Enzymatic Oxydate-Triggered Self-Illuminated Photoelectrochemical

Jan 29, 2016 - Schematic Illustration of (A) Digital Multimeter-Based Photoelectrochemical (PEC) Immunoassay Mode toward Target PSA by Coupling Enzyma...
5 downloads 11 Views 3MB Size
Download PDF
Article pubs.acs.org/ac

Enzymatic Oxydate-Triggered Self-Illuminated Photoelectrochemical Sensing Platform for Portable Immunoassay Using Digital Multimeter Jian Shu,† Zhenli Qiu,† Qian Zhou,† Youxiu Lin,† Minghua Lu,*,‡ and Dianping Tang*,† †

Key Laboratory of Analysis and Detection for Food Safety (MOE & Fujian Province), Institute of Nanomedicine and Nanobiosensing, Department of Chemistry, Fuzhou University, Fuzhou 350108, People’s Republic of China ‡ Institute of Environmental and Analytical Science, School of Chemistry and Chemical Engineering, Henan University, Kaifeng 475004, Henan, People’s Republic of China S Supporting Information *

ABSTRACT: Herein a novel split-type photoelectrochemical (PEC) immunosensing platform was designed for sensitive detection of low-abundance biomarkers (prostate-specific antigen, PSA, used in this case) by coupling a peroxyoxalate chemiluminescence (PO-CL) self-illuminated system with digital multimeter (DMM) readout. The PEC detection device consisted of a capacitor/DMM-joined electronic circuit and a PO-CL-based self-illuminated cell. Initially, reduced graphene oxide-doped BiVO4 (BiVO4-rGO) photovoltaic materials with good photoelectric properties was integrated into the capacitor/DMM-joined circuit for photocurrent generation in the presence of hydrogen peroxide (H2O2, as the hole-trapping reagent). A sandwich-type immunoreaction with target PSA was carried out in capture antibody-coated microplates by using glucose oxidase/detection antibody-conjugating gold nanoparticle (pAb2-AuNP-GOx). Accompanying the sandwiched immunocomplex, the labeled GOx could oxidize glucose to produce H2O2. The as-generated H2O2 could act as the coreaction reagent to trigger the chemiluminescence of the peroxyoxalate system and the PEC reaction of the BiVO4-rGO. Meanwhile, the self-illuminated light could induce photovoltaic material (BiVO4-rGO) to produce a voltage that was utilized to charge an external capacitor. With the switch closed, the capacitor could discharge through the DMM and provide an instantaneous current. Different from conventional PEC immunoassays, the asgenerated photoelectron was stored in the capacitor and released instantaneously to amplify the photocurrent. Under the optimal conditions, the transient current increased with the increasing target PSA concentration in the dynamic working range from 10 pg mL−1 to 80 ng mL−1 with a detection limit (LOD) of 3 pg mL−1. This work demonstrated for the first time that the peroxyoxalate CL system could be used as a suitable substitute of physical light source to apply in PEC immunoassay. In addition, this methodology afforded good reproducibility, precision, and high specificity, and the method accuracy matched well with the commercial PSA ELISA kit. Importantly, the developed split-type photoelectrochemical immunoassay could not only avoid the interfering of the biomolecules relative to the photovoltaic materials but also eliminate the need of an exciting light source and expensive instrumentation, thus representing a user-friendly and low-cost assay protocol for practical utilization in quantitative low-abundance proteins.

P

assays,13−15 but most have involved expensive instrumentations, external exciting light source, or low sensitivity and are unsuitable for routine use. Nowadays, an ongoing effort has been focused on the field of assay development to simplify the assay procedure while preserving the essential benefits in sensitivity, robustness, broad applicability, and suitability to automation.16,17 For successful development of PEC immunoassays, fabrication of the sensing platform is very crucial.18 In the conventional PEC immunoassays, the biomolecules (e.g., antigen and antibody) are

hotoelectrochemical (PEC) sensing, a promising analytical technique based on the relationship between the change of photocurrent and the concentration of target analyte, has gained increasing attention in recent years.1−4 In comparison with conventional electrochemistry, chemiluminescence (CL) and electrochemiluminescence methods that directly detect the changes of the electrical/optical signals caused by the electrochemical/or chemical reactions, the PEC sensing strategy records the photocurrent change under an external light source excitation.5−8 Because of its high selectivity/ sensitivity, short response time, minimal sample consumption, and low-background signal, the photoelectrochemical sensor has become a powerful tool in the fields of clinical diagnostics, food safety, and environmental monitoring.9−12 Methods based on PEC detection principle have been developed for immuno© 2016 American Chemical Society

Received: January 20, 2016 Accepted: January 29, 2016 Published: January 29, 2016 2958

DOI: 10.1021/acs.analchem.6b00262 Anal. Chem. 2016, 88, 2958−2966

Article

Analytical Chemistry

Scheme 1. Schematic Illustration of (A) Digital Multimeter-Based Photoelectrochemical (PEC) Immunoassay Mode toward Target PSA by Coupling Enzymatic Oxydate (H2O2)-Triggered Peroxyoxalate Self-Illuminated System with an External Capacitor on the BiVO4-rGO Photovoltaic Material, (B) H2O2-Induced Peroxyoxalate Chemiluminescence (CL) Reaction, and (C) H2O2-Triggered Photoelectron Generation Mechanism on the BiVO4-rGO

for label-free detection of mycotoxins that utilized a digital multimeter with computer connection for data logging.31 The Crooks’s group designed a DMM-based electrochemical microfluidic detection device on an aptamer-based origami paper.32 Owing to the capability of intuitively evaluating electrical signal, the DMM can be applied in the fields of analytical chemistry.33−35 To this end, our motivation in this work was to integrate the user-friendly, low-cost, and portable digital multimeter with the split-type mode for advanced development of PEC immunoassays. Another aspect on the improvement of photoelectrochemical sensors lies in the design of the exciting light source. Conventional PEC analytical methods not only need the electrochemical apparatus but also include an external light instrument that excites the photovoltaic materials.36,37 This makes the detection device complicated. Inspiringly, electrochemical and luminescent methods as well as their combination have been expended to electrochemiluminescence-based detection systems.38−40 Unlike fluorescence/phosphorescence, electronic excited state of chemiluminescence (CL) is the product of a chemical reaction rather than the absorption of a photon.41 It is the antithesis of a photochemical reaction, in which light is used to drive an endothermic chemical reaction.42 Luminol-based CL system, one of the most classical CL systems, can be utilized as the excitation light to substitute the physical light source.43 Ge et al. utilized the N-(aminobutyl)-N(ethylisoluminol)-functionalized gold nanoparticles as the internal CL light source for PEC ATP detection with DMM readout.44 Unfortunately, luminol-based CL systems are often carried out at a constant emission wavelength of ∼432 nm,

immobilized onto the photovoltaic-material-based sensing interface via adsorption, encapsulation, self-assembly, and covalent binding.19−23 One major disadvantage of using these techniques is the instability of the immobilized proteins during continuous use, the additional diffusion barrier resulting from the entrapped materials, or the decrease of protein bioactivity, thus resulting in unsatisfactory results with the photovoltaic characteristics.13,24,25 Moreover, the intrinsic strong oxidation characteristics of holes generated in photovoltaic materials have inevitably harmful effects on biomolecules even though the materials that respond to visible light were widely used in these systems to avoid irradiation with ultraviolet light.4,26,27 Recently, our group constructed a new split-type PEC immunosensing platform based on target-induced nanoenzyme reactor mediated hole-trapping strategy.13 Results indicated that introduction of the split-type assay format effectively avoided the damage of the proteins by physical light source because the antigen−antibody reaction and photocurrent generation were implemented into two cells, respectively. Unfavorably, this method involved an expensive electrochemical workstation and external exciting source system. In contrast, the emergence of digital multimeter (DMM) opens a new horizon for the development of advanced PEC sensors. Typically, a digital multimeter, one of the most common items of test equipment used in the electronic industry today, is able to provide excellent readings of the basic measurements of amps, volts, and ohms.28 Digital multimeters are now far more common, but analog multimeters are still preferable in some cases, for example, when monitoring a rapidly varying value.29,30 Yang et al. previously described a new lab-on-chip 2959

DOI: 10.1021/acs.analchem.6b00262 Anal. Chem. 2016, 88, 2958−2966

Article

Analytical Chemistry

Figure 1. Typical SEM images of (A) BiVO4 and (B) BiVO4-rGO (inset: magnification image of reduced graphene oxide); (C) EDS analysis of BiVO4-rGO [inset: Zeta potential profiles of (a) GO, (b) BiVO4 and (c) aminated BiVO4]; (D) Raman spectra of BiVO4 and BiVO4-rGO [insets: D and G bands of (top) graphene oxide and (bottom) BiVO4-rGO]; (E) XRD patterns of BiVO4 and BiVO4-rGO; (F) UV−vis absorption spectra of (a) BiVO4 and (b) BiVO4-rGO [insets: photographs of (left-a) BiVO4 and (left-b) BiVO4-rGO; magnification of UV−vis absorption spectra of (right-a) BiVO4 and (right-b) BiVO4-rGO].

a low-cost and portable PEC immunoassay by using a PO-CL exciting light source with DMM readout.

which limits the selectable photovoltaic materials. Thus, searching for a novel CL system with high luminescent efficiency and multiwavelength selectivity to substitute a physical light source would be advantageous. Peroxyoxalate chemiluminescence (PO-CL) system with the quantum yield up to 50% was first reported by Rauthut in 1967,45 and has been widely applied in the CL fields.46,47 The emission is generated by reaction of an oxalate ester with hydrogen peroxide in the presence of a suitably fluorescent energy acceptor.48 Compared with traditional luminol-based CL systems, one promising merit of using the PO-CL system is that one can control/tailor the reaction conditions and activate their different fluorophores in a very predictable wavelength to meet the needs of specific PEC applications. To the best of our knowledge, there is no report focusing on PEC immunoassays by using the PO-CL system as the exciting light source. Herein we report the proof-of-concept of a novel and splittype DMM-based PEC immunoassay for sensitive detection of prostate-specific antigen (PSA; a glycoprotein enzyme encoded in humans by KLK3 gene which is often elevated in the presence of prostate cancer or other prostate disorders) using peroxyoxalate self-illuminated CL system (Scheme 1). The photocurrent is generated on a highly efficient photovoltaic nanomaterial (reduced graphene oxide-doped BiVO4, BiVO4rGO) in a homemade photoelectrochemical detection device. The coreaction reagent (H2O2) for PO-CL system is acquired classical enzyme immunoassay: glucose oxidase/detection antibody-conjugating gold nanoparticle (pAb2-AuNP-GOx) toward the oxidization of glucose, in capture antibody-coated microplate with a sandwich-type assay format. The selfilluminated light by PO-H2O2 system can trigger BiVO4-rGO to produce a voltage and charge the external capacitor. With the switch closed, the capacitor discharges through the DMM to acquire an instantaneous current. By monitoring the change in the current, we can quantitatively determine the concentration of target PSA in the sample. The aim of this study is to explore

■

EXPERIMENTAL SECTION Photoelectrochemical Measurement. Scheme 1 gives the immunoreaction process toward target PSA and the generation mechanism of photocurrent coupling with the POCL self-illuminated system in the manmade PEC detection cell. Partial Experimental Sections was described in detail in the Supporting Information. In the presence of target PSA, the sandwich-type immunoreaction was carried out on anti-PSA capture antibody (mAb1)-coated microplates by using pAb2AuNP-GOx as detection antibody. The plates were washed again. 110 μL of pH 6.0 PBS (50 mM) containing 4 mM glucose was added to each well, and reacted for 12 min at 37 °C. During this process, glucose was oxidized to D-glucono-βlactone and hydrogen peroxide. After that, 50 μL of the reaction solution including H2O2 was injected into the PEC detection cell from the microplate by using micro syringe. The plastic rip equipped with two-electrode system was closed (note: the two electrodes must be immersed into the solution by controlling electrode height). Following that, another 50 μL of the reaction solution was injected into the CL reaction cell through the inlet channel. During this process, the generated H2O2 could act as the coreaction reagent to trigger the selfchemiluminescence of PO-CL system. Meanwhile, H2O2 in the PEC detection cell could induce the photoelectrochemical reaction of the immobilized BiVO4-rGO on the GCE under the PO-CL light irradiation. The as-produced charges between two electrodes were accumulated on the external capacitor (∼within 30 s) under the switch opened. With the switch closed, the stored electrons were discharged from the capacitor, and flowed through the digital multimeter, thus producing an instantaneous current. The current was recorded and registered as the signal of the immunoassay. Currents against the logarithm of PSA concentration calibration plots were then constructed. 2960

DOI: 10.1021/acs.analchem.6b00262 Anal. Chem. 2016, 88, 2958−2966

Article

Analytical Chemistry

for the aggregation of negatively charged graphene oxide. The doped graphene oxide in the BiVO4 was easily reduced to the reduced graphene oxide during the hydrothermal process.49 Introduction of reduce graphene oxide was not only expected to improve the conductivity of the nanocomposites but also was favorable for electron communication from BiVO4 to rGO under the light excitation.50 Logically, one puzzling question arises to whether the doped graphene oxide was readily reduced during the hydrothermal process. To verify this point, we used Raman spectroscopy (a useful tool to characterize the crystallization, stereochemical purity, and electronic properties) to investigate the nanostructures. Figure 1D represents Raman spectra of BiVO4 and BiVO4-rGO, respectively. The peaks at 126 and 212 cm−1 derived from external vibration modes of BiVO4. The corresponding peaks assigned to VO43− bending vibration (324 and 368 cm−1) and V−O stretching vibration (706 and 830 cm−1) were also observed.51 The Raman shifts indicated between BiVO4 and BiVO4-rGO corresponded to the typical bands of monoclinic BiVO4.52 Meanwhile, BiVO 4-rGO exhibited obvious Raman shifts at 1335 and 1604 cm−1 that was absent in pure BiVO4 spectrum, which was ascribed to D and G bands of graphene oxide. Particular interesting was the ratio of D band to G band in the intensity. As shown from the inset in Figure 1D, the ratio of ID/IG for graphene oxide was 1.12, while that of BiVO4-rGO decreased to 0.92, indicating the formation of reduced graphene oxide in BiVO4-rGO.49 Certainly, another issue to be answered was whether introduction of reduced graphene oxide caused the change of BiVO4 in the crystal phase transition. To clarify this issue, we used XRD to investigate BiVO4 before and after doping with reduced graphene oxide. As shown in Figure 1E, BiVO4 and BiVO4-rGO possessed highly similar XRD patterns and diffraction peaks at 18.7°, 18.9°, 28.9°, 30.5°, 35.2° 39.8°, 46.7°, 53.3°, 58.5°, and 59.3° for (110), (011), (121), (040), (002), (211), (042), (161), (321), and (123) crystal planes of monoclinic BiVO4, respectively, which was in accordance with standard card of no. 14-0688 (for monoclinic BiVO4 with high stability and activity in visible light range).53,54 Due to high crystallinity of BiVO4 and small-amount rGO (∼1.0 wt %) in BiVO4-rGO, no obvious diffraction peak of reduced graphene oxide was observed in XRD pattern. The results revealed that the as-prepared BiVO4-rGO could still maintain the crystal phase of BiVO4, which provided a favorable condition for the development of BiVO4-rGO-based PEC sensing platform. In addition, UV−vis adsorption spectroscopy was employed for characterization of BiVO4-rGO. As indicated from Figure 1F, BiVO4-rGO exhibited similar UV−vis absorption spectra with that of BiVO4 alone in the region of 300−800 nm. However, the absorbance of BiVO4-rGO was higher than that of BiVO4 and displayed a slight bathochromic shift relative to BiVO4 (Figure 1F, right inset), which was ascribed to the bonding interaction between BiVO4 and graphene oxide.55 Such a high absorbance was conducive for the generation of photocurrent under the visible light. BiVO4-rGO-Based Electrochemical and Photoelectrochemical Characteristics. To fabricate a highly efficient PEC immunoassay, design of photovoltaic materials is very crucial for photocurrent generation. In this study, the photoelectron is produced by semiconductive BiVO4, which is stored into the external capacitor. One concern to be addressed was whether the doped rGO was favorable for electron transfer from BiVO4 to the electrode. To elucidate this point, we used electro-

After each run, the PEC detection cell was removed and the solution in the CL reaction cell was flowed out from the outlet channel. Meanwhile, the two electrodes with the plastic lid and CL reaction cell were washed. Subsequently, the new PEC detection cell was inserted in the CL reaction cell that containing 0.5 mL of fresh PO-CL solution for next PEC measurement.

■

RESULTS AND DISCUSSION Characterization of BiVO4-rGO-Based Photovoltaic Nanomaterials. Scheme 1 represents the DMM-based PEC immunosensing platform by coupling an enzymatic oxidateinduced peroxyoxalate self-illuminated system with a BiVO4rGO-based photocurrent generation process. Compared with conventional PEC immunoassays, the exciting light source is generated through H2O2-triggered peroxyoxalate chemiluminescence in the chemiluminescence device instead of traditional physical light source, while the photocurrent is achieved by H2O2-induced BiVO4-rGO photovoltaic material under the exciting light irradiation in this work. The as-produced photocurrent is read on a portable digital multimeter in place of conventional electrochemical workstation. In the presence of target PSA, the sandwich-type immunoreaction is carried out into monoclonal anti-PSA (mAb1) capture antibody-coated microplate using GOx and polyclonal anti-PSA detection antibody-labeled AuNP (GOx-AuNP-pAb2) (Note: use of AuNP can increase the labeled amount of GOx and enhance the detectable sensitivity according to our previous works.13,27) Accompanying the antigen−antibody reaction, the conjugated GOx molecules in the microplate can catalytically oxidize glucose to produce hydrogen peroxide. Upon the oxidate introduction into the PEC detection cell, the resulting H2O2 can trigger the photocurrent generation as described above. The as-produced photocurrent indirectly depends on concentration of target PSA in the sample. To achieve a high sensitivity for the development of PEC immunoassay, the successful preparation of the BiVO4-rGO photovoltaic material is very important. Initially, we used scanning electron microscopy (SEM) to characterize the as-synthesized BiVO4 and BiVO4-rGO. The as-prepared BiVO4 nanostructures exhibited blocky morphology with smooth surface, and the average size was 100−400 nm (Figure 1A). In contrast, the surface became rougher upon the doped graphene oxide in the BiVO4 (Figure 1B). Vaguely, a few of slightly transparent graphene nanosheets could be also observed at BiVO4-rGO. Obviously, the doped rGO with BiVO4 did not cause the change of BiVO4 in the morphology and size distribution (Figure 1A,B). The inset in Figure 1B also exhibited that the doped rGO in the BiVO4-rGO was a transparent and thin layer structure with the folded edge. To further verify the formation of BiVO4-rGO, energy-dispersive X-ray spectroscopy (EDS) was also utilized. The four elements including Bi, V, O, and C appeared in BiVO4-rGO (Figure 1C). Moreover, the atomic ratio value for Bi and V was ∼1.12, which was almost in accordance with the theoretical result. The reduced graphene oxide was grown on the aminated BiVO4 by the electrostatic reaction herein. As shown from inset in Figure 3C, the zeta potentials of graphene oxide and the as-synthesized BiVO4 nanostructures were −59.7 mV (curve a) and −22.1 mV (curve b), respectively, which mainly derived from the dissociated − OH/−COOH. When BiVO4 was modified with the (3aminopropyl) triethoxysilane, however, a positive potential of +9.5 mV was acquired (curve c), which provided a precondition 2961

DOI: 10.1021/acs.analchem.6b00262 Anal. Chem. 2016, 88, 2958−2966

Article

Analytical Chemistry

(curve d) alone. However, introduction of rGO enhanced the photocurrents of BiVO4 in the absence of H2O2. Significantly, the photocurrents of BiVO4-rGO were more stable than those of BiVO4 under repeated irradiation. Upon addition of H2O2 into pH 8.5 PBS, the photocurrents of BiVO4-rGO and BiVO4 increased thereupon (curves a−b). Relative to curves c−d, introduction of H2O2 could cause a 10-fold signal increase of using BiVO4-rGO (curve a vs curve c), while the signal only increased to 6-fold of using BiVO4 (curve b vs curve d). Such a signal enhancement mainly originated from the synergistic effect between BiVO4 and rGO. The possible mechanism of the photogenerated electron can be depicted as follows:

chemical impedance spectroscopy (EIS) to monitor BiVO4rGO-modified glassy carbon electrode (GCE) [Note: for comparison, graphene oxide (GO) and BiVO4 alone were immobilized on the GCE by using the same method]. Figure 2A gives the corresponding Nyquist diagrams (inset: equivalent

BiVO4 + hv → BiVO4 (h+ + e−)

(1)

H 2O2 + BiVO4 (h+) → O2 + H 2O + BiVO4

(2)

BiVO4 (e−) + graphene → BiVO4 + graphene(e−)

(3)

Narrow gap BiVO4-rGO generated the electron hole pairs irradiated by visible light (eq 1). In the presence of H2O2, the adsorbed H2O2 acting as a hole scavenger would replace the water to effectively consume the hole of semiconductors (eq 2). Because rGO worked as a good conductive mediator and the supporting material to contact with BiVO4, the photogenerated electrons on the BiVO4 surface transferred to rGO instantly via a percolation mechanism (eq 3),58 and moved to the electrode. Figure 2D gives typical I−V (current vs potential) curves of BiVO4-rGO-modified GCE in the absence and presence of H2O2 and light source irradiation. An anodic photocurrent was achieved within the applied potential, while the current increased with the potential shift from negative to positive. The reason might be most likely as a consequence of the fact that the separation of photogenerated carriers and charge transport of BiVO4-rGO were more efficient at the electrode subjecting positive potentials.59 Also, clearly the currents in the presence of H2O2 (curves a−b) were stronger than those in the absence of H2O2 (curves c−d). The photocurrents of using BiVO4-rGO increased with the increasing H2O2 concentration and allowed the detection of H2O2 as low as 0.1 μM at 0 V (Figure S2), which facilitated the construction of BiVO4-rGObased PEC sensing system. Characteristics of PO-CL Self-Illuminated System with BiVO4-rGO-Based PEC Reaction. As the exciting light source of BiVO4-rGO-based PEC reaction, the high-efficient selfilluminated system would be preferable. As shown from curves a−b in Figure 3A, the bis(2,4,5- trichloro-6- carbopentoxyphenyl) oxalate (CPPO) and 10-diphenylanthracene (DPA) system exhibited the same maximum emission wavelength at 434 nm with the different emission intensities toward variousconcentration H2O2. Using these PO-CL-based self-illuminated systems as light source, we monitored the photocurrents of BiVO4-rGO on an electrochemical workstation, and the photocurrents increased with the increasing H2O2 concentration (Figure 3A, inset, curves a−b). This should attribute to the fact that the high-concentration H2O2 enhanced the chemiluminescence intensity, thereby increasing the corresponding photocurrents. Figure 3B showed the photocurrent changes toward various concentration of H2O2 irradiated by this CL light source with a constant intensity (Figure 3C). For comparison, another type of fluorescent agent, 9,10-bis(phenylethynyl) anthracene (BPEA) (λEM = 521 nm) was studied by using the same assay mode. The similar results with the CPPO−DPA system could be found in the chemilumi-

Figure 2. (A) Nyquist diagrams EIS of (a) GCE, (b) BiVO4-rGO/ GCE, (c) GO/GCE, and (d) BiVO4/GCE (inset: equivalent circuit; Rs: electrolyte resistance, Cdl: lipid bilayer capacitance, Rct: charge transfer resistance, Zw: Warburg element); (B) Fluorescence spectra of BiVO4 and BiVO4-rGO; (C) Photocurrents of (a,c) BiVO4-rGO and (b,d) BiVO4 in the (a,b) presence and (c,d) absence of H2O2; and (D) I−V curves of BiVO4-rGO-modified GCE in the (a,b) presence and (c,d) absence of H2O2 with (a,d) or without (c,b) light source irradiation (note: PEC measurements were performed with a threeelectrode system and a 300 W xenon lamp).

circuit). The parameter (Rct) represents the charge transfer resistance of redox couple and reflects the changes of the electrode surface, which can be read from the semicircle diameter at low frequency.56 Obviously, the resistance of using BiVO4-rGO/GCE (plots b) was lower than those of GO/GCE (plots c) and BiVO4/GCE (plots d), suggesting that the asprepared BiVO4-rGO could accelerate the electron communication between the solution and the base electrode. Vice versa, we could speculate that the photogenerated electron by BiVO4 could transfer to the electrode through reduced graphene oxide. To further demonstrate the above-mentioned hypothesis on the photogenerated electron transfer, fluorescence spectra of BiVO4 and BiVO4-rGO were monitored under the same conditions (λEX = 378 nm) (Figure 2B). Both BiVO4 and BiVO4-rGO exhibited the emission at ∼529 nm, originating from the excited electrons returning to ground state and recombined with the holes. Favorably, the intensity of BiVO4rGO was lower than that of BiVO4. The quenching effect of rGO was ascribed to the excited electron transferring from conduction band of BiVO4 to rGO because rGO served as an excellent electron acceptor and mediator,57 thus greatly inhibiting recombination of excited electron and hole. On the basis of the strong light absorption and efficient separation of electron−hole pairs, we reasonably speculated that BiVO4-rGO could significantly enhance photocurrent response and stability under the visible light irradiation, which could be directly verified by the photocurrent measurement. As shown from Figure 2C, the relatively low photocurrents were achieved at BiVO4-rGO (curve c) and BiVO4 2962

DOI: 10.1021/acs.analchem.6b00262 Anal. Chem. 2016, 88, 2958−2966

Article

Analytical Chemistry

Figure 4. (A,B) Effects of (A-a) GOx catalytic time, (A-b) the antigen−antibody reaction time, (B-a) pH of GOx catalytic solution, and (B-b) pH of PBS in the PEC detection cell on the photocurrent of self-illuminated PEC immunoassay (0.5 ng mL−1 PSA used in the case), (C) calibration curves of self-illuminated PEC immunoassay toward various-concentration PSA standards in the (a) presence and (b) absence of external capacitor, and (D) the specificity of PEC immunoassay against of CEA (10 ng mL−1), CA 125 (10 U mL−1), CA 15−3 (10 U mL−1) and AFP (10 ng mL−1).

Figure 3. (A) CL spectra of (a,b) CPPO−DPA system and (c,d) CPPO-BPEA system in the presence of (a,c) high- or (b,d) lowconcentration H2O2 (insets: the photocurrents of BiVO4-rGO irradiated by different CL sources: (a,b) CPPO−DPA-based selfilluminated system and (c,d) CPPO-BPEA-based self-illuminated system, and the corresponding photographs; (B) photocurrent of BiVO4-rGO relative to different-concentration H2O2 under constant intensity CL irradiation; (C) photograph of PO-based self-illuminated system; and (D) photocurrents of BiVO4-rGO-modified GCE for repeated determination of same-concentration H2O2 under the same conditions.

photocurrent (note: To avoid confusion, the incubation times of mAb1 with target PSA were paralleled with those of mAb1PSA with GOx-AuNP-pAb2). As seen from Figure 4A, the optimal photocurrents were obtained after 12 min for GOxcatalyzed glucose to produce H2O2 (curve a) and 40 min for immunoreaction (curve b). To save the assay time, 12 and 40 min were utilized for H2O2 generation and the antigen− antibody reaction. Apart from the above-mentioned reaction time, the pH of glucose solution for GOx catalytic reaction also affects the H2O2 generation because of pH-interfered GOx bioactivity. A high or low pH would lower the catalytic activity of GOx. As shown from curve a in Figure 4B, a maximum photocurrent was acquired at pH 6.0 PBS containing 4 mM glucose. By the same token, we also found that pH of PBS in the PEC detection cell (upper solution) influenced the photocurrent of BiVO4-rGO. Results indicated that the photocurrent increased with the increasing pH value, and the strong photocurrents were present at pH ≥ 8.5. The reason might be the fact that the alkaline solution is more beneficial to nucleophilic reaction under light irradiation60 and that the acidic solution could dissolve the chitosan membrane to some extent. However, a too basic solution would damage the aggregation of BiVO4-rGO. Considering these concerns, pH 6.0 PBS and pH 8.5 PBS were used for enzymatic oxidization and PEC generation, respectively, throughout this work. Analytical Performance of Self-Illuminated PEC Immunoassay toward Target PSA. Under optimal conditions, the sensitivity and detectable ability of self-illuminated PEC immunoassay were evaluated by assaying differentconcentration PSA standards. As shown in Figure 4C, the transient photocurrent obtained by digital multimeter increased with the increasing target PSA in the sample. A good linear dependence between the photocurrent and the logarithm of PSA concentration was achieved within the dynamic working range from 10 pg mL −1 to 80 ng mL−1 in the presence of the

nescence characteristics toward H2O2, as seen from curves c−d in Figure 3A. However, the photocurrents by using CPPOBPEA were not almost changed regardless of the presence of high- or low-level H2O2 (Figure 3A, inset, curves c−d). The results indicated that (i) peroxyoxalate self-illuminated system could be used as the chemiluminescence light source to excite BiVO4-rGO for the photocurrent generation and that (ii) peroxyoxalate self-illuminated system with BiVO4-rGO could be utilized for quantitative monitoring of H2O2 concentration. After each run, two electrodes with the rubber lid were washed with distilled water and used for next PEC measurement. Meanwhile, the used PEC detection cell was removed and the new one was inserted in the pedestal. To verify that the BiVO4-rGO-modified GCE could be repeatedly utilized for PEC measurements, the photocurrents of the PEC detection device were determined under the same conditions before and after addition of the same-concentration H2O2. As indicated from Figure 3D, the background/response photocurrents were stable, and the relative standard deviation (RSD) was below 10% (i.e., 4.7% for background signals vs 7.8% for response signals, n = 7). Thus, the BiVO4-rGO-modified GCE could be employed for repeated usage during PEC measurements. Optimization of Experimental Conditions. To acquire an optimal analytical performance for immunoassay development, some possible factors influencing the immunoreaction and photocurrent generation should be studied. In this case, 0.5 ng mL−1 PSA was used as an example. As a split-type detection protocol, we first optimized the factors affecting the immunoreaction, e.g., GOx catalytic time toward glucose and incubation time for the antigen−antibody reaction (Figure 4A). Herein, the photocurrent was generated via enzymatic oxidate (H2O2)triggered PEC reaction after the antigen- antibody reaction. Hence, the time for GOx-catalyzed glucose to produce H2O2 and immunoreaction time would directly affect the following 2963

DOI: 10.1021/acs.analchem.6b00262 Anal. Chem. 2016, 88, 2958−2966

Article

Analytical Chemistry

Table 1. Method Accuracy and Recovery Evaluation of the Developed PEC Immunoassay for (Undiluted or Diluted) Human Serum Samples with the Referenced PSA ELISA kit method accuracy; conc. [mean ± SD (RSD), ng mL−1, n = 3]; t-test

recovery evaluation

sample no.

PEC immunoassay

PSA ELISA kit

texpb

added (ng mL−1)c

found (ng mL−1)

recovery (%)d

1 2 3 4 5 6 7 8 9 10

34.5 ± 3.2 (9.3%) 3.25 ± 0.29 (8.9%) 8.57 ± 0.72 (8.4%) 16.61 ± 1.21 (7.3%) 21.82 ± 2.01 (9.2%) 0.37 ± 0.04 (10.8%) 0.082 ± 0.007 (8.5%) 1.32 ± 0.11 (8.3%) 0.98 ± 0.07 (7.2%) 2.31 ± 0.13 (5.6%)

36.4 ± 2.1 4.02 ± 0.39 9.24 ± 0.87 15.33 ± 1.11 20.46 ± 1.67 0.41 ± 0.03 0.075 ± 0.006 1.43 ± 0.09 1.02 ± 0.03 2.16 ± 0.19

0.86 2.74 1.03 1.35 0.91 1.39 1.32 1.34 0.91 1.13

0.5 3.0 10.0 15.0 20.0 0.5 3.0 10.0 15.0 20.0

35.1 ± 3.1 6.36 ± 0.56 18.36 ± 1.23 31.98 ± 2.78 40.87 ± 3.96 0.92 ± 0.08 3.11 ± 0.24 11.2 ± 1.02 16.3 ± 0.15 24.6 ± 1.79

120 103.7 97.9 102.5 95.3 110 100.9 98.8 102.2 111.5

a

a

Sample nos. 1−5 were determined without dilution, and sample nos. 6−10 were achieved by dilution of sample no. 1 to different concentrations. The texp values were obtained by using a t-test statistical analysis between two methods. cPSA standards with different concentrations were added into the corresponding initial human serum samples, respectively. dThe recovery was calculated by the value (The found concentration subtracts that of PEC immunoassay) relative to the added PSA level. b

capacitor. The linear regression equation was i (nA) = 228.99 × lg C[PSA] − 213.23 (pg mL−1, R2 = 0.989, n = 7) with a limit of detection (LOD) of 3.0 pg mL−1 (estimated from the expression of 3S/K, where S is the standard deviation for 11 determinations of blank solution and K is the slope of the calibration plot) (Figure 4C-a). Each data point represents the average value obtained from three different measurements. The maximum relative standard deviation (RSD) was 7.3%, indicating a good reproducibility and precision of our strategy. For comparison, we investigated the analytical properties of DMM-based PEC immunoassay without the capacitor (Figure 4C-b). The linear range, regression equation and LOD were 0.5−150 ng mL−1, i (nA) = 17.94 × lgC[PSA] − 36.06 (pg mL−1, R2 = 0.994, n = 6) and 0.1 ng mL−1 PSA, respectively. Although the PEC immunoassay without the capacitor could be used for the detection of high-concentration PSA, the sensitivity (i.e., the slope of the regression equation) was heavily lower than that of with the capacitor. Such a high sensitivity was suitable for distinguishable too-close PSA concentrations, especially at a low-concentration target PSA. To monitor the specificity of DMM-based PEC immunoassay, some possible biomarkers existing in normal human serum, e.g., carcinoembryonic antigen (CEA), alpha fetoprotein (AFP), cancer antigen 125 (CA 125) and cancer antigen 15−3 (CA 15−3), were studied with and without target PSA, respectively. As indicated from Figure 4D, only the presence of target PSA could cause a stronger photocurrent in comparison with background signal, regardless of mixture with target analyte or alone. Such a high selectivity might be attributed to the specific antigen−antibody reaction. Analysis of Real Human Serum Samples. To investigate the accuracy of the newly developed PEC immunoassay with a commercialized PSA ELISA kit, we collected five human serum samples from the local hospital. The evaluation was carried out by using a t-test statistical analysis between two methods and the recovery experiments (Table 1). First, these samples were detected using our developed PEC immunoassay and the PSA ELISA kit, respectively. As shown in Table 1, all texp values in these cases were less than tcrit (tcrit[4,0.05] = 2.77), indicating no significant differences at the 0.05 significance level between two method for the detection of PSA real samples. Further, we also used the standard addition method to evaluate the feasibility and applicability of our strategy. Experimental results indicated

that the recovery was 95.3−120%. Therefore, the DMM-based PEC immunoassay with peroxyoxalate self-illuminated system could be used as an optimal scheme for quantitative determination of PSA concentration in the clinical serum samples.

■

CONCLUSIONS

In summary, we successfully demonstrated an innovative and portable PEC immunoassay coupling with a spilt-type detection mode and a peroxyoxalate self-illuminated system via enzymeoxidized product. Experimental results indicated that the assynthesized BiVO4- rGO nanocomposites could exhibit good photovoltaic properties, sensitive to H2O2 and excellent stability under the visible light irradiation. Meanwhile, the portable split-type self-illuminated PEC immunoassay could be readily carried out by using low-cost microplate, highly efficient peroxyoxalate chemiluminescence system and user-friendly digital multimeter. Compared with conventional PEC immunoassays, the assay used digital multimeter as the readout, and employed peroxyoxalate chemiluminescence to replace the physical light source. To the best of our knowledge, this is the first work by using peroxyoxalate chemiluminescence as the exciting light source for the development of split-type PEC immunoassay. Although the detection limit of the present work (3.0 pg mL−1) was higher than that of our previous work (0.32 pg mL−1),13 our designed DMM-based PEC immunoassay could eliminate the damage of biomolecules by the physical light source, maintain the bioactivity of the proteins, and adequately promote the efficiency of the photovoltaic materials. Because of its simple operation, low cost, wide availability of digital multimeter without the needs of computer, an electrochemical workstation, and a physical light source, the methodology demonstrated in this work can be utilized by the public for quantitative monitoring of low-abundant proteins or biomarkers by controlling the corresponding antibodies, thereby representing a versatile detection protocol. Nevertheless, only one disadvantage of our strategy is that the product through the immunoreaction was artificially injected into the PEC detection cell. This could potentially be improved by coupling with a magnetic bead-based immunoassay and the microfluidic device. 2964

DOI: 10.1021/acs.analchem.6b00262 Anal. Chem. 2016, 88, 2958−2966

Article

Analytical Chemistry

■

(20) Haddour, N.; Cosnier, S.; Gondran, C. Chem. Commun. 2004, 40, 2472−2473. (21) Haddour, N.; Chauvin, J.; Gondran, C.; Cosnier, S. J. Am. Chem. Soc. 2006, 128, 9693−9698. (22) Wang, G.; Yu, P.; Xu, J.; Chen, H. J. Phys. Chem. C 2009, 113, 11142−11148. (23) Kang, Q.; Yang, L.; Chen, Y.; Luo, S.; Wen, L.; Cai, Q.; Yao, S. Anal. Chem. 2010, 82, 9749−9754. (24) Zhao, W.; Shan, S.; Ma, Z.; Wan, L.; Xu, J.; Chen, H. Anal. Chem. 2013, 85, 11686−11690. (25) Zang, Y.; Lei, J.; Zhang, L.; Ju, H. Anal. Chem. 2014, 86, 12362− 12368. (26) Han, Q.; Wang, K.; Xu, L.; Yan, X.; Zhang, K.; Chen, X.; Wang, Q.; Zhang, L.; Pei, R. Analyst 2015, 140, 4143−4147. (27) Shu, J.; Qiu, Z.; Zhuang, J.; Xu, M.; Tang, D. ACS Appl. Mater. Interfaces 2015, 7, 23812−23818. (28) Tang, D.; Zhang, B.; Liu, B.; Chen, G.; Lu, M. Biosens. Bioelectron. 2014, 55, 255−258. (29) Pacardo, D.; Neupane, B.; Wang, G.; Gu, Z.; Walker, G.; Ligler, F. Anal. Bioanal. Chem. 2015, 407, 719−725. (30) Victoria, L.; Arenas, A.; Ortega, M.; Barzanallana, R.; Ibanez, J. Anal. Chim. Acta 1990, 229, 153−155. (31) Yang, M.; Sun, S.; Bruck, H.; Kostov, Y.; Rasooly, A. Lab Chip 2010, 10, 2534−2540. (32) Liu, H.; Xiang, y.; Lu, Y.; Crooks, R. Angew. Chem. 2012, 124, 7031−7034. (33) Lou, B.; Zhou, Z.; Du, Y.; Dong, S. Electrochem. Commun. 2015, 59, 64−67. (34) Han, Y.; Park, Y.; Chun, H.; Yoon, H. Sens. Actuators, B 2015, 220, 233−242. (35) Duk Han, Y.; Jin Chun, H.; Yoon, H. Biosens. Bioelectron. 2014, 59, 259−268. (36) Wang, J.; Liu, Z.; Hu, C.; Hu, S. Anal. Chem. 2015, 87, 9368− 9375. (37) Dai, H.; Zhang, S.; Gong, L.; Li, Y.; Xu, G.; Lin, Y.; Hong, Z. Biosens. Bioelectron. 2015, 72, 18−24. (38) Delaney, J.; Hogan, C.; Tian, J.; Shen, W. Anal. Chem. 2011, 83, 1300−1306. (39) Shi, C.; Shan, X.; Pan, Z.; Xu, J.; Lu, C.; Bao, N.; Gu, H. Anal. Chem. 2012, 84, 3033−3038. (40) Wang, S.; Dai, W.; Ge, L.; Yan, M.; Yu, J.; Song, X.; Ge, S.; Huang, J. Chem. Commun. 2012, 48, 9971−9973. (41) Zhao, W.; Wang, J.; Zhu, Y.; Xu, J.; Chen, H. Anal. Chem. 2015, 87, 9520−9531. (42) Adam, W.; Kazakov, D.; Kazakov, V. Chem. Rev. 2005, 105, 3371−3387. (43) Samadi-Maybodi, A.; Akhoondi, R. Luminescence 2015, 30, 775− 779. (44) Ge, L.; Wang, P.; Ge, S.; Li, N.; Yu, J.; Yan, M.; Huang, J. Anal. Chem. 2013, 85, 3961−3970. (45) Rauhut, M.; Bollyky, L.; Roberts, B.; Loy, M.; Whitman, R.; Iannotta, A.; Semsel, A.; Clarke, R. J. Am. Chem. Soc. 1967, 89, 6515− 6522. (46) Biparva, P.; Abedirad, S.; Kazemi, S. Spectrochim. Acta, Part A 2015, 145, 454−460. (47) Chaichi, M.; Azizi, S.; Alijanpour, O.; Heidarpour, M.; Qandalee, M. J. Lumin. 2013, 138, 65−71. (48) Tu, W.; Wang, W.; Lei, J.; Deng, S.; Ju, H. Chem. Commun. 2012, 48, 6535−6537. (49) Wang, Y.; Wang, W.; Mao, H.; Lu, Y.; Lu, J.; Huang, J.; Ye, Z.; Lu, B. ACS Appl. Mater. Interfaces 2014, 6, 12698−12706. (50) Ng, Y.; Iwase, A.; Kudo, A.; Amal, R. J. Phys. Chem. Lett. 2010, 1, 2607−2612. (51) Yu, J.; Zhang, Y.; Kudo, A. J. Solid State Chem. 2009, 182, 223− 228. (52) Yin, C.; Zhu, S.; Chen, Z.; Zhang, W.; Gu, J.; Zhang, D. J. Mater. Chem. A 2013, 1, 8367−8378. (53) Tokunaga, S.; Kato, H.; Kudo, A. Chem. Mater. 2001, 13, 4624− 4628.

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.analchem.6b00262. Experimental section, DLS curve of AuNP and GOxAuNP-pAb2 (Figure S1) and photocurrents of BiVO4rGO- modified GCE toward various-concentration H2O2 (Figure S2) (PDF)

■

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. Phone: +86-591-2286 6125. Fax: +86-591-2286 6135. *E-mail: [email protected]. Notes

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was financially supported by the National Natural Science Foundation of China (Grant nos.: 41176079 and 21475025), the National Science Foundation of Fujian Province (Grant no.: 2014J07001), and the Program for Changjiang Scholars and Innovative Research Team in University (Grant no.: IRT15R11).

■

REFERENCES

(1) Li, H.; Li, J.; Yang, Z.; Xu, Q.; Hu, X. Anal. Chem. 2011, 83, 5290−5295. (2) Freeman, R.; Girsh, J.; Willner, I. ACS Appl. Mater. Interfaces 2013, 5, 2815−2834. (3) Zeng, X.; Tu, W.; Li, J.; Bao, J.; Dai, Z. ACS Appl. Mater. Interfaces 2014, 6, 16197−16203. (4) Wang, W.; Hao, Q.; Wang, W.; Bao, L.; Lei, J.; Wang, Q.; Ju, H. Nanoscale 2014, 6, 2710−2717. (5) Chen, D.; Feng, H.; Li, J. Chem. Rev. 2012, 112, 6027−6053. (6) Shu, J.; Qiu, Z.; Wei, Q.; Zhuang, J.; Tang, D. Sci. Rep. 2015, 5, 15113. (7) Yang, L.; Guan, G.; Wang, S.; Zhang, Z. J. Phys. Chem. C 2012, 116, 3356−3362. (8) Long, Y.; Bao, L.; Zhao, J.; Zhang, Z.; Pang, D. Anal. Chem. 2014, 86, 7224−7228. (9) Wen, G.; Wen, X.; Choi, M.; Shuang, S. Sens. Actuators, B 2015, 221, 1449−1454. (10) Han, D.; Ma, Z.; Zhao, W.; Xu, J.; Chen, H. Electrochem. Commun. 2013, 35, 38−41. (11) Li, Y.; Ma, M.; Zhu, J. Anal. Chem. 2012, 84, 10492−10499. (12) Zhao, W.; Xu, J.; Chen, H. Chem. Rev. 2014, 114, 7421−7441. (13) Zhuang, J.; Tang, D.; Lai, W.; Xu, M.; Tang, D. Anal. Chem. 2015, 87, 9473−9480. (14) Yu, X.; Wang, Y.; Chen, X.; Wu, K.; Chen, D.; Ma, M.; Huang, Z.; Wu, W.; Li, C. Anal. Chem. 2015, 87, 4237−4244. (15) Zhao, W.; Xu, J.; Chen, H. Chem. Soc. Rev. 2015, 44, 729−741. (16) Akhavan-Tafti, H.; Binger, D.; Blackwood, J.; Chen, J.; Creager, R.; de Silva, R.; Eickholt, R.; Gaibor, J.; Handley, R.; Kapsner, K.; Lopac, S.; Mazelis, M.; McLernon, T.; Mendoza, J.; Odegaard, B.; Reddy, S.; Salvati, M.; Schoenfelner, B.; Shapir, N.; Shelly, K.; Todtleben, J.; Wang, G.; Xie, W. J. Am. Chem. Soc. 2013, 135, 4191− 4194. (17) Tang, D.; Lin, Y.; Zhou, Q.; Lin, Y.; Li, P.; Niessner, R.; Knopp, D. Anal. Chem. 2014, 86, 11451−11458. (18) Wild, D. The Immunoassay Handbook, 3rd ed.; Elsevier, Amsterdam, 2005; p 194. (19) Cohen, C.; Weber, S. Anal. Chem. 1993, 65, 169−175. 2965

DOI: 10.1021/acs.analchem.6b00262 Anal. Chem. 2016, 88, 2958−2966

Article

Analytical Chemistry (54) Yin, W.; Wang, W.; Zhou, L.; Sun, S.; Zhang, L. J. Hazard. Mater. 2010, 173, 194−199. (55) Zhang, Y.; Tang, Z.; Fu, X.; Xu, Y. ACS Nano 2010, 4, 7303− 7314. (56) Wan, Y.; Lin, Z.; Zhang, D.; Wang, Y.; Hou, B. Biosens. Bioelectron. 2011, 26, 1959−1964. (57) Lightcap, I.; Kosel, T.; Kamat, P. Nano Lett. 2010, 10, 577−583. (58) Fu, Y.; Sun, X.; Wang, X. Mater. Chem. Phys. 2011, 131, 325− 330. (59) Kongkanand, A.; Tvrdy, K.; Takechi, K.; Kuno, M.; Kamat, P. J. Am. Chem. Soc. 2008, 130, 4007−4015. (60) Long, M.; Cai, W. Chin. J. Catal. 2008, 29, 881−883.

2966

DOI: 10.1021/acs.analchem.6b00262 Anal. Chem. 2016, 88, 2958−2966