Heavy metal detection method and device based on surface microporous structure substrate

Abstract

The invention provides a heavy metal detection method and device based on a surface microporous structure substrate, and relates to the field of heavy metal detection. The heavy metal detection method based on the surface microporous structure substrate comprises the following steps: focusing a Bessel beam on a to-be-detected sample on a light absorption substrate, and exciting the to-be-detected sample to generate plasma; wherein the Bessel beam is formed by converting Gaussian pulse laser; and obtaining the content of heavy metal elements in the to-be-detected sample according to the spectral signal of the plasma. According to the heavy metal detection method, the accuracy and efficiency of heavy metal detection are remarkably improved by designing the special surface microporous structure substrate, shaping the light beam and adopting the Gaussian mixture regression model.

Description

Technical Field

[0001] The present application relates to the field of heavy metal detection, and more specifically, to a heavy metal detection method and device based on a substrate with a surface microporous structure. Background Art

[0002] Heavy metals such as lead, mercury, cadmium, and arsenic pose a serious threat to the environment and human health and may cause a range of health problems, including neurological damage, kidney disease, and cancer; therefore, heavy metals need to be detected. Common heavy metal detection methods include atomic absorption spectroscopy (AAS), inductively coupled plasma mass spectrometry (ICP-MS), X-ray fluorescence spectroscopy (XRF), and laser induced breakdown spectroscopy (LIBS). These methods are able to detect different sample types and have different detection sensitivities, and are widely used in environmental monitoring, food safety, and industrial emissions.

[0003] Laser induced breakdown spectroscopy (LIBS) is an analytical method based on laser technology. It generates local plasma on the sample surface by focusing high-energy laser pulses, thereby exciting the elements in the sample and emitting characteristic spectra. Analyzing these spectral signals can obtain the type and content of each element in the sample. It is particularly suitable for real-time analysis of complex samples and is widely used in environmental monitoring, materials science, chemical analysis and other fields.

[0004] However, the current laser induced breakdown spectroscopy (LIBS) method has limited accuracy in heavy metal detection due to interference from the interaction between laser and complex samples, insufficient sample enrichment efficiency, and insufficient accuracy and stability in spectral signal processing and analysis, and needs to be further improved. Summary of the invention

[0005] The purpose of the embodiments of the present application is to provide a heavy metal detection method and device based on a surface microporous structure substrate. The heavy metal detection method significantly improves the accuracy and efficiency of heavy metal detection by designing a special surface microporous structure substrate, beam shaping and using a Gaussian mixture regression model.

[0006] In a first aspect, an embodiment of the present application provides a heavy metal detection method based on a surface microporous structure substrate. The method includes: focusing a Bessel light beam on a sample to be tested on a light absorbing substrate to excite the sample to be tested to generate plasma; wherein the Bessel light beam is converted from a Gaussian pulse laser; and obtaining the heavy metal element content in the sample to be tested based on the spectral signal of the plasma.

[0007] In the above implementation process, the heavy metal detection method provided in the embodiment of the present application improves the plasma generation efficiency by converting Gaussian pulse laser into Bessel beam to enhance the interaction between laser and sample, and uses spectral signals for quantitative analysis, thereby significantly improving the detection sensitivity and accuracy of heavy metal elements.

[0008] Optionally, in an embodiment of the present application, the method for preparing the light absorbing substrate includes: using a femtosecond laser to process the surface of a metal substrate to form a surface microporous structure substrate having a plurality of deep holes; wherein the deep holes are periodically and regularly arranged on the surface microporous structure substrate.

[0009] In the above implementation process, the surface microporous substrate provided in the embodiment of the present application can enhance the LIBS detection sensitivity through light absorption. At the same time, due to the deep hole structure of the surface microporous structure substrate, the spectral signal of the metal oxide colloidal particle pollutants can be further amplified, resulting in a more conducive to enhancing the LIBS signal than a simple ordinary micro-nano structure, thereby improving the sensitivity and accuracy of heavy metal element detection.

[0010] Optionally, in the embodiment of the present application, before focusing the Bessel light beam on the sample to be tested on the light absorbing substrate, the heavy metal detection method further includes: adding the sample to be tested to the light absorbing substrate; and drying the sample to be tested to obtain the sample to be tested.

[0011] In the above implementation process, based on the specific structure and size of the surface microporous structure substrate, and the beneficial effect of amplifying the heavy metal spectral signal of the metal oxide colloidal particle pollutant, the structure can be combined with the droplet deposition method to directly and quickly deposit the sample evenly on the substrate surface, and prepare the sample to be tested by heating. Compared with the traditional sample processing process, this method reduces the early filtration, precipitation, centrifugation and other operations, further reducing the difficulty of operation and time cost.

[0012] Optionally, in an embodiment of the present application, obtaining the content of heavy metal elements in the sample to be tested according to the spectral signal of the plasma includes:

[0013] The spectral signal is converted into spectral data; the spectral data is input into the quantitative prediction model to obtain the heavy metal element content.

[0014] In the above-mentioned implementation process, the heavy metal detection method provided by the embodiment of the present application effectively converts the plasma light signal into accurate spectral data through a high-resolution spectrometer, and inputs the spectral data into a quantitative prediction model, which can accurately predict the content of heavy metal elements in the sample. In the embodiment of the present application, the use of a quantitative prediction model can significantly improve the accuracy and efficiency of heavy metal detection, and support fast and reliable sample analysis. Its flexibility and adaptability enable the heavy metal detection method provided by the embodiment of the present application to be widely used in the heavy metal content assessment of various complex water samples.

[0015] Optionally, in an embodiment of the present application, the quantitative prediction model determination process includes: determining the number of Gaussian components of the Gaussian mixture model, defining the mean vector and covariance matrix of each Gaussian component to describe the clustering characteristics in the spectrum; calculating the posterior probability that each data point in the spectral data belongs to each Gaussian distribution to evaluate the degree of match between each data point and each Gaussian component; based on the posterior probability, updating the weight, mean and covariance of the Gaussian distribution until the parameters of the Gaussian mixture model converge; based on samples with known heavy metal concentrations, calculating the integral area of ​​each Gaussian distribution, and establishing a quantitative relationship between the integral area and the concentration of heavy metal elements.

[0016] In the above implementation process, in the process of determining the quantitative prediction model of the heavy metal detection method provided by the embodiment of the present application, prior knowledge is introduced, and the characteristic spectrum information of common heavy metal elements in industrial wastewater is combined to initialize the parameters of the GMR model to make it more suitable for practical application scenarios. For large-scale high-dimensional spectral data, the convergence speed and stability of the EM algorithm are improved, and parallel computing or approximate inference methods are introduced to accelerate the model training process. In addition, industrial wastewater samples usually contain a variety of metal oxides, and each oxide produces different emission lines in the laser induced breakdown spectrum. The Gaussian mixture model can decompose these complex mixed spectra into multiple Gaussian distributions, each distribution corresponding to a characteristic spectrum line of a metal element. This helps to more accurately identify and quantify the concentration of different heavy metal elements. In the case of coexistence of characteristic peaks of multiple metal elements, the spectral characteristic peaks may overlap, making it difficult to distinguish different components. The Gaussian mixture model can decompose the overlapping spectral peaks by fitting multiple Gaussian distributions, thereby improving the quantitative analysis accuracy of each heavy metal element. In the embodiment of the present application, the Gaussian mixture model uses multiple Gaussian distributions to fit the emission line shape in the spectrum. Compared with simple linear or nonlinear regression models, Gaussian distribution can more accurately fit the characteristic spectral peaks of metal elements, thereby obtaining more accurate concentration predictions. Furthermore, a quantitative relationship between the characteristic spectral intensity and content of each metal element is established. Since Gaussian distribution can accurately distinguish the characteristics of emission peaks, the model can better predict the concentration of metal oxides in samples.

[0017] In the second aspect, an embodiment of the present application provides a heavy metal detection device based on a surface microporous structure substrate, characterized in that the heavy metal detection device includes: a LIBS detection module and a data processing module; the LIBS detection module is used to focus the Bessel light beam on the sample to be tested on the light absorbing substrate to excite the sample to be tested to generate plasma; wherein the Bessel light beam is converted from a Gaussian pulse laser; the data processing module is used to obtain the heavy metal element content in the sample to be tested according to the spectral signal of the plasma.

[0018] Optionally, in an embodiment of the present application, the LIBS detection module includes: a laser emitting component and a beam adjustment mirror group; the laser emitting component is used to send a pulsed laser to the beam adjustment mirror group; the beam adjustment mirror group is used to convert the Gaussian pulse laser into a Bessel beam, and focus the Bessel beam on the sample to be tested on the light absorbing substrate, to excite the sample to be tested to generate plasma.

[0019] Optionally, in an embodiment of the present application, the beam adjustment mirror group includes: a reflector, an axle and a focusing lens; the reflector is used to receive the pulsed laser and change the direction of the pulsed laser to be perpendicular to the plane where the light absorbing substrate is located; the axle is used to convert the pulsed laser into a Bessel beam; and the focusing lens is used to focus the Bessel beam on the sample to be tested.

[0020] Optionally, in an embodiment of the present application, the data processing module includes a spectral signal acquisition component, a spectral signal conversion component and a quantitative prediction unit; the spectral signal acquisition component is used to collect plasma light signals and import them into the spectral signal conversion component; the spectral signal conversion component is used to convert spectral signals into spectral data; the quantitative prediction unit is configured to obtain the heavy metal element content in the sample to be tested according to the spectral data.

[0021] Optionally, in an embodiment of the present application, the heavy metal detection device also includes a light absorbing substrate processing device, which is used to use a femtosecond laser to process the surface of a metal substrate to form a surface microporous structure substrate with multiple deep holes; wherein the deep holes are periodically and regularly arranged on the surface microporous structure substrate.

[0022] In the above implementation process, the heavy metal detection method and device based on the surface microporous structure provided by the embodiment of the present application designs a surface microporous structure substrate with a specific structure to enhance the detection ability of metal oxides; by converting the Gaussian beam into a Bessel beam, the sample pretreatment steps of the traditional droplet drying method are simplified, and the spectral data quality is optimized; finally, the Gaussian mixture regression model is used to process the collected spectral data, and the overlapping peaks are decomposed into independent characteristic peaks, which significantly improves the quantitative prediction ability of heavy metal elements. The accuracy and efficiency of heavy metal detection in industrial wastewater are comprehensively improved from multiple aspects.

[0023] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory and a processor, wherein the memory stores program instructions, and when the processor reads and runs the program instructions, it executes the steps in any of the above implementation methods.

[0024] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores computer program instructions, and when the computer program instructions are read and executed by a processor, the steps in any of the above-mentioned implementation methods are executed. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments of the present application will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 A first flow chart of heavy metal detection provided in an embodiment of the present application;

[0027] Figure 2 An example of the preparation process of the light absorbing substrate provided in the embodiment of the present application;

[0028] Figure 3 Schematic diagram of a substrate with a surface microporous structure provided in an embodiment of the present application;

[0029] Figure 4 A second flow chart of heavy metal detection provided in the embodiment of the present application;

[0030] Figure 5 A flow chart for determining a quantitative prediction model provided in an embodiment of the present application;

[0031] Figure 6 The Gaussian mixture model provided in the embodiment of the present application is optimized by the expected maximum algorithm EM to convert overlapping peaks into multiple branch peak instance graphs;

[0032] Figure 7 A schematic diagram of the structure of a heavy metal detection device provided in an embodiment of the present application;

[0033] Figure 8 A light absorbing substrate processing device provided in an embodiment of the present application;

[0034] Fig. 9 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application;

[0035] Reference numerals: heavy metal detection device-1000; LIBS detection module-100; laser emission assembly-110; nanosecond laser-111; timing pulse controller-112; beam adjustment mirror group-120; reflector-121; conical lens-122; focusing lens-123; first focusing lens-1231; second focusing lens-1232; data processing module-200; spectral signal acquisition assembly-210; acquisition lens-211; optical fiber probe-212; optical fiber-2 13; spectral signal conversion component-220; quantitative prediction unit-230; computer-240; XY displacement platform-300; light absorption substrate processing device-400; femtosecond laser-410; spatial light modulator-420; third focusing lens-430; objective lens-440; metal substrate-450; displacement platform controller-460; XYZ displacement platform-470; mechanical switch-490; deep hole-H; electronic device-500; processor-501; memory-502. DETAILED DESCRIPTION

[0036] The technical solutions in the embodiments of the present application will be described below in conjunction with the accompanying drawings in the embodiments of the present application. For example, the flowcharts and block diagrams in the accompanying drawings show the possible architectures, functions and operations of the systems, methods and computer program products according to multiple embodiments of the present invention. In this regard, each box in the flowchart or block diagram may represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical functions. It should also be noted that in some alternative implementations, the functions marked in the box may also occur in an order different from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they may sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram and / or the flowchart, and the combination of boxes in the block diagram and / or the flowchart, can be implemented by a dedicated hardware-based system that performs a specified function or action, or can be implemented by a combination of dedicated hardware and computer instructions. In addition, the functional modules in the various embodiments of the present invention can be integrated together to form an independent part, or each module can exist separately, or two or more modules can be integrated to form an independent part.

[0037] Laser induced breakdown spectroscopy (LIBS) is an efficient heavy metal content analysis technology widely used in materials science, environmental monitoring and chemical analysis. This method focuses high-energy laser pulses onto the sample surface to instantly generate local plasma, excite the elements in the sample and emit characteristic spectra. By analyzing these spectral signals, the type and content of each element in the sample can be quickly obtained, making it particularly suitable for instant analysis of complex samples.

[0038] At present, the heavy metal detection method based on light absorption microstructure is to use femtosecond laser to prepare periodic micro-nanostructure on aluminum foil material to form a light absorption substrate. The substrate enhances the light absorption characteristics through its micro-nanostructure, thereby improving the interaction effect between laser and sample and enhancing the spectral signal. Samples are extracted from agricultural ditch water or polluted water samples, and the supernatant is added to the surface of the light absorption substrate after centrifugation. After drying, the sample to be tested is obtained. The surface of the sample to be tested is ablated by pulsed laser, the sample is excited to generate plasma, and the plasma spectral signal is collected. The pulsed laser is processed by a photodetector and a high-resolution spectrometer to obtain the characteristic spectrum of plasma. The collected spectral data is processed, firstly outliers are removed and spectral baseline correction is performed. Then, the partial least squares method (PLS) combined with the simulated annealing algorithm is used for quantitative analysis. The simulated annealing algorithm is used to optimize the characteristic wavelengths in the spectral data, and the best characteristic wavelength combination is found through multiple iterations, thereby improving the accuracy of heavy metal content prediction.

[0039] In the above implementation process, there is a problem of insufficient sample pretreatment. Although the centrifugation, supernatant aspiration and drying methods are simple and easy to operate, they do not particularly optimize the enrichment and deposition effects when dealing with trace heavy metals. Since the concentration of heavy metal elements is usually low, this pretreatment method may not be able to effectively capture and concentrate these trace components, resulting in inaccurate or insufficient sensitivity in the final test results. In addition, the adaptability of this method in processing different samples is also limited, especially when the sample contains high viscosity or suspended particles, which may affect the efficiency of supernatant acquisition and further reduce the detection ability of heavy metals.

[0040] The obtained spectral data are processed by partial least squares (PLS) combined with simulated annealing algorithm. However, its accuracy needs to be improved when processing more complex spectral data.

[0041] In the above scheme, the way to improve the signal intensity of laser-induced breakdown spectroscopy mainly relies on the light absorption capacity of the microstructure; therefore, the accuracy of heavy metal content detection is limited by the characteristics of the substrate material and the design of the microstructure. In addition, if the optical properties of the substrate material are not ideal, or the design of the microstructure cannot effectively adapt to different types of samples, the signal enhancement effect may be insufficient. In addition, the manufacturing process of the microstructure is complex and costly, which may limit its promotion in large-scale applications.

[0042] Based on this, the present application proposes a heavy metal detection method and device based on a surface microporous structure substrate. The heavy metal detection method converts a Gaussian beam into a Bessel beam, optimizes the interaction between the laser and the sample, improves the quality of the spectral signal and the detection sensitivity, and improves the overall accuracy of heavy metal detection.

[0043] Please see Figure 1 , Figure 1 The first flow chart of heavy metal detection provided in the embodiment of the present application; The present application provides a heavy metal detection method, which can be Figure 8 The heavy metal detection method includes:

[0044] Step S100: focusing the Bessel beam on the sample to be tested on the light absorbing substrate to excite the sample to be tested to generate plasma, wherein the Bessel beam is converted from a Gaussian pulse laser.

[0045] In the above step S100, a Bessel beam is generated to excite the sample to be tested so that the sample to be tested generates plasma. Exemplarily, the Gaussian pulse laser is converted into a Bessel beam through beam shaping, and then focused onto the sample to be tested on the light absorbing substrate, and the interaction between the laser and the sample causes high-temperature plasma to form on the sample surface.

[0046] The light-absorbing substrate refers to a specific structure that provides a place for the interaction between the sample to be tested and the laser.

[0047] It should be noted that the Bessel beam is a special form of beam with constant intensity distribution and large depth tolerance. The non-diffraction property of the Bessel beam is one of its most important characteristics. The traditional Gaussian beam will diffract during propagation, resulting in the expansion of the focal area and energy dispersion, while the Bessel beam can maintain its cross-sectional shape unchanged over a longer propagation distance.

[0048] On the one hand, due to the non-diffraction characteristics of the Bessel beam, the focus of the laser on the sample can remain stable over a long distance, ensuring the stability and consistency of the breakdown plasma generation.

[0049] On the other hand, the telephoto region of the Bessel beam can penetrate deeper into the sample, which can enhance the detection capability of heavy metal colloidal particle pollutants in industrial wastewater dried on the surface microporous structure substrate.

[0050] On the other hand, the traditional laser beam has a shallow focusing depth, which may cause measurement errors due to surface and local changes of the sample. In the embodiment of the present application, since the traditional water sample pretreatment method of filtration, centrifugation and precipitation is not adopted, the focusing depth is increased by using a Bessel beam, which reduces the error caused by the irregularity of the sample surface.

[0051] Also, by distributing the energy more evenly, the Bessel beam is able to generate a more stable and uniform plasma on the sample surface, thereby enhancing the repeatability and stability of the spectral signal.

[0052] Step S200: obtaining the content of heavy metal elements in the sample to be tested according to the spectrum signal of the plasma.

[0053] In the above step S200, after exciting the sample, the generated plasma will emit a specific spectral signal, and the light signal emitted by the plasma contains the characteristic spectral lines of different elements in the sample. By collecting and analyzing the plasma light signal, the heavy metal elements and their concentrations in the sample can be identified. Using spectral analysis methods, such as partial least squares (PLS), combined with standard samples of known concentrations, the content of heavy metal elements in the sample to be tested can be predicted.

[0054] pass Figure 1 It can be seen that the heavy metal detection method provided in the embodiment of the present application improves the plasma generation efficiency by converting Gaussian pulse laser into Bessel beam to enhance the interaction between laser and sample, and uses spectral signals for quantitative analysis, thereby significantly improving the detection sensitivity and accuracy of heavy metal elements.

[0055] In an optional embodiment, the preparation method of the light absorbing substrate includes: using a femtosecond laser to process the surface of a metal substrate to form a surface microporous structure substrate having a plurality of deep holes; wherein the deep holes are periodically and regularly arranged on the surface microporous structure.

[0056] For example, please refer to Figure 2 and Figure 3 , Figure 2 An example of the preparation process of the light absorbing substrate provided in the embodiment of the present application; Figure 3 Schematic diagram of a substrate with a surface microporous structure provided in an embodiment of the present application; the preparation process of the light absorbing substrate includes:

[0057] Step S1: preparing a metal substrate and removing impurities on the surface of the metal substrate.

[0058] For example, an aluminum (Al) plate with a purity of 99.99%, a size of 5 cm×5 cm, and a thickness of 1 mm was used, and ultrasonic cleaning was performed with a power of 30 W for 30 minutes using anhydrous ethanol to remove surface impurities.

[0059] Step S2: Using a femtosecond laser to process the surface of a metal substrate to form a surface microporous structure substrate having a plurality of deep holes.

[0060] For example, the power of the laser is set to 2W, and the computer synchronously controls the drive and the mechanical switch to complete the high-precision displacement and selective processing of the substrate: the displacement stage moves 1μm, and the mechanical switch remains closed during the movement. After the movement is completed, the mechanical switch is turned on to ablate the substrate. This is the processing cycle to complete the processing of a 4.5cm×4.5cm area, forming a unit structure with a length and width of about 600nm and a depth of about 600nm. Figure 3 The deep pore surface microporous structure substrate shown.

[0061] Step S3: removing nanoparticles generated by laser processing.

[0062] Exemplarily, ultrasonic cleaning is performed again after processing, with a cleaning power of 30 W for 20 minutes using ultrapure water to remove nanoparticles generated by laser processing.

[0063] In the above implementation process, the surface microporous substrate is processed by femtosecond laser to form a micro-nano structure, which can significantly improve the absorption effect of the substrate on the laser. The periodicity and regularity of the surface micropores increase the interaction area between the laser and the substrate, thereby enhancing the efficiency of laser-induced plasma generation and the signal intensity of laser-induced breakdown spectroscopy (LIBS).

[0064] Furthermore, the microporous structure of the surface of the microporous substrate is conducive to the enrichment and deposition of industrial wastewater samples. The deposition on the substrate surface is more uniform and effective, thereby improving the stability of spectral analysis.

[0065] The microporous structure of the substrate with surface microporous structure (length, width and depth are all 600nm) increases the specific surface area of ​​the sample, making the interaction between light and matter more complete. At the same time, the surface microstructure helps to generate a stronger local electric field under the action of laser. The enhanced electric field increases the temperature and density of the plasma on the sample surface, causing the high-temperature plasma to release a stronger spectral signal, thereby improving the sensitivity and signal-to-noise ratio of LIBS detection.

[0066] In addition, the laser has multiple reflections and absorptions in the groove structure of the surface microporous structure substrate. The surface microporous structure substrate can effectively absorb and concentrate laser energy, making it easier to excite trace elements in the sample. By improving the utilization efficiency of laser energy, the detection capability of trace elements is enhanced, and the characteristic spectral signal of metal oxide colloidal particles is strengthened.

[0067] Finally, the embodiment of the present application uses aluminum as the base material of the surface microporous substrate, which can reduce the overall manufacturing cost. Aluminum itself is a cheap and easily available material, and the micro-nano structure formed by femtosecond laser processing technology has good replaceability. The substrate can be cut or replaced according to actual needs, which is suitable for large-scale and conventional detection.

[0068] pass Figure 2 and Figure 3It can be seen that the surface microporous substrate provided in the embodiment of the present application can enhance the LIBS detection sensitivity through light absorption. At the same time, due to the deep hole structure of the surface microporous structure substrate, the spectral signal of the metal oxide colloidal particle pollutants can be further amplified, which is more conducive to enhancing the LIBS signal than a simple ordinary micro-nano structure, thereby improving the sensitivity and accuracy of heavy metal element detection.

[0069] In an optional embodiment, before focusing the Bessel light beam on the sample to be tested on the light absorbing substrate, the heavy metal detection method further includes:

[0070] The sample to be tested is added to a light absorbing substrate, and the sample to be tested is dried to obtain a sample to be tested.

[0071] Exemplarily, 400 μL of collected industrial wastewater is sucked onto the surface microporous substrate using a pipette, and the surface microporous substrate dripped with the industrial wastewater is heated and dried by a constant temperature heating plate to obtain a sample to be tested.

[0072] It can be seen that based on the specific structure and size of the surface microporous substrate, and the beneficial effect of amplifying the spectral signal of metal oxide colloidal particle pollutants, this structure can be combined with the droplet deposition method to directly and quickly deposit the sample evenly on the substrate surface, and prepare the sample to be tested by heating. Compared with the traditional sample processing process, this method reduces the early filtration, precipitation, centrifugation and other operations, further reducing the difficulty of operation and time cost.

[0073] Please see Figure 4 , Figure 4 The second flow chart of heavy metal detection provided in the embodiment of the present application; in the above step S200, obtaining the heavy metal element content in the sample to be tested according to the spectral signal of the plasma can be achieved in the following way:

[0074] Step S210: converting the spectral signal into spectral data.

[0075] In the above step S210, the spectral signal is converted into spectral data. For example, the generated plasma light signal can be focused onto the optical fiber probe through a collection lens and transmitted to the high-resolution spectrometer through the optical fiber probe; the high-resolution spectrometer converts the light signal into an electrical signal, and the computer converts the electrical signal into spectral data.

[0076] In some embodiments, the above step S210 converts the spectral signal into spectral data, including: preprocessing the first spectral data set transmitted by the computer to the quantitative prediction model to obtain second spectral data.

[0077] In the above process, preprocessing, for example, normalizes the first spectral data, removes noise through a smoothing and denoising algorithm to ensure data quality, and finally outputs the second spectral data. Normalization is because the spectral intensity may be affected by laser power fluctuations or slight disturbances in environmental conditions. The spectrum is normalized using a standardization method so that the intensity of each spectral peak is on the same scale. Smoothing and denoising is to further reduce the impact of laser energy fluctuations or environmental interference on spectral data through smoothing.

[0078] Step S220: inputting the spectral data into the quantitative prediction model to obtain the heavy metal element content.

[0079] In the above step S220, the spectral data is input into the quantitative prediction model to obtain the content of heavy metal elements in the sample to be tested.

[0080] First, it is necessary to select an appropriate quantitative prediction model, for example, an algorithm based on machine learning (such as support vector machine, random forest, etc.), or a classic statistical method (such as partial least squares PLS). Quantitative prediction models are usually trained based on previously known standard sample data to establish the association between input spectral features and heavy metal content.

[0081] After the spectral data acquired in step S210 is input into the quantitative prediction model, the quantitative prediction model processes the data, extracts spectral features, and identifies characteristic wavelengths related to the heavy metal content.

[0082] pass Figure 4 It can be seen that the heavy metal detection method provided by the embodiment of the present application effectively converts the plasma light signal into accurate spectral data through a high-resolution spectrometer, and inputs the spectral data into a quantitative prediction model, which can accurately predict the content of heavy metal elements in the sample. In the embodiment of the present application, the use of a quantitative prediction model can significantly improve the accuracy and efficiency of heavy metal detection, and support fast and reliable sample analysis. Its flexibility and adaptability enable the heavy metal detection method provided by the embodiment of the present application to be widely used in the heavy metal content assessment of various complex samples.

[0083] Please see Figure 5 , Figure 5 A flowchart for determining a quantitative prediction model provided in an embodiment of the present application; in an optional implementation of an embodiment of the present application, the determination process of the quantitative prediction model includes:

[0084] Step S221: determine the number of Gaussian components of the Gaussian mixture model, and define the mean vector and covariance matrix of each Gaussian component to describe the clustering characteristics in the spectrum.

[0085] In the above step S221, it is first necessary to determine the number of Gaussian components (K) contained in the Gaussian mixture model, each Gaussian component represents a feature or cluster in the spectrum. According to the actual data and the expected types of heavy metals, select a suitable K value. For example, if we expect three heavy metals (such as lead, mercury, and cadmium) in the sample, K=3 can be selected. Subsequently, the mean vector (μ) and covariance matrix (Σ) of each Gaussian component are defined, which will help the model describe the clustering characteristics of the spectrum.

[0086] It should be noted that the Gaussian mixture model overcomes the problem of insufficient descriptive ability of a single Gaussian distribution. It uses multiple Gaussian distributions to describe the distribution of a set of data, and introduces latent variables to describe which Gaussian distribution a certain sample point belongs to. As a result, the parameter estimation problem becomes a parameter estimation problem with latent variables.

[0087] Based on the Gaussian peak distribution, a high-dimensional Gaussian mixture model is derived accordingly. Suppose that the points of the approximate ellipsoid distribution in the high-dimensional (d-dimensional) space are described by the probability density function as follows:

[0088]

[0089] Among them, μ is the mean vector, Σ is the covariance matrix of this density function, and these parameters determine the center point, width, direction, etc. of the function shape. However, the spectrum is a one-dimensional plane and there is a superposition form.

[0090] At this time, it can be represented by the weighted average of multiple single Gaussian probability functions, that is, the Gaussian mixture model:

[0091]

[0092] w k is the weight of each branch, and the sum of branch weights is 1.

[0093] The spectrum can be expressed as:

[0094] At this time, the residual σ between the fitted overlapping peak and the actual spectrum Yi is:

[0095]

[0096] For example, please refer to Figure 6 , Figure 6 The Gaussian mixture model provided in the embodiment of the present application is optimized by the expected maximum algorithm EM to convert overlapping peaks into multiple branch peak instance graphs; Figure 6The original data in the figure are the 236-239nm spectral data collected by the spectrometer, and the Al element branch peaks at Al I 236.93nm, Al I 237.31nm, Al I 237.63nm, and Al I 237.84nm are obtained by combining the Gaussian mixture model with the authoritative database. In order to resolve these overlapping spectral peaks, the Gaussian mixture model (GMM) is used to fit the spectral data.

[0097] Initialize the Gaussian mixture model parameters, the number of Gaussian components K = 4, corresponding to the 4 characteristic emission lines of aluminum. According to the characteristic peak positions in the NIST database, μ1 = 236.93nm, μ2 = 237.31nm, μ3 = 237.63nm, μ4 = 237.84nm. Assuming that the resolution of the instrument results in a standard deviation of each peak of σ = 0.05nm, the covariance Σ k =σ 2 =(0.05) 2 =0.0025. Weight w k , initially assume that the weight of each Gaussian component is equal, therefore: w1=w2=w3=w4=0.25; the probability density function of the Gaussian mixture model can be expressed as formula (1), and the probability distribution of the Gaussian mixture model can be expressed as formula (2).

[0098] Step S222: Calculate the posterior probability that each data point in the spectral data belongs to each Gaussian distribution, so as to evaluate the matching degree between each data point and each Gaussian component.

[0099] Step S223: Update the weight, mean and covariance of the Gaussian distribution according to the posterior probability until the parameters of the Gaussian mixture model converge.

[0100] In the above steps S222-S223, the posterior probability of each data point in the spectral data belonging to each Gaussian distribution is calculated using Bayes' theorem to evaluate the matching degree of each data point with each Gaussian component. Based on the calculated posterior probability, the model iteratively updates the weight, mean and covariance of each Gaussian component. This process continues until the model parameters converge, that is, the updated parameter change is less than the set threshold.

[0101] In the above implementation process, the expectation maximization algorithm is applied to optimize the parameters in the Gaussian mixture model GMR so that the model best fits the spectral data. The probability that each spectral point belongs to each Gaussian component is calculated. This step assumes that the current model parameters are correct, and calculates the matching degree of each data point with each Gaussian distribution (i.e., the posterior probability), and recalculates the parameters of each Gaussian distribution, i.e., the mean vector μ, the covariance matrix Σ, and the weights w of each branch. k, and iterates repeatedly until the model parameters converge or the maximum number of iterations is reached. The Gaussian mixture model parameters optimized by the EM algorithm include the final mean, variance, and weight of each component.

[0102] Exemplarily, the expectation-maximization algorithm is applied to optimize the model parameters.

[0103] E step, expectation step: calculate each data point x i The posterior probability γ belonging to each Gaussian distribution ik ,

[0104]

[0105] M step, maximization step: update model parameters according to the posterior probability:

[0106] Update weight w k :

[0107] Update mean μ k :

[0108] Update covariance∑ k :

[0109]

[0110] Repeat the E and M steps until the parameters converge, that is, the change in each iteration is lower than the set threshold.

[0111] Through the optimization of the EM algorithm, the mean, covariance and weight of each Gaussian component are finally obtained. The optimized result can fit the actual spectral data more accurately.

[0112] Step S224: Based on the samples with known heavy metal concentrations, the integrated area of ​​each Gaussian distribution is calculated, and a quantitative relationship between the integrated area and the heavy metal element concentration is established.

[0113] In the above step S224, the accuracy of the model is verified using sample spectral data of known concentrations, and the prediction error of the model (such as mean square error) is evaluated. Based on the verification results, K (number of components) or other GMM parameters are readjusted to improve the accuracy and robustness of the model. Based on the output of the final model, the content of heavy metal elements in industrial wastewater is predicted.

[0114] For example, the integral area A under each Gaussian distribution can be calculated by fitting the parameters obtained by the model. k , which is proportional to the concentration of aluminum. That is, Establish the integral area A k After the quantitative relationship between the concentration of aluminum element C is established, the calibration curve function f(Ak ) Calculate the concentration of aluminum.

[0115] pass Figure 5 It can be seen that in the process of determining the quantitative prediction model of the heavy metal detection method provided in the embodiment of the present application, prior knowledge is introduced, and the characteristic spectrum information of common heavy metal elements in industrial wastewater is combined to initialize the parameters of the GMR model to make it more suitable for practical application scenarios. For large-scale high-dimensional spectral data, the convergence speed and stability of the EM algorithm are improved, and parallel computing or approximate inference methods are introduced to accelerate the model training process. In addition, industrial wastewater samples usually contain a variety of metal oxides, and each heavy metal element produces different emission lines in the laser induced breakdown spectrum. The Gaussian mixture model can decompose these complex mixed spectra into multiple Gaussian distributions, each distribution corresponding to the spectral characteristics of a metal element. This helps to more accurately identify and quantify the concentrations of different heavy metal elements. In the case of coexistence of characteristic peaks of multiple metal elements, the spectral characteristic peaks may overlap, making it difficult to distinguish different components. The Gaussian mixture model can decompose the overlapping spectral peaks by fitting multiple Gaussian distributions, thereby improving the quantitative analysis accuracy of each metal oxide. In the embodiment of the present application, the Gaussian mixture model uses multiple Gaussian distributions to fit the emission line shape in the spectrum. Compared with simple linear or nonlinear regression models, Gaussian distribution can more accurately fit the characteristic spectral peaks of metal elements, thereby obtaining more accurate concentration predictions. Furthermore, a quantitative relationship between the characteristic spectral intensity and content of each metal element is established. Since Gaussian distribution can accurately capture the characteristics of the emission peak, the model can better predict the concentration of metal oxides in the sample.

[0116] Please see Figure 7 , Figure 7 This is a structural schematic diagram of a heavy metal detection device 1000 provided in an embodiment of the present application; an embodiment of the present application provides a heavy metal detection device 1000, which includes a LIBS detection module 100 and a data processing module 200.

[0117] The LIBS detection module 100 is used to focus the Bessel beam on the sample to be tested on the light absorbing substrate to excite the sample to be tested to generate plasma; wherein the Bessel beam is converted from a Gaussian pulse laser.

[0118] The data processing module 200 is used to obtain the content of heavy metal elements in the sample to be tested according to the spectrum signal of the plasma.

[0119] Please continue to see Figure 7 , the LIBS detection module 100 includes: a laser emitting component 110 and a beam adjustment lens group 120;

[0120] The laser emitting assembly 110 is used to send pulsed laser to the beam adjusting lens assembly 120. Figure 7 As shown, the laser emitting assembly 110 includes a nanosecond laser 111 and a timing pulse controller 112 which are connected to each other.

[0121] The laser wavelength emitted by the nanosecond laser 111 is 532nm, the pulse frequency is 1Hz, the energy is 30mJ, and the beam diameter is 7mm. The laser emitted by the nanosecond laser 111 is focused on the sample surface through the beam adjustment lens group 120 to generate plasma.

[0122] The timing pulse controller 112 can generate pulses of TTL amplitude and digitally control the delay time for output, with a delay resolution of 5ps, and is used to accurately control the light emission time of the nanosecond laser 111.

[0123] The beam adjustment lens group 120 is used to convert the Gaussian pulse laser into a Bessel beam, and focus the Bessel beam on the sample to be tested on the light absorbing substrate to excite the sample to be tested to generate plasma.

[0124] Please continue to see Figure 7 The beam adjustment lens assembly 120 includes: a reflecting mirror 121, an aconic lens 122, a focusing lens 123 and a focusing lens 123.

[0125] The reflector 121 is used to receive the pulsed laser and change the direction of the pulsed laser to be perpendicular to the plane where the light absorbing substrate is located. Figure 7 As shown, the laser beam output by the nanosecond laser 111 is in the same plane as the reflector 121, and the reflector 121 is located at 45° in the horizontal direction of the nanosecond laser 111; the reflector 121 can convert the horizontal light emitted by the laser into vertical light.

[0126] The axicon lens 122 is used to convert the pulsed laser into a Bessel beam. For example, the axicon lens 122 provided in the embodiment of the present application has a base angle of 5° and is made of fused silica.

[0127] It should be noted that the Bessel beam can generally be generated by converting a Gaussian beam through an axicon 122. A Bessel beam is a beam whose light amplitude is described by a first-order Bessel function. Its greatest feature is that it is non-diffractive, that is, the radial light field distribution of the Bessel beam does not change within a certain range as the propagation distance increases. Therefore, the Bessel beam is also called a non-diffraction beam. Figure 7 As shown, the axicon 122 is a lens composed of a conical surface and a plane. Due to the conical geometric phase distribution of the axicon 122, the plane wave is transformed into a conical surface wave with a certain cone angle after phase modulation by the axicon 122, thereby generating a Bessel beam.

[0128] The diffraction-free distance length and main lobe width of the Bessel beam mainly depend on the cone angle and refractive index of the conical lens 122. The refraction of the plane wave by the conical lens 122 follows Snell's law, so the cone angle of the Bessel beam is expressed as: β = arcsin (nsinα) - α; where α is the bottom angle of the conical lens 122, and n is the refractive index of the material of the conical lens 122. The cone angle β of the Bessel beam is also called the deflection angle of the conical lens 122. The diffraction-free distance of the Bessel beam generated by the conical lens 122 is closely related to the incident beam waist radius ω0. The diffraction-free distance of the Bessel beam generated by the conical lens 122 can be expressed as: Since the cross-sectional amplitude distribution of the Bessel beam follows the 0th-order Bessel function, the half-width of the central main lobe of the Bessel beam generated by the axicon 122 mirror propagating in free space is expressed as: K is the wave vector of light propagating in free space. Since the axicon 122 generates the Bessel beam by phase modulation, there is almost no energy loss in the process, so the conversion efficiency of using the axicon 122 to generate the Bessel beam is very high.

[0129] The focusing lens 123 is used to focus the Bessel beam on the sample to be tested. Figure 7 As shown, the focusing lens 123 of the heavy metal detection device 1000 provided in the embodiment of the present application includes a first focusing lens 1231 and a second focusing lens 1232. The first focusing lens 1231 is used to change the direction of the light passing through the conical lens 122, and forms a 4f system with the second focusing lens 1232 to expand the transmission distance and the working distance, and complete the beam reduction and focusing. The second focusing lens 1232 changes the direction of the light passing through the first focusing lens 1231, and forms a system with the first lens to expand the transmission distance and the working distance, and complete the beam reduction and focusing.

[0130] like Figure 7 As shown, the aconic lens 122, the first focusing lens 1231, and the second focusing lens 1232 are arranged directly below the reflector 121, and the front focus of the second aconic lens 122 and the rear focus of the first lens are ensured to coincide.

[0131] Please continue to see Figure 7 The data processing module 200 includes a spectral signal acquisition component 210 , a spectral signal conversion component 220 and a quantitative prediction unit 230 .

[0132] The spectrum signal acquisition component 210 is used to acquire spectrum signals. Figure 7 As shown, the spectrum signal acquisition component 210 in the heavy metal detection device 1000 provided in the embodiment of the present application includes: a collection lens 211, a fiber optic probe 212 and an optical fiber 213. Figure 7As shown, the optical fiber probe 212 is placed on the left side of the collection lens 211, and the optical fiber probe 212 is connected to the high-resolution spectrometer through the optical fiber 213 to transmit the plasma optical signal. Among them, the focal length of the collection lens 211 is 5cm, and the generated plasma optical signal is focused on the optical fiber probe 212; the optical fiber probe 212 can collect the plasma optical signal focused by the collection lens 211, and the optical fiber 213 is used to transmit the plasma optical signal to the high-resolution spectrometer.

[0133] The spectral signal conversion component 220 is used to convert the spectral signal into spectral data. Figure 7 As shown, the spectrum signal conversion component 220 in the embodiment of the present application includes: a high-resolution spectrometer, which is connected to the timing pulse controller 112, and the timing pulse controller 112 controls the time of the high-resolution spectrometer. The wavelength range of the high-resolution spectrometer in the embodiment of the present application is 200-1100nm, and the spectrum resolution is 0.4nm; it can convert the plasma optical signal transmitted by the optical fiber 213 into an electrical signal.

[0134] The quantitative prediction unit 230 is configured to obtain the heavy metal element content in the sample to be tested according to the spectral data. Figure 7 As shown, the quantitative prediction unit 230 provided in the embodiment of the present application is a computer 240 including a quantitative prediction model, which further converts the electrical signal converted from the high-resolution spectrum into spectral data. The high-resolution spectrometer is electrically connected to the computer 240 to transmit the electrical signal converted from the plasma optical signal; the computer 240 transmits the first spectral data to the given quantitative prediction model. The quantitative prediction model in the embodiment of the present application is built with a Gaussian Mixture Regression (GMR) model optimized for heavy metal detection in industrial wastewater. The model combines the characteristic spectral information of common metal elements in industrial wastewater, and performs targeted initialization and optimization of model parameters. Its function is to quantitatively analyze the spectral data obtained by the computer 240, and improve the ability to analyze complex spectral data by adaptively adjusting the number of Gaussian components and optimizing the solution algorithm, and finally obtain the content of metal elements in industrial wastewater.

[0135] like Figure 7 As shown, the heavy metal detection device 1000 also includes an XY displacement platform 300, on the surface of which a surface microporous substrate with a dried sample can be placed, and the X and Y directions can be accurately moved to ensure that the laser uniformly hits the surface of the surface microporous substrate and the consistency of the defocus amount during the spectrum acquisition process moves along the specified path according to the set linear velocity and acceleration.

[0136] Please see Figure 8 , Figure 8The light absorbing substrate processing device 400 provided in the embodiment of the present application, in an optional implementation of the embodiment of the present application, the heavy metal detection device 1000 also includes a light absorbing substrate processing device 400, and the light absorbing substrate processing device 400 is used to use a femtosecond laser to process the surface of the metal substrate 450 to form a surface microporous structure substrate having multiple deep holes H; wherein the deep holes H are periodically and regularly arranged on the surface microporous structure substrate.

[0137] like Figure 8 As shown, the light absorbing substrate processing device 400 includes a femtosecond laser 410, a spatial light modulator 420, a third focusing lens 430, an objective lens 440, a metal substrate 450, a displacement platform controller 460, an XYZ displacement platform 470, a computer 490 and a deep hole-H.

[0138] The femtosecond laser 410 emits an ultrashort pulse laser with a pulse width of femtoseconds, which is used to process the substrate into a substrate with a surface microporous structure. Exemplarily, the emission wavelength is 515nm, the repetition frequency is 1MHz, the beam diameter is 2mm, and the pulse width is 150fs.

[0139] The mechanical switch 490 is controlled by the computer 240 to adjust the output of the laser to selectively ablate the substrate.

[0140] The spatial light modulator 420 performs phase modulation on the femtosecond Gaussian beam to generate a Bessel beam.

[0141] The third focusing lens 430 can change the direction of the light passing through the spatial light modulator 420, and form a 4f system with the objective lens 440 to expand the transmission distance and the working distance, and complete the beam reduction and focusing.

[0142] The objective lens 440 is used to change the direction of the light passing through the third focusing lens 430, and forms a 4f system with the third focusing lens 430 to expand the transmission distance and the working distance, thereby completing the beam reduction and focusing of the light beam.

[0143] The displacement platform controller 460 is used to receive instructions from the computer 240 to drive the high-precision displacement platform to complete sample processing.

[0144] The surface of the high-precision XYZ displacement platform 470 can be used to place a surface microporous substrate with a dried sample. The distance from the focus to the sample surface can be precisely controlled in the Z direction, and precise movement in the X and Y directions can be performed to ensure that the laser evenly hits the surface of the surface microporous substrate according to the set linear velocity and acceleration during the spectrum acquisition process.

[0145] The computer 240 synchronously controls the mechanical switch 490 and the displacement platform driver 460 to complete high-precision displacement and selective ablation of the substrate to form a preset processed substrate pattern.

[0146] In general, the heavy metal detection method and device based on the surface microporous structure substrate provided by the embodiment of the present application designs a surface microporous structure substrate with a specific structure to enhance the detection ability of metal oxides; by converting the Gaussian beam into a Bessel beam, the sample pretreatment steps of the traditional droplet drying method are simplified, and the spectral data quality is optimized; finally, the Gaussian mixture regression model is used to process the collected spectral data, and the overlapping peaks are decomposed into independent characteristic peaks, which significantly improves the quantitative prediction ability of heavy metal elements. The accuracy and efficiency of heavy metal detection in industrial wastewater are comprehensively improved from multiple aspects.

[0147] See also Fig. 9 , Fig. 9 The electronic device 500 provided in the embodiment of the present application includes: a processor 501 and a memory 502, wherein the memory 502 stores machine-readable instructions executable by the processor 501, and the machine-readable instructions are executed by the processor 501 to perform the steps in any implementation of the above-mentioned heavy metal detection method.

[0148] Based on the same inventive concept, an embodiment of the present application also provides a computer-readable storage medium, in which computer program instructions are stored. When the computer program instructions are read and executed by a processor, the steps in any implementation method of the above-mentioned heavy metal detection method are executed.

[0149] The computer-readable storage medium may be a random access memory (RAM), a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or any other medium that can store program code.

[0150] In this article, relational terms such as first and second, etc. are used only to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the statement "include..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0151] The above description is only an embodiment of the present application and is not intended to limit the protection scope of the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A heavy metal detection method based on a surface microporous structure substrate, characterized in that: The method comprises: Focusing the Bessel beam on the sample to be tested on the light absorbing substrate to excite the sample to be tested to generate plasma; wherein the Bessel beam is converted from a Gaussian pulse laser; The content of heavy metal elements in the sample to be tested is obtained according to the spectrum signal of the plasma.

2. The method according to claim 1, characterized in that The method for preparing the light absorbing substrate comprises: A surface microporous structure having a plurality of deep holes is formed on the surface of a metal substrate by using a femtosecond laser; wherein the deep holes are periodically and regularly arranged on the surface microporous structure substrate.

3. The method according to claim 1, characterized in that Before focusing the Bessel light beam on the sample to be tested on the light absorbing substrate, the heavy metal detection method further includes: adding a sample to be tested to the light absorbing substrate; The sample to be tested is dried to obtain the sample to be tested.

4. The method according to claim 1, characterized in that The step of obtaining the heavy metal element content in the sample to be tested according to the spectral signal of the plasma comprises: converting the spectral signal into spectral data; The spectral data is input into a quantitative prediction model to obtain the heavy metal element content.

5. The method according to claim 4, characterized in that in, The quantitative prediction model determination process includes: Determine the number of Gaussian components of the Gaussian mixture model, and define the mean vector and covariance matrix of each Gaussian component to describe the clustering characteristics in the spectrum; Calculating the posterior probability that each data point in the spectral data belongs to each Gaussian distribution to evaluate the matching degree of each data point with each Gaussian component; According to the posterior probability, the weight, mean and covariance of the Gaussian distribution are updated until the parameters of the Gaussian mixture model converge; Based on samples with known heavy metal concentrations, the integral area of ​​each Gaussian distribution was calculated, and a quantitative relationship between the integral area and the heavy metal element concentration was established.

6. A heavy metal detection device based on a substrate with a surface microporous structure, characterized in that: The heavy metal detection device comprises: a LIBS detection module and a data processing module; The LIBS detection module is used to focus the Bessel beam on the sample to be tested on the light absorbing substrate to excite the sample to be tested to generate plasma; wherein the Bessel beam is converted from a Gaussian pulse laser; The data processing module is used to obtain the content of heavy metal elements in the sample to be tested according to the spectrum signal of the plasma.

7. The heavy metal detection device according to claim 6, characterized in that: The LIBS detection module includes: a laser emission component and a beam adjustment lens group; The laser emission assembly is used to send Gaussian pulse laser to the beam adjustment mirror group; The beam adjustment mirror group is used to convert the Gaussian pulse laser into a Bessel beam, and focus the Bessel beam on the sample to be tested on the light absorbing substrate to excite the sample to be tested to generate plasma.

8. The heavy metal detection device according to claim 7, characterized in that: The beam adjustment lens assembly comprises: a reflector, an aconic lens and a focusing lens; The reflector is used to receive the pulse laser and change the direction of the Gaussian pulse laser to be perpendicular to the plane where the light absorbing substrate is located; The axicon is used to convert the Gaussian pulse laser into the Bessel beam; The focusing lens is used to focus the Bessel light beam on the sample to be tested.

9. The heavy metal detection device according to claim 6, characterized in that: The data processing module includes a spectral signal acquisition component, a spectral signal conversion component and a quantitative prediction unit; The spectrum signal acquisition component is used to collect the plasma light signal and introduce it into the spectrum signal conversion component; The spectral signal conversion component is used to convert the spectral signal into spectral data; The quantitative prediction unit is configured to obtain the content of heavy metal elements in the sample to be tested according to the spectral data.

10. The heavy metal detection device according to claim 6, characterized in that: The heavy metal detection device also includes a light absorbing substrate processing device, which is used to use a femtosecond laser to process the surface of a metal substrate to form a surface microporous structure substrate with multiple deep holes; wherein the deep holes are periodically and regularly arranged on the surface microporous structure substrate.

Images