Next Article in Journal
DRL-Based Beam Split Alleviation for Movable Antenna-Enabled Near-Field Wideband Communications
Previous Article in Journal
A Dual-Channel and Multi-Sensor Fusion Framework for Coal Mine Image Dehazing
Previous Article in Special Issue
Dynamic Sleep-Derived Heart Rate and Heart Rate Variability Features Associated with Glucose Metabolism Status: An Exploratory Feature-Selection Study Using Consumer Wearables
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Sensors
Volume 26
Issue 10
10.3390/s26103175
sensors-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite question_answer Discuss in SciProfiles
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

A Single Polyoxometalate Nanozyme for Cross-Reactive Thiol Array Sensing via pKa-Driven and Enrichment-Synergistic Strategy

by
Baohong Sun
†,
Ying Liu
†,
Xinxin Tian
,
Yu Fang
,
Jinpei Mei
,
Yang Chen
* and
Tao Ma
*
Anhui Engineering Technology Research Center of Biochemical Pharmaceutical, School of Pharmacy, Bengbu Medical University, Bengbu 233030, China
*
Authors to whom correspondence should be addressed.
†
These authors contributed equally to this work.
Sensors 2026, 26(10), 3175; https://doi.org/10.3390/s26103175
Submission received: 14 April 2026 / Revised: 5 May 2026 / Accepted: 9 May 2026 / Published: 17 May 2026
(This article belongs to the Special Issue Biosensors for Biomedical, Environmental and Food Applications)
Browse Figures
Versions Notes

Highlights

What are the main findings?
  • A single-material three-channel colorimetric array sensor based on Fe4P2W18 polyoxometalate nanozyme achieves 100% discrimination accuracy for glutathione, cysteine, and homocysteine within 5 min.
  • By exploiting pH-dependent stepped catalytic responses (pH 3.5, 4.0, 4.5), the sensor amplifies pKa-driven differences in biothiol molecular charge and structure, enabling a detection limit as low as 0.1 μM and quantification over 1–50 μM.
What are the implications of the main findings?
  • This strategy eliminates the need for multi-material sensor arrays while providing mechanistic clarity, offering a simple, rapid, and reliable tool for biothiol analysis in complex biological samples.
  • The sensor successfully distinguishes different cell types (HUVEC, HeLa, A549) and discriminates serum samples from cardiovascular disease patients, demonstrating strong potential for early disease diagnosis.

Abstract

Glutathione (GSH), cysteine (Cys), and homocysteine (Hcy) are essential biothiols involved in redox homeostasis and cell signaling. Yet, their structural similarity poses a significant challenge for accurate discrimination in complex biological samples. Herein, we report a polyoxometalate nanozyme (Fe4P2W18) with an atomically precise structure synthesized via a mild one-step method. Benefiting from its negatively charged surface and multi-electron reduction centers, the nanozyme effectively enriches biothiols under acidic conditions and exhibits excellent peroxidase-like activity. By leveraging its stepped catalytic response across pH 3.5, 4.0, and 4.5, which amplifies the pKa-driven differences in the overall molecular charge and electrostatic properties of biothiols, we constructed a single-material three-channel colorimetric array sensor that generates cross-reactive fingerprints. This sensor achieves 100% discrimination accuracy for GSH, Cys, and Hcy within 5 min, with a detection limit of 0.1 μM. It enables quantitative detection of thiol concentrations between 1 and 50 μM and accurately identifies binary and ternary thiol mixtures. Furthermore, the sensor successfully distinguishes cell types (HUVEC, HeLa, A549) and discriminates serum samples from cardiovascular disease patients. This strategy eliminates the complexity of multi-material arrays while leveraging the well-defined structure of Fe4P2W18 to provide mechanistic clarity, offering a simple, rapid, and reliable tool for biothiol analysis and early disease diagnosis.

1. Introduction

GSH, Cys, and Hcy are key low-molecular-weight biothiols involved in redox homeostasis, protein modification, and cell signaling [1,2]. Their abnormal levels are linked to various diseases: In human serum, normal physiological concentrations of GSH and Hcy typically range from 1 to 15 μM and 5 to 15 μM, respectively, while Cys is present at higher levels (~200–300 μM). Elevated Hcy (>15 μM) is an established independent risk factor for cardiovascular disease, and depleted GSH levels are associated with cancer and neurodegenerative disorders [3,4]. Despite their physiological significance, these thiols share high structural similarity, differing only in carbon chain length and side chains, making their accurate discrimination in complex samples challenging [5,6]. Traditional methods such as high-performance liquid chromatography and mass spectrometry provide sensitive quantification but require expensive equipment and labor-intensive pretreatment, thereby limiting point-of-care use [7,8]. Thus, simple, rapid sensing strategies enabling effective discrimination of structurally similar thiols are of great value for disease diagnosis.
Colorimetric nanozyme sensors have garnered widespread attention in biothiol detection in recent years due to their advantages, including simple operation, rapid response, and visual readout [9,10,11]. However, conventional nanozyme sensors generally follow a “lock-and-key” recognition mode, in which one sensing element selectively responds to a single analyte [12]. Owing to the high structural similarity among GSH, Cys, and Hcy, effective discrimination among the three using a single nanozyme sensing unit remains challenging [13]. To overcome this limitation, nanozyme-based array sensors can generate unique response “fingerprints” through an inhibition mechanism [14]. For instance, Zheng et al. developed a nanozyme array based on two types of metal ion-doped carbon dots, which improved the detection sensitivity for biothiols [15]. However, most reported array sensors for biothiol detection are based on strategies such as multi-material combinations, synergistic integration of single materials with environmental regulation, and confinement/microenvironment enhancement [16,17,18,19], with few strategies focusing on the synergistic regulation of nanozyme catalysis and thiol protonation states.
Among nanozyme materials, polyoxometalates (POMs) have recently emerged as promising sensing platforms owing to their unique combination of properties. Unlike many conventional nanozymes whose active sites and surface chemistry are poorly defined, POMs possess atomically precise molecular structures, well-characterized multi-electron redox centers, and inherently high negative charge density. These features confer several advantages for biothiol sensing: (i) the structural precision facilitates mechanistic understanding of catalytic and recognition processes; (ii) the highly negatively charged surface enables electrostatic enrichment of positively charged target molecules, enhancing sensitivity; and (iii) the peroxidase-like activity can be modulated by pH and composition. Despite these advantages, the use of a single POM nanozyme to construct a cross-reactive sensor array for discriminating structurally similar biothiols has not been reported.
The thiol (-SH) group of biothiols (GSH, Cys, Hcy) is weakly acidic and exists in a protonation-deprotonation equilibrium [20]. Under neutral or alkaline conditions, the proportion of thiolate (-S−) is high, leading to significantly enhanced nucleophilicity and stronger interactions with nanozymes, thereby resulting in a more pronounced inhibition effect [21,22]. In contrast, under acidic conditions, although the nucleophilicity (reducing ability) of the thiol group is diminished, making detection more challenging, the differences in protonation states among the three thiols become more pronounced due to their distinct pKa values [23]. These differences are further amplified across three pH gradients, generating highly distinctive “fingerprint responses” [24]. At these acidic pH values, the thiol groups remain predominantly protonated; the discriminatory power arises instead from the distinct net molecular charges and sizes governed by the protonation states of the α-amino, α-carboxyl, and side-chain groups. This necessitates that the nanozyme employed for array construction possesses a specific structure-activity relationship toward biothiols to enhance sensitivity under acidic conditions.
Herein, we synthesized a polyoxometalate nanozyme (Fe4P2W18) with an atomically precise structure via a mild one-step method (Scheme 1A). Due to the anion clusters and multi-electron reduction centers of the polyoxometalate framework, it can enrich biothiols through surface negative charges and electron transfer, thereby enhancing the specific detection capability for biothiols under acidic conditions (Scheme 1B). Leveraging the stepped catalytic response of this nanozyme under different pH conditions, we constructed a three-channel colorimetric array sensor using a single material. We employed three sensing units at pH 3.5, 4.0, and 4.5 to generate cross-reactive fingerprints, achieving 100% discrimination accuracy for GSH, Cys, and Hcy within 5 min. The resulting three-channel response vectors were processed by Linear Discriminant Analysis (LDA), a supervised pattern recognition algorithm that maximizes inter-class separation while minimizing intra-class variance, and Hierarchical Cluster Analysis (HCA), which provides complementary unsupervised grouping evidence (Scheme 1C). The response speed outperforms most previously reported colorimetric array sensors. This strategy eliminates the cumbersome steps associated with multi-material synthesis and formulation required by conventional array sensors. Meanwhile, the well-defined structure of Fe4P2W18 provides a reliable foundation for elucidating the catalytic and sensing mechanisms. Furthermore, this sensor was successfully applied to identify cell types (HUVEC, HeLa, A549) and to discriminate serum samples from patients with cardiovascular disease, demonstrating its promising potential for early disease diagnosis.

2. Materials and Methods

2.1. Chemical and Materials

Sodium tungstate dihydrate (Na2WO4·2H2O, ≥99%), phosphoric acid (H3PO4, 85%), acetic acid (CH3COOH, ≥99.5%), ferrous chloride (FeCl2, ≥98%), potassium chloride (KCl, ≥99.5%), and hydrogen peroxide (H2O2, 30%) were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). 3,3′,5,5′-Tetramethylbenzidine (TMB, ≥99%) was obtained from Aladdin Industrial Corporation (Shanghai, China). Glutathione (GSH, ≥98%), cysteine (Cys, ≥98%), and homocysteine (Hcy, ≥95%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Terephthalic acid (TA, ≥99%) was used as a fluorescence probe for hydroxyl radical (·OH) detection. All reagents were of analytical grade and used without further purification. Deionized water (18.2 MΩ·cm) was used throughout all experiments.

2.2. Characterization

The morphology of the Fe4P2W18 nanozyme was characterized by scanning electron microscopy (SEM, JSM-7800F, JEOL, Tokyo, Japan). X-ray diffraction (XRD) patterns were recorded on a D8 Advance diffractometer (Bruker, Karlsruhe, Germany) using Cu Kα radiation (λ = 1.5418 Å). Zeta potential measurements were performed on a Zetasizer Nano ZS90 (Malvern Instruments, Malvern, UK). Fourier-transform infrared (FT-IR) spectra were collected on a Nicolet iS10 spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) using KBr pellets. X-ray photoelectron spectroscopy (XPS) analysis was conducted on an ESCALAB 250Xi spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) with Al Kα radiation to determine elemental composition and chemical states. Energy-dispersive X-ray spectroscopy (EDS) was employed to verify the elemental distribution.

2.3. Synthesis of Na8[HPW9O34]·24H2O

The precursor Na8[HPW9O34]·24H2O was synthesized according to a previously reported method with slight modifications [25]. Typically, 120 g of Na2WO4·2H2O was dissolved in 150 mL of H2O at 80 °C under vigorous stirring. Subsequently, 3 mL of 85% H3PO4 was added, followed by the dropwise addition of 22 mL of CH3COOH within 30 min. After the formation of a white precipitate, the mixture was filtered. The precipitate was washed with 4 M NaCl solution, collected by filtration, and dried at 60 °C to obtain the final product.

2.4. Synthesis of Fe4P2W18 Nanozyme

The Fe4P2W18 polyoxometalate nanozyme was prepared via a mild one-step method. In a typical procedure, FeCl2 was dissolved in 40 mL of deionized water, and the pH was adjusted to 6.0. Na8[HPW9O34]·24H2O was then added at a molar ratio of FeCl2 to Na8[HPW9O34]·24H2O of 2.5:1, based on preliminary optimization of catalytic activity. The mixture was refluxed at 140 °C for 90 min. After cooling to room temperature, 7 g of KCl was added and stirred for 20 min. The resulting precipitate was collected by vacuum filtration, washed three times with deionized water to remove residual KCl and soluble impurities, and finally dried under vacuum at 60 °C until constant weight.

2.5. Peroxidase-like Activity Assay

The peroxidase-like activity of Fe4P2W18 was evaluated using the classic TMB chromogenic reaction. Typically, Fe4P2W18 (12 μg/mL), TMB (2 mM), and H2O2 (1 mM) were mixed in acetate buffer (pH 3.5) and incubated at 45 °C for 5 min. The absorbance of the reaction mixture was recorded at 652 nm using a UV-vis spectrophotometer (UV-2600, Shimadzu, Japan). Steady-state kinetic studies were performed by varying the concentration of TMB or H2O2 while keeping the other substrate constant. The kinetic parameters (Km and Vmax) were calculated using the Lineweaver-Burk double-reciprocal plot.

2.6. Detection Mechanism Investigation

The generation of hydroxyl radicals (·OH) during the catalytic process was confirmed using terephthalic acid (TA) as a fluorescent probe and electron paramagnetic resonance (EPR) spectroscopy. For the TA assay, Fe4P2W18, TA, and H2O2 were incubated in acetate buffer, and the fluorescence emission intensity was measured at 425 nm with excitation at 315 nm. EPR spectra were recorded on a Bruker EMXplus spectrometer (Bruker, Billerica, MA, USA) using 5,5-dimethyl-1-pyrroline N-oxide (DMPO) as the spin-trapping agent.

2.7. Colorimetric Array Sensor Construction and Thiol Discrimination

A single-material three-channel array sensor was constructed based on the pH-dependent catalytic activity of Fe4P2W18. The detection was performed in acetate buffer at three different pH values (3.5, 4.0, and 4.5). For each channel, Fe4P2W18 (12 μg/mL), TMB (2 mM), H2O2 (1 mM), and varying concentrations of biothiols (GSH, Cys, or Hcy) were mixed and incubated at 45 °C for 5 min. The absorbance at 652 nm was recorded. The signal response (ΔA) was calculated as ΔA = A0 − A, where A0 and A represent the absorbance in the absence and presence of biothiols, respectively. The three-channel response patterns were analyzed by linear discriminant analysis (LDA) and hierarchical cluster analysis (HCA) using SPSS Statistics version 26.0 (IBM, Armonk, NY, USA).

2.8. Analytical Performance Evaluation

The stability, reproducibility, repeatability, and selectivity of the Fe4P2W18 sensor array were systematically evaluated. For stability, the nanozyme was stored at room temperature, and its catalytic activity was measured daily over 7 days using the standard TMB assay (Fe4P2W18: 12 μg/mL, TMB: 2 mM, H2O2: 1 mM, acetate buffer pH 3.5, 45 °C, 5 min). Reproducibility was assessed by synthesizing five independent batches of Fe4P2W18 nanozyme, and each batch was tested for the detection of GSH, Cys, and Hcy at 10 μM under the optimized conditions; the RSD of the ΔA values across the five batches was calculated for each analyte. Repeatability was determined by performing ten successive measurements of GSH (10 μM) using a single sensor, and the RSD was calculated. For selectivity, a mixed interferent solution containing glucose, glycine, lysine, ascorbic acid, Na+, and K+ (100 μM each) and BSA (100 μg/mL) was prepared in acetate buffer. The signal response (ΔA) of GSH, Cys, and Hcy (10 μM each) was first measured individually, and then each analyte was spiked into the interferent mixture and measured again. The recovery was calculated as (ΔAin mixture/ΔAalone) × 100%. All experiments were performed in triplicate.

2.9. Real Sample Analysis

All procedures involving human samples were approved by the Medical Research Ethics Committee of Bengbu Medical University (approval No. 2025476) and were performed in accordance with the principles of the Declaration of Helsinki. Written informed consent was obtained from all participants and donors. Human serum samples were collected from healthy volunteers and patients with cardiovascular disease. Cell lines, including human umbilical vein endothelial cells (HUVEC), HeLa cells, and A549 cells, were cultured and processed to obtain cell lysates. The prepared sensor array was applied to these real samples following the same protocol described above. Each experiment was performed in triplicate to ensure reproducibility. Human serum and HeLa cell lysate were diluted 10-fold with acetate buffer (0.1 M, pH 3.5). Each sample was spiked with GSH, Cys, or Hcy at a final added concentration of 10 μM. The three-channel sensor array was applied under optimized conditions (Fe4P2W18: 12 μg/mL, TMB: 2 mM, H2O2: 1 mM, pH 3.5/4.0/4.5, 45 °C, 5 min), and the ΔA response was recorded. Measured concentrations were derived from the LDA1 calibration curves. Recovery was calculated as (Cmeasured − Cunspiked)/Cspiked × 100%.

3. Results and Discussion

3.1. Synthesis and Characterization of Fe4P2W18 Nanozyme

The Fe4P2W18 polyoxometalate nanozyme was synthesized via a mild one-step method using Na8[HPW9O34]·24H2O and FeCl2 as precursors. To obtain a nanozyme with high catalytic activity, the molar ratio of FeCl2 to Na8[HPW9O34] was first optimized. As shown in Figure 1A, when the ratio reached 2.5:1, the absorbance at 652 nm increased rapidly, while further increasing the FeCl2 content resulted in negligible improvement. Considering both catalytic efficiency and economic cost, a ratio of 2.5:1 was selected for subsequent experiments. The morphology and composition of Fe4P2W18 were characterized by SEM, XRD, Zeta potential analysis, and FT-IR spectroscopy. The SEM image (Figure 1B) reveals that the Fe4P2W18 is composed of numerous small nanocrystals with particle sizes in the range of several hundred nanometers. The XRD pattern of Fe4P2W18 (Figure 1C) exhibits a distinct sharp peak at approximately 2θ = 8°, indicative of a well-defined periodic arrangement of the polyoxometalate clusters, which is consistent with its atomically precise structure [26]. Meanwhile, a broad hump in the 2θ range of 20–35° is observed, suggesting the nanocrystalline nature of the material [27]. This combined feature is typical for nanoscale polyoxometalate aggregates, where the primary clusters maintain long-range order along specific crystallographic directions. Meanwhile, the overall particle size and surface disorder lead to peak broadening at higher diffraction angles [28]. This unique structural characteristic provides both high catalytic activity due to the exposed active sites and structural stability from the cluster framework. The precursor Na8[HPW9O34] exhibits a highly negative Zeta potential (approximately −30 mV), confirming the strong electrostatic driving force that is retained in the final Fe4P2W18 nanozyme (Figure 1D, −18 mV). This negatively charged surface enhances the affinity toward positively charged substrates, thereby improving its peroxidase-like catalytic performance. FT-IR spectroscopy (Figure 1E) further verified the chemical composition, with characteristic peaks at 779, 866, 935, and 1050 cm−1 corresponding to W–Oc–W, W–Ob–W, W1/4Od, and P–Oa bonds, respectively [25]. The presence of these characteristic peaks, together with the negatively charged surface revealed by Zeta potential, confirms the structural integrity of Fe4P2W18 and provides the necessary chemical basis for its enrichment of biothiols. The hydrolytic stability of the Fe4P2W18 framework across the working pH range was confirmed by FT-IR spectra recorded after incubation at pH 3.5, 4.0, and 4.5, which showed no changes in the characteristic POM vibrational bands (Figure 1F).
XPS analysis was employed to precisely determine the elemental composition and valence states of Fe4P2W18 (Figure 2A). The survey spectrum revealed characteristic peaks corresponding to P 2p, W 4f, Fe 2p, K 2p, and O 1s, which are in good agreement with the theoretical composition of the target compound. Quantitative atomic percentages were derived from the high-resolution narrow-scan analyses (Table 1). High-resolution XPS spectra were further acquired to elucidate the chemical states of tungsten and iron. The W 4f spectrum (Figure 2B) exhibits two distinct peaks at binding energies of approximately 35.6 eV and 37.7 eV, assigned to W 4f7/2 and W 4f5/2, respectively, with a spin–orbit splitting of 2.1 eV. These values are characteristic of W6+ in a typical polyoxometalate framework, confirming the fully oxidized state of tungsten. The Fe 2p spectrum (Figure 2C) shows two main peaks at approximately 711.2 eV and 724.8 eV, corresponding to Fe 2p3/2 and Fe 2p1/2, respectively, with a spin–orbit splitting of 13.6 eV (Figure 2C). Notably, the presence of a satellite peak at approximately 718.5 eV is a fingerprint of high-spin Fe3+ species, confirming that iron exists in the +3 oxidation state. These binding energy assignments are essential for the peroxidase-like catalytic activity of Fe4P2W18, as they confirm the retention of redox-active centers within the polyoxometalate structure [29]. The C 1s signal primarily serves as a binding energy reference for charge correction (Figure 3A). The K 2p doublet exhibits a characteristic spin–orbit splitting of approximately 2.8 eV, confirming the presence of K+ as the charge-balancing counterion within the polyoxometalate structure (Figure 3B). The P 2p signal, typically appearing as a single peak at around 133.5 eV for P5+, verifies the incorporation of phosphorus into the framework without the presence of reduced phosphorus species (Figure 3C). These features provide key evidence for the stability and structural integrity of the heteropoly Fe4P2W18 compound. The oxygen bonding environments, which would ideally be resolved by deconvolution of the O 1s XPS spectrum, were instead characterized by FT-IR spectroscopy (Figure 1E), where the distinct W–Oc–W, W–Ob–W, W=Od, and P–Oa vibrations provide unambiguous bond-level structural information complementary to the XPS data.
High-resolution transmission electron microscopy (HRTEM) was performed to further investigate the detailed microstructure of Fe4P2W18. As shown in Figure 2D, the nanozyme exhibits a distinct nanosheet-like morphology with lateral dimensions ranging from several hundred nanometers to approximately one micrometer. The nanosheets are composed of densely packed nanocrystalline domains with sizes of approximately 5–10 nm, consistent with the broad diffraction features observed in the XRD pattern. Notably, HRTEM imaging (Figure 2D) reveals well-resolved lattice fringes within these nanocrystalline regions, with an interplanar spacing of approximately 0.38 nm, corresponding to the characteristic spacing of the polyoxometalate framework. The coexistence of ordered lattice fringes and amorphous-like regions reflects the partially crystalline nature of the material, which is typical for nanoscale polyoxometalate aggregates. EDS mapping analysis (Figure 2E) further corroborated the structural homogeneity at the nanoscale, revealing the uniform distribution of W, O, Fe, and P elements, thereby demonstrating the successful construction of the W–O–Fe–P network within the polyoxometalate framework [30]. The corresponding energy-dispersive X-ray spectroscopy (EDS) spectrum exhibited characteristic X-ray peaks for W, O, Fe, and P, further confirming the elemental composition of the nanozyme (Figure 2F). Collectively, these characterizations confirm the successful synthesis of Fe4P2W18, with an atomically precise, well-defined structure.

3.2. POD-like Activity of Fe4P2W18 Nanozyme

The peroxidase-like activity of Fe4P2W18 was evaluated using the classic TMB chromogenic reaction. As shown in Figure 4A, when Fe4P2W18, TMB, and H2O2 were all present, a distinct absorption peak at 652 nm emerged, corresponding to the π–π* transition of oxidized TMB (oxTMB) [31]. In contrast, control experiments lacking any one of the three components (Fe4P2W18, TMB, or H2O2) exhibited negligible absorbance at 652 nm, confirming that the catalytic oxidation of TMB requires the simultaneous presence of the nanozyme and H2O2. Concurrently, the reaction solution turned from colorless to deep blue, providing a clear visual indication that Fe4P2W18 can utilize H2O2 as a substrate to generate highly reactive hydroxyl radicals (·OH) that oxidize TMB [32]. The effect of Fe4P2W18 concentration on catalytic activity was subsequently investigated. As illustrated in Figure 4B, the absorbance at 652 nm increased progressively with increasing nanozyme concentration in the range of 0–20 μg/mL, indicating a concentration-dependent catalytic response. A concentration of 12 μg/mL was selected as the optimal dosage for subsequent experiments, providing a suitable balance between high catalytic activity and economic efficiency.
To clearly understand the catalytic process of H2O2 by the Fe4P2W18 nanomaterial, we conducted a series of experiments. First, we selected terephthalic acid (TA), a fluorescent probe with high specificity for hydroxyl radicals (·OH). As shown in Figure 5A, significant fluorescence emission was observed only when Fe4P2W18, TA, and H2O2 were present simultaneously. In contrast, no obvious fluorescence was detected in the control groups (including Fe4P2W18 alone, TA alone, and H2O2 alone). This phenomenon strongly demonstrates that Fe4P2W18 effectively catalyzes the decomposition of H2O2 to generate highly reactive ·OH [33]. These ·OH subsequently react with TA to produce 2-hydroxyterephthalic acid (TAOH), which exhibits strong fluorescence characteristics. To further confirm this result, we performed electron paramagnetic resonance (EPR) measurements on the reaction system. As shown in Figure 5B, a distinct quartet signal was observed in the EPR spectrum, which is consistent with the standard spectrum of ·OH. These experimental results not only confirm that Fe4P2W18 generates ·OH during the catalytic decomposition of H2O2 but also provide an important experimental foundation for further in-depth investigation of the reaction mechanism of this catalytic system.

3.3. Optimization of Conditions

The influence of various parameters on the peroxidase-like activity of Fe4P2W18 was systematically investigated to determine the optimal reaction conditions. As shown in Figure 6A, the absorbance increased with pH from 3.0 to 3.5, reaching a maximum at pH 3.5, and then decreased at higher pH values. This indicates that Fe4P2W18 exhibits optimal catalytic activity at pH 3.5, likely due to pH-induced changes in the structure and active sites of the nanozyme, which subsequently affect its interaction with substrates. The catalytic activity increased with temperature from 30 to 60 °C, peaking at 60 °C, followed by a slight decrease at higher temperatures (Figure 6B). Notably, Fe4P2W18 maintained high catalytic efficiency across the entire temperature range of 30–70 °C, demonstrating broad temperature adaptability. To ensure the stability of the analytes, a reaction temperature of 45 °C was selected for subsequent experiments. As depicted in Figure 6C, the catalytic reaction proceeded rapidly, reaching a plateau within 5 min. This fast catalytic kinetics meets the requirements for rapid detection. Optimization of TMB and H2O2 concentrations (Figure 6D,E) revealed that the optimal concentrations were 2 mM for TMB and 1 mM for H2O2, beyond which no significant increase in activity was observed. Collectively, the optimal reaction conditions were determined as follows: pH 3.5, temperature 45 °C, reaction time 5 min, TMB concentration 2 mM, and H2O2 concentration 1 mM. All optimization experiments were performed in triplicate to ensure data reliability.

3.4. Kinetic Analysis of Fe4P2W18 Nanozyme

Steady-state kinetic experiments were conducted to evaluate the catalytic performance of Fe4P2W18. The absorbance at 652 nm increased progressively with increasing concentrations of H2O2 and TMB, demonstrating the dependence of the catalytic reaction on substrate concentration. To quantitatively assess the catalytic efficiency, kinetic parameters were calculated by fitting the experimental data to the Michaelis–Menten model. The corresponding Lineweaver–Burk double-reciprocal plots (Figure 7A–D) yielded excellent linear fits, confirming that the catalytic behavior follows classical Michaelis–Menten kinetics [34]. For H2O2 as the substrate, the Michaelis constant (Km) and maximum reaction velocity (Vmax) were determined to be 0.44 mM and 3.325 × 10−7 M s−1, respectively. The relatively low Km value indicates a high affinity of Fe4P2W18 toward H2O2. For TMB as the substrate, the Km and Vmax values were 0.863 mM and 3.281 × 10−7 M s−1, respectively. These results demonstrate that Fe4P2W18 exhibits favorable affinity and catalytic efficiency toward both substrates. When compared with natural horseradish peroxidase (HRP) (Table 2), the catalytic efficiency of Fe4P2W18 is comparable, further validating its potential as an effective peroxidase mimic for practical applications.

3.5. pKa-Driven and Enrichment-Synergistic Strategy for Biothiol Determination

A colorimetric sensor array was constructed based on the pH-dependent peroxidase-like activity of Fe4P2W18, leveraging a pKa-driven and enrichment-synergistic strategy for biothiol discrimination. The polyoxometalate framework contains anionic clusters and multi-electron reduction centers, enabling the nanozyme to enrich biothiols via both surface-negative-charge-mediated electrostatic interactions and electron-transfer processes [35]. The maximum adsorption capacity (Qmax) of Fe4P2W18 toward biothiols was determined to be approximately 0.68 μmol/mg, reflecting its efficient enrichment capability (Figure 8). Meanwhile, the thiol groups of GSH, Cys, and Hcy are weakly acidic and exhibit distinct pKa values (approximately 8.3 for Cys, 8.9 for Hcy, and 9.2 for GSH). Under the acidic pH conditions employed in this study (3.5, 4.0, and 4.5), these three biothiols exist predominantly in their protonated forms but with systematically varying protonation states due to their pKa differences. At these acidic pH values, the thiol groups remain predominantly protonated; the differential responses arise instead from the distinct net molecular charges and sizes governed by the protonation states of the α-amino, α-carboxyl, and side-chain groups. This variation leads to differential interactions with the nanozyme surface and distinct inhibition efficiencies toward the catalytic oxidation of TMB. By combining the enrichment capability of Fe4P2W18 with the pKa-driven differential responses of the three thiols across three finely tuned pH channels, the sensor array generates highly distinctive cross-reactive fingerprints (Scheme 2). The detection was performed at three different pH values (3.5, 4.0, and 4.5) to generate three sensing channels. Biothiols (GSH, Cys, and Hcy) inhibit the nanozyme-catalyzed TMB oxidation by scavenging reactive oxygen species (ROS), leading to a decrease in absorbance at 652 nm. The signal response (ΔA) was calculated as ΔA = A0 − A, where A0 and A represent the absorbance in the absence and presence of biothiols, respectively [36]. As shown in Figure 9A,B, the three biothiols exhibited distinct response patterns (bar chart and heat map) across the three pH channels, confirming the effectiveness of the sensor array for biothiol identification. Notably, at concentrations of 100, 10, 1, and 0.1 μM, all three biothiols were accurately discriminated without any misclassification (Figure 9C–F). The limit of detection (LOD) was calculated using the 3σ/slope method, where σ is the standard deviation of the blank signal (n = 3), and the slope was derived from the linear portion of the response–concentration curve. Even at an ultralow concentration of 0.1 μM, the sensor successfully distinguished GSH, Cys, and Hcy. This detection limit is significantly lower than the physiological concentrations of biothiols in serum, demonstrating the applicability of this array for real sample analysis. To rigorously validate the claimed 100% discrimination accuracy and exclude potential model overfitting, the LDA classification results were examined using confusion matrices and leave-one-out cross-validation (LOOCV). As shown in Table 3, at all four tested concentrations (100, 10, 1, and 0.1 μM), each of the three biothiols was correctly classified in all replicates, yielding a standard LDA accuracy of 100% (36/36). LOOCV was further performed by iteratively holding out one sample as the test set while training the model on the remaining 35 samples. The LOOCV-confirmed accuracy was also 100% at every concentration (Table 4), demonstrating that the perfect classification reflects genuine inter-class separability rather than overfitting to the training data.

3.6. Analytical Performance of the Sensor Array

The analytical performance of the Fe4P2W18 sensor array was systematically evaluated in terms of stability, reproducibility, repeatability, and selectivity. As summarized in Table 5, the nanozyme retained 98.4% of its initial catalytic activity after storage at room temperature for 7 days, demonstrating excellent stability. The batch-to-batch reproducibility was assessed using five independently synthesized batches, yielding relative standard deviations (RSD) of 2.1%, 1.9%, and 3.3% for GSH, Cys, and Hcy (10 μM each), respectively, all within the acceptable threshold of 5%. The repeatability, evaluated by ten successive measurements of GSH (10 μM) using a single sensor, gave an RSD of 1.8%, confirming the high precision of the measurement. Selectivity was evaluated by spiking each biothiol (10 μM) into a mixed interferent solution containing glucose, glycine, lysine, ascorbic acid, Na+, K+ (100 μM each), and BSA (100 μg/mL). The recoveries were 98.5%, 97.2%, and 96.8% for GSH, Cys, and Hcy, respectively, all within the 90–110% acceptance range, confirming the excellent anti-interference capability of the sensor array.

3.7. Quantitative Determination of GSH, Cys, and Hcy

To evaluate the quantitative performance of the sensor array, biothiol samples at varying concentrations (1, 10, 20, 30, 40, and 50 μM) were systematically analyzed. As shown in Figure 10A, different concentrations of GSH were well separated in the two-dimensional LDA plot, with the first discriminant function (LDA1) accounting for 98.42% of the total variance—a value substantially higher than the acceptable threshold of 60%, indicating that LDA1 effectively captures the concentration-dependent variation [37]. A good linear relationship was established between LDA1 scores and GSH concentrations in the range of 1–50 μM (Figure 10B), with a regression equation of y = 0.032x − 0.518 (R2 = 0.994). This excellent linearity confirms the reliable quantitative detection of GSH using the sensor array. Similarly, quantitative discrimination and linear relationships were also achieved for Hcy (Figure 10C,D) and Cys (Figure 10E,F). For Hcy, the LDA1 contribution was 97.63% with a regression equation of y = 0.028x − 0.447 (R2 = 0.991); for Cys, the LDA1 contribution was 96.89% with a regression equation of y = 0.030x − 0.492 (R2 = 0.992). These results collectively demonstrate that the sensor array not only enables qualitative discrimination but also provides excellent quantitative capability for biothiol detection across a broad concentration range, making it suitable for practical applications where precise concentration determination is required. The predictive accuracy of the calibration models was further validated by leave-one-out cross-validation. The RMSEP values were determined to be 1.6, 1.8, and 2.0 μM for GSH, Cys, and Hcy, respectively, corresponding to 3.3%, 3.7%, and 4.1% of the calibration range (1–50 μM) (Table 6), confirming that the models are not overfitted and possess reliable quantitative predictive power.

3.8. Discrimination of Binary and Ternary Thiol Mixtures

To evaluate the capability of the sensor array for identifying thiol mixtures in complex scenarios, binary and ternary mixtures containing different molar ratios of GSH, Cys, and Hcy were analyzed. This investigation is particularly important because biological samples typically contain multiple biothiols simultaneously rather than a single species, and the ability to discriminate mixtures is essential for practical diagnostic applications [38]. Three binary mixtures (Cys/Hcy = 50:50, Hcy/GSH = 50:50, and Cys/GSH = 50:50) and three ternary mixtures (Cys/Hcy/GSH = 25:25:50, 50:25:25, and 25:50:25) were prepared, each with a total concentration of 100 μM. As shown in Figure 11A–D, all binary and ternary mixtures were clearly distinguished in the two-dimensional LDA plots, with distinct clustering centers showing no overlap. The cumulative variance explained by the first two discriminant functions exceeded 97% in all cases, confirming the high discrimination power of the sensor array [39]. The classification accuracy achieved for both binary and ternary mixtures was 100%, with no misclassification observed across all replicates, outperforming previously reported nanozyme array sensors (Table 7). These results demonstrate that the sensor array not only distinguishes individual biothiols but also accurately identifies their mixtures at various molar ratios, which is crucial for practical applications where biological samples typically contain multiple biothiols simultaneously. This capability, combined with the quantitative detection performance described above, positions the sensor array as a versatile tool for comprehensive biothiol analysis in complex biological matrices.

3.9. Real Sample Analysis

The practical applicability of the sensor was validated by analyzing real biological samples, representing a critical step toward translation of this single-material array strategy from proof-of-concept to real-world diagnostic applications. Cancer cells often exhibit elevated intracellular GSH levels compared to normal cells due to altered glutathione metabolism [42]. Based on this established correlation, the sensor was employed to distinguish three cell types: HUVEC (normal), HeLa (cervical cancer), and A549 (lung cancer). As shown in Figure 12A, the three cell types were clearly separated in the LDA plot, with normal cells concentrated on the left and cancer cells on the right, demonstrating that the sensor array can effectively translate intracellular GSH level differences into distinct fingerprint responses. The violin plot (Figure 12B) further revealed significant differences between normal and cancer cells, as well as between the two cancer cell lines, highlighting the high sensitivity of the sensor to subtle variations in GSH levels. Hierarchical cluster analysis (HCA) confirmed that all cell samples were accurately classified without any misclassification (Figure 12C), achieving 100% accuracy. This successful cell typing underscores the key innovation of our strategy: by leveraging the pKa-driven differential responses of biothiols across three pH gradients, a single structurally well-defined Fe4P2W18 nanozyme can generate sufficiently orthogonal fingerprints to resolve complex biological samples, eliminating the need for multiple sensing materials.
The sensor was further applied to analyze Hcy levels in human serum samples, a clinically relevant application given the established link between elevated Hcy and cardiovascular disease [43]. Normal serum Hcy concentrations typically range from 5 to 15 μM, while patients with cardiovascular disease often exhibit significantly higher levels [44]. Four serum samples with different Hcy concentrations (9.1, 18.8, 33.7, and 55.9 μM) were analyzed. LDA revealed a clear distribution along the negative-to-positive axes, with low-concentration samples on the left and high-concentration samples on the right, without any overlap (Figure 12D). The violin plot (Figure 12E) showed increasing Euclidean distance with increasing Hcy concentration, clearly distinguishing normal from abnormal samples. HCA results further confirmed that the four serum samples were classified into four distinct groups, with normal and mild cases grouped as healthy (Figure 12F). These results validate the potential of the sensor for early diagnosis of cardiovascular disease. Importantly, the successful discrimination of serum samples with varying Hcy concentrations in a complex matrix such as human serum demonstrates the robustness of the sensor array against potential interferences, a direct benefit of the cross-reactive fingerprinting mechanism enabled by the single-material pH-gradient design. Although POMs may adsorb serum proteins, the cross-reactive fingerprinting strategy inherently tolerates such systematic background signals, as LDA extracts only the analyte-dependent differential responses across the three pH channels. Collectively, these real-sample validations highlight that our pKa-driven and enrichment-synergistic strategy not only simplifies sensor fabrication but also delivers reliable diagnostic information in clinically relevant contexts. To further validate the accuracy of the sensor array for quantitative analysis in complex biological matrices, spike-recovery experiments were performed. Human serum and HeLa cell lysate were spiked with GSH, Cys, and Hcy at a concentration of 10 μM. As shown in Table 8, the recoveries ranged from 91.0% to 108.0%, with RSD values between 2.9% and 4.8%, confirming the reliability of the sensor for accurate biothiol determination in real samples.

4. Conclusions

This work presents a conceptually distinct strategy for biothiol discrimination by synergistically regulating both nanozyme catalysis and thiol protonation states. Unlike conventional array sensors that rely on multi-material combinations or physical confinement, this approach utilizes a single polyoxometalate nanozyme (Fe4P2W18) with an atomically precise structure, which enables effective enrichment of biothiols through its negatively charged surface and multi-electron reduction centers. By leveraging the stepped catalytic response across three acidic pH gradients (3.5, 4.0, and 4.5), the system amplifies the inherent pKa-governed differences in biothiol molecular charge and structure, generating highly distinctive fingerprint responses. This single-material design eliminates the cumbersome synthesis and poor structural definition associated with multi-component arrays. At the same time, the well-defined molecular structure of Fe4P2W18 provides a reliable foundation for elucidating the catalytic and sensing mechanisms. The resulting sensor achieves 100% discrimination accuracy for GSH, Cys, and Hcy within 5 min, with successful application to cell typing and cardiovascular disease serum analysis, demonstrating a shift from multi-material complexity to single-material, mechanism-driven array sensing for early disease diagnosis.

Author Contributions

Conceptualization, T.M. and Y.C.; methodology, Y.L. and B.S.; software, Y.L.; validation, Y.L., X.T., and Y.F.; formal analysis, J.M.; investigation, Y.L.; resources, T.M.; data curation, X.T.; writing—original draft preparation, Y.L.; writing—review and editing, T.M.; visualization, B.S.; supervision, T.M.; project administration, T.M.; funding acquisition, T.M. and Y.C. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Natural Science Research Project of the Anhui Educational Committee (No. 2022AH040220), the Natural Science Key Project of Bengbu Medical University (Grant No. 2023byzd018), the National Natural Science Foundation of China (No. 82500009), and Anhui Provincial Higher Education Quality Engineering Project of 2015 (2015zy034).

Institutional Review Board Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Medical Research Ethics Committee of Bengbu Medical University (protocol code 2025476 and 11 March 2025).

Informed Consent Statement

Informed consent was obtained from all subjects involved in the study. Written informed consent has been obtained from the patient(s) to publish this paper.

Data Availability Statement

The data supporting this article are available in the article and in its online SI.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Yu, T.; Li, Y.; Li, J.; Gan, Y.; Long, Z.; Deng, Y.; Zhang, Y.; Li, H.; Yin, P.; Yao, S. Multifunctional fluorescent probe for simultaneous detection of ATP, Cys, Hcy, and GSH: Advancing insights into epilepsy and liver injury. Adv. Sci. 2025, 12, 2415882. [Google Scholar] [CrossRef] [PubMed]
  2. McGee, C.C.; Bandyopadhyay, T.; McCracken, C.N.; Talib, E.; Patterson, C.E.; Outten, C.E. Cysteine import via the high-affinity GSH transporter Hgt1 rescues GSH auxotrophy in yeast. J. Biol. Chem. 2025, 301, 108131. [Google Scholar] [CrossRef]
  3. Xia, C.; Xing, X.; Zhang, W.; Wang, Y.; Jin, X.; Tian, M.; Ba, X.; Hao, F. Cysteine and homocysteine can be exploited by GPX4 in ferroptosis inhibition independent of GSH synthesis. Redox Biol. 2024, 69, 102999. [Google Scholar] [CrossRef]
  4. Hou, Y.; Lv, B.; Du, J.; Ye, M.; Jin, H.; Yi, Y.; Huang, Y. Sulfide regulation and catabolism in health and disease. Signal Transduct. Target. Ther. 2025, 10, 174. [Google Scholar] [CrossRef]
  5. Long, Z.; Li, Y.; Yu, T.; Li, T.; Wang, J.; Zhang, Y.; Li, H.; Yin, P.; Zhou, H.; Yao, S. Novel Coumarin-based fluorescent probe for tracking Cys, GSH, and hcy Dynamics in Lysosomes under oxidative and Hyperosmotic stress. Anal. Chem. 2025, 97, 14620–14628. [Google Scholar] [CrossRef] [PubMed]
  6. Wu, X.-X.; Guo, C.-W.; Lu, W.-Y.; Li, H.-J.; Wu, Y.-C. A coumarin-based dual-channel fluorescent probe for discriminating Hcy and GSH in live cells. Anal. Methods 2026, 18, 109–114. [Google Scholar] [CrossRef] [PubMed]
  7. Qi, Y.; Xu, M.; Lu, H.; Wang, X.; Peng, Y.; Wang, Z.; Liang, F.; Jiang, X.; Du, B. Hepatic biotransformation of renal clearable gold nanoparticles for noninvasive detection of liver glutathione level via urinalysis. Angew. Chem. Int. Ed. 2024, 63, e202409477. [Google Scholar] [CrossRef]
  8. Zhong, Z.-J.; Ling, J.; Yao, Z.-P.; Liu, L.-F.; Zheng, J.-Y.; Xin, G.-Z. Targeted quantification of glutathione/arginine redox metabolism based on a novel paired mass spectrometry probe approach for the functional assessment of redox status. Anal. Chem. 2024, 96, 9885–9893. [Google Scholar] [CrossRef]
  9. Chigozie, A.E.; Ravikumar, A.; Yang, X.; Tamilselvan, G.; Deng, Y.; Arunjegan, A.; Li, X.; Hu, Z.; Zhang, Z. A metal–phenolic coordination framework nanozyme exhibits dual enzyme mimicking activity and its application is effective for colorimetric detection of biomolecules. Anal. Methods 2024, 16, 3530–3538. [Google Scholar] [CrossRef]
  10. Panferov, V.G.; Zhang, X.; Wong, K.Y.; Lee, J.H.; Liu, J. Biomedical applications of nanozymes: An enzymology perspective. Angew. Chem. Int. Ed. 2025, 64, e202512409. [Google Scholar] [CrossRef]
  11. Wang, H.; Zhao, L.; Gao, Z.; Cui, A.; Mu, X.; Wang, X.; Shan, G.; Chen, Y. CuS@ CDs nanozymes with photothermal-enhanced peroxidase-like activities for colorimetric/fluorescent detection of glutathione and ferric ions in the coexisting system. Sens. Actuators B Chem. 2024, 398, 134726. [Google Scholar] [CrossRef]
  12. Sun, B.; Cui, X.; Zhang, J.; Tang, Y.; Sun, H. Highly sensitive hydrolytic nanozyme–based sensors for colorimetric detection of aluminum ions. Anal. Bioanal. Chem. 2024, 416, 5985–5992. [Google Scholar] [CrossRef]
  13. Cai, C.; Zhu, C.; Lv, L.; Huang, P.; Mao, J.; Wu, F.-Y.; Deng, K.-Y. Distinct dual enzyme-like activities of Fe–N–C single-atom nanozymes enable discriminative detection of cellular glutathione. Chem. Commun. 2023, 59, 11252–11255. [Google Scholar] [CrossRef] [PubMed]
  14. Wei, M.; Yang, M.; Leng, H.; Shu, Y. A novel intelligent sensing strategy: Integration of metal-doped carbon dots nanozymes and machine learning for rapid screening of biothiols in disease. Sens. Actuators B Chem. 2025, 444, 138304. [Google Scholar] [CrossRef]
  15. Zhou, X.; Li, L.; Wang, Y.; Kong, T.; Cao, Z.; Xie, H.; Liang, W.; Wang, Y.; Qian, S.; Chao, J. Nanozyme inhibited sensor array for biothiol detection and disease discrimination based on metal ion-doped carbon dots. Anal. Chem. 2023, 95, 8906–8913. [Google Scholar] [CrossRef]
  16. Tabibi, T.; Dashtian, K.; Zare-Dorabei, R.; Nejati, M.; Esmaeili, A.; Noroozifar, M.; Kerman, K.; Mahdavi, M. Smartphone-integrated multifunctional nanozyme-based colorimetric sensor array for rapid and selective biothiol detection in point-of-care applications. Chem. Eng. J. 2025, 519, 165378. [Google Scholar] [CrossRef]
  17. Chen, X.; Liu, J.; Tang, Z.; Liu, S.; Peng, J.; Liang, H.; Niu, X. Machine learning-enabled time-resolved nanozyme-encoded recognition of endogenous mercaptans for disease diagnosis. Anal. Chem. 2025, 97, 10463–10473. [Google Scholar] [CrossRef]
  18. Wang, X.; Xu, R.; Wang, Y.; Li, M.; Wei, H.; Qin, G.; Li, Y.; Wei, Y. Self-supplying of hydrogen peroxide nanozyme-based colorimetric sensing array as electronic tongue for biothiol detection and disease discrimination. Talanta 2025, 288, 127727. [Google Scholar] [CrossRef] [PubMed]
  19. Zhang, T.; Liu, Y.; Wu, Q.; Zhou, L.; Liang, H.; Lu, Y.; Dong, M.; Liu, N. MoS2 decorated with Pd nanoparticles as a novel oxidase nanozyme for biothiol detection and cancer cell discrimination. Microchem. J. 2025, 215, 114423. [Google Scholar] [CrossRef]
  20. Ribson, R.D.; Follmer, A.H.; Babicz, J.T., Jr.; Sosa Alfaro, V.; Hadt, R.G.; Hunter, M.S.; Wilson, M.A.; Sokaras, D.; Alonso-Mori, R. Determination of Thiol Protonation States by Sulfur X-ray Spectroscopy in Biological Systems. J. Phys. Chem. Lett. 2025, 16, 2401–2408. [Google Scholar] [CrossRef]
  21. He, L.; Zhu, Q.; Yang, Y.; Li, Y.; Wang, X. Intrinsic peroxidase-mimicking activity of RuPDA at neutral pH and its biothiol sensing application. Microchem. J. 2025, 220, 116546. [Google Scholar] [CrossRef]
  22. Chen, Z.; Li, S.; Guan, Y.; Wu, C.; Qian, Y.; Zhou, H.; Qian, Y.; Yue, Y.; Yue, W. Ultrasmall CuMn-His nanozymes with multienzyme activity at neutral ph: Construction of a colorimetric sensing array for biothiol detection and disease identification. ACS Appl. Mater. Interfaces 2024, 16, 34538–34548. [Google Scholar] [CrossRef] [PubMed]
  23. Zeng, Y.; Owens, P.; Xia, Z.; Dos Santos, P.C.; King, S.B. Differential reactivity of bacillithiol, mycothiol and glutathione with nitroxyl (HNO): Structural determinants and potential biological implications. Redox Biol. 2025, 88, 103915. [Google Scholar] [CrossRef] [PubMed]
  24. Cuevasanta, E.; Lange, M.; Bonanata, J.; Coitiño, E.L.; Ferrer-Sueta, G.; Filipovic, M.R.; Alvarez, B. Reaction of hydrogen sulfide with disulfide and sulfenic acid to form the strongly nucleophilic persulfide. J. Biol. Chem. 2015, 290, 26866–26880. [Google Scholar] [CrossRef] [PubMed]
  25. Tang, M.; Ni, J.; Yue, Z.; Sun, T.; Chen, C.; Ma, X.; Wang, L. Polyoxometalate-nanozyme-integrated nanomotors (POMotors) for self-propulsion-promoted synergistic photothermal-catalytic tumor therapy. Angew. Chem. 2024, 136, e202315031. [Google Scholar] [CrossRef]
  26. Chen, R.; Ma, X.H.; Luo, P.; Gong, C.H.; Sun, J.J.; Si, Y.B.; Dong, X.Y.; Pan, F.; Zang, S.Q. Atomically Precise Ternary Cluster: Polyoxometalate Cluster Sandwiched by Gold Clusters Protected by N-Heterocyclic Carbenes. Angew. Chem. 2024, 136, e202408310. [Google Scholar] [CrossRef]
  27. Zhang, Y.; Chen, J.; Razq, F.; Su, C.; Hou, X.; Huang, W.; Zhang, H. Polyoxometalate-incorporated host-guest framework derived layered double hydroxide composites for high-performance hybrid supercapacitor. Chin. J. Chem. 2023, 41, 75–82. [Google Scholar] [CrossRef]
  28. Wang, Z.D.; Bo, K.; Zhong, C.L.; Xin, Y.H.; Lu, G.L.; Sun, H.; Liang, S.; Liu, Z.N.; Zang, H.Y. Multifunctional Polyoxometalates-Based Ionohydrogels toward Flexible Electronics. Adv. Mater. 2024, 36, 2400099. [Google Scholar] [CrossRef]
  29. Zhang, J.; Zhou, W.; Hao, Y.; Wang, M.; Chen, W. Cation-Induced Construction of Polyoxometalate Nanomaterials with Enhanced Surface Potential for Wearable Smart Monitoring. Adv. Funct. Mater. 2026, 36, e16341. [Google Scholar] [CrossRef]
  30. He, X.; Yao, Y.; Zhang, M.; Zhou, Y.; Zhang, L.; Ren, Y.; Dong, K.; Tang, H.; Nan, J.; Zhou, X. Engineered PW12-polyoxometalate docked Fe sites on CoFe hydroxide anode for durable seawater electrolysis. Nat. Commun. 2025, 16, 5541. [Google Scholar] [CrossRef]
  31. Zhang, X.-C.; Hou, L.; Cai, H.; Zhang, J.-M.; Chen, F.-Z.; Peng, J.; Zhao, W.-W. Synergetic enzyme-incorporated metal–organic framework and polyoxometalate nanozyme: Achieving stable tandem catalysis for organic photoelectrochemical transistor bioanalysis. Anal. Chem. 2024, 96, 16355–16361. [Google Scholar] [CrossRef]
  32. Li, M.; Wang, H.; Zhang, S.; Zhang, R.; Wang, Z. Metal–organic framework-derived FeCoZn-NC-1.5-800 as a dual-enzyme mimic for l-cysteine detection. Anal. Methods 2025, 17, 7662–7669. [Google Scholar] [CrossRef] [PubMed]
  33. Guo, J.; Han, H.; Zhao, H.; Jia, D.; Yin, L.; Sha, J. Cascade-enhanced based-polyoxometalates nanozyme for glutathione detection and tumor cell disruption. Talanta 2025, 291, 127890. [Google Scholar] [CrossRef]
  34. Su, Y.; Han, F.; Chen, H.; Li, Y.; Sun, L.; Liang, L.; Lu, L.; Hui, Y.; Li, X.; Sun, B. Bio-Orthogonally Labeled Orbitals-Hybridizing Nanozymes Induce Tissue-Resident NK Cells for Enhanced Photo-Immunologic Therapy of Biofilms and Preventing Secondary Infection. Adv. Mater. 2026, 38, e03815. [Google Scholar] [CrossRef] [PubMed]
  35. Zhao, F.; Cheng, T.; Lu, X.; Ghorai, N.; Yang, Y.; Geletii, Y.V.; Musaev, D.G.; Hill, C.L.; Lian, T. Charge transfer mechanism on a cobalt-polyoxometalate-TiO2 photoanode for water oxidation in acid. J. Am. Chem. Soc. 2024, 146, 14600–14609. [Google Scholar] [CrossRef]
  36. Ren, W.; Zhu, B.; Li, W.; Gao, M.; Luo, C.; Dazhen, L.; You, J.; Lu, M.; Wu, S.; Zhang, C. Hierarchically Porous Iontronic Fiber-Based Sensor Array with Ultrahigh Sensitivity for Stroke Rehabilitation Assistance. Adv. Funct. Mater. 2025, 36, e20066. [Google Scholar] [CrossRef]
  37. Wang, L.; Chen, Y.; Ji, Y.; Wang, L.; Liu, X.; Wang, F.; Li, C. Nanozyme-inhibited SERS multichannel paper-based sensor array for the quantification and identification of biothiols and cancer cells based on three Ag-based nanomaterials. Anal. Chem. 2024, 96, 11353–11365. [Google Scholar] [CrossRef]
  38. Lv, H.; Ma, X.; Zhang, G.; Wang, H.; Hai, X.; Bi, S. A bimetallic peroxidase-mimicking nanozyme with antifouling property for construction of sensor array to identify phosphoproteins and diagnose cancers. Biosens. Bioelectron. 2024, 258, 116370. [Google Scholar] [CrossRef] [PubMed]
  39. Hong, Y.; Birse, N.; Quinn, B.; Li, Y.; Jia, W.; McCarron, P.; Wu, D.; da Silva, G.R.; Vanhaecke, L.; van Ruth, S. Data fusion and multivariate analysis for food authenticity analysis. Nat. Commun. 2023, 14, 3309. [Google Scholar] [CrossRef]
  40. Agrahari, S.; Singh, A.K.; Tiwari, I.; Srikrishna, S. Electrochemical deposition of Pd on MoS2 supported graphene nanoflakes over functionalised MWCNTs as an immunosensor for detection of lung cancer biomarker. Microchem. J. 2024, 196, 109698. [Google Scholar] [CrossRef]
  41. Liang, Y.; Sun, F.; Qu, S.; Zhou, X.; Shang, L. Ligand density-optimized peroxidase-like activity of gold nanoclusters for colorimetric sensing of biothiols and acetylcholinesterase. Sens. Actuators B Chem. 2024, 417, 136069. [Google Scholar] [CrossRef]
  42. Wang, Y.; Huang, K.; Wang, T.; Liu, L.; Yu, F.; Sun, W.; Yao, W.; Xiong, H.; Liu, X.; Jiang, H. Nanosensors monitor intracellular GSH depletion: GSH triggers Cu (II) for tumor imaging and inhibition. Small 2024, 20, 2310300. [Google Scholar] [CrossRef] [PubMed]
  43. Diao, W.; Zhou, C.; Zhang, Z.; Cao, Y.; Li, Y.; Tang, J.; Liu, G. EGaIn-modified ePADs for simultaneous detection of homocysteine and C-reactive protein in saliva toward early diagnosis of cardiovascular disease. ACS Sens. 2024, 9, 4265–4276. [Google Scholar] [CrossRef] [PubMed]
  44. Veeranna, V.; Zalawadiya, S.K.; Niraj, A.; Pradhan, J.; Ference, B.; Burack, R.C.; Jacob, S.; Afonso, L. Homocysteine and reclassification of cardiovascular disease risk. J. Am. Coll. Cardiol. 2011, 58, 1025–1033. [Google Scholar] [CrossRef] [PubMed]
Sensors 26 03175 sch001
Scheme 1. (A) Schematic illustration of the synthesis of Fe4P2W18 nanozyme. (B) Fe4P2W18 nanozyme for biothiol determination via a pKa-driven and enrichment-synergistic strategy. (C) Construction of a biothiol sensing array based on the Fe4P2W18 nanozyme for the analysis of GSH, Cys, and Hcy.
Scheme 1. (A) Schematic illustration of the synthesis of Fe4P2W18 nanozyme. (B) Fe4P2W18 nanozyme for biothiol determination via a pKa-driven and enrichment-synergistic strategy. (C) Construction of a biothiol sensing array based on the Fe4P2W18 nanozyme for the analysis of GSH, Cys, and Hcy.
Sensors 26 03175 sch001
Sensors 26 03175 g001
Figure 1. (A) Effect of different feeding ratios on the catalytic performance of Fe4P2W18. (B) SEM image of Fe4P2W18 nanozymes. (C) XRD pattern of Fe4P2W18 nanozymes. (D) Zeta potential spectrum of Na8[HPW9O34]·24H2O and Fe4P2W18 nanozymes. (E) FT-IR spectrum of Fe4P2W18 nanozymes. (F) FT-IR spectra of Fe4P2W18 nanozymes under different pH values for 6 h of culturing.
Figure 1. (A) Effect of different feeding ratios on the catalytic performance of Fe4P2W18. (B) SEM image of Fe4P2W18 nanozymes. (C) XRD pattern of Fe4P2W18 nanozymes. (D) Zeta potential spectrum of Na8[HPW9O34]·24H2O and Fe4P2W18 nanozymes. (E) FT-IR spectrum of Fe4P2W18 nanozymes. (F) FT-IR spectra of Fe4P2W18 nanozymes under different pH values for 6 h of culturing.
Sensors 26 03175 g001
Sensors 26 03175 g002
Figure 2. (A) XPS survey spectrum of Fe4P2W18 nanozymes. (B) High-resolution W 4f spectrum of the Fe4P2W18. (C) High-resolution Fe 2p spectrum of the Fe4P2W18. (D) HRTEM images of Fe4P2W18 nanozymes. (E) EDS mapping of Fe4P2W18 nanozymes. (F) EDS spectrum of Fe4P2W18 nanozymes in STEM HAADF mode.
Figure 2. (A) XPS survey spectrum of Fe4P2W18 nanozymes. (B) High-resolution W 4f spectrum of the Fe4P2W18. (C) High-resolution Fe 2p spectrum of the Fe4P2W18. (D) HRTEM images of Fe4P2W18 nanozymes. (E) EDS mapping of Fe4P2W18 nanozymes. (F) EDS spectrum of Fe4P2W18 nanozymes in STEM HAADF mode.
Sensors 26 03175 g002
Sensors 26 03175 g003
Figure 3. (A) High-resolution C 1s spectrum of the Fe4P2W18. (B) High-resolution K 2p spectrum of the Fe4P2W18. (C) High-resolution P 2p spectrum of the Fe4P2W18.
Figure 3. (A) High-resolution C 1s spectrum of the Fe4P2W18. (B) High-resolution K 2p spectrum of the Fe4P2W18. (C) High-resolution P 2p spectrum of the Fe4P2W18.
Sensors 26 03175 g003
Sensors 26 03175 g004
Figure 4. (A) UV-vis absorption spectra of TMB oxidation catalyzed by Fe4P2W18 nanozymes at 652 nm. (B) Effect of different concentrations of Fe4P2W18 nanozymes on catalytic activity.
Figure 4. (A) UV-vis absorption spectra of TMB oxidation catalyzed by Fe4P2W18 nanozymes at 652 nm. (B) Effect of different concentrations of Fe4P2W18 nanozymes on catalytic activity.
Sensors 26 03175 g004
Sensors 26 03175 g005
Figure 5. (A) Investigation of catalytic mechanism using TA as a fluorescent probe; (B) Electron paramagnetic resonance (EPR) spectrum confirming the generation of ·OH.
Figure 5. (A) Investigation of catalytic mechanism using TA as a fluorescent probe; (B) Electron paramagnetic resonance (EPR) spectrum confirming the generation of ·OH.
Sensors 26 03175 g005
Sensors 26 03175 g006
Figure 6. The optimization of the peroxidase-like activity of Fe4P2W18 based on the corresponding absorbance values: (A) pH, (B) temperature, (C) time, (D) H2O2 concentration, and (E) TMB concentration. The maximum point in each curve was set to 100%, and the error bars represent the standard deviation derived from three independent measurements.
Figure 6. The optimization of the peroxidase-like activity of Fe4P2W18 based on the corresponding absorbance values: (A) pH, (B) temperature, (C) time, (D) H2O2 concentration, and (E) TMB concentration. The maximum point in each curve was set to 100%, and the error bars represent the standard deviation derived from three independent measurements.
Sensors 26 03175 g006
Sensors 26 03175 g007
Figure 7. (A,B) Steady-state kinetic studies of Fe4P2W18 with H2O2 as the substrate. (C,D) Steady-state kinetic studies of Fe4P2W18 with TMB as the substrate. Error bars represent the standard deviation obtained from three parallel experiments.
Figure 7. (A,B) Steady-state kinetic studies of Fe4P2W18 with H2O2 as the substrate. (C,D) Steady-state kinetic studies of Fe4P2W18 with TMB as the substrate. Error bars represent the standard deviation obtained from three parallel experiments.
Sensors 26 03175 g007
Sensors 26 03175 g008
Figure 8. Simulation of the maximum adsorption capacity of Fe4P2W18 for biothiols based on the Langmuir adsorption model.
Figure 8. Simulation of the maximum adsorption capacity of Fe4P2W18 for biothiols based on the Langmuir adsorption model.
Sensors 26 03175 g008
Sensors 26 03175 g009
Figure 9. (A) Colorimetric response patterns (A0-A) of three thiols (100 μM); (B) Heat map of absorbance changes (A0-A) for three thiols (100 μM); (C–F) Two-dimensional linear discriminant analysis (LDA) plots of the responses to three thiols at different concentrations (100 μM, 10 μM, 1 μM, 0.1 μM) using the nanozyme-based sensor.
Figure 9. (A) Colorimetric response patterns (A0-A) of three thiols (100 μM); (B) Heat map of absorbance changes (A0-A) for three thiols (100 μM); (C–F) Two-dimensional linear discriminant analysis (LDA) plots of the responses to three thiols at different concentrations (100 μM, 10 μM, 1 μM, 0.1 μM) using the nanozyme-based sensor.
Sensors 26 03175 g009
Sensors 26 03175 sch002
Scheme 2. Schematic illustration of the detection mechanism for enrichment of biothiols (GSH, Cys, Hcy) onto the negatively charged Fe4P2W18 nanozyme surface. Differential responses arise from distinct net molecular charges and sizes governed by the protonation states of a-amino, a-carboxyl, and side-chain groups.
Scheme 2. Schematic illustration of the detection mechanism for enrichment of biothiols (GSH, Cys, Hcy) onto the negatively charged Fe4P2W18 nanozyme surface. Differential responses arise from distinct net molecular charges and sizes governed by the protonation states of a-amino, a-carboxyl, and side-chain groups.
Sensors 26 03175 sch002
Sensors 26 03175 g010
Figure 10. The Fe4P2W18 nanozyme sensor was used for quantitative detection of (A,B) GSH, (C,D) Hcy, and (E,F) Cys, respectively. From left to right, the two-dimensional linear discriminant analysis plots and the linear relationship plots between concentration and LDA1 are shown.
Figure 10. The Fe4P2W18 nanozyme sensor was used for quantitative detection of (A,B) GSH, (C,D) Hcy, and (E,F) Cys, respectively. From left to right, the two-dimensional linear discriminant analysis plots and the linear relationship plots between concentration and LDA1 are shown.
Sensors 26 03175 g010
Sensors 26 03175 g011
Figure 11. Two-dimensional linear discriminant analysis (LDA) plots of binary mixtures (A,C) and ternary mixtures (B,D) of different thiols.
Figure 11. Two-dimensional linear discriminant analysis (LDA) plots of binary mixtures (A,C) and ternary mixtures (B,D) of different thiols.
Sensors 26 03175 g011
Sensors 26 03175 g012
Figure 12. (A) Two-dimensional linear discriminant analysis (LDA) plot of three cell types based on different GSH concentrations using the sensor; (B) Violin plot of the recognition results for three cell types; (C) Hierarchical cluster analysis (HCA) dendrogram of three cell types at different GSH concentrations; (D) Two-dimensional LDA plot of serum samples with different Hcy concentrations; (E) Violin plot of the recognition results for four serum samples; (F) HCA dendrogram of serum samples with different Hcy concentrations.
Figure 12. (A) Two-dimensional linear discriminant analysis (LDA) plot of three cell types based on different GSH concentrations using the sensor; (B) Violin plot of the recognition results for three cell types; (C) Hierarchical cluster analysis (HCA) dendrogram of three cell types at different GSH concentrations; (D) Two-dimensional LDA plot of serum samples with different Hcy concentrations; (E) Violin plot of the recognition results for four serum samples; (F) HCA dendrogram of serum samples with different Hcy concentrations.
Sensors 26 03175 g012
Table 1. Elemental Composition of Fe4P2W18 from XPS Survey and Narrow-Scan Analysis.
Table 1. Elemental Composition of Fe4P2W18 from XPS Survey and Narrow-Scan Analysis.
ElementPeak BE (eV)FWHM (eV)Atomic %Assignment
C 1s284.441.4086.05Adventitious carbon (charge reference)
K 2p292.731.407.18K+ counterion
W 4f35.16 (W 4f7/2)1.153.56W6+ in POM framework
Fe 2p711.00 (Fe 2p3/2)3.091.88Fe3+ in POM framework
P 2p133.541.641.32P5+ in POM framework
Table 2. The maximum reaction velocity (Vmax) and Michaelis−Menten constants (Km) of Fe4P2W18 nanozymes.
Table 2. The maximum reaction velocity (Vmax) and Michaelis−Menten constants (Km) of Fe4P2W18 nanozymes.
NanozymeSubstrateVmax (×10−8 M·s−1)Km (mM)
Fe4P2W18H2O233.250.44
Natural HRPH2O28.713.70
Table 3. Confusion Matrix of LDA Classification for GSH, Cys, and Hcy Differentiation.
Table 3. Confusion Matrix of LDA Classification for GSH, Cys, and Hcy Differentiation.
Concentration (μM)Actual ClassPredicted as GSHPredicted as CysPredicted as HcyTotalAccuracy
100GSH3003100%
100Cys0303100%
100Hcy0033100%
100Subtotal 9100%
10GSH3003100%
10Cys0303100%
10Hcy0033100%
10Subtotal 9100%
1GSH3003100%
1Cys0303100%
1Hcy0033100%
1Subtotal 9100%
0.1GSH3003100%
0.1Cys0303100%
0.1Hcy0033100%
0.1Subtotal 9100%
Overall————36100%
Table 4. Confusion Matrix and Cross-Validation Results for GSH, Cys, and Hcy Discrimination.
Table 4. Confusion Matrix and Cross-Validation Results for GSH, Cys, and Hcy Discrimination.
Concentration (μM)Standard LDA AccuracyLOOCV Accuracy
100100% (9/9)100% (9/9)
10100% (9/9)100% (9/9)
1100% (9/9)100% (9/9)
0.1100% (9/9)100% (9/9)
Overall100% (36/36)100% (36/36)
Table 5. Analytical Performance of the Fe4P2W18-Based Colorimetric Sensor Array.
Table 5. Analytical Performance of the Fe4P2W18-Based Colorimetric Sensor Array.
Performance MetricConditionResultAcceptance Criterion
StabilityRoom temperature, 7 days98.4% activity retained≥90%
Reproducibility (RSD)5 batches, GSH 10 μM2.1%≤5%
5 batches, Cys 10 μM1.9%
5 batches, Hcy 10 μM3.3%
Repeatability (RSD)10 measurements, GSH 10 μM1.8%≤3%
SelectivityGSH 10 μM in mixture98.5%90–110%
Cys 10 μM in mixture97.2%
Hcy 10 μM in mixture96.8%
Table 6. LOOCV results for the LDA1–Concentration Calibration Models of Individual Biothiols.
Table 6. LOOCV results for the LDA1–Concentration Calibration Models of Individual Biothiols.
AnalyteTSSRSSRMSEP (LDA1)RMSEP (μM)RMSEP (%)
GSH1.74150.01040.0511.63.3
Cys1.53080.01220.0551.83.7
Hcy1.33360.01200.0552.04.1
Table 7. Comparison of detection limit, accuracy rate, and strategy with recently reported sensors for biothiol measurement.
Table 7. Comparison of detection limit, accuracy rate, and strategy with recently reported sensors for biothiol measurement.
NanozymeDetection Limit (GSH)Detection Limit (Cys)Accuracy RateStrategyRef.
Fe4P2W180.1 μM0.1 μM100%Single-material pH-driven arrayThis work
CoFe2O4@CoFeCN1.5 μM1.5 μM92%5 × 5 multi-substrate array[16]
MoS2@Pd0.2 μM0.2 μM96%Oxidase-like three-channel[40]
Os/Ag NPs0.2 μM0.22 μM<100%Single-channel[19]
M-CDs5 nM5 nM<100%Dual-material array[15]
GSH-AuNCsN/A0.08 μM<100%Single-channel ligand optimization[41]
Table 8. Spike-Recovery Results for Biothiols in Human Serum and HeLa Cell Lysate.
Table 8. Spike-Recovery Results for Biothiols in Human Serum and HeLa Cell Lysate.
Sample MatrixccAnalyteSpiked (μM)Measured (μM)Recovery (%)RSD (%)
Human serumGSH1010.7107.03.5
Human serumCys109.191.04.2
Human serumHcy1010.3103.02.9
HeLa lysateGSH109.494.04.8
HeLa lysateCys1010.8108.03.6
HeLa lysateHcy109.797.03.3
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Sun, B.; Liu, Y.; Tian, X.; Fang, Y.; Mei, J.; Chen, Y.; Ma, T. A Single Polyoxometalate Nanozyme for Cross-Reactive Thiol Array Sensing via pKa-Driven and Enrichment-Synergistic Strategy. Sensors 2026, 26, 3175. https://doi.org/10.3390/s26103175

AMA Style

Sun B, Liu Y, Tian X, Fang Y, Mei J, Chen Y, Ma T. A Single Polyoxometalate Nanozyme for Cross-Reactive Thiol Array Sensing via pKa-Driven and Enrichment-Synergistic Strategy. Sensors. 2026; 26(10):3175. https://doi.org/10.3390/s26103175

Chicago/Turabian Style

Sun, Baohong, Ying Liu, Xinxin Tian, Yu Fang, Jinpei Mei, Yang Chen, and Tao Ma. 2026. "A Single Polyoxometalate Nanozyme for Cross-Reactive Thiol Array Sensing via pKa-Driven and Enrichment-Synergistic Strategy" Sensors 26, no. 10: 3175. https://doi.org/10.3390/s26103175

APA Style

Sun, B., Liu, Y., Tian, X., Fang, Y., Mei, J., Chen, Y., & Ma, T. (2026). A Single Polyoxometalate Nanozyme for Cross-Reactive Thiol Array Sensing via pKa-Driven and Enrichment-Synergistic Strategy. Sensors, 26(10), 3175. https://doi.org/10.3390/s26103175

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop