Flexible armored super-hydrophobic film, preparation method and application

Abstract

The invention discloses a flexible armored super-hydrophobic film, a preparation method and application. Relates to the technical field of metal element detection. According to the invention, improvement is carried out on the basis of the prior art, the flexible armored super-hydrophobic film is used as a solid-liquid conversion material, the problem of unstable detection caused by damage of a super-hydrophobic surface in the prior art is solved through the mechanical reliability and durability of the flexible armored super-hydrophobic film, the flexible armored super-hydrophobic film can be stably used for a long time, and the detection accuracy is improved. A complicated processing technology is not needed; according to the scheme, a liquid-phase sample is converted into a solid-phase sample based on a flexible armored super-hydrophobic film, and the solid-phase sample is detected by an LIBS technology to obtain an original element spectrum; establishing a quantitative model of the target metal element, wherein the quantitative model is used for performing data processing on the original element spectrum to obtain a target metal element detection result; on-site rapid detection of a liquid sample is realized based on the flexible armored super-hydrophobic film, equipment operation is simple, and complex pretreatment is not needed.

Description

Technical Field

[0001] The present invention relates to the technical field of metal element detection, and in particular to a flexible armored super-hydrophobic film, a preparation method and an application thereof. Background Art

[0002] Under the influence of industrial and agricultural activities, metals such as strontium, beryllium, manganese and barium continue to accumulate in surface water, posing a major threat to ecosystems and human health. Therefore, rapid and accurate methods for detecting these metal elements are essential to strengthen environmental regulation and protect human health. Traditional techniques for detecting metal elements in liquids mainly include inductively coupled plasma mass spectrometry (ICP-MS), inductively coupled plasma optical emission spectroscopy (ICP-OES) and microwave plasma torch (MPT). However, these methods usually require complex sample pretreatment steps and require bulky instruments and equipment, which limits their application in rapid on-site analysis.

[0003] Laser Induced Breakdown Spectroscopy (LIBS) is an element detection method that focuses a pulsed laser beam on the sample surface to form a laser-induced plasma containing the elements that make up the sample, and ultimately achieves qualitative and quantitative element detection through spectral analysis. It has unique advantages such as miniaturization of the instrument, simple pretreatment, and rapid on-site analysis. It is very suitable for rapid detection of trace metal elements and has been widely used in the detection of metal elements in solid matrices.

[0004] Compared with the detection of solid matrices, laser induced breakdown spectroscopy technology faces greater challenges when detecting and analyzing liquid matrices. Since the oscillation and splashing of the liquid surface will affect the generation and stability of plasma, the sensitivity and stability of laser induced breakdown spectroscopy technology for detecting liquid samples are not as good as those for solid samples. Its detection limit (LOD) is one to two orders of magnitude higher (mg / L) than that of traditional spectroscopic methods, and it is often difficult to reach the limit values ​​of metal elements specified in the sanitary standards for drinking water. Summary of the invention

[0005] The technical problem to be solved by the present invention is that when the traditional laser induced breakdown spectroscopy is used for liquid sample detection, its sensitivity and stability are low, and its detection limit is difficult to reach the limit value of metal elements specified in the sanitary standard of drinking water; the purpose of the present invention is to provide a flexible armored super-hydrophobic film, a preparation method and application, improve on the basis of the prior art, provide a new matrix conversion material and a preparation method thereof, on the basis of the matrix conversion technology, use the flexible armored super-hydrophobic film as the matrix conversion material, through the mechanical reliability and durability of the flexible armored super-hydrophobic film, solve the problem of unstable detection caused by damage to the super-hydrophobic surface in the prior art; at the same time, the flexible armored super-hydrophobic film can be used stably for a long time without complicated processing technology, based on the flexible armored super-hydrophobic film can realize on-site rapid detection of liquid samples, the equipment operation is simple, and no complicated pre-treatment is required.

[0006] The present invention is achieved through the following technical solutions:

[0007] The present invention provides a flexible armored super-hydrophobic film for matrix conversion, comprising hydrophobic particles and a film; the hydrophobic particles are used to adhere to the surface of the film to form a super-hydrophobic layer; the film has pockets of 10-50 μm.

[0008] A further optimization scheme is that the diameter of the hydrophobic particles is 1-100 nm, and the thickness of the flexible armored super-hydrophobic film is 50-500 μm.

[0009] A further optimized solution is that the material of the film is made of at least one of the following materials: PC, PET; and the hydrophobic particles are hydrophobic nano-silicon dioxide or hydrophobic nano-titanium dioxide.

[0010] This solution also provides a method for preparing the flexible armored super-hydrophobic film described in the above solution, comprising:

[0011] Wipe the film with micron-sized pockets and the transparent carrier substrate clean and dry;

[0012] Evenly sticking the film on a transparent supporting substrate;

[0013] 100-300 μL of super-hydrophobic nanoparticles are attached to the surface of the film, and a flexible armored super-hydrophobic film is formed after drying.

[0014] A further optimized solution is that the method of attaching 100-300 μL of super-hydrophobic nanoparticles to the surface of the film includes: spraying, coating, spin coating, or vapor deposition.

[0015] This solution also provides the application of the flexible armored super-hydrophobic film described in the above solution to realize liquid sample analysis based on LIBS technology; the method for liquid sample analysis based on LIBS technology includes:

[0016] Prepared a flexible armored super-hydrophobic film;

[0017] The liquid phase sample is converted into a solid phase sample based on a flexible armored super-hydrophobic film, and the solid phase sample is detected based on the LIBS technology to obtain the original element spectrum;

[0018] A quantitative model of the target metal element is established to process the original element spectrum to obtain the target metal element detection result.

[0019] A further optimization scheme is to convert the liquid phase sample into a solid phase sample based on a flexible armored super-hydrophobic film, and detect the solid phase sample based on the LIBS technology to obtain the original element spectrum; including the following methods:

[0020] Dropping a liquid sample onto the flexible armored super-hydrophobic film, and forming a detection point after the liquid sample is dried;

[0021] The detection point is placed at the laser focus of the LIBS spectrometer for detection to obtain the original element spectrum of the detection point.

[0022] A further optimization scheme is to establish a quantitative model of the target metal element for performing data processing on the original element spectrum to obtain the target metal element detection result; including the method:

[0023] Confirm the spectral range of the target metal element according to the spectral database, and extract the target spectral data from the original element spectrum based on the spectral range;

[0024] Preprocessing the target spectral data, and performing principal component analysis on the preprocessed target spectral data to extract characteristic data;

[0025] The characteristic data is quantitatively analyzed based on the XGBoost machine learning algorithm to obtain a quantitative model of the target metal element concentration and spectral intensity, thereby obtaining the detection result of the target metal element in the liquid sample.

[0026] A further optimization scheme is that the target spectral data is preprocessed, and the preprocessed target spectral data is subjected to principal component analysis to extract characteristic data; including the method:

[0027] The target spectral data is normalized and linearly transformed to [0–1]. The normalized target spectral data is baseline corrected based on the fourth-order polynomial fitting method, and then the average spectrum of the target spectral data after baseline correction is used as the basic spectrum for multivariate scattering correction.

[0028] The preprocessed target spectral data are divided into a training set, a test set and a validation set, and characteristic data are extracted from the target spectral data based on a principal component analysis method, wherein the principal component analysis method comprises the following steps in sequence: calculating a covariance matrix, eigenvalue decomposition, and extracting principal components; wherein the principal component analysis method extracts principal components based on a cumulative variance contribution rate method, ensuring that the total variance of the first k principal components accounts for ≥95% of the total variance of the original data.

[0029] A further optimization scheme is that the XGBoost machine learning algorithm uses the average value of all samples as the initial model. In each round of iteration, the residual of the current model is first calculated, and a new decision tree is trained with the residual as the target variable to fit the residual, and the output of the new decision tree is superimposed on the current model for updating the predicted value; the XGBoost machine learning algorithm finds the optimal decision tree structure and parameter configuration by minimizing the objective function, and a regularization term is added to the objective function.

[0030] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0031] 1. The present invention provides a flexible armor super-hydrophobic film, a preparation method and an application thereof; based on the prior art, the invention makes improvements by using the flexible armor super-hydrophobic film as a matrix conversion material, and solves the problem of unstable detection caused by damage to the super-hydrophobic surface in the prior art through the mechanical reliability and durability of the flexible armor super-hydrophobic film; at the same time, the flexible armor super-hydrophobic film can be used stably for a long time without complicated processing technology;

[0032] 2. A flexible armored super-hydrophobic film, preparation method and application provided by the present invention; using the flexible armored super-hydrophobic film as a matrix conversion material to achieve on-site rapid detection of liquid matrix, the equipment is simple to operate and does not require complex pre-treatment;

[0033] 3. A flexible armored super-hydrophobic film, preparation method and application provided by the present invention; based on the directional enrichment and matrix conversion function of the flexible armored super-hydrophobic film, the transmission of laser energy is optimized, the emission spectrum signal of the element is enhanced, thereby significantly improving the detection sensitivity;

[0034] 4. The present invention provides a flexible armored super-hydrophobic film, a preparation method and an application thereof; by combining the principal component analysis method and the XGBoost machine learning algorithm to establish a quantitative model, the problem of high uncertainty in single variable analysis prediction is effectively solved, and the detection accuracy is improved; rapid and reliable simultaneous detection of multiple metal elements in liquids is achieved, which is widely applicable to environmental monitoring, drinking water safety, food quality control and other fields. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the following briefly introduces the drawings required for use in the embodiments. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be regarded as limiting the scope. For those of ordinary skill in the art, other relevant drawings can be obtained based on these drawings without creative work. In the drawings:

[0036] Figure 1 The figure is a schematic diagram of the process for preparing a flexible armor super-hydrophobic film;

[0037] Figure 2 Schematic diagram of the principle of hydrophobic surface formation;

[0038] Figure 3 This is a scanning electron microscope image of the micrometer-scale pocket structure on the film surface;

[0039] Figure 4 A scanning electron micrograph of micrometer-sized pockets filled with superhydrophobic nanoparticles.

[0040] Figure 5 Schematic diagram of the mechanical robustness of the water contact angle of the flexible armored superhydrophobic film;

[0041] Figure 6 This is a schematic diagram of the application process of the flexible armor super-hydrophobic film;

[0042] Figure 7 It is a schematic diagram of the comparison between the pre-processed target spectrum data and the original spectrum data;

[0043] Figure 8 Schematic diagram of the prediction results of the training set and test set for each element;

[0044] Fig. 9 Schematic diagram of the prediction results of the validation set for each element;

[0045] Fig.10 It is a schematic diagram of the calibration results of beryllium and strontium in the standard solution;

[0046] Fig.11 This is a schematic diagram of the spiked detection results of the second example of liquid sample analysis. DETAILED DESCRIPTION

[0047] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with embodiments and drawings. The exemplary embodiments of the present invention and their description are only used to explain the present invention and are not intended to limit the present invention.

[0048] Example 1

[0049] This embodiment provides a flexible armored super-hydrophobic film for matrix conversion, comprising hydrophobic particles and a film; the hydrophobic particles are used to adhere to the surface of the film to form a super-hydrophobic layer; the film has pockets of 10-50 μm.

[0050] The diameter of the hydrophobic particles is 1-100 nm, and the thickness of the flexible armored super-hydrophobic film is 50-500 μm.

[0051] The film is made of at least one of the following materials: PC, PET; the hydrophobic particles are hydrophobic nano-silicon dioxide or hydrophobic nano-titanium dioxide.

[0052] Example 2

[0053] This embodiment provides a method for preparing the flexible armored super-hydrophobic film described in Example 1, such as Figure 1 As shown, including:

[0054] Step 1: Wipe and dry the film with micron-sized pockets and the transparent supporting substrate; the transparent supporting substrate is made of at least one of the following transparent materials: glass, quartz, plastic; specifically, use ultrapure water and ethanol to clean the transparent supporting substrate, and after cleaning, put the transparent supporting substrate into an oven for drying to ensure that there are no impurities on the surface of the transparent supporting substrate.

[0055] Step 2: Evenly stick the film on a transparent carrier substrate; evenly stick the flexible PET film with micron-sized pockets on the surface of the transparent carrier substrate to ensure that the film is evenly attached to ensure the uniformity and consistency of subsequent coatings.

[0056] Step 3: Attach 100-300 μL of super-hydrophobic nanoparticles to the surface of the film, and form a flexible armored super-hydrophobic film after drying. In this step, the method of attaching 100-300 μL of super-hydrophobic nanoparticles to the surface of the film includes: spraying, coating, spin coating, or vapor deposition. In this embodiment, after coating 200 μL of super-hydrophobic silica nanoparticles on the surface of the film, after about 5 minutes of drying, a film interface with super-hydrophobicity is formed.

[0057] In this example, a control group was also prepared: a blank glass carrier was coated with an equal amount of super-hydrophobic silica nanoparticles and dried for about 5 minutes to form a super-hydrophobic interface.

[0058] In order to verify the mechanical robustness of the prepared flexible armor super-hydrophobic film, this embodiment uses a sandpaper wear test to test it. The specific operation is: a 100g weight is placed on the surface of the flexible armor super-hydrophobic film, and the surface is facing down to contact with 800-mesh sandpaper, and a cycle moves 23cm. Through multiple wear cycles, the super-hydrophobic performance changes of the film under different wear times are evaluated.

[0059] Schematic diagram of the principle of forming a hydrophobic surface based on micron-sized pockets Figure 2 As shown, Figure 3 The SEM image shows the structure of micron-scale pockets on the surface of the film, which can effectively protect the Figure 4 The superhydrophobic silica particles shown greatly enhance the mechanical robustness of the film; Figure 5 As shown in the figure, the hydrophobic nanostructures deposited on the glass are very fragile. Without protection, their superhydrophobicity is significantly destroyed after only a few wear cycles (about 23 cm); while the flexible armored superhydrophobic film still maintains excellent superhydrophobic properties after 37 wear cycles (about 850 cm of sandpaper wear distance), with a water contact angle of at least 150°, which is significantly better than the control group. This shows that the micron-sized pockets of the basement membrane have a good protective effect on the hydrophobic nanostructures, enhance the mechanical robustness of the film, and ensure its stability and long-term use in different environments.

[0060] Example 3

[0061] This embodiment provides an application of a flexible armored super-hydrophobic film for realizing liquid sample analysis based on LIBS technology; Figure 6 As shown, the method for analyzing liquid samples based on LIBS technology includes:

[0062] S3-1, preparation of flexible armored super hydrophobic film;

[0063] S3-2, converting the liquid phase sample into a solid phase sample based on the flexible armored super-hydrophobic film, and detecting the solid phase sample based on the LIBS technology to obtain the original element spectrum; this step specifically includes the following method:

[0064] S3-21, dropping a liquid sample onto the flexible armored super-hydrophobic film, and forming a detection point after the liquid sample is dried;

[0065] S3-22, placing the detection point at the laser focus of the LIBS spectrometer for detection to obtain the original element spectrum of the detection point.

[0066] S3-3, establishing a quantitative model of the target metal element, for performing data processing on the original element spectrum to obtain the target metal element detection result; this step specifically includes the following method:

[0067] S3-31, confirming the spectral range of the target metal element according to the spectral database, and extracting the target spectral data from the original element spectrum based on the spectral range; the spectral database used in this embodiment is NIST Atmospheric Spectra Database;

[0068] S3-32, preprocessing the target spectrum data, and performing principal component analysis on the preprocessed target spectrum data to extract characteristic data; this step specifically includes the following method:

[0069] S3-321, normalize the target spectral data, and linearly transform and map the target spectral data to [0, 1]; perform baseline correction on the normalized target spectral data based on a fourth-order polynomial fitting method, and then use the average spectrum of the target spectral data after baseline correction as a basic spectrum to perform multivariate scattering correction;

[0070] S3-322, divide the preprocessed target spectral data into a training set, a test set and a validation set (where the training set accounts for 70% of the total data and is used for model training and parameter optimization; the test set accounts for 20% of the total data and is used to evaluate the performance and generalization ability of the model; the validation set accounts for 10% of the total data and is used to tune hyperparameters and verify the performance of the model on unknown data.), and extract characteristic data from the target spectral data based on the principal component analysis method, the principal component analysis method includes the following steps in sequence: calculating the covariance matrix, eigenvalue decomposition, and extracting principal components; wherein the principal component analysis method extracts the principal components based on the cumulative variance contribution method to ensure that the total variance of the first k principal components accounts for ≥95% of the total variance of the original data.

[0071] The specific covariance matrix Q is calculated according to the following formula: in represents the mean vector of spectral data; S i represents the i-th spectral vector; N represents the number of spectral data vectors; the superscript T represents transposition;

[0072] Select the first k eigenvectors to achieve dimensionality reduction, that is: S PCA =W T S normalized ; Where W represents the eigenvector matrix of principal component analysis; S PCA Spectral data after dimension reduction; S normalized is the spectral data after normalization;

[0073] S3-33, based on the XGBoost machine learning algorithm, the characteristic data is quantitatively analyzed to obtain a quantitative model of the target metal element concentration and spectral intensity, thereby obtaining the detection result of the target metal element in the liquid sample.

[0074] In this step, the XGBoost machine learning algorithm uses the average value of all samples as the initial model. In each round of iteration, the residual of the current model is first calculated, and a new decision tree is trained with the residual as the target variable to fit the residual, and the output of the new decision tree is superimposed on the current model for updating the predicted value; the XGBoost machine learning algorithm finds the best decision tree structure and parameter configuration by minimizing the objective function, and a regularization term is added to the objective function to control the complexity of the model and prevent overfitting. Through multiple rounds of iterations, the prediction accuracy of the model is continuously optimized.

[0075] The XGBoost machine learning algorithm finds the best tree structure and parameter configuration by minimizing the objective function. That is:

[0076]

[0077] Where L(θ) represents the objective function, represents the loss function; y i Indicates actual value; represents the predicted value; k represents the kth decision tree; K represents the number of decision trees. The optimal parameters of the XGBoost machine learning algorithm are searched by the GridSearchCV function after 5 cross-validation evaluations.

[0078] Example 4

[0079] This embodiment provides a first example of liquid sample analysis based on LIBS technology, which includes the following steps:

[0080] S4-1, preparation of flexible armored superhydrophobic film;

[0081] S4-2, the liquid sample is converted into a solid sample based on a flexible armored super-hydrophobic film, and the solid sample is detected based on the LIBS technology to obtain the original element spectrum; the liquid sample contains four target elements: beryllium (Be), strontium (Sr), barium (Ba) and manganese (Mn); a portion of the liquid sample (50 μl) is drawn with a pipette and dropped on the flexible armored super-hydrophobic film, which is then placed in a heating chamber until the droplet is completely dry.

[0082] In this embodiment, the LIBS technology detection equipment uses a Nd:YAG laser to generate a laser wavelength of 1064nm, a pulse energy of 100mJ, and a pulse width of 13ns. The laser is focused on the sample surface by a lens to generate plasma. The emission light of the plasma is coupled to a spectrometer with a wavelength range of 185 to 780nm. The measurement adopts a single-shot mode with a delay time of 500ns to avoid the continuous spectrum caused by bremsstrahlung and compound radiation in the early stage of plasma evolution; the converted solid phase sample is placed at the laser focus of the instrument and detected to obtain the original spectrum data.

[0083] S4-3, finally, establish a quantitative model and perform quantitative analysis on the target element: first, find the characteristic spectral peak corresponding to the target element in the NISTAtomic SpectraDatabase, and then perform preprocessing processes such as normalization, baseline correction, and multivariate scattering correction on the original spectral data of the target element. Then, perform principal component analysis and XGBoost quantitative analysis on the spectral data to establish a quantitative model. Quantitatively analyze the concentration of the target element.

[0084] The specific steps include:

[0085] S4-31, search for characteristic peaks of beryllium, strontium, barium and manganese in the NISTAtomic Spectra Database. Due to the need for subsequent analysis, if full spectrum analysis is used, the time pressure of model training will be greatly increased, so the partial spectrum containing the characteristic peak of each target element is used for analysis. The partial spectrum range of each target element is: beryllium (300-340nm), strontium (450-480nm), barium (480-510nm), and manganese (390-420nm).

[0086] S4-32, the full spectrum is spectrally preprocessed, and the spectrum normalization adopts deviation standardization to linearly transform the original data so that the result value is mapped to [0–1]. The full spectrum data is baseline corrected using the fourth-order polynomial fitting method. The average spectrum of the spectral data after the baseline correction is used as the basic spectrum of the multivariate scattering correction, and the final corrected spectral data is obtained. The results are shown in Figure 7 As shown in the figure, it can be seen that the noise in the spectral data has been suppressed to a certain extent, while the peak value has not been affected much. This shows that the spectral data can be made easier to analyze through the preprocessing process.

[0087] S4-33, perform principal component analysis to extract features: divide the selected partial spectra of each target element into three parts: training set, test set and validation set in a ratio of 7:2:1, and reduce the dimension of the spectral data based on principal component analysis. Calculate the covariance matrix, perform eigenvalue decomposition, and extract the principal components.

[0088] In this embodiment, the number of principal components selected by principal component analysis is selected by the cumulative variance contribution method to ensure that the total variance explained by the first k principal components accounts for the proportion of the total variance of the original data. The number of principal components selected for each target element is beryllium (51), strontium (40), barium (32), and manganese (37).

[0089] S4-34, the XGBoost algorithm builds a quantitative model based on the feature data, as follows:

[0090] The average value of all samples is used as the initial model. In each iteration, the residual of the current model is first calculated, that is, the difference between the predicted value and the actual value. Then, using the residual as the target variable, a new decision tree is trained to fit these residuals, and the output of the new tree is added to the current model to update the predicted value. In addition, a regularization term is added to the objective function of XGBoost to control the complexity of the model and prevent overfitting. Through multiple rounds of iterations, the prediction accuracy of the model is continuously optimized.

[0091] The optimal parameters of each element in this embodiment are beryllium (number of decision numbers: 100, learning ratio: 0.1, maximum depth: 3), strontium (number of decision numbers: 100, learning ratio: 0.2, maximum depth: 3), barium (number of decision numbers: 100, learning ratio: 0.1, maximum depth: 3), and manganese (number of decision numbers: 100, learning ratio: 0.1, maximum depth: 3).

[0092] The performance of the established quantitative model was then evaluated using the coefficient of determination (R 2 ) and root mean square error (RMSE). 2 The closer the value is to 1, the better the model fits the observed data. The lower the RMSE, the more negligible the difference between the predicted value and the actual value, which means the higher the prediction accuracy and credibility of the model. Figure 8 As shown in Figure 2, the prediction results of the training set and test set of beryllium are as follows: Figure 8 As shown in b, the prediction results of the training set and test set of strontium are as follows Figure 8 As shown in (d), the prediction results of the training set and test set of barium are as follows Figure 8 As shown in c, the prediction results of the training set and test set of manganese are as follows Figure 8 As shown in Figure a, the performance of the established quantitative model is good, and the performance of the model in the prediction and training data sets is satisfactory. In order to test the generalization performance and prediction ability of the model, the model is then used to predict the validation set data. The prediction results of each element validation set are shown in Fig. 9 As shown; the prediction results of the beryllium element validation set are as follows Fig. 9 As shown in b, the prediction results of the strontium element validation set are as follows Fig. 9 As shown in (d), the prediction results of the manganese element validation set are as follows Fig. 9 As shown in a, the prediction results of the barium element validation set are as follows Fig. 9 As shown in c; R of the fitting curve of each element 2 They are all close to 1, indicating that the linearity of the prediction results of each element is good. When the slope of the fitting curve is very close to 1, the intercepts of each curve are also close to 0, which means that the prediction results match the actual results well and the model is more accurate.

[0093] Example 5

[0094] This embodiment also provides a second example of liquid sample analysis based on LIBS technology, which specifically includes the following steps:

[0095] S5-1, preparation of flexible armored super hydrophobic film;

[0096] S5-2, convert the liquid sample into a solid sample: select commercially available bottled drinking water as the spiked sample, add a mixed standard solution of four target elements to the sample, with an added concentration of 10% and a concentration of 5 μg / L; the rest is the same as the first example.

[0097] S5-3, establish a single variable quantitative model, bring the actual sample data into the quantitative model obtained in the first example for prediction. Specifically include: Establishment of a single variable model: establish a single variable model of the target element spectrum and concentration based on the characteristic spectrum peak of the target element spectrum, and quantitatively analyze the concentration of the target element. The content of the substance to be measured in the spectrum and the concentration of the substance satisfy the Saber-Romagin formula:

[0098] I=aC b ;

[0099] Wherein C is the concentration of the element to be measured, a represents a coefficient related to the plasma excitation process, a can be determined by the slope of the standard curve, and b is the self-absorption coefficient; b=1 is generally selected in laser induced breakdown spectroscopy; I is the peak value of the element spectrum curve of the element to be measured; the least squares method is used to fit the linear curve of the peak value and the concentration; that is, the a in the Saber-Romagin formula is obtained by fitting the least squares method; thereby the corresponding linear curve is fitted; according to the obtained linear curve, the spectral intensity is brought into the linear curve to obtain the concentration of the element to be measured. Fig.10 The results of the calibration of Be and Sr in standard solutions are shown. 2 The values ​​of 0.982 and 0.973 respectively indicate that it has a certain predictive ability. However, it should also be noted that the error bars of the data are larger with the increase of its concentration, which reflects the aforementioned instability.

[0100] Finally, the obtained spectral data were brought into the univariate model and quantitative model for prediction. The results are as follows: Fig.11 As shown in the figure, the RSD of the results obtained by the quantitative model in the first example is greatly reduced compared with the results obtained by directly using the univariate model. The RSD of the Be element is reduced from 37.18% to 6.17%, and the RSD of the Sr element is reduced from 31.53% to 8.36%. This is enough to prove that the use of this method can effectively improve the stability of LIBS for trace metal detection in water.

[0101] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A flexible armored super-hydrophobic film, characterized in that: Used for matrix conversion, comprising hydrophobic particles and a film; the hydrophobic particles are used to adhere to the surface of the film to form a super hydrophobic layer; the film has pockets of 10-50 μm.

2. A flexible armored super-hydrophobic film according to claim 1, characterized in that: The diameter of the hydrophobic particles is 1-100 nm, and the thickness of the flexible armored super-hydrophobic film is 50-500 μm.

3. A flexible armored super-hydrophobic film according to claim 1, characterized in that: The film is made of at least one of the following materials: PC, PET; the hydrophobic particles are hydrophobic nano-silicon dioxide or hydrophobic nano-titanium dioxide.

4. The method for preparing a flexible armored super-hydrophobic film according to any one of claims 1 to 3, characterized in that: include: Wipe the film with micron-sized pockets and the transparent carrier substrate clean and dry; Evenly sticking the film on a transparent supporting substrate; 100-300 μL of super-hydrophobic nanoparticles are attached to the surface of the film, and a flexible armored super-hydrophobic film is formed after drying.

5. The method for preparing the flexible armored super-hydrophobic film according to claim 4, characterized in that: The method of attaching 200 μL of super-hydrophobic nanoparticles to the surface of the film includes: spraying, coating, spin coating, or vapor deposition.

6. The use of the flexible armored super-hydrophobic film according to any one of claims 1 to 3, characterized in that: Used to implement liquid sample analysis based on LIBS technology; the method for liquid sample analysis based on LIBS technology includes: Prepared a flexible armored super-hydrophobic film; The liquid phase sample is converted into a solid phase sample based on a flexible armored super-hydrophobic film, and the solid phase sample is detected based on the LIBS technology to obtain the original element spectrum; A quantitative model of the target metal element is established to process the original element spectrum to obtain the target metal element detection result.

7. The use of the flexible armored super-hydrophobic film according to claim 6, characterized in that: The liquid phase sample is converted into a solid phase sample based on a flexible armored super-hydrophobic film, and the solid phase sample is detected based on the LIBS technology to obtain the original element spectrum; Included methods: Dropping a liquid sample onto the flexible armored super-hydrophobic film, and forming a detection point after the liquid sample is dried; The detection point is placed at the laser focus of the LIBS spectrometer for detection to obtain the original element spectrum of the detection point.

8. The use of the flexible armored super-hydrophobic film according to claim 6, characterized in that: The quantitative model of the target metal element is established to process the original element spectrum to obtain the target metal element detection result; including method: Confirm the spectral range of the target metal element according to the spectral database, and extract the target spectral data from the original element spectrum based on the spectral range; Preprocessing the target spectral data, and performing principal component analysis on the preprocessed target spectral data to extract characteristic data; The characteristic data is quantitatively analyzed based on the XGBoost machine learning algorithm to obtain a quantitative model of the target metal element concentration and spectral intensity, thereby obtaining the detection result of the target metal element in the liquid sample.

9. The use of the flexible armored super-hydrophobic film according to claim 8, characterized in that: The target spectrum data is preprocessed, and the preprocessed target spectrum data is subjected to principal component analysis to extract characteristic data; Included methods: The target spectrum data is normalized, and the baseline correction is first performed on the normalized target spectrum data based on the fourth-order polynomial fitting method, and then the average spectrum of the target spectrum data after baseline correction is used as the basic spectrum for multivariate scattering correction; The preprocessed target spectral data is divided into a training set, a test set and a validation set, and characteristic data is extracted from the target spectral data based on a principal component analysis method, wherein the principal component analysis method sequentially comprises the following steps: calculating a covariance matrix, eigenvalue decomposition, and extracting principal components; The principal component analysis method is based on the cumulative variance contribution method to extract the principal components, ensuring that the total variance of the first k principal components accounts for ≥ 95% of the total variance of the original data.

10. The use of a flexible armored super-hydrophobic film according to claim 8, characterized in that: The XGBoost machine learning algorithm uses the average value of all samples as the initial model. In each round of iteration, the residual of the current model is first calculated, and a new decision tree is trained with the residual as the target variable to fit the residual, and the output of the new decision tree is superimposed on the current model for updating the predicted value; the XGBoost machine learning algorithm finds the optimal decision tree structure and parameter configuration by minimizing the objective function, and a regularization term is added to the objective function.

Images