A simple “on–off–on” ECL sensor for glucose determination based on Pd nanowires and Ag doped g-C₃N₄ nanosheets

Abstract

One-dimensional Pd nanowires (PdNWs) and Ag doped g-C₃N₄ nanosheets (Ag-g-C₃N₄ NSs) were prepared and applied to construct a simple “on–off–on” ECL sensor for glucose detection. Compared to some complex and time-consuming synthetic procedures, in this work, Ag-g-C₃N₄ NSs were prepared via one step thermal polymerization of melamine and AgNO₃ without any handling to get the layered structure. PdNW-functionalized Ag-g-C₃N₄ NSs were used as luminophores, which displayed prominent ECL performance and superior efficient electron transport capacity. Combining PdNW-functionalized Ag-g-C₃N₄ NSs with the competitive binding reaction of the glucose–ConA–dextran system, the prepared “on–off–on” sensor showed high sensitivity and wide linearity for glucose detection. The linear range was 1.0 × 10⁻¹¹ M to 1.67 × 10⁻⁶ M and the detection limit was 3.0 × 10⁻¹² M (S/N = 3).

---

1. Introduction

One-dimensional (1D) nanomaterials with distinctive structural and electronic properties, such as nanowires (NWs), are receiving considerable attention because of their importance in basic research and potential applications.^1–3 In particular, palladium nanowires (PdNWs), a widely applied metal nanomaterial in catalysis,^4,5 have stimulated great interest for energy storage⁶ and sensors.⁷ Many studies have suggested that this intrinsic catalytic activity of PdNWs emerges from their inherent unique anisotropic structure and the synergistic effect between their different physical and chemical properties.^8–10 To date, considerable effort has been made for preparing 1D nanowires, such as by the template synthesis,¹¹ electrospinning method,¹² wet-chemical method,¹³etc. However, it is still essential to introduce a simple and mild method to fabricate PdNWs. In this work, we proposed a one pot wet-chemical reduction route for synthesizing PdNWs using K₂PdCl₄, polyvinylpyrrolidone (PVP), KI and ultrapure water.

PdNWs have attracted more and more attention due to their inherent advantages such as the low background signal, high detection sensitivity, and short time consumption to drive the dreams to explore their possible applications in electrochemiluminescence (ECL).^14,15 Few studies have focused on PdNW-based ECL sensors.^16–18 Wang et al.¹⁶ established a sandwiched and sensitive ECL immunosensor for carcinoembryonic antigen (CEA) detection based on Pd nanowires obtained by a green approach. Ran et al.¹⁷ used platinum and palladium nanowires (Pt–Pd NWs) as signal amplification materials for the detection of L-alanine. Ye et al.¹⁸ proposed one-dimensional Pd–Au nanowires as an immobilization platform for the ECL detection of acetylcholinesterase (AChE) activity.

In recent years, a switch strategy in the ECL assay named “on–off–on” has attracted increasing attention due to its low background signal and high sensitivity compared to the “on–off” mode and “off–on” mode.^19–21 Lei et al.²² constructed an “on–off–on” ECL biosensor for copper ion (Cu²⁺) detection. The first signal-on state with remarkably high ECL intensity was achieved by supramolecular nanorods with a co-reaction accelerator. Later, the signal-off state was achieved by ECL quencher probes (Fc-NH₂/Cu-Sub/nanoAu). As expected, the second signal-on state was the ECL signal recovery when the Cu²⁺-specific DNAzyme irreversibly cleaved, causing the release of the quencher probes from the sensor interface. Zhao et al.²³ designed a label-free ECL aptasensor via a novel “on–off–on” switch system for highly sensitive determination of kanamycin by the quenching effect of hemin/G-quadruplex DNAzymes towards the S₂O₈²⁻–O₂ system. Inspired by the above features of the “on–off–on” mode, finding a high initial ECL emitter as the first signal-on state is necessary to achieve a sensitive analysis.

Graphite-like carbon nitride (g-C₃N₄) nanosheets, which present intriguing advantages such as better film-forming properties, higher fluorescence quantum yield and lower toxicity than their bulk counterparts,²⁴ have become a focus of attention to be effective, strong, stable and promising ECL luminophores. Many attempts have been made to synthesize g-C₃N₄ nanosheets from bulk g-C₃N₄ such as by the ultrasound process. Sha et al.²⁵ obtained g-C₃N₄–COOH nanosheets by placing bulk g-C₃N₄ in mixed acid (VH₂SO₄ : VHNO₃ = 3 : 1) and sonicating it for 12 hours. Zhang et al.²⁶ prepared g-C₃N₄ nanosheets by liquid phase exfoliation in water by sonicating bulk g-C₃N₄ powder for 16 hours. And also many reports investigated hybridizing g-C₃N₄ with some nanomaterials (such as Au and Ag nanoparticles, carbon materials and polymers) to improve the ECL properties of g-C₃N₄.^27–30 All these above processes were related to relatively complex and time-consuming synthetic procedures, which greatly limit the further applications of g-C₃N₄. To circumvent this challenge, we introduced a method³¹ to prepare Ag doped g-C₃N₄ nanosheets (Ag-g-C₃N₄ NSs) via one step thermal polymerization of melamine and AgNO₃ without any handling to get the layered structure. The ECL intensity of g-C₃N₄ could be enhanced by chemical doping and modification of the luminophore.

In this study, we have taken advantage of PdNW-functionalized Ag-g-C₃N₄ NSs and the “on–off–on” switch system to construct a highly sensitive non-enzymatic ECL sensor for the determination of glucose, which plays a crucial role in reducing health risks by monitoring its level in biological fluids. Firstly, PdNW-functionalized Ag-g-C₃N₄ NSs, which possess a large surface area, excellent electrocatalytic performance, and satisfactory biocompatibility, are modified onto a glass carbon electrode (GCE) for immobilizing phenoxy dextran (DexP) through π–π interaction to achieve a high initial intensity as the first “signal-on” state. Next, the “signal-off” state was obtained by incubating ConA onto the electrode via the special binding between ConA and DexP which was due to the fact that ConA protein on the electrode could hinder electron transfer. Finally, the modified electrode was immersed into the glucose solution, the glucose would compete with DexP for ConA binding sites due to the higher binding affinity of glucose for ConA than DexP, and thus ConA would be taken away from the electrode. Therefore, the hindering effect from ConA incubated on the electrode would be weakened, resulting in an increase in ECL intensity and achieving the “signal-on” state (Scheme 1). The proposed biosensor with a wide linear response range, low detection limit and good stability, was successfully applied to the determination of glucose in samples of bovine serum albumin.

Scheme 1 Illustration of the fabrication process and the detection principle of the glucose sensor.

2. Materials and methods

2.1. Chemicals

Melamine (2,4,6-triamino-1,3,5-triazine, 99%), 1,2-epoxy-3-phenoxypropane (Epoxy), Concanavalin A (ConA) from Canavalia ensiformis (jack bean), Tween20, potassium palladium(II) tetrachloride (K₂PdCl₄, 99.998%), and poly(vinyl pyrrolidone) (PVP, MW = 29 000) were received from Sigma-Aldrich (St. Louis, MO, USA). Dextran was purchased from Sinopharm Chemical Reagent Co. (Shanghai, China). Phenoxy dextran (DexP) was prepared according to the literature.³² Phosphate buffered saline (PBS) solutions with pH 7.4 were prepared using 0.10 M KCl, 0.10 M K₂HPO₄ and 0.10 M KH₂PO₄. Concanavalin A with various concentrations were prepared in 0.10 M PBS (pH 7.4) with 0.50 mM CaCl₂ and 0.50 mM MnCl₂ to activate the ConA conformation. All the other chemical reagents were of analytical grade and used in the convenient way. Ultrapure water (≥18 MΩ) was employed throughout the experiments.

2.2. Apparatus

The ECL and cyclic voltammetry (CV) measurements were respectively performed with an MPI-A electrochemiluminescence analyzer (Xi'an Remax, China) and a CHI 600D electrochemical work station (Shanghai CH Instruments, China). The voltage of the photomultiplier tube (PMT) of the ECL analyzer was set at 600 V in the process of detection. X-ray photoelectron spectroscopy (XPS) was performed on a Thermo Scientific Escalab 250Xi. Scanning electron microscopy (SEM) was carried out on an SU8010 (Hitachi Instruments Co., Japan). Transmission electron microscopy (TEM) was carried out on a Tecnai G2 F20 S-TWIN (Thermo Fisher Scientific, America).

2.3. Synthesis of Ag-g-C₃N₄ nanosheets

The Ag-doped g-C₃N₄ NSs were synthesized from melamine powder and AgNO₃ according to Hu's method.³¹ Firstly, 151 mg AgNO₃ was dissolved in 5 mL ultrapure water with stirring. Then, 5.0 g melamine powder was added to the above solution with stirring to make a slurry. The slurry was put into an alumina crucible with a cover and then heated at 550 °C for 4 h, and then cooled to room temperature to generate a yellow powder. The obtained yellow powder was Ag-doped g-C₃N₄ NSs.

2.4. Synthesis of Pd nanowires

PdNWs were prepared by a simple one-step hydrothermal method with a modification.¹⁰ Initially, 32.6 mg K₂PdCl₄ powder, 12 mL ultrapure water and 0.8 g PVP (29 000) were mixed and stirred for 15 min. Then, KI (1.8 mL, 166 g L⁻¹) was dropped into the above mixture with stirring. Next, the solution was transferred into a Teflon-lined stainless steel autoclave and then heated at 200 °C for 3 h. After cooling to room temperature, the PdNWs were precipitated by adding 15 mL isopropanol, then separated via centrifugation at 5000 rpm for 15 min, and purified three times with an isopropanol–ethanol mixed solution (1 : 1) and ultrapure water. Finally, the product was dried at 60 °C under vacuum.

2.5. Preparation of PdNWs–Ag-g-C₃N₄

Firstly, 3 mg Ag-g-C₃N₄ and 100 μL 1% chitosan solution were added into 10 mL ultrapure water with vigorous stirring for 8 h to get a homogeneous dispersion. Thus, the Ag-g-C₃N₄ had a better dispersion and membrane-forming properties with introducing chitosan solution. And also chitosan could provide –NH₂ for binding PdNWs. After that, 2 mg PdNWs were added into the above solution with stirring overnight. Finally, the prepared PdNWs–Ag-g-C₃N₄ composite was centrifuged and washed several times in ethanol and ultrapure water, and then redispersed in ultrapure water for future use.

2.6. Construction of the sensor

First, a bare glassy carbon electrode (GCE, Φ = 4.0 mm) was polished to obtain a mirror-like surface according to previous literature.¹⁵ Then, it was coated with 10 μL of PdNWs–Ag-g-C₃N₄ suspension. After drying, 10 μL as-prepared phenoxy dextran (DexP) solution was incubated on the PdNWs–Ag-g-C₃N₄/GCE for 1 h via π–π interaction. After the nonspecific adsorption sites were blocked with 0.1% Tween20 (T20), 10 μL 3.5 mg mL⁻¹ ConA was incubated on the electrode for 1 h. Subsequently, the sensor was incubated in glucose solution for 5 min and then washed carefully with ultrapure water, and the ECL response of the modified electrodes was measured in 3.0 mL 0.10 M PBS (pH 7.4). During the ECL measurement, the following parameters were set: the voltage of 600 V for the photomultiplier tube, and the scan rate of 0.3 V s⁻¹.

3. Results and discussion

3.1. Characterization of the synthesized nanomaterials

The morphologies of PdNWs, Ag-g-C₃N₄ NSs and PdNWs–Ag-g-C₃N₄ were characterized by SEM. The SEM images shown in Fig. 1A and B indicate that PdNWs have a nanowire-like structure and Ag-g-C₃N₄ NSs have a platelet-like structure. As expected, for PdNWs–Ag-g-C₃N₄ (Fig. 1C), both the layered structure and nanowire structure were clearly observed. And the PdNWs blended with Ag-g-C₃N₄ perfectly, indicating the successful preparation of PdNWs–Ag-g-C₃N₄ through the interaction between the amino groups from chitosan and PdNWs. The EDS spectrum of PdNWs–Ag-C₃N₄ is shown in Fig. 1D. The EDS spectrum of the PdNWs–Ag-C₃N₄ clearly showed the existence of Ag, C, N and Pd elements, also revealing the successful synthesis of PdNWs–Ag-C₃N₄. TEM and high resolution TEM were performed on the synthesized PdNWs–Ag-C₃N₄ nanocomposite to obtain information on the size and lattice structure of PdNWs. As seen from Fig. 1E and F, a typical Ag-C₃N₄ morphology of small thin nanosheets with wrinkles and an irregular shape and a typical PdNW morphology of nanowires were both observed, which was in accordance with previous literature.³¹

Fig. 1 SEM images of (A) PdNWs, (B) Ag-g-C₃N₄, (C) PdNWs–Ag-g-C₃N₄, and the EDS spectrum of (D) PdNWs–Ag-g-C₃N₄. Typical TEM image (E) and the HRTEM image (F) of PdNWs–Ag-g-C₃N₄.

In order to further prove the successful synthesis of PdNWs–Ag-g-C₃N₄, XPS was used for elemental analysis. As expected, characteristic peaks for Pd 3d, Ag 3d, C 1s and N 1s core-level regions are clearly observed in the survey spectrum of PdNWs–Ag-g-C₃N₄ (Fig. 2A). Fig. 2B presents the XPS signature of the Ag 3d_5/2 and Ag 3d_3/2 peaks, separately located at 367.48 and 373.48 eV. And the Pd 3d_3/2 and Pd 3d_5/2 peaks are respectively located at 339.98 eV and 334.58 eV, as seen in Fig. 2C. The peaks at 287.88 eV and 284.78 eV assigned to C 1s, and 398.48 eV assigned to N 1s are respectively observed in Fig. 3D–E. XPS results confirm the successful preparation of PdNWs–Ag-g-C₃N₄.

Fig. 2 XPS spectra of the (A) full region for PdNWs–Ag-g-C₃N₄, (B) Ag 3d region, (C) Pd 3d region, (D) C 1s region and (E) N 1s region.

Fig. 3 (A) ECL responses in PBS (pH 7.4) containing 0.10 M K₂S₂O₈ and (B) CVs in 5.0 mM K₃[Fe(CN)₆]/K₄[Fe(CN)₆] at the bare GCE (a), PdNWs–Ag-g-C₃N₄/GCE (b), DexP/PdNWs–Ag-g-C₃N₄/GCE (c), Tween20/DexP/PdNWs–Ag-g-C₃N₄/GCE (d), and ConA/Tween20/DexP/PdNWs–Ag-g-C₃N₄/GCE (e). Scan rate of 300 mV s⁻¹ for ECL and 100 mV s⁻¹ for CV measurements.

3.2. ECL and CV characterization of the sensor

As shown in Fig. 3A, the bare GCE in 0.1 M K₂S₂O₈ solution produced a low ECL intensity (curve a). When the PdNWs–Ag-g-C₃N₄ was modified onto the surface of the GCE, a high ECL signal was obtained (curve b), indicating the successful immobilization of PdNWs–Ag-g-C₃N₄ on the electrode. The plots of ECL intensity versus the potential of PdNWs–Ag-C₃N₄ are provided in Fig. S2 of ESI.† From curves c to e, it is observed that the ECL intensity dropped sequentially with the successive modification of DexP, T20 and ConA onto the electrode, which was attributed to their non-electroactive properties hindering electron transfer.

The cyclic voltammograms (CVs) of different modified electrodes were recorded in 5.0 mM K₃[Fe(CN)₆]/K₄[Fe(CN)₆], and the results are displayed in Fig. 3B. A couple of quasi-reversible redox peaks of [Fe(CN)₆]⁴⁻/³⁻ were observed on the bare GCE (curve a). When the PdNWs–Ag-g-C₃N₄, DexP, T20 and ConA were successively modified onto the bare GCE (curves b–e), the redox peak currents continuously decreased due to the electron transmission insulation of PdNWs–Ag-g-C₃N₄, DexP, T20 and ConA.

3.3. Optimization of measurement conditions

Among the factors influencing the performance of the biosensor, the concentration of ConA and K₂S₂O₈, and the incubation time of glucose were important in this experiment. The effect of the concentration of ConA on the ECL emission of the modified electrode was investigated. As shown in Fig. 4A, with the concentration of ConA increasing from 1 to 3.5 mg mL⁻¹, the ECL intensity decreased. When the concentration of ConA exceeded 3.5 mg mL⁻¹, the intensity changed slowly, and thus the concentration of ConA was chosen as 3.5 mg mL⁻¹.

Fig. 4 (A) Effect of ConA concentration on the ECL responses in the presence of 0.10 M K₂S₂O₈. (B) Effect of K₂S₂O₈ concentration on the ECL responses. (C) Effect of incubation time on the ECL responses in PBS (pH 7.4) containing 0.10 M K₂S₂O₈. Scan rate: 300 mV s⁻¹. Scan voltage: −1.30–0.0 V.

Fig. 4B and C depict the influence of K₂S₂O₈ concentration (0.06–0.13 M) and the influence of incubation time of glucose (1–9 min) on the ECL response of the sensor incubated with 2.0 × 10⁻⁸ M glucose respectively. As shown in Fig. 4B and C, the maximum ECL signal was obtained at 0.10 M K₂S₂O₈ and 5 min of incubation time of glucose respectively. Taking into account the possible application of the sensor, 0.10 M K₂S₂O₈ and 5 min of incubation time of glucose were selected in further experiments.

3.4. The ECL mechanism of the “on–off–on” switch system

To confirm the function of PdNWs and Ag-g-C₃N₄ in this construction strategy, the effect of PdNWs and Ag-g-C₃N₄ on the ECL signal in 0.10 M K₂S₂O₈ solution was investigated and the results are shown in Fig. 5A. As seen from Fig. 5A, compared with the bare electrode (blue curve), an obvious increase in ECL intensity by about 14 365 a.u. was observed at the Ag-g-C₃N₄/GCE (red curve), indicating that Ag-g-C₃N₄ was a nice luminophore. The possible electrochemical reaction of Ag-g-C₃N₄ and co-reactant S₂O₈²⁻ on the electrode is as follows:³³

Ag-g-C₃N₄ + e⁻ → Ag-g-C₃N₄˙⁻ (1)

O₂ + 2H₂O + 2e⁻ → 2H₂O₂ (2)

H₂O₂ + e⁻ → OH⁻ + OH˙ (3)

OH˙ + Ag-g-C₃N₄˙⁻ → Ag-g-C₃N₄* + OH⁻ (4)

S₂O₈²⁻ + e⁻ → SO₄²⁻ + SO₄˙⁻ (5)

Ag-g-C₃N₄˙⁻ + SO₄˙⁻ → Ag-g-C₃N₄* + SO₄²⁻ (6)

and/or

Ag-g-C₃N₄ + SO₄˙^─ → Ag-g-C₃N₄⁺ + SO₄²⁻ (7)

Ag-g-C₃N₄⁺ + Ag-g-C₃N₄˙^─ → Ag-g-C₃N₄* + Ag-g-C₃N₄ (8)

finally,

Ag-g-C₃N₄* → Ag-g-C₃N₄ + hν (9)

Fig. 5 (A) The comparison of ECL responses at the biosensor with different nanomaterials: bare GCE (blue curve), Ag-g-C₃N₄/GCE (red curve), and (c) PdNWs–Ag-g-C₃N₄/GCE (black curve). (B) ECL responses of the “on–off–on” mechanism.

For the sensor incubated with the PdNWs–Ag-g-C₃N₄, the ECL emission reached 17 292 a.u. (black curve), which was greater than that in the case of Ag-g-C₃N₄ (red curve), indicating a signal amplification role of PdNWs in this sensor. Such an increase might be due to the fact that PdNWs had good electroconductivity, which accelerated the electron transfer between S₂O₈²⁻ and the electron surface (Fig. S1†). Additionally, the large specific surface area of PdNWs–Ag-g-C₃N₄ served as the support to achieve a high loading of recognition element DexP on the PdNWs–Ag-g-C₃N₄ for binding ConA.

The PdNWs–Ag-g-C₃N₄ was employed to effectively improve the ECL intensity, achieving the first “signal-on” state in this work (Fig. 5B blue curve). When the resultant electrode was reacted with DexP and T20, the ECL response decreased. The “signal-off” state (black curve) was obtained by incubating ConA onto the electrode via the special binding between ConA and DexP, and the ECL signal reduced due to the non-electroactive properties hindering electron transfer. Finally, the modified electrode was immersed into the glucose solution, the glucose would compete with DexP for ConA binding sites due to the higher binding affinity of glucose for ConA than DexP, and thus ConA would be taken away from the electrode. Therefore, the hindering effect from ConA incubated on the electrode would be weakened, resulting in an increase in ECL intensity and achieving the second “signal-on” state (red curve).

3.5. Performance of the sensor

3.5.1. Calibration curve. The quantitative measurements of glucose were performed by using the proposed sensor under optimal conditions in the “on–off–on” mode. As illustrated in Fig. 6, the ECL signal increased with the increase of glucose concentrations from 1.0 × 10⁻¹¹ M to 1.67 × 10⁻⁶ M (Fig. 6A, curves a–h), and the I_ECL exhibited a brilliant correlation with the logarithmic value of glucose concentration (lg c) (Fig. 6B). The regression equation was I = 22 849.24 + 1830.81 lg  c with an R value of 0.995. Additionally, the limit of detection (LOD) was calculated to be 3.0 × 10⁻¹² M (S/N = 3). Compared to other reported sensors listed in Table 1, the developed ECL sensor exhibited high sensitivity for glucose determination with a relatively low LOD. In this work, our prepared sensor indeed exhibited a high sensitivity and wide linearity, which may be ascribed to the combination of the high ECL efficiency of PdNWs–Ag-g-C₃N₄ and the competitive binding reaction of the glucose–ConA–dextran system. The reasonable explanation is as follows. First, the ECL technique itself exhibited a high sensitivity. Second, Ag-g-C₃N₄ itself exhibited an excellent ECL behavior and can generate a strong ECL intensity at cathodic potentials. With introducing PdNWs into the Ag-g-C₃N₄, a higher initial intensity was achieved. Third, the competitive binding reaction of the glucose–ConA–dextran system may provide the proposed sensor with excellent performance.³⁴

Fig. 6 (A) ECL responses of the sensor to 1.0 × 10⁻¹¹ (a), 1.1 × 10⁻¹⁰ (b), 1.11 × 10⁻⁹ (c), 6.11 × 10⁻⁹ (d), 1.61 × 10⁻⁸ (e), 6.61 × 10⁻⁸ (f), 1.66 × 10⁻⁷ (g) and 1.67 × 10⁻⁶ (h) glucose in PBS (pH 7.4) containing 0.10 M K₂S₂O₈. Scan rate: 300 mV s⁻¹. Scan voltage: −1.30–0.0 V. (B) The linear calibration curve.

Table 1 Comparison of different glucose sensors

Electrode	Methods	Linear range/M	Detection limit/M	Ref.
a SWNTs: single-walled carbon nanotubes. b GE: gold electrode. c GO: graphene oxide. d Att: attapulgite. e GOx: glucose oxidase. f CdS: cadmium sulfide.
ConA/DexP/SWNTs^a/GE^b	Chemiresist	1.0 × 10^−12—1.0 × 10^−9	—	34
ConA/DexP/GO^c/GCE	EIS	5.0 × 10^−6—9.0 × 10^−3	3.4 × 10^−7	35
Att^d-TiO2/GCE	ECL	1.0 × 10^−9—1.0 × 10^−3	1.0 × 10^−11	36
GOx^e/CdS^f/GCE	ECL	4.0 × 10^−13—4.0 × 10^−10	2.0 × 10^−13	37
ConA/DexP/PdNWs- Ag-g-C3N4/GCE	ECL	1.0 × 10^−11—1.67 × 10^−6	3.0 × 10^−12	This work

---

3.5.2. Stability and selectivity of the sensor. The stability of the sensor under consecutive cyclic potential scans was evaluated in the presence of 2.0 × 10⁻⁸ M glucose. As seen from Fig. 7A, after consecutive cyclic scans for 15 cycles, the ECL signal revealed inconspicuous undulation with a relative standard deviation (RSD) of 0.87%, showing a nice stability of the sensor. Additionally, Fig. S3† shows the long-term storage stability of our proposed sensor. The ECL response of the sensor towards 2.0 × 10⁻⁸ M glucose was tested every day. After 12 days, the ECL intensity decreased 15.2%, demonstrating that the sensor possessed good stability.

Fig. 7 (A) Operational stability of the sensor incubated with 2.0 × 10⁻⁸ M glucose in PBS (pH 7.4) containing 0.10 M K₂S₂O₈. (B) The ECL response of the sensor to different samples: (a) ascorbic acid (0.10 mM), (b) maltose (0.10 μM), (c) lactose (0.10 μM), (d) serine (1.0 mg mL⁻¹), (e) Hg²⁺ (0.10 μM), (f) Cu²⁺ (0.10 μM), (g) blank, (h) glucose (2.0 × 10⁻⁸ M) and (i) a mixture containing the above samples.

In addition, ascorbic acid (0.10 mM), maltose (0.10 μM), lactose (0.10 μM), serine (1.0 mg mL⁻¹), Hg²⁺ (0.10 μM) and Cu²⁺ (0.10 μM) acting as interfering substances were used for evaluating the selectivity of the sensor in glucose determination. According to Fig. 7B, a low ECL signal was observed when incubated with ascorbic acid, dopamine, uric acid, maltose, lactose and serine, respectively, which showed little fluctuations compared with the blank experiment, while the target 2.0 × 10⁻⁸ M glucose and the mixture containing the above interfering substances and 2.0 × 10⁻⁸ M glucose showed significant ECL signals. The results above verified that ascorbic acid, dopamine, uric acid, maltose, lactose and serine cannot cause interference to the proposed sensor, indicating an outstanding specificity of the sensor.

3.5.3. Application of the sensor. The application potential of the sensor was determined by standard addition methods in bovine serum albumin. The results are shown in Table 2 and the recoveries ranged from 91.3% to 108%, which verified the outstanding analysis accuracy and promising application potential of this ECL sensor.

Table 2 Recoveries of glucose by the biosensor in bovine serum albumin samples

Samples	Added (mol L^−1)	Found^a (mol L^−1)	Recovery (%)
a Average of three determinations ± SD, n = 3.
1	1.00 × 10^−9	9.13 × 10^−10 ± 8.68 × 10^−11	91.3
2	1.00 × 10^−8	9.43 × 10^−9 ± 1.48 × 10^−9	94.3
3	1.00 × 10^−7	1.08 × 10^−7 ± 1.70 × 10^−8	108

---

4. Conclusions

In conclusion, we designed a novel ECL sensor for ultrasensitive detection of glucose. Compared with conventional glucose detection methods, the proposed ECL sensor exhibited its own advantages. First, it could be expediently constructed by a simple “on–off–of” strategy. Second, it combined the advantages of the PdNWs–Ag-g-C₃N₄ hybrid and the competition reaction between glucose and dextran for ConA binding sites. Third, it improved the detection sensitivity with a relatively low LOD. With these advantages, the proposed “on–off–on” strategy could be utilized as a versatile and effective method for detection of other targets.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was financially supported by the NNSF of China (Grant no. 81603242) and the Discipline Talent Promotion Program of “Xinglin Scholars” of Chengdu University of TCM (Grant no. ZRQN2018012).

References

1. J. Wang, P. N. Zhang, Y. J. Xiahou, D. Y. Wang, H. B. Xia and H. Möhwald, Simple synthesis of Au-Pd alloy nanowire networks as macroscopic, flexible electrocatalysts with excellent performance, ACS Appl. Mater. Interfaces, 2018, 10, 602–613.
2. C. Koenigsmann and S. S. Wong, Tailoring chemical composition to achieve enhanced methanol oxidation reaction and methanol-tolerant oxygen reduction reaction performance in palladium-based nanowire systems, ACS Catal., 2013, 3, 2031–2040.
3. J. H. Shim, J. Kim, C. Lee and Y. Lee, Porous Pd layer-coated Au nanoparticles supported on carbon: synthesis and electrocatalytic activity for oxygen reduction in acid media, Chem. Mater., 2011, 23, 4694–4700.
4. T. Yuan, H. Y. Chen, X. H. Ma, J. J. Feng, P. X. Yuan and A. J. Wang, Simple synthesis of self-supported hierarchical AuPd alloyed nanowire networks for boosting electrocatalytic activity toward formic acid oxidation, J. Colloid Interface Sci., 2018, 513, 324–330.
5. L. F. Zhang, S. L. Zhong and A. W. Xu, Highly branched concave Au/Pd bimetallic nanocrystals with superior electrocatalytic activity and highly efficient SERS enhancement, Angew. Chem., Int. Ed., 2013, 52, 645–649.
6. M. K. Debe, Electrocatalyst approaches and challenges for automotive fuel cells, Nature, 2012, 486, 43–51.
7. X. Q. Zeng, M. L. Latimer, Z. L. Xiao, S. Panuganti, U. Welp, W. K. Kwok and T. Xu, Hydrogen gas sensing with networks of ultrasmall palladium nanowires formed on filtration membranes, Nano Lett., 2011, 11, 262–268.
8. L. L. Wei, R. Gu and J. M. Lee, Highly efficient reduction of hexavalent chromium on amino-functionalized palladium nanowires, Appl. Catal., B, 2015, 176, 325–330.
9. X. Y. Yang, J. Quan, S. Y. Liu and F. S. Liu, The release behavior of helium in palladium nanowire: A molecular dynamics study, Comput. Mater. Sci., 2016, 111, 190–196.
10. Y. F. He, J. T. Feng, Y. Y. Du and D. Q. Li, Controllable synthesis and acetylene hydrogenation performance of supported Pd nanowire and cuboctahedron catalysts, ACS Catal., 2012, 2, 1703–1710.
11. Y. J. Song, R. M. Garcia, R. M. Dorin, H. R. Wang, Y. Qiu, E. N. Coker, W. A. Steen, J. E. Miller and J. A. Shelnutt, Synthesis of platinum nanowire networks using a soft template, Nano Lett., 2007, 7, 3650–3655.
12. Y. S. Kim, H. J. Kim and W. B. Kim, Composited hybrid electrocatalysts of Pt-based nanoparticles and nanowires for low temperature polymer electrolyte fuel cells, Electrochem. Commun., 2009, 11, 1026–1029.
13. S. Sun, F. Jaouen and J. P. Dodelet, Controlled growth of Pt nanowires on carbon nanospheres and their enhanced performance as electrocatalysts in PEM fuel cells, Adv. Mater., 2008, 20, 3900–3904.
14. J. H. Luo, Q. Li, S. H. Chen and R. Yuan, Coreactant-free dual amplified electrochemiluminescent biosensor based on conjugated polymer dots for the ultrasensitive detection of microRNA, ACS Appl. Mater. Interfaces, 2019, 11, 27363–27370.
15. C. Zhang, Y. Fan, H. Zhang, S. H. Chen and R. Yuan, An ultrasensitive signal-on electrochemiluminescence biosensor based on Au nanoclusters for detecting acetylthiocholine, Anal. Bioanal. Chem., 2019, 411, 905–913.
16. H. J. Wang, Y. L. Yuan, Y. Q. Chai and R. Yuan, Self-enhanced electrochemiluminescence immunosensor based on nanowires obtained by a green approach, Biosens. Bioelectron., 2015, 68, 72–77.
17. P. Y. Ran, S. Zhu, J. Y. Song, J. L. Wu, F. J. Mo and Y. Z. Fu, A widened emission window of the peroxydisulfate-oxygen system for the detection of L-alanine, Colloids Surf., B, 2018, 169, 418–421.
18. C. Ye, M. Q. Wang, X. Zhong, S. H. Chen, Y. Q. Chai and R. Yuan, Highly sensitive electrochemiluminescence assay of acetylcholinesterase activity based on dual biomarkers using Pd-Au nanowires as immobilization platform, Biosens. Bioelectron., 2016, 79, 34–40.
19. Y. Zhuo, H. Wang, Y. Lei, P. Zhang, J. Liu, Y. Chai and R. Yuan, Electrochemiluminescence biosensing based on different modes of switching signal, Analyst, 2018, 3230–3248.
20. X. J. Du, D. Jiang, N. Hao, Q. Liu, J. Qian, L. M. Dai, H. P. Maoa and K. Wang, An ON¹–OFF–ON² electrochemiluminescence response: combining the intermolecular specific binding with a radical scavenger, Chem. Commun., 2015, 51, 11236–11239.
21. Y. Zhou, H. J. Wang, Y. Zhuo, Y. Q. Chai and R. Yuan, A highly-efficient electrochemiluminescent Ag nanoclusters/TiO₂ nanomaterials as signal probe for ferrocene-driven light switch bioanalysis, Anal. Chem., 2017, 89, 3732–3738.
22. Y. M. Lei, M. Zhao, A. Wang, Y. Q. Yu, Y. Q. Chai, R. Yuan and Y. Zhuo, Electrochemiluminescence of supramolecular nanorods and their application in the “on-off-on” detection of copper ions, Chem.–Eur. J., 2016, 22, 8207–8214.
23. M. Zhao, Y. Zhuo, Y. Q. Chai and R. Yuan, Au nanoparticles decorated C₆₀ nanoparticle-based label-free electrochemiluminesence aptasensor via a novel “on–off–on” switch system, Biomaterials, 2015, 52, 476–483.
24. H. F. Xu, S. J. Liang, X. Zhu, X. Q. Wu, Y. Q. Dong, H. S. Wu, W. X. Zhang and Y. W. Chi, Enhanced electrogenerated chemiluminescence behavior of C₃N₄ QDs@C₃N₄ nanosheet and its signal-on aptasensing for platelet derived growth factor, Biosens. Bioelectron., 2017, 92, 695–701.
25. H. F. Sha, Y. Zhang, Y. F. Wang, H. Ke, X. Xiong and N. Q. Jia, Electrochemiluminescence resonance energy transfer biosensor between the glucose functionalized MnO₂ and g-C₃N₄ nanocomposites for ultrasensitive detection of concanavalin A, Biosens. Bioelectron., 2019, 124, 59–65.
26. X. D. Zhang, X. Xie, H. Wang, J. J. Zhang, B. C. Pan and Y. Xie, Enhanced Photoresponsive Ultrathin Graphitic-Phase C₃N₄ Nanosheets for Bioimaging, J. Am. Chem. Soc., 2013, 135, 18–21.
27. X. L. Fu, F. Hou, F. R. Liu, S. W. Ren, J. T. Cao and Y. M. Liu, Electrochemiluminescence energy resonance transfer in 2D/2D heterostructured g-C₃N₄/MnO₂ for glutathione detection, Biosens. Bioelectron., 2019, 129, 72–78.
28. Y. F. Wang, W. W. Guo and N. Q. Jia, High-sensitivity electrochemiluminescence biosensor based on silver-carbon nitride for the detection of dopamine utilizing enhancement effects, ChemElectroChem, 2018, 5, 1–8.
29. R. P. Liang, L. D. Yu, Y. J. Tong, S. H. Wen, S. P. Cao and J. D. Qiu, An ultratrace assay of arsenite based on the synergistic quenching effect of Ru(bpy)₃²⁺ and arsenite on the electrochemiluminescence of Au-g-C₃N₄ nanosheets, Chem. Commun., 2018, 54, 14001–14004.
30. R. F. Zhu, Y. H. Zhang, X. Fang, X. Q. Cui, J. Wang, C. C. Yue, W. H. Fang, H. Zhao and Z. X. Li, In situ sulfur-doped graphitic-like carbon nitride nanosheets with enhanced electrogenerated chemiluminescence and used for sensitive and selective sensing of L-cysteine, J. Mater. Chem. B, 2019, 7, 2320–2329.
31. S. W. Hu, L. W. Yang, Y. Tian, X. L. Wei, J. W. Ding, J. X. Zhong and P. K. Chu, Simultaneous nanostructure and heterojunction engineering of graphitic carbon nitride via in situ Ag doping for enhanced photoelectrochemical activity, Appl. Catal., B, 2015, 163, 611–622.
32. P. W. Barone and M. S. Strano, Reversible control of carbon nanotube aggregation for a glucose affinity sensor, Angew. Chem., Int. Ed., 2006, 45, 8138–8141.
33. Y. Fan, X. R. Tan, X. Ou, Q. Y. Lu, S. H. Chen and S. P. Wei, A novel “on-off” electrochemiluminescence sensor for the detection of concanavalin A based on Ag-doped g-C₃N₄, Electrochim. Acta, 2016, 202, 90–99.
34. L. N. Cella, W. Chen, N. V. Myung and A. Mulchandani, Single-walled carbon nanotube-based chemiresistive affinity biosensors for small molecules: ultrasensitive glucose detection, J. Am. Chem. Soc., 2010, 132, 5024–5026.
35. J. Huang, L. Zhang, R. P. Liang and J. D. Qiu, “On-off” switchable electrochemical affinity nanobiosensor based on graphene oxide for ultrasensitive glucose sensing, Biosens. Bioelectron., 2013, 41, 430–435.
36. Y. Z. Wang, H. Zhong, X. R. Li, G. Q. Liu, K. Yang, M. Ma, L. L. Zhang, J. Z. Yin, Z. P. Cheng and J. K. Wang, Nonenzymatic electrochemiluminescence glucose sensor based on quenching effect on luminol using attapulgite-TiO₂, Sens. Actuators, B, 2016, 230, 449–455.
37. B. Qiu, M. Miao, L. S. She, X. F. Jiang, Z. Y. Lin and G. N. Chen, An ultrasensitive biosensor for glucose based on solidstate electrochemiluminescence on GOx/CdS/GCE electrode, Anal. Methods, 2013, 5, 1941–1945.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c9ay02357g

---