Highly selective enrichment of aflatoxin B1 from edible oil using polydopamine-modified magnetic nanomaterials

kdz1011@just.edu.cn Abstract Aflatoxin B 1 (AFB 1 ) is a highly toxic mycotoxin that enters the human body through the food chain and poses a serious threat to human health. In this paper, polydopamine (PDA)-coated Fe 3 O 4 magnetic nanoparticles (Fe 3 O 4 @PDA MNPs) were prepared by the co-precipitation method to enrich aflatoxin from edible oil. Transmission electron microscopy (TEM), Fourier transform infrared spectroscopy, X-ray photoelectron spectroscopy, and vibrating sample magnetometer were used to characterize the Fe 3 O 4 @PDA MNPs. Using the obtained Fe 3 O 4 @PDA MNPs as an adsorbent, a simple method for enriching AFB 1 from samples by magnetic solid phase extraction (MSPE) combined with fluorescence rapid detection was developed. The effects of the ratio of Fe 3 O 4 MNPs to PDA, adsorption dosage, sample volume, adsorption time, and elution time on enrichment of AFB 1 were investigated to determine the optimal experimental conditions. This method has good intraday and daytime . Practical Application: Polydopamine was used for Fe 3 O 4 magnetic nanoparticles coating by the co-precipitation method. This material can be used for selective enrichment of AFB 1 in edible oil with high enrichment factor and has good intraday precision and daytime


Introduction
Mycotoxins are toxic secondary metabolites produced by a variety of molds that can cause serious harm to human health by contaminating various foods and animal feeds (Abia et al., 2013). In recent years, aflatoxins have become one of the most important and highly toxic groups of mycotoxins; they have been frequently detected in agriculture, attracting global attention (Pietri et al., 2016). Among the aflatoxins, aflatoxin B 1 (AFB 1 ) is the most toxic, and was listed as the first class of carcinogens by the International Agency for Research on Cancer (Lee et al., 2015). It is primarily produced by Aspergillus flavus and Aspergillus parasiticus. Many studies have shown that AFB 1 is genotoxic, carcinogenic, embryotoxic, teratogenic, and immunotoxic, (Kew, 2013). AFB 1 is widely distributed in nature and food, especially in peanuts, corn, rice, sorghum, milk, and oil (Li et al., 2018). Edible oil contaminated by AFB 1 has been widely distributed throughout the human population. This kind of pollution is difficult to remove and seriously threatens human health and safety (Dai et al., 2017). Therefore, a safe and effective strategy is need to detect and degrade AFB 1 in food.
Magnetic solid phase extraction (MSPE) is a new type of sample preparation technology widely used in the detection of organic pollutants Zheng et al., 2014), metal ions (Xiang et al., 2013), and biologically active substances (Xu et al., 2016). Because Fe 3 O 4 submicron particles coated with polydopamine (PDA, Fe 3 O 4 @PDA) are magnetic, have a large surface area, strong adsorption capacity, hydrophilicity, and are easily separated, they are considered an ideal adsorption material. In this paper, the magnetic adsorbent Fe 3 O 4 and its modification were prepared to enrich AFB 1 in samples using the MSPE method under the auxiliary conditions of oscillation or ultrasound. The magnetic adsorbent containing AFB 1 was separated from the sample matrix by an external magnetic field. AFB 1 eluted from the magnetic adsorbent was rapidly detected by fluorescent immunization. MSPE technology is much faster to use than traditional solid phase extraction technology in column or filtration operations. Moreover, when the contact area of the magnetic adsorbent and the target analyte in the extraction process is sufficiently large, the phase transfer of the target analyte can be completed quickly with a high extraction efficiency.

Apparatus
A magnetic force heating mixer was obtained from Changzhou Putian Instrument Manufacturing Co. Ltd. (Changzhou, Wuxi, China). A KQ-200KDB ultrasonic cleaner was obtained from Kunshan Ultrasonic Instrument Co. Ltd.. An S-4800IIFESE scanning electron microscope was obtained from High-Technologies Corporation (Japan). A Tecnai 12 transmission electron microscope was obtained from Philips Company (Netherlands). An IS410 Fourier transform infrared spectrometer was obtained from ThermoFisher. A PPMS-9 MPMS-XL vibrating sample magnetometer was obtained from Quantum Design. An ESCALAB 25 X-ray photoelectron spectroscope was obtained from Thermo Scientific (USA). A DHG-9101-1S electrothermal blowing dry box was obtained from Changzhou Putian Instrument Manufacturing Co. Ltd. (Changzhou, Wuxi, China). An FD-100 fluorescent quantitative immunoanalyzer was obtained from Shanghai Feice Biotechnology Co. Ltd..

Preparation of Fe 3 O 4 @PDA MNPs
First, 1.350 g of FeCl 3 ·6H 2 O, 3.854 g of NH 4 Ac and 0.4 g of sodium citrate were dissolved in 70 mL of EG, then stirred at 25°C for 10 min to dissolve the reactants completely. The mixed solution was heated for 1 h at 170°C, and then placed in a stainless-steel high-pressure autoclave equipped with a polytetrafluoroethylene lining. The reaction kettle was sealed at 200°C for 12 h, and then cooled to room temperature. Magnetic products were separated and collected by magnets, washed with EtOH and DDW three times, and then dried in a vacuum dryer at 60 °C for 24 h to obtain pure Fe 3 O 4 MNPs.
The synthesized Fe 3 O 4 MNPs and dopamine hydrochloride were dissolved in Tris-HCl (pH=8.5) solution in a certain proportion, and mechanically stirred at room temperature for 24 h. Then the product was separated and collected by magnets and washed with EtOH and DDW three times. Finally, it was dried in a vacuum dryer at 60°C for 24 h to obtain a pure core-shell structured Fe 3 O 4 @PDA MNPs.

Fluorescence rapid detection of AFB 1
One milliliter of the AFB 1 -containing oil sample and the sample extract in a 1:5 ratio were added to a 10-mL centrifuge tube and placed on a shaking incubator for 8 min. After extracting, the tube was centrifuged for 2 min at 4000 rpm. Then 100 µL of the supernatant was added to 600 µL of the sample dilution solution. After mixing, 100 µL sample solution was added to the AFB 1 fluorescent quantitative rapid detection reagent strip sample hole by pipette, incubated for 8 min, and then the reagent strip was inserted into the fluorescence reader to determine the concentration of AFB 1 in the oil sample.

Enrichment and elution of AFB 1
After determining the AFB 1 concentration, 25 mL of sample solution containing 3.6 µg/L of AFB 1 was transferred into a 100 mL beaker. Then, 0.03 g of activated Fe 3 O 4 @PDA MNPs was added, and the suspension was oscillated to facilitate the adsorption of AFB 1 onto the surface of the adsorbent. Then, the mixture was placed on a super magnet and magnetically separated into a solution and solid Fe 3 O 4 @PDA MNPs. The concentration of AFB 1 in the separated solution was determined, then 2.0 mL of a mixture of Me 2 CO/ACN/CH 2 Cl 2 (1:1:2, v/v/v) was added and the mixture was subjected to ultrasonication for 10 min. After desorption, the eluent was separated by magnetic decantation and evaporated to dryness under nitrogen gas flow at room temperature. The dry residue was dissolved in 2.0 mL of 0.5 mM Triton X-100 in 15% (v/v) ACN/water and the solution was oscillated for 5 min. The final solution was evaporated to 300 µL under nitrogen and the concentration of AFB 1 was determined.

Method validation
Intraday precision was evaluated by spiking samples with three different AFB 1 concentrations (1.8, 3.6, or 7.2 µg/L). Five replicates of each concentration were analyzed on the same day to determine the accuracy of the method. For interday precision, samples spiked with the same amount of AFB 1 were analyzed on three consecutive days. The precisions were expressed as the percentage relative standard deviation (RSD). The enrichment factor (EF) of the method and recovery was calculated using the following equation: where S s = sample volume, Sel = elution volume, and R% = percent recovery.

Electron microscopy
In the weakly alkaline solution Tris-HCl, dopamine self-polymerized and adsorbed onto the surface of the Fe 3 O 4 MNPs. The magnetic nanoparticles (Fe 3 O 4 @PDA MNPs) coated with polydopamine were obtained. The shape and size of the Fe 3 O 4 MNPs and Fe 3 O 4 @PDA MNPs were visualized and characterized by TEM and SEM images. As shown in Figure 1a and b, the synthesized magnetic Fe 3 O 4 MNPs were homogeneous particles that were slightly aggregated and approximately 250 nm in size. Figure 1c is the TEM diagram of dopamine-coated Fe 3 O 4 @PDA MNPs. The individual particles were approximately 350 nm. The polymer coating formed by polydopamine adsorbed onto the surface of Fe 3 O 4 MNPs was formed by hydroxyl-iron chemical interaction.

X-ray diffraction of MNPs.
In order to verify the Fe 3 O 4 @PDA MNPs were properly prepared in this assay, the crystal structure and phase composition β = 0.00137 rad, 2θ = 35.5° at the strongest peak, and the calculated magnetic core particle size is 220.44 nm, which is similar to the average particle diameter of the magnetic core measured by TEM.
The Fourier transform infrared spectra of the prepared Fe 3 O 4 MNPs and Fe 3 O 4 @PDA MNPs are shown in Figure 3a.
The absorption peak at 3430 cm -1 is the stretching vibration peak of the OH functional group, and the corresponding bending vibration peak is at 1623 cm -1 . There is a strong absorption band near 575 cm -1 , which is the stretching vibration peak of the Fe-O-Fe bond and the characteristic absorption peak of Fe 3 O 4 (Wei et al., 2010). The PDA spectrum shows a large relative in the area of 1500-1100 cm -1 . Absorbance, which is due to the formation of polymers, is primarily attributable to CO and CN functional groups (Si & Yang, 2011). The peak at 1507 cm -1 indicates the presence of the N-H bending vibration; at 1435 cm -1 is the C-C tensile vibration, and the weaker peak at 1281 cm -1 indicates the presence of C-O tensile vibration.

Vibrating sample magnetometer of MNPs
In order to achieve rapid solid-liquid separation of magnetic nanomaterials from aqueous solution, Fe 3 O 4 @PDA MNPs must have sufficient magnetic strength. The magnetic properties of Fe 3 O 4 MNPs and Fe 3 O 4 @PDA MNPs were investigated by vibrating sample magnetometer. The hysteresis loop of Fe 3 O 4 @PDA MNPs is shown in Figure 4. The maximum saturation magnetization of Fe 3 O 4 MNPs and Fe 3 O 4 @PDA MNPs was 90 and 45 emu/g, respectively. The coercivity and residual magnetization of the two MNP types were close to zero, which is characterized by paramagnetism. Compared with unmodified Fe 3 O 4 MNPs, the saturation magnetic strength of dopamine-modified Fe 3 O 4 @ PDA MNPs was significantly weakened due to the non-magnetic polymer coating, but the maximum saturation magnetic strength It has been reported that after the surface polymerization of dopamine, some absorption bands in the infrared spectrum are slightly changed (Zeng et al., 2013).

X-ray photoelectron spectroscopy of MNPs
X-ray photoelectron spectroscopy was used to investigate the elemental composition of the surface of magnetic nanomaterials. As shown in Figure 3b, the surface of Fe 3 O 4 MNPs was mainly composed of Fe and O elements. After dopamine modification, the surface of the magnetic nanomaterials also contained N and C elements, and the intensity of the Fe element signal peak is significantly weakened. The result indicates that polydopamine was successfully coated onto the surface of Fe 3 O 4 MNPs .  It can also be seen from the photograph that the Fe 3 O 4 @PDA MNPs were dispersed in water to form a uniform suspension. Under the action of an external magnetic field, the Fe 3 O 4 @PDA MNPs were separated from the water and gathered around the magnet, and the solution was transparent. When the applied magnetic field was removed, the Fe 3 O 4 @PDA MNPs were evenly dispersed in the water. This process was repeated to demonstrate that the prepared Fe 3 O 4 @PDA MNPs were superparamagnetic.

Optimization of the MSPE procedure
In order to study the effect of different ratios of Fe 3 O 4 MNPs and PDA on their AFB 1 extraction efficiency, Fe 3 O 4 @PDA MNPs were prepared with different ratios (1:1, 1:2, 1:3, 1:4, 1:5). As shown in Figure 5a, the most effective ratio was 1:3; excessive

Conclusions
In this study, polydopamine-coated magnetic nanomaterials Fe 3 O 4 @PDA MNPs which have a high affinity for aflatoxins were prepared, forming simple and effective magnetic solid particles that can be extracted from large-volume liquid samples. Using Fe 3 O 4 @PDA MNPs as an adsorbent, the microbial aflatoxin AFB 1 in edible oil samples was analyzed by fluorescence immunoassay, and the enrichment effect and optimal experimental conditions were determined. The enrichment factor was 84. The ratio of Fe 3 O 4 MNPs to PDA was 1:3, the amount of Fe 3 O 4 @PDA MNPs was 0.03 g, the sample volume was 25 mL, the adsorption time was 60 min, and the elution time was 10 min. The method has good intraday precision and daytime precision. The RSDs are 2.39% ~ 2.48% and 1.56% ~ 2.57%, respectively, and the recovery rate is in the range of 98.17% ~ 101.67%. thickness of the PDA coating reduced the magnetic properties of the material and impact the enrichment effect.
To determine the effect of adsorbent amount on the extraction efficiency of AFB 1 , different amounts of adsorbent Fe 3 O 4 @PDA MNPs (0.005, 0.01, 0.02, 0.03, 0.04 and 0.05 g) were added to 25 mL of 3.60 µg/L AFB 1 extract. In Figure 5b, the concentration increased with the amount of the magnetic adsorbent until reaching 0.3 g, and then remained unchanged. The large specific surface area of the nanosorbent may explain the low mass of adsorbent required. Therefore, we chose 0.03 g as the best adsorbent amount for subsequent experiments.
To evaluate the possibility of enriching low concentrations of AFB 1 from large volumes of sample, 5 mL of 3.60 µg/L AFB 1 extract was diluted to 10, 15, 20, 25, 30, 40, or 50 mL. Figure 5c indicates that a quantitative recovery was available at 25 mL. As previously described, the final amount of analyte was 300 µL. Therefore, the theoretical enrichment factor was 84, which verifies the feasibility of determining AFB 1 at different concentrations.
The remaining experimental conditions were kept unchanged, and the shaking adsorption times tested were 15, 30, 45, 60, 75, and 90 min. As show in Figure 5d, adsorption efficiency was maximized when the adsorption time was 60 min and remained unchanged over longer adsorption times. To shorten the experiment time, the optimal adsorption time was 60 min. Similarly, from Figure 5e, when the elution times tested were 5, 10, 15, and 20 min in the ultrasound system, the highest analytical effect of the elution time was obtained at 10 min. After the time exceeded 10 min, the elution efficiency began to decrease; thus, 10 min was chosen as the best resolution time.

Method validation
The accuracy of determining the AFB 1 concentration of the method was evaluated by preparing standards of different concentrations of AFB 1 . Table 1 summarizes the results of intraday precision analysis. Table 2 summarizes the results of daytime precision analysis. The RSDs are 2.39% ~ 2.48% and 1.56% ~ 2.57%, respectively. Intraday and daytime changes indicate that the method has better accuracy. As shown in Table 1, a good recovery rate was in the range of 98.17% ~ 101.67%.