Introduction

Electrochemiluminescence (ECL) represents a marriage between electrochemistry and spectroscopy. Since it does not require the use of an external light source, problems derived by light scattering in spectroscopic-based techniques are avoided, improving the signal-to-noise ratio dramatically. In addition, the time and position of light-emitting reactions can be accurately controlled by the potential applied at an electrode surface. Due to its simplicity, rapidity, excellent detection sensitivity, high selectivity, easy controllability and flexibility, ECL detection has emerged as a powerful tool in analytical chemistry related areas such as medical diagnostics, environmental and food monitoring, and biowarfare agent detection [1–4].

One of the key tasks of ECL research is to identify new luminophore species. Quantum dots (QDs) have been extensively used in ECL-based detection [3, 5, 6]. Compared with conventional molecular emitters, QDs own several distinctive merits like size/surface-trap controlled luminescence and good stability against photobleaching. However, up to now, most QDs used as luminophore species are toxic heavy metals-based semiconductors, such as CdSe, CdS, and PbS. Obviously, the intrinsic toxicity of these QDs would limit their applications in ECL sensors, especially in bioanalysis. Therefore, it is of considerable interest in developing environment-friendly ECL luminophores for biosensing. To explore novel “green” luminophores with high ECL efficiency, the family of nanoemitters has been enlarged from exclusively QDs to other miscellaneous nanomaterials with various compositions, sizes and shapes, such as Ag2Se QDs [7], carbon nanodots [8], Ag-carbon nanodots [9], and graphene QDs [10]. Emerging as a novel type of nontoxic and harmless luminescent nanomaterial, gold nanoclusters (AuNCs) have drawn wide attention [11–16]. Their size falls within the range of the Fermi wavelength of conduction electrons and results in size-dependent HOMO–LUMO transition and luminescence. Furthermore, AuNCs show fantastic characteristics including remarkable water-solubility, high colloidal stability, outstanding catalytic performance, and unusual photophysical properties, which make them promising in biosensing and labeling applications [17–23]. Although the fluorescent properties of AuNCs have been widely studied, there are only a few researches about the AuNC-based ECL sensing, even less in aqueous system and by using immobilization mode [24, 25].

Herein, we reported of the ECL property of methionine stabilized AuNCs (Met-AuNCs). The Met-AuNCs were synthesized by using methionine both as a reductant and a stabilizer, and the ECL properties of the Met-AuNC-modified glassy carbon electrode (Met-AuNC/GCE) were studied. The excellent cathodic ECL behavior has been observed with K2S2O8 as coreactant and the ECL mechanism has been discussed. Moreover, dopamine (DA) was taken as a model analyte to evaluate the potential application of the Met-AuNC-based ECL sensor. The practicality of the Met-AuNC/GCE was validated by determining the DA released from PC12 cells via exocytosis through stimulation of potassium ions. This ECL sensor is environmental friendly, highly selective and sensitive, which can be widely applied in biological analysis. In particular, this approach would open new routes to apply other NCs ECL in different aspects of biochemistry.

Materials and methods

Materials

HAuCl4·4H2O, methionine, and NH3·H2O were purchased from Aladdin Reagent Company (https://www.sigmaaldrich.com). Dopamine was obtained from Alfa Aesar (http://alfaaesar.lookchem.com/). PC12 cells were purchased from Beijing DingGuo ChangSheng Biotechnology Co. Ltd. (http://www.dlcs100.com/). Phosphate buffer (PB, 0.2 M) with different pH were prepared by changing the ratio of Na2HPO4 and NaH2PO4. All other regents were of analytical grade and used as received without further purification. All solutions were prepared with ultrapure water. PB with elevated K+ included 104.2 mM KCl, 50 mM NaCl, 2 mM CaCl2, 0.7 mM MgCl2, 1 mM NaH2PO4, and 10 mM HEPES. Culture media and all the solutions were adjusted to pH 7.4 by adding the appropriate volumes of 1 M NaOH.

Apparatus

Cyclic voltammetric (CV) and ECL measurements were obtained using a model MPI-A electro-chemiluminescence analyzer (Xi’an Remex Analysis Instrument Co. Ltd. Xi’an, China). The ECL spectra were obtained by collecting the ECL data during the cyclic potential sweep with a series of optical filters at wavelengths of 425, 440, 460, 490, 520, 535, 555, 575, 590, and 620 nm. The photoluminescence spectrum was measured by Cary Eclipse fluorescence spectrophotometer (Agilent, USA). The transmission electron microscope (TEM) was performed on JEOL 2010 transmission electron microscopy operated at an accelerating voltage of 200 kV.

Synthesis of Met-AuNCs

The Met-AuNCs were prepared according to our previous method [26]. Briefly, 4 mL of 0.1 M methionine and 0.6 mL of 0.5 M NaOH were added into 0.4 mL of 20 mg mL−1 HAuCl4. The mixture solution was incubated at 37 °C for 6 h. Then, 0.5 mL of 1 M H2SO4 was added into the mixture to precipitate AuNCs. The product was collected with centrifugation at 6000 rpm (or with relative centrifugal force of 1815×g) for 2 min. Then the AuNCs were dissolved in a solution of NH3·H2O (1.4%) and incubated at 70 °C for 30 min. The resulting solution of Met-AuNCs was stored in the dark at 4 °C.

Fabrication of the AuNC-modified GCE

Prior to use, the GCE was polished sequentially with 1.0, 0.3, and 0.05 μm alumina slurry, followed by washing thoroughly with water. Then, the electrode was washed ultrasonically with HNO3, ethanol, and water, respectively. 100 μL of Met-AuNCs was mixed with 20 μL nafion (1.5%). 8 μL of the mixture was dropped on the GCE surface (3 mm in diameter, the surface area of the electrode is 0.0707 cm2). After being dried in air at room temperature, the Met-AuNC-modified GCEs were obtained. For comparison experiments, the BSA-AuNCs were prepared (see the Supplemental data), and the final concentration of the BSA-AuNCs was the same as the Met-AuNCs. The BSA-AuNC-modified GCEs were prepared by the same procedure. To prepare the polydopamine (PDA)/Met-AuNC/GCE, 5 mg DA was dissolved in 1 mL of 5 mM Tris-HCl solution (pH 8.5), then Met-AuNC/GCE was immersed in it and stored in the 4 °C refrigerator for 2 h. Finally, the obtained PDA/Met-AuNC/GCE was washed by ultrapure water and stored in the 4 °C when not used.

Procedures

The ECL detection was performed at room temperature in the electrolyte containing K2S2O8. The conventional three-electrode system was used through detection. A modified GCE was used as the working electrode, while a Pt wire as the auxiliary electrode, and an Ag/AgCl (saturated with KCl) electrode as the reference electrode. The potential was cycled between 0 and −2.0 V with a scan rate of 100 mV⋅s−1, and the photomultiplier tube (PMT) was biased at 800 V in the experiments.

K+-stimulated DA release from PC12 cells

The details of PC12 cell culture method were described as follows. PC12 cells were maintained in RPMI-1640 media supplemented with 10% fetal bovine serum (FBS), and 0.5% penicillin-streptomycin solution in a humidified atmosphere with 5% CO2 at 37 °C. The medium was replaced every day throughout the lifetime of all cultures. ECL was applied to detect the K+-stimulated release of DA from PC 12 cells [27]. Briefly, 0.5 mL of trypsin was added into the culture dishes for 1 min, and the PC12 cells in the culture dishes were suspended by culture medium. After that, the PC12 cells suspension on the culture dishes was centrifuged for 5 min at 1000 rpm, and washed 3 times with PB. Then, 5 mL of the elevated-K+ PB were placed in the culture dishes (5 × 105 cells mL−1), and incubated in the incubator for 10 min. After centrifuged, the supernatant was collected, and the extracellular fluids were transferred to sterilized beakers. Finally, the concentrations of DA in the mixtures with 0.1 M K2S2O8 were determined by ECL.

Results and discussion

Choice of materials

Herein, to explore the ECL properties of the Met-AuNCs, different electrodes, such as Au electrode, Pt electrode, GCE and ITO electrode and different co-reactants such as H2O2, benzoyl peroxide (BPO), tri-npropylamine (TPrA) and K2S2O8 were investigated. The results indicated that the Met-AuNC-modified GCE with K2S2O8 as co-reactant in aqueous solution exhibited the most stable and efficient cathodic ECL signal. Therefore, GCE was used as the working electrode, and S2O8 2− was chosen as the co-reactant to study the ECL properties of the Met-AuNCs.

Characterization of the Met-AuNCs

Transmission electron microscopy (TEM) revealed an average diameter of 2.5 ± ss0.3 nm for the Met-AuNCs (Fig. 1a). The inset high-resolution TEM image shows lattice planes separated by 0.236 nm, corresponding to the (111) lattice spacing of face-centered cubic Au [28]. The aqueous AuNCs solution emitted strong yellow fluorescence under UV light irradiation. By fluorescence spectrophotometry (excited at 420 nm), it was observed that the Met-AuNCs solution exhibited strong photoluminescence (PL) at around 535 nm (Fig. 1b). These phenomena corresponded to our previous results, indicating the successfully synthesized the Met-AuNCs [26].

Fig. 1
figure 1

a TEM image of Met-AuNCs. b Photoluminescence excitation (emission at 535 nm) and emission (excitation at 420 nm) spectra of Met-AuNCs

Full size image

ECL behaviors of Met-AuNC-modified GCE

ECL is a useful technique for both fundamental study and analytical applications. It is of considerable interest in exploring the ECL properties of the Met-AuNCs with excellent water-solubility. ECL behaviors of Met-AuNCs at GCE were investigated with cathodic coreactant K2S2O8. Figure 2a shows the ECL curves on bare GCE and Met-AuNC/GCE at a scan rate of 100 mV⋅s−1 within the potential range of 0 to −2.0 V. Weak ECL emission can be observed on Met-AuNC/GCE in the absence of K2S2O8 (Fig. 2a, curve a). After the addition of 0.1 M K2S2O8, a strong ECL signal was obtained with the peak potential of −1.86 V (Fig. 2a, curve b). The CVs were also recorded simultaneously (Fig. S2). The cathodic peak at −1.0 V on Met-AuNC/GCE was corresponding to the reduction of S2O8 2− [24, 29]. To gain a better understanding of the ECL generation, we also performed the experiment in the absence of Met-AuNCs. The background ECL signal from the GCE without Met-AuNCs was very low (Fig. 2a, curve c), indicating that the ECL emission was originated from the Met-AuNCs. On the other hand, positive shift of the ECL onset potential to −0.7 V was observed on the Met-AuNC/GCE. This is further confirmed by potential step experiment. As shown in Fig. 2b, light emission was generated on Met-AuNC/GCE when the potential step from 0 V to −0.7 V, while no light emission was observed on GCE. These results suggested that Met-AuNCs accelerated electron transfer between the electrode and S2O8 2− [24]. Figure 2c shows the ECL emission of Met-AuNC/GCE with 0.1 M S2O8 2− under continuous cyclic voltammetry between 0 and −2.0 V of 20 cycles. Strong and stable ECL signals were observed with the relative standard deviation (RSD) of 1.9%, demonstrating that the ECL emission of Met-AuNC/GCE was high repeatable for further analytical application.

Fig. 2
figure 2

a ECL intensity versus potential. (a) On the Met-AuNC/GCE in 0.2 M PB (pH 7.4) containing 0.1 M KCl; (b) on the Met-AuNC/GCE in 0.2 M PB (pH 7.4) containing 0.1 M K2S2O8 and 0.1 M KCl; (c) on the bare GCE in 0.2 M PB (pH 7.4) containing 0.1 M K2S2O8 and 0.1 M KCl. Scan rate, 100 mV⋅s−1. b ECL signal from (a) bare GCE and (b) Met-AuNC/GCE in 0.2 M PB (pH 7.4) containing 0.1 M K2S2O8 and 0.1 M KCl generated by potential steps between 0 V and −0.7 V. c Time-dependent ECL signals of the Met-AuNC/GCE in 0.2 M PB (pH 7.4) containing 0.1 M K2S2O8 and 0.1 M KCl. The potential was cycled between 0 and −2.0 V with a scan rate of 100 mV⋅s−1. d ECL signals of (a) BSA-AuNC/GCE and (b) Met-AuNC/GCE in 0.2 M PB (pH 7.4) containing 0.1 M K2S2O8 and 0.1 M KCl. The potential was cycled between 0 and −2.0 V with a scan rate of 100 mV⋅s−1. e ECL spectrum of Met-AuNC/GCE in 0.2 M PB (pH 7.4) containing 0.1 M K2S2O8 and 0.1 M KCl

Full size image

To our best knowledge, the AuNCs protected by bovine serum albumin (BSA-AuNCs) are the most frequently used AuNC-based ECL luminophore [24, 25]. Herein, we found that the ECL intensity of Met-AuNC/GCE was about five times higher than that of BSA-AuNC/GCE (Fig. 2d, Fig. S1), suggesting that the surface ligand of AuNCs played an important role in ECL reaction. On the other hand, due to the nonconductive protein shell, the BSA-AuNCs showed much higher electron-transfer resistance (R et) than that of Met-AuNCs, which resulted in the decrease of the electron transfer rate between the modified electrode and S2O8 2− (Fig. S3). Thus, methionine not only acts as reducing and stabilizing reagent for the synthesis of AuNCs, but also enhances the sensitivity of ECL detection.

The ECL spectrum of Met-AuNCs was measured by employ in a series of optical filters. Figure 2e showed a maximum ECL emission at 550 nm with a red shift of 20 nm compared to that in the PL. This red shift between PL and ECL emission was also observed in the previously reported ECL of CdSe nanocrystals [30], Si nanocrystals [31], and peptide nanovesicles [32], which was explained to the strong role of surface states on the ECL process of the luminophore species.

All the above results implied that Met-AuNCs and S2O8 2− can form an excellent ECL system (Met-AuNCs/S2O8 2−), in which Met-AuNCs are the luminophores while S2O8 2− acts as the coreactant. The ECL emission mechanism was caused by electron transfer annihilation between an anionic quantum dot radical (R•−) and the electrogenerated SO4 •−. The possible ECL mechanism is described in Scheme 1 with the following Eqs. [10]:

$$ \mathrm{Met}-AuNC+{\mathrm{e}}^{-}\to \kern0.2em \mathrm{Met}-{AuNC}^{\bullet -} $$
(1)
$$ {\mathrm{S}}_2{\mathrm{O}}_8{{}^2}^{-}+{\mathrm{e}}^{-}\to {\mathrm{S}\mathrm{O}}_4{{}^2}^{-}+{{\mathrm{S}\mathrm{O}}_4}^{\bullet -} $$
(2)
$$ \mathrm{Met}-{AuNC}^{\bullet -}+{{\mathrm{SO}}_4}^{\bullet -}\to \mathrm{Met}-AuN{C}^{\ast }+{\mathrm{SO}}_4{{}^2}^{-} $$
(3)
$$ \mathrm{Met}-AuN{C}^{\ast}\to \mathrm{Met}-AuNC+h\nu $$
(4)
Scheme 1
scheme 1

Schematic diagram for the ECL sensor based on Met-AuNCs

Full size image

Electrochemiluminescence efficiency

The ECL efficiency (Φ ECL) for an annihilation system is defined as the number of photons emitted per redox event. It is the product of the efficiency of populating the excited state and the quantum yield of emission from that excited state. Eq. (5) was used to determine Φ ECL of the Met-AuNCs using [Ru(bpy)3]2+ in the solution as a relative standard [33, 34]:

$$ {\varPhi}_{ECL}={\varPhi^{{}^{\circ}}}_{ECL}\left({IQ^{{}^{\circ}}}_f/{Q}_f{I}^{{}^{\circ}}\right) $$
(5)

where Φ° ECL is the ECL efficiency of [Ru(bpy)3]2+ (1 mM and 0.1 M TBABF/ACN), taken as 5.0% [35], Q f and Q° f are the corresponding faradaic charge passed for the Met-AuNC film and [Ru(bpy)3]2+ respectively, and I and I o are the integrated photomultiplier tube responses for them respectively. The Φ ECL of Met-AuNCs was estimated to be 2.23%, which was much higher than that of BSA-AuNCs (0.325%), Polyaniline-[Ru(bpy)2PIC]2+ (1.00%) [36], nafion-[Ru(bpy)3]2+ (0.089%) [34] and [Ru(bpy)2(poly-4-vinyl pyridine)10]2+ (0.152%) [33].

Optimization of method

In order to improve the performance of the ECL sensor, the following parameters were optimized: (a) amount of Met-AuNCs, (b) K2S2O8 concentration and (c) pH of the solution. Respective data and Figures are given in the Electronic Supporting Material. We found the following experimental conditions to give best results: (a) An amount of Met-AuNCs of 6.12 μg; (b) a pH of 7.4; (c) a K2S2O8 concentration of 0.1 M (Fig. S4).

Analytical performance of the ECL sensor

DA is an important neurotransmitter of catecholamine in the human brain. The temporal fluctuation of the DA concentrations in the human brain has a critical effect on several neurological disorders such as Harrington’s disease and Parkinson’s disease [37–40]. The efficient determination of DA becomes increasingly significant in the field of clinical disease diagnosis and the research of physiological functions [41–53]. Thus, DA was taken as a model analyte to evaluate the potential application of the Met-AuNC-based ECL sensor. When a trace of DA was added into the detection solution, the cathodic ECL emission of Met-AuNC/GCE was quenched. The ECL intensity decreased with the increasing concentration of DA (Fig. 3a), leading to an analytical method for the detection of quenchers and quencher relatives based on the ECL of Met-AuNCs. The quenching mechanism of DA to ECL might be as follows (Scheme 1): SO4 •−, a strong radical oxidant produced by electro-reduction of S2O8 2−, oxidized DA to produce o-benzoquinone species. The energy transfer between Met-AuNCs excited state and o-benzoquinone species might be responsible for the quenching effect [41]. To further demonstrate the speculation, the ECL intensity of the PDA (the oxidized dopamine)/Met-AuNC-modified GCE was studied. As expected, compared with the Met-AuNC/GCE, the ECL intensity of the PDA/Met-AuNC/GCE decreased obviously (Fig. S5), which indicated that the oxidized DA can mainly lead to the ECL quenching of the Met-AuNC/GCE. With the increasing amount of DA ranging from 0.1 to 4 μM, the plot of I 0 /I vs. the concentration of DA showed a good correlation (Fig. 3b), which hinted that the quenching followed the dynamic principle described by the Stern-Volmer equation,

$$ {I}_0/I=1+{K}_{SV}\left[Q\right] $$
(6)
Fig. 3
figure 3

a ECL profiles of the Met-AuNC/GCE in the presence of DA in 0.2 M PB (pH 7.4) containing 0.1 M K2S2O8 and 0.1 M KCl. The potential was cycled between 0 and −2.0 V with a scan rate of 100 mV⋅s−1. b Linear calibration plot for DA detection

Full size image

where I 0 is the initial ECL intensity, I is the ECL intensity at a given concentration of quencher [Q], and K SV is the quenching constant.

Under the optimal conditions, the limit of detection was 32 nM (S/N = 3). The linear range and detection limit of this sensor was compared with other methods reported previously (Table 1), showing that this ECL sensor has much lower detection limit than those of some other ECL sensors [7, 24, 42, 43], electrochemistry sensors [44–49] and fluorescence sensors [50, 51]. The results presented in Table 1 indicated that the analytical performance of this sensor was better than that of the other existing sensors for detection of DA. Thus, this method not only expanded the application of Met-AuNCs, but also opened new doors toward the rapid detection of DA.

Table 1 Comparison of different methods for the detection of DA
Full size table

The selectivity is one of the major concerns for ECL sensors. An interference investigation was performed by using the solution containing 0.5 μM DA and 5.0 μM common co-existing substances such as uric acid (UA), ascorbic acid (AA), cysteine, glucose (Glu), glycine, lysine, citric acid, fructose, maltose, H2O2, CO3 2−, K+, Na+, Mg2+, Cl−, and NO3 −, respectively [32, 49, 54–57]. It was found that the above interfering substances do not cause a noticeable interference to the detection of DA (Fig. 4). Thus, the Met-AuNC/GCE has high selectivity for the determination of DA. The good selectivity for DA might be attributed to the efficient energy transfer between Met-AuNCs excited state and o-benzoquinone species produced by the oxidation of DA [37].

Fig. 4
figure 4

Selectivity of the Met-AuNC-modified GCE/K2S2O8 system based sensor for DA in the solution containing analytes (left) and 0.5 μM DA + 5 μM interferences (right): (a) blank, (b) UA, (c) AA, (d) cysteine, (e) Glu, (f) glycine, (g) lysine, (h) citric acid, (i) fructose, (j) maltose, (k) H2O2, (l) CO3 2−, (m) K+, (n) Na+, (o) Mg2+, (p) Cl−, (q) NO3 −

Full size image

Detection of dopamine released by cells

Another attractive feature of this method is that it can be used for assaying the K+-stimulated DA release from PC12 cells. To validate the practicality of the present method, three samples containing certain amounts of DA in the elevated-K+ PB solution containing 0.1 M K2S2O8 were prepared, and the amounts of DA in the mixtures were measured using this method. The results presented in Table S1 indicated that this approach had good accuracy in measuring K+-stimulated DA samples that release from PC12 cells by the present method. Furthermore, Fig. 5a shows representative ECL profiles of PC12 cells that had been stimulated with K+ ions and spiked with standard DA solutions at various concentrations. From the linear plot, we estimated the concentration of DA released from the PC12 cells to be 0.75 μM (Fig. 5b), which is in agreement with the literature [27]. These observations substantially demonstrated that the method can be used for detection of DA in real biological samples.

Fig. 5
figure 5

a ECL profiles of (a) elevated-K+ PB solution containing 0.1 M K2S2O8 and (b)–(g) extracellular fluid of PC12 cells after incubation in elevated-K+ PB. The fluids were spiked with DA at the concentrations of (b) 0, (c) 0.2, (d) 0.4, (e) 0.6, (f) 0.8, and (g) 1.0 μM. b Calibration curve for detection of DA released from PC12 cells

Full size image

Conclusions

In summary, an ECL sensor based on Met-AuNC-modified GCE has been developed. The Met-AuNCs with good water dispersibility and excellent conductivity were synthesized by using methionine both as a reductant and a stabilizer. The excellent cathodic ECL behavior of Met-AuNC/GCE has been observed with K2S2O8 as coreactant. Fabricated with different shell ligand, the Met-AuNCs showed not only different spectroscopic characterization of the ECL (yellow color luminescence) from that of BSA-AuNCs (red color luminescence) but also significantly enhanced signal as compared to BSA-AuNCs. These phenomena prompt that AuNC-based ECL probes with tunable luminescent properties may be obtained by varying the shell ligand and have potential applications in bioanalysis such as multi-labeling techniques. In addition, the ECL system showed acceptable accuracy and precision for the detection for the model analyte DA released from PC12 cells. Therefore, this study not only enriches the foundation study about the ECL properties of AuNCs but also opens up a new avenue to design and development of ECL devices from other functional-metal based NCs. We believe that the use of the bioinspired AuNCs with excellent properties would bring huge potential in the ECL bioanalytical applications.