Fingertip-chip sensor based on Pd nanocluster sensitized 3D NiO nanotube arrays for real-time, selective methane detection

Kunmei Yang† ^b, Yue Kang†^a, Jia Yan*^a, Weihao Fang^c, Jiazhen Zhang^b and Zhilong Song*^a
aInstitute for Energy Research, Jiangsu University, Zhenjiang, Jiangsu 212013, China. E-mail: yanj@ujs.edu.cn; songzl@ujs.edu.cn
^bSchool of the Environment and Safety Engineering, Jiangsu University, Zhenjiang, Jiangsu 212013, China
^cResearch Center of Fluid Machinery Engineering and Technology, Jiangsu University, Zhenjiang, Jiangsu 212013, China

First published on 7th August 2025

---

---

Introduction

Methane (CH₄) is a significant greenhouse gas^1,2 and a potential safety hazard in mining environments,^3,4 where it poses a serious risk of explosive incidents.^5,6 Typical underground CH₄ concentrations in coal mines are often in the range of 500 to 10000 ppm, which corresponds to a volume fraction of 0.05 to 1 per cent. According to China's coal mine safety regulations, the alarm threshold for CH₄ sensors in high-risk areas, such as mining faces, is 0.5 per cent by volume, or 5000 ppm.^7–9 In contrast, natural gas pipeline leaks have an initial concentration of less than 100 ppm, making the detection of such small leaks critical to operational safety and environmental protection. Despite OSHA allows an exposure limit of 5000 ppm, early warning requires sensors capable of detecting sub-ppm concentrations.^7,10

The presence of interfering gases in coal mines significantly impacts the technical complexity of selective methane detection. This is particularly evident with trace gases (Table S1) such as hydrogen sulfide (H₂S), which has a safety limit of only 6.6 ppm, and carbon monoxide (CO), which has a safety limit of 24 ppm.^11,12 Optical sensing technology, based on the infrared characteristic absorption of CH₄, has shown high selectivity and stability, achieving ppb-level detection limits.^13–15 However, issues such as large size, high cost, and significant power consumption often hinder its practical application. Electrochemical sensors, while offering low power consumption and miniaturization, face challenges related to stability and selectivity.^16,17 On the other hand, metal oxide-based gas sensors,^18–22 such as those utilizing nickel oxide (NiO), are promising candidates due to their high surface area, cost-effectiveness, and scalability.^23–26 However, these sensors still face significant challenges in terms of selectivity and sensitivity, especially when detecting trace gases like CH₄. Recent efforts have focused on improving sensor performance by incorporating noble metal catalysts, such as palladium (Pd), platinum (Pt), and gold (Au), known for enhancing gas adsorption and improving selectivity.^27–30 Zhang et al. prepared PdO/Au/NiO porous nanosheet composites, in which Au dissociates oxygen from air through the “spillover effect” to generate more reactive oxygen species, which preferentially oxidize CH₄ and inhibit the adsorption of other gases.³¹ Similarly, Chen et al. improved sensing performance by adding a Pt layer to NiO, which improved charge transfer and reduced interference, thereby increasing selectivity for ammonia.³² Despite these advancements, the integration of noble metals onto NiO surfaces presents challenges. Despite these advancements, the integration of noble metals onto NiO surfaces presents challenges. One major issue is the complexity of fabrication methods. Techniques such as physical vapor deposition (PVD), chemical vapor deposition (CVD), electrochemical deposition, and the sol–gel method have been employed to deposit metal clusters onto metal oxides.^33–35 However, these methods often struggle with controlling the uniformity of metal deposition, leading to inconsistent device performance. Additionally, metal-modified NiO sensors typically require high operating temperatures to achieve optimal performance, limiting their suitability for real-time, low-energy applications.^36–39

To address these limitations, recent advancements in nanomaterial fabrication, particularly atomic layer deposition (ALD), to address these limitations, recent advancements in nanomaterial fabrication, particularly atomic layer deposition (ALD), have allowed for precise control over the deposition of metal clusters at the atomic level. For instance, ALD has been successfully used to deposit Pd clusters on TiO₂ nanostructures, significantly enhancing the performance of gas sensors by improving the metal's dispersion and its interaction with the semiconductor surface.⁴⁰ Similarly, ALD-based deposition of Rh clusters on SnO₂ nanotubes has shown improvements in both sensitivity and selectivity for formaldehyde detection. The atomic-level control provided by ALD enables a more uniform and stable interaction between the metal and the semiconductor, leading to enhanced sensor performance.⁴¹

In this study, we present a novel methane (CH₄) sensor based on a three-dimensional (3D) NiO nanotube array, sensitized by Pd nanoclusters using ALD technology. The 3D NiO structure offers a large surface area, which is crucial for enhancing gas adsorption, while the Pd nanoclusters serve as catalytic sites, promoting the interaction between CH₄ molecules and the NiO surface. We conducted a comprehensive series of tests to evaluate the sensor's sensitivity, selectivity, and long-term stability. Furthermore, to improve the practical utility of the sensor, we integrated it into a wireless monitoring system featuring fingertip-chip sensors, enabling real-time CH₄ monitoring in dynamic environments, where fast and accurate detection is critical for safety. This work presents a promising approach for the development of highly sensitive and selective sensors suitable for a wide range of industrial and environmental applications.

Materials and methods

Sensor film deposition

The Pd/NiO sensors were fabricated using a two-step atomic layer deposition (ALD) method. First, NiO was deposited onto a three-dimensional (3D) anodic aluminum oxide (AAO) substrate at 250 °C. The Ni precursor was held in a bubbler at 130 °C, and ozone (O₃) was introduced into the chamber, with high-purity nitrogen (N₂, 99.9999%) serving as the carrier and purge gas. The Ni ALD pulse sequence (1 cycle) included Ni precursor injection (2000 ms), followed by a purge (30 s) and wait time (5 s), then O₃ injection (2000 ms), followed by a purge (30 s) and wait time (5 s). The thickness of the NiO layer was controlled using the ALD recipe for 150 cycles and then annealed at 500 °C for 12 hours. For the palladium (Pd) overlayer decoration, the pulse sequence (1 cycle) involved Pd precursor injection (1000 ms), followed by a purge (30 s) and wait time (5 s), then hydrazine (N₂H₄) injection (200 ms), followed by a purge (30 s) and wait time (5 s). The Pd thickness was controlled using the ALD recipe for 5, 10, and 20 cycles, respectively.

Characterization

The morphology and size of the thin films were characterized using a scanning electron microscope (SEM, JSM-7100F, JEOL) and high-resolution transmission electron microscope (HR-TEM, JEM 2010F, JEOL). SEM images were obtained at an accelerating voltage of 10 kV, and HR-TEM images were captured at 200 kV. X-ray diffraction (XRD) measurements (MAXima_X XRD-7000, Shimadzu, Japan) were performed on a Bruker D8 Advance X-ray diffractometer with Cu Kα radiation (λ = 1.5406 Å) in the 2θ range of 10–70°. The density functional theory (DFT) calculations were performed using the generalized gradient approximation (GGA) functional with a 3 × 3 supercell model to optimize the structures of NiO and noble metal (Pt, Pd, and Au) decorated NiO for CH₄ adsorption.

Gas-sensing measurement

The sensing measurements were conducted in a dynamic gas chamber (2 cm × 2 cm × 2 cm), the controlled amount of CH₄, H₂S, and CO were introduced using mass flow controllers (MFC1-3), respectively, and each gas was intermixed with air injected from MFC4 to control the gas concentration. The total flow rate was controlled at 500 sccm. The sensor resistance was recorded as a function of time at the fixed voltage of 3.3 V. Response definition: S = R_a/R_g − 1, with R_a representing the resistance in the presence of clean air and R_g denoting the resistance in the presence of the target gas.

Results and discussion

Design of potential Pd/NiO sensors

To optimize sensor performance, density functional theory (DFT) calculations were performed to investigate the interaction between methane (CH₄) and various noble metal-modified NiO surfaces (Pd, Pt, Au). The calculations utilized spin-polarized DFT within the generalized gradient approximation (GGA) with the Perdew–Burke–Ernzerhof (PBE) functional.⁴² The ionic cores were modeled using projected augmented wave (PAW) potentials, and a plane wave basis set with a kinetic energy cutoff of 450 eV was employed.⁴³ The Monkhorst–Pack k-point sampling was generated with a 3 × 3 × 1 grid. Convergence thresholds for the maximum force and maximum energy change were set to 0.02 eV Å⁻¹ and 1.0 × 10⁻⁵ eV per atom, respectively. A vacuum layer of 15 Å thickness was added along the z-direction to prevent interactions between different surfaces. The adsorption energy of CH₄ on NiO, Pt/NiO, Pd/NiO, and Au/NiO was calculated using the following equation:

Eads = Etotal – (Eslab + Egas)	(1)

where E_ads is the adsorption energy, E_total is the total energy for the adsorption state, E_slab is the energy of the clean surface of NiO and Pd/NiO, and E_gas is of the free CH₄. The slab model was constructed on the (100) facet of the optimized bulk NiO, and a 2 × 2 × 1 supercell was used for subsequent computations.

Unit cells of Pt₁₃, Pd₁₄ and Au₁₃ clusters were respectively set to be adsorbed on the NiO surface. This work is based on the concept that noble metal catalysts can enhance selectivity by accelerating specific reactions towards the target gas while suppressing interference from other gases.⁴⁴ Pt, Pd, and Au are chosen for NiO decoration, with optimized molecular structures for CH₄ adsorption shown in Fig. 1a. The activity for selectivity of the sensors was evaluated by measuring the C–H bond length of CH₄ adsorbed on the surface of the catalysts. As illustrated in Fig. 1b, the Pd/NiO structure exhibited the largest C–H bond length change (Δλ = 0.021 Å) compared to NiO (Δλ = 0.002 Å), Pt/NiO (Δλ = 0.011 Å), and Au/NiO (Δλ = 0.002 Å). The results indicate that Pd/NiO provides the most favorable adsorption for CH₄, suggesting that Pd is the optimal for CH₄ detection.

Sensor preparation and structural characterization

The Pd/NiO sensor was prepared using a two-step ALD method, with 3D nanoporous single-sided open-pore AAO serving as the substrate material (Fig. 2a). The 3D AAO structure facilitates directional gas molecule diffusion through a network of regularly arranged nanopores. Additionally, its high specific surface area, reaching up to several hundred m² g⁻¹, provides abundant active sites for the gas adsorption–desorption process. The Pd/NiO sensor consists of a thin layer of NiO deposited on the walls of the AAO, followed by the deposition of metallic Pd nanoclusters. The ALD process for sensor fabrication includes four steps (Fig. 2b): precursor injection, blowdown, reactant injection, and blowdown. During NiO deposition, the precursor nickel(II) 1-dimethylamino-2-methyl-2-butoxide (Ni(dmamb)₂) was first introduced into the reaction chamber, where it adsorbed onto the AAO surface. The chamber was then purged with inert nitrogen (N₂) to remove any unadsorbed precursor. Ozone (O₃) was introduced into the chamber, reacting with the adsorbed Ni precursor to form NiO. Finally, the chamber was purged again with N₂ to remove unreacted oxidizers and by-products. The sequence of surface chemical reactions during NiO deposition is outlined as follows:

(*OH) + Ni(dmamb)2 → (*O)Ni(dmamb) + H(dmamb)	(2)

(*O)Ni(dmamb) + O3 → NiO(s) + O2 + CO2 + H2O + N2	(3)

Subsequently, Pd nanoclusters were deposited onto the NiO surface. The Pd ALD process was identical to the NiO ALD process, using palladium(II) hexafluoroacetylacetonate (Pd(hfac)₂) and N₂H₄ as reaction precursors. The following surface chemistry reactions occurred during Pd deposition:

(*OH) + Pd(hfac)2 → (*O)Pd(hfac) + H(hfac)	(4)

(*O)Pd(hfac) + N2H4 → Pd(s) + HF + CO2 + H2O + N2	(5)

By precisely controlling the number of ALD cycles, the thickness of the NiO and Pd layers can be accurately calibrated.

Fig. 3a illustrates the SEM image of a Pd/NiO sensing material supported on an AAO substrate. In Fig. 3b, the enlarged cross-sectional view reveals a uniform Pd/NiO film anchored to the AAO wall, with a mean thickness of 15 nm. Fig. 3c–e display enlarged SEM images of the top, middle, and bottom regions, which are clearly distinguishable, demonstrating the uniformity of the material throughout the film. The HRTEM image (Fig. 3g) exhibits lattice distances of 0.241 nm and 0.209 nm, corresponding to the (110) and (012) planes of NiO, respectively. As demonstrated in Fig. 3h, XRD was utilized thoroughly investigate the structural characteristics of the Pd/NiO sensor. The XRD peaks of the Pd/NiO sample were fully indexed to the structure of nickel oxide (JCPDS #44-1159), confirming the presence of NiO. In contrast, the XRD spectrum of the Pd film (green sample) corresponds to the face-centered cubic (fcc) structure of palladium (JCPDS #46-1043). Finally, Fig. 3i presents an elemental mapping analysis, demonstrating the uniform distribution of oxygen (O), nickel (Ni), and palladium (Pd) on the substrate. This confirms the successful design and fabrication of the Pd/NiO sensor on the AAO substrate.

Gas-sensing performance

To thoroughly investigate the performance of the Pd/NiO sensor, systematic measurements of its response characteristics were conducted across different operating temperatures. The sensor was tested with methane (CH₄) concentrations ranging from 20 ppm to 10000 ppm to identify the optimal operating temperature. As shown in Fig. 4a, the response of the Pd/NiO sensor to CH₄ varied with temperature, with the best performance observed at an optimized operating temperature of 200 °C. Fig. 4b presents the results of four repetitive tests for sensors with Pd deposition cycles of 5, 10, and 20, respectively. After four test cycles, the change in the magnitude of the sensor's response to CH₄ remained consistent across cycles, and the sensor returned to its initial state after each measurement. This demonstrates the excellent repeatability of the Pd/NiO sensors. Additionally, the Pd/NiO sensor with 10 Pd deposition cycles (Pd(10)/NiO) showed the best sensing response (Fig. 4b and c), and was selected for further gas-sensing property evaluation. The theoretical detection limit (LOD) is defined as three times the relative standard deviation (RSD) (σ) of the sensor noise divided by the slope of the linear fit:^45,46

LOD(ppb) = 3 × σ/slope	(6)

For the detailed calculation procedures, refer to Fig. S1 in the SI. The LOD was calculated to be as low as 70 ppb for CH₄, which is ideal for detecting methane in industrial and environmental applications. Based on these results, future experiments with the Pd(10)/NiO sensor will be conducted at an operating temperature of 200 °C.

In Fig. 4d, the sensor's response to CH₄ at concentrations of 500 ppm and 10000 ppm was evaluated over a humidity range of 10–85%. The response showed only small variations, indicating that the Pd(10)/NiO sensor demonstrates excellent resistance to humidity. Fig. 4e and f show that the Pd(10)/NiO sensor's response to 500 ppm of CH₄ remained stable over a period of 2400 min and through multiple cycles, highlighting the sensor's good long-term stability. A slight dip in the response during the first 400 min, observed in Fig. 4e, is likely due to the sensor stabilization process. However, the change in response was minimal.

The selectivity of the Pd/NiO sensor for detecting CH₄ was quantified by comparing the response to CH₄ (S_CH4) with the response to interfering gases such as CO, H₂S, SO₂, and ethanol (Fig. S2). A ratio S_CH4/S_x, where x represents the interfering gas, was used to assess selectivity. A ratio S_CH4/S_x > 1 indicates selective detection of CH₄ over the interfering gases, with a higher ratio signifying better selectivity. The Pd/NiO sensor exhibited optimal selectivity for CH₄ at 200 °C, with response ratios of 14.6 (S_CH4/S_CO), 12.1 (S_CH4/S_H2S), 15.4 (S_CH4/S_SO2) and 11.4 (S_CH4/S_ethanol), respectively. To further assess selectivity, the sensor's response to CH₄ mixed with interfering gases (H₂S and CO) was tested. As demonstrated in Fig. 5a and c, the sensing response was able to recover to a level close to the baseline after each change in the concentration of the interfering gases, indicating the sensor's recovery capability. Fig. 5b and d demonstrate that the response value stabilized with minimal change in the concentration of interfering gases, suggesting the sensor's strong anti-interference performance. Additionally, the Pd/NiO sensor exhibits very competitive CH₄ sensing properties compared to typical metal oxide gas sensors (see Table S2 in the SI).

Sensing mechanism

The enhanced methane sensing performance of the Pd/NiO sensor can be attributed to the catalytic effect of Pd nanoclusters, which reduce the activation energy required for methane adsorption and reaction on the NiO surface. This interaction significantly improves the sensor's sensitivity and selectivity. To further understand the underlying mechanism, DFT calculations were conducted to investigate the interaction between the sensing materials and the gases. Fig. 6a and b present the adsorption configurations and adsorption energies (E_ads) of CH₄ on NiO and Pd/NiO, respectively. Fig. 6c shows the differential charge density and Bader charges (Δq). The calculated differential charge density results indicate that the Pd/NiO surface exhibits a larger Bader charge of 0.18 e. These results suggest that CH₄ is strongly adsorbed on the Pd/NiO surface, which aligns with the observed sensitive and selective CH₄ sensing behavior.

Real-time application for CH₄ detection

A portable wireless CH₄ detection system was developed based on a Pd/NiO sensor, designed to demonstrate its practical application in real-time environmental monitoring. The system allows seamless wireless communication with a smartphone, providing users with real-time CH₄ concentration updates. This capability makes it particularly useful for applications such as mine safety monitoring and natural gas leak detection. The complete sensor system is integrated into a compact, portable wireless device, constructed on a printed circuit board (PCB) (Fig. 7a), with the detailed circuit design shown in Fig. 7b. While the optimized operating temperature for the Pd/NiO sensor is 200 °C, which ensures optimal performance, we have taken care to design the device to maintain its portability. The device's compact size (5 × 5 mm) and the localized heating area (0.5 × 0.5 mm) of the sensor minimize power consumption and thermal load. This small, focused heating region ensures energy efficiency, keeping the rest of the device at a lower temperature and preserving portability without compromising sensor performance. In addition, the sensor system incorporates several key components: an 800 mAh Li-ion battery for power supply, a fingertip-sized sensor module, a microcontroller (MCU) equipped with analog readout circuitry, and a wireless data transmission module (MCU & Wi-Fi unit). Noting that the fingertip-sized sensor module was anchored on the back side of the PCB, as shown in Fig. 7b. Upon detection of CH₄ gas, the resistance of the sensor's sensitive resistance undergoes a significant change. This variation is accurately captured by the built-in analog-to-digital converter (ADC) of the STM32 microcontroller. The MCU processes the data and computes the CH₄ concentration based on the resistance change. Subsequently, the CH₄ concentration data is transmitted to a cloud platform via the Wi-Fi module. The smartphone application, connected to the cloud platform, receives real-time updates on CH₄ concentration levels, enabling users to monitor gas levels in their environment at any time. If the CH₄ concentration exceeds a user-defined safety threshold set within the app, the system automatically activates an alarm via the device's buzzer, alerting the user to take necessary precautions (Video S1, showing the sensor's rapid response at three validation concentrations (val. conc.) 100, 500, and 1000 ppm concentrations with real-time display). The results demonstrate that the developed sensor is highly effective for applications such as mine safety monitoring and natural gas leak detection.

Conclusions

This study presents a Pd nanocluster-sensitized 3D NiO nanotube array sensor that demonstrates exceptional sensitivity and selectivity for CH₄ detection. The sensor exhibits a low detection limit, rapid response time, and superior selectivity in the presence of interferent gases such as H₂S and CO, making it particularly effective for CH₄ monitoring in both mining environments and natural gas leak detection applications. The design of the Pd/NiO sensor, which leverages the catalytic properties of Pd and the high surface area of NiO nanotubes, provides a robust and reliable solution for real-time CH₄ detection. Moreover, this sensor offers a low-cost and scalable approach, making it a promising candidate for widespread deployment in environmental monitoring and safety-critical application.

Author contributions

J. Y. and Z. S.: funding acquisition, supervision, conceptualization, writing – review and editing; K. Y. and Y. K.: conceptualization, methodology, investigation, validation, formal analysis, writing – original draft; W. F., J. Z.: conceptualization, methodology, investigation.

Conflicts of interest

All authors have given their approval for the final version of the manuscript and declare no competing financial interest.

Data availability

Supplementary information is available: The theoretical limit of detection (LOD) calculation (Fig. S1); selectivity against the interfering gases (Fig. S2); gas concentration range and sensor verification index (Table S1); CH₄ sensing performance comparison (Table S2). See DOI: https://doi.org/10.1039/D5LC00570A.

The subset of data may be included as SI within the manuscript. For further data access, please contact the corresponding author.

Acknowledgements

The research described in this paper was supported by the Senior Talent Fund of Jiangsu University (23JDG011 and 23JDG012). We thank the Analytical and Testing Center of Jiangsu University for the characterization support and shiyanjia Lab (http://www.shiyanjia.com) of DFT analysis and TEM characterization, and the MNT Micro and Nanotech Co., LTD at Wuxi, Jiangsu, China for the ALD technology support.

Notes and references

1. Z. Wang, T. Xia and X. Hu, ACS Appl. Nano Mater., 2024, 7, 20994–21004 CrossRef CAS .
2. D. S. Reay, P. Smith, T. R. Christensen, R. H. James and H. Clark, Annu. Rev. Environ. Resour., 2018, 43, 165–192 CrossRef .
3. W. Zhang, T. Yuan, X. Wang, Z. Cheng and J. Xu, Sens. Actuators, B, 2022, 354, 131004 CrossRef CAS .
4. T. Liu, Y. Wei, G. Song, B. Hu, L. Li, G. Jin, J. Wang, Y. Li, C. Song, Z. Shi, L. Zhao, J. Hu, W. Zhao, M. Hou, R. Li and J. Wang, Measurement, 2018, 124, 211–223 CrossRef .
5. R. Rivera-Martinez, P. Kumar, O. Laurent, G. Broquet, C. Caldow, F. Cropley, D. Santaren, A. Shah, C. Mallet, M. Ramonet, L. Rivier, C. Juery, O. Duclaux, C. Bouchet, E. Allegrini, H. Utard and P. Ciais, Atmos. Meas. Tech., 2024, 17, 4257–4290 CrossRef CAS .
6. K. Zhou, F. Li, H. Cai, Y. Jing, J. Zhuang, M. Li and Z. Xing, Energy Build., 2022, 254, 111645 CrossRef .
7. G. De Smedt, F. de Corte, R. Notelé and J. Berghmans, J. Hazard. Mater., 1999, 70, 105–113 CrossRef CAS PubMed .
8. H. F. Coward and G. W. Jones, J. Franklin Inst., 1927, 203, 161 CrossRef .
9. Y. Wang, X. Li, W. Wang and Z. Guo, Int. J. Hydrogen Energy, 2017, 42, 18728–18733 CrossRef CAS .
10. X. Sun, M. Tang, M. Yu, Y. Fan, C. Qin, J. Cao and Y. Wang, Mater. Today Commun., 2024, 40, 109551 CrossRef CAS .
11. W. Xie, H. Wang, M. Wang and Y. He, Sci. Rep., 2021, 11, 5712 CrossRef CAS PubMed .
12. X. Shi, Y. Zhang, X. Chen, Y. Zhang and T. Ma, Fuel Process. Technol., 2021, 213, 106671 CrossRef CAS .
13. Y. Zhang, X. Du, J. Zhai and X. Xie, Analysis Sensing, 2023, 3, 202300002 CrossRef .
14. K. Waich, T. Mayr and I. Klimant, Meas. Sci. Technol., 2007, 18, 3195 CrossRef CAS .
15. F. Fu, X. Zhang, W. Wang and X. Xie, ACS Sens., 2025, 10, 3757–3762 CrossRef CAS PubMed .
16. G. Maduraiveeran, M. Sasidharan and V. Ganesan, Biosens. Bioelectron., 2018, 103, 113–129 CrossRef CAS PubMed .
17. H. Wan, H. Yin, L. Lin, X. Zeng and A. J. Mason, Sens. Actuators, B, 2018, 255, 638–646 CrossRef CAS PubMed .
18. Z. Song and J. Yan, Chem. Mater., 2023, 35, 7750–7760 CrossRef CAS .
19. S. Zhang, J. Li, L. Han, B. Zhang, Y. Wang and Z. Zhang, Mater. Res. Bull., 2023, 165, 112332 CrossRef CAS .
20. R. G. Anjitha and P. K. Basu, IEEE Sens. J., 2025, 25, 159–166 Search PubMed .
21. W. Chen, L. X. Sui, X. Long, J. X. Liao, S. Z. Wang and X. B. Wei, Int. J. Hydrogen Energy, 2024, 50, 973–991 CrossRef CAS .
22. R. Binowesley, K. Savarimuthu, M. G. N. Alsath and M. Ramya, J. Mater. Sci.: Mater. Electron., 2024, 35, 2288 CrossRef CAS .
23. Y. Jiang, K. Hu, J. Zhang, Y. Hu, Y. Yang, X. Yin and P. Huang, Ceram. Int., 2024, 50, 36428–36439 CrossRef CAS .
24. N. P. Hung, N. V. Duy, C. T. Xuan, D. T. T. Le, C. M. Hung, H. Jin and N. D. Hoa, RSC Adv., 2024, 14, 12438–12448 RSC .
25. A. Stanoiu, A. C. Kuncser, D. Ghica, O. G. Florea, S. Somacescu and C. E. Simion, Chemosensors, 2021, 9, 125 CrossRef CAS .
26. J. Ma, W. Meng, Y. Li, L. Dai and L. Wang, Sens. Actuators, B, 2024, 401, 135058 CrossRef CAS .
27. Z. Song, W. Fang, B. Zhu and J. Yan, Anal. Methods, 2024, 16, 5954–5958 RSC .
28. R. G. Anjitha and P. K. Basu, IEEE Sens. J., 2023, 23, 143–153 CAS .
29. L. Han, S. Zhang, B. Zhang, B. Zhang, Y. Wang, H. Bala and Z. Zhang, J. Materiomics, 2022, 8, 545–555 CrossRef .
30. Y. Xia, J. Wang, L. Xu, X. Li and S. Huang, Sens. Actuators, B, 2020, 304, 127334 CrossRef CAS .
31. S. Zhang, M. Xing, J. Li, N. Luo, B. Zhang and Z. Zhang, Mater. Res. Bull., 2024, 174, 112706 CrossRef CAS .
32. H.-I. Chen, C.-Y. Hsiao, W.-C. Chen, C.-H. Chang, T.-C. Chou, I. P. Liu, K.-W. Lin and W.-C. Liu, Sens. Actuators, B, 2018, 256, 962–967 CrossRef CAS .
33. Z. Han, W. Ahmad, Y. Rong, X. Chen, S. Zhao, J. Yu, P. Zheng, C. Huang and H. Li, Foods, 2024, 13, 1721 CrossRef CAS PubMed .
34. H. Mao, X. Du, Y. Yan, X. Zhang, G. Ma, Y. Wang, Y. Liu, B. Wang, X. Yang and Q. Shi, Int. J. Agric. Biol. Eng., 2022, 15, 180–188 Search PubMed .
35. X. Huang, W. Sun, Z. Li, J. Shi, N. Zhang, Y. Zhang, X. Zhai, X. Hu and X. Zou, Food Chem., 2022, 396, 133654 CrossRef CAS PubMed .
36. J. Yan and Z. Song, J. Alloys Compd., 2024, 976, 173075 CrossRef CAS .
37. L. Dai, Y. Gao, W. Meng, Y. Li, Z. He and L. Wang, Sens. Actuators, B, 2024, 401, 134970 CrossRef CAS .
38. S. Park, P. M. Bulemo, W.-T. Koo, J. Ko and I.-D. Kim, Sens. Actuators, B, 2021, 343, 130137 CrossRef CAS .
39. D. Hamad, N. M. Shaalan, A. M. Abdelraheem and A. M. Abd-Elnaiem, J. Environ. Chem. Eng., 2024, 12, 113464 CrossRef CAS .
40. B. Bawab, S. M. Thalluri, J. Rodriguez-Pereira, H. Sopha, R. Zazpe and J. M. Macak, Electrochim. Acta, 2022, 429, 141044 CrossRef CAS .
41. L. Zhou, X. Chang, W. Zheng, X. Liu and J. Zhang, Chem. Eng. J., 2023, 475, 146300 CrossRef CAS .
42. G. Kresse and J. Furthmüller, Phys. Rev. B: Condens. Matter, 1996, 54, 11169–11186 CrossRef CAS PubMed .
43. Q. Fu, L. Yan, D. Liu, S. Zhang, H. Jiang, W. Xie, L. Yang, Y. Wang, H. Wang and X. Zhao, Appl. Catal., B, 2024, 343, 123501 CrossRef CAS .
44. L. Xue, Y. Ren, Y. Li, W. Xie, K. Chen, Y. Zou, L. Wu and Y. Deng, Small, 2023, 19, 2302327 CrossRef CAS PubMed .
45. Z. Song, Z. Wei, B. Wang, Z. Luo, S. Xu, W. Zhang, H. Yu, M. Li, Z. Huang, J. Zang, F. Yi and H. Liu, Chem. Mater., 2016, 28, 1205–1212 CrossRef CAS .
46. M. Li, D. Zhou, J. Zhao, Z. Zheng, J. He, L. Hu, Z. Xia, J. Tang and H. Liu, Sens. Actuators, B, 2015, 217, 198–201 CrossRef CAS .

---

Footnote

† These authors contributed equally to this work.

This journal is © The Royal Society of Chemistry 2025