Aerosol analysis method based on high-repetition-rate LIBS and two-dimensional correlation spectroscopy

Abstract

The application discloses an aerosol analysis method based on high-repetition-frequency LIBS and two-dimensional correlation spectroscopy, and belongs to the technical field of multispectral measurement. The method repeatedly irradiates an aerosol sample by using pulsed laser with a repetition frequency of more than 1 kHz to excite plasma; the light signals emitted by the plasma are collected multiple times under different repetition frequencies to form original LIBS spectral data; the original LIBS spectral data are preprocessed to form a two-dimensional LIBS spectral data matrix, and the cross-correlation coefficients between each spectral component and two-dimensional correlation synchronous spectrum are calculated; by identifying the autocorrelation peaks and cross peaks in the two-dimensional correlation synchronous spectrum, the characteristic spectral lines of particulate matter elements in the aerosol sample are extracted. The application combines kilohertz high-repetition-frequency pulsed laser with two-dimensional correlation spectroscopy analysis to realize high-dimensional spectral information acquisition and analysis of discrete heterogeneous aerosol samples, and improves the aerosol LIBS detection flux and analysis performance.

Description

TECHNICAL FIELD

[0001] The present application belongs to the technical field of multispectral measurement, and particularly relates to an aerosol analysis method based on high-repetition-rate LIBS and two-dimensional correlation spectroscopy. BACKGROUND

[0002] An aerosol refers to a gaseous dispersion system composed of solid or liquid particles suspended in a gaseous medium, and atmospheric aerosols have the characteristics of multiple sources and mixed forms. The atmospheric aerosols can affect the earth's radiation balance by absorbing or scattering sunlight and condensing into clouds, and participate in global atmospheric processes and material transfer, which are necessarily related to the frequent extreme abnormal weather in recent years. Therefore, high-throughput chemical information measurement of atmospheric aerosols has important scientific research and environmental values, and can provide an original database for establishing a global weather system model.

[0003] Existing aerosol analysis technologies mainly include the following three types:

[0004] Firstly, the traditional filter membrane technology. That is, a technology for filtering, adsorbing and sampling various properties of aerosols in the air by using a filter membrane. This technology is not suitable for online aerosol measurement, and therefore has the disadvantage of failing to preserve the original characteristics of the aerosol.

[0005] Secondly, the online aerosol mass spectrum detection technology, such as AMS, ATOFMS, SPAMS, etc. That is, a technology for ionizing particles in the aerosol into positive and negative ions, collecting the positive and negative ions to form a mass spectrum, and analyzing the composition and concentration of chemical substances in the aerosol according to the position and intensity of the characteristic peaks in the mass spectrum. This technology overcomes the disadvantage of the traditional filter membrane sampling technology that fails to preserve the original characteristics of the aerosol, but lacks the ability to detect element component information matching organic matter detection, and is difficult to perform high temporal and spatial resolution measurement.

[0006] Thirdly, Laser-Induced Breakdown Spectroscopy (LIBS) technology. That is, using high-energy pulse laser and aerosol interaction to produce transient plasma, through the analysis of plasma emission spectrum, realizing the qualitative and quantitative detection of chemical elements in aerosol spectrum technology. This technology has the advantages of simultaneous multi-element detection, no sample pretreatment, etc., and has been widely used in the analysis of atmospheric, biological and industrial emissions and other aerosols. However, so far, LIBS mostly uses low pulse frequency (≤10Hz) analysis system, and the low-frequency LIBS technology has very low particle detection flux and sampling efficiency, which is difficult to conduct qualitative and quantitative analysis of natural discrete aerosol samples with mathematical statistical significance. Moreover, laser-induced breakdown and plasma expansion process will atomize background gas and particles at the same time, which will easily lead to the characteristic spectrum of particles being submerged in strong continuous background and gas element characteristic radiation, so it is difficult to meet the complex spectral analysis requirements of natural discrete aerosol samples. SUMMARY

[0007] The purpose of the present application is to provide an aerosol analysis method based on high-repetition-rate LIBS and two-dimensional correlation spectroscopy, to solve the problems of low sampling rate, poor spectral resolution and difficulty in effective statistical analysis of traditional aerosol low-frequency LIBS technology.

[0008] To solve the above technical problems, the present application adopts the following technical solutions:

[0009] An aerosol analysis method based on high-repetition-rate LIBS and two-dimensional correlation spectroscopy, comprising:

[0010] Using a pulsed laser with a repetition frequency of more than 1 kHz to repeatedly irradiate the aerosol sample to generate plasma;

[0011] Setting n repetition frequency points, and collecting the light signal emitted by the plasma m times under each repetition frequency condition to form n*m spectrum records and constitute original LIBS spectrum data;

[0012] Pretreating the original LIBS spectrum data to form a two-dimensional LIBS spectrum data matrix;

[0013] Calculating the cross-correlation coefficient between each spectrum component in the two-dimensional LIBS spectrum data matrix and the two-dimensional correlation synchronous spectrum;

[0014] Identifying the autocorrelation peak and cross peak in the two-dimensional correlation synchronous spectrum to extract the characteristic spectrum of the particles in the aerosol sample.

[0015] In some embodiments of the present application, the frequency range involved in the n heavy frequency points should meet the condition that the perturbation range of the pulse laser repetition frequency can produce both mutually independent events and related events between pulses; wherein the mutually independent events are the case that the plasma generated by the previous pulse laser excitation has been quenched when the next pulse laser arrives; the related events are the case that the plasma generated by the previous pulse laser excitation still exists when the next pulse laser arrives. When the laser pulse repetition frequency is high enough, i.e. the interval between two pulse lasers is small enough, the intermediate products after the previous plasma de-excitation can be detected, and the chemical information mapped by these intermediate products is completely different from that of the original sample, because the high repetition rate pulse laser action makes the originally isolated events become related events. Unlike the perturbation mechanism of conventional parameters such as laser pulse energy, the perturbation range of laser pulse repetition frequency from kilohertz to tens of kilohertz can produce both mutually independent events and related events between pulses. This strategy of pulse repetition frequency modulation is equivalent to providing higher dimensional measurements, which is more conducive to the identification of particle element characteristic radiation and air element characteristic radiation in complex aerosol systems.

[0016] In some embodiments of the present application, the perturbation range of the pulse laser repetition frequency is preferably set to be between 1 kHz and 40 kHz, so that the shortest interval between adjacent pulses can be 25 μs. Since the dissociation time of aerosol particles is tens of microseconds, the single-pulse LIBS isolated events under high repetition rate pulse excitation can be changed into related events, which not only linearly increases the LIBS detection flux and statistical effectiveness, but also produces a nonlinear enhancement effect similar to double pulses.

[0017] In some embodiments of the present application, the process of preprocessing the original LIBS spectral data includes:

[0018] Screening effective sampling spectral data from the original LIBS spectral data, the effective sampling spectral data being the LIBS spectrum containing particle element characteristic radiation, i.e. the LIBS event corresponding to effective ablation and excitation of aerosol particles by laser;

[0019] Performing background subtraction processing on the screened effective sampling spectral data to eliminate the interference of LIBS continuous background radiation;

[0020] Taking the average of the effective sampling spectral data under the same heavy frequency point condition after the background subtraction processing to form n groups of average spectral data;

[0021] Using the n groups of average spectral data to form a two-dimensional LIBS spectral data matrix to facilitate the calculation of cross-correlation coefficients and the formation of two-dimensional correlation synchronous spectrum.

[0022] In some embodiments of the present application, when screening effective sampling spectrum data, it is preferred to screen with a threshold of signal-to-noise ratio of low-energy state radiation spectral line of major elements of particulate matter in the spectrum being higher than 3.

[0023] In some embodiments of the present application, the two-dimensional LIBS spectrum data matrix is a two-dimensional data matrix with wavelength and repetition frequency of plasma emission light as independent variables. That is, rows and columns of the two-dimensional data matrix are respectively wavelength and perturbation factor, each column is an average spectrum under a certain repetition frequency value, and each row is a response of a certain spectral component with repetition frequency.

[0024] In some embodiments of the present application, the process of calculating the cross-correlation coefficient between each spectral component and the two-dimensional correlation synchronous spectrum in the two-dimensional LIBS spectrum data matrix comprises:

[0025] The dynamic spectrum of each spectral component disturbed by the repetition frequency of laser pulses is calculated, and the calculation formula is:

[0026] ;

[0027] wherein, represents the dynamic spectrum; represents the i th value of the repetition frequency; represents the spectral intensity value measured under the condition of wavelength λ and repetition frequency ; represents the average value of spectral intensity at wavelength λ under different repetition frequencies;

[0028] The cross-correlation coefficient between each spectral component is calculated by using fast discrete Fourier transform, and the calculation formula is:

[0029] , ;

[0030] wherein, represents the cross-correlation coefficient of two spectral components and with repetition frequency x; F represents fast discrete Fourier transform; and respectively represent the dynamic spectral components disturbed by the repetition frequency x at wavelength and ; represents Hadamard product; * represents complex conjugate; represents a complex set;

[0031] The inverse Fourier transform of the cross-correlation coefficient between each spectral component is performed , take the real part of the first element of each vector after inverse Fourier transform, and form the two-dimensional correlation synchronous spectrum.

[0032] In some embodiments of the present application, the feature spectrum extraction process comprises:

[0033] Extracting spectral intensity values on the diagonal of the two-dimensional correlation synchronous spectrum, and removing random noise to obtain an autocorrelation peak.

[0034] Extracting a tangent line in the two-dimensional correlation synchronous spectrum with a wavelength equal to the typical characteristic peak of the particulate matter element in the aerosol sample to obtain a cross peak.

[0035] The element characteristic spectrum in the aerosol particle is separated from the element characteristic spectrum in the air by the cross peak.

[0036] Compared with the prior art, the advantages and positive effects of the present application mainly lie in:

[0037] 1. The present application uses high-repetition-rate pulsed laser as excitation light source (from less than 10Hz in the prior art to kilohertz to tens of kilohertz), which increases the interaction frequency between laser and aerosol sample. On the one hand, it can linearly improve the system detection flux and statistical effectiveness, and improve the sampling efficiency of LIBS for discrete aerosol samples. On the other hand, it can utilize the mutual cooperation between consecutive pulses to re-excite the intermediate products induced by the previous pulse, and obtain nonlinear enhancement effect.

[0038] 2. The present application uses the repetition frequency of pulsed laser as a disturbance factor, and sets the value range of the repetition frequency to 1kHz-40kHz to produce independent events and related events between LIBS pulses, so that the characteristic element radiation of aerosol particles and the background radiation show different trends with the repetition frequency, and thus the two-dimensional correlation spectrum analysis method can accurately identify the characteristic element radiation of aerosol particles. Therefore, the problem that the two-dimensional correlation spectrum analysis method in the prior art uses low-repetition-rate laser pulse energy as a disturbance factor, and it is difficult to identify the characteristic element radiation of aerosol particles from the background radiation because the characteristic element radiation of aerosol particles and the background radiation intensity have the same linear enhancement trend.

[0039] 3. The present application organically integrates high-repetition-rate LIBS excitation method and two-dimensional correlation spectrum analysis method, and the two-dimensional correlation spectrum analysis method can analyze the linear and nonlinear interaction process between high-repetition-rate laser pulses and discrete aerosol particles in the system, thereby greatly improving the information abundance and reliability of aerosol LIBS spectral data.

[0040] Other features and advantages of the present application will become more apparent after reading the detailed description of the embodiments of the present application in conjunction with the accompanying drawings. Attached Figure Description

[0041] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0042] Figure 1 This is a flowchart of an embodiment of the aerosol analysis method based on high repetition rate LIBS and two-dimensional correlation spectroscopy proposed in this invention;

[0043] Figure 2 This is a schematic diagram of the overall architecture of one embodiment of the spectral acquisition experimental device;

[0044] Figure 3 This is a two-dimensional LIBS spectrum of aerosols after pretreatment;

[0045] Figure 4 It is a two-dimensional correlation synchronous spectrum of aerosol LIBS;

[0046] Figure 5 It is by Figure 4 The waveform of the autocorrelation peak extracted from the diagonal of the LIBS two-dimensional correlation synchronization spectrum;

[0047] Figure 6 It is by Figure 4 The spectral component extracted from the LIBS two-dimensional correlation synchronous spectrum is the cross-peak waveform at 589.0 nm.

[0048] In the diagram, 1. Laser; 2. Half-wave plate; 3. Polarizing beam splitter prism; 4. First converging lens; 5. Photoelectric converter; 6. Oscilloscope; 7. Spectrometer; 8. Digital delay signal generator; 9. Processor; 10. Sample container; 11. Aerosol inlet; 12. Aerosol outlet; 13. Collecting lens; 14. Second converging lens; 15. Optical fiber. Detailed Implementation

[0049] The specific embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0050] It should be noted that in the description of this invention, terms such as "front" and "rear," which indicate direction or positional relationship, are based on the direction or positional relationship shown in the accompanying drawings. This is merely for ease of description and does not indicate or imply that the device or element must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, it should not be construed as a limitation of this invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.

[0051] Furthermore, it needs to be explained that in the description of the present application, unless otherwise explicitly specified and limited, the terms "connected", "connected" should be understood broadly. For example, it can be fixedly connected, or it can be detachably connected or integrally connected; it can be directly connected, or it can be indirectly connected through an intermediate medium, and it can be connected inside two elements. For those skilled in the art, the specific meaning of the above terms in the present application can be understood according to the specific circumstances.

[0052] The aerosol analysis method of the embodiment, aiming at the problems of low sampling rate and poor spectral resolution of traditional aerosol low-frequency LIBS technology, first uses high-repetition-rate pulsed laser to cooperatively sample and excite aerosol particles, which can linearly improve the LIBS detection flux and nonlinearly improve the sampling efficiency; secondly, aiming at the linear and nonlinear enhancement effect of high-repetition-rate pulsed laser acting on aerosol particles, a two-dimensional correlation spectrum analysis method is introduced to analyze the high-dimensional spectrum of linear and nonlinear effects, so as to selectively extract the characteristic radiation spectrum of particulate matter in aerosol. From the aspects of spectral excitation and spectral analysis, the real-time analysis performance of aerosol LIBS is jointly improved.

[0053] In order to excite, dissociate and collect spectra of the aerosol sample, the embodiment first designs a spectral collection experimental device, as shown in Figure 2 The device mainly includes a laser 1, a spectrometer, a photoelectric converter 5, an oscilloscope 6, a spectrometer 7, a digital delay signal generator 8, a processor 9, a sample container 10 and the like.

[0054] Among them, the laser 1 preferably uses a semiconductor-pumped solid-state laser to generate pulsed laser with adjustable repetition frequency. A half-wave plate 2 and a polarization beam splitter prism 3 are arranged in front of the laser 1 in turn to form a spectrometer, which is used to divide the pulsed laser emitted by the laser 1 into two paths: one path is focused by a first converging lens 4, irradiates the aerosol sample in the sample container 10, and makes the energy density at the laser focal point reach the aerosol particle breakdown threshold, realizing the ablation, dissociation and excitation of particulate matter; the other path is collected by a photoelectric converter 5, and converted into an electrical signal, then sent to an oscilloscope 6 for monitoring the energy and repetition frequency of the pulsed laser. In the embodiment, the photoelectric converter 5 can use a photodiode to realize the conversion of optical signal to electrical signal.

[0055] An aerosol sample can be injected into the sample container 10 through the aerosol injection port 11 at the top of the sample container 10, and the sample container 10 is required to be filled with the aerosol sample to avoid the aerosol analysis result being inaccurate due to the remaining of other gases in the sample container 10. An aerosol outlet 12 can be arranged at the bottom of the sample container 10, and the aerosol in the sample container 10 can be completely discharged through the aerosol outlet 12 after the experiment is completed to prevent residue.

[0056] After the interaction between the pulsed laser and the aerosol sample, a transient plasma is generated. The light signal emitted by the plasma is formed into parallel light by the collection lens 13, and then propagates in a straight line to the second converging lens 14. After being converged by the second converging lens 14, the light signal is transmitted to the spectrometer 7 through the optical fiber 15 to generate spectral data and a spectrum. The processor 9 reads the spectral data generated by the spectrometer 7 for analyzing the element components of the aerosol sample.

[0057] The digital delay signal generator 8 is connected to the laser 1, the oscilloscope 6 and the spectrometer 7 for synchronously triggering the three instruments to complete the laser emission and the synchronous collection of the light signal. The processor 9 can be configured to communicate with the digital delay signal generator 8 for adjusting the repetition frequency of the pulsed laser output by the laser 1.

[0058] The aerosol analysis method of the embodiment will be described in detail below. Figure 1

[0059] S101, high repetition frequency LIBS spectrum excitation.

[0060] In order to improve the sampling efficiency and detection flux of LIBS, the aerosol sample is repeatedly irradiated by pulsed laser with a frequency of more than 1 kHz to generate plasma.

[0061] As a preferred embodiment, the pulsed repetition frequency can be set in the range of kilohertz to tens of kilohertz, for example, between 1 kHz and 40 kHz. The pulsed laser with adjustable high repetition frequency is focused on the aerosol sample, and then the particulate matters in the aerosol are excited, dissociated and light signal collected under the combined action of the pulsed laser, the laser-induced plasma and the shock wave.

[0062] Unlike the perturbation mechanism of the conventional parameters such as the pulsed laser energy, the perturbation range of the kilohertz to tens of kilohertz pulsed laser repetition frequency can generate two cases of independent events and related events between the pulses. This strategy of pulsed repetition frequency modulation is equivalent to providing higher dimensional measurement, which is more conducive to the identification of the characteristic radiation of the particulate elements and the characteristic radiation of the air elements in the complex aerosol system.

[0063] ​S102, two-dimensional LIBS spectral data acquisition.

[0064] In this embodiment, the repetition frequency of the pulsed laser is taken as the disturbance factor. The repetition frequency of the pulsed laser is modulated, and the light signals emitted by the plasma under different repetition frequency conditions are collected to obtain two-dimensional LIBS spectral data.

[0065] In some embodiments, n repetition frequency points can be set in a frequency range of kilohertz to tens of kilohertz, and the light signals emitted by the plasma are collected m times under each repetition frequency condition to form n*m spectral records, which in turn constitute the original LIBS spectral data.

[0066] For example, 9 repetition frequency points can be set in a frequency range of 1 kHz to 40 kHz, i.e., n=9, which are 1 kHz, 5 kHz, 10 kHz, 15 kHz, 20 kHz, 25 kHz, 30 kHz, 35 kHz, and 40 kHz. The laser pulse energy is set to 850 μJ, and the spectrometer 7 uses the same exposure time of 40 ms. 1000 LIBS spectral data are measured under each repetition frequency condition, i.e., m=1000, and 9 groups are collected, thereby forming 9*1000 spectral records, which constitute the original LIBS spectral data.

[0067] After the pulsed laser excites the aerosol to generate the plasma, the emission light of the plasma is complex light, which radiates in the wavelength range of 300 nm to 850 nm. A plurality of sampling points can be set in the wavelength range of 300 nm to 850 nm, for example, 4094 sampling points are set. Thus, 1000*4094 spectral data can be formed in each spectral record formed by collecting the emission light of the plasma in the wavelength range of 300 nm to 850 nm using the spectrometer. The wavelength range of 300 nm to 850 nm can cover the entire visible light band, and the spectra of elements Na and Cl in the aerosol and elements O, N, and H in the air are basically within this wavelength range.

[0068] The LIBS spectra of the aerosol sample are collected under multiple pulsed laser repetition frequency conditions, thereby expanding the one-dimensional spectrum into a two-dimensional spectrum. When the repetition frequency of the pulsed laser is high enough, i.e., the emission time interval of two pulsed lasers is small enough, the intermediate products after the de-excitation of the previous plasma can be detected. The chemical information mapped by these intermediate products is completely different from that of the original sample. Due to the action of the high repetition frequency pulse, the originally isolated events become related events.

[0069] Different from the perturbation mechanism of conventional parameters such as pulse laser energy, the perturbation range of kilohertz to tens of kilohertz pulse laser repetition frequency can produce two cases of independent events and related events between pulses. The strategy of pulse repetition frequency modulation is equivalent to providing higher-dimensional measurement, which is more conducive to the identification of particle element characteristic radiation and air element characteristic radiation in a complex aerosol system.

[0070] Here, the independent event refers to the case that the plasma generated by the previous pulse laser excitation has been quenched when the next pulse laser arrives. The related event refers to the case that the plasma generated by the previous pulse laser excitation still exists when the next pulse laser arrives.

[0071] Assuming that the lifetime of the plasma is 40 microseconds, the pulse intervals of 30 kHz, 35 kHz, and 40 kHz are less than 40 microseconds, it is considered that the events between the pulses of the pulse laser acting on the aerosol sample are related, i.e., related events, because the plasma generated by the previous laser pulse still exists when the next laser arrives. When the repetition frequency of the pulse laser is less than 25 kHz, the plasma generated by the previous laser pulse has been quenched when the next laser arrives, it is considered that the process of the pulse laser acting on the aerosol sample is an independent event.

[0072] In this embodiment, when n frequency points are set, the frequency range involved in the n frequency points should cover both the independent events and the related events between pulses, so that not only the LIBS detection flux and statistical effectiveness can be linearly increased, but also a nonlinear enhancement effect similar to double pulses can be generated.

[0073] S103, two-dimensional LIBS spectral data preprocessing.

[0074] The original LIBS spectral data collected are preprocessed to form a two-dimensional LIBS spectral data matrix.

[0075] The two-dimensional LIBS spectral data preprocessing process of the embodiment mainly includes the following three parts:

[0076] First, the selection process of effective sampling spectral data;

[0077] That is, the process of selecting LIBS spectral data containing particle element characteristic radiation from original LIBS spectral data.

[0078] Specifically, the spectral data of repeated measurements under the same repetition frequency condition can be screened for "effective sampling spectrum". The effective sampling spectrum here refers to the LIBS spectrum containing particle element characteristic radiation, i.e., the LIBS event corresponding to the effective ablation and excitation of aerosol particles by pulse laser.

[0079] In some embodiments, the effective sampling spectrum data can be screened with a threshold of signal-to-noise ratio of the low-energy state radiation spectrum line of the main element of the particulate matter in the spectrum being higher than 3.

[0080] For example, for the case of a solid-phase NaCl aerosol sample, the threshold condition can be whether the signal-to-noise ratio of the low-energy state radiation spectrum line Na I 589.0 nm of the Na element exceeds 3, and the spectrum data with a signal-to-noise ratio greater than 3 is screened to form the effective sampling spectrum data.

[0081] Secondly, a background reduction process;

[0082] That is, the effective sampling spectrum data screened is subjected to a background reduction process to eliminate the interference of the continuous background radiation of LIBS.

[0083] In some embodiments, for the screened effective sampling spectrum, the background reduction process can be performed using the msbackadj function provided by Matlab.

[0084] Thirdly, an averaging process;

[0085] That is, the effective sampling spectrum data under the same repetition frequency point condition after the background reduction process is averaged.

[0086] Specifically, under each value condition of the repetition frequency, for example, when the repetition frequency is 1 kHz, the average value of all the effective sampling spectra after the background reduction process is calculated to form the average spectrum data under the condition of the repetition frequency.

[0087] The average spectrum data under each value condition of the repetition frequency is used to form a two-dimensional LIBS spectrum data matrix, with wavelength and repetition frequency as the independent variables, that is, each column is an average spectrum under a certain value condition of the repetition frequency, and each row is the response of a certain spectral component with the repetition frequency.

[0088] Figure 3 An aerosol two-dimensional LIBS spectrum graph formed by the two-dimensional LIBS spectrum data matrix is shown. Figure 3 It can be seen that when the pulse laser repetition frequency is increased from 1 kHz to 30 kHz, the spectral intensity is increased by about 6 times, indicating that high repetition frequency pulse laser has a nonlinear enhancement effect on aerosol LIBS.

[0089] S104, two-dimensional correlation synchronous spectrum calculation.

[0090] The two-dimensional data matrix formed by the preprocessed aerosol LIBS spectrum is used to calculate the dynamic spectrum of each spectral component disturbed by the repetition frequency, and the cross-correlation coefficient between each two spectral components is calculated through fast discrete Fourier transform, and then a two-dimensional correlation synchronous spectrum graph is obtained.

[0091] The process is specifically as follows:

[0092] The dynamic spectrum of each spectral component disturbed by the repetition frequency of laser pulse is calculated, and the calculation formula is:

[0093]

[0094] Among them, represents the dynamic spectrum; represents the i-th value of the repetition frequency, i.e. the i-th repetition frequency point; represents the spectral intensity value measured under the condition that the wavelength is λ and the repetition frequency is ; represents the average value of the spectral intensity measured at the wavelength λ under different repetition frequencies.

[0095] The cross-correlation coefficient between each two spectral components is calculated by using fast discrete Fourier transform, and the calculation formula is:

[0096] ;

[0097] Among them, represents the cross-correlation coefficient of two spectral components and varying with the repetition frequency x; F represents fast discrete Fourier transform; and respectively represent the dynamic spectral components disturbed by the repetition frequency x at the wavelengths and ; represents Hadamard product; * represents complex conjugate; represents a complex set.

[0098] The inverse Fourier transform of the cross-correlation coefficient between each two spectral components is performed, i.e.

[0099] ;

[0100] The real part of the first element of each vector after the inverse Fourier transform is taken to form a two-dimensional correlation synchrospectrum. Figure 4 The two-dimensional correlation synchrospectrum is shown.

[0101] S105, two-dimensional correlation spectrum feature extraction.

[0102] The autocorrelation peak and cross peak in the two-dimensional correlation synchrospectrum are identified, and the characteristic spectral line of the particulate matter element in the aerosol sample is extracted.

[0103] The specific process is as follows:

[0104] ​​The spectral intensity values ​​on the diagonal of the two-dimensional correlation synchronous spectrum are extracted to obtain the autocorrelation peaks, as shown in the figure. Figure 5 As shown. Elemental characteristic radiation lines are extracted from the autocorrelation peaks on the diagonal of the two-dimensional correlation synchronization spectrum, which can eliminate random noise. Figure 5 As can be seen, compared with the average spectrum (gray dotted line) after only preprocessing, the signal-to-noise ratio of the obtained autocorrelation peak (black solid line) is improved by two orders of magnitude.

[0105] Tangents at wavelengths equal to the typical characteristic peaks of particulate matter elements in the aerosol sample are extracted from the two-dimensional correlation synchronous spectrum to obtain cross-peaks. In this embodiment, if NaCl salt particles are used as the aerosol sample, the tangent at 589.0 nm of Na I is extracted to obtain the cross-peak, as shown in the figure below. Figure 6 As shown. If other aerosol samples are used, the tangent wavelength can be selected according to the sample composition.

[0106] Depend on Figure 6 As can be seen, the cross-peaks in the two-dimensional correlation synchronous spectrum completely separate the characteristic spectral lines of aerosol particles (Na, Cl) from those of air elements (N, O, H). This demonstrates that the aerosol analysis method of this embodiment can effectively improve the spectral resolution and analytical performance of aerosol LIBS.

[0107] This invention not only significantly improves the sampling efficiency and statistical analysis effectiveness of LIBS technology for aerosol particles, but also enables high-dimensional, nonlinear analysis of complex aerosol LIBS spectra, offering advantages such as real-time in-situ processing, high throughput, high sensitivity, and high selectivity. Its analytical targets can be aerosol samples from sources such as the atmosphere, ocean, biological sources, and industrial emissions, and can also be extended to other discrete heterogeneous samples.

[0108] Of course, the above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. An aerosol analysis method based on high-repetition-rate LIBS and two-dimensional correlation spectroscopy, characterized in that, The application relates to a method for detecting aerosol particles by using laser-induced breakdown spectroscopy (LIBS). The pulse repetition frequency is set between 1 kHz and 40 kHz, a high-repetition-frequency adjustable pulse laser with an energy of 850 muJ is focused on an aerosol sample, and then the particles in the aerosol are excited and dissociated under the joint action of the pulse laser, laser-induced plasma and shock wave to generate plasma. n frequency points are set in the frequency range of 1 kHz to 40 kHz, so that two conditions of independent events and related events between pulses are generated; the independent event is that the plasma generated by the previous pulse laser excitation has been quenched when the next pulse laser arrives; the related event is that the intermediate product after the previous pulse laser dissociates the aerosol particles still exists when the next pulse laser arrives. Under each frequency condition, the light signals emitted by the plasma are collected m times to form n*m spectral records; a plurality of sampling points are set in the wavelength range of 300 nm to 850 nm, and the emission light of the plasma is collected at each sampling point by using a spectrometer to form spectral data in each spectral record, thereby constituting original LIBS spectral data; effective sampling spectral data is selected from the original LIBS spectral data, and the effective sampling spectrum is a LIBS spectrum containing element characteristic radiation of the particles; The background of the effective sampling spectral data is removed to eliminate the interference of the LIBS continuous background radiation; The effective sampling spectral data under the same frequency condition after the background removal is averaged, and n groups of average spectral data are used to form a two-dimensional LIBS spectral data matrix; The cross-correlation coefficient between each spectral component in the two-dimensional LIBS spectral data matrix and the two-dimensional correlation synchrospectrum are calculated; The spectral intensity value on the diagonal line of the two-dimensional correlation synchrospectrum is extracted to obtain an autocorrelation peak; A tangent line with a wavelength equal to the characteristic spectral line of the element in the aerosol sample is extracted from the two-dimensional correlation synchrospectrum to obtain a cross peak; The element characteristic spectral line in the aerosol particles is separated from the element characteristic spectral line in the air by the cross peak. 2.The aerosol analysis method based on high-repetition-rate LIBS and two-dimensional correlation spectroscopy according to claim 1, wherein, When the effective sampling spectral data is selected, the signal-to-noise ratio of the low-energy-state radiation spectral line of the main element of the particles in the spectrum is higher than 3, and the selection is performed by taking the signal-to-noise ratio as a threshold. 3.The aerosol analysis method based on high-repetition-rate LIBS and two-dimensional correlation spectroscopy according to claim 1, wherein, The two-dimensional LIBS spectral data matrix is a two-dimensional data matrix with the wavelength of the plasma emission light and the repetition frequency as independent variables.

4. The aerosol analysis method based on high-repetition-rate LIBS and two-dimensional correlation spectroscopy according to any one of claims 1 to 3, characterized in that, The process of calculating the cross-correlation coefficient between each spectral component in the two-dimensional LIBS spectral data matrix and the two-dimensional correlation synchrospectrum includes: The dynamic spectrum of each spectral component generated by the laser pulse repetition frequency disturbance is calculated, and the calculation formula is: wherein, denotes the dynamic spectrum; x i denotes the ith value of the repetition frequency; I(λ, x i ) denotes the measured spectral intensity value at a wavelength of λ, a repetition frequency of x i ; denotes the average of the spectral intensity values measured at a wavelength of λ at different repetition frequencies; The cross-correlation coefficient between each spectral component is calculated by using fast discrete Fourier transform, and the calculation formula is: where γ(λ1, λ2) denotes the cross-correlation coefficient of the two spectral components λ1and λ2as a function of the repetition frequency x; F denotes the fast discrete Fourier transform; and denote the dynamic spectral components at wavelengths λ1and λ2perturbed by the repetition frequency x, respectively; o denotes the Hadamard product; * denotes the complex conjugate; C denotes the complex set. The cross-correlation coefficients between each of the resulting spectral components are subjected to an inverse Fourier transform F -1 {γ(λ1,λ2)}, taking the real part of the first element of each vector after inverse Fourier transform, to form the two-dimensional correlation synchrospectrum.

Images