A dual-potential electrochemiluminescence ratiometric sensor for sensitive detection of dopamine based on graphene-CdTe quantum dots and self-enhanced Ru(II) complex

Biosensors and Bioelectronics 90 (2017) 61–68

Contents lists available at ScienceDirect

Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios

crossmark

A dual-potential electrochemiluminescence ratiometric sensor for sensitive detection of dopamine based on graphene-CdTe quantum dots and selfenhanced Ru(II) complex Xiaomin Fua, Xingrong Tanb, Ruo Yuana, Shihong Chena,

⁎

a Key Laboratory of Luminescent and Real-Time Analytical Chemistry (Southwest University), Ministry of Education, College of Chemistry and Chemical Engineering, Southwest University, Chongqing 400715, PR China b Department of Endocrinology, 9 th People's Hospital of Chongqing, Chongqing 400700, PR China

A R T I C L E I N F O

A BS T RAC T

Keywords: Electrochemiluminescence Ratiometric sensor Graphene-CdTe quantum dots Self-enhanced Ru(II) complex

A novel dual-potential ratiometric electrochemiluminescence (ECL) sensor was designed for detecting dopamine (DA) based on graphene-CdTe quantum dots (G-CdTe QDs) as the cathodic emitter and selfenhanced Ru(II) composite (TAEA-Ru) as the anodic emitter. TAEA-Ru was prepared by linking ruthenium(II) tris(2,2′-bipyridyl-4,4′-dicarboxylato) with tris(2-aminoethyl)amine. Firstly, 3-aminopropyltriethoxysilane founctionalized G-CdTe QDs was used as the substrate for capturing target DA via the specific recognition of the diol of DA to the oxyethyl group of APTES. Then, Cu2O nanocrystals supported TAEA-Ru was further bound by the strong interaction between amino groups of DA and carboxyl groups of the Cu2O-TAEA-Ru. With the increase in DA concentration, the loading of Cu2O-TAEA-Ru at the electrode increased. As a result, the anodic ECL signal from TAEA-Ru increased, and the cathodic ECL signal from G-CdTe QDs/O2 system decreased correspondingly. Such a decrease was resulted from the ECL resonance energy transfer (RET) from G-CdTe QDs to TAEA-Ru as well as the dual quenching effects of Cu2O to G-CdTe QDs, namely the ECL-RET from G-CdTe QDs to Cu2O and the consumption of coreactant O2 by Cu2O. Based on the ratio of two ECL signals, the determination of DA was achieved with a linear range from 10.0 fM to 1.0 nM and a detection limit low to 2.9 fM (S/N=3). The combination of G-CdTe QDs/O2 and TAEA-Ru would break the limitation of the same coreatant shared in previous ECL ratiometric systems and provide a potential application of ECL ratiometric sensor in the detection of biological small molecules with the assistance of the dual molecular recognition strategy.

1. Intoduction Electrochemiluminescence (ECL) is a new technique which has been of great interest in pharmaceutical analysis, clinical diagnosis, environmental and food analysis as well as immunoassay (Zhang et al., 2014). The detective results are mainly based on the change of single signal intensity. These single signal based ECL systems may introduce false positive or negative errors due to instrumental efficiency or some environmental changes, such as the concentration of coreactant, pH, etc. (Zhang et al., 2013a). The ratiometric ECL assay relies on the ratio of ECL intensities at two excitation potentials instead of the single signal, which is an ideal choice to reduce the false-positive or negative signals and make the detective results more credibility (Hao et al., 2014). Currently, the ECL ratiometric systems have been constructed for detecting microRNA, aptamer-related metal ion, DNA and cancer

⁎

cells (Zhang et al., 2013a; Hao et al., 2014; Cheng et al., 2014; Wang et al., 2016). However, the ECL ratiometric assays are faced with the following challenges. (1) It is difficult to find two suitable potentialresolved ECL emitters in the case of the share of the same coreactant such as H2O2 or S2O82- (Wu et al., 2016; Lei et al., 2015). Usually, two emitting states in reported ECL ratiometry were confined to CdS nanocrystals (Hao et al., 2014), graphene quantum dots (Zhao et al., 2015) or G-C3N4 nanosheets (Wang et al., 2016) as cathodic and luminol as anodic ECL emitters, and the shared coreactant is H2O2. (2) In previous ECL ratiometric systems, the second signal probe was usually introduced or released from the electrode through the formation of sandwich immune model, DNA hybridization or replacement model (Wang et al., 2016; Lei et al., 2015; Huang et al., 2016). Such models are difficultly achieved in the case of biological small molecules, resulting in a fact that almost no study concerning ECL ratiometric

Corresponding author. E-mail address: [email protected] (S. Chen).

http://dx.doi.org/10.1016/j.bios.2016.11.025 Received 17 August 2016; Received in revised form 6 November 2016; Accepted 7 November 2016 Available online 11 November 2016 0956-5663/ © 2016 Elsevier B.V. All rights reserved.

Biosensors and Bioelectronics 90 (2017) 61–68

X. Fu et al.

2. Experimental

sensor for the determination of biological small molecule has been reported until now. In our previous work, various self-enhanced complexes have been synthetized and applied in the single signal based bio-analysis (Wang et al., 2015a; Chen et al., 2015). Self-enhanced ECL complexes, which integrated the luminophores and their coreactants in the same molecular structure or composite, not only showed an improved luminous stability and luminous efficiency but also avoided the addition of exogenous coreactant into testing solution (Wang et al., 2014, 2015a). And, dissolved O2 as the coreactant also can avoid the addition of exogenous reagent. Thus, by coupling with self-enhanced Ru complexes and the ECL emitters with dissolved O2 as the coreactant, the resultant ECL ratiometric system may well avoid the troubles such as the interaction of two coreactants or the cross impact of two coreactants to ECL emitters in the case of the coexistence of two coreactants in testing solution. This strategy could overcome the limitation of the same coreatant shared in previous system, thus expanding the application of ECL ratiometric systems. In addition to the luminous material and coreatant, the suitable quencher or enhancer is required for constructing a dual-potential ECL sensor. Metal nanoparticles (NPs) such as Pt NPs (Zhang et al., 2013a), Au NPs (Hao et al., 2014), and Ag NPs (Wang et al., 2016), and cyanine dye (Cheng et al., 2014) have been employed to enhance or quench the ECL intensity based on the ECL energy transform (ECL-RET) or surface plasmon resonance in dual-potential ECL sensors. Besides, luminophor luminol also served as a quencher through competitively consuming hydrogen peroxide as the coreactant of both luminol and CdS (Huang et al., 2016). Nevertheless, above strategies suffered from following negative factors. (1) ECL-RET based enhancers or quenchers, which need a sufficient spectra overlap between ECL energy donor and acceptor, didn’t display universal applicability. (2) In the case of luminol as quencher, it is necessary for luminol to exhibit a stronger competition ability to the coreactant than the other luminophor. This situation is also not widely applicable. Therefore, it is desirable to seek a universal enhancer or quencher to construct the dual-potential ECL sensors. Since cuprous oxide (Cu2O), a low cost and eco-friendly p-type semiconductor, has been proved to could efficiently catalyze the reduction reaction of oxygen, it would be a superior and universal quencher of the ECL system with dissolved O2 as the coreactant (Wang et al., 2015b). In this work, using dopamine (DA) as the model target, a novel ECL ratiometric sensor was designed for detecting biological small molecule DA. Here, graphene-CdTe quantum dots (G-CdTe QDs) and selfenhanced Ru complexes (TAEA-Ru) were chosen as the cathodic and anodic ECL emitters, respectively. 3-aminopropyltriethoxysilane (APTES) founctionalized G-CdTe QDs was used as the substrate for capturing target DA, which could further bind Cu2O nanocrystals supported TAEA-Ru. With increasing the DA concentration, the loading of Cu2O-TAEA-Ru at the electrode increased, resulting in an increase in anodic ECL signal from TAEA-Ru and a decrease in cathodic ECL signal from G-CdTe QDs/O2 system. Such a decrease in cathodic ECL signal was ascribed to the ECL-TER from G-CdTe QDs to TAEA-Ru as well as the dual quenching effects of Cu2O to G-CdTe QDs, namely the ECL-TER from G-CdTe QDs to Cu2O and the consumption of coreactant O2 by Cu2O. By detecting the ratio of two ECL intensities, ratiometric detection of DA was achieved with the detection limit down to fM level. This work exhibits following advantages. (1) A new group of luminescent reagents (Ru luminophore and CdTe QDs) was introduced into the ECL ratiometric assay. (2) This work would provide a novel strategy for constructing two coreatants based dual-potential ECL system. (3) The dual molecular recognition strategy not only easily achieved the immobilization of small biomolecule DA onto the electrode, but also conveniently introduced the second ECL signal probes via a specific recognition between the small biomolecule and the functionalized second signal probes, thus achieving the ECL ratiometric assays of biological small molecules.

2.1. Reagents and chemicals Dopamine (DA), copper chloride (CuCl2), sodium hydroxide (NaOH), poly(ethylene glycol) (PEG) and ascorbic acid (AA) were purchased from Sigma Chemical Co. (St. Louis, MO, USA). Sodium tellurite(IV) (Na2TeO3) and cadmium chloride hemipentahydrate (CdCl2·2.5H2O) were supplied from Alfa Aesar Chemical Co., Ltd. (Tianjin, China). Tris(2,2′-bipyridyl-4,4′-dicarboxylato)ruthenium(II) was received from Suna Technology Inc. (Suzhou, China). NaBH4 and trisodium citrate dihydrate were obtained from Kelong Chemical Co. (Chengdu, China). 3-Aminopropyltriethoxysilane (APTES) was received from Lian Gang Dyestu Chemical Industry Co. Ltd. (Liaoning, China). Tris(2-aminoethyl)amine (TAEA) was obtained from Tokyo Chemicals Industry Development Co. Ltd. (Tokyo, Japan). Graphene oxide (GO) was gotten from Pioneer Nanotechnology Co. (Nanjing, China). N-hydroxysulfosuccinimide (NHS) and 1-ethyl-3-[3-(dimethylamino) propyl] carbodiimide hydrochloride (EDC) were purchased from Shanghai Medpep Co. Ltd. (Shanghai, China). 0.10 M phosphate buffer solutions (PBS) with various pH were obtained by mixing the stock solution of 0.10 M KH2PO4 and Na2HPO4 containing 0.10 M KCl as a supporting electrolyte. Doubly distilled water was used throughout the experiments. 2.2. Apparatus ECL measurements were carried out on a MPI-A electrocheminescence analyzer (Xi’an Remax Electronic Science and Technology Co., Xi’an, China) with the voltage of the photomultiplier tube (PTM) setting at 800 V in the process of detection. A CHI 600B electrochemical work-station (Shanghai CH Instruments, China) was used for cyclic voltammetry (CV) measurements. The morphologies of various nanomaterials were obtained by transmission electron microscopy (TEM) on JEM 1200EX Instrument (JEOL, Tokyo, Japan). UV–vis spectra were measured with a UV-2450 UV–vis spectrophotometer (Shimadzu, Tokyo, Japan). The X-ray photoelectron spectroscopy (XPS) was performed on Thermo Scientific Escalab 250Xi Instrument (Thermoelectricity Instruments, USA). X-ray diffraction (XRD) was measured with Purkinje General Instrument XD-3 with Cu-Kα radiation (λ=0.15406 nm). A conventional three-electrode system was used with a modified glassy carbon electrode (GCE, Φ=4.0 mm) as the working electrode, a platinum wire as the auxiliary, and an Ag/AgCl (saturated KCl solution) reference electrode. 2.3. Preparation of G-CdTe QDs Graphene-CdTe QDs (G-CdTe QDs) were synthesized according to reported reflux routes method (Yu et al., 2016). Typically, 220.0 μL graphene oxide (1.0 mg/mL) was injected into 50.0 mL of deionized water containing 36.89 mg of CdCl2 under stirring for 1 h. Next, 1.0 mL of Na2TeO3 (0.010 M), 50.0 mg of trisodium citrate dihydrate, 33.0 μL of MPA, and 100.0 mg of NaBH4 were added into above mixture solution under stirring at the room temperature. At last, the resulting solution was subjected to reflux at 130 °C for 10 h. The final solution was purified, sedimented in 1:1 (v/v) ethanol/water, and centrifuged at 8000 rpm for 5 min. The final product was dissolved in 4.0 mL deionized water and stored at 4 °C prior to use. 2.4. Preparation of APTES-G-CdTe QDs 1.0 mL of prepared G-CdTe QDs solution reacted with 0.5 mL of EDC/NHS (molar ratio 4:1) for 1 h at room temperature to active the carboxyl groups of G-CdTe QDs. Then 100.0 μL APTES (11.0 mM) was injected into the above mixture and stirred at 4 °C for 12 h. The APTES-G-CdTe QDs were obtained by centrifugation and washed with 62

Biosensors and Bioelectronics 90 (2017) 61–68

X. Fu et al.

slurry and then cleaned thoroughly in an ultrasonic cleaner with alcohol and water sequentially. 10.0 μL of APTES-G-CdTe QDs was coated on the cleaned GCE and dried at room temperature for 12 h. Then, 10.0 μL of DA with various concentrations was incubated on the electrode surface for 2 h. Ultimately, the modified electrode was incubated with 10.0 μL Cu2O-TAEA-Ru at room temperature for 6 h. Nonspecifically adsorbed species were removed by washing the electrode with doubly distilled water after each step. The schematic graph of the fabrication process for the ECL sensor is outlined in Scheme 1C. 2.8. Measurement procedure ECL response of the sensor was investigated with an MPI-A ECL analyzer in a detection cell containing 3.0 mL PBS (0.10 M, pH 7.0) solution at room temperature. The voltage of the photomultiplier tube (PMT) was set at 800 V and the potential scan was from −1.7V to 1.6 V (vs SCE) with a scan rate of 500 mV s−1 in the process of detection. The measurement was based on the ratio of cathode to anode ECL peak intensity (Ic /Ia). 3. Results and discussion 3.1. Characterization of nanomaterials Scheme 1. The preparation procedures of (A) APTES-G-CdTe QDs and (B) Cu2OTAEA-Ru. (C) Schematic illustration of the ECL ratiometric sensor.

Fig. 1A shows the TEM image of the G-CdTe QDs. A large number of CdTe QDs were uniformly distributed on the graphene surfaces. Nearly free CdTe QDs or pure graphene sheets were observed, revealing a strong bonding between the graphene and CdTe QDs. The TEM image of APTES-G-CdTe QDs (Fig. 1B) exhibited a similar morphology to CdTe QDs, indicating that the modification of small molecular APTES did not affect the morphology of G-CdTe QDs. As seen in Fig. 1C, the TEM image of Cu2O nanocrystals showed a cubic structures with the average diameter of 50 nm. Compared with Cu2O, Cu2O-TAEA-Ru (Fig. 1D) retained a similar cubic structures, but its interface became more blurred due to the modification of TAEA-Ru, suggesting that Cu2O-TAEA-Ru was successfully obtained. Fig. S1A displays the UV–vis absorption spectra of (a) GO, (b) CdTe QDs, and (c) G-CdTe QDs. For GO (curve a), a sharp absorption peak at 229 nm was observed, which was attributed to the characteristic of π→ π* transition for C˭C. For CdTe QDs (curve b), the broad characteristic absorption at around 542 nm from the first electronic transition was observed (Ge et al., 2008., Bao et al., 2006). In the UV–vis spectrum of

water. Finally, APTES-G-CdTe QDs were dispersed in 1.0 mL deionized water and stored in refrigerator prior to use. The preparation process of APTES-G-CdTe QDs is shown in Scheme 1A. 2.5. Preparation of amino-functionalized Cu2O nanocrystals Cu2O nanocrystals were prepared by following procedure (Gou and Murphy, 2004). Typically, poly (ethylene glycol) (16.0 mL, 62.50 mM) was added in CuCl2 aqueous solution (4.0 mL, 0.010 M). Then, 10.0 mL of NaOH (0.12 M) and 10.0 mL of ascorbic acid (4.76 mM) were added into above solution. After stirring for 5 min in air, the solution was remained undisturbed under the nitrogen atmosphere for 30 min to reaction completely. Finally, the formed nanoparticles were washed with double-distilled water and ethanol for three times, and then centrifuged at 12000 rpm for 10 min to obtain the precipitation of Cu2O nanocrystals. Amino-functionalized Cu2O nanocrystals were synthesized according to previous report (Ma et al., 2016). 0.010g of Cu2O nanocrystals was dispersed in 5.0 mL of ethanol containing 10.0 μL of APTES (5.56 M). The mix solution was stirred for 1.5 h at 70 °C. The final product was obtained by centrifugation and washed with ethanol. 2.6. Preparation of Cu2O-TAEA-Ru The preparation process of Cu2O-TAEA-Ru is shown in Scheme 1B. 3.0 mg amino-functionalized-Cu2O nanocrystals were dispersed in 1.0 mL distilled water. With the assistance of glutaraldehyde (GA), tris(2-aminoethyl) amine (TAEA) (50.0 mM, 100.0 μL) was decorated on the surface of Cu2O nanocrystals. 500.0 μL of activated tris(2,2′bipyridyl-4,4′-dicarboxylato)ruthenium(II) (Ru(dcbpy)32+) solution (2.0 mM) was modified on the as-obtained Cu2O-TAEA by using EDC and NHS as coupling agents to form Cu2O-TAEA-Ru composite. Followed by centrifugation at 9000 rpm for 6 min at 4 °C to remove unreacted reagents, Cu2O-TAEA-Ru composite was obtained. 2.7. Fabrication of ratiometric ECL sensor Before the fabrication, the glassy carbon electrode (GCE, Φ=4.0 mm) was polished sequentially with 0.3 and 0.05 µm alumina

Fig. 1. TEM images of (A) G-CdTe QDs (B) APTES-G-CdTe QDs, (C) Cu2O nanocrystals, and (D) Cu2O-TAEA-Ru. Insert of A: the enlarged image of G-CdTe QDs.

63

Biosensors and Bioelectronics 90 (2017) 61–68

X. Fu et al.

180.0 K

O1s

A

B

300.0K

405.5

572.1 412.0

583.5

C1s

Counts / s

Counts / s

120.0 K

60.0 K

200.0 K 405.0

412.5

572

Cd 100.0 K

O1s

583

Te

C1s

0.0 K 0.0K

0

200 400 600 Binding Energy / eV

800

0

111

C

150.0 k

200 400 600 Binding Energy / eV D

Counts/s

Intensity / a.u

100.0 k

200

220

Cu2p Ru3p O1s

50.0 K

N1s

311

C1s

110

20

800

0.0k

40 2 degree

60

0

80

200

400 600 800 Binding Energy / eV

1000

Fig. 2. XPS spectra of (A) GO and (B) G-CdTe (insets: the high-resolution curves of Cd and Te areas); (C) XRD pattern of Cu2O; (D) XPS spectra of Cu2O-TAEA-Ru (insets: the highresolution curves of Cu areas).

prepared.

G-CdTe QDs composite, the absorption peak from GO at 229 nm red shift to 258 nm, which was ascribed to the reduce of GO (curve c) (Yu et al., 2016). And the absorption band of CdTe QDs couldn’t be observed in the UV–vis spectrum of G-CdTe QDs, which may be ascribed to that the absorption of the QDs was strongly screened by the broad background absorption of graphene (Yu et al., 2016). To investigate the formation of G-CdTe QDs, the elemental analysis was studied with X-ray photoelectron spectroscopy (XPS) and the results are shown in Fig. 2A and B. For pure GO (Fig. 2A), O1s peak at 532.4 eV and C1s peak at 286. 6 eV were observed. For G-CdTe QDs (Fig. 2B), the characteristic peaks of O1s (532.4 eV), C1s (286.6 eV), Cd 3d (Cd 3d5/2 at 405.5 eV and Cd 3d3/2 at 412.0 eV) and Te 3d (Te 3d5/2 at 572.1 eV and Te 3d3/2 at 583.5 eV) were clearly observed in XPS spectrum of G-CdTe QDs (Vesely and Langer, 1971). XRD spectrum measurement was performed to characterize the structure of Cu2O nanocrystals. As show in Fig. 2C, the XRD pattern of Cu2O showed the peaks at 29.46, 36.5, 42.3, 61.4, and 73.5°, which corresponded to the(110), (111), (200), (220), and (311) planes of Cu2O, respectively. The results are consistent with the literature (Zhang et al., 2007; Zhou et al., 2014a). To confirm the formation of Cu2O-TAEA-Ru, UV–vis and XPS spectra were measured and the results are shown in Figs. S1B and 2D, respectively. The UV–vis spectrum of Cu2O nanocrystals showed a wide absorption band at about 510 nm (Fig. S1B, curve a) (Zhang et al., 2006). Ru(dcbpy)32+ displayed the characteristic absorption peaks at 302 nm and 465 nm (curve b) (Zhou et al., 2014b). After Ru(dcbpy)32+ was modified onto the surface of Cu2O nanocrystals by TAEA, the characteristic absorption peak of Ru(dcbpy)32+ at 304 nm was observed. However, the absorption peak of Cu2O nanocrystals at ~510 nm shifted to 480 nm (curve c). The XPS spectra of Cu2OTAEA-Ru are shown in Fig. 2D. As expected, the peaks of O1s (530.9, eV), Ru3p (490.78 eV), N1s (400.5 eV), and C1s (284.6 eV) were observed, which demonstrated the presence of TAEA and Ru(II) luminophor with carboxy groups (Wang et al., 2015a). Moreover, in the inset of Fig. 2D, the peaks at 933.4 eV and 953.3 eV are corresponding to Cu 2p3/2 and Cu 2p1/2, respectively, which are assigned to Cu1+ in Cu2O-TAEA-Ru (Chakravarty et al., 2015). Above results proved that the Cu2O-TAEA-Ru complex was successfully

3.2. Characterization of the stepwise fabrication of the sensor To characterize the fabrication of the sensor, CV curves at each modification step were recorded in PBS (0.10 M, pH 7.0) containing 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6]. As shown in Fig. S2, compared with bare GCE (curve a), the redox peak current of APTES-G-CdTe QDs/ GCE decreased due to the blocking effect of semiconductor QDs on the electron transfer (curve b). When DA was loaded onto the APTES-GCdTe QDs modified electrode surface (curve c), the peak currents increased due to the fact that DA could accelerate electron transfer. Finally, with the incubation of Cu2O-TAEA-Ru onto the electrode, the peak currents decreased because of the poor conductivity of Cu2OTAEA-Ru. Electrochemical impedance spectroscopy (EIS) was also carried to characterize the fabrication of the sensor and the typical impedance spectra (presented in the form of the Nyquist plot) are shown in Fig. 3A. The semicircle diameter in the impedance spectrum equals to the electron transfer resistance (Ret), which controls the electron transfer kinetics of the redox probe at the electrode interface. Compared with the bare GCE (curve a), an obviously increased Ret was discovered at the APTES-G-CdTe QDs modified GCE (curve b), ascribing to the poor conductivity of APTES-G-CdTe QDs. Then, a decreased Ret was obtained after the modification of DA (curve c), indicating that DA could accelerate the electron transfer between the redox probe and the electrode. With the immobilization of Cu2OTAEA-Ru, the Ret was increased (curve d), which could be ascribe to the fact that Cu2O-TAEA-Ru modified onto the electrode could hinder the electron transfer between the redox probe [Fe(CN)6]4−/3− and the electrode. The EIS and CV results confirmed the stepwise fabrication of the sensor. ECL-potential responses at each modification step were recorded in 3.0 mL 0.10 M PBS solution (pH 7.0). As shown in Fig. 3B, the APTESG-CdTe QDs modified electrode showed an intensive cathode ECL emission peak at −1.63 V (curve b). The possible ECL route is proposed as follows (Liu et al., 2015): 64

Biosensors and Bioelectronics 90 (2017) 61–68

X. Fu et al.

A

9000 ECL intensity / a.u

-360

-Z

-240 -120 0 a

0

200

c

400 Z

b

6000 d c

3000

a

d b

600

B

0 -2

800

-1

0 Potential / V

1

2

Fig. 3. (A) Nyquist diagrams of electrochemical impedance response in PBS (pH 7.0) containing 5.0 mM K3[Fe(CN)6]/K4[Fe(CN)6] and (B) ECL responses in 0.10 M PBS buffer (pH 7.0) of different modified electrodes: (a) bare GCE, (b) APTES-G-CdTe QDs/GCE, (c) DA/APTES-G-CdTe QDs/GCE, and (d) Cu2O-TAEA-Ru/DA/APTES-G-CdTe QDs/GCE.

G-CdTe QDs+ne− → nG-CdTe QDs•−

TAEA-Ru constant in 0.10 M PBS (pH 7.0), we measured the ECL response of bare GCE with varied concentration of G-CdTe QDs in 0.10 M PBS (pH 7.0). As expected, with the increase of G-CdTe QDs concentration, anode ECL signal from TAEA-Ru increased (Fig. 4D). The decrease in ECL intensity from donor (Fig. 4C) with increasing the concentration of acceptor, and the increase in ECL intensity from acceptor with increasing donor concentration (Fig. 4D), are enough to confirm the ECL-RET from donor (G-CdTe QDs) to acceptor (TAEARu). Furthermore, it has been reported that CdTe QDs can enhance the ECL emission of Ru(bpy)32+ by energy transfer (Wang et al., 2016). Additionally, in order to investigate the quenching effect of Cu2O nanocrystals, different amount of Cu2O nanocrystals has been added in detection cell. As seen in Fig. 4E, an obvious quenching signal at APTES-G-CdTe QDs/GCE was observed with increasing the amount of Cu2O nanocrystals. The reason is as following Eq. (1). Cu2O nanocrystals could effectively promote the reduction reaction of O2 (Wang et al., 2015b). In the presence of Cu2O nanocrystals, dissolved O2, as the coreactant of G-CdTe QDs, was consumed, thus resulting in the quenching of ECL signal of G-CdTe QDs/O2.

O2+2G-CdTe QDs•−+2H+ → H2O2+2G-CdTe QDs* G-CdTe QDs* → G-CdTe QDs+hν After DA was immobilized on the APTES-G-CdTe QDs/GCE via the specific recognition of the diol of DA to the oxyethyl group of APTES, the cathode ECL intensity showed a negligible change (curve c). With the incubation of Cu2O-TAEA-Ru onto the modified electrode, cathode ECL intensity showed an obvious decrease and another new anode ECL peak can be observed at +1.55 V (curve d) at the same time. The new anode ECL peak was from the self-enhanced TAEA-Ru(II) compounds, and the possible luminous mechanisms of TAEA-Ru (II) were expressed as the following: TAEA-Ru (II) −2 e− → TAEA•+-Ru(III) TAEA•+-Ru(III) → TAEA•-Ru(III)+H+ TAEA•-Ru(III) → TAEA-Ru* (II)

Cu2O+O2+H+ → Cu2++H2O

•

(1)

On the other hand, the spectral overlap between the UV–vis absorption spectrum of Cu2O (as energy acceptor) and the ECL spectrum of G-CdTe QDs (as energy donor) was investigated. As shown in Fig. 4F, an excellent spectral overlap between them was observed, indicating the possibility of ECL-RET from G-CdTe QDs to Cu2O. Hence, Cu2O exhibited a dual quenching effect to G-CdTe QDs, including ECL-RET from G-CdTe QDs to Cu2O and the consumption of coreatant O2 by Cu2O.

TAEA-Ru* (II) → TAEA -Ru(II)+hν

3.3. The quenching mechanism of ratiometric ECL sensor In order to speculate the energy transfer between G-CdTe QDs and TAEA-Ru, the following experiments were conducted. Firstly, the ECL spectra of G-CdTe QDs and TAEA-Ru were detected to confirm the energy donor and acceptor since the luminophore emits short wavelength as the energy donor and that emits long wavelength as the energy acceptor. As seen in Fig. 4A, the maximum emission wavelengths of G-CdTe QDs and TAEA-Ru appeared at 606 nm and 638 nm, respectively. Obviously, G-CdTe QDs and TAEA-Ru would serve as the donor and acceptor in ECL energy transfer, respectively. Secondly, the spectral overlap between the ECL spectrum of donor (G-CdTe QDs) and the UV–vis absorption spectrum of acceptor (TAEA-Ru) was investigated and the results are presented in Fig. 4B. Obviously, the UV–vis absorption spectrum of TAEA-Ru partly overlapped with the ECL spectral G-CdTe QDs, indicating the possibility of ECL-RET between TAEA-Ru and G-CdTe QDs. Thirdly, keeping the amount of G-CdTe QD constant at the modified electrode, we measured the ECL responses of G-CdTe QDs/GCE in 0.10 M PBS (pH 7.0) containing varied concentration of Ru-TAEA. As shown in Fig. 4C, with increasing the concentration of TAEA-Ru, cathode ECL intensity from G-CdTe QDs decreased and another new anode ECL peak from TAEA-Ru at +1.55 V increased. On the other hand, TAEA-Ru exhibited an excellent water solubility, thus was added in the detection cell instead of immobilization on electrode surface. Keeping the concentration of

3.4. Calibration curve for DA detection As shown in Fig. 5A, under the optimized experimental conditions (Fig. S3), with the increase of the concentrations of DA incubated onto the sensor, the cathode ECL intensity around −1.63 V descended while the anodic ECL intensity around +1.55 V increased. As a result (Fig. 5B), a good linear relationship was obtained between the logarithmic value of DA concentration and the ratio of cathode to anode ECL peak intensity (Ic/Ia) in the concentration range from 10.0 fM to 1.0 nM. The regression equation was Ic/Ia=−1.749−0.232logc (R2=0.999), here, c was the concentration of DA. The detection limit (defined as LOD=3SB/m) was estimated to 2.9 fM, where SB is the standard deviation of the blank and m is the slope of the corresponding calibration curve. Compared with reported ECL test platforms for DA (Table S1) (Zhang et al., 2013b; Huang et al., 2015; Feng et al., 2016; Fu et al., 2015; Tang et al., 2015), our designed ECL system in this work showed a wider linear range and a lower detection limit for DA. This could be ascribed to the combination of ratiometric sensing with dual molecular recognition strategy, and the dual quenching effect of 65

Biosensors and Bioelectronics 90 (2017) 61–68

X. Fu et al.

606 nm 638 nm

a

600 650 Wavelength / nm

700

750

400

C a

6000 d

4000 d

2000

a

d

3000 a

1500 0

-0.8

0.0 0.8 Potential / V

1.6

-1.6

E

-0.8

0.0 0.8 Potetial / V

1.6

F a

6000

Cu2O Absorbance

(Acceptor)

CdTe QD (Donor)

4000 d

2000

ECL intensity /a.u

-1.6

ECL intensity / a.u

800

4500

0

8000

600 Wavelength / nm

D

ECL intensity /a.u.

8000

550

CdTe QD (Donor)

Absorbance

b

500

ECL intensity / a.u

Ru (Acceptor)

ECL intensity /a.u

B

ECL intensity / a.u

A

0 -1.5

-1.0 -0.5 Potential / V

0.0

200

400 600 wavelength / nm

800

Fig. 4. (A) ECL spectra of (a) G-CdTe QDs and (b) Ru-TAEA in 0.10 M PBS (pH 7.0). (B) ECL spectra of G-CdTe QDs and UV–vis absorption spectra of Ru-TAEA. (C) ECL response curves of G-CdTe QDs/GCE in 0.10 M PBS (pH 7.0) containing Ru-TAEA from (a) 0 to (d) high concentration. (D) ECL response curves of bare GCE in 0.10 M PBS (pH 7.0) containing constant concentration of TAEA-Ru with increasing G-CdTe QDs concentration from (a) 0 to (d) high concentration. (E) ECL response curves of G-CdTe QDs/GCE in 0.10 M PBS (pH 7.0) containing varied concentration of Cu2O from (a) 0 to (d) high concentration. (F) ECL spectra of G-CdTe QDs and UV–vis absorption spectra of Cu2O.

lesterol and epinephrine were tested in the case of the approximate concentration of these interfering substances in a real sample. The interference experiments were performed using 4.0 mM cholesterol, 1.0 mM lactic acid, 0.3 mM uric acid, 30 mM glucose and 70 nM epinephrine. As shown in Fig. 5D, above interferences didn’t cause an obvious signal changes. However, in the case of 3.0×10−11 M DA, the ECL intensity ratio (Ic/Ia) decreased substantially. The reasons were following. Since uric acid, lactic acid and cholesterol haven’t the diol, they could not be captured onto the electrode, and thus the second signal probe could not be incubated onto the electrode. As for the epinephrine and glucose, although they have a similar the diol as DA, it is difficult or impossible for them to react with carboxyl due to the absence of primary amine species, and thus the second signal probe also could not be incubated onto the electrode. Above results suggested that the proposed ratiometric ECL sensor exhibited a high selectivity toward DA.

Cu2O to G-CdTe QDs as well as the ECL-RET from G-CdTe QDs to TAEA-Ru. The low detection limit of the sensor would exhibit potential applications for detecting lower concentrations DA in unconventional body fluids such as brain extracellular DA, because the basal level of brain extracellular DA was previously reported to be within 5–20 nM (Lin et al., 2010). Besides, the low detection limit may show the potential application in figure blood which was usually detected by diluting real samples with many times. 3.5. Stability and interference determination of the sensor The stability of each step was investigated in Supplementary materials and results are showed in Fig. S4. Additionally, the stability of the ECL ratiometric sensor was demonstrated under consecutive cyclic potential scans for seven cycles at 4.0×10−11 M of DA. As depicted in Fig. 5C, the ECL intensity did not show any significant changes. The probably reasons for this satisfying stability were presented as follows. G-CdTe QDs and self-enhanced ECL complex showed an excellent luminous stability. Simultaneously, the stability of chemical bond in the dual molecular recognition strategy was satisfactory. In order to evaluate the selectivity of the proposed ECL sensor, possible interferences concluding uric acid, glucose, lactic acid, cho-

3.6. Recovery experiments Abnormal DA level in serum is usually used to indicate various pathosis, such as adrenal medulla tumor, hypertension, myocardial infarction and so on. In this work, human serum samples obtained from the Ninth People's Hospital (Chongqing, China) were selected as 66

Biosensors and Bioelectronics 90 (2017) 61–68

X. Fu et al.

9000 1.6

B

f

6000

a

1.2 Ic / Ia

ECL intensity / a.u

A

a

3000

0.8

f 0.4

0 -1.6

-0.8

0.0 0.8 Potentive / V

-13.5

1.6

-12.0 -10.5 log (cDA / M)

-9.0

80 D

60

4000

Ic / Ia

ECL intensity / a.u

6000

C

2000

40 20

0

0

25

50 75 Time / s

100

DA terol acid hrine c acid cose les actic glu uri nep l cho epi

Fig. 5. (A) ECL responses of the sensor towards different concentrations of DA in air-saturated pH 7.0 PBS from (a) to (f): 1.0×10−14 M, 1.0×10−13 M, 5.0×10−12 M, 4.0×10−11 M, 1.0×10−10 M, 1.0×10−9 M, respectively. (B) Linear calibration between logarithmic value of DA concentration and the ratio of cathode to anode ECL peak intensity. (C) Continuous cyclic scans of the sensor with 4.0×10−11 M DA in air-saturated pH 7.0 PBS buffer. (D) Selectivity of the proposed sensor to 3.0×10–11 M of DA, 4.0 mM cholesterol, 1.0 mM lactic acid, 70 nM epinephrine, 0.3 mM uric acid and 30 mM glucose.

the dual molecular recognition strategy was firstly applied to detect biological small molecule DA. The prepared sensor showed a wide linear response range and a low detection limit for DA. In previous ECL ratiometric assays, the cathodic and anodic ECL emitters generally shared the same coreactant. The introduction of two coreatants without cross interference would break this restriction, thus holding a new perspective for constructing an efficient and convenient ratiometric ECL system.

Table 1 Recoveries of DA in diluted serum samples using the proposed ratiometric ECL sensor. Sample

Added (pM)

Found (pM)a

Diluted serum 1

0 6.0 9.0 0 4.0 8.0

4.3 ± 0.3 10.0 ± 0.2 13.8 ± 0.3 5.1 ± 0.4 9.3 ± 0.2 13.7 ± 0.5

Diluted serum 2

Recovery (%)

95 106 105 108

Acknowledgements a

Mean ± SD, n = 3.

This work was supported by National Natural Science Foundation of China (51473136 and 21575116), Natural Science Foundation Project of Chongqing City (CSTC-2014JCYJA20005).

the real samples. It was reported that the concentration of DA in human serum sample was below 0.9 nM. Since the linear range of our sensor was from 10.0 fM to 1.0 nM, the dilution of serum samples is necessary. 100 fold diluted serum was applied, and the DA concentration of 4.3 ± 0.3 pM and 5.1 ± 0.4 pM were obtained in two diluted serum samples with the proposed ECL sensor, respectively, and results are shown in Table 1. For undiluted serum samples, the concentration of DA were estimated to 0.43 ± 0.03 nM and 0.51 ± 0.04 nM, respectively, which are broadly consistent with the literature (Chen et al., 2007). Additionally, recovery experiments were also used as a preliminary evaluation to illustrate the real application of our sensors. Recovery experiments were performed by the standard addition method. As presented in Table 1, the recoveries ranged from 95% to 108%, indicating an acceptable accuracy of our proposed sensor.

Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at doi:10.1016/j.bios.2016.11.025. References Bao, H.F., Wang, E., Dong, S.J., 2006. Small 2, 476–480. Cheng, Y., Huang, Y., Lei, J.P., Zhang, L., Ju, H.X., 2014. Anal. Chem. 86, 5158–5163. Chen, A.Y., Gui, G.F., Zhuo, Y., Chai, Y.Q., Xiang, Y., Yuan, R., 2015. Anal. Chem. 87, 6328–6334. Chen, F.N., Zhang, Y.X., Zhang, Z.J., 2007. Chin. J. Chem. 25, 942–946. Chakravarty, A., Bhowmik, K., Mukherjee, A., De, G., 2015. Langmuir 31, 5210–5219. Feng, Y.Q., Dai, C.H., Lei, J.P., Ju, H.X., Cheng, Y.X., 2016. Anal. Chem. 88, 845–850. Fu, X.M., Feng, J.H., Tan, X.R., Lu, Q.Y., Yuan, R., Chen, S.H., 2015. RSC Adv. 5, 42698–42704. Gou, L.F., Murphy, C.J., 2004. J. Mater. Chem. 14, 735–738. Ge, C.W., Xu, M., Liu, J., Lei, J.P., Ju, H.X., 2008. Chem. Commun., 450–452. Hao, N., Li, X.L., Zhang, H.R., Xu, J.J., Chen, H.Y., 2014. Chem. Commun. 50, 14828–14830. Huang, Y., Lei, J.P., Cheng, Y., Ju, H.X., 2016. Biosens. Bioelectron. 77, 733–739.

4. Conciusion A new group of luminescent reagents, CdTe QDs and Ru(dcbpy)32+, was introduced to ratiometric ECL system. Based on G-CdTe QDs/O2 as cathodic emitter and self-enhanced ECL Ru complex as anodic ECL emitter, a novel dual-potential ratiometric ECL sensor combined with 67

Biosensors and Bioelectronics 90 (2017) 61–68

X. Fu et al.

Wang, H.J., Zhang, J., Yuan, Y.L., Chai, Y.Q., Yuan, R., 2015b. RSC Adv. 5, 58019–58023, (b). Wu, P., Hou, X.D., Xu, J.J., Chen, H.Y., 2016. Nanoscale 8, 8427–8442. Yu, Y.Q., Zhang, H.Y., Chai, Y.Q., Yuan, R., Zhuo, Y., 2016. Biosens. Bioelectron. 8, (58−15). Zhang, H.R., Wu, M.S., Xu, J.J., Chen, H.Y., 2014. Anal. Chem. 86, 3834–3840. Zhang, H.R., Xu, J.J., Chen, H.Y., 2013a. Anal. Chem. 85, 5321–5325, (a). Zhang, L., Cheng, Y., Lei, J.P., Liu, Y.T., Hao, Q., Ju, H.X., 2013b. Anal. Chem. 85, 8001–8007, (b). Zhang, J.T., Liu, J.F., Peng, Q., Wang, Xun, Li, Y.D., 2006. Chem. Mater. 18, 867–871. Zhang, H.G., Zhu, Q.S., Zhang, Y., Wang, Y., Zhao, L., Yu, B., 2007. Adv. Funct. Mater. 17, 2766–2771. Zhao, H.F., Liang, R.P., Wang, J.W., Qiu, J.D., 2015. Chem. Commun. 51, 12669–12672. Zhou, D.L., Feng, J.J., Cai, L.Y., Fang, Q.X., Chen, J.R., Wang, A.J., 2014a. Electrochim. Acta 115, 103–108, (a). Zhou, Y., Zhuo, Y., Liao, N., Chai, Y.Q., Yuan, R., 2014b. Talanta 129, 219–226, (b).

Huang, C.X., Chen, X., Lu, Y.L., Yang, H., Yang, W.S., 2015. Biosens. Bioelectron. 63, 478–482. Liu, L.L., Ma, Q., Li, Y., Liu, Z.P., Su, X.G., 2015. Biosens. Bioelectron. 63, 519–524. Lei, Y.M., Huang, W.X., Zhao, M., Chai, Y.Q., Yuan, R., Zhuo, Y., 2015. Anal. Chem. 87, 7787–7794. Lin, Y.Q., Zhang, Z.P., Zhao, L.Z., Wang, X., Yu, P., Su, L., Mao, L.Q., 2010. Biosens. Bioelectron. 25, 1350–1355. Ma, H.M., Li, Y., Wang, Y.L., Hu, L.H., Zhang, Y., Fan, D.W., Yan, T., Wei, Q., 2016. Biosens. Bioelectron. 78, 167–173. Tang, L.J., Li, S., Han, F., Liu, L.Q., Xu, L.G., Ma, W., Kuang, H., Li, A., Wang, L.B., Xu, C.L., 2015. Biosens. Bioelectron. 7, (17–12). Vesely, C.J., Langer, D.W., 1971. Phys. Rev. B: Solid State 4, 451–462. Wang, H.J., He, Y., Chai, Y.Q., Yuan, R., 2014. Nanoscale 6, 10316–10322. Wang, Y.Z., Hao, N., Feng, Q.M., Shi, H.W., Xu, J.J., Chen, H.Y., 2016. Biosens. Bioelectron. 77, 76–82. Wang, H.J., Yuan, Y.L., Chai, Y.Q., Yuan, R., 2015a. Small 11 (30), 3703–3709, (a).

68