Axially Coordinated Gold Nanoclusters Tailoring Fe–N–C Nanozymes for Enhanced Oxidase‐Like Specificity and Activity

Abstract Metal–organic frameworks (MOF) derived nitrogen‐doped carbon‐supported monodisperse Fe (Fe–N–C) catalysts are intensively studied, but great challenges remain in understanding the relationship between the coordination structure and the performance of Fe–N–C nanozymes. Herein, a novel nanocluster ligand‐bridging strategy is proposed for constructing Fe‐S1N4 structures with axially coordinated S and Au nanoclusters on ZIF‐8 derived Fe–N–C (labeled Aux/Fe‐S1N4‐C). The axial Au nanoclusters facilitate electron transfer to Fe active sites, utilizing the bridging ligand S as a medium, thereby enhancing the oxygen adsorption capacity of composite nanozymes. Compared to Fe‐N‐C, Aux/Fe‐S1N4‐C exhibits high oxidase‐like specificity and activity, and holds great potential for detecting acetylcholinesterase activity with a detection limit of 5.1 µU mL−1, surpassing most reported nanozymes.

-weighted Fourier-transformed spectra.g) Fitting curves of the EXAFS in the r-space and k-space (inset of (g)).

Structure Characterization
As illustrated in Figure 1a, to construct Au x /Fe-S 1 N 4 -C nanozymes, negatively charged Au 25 (L-Cys) 18 (denoted as Au 25 ) nanoclusters were anchored onto positively charged ZIF-8 derived Fe-N-C matrix of rhombohedral dodecahedron morphology through impregnation and pyrolysis at given temperature (Figures S1and S2, Supporting Information).For convenience, the final samples were labeled as Au x /Fe-N-C-T (where T represents the pyrolysis temperature).The sample obtained at 300 °C was specially designated as Au x /Fe-S 1 N 4 -C to highlight the coordination structure of its active sites.
Figure 1b shows a representative bright-field transmission electron microscopy (TEM) image of Au x /Fe-S 1 N 4 -C.In contrast to Au 25 /Fe-N-C without undergoing re-pyrolysis (Figure S3, Supporting Information), highly dispersed and uniformly sized Au nanoclusters were more clearly observed on the Fe-N-C matrix, as revealed in the high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image of Au x /Fe-S 1 N 4 -C (Figure 1c).Remarkably, the average size of Au nanoclusters was only 1.7 nm, akin to the pristine Au 25 (L-Cys) 18 .Furthermore, aberration-corrected high-angle annular dark-field scanning transmission electron microscopy (AC-HAADF-STEM, Figure S4, Supporting Information) shows that pyrolysis temperature of 300°C has little effect on the Au core of the clusters and does not produce individual atoms of Au.Elemental mapping confirmed the uniform distribution of Fe, Au and S element on the Fe-N-C matrix, corroborating the high dispersion of Au nanoclusters (Figure 1d).Thermal gravity and differential thermal analysis (Figure S5, Supporting Information) revealed that Au 25 nanoclusters experienced partial decomposition during pyrolysis at 300 °C, which resulted in some of the S atoms in the L-Cys ligands bonding to Fe-N-C matrix, thereby enhancing the interaction between Au 25 nanoclusters and Fe-N-C matrix.Moreover, the graphitization degree of Fe-N-C matrix was improved to some extent (Figure S6, Supporting Information).
X-ray absorption fine structure (XAFS) analysis was used to further probe into the chemical state and coordination structure of Fe sites within Au x /Fe-S 1 N 4 -C.The analysis on Fe K-edge x-ray absorption near-edge structure (XANES) of Au x /Fe-S 1 N 4 -C (Figure 1e) suggested that the average oxidation state of Fe in Au x /Fe-S 1 N 4 -C was between +2 and +3.The Fouriertransformed k 3 -weighted extended XAFS (FT-EXAFS) spectrum of Au x /Fe-S 1 N 4 -C at Fe K-edge presented a main peak at 1.63 Å, FePc at 1.50 Å and Fe-S bond at 1.87 Å, while no Fe─O bond (1.47 Å) or Fe─Fe bond (2.18 Å) was detected (Figure 1f; Table S1, Supporting Information).Besides, the wavelet transforms (WT) contour plot (Figure S7, Supporting Information) of Au x /Fe-S 1 N-C exhibited higher intensity maximum (5.2 Å) than FePc (2.7 Å), which is attributed to the Fe-S bonding.The above results indicate that Fe atoms are atomically dispersed and coordinated by N and S atoms.The corresponding EXAFS fitting of the first coordination shell was performed to ascertain the structural parameters and quantitative chemical configuration of Fe atoms (Figure 1g).The coordination number of Fe atom was close to five, comprising four N atoms and one S atom (Table S1, Supporting Information), further affirming the formation of Fe-S 1 N 4 moieties in Au x /Fe-S 1 N 4 -C.The above analysis indicated that Au x /Fe-S 1 N-C was characterized by Fe-S 1 N 4 sites, where Au x nanoclusters were anchored onto Fe-N 4 sites through the axial coordination of S atoms with Fe-N 4 .
Of note, the re-pyrolysis temperature exerted a significant impact on the size and distribution of the Au x species anchored onto the Fe-N-C matrix.With the re-pyrolysis temperature increasing progressively from 300 °C to 400, 500, and 600 °C, the average particle size of Au species increased from 1.7 nm to 3.4, 3.9, and 5.1 nm, correspondingly (Figure S8, Supporting Information).Once the re-pyrolysis temperature exceeded 400 °C, characteristic diffraction peaks attributed to metallic Au became discernible in the X-ray diffraction (XRD) patterns of Au x /Fe-N-C-T samples, and their intensity increased with the rise of temperature (Figure 2a).To reveal the influence of the re-pyrolysis on the chemical structure and electron structure of Fe-N-C matrix as well as Au x species anchored on Fe-N-C matrix, X-ray photoelectron spectroscopy (XPS) analysis were conducted.The metal content of each sample was similar (Table S2, Supporting Information).The N 1s XPS spectrum of Au x /Fe-S 1 N 4 -C can be deconvoluted to pyridinic-N (397.9 eV), Fe-N (398.5 eV), pyrrolic-N (399.3 eV), and graphitic-N (400.8 eV) species (Figure 2b).Furthermore, the interconversion and stability of the structure with increasing temperature [8] resulted in an increase in the graphitic nitrogen content, a decrease in the pyrrolic nitrogen content, and an initial increase and then a decrease in the pyridinic nitrogen (Figure S9 and Table S3, Supporting Information).The metal-S (Fe-S and Au-S) species at 161.9 eV appeared in the S 2p XPS spectrum of Au x /Fe-S 1 N 4 -C (Figure 2c).Interestingly, the proportion of metal-S bonds increased and then decreased with the increase of the re-pyrolysis temperature, and Au x /Fe-S 1 N 4 -C presented the highest ratio of metal-S bonds (Figure 2d, see Figure S10 and Table S4, Supporting Information for details).On the other hand, the Au 4f peak gradually shifted to lower binding energies with increasing re-pyrolysis temperature (Figure 2e).This indicated that the loss of thiolate ligands leads to the reduction of positive monovalent Au on the surface to zero valence, and the Au─S bond weakened with re-pyrolysis temperature elevation.Taking all the results together, the increase of metal-S bond in the Au x /Fe-S 1 N 4 -C should be attributed to the increase of Fe─S bond.That's to say, S achieves optimal binding to Au and Fe (Au─S─Fe bonds) at 300 °C.

Specific Oxidase-Like Activity Tests
The peroxidase and oxidase-like activities of both Fe-N-C and Au x /Fe-S 1 N 4 -C were tested with 3,3′,5,5′-tetramethylbenzidine (TMB) as the chromogenic substrate (Figure 3a; Figures S11 and S12, Supporting Information).While the peroxidase activity of Fe-N-C marginally exceeded its oxidase activity, Au x /Fe-S 1 N 4 -C exhibited the remarkable selectivity for oxidase-like activity.The oxidase-like activity Au x /Fe-S 1 N 4 -C was 12 times that of its peroxidase activity.To reveal the role of Fe-N 4 sites and Au nanoclusters, the oxidase-like activities of the Fe-free Au x /N-C-300 (Figure S13, Supporting Information), Au NP/Fe-N-C (Figure S14, Supporting Information), L-Cys/Fe-N-C-300 (Figure S15, Supporting Information), Au 25 and the Au x /Fe-N-C-T samples obtained at different re-pyrolysis temperatures (i.e., 200, 300, 400, 500, and 600 °C) were also tested (Figure 3b; Figure S16, Supporting Information).Among all the samples tested, Au x /Fe-S 1 N 4 -C obtained at 300 °C showed the best selectivity and catalytic activity, with a 2.4-fold and 6.7-fold increase in oxidase-like activity compared to Au x -free sample (Fe-N-C) and the Fe-free sample (Au x /N-C-300), respectively.The poor activity of Au x /N-C-300 proved that the active centers are attributed to Fe sites rather than Au x .As shown in Figure 3b, Au NP/Fe-N-C (Figure S14, Supporting Information) exhibits lower activity compared to Fe-N-C due to the partial occupancy of the active site by Au.After calcination at 300 °C, L-Cys in L-Cys/Fe-N-C-300 is essentially residue-free (Figure S15, Supporting Information), resulting in similar oxidase-like activity to Fe-N-C.Notably, Au 25 shows almost no oxidase-like activity.Comparative analysis of the oxidase-like activities of these samples reveals that Au 25 , L-Cys, and Au NP are not active centers.In addition, the oxidase-like activities of Au x /Fe-N-C-T showed a volcanic relationship with the re-pyrolysis temperature.Clearly, the re-pyrolysis temperature exerted a significant effect on the  1).

Catalytic Mechanism
An in-depth exploration of the catalytic mechanisms and reaction pathways offers promising insights for the rational design of MOF-derived Fe-N-C nanozymes.The investigation of intermediate radicals was conducted through in situ electron paramagnetic resonance (EPR), using DMPO as a spin trapping agent (Figure 4a − compared to the Fe-N-C samples.To further elucidate the O 2 activation at the Au x /Fe-S 1 N 4 sites, in situ fourier transform infrared (FTIR) spectroscopy was employed.In the presence of O 2 , the signal at 1100 cm −1 on Fe-N-C gradually increased, attributed to * OH (Figure S17, Supporting Information).For the Au x /Fe-S 1 N 4 -C (Figure 4b), the absorption peak exhibited a subtle shift compared with Fe-N-C, indicating that the formation of Au x /Fe-S 1 N 4 sites affects the electronic structure of Fe active sites.According to the literature, the absorption peaks at 930, 1169, and 1460 cm −1 represented the adsorption of * O 2 , * OOH, * O 2 − on Fe-S 1 N 4 sites, [23] respectively, which was consistent with the EPR results.This affirms that the catalytic mechanism of Au x /Fe-S 1 N 4 -C diverges from that of Fe-N-C.The introduction of Au x cluster-S by axial coordination with Fe-N 4 changes the electronic structures of Fe-N-C catalyst, which endows Au x /Fe-S 1 N 4 -C with superior oxygen adsorption capacity compared to Fe-N-C, thus exhibiting excellent oxidase activity and selectivity.
To understand the role of Au x -S axial coordination in enhancing the activity of Au x /Fe-S 1 N 4 -C oxidase-like enzymes, we investigated the adsorption energies and associated reaction barrier via density-functional theory calculations.Given that the metal core of the Au 25 cluster is Au 13 , [24] which has the greatest impact on the electronic structure of S, and considering the computational requirements, we employed Au 13 as the representative model to present Au x species formed on the Fe-N-C after repyrolysis (Figure 4c

Acetylcholinesterase Activity Detection
As a proof-of-concept application, the as-designed highly specific oxidase-like Au x /Fe-S 1 N 4 -C was applied to a colorimetric assay for sensitive, rapid and efficient detection of AChE.As shown in Figure 5a, AChE first catalyze the hydrolysis of acetylthiocholine (ATCh) into thiocholine (TCh).TCh as a sulfhydryl molecule tends to coordinate with metal atoms, severely blocking the active sites of Au x /Fe-S 1 N 4 -C and consequently inhibiting its oxidase-like activity.It is well-established that certain sulfhydryl biomolecules, such as L-Cys and TCh, are directly implicated in several diseases. [12]Here, we initially explored the influence of L-Cys on the activity of Au x /Fe-S 1 N 4 -C nanozymes.As expected, the absorbance at 652 nm substantially decreased upon L-Cys addition (Figure 5b).This implies that the active sites of the nanozymes are blocked, and the antioxidant properties of L-Cys   scavenge the free radicals, leading to a reduction in color intensity.The calculated limit of detection (LOD) is as low as 11.5 nM (Figure 5c) according to the 3 rules (LOD = 3/S, where  is the relative standard deviation (RSD), S is the slope of the calibration curve for the linear range portion), surpassing most reported nanozymes. [13]e further investigated the sensitivity of Au x /Fe-S 1 N 4 -C for the detection of AChE activity in the presence of ATCh (5 mm) across AChE concentrations ranging from 1 to 10 mU mL −1 .As AChE increased, the activity of Au x /Fe-S 1 N 4 -C decreased, resulting in lower absorbance values (Figure 5d).The relationship between calculated absorbance variations and the AChE activity is described in Figure 5e.The results indicate that the Au x /Fe-S 1 N 4 -C based biosensor can be used to assess AChE activity with a LOD of 5.1 μU mL −1 .Clearly, the Au x /Fe-S 1 N 4 -C+TMB system exhibited superior performance compared to previous reports (Table S5, Supporting Information).
Furthermore, a series of potential interfering substances were used to examine the anti-interference ability of the biosensor (Figure S20, Supporting Information).Negligible changes were observed in the detection of AChE.These results highlight the promising application of Au x /Fe-S 1 N 4 -C in AChE activity detection.Sustaining stability and obtaining reproducible experimental results are crucial factors for achieving practical applications of the AChE biosensor.As depicted in Figure S21 (Supporting Information), the oxidase-like activity of the AChE biosensor slightly decreased after storage at room temperature, indicating its remarkable stability.Furthermore, the RSD of five independently prepared AChE biosensors was 3.06%.These results demonstrate that the developed Au x /Fe-S 1 N 4 -C oxidaselike nanozymes-based sensor has excellent long-term stability and reproducibility, making it highly potential for practical applications.
To confirm the feasibility of the proposed method in real biological samples, we applied the standard addition method to detect L-Cys (Table S6, Supporting Information) and AChE (Table S7, Supporting Information) activity using diluted normal human serum (5%) as the matrix. [25]The different concentrations of L-Cys and AChE solutions were added to the two diluted serum samples, respectively.It can be noticed that the method detection of L-Cys and AChE activity exhibited satisfactory recovery ranging (100.21-99.66%,101.00-99.67%)and low RSD (0.13-0.60%, 0.25-1.98%).This suggests that our colorimetric assay has significant potential for application in real samples.

Conclusion
In summary, a novel Au x /Fe-S 1 N 4 -C nanozyme containing Fe-S 1 N 4 structures with axially coordinated S and Au cluster was successfully constructed via our proposed cluster ligand-bridging strategy.Our study demonstrates that the axial Au nanoclusters facilitate electron transfer to Fe sites via the bridging ligand S, greatly enhancing the oxygen adsorption capacity of Au x /Fe-S 1 N 4 -C.Consequently, compared to Fe-N-C, Au x /Fe-S 1 N 4 -C exhibited remarkable selectivity, with an oxidase-like activity 12 times higher than its peroxidase.Notably, Au x /Fe-S 1 N 4 -C demonstrated a lower LOD and better anti-interference ability in AChE detection, which has great potential for application in real samples.This work helps us to understand the structureselectivity relationship of peroxidase and oxidase-like enzymes, laying the foundation for the rational design and realization of highly selective oxidase-like nanozymes.

Experimental Section
Synthesis of Fe-N-C: Typically, zinc nitrate hexahydrate (2.38 g, 8 mmol) and ferrocene (0.465 g, 2.5 mmol) were dissolved in methanol (60 mL).The solution was then mixed with a solution of 2-methylimidazole (2.628 g, 15 mmol) in methanol (60 mL).The resulting mixture was stirred for 6 h at room temperature.Afterward, Fe@ZIF-8 was obtained by centrifugation and washed thoroughly with methanol, and dried at 70 °C under vacuum.The Fe@ZIF-8 power was transferred into a ceramic boat, placed in a tube furnace, and heated to 900 °C for 2 h with a heating rate of 5 °C min −1 under H 2 /Ar (5%:95%), followed by natural cooling to room temperature.The resulting powder was then washed in 0.5 mol L −1 sulfuric acid at 80°C for 10 h to remove unstable species.

Synthesis of Au x /Fe-N-C-T:
To form L-Cys-Au(I) complexes, aqueous solutions of HAuCl 4 (23.4 mm, 0.4 mL) and L-Cys (5 mm, 4 mL) were mixed in water (4.7 mL).An aqueous NaOH solution (1 M, 0.1 mL) was then introduced into the reaction mixture, followed by the addition of 0.2 mL of NaBH 4 solution (prepared by dissolving 43 mg of NaBH 4 powder in 10 mL of 0.2 m NaOH solution).The L-Cys -Au 25 nanoclusters were obtained after 3 h for further synthesis.Subsequently, 25 mg Fe-N-C were dissolved in water (5 mL), which was added to the above L-Cys-Au 25 aqueous solution.After stirring for 20 h at room temperature, the Au 25 /Fe-N-C was obtained by centrifugation, washed thoroughly with water, and dried at 60 °C under vacuum.The resulting Au 25 /Fe-N-C power was then transferred into a ceramic boat and heated to 200-600 °C for 2 h with a heating rate of 5 °C min −1 under N 2 , followed by natural cooling to room temperature.

Figure 1 .
Figure 1.Synthesis and characterizations of Au x /Fe-S 1 N 4 -C.a) Schematic preparation process.b) TEM and c) HAADF-STEM images.Inset in (c) is corresponding particle-size histograms of Au nanoclusters supported on Fe-N-C.d) EDS elemental mapping.e) Normalized XANES spectra at Fe Kedge and f) the corresponding k 3 -weighted Fourier-transformed spectra.g) Fitting curves of the EXAFS in the r-space and k-space (inset of (g)).

Figure 2 .
Figure 2. Characterizations of Au 25 /Fe-N-C and Au x /Fe-N-C-X obtained with different re-pyrolysis temperatures.a) XRD patterns.b) High-resolution XPS spectra of N 1s and c) S 2p of Au x /Fe-S 1 N 4 -C.d) Metal-S contents calculated by XPS spectra.(e) XPS spectra of Au 4f.
).The Au x /Fe-S 1 N 4 -C exhibited robust characteristic signals corresponding to DMPO-•O 2 − , while the signals from the Fe-N-C was much weaker.This observation suggests that Au x /Fe-S 1 N 4 -C exhibits enhanced capacity in activating O 2 to generate •O 2

Figure 3 .
Figure 3. Mimic enzyme performance of Fe-N-C and Au x /Fe-N-C-T obtained with different re-pyrolysis temperatures.a) Comparison of peroxidase-like and oxidase-like activities.b) The absorbance of different samples at 652 nm.c) Steady-state kinetic assay toward TMB.d) Michaelis-Menten curves toward TMB.
; Figure S18, Supporting Information).Significantly, the rate-determining step for Fe-N-C corresponded to the initial process (O 2 + H + + e − → * OOH) due to the weak O 2 adsorption on Fe-N-C (E ads = −0.41eV, charge = 0.254 |e|) (Figure 4d).In contrast, Au x /Fe-S 1 N 4 -C exhibited significantly enhanced O 2 adsorption (E ads = −0.67 eV, charge = 0.3723 |e|) through axial Au x -S coordination Fe sites, promoting four-electron oxygen reduction.In short, the additional axially-coordinated Au x -S give Au x /Fe-S 1 N 4 -C a strong electron push effect, which activates O 2 and facilitates the cleavage of the O-O band, ultimately leading to significantly enhanced oxidase-like activity.

Figure 4 .
Figure 4. Catalytic Mechanism of Fe-N-C and Au x /Fe-S 1 N 4 -C.a) EPR spectra b) In situ FTIR recorded under an O 2 atmosphere.c) The reaction pathways of O 2 reduction to H 2 O. d) Free energy diagram for oxygen reduction.Inset in (d) is structure of O 2 adsorption configuration, differential charge density and Bader charge.(The purple and green areas show the accumulation and depletion of charges with the iso-surface value of 1.0 × 10 −4 e/bohr 3 .The bronze, light blue, brown, yellow, gold, red, and white spheres represent the C, N, Fe, S, Au, O, and H atoms, respectively.).

Figure 5 .
Figure 5. Au x /Fe-S 1 N 4 -C +TMB system for colorimetric detection.a) Schematic illustration.b) UV-vis spectra in the presence of L-Cys.c) Linear range for detection of L-Cys.d) UV-vis spectra in the presence of AChE.e) The linear relationship for detection of AChE.

Table 1 .
Comparison of the kinetic parameters of the oxidase-like nanozymes.