In this article, an N-rich covalent organic framework (COFTFPB-TZT) was successfully synthesized using 4,4′,4′-(1,3,5-triazine-2,4,6-triyl) trianiline (TZT), and 4-[3,5-bis (4-formyl-phenyl) phenyl] benzaldehyde (TFPB). The as-prepared COFTFPB-TZT possesses irregular cotton wool patches with a large specific surface area. A novel selective electrode based on COFTFPB-TZT was used for the determination of Mercury ions. The abundance of N atoms in COFTFPB-TZT provides more coordination sites for Hg2+ adsorption, resulting in a change in the surface membrane potential of the electrode to selectively recognize Hg2+. Under optimal experimental conditions, the ion-selective electrode shows a good potential response to Hg2+, with a linear range of 1.0 × 10−9∼1.0 × 10−4, a Nernst response slope of 30.32 ± 0.2 mV/-PC at 25°C and a detection limit of 4.5 pM. At the same time, the mercury-ion electrode shows a fast response time of 10 s and good reproducibility and stability. The selectivity coefficients for Fe2+, Zn2+, As3+, Cr6+, Cu2+, Cr3+, Al3+, Pb2+, NH4+, Ag+, Ba2+, Mg2+, Na+, and K+ are found to be small, indicating no interference in the detection system. The proposed method can be successfully applied to the determination of Hg2+ in 3 typical environmental water samples, with a recovery rate of 98.6–101.8%. In comparison with the spectrophotometric method utilizing dithizone, the proposed method is simple and fast and holds great potential application prospects in environmental water quality monitoring and other fields.

1. Introduction

Heavy metal pollution poses a great threat to human health and causes serious environmental problems. Mercury (Hg) is a highly toxic heavy metal element widely distributed in air, water, and soil [1]. Trace amounts of Hg2+ can produce strong toxicity and accumulate in the environment and organisms and can be transferred to the human body through the food chain, directly damaging the human brain, nervous system, kidney, and endocrine system [2, 3]. Therefore, the determination of Mercury ions is necessary in environmental monitoring and clinical analysis, and it is of great significance to improve the quality of human life and protect the Earth’s environment. At present, mature methods for mercury-ion detection are as follows: spectrophotometry [4], inductively coupled plasma–mass spectrometry (ICP–MS) [5], atomic absorption/emission spectrometry (AAS/AES) [6], atomic fluorescence spectrometry (AFS) [7], high-performance liquid chromatography (HPLC) [8], inductively coupled plasma atomic emission spectrometry (ICP–AES) [9], and fluorescence probe detection [10]. Although these methods show excellent detection performance, most of them require expensive precision instruments, complex sample preparation processes, and skilled operators and cannot be conveniently used online or outdoors. Therefore, it is very important to find a simple, rapid, and inexpensive method for the determination of mercury-ion content.

The ion-selective electrode (ISE) method, as a rapid analytical tool, has been widely developed in the fields of the environment, medicine, and agriculture due to its advantages of fast speed, simple preparation, low cost, and high sensitivity [11]. In the process of preparing ion-carrier membranes with different selective electrodes, many organic, and inorganic complexes can be used as selective carriers of Mercury ions, such as crown ether [12], amide [13], glycidyl diamine [14], metalloporphyrin [15], Schiff base [16], calixarene derivatives [17, 18], and alkylamine Mercury salt complexes [19]. Therefore, there is an urgent need to prepare Mercury ion-selective electrodes with good selectivity and a low detection limit. A large number of studies have proven that compounds containing N, O, and S atoms can form coordination compounds with heavy metal ions [20]. Mercury (II) shows a strong affinity with ligands containing sulphur atoms, such as sulfhydryl compounds [21], sulphur-containing heterocyclic rings [22], and thiourea derivatives [23], and its interaction relationship has been further verified by density functional theory calculations [24]. Many compounds containing sulphur atoms have been successfully used as active carriers for Mercury ion-selective electrodes [25]. Li et al. [26] prepared a simple and rapid Mercury ion selective electrode based on a 1-undecanethiol (1-UDT) PVC film electrode for the selective determination of Hg2+ ions, with a Nernst-response range and a detection limit of 4.5 nM. Fang et al. [27] reported a portable Hg2+ nanosensor using the CuS nanozyme functions as a Hg2+ recognition unit, which exhibits high sensitivity with a minimum detectable Hg2+ concentration of 50 ppt. Diamantis [28] et al. reported a microporous 8-connected Zr4+ metal-organic framework (MOF) based on a terephthalate ligand decorated with a chelating 2-picolylamine side group (dMOR-2), which shows a highly efficient fluorescence sensing and sorption of Mercury ions (Hg2+), and the limits of detection were determined to be below 2 ppb for Hg2+.

The covalent organic framework (COF) is a porous conjugated polymer that shows a large surface area, adjustable pore order, and easy functionalization [2931]. In this study, a COF with an irregular cotton wool patch structure was synthesized from 4,4′,4′-(1,3,5-triazine-2,4,6-triyl) trianiline (TZT), and 4-[3,5-bis (4-formyl-phenyl) phenyl] benzaldehyde (TFPB). The as-prepared COF has ordered holes and a large specific surface area, and more N atoms will be exposed on the inner wall of the holes and the specific surface area of COF, providing more adsorption sites for the binding of Hg2+. SEM and TEM characterization also prove the existence of such a recognition effect. The experimental results show that the electrode has high sensitivity, selectivity, reproducibility, and stability. The electrode is used for the rapid determination of Mercury ions in 3 typical environmental water samples and the results are compared with the spectrophotometric method with dithizone. The as-prepared COFTFPB-TZT shows excellent selectivity for Hg2+, holding important application value in environmental monitoring, biomedical and agricultural product safety detection, and other fields.

2. Experimental Section

2.1. Reagents and Instruments

4,4′,4′-(1,3,5-triazine-2,4,6-triyl) trianiline (TZT) and 4-[3,5-bis (4-formyl-phenyl) phenyl] benzaldehyde (TFPB) were obtained from Jilin Chinese Academy of Sciences Technology Co., Ltd. (Beijing, China). o-Dichlorobenzene (o-DCB), butyl alcohol (n-BuOH), dichloromethane (DCM), tetrahydrofuran (THF), N, N-dimethylformamide (DMF), acetic acid (AcOH), and Mercury chloride (Hg2Cl) were purchased from Aladdin Biochemical Technology Co. Ltd. (Shanghai, China). Potassium ferricyanide (K3 [Fe (CN)6]), potassium ferrocyanide (K4 [Fe (CN)6]), sodium chloride (NaCl), disodium hydrogen phosphate (Na2HPO4·7H2O), sodium dihydrogen phosphate (NaH2PO4·7H2O), sodium sulphate (Na2SO4), chlorine (HCl), sulfuric acid (H2SO4, 98%), and hydrogen peroxide (H2O2, 30%) were purchased from Sinopharm Chemical Reagents Co. Ltd. (Shanghai, China). All chemicals used in the experiment were of analytical grade. Ultrapure water with a specific resistance larger than 18.3 MΩ·cm−1 was used throughout the experiments.

The pH values were measured with a REXPHSJ-4A pH Meter (INESA Instrument Co. Ltd., China). All electrochemical measurements were performed using a CHI-760B workstation (Shanghai Chenhua Instruments Co., China). The potential values were detected with an Ollie Dragon Model 868 pH/mV meter (Thermo Orion, USA). Scanning electron microscopy (SEM) images were obtained using a Nova Nano230 (Thermo Fisher Scientific Co. Ltd., USA). Transmission electron microscopy (TEM) images were obtained using an FEI (Tecnai G2 F20 TMP, USA). The X-ray powder diffraction (XRD) spectrum was measured using a D/max-2500 diffractometer with a Cu Kα radiation source (λ = 1.54056 Å), Rigaku Co., Ltd., Tokyo), and a gas adsorption instrument (Micromeritics ASAP 2460 Sorptometer, MAC Instruments Co., Ltd., USA) was used to characterize the samples.

2.2. Synthesis of COFTFPB-TZT Material

First, 0.1 mol of TFPB and 0.1 mol of TZT were dissolved in 3 mL of dichloromethane (DCM)/n-butanol mixed solvent (1 : 1), followed by ultrasound treatment for 30 s. Next, 0.2 mL of 6.0 M acetic acid (AcOH) was added and poured into a Teflon-lined autoclave. Next, the reactor was heated at 120°C for 3 days. The precipitate obtained was repeatedly washed with dimethylformamide until the upper liquid became colourless. Meanwhile, the product was further purified by soaking in THF for 6 hours.

2.3. Electrode Preparation of COFTFPB-TZT

A “piranha” solution was prepared by mixing 98% H2SO4 and 30% H2O2 at a volume ratio of 3 : 1 before being poured onto the surface of a gold plate electrode (GPE) for 1 min. The GPE was cleaned with ultrapure water and ethanol and dried with nitrogen for later use. Ethanol was ultrasonically dispersed with 1.5 mg COFTFPB-TZT, and then the gold surface was immersed in 1.5 g/L COFTFPB-TZT ethanol solution and self-assembled at 4°C for 24 h. Finally, the COFTFPB-TZT-modified electrode was washed with ethanol and ultrapure water, dried, and stored for later use.

3. Results and Discussion

3.1. Synthesis and Characterization of COFTFPB-TZT

The morphology and structure of as-prepared COFTFPB-TZT before and after reacting with Mercury ions (Hg2+) in solution was further characterized by transmission electron microscopy (TEM). As shown in Figure 1, COFTFPB-TZT exhibits irregular cotton wool patches (Figures 1(a) ∼1(c)). Upon the incorporation of Hg2+, an amount of target Hg2+ is adsorbed onto the surface of irregular cotton wool patches via chemical chelation (Figures 1(d)–1(f)), and the irregular cotton wool patches are regathered into stacked layers of nano cotton. The change in the morphology of COFTFPB-TZT before and after the addition of Hg2+ is ascribed to the interaction between COFTFPB-TZT and Hg2+ via the chelation of chemical bonds. In addition, the surface morphology characteristics of the GPE/COFTFPB-TZT electrode before and after reacting with Mercury ions were tested by SEM, as shown in Figure 1(g) ∼1(i). The naked gold plate electrode surface is observed to be very smooth (Figure 1(g)), and after the self-assembly of COFTFPB-TZT, the surface coverage becomes rough (Figure 1(h)). After testing in mercury-ion solution, the surface of the as-prepared COFTFPB-TZT changes significantly, and the electrode appears as larger agglomerates (Figure 1(i)). As shown in Scheme 1, the change in the structure can be ascribed to Mercury ions and amino functional groups of COFTFPB-TZT in the formation of strong coordination bonds, resulting in the complexes gathering strongly and changing the shape of the surface. Thus, different electron microscope images are presented.

The crystalline structure of COFTFPB-TZT was evidenced by X-ray diffraction (XRD) (Figure 2(a)), and the corresponding characteristic diffraction peaks for COFTFPB-TZT for the (100), (111), (021), (022), (023) and (203) planes are observed, which is in agreement with the simulated spectrum. In addition, adsorption-desorption isotherms (BET) were obtained to evaluate the specific surface area and pore size of the as-prepared COFTFPB-TZT. The specific surface area of COFTFPB-TZT is deduced to be 80.9 m2·g−1, and the pore size of COFTFPB-TZT is 0.35 nm (Figure 2(b)), providing more active sites for Hg2+ ions. The XRD and BET results confirm that the as-prepared COFTFPB-TZT possesses a highly crystalline structure and ordered holes with a large specific surface area, providing more adsorption sites for the binding of Hg2+. The N-atom rich composition of COFTFPB-TZT is responsible for the formation of COFTFPB-TZT/Hg2+, as evidenced by the FT-IR spectrum. As shown in Figure 3, the characteristic peak at 2390.49 cm−1 further proves the coordination relationship between the N element and Hg2+. These results confirm that an N-rich covalent organic framework (COFTFPB-TZT) can be successfully synthesized with 4,4′,4′-(1,3,5-triazine-2,4,6-triyl) trianiline (TZT), and 4-[3,5-bis (4-formyl-phenyl)phenyl]benzaldehyde (TFPB) reagent through the dehydration condensation reaction of amine aldehydes.

3.2. Electrochemical Characterization of COFTFPB-TZT

To verify the interaction between the carrier and Mercury ion, the electrochemical behaviour of the different modified electrodes above was investigated by electrochemical impedance spectroscopy (EIS) and cyclic voltammetry (CV). As shown in Figure 4(a), curve a represents the bare gold plate electrode, and the impedance value is very small, indicating that the pretreated bare gold plate electrode has a strong ability to transfer electrons. Curve b represents the COFTFPB-TZT-modified gold plate electrode. The semicircle appears in the high frequency part, and the impedance value increases significantly, indicating that COFTFPB-TZT forms a nonconductive monomolecular self-assembled film on the gold surface through the sulfhydryl group at one end, which hinders the electron conduction of [Fe (CN)6]3-/4- on the electrode surface. After combining Hg2+ (1.0 × 10−4 mol/L), the impedance value of the electrode decreases correspondingly (curve c), which is due to the strong coordination between the N element group and Hg2+ at the other end of COFTFPB-TZT, which adsorbs positively charged Mercury ions and enhances the electronic conductivity, resulting in an electrochemical conduction current and reduced interfacial impedance. The variation trend for the impedance can also be verified by the corresponding cyclic voltammetry (Figure 4(b)). A pair of obviously reversible redox peaks in curve a is obtained, suggesting fast electron transfer at the bare gold plate electrode surface. Upon modification with COFTFPB-TZT, the sharply decreased peak current in curve b indicates that the COFTFPB-TZT materials can hamper electron transfer to the electrode surface. When the COFTFPB-TZT-modified electrode is treated with Hg2+ (curve c), the peak currents are slightly enhanced, which can be ascribed to the coordination of Mercury (II). The CV results are consistent with the above EIS results, indicating that this method is feasible for the identification and detection of Mercury ions.

3.3. Optimal pH Selection

In the experiment, the relationship between the electrode potential and the concentration of the measured ionic solution at pH 3.0, 4.0, 5.0, 6.0, 7.0, 8.0, and 9.0 is discussed. The response slope was calculated and the relationship between the slope and pH was determined, as shown in Figure 3. As shown in Figure 5, the electrode presents a Nernst response to Hg2+ under acidic conditions (pH = 3.0∼6.0). The electrode response slope shows the highest value at pH = 6.0, with a slope value of 30.32 ± 0.2 mV/-PC at 25°C, which is close to the theoretical value of the Nernst response slope of 29.58 mV/-PC [2931]. The electrode slope at pH 3.0 and 4.0 is slightly smaller, which can be regarded as the nearly equal near-Nernst response slope. However, the slope of the response gradually decreases as the pH exceeds 6.0, which can be ascribed to the adverse reaction in the solution: Hg2+ + OH ⟶ Hg (OH)+. The decrease in slope originates from a small amount of Hg2+ being converted to Hg (OH)+ for the formation of insoluble salts, leading to a reduction in the concentration of Mercury ions at pH > 6.0. In addition, an insoluble Mercury salt produced in an alkaline solution (pH > 7.0) is likely to adhere to the electrode surface, leading to electrode damage. Therefore, the optimal pH value was selected to be 6.0.

3.4. Electrode Response Performance

The test response performance of GPE/COFTFPB-TZT to Hg2+ was investigated in the experiment. Figure 6 shows the potential response curve of the electrode combined with different concentrations of Hg2+ in Tris-HCl buffer solution (pH = 6.0). As shown in Figure 6, the electrode potential gradually increases with increasing Hg2+ concentration, indicating that the bonding of Hg2+ on the modified electrode surface increases. In addition, the electrode shows a good linear response to Hg2+ ions in the concentration range of 1.0 × 10−9∼1.0 × 10−4 mol/L in Tris-HCl buffer solution at pH = 6.0. The linear equation fitted by the least square method is δ E = 276.08 + 26.82 log C (R2 = 0.9873). The lower limit of detection can be deduced to be 4.5 × 10−12 mol/L according to the plotting method. Comparing the GPE/COFTFPB-TZT with other modified electrodes reported using different nanomaterials [3237], as shown in Table 1, it can be seen that the COFTFPB-TZT-modified biosensor in this work has better properties with a wider linear range and higher sensitivity.

3.5. Determination of Response Time, Stability, and Reproducibility

The response time and stability of the GPE/COFTFPB-TZT-modified gold plate electrode for Hg2+ detection were investigated (Figure 7). The change in the dynamic potential curve after the addition of different concentrations of Hg2+ ions to Tris-HCl buffer solution was determined, namely, continuous measurements were carried out from low concentration to high concentration in the range of 1.0 × 10−9∼1.0 × 10−4 mol/L and the change in the potential value over time was recorded. The reaction time for the electrode to reach equilibrium in the whole concentration range is very short, which is calculated as 95% of the maximum potential response, i.e., 20 s, indicating that the electrode has a fast response speed to Mercury ions. Then, the potential gradually tends to be stable with increasing time, indicating that the response performance of the electrode is stable. At the same time, the electrode was continuously monitored on a 1.0 × 10−7 mol/L Hg2+ sample for 25 min with a potential drift of +1.0 mV. The standard deviation for the potential data obtained is ± 2.22 mV, and the relative standard deviation is 1.72% (n = 10), indicating that the potential sensor has good stability. After the electrode was tested on Hg2+ samples for 15 days, the response slope of the electrode was changed to 20.5 mV/-PC, which decreased by 30.8%, indicating that the potential sensor can be used for at least half a month and has a long service life.

The potential response reproducibility of the GPE/COFTFPB-TZT electrode combined with different concentrations of Hg2+ samples was also investigated (Table 2); that is, the potential values of the 1.0 × 10−7 mol/L and 1.0 × 10−4 mol/L Hg2+ samples were measured back and forth 10 times. The relative standard deviations are 0.54% and 0.45%, respectively. The relative standard deviation is small, indicating that the electrode has good reproducibility.

3.6. Selectivity of the Electrode

In this experiment, the ion selectivity coefficient of the electrode was measured by the fixed interference ion concentration method (Figure 8), and the selectivity coefficient was calculated by using the Nicolskii–Eisenman formula [18].where represents the selectivity coefficient of Mercury ions, represents mercury-ion activity, and represents the activity of interfering ions. In the actual calculation, the ionic strength coefficient is ignored, and the activity is approximately replaced by the concentration. The selectivity coefficients for different metal ions (such as Fe2+, Zn2+, As3+, Cr6+, Cu2+, Cr3+, Al3+, Pb2+, NH4+, Ag+, Ba2+, Mg2+, Na+, and K+) were investigated. The results show that the selectivity coefficients for these metal ions and oxidizing ions are relatively small, which does not interfere with the determination of Hg2+ by the electrode, indicating that the electrode shows good selectivity.

3.7. Determination and Analytical Application of the Recovery Rate

To verify the practical applicability of the GPE/COFTFPB-TZT self-assembled gold plate electrode, the fabricated electrodes were used to detect Hg2+ in 3 typical water samples (river water, DI water, and polluted water) and compared with the spectrophotometric method with dithizone. As shown in Table 3, the concentration of Hg2+ measured by the GPE/COFTFPB-TZT self-assembled gold plate is consistent with that obtained by the spectrophotometric method with dithizone. Moreover, the recovery rates were found to vary in the range from 98.6% to 101.8%. According to the principle of the F test and T test in error analysis, the F value and T value for 6 samples were calculated, respectively, and the F value was determined to be 1.14, 2.14, 1.00, 2.86, 1.15, and 2.12, and the T value was determined to be 1.04, 1.06, 0.64, 0.30, 0.96, and 0.56, respectively. The above-given results indicate that the GPE/COFTFPB-TZT self-assembled gold plate electrode can be used to determine Mercury ions in actual water samples. The spectrophotometric method with dithizone is a standard method for the detection of metal ions but requires a complex pretreatment process and involves organic reagents. In comparison with the spectrophotometric method utilizing dithizone, the proposed method is simple, easy to operate, fast, and holds great promise for environmental monitoring, biological medicine, and agricultural product safety testing.

4. Conclusion

A novel N-rich COFTFPB-TZT is prepared with the dehydration condensation reaction of amine aldehydes. A Hg2+−selective electrode based on GPE/COFTFPB-TZT is developed based on a self-assembly method. In particular, COFTFPB-TZT contains many N elements, which provide abundant coordination sites for Hg2+. The experimental results show that the GPE/COFTFPB-TZT electrode shows excellent performance with a wide linear response range of 1.0 × 10−9 ∼1.0 × 10−4 mol/L and a short response time of 10 s, and the detection limit is determined to be 4.5 pM. In Tris-HCl buffer solution with pH = 6.0, the electrode potential electrode shows a good potential response to Hg2+ that meets the Nernst response. In comparison with the spectrophotometric method utilizing dithizone, the proposed method is simple and fast and holds great potential application prospects in environmental detection and biomedicine.

Data Availability

The data used to support the study are included in the paper.

Written informed consent for publication was obtained from all authors.

Conflicts of Interest

The authors declare that there are no conflicts of interest.


This work was financially supported by the following funds: Scientific Research Project of the Education Department of Hunan Province (Grant no. 18C0288)/Youth Scientific Research Fund Project of Central South University of Forestry and Technology (Grant no. 2016QY010)/Hunan Provincial Natural Science Foundation of China (Grant no. 2020JJ4950).