A reusable sandwiched electrochemical piezoelectric immunosensor has been developed for aflatoxin B1 (AFB1) detection using gold coated iron oxide core-shell (Au-Fe3O4) nanostructure. The monoclonal anti-aflatoxin antibody (aAFB1) was immobilized on self-assembled monolayer of 4-aminothiophenol on gold coated quartz crystal to fabricate immunoelectrode (BSA/aAFB1/4-ATP/Au). In addition, secondary rabbit-immunoglobulin antibodies (r-IgGs) functionalized with Au-Fe3O4 NPs via cysteamine (r-IgG-Cys-Au-Fe3O4) were allowed to interact with AFB1. Both competitive and noncompetitive strategies were employed and a competition between coated AFB1 and free AFB1 was carried out. The competitive mode shows higher linear range (0.05 to 5 ng mL−1) than the noncompetitive one (0.5 to 5 ng mL−1), high sensitivity 335.7 µA ng−1 mL cm−2, and LOD 0.07 ng mL−1. The fabricated immunosensor has been tested using cereal samples spiked with different concentrations of AFB1. The developed competitive immunoelectrode displays good reproducibility, and storage stability and regenerated with negligible loss in activity through removal of the r-IgG-Cys-Au-Fe3O4 conjugate using a strong external magnet.

1. Introduction

Aflatoxin B1 (AFB1), which is a low molecular mass compound mainly produced by the moulds Aspergillus flavus and Aspergillus parasiticus, can contaminate several important crops (e.g., corn, sorghum, peanuts, fruits, dried fruits, cocoa, and spices) under favorable environmental conditions [1, 2]. Owing to high toxicity and carcinogenicity, AFB1 is of major concern for food producers, the food processing industry, and consumers [3]. The European Commission has set the maximum permissible residue levels for AFB1 in corn products ready for retail sale at 2 μg kg−1 [47]. Compared to conventional techniques [810], immunosensors are an interesting alternative that can be used to quantify and detect selectively toxin molecules [1114]. Some of the biosensing techniques used for AFB1 detection include electrochemical biosensors [15, 16], surface plasmon resonance [17, 18], fluorescent biosensors [19, 20], and quartz crystal (piezoelectric) microbalance based sensors [21, 22]. Among the signal detection techniques, quartz crystal microbalance (QCM) based detection systems are considered to be the most promising for immunointeraction owing to their affordable cost and real-time and label-free compatibility with miniaturization, portability, and high sensitivity [2325].

The immunosensors which employ monoclonal antibody, synthesized from animal source as a receptor, are fast and efficient technique no doubt but are expensive and used for one time only. Therefore it will be advantageous to develop reusable immunosensor which can be regenerated mechanically without any chemical as the chemicals affect biological activity of the receptor antibody [26, 27]. In 2009, Wang and Gan have used magnetic core-shell Fe3O4/SiO2 composite nanoparticles to regenerate the QCM crystal. In this regard, magnetic nanoparticles (MNP) can be considered as one of the potential tools to regenerate immunosensor using strong external magnet [28]. The MNP is commonly used in the coated form to gain biocompatibility and stability [29, 30]. Since AFB1 is a low molecular weight (Mwt. 312) molecule, competitive mode of detection could provide the desired detection limit and sensitivity [31, 32]. The competition occurs between coated toxin (toxin-protein conjugate) and free toxin. The few binding sides of coated toxin (antigen-protein complex) are partially blocked through the toxin-protein binding. Therefore, the secondary antibody captures free toxin leaving heavy toxin-protein complex.

In this study, a sandwich type electrochemical quartz crystal microbalance (EQCM) based reusable immunosensor (BSA/aAFB1/4-ATP/Au) was fabricated using self-assembled monolayer of 4-aminothiophenol on gold coated quartz crystal. The gold coated magneto nanoparticles attached to the secondary antibody were used as a signal enhancing agent and for regeneration of the immunoelectrode through external magnet. Both competitive and noncompetitive strategy are studied. Here, competition occurs between free AFB1 and coated AFB1 (with no protein bonded). Interestingly, we have observed that competition mode offers wider linearity, lower detection range, and higher sensitivity. The constructed immunosensor can be used for estimation of aflatoxin B1 from sample and regenerated with negligible loss of activity using a strong external magnet.

2. Experimental Methods

2.1. Chemicals and Reagents

Monoclonal anti-aflatoxin B1 (aAFB1) antibodies, aflatoxin B1 (AFB1), bovine serum albumin (BSA), polyclonal IgG antibodies from rabbit (r-IgG), 4-aminothiophenol (4-ATP), N-ethyl-N-(3-dimethylaminopropyl) carbodiimide (EDC), N-hydroxysuccinimide (NHS), ferrous chloride hexahydrate (FeCl2·6H2O), ferric chloride tetrahydrate (FeCl3·4H2O), sodium hydroxide (NaOH), and chloroauric acid (HAuCl4·H2O) were procured from Sigma-Aldrich. All reagents were of analytical grade and used without further purification, and deionized water (18 MΩ cm) was used for the preparation of solutions. The gold coated (diameter: 6.7 mm) quartz resonator (AT cut quartz crystal, 13.7 mm dia, 6 MHz) was procured from Autolab, Netherlands.

2.2. Solution Preparation

Anti-AFB1 antibody (1 mg mL−1) solution was prepared in 50 mM phosphate buffer (PBS), 50 mM, pH 7.4, and a 0.15 M NaN3 was used as a preservative. r-IgG antibody (2 mg mL−1) solution was prepared in 50 mM PBS (pH 7.4). The stock solution of AFB1 was prepared in PBS (50 mM, pH 7.4) with 10% methanol and diluted in different working concentrations and stored at −20°C. A solution of bovine serum albumin (BSA, 1 mg mL−1) was prepared in PBS (50 mM, pH 7.0) and used as blocking agent for nonspecific binding sites.

2.3. Pretreatment of Quartz Crystals

The quartz crystals were immersed in 1 M NaOH for 5 min and 1 M HCl for 2 min in a sequence. Then, freshly prepared piranha solution 1 : 3 (30% v/v) was dropped on the gold surface for 2 min, with special care to avoid the contamination of the electrode leads. The quartz crystals were rinsed twice with deionized water followed by ethanol and dried in a stream of nitrogen after each pretreatment and then the initial resonance frequency () was recorded. After the above cleaning procedure, the quartz crystal was ready for surface modification and antibody immobilization.

2.4. Synthesis of Fe3O4 and Au Coated Fe3O4

Fe3O4 NPs were synthesized simply by the coprecipitation method reported earlier [33] with some modification. Solution of 0.07 M FeCl3·6H2O and 0.04 M FeCl2·4H2O (2 : 1, w/w ratio) was dissolved in 25 mL deionized water and then this mixture was added dropwise to the 100 mL solution of 0.15 mM NaOH with stirring under N2 atmosphere at room temperature. A black precipitate of Fe3O4 NPs was obtained. The black precipitate of Fe3O4 NPs thus obtained was dissolved in 20 mL citrate buffer (1.6 gm citric acid and 0.8 gm trisodium citrate) to stabilize ferrofluid in solution at a pH around 6.3.

The Au-Fe3O4 core-shell NPs were prepared using 3 mL of the synthesized colloidal Fe3O4 nanosuspension (0.1 M), boiled with 25 mL of ultrapure water under vigorous stirring condition. Then 0.2 mM HAuCl4 was added, followed by the addition of 10 mM trisodium citrate, and the reaction mixture was kept boiling and stirring for 15 min till the color of the solution turned red from black. The gold coated Fe3O4 NPs (Au- Fe3O4 NPs) solution was allowed to cool and stored in a dark glass bottle at 4°C before use.

2.5. Synthesis of r-IgG-Cys-Au-Fe3O4

Synthesized Au-Fe3O4 nanosuspension was treated with 10−3 M cysteamine and HCl in 1 : 15 volume ratio for 12 h at 25°C. Cysteamine functionalized Au-Fe3O4 NPs (Cys-Au-Fe3O4) were separated and purified by centrifugation at 10,000 rpm for 10 min. The purification and centrifugation process were repeated 4-5 times for removing nonbonded cysteamine. Then cysteamine functionalized Au-Fe3O4 core-shell NPs were redispersed in PB (50 mM, pH 7.0) solution. Cysteamine forms a self-assembled layer on Au-Fe3O4 NPs which provides NH2 group to bind favorably with COOH functional group of the polyclonal r-IgG antibodies during immobilization. The r-IgG antibodies are mixed with Cys-Au-Fe3O4 solution in 1 : 3 (v/v ratio) [34], followed by addition of 0.2 M EDC and 0.05 M NHS for the activation of –COOH group present in antibody. Further to block the nonspecific sites on the r-IgG-Cys-Au-Fe3O4 conjugates, 100 μL BSA (1 mg mL−1) was added and incubated for 2 h at 25°C. The mixture was centrifuged at 10,000 rpm for 10 min and washed for 4-5 times. Finally, the r-IgG-Cys-Au-Fe3O4 conjugate was resuspended in PB (50 mM, pH 7.4) and stored at 4°C until use.

2.6. Fabrication of AFB1/BSA/aAFB1/4-ATP/Au Immunosensor

Pretreated quartz crystal was immersed in 2 mM solution of 4-ATP in ethanol for 24 h at 25°C for SAM formation. However, a uniform and steady 4-ATP film was obtained [35]. The crystal was subsequently washed with ethanol followed by rinsing with water to remove any unbound ATP molecules. 10 μL of the monoclonal anti-aflatoxin B1 (aAFB1) antibody, activated with 0.2 M EDC and 0.05 M NHS for about 2 h, was spread over the electrode and incubated overnight at 4°C for the amide bond formation between aAFB1 and 4-ATP. In this study, optimized concentration of 40 μg mL−1 of aAFB1 was used. The nonspecific sites of fabricated immunoelectrodes were blocked with BSA (1 mg mL−1). These fabricated BSA/aAFB1/4-ATP/Au immunoelectrodes were exposed to saturated concentration of AFB1 (5 ng mL−1) for 35 min at 25°C.

2.7. Pretreatment and Analysis of Cereal Samples

The cereal samples (corn flakes) were spiked after the treatment. Corn flakes samples were crushed to powder using a hand-held blender. 2 g of powdered cereals was added to methanol : water (7 : 3, v/v) solution on a sonication bath for 45 min. The extract was centrifuged for 7 min at 5000 rpm to remove the solids. The supernatants were collected and allowed to evaporate to dryness under nitrogen at 25°C. The residues were resuspended in 5 mL PBS and filtered through 0.45 μm nylon membranes [36]. Finally, extract was spiked with the different concentrations of 0.05, 2, and 5 ng mL−1 of aAFB1.

2.8. Instrumentation

The resonant frequency of quartz crystal and electrochemical studies were monitored by Autolab Potentiostat/Galvanostat Model AUT83945 (PGSTAT302N). The electrochemical quartz crystal cyclic voltammetric (EQCM-CV) studies were carried out in a three-electrode cell using modified quartz crystal as the working electrode, gold wire as the auxiliary electrode, and saturated Ag/AgCl as the reference electrode in PBS (50 mM, pH 7.4, 0.9% NaCl) containing 5 mM [Fe(CN)6]3−/4− as a redox species. The Au-Fe3O4 core-shell magnetic nanoparticles were characterized by scanning electron microscopy (ZEISS EVO-18), vibrating sample magnetometer (VSM) (Microsense, ADE-Model EV9), transmission electron microscopy (JEOL JEM (Model 1200F)), and X-ray diffractometer from Bruker AXS (XRD). The structural and surface morphological characterizations of 4-ATP/Au, aAFB1/4-ATP/Au, BSA/aAFB1/4-ATP/Au, and AFB1/BSA/aAFB1/4-ATP/Au electrode were carried out using Fourier transform infrared spectroscopy (FT-IR, Perkin-Elmer, Model 2000), scanning electron microscopy (ZEISS EVO-18), and Autolab Potentiostat/Galvanostat Model AUT83945 (PGSTAT302N).

3. Results and Discussion

3.1. Characterization of Fe3O4 and Au-Fe3O4 Core-Shell Structure

Figure 1(a) shows the UV-Visible absorption spectrum of pure magnetic Fe3O4 NPs, Au NPs, and Au-Fe3O4 (curve (c)) core-shell NPs. Typical absorption spectra of pure Fe3O4 NPs (curve (a)) exhibit absorption edge at ~340 nm. The absorption peak seen at 324 nm and the sharp absorption maxima at 527 nm (curve (b)) are assigned for pure Au NPs exhibiting strong absorption that is dependent on the size and shape of particles. For spherical nanoparticles, the absorption band maximum generally falls between about 520 and 532 nm [37]. The UV-Visible absorption spectrum of Au-Fe3O4 core-shell (curve (c)) structure shows a broad peak at 532 nm. The shifting of peak position towards longer wavelengths (red shift) and disappearance of peak edge arise due to Fe3O4, indicating the formation of bimetallic core-shell structure with the existence of Fe3O4 as core. Au covers the Fe3O4 NPs surface and provides a broad shifted peak at 532 nm due to inherent surface plasmon resonance property of Au NPs.

The magnetic properties of Fe3O4 NPs and Au-Fe3O4 core-shell structure were analyzed by vibrating sample magnetometer (VSM) at 17 K. Figure 1(b) shows the hysteresis loop measured for the Fe3O4 NPs (curve (a)), Fe3O4 NPs in citrate buffer (curve (b)), and Au-Fe3O4 core-shell structure (curve (c)). The values of saturated magnetization from the hysteresis curve of the pure Fe3O4 NPs and Fe3O4 NPs in buffer were found to be 0.0028 and 0.0085 emu/g, respectively, at 17 K. The saturated magnetization of Fe3O4 NPs dispersed in citrate buffer was increased by ~4 times compared with the precipitated Fe3O4 NPs indicating uniform dispersion of Fe3O4 particles in citrate buffer. In the dispersed form, each NP acts like a tiny magnet, resulting in a higher magnetic moment density than that of precipitated Fe3O4 NPs. The saturated specific magnetization of Au-Fe3O4 core-shell structure decreases to 0.0022 emu/g. This decrease may be due to the fact that the gold is a nonmagnetic material, which could decrease the saturated specific magnetization [38] indicating that the gold was successfully coated on Fe3O4 NPs to form Au-Fe3O4 core shell.

EQCM-CV (Figure 1(c)) of Fe3O4 NPs dispersion and Au-Fe3O4 NPs were studied in PBS buffer (50 mM, pH 7.4, 0.9% NaCl) containing 5 mM [Fe(CN)6]3−/4−. 100 μL of NPs dispersion was added to buffer to conduct CV at a scan rate of 100 mV/s in the potential range of −0.2 to 0.8 V Figure 1(c) (a, b, and c. Curve (a) represents the EQCM-CV of [Fe(CN)6]3−/4− redox system in PBS buffer. The magnitude of the peak current increases after adding Fe3O4 NPs (curve (b)) which further increases on adding Au-Fe3O4 core shell (curve (c)) showing an enhanced electron transform rate through the medium to surface of electrode and confirming the Au is successfully coated onto Fe3O4 NPs.

To confirm the formation of Au-Fe3O4 NPs, EDX analysis has been studied for elemental composition in Fe3O4 and Au-Fe3O4 NPs. Figure 2(a) (images (A) and (B) for Fe3O4 and Au-Fe3O4 NPs) shows the presence of Fe peak at 6.8 keV and absence of Au peak in image (A), while image (B) shows peaks both for Au at 2.4 keV and 9.5 keV and for Fe at 0.58 keV, 6.5 keV, and 7.1 keV, respectively. The weight percentage of these elements, shown as insets of respective images, indicates the presence of corresponding elements.

Figure 2(b) shows the TEM images of Fe3O4 NPs, Fe3O4 NPs in citrate buffer, and Au-Fe3O4 NPs. The average particle size of Fe3O4 NPs, Fe3O4 NPs in citrate buffer, and Au-Fe3O4 NPs was ~8 nm, ~13 nm, and ~19 nm, respectively. Figure 2(b), image (A), shows Fe3O4 NPs overlap each other, while image (B) shows the uniform distribution of Fe3O4 NPs in ionic citrate buffer. It reveals that in buffer solution ionic citrate layer surrounds the Fe3O4 NPs. Image (C) presents the Au-Fe3O4 core-shell NPs having bilayer structure with a dark center surrounded by a lighter layer. The molecular spacing is 0.48 nm for darker part and 0.23 nm (from ImageJ software) for light part of shell. The lattice distances measured for the shell correspond to the known Au lattice parameters for the (111) plane and those measured for the core match well the Fe3O4 lattice parameters for the (311) plane. The presence of these two phases is also confirmed by X-ray diffraction (XRD) analysis. The XRD pattern of the Fe3O4 NPs and Au-Fe3O4 core-shell NPs, demonstrated in Figures 2(c) (a) and 2(c) (b), shows diffraction peaks at 2θ 30.15°, 35.76°, 43.2°, 53.6°, 57.6°, and 62.96° for Fe3O4 which exhibit indexed (220), (311), (400), (422), (511), and (440). For Au-Fe3O4 (graph (b)), the characteristic peaks seen at 38.28°, 44.43°, 59.1°, 64.70°, and 77.81°, marked by their indices (111), (200), (220), (311), and (222), are observed indicating that the Fe3O4 in Au-Fe3O4 NPs resembles pure Fe3O4 with a spinal hexagonal structure [39, 40].

3.2. Characterization of r-IgG-Cys-Au-Fe3O4

The conjugate formation of r-IgG-Cys-Au-Fe3O4 was confirmed by UV-absorption spectroscopy (Figure 3(a)). The absorbance maxima for the pure r-IgG antibody solution appear at 250 nm (curve (a)) whereas the r-IgG-Cys-Au-Fe3O4 conjugate shows two peaks observed at 258 nm for r-IgG and 529 nm for Au-Fe3O4 (curve (b)). Broadening in peaks and slight red shift are also observed in both peaks due to increased size of NPs. These results indicate that the r-IgG antibodies are immobilized onto the surface of the Cys-Au-Fe3O4 core-shell structure. Scheme 1 represents the formation of r-IgG-Cys-Au-Fe3O4 conjugate.

The EQCM-CV (Figure 3(b)) of cysteamine functionalized Au-Fe3O4 NPs and r-IgG- Au-Fe3O4 conjugate corroborates the fabrication of secondary antibody conjugate with Au-Fe3O4 NPs. EQCM-CV was studied in PBS buffer (50 mM, pH 7.4, 0.9% NaCl) containing 5 mM [Fe(CN)6]3−/4− at a scan rate of 100 mV/s in the potential range of −0.2 to 0.8 V Figure 3(b) (a, b, and c. 100 μL of conjugate dispersion was added to the buffer to conduct CV of conjugates. Curve (a) represents the EQCM-CV of Au-Fe3O4 nanoparticles in [Fe(CN)6]3−/4− redox system in PBS buffer. The magnitude of the peak current decreases after functionalization of Au-Fe3O4 nanoparticles with cysteamine (curve (b)); decrease in current occurs due to the insulating nature of cysteamine. The EQCM-CV of r-IgG-Au-Fe3O4 conjugate (curve (c)) in [Fe(CN)6]3−/4− redox system results in increase in current, due to the presence of carboxyl and amine group throughout the IgG antibodies indicating the formation of r-IgG-Cys-Au-Fe3O4 conjugates.

3.3. Characterization of the Immunosensor
3.3.1. Electrochemical Characterization of the Immunoelectrode

EQCM-CV shows both frequency and current changes. Figure 4(a) represents the change in frequency of QCM () after the SAM deposition of 4-ATP on the bare gold crystal surface obtained at a scan rate of 100 mV/s in the potential range of −0.2 to 0.8 V. The cleaned bare quartz crystal was taken as a reference. The frequency changes increase as mass increases on the electrode surface; the continuous increase in mass reveals the successive deposition of electrode.

Table 1 shows the list of mass deposition over the electrode at each layer. The successive deposition of various layers on Au quartz crystal is understood by mass change of 206.50 ng cm−2 for 4-ATP, 268.71 ng cm−2 for aAFB1 antibodies, and 396.196 ng cm−2 for BSA/aAFB1/4-ATP/Au electrode, respectively. After the competition the electrode surface mass increases drastically due to the formation of sandwiched between secondary antibody conjugate and monoclonal antibodies (r-IgG-Cys-Au-Fe3O4/AFB1/BSA/aAFB1/4-ATP/Au) over the electrode (Table 1). These results are supported with EQCM-CV.

EQCM-CV (Figure 4(b)) was conducted in PBS (50 mM, pH 7.4, 0.9% NaCl) containing 5 mM [Fe(CN)6]3−/4− as a redox species at a scan rate of 100 mV/s in the potential range of −0.2 to 0.8 V. Figure 4(b) shows CV of [Fe(CN)6]3−/4− of a bare Au QCM electrode (curve (a)), 4-ATP/Au electrode (curve (b)), aAFB1/4-ATP/Au electrode (curve (c)), and BSA/aAFB1/4-ATP/Au electrode (curve (d)). After the SAM deposition, the magnitude of anodic peak current (2.49 × 10−4 A) decreases (curve (b)), due to the insulating properties of the thin layer of thiol which hinders the electron movement through electrode of the gold surface. The presence of aAFB1 over the surface enhances the peak current up to 6.48 × 10−4 A (curve (c)) due to the presence of polar groups of antibody such as carboxyl and amine moieties. There is a slight decrease in the magnitude of current response (6.17 × 10−4 A) after immobilization of BSA onto the surface of AFB1/4-ATP/Au electrode, indicating enhanced electron-transfer barriers introduced upon assembly of BSA layer. After the competition, the magnitude of current increases (8.31 × 10−4) due to the presence of r-IgG-Cys-Au-Fe3O4 conjugate over the surface of AFB1/BSA/aAFB1/4-ATP/Au immunoelectrode. It reveals that the IgG antibodies interact with AFB1 coated over the surface.

3.3.2. SEM Studies

The surface morphological studies of aAFB1/4-ATP/Au, AFB1/BSA/aAFB1/4-ATP/Au, and r-IgG-Cys-Au-Fe3O4/AFB1/BSA/aAFB1/4-ATP/Au electrodes and immunoelectrodes were examined by scanning electron microscopy (SEM). Figure 4(c) represents the surface morphology of aAFB1/4-ATP/Au electrode (image (A)), showing highly dense globular morphology with bright streaks confirming the immobilization of aAFB1 onto 4-ATP/Au surface. Figure 4(c), image (B), shows the surface morphology of AFB1/BSA/aAFB1/4-ATP/Au electrode surface. The morphological changes in the SEM image after incubation of AFB1 indicate the binding of AFB1 to the aAFB1/4-ATP/Au electrode surface. Image (C) shows the surface morphology of r-IgG-Cys-Au-Fe3O4/AFB1/BSA/aAFB1/4-ATP/Au immunoelectrode and the surface is saturated with r-IgG-Cys-Au-Fe3O4 conjugate. Inset image (image (C)) shows the morphology of r-IgG-Cys-Au-Fe3O4/AFB1/BSA/aAFB1/4-ATP/Au at high magnification. It reveals that the secondary antibody interacts with coated antigen and forms sandwich like structure. Further, this fact is confirmed by EDX of this surface (Figure 4(c), image (D)) containing Fe and Au metal along with other elements and the weight% shown in image (D) (inset table). The presence of these elements confirms the interaction of secondary antibody r-IgG-Cys-Au-Fe3O4 conjugate over the AFB1/BSA/aAFB1/4-ATP/Au immunoelectrode.

3.3.3. Atomic Force Microscopy (AFM)

Topographic images were taken by AFM in noncontact mode (1 μm × 1 μm surface). In order to compare the topologies of each surface, surface roughness () and root mean square roughness () were estimated from the AFM images [41]. Figure 5 represents surface morphology of the bare Au (Figure 5(a)), 4-ATP/Au (Figure 5(b)), aAFB1/4-ATP/Au (Figure 5(c)), AFB1/BSA/aAFB1/4-ATP/Au (Figure 5(d)), and r-IgG-Cys-Au-Fe3O4/AFB1/BSA/aAFB1/4-ATP/Au (Figure 5(e)) based immunosensors. The and values and surface topology for the bare gold surface and 4-ATP/Au electrode are similar. Therefore, it can be assumed that a highly ordered and densely packed self-assembled layer of 4-ATP appears on the gold surface [42]. The drastic increase of surface roughness of aAFB1/4-ATP/Au (Figure 4(c)) as indicated by the values of (6.669 nm from 0.934 nm) and (2.109 nm from 0.195 nm) clearly demonstrates immobilization of the monoclonal aAFB1 antibody onto 4-ATP/Au electrode surface. The increase in and of aAFB1/4-ATP/Au immunosensor is attributed to the configuration and presence of paratope on antibody. After immobilization of BSA to block nonspecific sites, surface roughness further increases as observed by and values of 7.217 nm and 2.6 nm, respectively (image not shown), which was also supported by SEM image. The and values have been found to be reduced to 1.895 nm and 0.563 nm of AFB1/BSA/aAFB1/4-ATP/Au (image 5(d)) after the interaction of antigen with the immobilized antibody. Therefore, sandwiched structure (image 5(e)) of Au NP functionalized secondary antibody, antigen, and monoclonal antibody (r-IgG-Cys-Au-Fe3O4/AFB1/BSA/aAFB1/4-ATP/Au) is demonstrated by the rough surface topology and the (5.25 nm) and (6.39 nm) values.

3.3.4. FT-IR Studies

Figure 6 demonstrates FT-IR spectrum of the thiol monolayer between 500 and 3000 cm−1. The formation of a covalent gold-sulfur bond reveals the presence of the large band in the range of 624–640 cm−1 assigned to C–S stretching mode (spectrum (a)). The observed bands at 804 cm−1 and 1461 cm−1 and broad band at 3340 cm−1 due to =C–H deformation of the benzene ring, aromatic –C=C– in-plane vibrations, and N–H vibration of NH2 confirm the presence of 4-ATP on Au surfaces. After immobilization of aAFB1 on 4-ATP/Au electrode surface, the appearance of intense amide bands is characteristic of protein adsorption, amide I band at 1687 cm−1 corresponding to carbonyl C=O stretching vibration and amide II band at 1594 cm−1 due to the coupled C–N stretching, and –N–H bending mode indicates successful immobilization of monoclonal antibodies [43] (Figure 6, spectrum (b)). Figure 6 (spectrum (c)) represents FT-IR spectrum of AFB1/BSA/aAFB1/4-ATP/Au, that is, after recognition of AFB1 by BSA/aAFB1/4-ATP/Au immunosensor. The presence of 1474 cm−1 for methyl adjacent to epoxy ring, 1308 cm−1 for in-plane –CH bending of phenyl in Figure 6 (spectrum (c)), clearly indicates the presence of AFB1 on the surface of aAFB1/4-ATP/Au surface [42]. The bands at 1098 cm−1 for symmetric stretching of =C–O–C or symmetric bending of phenyl and 938 cm−1 for possibly isolated H further confirm interaction between coated AFB1-aAFB1 on the immunosensor surface. The band at 3414 cm−1 due to –N–H stretching of NH2 group of the antibody almost disappears in the spectrum of AFB1/BSA/aAFB1/4-ATP/Au (Figure 6, (c)) indicating strong interaction between the antigen epitope and paratope of the antibody.

3.3.5. Response Studies of the Immunosensor

The sensitivity and detection limit of an immunosensor depend on antibody loading. Prior to sensing studies, we have optimized all the parameters like the concentration of aAFB1 (40 μg mL−1), incubation time (30–35 min), and pH (7.4) in our previous study [35]. Further the concentration of r-IgG antibodies was optimized at 10–50 μg mL−1 (Figure 7(a)). Figure 7(a) shows the current response of AFB1/BSA/aAFB1/4ATP/Au electrode with varying (different solution, 10–50 μg mL−1) concentration of r-IgG-Cys-Au-Fe3O4 conjugate, current response increasing from 10 to 30 μg mL−1 after that decrease in current was observed, due to steric hindrance of conjugate or overloading of conjugates. Finally, 30 μg mL−1 concentration of r-IgG-Cys-Au-Fe3O4 conjugate was applied over the AFB1/BSA/aAFB1/4ATP/Au electrode for all the experiments.

Both competitive and noncompetitive strategies have been investigated. In case of competitive mode, the BSA/aAFB1/4-ATP/Au immunoelectrodes were fully covered with saturated concentration of AFB1 solution. Then the same (AFB1/BSA/aAFB1/4-ATP/Au) was allowed to interact with optimized concentration (30 μg mL−1) of secondary antibody conjugate (r-IgG-Cys-Au-Fe3O4) and free AFB1 with varying concentration. Finally, the response in the sandwiched form was recorded with EQCM-CV after washing with PB in N2 atmosphere. During the competition process, secondary antibodies easily access free AFB1, while the rest of the r-IgG-Cys-Au-Fe3O4 conjugates form sandwiched structure with the coated AFB1, respectively. Figure 7(b) shows calibration curve as a function of AFB1 concentration, linear range obtained from 0.05–5 ng mL−1 after which it decreases revealing that at 5 ng mL−1 concentration becomes saturated. The inset of Figure 7(b) shows the peak current intensity of the redox mediator is inversely proportional to the amount of free AFB1 in the sample and the peak current decreases with increase in concentration of AFB1. Scheme 2 represents the formation of this competitive sandwich type immunoelectrode. Each value is obtained in triplicate experiments and regression equation is obtained with a regression coefficient of ca. 0.98: This corresponds to the sensitivity of ca. 335.7 μA ng−1 mL cm−2 for AFB1 with a calculated detection limit of 0.07 ng mL−1.

For noncompetitive mode, BSA/aAFB1/4-ATP/Au immunoelectrode was allowed to interact with increasing concentration of AFB1. After the interaction, the same (AFB1/BSA/aAFB1/4-ATP/Au) electrode was again exposed to optimum concentration of r-IgG-Cys-Au-Fe3O4 conjugate to form sandwiched structure (r-IgG-Cys-Au-Fe3O4/AFB1/BSA/aAFB1/4-ATP/Au). Figure 7(c) shows calibration curve as a function of AFB1 concentration, linear range obtained from 0.5–5 ng mL−1, after which it decreases revealing that at 5 ng mL−1 concentration it becomes saturated. In this case, the peak current increases (Figure 7(c) inset) with increasing concentration of AFB1 as the concentration of r-IgG-Cys-Au-Fe3O4 conjugate also increases enhancing the electron transfer through the medium by virtue of the presence of Au-Fe3O4 nanoparticles on conjugate. All experiments were performed with triplicate measurements and the temperature was controlled at 25°C. The calibration curve shows range of 0.5 to 5 ng mL−1 with calculated LOD of 0.9 ng mL−1 for the noncompetitive mode. However, the regression coefficient is 0.933 and the linear equation is as follows: This corresponds to the sensitivity of ca. 123.9 μA ng−1 mL cm−2. It reveals that the competitive mode offers a wider linear range, higher sensitivity, and lower LOD and corresponds to higher regression coefficient in spite of identical nature of coated AFB1 and free AFB1.

3.3.6. Real Sample Testing and Selectivity of Immunoelectrode

To evaluate the applicability of the developed immunosensor to real sample analysis, corn flakes samples were spiked with various concentrations of AFB1. For this the corn flakes sample extracted with a methanolic solution of potassium bicarbonate was exploited as a real sample. Evaporation to dryness and final reconstitution in PBS buffer were necessary to avoid the inhibition of the antibody-antigen binding caused by methanol. The extract sample was spiked with three different concentrations of AFB1 (0.05, 2, and 5 ng mL−1) to examine the applicability of the proposed probe. During these experiments, the AFB1/BSA/aAFB1/4-ATP/Au immunosensor was dipped in the cell containing a mixture of different concentration of AFB1 spiked extracted sample and fixed amount of r-IgG-Cys/Au-Fe3O4 in PBS and incubated for 35 min. The EQCM-CV of AFB1/BSA/aAFB1/4-ATP/Au immunosensor was examined with the corn flakes extract in PBS, which reflects the minimum interference. Variation of peak current in blank and corn flakes extract is [Figure 8(a) (bars (1) and (2))] within 5%. It indicates that the AFB1/BSA/aAFB1/4-ATP/Au immunosensor shows the specificity towards the AFB1 antigen, not inflated with other constituents present in the corn flakes sample. However, the response of the AFB1/BSA/aAFB1/4-ATP/Au immunosensor changes when the corn flakes sample contains AFB1 and magnitude of the anodic peak further decreases as the AFB1 concentration increases in the corn flakes sample. The results obtained with this extracted solution in PBS and result from standard PBS within 3–5% variations show that the developed immunosensor is highly specific to AFB1; it avoids interference of other materials present in corn flakes extracts.

3.3.7. Reproducibility, Shelf Life, and Regeneration of Immunoelectrode

The reproducibility of the proposed immunoelectrode was estimated by repetitive measurement of immunoelectrode with current response using 2 ng mL−1 standard AFB1 solutions in PBS (50 mM, pH 7.4, 0.9% NaCl, containing 5 mM [Fe(CN)6]3−/4−). The results obtained in 5 repeated measurements show a relative standard deviation (RSD) of 2–3%, indicating that the obtained data are reproducible. These results demonstrate the acceptable reproducibility and precision of the proposed immunosensor. In addition, the immunosensor could be stored at 4°C for shelf life study. The stability of the BSA/aAFB1/4-ATP/Au immunoelectrode was evaluated by EQCM-CV study and the current response in the presence of 2 ng mL−1 standard AFB1 solution in PBS (50 mM, pH 7.4, 0.9% NaCl) was monitored at a regular interval of 7 days (Figure 8(b)). The immunoelectrode retains its activity up to 28 days with 5–7% decrease in activity. 95% of the initial response was left remaining after 1 week and 90% of the initial response was left remaining after 1 month, indicating acceptable stability.

The immunosensor can be regenerated (Scheme 2) using an external strong magnet to remove the immuno-r-IgG-Cys-Au-Fe3O4 conjugate (i.e., r-IgG-Cys-Au-Fe3O4). It was observed that the reagent-free regeneration method could regenerate the immunosensor up to 15-16 times with 2-3% loss in activity using a fixed concentration of AFB1.

4. Conclusions

In the present study, a reusable immunoelectrode is developed using self-assembled 4-ATP on quartz crystal electrode. EQCM measurement technique is applied to determine the response current of sandwiched structure comprising BSA/aAFB1/4-ATP/Au immunoelectrode, AFB1, and r-IgG-Cys-Au-Fe3O4. The regeneration of the immunoelectrode is done after removing the AFB1 attached with r-IgG-Cys-Au-Fe3O4 through external magnet. The immunosensor can be regenerated about 15-16 times with 2-3% loss of activity. We have compared the competitive and noncompetitive methods for the determination of the AFB1. It has been observed that the competitive mode has offered wider linear range of 0.05–5 ng mL−1 with the limit of detection of 0.07 ng mL−1 and higher sensitivity than the noncompetitive one while coated AFB1 and free AFB1 are identical in chemical structure. This immunosensor is found to be highly promising for detection of AFB1 in corn flakes samples. Therefore, the same principle can be utilized for detection of other food toxins such as OTA, OTB, fumonisins, and zearalenone and other small molecules also.

Conflict of Interests

The authors declare that there is no conflict of interests regarding the publication of this paper.


The authors thank Dr. Ashok Kumar Chauhan (Founder President, Amity University, Uttar Pradesh) for providing the facilities. They also thank Dr. (Mrs.) Balvinder Shukla, Vice Chancellor, Amity University, Uttar Pradesh, and Professor L. M. Bharadwaj, Director, AINT. The authors also express their deep gratitude to Ms. Shuvra Singha, Department of Chemistry, University of Hyderabad, India, for her help in conducting the XRD studies.