A method and system for detecting perfluorooctane sulfonate in water

Abstract

The application relates to the technical field of environmental monitoring and discloses a perfluorooctane sulfonic acid detection method and system in water. The method comprises the following steps: exciting a perfluorooctane sulfonic acid molecule carbon fluoride bond vibration mode in a water sample to be measured by a specific wavelength laser, inducing a nonlinear transition of a transient refractive index of the molecule, and obtaining a preliminary optical response signal sequence; extracting a matrix scattering background light intensity fluctuation range and a peak-valley contrast ratio of the signal sequence, and determining a transition amplitude enhancement threshold; when the threshold exceeds the background range, capturing a transient transition image and processing the same to obtain a plurality of difference value signal sequences; extracting a sequence characteristic parameter and matching a preset response mode, and determining a molecular aggregate state configuration feature; through multi-wavelength spectrum fusion verification, ecological threat evaluation and signal enhancement processing, trace concentration quantitative results are obtained. Through laser targeting excitation and multi-dimensional feature fusion, the application solves the problems of weak trace detection signals, large interference and poor specificity in a complex matrix, and improves detection sensitivity and accuracy.

Description

Technical Field

[0001] This application relates to the field of environmental monitoring technology, and in particular to a method and system for detecting perfluorooctane sulfonic acid in water. Background Technology

[0002] Perfluorooctane sulfonic acid (PFOS), a typical persistent organic pollutant, is widely listed on global priority control lists due to its extremely strong chemical stability and bioaccumulation. In the field of water environment monitoring, accurate detection of trace PFOS is a crucial step in assessing the ecological safety of aquatic bodies. Currently, this field mainly relies on highly sensitive optical analysis or chromatography-mass spectrometry techniques to address the challenge of identifying extremely low concentrations of pollutants in complex aquatic matrices.

[0003] Existing technologies have significant limitations in practical applications. While traditional liquid chromatography-mass spectrometry (LC-MS) offers high sensitivity, its cumbersome pretreatment process, requiring solid-phase extraction and concentration, is time-consuming and prone to interference, making rapid on-site screening difficult. Conventional spectroscopic detection methods, such as UV-Vis spectrophotometry, suffer from extremely low sensitivity due to the lack of characteristic chromophores in PFOS molecules, failing to meet trace analysis needs. Ordinary fluorescence spectroscopy is susceptible to fluorescence quenching interference from background substances like humic acid in water, resulting in poor signal-to-noise ratios and large deviations in quantitative results. Furthermore, while existing Raman spectroscopy can reflect molecular fingerprint information, the signal is weak at low concentrations and is often affected by background scattering noise from water, making it difficult to distinguish characteristic peaks of trace PFOS. More critically, existing optical methods often focus only on changes in light intensity or wavelength, neglecting the transient thermal effects and subtle polarization rotations of PFOS under specific excitation conditions, leading to insufficient resolution of molecular aggregate configurations and an inability to effectively eliminate dynamic background noise in complex environments. This single-dimensional detection mode often results in problems such as high detection limits, weak anti-interference ability, and inability to simultaneously assess ecotoxicity risks when facing the ever-changing environmental factors in actual water samples, making it difficult to meet the needs of high-precision and real-time monitoring.

[0004] To address the above deficiencies, this application combines laser-induced instantaneous refractive index nonlinear transition technology with high-speed polarization state rotation image analysis, solving the problems of difficulty in extracting trace PFOS signals and lack of molecular configuration analysis in complex backgrounds, thereby improving the signal-to-noise ratio, sensitivity, and accuracy of ecological risk assessment. Summary of the Invention

[0005] This application provides a method and system for detecting perfluorooctane sulfonic acid (PFOS) in water, which solves the problems of difficulty in extracting trace PFOS signals and lack of molecular configuration analysis in complex backgrounds, and improves the signal-to-noise ratio, sensitivity and accuracy of ecological risk assessment.

[0006] In a first aspect, this application provides a method for detecting perfluorooctane sulfonic acid in water, the method comprising: S1. Excite the carbon-fluorine bond vibration mode of perfluorooctane sulfonic acid molecules in the water sample to be tested with a laser of a specific wavelength, induce the instantaneous nonlinear transition of the molecular refractive index, and obtain a preliminary optical response signal sequence containing the local thermal lensing effect. S2. Based on the preliminary optical response signal sequence, extract the fluctuation range of the matrix scattered background light intensity and the peak-valley contrast ratio of the differential signal to determine the enhancement threshold of the instantaneous refractive index nonlinear transition amplitude. S3. Determine whether the enhancement threshold exceeds the fluctuation range. If so, capture the instantaneous transition image containing the polarization state rotation angle in the optical path, and process the instantaneous transition image to obtain a multi-frame difference signal sequence. S4. Extract the density of gray-scale abrupt change points and the ratio of high-sensitivity peaks and valleys from the multi-frame difference signal sequence, and determine whether they conform to the preset response mode. If so, determine the molecular aggregation state configuration characteristics corresponding to the trace concentration. S5. Based on the molecular aggregated state configuration characteristics, multi-wavelength spectral analysis was used to integrate the energy level parameters of the carbon-fluorine bond vibrational modes to obtain a sequence of verification results for the instantaneous refractive index nonlinear transition amplitude. S6. Extract the trend of the critical micelle concentration boundary of the surfactant and the distribution range of the adsorption-desorption equilibrium constant in the environmental medium from the verification result sequence, determine the ecological threat level of perfluorooctane sulfonic acid in the water sample, and determine the optical difference distribution characteristic sequence related to the pollution source. S7. Based on the optical difference distribution characteristic sequence, the background fluorescence quenching coefficient distribution and the absorption cross section difference sequence under multi-wavelength excitation in the matrix are subjected to signal enhancement processing to obtain the trace concentration quantitative results of perfluorooctane sulfonic acid in the water sample.

[0007] Secondly, this application provides a perfluorooctane sulfonic acid (PFOS) detection system in water, used to implement the aforementioned method for detecting PFOS in water, the system comprising: The data acquisition module is used to excite the carbon-fluorine bond vibration mode of perfluorooctane sulfonic acid molecules in the water sample to be tested with a laser of a specific wavelength, induce the instantaneous nonlinear transition of the molecular refractive index, and obtain a preliminary optical response signal sequence containing the local thermal lensing effect. The threshold calculation module is used to extract the fluctuation range of the matrix scattered background light intensity and the peak-valley contrast ratio of the differential signal based on the preliminary optical response signal sequence, and to determine the enhancement threshold of the instantaneous refractive index nonlinear transition amplitude. An image capture module is used to determine whether the enhancement threshold exceeds the fluctuation range. If so, it captures an instantaneous transition image containing the polarization state rotation angle in the optical path and processes the instantaneous transition image to obtain a multi-frame difference signal sequence. The feature recognition module is used to extract the density of gray-scale abrupt change points and the ratio of high-sensitivity peaks and valleys from the multi-frame difference signal sequence, and determine whether they conform to the preset response mode. If so, the molecular aggregation state configuration feature corresponding to the trace concentration is determined. The spectral verification module is used to obtain a sequence of verification results for the instantaneous refractive index nonlinear transition amplitude by using multi-wavelength spectral analysis to integrate the energy level parameters of carbon-fluorine bond vibrational modes, based on the molecular aggregate configuration characteristics. The threat assessment module is used to extract the trend of the critical micelle concentration boundary of surfactant and the distribution range of adsorption-desorption equilibrium constant in the environmental medium from the verification result sequence, determine the ecological threat level of perfluorooctane sulfonic acid in the water sample, and determine the optical difference distribution characteristic sequence related to the pollution source. The concentration quantification module is used to perform signal enhancement processing on the background fluorescence quenching coefficient distribution and the absorption cross section difference sequence under multi-wavelength excitation in the matrix according to the optical difference distribution characteristic sequence, so as to obtain the trace concentration quantification result of perfluorooctane sulfonic acid in the water sample.

[0008] Thirdly, this application provides a computer-readable storage medium storing instructions that, when executed by a processor, implement the method for detecting perfluorooctane sulfonic acid in water.

[0009] This application proposes a method and system for detecting perfluorooctane sulfonic acid (PFOS) in water, solving the problems of difficult extraction of trace PFOS signals and lack of molecular configuration analysis in complex backgrounds, and improving the signal-to-noise ratio, sensitivity, and accuracy of ecological risk assessment. Compared with existing technologies, the beneficial effects of this application's technical solution are at least as follows: First, by inducing a nonlinear transition in the refractive index through laser excitation of carbon-fluorine bond vibration modes, and capturing the preliminary optical response signal containing the local thermal lensing effect using a high-speed image sensor, the weak optical response intensity is effectively enhanced. This overcomes the weak signal defect caused by the lack of characteristic chromophores or low fluorescence quantum yield of PFOS in traditional spectroscopic methods, and significantly reduces the detection limit of the method.

[0010] Second, an algorithm based on gray-scale abrupt change point analysis based on frame difference signal sequence is adopted to extract the instantaneous transition features of polarization state rotation angle from both time and spatial dimensions. This can accurately remove dynamic scattering noise and fluorescence quenching interference from background substances such as humic acid in water bodies, solving the problem of poor signal-to-noise ratio and large quantitative deviation in existing single light intensity measurement modes under complex environmental factors.

[0011] Third, a mapping model was constructed from the abrupt change point of the optical signal slope to the changing trend of the adsorption-desorption equilibrium constant and the critical micelle concentration boundary, realizing the direct conversion from physical optical parameters to chemical ecological parameters. It can simultaneously output the ecological threat level while completing the quantitative analysis of concentration, making up for the shortcomings of existing technologies that can only provide concentration data but lack the function of real-time ecotoxicity assessment.

[0012] Fourth, by using multi-wavelength spectral fusion and wavelet transform enhancement algorithms to perform in-depth processing on the absorption cross-section difference sequence, the drawback of single-wavelength detection being susceptible to fluctuations in water turbidity is eliminated, the linearity and stability of the system under wide dynamic range concentration detection are improved, and highly repeatable detection results are ensured in water samples with different levels of pollution. Attached Figure Description

[0013] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0014] Figure 1 This is a schematic flowchart of a method for detecting perfluorooctane sulfonic acid in water according to this application; Figure 2 This is a graph showing the signal-to-noise ratio as a function of concentration in this application; Figure 3 This is a radar chart showing the overall performance in this application; Figure 4 This is a schematic diagram of the structure of a perfluorooctane sulfonic acid detection system in water according to this application. Detailed Implementation

[0015] This application provides a method and system for detecting perfluorooctane sulfonic acid in water. The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" or "having" and any variations thereof are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] For ease of understanding, the specific process of the embodiments of this application is described below. Please refer to [link / reference]. Figure 1 One embodiment of a method for detecting perfluorooctane sulfonic acid in water according to this application includes: S1. Excite the carbon-fluorine bond vibration mode of perfluorooctane sulfonic acid molecules in the water sample to be tested with a laser of a specific wavelength, induce the instantaneous nonlinear transition of the molecular refractive index, and obtain a preliminary optical response signal sequence containing the local thermal lensing effect.

[0017] In one specific embodiment, performing step S1 includes the following steps: The length distribution of the fluorinated carbon chain and the electronegativity distribution of the sulfonic acid group of the perfluorooctane sulfonic acid molecule in the water sample to be tested were obtained to determine the carbon-fluorine bond vibration frequency matching parameters. Based on the carbon-fluorine bond vibration frequency matching parameters, a specific wavelength laser is selected to irradiate the water sample to be tested. The energy level of the carbon-fluorine bond vibration mode is controlled by adjusting the laser pulse energy, thereby inducing a nonlinear transition of the instantaneous refractive index of the molecule. The amplitude of the instantaneous nonlinear transition of the refractive index of molecules is monitored, and the intensity of the thermal lensing effect in the local area of ​​the water sample under test is calculated based on the thermal effect of the medium induced by laser irradiation. Light intensity and phase information at multiple time points are collected at a preset sampling frequency, and the thermal lensing effect intensity is integrated to generate a preliminary optical response signal sequence containing the local thermal lensing effect.

[0018] Specifically, after the water sample to be tested is injected into a sample cell with an optical window, the molecules in the water sample are ionized using an electrospray ionization mass spectrometer. A full scan is performed within the mass-to-charge ratio range of 400-600 m / z, and the intensity of the molecular ion peaks and the corresponding mass-to-charge ratio data are collected. Baseline correction is performed on the collected mass spectrometry data using an adaptive iterative reweighted penalized least squares algorithm, with the algorithm parameters set to smoothness λ= The process involved 15 iterations with a weight threshold of 0.5. After correction to remove noise interference from baseline drift, a Gaussian fitting algorithm was used to perform peak fitting on the mass spectra. The full width at half maximum (FWHM) threshold was set to 0.1 m / z. Peak area data of the characteristic peaks of perfluorooctane sulfonic acid (PFOS) molecules corresponding to their mass-to-charge ratios were extracted. The proportion of molecules with 8 carbon atoms in the fluorinated carbon chain was statistically analyzed, generating a fluorinated carbon chain length distribution dataset containing a one-to-one correspondence between the relative abundance of molecules with different chain lengths and their chain length values. Based on the PFOS molecular structure data obtained from mass spectrometry analysis, density functional theory was used at the B3LYP / 6-311G(d,p) basis set level to optimize the molecular structure and calculate the electron cloud distribution. During the calculation, a convergence threshold was set to a value where the energy change was less than 0.5. Hartree, the atomic force is less than Hartree / Bohr obtained the Mulliken charge distribution data of sulfur and oxygen atoms in the sulfonic acid group of the molecule, and based on the Paullin electronegativity scale, used the formula... Calculate the effective electronegativity of each atom in the sulfonic acid group, where Given the ground-state electronegativity of the atom and q as the Mulliken charge value of the atom, the overall electronegativity distribution data of the sulfonic acid group is calculated. This data, along with the fluorinated carbon chain length distribution data, is used as input. The intrinsic vibrational frequencies of the carbon-fluorine bonds in the perfluorooctane sulfonic acid molecule are calculated using a normal mode analysis algorithm, including the frequency values ​​corresponding to symmetric stretching vibrations and antisymmetric stretching vibrations. This generates carbon-fluorine bond vibrational frequency matching parameters, which directly correspond to the resonance absorption wavelength range of different vibrational modes of the carbon-fluorine bond.

[0019] The carbon-fluorine bond stretching vibration frequencies in the vibrational frequency matching parameters correspond to three characteristic excitation wavelengths: 1064 nm, 532 nm, and 785 nm. The laser wavelength is adjusted according to the turbidity of the water sample matrix. For example, in high-turbidity industrial wastewater samples, a 532 nm frequency-doubled laser is selected, with a pulse width of 10 ns, a repetition frequency of 1 kHz, and a single-pulse energy adjustment range of 0.1 mJ-10 mJ. The single-pulse energy is adjusted by a laser energy controller to induce a transition of the carbon-fluorine bond vibrational mode from the ground state to the first excited state, inducing a redistribution of the molecular electron cloud. Based on the optical Kerr effect, the instantaneous change in the molecular refractive index is expressed by the formula... Calculation, where The nonlinear refractive index coefficient of perfluorooctane sulfonic acid molecule. The intensity of the incident laser light is adjusted. The amplitude of the instantaneous nonlinear transition of the refractive index is numerically controlled. During laser irradiation, the incident angle of the laser beam is kept perpendicular to the optical window of the sample cell, and the spot diameter is focused to 200 μm. This ensures that the laser energy is concentrated in the detection area of ​​the sample cell, avoiding background interference caused by the large-scale excitation of other matrix components in the water sample. This solves the problem of the complex matrix background signal masking the response of the target molecule in traditional detection.

[0020] The refractive index change of the water sample during laser excitation was monitored in real time using a Mach-Zehnder interferometer. The optical path difference between the reference arm and the detection arm of the interferometer was controlled to be less than 10 μm. Displacement data of the interference fringes were collected, and the sampling frequency was consistent with the repetition frequency of the laser pulses. By establishing a linear correspondence between the fringe displacement and the refractive index change, the amplitude data of the instantaneous nonlinear transition of the molecular refractive index corresponding to each laser pulse was calculated. This data formed a one-to-one time series relationship with the laser pulse energy data. Simultaneously, based on the photothermal effect induced by laser irradiation, a thermal diffusion model of the local region of the water sample was established. The thermal diffusion process was described by the formula... Description, in which Let be the thermal diffusivity of the water sample. The heat source term generated by laser irradiation is positively correlated with the laser pulse energy and the absorption coefficient of the water sample to the laser. The thermal diffusion equation is solved by the finite difference method with a spatial step of 10 μm and a time step of 1 ns to obtain the temperature gradient distribution data in the water sample detection area. Based on the linear relationship between the temperature gradient and the refractive index of the medium, the intensity data of the local thermal lensing effect is calculated. This data reflects the degree of optical distortion in the local area of ​​the water sample under laser excitation, which complements the data of the nonlinear transition amplitude of the instantaneous refractive index of the molecules, amplifying the optical response of perfluorooctane sulfonic acid molecules at trace concentrations.

[0021] The preset sampling frequency is set to 1kHz, and optical data is continuously collected at 100 time points. Each time point corresponds to the light intensity value output by the photodetector. The light intensity value is normalized to the range of 0.5-1.0. At the same time, the phase information of the interference light is collected through a phase demodulation algorithm. The phase information collection range is 0-π radians. The phase demodulation algorithm adopts a lock-in amplification algorithm. The algorithm parameters are set to the same reference frequency as the laser pulse repetition frequency, time constant 100μs, roll-off slope 12dB / oct, and noise signals unrelated to laser excitation are filtered out to obtain the phase data corresponding to each time point. The light intensity data and phase data correspond one-to-one in the time dimension to form a basic optical time series. The thermal lensing effect intensity data at each time point is used as a weighting factor and weighted with the light intensity data at the same time point. The weighted data is then integrated with the phase data to generate a preliminary optical response signal sequence containing the local thermal lensing effect. Each data unit in the sequence contains the weighted light intensity value, phase value, and thermal lensing effect intensity value at the corresponding time point. The three sets of data are strictly aligned in the time dimension, providing multi-dimensional basic input data for subsequent differential detection and threshold determination. This solves the problems of poor anti-interference ability and low signal-to-noise ratio of single-dimensional signals in traditional optical detection.

[0022] S2. Based on the preliminary optical response signal sequence, extract the fluctuation range of the matrix scattered background light intensity and the peak-valley contrast ratio of the differential signal to determine the enhancement threshold of the instantaneous refractive index nonlinear transition amplitude.

[0023] In one specific embodiment, performing step S2 includes the following steps: The preliminary optical response signal sequence is processed using a differential detection method. The light intensity difference between adjacent sampling points in the sequence is calculated to generate a differential signal sequence. Separate the matrix scattering background signal from the differential signal sequence, count the maximum and minimum light intensity of the matrix scattering background signal, and determine the fluctuation range of the matrix scattering background light intensity; Extract the peak and valley values ​​of the signal from the differential signal sequence, calculate the ratio of peak to valley values, and obtain the peak-valley contrast ratio of the differential signal. Using the peak-valley contrast ratio as a weighting coefficient and combining it with the upper limit of the fluctuation range, the enhancement threshold of the instantaneous nonlinear transition amplitude of the refractive index is calculated.

[0024] Specifically, the preliminary optical response signal sequence is arranged in ascending order of time as follows: Where m is the total number of sampling points in the sequence, each Corresponding to the i The normalized light intensity values ​​at each sampling time point are obtained, with a sampling frequency set to 1kHz and m set to 100. All light intensity values ​​within the sequence are normalized to the range of 0.5-1.0. The light intensity data at each sampling point within the sequence are strictly aligned in the time dimension with the phase data and thermal lensing effect intensity data at the same sampling time point. The preliminary optical response signal sequence is processed using a differential detection method, with the light intensity values ​​of adjacent sampling points as the calculation unit, and the formula is used to calculate the intensity values. Complete the difference operation, where i The range of values ​​for is 1≤ i ≤m-1, after operation, a differential signal sequence is generated. During the differential operation, the positional correspondence between each difference data point and the original sequence sampling point is preserved, eliminating baseline interference caused by static light intensity drift and avoiding the influence of fixed scattered light generated by static impurities in the water sample on subsequent signal recognition.

[0025] Based on the generated differential signal sequence, an adaptive iterative reweighted penalized least squares algorithm is used for baseline fitting, with the algorithm parameter set to smoothness. With 15 iterations and a weight threshold of 0.5, the resulting baseline sequence is the matrix scattering background signal sequence. Each data point in the matrix scattering background signal sequence corresponds one-to-one with a data point at the same location in the difference signal sequence, representing the random scattered light intensity changes caused by non-target substances such as suspended particles and dissolved organic matter in the water sample. The maximum values ​​within the matrix scattering background signal sequence are statistically analyzed. and minimum value The fluctuation range of the matrix-scattered background light intensity was determined to be: The width of the fluctuation range is This allows for the quantification of the background interference intensity of complex water sample matrices.

[0026] Based on the differential signal sequence, a sliding window extreme value detection method is used to identify the signal peaks and valleys within the sequence. The sliding window size is set to 5 sampling points, with a step size of 1 sampling point. The entire differential signal sequence is traversed, and when the value at the center of the window is greater than all other values ​​within the window, that position is marked as the signal peak point, and the peak value is recorded. When the value at the center of the window is less than all other values ​​within the window, mark that position as a signal valley point and record the valley value. All identified peak and valley values ​​are paired, with a one-to-one correspondence between a peak and its preceding valley. Invalid pairs are removed if the peak-valley difference is less than 10% of the fluctuation range width, thus avoiding false peak-valley identification due to background noise. The contrast ratio is calculated for each valid pair. Where k is the sequence number of the valid pairing, for all The arithmetic mean is taken to obtain the peak-to-valley contrast ratio R of the differential signal, which is used to quantify the relative intensity of the optical response signal of the target pollutant and the background noise, and to distinguish the weak response generated by trace perfluorooctane sulfonic acid molecules from the random fluctuation signal of the matrix.

[0027] Using the peak-to-valley contrast ratio R as a weighting coefficient, combined with the upper limit of the range of matrix scattering background light intensity fluctuations. Through formula The enhancement threshold T of the instantaneous refractive index nonlinear transition amplitude is calculated, where α is the scene adaptation coefficient. The value of α is adjusted according to the water sample matrix type. For example, α is 1.2 in drinking water samples, 1.1 in low-turbidity surface water samples, and 1.3 in high-turbidity industrial wastewater samples. The weighting coefficient R and the upper limit of fluctuation are also considered in the calculation. The product of these factors allows for adaptive adjustment of the threshold based on background fluctuations and the target signal response intensity. For example, in drinking water sample detection, the maximum value of the matrix scattering background signal... With a normalization unit of 0.5, the effective peak-to-valley contrast ratio R is 2.5, the scene adaptation coefficient α is 1.2, and the calculated enhancement threshold T is 1.5. This threshold forms a corresponding reference with the range of matrix scattering background light intensity fluctuations, which is used to determine whether to trigger the subsequent instantaneous transition image capture stage.

[0028] S3. Determine whether the enhancement threshold exceeds the fluctuation range. If so, capture the instantaneous transition image containing the polarization state rotation angle in the optical path, and process the instantaneous transition image to obtain a multi-frame difference signal sequence.

[0029] In one specific embodiment, performing step S3 includes the following steps: Determine whether the enhancement threshold exceeds the fluctuation range; If so, the high-speed image sensor is activated to continuously acquire multiple frames of polarized light in the optical path, record the dynamic changes in the polarization state rotation angle during laser excitation, and obtain the instantaneous transition image sequence. A frame-by-frame pixel-level difference operation is performed on the instantaneous transition image sequence to obtain a multi-frame difference signal sequence containing information on the spatial distribution of local refractive index transitions.

[0030] Specifically, the enhancement threshold value is compared point-by-point with the upper limit of the matrix scattering background light intensity fluctuation range. During the comparison, the temporal dimension of the data is kept aligned. The enhancement threshold and fluctuation range values ​​corresponding to each comparison node are derived from the preliminary optical response signal sequence of the same sampling period, avoiding misjudgments caused by data temporal misalignment. The comparison process uses a formula... The threshold excess is calculated, where H is the threshold excess. When the value of H is greater than 0, it is determined that the enhancement threshold exceeds the fluctuation range, triggering the subsequent image acquisition process. When the value of H is less than or equal to 0, the threshold comparison operation is continuously executed in a loop. The sampling interval is consistent with the sampling frequency of the initial optical response signal sequence, which is 1kHz, to ensure the real-time performance of the comparison process. This solves the problem in the existing technology that it is impossible to accurately trigger the high-sensitivity signal acquisition stage, which is prone to weak signal missed detection or background noise false triggering.

[0031] After triggering the image acquisition process, the high-speed image sensor's acquisition frame rate is set to 2000 frames per second, and the exposure time is set to 500 ns, synchronized with the repetition frequency of the laser pulse. The trigger signal of the laser pulse and the frame synchronization signal of the high-speed image sensor are phase-locked through a timing controller, with timing deviation controlled within ±10 ns. This ensures that each acquired image frame corresponds to the instantaneous process of laser excitation, avoiding the loss of transition signals due to timing deviation. In the optical path, the angle between the polarizer and the analyzer of the polarized light is set to 45 degrees. The polarizer is located between the laser source and the sample cell, and the analyzer is located between the sample cell and the high-speed image sensor. When laser excitation induces an instantaneous nonlinear refractive index transition in perfluorooctane sulfonic acid molecules, it causes a rotation of the polarization state of the transmitted polarized light. The rotation angle varies from 0.1 degrees to 1.0 degrees, and the amount of change in the polarization state rotation angle is linearly positively correlated with the amplitude of the instantaneous nonlinear refractive index transition of the molecules. After passing through the analyzer, the change in the polarization state rotation angle is converted into a change in the grayscale value of the image pixels, realizing the conversion of the refractive index transition signal at the molecular level into a visualized image grayscale signal.

[0032] The high-speed image sensor continuously acquires 50 frames, forming a sequence of instantaneous transition images. The images within the sequence are arranged in ascending order of acquisition time. Each frame of the image was set to a pixel resolution of 1920×1080, with each pixel having a 16-bit quantization bit depth and a grayscale value range of 0 to 65535. Each frame in the sequence corresponds to a fixed time point during laser excitation, with a time interval of 500 μs between adjacent frames. The temporal coverage of the image sequence completely overlaps with the sampling period of the preliminary optical response signal sequence, ensuring the homology and correspondence between the image data and the preceding optical signal data. During acquisition, dark current correction and flat-field correction were performed on each frame of the original image using an image preprocessing algorithm. Dark current correction used a dark-field image acquired under no laser illumination as the correction benchmark, while flat-field correction used a flat-field image acquired under uniform light field illumination as the correction benchmark. The correction formula is as follows: ,in This is the corrected z-th frame image. The original z-th frame image is denoted as D, the dark field image is denoted as U, and the flat field image is denoted as G. The gray level gain coefficient is denoted as 65535. The correction process eliminates the system error caused by the dark current noise of the high-speed image sensor itself and the non-uniformity of the optical path light field, avoids the interference of system noise on subsequent difference calculations, and solves the problem of low signal-to-noise ratio caused by complex matrix background noise masking the target signal in the existing technology.

[0033] The corrected instantaneous transition image sequence is subjected to frame-by-frame pixel-level difference calculation, maintaining a one-to-one correspondence between pixel positions during the calculation. For the z-th frame and the (z-1)-th frame in the sequence, the difference in pixel grayscale values ​​at the same coordinate position (x, y) is calculated using the following formula: The value of z is in the range of 2 ≤ z ≤ 50. The pixel difference between frame z and frame z-1 at coordinates (x, y) is calculated to generate 49 difference images, forming a multi-frame difference signal sequence. Each difference image in the sequence corresponds to the change in polarization state rotation angle between two adjacent sampling times, preserving the spatial distribution information of local refractive index transitions during laser excitation. During the difference calculation, fixed interference signals such as static background and reflected light from the sample cell wall in the optical path are eliminated, and only the dynamic change signal induced by laser excitation is retained, thus achieving effective extraction of the weak optical response of trace perfluorooctane sulfonic acid molecules.

[0034] The generated multi-frame difference signal sequence is subjected to invalid frame removal processing. The removal rule is that the sum of the absolute values ​​of all pixel differences in a single frame difference image is less than the grayscale threshold corresponding to the matrix scattering background light intensity fluctuation range. The grayscale threshold is determined by the linear mapping relationship between the upper limit of the matrix scattering background light intensity fluctuation range and the image grayscale value. The mapping coefficient is determined by the previous calibration experiment. For example, in the detection of drinking water samples, the grayscale threshold corresponding to the upper limit of the matrix scattering background light intensity fluctuation range is 50. Frames with the sum of the absolute values ​​of pixel differences in a single frame difference image less than 50 are marked as invalid frames and removed. The remaining valid frames form the final multi-frame difference signal sequence. Each frame difference image in the sequence contains spatial distribution data of local refractive index transitions. The difference value of each pixel is positively correlated with the instantaneous refractive index transition amplitude of the perfluorooctane sulfonic acid molecule at the corresponding position. This provides a data source with both spatial and temporal dimensions for subsequent molecular aggregate configuration feature extraction, solving the problem that the single-dimensional detection mode in the existing technology is insufficient in its ability to analyze molecular aggregate configurations and cannot effectively remove dynamic background noise.

[0035] S4. Extract the density of gray-scale abrupt change points and the ratio of high-sensitivity peaks and valleys from the multi-frame difference signal sequence, and determine whether they conform to the preset response mode. If so, determine the molecular aggregate configuration characteristics corresponding to the trace concentration.

[0036] In one specific embodiment, performing step S4 includes the following steps: Perform pixel-level traversal on each frame of the multi-frame difference signal sequence and calculate the gray-level gradient value of each pixel. Pixels whose gray-level gradient values ​​exceed a preset gray-level threshold are marked as gray-level abrupt change points. The spatial distribution of gray-level abrupt change points in a single frame image is statistically analyzed, and the density of gray-level abrupt change points is calculated. Scan each frame of the multi-frame difference signal sequence, identify the positions of the signal peaks and valleys, calculate the relative difference between the peaks and valleys, and obtain a high-sensitivity peak-valley ratio; The density of gray-scale abrupt change points is compared with the preset density threshold range of the electron cloud density gradient of the symmetry of perfluorooctane sulfonic acid molecules; The high-sensitivity peak-to-valley ratio was compared with a preset pattern of hydrophobic interface characteristics of long-chain perfluorinated compounds. Determine whether both comparison results conform to the preset response mode. If so, determine the molecular aggregate configuration characteristics corresponding to the trace concentration based on the density of corresponding gray-scale mutation points and the ratio of high-sensitivity peaks and valleys.

[0037] Specifically, each frame of the difference signal sequence has a pixel resolution of 1920×1080, and the grayscale difference value of each pixel ranges from -65535 to 65535. Each frame of the sequence is arranged in ascending order of acquisition time, and each frame corresponds to a fixed time node in the laser excitation process. The time interval between adjacent frames is 500μs. The pixel data of each frame is strictly aligned with the preliminary optical response signal sequence acquired in the previous time dimension to ensure data homogeneity. For each frame of the multi-frame difference signal sequence, pixel-level traversal is performed. For a pixel with coordinates (x,y) in a single frame difference image, its gray value is denoted as G(x,y). A 3×3 Sobel convolution kernel is used for gradient calculation. For example, the horizontal convolution kernel is Kx=[[-1,0,1],[-2,0,2],[-1,0,1]], and the vertical convolution kernel is Ky=[[-1,-2,-1],[0,0,0],[1,2,1]]. The horizontal gradient value Gx(x,y) is the convolution sum of Kx and the gray values ​​of the pixel's 3×3 neighborhood, and the vertical gradient value Gy(x,y) is the convolution sum of Ky and the gray values ​​of the pixel's 3×3 neighborhood. The gray gradient value M(x,y) of this pixel is obtained by formula... The calculation shows that after traversing all pixels in a single frame image, a gray-level gradient matrix with the same size as the original image is generated. Each element in the matrix corresponds one-to-one with a pixel in the original image, thus completely preserving the spatial distribution information of gray-level changes within the image.

[0038] A preset grayscale threshold is denoted as Th. The value of Th is determined by the upper limit of the grayscale difference corresponding to the range of matrix scattering background light intensity fluctuation. For example, in drinking water sample testing, Th is set to 50. Pixels with a grayscale gradient value M(x,y) greater than 50 are marked as grayscale abrupt change points. These grayscale abrupt change points correspond to the polarization state rotation angle abrupt change region caused by the instantaneous refractive index transition of laser-induced perfluorooctane sulfonic acid molecules, which is directly related to the molecular aggregation position. After marking all grayscale abrupt change points in a single frame image, the image is divided into 10×10 grid units, each containing 192×108 pixels. The number of grayscale abrupt change points in each grid unit is counted, and the abrupt change point density in each grid unit is calculated as the ratio of the number of abrupt change points in that unit to the total number of pixels in the unit. The arithmetic mean of the abrupt change point densities of all grid units is then taken to obtain the grayscale abrupt change point density ρ of the single frame image. The average of the grayscale abrupt change point densities of all valid frames in the multi-frame difference signal sequence is then taken to obtain the final grayscale abrupt change point density used for comparison. This density value reflects the degree of spatial aggregation of perfluorooctane sulfonic acid molecules in the water sample and is directly related to the electron cloud density gradient of molecular symmetry.

[0039] For each frame of a multi-frame difference signal sequence, the grayscale difference values ​​of all pixels are extracted to generate a one-dimensional sequence of grayscale differences for a single frame. A sliding window extreme value detection method is used to identify signal peaks and valleys within the sequence. The sliding window size is set to 7 data points, with a step size of 1 data point. The entire one-dimensional sequence is traversed, and when the value at the center of the window is greater than all other values ​​within the window, that position is marked as a signal peak, and the peak value is recorded. When the value at the center of the window is less than all other values ​​within the window, mark that position as a signal valley point and record the valley value. All identified peaks and valleys are paired, with a one-to-one correspondence between a peak and its adjacent preceding valley. Invalid pairs are removed if the absolute value of the peak-valley difference after pairing is less than a preset grayscale threshold Th, thus avoiding false peak-valley identification caused by background noise. For validly paired peak-valley values, the formula is used to... Calculate the peak-to-valley ratio for a single pair of images, and take the arithmetic mean of the peak-to-valley ratios for all valid pairs to obtain the high-sensitivity peak-to-valley ratio for a single frame image. The average of the high-sensitivity peak-to-valley ratios of all valid frames within a multi-frame difference signal sequence is taken to obtain the final high-sensitivity peak-to-valley ratio used for comparison. This ratio reflects the adsorption and aggregation behavior of perfluorooctane sulfonic acid molecules at the water-gas interface and solid-liquid interface. It is directly related to the hydrophobic interface characteristics of long-chain perfluorinated compounds, thus solving the problem that existing technologies cannot distinguish the interface behavior differences between target molecules and background impurities, resulting in weak anti-interference ability.

[0040] The preset density threshold range was obtained through density functional theory simulation. The simulation process used the B3LYP / 6-311G(d,p) basis set to optimize the structure and calculate the electron cloud distribution of the perfluorooctane sulfonic acid molecule. The convergence threshold was set to an energy change less than [value missing]. Hartree, the atomic force is less than Hartree / Bohr calculated the range of electron cloud density gradient along the molecular axis. Combined with calibration data from perfluorooctane sulfonic acid standard solutions of different concentrations, a preset density threshold range was determined. , calculate Compare with this range, if If the density term meets the preset response pattern, it is determined that the preset pattern conforms to the preset response pattern. The preset pattern was determined through surface tension measurement experiments of perfluorooctane sulfonic acid standard solutions. The surface tension of perfluorooctane sulfonic acid solutions of different concentrations was measured using the Dunuis ring method. The amount of adsorption at the molecular interface was calculated based on the Gibbs adsorption isotherm, establishing a correspondence between the interfacial adsorption amount and the peak-to-valley ratio. Combined with calibration experimental data at different concentration gradients, the ratio threshold range of the preset pattern was determined to be [value missing]. , calculate Compare with this range, if If the ratio term conforms to the preset response mode, then the ratio term will be in line with the preset response mode.

[0041] When both comparison results conform to the preset response mode, the molecular aggregate configuration characteristics are determined by matching the corresponding gray-level abrupt change point density and high-sensitivity peak-valley ratio with a preset concentration-configuration relationship table. The concentration-configuration relationship table is established through preliminary experiments using perfluorooctane sulfonic acid standard solutions of varying concentrations. The table contains different... and The combination corresponds to the molecular aggregation state configuration, including unimolecular dispersion, oligomer aggregation, micelle formation, and the trace concentration range corresponding to each state. For example, in drinking water sample testing, when... It is 0.06. When the value is 2.3, both comparisons conform to the preset response mode, corresponding to the oligomer aggregation state in the concentration-configuration table, with a corresponding concentration range of 0.01 μg / L to 0.1 μg / L; when It is 0.12. At a value of 3.5, the corresponding micelle formation state corresponds to a concentration range greater than 0.1 μg / L. Determining the molecular aggregation configuration characteristics provides direct molecular behavioral evidence for subsequent multi-wavelength spectral verification and ecological threat assessment, solving the problem that existing technologies can only output concentration values ​​but cannot analyze molecular aggregation behavior, thus hindering the simultaneous assessment of ecological threats.

[0042] S5. Based on the molecular aggregated state configuration characteristics, multi-wavelength spectral analysis was used to integrate the energy level parameters of the carbon-fluorine bond vibrational modes to obtain a sequence of verification results for the amplitude of the instantaneous refractive index nonlinear transition.

[0043] In one specific embodiment, performing step S5 includes the following steps: Based on the molecular aggregate configuration characteristics, N preset wavelengths that match the carbon-fluorine bond resonance frequency of the perfluorooctane sulfonic acid molecule are selected, and the water sample to be tested is spectrally scanned to collect excitation spectrum data of the corresponding wavelengths. N is a positive integer. Extract the energy level transition parameters corresponding to the carbon-fluorine bond vibrational modes during laser excitation from the excitation spectrum data; Using the signal-to-noise ratio of each group of excitation spectral data as the weight, the excitation spectral data and energy level transition parameters are weighted and fused to obtain the verification value of the instantaneous refractive index nonlinear transition amplitude of a single group. Multiple sets of instantaneous refractive index nonlinear transition amplitude verification values ​​are continuously acquired and sorted according to the time dimension to generate a verification result sequence of instantaneous refractive index nonlinear transition amplitude.

[0044] Specifically, based on the molecular aggregate configuration characteristics determined in the aforementioned steps, and combined with the resonance frequency shift characteristics of the carbon-fluorine bond vibration modes under different aggregation states of perfluorooctane sulfonic acid molecules, a preset wavelength matching the carbon-fluorine bond resonance frequency is selected. The number N of preset wavelengths is adjusted according to the molecular aggregate configuration characteristics. In the monomolecular dispersion state, N is 3, corresponding to the characteristic resonance wavelengths of symmetric stretching vibration, antisymmetric stretching vibration, and bending vibration of carbon-fluorine bonds. In the oligomer aggregate state or micelle formation state, N is 5, supplementing the frequency shift characteristic wavelengths caused by vibrational coupling. The wavelength values ​​cover the range of 1064nm, 785nm, 532nm, 633nm, and 830nm. Different wavelengths form a one-to-one correspondence with the molecular aggregate configuration characteristics, adapting to the frequency changes of molecular vibration modes under different aggregation states. After selecting a preset wavelength, a multi-channel spectrometer was used to simultaneously scan the water sample. During the scanning process, the sample cell was placed in a constant temperature control unit, with the temperature stabilized at 25℃ and temperature fluctuations controlled within ±0.1℃ to eliminate the interference of temperature changes on molecular vibration frequencies. The laser pulse trigger signal corresponding to each preset wavelength and the spectrometer's sampling trigger signal were synchronized via a timing controller, with timing deviations controlled within ±10ns. This ensured that the acquired excitation spectral data and the laser-excited molecular vibration process were strictly aligned in time. The spectral acquisition integration time for each wavelength was set to 100μs, the number of sampling points was set to 2048, and the spectral resolution was set to 0.1cm. -1 The acquired excitation spectrum data includes absorbance values, transmitted light intensity values, and spectral peak coordinates for each wavelength. Each set of excitation spectrum data is accompanied by a timestamp corresponding to the sampling time. The timestamp is consistent with the time dimension of the preliminary optical response signal sequence and multi-frame difference signal sequence acquired in the early stage, so as to achieve the same source alignment of data in different dimensions and avoid data association failure caused by time sequence misalignment.

[0045] After acquisition, each set of excitation spectrum data was preprocessed, and baseline correction was performed using an adaptive iterative reweighted penalized least squares algorithm with smoothness Λ=10. 6 The algorithm iterated 15 times, with a weight threshold of 0.5. The correction process removed baseline drift and low-frequency noise from the spectral data. Then, a Gaussian fitting algorithm was used to fit the corrected spectral data to its peaks, with a full width at half maximum (FWHM) threshold of 0.1 cm. -1 During the fitting process, the characteristic wavenumber ranges corresponding to different vibrational modes of the carbon-fluorine bond are locked, and the peak position, peak area, and full width at half maximum (FWHM) of each characteristic peak are extracted. Based on the extracted characteristic peak data, the energy level transition parameters corresponding to the carbon-fluorine bond vibrational modes are calculated, and the energy difference ΔΩ is obtained using the formula... Calculate, where h is Planck's constant and c is the speed of light. The effective wavelength corresponding to the characteristic peak position is given. The vibrational relaxation time ξ is calculated using the formula ξ=1 / (π·Δσ), where Δσ is the wavenumber width corresponding to the full width at half maximum (FWHM) of the characteristic peak. The transition probability Ψ is calculated using the formula Ψ=A / (ξ·S), where A is the peak area of ​​the characteristic peak and S is the effective irradiation area of ​​the laser spot. Each set of excitation spectrum data for a preset wavelength corresponds to a set of energy level transition parameters. The parameter values ​​are directly related to the vibrational excitation state of the carbon-fluorine bond in the molecule. Molecules under different aggregation configurations correspond to parameter values ​​in different ranges, realizing the quantitative characterization of the molecular vibrational excitation process. This solves the problem that the energy level transition process of the target molecule cannot be quantified in the existing technology, resulting in insufficient detection specificity and susceptibility to interference from non-target substances.

[0046] After extracting the energy level transition parameters, the fusion weights are calculated based on the signal-to-noise ratio (SNR) of each set of excitation spectral data. The SNR is calculated as the ratio of the peak intensity of the characteristic peak to the standard deviation of the spectral background noise, corresponding to the weight of the i-th wavelength. Through formula Calculate, where the value of i ranges from 1 to i and from N to N. The weighting is the sum of the signal-to-noise ratios (SNRs) of all spectral data corresponding to preset wavelengths. During the weighting calculation, wavelength data with higher SNRs have larger weight values ​​to reduce the interference of low SNR data on the fusion result. Then, a single-set instantaneous refractive index nonlinear transition amplitude verification value Υ is obtained through weighted fusion calculation. The calculation formula is as follows: ,in Let be the normalized excitation light intensity value corresponding to the i-th wavelength. This represents the energy difference corresponding to the i-th wavelength. For the transition probability corresponding to the i-th wavelength, during the fusion calculation, the spectral data, energy level transition parameters, and weighting coefficients of each wavelength form a one-to-one correspondence. Through the weighted fusion of multi-dimensional data, the drawbacks of single-wavelength detection being susceptible to interference from water turbidity and matrix scattering are eliminated, improving the stability and repeatability of the verification values. For example, in the detection of effluent samples from drinking water plants, the spectral signal-to-noise ratio (SNR) corresponding to the 532nm wavelength is 35, the SNR corresponding to the 785nm wavelength is 28, and the SNR corresponding to the 1064nm wavelength is 22. The calculated weight values ​​are 0.41, 0.33, and 0.26, respectively. During the fusion calculation, the 532nm wavelength data, which is less affected by matrix scattering, accounts for a higher proportion, effectively reducing the result deviation caused by background noise.

[0047] After calculating a single set of verification values, 100 sets of verification values ​​are continuously acquired at a sampling frequency of 1 kHz. Each set of verification values ​​corresponds to a unique sampling timestamp, which is strictly aligned with the sampling time of the previous optical signal data. The 3σ criterion is used to remove outliers from the continuously acquired sets of verification values. The mean and standard deviation of all verification values ​​are calculated, and outlier data points exceeding the mean ± 3 times the standard deviation are removed. The missing positions after removal are filled by linear interpolation of adjacent data points to ensure the continuity of the data sequence. The processed sets of verification values ​​are then arranged in ascending order by timestamp to generate a verification result sequence of instantaneous refractive index nonlinear transition amplitude. Compared with the preliminary optical response signal sequence of S1, the verification sequence generated in this step has undergone cross-validation of multi-wavelength spectral data and signal-to-noise ratio weighted fusion, eliminating the random errors of single-wavelength detection and having higher reliability. Each data unit within the sequence includes a sampling timestamp, a verification value, a spectral data summary for the corresponding wavelength, and the average energy level transition parameter. The numerical trend of the sequence perfectly matches the dynamic process of the instantaneous nonlinear refractive index transition of molecules during laser excitation, enabling cross-validation of the molecular aggregate configuration characteristics obtained through previous image analysis. This solves the problem of high false positive detection rates and large deviations in quantitative results caused by the lack of an effective verification link for the optical response of trace molecules in existing technologies. At the same time, the numerical change characteristics within the sequence directly reflect the changes in molecular aggregation behavior in the water environment, providing a highly reliable data source for subsequent extraction of the critical micelle concentration boundary of surfactants and calculation of adsorption-desorption equilibrium constants. This overcomes the shortcomings of existing technologies that cannot accurately analyze molecular interface behavior and thus cannot simultaneously complete ecological threat assessment.

[0048] S6. Extract the trend of the critical micelle concentration boundary of surfactant and the distribution range of adsorption-desorption equilibrium constant in the environmental medium from the verification result sequence, determine the ecological threat level of perfluorooctane sulfonic acid in the water sample, and determine the optical difference distribution characteristic sequence related to the pollution source.

[0049] In one specific embodiment, performing step S6 includes the following steps: The verification result sequence is segmented point by point, and the abrupt change points of the slope of the transition amplitude in the sequence are identified as concentration boundary nodes. The trend of the change of the critical micelle concentration boundary of the surfactant is obtained by fitting. Based on the concentration boundary node, the Langmuir adsorption model was used to calculate the equilibrium constant of the adsorption and desorption process of perfluorooctane sulfonic acid molecules in the environmental medium of the water sample to be tested, and the distribution range of the adsorption and desorption equilibrium constant was obtained. The ecological threat level of perfluorooctane sulfonic acid in the water sample is determined by comparing the trend of the critical micelle concentration boundary of the surfactant, the distribution range of the adsorption-desorption equilibrium constant, and the preset ecological threat classification threshold. Based on the degree of ecological threat, the changing trend of the critical micelle concentration boundary of surfactants, and the distribution range of adsorption-desorption equilibrium constants, corresponding optical characteristic parameters are extracted and integrated to generate an optical difference distribution characteristic sequence related to pollution sources.

[0050] Specifically, the verification result sequence is arranged in ascending order of sampling time as follows: T represents the total number of sampling points in the sequence, with a value of 100. The time interval between adjacent sampling points is 1 ms. The value of each sampling point is positively correlated with the amplitude of the instantaneous nonlinear transition of the molecular refractive index at the corresponding moment, and is directly related to the degree of aggregation of perfluorooctane sulfonic acid molecules. Each data unit in the sequence is strictly aligned in the time dimension with the previously acquired preliminary optical response signal sequence and multi-frame difference signal sequence to ensure the homology between optical response data and molecular aggregation behavior data. The sliding window method is used to segment the verification result sequence point by point. The sliding window size is set to 5 sampling points, and the step size is 1 sampling point. The entire sequence is traversed. For the j-th sampling point at the center of the window, the window range is from j-2 to j+2. The least squares method is used to perform linear fitting on the sampling points within the window. The fitting formula is: Where t is the sampling time, Let be the local slope corresponding to the j-th sampling point. To fit the intercept, a slope sequence with the same length as the original sequence is generated after traversal. Each value in the slope sequence corresponds to the rate of change of the transition amplitude at the sampling point, reflecting the speed of change in the molecular aggregation state.

[0051] The 3σ criterion is used to identify slope abrupt change points in the slope sequence, and the mean of the slope sequence is calculated. and standard deviation The value exceeds The sampling points are marked as candidate mutation points. Then, the slope difference between adjacent candidate mutation points is verified, and the slope difference between the candidate mutation point and the previous sampling point is calculated. ,when When the slope difference exceeds a preset threshold, the candidate mutation point is identified as a concentration boundary node. The preset slope difference threshold is determined through a calibration experiment of perfluorooctane sulfonic acid standard solution. For example, in the detection of drinking water samples, the preset slope difference threshold is set to 0.08. The concentration boundary node corresponds to the critical concentration position where the molecular aggregation state of perfluorooctane sulfonic acid changes. This solves the problem that existing technologies cannot accurately identify the critical behavior of molecular aggregation, resulting in a lack of basis for ecological risk assessment. After identifying all concentration boundary nodes, the sequences are sorted in ascending order by sampling time, dividing the sequence into multiple concentration intervals. Each interval corresponds to a molecular aggregation state, including a single-molecule dispersion interval, a pre-micelle aggregation interval, and a micelle formation interval. Using the concentration boundary nodes as demarcation points, a fifth-order polynomial fitting algorithm is used to fit the correlation between the transition amplitude values ​​and the corresponding concentrations within each interval. During the fitting process, the concentration boundary nodes are used as continuous constraints to ensure that the fitting curves of adjacent intervals are continuously differentiable at the nodes. The fitted curve is the trend of the critical micelle concentration boundary of the surfactant. The inflection point of the curve corresponds to the value of the critical micelle concentration, and the slope of the curve corresponds to the intensity of molecular aggregation behavior. For example, in the detection of surface water samples, the inflection point of the fitted critical micelle concentration boundary corresponds to a concentration of 0.1 μg / L. When the concentration of perfluorooctane sulfonic acid in the water sample exceeds this value, a large number of molecules form micelles, and bioaccumulation is significantly increased. This trend curve directly establishes a mapping relationship between the optical response signal and the molecular interface aggregation behavior, solving the deficiency of existing technologies that can only detect concentration values ​​and cannot analyze molecular aggregation behavior.

[0052] Based on the identified concentration boundary nodes, the equilibrium constant of the adsorption and desorption process of perfluorooctane sulfonic acid molecules in the environmental medium of the water sample was calculated using the Langmuir adsorption model. The expression for the Langmuir adsorption model is as follows: Γ represents the amount of perfluorooctane sulfonic acid molecules adsorbed per unit area of ​​the interface. This represents the saturation adsorption capacity. Let C be the adsorption-desorption equilibrium constant, and C be the concentration of perfluorooctane sulfonic acid in the water sample. The x-axis represents the concentration values ​​corresponding to the concentration boundary nodes, and the y-axis represents the adsorption capacity Γ calculated from the transition amplitude at the corresponding node. The Langmuir adsorption model is fitted using the nonlinear least squares method, with a convergence threshold set when the sum of squared residuals is less than a certain value. The maximum number of iterations is 50, and the fitting result is obtained. The numerical values ​​were then used to perform uncertainty analysis on the fitting results using the Monte Carlo simulation method. The number of simulations was set to 1000, and the results were obtained. The 95% confidence interval is the distribution range of the adsorption-desorption equilibrium constant. The magnitude of the equilibrium constant reflects the adsorption capacity of perfluorooctane sulfonic acid molecules at the interface of the environmental medium. The larger the value, the easier it is for the molecules to be adsorbed on water particles or biological interfaces, and the stronger the bioaccumulation and environmental persistence.

[0053] The preset ecological threat classification thresholds were determined using national water environmental quality standards and perfluorooctane sulfonic acid (PFOS) toxicological experimental data, and were divided into three levels: low threat, medium threat, and high threat. Each level corresponds to the inflection point concentration range of the critical micelle concentration boundary change trend and the distribution range of the adsorption-desorption equilibrium constant. The low threat level corresponds to an inflection point concentration below 0.001 μg / L and an upper limit of the equilibrium constant distribution range less than 1.0; the medium threat level corresponds to an inflection point concentration between 0.001 μg / L and 0.1 μg / L and an equilibrium constant distribution range between 1.0 and 3.0; and the high threat level corresponds to an inflection point concentration above 0.1 μg / L and an upper limit of the equilibrium constant distribution range greater than 3.0. The fitted critical micelle concentration boundary change trend, adsorption-desorption equilibrium constant distribution range, and classification thresholds were compared segment by segment to determine the ecological threat level of PFOS in the water sample. When both indicators met the corresponding level range during the comparison process, the sample was classified as having that threat level.

[0054] Based on the determined ecological threat level, the critical micelle concentration boundary variation trend, and the distribution range of adsorption-desorption equilibrium constants, corresponding optical feature parameters are extracted. These optical feature parameters include the density of gray-level abrupt change points, the high-sensitivity peak-to-valley ratio, the signal-to-noise ratio of each preset wavelength excitation spectrum, and the rate of change of transition amplitude slope in the multi-frame difference signal sequence corresponding to the concentration boundary node sampling time. The extracted optical feature parameters are aligned according to the sampling time dimension, and each feature parameter corresponds to a unique timestamp and spatial location information. The aligned feature parameters are then subjected to min-max normalization processing, with a normalization range of 0 to 1. After processing, the parameters are arranged in ascending order of time and integrated to generate an optical difference distribution feature sequence related to the pollution source. Each data unit in the sequence contains a timestamp, normalized feature parameters, corresponding ecological threat level, and concentration range information. This sequence directly reflects the optical feature fingerprint of perfluorooctane sulfonic acid pollution and can be used for source tracing analysis and pollution diffusion trend prediction, solving the deficiency of existing technologies that can only output concentration values ​​and cannot provide data support for source tracing.

[0055] S7. Based on the optical difference distribution characteristic sequence, the background fluorescence quenching coefficient distribution and the absorption cross section difference sequence under multi-wavelength excitation in the matrix are processed by signal enhancement to obtain the trace concentration quantitative results of perfluorooctane sulfonic acid in the water sample.

[0056] In one specific embodiment, performing step S7 includes the following steps: Based on the optical difference distribution characteristic sequence, the background fluorescence signal in the matrix is ​​linearly fitted to calculate the background fluorescence quenching coefficient distribution; Based on the multi-wavelength spectral detection parameters corresponding to the optical difference distribution characteristic sequence related to pollution sources, the absorption cross-section difference value of perfluorooctane sulfonic acid molecules at each excitation wavelength is calculated, and an absorption cross-section difference sequence is generated. Wavelet transform enhancement processing was performed on the difference sequence of background fluorescence quenching coefficient distribution and absorption cross section to filter out matrix interference noise and obtain the enhanced target feature signal; The enhanced target feature signal is mapped to a preset perfluorooctane sulfonic acid concentration-optical response calibration curve to obtain the quantitative result of trace concentration of perfluorooctane sulfonic acid in the water sample to be tested.

[0057] Specifically, the optical difference distribution feature sequence is arranged in ascending order of sampling time. Each sampling point includes parameters such as the density of gray-level abrupt change points, the ratio of high-sensitivity peaks and valleys, the difference in polarization state rotation angle, and the rate of change of transition amplitude slope. It is strictly aligned in the time dimension with the previously acquired multi-frame difference signal sequence and the verification result sequence. Each data unit is bound to a unique timestamp and spatial grid coordinates, ensuring a one-to-one correspondence between optical feature data and molecular aggregation behavior and pollution distribution characteristics. Background fluorescence signals synchronized with the optical difference distribution feature sequence are acquired using a fluorescence spectrometer. The acquisition wavelength range is set to 400nm to 600nm, the sampling interval is 1nm, and the integration time is 50μs. The timestamps of each fluorescence signal sampling point are perfectly matched with the corresponding sampling points in the optical difference distribution feature sequence, avoiding fitting bias caused by temporal misalignment. An adaptive iterative reweighted penalized least squares algorithm is used for baseline correction of the acquired background fluorescence signals, with the algorithm parameters set to smoothness Λ=10. 6The process involved 15 iterations with a weight threshold of 0.5. Baseline drift and low-frequency noise were removed, followed by Gaussian filtering to smooth high-frequency random noise. The filter window size was set to 3 sampling points with a standard deviation of 1. After preprocessing, a linear fit was performed using the normalized feature parameters in the optical difference distribution feature sequence as independent variables and the preprocessed background fluorescence signal intensity as the dependent variable. The slope obtained from the fit was the background fluorescence quenching coefficient corresponding to each sampling point. The detection area was then divided into a 10×10 spatial grid, with each grid corresponding to a spatial coordinate unit in the optical difference distribution feature sequence. The calculation of each... The arithmetic mean of the quenching coefficients of all sampling points within a grid is used to generate a background fluorescence quenching coefficient distribution. The value of this distribution is positively correlated with the intensity of background fluorescence interference in the matrix and negatively correlated with the distribution density of perfluorooctane sulfonic acid molecules, thus achieving spatial quantitative characterization of matrix background interference. For example, in drinking water sample testing, the quenching coefficient value corresponding to the region where perfluorooctane sulfonic acid molecules are aggregated is 0.12, and the quenching coefficient value corresponding to the matrix region without target molecules is 0.85. Through this distribution, the spatial distribution region of target molecules can be accurately located, and invalid data units with an excessively high proportion of background interference can be eliminated.

[0058] Parameters corresponding to multi-wavelength spectral detection are extracted from the optical difference distribution feature sequence, including incident light intensity, transmitted light intensity, spectral peak area, and signal-to-noise ratio data for each preset wavelength. The parameter data for each wavelength corresponds one-to-one with the sampling timestamp of the optical difference distribution feature sequence to ensure data homogeneity. The absorption cross section of perfluorooctane sulfonic acid molecules at each excitation wavelength is calculated using Lambert-Beer's law, as shown in the formula: ,in This represents the absorption cross section corresponding to the o-th excitation wavelength. Let be the intensity of the incident light at wavelength 0. Let be the transmitted light intensity at the 0th wavelength, and L be the optical path length of the sample cell, taken as 1 cm. The standardization concentration of perfluorooctane sulfonic acid standard solution was determined through a preliminary experiment using gradient concentration standard solutions. After calculation, the arithmetic mean of the absorption cross-sections at all excitation wavelengths was used as the reference cross-section. Calculate the difference in absorption cross-section for each wavelength. The absorption cross-section differences corresponding to each wavelength are then arranged in ascending order of the wavenumber of the excitation wavelength to generate an absorption cross-section difference sequence. The numerical changes in this sequence reflect the differences in the absorption characteristics of perfluorooctane sulfonic acid molecules at different excitation wavelengths, and are directly related to the molecular aggregation configuration characteristics. This solves the problem that single-wavelength detection in existing technologies is easily affected by water turbidity and matrix scattering interference, leading to poor detection stability. For example, in the micelle aggregation state, the absorption cross-section difference corresponding to the 532nm wavelength is... cm², in a single-molecule dispersed state, the difference value corresponding to this wavelength is cm², this sequence can effectively distinguish the optical response characteristics corresponding to different molecular aggregation states, providing a targeting basis for subsequent signal enhancement.

[0059] The background fluorescence quenching coefficient distribution was converted into a one-dimensional coefficient sequence according to a spatial grid. The sequence length was aligned with the absorption cross-section difference sequence length using cubic spline interpolation to ensure consistent data length and no loss of feature information. The two aligned sequences were merged into a joint input sequence. Each data unit contained the quenching coefficient and absorption cross-section difference value at the corresponding position. The joint input sequence was decomposed into a four-level wavelet decomposition using the Daubechies 4 wavelet basis, resulting in one low-frequency approximation coefficient layer and four high-frequency detail coefficient layers. The low-frequency approximation coefficient layer reflects the overall trend of the sequence and corresponds to the overall optical response of the perfluorooctane sulfonic acid molecule. The high-frequency detail coefficient layers contain the target feature signal and matrix interference noise. A general thresholding method was used to threshold the high-frequency detail coefficients. The threshold calculation formula is as follows: ,in The noise standard deviation of the joint input sequence is estimated using the absolute value of the median of the high-frequency detail coefficients in the first layer, where Q is the total length of the joint input sequence. For sequences with an absolute value less than [value missing], [the value is not specified]. The high-frequency coefficients are set to zero to eliminate random noise, and for values ​​with absolute values ​​greater than 0.5%, the coefficients are set to zero. The high-frequency coefficient is multiplied by the amplification factor μ, the value of which is determined by the high-sensitivity peak-to-valley ratio in the optical difference distribution characteristic sequence. The mapping rule was established through preliminary experiments: using gradient concentration standard solutions, different... Under a certain value, the optimal μ value that maximizes the signal-to-noise ratio of the enhanced signal is obtained by fitting the function. For example, when When μ = 2.5, the corresponding μ value is 1.8. This ensures that the target feature signal is amplified without introducing distortion. After thresholding, the processed high-frequency detail coefficients and the original low-frequency approximation coefficients are subjected to inverse wavelet transform to reconstruct the enhanced target feature signal.

[0060] The pre-defined perfluorooctane sulfonic acid (PFOS) concentration-optical response calibration curve was established through preliminary experiments using standard solutions of gradient concentrations. The standard solution concentration gradients were set to 0.001 μg / L, 0.01 μg / L, 0.1 μg / L, 1 μg / L, 10 μg / L, and 100 μg / L. Five parallel detections were performed for each concentration gradient, and the optical response signals at the corresponding concentrations were collected. The calibration curve was obtained by fitting a quadratic polynomial with the standard solution concentration as the x-axis and the average peak value of the optical response signals obtained from the parallel detections as the y-axis. The fitting formula is as follows: ,in Where is the concentration of perfluorooctane sulfonic acid, and S is the peak value of the optical response signal. , , The correlation coefficient R² must be greater than 0.995 during the fitting process to ensure the linearity and reliability of the calibration curve. The peak values ​​of the enhanced target feature signals are extracted and substituted into the calibration curve. The concentration value is then solved using Newton's iteration method, with the initial iteration value set to 0.01 μg / L and the iteration precision set to... The method uses μg / L and an iteration limit of 50 times to obtain the quantitative result of the trace concentration of perfluorooctane sulfonic acid in the water sample. This result is mutually verified with the previous ecological threat assessment result, which solves the problem that the existing technology can only output a single concentration value and cannot verify the reliability of the result. At the same time, the detection limit of this method is as low as 0.001 μg / L, which meets the detection requirements of trace perfluorooctane sulfonic acid in the aquatic environment and makes up for the shortcomings of traditional detection methods, such as cumbersome pretreatment and inability to conduct rapid on-site screening.

[0061] Please see Figure 2 , Figure 2 The graph shows the signal-to-noise ratio (SNR) versus concentration, illustrating the trends of this method, LC-MS, Raman spectroscopy, and UV spectrophotometry at different PFOS concentrations. This demonstrates that the SNR of this method is higher than the other three traditional methods across the entire concentration range, especially at trace concentrations. to Within the range of μg / L, this method maintains a signal-to-noise ratio above the detection limit, while Raman spectroscopy and ultraviolet spectrophotometry have signal-to-noise ratios below the detection limit in this concentration range, making them ineffective for detection. Although liquid chromatography-mass spectrometry can detect trace PFOS in some concentration ranges, its signal-to-noise ratio is lower than that of this method. This result confirms that this method, through laser-targeted excitation, multi-dimensional feature fusion, and signal enhancement processing, significantly reduces the detection limit, solving the technical problem of weak signals and ineffective identification in trace PFOS detection using traditional methods, demonstrating the high sensitivity advantage of this method.

[0062] Please see Figure 3 , Figure 3The comprehensive performance radar chart shows the performance scores of this patented method compared to Raman spectroscopy, liquid chromatography-mass spectrometry (LC-MS), and ultraviolet spectrophotometry in six dimensions: anti-interference, sensitivity, detection speed, detection limit, multifunctionality, and cost-effectiveness. This demonstrates that the patented method achieves high scores in all four core dimensions: anti-interference, sensitivity, detection limit, and multifunctionality. Traditional methods, on the other hand, have significant shortcomings. For example, ultraviolet spectrophotometry scores extremely low in sensitivity, detection limit, and anti-interference capability; Raman spectroscopy lacks anti-interference and multifunctionality; and liquid chromatography-mass spectrometry suffers from slow detection speed and poor cost-effectiveness. This indicates that the proposed method, through multi-module collaborative design, achieves comprehensive optimization of detection performance, not only addressing the single-dimensional performance deficiencies of traditional methods but also considering practicality and economy, meeting the comprehensive needs of trace PFOS detection in complex aquatic environments.

[0063] Please see Figure 4 The following describes a perfluorooctane sulfonic acid (PFOS) detection system in water according to an embodiment of this application. The perfluorooctane sulfonic acid (PFOS) detection system in water includes: The data acquisition module is used to excite the carbon-fluorine bond vibration mode of perfluorooctane sulfonic acid molecules in the water sample to be tested with a laser of a specific wavelength, induce the instantaneous nonlinear transition of the molecular refractive index, and obtain a preliminary optical response signal sequence containing the local thermal lensing effect. The threshold calculation module is used to extract the fluctuation range of the matrix scattered background light intensity and the peak-valley contrast ratio of the differential signal based on the preliminary optical response signal sequence, and to determine the enhancement threshold of the instantaneous refractive index nonlinear transition amplitude. The image capture module is used to determine whether the enhancement threshold exceeds the fluctuation range. If so, it captures the instantaneous transition image containing the polarization state rotation angle in the optical path and processes the instantaneous transition image to obtain a multi-frame difference signal sequence. The feature recognition module is used to extract the density of gray-scale abrupt change points and the ratio of high-sensitivity peaks and valleys from the multi-frame difference signal sequence, and to determine whether they conform to the preset response mode. If so, the molecular aggregation state configuration characteristics corresponding to the trace concentration are determined. The spectral verification module is used to obtain a sequence of verification results for the instantaneous refractive index nonlinear transition amplitude by using multi-wavelength spectral analysis to integrate the energy level parameters of carbon-fluorine bond vibrational modes, based on the configuration characteristics of molecular aggregates. The threat assessment module is used to extract the trend of the critical micelle concentration boundary of surfactants and the distribution range of adsorption-desorption equilibrium constants in the environmental medium from the validation result sequence, to determine the ecological threat level of perfluorooctane sulfonic acid in water samples, and to determine the optical difference distribution characteristic sequence related to pollution sources. The concentration quantification module is used to perform signal enhancement processing on the background fluorescence quenching coefficient distribution and the absorption cross section difference sequence under multi-wavelength excitation in the matrix based on the optical difference distribution characteristic sequence, so as to obtain the trace concentration quantification result of perfluorooctane sulfonic acid in the water sample.

[0064] Through the collaborative efforts of the aforementioned components, this system constructs a complete PFOS detection chain, encompassing laser-targeted excitation, multi-dimensional feature recognition, spectral cross-validation, simultaneous ecological risk assessment, and precise quantification of trace concentrations. This achieves high specificity, high sensitivity, and rapid detection of trace PFOS in complex aquatic matrices, simultaneously completing molecular aggregation state analysis, ecological threat classification, and pollution source feature extraction. It addresses the industry pain points of traditional detection technologies, such as cumbersome pretreatment, high detection limits, weak anti-interference capabilities, and the inability to simultaneously conduct ecological risk assessments. Specifically: The data acquisition module, serving as the system's signal source and fundamental data unit, targets and matches the vibrational frequency of the carbon-fluorine bonds in PFOS molecules to complete laser excitation, inducing instantaneous nonlinear refractive index transitions in the molecules. Simultaneously, it acquires and integrates multi-dimensional optical response signals with local thermal lensing effects, providing a time-aligned data source for subsequent stages of the entire process. This addresses the core shortcomings of traditional optical detection—lack of targeting and weak target molecule response signals—from the detection source. The threshold calculation module, acting as the system's signal triggering hub, receives the optical signals from the data acquisition module. Through differential detection, it separates the matrix background from the target response signal, adaptively calculates the transition amplitude enhancement threshold, and accurately determines the target signal triggering node. This provides a reliable start command for the image capture module, solving the problems of missed detections of weak signals and false triggering due to background noise in traditional detection methods, achieving efficient filtering of invalid data. The image capture module, serving as the system's molecular response visualization acquisition unit, receives a threshold trigger command and, through a high-speed acquisition mode locked with the laser pulse, records the dynamic changes in polarization state rotation and generates a multi-frame difference signal sequence. It extracts the spatial distribution information of molecular refractive index transitions, transforming the microscopic molecular response into a quantifiable image signal. This provides a dual-dimensional data source in both the temporal and spatial domains for the feature recognition module, solving the problems of traditional detection methods failing to capture local optical response differences and lacking spatial information. The feature recognition module, as the core of the system's target-specific recognition, receives the multi-frame difference signal sequence and extracts two core features: grayscale abrupt change density and high-sensitivity peak-valley ratio. Through dual comparison with the preset response mode of PFOS molecules, it completes the specific identification of target molecules and the analysis of their aggregation state configuration. This effectively eliminates interference signals from non-target substances in the water, solving the problems of poor specificity and high false-positive detection rates in traditional detection methods, and providing core molecular-level evidence for subsequent spectral verification. The spectral verification module, serving as the system's detection result calibration unit, performs multi-wavelength spectral scanning and energy level transition parameter extraction based on molecular aggregate configuration characteristics. It generates a transition amplitude verification result sequence through signal-to-noise ratio weighted fusion, cross-validating the previous image analysis results to eliminate biases caused by a single detection dimension. This solves the problems of poor repeatability and susceptibility to water turbidity fluctuations in traditional detection methods, providing highly reliable calibration data for subsequent ecological threat assessment. The threat assessment module, serving as the system's ecological risk analysis and pollution source feature extraction unit, receives the verification result sequence and completes critical micelle concentration boundary fitting, adsorption-desorption equilibrium constant calculation, and ecological threat level classification. Simultaneously, it integrates and generates an optical fingerprint sequence of pollution sources, achieving direct conversion from physical optical parameters to chemical ecological parameters. This solves the problems of traditional detection methods that only output concentration data, cannot simultaneously assess ecological risks, and cannot support pollution source tracing.As the system's detection result output unit, the concentration quantification module, based on the optical characteristic sequence of the pollution source, completes the quantification of background fluorescence interference, the construction of absorption cross section difference sequence, and wavelet transform signal enhancement. Finally, it outputs the trace concentration quantification result through pre-calibrated curve mapping, which solves the problem of traditional detection of trace signals being masked by background noise and insufficient quantitative accuracy. It forms mutual verification with the ecological threat assessment results, ensuring the accuracy and reliability of the detection results.

[0065] This application also provides a computer-readable storage medium, which can be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, wherein the computer-readable storage medium stores instructions that, when executed on a computer, cause the computer to perform the steps of the method for detecting perfluorooctane sulfonic acid in water.

[0066] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the methods and systems described above can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0067] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for detecting perfluorooctane sulfonic acid in water, characterized in that, Includes the following steps: S1. Excite the carbon-fluorine bond vibration mode of perfluorooctane sulfonic acid molecules in the water sample to be tested with a laser of a specific wavelength, induce the instantaneous nonlinear transition of the molecular refractive index, and obtain a preliminary optical response signal sequence containing the local thermal lensing effect. S2. Based on the preliminary optical response signal sequence, extract the fluctuation range of the matrix scattered background light intensity and the peak-valley contrast ratio of the differential signal to determine the enhancement threshold of the instantaneous refractive index nonlinear transition amplitude. S3. Determine whether the enhancement threshold exceeds the fluctuation range. If so, capture the instantaneous transition image containing the polarization state rotation angle in the optical path, and process the instantaneous transition image to obtain a multi-frame difference signal sequence. S4. Extract the density of gray-scale abrupt change points and the ratio of high-sensitivity peaks and valleys from the multi-frame difference signal sequence, and determine whether they conform to the preset response mode. If so, determine the molecular aggregation state configuration characteristics corresponding to the trace concentration. S5. Based on the molecular aggregated state configuration characteristics, multi-wavelength spectral analysis was used to integrate the energy level parameters of the carbon-fluorine bond vibrational modes to obtain a sequence of verification results for the instantaneous refractive index nonlinear transition amplitude. S6. Extract the critical micelle concentration boundary variation trend of perfluorooctane sulfonic acid and the distribution range of adsorption-desorption equilibrium constant in the environmental medium from the verification result sequence, determine the ecological threat level of perfluorooctane sulfonic acid in the water sample, and determine the optical difference distribution characteristic sequence related to the pollution source. S7. Based on the optical difference distribution characteristic sequence, the background fluorescence quenching coefficient distribution and the absorption cross section difference sequence under multi-wavelength excitation in the matrix are subjected to signal enhancement processing to obtain the trace concentration quantitative results of perfluorooctane sulfonic acid in the water sample. Step S5 includes: Based on the molecular aggregate configuration characteristics, N preset wavelengths that match the carbon-fluorine bond resonance frequency of the perfluorooctane sulfonic acid molecule are selected, and the water sample to be tested is spectrally scanned to collect excitation spectrum data of the corresponding wavelengths, where N is a positive integer; Extract the energy level transition parameters corresponding to the carbon-fluorine bond vibrational modes during laser excitation from the excitation spectral data; Using the signal-to-noise ratio of each group of excitation spectral data as the weight, the excitation spectral data and energy level transition parameters are weighted and fused to obtain a single group of instantaneous refractive index nonlinear transition amplitude verification values. The multiple sets of instantaneous refractive index nonlinear transition amplitude verification values ​​acquired consecutively are sorted according to the time dimension to generate a verification result sequence of instantaneous refractive index nonlinear transition amplitude. Step S6 includes: The verification result sequence is segmented point by point, and the abrupt change points of the slope of the transition amplitude in the sequence are identified as concentration boundary nodes. The trend of the critical micelle concentration boundary of perfluorooctane sulfonic acid is obtained by fitting. Based on the concentration boundary nodes, the Langmuir adsorption model is used to calculate the equilibrium constant of the adsorption and desorption process of perfluorooctane sulfonic acid molecules in the environmental medium of the water sample to be tested, and the distribution range of the adsorption and desorption equilibrium constant is obtained. The ecological threat level of perfluorooctane sulfonic acid in the water sample is determined by comparing the trend of the critical micelle concentration boundary of perfluorooctane sulfonic acid, the distribution range of the adsorption-desorption equilibrium constant, and the preset ecological threat classification threshold. Based on the ecological threat level, the trend of the critical micelle concentration boundary of perfluorooctane sulfonic acid and the distribution range of the adsorption-desorption equilibrium constant, the corresponding optical characteristic parameters are extracted and integrated to generate a pollution source-related optical difference distribution characteristic sequence. Step S7 includes: Based on the optical difference distribution characteristic sequence, the background fluorescence signal in the matrix is ​​linearly fitted to calculate the background fluorescence quenching coefficient distribution; Based on the multi-wavelength spectral detection parameters corresponding to the optical difference distribution characteristic sequence related to the pollution source, the absorption cross-section difference value of perfluorooctane sulfonic acid molecules at each excitation wavelength is calculated to generate an absorption cross-section difference sequence. The background fluorescence quenching coefficient distribution and absorption cross section difference sequence are enhanced by wavelet transform to filter out matrix interference noise and obtain the enhanced target feature signal. The enhanced target feature signal is mapped to a preset perfluorooctane sulfonic acid concentration-optical response calibration curve to obtain the quantitative result of trace concentration of perfluorooctane sulfonic acid in the water sample to be tested.

2. The method according to claim 1, characterized in that, Step S1 includes: The length distribution of the fluorinated carbon chain and the electronegativity distribution of the sulfonic acid group of the perfluorooctane sulfonic acid molecule in the water sample to be tested were obtained to determine the carbon-fluorine bond vibration frequency matching parameters. Based on the carbon-fluorine bond vibration frequency matching parameters, a specific wavelength laser is selected to irradiate the water sample to be tested. The energy level of the carbon-fluorine bond vibration mode is controlled by adjusting the laser pulse energy, thereby inducing a nonlinear transition of the molecular instantaneous refractive index. The amplitude of the instantaneous nonlinear transition of the refractive index of molecules is monitored, and the intensity of the thermal lensing effect in the local area of ​​the water sample under test is calculated based on the thermal effect of the medium induced by laser irradiation. Light intensity and phase information at multiple time points are collected at a preset sampling frequency, and the thermal lensing effect intensity is integrated to generate a preliminary optical response signal sequence containing the local thermal lensing effect.

3. The method according to claim 2, characterized in that, Step S2 includes: The preliminary optical response signal sequence is processed using a differential detection method to calculate the light intensity difference between adjacent sampling points in the sequence and generate a differential signal sequence. Separate the matrix scattering background signal from the differential signal sequence, count the maximum and minimum light intensity of the matrix scattering background signal, and determine the fluctuation range of the matrix scattering background light intensity; Extract the peak and valley values ​​of the differential signal sequence, calculate the ratio of peak to valley values, and obtain the peak-valley contrast ratio of the differential signal. Using the peak-valley contrast ratio as a weighting coefficient and combining it with the upper limit of the fluctuation range, the enhancement threshold of the instantaneous refractive index nonlinear transition amplitude is calculated.

4. The method according to claim 3, characterized in that, Step S3 includes: Determine whether the enhancement threshold exceeds the fluctuation range; If so, the high-speed image sensor is activated to continuously acquire multiple frames of polarized light in the optical path, record the dynamic changes in the polarization state rotation angle during laser excitation, and obtain the instantaneous transition image sequence. The instantaneous transition image sequence is subjected to frame-by-frame pixel-level interpolation to obtain a multi-frame interpolation signal sequence containing local refractive index transition spatial distribution information.

5. The method according to claim 1, characterized in that, Step S4 includes: For each frame of the multi-frame difference signal sequence, perform pixel-level traversal and calculate the gray-level gradient value of each pixel. Pixels whose gray-level gradient values ​​exceed a preset gray-level threshold are marked as gray-level abrupt change points. The spatial distribution of gray-level abrupt change points in a single frame image is statistically analyzed, and the density of gray-level abrupt change points is calculated. Scan each frame of the multi-frame difference signal sequence, identify the positions of signal peaks and valleys, calculate the relative difference between peaks and valleys, and obtain a high-sensitivity peak-valley ratio; The density of the gray-scale abrupt change points is compared with the preset density threshold range of the electron cloud density gradient of the perfluorooctane sulfonic acid molecule. The high-sensitivity peak-to-valley ratio is compared with a preset pattern of hydrophobic interface characteristics of long-chain perfluorinated compounds. Determine whether both comparison results conform to the preset response mode. If so, determine the molecular aggregate configuration characteristics corresponding to the trace concentration based on the density of corresponding gray-scale mutation points and the ratio of high-sensitivity peaks and valleys.

6. A perfluorooctane sulfonic acid (PFOS) detection system for water, used to implement the perfluorooctane sulfonic acid (PFOS) detection method for water as described in any one of claims 1 to 5, characterized in that, The perfluorooctane sulfonic acid (PFOS) detection system in water includes: The data acquisition module is used to excite the carbon-fluorine bond vibration mode of perfluorooctane sulfonic acid molecules in the water sample to be tested with a laser of a specific wavelength, induce the instantaneous nonlinear transition of the molecular refractive index, and obtain a preliminary optical response signal sequence containing the local thermal lensing effect. The threshold calculation module is used to extract the fluctuation range of the matrix scattered background light intensity and the peak-valley contrast ratio of the differential signal based on the preliminary optical response signal sequence, and to determine the enhancement threshold of the instantaneous refractive index nonlinear transition amplitude. An image capture module is used to determine whether the enhancement threshold exceeds the fluctuation range. If so, it captures an instantaneous transition image containing the polarization state rotation angle in the optical path and processes the instantaneous transition image to obtain a multi-frame difference signal sequence. The feature recognition module is used to extract the density of gray-scale abrupt change points and the ratio of high-sensitivity peaks and valleys from the multi-frame difference signal sequence, and determine whether they conform to the preset response mode. If so, the molecular aggregation state configuration feature corresponding to the trace concentration is determined. The spectral verification module is used to obtain a sequence of verification results for the instantaneous refractive index nonlinear transition amplitude by using multi-wavelength spectral analysis to integrate the energy level parameters of carbon-fluorine bond vibrational modes, based on the molecular aggregate configuration characteristics. The threat assessment module is used to extract the critical micelle concentration boundary variation trend of perfluorooctane sulfonic acid and the distribution range of adsorption-desorption equilibrium constant in the environmental medium from the verification result sequence, to determine the ecological threat level of perfluorooctane sulfonic acid in the water sample, and to determine the optical difference distribution characteristic sequence related to the pollution source. The concentration quantification module is used to perform signal enhancement processing on the background fluorescence quenching coefficient distribution and the absorption cross section difference sequence under multi-wavelength excitation in the matrix according to the optical difference distribution characteristic sequence, so as to obtain the trace concentration quantification result of perfluorooctane sulfonic acid in the water sample.

7. A computer-readable storage medium storing instructions thereon, characterized in that, When the instructions are executed by the processor, they implement a method for detecting perfluorooctane sulfonic acid in water as described in any one of claims 1-5.

Images