An electrogenerated chemiluminescent biosensor based on a g-C₃N₄–hemin nanocomposite and hollow gold nanoparticles for the detection of lactate

Abstract

In this article, a new electrochemiluminescent (ECL) biosensor based on a g-C₃N₄–hemin nanocomposite and hollow gold nanoparticles (HGNPs) was constructed to detect lactate. Firstly, the g-C₃N₄ nanosheets were prepared through ultrasonication-assisted liquid exfoliation of bulk g-C₃N₄, which was obtained through polymerizing melamine under 600 °C. Then, the nanocomposites of g-C₃N₄ nanosheets and hemin were prepared to modify a glassy carbon electrode. Subsequently, HGNPs were self-assembled onto the electrode for adsorbing lactate oxidase to achieve a lactate biosensor. Due to the excellent catalytic effect of g-C₃N₄–hemin and HGNPs on the luminol/H₂O₂ ECL system, the as-prepared biosensor exhibited a good response performance to lactate with a linear range of 1.7 × 10⁻⁸ to 5.0 × 10⁻⁴ M and a detection limit of 5.5 × 10⁻⁹ M. In addition, the prepared ECL biosensor exhibited satisfying reproducibility and stability. The g-C₃N₄–hemin nanocomposite might have great potential application in a luminol/H₂O₂ ECL system.

---

Introduction

Lactate is a product of the anaerobic metabolism of glucose. A high level of lactate in the blood will cause lactate poisoning. What's more, as a biomarker, lactate excreted in blood¹ and in sweat² could be used for monitoring various diseases, such as liver diseases,³ heart failure,¹ drug toxicity,⁴ metabolic disorders,⁵ and mortality in ventilated infants.⁶ It is necessary to develop a sensitive, efficient, and reliable method to detect lactate because of the significance of lactate in clinical medicine,⁷ sports medicine,⁸ and food industry.⁹ Many techniques have been used to measure lactate concentration, such as high performance liquid chromatography,¹⁰ refractive index detection,^11,12 and ultraviolet spectrophotometry.¹³ However, these techniques have some limitations such as cumbersome process, time consumption or cost effectiveness.

Recently, electrogenerated chemiluminescence (ECL) technique has been widely used in analytical chemistry,^14,15 food safety fields,¹⁶ and immunoassay.^17,18 When ECL technique was used in analytical detection, this analytical method exhibited low detection limit, simple instrumentation, low background signal and high sensitivity. Luminol, as a fluorophoreas, generates strong ECL signals in the presence of hydrogen peroxide (H₂O₂).¹⁹ Most oxidase (glucose oxidase, lactate oxidase, or cholesterol oxidase) would catalyze the oxidation of corresponding substrate to generate H₂O₂, which can enhance the ECL signals of luminol, thus achieving the detection of corresponding substrates. Currently, biosensors based on lactate oxidase (LOX) have been widely used for the determination of lactate in luminol/H₂O₂ ECL system. For example, Haghighi and partners fabricated a highly sensitive ECL biosensor using ZnO nanoparticles decorated multiwalled carbon nanotubes for detecting lactate in human blood plasma samples.²⁰

Luminol generates strong ECL signals in alkaline media, and weak ECL signals in physiological pH conditions. However, most of the enzymes have good catalytic effects in physiological pH. This fact would limit the high sensitive detection in this system. Therefore, it needs to look for appropriate materials to catalyze substrate for signal amplification in physiological conditions. Gold nanoparticles (AuNP) have been widely used as catalysts in ECL reactions to enhance the sensitivity and expand new applications of AuNP in biosensor²¹ and diagnostics²² owing to the unique chemical and physical properties, including their functional versatility,²³ biocompatibility,²⁴ and low toxicity.^25,26 As a unique shape of AuNP, hollow gold nanospheres (HGNPs) have attracted more attention for their tunable interior and exterior diameters. Compared with AuNP, the hollow structure of HGNPs allow more molecules to adsorb on the inside surface of the walls as well on the outside surface.²⁷ Previous studies have proved that HGNPs are promising for chemical and biological sensing applications.

Recently, a novel metal-free graphite carbon nitride (g-C₃N₄) has attracted considerable attention and widely used in the fields such as catalysis, degradation and sensor,^28–30 owing to the remarkable features including outstanding electronic and catalytic properties, luminescence performance and high thermal and chemical stability. However, the water-solubility of bulk g-C₃N₄ is poor, which limited its applications in aqueous solution. Fortunately, g-C₃N₄ nanosheets could broaden the application of g-C₃N₄ since nanosheets exhibit good water-solubility,³¹ coordination points to metal ion,³² and superior physicochemical properties, such as shorter photoinduced charger carriers transferring distance and higher surface-area-to-volume ratio. Tian and his co-workers firstly demonstrated that g-C₃N₄ nanosheets exhibited high electrocatalytic activity toward the reduction of H₂O₂, and further investigated its application for electrochemical glucose biosensing.³³ In addition, the Fe–g-C₃N₄ hybrids were used to construct a non-enzyme glucose sensor for the determination of glucose with high stability and selectivity.³⁴ All the facts indicated that the g-C₃N₄ would provide a promising application in the sensing system. However, up to now, to the best of our knowledge, g-C₃N₄ nanosheets have not been reported in luminol/H₂O₂ ECL system.

Inspired by above observation, our aim in this work is to integrate the g-C₃N₄, hemin and HGNPs to achieve a highly sensitive lactate biosensor. Firstly, g-C₃N₄ nanosheets serve as a carrier to combine with hemin by π–π stacking to form g-C₃N₄–hemin nanocomposite. Then, a lactate biosensor was constructed based on the combination of g-C₃N₄–hemin and HGNPs. Due to the fact that both g-C₃N₄–hemin nanocomposite and HGNPs could enhance the ECL intensity of luminol/H₂O₂ system, the prepared biosensor exhibited a low detection limit, good accuracy, and high sensitivity for the lactate. Details of the preparation, characterization, optimization of conditions and performance of biosensor were described as follows.

Experimental

Reagents and materials

Melamine (C₃H₆N₃, 2,4,6-triamino-1,3,5-trazine, 99%), sodium citrate (Na₃C₆H₅O₇·2H₂O), sodium borohydride (NaBH₄), cobalt dichloride (CoCl₂·6H₂O), chitosan, lactate and LOX were obtained from Aladdin Ltd. (Shanghai, China). Gold chloride tetrahydrate (HAuCl₄·4H₂O) and hemin were obtained from Sigma Chemical Co. (St. Louis, MO, USA). 0.10 M phosphate-buffered saline (PBS) solutions with various pH were produced with 0.10 M Na₂HPO₄ and 0.10 M KH₂PO₄. The supporting electrolyte was 0.10 M KCl. Ultra pure water was used throughout the whole experimental process. All other chemicals were of analytical grade without further purification.

Apparatus

The ECL emission was monitored by a MPI-E electrochemical analyser (Xi'an Remax Analyse Instrument Co. Ltd, Xi'an, China) with the voltage of the photomultiplier tube (PMT) set at 800 V in the detection process. Cyclic voltammetry (CV) was performed with a CHI 600D electrochemical work station (Shanghai CH Instruments Co., China). The transmission electron microscopy (TEM) was carried out on TECNAI 10 (Philips Fei Co., Hillsboro). Fourier Transform Infrared spectroscopy (FT-IR) was performed on a Nexus 670 FT-IR spectrophotometer (Nicolet Instruments) using a KBr pellets. The topographs of different modified films were investigated with atomic force microscopy (AFM, MFD-3D, AR sylum Research). X-ray photoelectron spectroscopy (XPS) was carried out on Thermo ESCALAB 250 spectrometer (SID-Molecular). The crystal structure of the samples was investigated using powder X-ray diffraction (XRD) (Purkinje General InstrumentXD-3, Japan) with Cu Kα radiation (λ = 0.15406 nm). The morphologies and sizes of various nanomaterials were analyzed using a scanning electron microscopy (SEM) (SEM, S-4800, Hitachi, Japan) with an acceleration voltage of 10 kV.

Preparation of g-C₃N₄ nanosheets

The bulk g-C₃N₄ was prepared by direct pyrolysis of melamine in the semiclosed system according to a previous reported method in the literature with a little modification.³² In detail, 20 g melamine was added into an alumina crucible and heated at 600 °C for 2 h under air condition with a ramp rate of about 3 °C min⁻¹. The obtained yellow products were bulk g-C₃N₄. Then, 200 mg bulk g-C₃N₄ power was dispersed in 200 mL of ultra pure water with ultrasound. The obtained suspension was centrifuged at 5000 rpm to remove the residual unexfoliated g-C₃N₄. Then, the supernatant was dried in air to get g-C₃N₄ nanosheets. The corresponding synthesis process is shown in Scheme 1.

Scheme 1 Preparation of g-C₃N₄–hemin composite and fabrication process of the biosensor.

Preparation of hollow gold nanoparticles (HGNPs)

HGNPs were synthesized according to the literature²⁷ with some modification. The containers used in the synthesis process were cleaned using chromic acid solution. 200 μL of sodium citrate solution (0.10 M) was added into 50 mL of ultra pure water with rapid magnetic stirring under the nitrogen (N₂) atmosphere. Following that, 200 μL of freshly made NaBH₄ solution (1.0 M) was added. Subsequently, 100 μL CoCl₂ solution (0.50 M) was added into above solution under stirring. The color of the solution changed from pale pink to gray. Finally, 150 μL HAuCl₄ solution (0.10 M) was slowly added in drops. The color of the solution changed from dark brown to amaranth. The remaining cobalt nanoparticles left in the solution were oxidized under ambient conditions with rapid stirring for 1 h to obtain the HGNPs, which were collected by centrifugation and washed several times with ultra pure water. The resultant HGNPs were dispersed in 5 mL ultra pure water.

Fabrication of the biosensor

g-C₃N₄ nanosheets serve as a nanocarrier for hemin to achieve g-C₃N₄–hemin composite by π–π stacking. The details are as follows. First, 2 mg g-C₃N₄ and 2 mg hemin were dispersed in 1 mL pure water by stirring at 65 °C for 8 h. Then, the mixture was centrifuged several times at 12 000 rpm to obtain g-C₃N₄–hemin composite, which were redispersed in 1.0 mL chitosan solution (0.2%) in ultrasound conditions for further use.

The glassy carbon electrode (GCE, ϕ = 4 mm) was polished cleanly with 0.3 and 0.05 μm alumina slurry, and then rinsed with ethanol and water by sonication, respectively. Then, 10 μL dispersion of g-C₃N₄–hemin was cast onto a pretreated GCE surface. After dried in the air, the electrode was soaked in HGNPs colloid solution for 10 h to self-assembly HGNPs. After that, 10 μL LOX was dropped on the surface of electrode to structure a lactate biosensor (noted as LOX/HGNPs/g-C₃N₄–hemin/GCE). The corresponding fabrication process of the biosensor is shown in Scheme 1.

Experimental determination

The electrochemical measurements (CV and EIS) were performed in 3.0 mL 5.0 mM ferricyanide solution. The ECL detection was performed in 3.0 mL PBS solution containing luminol. The whole measurement process was supported by a conventional three electrode system at room temperature. A modified glassy carbon electrode was used as working electrode, a platinum wire as the auxiliary electrode, and a saturated calomel electrode (SCE) as reference electrode for electrochemical measurements and Ag/AgCl (saturated KCl) as reference electrode for ECL detection. The ECL determination was based on the change in ECL intensity (ΔI = I₀ − I_t). Herein, I_t and I₀ are the ECL signals with and without lactate, respectively.

Previous reports^35–37 have proved that the ECL of luminol in luminol/H₂O₂ system is related to various reactive oxygen species such as OH˙, O₂˙⁻, and other radical derivatives, which can oxidize luminol to generate the excited state. When the excited state fell to the ground state, and light was emitted. In our experiments, LOX modified on the biosensor would catalyze the oxidation of substrate lactate to generate H₂O₂, which further generated various reactive oxygen species (ROS). These ROS could oxidize luminol to produce the ECL signal. Under positive potential, with the help of g-C₃N₄–hemin composite, the luminol produced stronger ECL signals since both g-C₃N₄ and hemin could act as catalyst to enhance the ECL signal. The change of the ECL intensity was directly proportional to the concentration of lactate. The detection principle of the prepared biosensor is as follow.

Luminol + H₂O₂ → 3-aminophetalate + N₂ + hv

Results and discussion

The characterization of different nanomaterials

Fig. 1A shows the XRD patterns of bulk g-C₃N₄, g-C₃N₄ nanosheets, and g-C₃N₄–hemin nanocomposite. Bulk g-C₃N₄ (curve a) had two distinct diffraction peaks at 27.74° and 13.11°, which could be indexed for graphitic materials as the (002) and (100) peaks in JCPDS 87-1526.³⁸ These results were in good agreement with the report in the literature.^39,40 Compared with bulk g-C₃N₄, the g-C₃N₄ nanosheets (curve b) only showed a much weaker (002) peak at 27.74°, suggesting the exfoliation of bulk g-C₃N₄ after ultrasonication treatment. For the g-C₃N₄–hemin nanocomposite (curve c), a noticeably peak at 24°was noticed, which agreed with hemin.⁴¹ These observations indicated the formation of g-C₃N₄–hemin nanocomposite.

Fig. 1 (A) XRD of (a) bulk g-C₃N₄, (b) g-C₃N₄ nanosheets, and (c) g-C₃N₄–hemin nanocomposite; (B) FT-IR spectra of (a) hemin, (b) g-C₃N₄ nanosheets, and (c) g-C₃N₄–hemin nanocomposite.

Fig. 1B shows the FT-IR spectra of hemin, g-C₃N₄ nanosheets, and g-C₃N₄–hemin nanocomposite. As shown in curve a, the strong sharp peaks at 1700 cm⁻¹ and 3445 cm⁻¹ were attributed to C=O and –OH of hemin. For the IR spectrum of g-C₃N₄ nanosheets (curve b), the sharp peak at around 810 cm⁻¹ was originated from heptazine ring system. The peaks in the region from 900 to 1800 cm⁻¹ were attributed to either trigonal C–N(–C)–C (full condensation) or bridging C–NH–C units. The broad peaks between 3000 and 3600 cm⁻¹ were contributed by N–H stretching. The peaks at 1150 cm⁻¹ and 3073 cm⁻¹ were observed in the IR spectrum of g-C₃N₄–hemin nanocomposite (curve c), which were attributed to CH and CH=C of hemin. The peak at 810 cm⁻¹ was observed, which was due to the g-C₃N₄. The strong sharp band of the carbonyl acid groups of hemin at 1700 cm⁻¹ was disappeared and replaced by the carboxylate bands, v_as at 1543 cm⁻¹, due to the interactions of the carboxylic groups with basic ligands.

TEM was performed to characterize HGNPs, g-C₃N₄ nanosheets and g-C₃N₄–hemin composite. As shown in Fig. 2A, the hollow structure of HGNPs was clearly recognized in TEM image. The lamellar structure of nanosheets was clearly observed in the TEM image of g-C₃N₄ nanosheets (Fig. 2B). Compared with g-C₃N₄ nanosheets, the TEM image of g-C₃N₄–hemin composite (Fig. 2C) obviously changed and it presented an irregular, non-transparent and thicker flake structure, indicating the successful synthesis of g-C₃N₄ nanosheets and g-C₃N₄–hemin composite.

Fig. 2 TEM images of (A) HGNPs, (B) g-C₃N₄ nanosheets and (C) g-C₃N₄–hemin composite.

The XPS spectra of g-C₃N₄ nanosheets and g-C₃N₄–hemin composite samples were measured to further confirm the successful synthesis of the g-C₃N₄–hemin composite. The XPS survey of g-C₃N₄ nanosheets sample (Fig. 3A) showed that the g-C₃N₄ nanosheets mainly composed of C and N elements, which was consistent with the C₃N₄ stoichiometry. The signal of O1s was observed, which may be due to the adsorption of oxygen during the progress of polymerization or preparation of ultrathin g-C₃N₄ nanosheets. Besides the C1s and N1s, the Cl2p peak at 199.2 eV and the Fe2p peak at 710.8 eV, which were due to hemin, were observed in the XPS survey of g-C₃N₄–hemin composite (Fig. 3B), indicating the successful synthesis of g-C₃N₄–hemin composite.

Fig. 3 Wide scan survey of XPS spectra of (A) g-C₃N₄ nanosheets and (B) g-C₃N₄–hemin composite.

Characterization of the biosensor fabrication

The g-C₃N₄ nanosheets, g-C₃N₄–hemin, HGNPs/g-C₃N₄–hemin, and LOX/HGNPs/g-C₃N₄–hemin modified films were characterized by SEM. As shown in Fig. 4A, the nanosheet structure of g-C₃N₄ nanosheets was obviously observed. Fig. 4B exhibits the SEM image of g-C₃N₄–hemin nanocomposite modified film. The change in sheet structure indicated that g-C₃N₄–hemin nanocomposite was obtained. For the SEM image of HGNPs/g-C₃N₄–hemin (Fig. 4C), HGNPs were uniformly distributed on the whole surface of g-C₃N₄–hemin nanocomposite. The sizes of HGNPs were about 50–100 nm. Compared with Fig. 4C, the whole image became dimly in Fig. 4D, which was due to the nonconductive property of LOX coated on the HGNPs. The SEM images provided the proof of the successful synthesis of g-C₃N₄ nanosheets and g-C₃N₄–hemin composite, and the gradual modification of g-C₃N₄–hemin nanocomposite, HGNPs, and LOX on the electrode.

Fig. 4 SEM images of (A) g-C₃N₄ nanosheets, (B) g-C₃N₄–hemin, (C) HGNPs/g-C₃N₄–hemin, and (D) LOX/HGNPs/g-C₃N₄–hemin modified films.

The AFM technique was also employed to confirm the fabrication process of the biosensor. The AFM images of g-C₃N₄ nanosheets, g-C₃N₄–hemin nanocomposite, HGNPs/g-C₃N₄–hemin, and LOX/HGNPs/g-C₃N₄–hemin modified films are shown in Fig. 5. Fig. 5A clearly revealed the nanosheet structure of g-C₃N₄ nanosheets. Furthermore, it was observed that g-C₃N₄ nanosheets were well-separated from each other. Compared with Fig. 5A, the AFM image of g-C₃N₄–hemin composite (Fig. 5B) clearly changed, and the sheet structure for g-C₃N₄ changed to a stacked sheet-like structure for g-C₃N₄–hemin with an increased thickness due to the modification of hemin on the g-C₃N₄ by π–π stacking. After HGNPs were modified on the g-C₃N₄–hemin composite, the spherical particles of HGNPs were obviously observed at the corresponding AFM image (Fig. 5C). The AFM image of LOX/HGNPs/g-C₃N₄–hemin became more smooth (Fig. 5D) with the modification of LOX on HGNPs/g-C₃N₄–hemin due to LOX filling the interstitial places between HGNPs and HGNPs, and HGNPs and g-C₃N₄–hemin, suggesting that LOX was successfully immobilized on the surface of HGNPs/g-C₃N₄–hemin. With the gradual modification of hemin, HGNPs, and LOX, the corresponding thickness gradually increased, indicating the fabrication process of the biosensor.

Fig. 5 AFM images of (A) g-C₃N₄ nanosheets, (B) g-C₃N₄–hemin, (C) HGNPs/g-C₃N₄–hemin, and (D) LOX/HGNPs/g-C₃N₄–hemin modified films.

Electrochemical impedance spectroscopy (EIS) is also a useful tool for evaluating the interfacial changes of an electrode. The semicircle diameter of impedance spectrum equals the electron transfer resistance (R_et), which controls the electron transfer kinetics of the redox probe at the electrode interface. Fig. 6A presents the impedance spectrum of each modified stage in 5.0 mM K₃Fe(CN)₆/K₄Fe(CN)₆ (1 : 1). Compared with the bare GCE (Fig. 6A, curve a), the Nyquist semicircle of g-C₃N₄–hemin modified electrode (Fig. 6A, curve b) increased due to that the poor conductive properties of g-C₃N₄–hemin nanocomposite and chitosan would obstruct the electron transfer of the electrochemical probe. After the HGNPs were immobilized onto the electrode, the Nyquist semicircle further increased (Fig. 6A, curve c). The reason is as follows. On the one hand, HGNPs with large specific surface area and good conductivity would accelerate the electron transfer rate of the electrochemical probe. On the other hand, HGNPs with negative charge would reject [Fe(CN)₆]^3−/4− arriving at the electrode surface, which leaded to a decrease in the concentration of [Fe(CN)₆]^3−/4− on the electrode surface, thereby causing an increase in the Nyquist semicircle. The later dominated in the measurement process, therefore an increase in the Nyquist semicircle was observed. After the LOX was immobilized onto the electrode (Fig. 6A, curve d), the Nyquist semicircle further increased due to the hindrance caused by non-conductive LOX. The modification of the biosensor was also characterized by CV and the corresponding cyclic voltammograms were shown in Fig. 6B. As observed, with the gradual modification of g-C₃N₄–hemin nanocomposite, HGNPs, and LOX, the CV peak currents of the modified electrodes gradually reduced. The results of CV were consistent with those of EIS. The results of EIS and CV also proved that the enzyme LOX has been modified on the electrode.

Fig. 6 (A) EIS and (B) CV characterization of (a) bare GCE, (b) g-C₃N₄–hemin/GCE, (c) HGNPs/g-C₃N₄–hemin/GCE, and (d) LOX/HGNPs/g-C₃N₄–hemin/GCE in 5.0 mM K₃[Fe(CN)₆]/K₄[Fe(CN)₆] (1 : 1). Scan rate: 50 mV s⁻¹. The insert shows the amplified CV curves of (B).

Optimization of analytical conditions

The effect of the concentration of luminol, pH of PBS, and the incubating time of HGNPs and LOX on the intensity of ECL signal was studied to optimize the performance of the ECL biosensors for detecting lactate. As shown in Fig. S1A,† the ECL signal intensity increased with the increase in the concentrations of luminol from 0.05 to 0.30 mM, and decreased slightly with the concentration of luminol exceeding 0.30 mM (Fig. S1A†). Therefore, 0.30 mM was selected as the optimal concentration of luminol in this study.

To obtain a higher enzyme loading on the modified electrode, the incubating time of HGNPs was optimized since the quantity of LOX assembled onto the electrode depends on the quantity of HGNPs on the electrode. The influence of incubation time of HGNPs on the performance of the biosensor was investigated and the results are shown in Fig. S2B.† ECL signal intensity increased with the increase in incubated time of HGNPs from 2 h to 10 h, and decreased slightly with the incubated time of HGNPs over 10 h. Thus, the incubation time of 10 h was selected for the biosensor fabrication.

The incubation time of LOX was optimized since the concentration of the LOX modified on the electrode was influenced by the incubation time of LOX. The influence of incubation time of LOX on the performance of the biosensor was investigated and the results are shown in Fig. S1C.† As seen, the ECL signal intensity increased with the increase in incubated time of LOX from 2 h to 12 h, but increased slightly with the incubated time of LOX over 10 h. Thus, 10 h was chosen as the incubation time of LOX.

The pH as an important factor, affected the luminol's ECL signal and the enzymatic reactions for the following reasons. (1) The pH greatly influenced the ECL signal in luminol/H₂O₂ system. (2) The catalytic activity of LOX was affected by pH and the optimum pH for LOX activity was 6.0–7.0, and. The ECL signals were detected at the biosensor in PBS solutions with different pH, in the presence of 0.30 mM luminol and 0.028 mM lactate. The corresponding results are shown in Fig. S1D.† The ECL intensity increased with the pH from 6.0 to 9.0 and decreased after pH 9.0 for the reason that alkaline conditions were favorable for luminol luminescence, but too high pH also affected the enzyme activity. We mainly studied the application of the biosensor in the actual detection, so physiological pH 7.4 was selected to be used in our experiments.

Performance of the biosensor

In order to investigate the effect of hemin, g-C₃N₄ and HGNPs on the luminol/H₂O₂ ECL system, a comparative test was performed using different modified electrodes under the scanning potential in range of 0.2–0.8 V. Fig. S2A† shows the ECL behaviors of (a) bare GCE, (b) g-C₃N₄/GCE, (c) g-C₃N₄–hemin/GCE and (d) HGNPs/g-C₃N₄–hemin/GCE. As seen, compared with the control electrodes bare GCE, g-C₃N₄/GCE and g-C₃N₄–hemin/GCE, our prepared biosensor (HGNPs/g-C₃N₄–hemin/GCE) showed the maximum ECL intensity, indicating the promoting effect of g-C₃N₄, hemin and HGNPs on the ECL intensity in luminol/H₂O₂ system. The reasons may be as follows. (1) g-C₃N₄ nanosheets have outstanding electronic and catalytic properties and luminescence performance. (2) Hemin hold activity centre which could accelerate the electron transfer rate. (3) The HGNPs could amplify the ECL signal of luminol due to the good conductivity, large surface area, and excellent electroactivity for luminol. This control experiment suggested that hemin, g-C₃N₄ and HGNPs could directly enhance the ECL intensity of luminol/H₂O₂ system for the excellent catalytic performance.

In addition, it was reported that g-C₃N₄ nanosheets could generate strong ECL signal with various coreactant under the different potential. In order to investigate whether the g-C₃N₄ nanosheets have ECL response under our working potential (0.2–0.8 V), we studied the ECL response of g-C₃N₄ modified electrode in the range of −2.0–2.0 V. As seen in Fig. S2B,† g-C₃N₄ nanosheets modified electrode indeed generated strong ECL signal under the potential in the range of −1.0–2.0 V and 1.0–2.0 V, respectively. Obviously, the ECL signal resulted from g-C₃N₄ don't interfere with our determination at the working potential (0.2–0.8 V).

Fig. 7A reveals the ECL behaviors of the biosensor for the determination of lactate under optimal experimental conditions. As seen, the ECL signal increased with the addition of lactate, which was ascribed to the fact that H₂O₂ produced by the catalysis of LOX can enhance the ECL signal of luminol. The curve a of Fig. 7B shows the linear relationship of the biosensor between the ECL signal intensity and the concentration of lactate. The linear range was from 1.7 × 10⁻⁸ to 5.0 × 10⁻⁴ M and the limit of detection (S/N = 3) was 5.5 × 10⁻⁹ M. The regression equation was I (a.u.) = 874.79 + 9.28 × 10³c (mM) with a correlation coefficient of 0.993. To compare the ECL performance of HGNPs and AuNPs modified electrodes, the control electrode LOX/AuNPs/g-C₃N₄–hemin with AuNPs was constructed and its response to lactate was investigated (curve b, Fig. 7B). The linear range from 1.4 × 10⁻⁵ M to 3.8 × 10⁻⁴ M and the limit of detection of 3.7 × 10⁻⁶ M were obtained. Obviously, the biosensor with HGNPs exhibited a higher sensitivity, wider linear range and lower detection limit than AuNPs modified electrode. This was due to that the hollow structure of HGNPs allow more LOX to adsorb on the inside surface of the walls as well on the outside surface.

Fig. 7 (A) ECL responses of the biosensor to lactate with the concentrations of 0, 1.67 × 10⁻⁸, 3.33 × 10⁻⁷, 6.67 × 10⁻⁷, 1.17 × 10⁻⁶, 2.67 × 10⁻⁶, 4.33 × 10⁻⁶, 7.67 × 10⁻⁶, 1.27 × 10⁻⁵, 1.93 × 10⁻⁵, 2.77 × 10⁻⁵, 3.77 × 10⁻⁵, 5.43 × 10⁻⁵, 8.77 × 10⁻⁵, 1.38 × 10⁻⁴, 2.04 × 10⁻⁴, 2.88 × 10⁻⁴, 3.88 × 10⁻⁴, and 5.04 × 10⁻⁴ M (from bottom to top) in 0.10 M PBS (pH 7.4) containing 0.30 mM luminol. (B) Calibration curves of (a) LOX/HGNPs/g-C₃N₄–hemin/GCE and (b) LOX/AuNPs/g-C₃N₄–hemin/GCE.

In addition, the comparison between different detection methods for lactate was performed and the results are shown in ESI† Table 1. As seen, our biosensor exhibited a lower detection limit and higher sensitivity, which was due to the following reasons. (1) both g-C₃N₄–hemin nanocomposite and HGNPs could enhance the ECL intensity of luminol/H₂O₂ system. (2) More LOX was immobilized for the hollow structure of HGNPs.

To assess the applicability of the biosensor, the lactate in real serum samples was tested using the biosensor. Serum samples were diluted with pH 7.4 PBS and the results are shown in Table 1. Compared with the UV-vis spectrophotometric enzyme kinetics in hospital, the prepared biosensors showed an acceptable result and the corresponding relative deviation was −8.0%, 9.1%, 12% and −7.9%, respectively, demonstrating the practicability of the sensing protocol.

Table 1 Analytical results of lactate in human serum samples

Sample^a	Lactate concentration tested by hospital (mM)	Lactate concentration tested by tested by biosensor^b (mM)	Relative deviation
a The human serum sample from 9th People's Hospital of Chongqing, Chongqing.b Average of three determination ±SD, n = 4.
1	0.25	0.23 ± 0.02	8.0%
2	0.22	0.24 ± 0.01	9.1%
3	0.41	0.46 ± 0.04	12%
4	0.38	0.35 ± 0.01	−7.9%

---

The stability of the biosensor was also tested by monitoring its ECL response towards 0.028 mM lactate. The electrode was stored in 4 °C when not in use, the electrode was detected every two days. 20 days later, the response of the ECL biosensor decreased 8.9%, indicating an acceptable stability of the biosensor, as shown in Fig. S3.† The interfering substances, such as dopamine (0.83 mM), uric acid (0.83 mM), ascorbic acid (0.83 mM) and glucose (0.83 mM) were tested and it was found that the change of ECL signal resulted from interfering substances was negligible, which indicated that the biosensor had a good anti-interference ability and good selectivity.

Conclusion

In summary, an ECL biosensor was constructed based on g-C₃N₄–hemin nanocomposite and HGNPs for the detection of lactate. g-C₃N₄ nanosheets exhibited an excellent catalytic effect on luminol/H₂O₂ ECL system, and hemin and HGNPs could further improve its ECL properties. The combination of g-C₃N₄–hemin nanocomposite and HGNPs might be an efficient and promising nanomaterial for ECL sensing. This study provided a new, economic, effective and sensitive method to detect lactate, thus exhibiting great potential application in clinic.

Acknowledgements

This work was financially supported by the NNSF of China (21075100, 21275119, 21105081), Ministry of Education of China (Project 708073), Research Fund for the Doctoral Program of Higher Education (RFDP) (20110182120010), Natural Science Foundation of Chongqing City (CSTC-2011BA7003, CSTC-2014JCYJA20005, CSTC-2010BB4121), State Key Laboratory of Silkworm Genome Biology (sklsgb2013012) and the Fundamental Research Funds for the Central Universities (XDJK2013A008, XDJK2013A027, XDJK2014A012), China.

Notes and references

1. A. Loforte, A. Montalto, F. Ranocchi, P. L. Della Monica, G. Casali, A. Lappa, A. Menichetti, C. Contento and F. Musumeci, Artif. Organs, 2012, 36, E53–E61.
2. D. A. Sakharov, M. U. Shkurnikov, M. Y. Vagin, E. I. Yashina, A. A. Karyakin and A. G. Tonevitsky, Bull. Exp. Biol. Med., 2010, 150, 83–85.
3. G. Q. Zhou, L. L. Tang, Y. P. Mao, L. Chen, W. F. Li, Y. Sun, L. Z. Liu, L. Li, A. H. Lin and J. Ma, Int. J. Radiat. Oncol., Biol., Phys., 2012, 82, E359–E365.
4. S. V. Meethal and C. S. Atwood, Neurobiol. Aging, 2012, 33, 569–581.
5. J. Tonini, A. S. Michallet, P. Flore, H. Nespoulet, J. L. Pepin, B. Wuyam, P. Levy and R. Tamisier, Respir. Physiol. Neurobiol., 2011, 179, 287–293.
6. P. Labroo and Y. Cui, Biosens. Bioelectron., 2013, 41, 852–856.
7. N. Nesakumar, S. Sethuraman, U. M. Krishnan and J. B. B. Rayappan, J. Colloid Interface Sci., 2013, 410, 158–164.
8. N. G. Patel, A. Erlenkotter, K. Cammann and G.-C. Chemnitius, Sens. Actuators, B, 2000, 67, 134–141.
9. A. M. Herrero, T. Requena, A. J. Reviejo and J. M. Pingarron, Eur. Food Res. Technol., 2004, 219, 557–560.
10. H. Henry, N. Marmy Conus, P. Steenhout, A. Beguin and O. Boulat, Biomed. Chromatogr., 2012, 26, 425–428.
11. N. Nesakumar, S. Sethuraman, U. M. Krishnan and J. B. B. Rayappan, J. Colloid Interface Sci., 2013, 410, 158–164.
12. F. K. Sartain, X. Yang and C. R. Lowe, Anal. Chem., 2006, 78, 5664–5670.
13. A. C. B. Dias, R. A. O. Silva and M. A. Z. Arruda, Microchem. J., 2010, 96, 151–156.
14. P. Bertoncello and R. J. Forster, Biosens. Bioelectron., 2009, 24, 3191–3200.
15. Y. Y. Su and Y. Lv, RSC Adv., 2014, 4, 29324–29339.
16. V. R. Rivera, F. J. Gamez, W. K. Keener and M. A. Poli, Anal. Biochem., 2006, 353, 248–256.
17. L. Z. Hu and G. B. Xu, Chem. Soc. Rev., 2010, 39, 3275–3304.
18. W. J. Fei, F. F. Chen, L. Sun, Q. H. Li, J. P. Yang and Y. Wu, Microchim. Acta, 2014, 181, 419–425.
19. Z. Y. Lin, J. H. Chen, Y. W. Chi, B. Qui, J. M. Lin and G. N. Chen, Electrochim. Acta, 2008, 53, 6464–6468.
20. B. Haghighi and S. Bozorgzadeh, Talanta, 2011, 85, 2189–2193.
21. M. M. Cheng, G. Cuda, Y. L. Bunimovich, M. Gaspari, J. R. Heath, H. D. Hill, C. A. Mirkin, A. J. Nijdam, R. Terracciano, T. Thundat and M. Ferrari, Curr. Opin. Chem. Biol., 2006, 10, 11–19.
22. N. L. Rosi and C. A. Mirkin, Chem. Rev., 2005, 105, 1547–1562.
23. A. C. Templeton, W. P. Wuelfing and R. W. Murray, Acc. Chem. Res., 2000, 33, 27–36.
24. R. Bhattacharya and P. Mukherjee, Adv. Drug Delivery Rev., 2008, 60, 1289–1306.
25. M. De, P. S. Ghosh and V. M. Rotello, Adv. Mater., 2008, 20, 4225–4241.
26. E. E. Connor, J. Mwamuka, A. Gole, C. J. Murphy and M. D. Wyatt, Small, 2005, 1, 325–327.
27. A. M. Schwartzberg, T. Y. Oshiro, J. Z. Zhang, T. Huser and C. E. Talley, Anal. Chem., 2006, 78, 4732–4736.
28. C. M. Cheng, Y. Huang, X. Q. Tian, B. Z. Zheng, Y. Li, H. Y. Yuan, D. Xiao, S. P. Xie and M. M. F. Choi, Anal. Chem., 2012, 84, 4754–4759.
29. Y. P. Li, J. Zhan, L. Y. Huang, H. Xu, H. M. Li, R. X. Zhang and S. L. Wu, RSC Adv., 2014, 4, 11831–11839.
30. F. Goettmann, A. Fischer, M. Antonietti and A. Thomas, Angew. Chem., Int. Ed., 2006, 45, 4467–4471.
31. L. C. Chen, D. J. Huang, S. Y. Ren, T. Q. Dong, Y. W. Chi and G. N. Chen, Nanoscale, 2013, 5, 225–230.
32. J. Q. Tian, Q. Liu, A. Asiri, A. O. Al-Youbi and X. P. Sun, Anal. Chem., 2013, 85, 5595–5599.
33. J. Q. Tian, Q. Liu, C. J. Ge, Z. C. Xing, A. M. Asiri, A. O. Al-Youbi and X. P. Sun, Nanoscale, 2013, 5, 8921–8924.
34. J. Q. Tian, Q. Liu, A. M. Asiri, A. H. Qusti, A. O. Al-Youbi and X. P. Sun, Nanoscale, 2013, 5, 11604–11609.
35. S. F. Li, X. M. Zhang, W. X. Du, Y. H. Ni and X. W. Wei, J. Phys. Chem. C, 2009, 113, 1046–1051.
36. J. Wang, R. R. Zhao, M. Z. Xu and G. N. Chen, Electrochim. Acta, 2010, 56, 74–79.
37. H. Dai, Y. Y. Lin, G. F. Xu, L. S. Gong, C. P. Yang, X. L. Ma and G. N. Chen, Electrochim. Acta, 2012, 78, 508–514.
38. L. Ge, C. C. Han and J. Liu, J. Mater. Chem., 2012, 22, 11843–11850.
39. X. C. Wang, K. Maeda, X. F. Chen, K. Takanabe, K. Domen, Y. D. Hou, X. Z. Fu and M. Antonietti, J. Am. Chem. Soc., 2009, 131, 1680–1681.
40. Q. J. Xiang, J. G. Yu and M. Jaroniec, Nanoscale, 2011, 3, 3670–3678.
41. A. Paneque, J. F. Bertran, E. Reguera and H. Y. Madeira, Transition Met. Chem., 2001, 26, 76–80.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c4ra09616a

---