Heavy Metal Ion Detection by Carbonized Metal–Organic–Framework (MOF-C) Nanocomposite-Modified Electrochemical Sensors

Abstract

Efficient detection of heavy metal ions in complex marine environments is essential to the safety of marine organisms and human beings. This study developed a novel screen-printed-electrode (SPE) electrochemical sensor for rapid on-site determination of typical heavy metal ions such as Cu²⁺, Pb²⁺, Cd²⁺, and Hg²⁺ in seawater. The sensor employs a three-electrode system, with the working electrode modified with a composite of metal–organic framework-derived carbon (MOF-C) and multiwalled carbon nanotubes (MWCNTs), thereby significantly enhancing detection sensitivity and selectivity. By optimizing square-wave anodic stripping voltammetry (SWASV) parameters, detection limits of 0.83, 0.40, 1.05, and 0.30 μM for the detection of Cu²⁺, Pb²⁺, Cd²⁺, and Hg²⁺ ions were achieved. In mixed-ion detection, excellent peak separation and strong resistance to interferences were demonstrated. Experimental results demonstrate that the sensor exhibits good linear response, excellent interference resistance, and high practicality, providing a new approach for rapid on-site determination of heavy metal pollution in marine environments.

1. Introduction

Heavy metal ions such as lead (Pb²⁺), mercury (Hg²⁺), and cadmium (Cd²⁺) enter the aquatic environment through multiple pathways, including industrial wastewater discharge, ship antifouling coatings, ballast water transport and accumulation via the water cycle from land-based pollution [1,2]. These ions exhibit substantial toxicity, poor degradability, and high bioaccumulation potential, posing severe threats to marine ecosystems and human health [3,4,5,6]. They can readily accumulate in marine organisms and the human body through the food chain. Long-term exposure to heavy metal environments may lead to widespread and profound health hazards to the nervous, blood, and digestive systems, subsequently causing reduced immune function and immune system disorders, triggering major pollution incidents like Japan’s Minamata disease [7,8,9,10,11,12,13,14,15,16,17]. The development of highly efficient and sensitive heavy metal ion detection technologies for real-time monitoring and early warning of marine pollution is urgently needed for ecological security and sustainable development.

Currently, the most commonly used methods for heavy metal ion detection primarily involve high-end instruments such as atomic absorption spectroscopy (AAS), atomic fluorescence spectroscopy (AFS), and inductively coupled plasma mass spectrometry (ICP-MS) [18,19,20,21,22]. While these methods offer high sensitivity and accuracy [23,24,25,26,27,28,29,30,31,32], they generally suffer from significant drawbacks, including high cost, bulky instrumentation, limited portability, and cumbersome sample pretreatment processes. There is still a lack of convenient, rapid methods for detecting heavy metal ions, especially in complex aquatic environments with high precision. Electrochemical sensors, with their high sensitivity, fast response, and low cost, have been widely used to detect trace amounts of pollutants, particularly heavy metal ions [33].

Detection of typical toxic heavy metal ions in marine environments such as Cu²⁺, Pb²⁺, Cd²⁺, Hg²⁺, and common coexisting interfering ions like Zn²⁺, Cr³⁺, and As³⁺ by the electrochemical method has been reported [34], and many nanomaterials have been used to improve the performance of the sensors due to their three-dimensional nanostructure networks at the electrode/electrolyte interface, featuring high specific surface area, excellent charge-transport capability, and tunable surface chemistry [35]. For example, Bismuth Tungstate Nanocrystals (Bi₂WO₆) [36], metal–organic frameworks (MOFs) [37] and multiwalled carbon nanotubes (MWCNTs) [38,39,40] have been employed to functionalize electrode surfaces of ion sensors [41]. These materials enhance the performance of the sensors in different ways. For instance, Bi₂WO₆ nanocrystals exhibit outstanding electrocatalytic activity due to their layered semiconductor structure, which enables the catalysis of redox reactions for target molecules at low overpotentials while amplifying current signals [42]. Metal–organic frameworks (MOFs) and their derivatives possess porous structures and abundant coordination sites yet suffer from inherently low conductivity [43]. Metal/metal oxide nanomaterials can reduce overpotentials and enhance signal intensity through alloying effects but face challenges of agglomeration and insufficient stability [44]. MWCNT networks with high conductivity can form efficient electronic “highways” and are suitable for adsorbing catalytic carriers; their accelerated response rates make them ideal for use as electrode materials. However, MWCNT materials face challenges of insufficient selectivity, stability, and lifespan.

The development and optimization of new nanomaterial compositions have been used to improve the performance of electrochemical sensors for heavy metal ions. The combination of MOFs and MWCNTs would form a continuous conductive network in which MOFs provided pores and coordination sites for target ion adsorption and MWCNTs enhance the conductivity of the sensing electrodes and sensitivity of the sensors [45]. MOF and MWCNT nanocomposites for the detection of metal ions have been reported previously [46]. However, previous works only focus on either one specific ion detection, like Pb²⁺ [47], or two kinds of ions but using bulky sensing systems [48,49]. The use of electrochemical sensors for detecting multiple ions simultaneously in real seawater samples remains challenging, and investigations into sensing stability, selectivity, and practicality under the conditions of complex aquatic environments remain unexplored. On-site, real-time detection applications that require enhanced interference resistance in high-salinity seawater matrices are urgently needed.

This paper proposes investigating an SPE electrochemical sensor for the determination of coexisting marine heavy metal ions. A Carbonized Metal–Organic Framework–MWCNT (MOF-C@MWCNT) nanocomposite was functionalized on the Screen-Printed Electrode (SPE) to enhance sensor conductivity and specificity. Working parameters of the electrochemical sensor for detecting four typical heavy metal ions, i.e., Cu²⁺, Pb²⁺, Cd²⁺, and Hg²⁺, were optimized to improve the sensitivity and selectivity. The sensor’s performance was verified by detecting seawater samples spiked with target ions, as well as real samples obtained from different coastal locations.

2. Materials and Methods

2.1. Reagents and Materials

All reagents were used as received. Ion solutions of different concentrations were prepared as required by using dry reagent powder or standard ion solutions, in which copper sulfate (CuSO₄), magnesium chloride (MgCl₂), calcium chloride (CaCl₂), and potassium chloride (KCl) were purchased from Sinopharm Reagents (Shanghai, China), and standard lead nitrate titration solution, cadmium solution, mercury solution, and zinc chloride (ZnCl₂) solution were obtained from Tianjin Comeo Chemical Reagent Co., Ltd. (Tianjin, China). The polyvinyl chloride (PVC) substrate (Pinrui Plastic Industry, Wenzhou, China), MOF-C (Guangdong Tanyu New Materials Co., Ltd., Zhuhai, China), flower-shaped bismuth tungstate nanoparticles (Jiangsu Xianfeng Nanotechnology Co., Ltd., Suzhou, China), MWCNTs (Guangdong Kaisa Group New Materials Co., Ltd., Shenzhen, China), Ag/AgCl ink (GIE-4250), and carbon ink (G1-2067) purchased from Guangzhou Yinbiao Trading Co., Ltd. (Guangzhou, China) were applied for printing the sensor electrodes, and ethanol (Sinopharm Reagents) and slow-drying water (Tuye Sanx, Taiyuan, China) were used as cleaning solutions during the screen-printing process. Artificial seawater samples were ordered from the National Marine Standard Substance Center (Hangzhou, China). Real seawater samples obtained from Dalian, Zhenjiang, and Maoming were pretreated with 0.22 μm filter membranes and syringes to remove impurities. To enable high-selectivity detection of heavy metal ions, Nafion solution (DuPont, Wilmington, DE, USA) was applied to create cation-selective channels on the working electrodes. Potassium ferricyanide (Aladdin, Shanghai, China) was used as a substrate to optimize and verify the electrochemical sensors.

2.2. Design and Preparation of Electrochemical Sensors

The working steps for the design and preparation of the electrochemical sensor are described briefly as follows: firstly, the layout of the sensor was designed by AutoCAD 2018 software, followed by the creation of the screen-printing stencils by photolithography techniques; secondly, the electrodes of the sensor and the insulating layer were printed on the PVC substrate, sequentially followed by the functionalization of nanomaterials on the working electrode. PVC was chosen for its excellent hydrophobicity, biocompatibility, cost-effectiveness, ease of handling, and environmental friendliness.

Note that before screen-printing, carbon ink was diluted with a slow-drying solution at a 1:10 mass ratio to achieve the desired viscosity. After printing the carbon working electrode (WE) and counter electrode (CE), the devices were heated at 65 °C for 30 min to remove the solvent. The reference electrode (RE) was printed on the sensor using a fine-screen stencil, followed by a second heating step at 65 °C for 30 min to dry the RE. To achieve effective signal transmission, the printed electrodes were 5~10 μm thick. To ensure a stable potential response and durability across a wide potential window, Ag/AgCl ink was selected for printing RE. Finally, the three-electrode electrochemical sensor was encapsulated with a confined reaction zone of 5 mm in diameter using a photocurable insulating layer, which was irradiated with a 180 W UV lamp at 365 nm for 5 min, as shown in Figure S1 of the Supplementary Information (SI-1). The screen-printing method is cost-effective, and 105 chips can be produced for a few USD in an hour.

2.3. Functionalization and Optimization of the WE

To enhance the sensor’s detection limits and selectivity, the WE of the printed sensors was functionalized with MWCNTs and MOF-C nanomaterials. The modification was performed as follows: first, a solution was prepared by mixing 0.02 g of MOF-C with 2 mL of solvent (1:1 ethanol/deionized water) then adding 20 μL of Nafion solution as a binder. After mixing, the solution was sonicated for 20 min at 25 °C to achieve uniform dispersion and then pipetted onto the WE surface. To achieve full coverage of the WE, 5 μL solution was applied to each WE. The electrode was then heated at 50 °C for 10 min, then dried at room temperature for 20 min. This process allowed Nafion and the nanomaterials to form a continuous thin film, enhancing adhesion to the electrode and yielding the modified nanomaterial electrode chip. In addition, Bi₂WO₆ and MWCNTs were applied to modify the working electrode, in which 2 mg of Bi₂WO₆ or MWCNTs were dispersed in 1 mL of ethanol solution containing 0.1% Nafion and sonicated for 30 min to form a uniform suspension, followed by dip-coating the solution on the WE.

2.4. Data Acquisition and Processing

2.4.1. Parameter Setting and Data Acquisition

To characterize the performance of the SPE sensors, square-wave anodic stripping voltammetry (SWASV) and cyclic voltammetry (CV), which have been commonly used for electrochemical mechanism studies, were applied in this study [50,51,52]. Unless otherwise stated, CV scanning in the range of −0.8~1.0 V with a scan rate between 50 and 300 mV/s under the condition of 1 μA/V signal sensitivity was applied. All electrochemical signals were acquired using a portable PalmSens4 electrochemical workstation paired with a three-electrode system for the SPE electrochemical sensor. Before the experiments, the workstation was calibrated with K₃[Fe(CN)₆]/K₄[Fe(CN)₆] standard solution, which was freshly prepared before each experiment by dissolving 10 mM (K₃Fe(CN)₆) and 10 mM (K₄Fe(CN)₆) in 0.1 M KCl solution. Real-time current responses were recorded throughout the scans to assess electrode performance. All tests were conducted at a constant temperature of 25 ± 1 °C and a humidity of 50 ± 5% RH. Electrochemical Impedance Spectroscopy (EIS) in the range of 0.1~0.1 MHz was conducted by Randles to evaluate the conductivity of electrochemical sensors.

Square-wave anodic stripping voltammetry (SWASV) was employed in this study to quantify heavy metal ions because it combines enrichment and stripping of the target ions. Deposition potential and scanning parameters were optimized to achieve effective separation of multiple ion peaks with high specificity. For single heavy metal ion SWASV parameter optimization, an enrichment time of 120 s, square-wave amplitude of 25 mV, frequency of 15 Hz, and scan range of −0.2~0.5 V were applied during Hg²⁺ detection. Stripping peak currents were used to quantify the concentration of target ions. The detection procedures for the other three ions were the same but with different scanning ranges adjusted based on preliminary stripping potentials (−0.2~0.1 V, −0.8~−0.2 V, −1.0~−0.5 V, for Cu²⁺, Pb²⁺, and Cd²⁺, respectively). For mixed heavy metal ion detection, samples with ion concentration gradients of 1, 5, 8, 10, 50, and 80 μM were tested, and the working conditions of −0.60 V deposition voltage with an enrichment time of 150 s and a 15 Hz square-wave amplitude of 25 mV with a scanning range of −1.0~0.5 V were applied. To detect seawater, both spiked artificial seawater and real seawater were used. Because four metal ions are detected simultaneously in the mixed solution, −0.6 V is uniformly used as the deposition voltage. The SWASV operating parameters were the same as those used for detecting mixed ions. Each sample underwent three parallel measurements to ensure data reproducibility. After each SWASV scanning, the electrode was electrochemically cleaned by applying +0.80 V (vs. Ag/AgCl) for 60 s in the supporting electrolyte, followed by rinsing with DI water and drying, to minimize memory effects from residual metal deposition.

2.4.2. Raw Data Processing

The amperometry curves were processed using baseline correction and smoothing to improve visual comparison with commercial software (PSTrace). The baseline interval was selected within the potential range, avoiding significant elution peaks. The “Peak Analysis” function was applied to identify and extract key parameters of the stripping peaks, namely, peak potential E_p, peak current I_p and half-peak width W₁/₂. Note that peaks with a signal-to-noise ratio (S/N) < 5 were manually verified to avoid misidentification of false peaks [53]. Peak current data were exported and preprocessed using Origin software.

2.4.3. Quantitative Calculation and Statistical Analysis

During quantitative analysis, for each individual heavy metal ion, a linear fitting of the elution peak current (I_p) with the ion concentration (C, μM) was conducted. The limit of detection (LOD) of the sensor was calculated by using the 3σ method; i.e., LOD = 3σ/k, where σ is the standard deviation of peak current measurements from 10 blank samples, and k is the slope of the corresponding heavy metal ion calibration curve. For the detection of seawater samples, the spiked recovery rate was calculated using the following formula:

$$R\left(\%\right)={\displaystyle \frac{C_{found}-C_0}{C_{spike}}}\times 100\%$$

(1)

where $C_0$ is the concentration of the target heavy metal ion in the sample to be detected, $C_{found}$ is the target heavy metal ion concentration in the spiked sample, and $C_{spike}$ is the concentration of target heavy metal ions added to the sample. During statistical analysis, all experimental data were presented as “mean ± standard deviation” (mean ± SD, n = 3). The repeatability of the detection method was evaluated by calculating the relative standard deviation (RSD).

3. Results and Discussion

3.1. Device Fabrication and Characterization

The electrochemical sensor was fabricated using the screen-printing method, as previously reported [54]. The working procedure and the design of the chip can be found in Figures S1 and S2 of the Supplementary Information (SI). PVC film serves as the substrate due to its excellent corrosion resistance and insulation properties, capable of withstanding most chemical corrosions while maintaining highly stable physicochemical characteristics. The working steps are briefly described as shown in Figure 1A. The fabrication and modification of the SPE sensor involved sequentially printing layers of carbon electrodes (WE and CE), a Ag/AgCl electrode (RE), and a hydrophobic insulating layer onto the flexible PVC substrate to construct a confined sensing area for the testing. Thereafter, the WE of the sensor was functionalized by dip-coating with the MWCNT/MOF-C composite and then drying the device. At this stage, the sensor is ready to use. Figure 1B illustrates the operating mechanism of the sensor for detecting heavy metal ions, with Hg²⁺ as an example. The central circular carbon WE serves as a heavy metal ion enrichment electrode, and the outer carbon CE ensures a symmetrical electric-field distribution and efficient electrolyte exchange, while the Ag/AgCl RE provides spatial utilization and potential stability. During testing, an SWASV signal is applied between WE and CE. Hg²⁺ is first preconcentrated and reduced onto the WE surface by electrodeposition and then stripped back into solution during the applied potential sweep. The resulting current generated during the preconcentration and stripping step is recorded. The amplitude of the stripping current is proportional to the concentration of Hg²⁺ in the sample; consequently, the current can be used to quantify trace amounts of the target Hg²⁺ [55,56].

Figure 1C illustrates the screen stencils used for screen-printing, and Figure 1D shows an example of the printed electrochemical sensor array, with multiple sensors printed on PVC film. The screen-printing method offers significant process advantages for fabricating SPE sensors, including large-scale production and low cost. The high-precision screens enable high-resolution patterning, and multiple coating steps further enhance the uniformity and adhesion of the conductive ink layer as well as batch-to-batch consistency. Figure 1E presents a schematic of the testing system, in which a sensor is clamped in a three-electrode clip and an analyte solution is loaded onto the sensing area; an example of the resulting SWASV signal is shown on the right-hand side of the panel. The peak current is collected to calibrate and characterize the concentration of target ions in the sample. The chip electrode system modified with MOF-C@MWCNTs, which exhibited the strongest modification effect, was characterized using scanning electron microscopy (SEM) and Raman spectroscopy. Figure 1F illustrates an example of the WE surface after the modification of MOF-C@MWCNT/SPE, indicating the success of the modification of porous nanomaterial composite on the surface. Figure 1G–I shows the Raman spectrum of the MWCNTs (Figure 1G), MOF-C (Figure 1H), and the chip modified with MOF-C@MWCNT composite (Figure 1I), respectively.

The high-temperature carbonized MOF-C material exhibits D and G peaks at approximately 1350 cm⁻¹ and 1580 cm⁻¹, respectively, attributed to sp³-hybridized carbon-defect vibrations and sp²-graphitic-structure vibrations. The ratio of D peak intensity to G peak intensity is ~1.5. MWCNTs exhibit characteristic D and G peaks at 1350 cm⁻¹ and 1580 cm⁻¹, respectively, with a peak intensity ratio of ~3. The peak observed at 1350 cm⁻¹ in the MOF-C@MWCNT/SPE indicates the presence of numerous carbonized edges on the electrode surface. The ratio of the peak intensity at 1350 cm⁻¹ to that at 1580 cm⁻¹ is ~1.8, reflecting the differing MOF-C-to-MWCNT ratios on the electrode surface and resulting in a varying D/G peak intensity ratio. These peak positions confirm that the MOF-C@MWCNT composite has been successfully deposited onto the electrode surface. The characterization also supports the prior hypothesis that the MOF-C@MWCNT composite significantly outperforms individual nanomaterials in conductivity. The MOF-C@MWCNT-modified SPE constitutes a highly integrated sensing platform that enables rapid enrichment, uniform distribution, and electrochemical transformation of heavy metal ions at the working electrode surface. This rational chip design therefore provides a solid physical and electrochemical foundation for the high-performance detection of multiple heavy metal ions, as demonstrated in the following sections.

3.2. Parameter Optimization of the SPE Sensors

A K₃Fe(CN)₆/K₄Fe(CN)₆ redox system was used as the electrolyte to evaluate the effectiveness of the SPE electrochemical sensor via CV. Real-time current responses were recorded throughout the scans. As illustrated in Figure 2A, both the anodic and cathodic current peaks are highly symmetric across all scan rates, indicating a reversible redox process and good sensor performance. Figure 2B shows the linear fitting of the anodic and cathodic peak currents (I_pa and I_pc) and the square root of the scan rate. The peak potential remains nearly constant, and the resulting linear fitting coefficients (R²) are 0.997 and 0.989, respectively, indicating a diffusion-controlled electrode process. These results confirm that the SPE electrode exhibits efficient charge transfer kinetics and stable interfacial electrochemical behavior. The reliability of the SPE sensors was also verified by using commercial SPE sensors (Figure S1A), and the effectiveness of the quality control and the temporal stability of the fabricated chips were also verified by five randomly selected SPE sensors from the same batch that underwent uniformity testing (Figure S1B). More information is provided in the Supplementary Information (SI).

To identify the optimal nanomaterial for enhancing the performance of the fabricated SPE, MWCNTs, bismuth tungstate, MOF-C, and MOF-C@MWCNTs were also evaluated. CV curves and EIS curves obtained from the SPE sensors modified with these nanomaterials are compared with those of the bare SPE electrode. As shown in Figure 2C, the CV profiles indicate that the composite material combining MOF-C and MWCNTs exhibits a higher peak current than the other cases. Compared to the bare electrode, the peak current increased by approximately one-fold. This confirms the synergistic effect between MWCNTs and MOF-C. As verified in Figure 2D, the composite’s impedance lies between those of MWCNTs and MOF-C. Fitting with the Randles circuit reveals a surface resistance (Rct) of approximately 386 Ω, significantly lower than the bare electrode’s 750 Ω. This indicates efficient electron transfer at the electrode surface while maintaining selective adsorption. Consequently, in this study, MOF-C@MWCNTs are employed to modify SPE sensors.

The rational design and precise control of nanomaterial modification are pivotal in optimizing the detection performance of electrochemical sensors. In this study, the MOF-C-to-MWCNT ratio significantly influences the sensor’s conductivity and ion enrichment capability. A series of composite material groups (MOF-C: MWCNTs) was utilized to modify the chip surface and measured by CV scanning. As shown in Figure 2E, the optimal modification performance was achieved at an MOF-C: MWCNT ratio of 1:1, yielding the highest oxidation current peak of ~170 µA. This optimal ratio ensures balanced contributions from both components, thereby facilitating efficient charge-transfer pathways at the electrode surface while maximizing the number of active sites. Additionally, the volume of the deposited modification solution is another critical parameter affecting chip performance. With the MOF-C: MWCNT ratio fixed at 1:1, CV results presented in Figure 2F indicate that the peak current initially increases and then decreases as the deposition volume rises from 0.5 μL to 10 μL. The maximum peak oxidation current is observed at a deposition volume of 5 μL, reaching ~190 µA. This phenomenon can be attributed to the fact that an appropriate amount of nanomaterials enhances the electrode’s active surface area; however, excessive deposition can increase interfacial resistance and impede charge transport. This behavior arises because excessive loading of MOF-C/MWCNT produces an overly thick, densely packed film on the electrode, which increases interfacial resistance, lengthens electron-transfer pathways, and partially blocks ion diffusion channels, ultimately hindering charge transport and lowering the oxidation current, as reported in the literature [55].

3.3. Detection of Single Marine Heavy Metal Ions

As mentioned above, SWASV can detect multiple ions simultaneously at μM or even nM levels. As a result, SWASV was employed to detect Cu²⁺, Pb²⁺, Cd²⁺, and Hg²⁺ in this work. However, optimizing parameters such as the amplitude and frequency of the SWASV signal remains very tricky. For instance, exceeding amplitude thresholds will degrade sensing resolution, while improper frequencies will reduce the sensing signal-to-noise ratio [56,57]. To optimize the working parameters, herein, we take Hg²⁺ as an example. Preliminary experiments indicated that the stripping potential for Hg²⁺ is ~0.1 V. As the stripping peak positions of metal ions vary with different electrolyte/buffer systems, parameter optimization for SWASV was performed based on the preliminary deposition potential of Hg²⁺, using 0.1 M KCl as the supporting electrolyte to ensure consistent ionic strength and system stability.

This study optimized deposition voltages of Hg²⁺ at five potential values: 0 V, 0.1 V, 0.2 V, 0.3 V, and 0.4 V. The optimal deposition potential was selected based on peak current magnitude. Within the potential range of −0.2 to 0.5 V, scans were performed with an enrichment time of 120 s, a square-wave amplitude of 25 mV, and a frequency of 15 Hz. As shown in Figure 3A, the optimal deposition voltage for Hg²⁺ was ~0.2 V. Following the determination of the optimal deposition potential, the deposition time for SWASV was optimized. Deposition durations of 60 s, 90 s, 120 s, 150 s, and 180 s were tried. As shown in Figure 3B, the peaking current of depositing Hg²⁺ tends to stabilize at ~7.3 µA. To achieve sufficient Hg deposition on the WE, an optimized deposition voltage of +0.30 V and a deposition time of 60 s were used for Hg²⁺ detection.

To characterize the performance of the sensor for Hg²⁺ ion detection, 0.1 M KCl solution was selected as the supporting electrolyte, and samples containing Hg²⁺ in the range of 0.1, 0.5, 0.8, 1, 5, 8, 10, 50, 80, and 100 μM were prepared. During detection, Hg²⁺ enrichment was performed at a deposition potential of +0.30 V for 60 s, as described above. Subsequently, a square-wave stripping scan with a step size of 4 mV, amplitude of 25 mV, and 15 Hz signal was applied over a potential window of −0.2 V to 0.4 V. Figure 3D summarizes the stripping peak current for Hg²⁺ collected at +0.3 V scanning voltage, indicating that the stripping peak current increased almost continuously as the Hg²⁺ concentration increased from 0.1 to 100 μM. A linear fitting of the peak current yields an equation I_p = 1.438 C_Hg²⁺ + 0.166, with a correlation coefficient R² of 0.976, as shown in Figure 3E. After further subtracting the standard deviation of the blank current (σ = 0.142 µA), an LOD of 0.3 μM for the measurement of Hg²⁺ can be obtained. The value of σ was also used for the calibration of LOD of the other ion sensors. The sensor’s superb performance is attributed to the ion coordination in the MOF-C nanomaterial and the reduced interfacial resistance and polarization resulting from MWCNT modification [58]. To evaluate interference resistance, solutions of Mg²⁺, Cu²⁺, Zn²⁺, Cd²⁺, Pb²⁺, and Ca²⁺ at 50 times the Hg²⁺ concentration (10 μM) were spiked in the samples. As shown in Figure 3F, the measured Hg²⁺ peak current deviations were all within ±5%, indicating that, under these SWASV conditions, Hg²⁺ detection also exhibits high sensitivity and excellent interference resistance. Detection procedures for other ions follow the same steps as for Hg²⁺, and the results are demonstrated as follows. Repeated measurements of the same sensor over 7 days reveal consistent peak current responses, as illustrated in Figure 3F, with an RSD of approximately 2%, confirming the chip’s long-term stability. These performance tests confirm that the fabricated SPE electrochemical sensors exhibit outstanding electrochemical properties, thereby providing a robust platform for subsequent detection of heavy metal ions in marine samples.

Optimization of working parameters for the detection of the other ions, Cu²⁺, Cd²⁺, and Pb²⁺, was the same as that of Hg²⁺, as demonstrated in Figure S2. Figure 4A shows the results of detecting Cu²⁺, in which the peak voltage for Cu²⁺ is around −0.03 V; however, this value is more positive than those reported in the literature [59]. This is attributed to Cu²⁺ coordination in the porous surface framework, which reduces the desorption overpotential, and to accelerated electron transport facilitated by the conductive MWCNT network. It exhibits good linearity over the concentration range of 0.1 μM to 100 μM, which can be fitted by a linear equation Ip = 0.511$C_{{Cu}^{2+}}$ + 0.496 (R² = 0.976) (Figure 4B). Based on the linear fitting curve, the detection limit is ~0.83 μM, demonstrating high detection accuracy and sensitivity.

Figure 4D,G show the electrochemical signals collected during the detection of Cd²⁺ and Pb²⁺ over the 0.1~100 μM range, respectively. For both ions, peak currents show a good linear correlation with concentration. Figure 4G shows a distinct oxidation peak for Cd²⁺ ions at −0.75 V, with the stripping potential shifted by 50 mV to more positive values than predicted. This shift is likely due to the electrode surface’s chemical sites, the pore environment, and the thickness of the mass-transfer layer. Within the concentration range of 0.1~100 μM, the peak current increases linearly with the concentration, with the linear fitting equation of Ip = 0.407$C_{{Cd}^{2+}}$ + 0.436 (R² = 0.971) (Figure 4E). An LOD of 1.05 μM for the detection of Cd²⁺ can be obtained based on the equation. Similarly, a distinct oxidation peak for Pb²⁺ detection at −0.40 V is observed in Figure 4G, and a linear fit of the peak current is shown in Figure 4H. Based on the fitting curve (Ip = 1.069$C_{{Pb}^{2+}}$ + 0.249 (R² = 0.984)), an LOD of ~0.40 μM for Pb²⁺ detection can be obtained.

Interference resistance testing was also performed for the testing of Cu²⁺, Cd²⁺, and Pb²⁺, in which 0.1 M KCl solution was employed as the supporting electrolyte and interfering ions (including Zn²⁺, Hg²⁺, Mg²⁺, Ca²⁺) of 50-fold concentrations of the target ions were spiked in the samples. The results are presented in Figure 4C, Figure 4F, and Figure 4I, respectively. The results indicate that the interfering ions have little effect on the target ions’ signals, with <5% amplitude deviation. All four ions demonstrated high sensitivity, broad linear ranges (i.e., good linearity), and excellent resistance to matrix interference. Noted that cyclic voltammetry (CV) experiments were also conducted to detect heavy metal ions, and examples of the results are presented in Figure S5. The results show that unsatisfactory peak currents were obtained because only linear potential scanning was used without a concentration step. In contrast, SWASV enriches metal ions on the MOF-C@MWCNT composite layer via electrodeposition, thereby amplifying peak currents and improving detection limits.

3.4. Simultaneous Determination of Mixed Heavy Metal Ions

After completing individual detection of four typical heavy metal ions, Cu²⁺, Cd²⁺, Pb²⁺, and Hg²⁺, validation of the SPE electrochemical sensor for the detection of metal ion mixture samples was conducted. During testing, Cu²⁺, Pb²⁺, Cd²⁺, and Hg²⁺ were spiked into a 0.1 M KCl supporting electrolyte, and concentration gradients of 80 μM, 50 μM, 10 μM, 8 μM, 5 μM, and 1 μM were prepared. According to the previously optimized parameters, cadmium ions require the longest deposition time of 150 s. To accommodate the reduction enrichment requirements of all four ions (Cu²⁺, Pb²⁺, Cd²⁺, and Hg²⁺), an enrichment potential of −0.60 V for 150 s was applied to the MOF-C@MWCNTs WE. Thereafter, an SWASV scan with a 25 mV square-wave amplitude, 15 Hz frequency, and 4 mV step size was applied across the WE scanning window of −1.0~0.5 V. Current–voltage curves were recorded in real time throughout the SWASV process, with each scan taking approximately 4 min.

Figure 5 shows the results of detecting a mixture of four kinds of heavy metal ions (Cu²⁺, Cd²⁺, Pb²⁺, and Hg²⁺) spiked in KCl supporting electrolyte. Figure 5A reveals four successive peaks of the target ions from left to right: Cd²⁺ peak at ~−0.75 V, Pb²⁺ peak at ~−0.55 V, Cu²⁺ peak at ~−0.21 V, and Hg²⁺ peak at ~0.051 V. Compared to individual detection, the peaks for Cd²⁺ and Pb²⁺ show slight shifts but no significant changes. This occurs because, in a system where Cu²⁺, Pb²⁺, Cd²⁺, and Hg²⁺ coexist, their reduction enrichment at the electrode surface follows a first-come, first-served mechanism. Cu²⁺ and Hg²⁺ ions have higher reduction potentials than Cd²⁺ and Pb²⁺, resulting in thicker metal films that increase local resistance and polarization. At this stage, Cu²⁺ and Hg²⁺ ions occupy a large number of MOF sites, causing Cd²⁺ and Pb²⁺ ions to have fewer adsorption sites. Consequently, a more negative potential is required for desorption, indicating surface competition and adsorption effects [60]. The simultaneous presence of multiple metal ions reduces selectivity and results in higher detection limits than single-metal detection. In mixed-metal systems, ions compete for the limited electroactive sites during deposition, with higher-potential species such as Cu²⁺ and Hg²⁺ being preferentially reduced. Furthermore, co-deposition interferes with the stripping process, leading to peak broadening and partial overlap, thereby reducing the effective stripping current and increasing signal uncertainty. Therefore, the detection limit for mixed-metal detection is higher than that for single-metal detection [61,62]. Site competition reduces deposition rates, thereby lowering peak currents, whereas co-precipitation forms multimetallic layers that hinder electron-transfer efficiency. These synergistic effects result in lower actual deposition rates for each metal than in single-metal systems [63]. Ultimately, the simultaneous presence of multiple metal ions limits selectivity, resulting in higher detection limits in mixed-metal detection than in single-metal detection.

Figure 5B summarizes the fitting of the peak current for detecting these ions. Under mixed detection conditions, the detection limits for the four metals are 1.07 μM (Cu²⁺), 0.85 μM (Pb²⁺), 2.42 μM (Cd²⁺), and 0.35 μM (Hg²⁺), as obtained for the calculated ${\sigma }_{blank}$ = 0.142 μA and the curve slopes. Compared to the LODs in single-component detection, the detection limits for all ions increased in the mixed system. For the mixed-ion calibration, linear regression to obtain k was performed within the validated working range; any concentration points below the corresponding LOD (if shown as exploratory data) were excluded from the regression. The increase was relatively small for Cu²⁺ and Hg²⁺, approximately 1.2 to 1.3 times compared to those of the individual cases, while Pb²⁺ and Cd²⁺ increased to about 3.9 times and 2.3 times, respectively. This trend primarily stems from competitive adsorption at the electrode’s active sites during the coexistence of multiple ions, coupled with the combined effects of mass transfer and interfacial resistance. The more negative potential and longer mass-transfer path during the enrichment phase resulted in the most significant reduction in enrichment efficiency for Pb²⁺ and Cd²⁺. Conversely, Cu²⁺ and Hg²⁺ exhibit more positive reduction potentials and are less affected by competition, resulting in a smaller increase in detection limits during mixed detection. This result aligns with expectations of increased background noise and physicochemical effects during simultaneous multi-ion enrichment. Note that in seawater matrices, the presence of Cl⁻ ions leads to the formation of stable complexes with target analytes, resulting in an increase in detection limits. Although the sensor’s detection limit for mixed-ion detection remained at the μM level, it exhibited a good linear response range and strong resistance to interference, making it suitable for rapid on-site screening of heavy pollution.

3.5. Real Sample Validation and Stability

To verify the sensor’s practicability, artificial seawater was spiked with heavy metal ions and then tested with the sensor. One control group contained no spiked heavy metal ions, while other groups were spiked with standard solutions of Cu²⁺, Pb²⁺, Cd²⁺, and Hg²⁺ at concentrations of 5 μM, 10 μM, and 20 μM. The recovery ratios of testing were obtained through spiking experiments, as shown in

Table 1. Recovery ratio of heavy metal ions spiked in artificial seawater.

Group	Spiked (μM)	Detected (μM)	Recovery Rate (%)
Blank	Cd2+ = 0, Pb2+ = 0, Cu2+ = 0, Hg2+ = 0	Cd2+ = 0, Pb2+ = 0.12 ± 0.05 Cu2+ = 0, Hg2+ = 0.35 ± 0.08	None
1	Cd2+ = 5, Pb2+ = 5 Cu2+ = 5, Hg2+ = 5	Cd2+ = 4.65 ± 0.32, Pb2+ = 4.87 ± 0.25 Cu2+ = 4.78 ± 0.18, Hg2+ = 4.96 ± 0.45	93.33 95.25 95.62 92.27
2	Cd2+ = 10, Pb2+ = 10, Cu2+ = 10, Hg2+ = 10	Cd2+ = 9.87 ± 0.63, Pb2+ = 10.1 ± 0.54 Cu2+ = 9.87 ± 0.65, Hg2+ = 10.2 ± 0.88	98.73 99.12 98.23 98.43
3	Cd2+ = 20, Pb2+ = 20, Cu2+ = 20, Hg2+ = 20	Cd2+ = 18.8 ± 1.32, Pb2+ = 21.2 ± 1.15 Cu2+ = 20.8 ± 1.23, Hg2+ = 21.8 ± 1.96	93.42 106.24 104.35 107.25

.

In the blank control solution, no detectable stripping current peaks were observed for Cd²⁺, Pb²⁺, or Cu²⁺, while Hg²⁺ exhibited a faint signal of 0.35 ± 0.08 µA. As concentration increased, the stripping peak currents for all four metals increased with the amounts of spiked ions. Taking the 5 µM spiked group as an example, the peak currents for Cd²⁺, Pb²⁺, Cu²⁺, and Hg²⁺ were 4.65 ± 0.32, 4.87 ± 0.25, 4.78 ± 0.18, and 4.96 ± 0.45 µA, corresponding to recovery ratios of 93.1%, 95.2%, 95.6%, and 92.2%, indicating a slight decrease in enrichment efficiency at low concentrations. At 10 µM spiking, recovery rates for all four ions increased to 98~99%, while at 20 µM spiking, recovery ratios further approached or exceeded 100%. The increasing recovery ratio may result from more complete enrichment at high concentrations or slight deviations in background subtraction. Overall, this method achieved spiked recovery rates for Cu²⁺, Pb²⁺, Cd²⁺, and Hg²⁺ in artificial seawater within the reasonable range of 92–107%, with excellent repeatability (RSD < 5%), indicating that the SPE electrochemical sensor we developed maintains stable quantitative accuracy and superior interference resistance even in high-salinity matrices.

In addition, coastal seawater samples collected from three different locations, Dalian, Zhanjiang, and Maoming (Figure S7), were detected by the sensors we developed. Suspended particles were removed using a 0.22 μm PTFE filter as described above, and the samples were acidified to pH < 2 before refrigeration at 4 °C to remove sediment. Subsequently, 10 μM Cd²⁺, Pb²⁺, Cu²⁺, and Hg²⁺ was spiked in the water sample, respectively. Target ions (e.g., Pb²⁺, Cd²⁺) were electro-deposited and enriched onto the WE by applying a constant potential (e.g., −1.2 V vs. Ag/AgCl), followed by the application of an SWASV signal with a scan range of 0.1~0.8 V (frequency 25 Hz, amplitude 50 mV, step size 4 mV). Ion species were qualitatively identified by their elution peak potentials. The entire process was completed within 20 min. When combined with portable electrochemical workstation data-processing software, it enables rapid on-site detection of heavy metal ions in seawater samples.

Table 2. Detection of heavy metal ions in real seawater samples.

Water Sample	Blank (μM)	Spiked (μM)	Detected (μM)	Recovery Rate (%)
Dalian	Cd2+ = 0 Pb2+ = 0.96 ± 0.12 Cu2+ = 0 Hg2+ = 0.65 ± 0.13	Cd2+ = 10 Pb2+ = 10 Cu2+ = 10 Hg2+ = 10	Cd2+ = 9.76 ± 0.65 Pb2+ = 10.22 ± 0.54 Cu2+ = 10.36 ± 0.56 Hg2+ = 10.91 ± 0.89	97.63 98.13 92.64 102.62
Maoming	Cd2+ = 0 Pb2+ = 0 Cu2+ = 0 Hg2+ = 0.53 ± 0.19	Cd2+ = 10 Pb2+ = 10 Cu2+ = 10 Hg2+ = 10	Cd2+ = 10.12 ± 0.35 Pb2+ = 9.59 ± 0.24 Cu2+ = 9.86 ± 0.36 Hg2+ = 10.74 ± 0.75	101.25 95.95 98.67 102.15
Zhanjiang	Cd2+ = 0 Pb2+ = 0 Cu2+ = 0 Hg2+ = 0.57 ± 0.11	Cd2+ = 10 Pb2+ = 10 Cu2+ = 10 Hg2+ = 10	Cd2+ = 9.32 ± 0.34 Pb2+ = 9.87 ± 0.62 Cu2+ = 9.75 ± 0.23 Hg2+ = 106.6 ± 0.55	93.22 98.74 97.53 100.93

presents the results for the detection of heavy metal ions in real seawater samples from Dalian, Zhanjiang, and Maoming, spiked with the target ions. In this spiked recovery experiment, Cd²⁺ and Cu²⁺ concentrations were undetectable in unspiked seawater samples from Dalian, Maoming, and Zhanjiang. The concentrations of Pb²⁺ and Hg²⁺ in the Dalian sample are the highest, at approximately 0.96 ± 0.13 μM and 0.65 ± 0.13 μM, respectively. After spiking target ions of 10 μM, the measured concentrations of Cd²⁺, Pb²⁺, Cu²⁺, and Hg²⁺ are 9.32~10.12 μM (Cd²⁺), 9.59~9.87 μM (Pb²⁺), 9.75~10.36 μM (Cu²⁺), and 10.11~10.91 μM (Hg), respectively. Recovery rates of 93.2–103.6% and relative standard deviations below 6% were obtained (Table 2), demonstrating excellent quantitative accuracy and detection repeatability.

Based on detection data from actual seawater samples collected from these locations, the spatio-temporal distribution patterns of coastal heavy metal pollution can be preliminarily revealed. For example, the background concentrations of Pb²⁺ and Hg²⁺ in Dalian seawater samples are significantly higher than in the other two locations. This may be closely related to the dense shipbuilding and repair industries, port logistics, and historical industrial sediment releases in the surrounding areas. In contrast, the Hg²⁺ background levels in Zhanjiang and Maoming waters (0.53~0.57 μM) were lower than Dalian’s. This indicates that the Beibu Gulf region of the South China Sea is significantly affected by feed discharge from coastal aquaculture, wastewater discharge from the petrochemical industry, and atmospheric deposition carried by monsoons [64]. Notably, Cd²⁺ and Cu²⁺ were not detected in any of the three locations under non-spiked conditions, potentially reflecting stringent environmental controls imposed on electroplating and metallurgical industries in recent years [65]. The test results indicate potential improvements in the protection of the marine environment, thereby enhancing the safety of the ecosystem and marine organisms.

4. Conclusions

This study developed a highly sensitive, low-cost, and portable SPE electrochemical sensor for detecting heavy metal ions in the marine environment. Batch preparation of a three-electrode electrochemical sensor via screen-printing under conventional laboratory conditions was achieved. An optimized MOF-C@MWCNT composite was used to functionalize WE, thereby significantly enhancing conductivity and specificity for the detection of four typical marine heavy metal ions by SWASV. Simultaneous detection of heavy metal ion mixtures in seawater matrices was also achieved by optimizing deposition potential and scanning parameters. The sensor developed in this work demonstrates high accuracy and stability for detecting heavy metal ions in high-salinity matrices, with recoveries ranging from 92% to 107% and a relative standard deviation (RSD) below 5%. This study presents a facile method for fabricating high-performance SPE electrochemical sensors, particularly for heavy metal ion detection, via an innovative nanomaterial composite strategy. The SPE electrochemical sensor is compact and enables the entire detection process, from sample loading to result output, to be completed within 10 min using a handheld electrochemical reader. We believe that the advantages of portability, mass production, and simultaneous detection capabilities offered by the reported sensor would make it a practical tool for rapid on-site screening of heavy metal pollution, particularly in marine environments.