe 3 O 4 is a mesoporous nanomaterial.
The N-CDs/Fe 3 O 4 nanocomposite was characterized by UV-visible absorption spectroscopy. As shown in Figure 2b, the absorption curve of mesoporous Fe 3 O 4 exhibited a weak absorption peak at around 280 nm, which was attributed to the n-π* transition of the C=O bond in the precursor of sodium acetate. The spectra of N-CDs, showed the characteristic absorption peaks at~220 nm and 280 nm resulting from the π-π* transition of the C=C bond and the n-π* transition of the C=O bond, respectively. The N-CDs/Fe 3   The N-CDs/Fe3O4 nanocomposite was characterized by UV-visible absorption spectroscopy. As shown in Figure 2b, the absorption curve of mesoporous Fe3O4 exhibited a weak absorption peak at around 280 nm, which was attributed to the n-π* transition of the C=O bond in the precursor of sodium acetate. The spectra of N-CDs, showed the characteristic absorption peaks at ~220 nm and 280 nm resulting from the π-π* transition of the C=C bond and the n-π* transition of the C=O bond, respectively. The N-CDs/Fe3O4 nanocomposite exhibited the characteristic absorption peaks of both Fe3O4 and N-CDs, indicating the effective preparation of the nanocomposite. In addition, the absorption peak strength of the N-CDs/Fe3O4 nanocomposite at 280 nm was significantly higher than that of N-CDs and Fe3O4 at 280 nm. This might be due to the enrichment of N-CDs on the surface of the Fe3O4 nanoparticles.
The elemental and chemical groups on the surface of N-CDs/Fe3O4 were characterized by X-ray photoelectron spectroscopy (XPS). Figure 2c shows the XPS survey spectrum of N-CDs/Fe3O4, which revealed four elements including Fe, C, N, and O with the corresponding element contents of 1.8%, 64.8%, 7.9%, and 25.5%. ICP-OES was also applied to accurately analyze the Fe content in N-CDs/Fe3O4. The mass fraction ratio of Fe was 61.4%. This indicated that the surface of mesoporous Fe3O4 was enveloped by N-CDs. Thus, the iron content in the entire nanocomposite determined by ICP-OES was remarkably higher than the iron content on the surface determined by XPS. Figure 3a shows the high-resolution C1s spectrum of N-CDs/Fe3O4. The peaks with binding energy of 284.6 eV, 286.0 eV, and 288.0 eV correspond to the structure of graphite C (C-C=C, sp 2 carbon), the C-N/C-O bond, or the C=N/C=O bond. The corresponding high-resolution N1s spectrum in Figure 3b displays peaks with binding energy of 399.5 eV, 400.8 eV, and 401.5 eV, corresponding to the C-N-C bond, the C=N-C bond, and the N-H bond, respectively. Figure  3c shows the high-resolution O1s spectrum of N-CDs/Fe3O4. The peaks at 529.9 eV, 531.3 eV, 532.3 eV, and 533.6 eV are related to the Fe-O bond, the O-H bond, the C=O, and the C-OH bond, respectively. The characteristic peaks in the high-resolution Fe2p spectrum of N-CDs/Fe3O4 confirmed the presence of Fe 2+ and Fe 3+ in N-CDs/Fe3O4 (Figure 3d) [41,42]. The above results indicated that N-CDs were successfully modified on the surface of the mesoporous Fe3O4 nanomaterial, demonstrating the successful formation of the nanocomposite. The elemental and chemical groups on the surface of N-CDs/Fe 3 O 4 were characterized by X-ray photoelectron spectroscopy (XPS). Figure 2c shows the XPS survey spectrum of N-CDs/Fe 3 O 4 , which revealed four elements including Fe, C, N, and O with the corresponding element contents of 1.8%, 64.8%, 7.9%, and 25.5%. ICP-OES was also applied to accurately analyze the Fe content in N-CDs/Fe 3 O 4 . The mass fraction ratio of Fe was 61.4%. This indicated that the surface of mesoporous Fe 3 O 4 was enveloped by N-CDs. Thus, the iron content in the entire nanocomposite determined by ICP-OES was remarkably higher than the iron content on the surface determined by XPS. Figure 3a shows the high-resolution The functional groups of N-CDs/Fe3O4 were also characterized by Fourier transform infrared spectroscopy (FT-IR). As shown in Figure 4a, N-CDs/Fe3O4 exhibited the characteristic peaks of both N-CDs and Fe3O4. The absorption peak at around 3400 cm −1 was attributed to the stretching vibration of O-H, and the peak at 1628 cm −1 was attributed to the stretching vibration of C=O. The peaks at 1534 cm −1 and 1357 cm −1 resulted from the stretching vibration of C=N and C-N, respectively. The absorption peak around 580 cm −1 indicated the stretching vibration of Fe-O. The C=O stretching might be attributed to the residual acetate group on the surface of the Fe3O4 nanoparticles, which derived from the raw material acetate. The presence of these characteristic groups proved the successful preparation of the N-CDs/Fe3O4 nanocomposite. In addition, the abundant oxygen-containing and nitrogen-containing functional groups on N-CDs/Fe3O4 will favor the good dispersion of the magnetic nanocomposite in water.
The crystal structure of N-CDs/Fe3O4 was characterized by X-ray diffraction (XRD). The results are displayed in Figure 4b [39,42,43]. This phenomenon indicated that the modification of the Fe3O4 nanocomposite by N-CDs did not affect the inherent crystal structure of Fe3O4.  The ferromagnetic properties of N-CDs/Fe3O4 were characterized by measuring the hysteresis loop. As shown in Figure 4c, the saturation magnetization of N-CDs/Fe3O4 was 40.9 emu/g, indicating the magnetic properties of the N-CDs/Fe3O4 nanocomposite. The mass fraction of N-CDs in N-CDs/Fe3O4 was characterized by thermogravimetric analysis. As shown in Figure 4d, N-CDs/Fe3O4 exhibited good thermal stability below 550 °C. When the temperature changed from 550 °C to 800 °C, the mass of N-CDs/Fe3O4 decreased sharply to maintain an equilibrium. In comparison with Fe3O4, a mass ratio of 13.2% was obtained resulting from the thermal decomposition of N-CDs. The morphology and size of N-CDs/Fe3O4 were characterized by transmission electron microscopy (TEM). As shown in Figure 5, N-CDs/Fe3O4 appeared spherical in shape, with a particle size of ~130 nm.

Peroxide-Like Activity of N-CDs/Fe3O4
The peroxidase-like activity of N-CDs/Fe3O4 was investigated by catalyzing the oxidation of TMB by H2O2. Figure 6a shows the absorbance values of different solutions at 652 nm when the reaction time changed. As seen, the absorbance value of the solution

Peroxide-Like Activity of N-CDs/Fe3O4
The peroxidase-like activity of N-CDs/Fe3O4 was investigated by catalyzing the oxidation of TMB by H2O2. Figure 6a shows the absorbance values of different solutions at 652 nm when the reaction time changed. As seen, the absorbance value of the solution remained almost unchanged when only N-CDs/Fe3O4 and TMB were present in the solution, indicating no reaction between N-CDs/Fe3O4 and TMB. When TMB was mixed with H2O2, a slight increase in the absorbance of the solution was observed due to the slow oxidation of TMB by H2O2. When N-CDs/Fe3O4 was mixed with TMB and H2O2, the absorbance of the solution significantly increased. After 10 min of reaction, the absorption spectrum of the solution in Figure 6b also confirmed this phenomenon, proving that N-CDs/Fe3O4 can catalyze the oxidation of TMB by H2O2 to produce a blue TMB oxide (ox-TMB). To investigate the possible synergistic effect of N-CDs and the Fe3O4 nanoparticles, the catalytic activity of the N-CDs and Fe3O4 nanoparticles was investigated separately. To keep up with the content of N-CDs and Fe3O4 nanoparticles in the nanocomposite, the concentration of N-CDs and Fe3O4 nanoparticles was 13.2% and 86.8% of that of the nanocomposite, respectively. As shown in Figure 6c, when N-CDs and the Fe3O4 nanoparticles  Figure 6c, when N-CDs and the Fe 3 O 4 nanoparticles were present alone, they could catalyze the oxidation of TMB by H 2 O 2 , but the absorbance of the reaction solution at 652 nm was not high, indicating low nanozyme activity. It has been proven that the introduction of an N atom to the graphitic carbon structures of CDs can enhance the peroxidase-like activity owing to the enhanced electrical properties and affinity towards substrates [16,23,26]. At the same time, the catalytic activity of Fe 3 O 4 originates from Fe 2+ on the surface of the nanoparticles, which undergoes Fenton or Fenton-like reactions, resulting in the oxidation of TMB by H 2 O 2 [44]. Even if the two were mixed, the absorbance of the solution after reaction was significantly lower than that of the solution obtained using the N-CDs/Fe 3 O 4 nanocomposite. This phenomenon proved the coordinated effect of N-CDs and Fe 3 O 4 nanoparticles combined in a nanocomposite, leading to a significant improvement in the catalytic ability. This might be attributed to the spatial confinement of N-CDs by the mesoporous structure of the Fe 3 O 4 nanocomposite.
The Michaelis constant (Km) and maximum reaction rate constant (Vmax) are important parameters to evaluate the activity of nanozymes. Hydrogen peroxide and TMB were applied as substrates to measure the Km and Vmax of the nanozyme using the Lineweaver-Burk curves (Figure 7). The Km using TMB as the substrate was 0.607 mM, and the Vmax was 2.054 × 10 −7 M/s (Figure 7a,c). The Km value obtained using H 2 O 2 as the substrate was 0.719 mM, and the Vmax was 1.032 × 10 −8 M/s (Figure 7b,d).
confinement of N-CDs by the mesoporous structure of the Fe3O4 nanocomposite.
The Michaelis constant (Km) and maximum reaction rate constant (Vmax) are important parameters to evaluate the activity of nanozymes. Hydrogen peroxide and TMB were applied as substrates to measure the Km and Vmax of the nanozyme using the Lineweaver-Burk curves (Figure 7). The Km using TMB as the substrate was 0.607 mM, and the Vmax was 2.054 × 10 −7 M/s (Figure 7a,c). The Km value obtained using H2O2 as the substrate was 0.719 mM, and the Vmax was 1.032 × 10 −8 M/s (Figure 7b,d). Commonly, a high Vmax suggests high catalytic activity, whereas a small Km indicates a high affinity between enzymes and substrates. Table 1 reports the comparison between Km and Vmax of different nanozymes [44][45][46][47][48]. As shown, the Km obtained using the developed N-CDs/Fe 3 O 4 was lower than that obtained with the natural enzyme horseradish peroxidase (HRP) [44] when H 2 O 2 was applied as the substrate. The Km was also lower than that obtained using N-CD [46], Fe 3 O 4 [44], or the Fe-doped carbon-dot (Fe-CD) [48] nanozyme, but higher than that obtained using a metal oxide hybrid with nitrogen-doped carbon dots [45] or Fe-and N-co-doped CDs (Fe, N-CDs) [47]. In addition, the Vmax was the highest with TMB as the substrate.
Under optimal reaction conditions, the required amount of nanozyme was 1U when 1 µmol of the substrate was used and converted into the product. The specific activity (U/mg) refers to the number of activity units per unit mass of nanozyme. The specific activity of N-CDs/Fe 3 O 4 was 13.1 U/mg, which was remarkably higher than that of Fe 3 O 4 (0.215 U/mg). Thus, the modification of mesoporous Fe 3 O 4 using N-CDs could significantly enhance the nanozyme activity.

Optimization of the Detection Conditions
To obtain the optimal nanozyme reaction conditions, the reaction temperature and pH were optimized. As shown in Figure 8a, the absorbance value of the mixed solution (nanozyme + TMB + H 2 O 2 ) at 652 nm reached its maximum at pH 4, indicating that N-CDs/Fe 3 O 4 had the best peroxidase-like activity in this condition. The absorbance value increased when the reaction increased from 20 • C to 45 • C (Figure 8b). Although the absorbance of the mixture at 45 • C was greater than at 40 • C, the increase was not remarkably high. Considering that 40 • C is close to the physiological temperature, it was chosen for the subsequent experiments.

Colorimetric Detection of Glucose
Glucose is the main energy source of the human body. Long-term hyperglycemia is likely to lead to the occurrence of diabetes, increasing the risk of heart disease, stroke, blindness, renal failure, peripheral neuropathy, etc. Thus, the effective monitoring of blood glucose is of great significance for the diagnosis of hyperglycemia. In this work, Gox was applied to catalyze the oxidation of glucose to form gluconic acid and hydrogen peroxide. At the same time, H2O2 was decomposed into reactive oxygen species (ROS) under the catalysis of the peroxidase-like enzyme N-CDs/Fe3O4 which oxidized the colorless TMB to the blue ox-TMB. The reaction process was monitored by measuring the absorption of the mixture containing Gox, nanozyme, TMB, H2O2, and different concentrations of glucose.

Colorimetric Detection of Glucose
Glucose is the main energy source of the human body. Long-term hyperglycemia is likely to lead to the occurrence of diabetes, increasing the risk of heart disease, stroke, blindness, renal failure, peripheral neuropathy, etc. Thus, the effective monitoring of blood glucose is of great significance for the diagnosis of hyperglycemia. In this work, Gox was applied to catalyze the oxidation of glucose to form gluconic acid and hydrogen peroxide. At the same time, H 2 O 2 was decomposed into reactive oxygen species (ROS) under the catalysis of the peroxidase-like enzyme N-CDs/Fe 3 O 4 which oxidized the colorless TMB to the blue ox-TMB. The reaction process was monitored by measuring the absorption of the mixture containing Gox, nanozyme, TMB, H 2 O 2 , and different concentrations of glucose.

Selectivity of the Sensor and Real Sample Analysis
Selectivity is an important parameter for evaluating a sensor performance. To explore the selectivity of the constructed colorimetric sensor, several common substances in the serum were selected as possible interferences including uric acid (UA), urea, cysteine (Cys), glycine (Gly), lysine (Lys), ascorbic acid (AA), dopamine (DA), Mg 2+ , Na + , and K + .

Selectivity of the Sensor and Real Sample Analysis
Selectivity is an important parameter for evaluating a sensor performance. To explore the selectivity of the constructed colorimetric sensor, several common substances in the serum were selected as possible interferences including uric acid (UA), urea, cysteine (Cys), glycine (Gly), lysine (Lys), ascorbic acid (AA), dopamine (DA), Mg 2+ , Na + , and K + . As shown in Figure 9c, no significant changes in the absorbance at 652 nm was observed even when the concentration of each of the above substance was five times higher than that of glucose, indicating the constructed sensor has good selectivity for the colorimetric detection glucose.
To investigate the application of the constructed sensor in a real application, the glucose level in a healthy woman was determined. The detected glucose concentration (4.89 mM) was quite similar to that obtained using an automatic biochemical analyzer (Ci-8200, Abbott, Washington, DC, USA), indicating good accuracy.

Reusability of N-CDs/Fe 3 O 4
Because N-CDs/Fe 3 O 4 has high magnetic susceptibility, it can also be recovered by magnetic separation. The performance of the recovered N-CDs/Fe 3 O 4 in glucose detection is shown in Figure 9d. As seen, N-CDs/Fe 3 O 4 still maintained approximately 80% of the original catalytic activity after five cycles of use. Therefore, N-CDs/Fe 3 O 4 can be recycled and has high reusability.

Visual Glucose Detection Based on an Integrated Hydrogel
Visual detection has the advantages of being fast and providing intuitive results, demonstrating great potential in glucose detection. To explore the possibility of using the nanozyme in visual glucose detection, the N-CDs/Fe 3 O 4 nanozyme and natural glucose oxidase were integrated into an agarose gel. Multienzyme-mediated catalysis in a hydrogel can improve the catalytic efficiency because of the limited distribution in space and the relative high concentration of the substrates. More importantly, the detection operation could be greatly simplified since the identification unit and the signal module are all integrated in the hydrogel. When the integrated hydrogel reacted with glucose at different concentrations, photos were taken with a smartphone, and the red (R), green (G), and blue (B) optical primary colors were extracted. As shown in Figure 10, a good linear relationship between (G + B)/2R (y) and Glu concentration (C) was observed when the concentration of glucose ranged from 5 µM to 300 µM (y = 0.0048C + 0.93, R 2 = 0.991). The detection limit was 2.8 µM.
Molecules 2023, 27, x FOR PEER REVIEW 12 of 17 original catalytic activity after five cycles of use. Therefore, N-CDs/Fe3O4 can be recycled and has high reusability.

Visual Glucose Detection Based on an Integrated Hydrogel
Visual detection has the advantages of being fast and providing intuitive results, demonstrating great potential in glucose detection. To explore the possibility of using the nanozyme in visual glucose detection, the N-CDs/Fe3O4 nanozyme and natural glucose oxidase were integrated into an agarose gel. Multienzyme-mediated catalysis in a hydrogel can improve the catalytic efficiency because of the limited distribution in space and the relative high concentration of the substrates. More importantly, the detection operation could be greatly simplified since the identification unit and the signal module are all integrated in the hydrogel. When the integrated hydrogel reacted with glucose at different concentrations, photos were taken with a smartphone, and the red (R), green (G), and blue (B) optical primary colors were extracted. As shown in Figure 10, a good linear relationship between (G + B)/2R (y) and Glu concentration (C) was observed when the concentration of glucose ranged from 5 µM to 300 µM (y = 0.0048C + 0.93, R 2 = 0.991). The detection limit was 2.8 µM.

Characteriaztions and Instrumentations
The fluorescence excitation and emission spectra of N-CDs were measured using a fluorescence spectrometer (LuoroMax-4, Horiba, France). The size and morphology of N-CDs were characterized using transmission electron microscopy (TEM, JEM-2100, Japan Electronics Corporation, Tokyo, Japan). The chemical groups of the nanomaterials were determined by X-ray photoelectron spectroscopy (XPS, PHI5300, PE company, Boston, MA, USA). Fourier transform infrared spectrometry (Vertex 70, Bruker company in Bremen, Germany) was applied to characterize the chemical groups of the synthesized nanomaterials. The crystal structure of N-CDs/Fe 3 O 4 was characterized using a powder diffractometer (Bruker, Germany). The UV-visible absorption spectrum was determined using a UV-2450 spectrophotometer (UV-Vis, Shimadzu Corporation, Tokyo, Japan). The content of Fe in N-CDs/Fe 3 O 4 was measured by an Agilent 730 inductively coupled plasma emission spectrometer (ICP-OES, Agilent Corporation, Palo Alto, CA, USA). The mass composition of the N-CDs/Fe 3 O 4 components was characterized by scanning calorimetry and thermogravimetric synchrotron diffraction (TGA/DSC, Mettler Toledo, Zurich, Switzerland). The N 2 adsorption/desorption isotherm was obtained at 77 K using the ASAP2020 physical adsorption instrument. Before the test, the sample was degassed at 180 • C for 6 h under vacuum conditions. The pore size was obtained by isothermal adsorption branch analysis using the Barret-Joyner-Halenda (BJH) model. Thermogravimetric analysis was performed in a nitrogen atmosphere using GA/DSC1 scanning calorimetry and a thermogravimetric synchrometer (Mettler Toledo, Zurich, Switzerland) at a heating rate of 5 • C/min in the temperature range of 25 • C~800 • C.

Synthesis of the N-CDs/Fe 3 O 4 Nanocomposite
The preparation of N-CDs/Fe 3 O 4 involved two steps. The first step was the synthesis of mesoporous Fe 3 O 4 nanoparticles, and the second step was the in situ synthesis of N-CDs accompanied by loading N-CDs onto mesoporous Fe 3 O 4 nanoparticles. The mesoporous Fe 3 O 4 nanoparticles were prepared according to a method reported in the literature [39]. Briefly, FeCl 3 ·6H 2 O (1 g) was placed in a beaker containing ethylene glycol (20 mL). A uniformly yellow solution was obtained after stirring for 30 min. Then, 3 g of sodium acetate and 1,2-ethylenediamine (10 mL) were added under magnetic stirring, leading to a brownish solution. The obtained solution was then transferred to a polytetrafluoroethylene autoclave and reacted for 8 h at 200 • C. The black solid obtained from the reaction was thoroughly washed with ultrapure water and ethanol. Then, the mesoporous Fe 3 O 4 nanomaterials were obtained through magnetic separation.
To in situ synthesize N-CDs, L-histidine (0.1 g) was mixed with the prepared mesoporous Fe 3 O 4 nanomaterials (10 mg) in a beaker containing 10 mL of ultrapure water. Then, the mixture was transferred to a polytetrafluoroethylene autoclave and reacted at 180 • C for 10 h. After the reaction, a black solid was obtained through magnetic separation. After washing repeatedly with ultrapure water and ethanol, the N-CDs/Fe 3 O 4 nanozyme was obtained.

Peroxidase-Like Activity of N-CDs/Fe 3 O 4
The enzymatic properties of the N-CDs/Fe 3 O 4 magnetic nanozyme were studied using TMB or H 2 O 2 as substrates. The mixture of TMB (0.5 mM), H 2 O 2 (6.6 mM), and N-CDs/Fe 3 O 4 (10 µg/mL) in NaAc-HAc buffer (0.1 M, pH = 4.0) was reacted at room temperature for 10 min. Then, the UV-visible absorption spectrum of the mixed solution was recorded, and the absorbance value at 652 nm was measured as a function of time. The affinity between N-CDs/Fe 3 O 4 and the enzyme substrates was investigated by measuring the steady-state kinetic parameters. Under the optimal conditions, the concentration of TMB was fixed at 0.5 mM, and the concentration of H 2 O 2 was changed from 0.1 mM to 0.7 mM. When TMB was used as the substrate, the concentration of H 2 O 2 was fixed at 0.2 mM, and the concentration of TMB was chosen in the range from 0.1 mM to 0.7 mM. Using the absorbance curve of the system at 652 nm over time, the Michaelis constant (K m ) and the maximum initial reaction rate (V max ) of N-CDs/Fe 3 O 4 were calculated using the double reciprocal curve of the Michaelis equation. where b nanozyme represents the catalytic activity of the nanozyme; V is the total volume of the reaction solution (µL); ε is the molar absorption coefficient (TMB: ε 652nm = 39,000 M −1 cm −1 ); ∆ A/ ∆ T is the initial rate of change of the absorbance at 652 nm; m is the weight of the nanozyme in each measurement (mg).
The recyclability of N-CDs/Fe 3 O 4 was achieved by magnetic separation. The activity of the recycled nanozyme was investigated by measuring the absorbance value of the mixed solution at 652 nm.

Colorimetric Detection of Glucose
The medium for the colorimetric detection of glucose was the NaAc-HAc buffer solution (0.1 M, pH = 4.0) containing TMB (0.5 mM), N-CDs/Fe 3 O 4 (10 µg/mL), and Gox (0.1 mg/mL). Different concentrations of glucose were added to the above detection medium and incubated at 37 • C for 40 min. Subsequently, the UV-visible absorption spectrum of the mixed solution was recorded. To explore the selectivity of the colorimetric detection of glucose, several common substances and inorganic salt ions in serum were selected as the possible interferences, including Na + , Mg 2+ , K + , urea, cysteine, glycine, glutamate, ascorbic acid, and dopamine. The detection performance of glucose in the absence or presence of the interferences was measured. The concentration of the possible interfering substances used was 2 mM. To evaluate the accuracy of the sensing of glucose in a real sample, the serum of a healthy woman (provided by Guangxi Medical University Cancer Hospital) was diluted by a factor of 50 before the measurement.

Visual Glucose Detection Based on an Integrated Hydrogel
Visual glucose detection was investigated based on an agarose hydrogel integrated with the nanozyme (N-CDs/Fe 3 O 4 ), the biological enzyme (Gox), and TMB. Briefly, agarose (45 mg) was dissolved in 1 mL of boiling water. Then, TMB (1 mL, 5 mM), NaAc-HAc buffer (1 mL, 0.1 M, pH = 4), and N-CDs/Fe 3 O 4 (0.3 mL, 0.1 mg/mL) were added. After the solution was mixed, it was cooled to 45 • C, and then Gox (0.3mL, 1 mg/mL) was added before shaking well. Then, the solution was added to the mold (96 well plate, 200 µL per well), cooled, and molded to obtain the hydrogel. The integrated hydrogel was obtained after soaking in a 50% ethylene glycol solution for 60 min. For the visual detection of glucose, the integrated hydrogel was added to the solution of glucose at different concentrations and reacted at 40 • C for 30 min. After the reaction, the hydrogel was taken out, and photos were taken with a smartphone. Then, the data for optical primary colors including red (R), green (G), and blue (B) were extracted.

Conclusions
In this work, a magnetic nanozyme was synthesized using a simple two-step synthesis method by modifying mesoporous Fe 3 O 4 with N-CDs. The N-CDs/Fe 3 O 4 nanozyme contains abundant nitrogen and oxygen functional groups on its surface and possesses good peroxidase-like activity and high magnetic susceptibility. The investigation on the nanozyme parameters verified a high substrate specificity. Based on the excellent peroxidase activity of N-CDs/Fe 3 O 4 and the catalytic oxidation of glucose by glucose oxidase, a colorimetric sensor was constructed to achieve the sensitive detection of glucose. N-CDs/Fe 3 O 4 can be recycled using magnetic separation and exhibited high reusability. In addition, an integrated hydrogel containing the N-CDs/Fe 3 O 4 nanozyme, glucose oxidase, and TMB was prepared and exhibited great potential in the visual detection of glucose. When the synthesized magnetic nanozyme was combined with other enzymes such as cholesterol oxidase, a platform for detecting other metabolites can be constructed. Magnetic nanozymes are expected to demonstrate broad application in the colorimetric detection of metabolites.

Data Availability Statement:
The data presented in this study are available on request from the corresponding author.