Au NP-Decorated g-C3N4-Based Photoelectochemical Biosensor for Sensitive Mercury Ions Analysis.

Herein, an efficient and feasible photoelectrochemical (PEC) biosensor based on gold nanoparticle-decorated graphitic-like carbon nitride (Au NPs@g-C3N4) with excellent photoelectric performance was designed for the highly sensitive detection of mercury ions (Hg2+) . The proposed Au NPs@g-C3N4 was first modified on the surface of the electrode, which possessed a remarkable photocurrent conversion efficiency and could produce a strong initial photocurrent. Then, the thymine-rich DNA (S1) was immobilized on the surface of the modified electrode via Au-N bonds. Subsequently, 1-hexanethiol (HT) was added to the resultant electrode to block nonspecific binding sites. Finally, the target Hg2+ was incubated on the surface of the modified glassy carbon electrode (GCE). In the presence of target Hg2+, the thymine-Hg2+-thymine (T-Hg2+-T) structure formed due to the selective capture capability of thymine base pairs toward Hg2+, resulting in the significantly decrease of the photocurrent. Thereafter, the proposed PEC biosensor was successfully used for sensitive Hg2+ detection, as it possessed a wide linear range from 1 pM to 1000 nM with a low detection limit of 0.33 pM. Importantly, this study demonstrates a new method of detecting Hg2+ and provides a promising platform for the detection of other heavy metal ions of interest.

Introduction

In recent years, with the rapid development of industrialization, the accumulation of heavy metal ions in the environment has increased, which has drawn more and more attention to the pollution problem. In terms of environmental pollution, mercury ions (Hg2+) are some of the most toxic heavy metal ions, which are a great threat to human life. An excessive content of Hg2+ would cause genetic mutation, affect the cell inheritance, produce teratogenesis, and cause cancer.[1−6] Therefore, the sensitive detection of Hg2+ has been an important issue in the fields of environmental and human health. At present, many methods such as colorimetry,[7,8] fluorescence,[9,10] electrochemiluminescence[11−15] and electrochemistry[16,17] have been used to analyze of Hg2+. However, these methods still exhibit the problems of insufficient detection range, poor sensitivity, and expensive equipment. In order to overcome the limitations of the above methods, a novel, sensitive, and accurate Hg2+ detection method urgently needs to be established. As a new and developing analytical technology, photoelectrochemical (PEC) biosensors combine optical and electrochemical methods[18−25] and possess the advantages of high sensitivity, excellent selectivity, simple equipment, and low cost. This technology has been widely used in various fields such as biological analysis, the pharmaceutical industry, environmental monitoring, and food safety.[26−32] Given these advantages, the PEC assay might be a promising analytical approach for the development of a highly sensitive and accurate Hg2+ detection strategy.

Choosing appropriate photoelectric materials is the key to constructing a highly sensitive PEC sensing platform. Among numerous photoelectric materials, graphitic-like carbon nitride (g-C3N4) has a band gap of 2.7 eV, which has attracted wide attention due to g-C3N4’s excellent chemical stability, adjustable electronic structure, low price, convenient synthesis, and lack of toxicity.[33−38] Unfortunately, the photoelectric performance of g-C3N4 is limited by the quick recombination rate of photogenerated electron–hole pairs and its low specific surface area. In order to improve the photoelectric performance of g-C3N4, many researchers have focused on modifying g-C3N4 to boost the charge separation through element doping, sensitization, and semiconductor coupling.[39,40] In particular, a semiconductor modified with metal nanoparticles (NPs) could effectively improve the photoelectric performance of the semiconductor for two reasons. One is that metal nanoparticles, a type of conductive material with a large specific surface area and unique photoelectric performance, could promote electron capture and transfer in the semiconductor. More importantly, the surface plasmon resonance (SPR) phenomenon would occur with the introduction of metal nanoparticles, which could lead to the enhancement of the light absorption capacity and the charge transfer capacity of semiconductor.[41−45] Inspired by these, we tried to combine g-C3N4 with gold nanoparticles (Au NPs) to prepare the Au NPs@g-C3N4 complex. In consequence, we have found that as-prepared Au NPs@g-C3N4 possessed an excellent photoelectric performance and could be used as an appropriate and promising photoelectric material for the construction of a PEC biosensor.

In this study, Au NPs@g-C3N4 as the photoelectric material was employed to construct a PEC biosensor to realize the highly sensitive and selective detection of Hg2+, which is shown in Scheme . First, Au NPs@g-C3N4 was modified on the electrode surface to generate a strong initial photocurrent signal. Afterward, the thymine-rich DNA (S1) was immobilized on the modified glassy carbon electrode (GCE) surface via Au–N bonds. Next, 1-hexanethiol (HT) was coated on the electrode surface to block nonspecific adsorption sites. Ultimately, Hg2+ was added, and the stable T–Hg2+–T coordination complex formed through the substitution of a proton with the nitrogen atom at thymine position 3 in the DNA molecule. The obtained T–Hg2+–T structure could block electron transfer, leading to an obviously decreased photocurrent signal for the detection of Hg2+. As expected, the PEC biosensor based on Au NPs@g-C3N4 exhibited high sensitivity, excellent selectivity, and stability for the Hg2+ assay, paving a new pathway for sensitive detection of other heavy metal ions.

Scheme 1

Schematic Representation of the PEC Biosensor for Hg2+ Detection

Experimental Section

Materials and Reagents

Chemical Reagent Co. Ltd. (Chongqing, China) provided H2O2. HAuCl4·4H2O was obtained from China National Medicines Corporation Ltd. (Beijing, China). HT, Tris-HCl buffer, and K3[Fe(CN)6] were bought from Shanghai Macklin Biochemical Co., Ltd. (Shanghai, China). KCl, NaCl, CaCl2, MgCl2·6H2O, Pb(NO3)2, and Hg(NO3)2 were purchased from Chengdu Kelong Chemical Co., Ltd. (Chengdu, China). This experiment used 0.1 M phosphate buffer solutions (PBS) (pH 7.0), including 0.1 M KCl, 0.1 M Na2HPO4 and 0.1 M KH2PO4. Ultrapure water was employed to prepare the solutions in this work. The oligonucleotide (S1) was synthesized by Sangon Inc. (Shanghai, China), and its sequence was as follows: 5′-NH2-CAAATGAACTTTGGTTTCCCTTTTCATTTT-3′.

Apparatus

The PEC measurement was carried out on a CHI 660E electrochemistry workstation with the help of the TPEC10W LED light source. A three-electrode system, which contained a platinum wire counter electrode, a calomel (saturated KCl) reference electrode, and a glassy carbon working electrode (GCE, Φ = 4 mm), was utilized in this work. A CHI 660E electrochemistry workstation (Shanghai Chenhua Instrumission, China) was also used for electrochemical measurements. The morphology of the nanomaterial was characterized by scanning electron microscopy (SEM, JSM-7800F, Japan) and transmission electron microscopy (TEM, FEI talos F200X, United States). Elemental analysis was performed on an X-ray photoelectron spectrometer (XPS, ESCALAB 250Xi, United States).

Preparation of g-C3N4 and Au NPs@g-C3N4

g-C3N4 was prepared based on the literature.[46] First, 5 g of melamine was put into a crucible and dried for 24 h at 80 °C, then calcined in a muffle furnace for 3 h. The calcination temperature was 550 °C, and the heating rate was 5 °C·min–1. The obtained product was a powder, and its color was light yellow. Finally, the product was thoroughly washed with 0.1 M nitric acid and ultrapure water.

Au NPs@g-C3N4 could be synthesized by an in situ reduction method.[47] Into 5 mL of the 2.0 mg·mL–1 g-C3N4 solution was dropped 20 μL of 10 mM HAuCl4, and the solution was then stirred in the dark for 2 h. Next, the newly prepared NaBH4 (30 μL, 0.01 M) was added to the above solution drop by drop, and the mixture was stirred continuously. When the gas evolution stopped, the reaction was complete. Subsequently, the obtained solution was placed in a centrifuge and washed by centrifugation to remove impurities. Finally, Au NPs@g-C3N4 was stored at 4 °C for subsequent experiments.

Construction of the PEC Biosensor

The GCE was fully polished with α-Al2O3 powder and then washed with ultrapure water before modification. Then, the clean GCE was coated with 10 μL of the as-prepared Au NPs@g-C3N4 solution and dried at 37 °C To generate a uniform film. Subsequently, the Au NPs@g-C3N4-modified electrode was incubated with 5 μL of 2 μM thymine-rich DNA (S1) at 4 °C overnight. S1 could be stably immobilized on the surface of Au NPs@g-C3N4/GCE via a Au–N bond. After the sample was blocked with 10 μL of 1 mM HT for 40 min, 10 μL of Hg2+ solutions with different concentrations was dropped onto the modified electrode. The sample was incubated at 37 °C for 1 h. In the presence of Hg2+, the T-rich DNA (S1) could recognize and capture Hg2+ and quickly folded into the T–Hg2+–T structure, thus quenching the photocurrent signal. The construction process of the PEC biosensor was shown in Scheme .

PEC Measurement

The PEC measurement was performed in 5 mL of 0.1 M PBS containing 40 μL H2O2, where H2O2 was the electron donor. The LED lamp with a wavelength of 365–370 nm served as the excitation light source and was switched off, on, and off for 10, 20, and 10 s, respectively, under a 0.0 V potential.

Results and Discussion

Characterization of the Different Materials

The morphology of the prepared material was characterized by SEM. As displayed in Figure A, the SEM image showed that the sample of g-C3N4 presented a layered structure with a mist-like edge. The diameter of g-C3N4 was about 110 nm. The SEM image of Au NPs@g-C3N4 is displayed in Figure B. It was observed that the surface of g-C3N4 was coated with numerous Au NPs. Meanwhile, TEM was also used to morphologically characterize Au NPs@g-C3N4. As shown in Figure C, a large number of small black spots of Au NPs were evenly distributed on the surface of g-C3N4. After being linked with Au NPs, the average size increased by about 10 nm. Additionally, these results were in accordance with the literature,[48−50] which showed that g-C3N4 and Au NPs@g-C3N4 were successfully synthesized.

Figure 1

SEM images of (A) g-C3N4 and (B) Au NPs@g-C3N4. (C) TEM image of Au NPs@g-C3N4.

In addition, XPS was used for elemental analysis. As shown in Figure , the peaks at 397 and 286 eV might correspond to N 1s and C 1s, respectively. This was consistent with literature, indicating the presence of N and C elements.[51−53] Moreover, the peaks at 82 and 86 eV confirmed the presence of the Au element.[54,55] These results demonstrated the successful preparation of Au NPs@g-C3N4.

Figure 2

XPS analysis of (a) the full region of Au NPs@g-C3N4, (b) the N 1s region, (c) the C 1s region, and (d) the Au 4f region.

Condition Optimization

To obtain an excellent analytical performance, the concentration of H2O2 in 5 mL of PBS and the wavelength of emitted light were optimized. As illustrated in Figure A, when the concentration of H2O2 increased from 0.02 to 0.08 M, the photocurrent also increased. However, the photocurrent decreased when the concentration of H2O2 increased from 0.08 to 0.12 M. Therefore, the optimal concentration of H2O2 was 0.08 M, corresponding to the photocurrent of 3.9 μA. As shown in Figure B, no significant photocurrent was observed at irradiation wavelengths of 455–465 nm and the mixed white light of 5000–5500 K. A photocurrent of about 2.27 μA could be found at 395–405 nm. However, when the irradiation wavelength was 365–370 nm, the photocurrent was the highest at 3.9 μA. Therefore, the optimal irradiation wavelength in this work was 365–370 nm.

Figure 3

Effect of (A) the H2O2 concentration in the detection solution and (B) the irradiation wavelength on the photocurrent.

Comparison of PEC Signals of Different Materials

In order to prove the superiority of Au NPs@g-C3N4 as a photoelectric material, PEC signals of g-C3N4 and Au NPs@g-C3N4 were compared under the same experimental conditions. As illustrated in Figure , the PEC signal of g-C3N4 was 1.3 μA. Au NPs@g-C3N4 produced a higher PEC signal, which was almost three times larger than that of g-C3N4, because the existence of Au NPs could produce a strong SPR enhancement effect. From the result, it could be seen that Au NPs@g-C3N4 performed better than g-C3N4.

Figure 4

PEC signals of g-C3N4 and Au NPs@g-C3N4.

PEC Mechanism of the Biosensor

Figure A shows the electron transfer process after the incubation of Au NPs@g-C3N4 on the GCE surface. g-C3N4 absorbed the light energy, and the electrons were excited into its conduction band (CB). One part of the photogenerated electrons was transferred to the electrode, and the other part was transferred to Au NPs. At the same time, Au NPs would be excited by incident light, causing electrons to oscillate collectively due to the SPR effect. After that, these electrons could immediately move to electrode. In addition, H2O2 as an electron donor contributed electrons to the valence band (VB) of g-C3N4. A strong photocurrent signal would be generated, as seen in Figure A. Figure B is the electron transfer diagram of photocurrent quenching after the addition of Hg2+. The T–Hg2+–T structure formed by thymine-rich DNA (S1) and Hg2+ would produce the steric hindrance effect, which could reduce the ability of light to capture electrons and prevent H2O2 from providing electrons. In addition, the photogenerated electrons produced from g-C3N4 and Au NPs could flow to the T–Hg2+–T structure because Hg2+ possessed the electron-withdrawing property. As a result, the photocurrent signal was greatly quenched.

Figure 5

Mechanisms for (A) photocurrent generation and (B) photocurrent quenching.

PEC Characterization of the Biosensor

In order to confirm the successful construction of the PEC biosensor, photocurrent characterization was carried out step by step. As seen in Figure , after Au NPs@g-C3N4 were incubated on the bare GCE surface, a sharply enhanced photocurrent was observed (curve b) compared with that of the bare GCE (curve a), mainly due to the favorable photoelectric properties of Au NPs@g-C3N4. The photocurrent evidently decreased (curve c) with the immobilization of S1 due to the poor conductivity of S1. Afterward, the photocurrent decreased again with addition of HT (curve d). Finally, when Hg2+ was modified on the above GCE surface, the photocurrent decreased significantly (curve e) due to the formation of the T–Hg2+–T structure. These results confirmed that the PEC biosensor was constructed successfully.

Figure 6

Photocurrent of (a) bare GCE, (b) Au NPs@g-C3N4/GCE, (c) S1/Au NPs@g-C3N4/GCE, (d) HT/S1/Au NPs@g-C3N4/GCE, and (e) Hg2+/HT/S1/Au NPs@g-C3N4/GCE.

Electrochemical Characterization of the PEC Biosensor

In order to study the preparation process of the PEC biosensor, cyclic voltammetry (CV) was used to characterize the peak current of the PEC biosensor at each step. The corresponding electrode was placed in 2 mL of PBS (pH 7, 0.1 M) containing 5.0 mM [Fe(CN)6]3–/4– and 0.1 M KCl at a scanning rate of 50 mV·s–1. As shown in Figure A, the bare electrode (curve a) showed a couple of reversible redox peaks. However, the peak current (curve b) evidently decreased after the modification with Au NPs@g-C3N4, which was mainly attributed to the semiconductor properties of Au NPs@g-C3N4. With the addition of S1, the peak current (curve c) further decreased because S1 and [Fe(CN)6]3–/4– both had negative charges and like charges repel each other. When HT was modified on the surface of the electrode, the peak current (curve d) continued to decrease. Finally, after Hg2+ was incubated on the electrode’s surface, the peak current (curve e) decreased, which was again caused by the formation of T–Hg2+–T structures.

Figure 7

(A) CV and (B) EIS responses of (a) bare GCE, (b) Au NPs@g-C3N4/GCE, (c) S1/Au NPs@g-C3N4/GCE, (d) HT/S1/Au NPs@g-C3N4/GCE, and (e) Hg2+/HT/S1/Au NPs@g-C3N4/GCE.

Meanwhile, the processes of the PEC biosensor were also characterized by electrochemical impedance spectroscopy (EIS), which was carried out in 2 mL of PBS (pH 7.0, 0.1 M) with 5.0 mM [Fe(CN)6]3–/4– and 0.1 M KCl. The frequency range was from 10 kHz to 0.1 Hz, the alternating current potential was 5 mV, and the direct current potential was 0.22 V. As displayed in Figure B, compared with bare GCE (curve a), the Au NPs@g-C3N4 -incubated electrode exhibited a greater charge-transfer resistance (Ret) (curve b) because Au NPs@g-C3N4 could hinder electron transfer. After S1 was modified on the surface of Au NPs@g-C3N4/GCE, the Ret increased once more (curve c) due to the repulsion between negatively charged S1 and [Fe(CN)6]3–/4–. When nonconductive HT was added to the electrode, Ret further increased(curve d). Ret continued to increase after the sample was incubated with Hg2+ (curve e) because of the formation of T–Hg2+–T structures. These results demonstrated that the PEC biosensor was prepared successfully.

PEC Analysis of Hg2+ at the Developed Biosensor

On the basis of the optimal conditions, photocurrents of samples incubated with different concentrations Hg2+ were measured to evaluate the analytical performance of the PEC biosensor, as shown in Figure A. As the concentration of Hg2+ increased from 1 pM to 1000 nM, the photocurrent decreased sharply. Figure B shows the linear response curve between the photocurrent and the logarithm of Hg2+ concentrations. The regression equation was I = −0.209lg c + 2.07, with a correlation coefficient of 0.998 (where I is the photocurrent and c is Hg2+ concentration). The insert of Figure B shows the linear response curve between the photocurrent and the Hg2+ concentration at a low concentration range. The regression equation was I = −23.485c + 2.727. The detection limit (LOD) of 0.33 pM was calculated according to LOD = 3SB/m, where SB was standard deviation of the blank signals and m was the analytical sensitivity. Meanwhile, a comparison was made between the analytical performance of the proposed PEC biosensor and other reported methods, which is illustrated in Table . The PEC biosensor constructed in this work had a better sensitivity and a wider linear range, which indicated that this sensing strategy was an excellent potential method for Hg2+ detection.

Figure 8

(A) Photocurrent with various Hg2+ concentrations. (B) Linear relationship between the photocurrent and the logarithm of the Hg2+ concentration. The insert in panel B presents the calibration curve of the photocurrent value vs the concentration of Hg2+ at a low concentration range.

Table 1

Comparison of Different Methods for Hg2+ Detection

analytical method	detection limit	linear range	ref.
fluorescence	0.24 nM	10–600 nM	(56)
fluorescence	15.2 nM	0–350 nM	(57)
SERSa	0.1 nM	0.1–1000 nM	(58)
colorimetry	20 nM	50–1 mM	(59)
colorimetry	10.3 nM	0–20 μM	(60)
electrochemistry	0.13 nM	0.01–0.5 nM	(61)
PEC	5.7 pM	10 pM to 1.0 μM	(62)
PEC	0.33 pM	1 pM to 1000 nM	our work

Surface-enhanced Raman scattering spectra.

Selectivity and Stability of the PEC Biosensor

In order to explore the selectivity of the proposed PEC biosensor, several interfering substances were evaluated, including Pb2+, Na+, Ca2+, Mg2+, and K+. As shown in Figure A, after the sample was incubated with 100 nM interfering substances, the obvious photocurrents were obtained. The photocurrent of the interfering substance was almost the same as that of the blank sample. However, when 1 nM Hg2+ was incubated, the photocurrent significantly decreased. The result showed that the PEC biosensor had an excellent selectivity for Hg2+ detection because the T–T mismatch bases had a high affinity for Hg2+. In addition, the stability of this PEC biosensor was investigated by recording the photocurrent of 1 nM Hg2+-incubated PEC biosensor under nine consecutive cycles of “off–on–off” light. It can be seen in Figure B that the photocurrent was stable with a relative standard deviation (RSD) of 0.417%, which indicated that the proposed PEC biosensor possessed a good stability for Hg2+ detection.

Figure 9

(A) Exploration of the selectivity of the PEC biosensor. (B) Stability test of the biosensor at 1 nM Hg2+.

Preliminary Application of Hg2+ Detection

The standard addition method was applied to evaluate the applicability and reliability of the constructed PEC biosensor. Different concentrations of Hg2+ were added into water and detected using this biosensor. As can be seen in Table , the concentrations of Hg2+ were 500 nM, 50 nM, 500 pM, 50 pM, and 5 pM, and the corresponding recovery rates were 95.6%, 105.6%, 94.6%, 94.6%, and 103.2%. The above results indicated that the proposed PEC biosensor had great potential to detect Hg2+ in real samples.

Table 2

Detection of Hg2+ with different concentrations by PEC biosensor

sample number	added (nM)	found (nM)	recovery (%)
1	500	478	95.6
2	50.0	52.8	105.6
3	5.00 × 10–1	4.73 × 10–1	94.6
4	5.00 × 10–2	4.73 × 10–2	94.6
5	5.00 × 10–3	5.16 × 10–3	103.2

Conclusions

In summary, a novel and sensitive PEC biosensor has been developed for Hg2+ detection based on Au NPs@g-C3N4 as the photoelectric material. The prepared Au NPs@g-C3N4 could provide an excellent initial photocurrent signal because the introduction of Au NPs to the surface of g-C3N4 could lead to the SPR phenomenon, thus enhancing both the absorption of visible light and the electron transfer ability. Besides, the chemical coordination of T–Hg2+–T between S1 and Hg2+ could hinder electron transfer, leading to a significant reduction of the PEC signal for achieving the quantitative analysis of Hg2+. The as-proposed PEC biosensor displayed the advantages of simple preparation, sensitive detection, good stability, and high selectivity. Furthermore, this developed PEC strategy provides a modular platform for the analysis of various trace heavy metal ions, which is expected to be applied in environmental and disease detection.