Co3O4 Nanoparticles Uniformly Dispersed in Rational Porous Carbon Nano-Boxes for Significantly Enhanced Electrocatalytic Detection of H2O2 Released from Living Cells

A facile and ingenious method to chemical etching-coordinating a metal-organic framework (MOF) followed by an annealing treatment was proposed to prepare Co3O4 nanoparticles uniformly dispersed in rational porous carbon nano-boxes (Co3O4@CNBs), which was further used to detect H2O2 released from living cells. The Co3O4@CNBs H2O2 sensor delivers much higher sensitivity than non-etching/coordinating Co3O4, offering a limit of detection of 2.32 nM. The wide working range covers 10 nM-359 μM H2O2, while possessing good selectivity and excellent reproducibility. Moreover, this biosensor was used to successfully real-time detect H2O2 released from living cells, including both healthy and tumor cells. The excellent performance holds great promise for Co3O4@CNBs’s applications in electrochemical biomimetic sensing, particularly real-time monitor H2O2 released from living cells.


Introduction
H 2 O 2 is a reactive oxygen species (ROS) frequently used as a marker for oxidative stress analysis. It is a by-product of reactions catalyzed by most oxidase enzymes [1], and is also involved in numerous physiological processes including cell differentiation and mediating immune responses [2,3]. Excess H 2 O 2 will attack methionine residues and cysteine, which will cause cell damage and cytotoxicity. Owing to its peculiar capability, the concentration of H 2 O 2 can be used as an indicator of several diseases diagnoses, such as Parkinson's disease [4,5], cancer [6,7], diabetes [8] and acute myocardial infarction [9]. Thus, the determination of H 2 O 2 is of great significance in biomedical, industrial, and academic applications. The H 2 O 2 levels in the intracellular physiology range are from 0.001 to 0.7 µM [10]. Therefore, sensors with high sensitivity, specificity and broad working range are needed to probe the intracellular H 2 O 2 . The excellent detection of H 2 O 2 mainly depends on the detection method and material two aspects. Among the technique for accurate and reliable detection of cellular H 2 O 2 , such as colorimetry [11,12], fluorescence [13,14], chromatography [15] and chemiluminescence [16], electrochemical techniques increasingly attracted attention due to their high sensitivity, good selectivity, low cost, as well as rapid response. For electrochemical detection, natural enzymes are usually the choice of sensing materials due to their remarkable specificity and high sensitivity in catalyzing the decomposition of H 2 O 2 . However, the inherent defect of natural enzymes, such as instability and ease to reduce or even deactivate the activity, limited their further applications [17]. Thus, non-enzymatic electrochemical sensors were proposed to overcome the limitations of natural enzyme sensing platforms [18]. Various nanomaterials have been used in H 2 O 2 sensors, including transition metals oxides (e.g., Fe 3 O 4 , Co 3 O 4 , NiO, CuO) [19][20][21][22]. Transition metals have multiple oxidation states. They can absorb other substances onto their surface, meanwhile activating them in the process. These good abilities make them an excellent choice in synthesizing nanoenzymes [1]. Among these materials, Co 3 O 4, a kind of intrinsic p-type transition metal oxide, was reported in electrochemically detecting H 2 O 2 because of its high electrochemical stability, fair price, and environmentally friendly [23]. However, the close-packed structures and poor electronic conductivity of Co 3 O 4 could reduce their specific surface area and deteriorate its performance in H 2 O 2 detection.
Metal-organic framework (MOF) possesses the periodic network structures made by the self-assembly of organic linkers and inorganic metal-containing nodes [24]. Recently, the unique merits of crystalline porous structure, highly dispersed metal components, and adjustable pore size of MOFs grant them outstanding performances in various applications [25]. In addition, MOF-derived carbon materials overcome the aggregation of metal nanoparticles that is induced by a further pyrolysis process [26]. Hence, metalorganic framework (MOF)-derived Co 3 O 4 are promising in synthesis Co 3 O 4 with uniform morphology and good electronic conductivity.
Tannic acid (TA) is a plant polyphenol. The chemical structure of TA is usually a decagalloyl glucose (C 76 H 52 O 46 ) [27]. It widely exists in plant tissues such as tea, wood, and wine [28]. Its adhesive and reduction capability have been demonstrated in materials synthesis for lithium-ion batteries [29,30], dye remove [31], oil/water separation [32], catalytic [33], cell proliferation [34] and drug delivery [35]. As a kind of phenolic acid, TA is a weak organic acid and can release protons [36], which is applied in etching MOF materials to synthesize hollow structured materials [37].
In this study, to achieve sensitive and specific H 2 O 2 detection, we rationally designed an ingenious method to synthesize ZIF-67 MOF-derived Co 3 O 4 nanoparticles (NPs) dispersing in porous carbon nano-box (Co 3 O 4 @CNBs) as a H 2 O 2 nanozyme. The function of TA is to etch ZIF-67 while preserving the overall cubic architectures during thermal annealing process. The Co 3 O 4 nanoparticles uniformly dispersed in porous carbon nanoboxes (Co 3 O 4 @CNBs) was synthesized by delicately tuning TA concentration and thermal annealing temperature. The sensing performance of Co 3 O 4 @CNBs in H 2 O 2 sensing was characterized. The dispersion of Co 3 O 4 NPs in the porous carbon nano-boxes (CNBs) was further investigated for its enhancement mechanism toward the specific reduction of H 2 O 2 . Moreover, the application of the Co 3 O 4 @CNBs H 2 O 2 sensor was demonstrated in detecting H 2 O 2 released from living cells.

Characterizations
The morphologies of the synthesized materials were observed by field emitted scanning electron microscopy (FESEM, JSM-7800 F, JEOL, Tokyo, Japan) and transmission electron microscopy (TEM, JEM-2100, JEOL, Japan). HAADF-STEM characterization was conducted with TEM (JEM-2100, JEOL, Japan). The surface properties of the materials were characterized by X-ray photoelectron spectroscopy (XPS, ESCALAB 250Xi, Thermo Scientific, Waltham, MA, USA). The crystal structure was characterized by X-ray diffraction (XRD, MAXima-X XRD-7000, Shimadzu, Tokyo, Japan). The chemical groups of the samples were recorded by Fourier transform infrared spectroscopy (FTIR, Thermo-Nicolet 6700, Thermo Scientific, MA, USA) with air as a reference. Thermogravimetric analysis (TGA, TA Instruments Q50, TA Instruments, New Castle, DE, USA) was performed using a thermal analyzer under airflow (10 • C min −1 ). JW-BK300C (JWGB SCI. & TECH., Beijing, China) determined N 2 adsorption-desorption isotherms and pore-size distributions. All electrochemical measurements were performed at room temperature on a CHI 760D (Chenhua Instruments, Shanghai, China). PBS (0.01 M, pH = 7.4) was used as the electrolyte for all electrochemical measures except in detection with cell viability.

Preparation Co 3 O 4 @CNBs from ZIF-67
Synthesis of Co 3 O 4 @CNBs involves the following three-steps: Synthesis of ZIF-67 nanocubes (ZIF-67 NCs): ZIF-67 NCs were synthesized according to the previous works [38]. 580 mg of Co(NO 3 ) 2 •6H 2 O and 4 mg of etyltrimethylammonium bromide (CTAB) were dissolved in 20 mL of deionized water and marked as solution A. 9.08 g of 2-methyimidazole (2-MIM) was dissolved in 140 mL of deionized water, and marked as a solution B. Then the 20 mL solution A was rapidly injected into 140 mL solution B and stirred at room temperature for 20 min. The mixture was centrifuged at 10,000 rpm for 10 min. The collected precipitate (ZIF-67 NCs) was washed with ethanol several times and then dried in an oven at dried at 60 • C for 24 h.
Synthesis of TA-Co nano boxes (TA-Co NBs): The as-prepared ZIF-67 NCs were first dispersed into 10 mL of ethanol, then poured into 150 mL of ethanol and deionized water mixture solution (Volume ratio of H 2 O and ethanol = 1:1) containing different concentration of TA solution (0 mg/mL, 0.5 mg/mL, 1 mg/mL, 2 mg/mL) and stirred at room temperature for 5 min. The precipitate collected by centrifugation was washed with ethanol and then dried in an oven at dried at 60 • C for 24 h. The TA etched ZIF-67 was recorded as TA-Co nano boxes (TA-Co NBs).
Synthesis of Co 3 O 4 @CNBs: The as-prepared TA-Co NBs powder was first annealed at 200 • C for 30 min and then further annealed at different temperatures (500 • C, 600 • C, 700 • C, 800 • C) for 1 h with a heating rate of 1 • C min −1 under N 2 flow, and cooled down to room temperature naturally. After that, the powder was annealed at 200 • C for 6 h in air with a heating rate of 10 • C min −1 . The obtained materials named as Co 3 O 4 @CNBs. In comparison, pristine ZIF-67 without TA etching was thermally annealed with the same condition and recorded as Co 3 O 4 @carbon (Co 3 O 4 @C).

Preparation of Co 3 O 4 @CNBs Modified Electrode
All electrochemical measurements were performed on a CHI 760D electrochemical workstation (Chenhua Instruments, China). A conventional three-electrode cell was used with a modified glass carbon electrode as the working electrode, Ag/AgCl (in saturated KCl solution) as the reference electrode, and platinum wire as the counter electrode. Glassy carbon electrodes (GCE) were polished with 0.3 and 0.05 µm alumina slurry on a polishing cloth and cleaned sequentially through water and ethanol under sonication for 3 min and dried in nitrogen flow for further use. Next, 7 µL 2 mg/mL Co 3 O 4 @CNBs aqueous dispersion was dropped on it and dried for 3 h at room temperature. After that, 5 µL 0.05% Nafion were dropped on it successively and dried at room temperature. Nafion film acts as a protective layer, preventing the falling of the loaded Co 3 O 4 @CNBs from the electrode. The supporting electrolyte of PBS (0.01 M, pH = 7.4) was deoxygenated using nitrogen before use and kept inside a nitrogen atmosphere. The prepared working electrodes were activated by cyclic voltammetric (CV) scanning for 20 cycles in the potential range from −1.0 to 1.0 V at a scan rate of 50 mV·s −1 . Amperometric current-time curves (i-t) were collected at −0.22 V in 0.01 M 10 mL PBS by successive injecting H 2 O 2 at 50 s intervals.

Detection of H 2 O 2 Released from Living Cells
In this work, three types of living human cells, A549, 4T1 and HUVEC cell were cultured in DMEM containing 10% FBS, 1 × antibiotic antimycotic. All the cells were supplemented with 10% FBS in a humidified incubator (with 5% CO 2 atmosphere) at 37 • C and grown in polystyrene-coated T25 (25 cm 2 ) cell culture flasks. Cells were washed three times with 0.01 M PBS (pH 7.4), detached by 1% Trypsin, collected by centrifugation, and the number was calculated using a cell counter. The response of H 2 O 2 released from approximately 1.0 × 10 5 cells was measured by Co 3 O 4 @CNBs modified GCE at −0.22 V in 2 mL DMEM medium.

Co 3 O 4 NPs Dispersed in Porous Carbon Nano Boxes by TA Assisted Etchings
Synthesis of Co 3 O 4 @CNBs involves the following three-step reaction (Scheme 1). First, ZIF-67 was synthesized by using the co-precipitation method [38]. Next, TA was used to etch the ZIF-67 to form the unique Co 3 O 4 NBs. Last, the Co 3 O 4 NBs were thermal annealed to carbonize the TA and subsequent low-temperature oxidation in the air to form Co 3 O 4 @CNBs. The FESEM characterization found that the co-precipitation method synthesized ZIF-67 is uniform regular cubic with a smooth surface (Supplementary Information Figure S1). The size of the cubic is about 760 nm. In our approach, TA functions as green and facile etching agent to etch ZIF-67 directly without extra procedures and chemicals. We found that TA (0.5 mg/mL, 1.0 mg/mL, 2.0 mg/mL) treatment did not change the overall size and the surface morphology of the ZIF-67 cubic. As compared in in FESEM characterization ( Figure 1A), TA treated ZIF-67 cubic has a size of 760 nm with a smooth surface. However, TEM characterization ( Figure 1B) reveals that the cubic's inner structure changes significantly after treated by TA with different concentrations. Though TA etched the inside of ZIF-67 cubic, the wall thickness of the cubic had no significant difference as TA concentration changed. The main effect of TA concentration influences the degree of etching reaction inside the cubic. As shown in Figure 1B, with the TA concentration increase from 0 to 2 mg/mL TA, the inside of ZIF-67 was solid at first and then showed the yolk-shelled heterogeneous structure. Finally, the ZIF-67 cubic was completely etched to form hollow interior TA-Co NBs ( Figure 1B). Incubating ZIF-67 cubic in 2 mg/mL TA solution for 5 min resulted in a ZIF-67 NB with a wall thickness of about 80 nm.
The TA etching reaction is illustrated in Supplementary Information Figure S2. First, the protons released from TA etch the ZIF-67, releasing the Co 2+ and 2-MIM simultaneously. At the same time, Co 2+ and TA coordinate together quickly to form the TA-Co shell. The attached TA block the exposed surface of ZIF-67, thus protecting the outer parts of MOFs from further etching, resulting in internal etching of the ZIF-67 to form TA-Co NBs [29,37,39].
Next, TA-Co NBs were carbonized in an N 2 atmosphere to synthesize the Co 3 O 4 @CNBs. First, the thermal carbonization and subsequent low-temperature oxidation in air at 200 • C were conducted with ZIF-67 cubic without TA treatment. From the SEM images in Figure 2, we found that even though the ZIF-67 is solid cubic, the thermal annealing still caused the shrink towards the inner side at the middle portion of each side. The morphology and structure have undergone apparent changes to a certain extent. This phenomenon is in line with previous studies in which ZIF-67 crystals obtained by direct annealing methods usually have a rough surface because of the aggregation of the nanoparticles [20,[40][41][42].
However, as the FESEM images shown in Figure 2, the TA-Co NBs lost their structural integrity when the thermal carbonization was conducted at 500 • C, 600 • C, and 800 • C. Uniform cubic structures were observed from the products obtained at 700 • C. The SEM characterized morphology in Figure 2 confirms that TA-assisted etching successfully avoids the high-temperature carbonization induced cubic shrink. We speculated that the thermal carbonization caused structure changes could be induced by the partial collapse of pores on the nano-boxes. We examined the porosity of the products obtained from different temperatures. From the N 2 adsorption-desorption isotherms curves shown in Figure 3, the porosity of Co 3 O 4 @CNBs obtained at 500, 600, 700, and 800 • C was 18.9 m 2 /g, 255.9 m 2 /g, 297.2 m 2 /g, 107.6 m 2 /g, respectively. The highest BET surface area is from the Co 3 O 4 @CNBs obtained at 700 • C. According to the N 2 adsorption-desorption isotherms, the adsorption isotherm for 500 • C and 600 • C were similar to a BET type II isotherm. While 700 • C and 800 • C appears the BET type IV shape adsorption according to BET classification. It is worth noting that, as shown in Figure 2, materials at 800 • C has shown the collapse of the cubics. Collectively, the SEM ( Figure 2) and BET results suggested that the intact cubic after annealing at 700 • C benefit the preserving of nano-pores on the nano box.  The TA etching reaction is illustrated in Supplementary Information Figure S2. First, the protons released from TA etch the ZIF-67, releasing the Co 2+ and 2-MIM simultaneously. At the same time, Co 2+ and TA coordinate together quickly to form the TA-Co shell. The attached TA block the exposed surface of ZIF-67, thus protecting the outer parts of MOFs from further etching, resulting in internal etching of the ZIF-67 to form TA-Co NBs [29,37,39].  The TA etching reaction is illustrated in Supplementary Information Figure S2. First, the protons released from TA etch the ZIF-67, releasing the Co 2+ and 2-MIM simultaneously. At the same time, Co 2+ and TA coordinate together quickly to form the TA-Co shell. The attached TA block the exposed surface of ZIF-67, thus protecting the outer parts of MOFs from further etching, resulting in internal etching of The highest BET surface area is from the Co3O4@CNBs obtained at 700 °C. According to the N2 adsorption-desorption isotherms, the adsorption isotherm for 500 °C and 600 °C were similar to a BET type II isotherm. While 700 °C and 800 °C appears the BET type IV shape adsorption according to BET classification. It is worth noting that, as shown in Figure 2, materials at 800 °C has shown the collapse of the cubics. Collectively, the SEM ( Figure 2) and BET results suggested that the intact cubic after annealing at 700 °C benefit the preserving of nano-pores on the nano box.      The structures of Co 3 O 4 @C and Co 3 O 4 @CNBs were further compared through TEM characterization. Thermal annealing action caused the shrink of ZIF-67, resulting in a quadrangular star shape ( Figure 4A). From the HRTEM characterization, the (111) planes of the metallic Co can be differentiated from the packed Co 3 O 4 NPs ( Figure 4B). The HAADF-STEM images ( Figure 4C) and the corresponding elemental mapping images of C, Co, O, N elements in Co 3 O 4 @C ( Figure 4D) clearly show the shrinking towards the inside at the four corners, and the inside of the cubic packed with dense and aggregated Co 3 O 4 NPs. In comparison, with the action of TA, the dispersed Co 3 O 4 was preserved nicely within the nano box ( Figure 4E). HRTEM image of Co 3 O 4 @CNBs in Figure 4F shows the lattice fringe spacing is about 0.20 nm, corresponding to the (111) planes of Co 3 O 4 . The HAADF-STEM images of Co 3 O 4 @CNBs ( Figure 4G) and the corresponding elemental mapping images of C, Co, O, N elements in Co 3 O 4 @CNBs ( Figure 4H) confirmed that the Co 3 O 4 NPs are highly dispersed in nano box. The evenly distributed C element would ensure electron transfer during electrochemical detection. The C, Co, O, N elements mapping images of Co 3 O 4 @CNBs thermal annealed at 500 • C, 600 • C, and 800 • C were shown in Supplementary Information Figure S3. From the FESEM and elements mapping, we confirmed that the C and Co are evenly distributed on the cubic. As the temperature increases, the structure gradually collapses at 800 • C.
of TA, the dispersed Co3O4 was preserved nicely within the nano box ( Figure 4E). HRTEM image of Co3O4@CNBs in Figure 4F shows the lattice fringe spacing is about 0.20 nm, corresponding to the (111) planes of Co3O4. The HAADF-STEM images of Co3O4@CNBs ( Figure 4G) and the corresponding elemental mapping images of C, Co, O, N elements in Co3O4@CNBs ( Figure 4H) confirmed that the Co3O4 NPs are highly dispersed in nano box. The evenly distributed C element would ensure electron transfer during electrochemical detection. The C, Co, O, N elements mapping images of Co3O4@CNBs thermal annealed at 500 °C, 600 °C, and 800 °C were shown in Supplementary Information Figure S3. From the FESEM and elements mapping, we confirmed that the C and Co are evenly distributed on the cubic. As the temperature increases, the structure gradually collapses at 800 °C. Apart from tracking the reaction by morphological characterization (Figures  1-4), the crystalline materials were characterized by XRD, FTIR, and XPS. First, the C 1s, Co 2p peaks can be observed from the XRD spectra, confirming the success in synthesis ZIF-67 ( Figure 5A). The ZIF-67 precursor completely disappears after TA etching, indicating the completion of chemical transformation. Diffraction peaks of Co3O4@CNBs in XRD characterization perfectly match with the standard patterns of Co3O4 (PDF # 42-1467). The FTIR spectrum also supports the formation of Co3O4 ( Figure 5B). FTIR spectrum shows that the prominent peaks at 3400 cm −1 are attributed to the vibration and stretching bands of functional groups of TA, which on account of TA complete substitution of 2-methylimidazole during the etching process [37]. Another strong bands at 667 cm −1 is attributed to the stretching vibration mode of Co-O with Co 2+ [40].
XPS analysis was applied to reveal the elemental valence state of the Co3O4@CNBs. Compared with ZIF-67, TA-Co NBs present observable changes in C Apart from tracking the reaction by morphological characterization (Figures 1-4), the crystalline materials were characterized by XRD, FTIR, and XPS. First, the C 1s, Co 2p peaks can be observed from the XRD spectra, confirming the success in synthesis ZIF-67 ( Figure 5A). The ZIF-67 precursor completely disappears after TA etching, indicating the completion of chemical transformation. Diffraction peaks of Co 3 O 4 @CNBs in XRD characterization perfectly match with the standard patterns of Co 3 O 4 (PDF # 42-1467). The FTIR spectrum also supports the formation of Co 3 O 4 ( Figure 5B). FTIR spectrum shows that the prominent peaks at 3400 cm −1 are attributed to the vibration and stretching bands of functional groups of TA, which on account of TA complete substitution of 2-methylimidazole during the etching process [37]. Another strong bands at 667 cm −1 is attributed to the stretching vibration mode of Co-O with Co 2+ [40].
XPS analysis was applied to reveal the elemental valence state of the Co 3 O 4 @CNBs. Compared with ZIF-67, TA-Co NBs present observable changes in C and Co's contents, which are attributed to the introduction of TA and pyrolysis of organic ligands. As shown in Figure 5C, the spectrum of Co 2p can be best-fitted with two prominent peaks at binding energies by Co 2p 3/2 and Co 2p 1/2 peaks located at around 780.3 and 795.1 eV, corresponding to the state of Co 3 O 4 phase. According to the XPS analysis ( Figure 5D), the appearance of C=O, C-O, and C-C in a high-resolution spectrum of C 1 s are caused by the structure of TA [41]. As shown in Supplementary Information Figure S4, the TGA analysis reveals the weight content of Co 3 O 4 in the composite is about 46.3 wt%. The weight loss under 250 • C is attributed to the evaporation of water molecules and air absorbed by the sample surface [42]. By analyzing FESEM, TEM, XRD, XPS, and FTIR results, we confirmed that Co 3 O 4 nanoparticles well dispersed in Co 3 O 4 @CNBs synthesized from 2 mg/mL TA etching.
in Supplementary Information Figure S4, the TGA analysis reveals th tent of Co3O4 in the composite is about 46.3 wt%. The weight loss un attributed to the evaporation of water molecules and air absorbed b surface [42]. By analyzing FESEM, TEM, XRD, XPS, and FTIR results, that Co3O4 nanoparticles well dispersed in Co3O4@CNBs synthesized f TA etching.

Dispersed Co3O4 NPs in Porous Carbon Nano Box Facilitate the Sensitive Ele Detection of H2O2
CV measurements were conducted to compare the electroche mance of carbonized ZIF-67 (Co3O4@C) and Co3O4@CNBs modified electrode (Co3O4@C/GCE and Co3O4@CNBs/GCE) in H2O2 detection. the dotted line and solid line represent the Co3O4@C/GCE and Co3O4 respectively. The blue and purple lines represent the absence and add respectively. With the addition of 2 mM H2O2, a cathode current arou

Dispersed Co 3 O 4 NPs in Porous Carbon Nano Box Facilitate the Sensitive Electrochemical Detection of H 2 O 2
CV measurements were conducted to compare the electrochemical performance of carbonized ZIF-67 (Co 3 O 4 @C) and Co 3 O 4 @CNBs modified glassy carbon electrode (Co 3 O 4 @C/GCE and Co 3 O 4 @CNBs/GCE) in H 2 O 2 detection. In Figure 6A, the dotted line and solid line represent the Co 3 O 4 @C/GCE and Co 3 O 4 @CNBs/GCE, respectively. The blue and purple lines represent the absence and addition of H 2 O 2 , respectively. With the addition of 2 mM H 2 O 2 , a cathode current around the potential of −0.22 V can be observed clearly from Co 3 O 4 @CNBs/GCE. In contrast, as shown in Figure 6A, no noticeable change was observed from Co 3 O 4 @C/GCE, indicating that Co 3 O 4 @C is inactive for electrooxidation of H 2 O 2 . Figure 6B  According to the previous reports, the electrocatalysis of H 2 O 2 on the Co 3 O 4 @CNBs can be expressed by the following equation [43]: The obvious reduction current indicated that Co3O4@CNBs nanoco have an excellent electrocatalytic activity for H2O2. The CV cu Co3O4@CNBs/GCE were collected at different scan rates between −0.8 and 0.01 M PBS (pH = 7.4). The reduction peak currents were enhanced with in scan rates. The current was in good linear with the scan rates ( Figure 6C), ing that the H2O2 reduction on the Co3O4@CNBs/GCE's surface was a ty sorption control process. The different performance between Co3O4@C and Co3O4@CNBs towa sensing is discussed. The thermal annealing and subsequent low-tempera dation will cause the four edges to shrink inward pristine ZIF-67 ( Figure 4A structural changes are accompanied by the decrease of porosity (Supple Information Figure S5) because the porosity of Co3O4@C is 149.5 m 2 /g whi nificantly smaller than that of Co3O4@CNBs (297.2 m 2 /g). Furthermore, the aggregated Co3O4 nanoparticles in Co3O4@C ( Figure 4A-C) impact the a sites of Co3O4 to react with H2O2 and potentially reduce the specific reac contributing to the electrochemical reduction of H2O2 (Scheme 2A). The obvious reduction current indicated that Co 3 O 4 @CNBs nanocomposite have an excellent electrocatalytic activity for H 2 O 2 . The CV curves of Co 3 O 4 @CNBs/GCE were collected at different scan rates between −0.8 and 0.2 V in 0.01 M PBS (pH = 7.4). The reduction peak currents were enhanced with increasing scan rates. The current was in good linear with the scan rates ( Figure 6C), suggesting that the H 2 O 2 reduction on the Co 3 O 4 @CNBs/GCE's surface was a typical adsorption control process.
The different performance between Co 3 O 4 @C and Co 3 O 4 @CNBs towards H 2 O 2 sensing is discussed. The thermal annealing and subsequent low-temperature oxidation will cause the four edges to shrink inward pristine ZIF-67 ( Figure 4A,C). The structural changes are accompanied by the decrease of porosity (Supplementary Information Figure S5) because the porosity of Co 3 O 4 @C is 149.5 m 2 /g which is significantly smaller than that of Co 3 O 4 @CNBs (297.2 m 2 /g). Furthermore, the obvious aggregated Co 3 O 4 nanoparticles in Co 3 O 4 @C ( Figure 4A-C) impact the available sites of Co 3 O 4 to react with H 2 O 2 and potentially reduce the specific reaction area contributing to the electrochemical reduction of H 2 O 2 (Scheme 2A). For the Co 3 O 4 @CNBs obtained from TA etching, the TA layer balanced the shrinkage stresses at different directions applied on the cubic during the annealing process. The architecture integrity avoids pore-collapse induced Co 3 O 4 NPs aggregation (BET data in Figure 3 and TEM data in Figure 4). The porous structures would facilitate the transportation of H 2 O 2 into the Co 3 O 4 @CNBs during electrochemical measurement. In addition, the TA protective layer alleviated the "stresses induced orientation contraction", ensuring the uniform disperse of Co 3 O 4 in CNBs to react with H 2 O 2 (Scheme 2B).
OR PEER REVIEW 10 of 15 Scheme 2. Schematic illustrations of the reaction mechanism of (A) Co3O4@C and (B) Co3O4@CNBs towards H2O2 sensing.

Analytical Performance of Co3O4@CNBs Based H2O2 Sensors
To construct a sensitive H2O2 sensors, the electrochemical testing condition was optimized. The electrochemical behavior of Co3O4@CNBs/GCE was analyzed in 10 mL 0.01 M PBS (pH = 7.4). It is noted that the CV signal to H2O2 is affected by the concentration of Co3O4@CNBs and the adding volume. Supplementary Information Figure S6A,B show that 7 μL, 2 mg/mL Co3O4@CNBs leads to the highest signal. Hence, 7 μL 2 mg/mL Co3O4@CNBs were employed in the following study. The amperometric technique was employed to measure the response of Co3O4@CNBs modified electrode. The optimal working potential for detecting H2O2 was −0.22 V.
With the optimized Co3O4@CNBs loading and electrochemical working voltage, the sensitivity and working range of the Co3O4@CNBs H2O2 sensor were characterized. The electrochemical response was recorded when successive adding varying H2O2 concentrations into 10 mL 0.01 M PBS (pH 7.4) solution. As shown in

Analytical Performance of Co 3 O 4 @CNBs Based H 2 O 2 Sensors
To construct a sensitive H 2 O 2 sensors, the electrochemical testing condition was optimized. The electrochemical behavior of Co 3 O 4 @CNBs/GCE was analyzed in 10 mL 0.01 M PBS (pH = 7.4). It is noted that the CV signal to H 2 O 2 is affected by the concentration of Co 3 O 4 @CNBs and the adding volume. Supplementary Information Figure S6A,B show that 7 µL, 2 mg/mL Co 3 O 4 @CNBs leads to the highest signal. Hence, 7 µL 2 mg/mL Co 3 O 4 @CNBs were employed in the following study. The amperometric technique was employed to measure the response of Co 3 O 4 @CNBs modified electrode. The optimal working potential for detecting H 2 O 2 was −0.22 V.
With the optimized Co 3 O 4 @CNBs loading and electrochemical working voltage, the sensitivity and working range of the Co 3 O 4 @CNBs H 2 O 2 sensor were characterized. The electrochemical response was recorded when successive adding varying H 2 O 2 concentrations into 10 mL 0.01 M PBS (pH 7.4) solution. As shown in Figure 7A, the response was linear with H 2 O 2 concentrations from 0.01 to 359 µM with a correlation coefficient of R 2 = 0.995 (insert picture), and the regression equation was I (µA) = −5.42 × 10 −6 − 1.28 × 10 −6 C (µM). The detection limit calculated was 2.32 nM (ratio of signal-to-noise (S/N) = 3). (Table 1) showed that our electrocatalytic performance of Co 3 O 4 @CNBs was satisfactory and even better than previous sensors.  Anti-interference ability is one of the critical analytical indicators for nonenzymatic biosensors. The amperometric method was adopted to study the interference of major interfering substances on Co3O4@CNBs based H2O2 detecting. As shown in Figure 7B, there are negligible signal responding to 0.15 mM glucose (Glu), cysteine (Cys), dopamine (DA), uric acid (UA), ascorbic acid (AA), glycine (Gly), sucrose (SUC), glutathione (GSH), and urea. While 0.05 mM H2O2 can induce a significantly larger signal, suggesting good selectivity for reducing H2O2.

Comparison of Co 3 O 4 @CNBs based H 2 O 2 detection with other H 2 O 2 biosensors
The stability and reproducibility of the Co3O4@CNBs-based sensor were also tested. Supplementary Information Figure S7 shows the amperometric electrochemical response of five modified independent electrodes from different batches. After statistical analysis of the test results, the relative standard deviation (RSD) obtained by five parallel tests was 2.4%, demonstrating a good reproducibility. The actual concentration of H2O2 present and the detected concentration were further tested, and the results in Supplementary Information Table S1 showed that the recovery rate of the Co3O4@CNBs-based sensor is from 95.62% to 105.78%. For the stability experiment, the modified electrode was stored at 4 °C for 15 days, and the current response to H2O2 (2 mM) was recorded every three days. It can be seen that after 15 days of storage, the current response of the sensor is maintained at 92% of the initial current.

Real-Time Detection of H2O2 Secreted from Living Cells by Co3O4@CNBs
To investigate the capability in actual samples application, Co3O4@CNBs H2O2  Anti-interference ability is one of the critical analytical indicators for nonenzymatic biosensors. The amperometric method was adopted to study the interference of major interfering substances on Co 3 O 4 @CNBs based H 2 O 2 detecting. As shown in Figure 7B, there are negligible signal responding to 0.15 mM glucose (Glu), cysteine (Cys), dopamine (DA), uric acid (UA), ascorbic acid (AA), glycine (Gly), sucrose (SUC), glutathione (GSH), and urea. While 0.05 mM H 2 O 2 can induce a significantly larger signal, suggesting good selectivity for reducing H 2 O 2 .
The stability and reproducibility of the Co 3 O 4 @CNBs-based sensor were also tested. Supplementary Information Figure S7 shows the amperometric electrochemical response of five modified independent electrodes from different batches. After statistical analysis of the test results, the relative standard deviation (RSD) obtained by five parallel tests was 2.4%, demonstrating a good reproducibility. The actual concentration of H 2 O 2 present and the detected concentration were further tested, and the results in Supplementary Information Table S1 showed that the recovery rate of the Co 3 O 4 @CNBs-based sensor is from 95.62% to 105.78%. For the stability experiment, the modified electrode was stored at 4 • C for 15 days, and the current response to H 2 O 2 (2 mM) was recorded every three days. It can be seen that after 15 days of storage, the current response of the sensor is maintained at 92% of the initial current.

Real-Time Detection of H 2 O 2 Secreted from Living Cells by Co 3 O 4 @CNBs
To investigate the capability in actual samples application, Co 3 O 4 @CNBs H 2 O 2 sensor was explored to real-time detect H 2 O 2 from the living cell in a culture medium. The response of human epithelial cell HUVEC, mouse breast cancer cell 4T1, and human lung cancer cell A549 to PMA, a diester of phorbol, which can activate many cell types to produce H 2 O 2 , was studied. First, the potential cytotoxicity of Co 3 O 4 @CNBs was evaluated by the standard MTT assay. Supplementary Information Figure S8 reveals that no significant decrease in cell viability was observed from 10 to 50 µg·mL −1 Co 3 O 4 @CNBstreated HUVEC cells and HeLa cells, demonstrating its good biocompatibility. The response of cells to PMA stimulation was measured by amperometric signal recorded in DMEM at −0.22 V. PMA is an activator widely used in in vitro experiments, which can stimulate cells to produce H 2 O 2 . As shown in Figure 8, the current has no obvious change when only cells exist. A promote and sharp increase of current peak was observed from all three cells challenged by 2.5 µg/mL PMA. In contrast, injecting PMA and catalase (CAT), an enzyme that catalyzes the decomposition of H 2 O 2 into water and oxygen, at the same time will demolish the current change, which was observed in only PMA stimulation. Since CAT will decomposite the H 2 O 2 released by PMA treated cells, by adding the PMA and CAT, we confirmed the current changes observed from cell stimulated by PMA only were induced by cell-released H 2 O 2 . The amperometric signal ( Figure 8) recorded from the cells proves that the Co 3 O 4 @CNBs H 2 O 2 sensor can detect the H 2 O 2 released by living cells, highlighting its potential in studying cell metabolism. Next, the actual amount of H 2 O 2 released from living cells was calculated according to the current and the calibration curve shown in Figure 7B. First, the current of the point reaching the plateau was read from the reaction curve. Then the H 2 O 2 amount was calculated by placing the current value into the calibration curve equation. As shown in Figure 8, according to the current change and the calibration curve, the amount from three different cells was calculated at 0.16 µM (HUVEC), 0.26 µM (A549) and 0.19 µM (4T1), respectively. l. Sci. 2022, 23, x FOR PEER REVIEW 12 Co3O4@CNBs was evaluated by the standard MTT assay. Supplementary In mation Figure S8 reveals that no significant decrease in cell viability was obse from 10 to 50 μg·mL −1 Co3O4@CNBs-treated HUVEC cells and HeLa cells, dem strating its good biocompatibility. The response of cells to PMA stimulation measured by amperometric signal recorded in DMEM at −0.22 V. PMA is an vator widely used in in vitro experiments, which can stimulate cells to prod H2O2. As shown in Figure 8, the current has no obvious change when only exist. A promote and sharp increase of current peak was observed from all t cells challenged by 2.5 μg/mL PMA. In contrast, injecting PMA and catalase (C an enzyme that catalyzes the decomposition of H2O2 into water and oxygen, a same time will demolish the current change, which was observed in only P stimulation. Since CAT will decomposite the H2O2 released by PMA treated c by adding the PMA and CAT, we confirmed the current changes observed f cell stimulated by PMA only were induced by cell-released H2O2. The ampero ric signal ( Figure 8) recorded from the cells proves that the Co3O4@CNBs H2O2 sor can detect the H2O2 released by living cells, highlighting its potential in stud cell metabolism. Next, the actual amount of H2O2 released from living cells calculated according to the current and the calibration curve shown in Figur First, the current of the point reaching the plateau was read from the reaction cu Then the H2O2 amount was calculated by placing the current value into the cal tion curve equation. As shown in Figure 8, according to the current change and calibration curve, the amount from three different cells was calculated at 0.16 (HUVEC), 0.26 μM (A549) and 0.19 μM (4T1), respectively.

Conclusions
In conclusion, Co3O4@CNBs nanocomposites have been prepared using a facile green method. Their application in the determination of H2O2 has been explored. The TA improves the materials' specific surface area and provides more active sites, fu

Conclusions
In conclusion, Co 3 O 4 @CNBs nanocomposites have been prepared using a facile and green method. Their application in the determination of H 2 O 2 has been explored. The used TA improves the materials' specific surface area and provides more active sites, further enhancing its electrocatalysis to reduce H 2 O 2 . The Co 3 O 4 @CNBs/GCE exhibits a good selectivity and high sensitivity for the determination of H 2 O 2 . Furthermore, the Co 3 O 4 @CNBs H 2 O 2 sensor can detect the H 2 O 2 secreted by HUVEC cells and 4T1, A549 cancer cells, highlighting its potential in biosensing and catalysis and biomedicine.