Aflatoxin B1 detection method based on BIC terahertz metamaterial sensor

Abstract

The invention discloses an aflatoxin B1 detection method based on a BIC terahertz metamaterial sensor, and designs a continuous domain bound state (BIC) terahertz metamaterial sensor based on an electromagnetic theory and a time domain finite difference method. Terahertz enhanced spectrums of aflatoxin B1 solutions with different concentrations are collected, and analysis finds that the amplitude of a transmission peak of aflatoxin B1 at about 1.5 THz is gradually reduced along with the increase of the concentration of the aflatoxin B1 solution, and the frequency of the transmission peak gradually moves to a low frequency along with the increase of the concentration of the aflatoxin B1 solution; a PLS and LS-SVM quantitative model of aflatoxin is established, analysis finds that the LS-SVM model has a better effect, and the detection limit of aflatoxin B1 is 2.04 * 10 <-7 > mu g / ml. By adopting the method disclosed by the invention, rapid, nondestructive and high-sensitivity detection of the aflatoxin B1 in the grains can be realized.

Description

Technical Field

[0001] The present invention relates to the technical field of chemical detection, and in particular to a method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor. Background Art

[0002] Aflatoxin (AFT) is a metabolite with a bifuran ring and a naphthalene-oxanone structure, primarily produced by strains of Aspergillus flavus and Parasitic Aspergillus. Aflatoxins have over 20 derivatives, of which B1, B2, G1, and G2 are the most common and toxic. Aflatoxins can contaminate crops such as cotton, peanuts, corn, and peppers, significantly impacting human and animal health. Aflatoxin B1 is classified as a Class I carcinogen by the World Health Organization's International Agency for Research on Cancer (IARC). It is highly hepatotoxic, mutagenic, and carcinogenic. Long-term consumption can cause serious health problems such as liver cancer and immunosuppression. Due to its harmfulness, countries have established strict limit standards for the content of aflatoxin B1 in food and monitor it through a variety of detection technologies. The maximum allowable content of aflatoxin B1 in vegetable oils in China is 10μg / kg, and the maximum allowable content of aflatoxin B1 in Europe is 2μg / kg. Aflatoxin must not be detected in infant food. Therefore, it is crucial to establish a highly sensitive, rapid and non-destructive detection method for aflatoxin B1.

[0003] Currently, common methods for detecting aflatoxins include enzyme-linked immunosorbent assay (ELISA), colloidal gold immunochromatography, high-performance liquid chromatography (HPLC), liquid chromatography-tandem mass spectrometry (LC-MS / MS), and biosensors. These methods suffer from high costs, long cycles, poor reproducibility, complex experimental procedures, expensive instrumentation, and potential damage to samples. Therefore, there is an urgent need for a highly sensitive, rapid, and non-destructive detection method.

[0004] Terahertz (THz) waves refer to electromagnetic waves with a frequency in the range of 0.1 to 10 THz, which is between microwaves and infrared light. Due to its low photon energy, strong penetration, and high spectral resolution, terahertz waves have shown great potential in biomedicine, chemical testing, material analysis and other fields. In recent years, researchers have made certain progress in detecting biological molecules, chemical substances, etc. using terahertz spectroscopy. However, traditional terahertz technology still faces challenges in detecting extremely low concentrations of toxic and harmful substances. The main reason is that most substances have a weak electromagnetic response to terahertz waves, resulting in limited detection sensitivity. Therefore, how to use terahertz technology to quantitatively detect extremely low concentrations of aflatoxin B1 and improve detection sensitivity is a technical problem that technicians in this field need to solve. Summary of the Invention

[0005] In view of the above situation, the purpose of the present invention is to provide a method for detecting aflatoxin B1 based on terahertz metamaterial sensors and chemometrics, so as to develop a new method based on metamaterials and use the metamaterials to achieve quantitative detection of aflatoxin B1, providing a new approach for the quantitative detection of aflatoxin B1.

[0006] The present invention provides an aflatoxin B1 detection method based on a BIC terahertz metamaterial sensor. The BIC terahertz metamaterial sensor adopts a silicon substrate with a gold film on the surface. The BIC terahertz metamaterial sensor is composed of a plurality of periodic structural units. The periodic structural units are square structures. The width p of the periodic structural units is 48 μm. A rectangular non-gold-plated area is provided on the periodic structural unit. The width W of the rectangular non-gold-plated area is 4.5 μm. The rectangular non-gold-plated area satisfies the conditional formula: I=α*p, where α is an influence factor and I is the length of the rectangular non-gold-plated area.

[0007] The detection method comprises the following steps:

[0008] (1) Preparing multiple aflatoxin B1 standard solutions of different concentrations, dropping the aflatoxin B1 standard solutions onto the BIC terahertz metamaterial sensor, and placing the sensor in a drying oven for drying;

[0009] (2) Place the dried BIC terahertz metamaterial sensor into the terahertz system and use the transmission mode for measurement to collect the terahertz spectrum information of the corresponding sample;

[0010] (3) Using spectral feature variable extraction algorithm to extract features of terahertz spectrum information;

[0011] (4) Establish a quantitative detection model of aflatoxin B1 terahertz transmission spectrum based on the extracted features and LS-SVM;

[0012] (5) Obtaining terahertz spectrum information of the sample containing aflatoxin B1 to be tested and inputting it into the quantitative detection model to obtain the content of aflatoxin B1 in the sample to be tested.

[0013] Compared with the existing technology, the present invention provides a method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor. A continuous domain bound state (BIC) terahertz metamaterial sensor is designed based on electromagnetic theory and the finite-difference time-domain method (FDTD). Using this BIC terahertz metamaterial sensor, terahertz-enhanced spectra of aflatoxin B1 solutions of varying concentrations are collected. Analysis reveals that the amplitude of the transmission peak of aflatoxin B1 at around 1.5 THz gradually decreases with increasing aflatoxin B1 solution concentration, and the frequency of the transmission peak gradually shifts toward lower frequencies. A PLS and LS-SVM quantitative model for aflatoxin was established, and analysis revealed that the LS-SVM model performed better, resulting in a detection limit of 2.04×10 -7 μg / ml. This application verifies the feasibility of highly sensitive detection of aflatoxin B1 solutions using a BIC terahertz metamaterial sensor combined with THz-TDS technology. The present invention utilizes continuum bound state technology to enhance the electromagnetic response of aflatoxin B1 to terahertz waves, ultimately increasing detection sensitivity. This invention enables rapid, nondestructive, and highly sensitive detection of aflatoxin B1 in grain.

[0014] In addition, the aflatoxin B1 detection method based on the BIC terahertz metamaterial sensor provided by the present invention also has the following technical features:

[0015] The BIC terahertz metamaterial sensor is prepared by the following preparation method:

[0016] The structural parameters of the BIC terahertz metamaterial sensor are designed using the time-domain finite-difference method. The BIC mechanism is introduced into the terahertz metasurface to achieve high-Q resonance. A silicon substrate is selected, and the substrate surface is cleaned to remove impurities and oxidized to enhance the adhesion of the photoresist. Subsequently, a chromium layer is plated on the silicon substrate using thin-film deposition technology, followed by a metal layer to enhance the adhesion between the metal layer and the silicon substrate. The silicon wafer is then coated with a layer of photoresist with a thickness of 1-2 μm using a spin coating process. After coating, a pre-baking process is performed. Bake at 100°C for 120 seconds to remove the solvent in the photoresist, ensuring that it is hardened and ready for exposure; after exposure, enter the development process, and the development time is controlled at 30 seconds to remove the unexposed part of the photoresist, leaving the pattern; after development, use a solvent to perform a glue treatment to remove the residual photoresist; then, use reactive ion etching or ion beam etching to perform the etching operation; after etching is completed, remove the photoresist mask through acetone ultrasonic cleaning to ensure that there is no residue on the chip surface; finally, perform surface oxidation or metal deposition post-treatment to enhance conductivity and corrosion resistance.

[0017] Wherein, the step (1) specifically includes:

[0018] According to the preset ratio scheme, 6 aflatoxin B1 standard solutions with different concentration gradients were prepared by step-by-step dilution with a precision pipette. All samples were fully shaken on a vortex oscillator for 3 minutes. Multi-stage dilution operations were performed during the experiment. After the dilution was completed, the solutions were placed in volumetric flasks for storage.

[0019] Wherein, the step (2) specifically includes:

[0020] Before the experiment, the system was preheated by an air compressor with an air dryer, and dry air was continuously filled into the optical cavity. The humidity was monitored by a hygrometer to keep it below 10% and the temperature constant at 25±0.5℃. During the experiment, wait for 2 minutes after each sample was placed, and start measurement after the system stabilized. A pipette was used to draw 20μl of sample solution in the order of sample concentration from low to high and drop it on the metamaterial sheet. After each sample test, the metamaterial sheet was placed in a beaker and shaken to clean it. Before collecting the sample spectrum, the metamaterial sheet with the sample was placed in a drying oven for 15 minutes, and the drying oven temperature was set to 50℃. Finally, the metamaterial sheet with the sample was placed in the terahertz system and measured using the transmission mode. Five points were taken for each sample, and 10 measurements were performed at each point. A total of 300 spectra were obtained for the six concentration samples, and each spectrum was the average of 64 repeated tests. The KS algorithm was used to divide the samples into a calibration set and a prediction set in a ratio of about 3:1 for subsequent model building.

[0021] Wherein, in step (5), the established quantitative detection model of aflatoxin B1 terahertz transmission spectrum is evaluated using the prediction set correlation coefficient, the prediction set root mean square error, the correction correlation coefficient, and the correction root mean square error.

[0022] Among them, the spectral feature variable extraction algorithm is any one of a competitive adaptive reweighted sampling algorithm, an uninformative variable elimination algorithm, a continuous projection algorithm, a variable projection importance analysis method, and an iterative information variable retention algorithm.

[0023] Among them, the value of α is 0.7. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments with reference to the accompanying drawings, in which:

[0025] Figure 1 Schematic diagram of a periodic structural unit in the BIC terahertz metamaterial sensor;

[0026] Figure 2 The transmission spectra of metamaterial A with α of 0.3 are measured at different angles.

[0027] Figure 3 The transmission spectra of metamaterial B with α of 0.4 at different test angles are shown below.

[0028] Figure 4 The transmission spectra of metamaterial C with α of 0.5 corresponding to different angles tested;

[0029] Figure 5 The transmission spectra of metamaterial D with α of 0.6 corresponding to different angles tested;

[0030] Figure 6 The transmission spectra of metamaterial E with α of 0.7 corresponding to different angles tested;

[0031] Figure 7 Transmission spectra of metamaterials with different α at 90°;

[0032] Figure 8 Transmission spectra of aflatoxin B1 solutions with different concentrations on the surface of the metamaterial;

[0033] Figure 9 This is the fitting diagram of the predicted values and true values of the THz spectrum of aflatoxin B1 solution by PLS and LS-SVM models. DETAILED DESCRIPTION

[0034] To make the objects, features, and advantages of the present invention more readily apparent, the following detailed description of specific embodiments of the present invention is provided in conjunction with the accompanying drawings. The accompanying drawings illustrate several embodiments of the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a more thorough and comprehensive understanding of the present invention.

[0035] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this invention pertains. The terms used in this specification of the present invention are for the purpose of describing specific embodiments only and are not intended to limit the present invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0036] See also Figure 1One embodiment of the present invention provides a BIC terahertz metamaterial sensor. In this embodiment, the BIC terahertz metamaterial sensor uses a silicon substrate (refractive index 3.335) with a gold film on the surface. The BIC terahertz metamaterial sensor is composed of multiple periodic structural units. Each periodic structural unit has a square structure and a width p of 48 μm. Each periodic structural unit is provided with a rectangular non-gold-plated area. The non-gold-plated area refers to an area with only the silicon substrate and no gold film. The width W of the rectangular non-gold-plated area is 4.5 μm. The rectangular non-gold-plated area satisfies the conditional formula: I = α*p, where α is the influence factor and I is the length of the rectangular non-gold-plated area.

[0037] Bound continuum states (BICs) are completely localized resonances occurring within the radiation continuum frequency range. They exhibit extremely high quality factors (Q) and generate strong interactions between light and matter, a phenomenon crucial for the development of novel functional devices. Introducing BICs into terahertz metasurfaces offers a new approach for customizing high-Q resonances. The project proposes using polarized terahertz light as the input and introducing a perturbation factor. By adjusting the perturbation factor, the BIC resonances are expected to enhance the interaction between the terahertz spectrum and matter.

[0038] During the fabrication of the BIC terahertz metamaterial sensor, the present invention utilizes the finite-difference time-domain (FDTD) method to design the structural parameters of the BIC terahertz metamaterial sensor. Considering that metamaterials are composed of periodic structural units, the transmission or reflection properties of a single periodic structural unit can represent the spectral properties of the entire metamaterial. Therefore, during simulation, a single periodic structural unit can be used to simulate the metamaterial to increase simulation speed.

[0039] Specifically, a silicon substrate is first selected, the surface of the substrate is cleaned to remove impurities, and an oxidation treatment is performed to enhance the adhesion of the photoresist; then, a layer of chromium is plated on the silicon substrate using thin film deposition technology, and then a metal layer is deposited to enhance the adhesion of the metal layer to the silicon substrate; then, a layer of photoresist with a thickness of 1-2 μm is coated on the silicon wafer using a spin coating process; after coating, a pre-bake is performed at 100°C for 120 seconds to remove the solvent in the photoresist, ensuring that it is hardened and ready for exposure; after exposure, the development process is entered, and the development time is controlled at 30 seconds to remove the unexposed part of the photoresist, leaving a pattern; after development, a solvent is used for glue treatment to remove residual photoresist; then, reactive ion etching or ion beam etching is used for etching; after etching is completed, the photoresist mask is removed by acetone ultrasonic cleaning to ensure that there is no residue on the chip surface; finally, surface oxidation or metal deposition post-treatment is performed to enhance conductivity and corrosion resistance.

[0040] In this embodiment, α is taken as 0.3, 0.4, 0.5, 0.6, and 0.7, respectively, and I is obtained as 14.4, 19.2, 24, 28.8, and 33.2 μm, respectively.

[0041] In this embodiment, metamaterials with different α values are tested. For α values of 0.3, 0.4, 0.5, 0.6, and 0.7, they are named A, B, C, D, and E, respectively. Terahertz transmission spectra of ABCDE at different angles are measured. Figure 2 The transmission spectra of metamaterial A with α of 0.3 are tested at different angles. Figure 3 The transmission spectra of metamaterial B with α of 0.4 at different test angles are shown below. Figure 4 The transmission spectra of metamaterial C with α of 0.5 corresponding to different angles tested are as follows: Figure 5 The transmission spectra of metamaterial D with α of 0.6 corresponding to different angles tested are as follows: Figure 6 The transmission spectra of the metamaterial E with an α of 0.7 were tested at different angles. After comparison, it was found that the BIC metamaterial had the best effect at angles of 90° and 270°. Figure 7 The transmission spectrum of metamaterials with different α at 90°. Figure 7 It can be seen that the metamaterial E has the best effect when α is 0.7.

[0042] Another embodiment of the present invention provides a method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor, wherein the BIC terahertz metamaterial sensor is the above-mentioned BIC terahertz metamaterial sensor, and the detection method includes the following steps (1) to (5):

[0043] (1) Prepare multiple aflatoxin B1 standard solutions of different concentrations, drop the aflatoxin B1 standard solutions on the terahertz metamaterial absorber, and place it in a drying oven for drying.

[0044] Aflatoxin B1 standard solution was purchased from Prebon Biotechnology with an initial concentration of 25 μg / ml. To meet the requirements of trace detection, ultrapure deionized water (18.2 MΩ·cm, Merck Millipore, USA) was used to perform a gradient dilution of the standard solution. The specific operation was as follows: First, according to the preset ratio scheme, six standard solutions with different concentration gradients (the highest concentration was 5.376×10 -7 μg / ml, which is much lower than the solubility of aflatoxin B1 in water of 0.012g / L). All diluted samples were stored in 10ml volumetric flasks. To ensure the homogeneity of the solution, all samples were thoroughly shaken on a vortex shaker for 3 minutes. Multiple dilutions were performed during the experiment: after the first dilution, the concentration range of aflatoxin B1 solution was 4.8×10-5 μg / ml~3.84×10 -4 μg / ml; after the second dilution, the concentration range was adjusted to 1.92×10 -6 μg / ml~1.536×10 -5 μg / ml; finally, the ultra-low concentration series of standard solutions shown in Table 1 were obtained through the third dilution. The entire dilution process strictly followed the operating specifications of trace analysis to ensure the reliability of the experimental data.

[0045] Table 1 Aflatoxin B1 solution ratio table

[0046]

[0047] (2) The dried terahertz metamaterial absorber is placed in a terahertz system and measured using the transmission mode to collect the terahertz spectrum information of the corresponding sample.

[0048] The terahertz transmission spectrum of aflatoxin B1 solution was measured using a TAS7500 terahertz time-domain spectrometer from Advantest, Japan. The system consists of three main components: a femtosecond laser, a terahertz emitter, and a terahertz detector. The spectrum measurement range is 0.1-5.0 THz, and the instrument has a resolution of 7.6 GHz. To minimize environmental interference and ensure stable sample information, the system was preheated using an air compressor with an air dryer before the experiment. Dry air was continuously filled into the optical cavity, and a hygrometer was used to maintain humidity below 10% and temperature at a constant 25 ± 0.5°C. During the experiment, a two-minute wait was maintained after each sample was added, allowing the system to stabilize before measurement began. To minimize the influence of residual concentration on the metamaterial sheet, 20 μl of sample solution was pipetted onto the metamaterial sheet in ascending order of concentration. After each sample was tested, the metamaterial sheet was shaken in a beaker to clean it. Because water strongly absorbs terahertz waves, before collecting sample spectra, to ensure that the solution on the metamaterial sheet is fully dried, the metamaterial sheet with samples is placed in a drying oven for 15 minutes, and the drying oven temperature is set to 50°C. Finally, the metamaterial sheet with samples is placed in the terahertz system and measured using transmission mode. To reduce the impact of random errors on the experimental results, 5 points are taken for each sample, and each point is measured 10 times. A total of 300 spectra are obtained for the 6 concentration samples, and each spectrum is the average of 64 repeated tests. The KS (Kennard-Stone) algorithm is used to divide the collected terahertz spectra of all samples into a calibration set and a prediction set in a ratio of approximately 3:1 for subsequent model building.

[0049] Then, the terahertz spectrum parameters are extracted. First, the terahertz time domain spectra of aflatoxin with different concentrations are collected, and the frequency domain signal is obtained by fast Fourier transform (FFT), which is expressed as formula (1).

[0050]

[0051] Where A(ω) represents the electric field amplitude, The phase difference between the reference signal and the sample signal, E(t) is the terahertz time domain signal. The refractive index and absorption coefficient of the sample are obtained by formula (2) and formula (3).

[0052]

[0053] where n(ω) is the refractive index, α(ω) is the absorption coefficient, ω is the frequency, k(ω) is the extinction coefficient, ρ(ω) is the amplitude ratio function, d is the sample thickness, and c is the speed of light in a vacuum.

[0054] Then, the terahertz transmission spectrum of aflatoxin was extracted by comparing the sample spectrum with the reference spectrum using formula (4).

[0055] T=(A S / A R ) 2 (4)

[0056] Among them A S and A R are the amplitudes of the sample signal and the reference signal, respectively.

[0057] The BIC metamaterial combined with THz-TDS technology was used to detect and analyze aflatoxin B1 solution. Figure 8 This is the transmission spectrum of aflatoxin B1 solution at a partial concentration on the surface of the metamaterial absorber. Considering that there is a lot of noise interference in the frequency bands below 1 THz and above 3 THz, in order to ensure the accuracy of the measurement, the transmission spectrum of the 1-3 THz frequency band is intercepted. The 263 spectral data contained in this frequency band are used for subsequent modeling. Figure 8 The transmission peak mainly shows the transmission characteristics of the metamaterial itself, from which we can see that there are transmission peaks around 1.5THz. Figure 8The middle magnified portion shows a partial zoom of the transmission spectrum of aflatoxin B1 solution at 1.5 THz at a certain concentration. Although the aflatoxin B1 content on the BIC metamaterial surface is extremely low, after signal enhancement and sensing, the transmission peak around 1.5 THz exhibits a clear amplitude and frequency shift response pattern. Specifically, the amplitude of the transmission peak around 1.5 THz gradually decreases with increasing aflatoxin B1 solution concentration, while the frequency of the transmission peak gradually shifts toward lower frequencies. This provides a theoretical basis for further analysis of the response relationship between the metamaterial's transmission peak characteristics and aflatoxin B1 solution concentration.

[0058] (3) The spectral feature variable extraction algorithm is used to extract features of terahertz spectral information.

[0059] The spectral feature variable extraction algorithm can be any one of a competitive adaptive reweighted sampling algorithm, an uninformative variable elimination algorithm, a continuous projection algorithm, a variable projection importance analysis method, and an iterative information variable retention algorithm.

[0060] The Competitive Adaptive Reweighted Sampling (CARS) algorithm is a feature variable selection method that combines Monte Carlo sampling with adaptive reweighting. By simulating a "survival of the fittest" competitive mechanism, the algorithm dynamically selects information variables with significant spectral characteristics during an iterative process. The core process of the CARS algorithm consists of two key stages: first, the number of candidate variables is gradually reduced based on an exponentially decaying function (EDF), and then adaptive reweighted sampling (ARS) is used to retain important variables with large absolute values of regression coefficients. In each Monte Carlo sampling iteration, the algorithm calculates the regression coefficient of each variable and competitively selects the variables based on the absolute value of the coefficient to ultimately obtain the optimal variable subset.

[0061] Uninformative Variable Elimination (UVE) is a variable screening method based on statistical analysis. It eliminates redundant or uninformative spectral variables by evaluating the contribution of each variable to the model, thereby improving the predictive performance and robustness of the model. Stability is calculated using the standard deviation of the mean vector and coefficient matrix.

[0062] The Successive Projections Algorithm (SPA) is a forward-iterative feature variable selection method primarily used to address multicollinearity in spectral analysis. It effectively extracts the most representative spectral features by gradually constructing an optimal subset of variables.

[0063] Variable Importance Projection (VIP) is a variable screening method based on the partial least squares algorithm. It is primarily used to assess the importance of each variable in a dataset in explaining the target variable. This method assesses variable importance by calculating the projected contribution of each variable in principal component analysis or factor analysis.

[0064] The Iteratively Retaining Informative Variables (IRIV) algorithm is a variable selection method that combines iterative optimization with variable importance assessment. This method uses a recursive "retain-remove" bidirectional screening mechanism to gradually optimize a subset of feature variables. Its core concept is to consider both the information contribution and redundancy of variables in each iteration, achieving feature selection by dynamically evaluating the discriminative power of variables.

[0065] First, the collected terahertz transmission spectra were divided into a calibration set and a prediction set using the KS algorithm in a ratio of approximately 3:1. Feature extraction was then performed using five algorithms: an iterative information-retaining variable algorithm, an uninformative variable elimination method, a continuous projection algorithm, a variable importance projection method, and a competitive adaptive reweighting algorithm. Finally, these features were input into the PLS model. Table 2 shows the validation results of the full-concentration PLS model for the THz spectrum of aflatoxin B1 solution using different feature extraction methods.

[0066] Table 2 Verification results of different feature extraction methods for the PLS model of THz spectrum of aflatoxin B1 solution

[0067]

[0068] The model prediction correlation coefficient R after feature extraction by UVE and CARS algorithms P The PLS model after CARS variable extraction has the highest accuracy, and its prediction correlation coefficient R P The root mean square error (RMSEP) reaches 5.68×10 -8 .

[0069] The aflatoxin B1 solution content detection model is composed of the related coefficient of prediction set (RP), the root mean square error of prediction set (RMSEP), the related coefficient of correction set (R C ), root mean square error of correction set (RMSEC) was used to evaluate the model, where R P and R C Between 0 and 1, R P and R C The closer it is to 1, the higher the accuracy of the model. The closer and smaller the values of RMSEP and RMSEC are, the more stable the model is.

[0070] The limit of detection (LOD) is the minimum concentration at which a substance can be detected. The LOD can be calculated with a 99.86% confidence interval based on the slope of the true-to-predicted curve and the prediction error variance established by the THz transmission spectrum.

[0071]

[0072] Where σ is the standard error of the predicted concentration, m is the slope of the fitting curve, and the RMSEP in the model is equal to the σ value.

[0073] (4) A quantitative detection model of aflatoxin B1 terahertz transmission spectroscopy was established based on the extracted features and LS-SVM.

[0074] Partial Least Squares Regression (PLSR) is a multivariate statistical modeling method that integrates the advantages of multiple linear regression, canonical correlation analysis and principal component analysis.

[0075] The Least Squares Support Vector Machine (LS-SVM) is a machine learning algorithm based on the standard support vector machine (SVM). Its core feature is that it transforms inequality constraints in the SVM into equality constraints, thereby simplifying the quadratic programming problem to solving a system of linear equations. Its key parameters are the input vector, the type of kernel function, and its corresponding parameters. Commonly used kernel functions are the linear kernel (Lin-Kernel) and the radial basis kernel (RBF-Kernel).

[0076] The least squares support vector machine (LS-SVM) is a machine learning method developed based on statistical theory. Based on the above research, the KS algorithm was used to divide the samples into a calibration set and a prediction set at a ratio of about 3:1. Subsequently, the five algorithms IRIV, UVE, SPA, VIP and CARS were used for feature extraction, and finally input into the LS-SVM model respectively. Table 3 shows the verification results of the mixed LS-SVM model of THz spectrum of aflatoxin B1 solution by different feature extraction methods. The accuracy of the LS-SVM model using the RBF basis function is better than that of the Lin basis function. In the absence of feature extraction, the effect is the best, and the prediction correlation coefficient R of the model is 0. P The prediction root mean square error (RMSEP) reached 5.30×10 -8 , at this time its parameter combination is γ=4.982107、σ 2 =332.5579.

[0077] Table 3 Verification results of LS-SVM model for THz spectrum of aflatoxin B1 solution using different feature extraction methods

[0078]

[0079] Figure 9 The following is a fitting diagram of the predicted values and true values of the THz spectrum of aflatoxin B1 solution using the PLS and LS-SVM models. The results show that the LS-SVM model performs better than the PLS model overall. The LS-SVM model based on the RBF kernel function has the highest prediction accuracy, with a prediction correlation coefficient (RP) of 0.920 and a root mean square error (RMSEP) of 5.30×10 -8 The slope of the fitting curve reached 0.7787. According to the slope of the fitting curve of the predicted value and the predicted root mean square error variance, the detection limit was calculated to be 2.04×10 -7 μg / ml.

[0080] Based on electromagnetic theory, this paper designs a BIC terahertz metamaterial sensor. Preliminary experiments determine the metamaterial sheet corresponding to the optimal polarization angle and perturbation factor. Based on this, quantitative detection of aflatoxin B1 was carried out, verifying the feasibility of highly sensitive THz-TDS detection of aflatoxin B1 solutions using a metamaterial sensor. Analysis revealed that the amplitude of the terahertz transmission peak gradually decreases with increasing aflatoxin B1 solution concentration, and the frequency of the transmission peak gradually shifts toward lower frequencies. A quantitative model for aflatoxin B1 solutions was established, resulting in a detection limit of 2.04×10 -7 μg / ml, which provides a theoretical basis for the high-sensitivity detection of aflatoxin B1 solution by terahertz metamaterial sensors combined with THz-TDS technology.

[0081] (5) Obtaining terahertz spectrum information of the sample containing aflatoxin B1 to be tested and inputting it into the quantitative detection model to obtain the content of aflatoxin B1 in the sample to be tested.

[0082] In summary, the present invention provides a method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor. A continuous domain bound state (BIC) terahertz metamaterial sensor was designed based on electromagnetic theory and the finite-difference time-domain method (FDTD). Using this BIC terahertz metamaterial sensor, terahertz-enhanced spectra of aflatoxin B1 solutions of varying concentrations were collected. Analysis revealed that the amplitude of the transmission peak of aflatoxin B1 at approximately 1.5 THz gradually decreased with increasing aflatoxin B1 solution concentration, and the frequency of the transmission peak gradually shifted toward lower frequencies. PLS and LS-SVM quantitative models for aflatoxin were established, and analysis revealed that the LS-SVM model performed better, resulting in a detection limit of 2.04×10 -7 μg / ml. This application verifies the feasibility of highly sensitive detection of aflatoxin B1 solutions using a BIC terahertz metamaterial sensor combined with THz-TDS technology. The present invention utilizes continuum bound state technology to enhance the electromagnetic response of aflatoxin B1 to terahertz waves, ultimately increasing detection sensitivity. This invention enables rapid, nondestructive, and highly sensitive detection of aflatoxin B1 in grain.

[0083] The above-described embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

Claims

1. A method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor, characterized in that: The BIC terahertz metamaterial sensor uses a silicon substrate with a gold film on the surface. The BIC terahertz metamaterial sensor is composed of multiple periodic structural units. The periodic structural unit is a square structure. The width p of the periodic structural unit is 48 μm. A rectangular non-gold-plated area is provided on the periodic structural unit. The width W of the rectangular non-gold-plated area is 4.5 μm. The rectangular non-gold-plated area satisfies the conditional formula: I = α*p, where α is an influence factor and I is the length of the rectangular non-gold-plated area. The detection method comprises the following steps: (1) Preparing multiple aflatoxin B1 standard solutions of different concentrations, dropping the aflatoxin B1 standard solutions onto the BIC terahertz metamaterial sensor, and placing the sensor in a drying oven for drying; (2) Place the dried BIC terahertz metamaterial sensor into the terahertz system and use the transmission mode for measurement to collect the terahertz spectrum information of the corresponding sample; (3) Using spectral feature variable extraction algorithm to extract features of terahertz spectrum information; (4) Establish a quantitative detection model of aflatoxin B1 terahertz transmission spectrum based on the extracted features and LS-SVM; (5) Obtaining terahertz spectrum information of the sample containing aflatoxin B1 to be tested and inputting it into the quantitative detection model to obtain the content of aflatoxin B1 in the sample to be tested.

2. The method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor according to claim 1, characterized in that: The BIC terahertz metamaterial sensor is prepared by the following preparation method: The structural parameters of the BIC terahertz metamaterial sensor were designed using the finite-difference time-domain method. By introducing the BIC mechanism into the terahertz metasurface, high-Q resonance was achieved. A silicon substrate was selected, and the surface was cleaned to remove impurities and oxidized to enhance the adhesion of the photoresist. Subsequently, a chromium layer is plated on the silicon substrate using thin film deposition technology, followed by the deposition of a metal layer to enhance the adhesion between the metal layer and the silicon substrate; Then, the silicon wafer is coated with a layer of photoresist with a thickness of 1-2 μm by using a spin coating process; After coating, a pre-bake is performed at 100°C for 120 seconds to remove the solvent from the photoresist, ensuring that it is hardened and ready for exposure; After exposure, the development process begins. The development time is controlled within 30 seconds to remove the unexposed part of the photoresist, leaving the pattern. After development, a solvent is used for a sizing process to remove the residual photoresist; Then, an etching operation is performed by adopting reactive ion etching or ion beam etching; After etching is completed, the photoresist mask is removed by acetone ultrasonic cleaning to ensure that there is no residue on the chip surface; finally, surface oxidation or metal deposition post-treatment is performed to enhance conductivity and corrosion resistance.

3. The method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor according to claim 1, characterized in that: The step (1) specifically includes: According to the preset ratio scheme, 6 aflatoxin B1 standard solutions with different concentration gradients were prepared by step-by-step dilution with a precision pipette. All samples were fully shaken on a vortex oscillator for 3 minutes. Multi-stage dilution operations were performed during the experiment. After the dilution was completed, the solutions were placed in volumetric flasks for storage.

4. The method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor according to claim 1, characterized in that: The step (2) specifically includes: Before the experiment, the system was preheated by an air compressor with an air dryer, and dry air was continuously filled into the optical cavity. The humidity was monitored by a hygrometer to keep it below 10% and the temperature constant at 25±0.5℃. During the experiment, wait for 2 minutes after each sample was placed, and start measurement after the system stabilized. A pipette was used to draw 20μl of sample solution in the order of sample concentration from low to high and drop it on the metamaterial sheet. After each sample test, the metamaterial sheet was placed in a beaker and shaken to clean it. Before collecting the sample spectrum, the metamaterial sheet with the sample was placed in a drying oven for 15 minutes, and the drying oven temperature was set to 50℃. Finally, the metamaterial sheet with the sample was placed in the terahertz system and measured using the transmission mode. Five points were taken for each sample, and 10 measurements were performed at each point. A total of 300 spectra were obtained for the six concentration samples, and each spectrum was the average of 64 repeated tests. The KS algorithm was used to divide the samples into a calibration set and a prediction set in a ratio of about 3:1 for subsequent model building.

5. The method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor according to claim 1, characterized in that: In step (5), the prediction set correlation coefficient, the prediction set root mean square error, the correction correlation coefficient, and the correction root mean square error are used to evaluate the established quantitative detection model of carbendazim terahertz transmission spectrum.

6. The method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor according to claim 1, characterized in that: The spectral feature variable extraction algorithm is any one of a competitive adaptive reweighted sampling algorithm, an uninformative variable elimination algorithm, a continuous projection algorithm, a variable projection importance analysis method, and an iterative information variable retention algorithm.

7. The method for detecting aflatoxin B1 based on a BIC terahertz metamaterial sensor according to claim 1, characterized in that: The value of α is 0.7.

Images