A high-fidelity single-sample DOM particle-fluorescent component synchronous characterization system and method

Abstract

The application discloses a high-fidelity single-sample DOM particle-fluorescent component synchronous characterization system and method, and belongs to the field of water environment monitoring.The system comprises a gradient ultrafiltration enrichment unit, an intelligent FlFFF separation unit, a time sequence-particle size linkage collection unit, a multi-parameter collaborative correction EEM detection unit and a multi-dimensional data fusion analysis module, and realizes synchronous characterization through the steps of gradient ultrafiltration enrichment, intelligent separation, differential collection, collaborative correction detection and multi-dimensional analysis.The application has the advantages of high fidelity and high precision through the technical improvements of double-stage ultrafiltration pressure control and concentration self-adaptive flow field regulation, is suitable for various water environment samples, can output a DOM biological activity score, and provides support for ecological function research.

Description

A high-fidelity single-sample DOM particle size-fluorescence component simultaneous characterization system and method Technical Field

[0001] This invention relates to the field of water environment monitoring technology, specifically to a system and method for characterizing the particle size distribution and fluorescent components of dissolved organic matter (DOM) in aquatic environments, which is particularly suitable for high-fidelity synchronous characterization of single samples. Background Technology

[0002] Dissolved organic matter (DOM) is a complex mixture of various organic compounds in the aquatic environment. Its particle size distribution and fluorescent component composition directly determine the source, migration and transformation characteristics, and ecological functions of DOM, making it a core indicator for water environment quality assessment and ecological process research. In existing technologies, the method of combining flow field flow classification (FIFFF) with fluorescence excitation-emission matrix (EEM) and parallel factor analysis (PARAFAC) has begun to be applied to the characterization of single-sample DOM. However, in practical applications, there are still many technical limitations that restrict the reliability and application scope of the characterization results.

[0003] 1. The ultrafiltration enrichment process carries the risk of structural damage: Existing technologies mostly use ultrafiltration membranes with a single molecular weight cutoff for DOM enrichment, which cannot simultaneously achieve both impurity removal and enrichment efficiency; moreover, membrane fouling is prone to occur during ultrafiltration, leading to a sudden increase in pressure, which in turn damages the natural structure of DOM and affects the authenticity of characterization.

[0004] 2. Poor adaptability of FlFFF separation parameters: The focusing flow, crossflow and other parameters of the existing FlFFF separation unit are mostly fixed settings, which cannot be dynamically adjusted according to the actual concentration of DOM in the sample. For high-concentration samples, incomplete separation is likely to occur, while for low-concentration samples, over-separation may occur, resulting in low separation efficiency and poor repeatability.

[0005] 3. Key information is missing in subsample collection: Existing technology uses uniform time intervals to collect FLFFF eluted components, but the characteristic fluorescent components of DOM (such as 1-3 kDa in humic aggregates and >100 kDa in protein aggregates) are highly enriched in specific molecular weight ranges. Uniform collection cannot accurately capture information in these key ranges, resulting in insufficient accuracy in subsequent component identification.

[0006] 4. Insufficient interference control and poor modeling robustness in EEM detection: Existing EEM detection only eliminates the internal filtering effect through simple dilution, without specifically eliminating interference signals such as Raman scattering; PARAFAC modeling relies on the statistical characteristics of the data itself, and the factor initialization is not properly guided, resulting in slow model convergence and robustness that is easily affected by data fluctuations.

[0007] 5. Limited structural fidelity verification: Existing technologies only verify the fidelity of the separation process through EEM spectral similarity, lacking direct verification of the integrity of the DOM molecular structure, and cannot fully ensure that the characterization results reflect the natural state of the DOM.

[0008] Therefore, developing a technology that can overcome the above-mentioned defects and achieve high-fidelity and high-precision synchronous characterization of DOM particle size and fluorescence components in a single sample has become an urgent technical problem to be solved in this field. Summary of the Invention

[0009] To address the above problems, this invention provides a high-fidelity single-sample DOM particle size-fluorescence component simultaneous characterization system and method.

[0010] A high-fidelity single-sample DOM particle size-fluorescence component simultaneous characterization system includes:

[0011] The gradient ultrafiltration enrichment unit adopts a dual-stage regenerated cellulose ultrafiltration membrane series structure to sequentially complete the pre-removal of macromolecular impurities and the enrichment of colloidal DOM. Both stages of ultrafiltration membranes are equipped with online pressure monitoring modules.

[0012] The intelligent FLFF separation unit integrates an adaptive flow field control module, a pre-DOC detection module, and a temperature constant module. The adaptive flow field control module and the pre-DOC detection module are linked to achieve continuous molecular weight separation of DOM.

[0013] The time-particle size linkage collection unit is equipped with a timing module and a molecular weight prediction module, which is used to collect sub-samples differentially based on the real-time elution signal of FlFFF.

[0014] Multi-parameter collaborative calibration EEM detection unit, integrating a 254 An online monitoring module, an automatic dilution device, and a fluorescence scanning module are used to acquire high signal-to-noise ratio full-spectrum EEM data;

[0015] The multi-dimensional data fusion analysis module integrates a single-sample PARAFAC enhanced modeling unit, a molecular weight-fluorescence intensity mapping unit, and a structure fidelity verification unit to complete fluorescent component identification, particle size correlation, and structural integrity verification.

[0016] Preferably, the molecular weight cutoffs of the two-stage regenerated cellulose ultrafiltration membranes in the gradient ultrafiltration enrichment unit are 3 kDa and 1 kDa, respectively, the pressure threshold of the online pressure monitoring module is 0.1-0.3 MPa, and it is linked to the feed pump signal of the gradient ultrafiltration enrichment unit.

[0017] Preferably, the adaptive flow field control module of the intelligent FLFFFF separation unit includes a focusing flow pump, a cross-flow pump and an outflow pump, with an adjustable focusing flow rate range of 1.5-2.5 mL / min and a cross-flow rate range of 0.8-1.5 mL / min; the temperature constant module has a temperature control accuracy of ±0.5℃.

[0018] Preferably, the intelligent FlFFF separation unit further includes a standard protein automatic calibration module, which has three built-in standard protein reservoirs with different molecular weights, and can automatically perform dynamic calibration of log(MW) and elution time.

[0019] Preferably, the differentiated collection strategy of the time-particle size linkage collection unit is as follows: a collection interval of 0.3 min is used in the molecular weight range of 1-3 kDa and >100 kDa, and a collection interval of 1 min is used in other molecular weight ranges.

[0020] Preferably, the multi-parameter collaborative correction EEM detection unit's a 254 The online monitoring module is linked to the automatic dilution device; when subsample a is detected... 254 Automatic dilution is initiated when the concentration is ≥0.02; the excitation wavelength scanning range of the fluorescence scanning module is 240-460 nm, and the emission wavelength scanning range is 220-620 nm.

[0021] Preferably, the single-sample PARAFAC enhancement modeling unit of the multi-dimensional data fusion analysis module adopts a non-negative matrix factorization-machine learning hybrid algorithm; the structure fidelity verification unit achieves dual verification through EEM spectral similarity matching and infrared spectral fingerprint comparison.

[0022] Preferably, the multi-dimensional data fusion analysis module also integrates a DOM activity assessment unit, which is used to calculate the component ratio and fluorescence index ratio in different molecular weight ranges and output a DOM bioactivity potential score.

[0023] A high-fidelity single-sample DOM particle size-fluorescence component simultaneous characterization method includes the following steps:

[0024] (1) Gradient ultrafiltration enrichment: The pretreated water sample is passed into the gradient ultrafiltration enrichment unit, and the macromolecular impurities are removed and colloidal DOM is enriched by passing through the two-stage ultrafiltration membrane in sequence. The pressure is monitored in real time by the online pressure monitoring module.

[0025] (2) Intelligent FlFFF separation: The enriched DOM is injected into the intelligent FlFFF separation unit, the pre-dOC detection module detects the DOM concentration, the adaptive flow field control module automatically sets the flow field parameters according to the concentration, and the temperature constant module maintains the separation temperature stable.

[0026] (3) Time-size-particle-size linkage collection: The time-size-particle-size linkage collection unit predicts the molecular weight range based on the real-time elution signal of FlFFF and collects a series of sub-samples according to the differentiated collection strategy;

[0027] (4) Multi-parameter collaborative calibration EEM detection: Subsamples are passed into the multi-parameter collaborative calibration EEM detection unit, and after detection by a254 and dilution as needed, full-spectrum EEM data acquisition and background subtraction are completed;

[0028] (5) Multidimensional data fusion analysis: Fluorescent components are identified by single-sample PARAFAC enhancement modeling, the correlation between fluorescence intensity and molecular weight is established, and the DOM particle size-fluorescent component distribution results are output after structural fidelity verification.

[0029] Preferably, step (5) further includes a DOM bioactivity assessment step: calculating the ratio of protein / humic components and the BIX / HIX ratio in different molecular weight ranges, and outputting a DOM bioactivity potential score of 0-10; the criteria for structural fidelity verification are EEM spectral similarity ≥99% and infrared characteristic functional group peak difference ≤3%.

[0030] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0031] 1. High fidelity characteristics: Through dual-stage ultrafiltration pressure control, constant temperature separation, and dual structure verification by EEM and infrared, the natural structure of DOM is ensured to be intact, and the characterization results are true and reliable, with a main peak recovery rate of ≥99% and an infrared characteristic peak difference of ≤3%;

[0032] 2. High accuracy: Intensive collection of data in key molecular weight ranges improves the accuracy of feature information capture; multi-parameter collaborative correction improves EEM data quality; enhanced PARAFAC model improves component identification accuracy; and the error in calculating the relative contribution of components is ≤2%.

[0033] 3. Strong adaptability: The intelligent adaptive flow field control can adapt to samples with DOC concentrations of 200-800 μmol / L, and is suitable for various water environment samples such as rivers, lakes, and lagoons, without the need for manual parameter adjustment;

[0034] 4. High efficiency: The enhanced PARAFAC model has a convergence speed that is more than 40% faster, and the entire process is automated. The total time from sample preprocessing to result output is ≤6 hours, and the calibration cycle is ≤30 minutes.

[0035] 5. Functional Extension: The integration of DOM bioactivity assessment function enables a leap from "structural characterization" to "functional assessment," expanding the application scenarios of the technology and providing more comprehensive data support for water environment ecological assessment and pollutant migration risk prediction. Attached Figure Description

[0036] Figure 1 is a schematic diagram of the system of the present invention. Detailed Implementation

[0037] The following embodiments, in conjunction with the accompanying drawings, are merely for illustrating the technical solutions described in the claims and are not intended to limit the scope of protection of the claims.

[0038] Example 1

[0039] A high-fidelity single-sample DOM particle size-fluorescence component simultaneous characterization system includes: a gradient ultrafiltration enrichment unit employing a dual-stage regenerated cellulose ultrafiltration membrane tandem structure for sequentially performing pre-removal of macromolecular impurities and enrichment of colloidal DOM; both ultrafiltration membranes are equipped with online pressure monitoring modules; an intelligent FLFFF separation unit integrating an adaptive flow field control module, a pre-DOC detection module, and a temperature constant module; the adaptive flow field control module and the pre-DOC detection module are signal-linked to achieve continuous molecular weight separation of DOM; a time-series-particle size linkage collection unit equipped with a timing module and a molecular weight prediction module for differentially collecting subsamples based on real-time FLFFF elution signals; and a multi-parameter collaboratively calibrated EEM detection unit integrating a… 254 The online monitoring module, automatic dilution device, and fluorescence scanning module are used to acquire high signal-to-noise ratio full-spectrum EEM data; the multi-dimensional data fusion analysis module integrates a single-sample PARAFAC enhancement modeling unit, a molecular weight-fluorescence intensity mapping unit, and a structure fidelity verification unit to complete fluorescent component identification, particle size correlation, and structural integrity verification.

[0040] The gradient ultrafiltration enrichment unit features two-stage regenerated cellulose ultrafiltration membranes with molecular weight cutoffs of 3 kDa and 1 kDa, respectively. The 3 kDa ultrafiltration membrane is used to pre-remove large molecular organic impurities >3 kDa and some particulate matter from the water sample, while the 1 kDa ultrafiltration membrane is used to precisely retain and enrich colloidal DOM in the 1-3 kDa range. The online pressure monitoring module has a pressure threshold of 0.1-0.3 MPa and is linked to the feed pump signal of the gradient ultrafiltration enrichment unit. When the inlet and outlet pressure of the ultrafiltration membrane exceeds 0.3 MPa, the feed pump automatically stops to prevent membrane fouling from causing a sudden pressure surge that could damage the natural structure of the DOM.

[0041] The adaptive flow field control module of the intelligent FLFFFF separation unit includes a focusing flow pump, a cross-flow pump, and an effluent flow pump. The adjustable focusing flow rate ranges from 1.5 to 2.5 mL / min, the cross-flow rate ranges from 0.8 to 1.5 mL / min, and the rate gradient deceleration rate can be automatically adjusted according to the sample concentration within the range of 0.03 to 0.05 mL / min. The temperature constant module has a temperature control accuracy of ±0.5℃, and maintains the channel temperature at a stable 25℃ through a constant temperature jacket on the outer wall of the separation channel, thereby improving the repeatability of the separation of different batches of samples.

[0042] The intelligent FlFFF separation unit also includes an automatic standard protein calibration module, which has built-in standard protein reservoirs for three different molecular weights: cytochrome C (12.4 kDa), bovine serum albumin (66.4 kDa), and thyroglobulin (669 kDa). It can automatically start the calibration process before or after sample analysis, complete the dynamic calibration of log (MW) and elution time, with a calibration cycle of ≤30 min, and output an accurate molecular weight-elution time regression equation.

[0043] The differentiated collection strategy of the time-particle size linkage collection unit is as follows: a collection interval of 0.3 min is used in the molecular weight range of 1-3 kDa and >100 kDa, which are the feature enrichment ranges of humic substances and protein aggregates (DOMs), respectively. Intensive collection can improve the capture accuracy of feature component information; a collection interval of 1 min is used in other molecular weight ranges to balance information integrity and detection efficiency; the molecular weight coverage of the collected subsamples is 0.3 kDa to >100 kDa, and the total number of subsamples is ≥50.

[0044] Among them, the multi-parameter collaborative correction EEM detection unit a 254 The online monitoring module is linked to the automatic dilution device, a 254 The absorbance coefficient of the sample at a UV wavelength of 254 nm is given in cm⁻¹. - ¹, when subsample a is detected 254 ≥0.02 cm - ¹ Automatic dilution is initiated at this time, with a dilution factor of 1-5 times, until a 254 <0.02 cm - ¹, Eliminate the internal filtering effect in fluorescence detection; The excitation wavelength scanning range of the fluorescence scanning module is 240-460 nm with a scanning step size of 4 nm, and the emission wavelength scanning range is 220-620 nm with a scanning step size of 4 nm, ensuring full wavelength coverage of the characteristic fluorescence peaks of DOM.

[0045] The single-sample PARAFAC enhanced modeling unit of the multi-dimensional data fusion analysis module employs a non-negative matrix factorization-machine learning hybrid algorithm, introducing fluorescence prior feature libraries of humic and protein DOMs to optimize initialization and improve model convergence speed and robustness. The structure fidelity verification unit achieves dual verification through EEM spectral similarity matching and infrared spectral fingerprint comparison. EEM spectral similarity is determined by calculating the correlation coefficient between the reconstructed synthetic spectrum and the original sample spectrum. Infrared spectral fingerprint comparison is performed at 3400 cm⁻¹. - ¹(hydroxyl), 1600 cm - ¹(aromatic ring), 1050 cm - ¹Characteristic functional group peaks such as (polysaccharides) are observed.

[0046] The multi-dimensional data fusion analysis module also integrates a DOM activity assessment unit, which is used to calculate the ratio of protein / humic components and the BIX / HIX ratio in different molecular weight ranges. Combined with the relative contribution ratio of components in each range, it outputs a DOM bioactivity potential score of 0-10. The higher the score, the stronger the bioavailability of DOM.

[0047] Example 2

[0048] A high-fidelity single-sample DOM particle size-fluorescence component synchronous characterization method includes the following steps: (1) Gradient ultrafiltration enrichment: the pretreated water sample is passed into the gradient ultrafiltration enrichment unit, and the macromolecular impurities are removed and colloidal DOM is enriched by passing through a double-stage ultrafiltration membrane in sequence. The pressure is monitored in real time by the online pressure monitoring module; (2) Intelligent FlFFF separation: the enriched DOM is injected into the intelligent FlFFF separation unit, the pre-DOC detection module detects the DOM concentration, the adaptive flow field control module automatically sets the flow field parameters according to the concentration, and the temperature constant module maintains the separation temperature stability; (3) Time-series-particle size linkage collection: the time-series-particle size linkage collection unit predicts the molecular weight range according to the real-time elution signal of FlFFF, and collects a series of sub-samples according to the differentiated collection strategy; (4) Multi-parameter collaborative correction EEM detection: the sub-sample is passed into the multi-parameter collaborative correction EEM detection unit, and the pressure is monitored in real time by the online pressure monitoring module; (5) Time-series-particle size linkage collection: the time-series-particle size linkage collection unit predicts the molecular weight range according to the real-time elution signal of FlFFF, and collects a series of sub-samples according to the differentiated collection strategy; (6) Multi-parameter collaborative correction EEM detection: the sub-sample is passed into the multi-parameter collaborative correction EEM detection unit, and the pressure is monitored in real time by the online pressure monitoring module; (7) Time-series-particle size linkage collection: the time-series-particle size linkage collection unit predicts the molecular weight range according to the real-time elution signal of FlFFF, and collects a series of sub-samples according to the differentiated collection strategy; (8) Multi-parameter collaborative correction EEM detection: the sub-sample is passed into the multi-parameter collaborative correction EEM detection unit, and the pressure is monitored in real time by the online pressure monitoring module; (9 254 After detection and dilution as needed, complete the full spectrum EEM data acquisition and background subtraction; (5) Multi-dimensional data fusion analysis: identify fluorescent components by single sample PARAFAC enhancement modeling, establish the correlation between fluorescence intensity and molecular weight, and output DOM particle size-fluorescent component distribution results after structure fidelity verification.

[0049] In step (1), the water sample pretreatment is as follows: after collecting the water environment sample, the suspended particulate matter is removed by vacuum filtration through a 0.7 μm GF / F filter membrane to obtain a clear filtrate; the dual-stage ultrafiltration membrane needs to be pretreated before use. The pretreatment process is as follows: first, wash with 0.05 M NaOH solution for 30 min, and then rinse repeatedly with ultrapure water until the CDOM fluorescence characteristics of the effluent are consistent with those of the ultrapure water to avoid contamination of the sample by residual impurities in the membrane; the online pressure monitoring module collects the inlet and outlet pressure data of the dual-stage ultrafiltration membrane in real time, and the pressure is maintained in the range of 0.1-0.3 MPa. The retentate of the 1 kDa ultrafiltration membrane is collected to complete the enrichment of colloidal DOM.

[0050] In step (2), the pre-DOC detection module detects the concentration range of DOM after enrichment as 200-800 μmol / L. When the concentration is ≥500 μmol / L, the adaptive flow field control module sets the initial value of the focusing flow to 2.5 mL / min, the initial value of the crossflow to 1.5 mL / min, and the gradient deceleration rate to 0.05 mL / min. When the concentration is <500 μmol / L, the initial value of the focusing flow is set to 1.5 mL / min, the initial value of the crossflow is set to 0.8 mL / min, and the gradient deceleration rate to 0.03 mL / min. The standard protein automatic calibration module is started before sample separation, and the output log (MW) has an R² ≥ 0.99 with the regression equation of elution time.

[0051] In step (3), the molecular weight prediction module of the time-particle size linkage collection unit determines the range based on the UV254 absorbance signal and characteristic fluorescence signal (Ex / Em=350 / 450 nm, 275 / 340 nm) eluted by FlFFF in real time: when the peak value of the Ex / Em=350 / 450 nm signal is detected, it is determined to be in the 1-3 kDa range; when the peak value of the Ex / Em=275 / 340 nm signal is detected and the elution time is >20 min, it is determined to be in the >100 kDa range; the subsamples are collected according to the differentiated collection strategy, and the subsamples are stored in light-proof sample bottles and refrigerated at 4℃ for later use to ensure stable fluorescence characteristics.

[0052] In step (4), a 254 The online monitoring module performs pre-detection on each sub-sample, with a detection optical path of 1 cm and a detection time ≤10 s / sample; the automatic dilution device uses ultrapure water as the diluent, and the dilution process is a gradient dilution, with re-detection after each dilution. 254 until a is satisfied 254 <0.02 cm -¹ Requirements: During fluorescence scanning, the EEM background of the blank carrier solution at the corresponding elution time point must be subtracted. The blank carrier solution is the same elution medium as that used during sample separation to eliminate fluorescence interference from the carrier solution itself.

[0053] In step (5), the single-sample PARAFAC enhancement modeling is as follows: the EEM data of all sub-samples are vertically stacked in the elution order to form a three-dimensional data array, and the dimension of the data array is the number of excitation wavelength points × the number of emission wavelength points × the number of sub-samples; the robustness of the model is verified by split-half analysis, requiring TCC≥0.98, and the optimal component number is determined by core consistency diagnosis, requiring CORCONDIA≥85%; the molecular weight-fluorescence intensity mapping is as follows: based on the regression equation obtained by standard protein calibration, the elution time of each sub-sample is converted into the average molecular weight, the relative intensity of each fluorescent component in the sub-sample is extracted, the component intensity-molecular weight distribution curve is plotted, and the relative contribution ratio of each component in different molecular weight ranges is calculated; in the structure fidelity verification, when the EEM spectral similarity is ≥99% and the difference in peak intensity of infrared characteristic functional groups is ≤3%, the DOM structure is determined to be intact.

[0054] Step (5) also includes a DOM bioactivity assessment step: calculating the ratio of protein / humic components and the BIX / HIX ratio in different molecular weight ranges, where BIX is the biogenic index and HIX is the humification index. Combining the relative contribution ratio of each range of components, a DOM bioactivity potential score of 0-10 is output through weighted calculation. The criteria for structural fidelity verification are EEM spectral similarity ≥99% and infrared characteristic functional group peak difference ≤3%.

[0055] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. The basic principles and main features of the present invention have been described above with specific implementation schemes. Based on the present invention, some modifications or substitutions can be made, but these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of protection claimed by the present invention.

Claims

1. A high-fidelity single-sample DOM particle size-fluorescence component simultaneous characterization system, characterized in that, include: The gradient ultrafiltration enrichment unit adopts a dual-stage regenerated cellulose ultrafiltration membrane series structure to sequentially complete the pre-removal of macromolecular impurities and the enrichment of colloidal DOM. Both stages of ultrafiltration membranes are equipped with online pressure monitoring modules. The intelligent FLFFFF separation unit integrates an adaptive flow field control module, a pre-DOC detection module, and a temperature control module. The adaptive flow field control module and the pre-DOC detection module are signal-linked to achieve continuous molecular weight separation of DOM. The adaptive flow field control module of the intelligent FLFFFF separation unit includes a focusing flow pump, a cross-flow pump, and an effluent flow pump, with an adjustable focusing flow rate range of 1.5-2.5 mL / min and a cross-flow rate range of 0.8-1.5 mL / min. The temperature control module has a temperature control accuracy of ±0.5℃. A time-series-particle-size linked collection unit, equipped with a timing module and a molecular weight prediction module, is used to collect subsamples differentially based on the real-time FLFF elution signal. A multi-parameter collaboratively calibrated EEM detection unit integrates a… 254 The online monitoring module, automatic dilution device, and fluorescence scanning module are used to acquire high signal-to-noise ratio full-spectrum EEM data; the multi-dimensional data fusion analysis module integrates a single-sample PARAFAC enhancement modeling unit, a molecular weight-fluorescence intensity mapping unit, and a structure fidelity verification unit to complete fluorescent component identification, particle size correlation, and structural integrity verification.

2. The system according to claim 1, characterized in that, The gradient ultrafiltration enrichment unit has two-stage regenerated cellulose ultrafiltration membranes with molecular weight cutoffs of 3 kDa and 1 kDa, respectively. The pressure threshold of the online pressure monitoring module is 0.1-0.3 MPa, and it is linked to the feed pump signal of the gradient ultrafiltration enrichment unit.

3. The system according to claim 1, characterized in that, The intelligent FlFFF separation unit also includes a standard protein automatic calibration module, which has three built-in standard protein reservoirs with different molecular weights, and can automatically perform dynamic calibration of log(MW) and elution time.

4. The system according to claim 1, characterized in that, The differentiated collection strategy of the time-particle size linkage collection unit is as follows: a collection interval of 0.3 min is used in the molecular weight range of 1-3 kDa and >100 kDa, and a collection interval of 1 min is used in other molecular weight ranges.

5. The system according to claim 1, characterized in that, The multi-parameter collaborative correction EEM detection unit a 254 The online monitoring module is linked to the automatic dilution device; when subsample a is detected... 254 Automatic dilution is initiated when the concentration is ≥0.02; the excitation wavelength scanning range of the fluorescence scanning module is 240-460 nm, and the emission wavelength scanning range is 220-620 nm.

6. The system according to claim 1, characterized in that, The single-sample PARAFAC enhancement modeling unit of the multi-dimensional data fusion analysis module adopts a non-negative matrix factorization-machine learning hybrid algorithm; the structure fidelity verification unit achieves dual verification through EEM spectral similarity matching and infrared spectral fingerprint comparison.

7. The system according to claim 1, characterized in that, The multi-dimensional data fusion analysis module also integrates a DOM activity assessment unit, which is used to calculate the component ratio and fluorescence index ratio in different molecular weight ranges and output a DOM bioactivity potential score.

8. A high-fidelity single-sample DOM particle size-fluorescence component simultaneous characterization method, characterized in that, Includes the following steps: (1) Gradient ultrafiltration enrichment: The pretreated water sample is passed into the gradient ultrafiltration enrichment unit, and the macromolecular impurities are removed and colloidal DOM is enriched by passing through a two-stage ultrafiltration membrane in sequence. The pressure is monitored in real time by the online pressure monitoring module; (2) Intelligent FLFFF separation: The enriched DOM is injected into the intelligent FLFFF separation unit. The pre-DOC detection module detects the DOM concentration, the adaptive flow field control module automatically sets the flow field parameters according to the concentration, and the temperature constant module maintains the separation temperature stably; The adaptive flow field control module of the intelligent FLFFF separation unit includes a focusing flow pump, a cross-flow pump and an outflow pump. The adjustable focusing flow rate range is 1.5-2.5 mL / min, and the cross-flow rate range is 0.8-1.5 mL / min. mL / min; the temperature control accuracy of the constant temperature module is ±0.5℃; (3) Time-particle size linkage collection: the time-particle size linkage collection unit predicts the molecular weight range based on the real-time elution signal of FlFFF and collects a series of sub-samples according to the differentiated collection strategy; (4) Multi-parameter collaborative correction EEM detection: the sub-samples are passed into the multi-parameter collaborative correction EEM detection unit, and after a254 detection and as-needed dilution, the full spectrum EEM data acquisition and background subtraction are completed; (5) Multi-dimensional data fusion analysis: the fluorescent components are identified by single sample PARAFAC enhancement modeling, the correlation between fluorescence intensity and molecular weight is established, and the DOM particle size-fluorescent component distribution results are output after structural fidelity verification.

9. The method according to claim 8, characterized in that, Step (5) also includes a DOM bioactivity assessment step: calculating the ratio of protein / humic components and the BIX / HIX ratio in different molecular weight ranges, and outputting a DOM bioactivity potential score of 0-10; the criterion for structural fidelity verification is EEM spectral similarity ≥99%. The difference in peak values ​​of infrared characteristic functional groups is ≤3%.

Images