Preparation method of silicon-modified silver-based magnetic nano composite material and method for detecting organic arsonic acid in water body by applying silicon-modified silver-based magnetic nano composite material

Abstract

The invention discloses a preparation method of a silicon-modified silver-based magnetic nano composite material and a method for detecting organic arsonic acid in a water body by using the silicon-modified silver-based magnetic nano composite material, and aims to solve the problems of high cost and low precision of the existing organic arsonic acid detection method. The method for detecting the organic arsonic acid in the water body comprises the following steps: 1, collecting an environmental water sample or pure water; 2, adding a single organic arsonic acid or a mixture of a plurality of organic arsonic acids into an environmental water sample or pure water; 3, taking the silicon-modified silver-based magnetic nano composite material as a Raman detection substrate, dropwise adding a training water sample for Raman detection, and collecting a Raman spectrogram of the water sample; and 4, performing dimension reduction processing on the water sample Raman spectrogram data by adopting principal component analysis (PCA), training by adopting a support vector classification (SVM) model, and constructing a category classification model. According to the method, the Fe3O4 (at) SiO2 (at) Ag substrate is combined with a machine learning algorithm, so that the detection selectivity of SERS on similar molecules with high structural similarity in a complex environment is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the field of analysis and detection, and specifically relates to a method for preparing a silicon-modified silver-based magnetic nanocomposite material, and a method for detecting a variety of mixed organic arsenic acids in water using machine learning-assisted surface-enhanced Raman spectroscopy. Background Art

[0002] The methylation of inorganic arsenic by microorganisms and the widespread use of pesticides and veterinary drugs, such as the abuse of dimethylarsinic acid as a herbicide and defoliant, and the use of 3-nitro-4-hydroxyphenylarsonic acid (Roxarsone) as an animal feed additive into the animal body. Most of it cannot be absorbed and is eventually excreted from the body in the form of dimethylarsinic acid, spreading its toxicity throughout the environment and significantly increasing the risk of cancer, reproductive problems and cardiovascular diseases.

[0003] At present, the commonly used detection methods for organic arsenic acids are high performance liquid chromatography-atomic fluorescence spectrometry (HPLC-AFS) and high performance liquid chromatography-inductively coupled plasma (HPLC-ICP). Both methods can effectively detect organic arsenic acids in the environment. However, these technologies are expensive, often involve complex and time-consuming sample preparation, require sample pretreatment and professional instruments and labor, and cannot distinguish and detect multiple organic arsenic acids at the same time. These factors make these techniques unsuitable for in situ detection of organic arsenic acids. Given the low concentration and persistence of organic arsenic in the environment, there is an urgent need for effective, label-free, rapid and sensitive analytical methods to monitor these pollutants.

[0004] At present, the main methods for detecting organic arsenic acid are: high performance liquid chromatography and electrophoresis, etc. The steps of this kind of detection method are mostly complicated and time-consuming, requiring sample pretreatment and professional instruments and manpower, and it is impossible to distinguish and detect multiple organic arsenic acids at the same time. Surface enhanced Raman scattering (SERS) can identify specific compounds based on the spectral differences of chemical bond vibration modes in arsenic speciation analysis. At present, spectral analysis is mainly based on manual identification, and experienced personnel are required to identify characteristic peaks from SERS spectra, which leads to considerable uncertainty in the analysis results. Especially when there is a large amount of spectral data, manual analysis is time-consuming and error-prone, and it is difficult to generalize and make judgments. Summary of the invention

[0005] The purpose of the present invention is to solve the problems that the existing detection method of organic arsenic acid is high in cost and the surface enhanced Raman scattering spectrum analysis accuracy is not high, and to provide a preparation method of a silicon-modified silver-based magnetic nanocomposite material and the use of the same to detect a variety of mixed organic arsenic acids in water.

[0006] The preparation method of the silicon-modified silver-based magnetic nanocomposite material of the present invention is achieved by the following steps:

[0007] 1. Magnetic Fe 3 O 4 Synthesis of NPs:

[0008] Ferric chloride hexahydrate, sodium acetate and polyethylene glycol (PEG) were dissolved in ethylene glycol and then placed in a stainless steel autoclave lined with polytetrafluoroethylene for hydrothermal reaction at a temperature of 180-200°C. After washing and drying, magnetic Fe 3 O 4 Particles;

[0009] 2. Fe 3 O 4 @SiO 2 Synthesis of:

[0010] The magnetic Fe 3 O 4 The particles were dispersed in a mixed solution of anhydrous ethanol and water, and ultrasonically dispersed uniformly. Ammonia and tetraethoxysilane (TEOS) were added, and the mixture was stirred at room temperature for 2 to 5 hours. After washing and drying, Fe 3 O 4 @SiO 2 Material;

[0011] 3. Fe 3 O 4 @SiO 2 Synthesis of @Ag:

[0012] AgNO 3 Dissolve in deionized water containing ammonia, then add Fe 3 O 4 @SiO 2 The material is ultrasonically dispersed uniformly to obtain Fe 3 O 4 @SiO 2 The silver ammonia solution is prepared by dissolving polyvinyl pyrrolidone (PVP) in anhydrous ethanol to obtain a polyvinyl pyrrolidone solution, and then the polyvinyl pyrrolidone solution and the Fe 3 O 4 @SiO 2 The silicon-modified silver-based magnetic nanocomposite material is prepared by mixing the silicon-modified silver-based magnetic nanocomposite material with a silver ammonia solution, performing a solvent thermal reaction at a temperature of 100 to 140° C., and obtaining the silicon-modified silver-based magnetic nanocomposite material after washing and drying.

[0013] The method for detecting organic arsenic acid in water using a silicon-modified silver-based magnetic nanocomposite material is implemented by the following steps:

[0014] 1. Collect environmental water samples or pure water;

[0015] 2. Add a single organic arsenic acid to an environmental water sample or pure water to obtain a single organic arsenic acid water sample; add a plurality of organic arsenic acids to an environmental water sample or pure water to obtain a plurality of mixed organic arsenic acid water samples, and use the single organic arsenic acid water sample and the plurality of mixed organic arsenic acid water samples as training water samples;

[0016] 3. Using silicon-modified silver-based magnetic nanocomposite as a Raman detection substrate, placing the Raman detection substrate on a glass slide, dropping training water samples for Raman detection, and collecting the Raman spectrum of the water sample;

[0017] Fourth, the Raman spectrum of the water sample is denoised and smoothed, and the principal component analysis PCA (Principal Component Analysis) is used to reduce the dimension of the Raman spectrum data of the water sample. Then, the support vector classification SVM model is trained to construct a classification model for the types of organic arsenic acid in the water sample to complete the detection of organic arsenic acid in the water body.

[0018] The present invention first prepares a silicon-modified silver-based magnetic nanocomposite material, and utilizes the enrichment and Raman enhancement of the silicon-modified magnetic substrate on organic arsenic acid, so that the sensitivity and detection limit of the organic arsenic acid detection result are significantly improved compared with the traditional detection results. Secondly, the SERS spectral data is processed by a machine learning algorithm, which improves the selectivity of the SERS method when applied to the detection of multiple organic arsenic acids, and simultaneously identifies the type of a single organic arsenic acid or multiple mixed organic arsenic acids in the water body, solving the problem of using the SERS technology to detect complex mixed organic arsenic acids in water bodies.

[0019] The preparation method of the silicon-modified silver-based magnetic nanocomposite material of the present invention and its application in detecting organic arsenic acid in water bodies include the following beneficial effects:

[0020] 1. The present invention uses magnetic Fe 3 O 4 @SiO 2 Ag Raman enhanced substrate is used for detection, and its magnetic property is conducive to the enrichment of organic arsenic acid molecules. 2 The protective layer can shield the magnetic dipole interaction, prevent particle agglomeration, improve stability, and significantly increase the number of analyte molecules that can be captured on the surface of the substrate material. 2 The protective layer can provide more receptor sites for metal nanoparticles, so that dense nanogaps are formed when nanosilver grows, providing more strong hot spots, which is more conducive to the detection of organic arsenic acid.

[0021] 2. Fe in the present invention 3 O 4 @SiO 2@Ag substrate can be used as a Raman enhanced substrate to detect dimethylarsonic acid, roxarsone and phenylarsonic acid in complex real environments with a short detection time, which is of great significance for monitoring the residues of organic arsenic acid in water bodies.

[0022] 3. The present invention provides a sensitive label-free SERS method, which can 3 O 4 @SiO 2 The combination of @Ag substrate and machine learning algorithm greatly improves the detection selectivity of SERS for similar molecules with high structural similarity in complex environments, eliminating the subjective judgment of analytes using a single characteristic peak. The SVM model is used to identify different forms of mixed organic arsenic, and the accuracy of complex mixed arsenic acid forms in natural water bodies is as high as 99.1%. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 The present invention is a flow chart of a method for detecting organic arsenic acid in water using a silicon-modified silver-based magnetic nanocomposite material;

[0024] Figure 2 1 is the original and pre-treated SERS spectra of single and mixed organic arsenic acids in the actual water sample obtained in Example 1, wherein FIG. C is a three-dimensional PCA graph of the SERS spectra of three single organic arsenic acids and ternary mixed organic arsenic acids after pre-treatment after standardization;

[0025] Figure 3 The confusion matrix of the classification results of the SERS spectrum data after PCA dimensionality reduction in Example 2, including confusion matrix diagrams of the classification results of four organic arsenic acids: (A) PLS-DA classification, (B) RF classification, (C) KNN classification, and (D) SVM classification;

[0026] Figure 4 Fe in Comparative Example 1 3 O 4 @Ag and Fe in Example 1 3 O 4 @SiO 2 @SERS spectrum comparison of Ag substrate;

[0027] Figure 5 is the Fe of different organic arsenic acids in Comparative Example 1 3 O 4 @SiO 2 @SERS spectrum comparison of Ag substrate;

[0028] Figure 6 For the artificial identification of Fe in Comparative Example 2 3 O 4 @SiO 2Concentrations of dimethylarsonic acid and roxarsone in water samples with Ag substrate and 600 cm -1 -1750cm -1 The linear model obtained by Raman intensity at . DETAILED DESCRIPTION

[0029] Specific implementation method 1: The preparation method of the silicon-modified silver-based magnetic nanocomposite material in this implementation method is implemented according to the following steps:

[0030] 1. Magnetic Fe 3 O 4 Synthesis of NPs:

[0031] Ferric chloride hexahydrate, sodium acetate and polyethylene glycol (PEG) were dissolved in ethylene glycol and then placed in a stainless steel autoclave lined with polytetrafluoroethylene for hydrothermal reaction at a temperature of 180-200°C. After washing and drying, magnetic Fe 3 O 4 Particles;

[0032] 2. Fe 3 O 4 @SiO 2 Synthesis of:

[0033] The magnetic Fe 3 O 4 The particles were dispersed in a mixed solution of anhydrous ethanol and water, and ultrasonically dispersed uniformly. Ammonia and tetraethoxysilane (TEOS) were added, and the mixture was stirred at room temperature for 2 to 5 hours. After washing and drying, Fe 3 O 4 @SiO 2 Material;

[0034] 3. Fe 3 O 4 @SiO 2 Synthesis of @Ag:

[0035] AgNO 3 Dissolve in deionized water containing ammonia, then add Fe 3 O 4 @SiO 2 The material is ultrasonically dispersed uniformly to obtain Fe 3 O 4 @SiO 2 The silver ammonia solution is prepared by dissolving polyvinyl pyrrolidone (PVP) in anhydrous ethanol to obtain a polyvinyl pyrrolidone solution, and then the polyvinyl pyrrolidone solution and the Fe 3 O 4 @SiO 2The silicon-modified silver-based magnetic nanocomposite material is prepared by mixing the silicon-modified silver-based magnetic nanocomposite material with a silver ammonia solution, performing a solvent thermal reaction at a temperature of 100 to 140° C., and obtaining the silicon-modified silver-based magnetic nanocomposite material after washing and drying.

[0036] In step 1 of this embodiment, the amount of ethylene glycol used relative to the total mass of sodium acetate and ferric chloride hexahydrate is (5-15) mL / g.

[0037] Specific embodiment 2: This embodiment is different from specific embodiment 1 in that the mass ratio of sodium acetate to ferric chloride hexahydrate in step 1 is (1-3):1.

[0038] Specific implementation method three: This implementation method is different from specific implementation methods one or two in that in step one, the hydrothermal reaction is carried out at a temperature of 190° C. for 8 to 10 hours.

[0039] Specific embodiment 4: This embodiment is different from the specific embodiments 1 to 3 in that the relative magnetic Fe of tetraethoxysilane (TEOS) in step 2 3 O 4 The total mass of the particles is (1-4) mL / g.

[0040] This embodiment refers to Fe per gram 3 O 4 Add 1-4 mL of tetraethoxysilane to the particles.

[0041] Specific embodiment 5: This embodiment is different from specific embodiments 1 to 4 in that Fe 3 O 4 @SiO 2 Materials and AgNO 3 The mass ratio is (0.3~0.5):2.

[0042] Specific embodiment 6: This embodiment is different from specific embodiments 1 to 5 in that the solvothermal reaction is carried out at a temperature of 120° C. for 5 to 8 hours in step 3.

[0043] Specific implementation method seven: This implementation method uses silicon-modified silver-based magnetic nanocomposite materials to detect a variety of mixed organic arsenic acids in water bodies according to the following steps:

[0044] 1. Collect environmental water samples or pure water;

[0045] 2. Add a single organic arsenic acid to an environmental water sample or pure water to obtain a single organic arsenic acid water sample; add a plurality of organic arsenic acids to an environmental water sample or pure water to obtain a plurality of mixed organic arsenic acid water samples, and use the single organic arsenic acid water sample and the plurality of mixed organic arsenic acid water samples as training water samples;

[0046] 3. Using silicon-modified silver-based magnetic nanocomposite as a Raman detection substrate, placing the Raman detection substrate on a glass slide, dropping training water samples for Raman detection, and collecting the Raman spectrum of the water sample;

[0047] Fourth, the Raman spectrum of the water sample is denoised and smoothed, and the principal component analysis PCA (Principal Component Analysis) is used to reduce the dimension of the Raman spectrum data of the water sample. Then, the support vector classification SVM model is trained to construct a classification model for the types of organic arsenic acid in the water sample to complete the detection of organic arsenic acid in the water body.

[0048] The SERS spectral data preprocessing method used in the machine learning algorithm classification model construction in step 4 of this embodiment is to use adaptive iterative reweighted penalized least squares (airPLS) baseline correction, Savitzky-Golay (SG) filter smoothing and standard normal variable (SNV) for preprocessing; PCA is used to retain most of the original information while reducing the dimension of the data to simplify the data. Step 4 detects the types of organic arsenic acid in water samples through the classification model of organic arsenic acid in water samples.

[0049] Specific implementation eight: This implementation differs from specific implementation seven in that the organic arsenic acid described in step two is one or more of dimethylarsenic acid, 3-nitro-4-hydroxyphenylarsenic acid, and phenylarsenic acid.

[0050] Specific implementation method 9: The difference between this implementation method and specific implementation method 8 or 9 is that in step 3, the wavelength of the excitation light source for controlling Raman detection is 532 nm, the integration time is 5 s, and the laser power is 5 mW.

[0051] This embodiment optimizes the integration time. In Raman spectroscopy detection, the integration time refers to the length of time for collecting spectral signals. The integration time has an important influence on the signal intensity and signal-to-noise ratio of the Raman spectrum. The integration time is an important parameter in Raman spectroscopy detection and is optimized according to the properties of the sample and experimental requirements. A reasonable integration time can improve the signal intensity and signal-to-noise ratio of the spectrum.

[0052] Specific implementation method ten: The difference between this implementation method and specific implementation methods eight to ten is that the kernel function in the support vector classification SVM model in step four is linear.

[0053] Embodiment 1: The preparation method of the silicon-modified silver-based magnetic nanocomposite material of this embodiment is implemented according to the following steps:

[0054] 1. Magnetic Fe 3 O 4 Synthesis of NPs:

[0055] 2.1 g FeCl3 6H 2 O, 5.5 g sodium acetate and 1.5 g polyethylene glycol (PEG) were dissolved in 60 mL ethylene glycol, and then placed in a stainless steel autoclave with a polytetrafluoroethylene liner, and subjected to hydrothermal reaction at 190 ° C for 8 h, washed with water and anhydrous ethanol three times in sequence, and vacuum dried at 60 ° C for 6 h to obtain magnetic Fe 3 O 4 Particles;

[0056] 2. Fe 3 O 4 @SiO 2 Synthesis of:

[0057] 0.2g of magnetic Fe 3 O 4 The particles were dispersed in a mixed solution of 100 mL of anhydrous ethanol and 25 mL of water, and ultrasonically dispersed uniformly. 1.8 mL of 25% ammonia water and 0.6 mL of tetraethoxysilane (TEOS) were added, and the mixture was stirred at room temperature for 3 h. The mixture was washed with anhydrous ethanol for 3 times and vacuum dried at 60 ° C for 6 h to obtain Fe 3 O 4 @SiO 2 Material;

[0058] 3. Fe 3 O 4 @SiO 2 Synthesis of @Ag:

[0059] 0.2 g of AgNO 3 Dissolve in 4 mL of deionized water containing 400 μL of ammonia (mass concentration is 25%), and then add 50 mg of Fe 3 O 4 @SiO 2 The material was ultrasonically dispersed for 30 minutes to obtain Fe 3 O 4 @SiO 2 silver ammonia solution, 0.5g polyvinyl pyrrolidone (PVP) was dissolved in 26mL anhydrous ethanol to obtain a polyvinyl pyrrolidone solution, and then the polyvinyl pyrrolidone solution and the Fe 3 O 4 @SiO 2 The silver-ammonia solution was mixed for 30 minutes, and the solvent thermal reaction was carried out at a temperature of 120°C for 6 hours. After washing with water and anhydrous ethanol for 3 times each, the silicon-modified silver-based magnetic nanocomposite material (Fe 3 O 4 @SiO 2 @Ag).

[0060] Application Example: The method of using silicon-modified silver-based magnetic nanocomposite materials to detect organic arsenic acid in water is implemented according to the following steps:

[0061] 1. Collect environmental water samples and filter them through a 0.22μm filter membrane;

[0062] 2. Adding dimethylarsinic acid, 3-nitro-4-hydroxyphenylarsinic acid and phenylarsinic acid to the water sample respectively to obtain a single organic arsenic acid water sample; adding dimethylarsinic acid, 3-nitro-4-hydroxyphenylarsinic acid and phenylarsinic acid to the water sample at a molar ratio of 1:1:1 to obtain a multi-organic arsenic acid water sample;

[0063] 3. Using silicon-modified silver-based magnetic nanocomposite as the Raman detection substrate, 0.01 g of the Raman detection substrate was placed on a glass slide, 20 μL of the training water sample was added for Raman detection, and the Raman spectrum of the water sample was collected;

[0064] The spectrum acquisition conditions of this embodiment are as follows: using 532nm laser as the excitation source, scanning spectrum range of 200-2000cm-1, acquisition grating of 1200gr / mm, laser power of 5mW, acquisition time of 5s / time, accumulation times of 1 time, and collecting 100 spectra for different types of samples respectively;

[0065] 4. Remove background noise from the Raman spectrum of water samples, use adaptive iterative reweighted penalized least squares (airPLS) baseline correction, Savitzky-Golay filter smoothing, standard normal variable (SNV) processing, and use principal component analysis PCA (Principal Component Analysis) in SCIMA software to reduce the dimension of the water sample Raman spectrum data. Principal component analysis PCA uses the similarities and differences between spectra to map the original data to a new coordinate system using linear transformation. The coordinate with the largest variance is the first principal component PC1. Similarly, the coordinate corresponding to the second largest variance is the second principal component PC2, and the second principal component PC2 is orthogonal to PC1. The sum of the variances of all principal components is equal to 1. Select the three components PC1, PC2, and PC3 with the highest contribution rate to create a visualization model, such as Figure 2 C. By mapping high-dimensional data to low-dimensional space, the complexity of the data is reduced while retaining the main information of the data. At the same time, the correlation between features is eliminated and the redundant information in the data is reduced.

[0066] 5. Use the Raman spectrum of water samples to set up training data sets and test data sets, where the training data set is 70% of the collected data and the test data set is 30% of the collected data. Use a machine learning classifier model to classify the types of organic arsenic acid in water samples, thereby detecting organic arsenic acid in water bodies.

[0067] The machine learning classifier models used in step five of this embodiment are least squares discriminant analysis (PLS-DA), random forest (RF), K nearest neighbor (KNeighbors) and support vector classification (SVM) models. The parameters of the least squares discriminant analysis (PLS-DA) machine learning algorithm are: n_components is the number of sample categories, and the number of categories in this embodiment is 3; the parameters of the random forest (RF) machine learning algorithm are: n_estimators is the number of decision numbers, and the number of categories in this embodiment is 100. The parameters of the K nearest neighbor (KNN) machine learning algorithm are: n_neighbors is the number of sample categories, the number of categories in this embodiment is 5, and algorithm is auto; the parameters of the support vector machine (SVM) machine learning algorithm are: the kernel function is linear or RBF function, preferably linear function, and the penalty coefficient C of the objective function is 1000;

[0068] This embodiment uses different models of least squares discriminant analysis (PLS-DA), random forest (RF), K nearest neighbor (KNeigh bors) and support vector classification (SVM) models, and predicts the test data set after training. The accuracy is obtained according to the actual results and the predicted results. The model with the highest accuracy is the best model, where the accuracy is calculated as the proportion of the correct number of model predictions to the total number.

[0069] like Figure 3 The results showed that the accuracy of the least squares discriminant analysis (PLS-DA) algorithm was 84%, the accuracy of the random forest (RF) algorithm was 96%, the accuracy of the K nearest neighbor (KNeighbors) algorithm was 95%, and the accuracy of the SVM algorithm was 99.1%. Therefore, SVM was selected as the optimal model classifier SVM to identify the different forms of organic arsenic acid in lake water.

[0070] Comparative Example 1

[0071] Referring to Example 1, the difference between this example and Example 1 is that step 2 is not performed to prepare the corresponding Fe 3 O 4 @Ag substrate.

[0072] Referring to Example 1, the stirring time in step 2 was changed to 1 h, and the other conditions remained unchanged to obtain the corresponding Fe 3 O 4 @SiO 2 -2@Ag substrate.

[0073] Referring to Example 1, the stirring time in step 2 was changed to 5 h, and the other conditions remained unchanged to obtain the corresponding Fe 3 O 4@SiO 2 -3@Ag substrate.

[0074] The above-mentioned different substrates were subjected to SERS spectrum detection according to Example 2, and the corresponding spectrum data are as follows Figure 4 and Figure 5 shown.

[0075] It can be seen that in Fe 3 O 4 The Raman spectrum measured on the surface of the Ag substrate did not enhance the Raman intensity of the characteristic peaks of different organic arsenic acids. 3 O 4 Surface coating with silica is essential for effectively enhancing the Raman signal of the characteristic peak of organic arsenic acid. 2 The protective layer precursor is adjusted to prepare different substrates with different effects on the enhanced performance. Comparison of three substrates detected 1.33×10 -7 The SERS spectra of dimethylarsonic acid and roxarsone aqueous solutions of M showed that the silica coating not only effectively prevented the aggregation and chemical degradation of Fe3O4 particles in harsh liquid environments, but also provided a new source of Ag(NH 3 ) 2 ] + Ions adsorbed on Fe 3 O 4 @SiO 2 The NPs microspheres provide more sites, allowing silver ions to be further reduced by PVP into silver nanoparticles with uniform morphology and size. The number of "hot spots" depends on the size of the gaps between the silver nanoparticles. The smaller the gap, the more conducive to electromagnetic enhancement. The main characteristic peaks of dimethylarsonic acid and roxarsone are enhanced, and the intensity is much higher than that of the other two Fe 3 O 4 @SiO 2 @Ag substrate.

[0076] Comparative Example 2

[0077] Referring to the application example, the sample solution in step 2 is a single dimethylarsonic acid solution and a single roxarsone solution, with a concentration of 6.65×10 -8 M-1.33×10 -5 M.

[0078] Referring to the example, the sample solution is mixed with Fe 3 O 4 @SiO 2 The test sample liquid was mixed with Ag and then Raman detection was performed. A 532nm laser was used as the excitation source and the scanning spectrum range was 200-2000cm -1Each spectrum was exposed for 5s with a grating of 1200gr / mm and a power of 10%w, and accumulated 3 times. The surface enhanced Raman spectra of the organic arsenic acid solution were collected to determine the characteristic peak positions of two organic arsenic acids: dimethylarsonic acid and roxarsone;

[0079] Referring to the application example, the Raman spectrum data is subjected to cosmic ray elimination, background noise removal, baseline correction, and Savitzky-Golay smoothing using the 600 cm -1 -1750cm -1 A linear model was constructed by combining the characteristic peak at the position with the concentration of organic arsenic acid;

[0080] It can be seen that the use of artificial identification of SERS spectra (such as Figure 6 E and Figure 6 G), found at 666 cm -1 、803cm -1 The signal peaks at 617cm-1 are attributed to the stretching vibration of As-C and As-O, which are the unique Raman peak signals of DMA. -1 The signal peak of As-C stretching vibration appears at 797 cm-1, and the signal peak of As-O bending vibration appears at 797 cm-1. -1 1314cm -1 The signal peaks at 666cm are the symmetrical and asymmetrical stretching vibrations of the nitro group, both of which are unique Raman peak signals of ROX. -1 The linear relationship between the SERS intensity and concentration of As-C at R 2 =0.978, the detection limit LOD is 5.32×10 -8 M. According to ROX's 617cm -1 The linear relationship between the SERS intensity and concentration of As-C at R 2 =0.946, the detection limit LOD is 7.89×10 -9 Manual identification of characteristic peaks is subjective and has considerable uncertainty. The use of machine learning algorithms for accurate classification greatly makes up for the time-consuming and error-prone manual analysis of a large number of spectra.

[0081] The embodiments provided above are not intended to limit the scope of the present invention, and the steps described are not intended to limit the execution order thereof. Those skilled in the art may make obvious improvements to the present invention in combination with existing common knowledge, which also fall within the scope of protection defined by the claims of the present invention.

Claims

1. A method for preparing a silicon-modified silver-based magnetic nanocomposite material, characterized in that The preparation method of the silicon-modified silver-based magnetic nanocomposite material is achieved by the following steps:

1. Synthesis of magnetic Fe3O4 NPs: Ferric chloride hexahydrate, sodium acetate and polyethylene glycol are dissolved in ethylene glycol, and then placed in a stainless steel autoclave with a polytetrafluoroethylene liner, and subjected to a hydrothermal reaction at a temperature of 180 to 200° C., and magnetic Fe3O4 particles are obtained after washing and drying; 2. Synthesis of Fe3O4@SiO2: The magnetic Fe3O4 particles are dispersed in a mixed solution of anhydrous ethanol and water, and ultrasonically dispersed uniformly. Ammonia water and tetraethoxysilane are added, and the mixture is stirred and reacted at room temperature for 2 to 5 hours. After washing and drying, the Fe3O4@SiO2 material is obtained.

3. Synthesis of Fe3O4@SiO2@Ag: AgNO3 is dissolved in deionized water containing ammonia, and then Fe3O4@SiO2 material is added and ultrasonically dispersed uniformly to obtain a silver-ammonia solution containing Fe3O4@SiO2, polyvinyl pyrrolidone is dissolved in anhydrous ethanol to obtain a polyvinyl pyrrolidone solution, and then the polyvinyl pyrrolidone solution and the silver-ammonia solution containing Fe3O4@SiO2 are mixed, and a solvent thermal reaction is carried out at a temperature of 100 to 140°C. After washing and drying, a silicon-modified silver-based magnetic nanocomposite material is obtained.

2. The method for preparing the silicon-modified silver-based magnetic nanocomposite material according to claim 1, characterized in that In step 1, the mass ratio of sodium acetate to ferric chloride hexahydrate is (1-3):

1.

3. The method for preparing the silicon-modified silver-based magnetic nanocomposite material according to claim 1, characterized in that In step 1, a hydrothermal reaction is carried out at a temperature of 190° C. for 8 to 10 hours.

4. The method for preparing the silicon-modified silver-based magnetic nanocomposite material according to claim 1, characterized in that In step 2, the amount of tetraethoxysilane used relative to the total mass of the magnetic Fe3O4 particles is (1-4) mL / g.

5. The method for preparing the silicon-modified silver-based magnetic nanocomposite material according to claim 1, characterized in that In step three, the mass ratio of Fe3O4@SiO2 material and AgNO3 is (0.3~0.5):

2.

6. The method for preparing the silicon-modified silver-based magnetic nanocomposite material according to claim 1, characterized in that In step 3, a solvothermal reaction is carried out at a temperature of 120° C. for 5 to 8 hours.

7. A method for detecting organic arsenic acid in water using the silicon-modified silver-based magnetic nanocomposite material as claimed in claim 1, characterized in that The method for detecting organic arsenic acid in water is implemented according to the following steps:

1. Collect environmental water samples or pure water; 2. Add a single organic arsenic acid to an environmental water sample or pure water to obtain a single organic arsenic acid water sample; add a plurality of organic arsenic acids to an environmental water sample or pure water to obtain a plurality of mixed organic arsenic acid water samples, and use the single organic arsenic acid water sample and the plurality of mixed organic arsenic acid water samples as training water samples; 3. Using silicon-modified silver-based magnetic nanocomposite as a Raman detection substrate, placing the Raman detection substrate on a glass slide, dropping training water samples for Raman detection, and collecting the Raman spectrum of the water sample; Fourth, the Raman spectrum of the water sample is denoised and smoothed, and the principal component analysis (PCA) is used to reduce the dimension of the Raman spectrum data of the water sample. Then, the support vector classification (SVM) model is trained to construct a classification model for the types of organic arsenic acid in the water sample to complete the detection of organic arsenic acid in the water body.

8. The method for detecting organic arsenic acid in water using silicon-modified silver-based magnetic nanocomposite materials according to claim 7, characterized in that The organic arsenic acid described in step 2 is one or more of dimethylarsenic acid, 3-nitro-4-hydroxyphenylarsenic acid, and phenylarsenic acid.

9. The method for detecting organic arsenic acid in water using silicon-modified silver-based magnetic nanocomposite materials according to claim 7, characterized in that In step 3, the wavelength of the excitation light source for Raman detection is controlled to be 532 nm, the integration time is 5 s, and the laser power is 5 mW.

10. The method for detecting organic arsenic acid in water using silicon-modified silver-based magnetic nanocomposite materials according to claim 7, characterized in that The kernel function in the support vector classification SVM model in step 4 is linear.

Images