Axial modulation of Fe sites for boosted electrochemical oxidation

Abstract

Fe/NC single-atom catalysts have attracted extensive attention due to their maximal atomic utilization and tunable coordination environments. However, the structure–activity relationship of Fe single atoms in electrooxidation remains unclear. Herein, we report a defect engineering strategy to fine-tune the charge configuration of FeN₄ sites by introducing an axial N ligand and constructing FeN₅–Fe₁/NC. This asymmetric coordination environment enhances the catalytic activity for dopamine (DA) oxidation, delivering a 2.1-fold improvement over traditional Fe₁/NC. The FeN₅–Fe₁/NC biosensor exhibits a wide linear detection range of 0.05–500 μM with a low detection limit of 23 nM for DA. Additionally, theoretical calculations confirm that axial N coordination modulates the electronic structure of the Fe center, optimizes intermediate adsorption, and lowers the energy barrier for DA oxidation. This work provides valuable insights into the rational design of single-atom catalysts for high-performance electrochemical sensing and fundamental mechanistic studies at the atomic scale.

---

New concepts

Reasonable design and construction of catalysts are desirable for boosting their catalytic performances. Single-atom catalysts (SACs) with MN_x active sites have made significant strides in achieving signal amplification of sensing. However, the symmetrical electronic density in the MN₄ moiety might impede the adsorption and activation of reaction intermediates, limiting catalytic performance. The construction of FeN₄ structures with axially coordinated N ligand (FeN₅–Fe₁/NC) exhibit 2.1 times higher catalytic activity toward DA oxidation compared to Fe₁/NC. Theoretical calculations confirm that axial N coordination modulates the electronic structure of the Fe center, optimizes intermediate adsorption, and lowers the energy barrier for DA oxidation. This study opens a new avenue for constructing high-performance electrochemical sensors and further widens the application scope of electrochemical SAC-based sensing for complex biological systems.

Introduction

Dopamine (DA) is a crucial neurotransmitter involved in regulating pleasure, motivation, and motor control across the renal, cardiovascular, and central nervous systems.^1–4 Abnormal DA levels are closely linked to various pathological conditions, including Alzheimer's disease, pituitary tumors, and schizophrenia.^5–8 Therefore, accurate and sensitive detection of DA is essential for early diagnosis and effective biomedical monitoring.^9–11 For this purpose, numerous techniques, such as liquid chromatography, spectrophotometry, colorimetry fluorescence, and electrochemistry have been developed to detect DA.^12–15 Among them, electrochemical sensing technology with low cost, fast response time, and simple operation advantages has attracted considerable attention to meet the need for in vivo and continuous health monitoring.^16–18 At present, a variety of nanomaterials with low cost, favorable accessibility, and high stability have been investigated in electrochemical sensors and biosensors.^19–21 However, traditional metal oxides and carbon materials often suffer from low metal atom utilization and limited catalysis efficiency, leading to biosensors with insufficient sensitivity and selectivity.^22–25 Therefore, the development of advanced electrocatalysts with high catalytic performance is of great significance.

Single-atom catalysts (SACs) featuring maximized atom utilization have sparked widespread interest in advanced catalysis.^26–30 Compared to conventional nanomaterials, SACs with rich active sites, tunable electronic and geometric structures are favorable for exploring the electrochemical process and signal amplification of electrochemical sensing. According to the DA oxidation mechanism via a 2e⁻ pathway to form dopamine-quinone, the hydroxyl (OH) adsorption on MN₄ sites has been investigated to serve as a mediator to facilitate the O–H bond breakage of DA and removal of the dehydrogenation product.^31–35 Thus, modulation of the intrinsic activity of MN₄ sites is still necessary to optimize the adsorption strengths of intermediates and achieve satisfactory properties for sensing. Breaking the square planar symmetry of MN₄ sites is believed to modulate the electronic structure of active centers and adsorption/desorption intermediates, which plays a vital role in enhancing their catalytic properties.^27,28 The introduction of heteroatom dopants (e.g., P, S, B) into the carbon substrate can optimize the electronic structure of active sites through long-range delocalization, further improving adsorption energy with reaction intermediates.^36–39 Besides, the heterogonous metal introduction can regulate the electronic structure of the asymmetrical active sites, lowering the energy barrier of reaction intermediates.^40–42 However, precise ligand coordination around metal centers in such multi-heteroatom dopant systems remains challenging.⁴³ Thus, it is urgently needed to develop new strategies to prepare high-performance SACs for sensitive electrochemical sensing.

Herein, we report the construction of FeN₄ structures with axially coordinated N ligand (FeN₅–Fe₁/NC) to promote DA oxidation. Due to pore defects originating from a dual template, Fe₁/NC provide an adjustable substrate for improving the loading capability and electronic structure of active sites. FeN₅–Fe₁/NC exhibit 2.1 times higher catalytic activity toward DA oxidation compared to Fe₁/NC. Theoretical calculations reveal that the axial N coordination in FeN₅ sites optimizes the adsorption strength for intermediates, thereby facilitating DA oxidation at the Fe center and enhancing the intrinsic catalytic activity. Due to its abundant and efficient active sites, the fabricated FeN₅–Fe₁/NC biosensor achieves a wide detection range of 0.05–500 μM with a low detection limit of 23 nM. This work encourages the pursuit of SACs with asymmetric coordination environments for high-performance electrochemical sensing and catalysis.

Results and discussion

As shown in Fig. 1a, the atomically dispersed FeN₅–Fe₁/NC and FeN₄–Fe₁/NC were synthesized using a two-step doping strategy. The morphology of FeN₅–Fe₁/NC was characterized by scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Fig. S1 reveals that FeN₅–Fe₁/NC have an irregular stacked 3D structure. Fig. 1b shows a hierarchically spherical pore structure of FeN₅–Fe₁/NC with a size of around 20 nm. Significantly, the absence of Fe nanoparticles or nanoclusters on the surface of the carbon substrate is observed by the high-resolution transmission electron microscopy image (HRTEM), indicating that the addition of Fe atoms did not induce nanoparticle growth (Fig. 1c). Furthermore, the aberration-corrected scanning transmission electron microscope (AC-STEM) image further reveals that the highly isolated bright spots are homogeneously distributed on the entire support, indicating the existence of single Fe atoms (Fig. 1d). Besides, the energy-dispersive X-ray spectroscopy (EDS) mapping images demonstrate the uniform distribution of C, N, and Fe elements within FeN₅–Fe₁/NC (Fig. 1e). Inductively coupled plasma-mass spectrometry (ICP-MS), an elemental analysis technique for quantifying metal ions, was performed to determine the content of Fe in catalysts. The Fe contents of FeN₅–Fe₁/NC and FeN₄–Fe₁/NC are 0.96 wt% and 0.91 wt%, respectively, both higher than that of Fe₁/NC (0.65 wt%), confirming the successful introduction of secondary doping (Table S1). The N₂ adsorption–desorption test was performed to further study the specific surface area and pore structure of FeN₅–Fe₁/NC (Fig. S2). FeN₅–Fe₁/NC possess a mesoporous structure, and a large specific surface area (830 m² g⁻¹), which could facilitate rapid mass transport, thereby enhancing the catalytic activities. In the XRD pattern (Fig. S3), the as-prepared catalysts show well-defined carbon diffraction peaks, assigned to (002) and (101) crystal planes of carbon. No characteristic peaks of metallic Fe or its oxides are observed, revealing the absence of Fe-based crystalline phases.

Fig. 1 (a) Schematic diagram of the synthesis of FeN₅–Fe₁/NC and FeN₄–Fe₁/NC. (b) TEM, (c) HRTEM, (d) AC-STEM, and (e) EDS mapping images of FeN₅–Fe₁/NC.

X-ray photoelectron spectroscopy (XPS) was carried out to assess the compositions and chemical states of FeN₅–Fe₁/NC. Four characteristic peaks are observed in the full XPS spectra, revealing the existence of C, N, O, and Fe elements in FeN₅–Fe₁/NC (Fig. S4a). As for N 1s spectra of FeN₅–Fe₁/NC (Fig. 2a), five peaks can be observed at 398.3, 399.1, 400.3, 401.0, and 403.9 eV, assigned to pyridinic-N, FeN_x, pyrrolic-N, graphitic-N, and oxidized N, respectively.⁴⁴ The contents of these N species are shown in Table S2. In comparison to Fe₁/NC (3.2%), the percentage of FeN_x (6.9%) of FeN₅–Fe₁/NC is higher, consistent with the ICP-MS result. In addition, the pyridine-N content in FeN₅–Fe₁/NC is also higher than that in Fe₁/NC, which is conducive to improving the electrocatalytic activity of catalysts.⁴⁵ C 1s spectra of FeN₅–Fe₁/NC can be deconvoluted into four types of carbon peaks of C–C, C–N, C–O, and C=O, respectively (Fig. S4b). To gain more detailed information about the electronic structure and coordination environment of Fe atoms, X-ray absorption spectroscopy (XAS) images were recorded. As shown in Fig. 2b, X-ray absorption near-edge structure (XANES) spectra of FeN₅–Fe₁/NC show absorption edge positions between those of Fe foil and Fe₂O₃, revealing that the valence state of Fe species in FeN₅–Fe₁/NC lies between 0 and +3. Fourier-transform extended X-ray absorption fine structure (EXAFS) spectra of FeN₅–Fe₁/NC exhibit a broad peak at 1.93 Å (Fig. 2c), which is close to Fe–N. No obvious Fe–Fe bond is observed. The EXAFS fitting results (Fig. 2d and Table S3) show that an average coordination number of N is about 5.1 for FeN₅–Fe₁/NC, and 4.2 for FeN₄–Fe₁/NC, respectively.

Fig. 2 (a) N 1s XPS spectra of FeN₅–Fe₁/NC and Fe₁/NC. (b) Fe K-edge XANES and (c) EXAFS spectra of FeN₅–Fe₁/NC, Fe₂O₃, and Fe foil. EXAFS fitting curves at R space for (d) FeN₅–Fe₁/NC and (e) FeN₄–Fe₁/NC. (f) WT images of FeN₅–Fe₁/NC and Fe foil.

To investigate the catalytic property of FeN₅–Fe₁/NC and better understand the role of FeN₅–Fe₁/NC in the catalytic process, an electrochemical sensor was constructed for DA detection. The electrocatalytic activity of the as-prepared samples toward DA was measured in 0.1 M PBS (pH = 7.4). Fig. 3a displays the schematic diagram of the electrochemical responses toward DA. Differential pulse voltammetry (DPV) was carried out to further explore the electrochemical performance of the as-prepared catalysts in the presence of 500 μM DA. As shown in Fig. 3b, FeN₅–Fe₁/NC exhibits a DA oxidation peak current of 118 μA, which is 2.1 and 1.5 times higher than Fe₁/NC, and FeN₄–Fe₁/NC, respectively. This indicates that secondary atom doping can simultaneously adjust the coordination environment of the active site and enhance the metal loading capacity, thereby improving catalytic performance. The electrochemical behaviors of FeN₅–Fe₁/NC, FeN₄–Fe₁/NC, and Fe₁/NC were further investigated by cyclic voltammetry (CV). As shown in Fig. 3c, we use [Fe(CN)₆]^3−/4− as a redox probe to compare bare glass carbon electrode (GCE), Fe₁/NC, FeN₄–Fe₁/NC and FeN₅–Fe₁/NC, respectively. FeN₅–Fe₁/NC show higher electrochemical active areas than FeN₄–Fe₁/NC and Fe₁/NC, revealing that secondary doping is conducive to improving intrinsic conductivity. Electrochemical impedance spectroscopy (EIS) was performed to further estimate the electrochemical behaviors of catalysts. The semicircular region of the Nyquist diagram corresponds to the charge transfer resistance (R_ct) between the electrolyte and catalyst. As shown in Fig. 3d, FeN₅–Fe₁/NC exhibit the smallest semicircular portion, indicating that the secondary doping may enhance the electron transfer ability of FeN₅–Fe₁/NC. Besides, the linear sweep voltammetry (LSV) curves display that FeN₅–Fe₁/NC, and FeN₄–Fe₁/NC-modified electrodes with wide potential windows can ensure that the response signals of DA are not interfered with (Fig. 3e). As shown in Fig. S5, the redox peak currents increase with the increment of scan rates. The linear relationship between scan rates and anodic peak currents is obtained, indicating that DA oxidation is a diffusion-controlled process. In addition, the detection sensitivity of FeN₅–Fe₁/NC and FeN₄–Fe₁/NC toward DA was examined by using amperometric i-t curves at a constant potential of 0.25 V upon successive addition of DA. As shown in Fig. 3f, the response time of FeN₄–Fe₁/NC (10.1 s) is 6.3 times longer than that of FeN₅–Fe₁/NC (1.6 s) toward DA, indicating a faster DA diffusion process of FeN₅–Fe₁/NC.

Fig. 3 (a) Scheme of FeN₅–Fe₁/NC for electrochemical oxidation of DA. (b) DPV responses of FeN₅–Fe₁/NC, FeN₄–Fe₁/NC, and Fe₁/NC with DA in 0.1 M PBS. (c) CV and (d) EIS curves of GCE, Fe₁/NC, FeN₄–Fe₁/NC and FeN₅–Fe₁/NC in [Fe(CN)₆]^3−/4−. (e) LSV curves of GCE, FeN₄–Fe₁/NC, and FeN₅–Fe₁/NC. (f) Amperometric i–t responses of FeN₅–Fe₁/NC and FeN₄–Fe₁/NC for DA.

Based on the superior catalytic activity of FeN₅–Fe₁/NC, electrochemical sensing was established to detect DA via DPV in 0.1 M PBS. Fig. 4a and b show the linear response curve of the current with DA concentration ranging from 0.05 to 500 μM with a limit of detection (LOD) of 23 nM. The i–t responses of the FeN₅–Fe₁/NC/GCE sensor toward DA were recorded in Fig. S6. DA was successively added to a constantly stirred solution of 0.1 M PBS (pH 7.4) to obtain the response current. The electrochemical sensor could respond quickly to DA in the range of 0.05–1000 μM. The above results confirm that the fabricated FeN₅–Fe₁/NC sensor has excellent performance for DA detection. In addition, the selectivity and stability of the electrochemical FeN₅–Fe₁/NC sensors were assessed as indicators of practical application feasibility. As shown in Fig. 4c, the FeN₅–Fe₁/NC-based electrochemical sensor has a negligible current response to interferents (including Na⁺, Zn²⁺, NO₂⁻, Cl⁻, glucose (Glu), cysteine (Cys), glutathione (GSH), sucrose (Suc) and ascorbic acid (AA)), demonstrating its excellent selectivity for DA detection. In addition, the deviation of current responses for DA caused by other interferents is within ±5% (Fig. S7), indicating its excellent anti-interference ability for detecting DA. Furthermore, the FeN₅–Fe₁/NC-based electrochemical sensor maintains 90% of its initial activity after storage for 18 days, indicating good stability of this sensing (Fig. 4d). Compared with the other reported electrochemical sensing for DA detection, FeN₅–Fe₁/NC-based electrochemical sensing exhibits excellent electrochemical activity (Table S4). To further explore the feasibility of the as-proposed electrochemical sensor in practical application, the standard addition method was applied to determine DA in actual serum samples. The recoveries of DA range from 97.7–102.8% (Table S5), indicating that FeN₅–Fe₁/NC-based electrochemical sensors have good feasibility for the detection of DA in real biological samples. DA and uric acid (UA) typically coexist in physiological fluids and have similar structures. Their oxidation signals often overlap significantly due to the similarity in redox potentials on the bare electrode. Thus, the feasibility of FeN₅–Fe₁/NC-based electrochemical sensing to distinguish between DA and UA was explored. Fig. S8 displays the oxidation responses of DA and UA. The DPV response of FeN₅–Fe₁/NC towards DA and UA was higher than that of FeN₄–Fe₁/NC, revealing that FeN₅–Fe₁/NC have better catalytic performance than FeN₄–Fe₁/NC. On this basis, the target analytes (DA, UA, and AA) were detected to assess the differentiated performance of the proposed electrochemical sensors. As shown in Fig. 4e, when UA and AA were added in solution with 500 μM DA, the oxidation current of DA remained constant, indicating that FeN₅–Fe₁/NC can effectively detect DA in the presence of the other analytes. The training matrix was obtained according to the experimental data (Table S6), which was subsequently transformed into a heat map. The responses of different colors correspond to three target analytes (Fig. S9), illustrating that DA, UA, and AA are well-distinguished. Afterward, linear discriminant analysis (LDA) was performed by the three most significant discriminant factors obtained from the data of DA, UA, and AA detected by the electrochemical sensor. The results were presented by further converting LDA into a visualized 3D standardized score map, showing that the three can be well distinguished (Fig. S10). Additionally, a heat map illustrating the response of the electrochemical sensor to various concentrations of UA was plotted to verify its quantitative capability. With the increase in UA concentration, the response was significantly enhanced and recorded in the heat map (Fig. 4f and Table S7). To sum up, the fabricated FeN₅–Fe₁/NC-based electrochemical sensing exhibits favorable discrimination performance.

Fig. 4 (a) DPV responses of FeN₅–Fe₁/NC for DA with different concentrations (0.05, 2, 5, 25, 50, 100, 250 and 500 μM). (b) Calibration curve of FeN₅–Fe₁/NC for DA detection. (c) Selectivity and (d) stability of the FeN₅–Fe₁/NC-based electrochemical sensing for DA detection. (e) DPV responses of FeN₅–Fe₁/NC by adding 250 μM DA (pink curve), 500 μM UA (blue curve), and 500 μM AA (green curve), with (red curve) a mixture of 250 μM DA, 500 μM UA, and in 500 μM AA in 0.1 M PBS (pH = 7.4). (f) Heat map of FeN₅–Fe₁/NC- and FeN₄–Fe₁/NC-based sensors for DA.

To further understand the underlying mechanism of DA oxidation by FeN_x at the atomic level, density functional theory (DFT) calculations were proposed. The theoretical models of FeN₅ and FeN₄ were well-constructed. Fig. 5a and Fig. S11 illustrate the reaction pathways of DA oxidation in FeN₅ and FeN₄, respectively. Initially, the OH is absorbed by the Fe site in FeN_x to form *OH/FeN_x. Then, the DA molecule binds to *OH/FeN_x to generate *H₂O/FeN_x + DA_H by an O–H bond. After that, *H₂O/FeN_x + DA_H is transformed to *H₂O/FeN_x. Eventually, the active sites return to the initial state after the desorption of H₂O molecule. The Gibbs free energy diagrams of the DA oxidation process are shown in the Fig. 5b. The OH adsorption on FeN₅ exhibits a stronger affinity for DA adsorption (0.48 eV) than that of FeN₄ models (0.42 eV), despite both being an exothermic process. Especially, *OH/FeN₅ displays a much lower energy barrier (0.08 eV) than *OH/FeN₄ (0.61 eV) for the O–H broken owing to its highly active reactive intermediates during the endothermic process. What's more, the rate-determining step of FeN₄ models is *H₂O desorption, requiring a large energy input of 0.61 eV owing to the strong binding strength of the *H₂O species on FeN₄. As shown in Fig. 5c, the projected density of states (PDOS) exhibits that the d-band center of the Fe atom combined with the H₂O molecule in FeN₅ and FeN₄ models are located at −2.73 eV and −0.34 eV, respectively. The lower d-band center (farther away from the Fermi energy) suggests the weakened bonding interaction between active sites and *H₂O desorption due to the additional axial ligand on Fe active centers, consistent with the free energy calculation results. In addition, the charge density difference of *H₂O desorption for different models was investigated since *H₂O desorption is the main influencing factor for the DA oxidation process. As shown in Fig. 5d, the bonding charge distributions between the *H₂O and Fe atoms are significantly different. There is much charge accumulation and depletion in FeN₄, indicating the strong binding strength of *H₂O. The above results exhibit that benefiting from the suitable interaction between active sites and intermediates, FeN₅ models deliver prominent catalytic activity towards the DA oxidation process.

Fig. 5 (a) DA oxidation procedures on FeN₅. (b) Gibbs free energy diagram of DA oxidation on FeN₅ and FeN₄ models. (c) PDOS of d-orbitals of Fe atoms in FeN₅ and FeN₄ models. (d) The charge density differences of FeN₅ and FeN₄ during *H₂O desorption, where the positive and negative charges are shown in yellow and cyan, respectively.

Conclusions

In summary, we developed an electrochemical sensor based on FeN₅–Fe₁/NC with atomically dispersed FeN₅ sites to sensitively detect DA. Benefiting from improving metal loading and intrinsic activity, FeN₅–Fe₁/NC display higher catalytic properties than Fe₁/NC. DFT calculations further reveal that the additional axial ligand on active sites endows catalytic centers with the suitable interaction with intermediates and thus reduces the reaction energy barrier during the catalytic DA process. The electrochemical FeN₅–Fe₁/NC sensors exhibit a good linear relationship range from 0.05 to 500 μM with a LOD of 23 nM. In addition, the FeN₅–Fe₁/NC-based electrochemical sensors exhibit good selectivity, stability, and promising detection in serum samples for DA detection. Expectantly, the reasonable design of single-atom materials opens a new avenue for constructing high-performance electrochemical sensors and further widens the application scope of electrochemical SAC-based sensing for complex biological systems.

Experimental section

Synthesis of catalysts

Briefly, 2 g SiO₂, as the template, was dissolved in 20 mL of deionized water. Then, a mixture of D-glucosamine hydrochloride (DGH), ZnCl₂, and FeCl₃ was added to deionized water with stirring for 30 min. After that, the resulting mixture was freeze-dried and carbonized at 900 °C for 2 h under an Ar atmosphere. Next, Fe₁/NC were obtained by etching the remaining powder with 10 wt% HF for 12 h to remove the template. The as-prepared Fe₁/NC (400 mg) were then dispersed into the 40 mL of water and isopropanol with volume ratios (1 : 1) containing FeCl₃ and dicyandiamide with mass ratios (1 : 15). After ultrasonication (2 h) and stirred (5 h), the precursor powder was washed and collected with deionized water and ethanol several times, and were dried at 60 °C. FeN₅–Fe₁/NC were obtained by pyrolysis at 900 °C for 1 h with a heating rate of 3 °C min⁻¹ under the Ar atmosphere. For comparison, FeN₄–Fe₁/NC were synthesized through similar procedures with the addition of only Fe source in the second step.

Preparation of the electrochemical sensors

The electrochemical measurements were performed with a three-electrode system by using a CHI-760 electrochemical analyzer. In a three-electrode system, the catalyst-modified glassy carbon electrode (GCE) was used as the working electrode. Then, the saturated calomel electrode and platinum wire electrode served as the reference electrode and counter electrode, respectively. Before conducting electrochemical assessments, the bare GCE was polished with alumina powder and then ultrasonic washed with ethanol and ultrapure water three times to obtain a mirror-like surface. As for the preparation of ink, 1 mg of catalysts was dispersed in 10 μL of Nafion (5 wt%), 490 μL of ethanol, and 500 μL of deionized water by ultrasonic treatment for 30 min to obtain a homogeneous catalytic ink. Then, the catalytic ink was added to the GCE and dried at 50 °C to prepare the electrochemical sensors. The loading amount of catalysts on the GCE was 0.071 mg cm⁻².

Conflicts of interest

There are no conflicts to declare.

Data availability

The data supporting the findings of this study are available within the article and its SI files. See DOI: https://doi.org/10.1039/d5nh00476d

Additional datasets generated and analyzed during the current study are available from the corresponding author upon reasonable request.

Acknowledgements

The authors gratefully acknowledge the financial support of the National Natural Science Foundation of China (no. 22204045), the Open Project Funding of the Key Laboratory of Fermentation Engineering (Ministry of Education), Hubei University of Technology (no. 202409FE14). We thank Dr Xiaoli Cai from School of Public Health, Wuhan University of Science and Technology, for assistance with real sample measurements. All experiments were performed in accordance with the guidelines of Declaration of Helsinki, and experiments were approved by the ethics committees at Wuhan University of Science and Technology (Wuhan, China). Informed consents were obtained from human participants of this study.

References

1. S. J. Lee, B. Lodder, Y. Chen, T. Patriarchi, L. Tian and B. L. Sabatini, Nature, 2021, 590, 451.
2. R. P. Shukla, M. Aroosh, A. Matzafi and H. Ben-Yoav, Adv. Funct. Mater., 2021, 31, 2004146.
3. N. Wang, H. Ao, W. Xiao, W. Chen, G. Li, J. Wu and H. Ju, Biosens. Bioelectron., 2022, 201, 113959.
4. Y. Xing, X. Chen and H. Zhao, Anal. Chem., 2024, 96, 6037.
5. Z. Wang, X. Wang, X. Zhu, J. Lv, J. Zhang, Q. Zhu and Z. Dai, Sens. Actuators, B, 2019, 285, 438.
6. Z. Shi, X. Wu, Z. Zou, L. Yu, F. Hu, Y. Li, C. Guo and C. M. Li, Biosens. Bioelectron., 2021, 186, 113303.
7. Z. Wang, H. Guo, R. Gui, H. Jin, J. Xia and F. Zhang, Sens. Actuators, B, 2018, 255, 2069.
8. Y.-M. Cai, Y.-H. Li, Y. Xiao, Q. Meyer, Q. Sun, W.-J. Lai, S.-W. Zhao, J. Li, L.-J. Zhang, H. Wang, Z. Lin, J. Luo and L.-L. Han, Rare Met., 2024, 43, 5792.
9. C. Zhang, X. You, Y. Li, Y. Zuo, W. Wang, D. Li, S. Huang, H. Hu, F. Yuan, F. Shao and M. Yin, Sens. Actuators, B, 2022, 354, 131233.
10. D. Butler, D. Moore, N. R. Glavin, J. A. Robinson and A. Ebrahimi, ACS Appl. Mater. Interfaces, 2021, 13, 11185.
11. A. Mikhraliieva, A. R. S. Lima, C. L. Jost, M. Nazarkovsky, Y. Xing and V. Zaitsev, Small, 2024, 20, 2400650.
12. Y. Zhang, H. Wang, L. Jiao, N. Wu, W. Xu, Z. Wu, Y. Wu, P. Hu, W. Gu and C. Zhu, Chem. Eng. J., 2023, 466, 143075.
13. R. Jin, C. Zeng, M. Zhou and Y. Chen, Chem. Rev., 2016, 116, 10346.
14. I. Chakraborty and T. Pradeep, Chem. Rev., 2017, 117, 8208.
15. R. Eivazzadeh-Keihan, P. Pashazadeh-Panahi, B. Baradaran, A. Maleki, M. Hejazi, A. Mokhtarzadeh and M. de la Guardia, TrAC, Trends Anal. Chem., 2018, 100, 103.
16. M. Shen, Q. Liu, J. Sun, C. Liang, C. Xiong, C. Hou, J. Huang, L. Cao, Y. Feng and Z. Shang, J. Colloid Interface Sci., 2024, 673, 453.
17. C. Chen, H. Li, J. Chen, D. Li, W. Chen, J. Dong, M. Sun and Y. Li, Energy Environ. Mater., 2023, 6, e12346.
18. Y. Wang, B. Hui, Y. Zhang, K. Liu, J. Wang, D. Zhang, J. Liang, J. Li, K. Zhang and S. Yang, Chem. Eng. J., 2025, 520, 166355.
19. J. Sun, M. Shen, A. J. Chang, C. Liang, C. Xiong, C. Hou, J. Li, P. Wang, J. Li and J. Huang, Chem. Eng. J., 2025, 503, 158551.
20. L. Bi, X. Liang, W. Zhang, Z. Wu, J. Zhan, T. Xie, L. Zhang, W. Xu, J. Jiang and M. Wu, Colloids Surf., A, 2024, 683, 133103.
21. W. Xu, N. Apodaca, H. Wang, L. Yan, G. Chen, M. Zhou, D. Ding, P. Choudhury and H. Luo, ACS Catal., 2019, 9, 5074.
22. X. Luo, J. Zhao, M. Li, X. Zhao, X. Wei, Z. Luo, W. Gu, D. Du, Y. Lin and C. Zhu, Mater. Today, 2023, 64, 121.
23. Z. Qin, B. Tang, G. Zhang, C. Zhu, K. Jiang, B. Zhang and F.-Z. Xuan, Nano Res., 2024, 17, 7658.
24. R. Zhao, T. Wang, J. Li, Y. Shi, M. Hou, Y. Yang, Z. Zhang and S. Lei, Nano Res., 2023, 16, 8570.
25. M. Zhou, Y. Jiang, G. Wang, W. Wu, W. Chen, P. Yu, Y. Lin, J. Mao and L. Mao, Nat. Commun., 2020, 11, 3188.
26. X. Xie, D. P. Wang, C. Guo, Y. Liu, Q. Rao, F. Lou, Q. Li, Y. Dong, Q. Li, H. B. Yang and F. X. Hu, Anal. Chem., 2021, 93, 4916.
27. M. Yan, X. Zhu, C. Xiong, X. Han, Z. Xue and Y. Wu, Nano Res., 2024, 17, 7256.
28. R. Li, W. Guo, Z. Zhu, Y. Zhai, G. Wang, Z. Liu, L. Jiao, C. Zhu and X. Lu, Anal. Chem., 2023, 95, 7195.
29. C. Zhou, J. Shi, Z. Dong, L. Zeng, Y. Chen, Y. Han, L. Li, W. Zhang, Q. Zhang, L. Gu, F. Lv, M. Luo and S. Guo, Nat. Commun., 2024, 15, 6741.
30. W. Xu, Y. Wu, W. Gu, D. Du, Y. Lin and C. Zhu, Chem. Soc. Rev., 2024, 53, 137.
31. L. Chen, X. Wang, T. Lu, H. Pang, S. Zhang, L. Xu, G. Yang, Q. Zhou and Y. Tang, Appl. Catal., B, 2025, 366, 125007.
32. W. Xu, Y. Wu, X. Wang, Y. Qin, H. Wang, Z. Luo, J. Wen, L. Hu, W. Gu and C. Zhu, Angew. Chem., Int. Ed., 2023, 62, e202304625.
33. H. Zhang, J. Li, S. Xi, Y. Du, X. Hai, J. Wang, H. Xu, G. Wu, J. Zhang, J. Lu and J. Wang, Angew. Chem., Int. Ed., 2019, 58, 14871.
34. M. Wang, L. Wang, Q. Li, D. Wang, L. Yang, Y. Han, Y. Ren, G. Tian, X. Zheng, M. Ji, C. Zhu, L. Peng and G. I. N. Waterhouse, Small, 2023, 19, e2300373.
35. B. Zhou, F. Guo, X. Liu, H. Zhang, J. Cheng, H. Zhang, Z. Chen, L. Tan, K. Zeng, F. Jia, L. Zhang and Z. Ai, Appl. Catal., B, 2025, 371, 125230.
36. S. Ding, J. A. Barr, Q. Shi, Y. Zeng, P. Tieu, Z. Lyu, L. Fang, T. Li, X. Pan, S. P. Beckman, D. Du, H. Lin, J.-C. Li, G. Wu and Y. Lin, ACS Nano, 2022, 16, 15165.
37. Y. Han, Y. Wang, R. Xu, W. Chen, L. Zheng, A. Han, Y. Zhu, J. Zhang, H. Zhang, J. Luo, C. Chen, Q. Peng, D. Wang and Y. Li, Energy Environ. Sci., 2018, 11, 2348.
38. X. Li, L. Jiao, R. Li, X. Jia, C. Chen, L. Hu, D. Yan, Y. Zhai and X. Lu, Small, 2024, 20, e2405532.
39. Y. Zhao, H.-C. Chen, X. Ma, J. Li, Q. Yuan, P. Zhang, M. Wang, J. Li, M. Li, S. Wang, H. Guo, R. Hu, K.-H. Tu, W. Zhu, X. Li, X. Yang and Y. Pan, Adv. Mater., 2024, 36, 2308243.
40. X. Yang, L. Xu and Y. Li, Coord. Chem. Rev., 2024, 516, 215961.
41. X. Tang, Y. Zhang, S. Tang, D. Lützenkirchen-Hecht, K. Yuan and Y. Chen, ACS Catal., 2024, 14, 13065.
42. H. Song, D. H. Jung, Y. Cho, H. H. Cho, V. G. Panferov, J. Liu, J. H. Heo and J. H. Lee, Coord. Chem. Rev., 2025, 541, 216835.
43. X. Luo, S. Li, Y. Wu, F. Tan, C. Li, W. Gu, J. Liu and C. Zhu, TrAC-Trend Anal. Chem., 2025, 185, 118169.
44. X. Luo, X. Wei, H. Wang, W. Gu, T. Kaneko, Y. Yoshida, X. Zhao and C. Zhu, Nano-Micro Lett., 2020, 12, 163.
45. L. Lin, Q. Zhu and A.-W. Xu, J. Am. Chem. Soc., 2014, 136, 11027.

---