Core–shell magnetic Fe3O4/CNC@MOF composites with peroxidase-like activity for colorimetric detection of phenol

Rapid and accurate detection of phenolic wastewater from industries has created global concern. Herein, core–shell magnetic cellulose nanocrystal supported MOF (Fe3O4/CNC@ZIF-8) with robust peroxidase-like activity was synthesized with tannic acid as modifier and bridge. The peroxidase-mimic catalytic activity of as-prepared Fe3O4/CNC@ZIF-8 was further investigated using o-phenylenediamine (OPD) as peroxidase substrates in the presence of H2O2. Moreover, the experimental conditions were optimized and the kinetic analysis results showed that Fe3O4/CNC@ZIF-8 had higher affinity towards both the substrate OPD and H2O2 than horseradish peroxidase (HRP). Finally, a phenol colorimetric assay with a linear range of 2–200 µM and a detection limit of 0.316 µM was constructed. The catalytic mechanism of Fe3O4/CNC@ZIF-8 with phenol was further investigated by fluorescence test and the generated ·OH was proved to act a crucial role to produce quinoid radicals. Additionally, the synthesized magnetic material had excellent stability and recyclability and ease to separation. These results suggest that the Fe3O4/CNC@ZIF-8 may be one of the promising candidates as peroxidase mimic for colorimetric detection of phenol.


Introduction
Numerous toxic pollutants are accumulated in the environment with the intensification of human activities and the acceleration of industrialization, which resulting in acute environmental problems and harm to human health. As one of the environmental contaminants, phenolic compounds have been widely applied in metallurgy, papermaking and chemical industry because they are important industrial raw materials (Wang et al. 2020). Wastewater containing phenol can cause severe harm and is detrimental to human health and the natural growth of aquatic organisms even when the concentration is low (Mei et al. 2015). So far, various traditional analytical methods have been developed as means of phenol detection, such as chromatographic analysis (Arce et al. 2018), gas chromatography-mass spectrometry (GC-MS) (Meng et al. 2011), high performance liquid chromatography (HPLC) (Chen et al. 2013), fluorometry (Mitra and Saha 2016) and electrochemical analysis (Karimi-Maleh et al. 2014), etc. Although these methods have high sensitivity, disadvantages also have been revealed including complicated sample pretreatment procedures, expensive equipment and long-time to analysis. Colorimetric detection is a qualitative and semi-quantitative analysis of the target object based on color change, which does not need expensive equipment and can achieve on-spot detection (Wu et al. 2020). It has a great application prospect in the analysis of environmental pollutants and the process of medical care.
Phenol colorimetric detection is based on the catalytic reaction of the phenol with 4-aminoantipyrine (4-AAP) in the presence of the oxidant agent, forming pink color product through the naked eye observation (Zeng et al. 2015). Enzyme-mimics with nano-micron scale materials have been considered to facilitate the chromogenic reaction due to their high reactivity and multi-size effect. With the exploration of Fe 3 O 4 nanoparticles (NPs) as mimetic peroxidase (Gao et al. 2007), various nanomaterials with enzymelike activity have been developed to possess superior catalytic specificity, activity and stability. Notably, metal organic framework (MOF) possess adjustable pore diameter, high specific surface area and exposed metal sites, showing strong peroxidaselike activity (Wu et al. 2017). For instance, MOF-808 was reported to be a novel peroxidase-like catalyst for glucose colorimetric biosensing at neutral pH (Zheng et al. 2018). The inherent problems of MOF powders arising from difficult separation could be overcome by using magnetic nanoparticles. Encapsulating Fe 3 O 4 NPs in MOF to form a core-shell structure can easily separate the catalyst from the reaction system by an extra magnet with enhanced catalytic activity. However, the urgent problem of most magnetic nanoparticles is that they tend to be highly aggregated because of the large interaction of magnetic dipoles (Ye et al. 2012).
To avoid the above problems, the stable support and surface modification become necessary. Cellulose nanocrystals (CNCs), a type of low-cost and environmental friendly biomass material, have gained great interest as stable supports for nanoparticle immobilization due to their excellent biocompatibility, large aspect ratio and specific surface area (Batmaz et al. 2014). Moreover, the abundant hydroxyl groups on the surface of CNC can offer nucleating sites for the growth of nanoparticles (Low et al. 2018). Guo et al. (2017) have successfully synthesized amino-functionalized Fe 3 O 4 @CNC hybrids to further complex with Cu (II) ions to provide specific protein binding sites. Galland et al. (2013) prepared cellulose nanofibers decorated with magnetic ferrite nanoparticles, which resulted in a rather uniform particle distribution due to the presence of cellulose nanofibers. Thus, constructing a colorimetric detection platform based on magnetic cellulose nanocrystal/MOF composites will be expected to be a novel material of great potential for the efficient detection of phenol.
With this in mind, we designed a core-shell magnetic Fe 3 O 4 /CNC@MOF nanocomposite as enzyme-mimic colorimetric sensor and applied it for phenol detection. Tannic acid (TA) was utilized as adhesive agent to link Fe 3 O 4 /CNC and ZIF-8 NPs. The catechol and galloyl groups of TA are well known to bind with metal ions with high affinity to form metalphenolic networks (Hao et al. 2020). Using magnetic Fe 3 O 4 /CNC as the supporter for MOF-based enzyme mimics colorimetric sensor is conducive to fast separation and alleviate aggregation of MOF nanoparticles. As designed, the Fe 3 O 4 /CNC@ZIF-8 nanocomposites exhibited intrinsic peroxidase-like activity and showed sharp color change upon adding to phenol solution. This work demonstrated convenient and effective phenol detection method upon cellulose-based enzyme mimic composites.
Preparation of tannic acid modified magnetic cellulose composite Magnetic cellulose composite was prepared through the one-pot solvothermal method as follows. First, 2.0 g CNCs were dispersed in 20 mL ethylene glycol by ultrasonication with vigorous stirring for 30 min at room temperature (RT). Then, 0.8 g FeCl 3 Á6H 2 O and 2.0 g NH 4 Ac were added to the CNC dispersion in sequence. Next, another 20 mL ethylene glycol containing 0.75 g PEG was added to the mixed solution and kept stirring for another 30 min to form a homogeneous yellow dispersion. After that, the mixture was transferred in a Teflon-lined autoclave and maintained at 200°C for 11 h. After natural cooling, the product were collected with the help of a magnet and washed with deionized (DI) water and ethanol several times. Finally, the catalysts were dried in a vacuum at 60°C for 4 h. For comparison, the Fe 3 O 4 NPs were prepared by following the same steps as showed above in the absence of CNCs.
For the tannic acid functionalized Fe 3 O 4 /CNC, 100 mg Fe 3 O 4 /CNC was suspended in 20 mL DI water including 60 mg TA under gentle stirring at RT for 1 h. Then, the composites were magnetically isolated by an external magnet to obtain TA modified Fe 3 O 4 /CNC (Fe 3 O 4 /CNC/TA). The product was washed with DI water and followed by ethanol five times, respectively. Finally, the resulting mixture was placed in a 60°C oven under vacuum for 4 h to evaporate ethanol.

Preparation of the Fe 3 O 4 /CNC@ZIF-8 nanocomposites
Utilizing the chelation of the TA and metal ions, Zeolite imidazole framework-8 (ZIF-8) was synthesized on the surface of Fe 3 O 4 /CNC/TA. Briefly, 0.148 g of 2-methylimidazole was dissolved in 10 mL of methanol to form a clear solution. Then 50 mg as-prepared Fe 3 O 4 /CNC/TA was added to the above solution and stirred under the ultrasonication for 45 min at RT. Afterwards, 0.410 g of zinc nitrate hexahydrate dissolving 10 mL methanol was added to the mixture at 50°C and stirred for another 2 h. The obtained products named Fe 3 O 4 /CNC@ZIF-8 were separated by a magnet and washed with water and methanol several times, subsequently dried under vacuum at 60°C for 4 h.

Characterization
Transmission electron microscopy (TEM) on a FEI Tecnal G 2 F30 and scanning electron microscopy (SEM, Hitachi S-4800, Hitachi, Japan) with energy dispersive X-ray spectroscopy (EDS) were used to characterize the microstructure and morphology of the sample. Samples crystal structure were obtained by The powder X-ray diffraction analysis (XRD Bruker, D8-Advance, Karlsruhe, Germany) with Cu Ka radiation (k = 1.542 Å ) at 10-80°(2h). The chemical structure of composites was characterized by Fourier transform infrared (FT-IR, Bruker VERTEX 70, Karlsruhe, Germany) spectra with a wavelength range of 4000-400 cm -1 . Temperature-mass curve of the materials were tested by thermal gravimetric analysis (TG, STA449F3-1053-M, Germany), which was carried out in nitrogen atmosphere with a heating rate of 10°C min -1 from 30 to 800°C. The specific surface area was calculated by the Bruner-Emmett-Teller (BET) method. The magnetic properties of the obtained magnetic composite materials were evaluated using a vibrating sample magnetometer VSM 7304 (Lakeshore, Columbus, OH, USA) at room temperature. The chemical state of the elements in the sample was characterized by X-ray photoelectron spectroscopy (XPS, AXIS SUPRA, UK). Fluorescence spectra were recorded on an RF-5301PC fluorescence spectrophotometer (Shimadzu Co., Ltd, Japan) equipped with a Xenon lamp to detect free radicals in the catalytic reactions. Colorimetric experiments were performed on The UV-Vis spectra (China, Shanghai, Shimadzu UV-2501 PC spectrometer).

Evaluation of the peroxidase-mimic catalytic activity of Fe 3 O 4 /CNC@ZIF-8
The peroxidase-like activity of the Fe 3 O 4 /CNC@ZIF-8 was investigated in the sensing of H 2 O 2 by the oxidation reaction of a colorless substrate OPD to form the yellow product (2, 3-diaminophenazine, DAP) (Vetr et al. 2018). In a typical experiment, 4 mg Fe 3 O 4 /CNC@ZIF-8, 500 lL 0.5 mM OPD and different concentration of H 2 O 2 (0-200 lM) was added into 0.2 M NaAc-HAc buffer (pH 4.0) so as to make a total volume of 3.0 mL. Subsequently, the mixed solution incubated at RT for 20 min. UV-Vis measurement was used to determine the obtained yellow-colored solution by monitoring the absorbance at the wavelength of 450 nm. The peroxidase-like activity of Fe 3 O 4 /CNC@ZIF-8 at different temperatures and pH values were also studied.

Kinetic analysis of Fe 3 O 4 /CNC@ZIF-8 peroxidase-like activity
Kinetic analysis was performed using steady state kinetics following previous literature, which was conducted by changing the concentration of one substrate (OPD or H 2 O 2 ) at fixing the other substrate . All experiments were carried out in 1.0 mL cuvettes with a path length (l) of 1.0 cm at RT. The Beer Lambert's law can change the absorbance value to their corresponding concentration: where A is the absorbance at 450 nm, e is the molar absorptivity (e = 1.67 9 10 4 M -1 cm -1 for DAP at 450 nm) (Vetr et al. 2018), c represents the concentration of one substrate. For the kinetic determination of OPD, 8 mg Fe 3 O 4 / CNC@ZIF-8 was added to a NaAc-HAc buffer (0.2 M, pH 4.0) containing different concentrations of OPD (0.125-1.25 mM) in the presence of 70 lM H 2 O 2 with a total volume of 3.0 mL. The absorbance at 450 nm was immediately recorded at a 60 s interval within 10 min. As a comparison, the kinetic determination of H 2 O 2 was carried out by fixing the concentration of OPD (0.5 mM) and varying the H 2 O 2 concentration (10-80 lM). Then the ''absorbance vs time'' plots were obtained to calculate the initial point (Slope initial ) of each curve. The initial reaction velocity (m) was calculated with the formula: Slope initial / (e 9 l). All of the m against substrate concentration ([S]) plots were then fitted via nonlinear regression of the Michaelis-Menten equation. Finally, the Michaelis constant (K m ) was obtained from the Lineweaver-Burk double reciprocal plot generated from the Michaelis-Menten equation, where V max is the maximal reaction velocity: Colorimetric detection of phenol For phenol detection, a chromogenic reaction based on the oxidative coupling between the phenol and 4-AAP in the presence of the H 2 O 2 was conducted at RT. Briefly, 800 lL 4-AAP (80 mM), 900 lL H 2 O 2 (110 mM) and different concentration of phenol (0-200 lM) were added into NaAc-HAc buffer (0.2 M, pH 4.0). Then, 4 mg Fe 3 O 4 /CNC@ZIF-8 was rapidly added, and the mixture incubated for 20 min at RT until a pink mixture was obtained. The mixture was separated through external magnet to remove the precipitate, and the absorbance of the supernatant was monitored by UV-Vis spectrophotometer at the wavelength of 525 nm. All the measurements were carried out in triplicate and the results from the obtained parallel data were averaged.
In addition, the selectivity of Fe 3 O 4 /CNC@ZIF-8 toward phenol was investigated. Aqueous solutions of different interfering substances in DI water, such as acetone, ethanol, ethyl ether, imidazole, benzene, cyclohexane and sodium sulfite, were prepared and similarly tested under the same experimental conditions.

Results and discussion
Synthesis and characterization of Fe 3 O 4 / CNC@ZIF-8 Core-shell magnetic Fe 3 O 4 /CNC@ZIF-8 was successfully prepared through one-step solvothermal and insitu growth method with tannic acid as modifier and bridge, as shown in Scheme 1. SEM and TEM images were used to characterize to study the surface morphology and particle size of the Fe 3 O 4 /CNC and Fe 3 O 4 /CNC@ZIF-8 (Fig. 1). The rod-like structure of CNCs was measured with a length of 200-250 nm and a width of 15-20 nm, which was a good support to prevent the aggregation of Fe 3 O 4 NPs . SEM image of Fe 3 O 4 /CNC showed that Fe 3 O 4 NPs loading on CNCs had a nearly spherical morphology with a rough surface, and particle size ranged between 100 and 250 nm (Fig. 1a). TEM micrograph (Fig. 1b) evidenced that CNCs can interweave each other to form a web-like network and Fe 3 O 4 NPs were well dispersed in the CNC network after solvothermal process. In Fig. 1c, thin tannic acid shell layers had formed around the Fe 3 O 4 /CNC cores, which were utilized as the adhesive coating on the surface of Fe 3 O 4 /CNC to guarantee a homogeneous growth of ZIF-8. As seen from Fig. 1d,  The FT-IR spectra are illustrated in Fig. 2a to contrast the functional groups on Fe 3 O 4 , ZIF-8 and Fe 3 O 4 /CNC@ZIF-8 surface. Concretely, for pure CNCs, the peaks at 1060 and 2904 cm -1 were caused by the C-O-C stretching of pyranose and the C-H stretching vibration, respectively Xiong et al. 2013). The broad peak in the band of 3464 cm -1 corresponded to the O-H stretching vibration, while the absorbance band at 1609 cm -1 originated from O-H bending vibration (Jahan et al. 2010). In the case of Fe 3 O 4 /CNC, the peak derived from Fe-O bonds was expected at 582 cm -1 , indicating that the Fe 3 O 4 NPs were successfully immobilized on the CNCs. O-H stretching of CNCs became weak due to the interactions of the hydroxyl groups and Fe 3 O 4 NPs (Liu et al. 2015). The spectrum of Fe 3 O 4 / CNC after TA coating displayed some changes including the -OH stretching of the phenolic and methylol group that appeared at 3420 cm -1 . Additionally, the 1718 and 1080 cm -1 vibrational bands corresponded to the C = O and C-O stretching vibrations, respectively. The peaks at 1434 and 1346 cm -1 in the spectrum of TA belonged to the aromatic C-C and phenolic C-O stretching vibrations, respectively (Dutta and Dolui 2011). Moreover, onelayer coating TA-CNC was also obtained due to the hydrogen bond between TA and the C6 hydroxyl group of the CNCs. For the Fe 3 O 4 /CNC@ZIF-8, the significant observed bands around 3136 and 2928 cm -1 were ascribed to the stretching vibration of C-H in methyl and imidazole rings. It is Worthwhile to mention that the series of complex and compact observed bands in the spectra of 700-1350 cm -1 and 1350-1500 cm -1 can be attributed to the stretching and plane bending of imidazole ring (Zheng et al. 2014). The absorption peaks at 421 and 759 cm -1 were ascribed to the nature of the ZIF-8 structure confirming the formation of the Zn-N bond.
The XRD patterns for the different samples are shown in Fig. 2b. The broad and strong peaks at 14.84°, 16.68°and 22.9°were assigned to the (1-10), (110) and (200) Fig. 3a), and the isotherm of nanocomposites displayed a typical type IV isotherm. The pore width was 5.1 nm and the BJH pore volume was 0.31 cm 3 /g, indicating a typical mesoporous structure. The BET surface area of Fe 3 O 4 / CNC@ZIF-8 was 463.63 m 2 /g, which is larger than many detection materials (Table S1). These analyses suggested that Fe 3 O 4 /CNC@ZIF-8 possessed a high surface area and mesoporous structure, which would probably provide more active sites and thereby improve the catalytic activity. Additionally, the hysteresis curve of sample magnetization were tested by VSM instrument and the result are shown in Fig. 3b. The saturation magnetization (M s ) value of Fe 3 O 4 was about 79.11 emu/g. In contrast with Fe 3 O 4 , the relatively weak magnetism Fe 3 O 4 /CNC (59.65 emu/ g) and Fe 3 O 4 /CNC@ZIF-8 (44.01 emu/g) were mainly due to the TA coating and dielectric property of the outer shell. Nevertheless, the Fe 3 O 4 /CNC@ZIF-8 still held strong magnetism and responded rapidly to external magnetic field, which was conveniently separated in the post-treatment process of wastewater.
XPS analysis was further performed to investigate the element composition and the chemical states of the Fe 3 O 4 /CNC@ZIF-8. The XPS spectrum showed the surface composition of Fe 3 O 4 /CNC@ZIF-8  Fig. 4a. The binding energy of 285.5 eV was related to C-C in cellulose, which indicated the successful synthesis of Fe 3 O 4 /CNC (Hu et al. 2017). Additionally, the peaks at 398.8 eV in Fig. 4a represented N 1 s, supporting the presence of the C=N-C in the ZIF-8 (Yang et al. 2015). To amply reveal the microscopic interactions, the C 1 s XPS spectrum are shown in Fig. 4(b) and three peaks at 284.04, 284.65 and 285.2 eV were associated with C-C, C=N and C-N, respectively (Chen et al. 2021). For O 1 s in Fig. 4(c), the peak appeared at 531.91, 532.53 and 533.72 eV were in line with -OH, N-O and -COOH. One small Zn 2p peak observed at 1021.45 eV was assigned to zinc (? 2) oxide, and the spectra of Zn 2p were in good agreement with its oxidation state. On the basis of previous work, the two characteristic peaks locating at 1021.8 and 1044.8 eV in the high resolution spectrum of Zn were due to Zn 2p 3/2 and Zn 2p 1/2 , respectively, further indicating the presence of ZIF-8 ( Fig. S2) (Yang et al. 2015). Notably, the Fe 2p spectrum showed the main peaks at around 710.7 and 724.2 eV mainly corresponded to the Fe2p 3/2 and Fe2p 1/2 , respectively, obviously demonstrating the existence of Fe 3 O 4 (Fig. 4d)  NPs. Besides, ZIF-8 possesses a porous structure with a large number of exposed active sites, which coated on the surface of Fe 3 O 4 NPs to further improve the peroxidase-like activity of the composite. It is known that the peroxidase-like catalytic activity of Fe 3 O 4 /CNC@ZIF-8 nanocomposites depends on the pH values, temperature and catalyst dosage. Therefore, the catalytic activity of Fe 3 O 4 / CNC@ZIF-8 was investigated in detail through modulating these factors. In Fig. S3(a), with the catalyst dosage increased from 2 to 12 mg, the catalytic activity first rose and then gradually fell, exhibiting the optimum catalyst dosage at 8 mg. Moreover, the pH and temperature were important factors affecting the performance of peroxidase catalysts, and hence various pH from 2.0 to 6.0 and temperature from 20 to 60°C were studied (Fig. S3b, c). The catalytic activity achieved a higher level in acidic solutions (pH 3.5-4.5) than in strong acidic or alkaline solutions. The maximum catalytic activity was obtained in a solution at pH 4.0 similar to HRP (Qiao et al. 2014). In addition, Fe 3 O 4 /CNC@ZIF-8 exhibited quite commendable catalytic activity even the temperature at 55°C , indicating the catalytic possessed a good temperature resistance.
Under the optimal conditions, the Fe 3 O 4 / CNC@ZIF-8 was used for the sensitive assay of H 2 O 2 by colorimetry. The UV-Vis spectra of the different concentration of H 2 O 2 are displayed in Fig. 5b, which observed the absorbance increased gradually with the increasing of H 2 O 2 concentration at 450 nm. Figure 5c showed a good linear relationship (R 2 = 0.9954) between absorbance value and H 2 O 2 concentration in the range from 2 to 200 lM with a limit of detection (LOD) of 2.24 9 10 -7 M (S/N = 3). Correspondingly, a clear color change from colorless to dark yellow was obviously differentiated by the naked eyes shown in Fig. 5d. These results showed that the Fe 3 O 4 /CNC@ZIF-8 has great potential in the quantitative analysis of H 2 O 2 detection.

Detection of phenol and optimization of reaction conditions
As expected, a colorimetric detection method of phenol had been established, which was proven a high-throughput analytical method with fast and visual readout advantages. When phenol is present, it can be oxidized by the Fe 3 O 4 /CNC@ZIF-8 forming the pink-colored complexes in the presence of H 2 O 2 , exhibiting a characteristic absorption peak at 525 nm (Wu et al. 2020). As shown in Fig. 6a, the reaction system including Fe 3 O 4 /CNC@ZIF-8 and H 2 O 2 (curve e) appeared obvious pink compared to that of the control experiment (curve a and b). Furthermore, in comparison with experimental system consisted of either Fe 3 O 4 /CNC@ZIF-8 or H 2 O 2 (curve d and e), both of them existed in the system could enhance the color reaction and generate a strong absorption peak at 525 nm. These results indicated that Fe 3 O 4 / CNC@ZIF-8 possessed a higher catalytic activity in the presence of H 2 O 2 than that of Fe 3 O 4 /CNC and pure ZIF-8.
Various experimental parameters were examined and optimized as follows to find the suitable conditions for phenol detection. Since the reaction used 4-AAP as a chromogenic agent in the presence of an oxidant agent to produce a colored compound, changes in 4-AAP and H 2 O 2 concentrations directly affected the efficiency of reaction. The influence of the 4-AAP concentration was studied in the range of  (Fig. S4a), and the maximum absorbance was obtained with the 4-AAP concentration in 8.0 mM. As shown in Fig. S4b, an improvement in absorbance intensity was observed with the H 2 O 2 concentrations increasing from 0.01 to 0.1 M and decreased slowly after that. Therefore, H 2 O 2 concentration of 0.1 M was chosen for subsequent experiments. For temperature factor, the maximum catalytic activity for Fe 3 O 4 /CNC@ZIF-8 revealed that 50°C was the optimal incubation temperature (Fig. S4c). Figure 6b demonstrated that when the phenol concentration was 0-200 lM, as the phenol concentration increased, the absorbance value gradually increased. A linear response was obtained between the absorbance and the phenol concentration in the range of 2-200 lM with the coefficient of correlation (R 2 ) equal to 0.9908 (Fig. 6c). The LOD of the designed phenol analysis platform was 3.16 9 10 -7 M (S/N = 3). Comparing the detection ranges and LODs of different materials for detecting phenol, it indicated that the Fe 3 O 4 /CNC@ZIF-8 material has certain advantages in phenol detection (Table S2).

Steady-state kinetic assay of Fe 3 O 4 /CNC@ZIF-8
Based on these optimal conditions, the Michaelis-Menten behavior of the Fe 3 O 4 /CNC@ZIF-8 was evaluated by steady-state kinetics analysis with H 2 O 2 and OPD as substrates, respectively. The oxidation reaction process of Fe 3 O 4 /CNC@ZIF-8 followed the conventional enzymatic dynamic regulation of the Michaelis-Menten equation. In Fig. 7, the Lineweaver-Burk plots were used to calculate the  Table 1, the K m value of Fe 3 O 4 /CNC@ZIF-8 with OPD as substrate was 0.883 mM, two times lower than that of HRP (1.80 mM). It was confirmed that Fe 3 O 4 @ZIF-8 and Fe 3 O 4 /CNC@ZIF-8 had a better affinity for OPD than HRP due to the large surface area of catalysts and good dispersion of Fe 3 O 4 NPs Moreover, Fe 3 O 4 / CNC@ZIF-8 exhibited a higher affinity than that without CNCs, indicating the significant role of biomass carrier. The K m value of Fe 3 O 4 /CNC@ZIF-8 with H 2 O 2 was 0.171 mM, which was smaller than that of HRP (0.214 mM). It was indicated lower H 2 O 2 concentration for Fe 3 O 4 /CNC@ZIF-8 could achieve maximum activity than that of HRP. Furthermore, the Fe 3 O 4 /CNC@ZIF-8 had a much smaller K m and higher V max than those of HRP (Kergaravat et al. 2012;Qiao et al. 2014), demonstrating that Fe 3 O 4 / CNC@ZIF-8 possessed higher activity and catalytic efficiency than that for HRP.
To gain a better understanding of the mechanism towards the reaction of phenol and 4-AAP to form a pink quinone imine, p-phthalic acid (PTA) was chosen as a fluorescence probe. Terephthalic acid can react with ÁOH to produce a fluorescent product 2-hydroxy terephthalic acid (HTA), which can be observed through fluorescence spectroscopy by displaying a fluorescent emission peak at 410 nm (Barreto 1994). In detail, 4 mg Fe 3 O 4 /CNC@ZIF-8 was firstly added to the mixture containing 100 lL 0.1 M H 2 O 2 , 2.89 mL NaAc-HAc buffer (0.2 M, pH 4.0) for reaction. Then, 10 lL 5 mM terephthalic acid solution was added to the above solution. Subsequently, the mixed solution incubated for 20 min at room temperature, and the fluorescence spectrum was recorded with an excitation wavelength at 325 nm. As illustrated in Fig. 8a The changes of absorbance at 505 nm were monitored by UV-Vis spectrophotometer, and the experimental results were exhibited in Fig. 9a, almost no apparent absorbance at 525 nm along with slight color change were observed in the reaction solutions with interfering substances. However, the reaction solution revealed a remarkable color variation from colorless to pink accompanied with a strong absorbance at 525 nm. In addition, the reusability of the synthesized Fe 3 O 4 /CNC@ZIF-8 was investigated by cyclic experiments. UV-Vis spectrophotometer was used to record the absorbance of the reaction systems at 525 nm. After each cycle, the composite was separated from the reaction system with extra magnet, and washed by DI water and ethanol five times and dried, which was reused for the next cycle. In Fig. 9b, the absorbance values of the reaction system had negligible change after sextic tests. Therefore, these results further confirmed the excellent selectivity and reusability of Fe 3 O 4 /CNC@ZIF-8-based phenol assay platform, offering a feasible and promising strategy for detecting phenol simple, rapidly and sensitively in actual samples.

Conclusion
In summary, Fe 3 O 4 /CNC@ZIF-8 nanocomposites as a novel peroxidase mimic were successfully synthesized by simple one-step solvothermal and in-situ growth method. The as-prepared Fe 3 O 4 /CNC@ZIF-8 was demonstrated to possess outstanding intrinsic peroxidase-like activity compared to the Fe 3 O 4 /CNC and ZIF-8. Furthermore, the steady-state kinetic parameters (K m and V max ) also theoretically proved Fe 3 O 4 / CNC@ZIF-8 nanocomposites had stronger affinity for OPD and a fast catalytic rate. Characterization analyses found that the Fe 3 O 4 /CNC@ZIF-8 exhibited superior catalytic activity due to the unique core-shell structure and well-dispersed nanoparticles on the CNCs. On the basis of the intrinsic peroxidase-like activities, a simple colorimetric sensing system of phenol was constructed and 0.316 lM of the LOD was obtained in the linear range of 2-200 lM. The Fe 3 O 4 / CNC@ZIF-8 was magnetically separable along with the recycled up to 6 times and the detection mechanism was probed by fluorescence spectroscopy. Consequently, owing to rapid response, low cost, as well as robust catalytic properties, Fe 3 O 4 /CNC@ZIF-8 nanocomposites are competent to an excellent peroxidase mimics catalysts for colorimetric detection of phenol in environmental monitoring.