A heavy metal ion-oriented water pollution online monitoring method

Abstract

This invention discloses an online monitoring method for heavy metal ions in water pollution. Water samples are collected, and interference features are extracted simultaneously using photochemical colorimetry, electrochemical voltammetry, and conductivity and organic matter analysis to construct a multidimensional feature matrix. An interference level coefficient is calculated based on a support vector machine model, and the method automatically switches to a rapid electrochemical mode, an electrochemical-assisted enhanced spectral mode, or a spectrally-dominated anti-interference mode based on this coefficient to achieve online detection of heavy metal ions. Finally, the results are output and automatically calibrated periodically. This invention achieves synergistic effects between electrochemistry and spectroscopy through adaptive mode switching and deep coupling detection, improving anti-interference capability, sensitivity, and accuracy. It is suitable for online monitoring of surface water and industrial wastewater.

Description

Technical Field

[0001] This invention relates to the field of environmental protection monitoring technology, and in particular to an online monitoring method for water pollution targeting heavy metal ions. Background Technology

[0002] With rapid industrialization, wastewater containing heavy metals discharged from industries such as electroplating, mining, and chemicals poses a serious threat to surface water and groundwater. Heavy metal ions such as lead (Pb²⁺), cadmium (Cd²⁺), mercury (Hg²⁺), and arsenic (As³⁺) are biotoxic and non-degradable, and can accumulate through the food chain even at extremely low concentrations, ultimately harming human health. Therefore, online, rapid, and accurate monitoring of heavy metal ions in water bodies is of great significance for environmental pollution early warning and control.

[0003] Currently, existing online heavy metal monitoring technologies are mainly divided into two categories: electrochemical analysis and spectroscopic analysis. Electrochemical analysis, such as anodic stripping voltammetry, is widely used due to its simple equipment and high sensitivity. However, this method suffers from a significant "multi-ion interference effect" in practical applications: when multiple heavy metal ions or organic substances such as surfactants are present in the water, the stripping peak potentials of each ion may overlap or shift, leading to inaccurate ion identification and even false alarms. Furthermore, while traditional mercury film electrodes are stable, they contain highly toxic mercury, which easily causes secondary pollution. The electrode surface is also susceptible to contamination by organic matter in complex water bodies, leading to signal attenuation and frequent maintenance.

[0004] Another category is spectroscopic analysis methods, such as atomic emission spectroscopy and laser-induced breakdown spectroscopy (LIBS). These methods can achieve simultaneous detection of multiple elements, but the equipment is expensive and consumes a lot of power. Furthermore, when directly detecting liquid samples, they suffer from low sensitivity and significant matrix effects. For example, directly using LIBS to detect liquids not only presents the problem of liquid splashing, but water molecules also quench the spectral signal, making it difficult to meet the Class I water requirements of the surface water environmental quality standards.

[0005] To address the limitations of single technologies, existing techniques have proposed simple combinations, such as combining electrochemical enrichment with spectroscopic detection. However, this simple "enrich first, detect later" approach is a mechanical accumulation of features and does not solve the problem of collaborative operation between the two technologies in dynamic online monitoring scenarios. First, the conditions for electrochemical enrichment (such as enrichment potential and time) need to be preset based on the unknown ion species in the water sample. If the preset is inappropriate, it will lead to low enrichment efficiency for specific ions, which will still fail to be detected by subsequent spectroscopic detection. Second, the triggering timing of spectroscopic detection is usually disconnected from the electrochemical dissolution process, making it impossible to capture the transient, highest concentration signal peak, resulting in limited improvement in detection sensitivity. Finally, when water quality fluctuations cause a surge in the concentration of a certain interfering substance (such as surfactant), the electrochemical module may completely fail, and if the system is unaware of this and continues to operate according to the established process, it will continuously output erroneous data.

[0006] Therefore, existing technologies lack an online monitoring method capable of intelligently sensing water sample matrix interference, dynamically and adaptively adjusting detection strategies, and deeply coupling electrochemical and spectroscopic technologies to achieve mutual signal enhancement. This technological disconnect results in monitoring systems failing to meet practical early warning requirements in terms of stability, accuracy, and anti-interference capabilities when facing complex and variable aquatic environments. Summary of the Invention

[0007] To achieve the above objectives, the present invention provides an online monitoring method for water pollution targeting heavy metal ions, comprising the following steps: Step 1: Collect the water sample to be tested, divide the water sample into 3 equal parts and inject them into 3 independent detection chambers. Simultaneously collect the first set of multispectral response vectors, the first set of electrochemical response vectors, matrix conductivity and dissolved organic matter concentration through photochemical colorimetric sensor array, differential pulse stripping voltammetry scanning and conductivity and organic matter analyzer. Construct a multidimensional feature matrix characterizing the interference characteristics of the current water sample. Step 2: Input the multidimensional feature matrix into the pre-trained classification model, calculate a real number between 0 and 1 as the interference level coefficient, and automatically select and switch to one of the following modes for subsequent online monitoring: fast electrochemical mode, electrochemical-assisted enhanced spectroscopy mode, or spectroscopy-dominated anti-interference mode, based on the numerical range of the interference level coefficient. Step 3: If switching to the fast electrochemical mode, a square wave anodic stripping voltammetry scan is performed using a boron-doped diamond working electrode. The ion type is determined based on the peak potential of the stripping peak, and the ion concentration is calculated based on the peak current intensity. Step 4: If switching to the electrochemical-assisted enhanced spectroscopy mode, apply an anodic dissolution pulse voltage after electrochemical enrichment, and simultaneously trigger the laser-induced breakdown spectroscopy system to emit a laser pulse at the instant the anodic dissolution pulse voltage is applied, collect and analyze plasma characteristic spectral lines to calculate ion concentration; Step 5: If the spectrum-dominated anti-interference mode is switched to, the water sample to be tested is adsorbed through a chelating resin enrichment column, then eluted with dilute nitric acid solution, and the eluent is introduced into a microwave plasma torch atomizer to collect and analyze the emission spectrum to calculate the ion concentration.

[0008] Preferably, the specific process of simultaneously acquiring the first set of multispectral response vectors, the first set of electrochemical response vectors, matrix conductivity, and dissolved organic matter concentration in step 1 is as follows: In the first detection chamber, a fiber optic photochemical colorimetric sensor array containing three different heavy metal ion indicators—dithizone sensitive membrane, xylenol orange sensitive membrane, and pyridine azonaphthol sensitive membrane—was inserted. The sensor array was illuminated with white LED light, and a CCD image sensor continuously acquired color change images of the sensor array after reacting with the water sample for 60 seconds. The processor extracted the RGB value changes of each sensitive membrane region before and after the reaction and generated the first set of multispectral response vectors. In the second detection chamber, an electrochemical sensor system consisting of a working electrode made of boron-doped diamond, an Ag / AgCl reference electrode, and a platinum wire counter electrode is inserted. An acetate-sodium acetate buffer solution with a pH of 4.5 is added to the second detection chamber to make the volume ratio of the water sample to the buffer solution 10:1. Differential pulse stripping voltammetry is used to scan the potential range of -1.3V to +0.5V, and the voltammetric curve is recorded. The peak potential and peak current of each stripping peak are extracted from the voltammetric curve to generate the first set of electrochemical response vectors. In the third detection chamber, the matrix conductivity of the water sample was measured using a conductivity meter, and the concentration of dissolved organic matter in the water sample was measured using a total organic carbon analyzer.

[0009] Preferably, the specific process of constructing the multidimensional feature matrix representing the current water sample interference characteristics in step 1 and calculating a real number ranging from 0 to 1 as the interference level coefficient in step 2 is as follows: The processor merges the first set of multispectral response vectors, the peak potentials and peak currents in the first set of electrochemical response vectors, the matrix conductivity, and the concentration of dissolved organic matter to construct a multidimensional feature matrix. The multidimensional feature matrix is ​​input into a pre-trained support vector machine classification model, which outputs an interference level coefficient. The interference level coefficient is a real number between 0 and 1. The larger the value of the interference level coefficient, the more severe the interference caused by the coexistence of organic matter or multiple ions in the water sample.

[0010] Preferably, the specific process of using a boron-doped diamond working electrode to perform square-wave anodic stripping voltammetry scanning, determining the ion type based on the peak potential of the stripping peak, and calculating the ion concentration based on the peak current intensity in step 3 is as follows: Step 3.1: Continuously inject fresh water sample into the flow cell and stir. Apply a constant potential of -1.0V to the working electrode for enrichment. The enrichment time is 300 seconds, so that the heavy metal ions in the water sample are reduced and enriched on the surface of the boron-doped diamond working electrode. Step 3.2: Stop stirring and injection, let stand for 15 seconds, then use square wave anodic stripping voltammetry to scan from -1.0V to +0.5V at a scan rate of 50mV / s and record the stripping voltammetric curve; Step 3.3: By analyzing the dissolution peaks appearing at different potentials on the dissolution voltammetry curve, and comparing them with the built-in standard ion dissolution potential library, the ion types are determined; Step 3.4: Based on the peak current intensity of each dissolution peak, substitute it into the pre-stored standard curve equation of the ion on the boron-doped diamond working electrode to calculate the real-time concentration of each heavy metal ion in the water sample to be tested.

[0011] Preferably, the specific process in step 4, which involves applying an anodic dissolution pulse voltage after electrochemical enrichment and simultaneously triggering the laser-induced breakdown spectroscopy system to emit a laser pulse at the instant the anodic dissolution pulse voltage is applied, is as follows: Step 4.1: Collect a fresh water sample and inject it into the three-electrode flow cell. The bottom of the three-electrode flow cell is equipped with a high-purity graphite plate as the working electrode, and an optical quartz glass window is provided on the opposite side. Step 4.2: Add supporting electrolyte to the three-electrode flow cell to make the final concentration 0.1 mol / L, apply a constant potential of -1.2 V to the working electrode for enrichment, enrichment time is 600 seconds, and at the same time circulate the water sample at a flow rate of 1 mL / min using a micro peristaltic pump. Step 4.3: After enrichment is completed, stop the sample injection and empty the liquid in the three-electrode flow cell, and then fill it with high-purity argon as a protective gas; Step 4.4: Start the Q-switched laser of the laser-induced breakdown spectroscopy system. At the instant a anodic dissolution pulse voltage with an amplitude of +1.5V and a pulse width of 50ms is applied, the Q-switched laser is simultaneously triggered to emit a laser pulse with an energy of 80mJ and a wavelength of 1064nm. The laser pulse is focused on the surface of the high-purity graphite plate through the optical quartz glass window. Step 4.5: The plasma light signal generated by the laser breaking down the atomic cloud is collected by the fiber optic probe, and then dispersed and detected by the spectrometer under the conditions of detection wavelength 200nm to 800nm ​​and resolution 0.1nm. The processor collects the spectral data and analyzes the intensity of the characteristic spectral lines of heavy metal elements.

[0012] Preferably, the synchronous triggering in step 4.4 means that the synchronous delay time between the pulse signal generated by the high-voltage power supply and the laser pulse emitted by the Q-switched laser is controlled within 1 microsecond, so that the laser pulse is precisely applied to the cloud of heavy metal atoms that have just been released from the surface of the high-purity graphite plate by the anodic dissolution pulse.

[0013] Preferably, the specific process in step 5, which involves adsorbing the water sample to be tested through a chelating resin enrichment column, eluting it with dilute nitric acid solution, and then introducing the eluent into a microwave plasma torch atomizer, is as follows: Step 5.1: Collect fresh water samples to be tested, adjust the pH value of the water samples to 5.5, and then pass them through a micro-enrichment column containing chelating resin at a flow rate of 2 mL / min, so that the heavy metal ions in the water samples are adsorbed by the chelating resin. Step 5.2: After enrichment is completed, the micro-enrichment column is rapidly flushed in the reverse direction with 3 mL of 1 mol / L dilute nitric acid solution to elute the adsorbed heavy metal ions. Step 5.3: The eluent is directly introduced into the atomizer of the microwave plasma torch, and after being atomized by high-purity nitrogen, it forms an aerosol that enters the center of the plasma torch flame; Step 5.4: Start the microwave source and set the power to 1000W. After the heavy metal ions are atomized and excited in the high-temperature plasma torch, they emit characteristic spectra. The emission spectra are collected in the 200nm to 500nm band by a fiber optic spectrometer. Step 5.5: The processor performs peak finding and background subtraction on the emission spectrum, and calculates the concentration of heavy metal ions in the original water sample based on the emission spectral intensity of each element at a specific wavelength, combined with the enrichment factor of online elution and the standard curve.

[0014] Preferably, the system further includes step 6: transmitting the heavy metal ion types and concentration data calculated in steps 3, 4, or 5 to the data platform of the remote monitoring center in real time via a 4G wireless transmission module; simultaneously, after each 24 monitoring cycles, the system automatically extracts a standard mixed solution containing 50 μg / L each of Pb²⁺ and Cd²⁺, repeats the complete process from steps 1 to 5, calculates the drift coefficient after comparing the detected values ​​with the standard values, and automatically corrects the standard curve equation parameters in steps 3, 4, and 5 to eliminate zero-point drift and sensitivity drift caused by long-term system operation.

[0015] Preferably, the specific determination rule for automatically selecting and switching to the corresponding mode based on the numerical range of the interference level coefficient in step 2 is as follows: if the interference level coefficient is less than or equal to 0.3, the current water sample is determined to be a low-interference clean water body and switched to the rapid electrochemical mode; if the interference level coefficient is greater than 0.3 and less than 0.7, the current water sample is determined to be a moderately interfered complex water body and switched to the electrochemical-assisted enhanced spectral mode; if the interference level coefficient is greater than or equal to 0.7, the current water sample is determined to be a highly interfered polluted water body and switched to the spectral-dominated anti-interference mode.

[0016] Preferably, the specific process of collecting the water sample to be tested in step 1 is as follows: the water sample to be tested is collected from the water area to be monitored by an automatic sampling pump, and the collected water sample is divided into three equal parts of equal volume, which are injected into the first detection chamber, the second detection chamber and the third detection chamber in the pre-detection unit, which are independent of each other and not connected to each other.

[0017] The beneficial effects of this invention are: 1. This invention breaks down the technical barriers between electrochemical and spectroscopic methods. It is not a simple sequential combination, but rather achieves synergistic operation of the two methods through a complete technical chain of interference feature extraction, adaptive decision-making, and deep coupling triggering. In the electrochemical-assisted enhanced spectral mode, electrochemical enrichment creates high-concentration sample points for spectral detection, while precise spectral triggering precisely captures the instantaneous high-concentration signal from electrochemical dissolution. The synchronization of these two methods in the spatiotemporal dimensions improves the detection signal-to-noise ratio by more than two orders of magnitude compared to any single method or simple combination method, producing a significant synergistic effect of "1+1>2". 2. This invention, through the construction of a multi-dimensional feature matrix and an adaptive mode switching mechanism, enables the system to intelligently sense changes in water quality and automatically select the optimal detection mode. In high-interference environments, the system automatically switches to a spectral-dominated anti-interference mode, which, combined with chelating resin separation technology, effectively avoids the failure of electrochemical methods due to electrode poisoning. In low-interference environments, the system employs a rapid electrochemical mode to achieve high-sensitivity detection. In moderate-interference environments, an electrochemical-assisted enhanced spectral mode leverages its synergistic advantages. This hierarchical adaptive strategy ensures the continuity and reliability of monitoring data under different water quality conditions. 3. This invention achieves fully intelligent operation throughout the entire process. From water sample collection, interference identification, mode switching, online detection, to automatic calibration, everything is completed automatically by the system without manual intervention, significantly reducing on-site maintenance workload. The system's built-in automatic calibration program executes once every 24 monitoring cycles, dynamically correcting the parameters of the standard curve equation using a standard mixture, effectively eliminating zero-point drift and sensitivity drift caused by long-term operation. Boron-doped diamond electrodes replace traditional mercury film electrodes, eliminating secondary pollution and reducing the detection limits for elements such as lead and cadmium to as low as 0.05 micrograms per liter, meeting the Class I surface water quality monitoring requirements. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in this invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, those skilled in the art can obtain other drawings based on these drawings without creative effort.

[0019] Figure 1 This is a flowchart of the steps of the method of the present invention; Figure 2 This is a flowchart of step 3 of the method of the present invention; Figure 3 This is a flowchart of step 4 of the method of the present invention. Detailed Implementation

[0020] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. It should also be noted that, to make the embodiments more comprehensive, the following embodiments are the best and preferred embodiments, and those skilled in the art can use other alternative methods to implement some well-known technologies; moreover, the accompanying drawings are only for more specific description of the embodiments and are not intended to specifically limit the present invention.

[0021] Please see Figures 1-3This invention provides an online monitoring method for heavy metal ions in water pollution. This method achieves intelligent, high-precision, and highly interference-resistant online monitoring of heavy metal ions in water bodies through multi-mode pre-detection and interference feature extraction, adaptive monitoring mode decision-making and switching, and three deeply coupled online monitoring modes: rapid electrochemical mode, electrochemical-assisted enhanced spectroscopy mode, and spectrally-dominated anti-interference mode. The entire method consists of a series of ordered steps, each completed through precise control and data processing.

[0022] Step 1: Multi-mode pre-detection and interference feature extraction The purpose of step 1 is to simultaneously collect the physicochemical characteristics of the water sample under test using multiple sensors, and to construct a multidimensional feature matrix that can characterize the current level of interference in the water sample. The specific process is as follows: First, water samples are collected from the water area to be monitored using an automatic sampling pump. The collected water samples are then divided into three equal parts by a three-way splitter and injected into the first, second, and third detection chambers of the pre-detection unit, which are independent and not connected to each other.

[0023] In the first detection chamber, a fiber optic photochemical colorimetric sensor array is pre-installed. This sensor array consists of three independent sensitive membrane sites, each with a different sensitive membrane: a dithizone sensitive membrane, a xylenol orange sensitive membrane, and a pyridine azonaphthol sensitive membrane. The dithizone sensitive membrane exhibits a selective color reaction to heavy metal ions such as lead and cadmium, typically changing from red to purple; the xylenol orange sensitive membrane is sensitive to ions such as lead and zinc, changing from yellow to red; and the pyridine azonaphthol sensitive membrane is sensitive to ions such as copper and cobalt, changing from orange to green. The sensor array is vertically illuminated with white LED light, and a CCD image sensor continuously acquires images of the color changes after the sensor array reacts with the water sample. The reaction time is precisely controlled to 60 seconds. The processor extracts the change in RGB values ​​for each sensitive membrane region before and after the reaction, i.e., the values ​​of the red, green, and blue channels after the reaction minus the values ​​of the corresponding channels before the reaction. For example, for a dithizone-sensitive membrane, the RGB values ​​before the reaction are (220, 215, 200), and after the reaction they become (235, 190, 160), so the change is (+15, -25, -40). The processor calculates the RGB changes of the three sensitive membranes respectively, and combines these three sets of nine values ​​to generate the first set of multispectral response vectors.

[0024] In the second detection chamber, an electrochemical sensor system was inserted, consisting of a boron-doped diamond working electrode, an Ag / AgCl reference electrode, and a platinum wire counter electrode. First, an acetate-sodium acetate buffer solution, with its pH precisely controlled at 4.5, was automatically added to the second detection chamber. The addition volume was controlled by a precision peristaltic pump, ensuring a volume ratio of the water sample to the buffer solution of exactly 10:1. Then, differential pulse stripping voltammetry was used for scanning, with the scanning potential range set from -1.3 V to +0.5 V (relative to the Ag / AgCl reference electrode). The scanning parameters were: pulse amplitude 50 mV, pulse width 50 ms, and step potential 4 mV. After the scan, the complete voltammetric curve was recorded. The processor smoothed, baseline-corrected, and processed the second derivative of the voltammetric curve, identifying the peak potential and peak current of each stripping peak. The peak potential is the potential value corresponding to the maximum stripping peak current, and the peak current is the peak height after subtracting the baseline. These data are used to generate the first set of electrochemical response vectors, including the peak potentials and corresponding peak currents of all detected dissolution peaks.

[0025] In the third detection chamber, a miniature conductivity meter and a total organic carbon (TOC) analyzer are integrated. The conductivity meter's electrodes are directly immersed in the water sample, and the matrix conductivity of the water sample is determined by measuring the impedance between the two electrodes, typically in microsiemens per centimeter (μS / cm). The TOC analyzer uses a UV-catalytic non-dispersive infrared (NDIR) method. First, the organic matter in the water sample is catalytically oxidized to carbon dioxide under UV light irradiation. Then, the carbon dioxide concentration is measured using a NIR detector, from which the dissolved organic matter concentration is calculated, typically in milligrams per liter (mg / L). The processor reads the values ​​output from the conductivity meter and the TOC analyzer to obtain the matrix conductivity and dissolved organic matter concentration.

[0026] At this point, the processor has collected all of the following raw data: The first set of multispectral response vectors contains the RGB changes of the three sensitive films, for a total of nine values.

[0027] Peak potentials in the first set of electrochemical response vectors: For each detected dissolution peak, record its peak potential.

[0028] Peak currents in the first set of electrochemical response vectors: For each detected dissolution peak, record its peak current.

[0029] Matrix conductivity: a real value.

[0030] Dissolved organic matter concentration: a real value.

[0031] The processor merges this data to construct a multidimensional feature matrix. This matrix is ​​a vector containing multiple eigenvalues, the dimension of which depends on the number of detected dissolution peaks, but typically includes all of the aforementioned values. When the number of dissolution peaks is zero, the peak potential and peak current terms can be set to default values ​​(e.g., 0). The multidimensional feature matrix is ​​used for subsequent interference level determination.

[0032] Step 2: Decision-making and switching of adaptive monitoring mode The purpose of step 2 is to calculate the interference level coefficient based on the multidimensional feature matrix constructed in step 1 using a pre-trained classification model, and to automatically select the subsequent online monitoring mode based on the numerical range of the coefficient.

[0033] First, a support vector machine (SVM) classification model needs to be pre-trained. The training process is as follows: In a laboratory setting, different concentrations of standard heavy metal ions (such as lead (Pb²⁺) and cadmium (Cd²⁺) and different concentrations of typical interfering substances (such as humic acid representing natural organic matter and sodium dodecyl sulfate representing surfactants) are added to clean water to construct thousands of water samples with varying degrees of interference. For each sample, its multidimensional feature matrix is ​​collected according to the method in step 1. Simultaneously, experienced analysts use authoritative methods (such as inductively coupled plasma mass spectrometry) to determine the actual concentration of the target heavy metal ions in the water sample. Based on the concentration of interfering substances and the actual detection difficulty, a real number ranging from 0 to 1 is manually assigned to each sample as an interference level coefficient. The closer the interference level coefficient is to 0, the less interference the water sample experiences, and conventional electrochemical methods can accurately detect it; the closer the interference level coefficient is to 1, the more severe the interference, and conventional methods fail. Using these thousands of multidimensional feature matrices as input and the corresponding manually assigned interference level coefficient as output, a support vector machine regression model is trained. The kernel function of the support vector machine is a radial basis function, and the penalty parameter and kernel function parameter are optimized through cross-validation. The trained model can output a real number between 0 and 1, i.e., the interference level coefficient γ, based on the input multidimensional feature matrix.

[0034] In actual monitoring, the processor inputs the multidimensional feature matrix constructed in step 1 into the trained support vector machine classification model. After internal calculations, the model outputs an interference level coefficient γ, where γ is a real number between 0 and 1.

[0035] The system has the following built-in judgment rules, which are used to automatically select and switch to the corresponding online monitoring mode based on the γ value: If γ≤0.3, the current water sample is determined to be a low-interference clean water body, and the system switches to rapid electrochemical mode.

[0036] If 0.3 < γ < 0.7, the current water sample is determined to be a moderately disturbed and complex water body, and the system switches to the electrochemical-assisted enhanced spectral mode.

[0037] If γ≥0.7, the current water sample is determined to be a highly polluted water body, and the system switches to the spectrum-dominated anti-interference mode.

[0038] Step 3: Execute the rapid electrochemical mode When the system determines the water body to be clean and with low interference, step 3, the rapid electrochemical mode, is executed. This mode utilizes a boron-doped diamond working electrode for square-wave anodic stripping voltammetry detection. The specific process is as follows: Step 3.1: Enrichment. The system starts a peristaltic pump to continuously inject fresh water samples into a dedicated flow cell. The flow cell integrates a boron-doped diamond working electrode, an Ag / AgCl reference electrode, and a platinum wire counter electrode. The magnetic stir bar at the bottom of the flow cell is activated to accelerate mass transfer. A constant potential of -1.0 volts (relative to Ag / AgCl) is applied to the boron-doped diamond working electrode, and the enrichment time is precisely controlled to 300 seconds. During this process, heavy metal ions (such as Pb²⁺ and Cd²⁺) in the water sample are reduced and enriched on the electrode surface, forming a metal film or amalgam (since the boron-doped diamond electrode itself does not contain mercury, the metal film is formed directly through reduction).

[0039] Step 3.2: Dissolution and Detection. After enrichment, the system immediately stopped sample injection and stirring, allowing the solution to stand for 15 seconds to stabilize the diffusion layer on the electrode surface and reduce convection interference. Subsequently, square wave anodic stripping voltammetry was used for scanning. The scanning parameters were: onset potential -1.0 V, termination potential +0.5 V, scan rate 50 mV / s, square wave frequency 25 Hz, and square wave amplitude 25 mV. The system recorded the complete stripping voltammetric curve, i.e., the curve of current versus potential.

[0040] Step 3.3: Ion Species Identification. The processor analyzes the dissolution voltammetry curve to identify the dissolution peaks. A dissolution peak is a local maximum of current. The ion species is determined by comparing the potential of the dissolution peak with a built-in standard ion dissolution potential library. The standard ion dissolution potential library consists of standard dissolution peak potentials of various heavy metal ions pre-measured under identical experimental conditions (same electrode, supporting electrolyte, pH value, etc.). For example, the standard dissolution peak potential of cadmium ion (Cd²⁺) is -0.62 V ± 0.02 V, and the standard dissolution peak potential of lead ion (Pb²⁺) is -0.42 V ± 0.02 V. If a dissolution peak appears near -0.62 V, it is identified as a cadmium ion; if a dissolution peak appears near -0.42 V, it is identified as a lead ion.

[0041] Step 3.4: Concentration Calculation. For each identified ion, the concentration is calculated by substituting the peak current intensity (peak height) of its dissolution peak into the pre-stored standard curve equation for that ion on a boron-doped diamond working electrode. The standard curve equation is obtained by preparing a series of standard solutions of known concentrations, measuring their peak currents according to the same electrochemical detection procedure, and then performing a linear regression between concentration and peak current. For example, the standard curve equation for cadmium ions is... ,in Concentration (micrograms per liter). Peak current (microamps) and For the regression coefficient, is the regression coefficient. Similarly, the standard curve equation for lead ions is... By substituting the measured peak current into the corresponding equation, the real-time concentration of each heavy metal ion in the water sample can be calculated.

[0042] Step 4: Execute electrochemical-assisted enhanced spectroscopy mode When the system determines the water body to be moderately disturbed and complex, step 4, electrochemically assisted enhancement of the spectral mode, is executed. This mode achieves signal enhancement through the simultaneous triggering of electrochemical enrichment and laser-induced breakdown spectroscopy. The specific process is as follows: Step 4.1: Online Enrichment and Electrode Modification. A fresh water sample is collected and injected into a specially designed three-electrode flow cell. A high-purity graphite plate is embedded at the bottom of the flow cell as the working electrode, and an optical quartz glass window is positioned opposite it for laser incident light and spectral signal collection. A counter electrode and a reference electrode are also present in the flow cell. Potassium nitrate, the supporting electrolyte, is added to the flow cell to achieve a final concentration of 0.1 mol / L. A constant potential of -1.2 V (relative to the reference electrode) is applied to the working electrode for enrichment for 600 seconds. Simultaneously, a micro-peristaltic pump is activated, circulating the water sample at a flow rate of 1 mL / min between the flow cell and a storage bottle to continuously replenish the analyte ions and improve enrichment efficiency.

[0043] Step 4.2: Synchronous Triggering of Pulsed Dissolution and Laser-Induced Breakdown Spectroscopy. After enrichment, the system stops sample injection and the liquid in the flow cell is emptied using a precision syringe pump. Then, high-purity argon gas is introduced into the flow cell as a protective gas at a flow rate of 0.5 liters per minute for 30 seconds to remove air and prevent oxygen and nitrogen interference lines during laser excitation. Next, the Q-switched laser of the laser-induced breakdown spectroscopy system is preheated and put into standby mode. The key step is synchronous triggering: the processor simultaneously outputs a transistor-to-transistor logic (TTL) high-level signal at the instant a large-amplitude anodic dissolution pulse voltage is applied. The amplitude of this anodic dissolution pulse voltage is +1.5 volts, and the pulse width is 50 milliseconds. The TTL signal triggers the Q-switch of the Q-switched laser, causing it to emit a laser pulse within an extremely short delay time. The synchronization delay time between the pulse signal generated by the high-voltage power supply and the laser pulse is controlled within 1 microsecond. The laser pulse energy is 80 millijoules, and the wavelength is 1064 nanometers. This laser pulse is precisely focused onto the surface of the high-purity graphite plate through an optical quartz glass window. Because the anodic stripping pulse instantly oxidizes and dissolves the metal enriched on the electrode surface, it forms a localized high-concentration atomic cloud. The laser pulse arrives precisely at the moment when the atomic cloud concentration is at its highest, thus enabling instantaneous, in-situ detection of the target analyte and significantly improving the intensity of the spectral signal.

[0044] Step 4.3: Spectral Acquisition and Analysis. Laser penetration of atomic clouds generates high-temperature plasma, which emits characteristic spectral lines during cooling. These optical signals are collected by a fiber optic probe placed outside an optical quartz glass window and transmitted to the spectrometer via fiber optic cable. The spectrometer's detection band is set to 200 nm to 800 nm, with an optical resolution of 0.1 nm. The spectrometer converts the optical signals into electrical signals and records a complete spectrum. The processor performs baseline removal, peak finding, and other processing on the spectrum to identify the characteristic spectral lines of the heavy metal element to be measured. For example, cadmium has a strong characteristic spectral line at 228.8 nm (Cd I 228.8 nm), and lead has a strong characteristic spectral line at 405.7 nm (Pb I 405.7 nm). The processor measures the intensity (peak height or peak area) of these characteristic spectral lines. Based on the intensity of the characteristic spectral lines for each element, combined with a pre-established standard curve using the standard addition method, the system calculates the concentration of the corresponding heavy metal ion in the water sample. The process of establishing a standard curve using the standard addition method is as follows: take equal portions of water sample, add different amounts of standard solution to each sample, and measure them according to the same steps. Plot the added concentration as the abscissa and the spectral intensity as the ordinate, fit a straight line, and the intersection of the extrapolation with the abscissa is the sample concentration.

[0045] Step 5: Execute the spectrum-dominated anti-interference mode When the system determines that the water body is highly polluted, step 5, the spectral-dominated anti-interference mode, is executed. This mode uses chelating resin to enrich and separate interfering substances, and then combines this with microwave plasma emission spectroscopy for detection. The specific process is as follows: Step 5.1: Sample Pretreatment. The system collects a fresh water sample for testing. First, a pH adjustment unit automatically adds dilute nitric acid or sodium hydroxide solution to precisely adjust the pH of the water sample to 5.5. Then, the pH-adjusted water sample is passed through a micro-enrichment column packed with chelating resin at a flow rate of 2 ml / min. The chelating resin used is an iminodiacetic acid type chelating resin, which has a high selective adsorption capacity for heavy metal ions. Heavy metal ions (such as Pb²⁺, Cd²⁺) in the water sample are adsorbed by the resin, while a large amount of organic matter, alkali metal and alkaline earth metal ions, and other interfering substances in the water sample are discharged with the waste liquid.

[0046] Step 5.2: Online Elution and Atomization. The enrichment process continues for a sufficient time (e.g., 10 minutes) to achieve the desired enrichment factor. Afterward, the system switches the six-way valve and rapidly flushes the micro-enrichment column in the opposite direction with 3 mL of a 1 mol / L dilute nitric acid solution. The dilute nitric acid instantly elutes the heavy metal ions adsorbed on the resin, forming a high-concentration eluent. This eluent is directly introduced into the atomizer of the microwave plasma torch. In the atomizer, the eluent is atomized by high-purity nitrogen gas, forming a fine aerosol, which is then carried by the carrier gas into the central channel of the microwave plasma torch flame.

[0047] Step 5.3: Emission Spectroscopy Detection. The system starts the microwave source and stabilizes the microwave power at 1000 watts. In the high-temperature plasma torch exceeding 5000 K, heavy metal ions are rapidly atomized, excited, and emit characteristic spectra. Emission spectra are continuously acquired in the 200 nm to 500 nm wavelength range using a fiber optic spectrometer. The spectrometer resolution should be high enough (e.g., 0.1 nm) to resolve adjacent spectral lines. The characteristic emission peaks of the analyte will appear in the acquired spectra.

[0048] Step 5.4: Concentration Calculation. The processor performs peak finding and background subtraction on the acquired emission spectra. Based on the emission spectral intensity (peak height) of each element at a specific wavelength, combined with the enrichment factor of online elution and a pre-established standard curve, the concentration of heavy metal ions in the original water sample is calculated. The enrichment factor is equal to the ratio of the injection volume to the elution volume. The standard curve is the relationship between spectral intensity and concentration measured under the same experimental conditions using a series of standard solutions through the same enrichment-elution process.

[0049] Step 6: Monitoring Result Output and System Self-Calibration Step 6 includes real-time transmission of monitoring results and periodic automatic calibration of the system to ensure long-term operational accuracy.

[0050] First, the system transmits the heavy metal ion types and concentration data calculated in steps 3, 4, or 5, along with monitoring time, water sample temperature, and other information, to the data platform of the remote monitoring center in real time via a 4G wireless transmission module, according to an agreed communication protocol (such as Modbus), and records it in the database.

[0051] Secondly, the system has a built-in automatic calibration program. Every 24 monitoring cycles (typically corresponding to 24 hours, as each cycle is approximately 1 hour), the system automatically extracts a standard mixture from its built-in standard solution bottle. This standard mixture contains known concentrations of lead and cadmium ions, each at 50 micrograms per liter. The system repeats the complete process from steps 1 to 5 with this standard mixture, starting with pre-detection, undergoing adaptive mode selection, and finally calculating the concentration value. The detected concentration value is compared with the standard value (50 micrograms per liter) to calculate the drift coefficient. For example, if the detected value of lead ions is 48.5 micrograms per liter, the drift coefficient is 48.5 / 50 = 0.97; if the detected value of cadmium ions is 51.2 micrograms per liter, the drift coefficient is 51.2 / 50 = 1.024. The system automatically corrects the parameters of the standard curve equation used for concentration calculation in steps 3, 4, and 5. For the fast electrochemical mode, the slope k and intercept b of the standard curve equation are corrected; for the electrochemical-assisted enhanced spectroscopy mode, the LIBS standard curve is corrected; and for the spectrally-dominated anti-interference mode, the microwave plasma emission spectroscopy standard curve is corrected. The correction method is as follows: the original slope is divided by the drift coefficient (when the drift coefficient < 1) or multiplied by the reciprocal of the drift coefficient (when the drift coefficient > 1), and the intercept is recalculated based on the new slope or remains unchanged. This automatic calibration eliminates zero-point drift and sensitivity drift that may occur during long-term system operation, ensuring the accuracy of subsequent monitoring data.

[0052] The above steps constitute a complete online monitoring method for heavy metal ions in water pollution. This method achieves accurate, reliable, and online monitoring of heavy metal ions in complex water bodies by intelligently identifying the degree of interference in water samples and adaptively switching to the most suitable detection mode.

[0053] Example An environmental protection department in a northern industrial city installed an automatic water quality monitoring station at a downstream section of a major polluted river within its jurisdiction. Upstream of this section are electroplating plants, metal processing plants, and other businesses. The main potential pollutants in the water are lead ions (Pb²⁺) and cadmium ions (Cd²⁺), and during the rainy season, surface runoff may bring in humus and other organic matter, further contributing to the pollution. The monitoring station is equipped with an online monitoring system implementing the method of this invention.

[0054] Step 1: Multi-mode pre-detection and interference feature extraction At 5:00 AM on a spring morning, the system initiated a routine monitoring operation according to a preset program. The automatic sampling pump collected 500 ml of water sample from the river. The system then divided the collected water sample into three equal portions, each approximately 166.7 ml, via a three-way splitter. These portions were then injected into three independent and unconnected detection chambers in the pre-detection unit.

[0055] In the first detection chamber, a fiber optic photochemical colorimetric sensor array is pre-installed. This array consists of three independent sensitive membrane sites, each with a different sensitive membrane: a dithizone sensitive membrane, a xylenol orange sensitive membrane, and a pyridine azonaphthol sensitive membrane. These three membranes exhibit selective colorimetric reactions to different heavy metal ions. The system activates a white LED to vertically illuminate the sensor array. A high-resolution CCD image sensor begins continuously acquiring images of the color changes after the sensor array reacts with the water sample. After a precise 60-second reaction time, the processor extracts the RGB value changes for each sensitive membrane region before and after the reaction. For example, the dithizone sensitive membrane region changes from (R: 220, G: 215, B: 200) before the reaction to (R: 235, G: 190, B: 160) after the reaction, a change of (R: +15, G: -25, B: -40). Similarly, the processor calculates the RGB changes for the other two sensitive membrane regions and combines these three sets of changes to generate the first set of multispectral response vectors.

[0056] An electrochemical sensor system was inserted into the second detection chamber. This system consisted of a boron-doped diamond working electrode, an Ag / AgCl reference electrode, and a platinum wire counter electrode. The system first automatically added an acetate-sodium acetate buffer solution to the second detection chamber, with the pH precisely controlled at 4.5. The addition volume was precisely controlled by a peristaltic pump, ensuring a volume ratio of the water sample to the buffer solution of exactly 10:1. Subsequently, differential pulse stripping voltammetry was used for scanning, with the scanning potential range set from -1.3V to +0.5V. After the scan, the system recorded the complete voltammetric curve. The processor processed the curve using its second derivative to locate the peaks and extract the peak potential and peak current of each stripping peak. In this detection, only a tiny stripping peak was found at -0.43V, with a peak current of 0.32 μA. The processor generated these data into the first set of electrochemical response vectors.

[0057] In the third detection chamber, the system integrates a miniature conductivity meter and a total organic carbon (TOC) analyzer. The conductivity meter probe is directly immersed in the water sample, and the matrix conductivity of the current water sample is measured to be 342 microsiemens per centimeter. The TOC analyzer, using ultraviolet-catalyzed nondispersive infrared absorption spectrometry, measures the dissolved organic matter concentration of the water sample to be 2.8 milligrams per liter.

[0058] At this point, the processor has collected all of the following raw data: The first set of multispectral response vectors: the RGB changes of the three sensitive films.

[0059] Peak potential in the first group of electrochemical response vectors: -0.43V.

[0060] Peak current in the first group of electrochemical response vectors: 0.32 μA.

[0061] Matrix conductivity: 342 μS / cm.

[0062] Dissolved organic matter concentration: 2.8 mg / L.

[0063] The processor merges this data to construct a multidimensional feature matrix. This matrix is ​​a vector containing seven eigenvalues, for example, in the form of... .

[0064] Step 2: Decision-making and switching of adaptive monitoring mode The system has a pre-trained support vector machine classification model. This model was developed in the laboratory stage by adding different concentrations of standard heavy metal ions (lead, cadmium) and different concentrations of typical interfering substances (humic acid, sodium dodecyl sulfate) to clean water bodies, constructing thousands of water samples with different levels of interference. The multidimensional feature matrix of the samples was extracted according to the method in step 1. At the same time, experienced analysts manually assigned an interference level coefficient between 0 and 1 to each water sample as a label based on the detection results of authoritative methods such as chromatography-mass spectrometry.

[0065] The processor inputs the multidimensional feature matrix constructed in step 1 into this support vector machine classification model. After kernel function calculation and decision function mapping within the model, the model outputs an interference level coefficient γ. The calculated value of γ is 0.25.

[0066] The system has the following built-in judgment rules: If γ≤0.3, the current water sample is determined to be a low-disturbance clean water body.

[0067] If 0.3 < γ < 0.7, then the current water sample is determined to be a moderately disturbed and complex water body.

[0068] If γ ≥ 0.7, the current water sample is determined to be a highly polluted water body.

[0069] Based on the value of γ=0.25, the system determines that the current water sample is a low-interference clean water body and automatically switches to the rapid electrochemical mode to prepare for subsequent steps.

[0070] Step 3: Execute the rapid electrochemical mode The system commands execute a fast electrochemical mode. This mode is performed in a specially designed flow cell that integrates the same boron-doped diamond working electrode as in step 1.

[0071] Step 3.1: Enrichment. The system starts the peristaltic pump to continuously inject fresh water samples into the flow cell, and the magnetic stir bar at the bottom of the cell is activated for stirring. Simultaneously, a constant potential of -1.0V is applied to the boron-doped diamond working electrode, and the enrichment time is precisely controlled to 300 seconds. During this process, lead and cadmium ions in the water sample are reduced and enriched on the electrode surface, forming a metal film.

[0072] Step 3.2: Dissolution and Detection. After enrichment, the system immediately stopped injection and stirring, allowing the solution to stand for 15 seconds to stabilize the diffusion layer on the electrode surface. Subsequently, square wave anodic stripping voltammetry was used for scanning at a scan rate of 50 mV / s, with a scan range from -1.0 V to +0.5 V. The system recorded the complete stripping voltammetric curve.

[0073] Step 3.3: Ion Species Identification. The processor analyzes the dissolution voltammetry curve. Two distinct, symmetrical dissolution peaks appear at -0.62V and -0.43V, respectively. The processor compares these peak potentials with a built-in standard ion dissolution potential library. The library records that, under the same conditions, the standard dissolution peak potential for cadmium ions is -0.62V ± 0.02V, and for lead ions it is -0.42V ± 0.02V. Therefore, the system determines that the peak at -0.62V corresponds to cadmium ions, and the peak at -0.43V corresponds to lead ions.

[0074] Step 3.4: Concentration Calculation. The system measured the dissolution peak current of cadmium ions to be 1.21 μA and the dissolution peak current of lead ions to be 0.79 μA. For each ion, the system pre-stored a standard curve equation established on this boron-doped diamond electrode. The standard curve equation for cadmium ions is as follows: ,in Concentration (micrograms per liter). This represents the peak current (microamps). The standard curve equation for lead ions is: Substituting the values ​​into the calculation, we get: Micrograms per liter.

[0075] Micrograms per liter.

[0076] The system has now successfully obtained the lead and cadmium ion concentrations for this monitoring.

[0077] (Scene switch: Suppose it's during another monitoring session) Suppose that after a heavy rain, the river water carries a large amount of surface organic matter. The system executes steps 1 and 2 again. At this time, the color sensor response in the first detection chamber is significantly enhanced, the voltammetric curve in the second detection chamber shows multiple overlapping peaks and a raised baseline, and the dissolved organic matter concentration measured in the third detection chamber increases to 15.6 mg / L. After inputting the constructed multidimensional feature matrix into the model, the output interference level coefficient γ is 0.55. This value falls between 0.3 and 0.7, therefore the system determines it to be a moderately disturbed complex water body and switches to the electrochemically assisted enhanced spectral mode, executing step 4.

[0078] Suppose that after a heavy rain, the river water carries a large amount of surface organic matter. The system executes steps 1 and 2 again. At this time, the color sensor response in the first detection chamber is significantly enhanced, the voltammetric curve in the second detection chamber shows multiple overlapping peaks and a raised baseline, and the dissolved organic matter concentration measured in the third detection chamber increases to 15.6 mg / L. After inputting the constructed multidimensional feature matrix into the model, the output interference level coefficient γ is 0.55. This value falls between 0.3 and 0.7, therefore the system determines it to be a moderately disturbed complex water body and switches to the electrochemically assisted enhanced spectral mode, executing step 4.

[0079] Step 4: Execute electrochemical-assisted enhanced spectroscopy mode Step 4.1: Online Enrichment and Electrode Modification. A fresh water sample is collected and injected into a specially designed three-electrode flow cell. A high-purity graphite plate is embedded at the bottom of the flow cell as the working electrode, with an optical quartz glass window positioned directly above and opposite it to allow laser transmission. Potassium nitrate, the supporting electrolyte, is added to the flow cell to achieve a final concentration of 0.1 mol / L. A constant potential of -1.2V is applied to the working electrode, and the enrichment time is 600 seconds. Simultaneously, a micro-peristaltic pump is activated, circulating the water sample between the flow cell and a storage bottle at a flow rate of 1 mL / min to maximize enrichment efficiency.

[0080] Step 4.2: Synchronous Triggering of Pulsed Dissolution and Laser-Induced Breakdown Spectroscopy. After enrichment, the system stops sample injection and the liquid in the flow cell is emptied using a precision syringe pump. Then, high-purity argon gas is introduced into the flow cell at a flow rate of 0.5 liters per minute for 30 seconds to purge air and create an inert gas environment for laser excitation. Next, the system activates the Q-switched laser of the laser-induced breakdown spectroscopy system for preheating and standby. The crucial step is synchronous triggering: the processor simultaneously outputs a transistor-to-transistor logic (TTL) high-level signal at the instant a large-amplitude anodic dissolution pulse voltage is applied. The amplitude of this anodic dissolution pulse voltage is +1.5V, and the pulse width is 50 milliseconds. The TTL signal triggers the Q-switch of the Q-switched laser, causing it to emit a laser pulse with an energy of 80 millijoules and a wavelength of 1064 nanometers within a 1-microsecond delay. This laser pulse is precisely focused onto the surface of the high-purity graphite plate through an optical quartz glass window. Because the anodic stripping pulse instantly oxidizes and dissolves the metal enriched on the electrode surface, it forms a localized high-concentration atomic cloud. The laser pulse arrives precisely at the moment when the concentration of this atomic cloud is at its highest, enabling instantaneous, in-situ detection of the target analyte.

[0081] Step 4.3: Spectral Acquisition and Analysis. Laser penetration of atomic clouds generates high-temperature plasma. During the cooling process, each element emits spectral lines of characteristic wavelengths. These optical signals are collected by an optical fiber probe placed outside a quartz window and transmitted to a spectrometer via optical fiber. The spectrometer's detection band is set to 200 nm to 800 nm, with an optical resolution of 0.1 nm. The spectrometer converts the optical signals into electrical signals and records a complete spectrum. The processor performs baseline removal, peak finding, and other processing on the spectrum, identifying the characteristic spectral line of cadmium at 228.8 nm (Cd I 228.8 nm) and the characteristic spectral line of lead at 405.7 nm (Pb I 405.7 nm). Based on the intensity of these two characteristic spectral lines, the system combines a pre-established standard curve (e.g., using the standard addition method) with... The concentrations of lead and cadmium ions in the water sample were calculated.

[0082] (Scenario switch: Suppose we are monitoring under another extreme condition) Suppose that in the early morning of a certain day, an upstream company illegally discharges untreated wastewater with a high concentration of cadmium and a large amount of organic additives. After the system executes steps 1 and 2, due to extremely strong organic interference, the electrochemical signal is completely overwhelmed, the photochemical sensor shows drastic color changes, and the dissolved organic matter concentration reaches as high as 55.0 mg / L. The calculated interference level coefficient γ reaches 0.88, which is greater than 0.7. The system determines that the water body is highly polluted and switches to the spectrum-dominated anti-interference mode, executing step 5.

[0083] Step 5: Execute the spectrum-dominated anti-interference mode Step 5.1: Sample Pretreatment. The system collects a fresh water sample for testing. First, a pH adjustment unit automatically adds dilute nitric acid or sodium hydroxide solution to precisely adjust the pH of the water sample to 5.5. Then, the adjusted water sample is passed through a micro-enrichment column packed with iminodiacetic acid-type chelating resin at a flow rate of 2 ml per minute. Heavy metal ions (such as cadmium ions) in the water sample are efficiently adsorbed by the chelating resin, while a large amount of interfering substances such as organic matter, alkali metals, and alkaline earth metal ions are discharged with the waste liquid.

[0084] Step 5.2: Online Elution and Nebulization. The enrichment process lasts for 10 minutes to ensure a sufficient enrichment factor. Afterward, the system switches the six-way valve and rapidly flushes the micro-enrichment column in the opposite direction with 3 mL of a 1 mol / L dilute nitric acid solution. The dilute nitric acid instantly elutes the cadmium ions adsorbed on the resin, forming a high-concentration eluent. This eluent is directly introduced into the nebulizer of the microwave plasma torch.

[0085] Step 5.3: Emission Spectroscopy Detection. In the atomizer, the eluent is atomized by high-purity nitrogen gas, forming a fine aerosol, which enters the central channel of the microwave plasma torch flame along with the carrier gas. The system starts the microwave source and stabilizes the microwave power at 1000 watts. In the high-temperature plasma torch exceeding 5000K, cadmium ions are rapidly atomized, excited, and emit characteristic spectra. The emission spectra are continuously acquired in the 200 nm to 500 nm wavelength range using a fiber optic spectrometer. In the acquired spectra, strong emission peaks of cadmium at 228.8 nm and 326.1 nm can be clearly seen.

[0086] Step 5.4: Concentration Calculation. The processor performs peak finding and background subtraction on the emission spectrum. Based on the emission intensity of cadmium at 228.8 nm, combined with the enrichment factor of this enrichment process (e.g., the ratio of injection volume to elution volume) and a pre-established standard curve, the concentration of cadmium ions in the original water sample is calculated. Due to the separation and enrichment by chelating resin, the spectral detection is completely free from matrix interference from organic matter, resulting in accurate and reliable detection results.

[0087] Step 6: Monitoring Result Output and System Self-Calibration After completing the detection in any of the above modes, the system proceeds to step 6. The processor calculates the types of heavy metal ions (lead, cadmium) and their concentrations (e.g., 0.63 micrograms per liter and 0.34 micrograms per liter), along with monitoring time, water sample temperature, and other information, and packages them into a data packet. This packet is then transmitted in real time to the data platform of the remote environmental monitoring center via the built-in 4G wireless transmission module, following the Modbus communication protocol, and a monitoring record is generated in the platform's database.

[0088] In addition, the system has a built-in automatic calibration program. Every 24 monitoring cycles (approximately 24 hours), the system automatically extracts a standard mixture from its built-in standard solution bottle. This standard mixture contains known concentrations of lead and cadmium ions, each at 50 micrograms per liter. The system repeats the complete process of steps 1 to 5 with this standard mixture. Assuming that in this calibration, the system detects a lead ion concentration of 48.5 micrograms per liter and a cadmium ion concentration of 51.2 micrograms per liter, the system compares the detected values ​​with the standard values ​​(50 micrograms per liter), calculating the drift coefficient for lead ions as 48.5 / 50 = 0.97 and the drift coefficient for cadmium ions as 51.2 / 50 = 1.024. Subsequently, the system automatically corrects the parameters of the standard curve equations used for concentration calculations in steps 3, 4, and 5. For example, for lead ions, the original standard curve equation is... The system will calculate the slope. Divide by 0.97, intercept Make appropriate adjustments to eliminate zero-point drift and sensitivity drift that may occur during long-term system operation, and ensure the accuracy of subsequent monitoring data.

[0089] To verify the superiority of the method of this invention, a comparative experiment was conducted under laboratory conditions. A set of simulated water samples was prepared, with lead and cadmium ion concentrations both at 5.0 micrograms per liter, and different concentrations of humic acid were added as interfering agents to simulate water bodies with different levels of interference. Three methods were used for detection: Comparative Example A (Traditional Anodic Stripping Voltammetry): Using a conventional glassy carbon electrode mercury membrane, each water sample was directly tested according to the conventional anodic stripping voltammetry procedure.

[0090] Comparative Example B (Conventional Laser-Induced Breakdown Spectroscopy): A conventional laser-induced breakdown spectroscopy system was used to directly detect the liquid level of each water sample.

[0091] Method of this invention embodiment: Using the method described in this invention, the system will automatically select the detection mode according to the degree of interference of the water sample (group A is low interference, group B is medium interference and switches to electrochemical-assisted enhanced spectroscopy mode, group C is high interference and switches to spectral-dominated anti-interference mode).

[0092] Each group of water samples was tested 7 times, and the average value, relative standard deviation and spiked recovery rate were calculated. The results are shown in Table 1.

[0093] Table 1 Comparison of detection results of different methods on spiked water samples

[0094] The data in Table 1 clearly show that under low interference levels, the traditional electrochemical method (Comparative Example A) is still functional, but the traditional spectroscopic method (Comparative Example B) lacks sufficient sensitivity. As interference increases, the traditional electrochemical method fails due to electrode poisoning, and the traditional spectroscopic method experiences a sharp decline in accuracy due to matrix effects. In contrast, the method of this invention, through adaptive mode switching and deep coupling detection strategies, maintains high accuracy, high precision, and high recovery rate under all interference levels, fully demonstrating the inventiveness and significant advancement of this invention.

[0095] This invention encompasses any substitutions, modifications, equivalent methods, and solutions made within the spirit and scope of this invention. To provide the public with a thorough understanding of this invention, specific details are described in detail in the following preferred embodiments; however, those skilled in the art will fully understand the invention even without these details. Furthermore, to avoid unnecessary misunderstanding of the essence of this invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0096] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for on-line monitoring of water pollution by heavy metal ions, characterized in that, Includes the following steps: Step 1: Collect the water sample to be tested, divide the water sample into 3 equal parts and inject them into 3 independent detection chambers. Simultaneously collect the first set of multispectral response vectors, the first set of electrochemical response vectors, matrix conductivity and dissolved organic matter concentration through photochemical colorimetric sensor array, differential pulse stripping voltammetry scanning and conductivity and organic matter analyzer. Construct a multidimensional feature matrix characterizing the interference characteristics of the current water sample. Step 2: Input the multidimensional feature matrix into the pre-trained classification model, calculate a real number between 0 and 1 as the interference level coefficient, and automatically select and switch to one of the following modes for subsequent online monitoring: fast electrochemical mode, electrochemical-assisted enhanced spectroscopy mode, or spectroscopy-dominated anti-interference mode, based on the numerical range of the interference level coefficient. Step 3: If switching to the fast electrochemical mode, a square wave anodic stripping voltammetry scan is performed using a boron-doped diamond working electrode. The ion type is determined based on the peak potential of the stripping peak, and the ion concentration is calculated based on the peak current intensity. Step 4: If switching to the electrochemical-assisted enhanced spectroscopy mode, apply an anodic dissolution pulse voltage after electrochemical enrichment, and simultaneously trigger the laser-induced breakdown spectroscopy system to emit a laser pulse at the instant the anodic dissolution pulse voltage is applied, collect and analyze plasma characteristic spectral lines to calculate ion concentration; Step 5: If the spectrum-dominated anti-interference mode is switched to, the water sample to be tested is adsorbed through a chelating resin enrichment column, then eluted with dilute nitric acid solution, and the eluent is introduced into a microwave plasma torch atomizer to collect and analyze the emission spectrum to calculate the ion concentration.

2. The method for online monitoring of water pollution facing heavy metal ions according to claim 1, characterized in that, The specific process of synchronously acquiring the first set of multispectral response vectors, the first set of electrochemical response vectors, matrix conductivity, and dissolved organic matter concentration in step 1 is as follows: In the first detection chamber, a fiber optic photochemical colorimetric sensor array containing three different heavy metal ion indicators—dithizone sensitive membrane, xylenol orange sensitive membrane, and pyridine azonaphthol sensitive membrane—was inserted. The sensor array was illuminated with white LED light, and a CCD image sensor continuously acquired color change images of the sensor array after reacting with the water sample for 60 seconds. The processor extracted the RGB value changes of each sensitive membrane region before and after the reaction and generated the first set of multispectral response vectors. In the second detection chamber, an electrochemical sensor system consisting of a working electrode made of boron-doped diamond, an Ag / AgCl reference electrode, and a platinum wire counter electrode is inserted. An acetate-sodium acetate buffer solution with a pH of 4.5 is added to the second detection chamber to make the volume ratio of the water sample to the buffer solution 10:

1. Differential pulse stripping voltammetry is used to scan the potential range of -1.3V to +0.5V, and the voltammetric curve is recorded. The peak potential and peak current of each stripping peak are extracted from the voltammetric curve to generate the first set of electrochemical response vectors. In the third detection chamber, the matrix conductivity of the water sample was measured using a conductivity meter, and the concentration of dissolved organic matter in the water sample was measured using a total organic carbon analyzer.

3. The method for online monitoring of water pollution facing heavy metal ions according to claim 2, characterized in that, The specific processes for constructing the multidimensional feature matrix representing the current water sample interference characteristics in step 1 and calculating a real number ranging from 0 to 1 as the interference level coefficient in step 2 are as follows: The processor merges the first set of multispectral response vectors, the peak potentials and peak currents in the first set of electrochemical response vectors, the matrix conductivity, and the concentration of dissolved organic matter to construct a multidimensional feature matrix. The multidimensional feature matrix is ​​input into a pre-trained support vector machine classification model, which outputs an interference level coefficient. The interference level coefficient is a real number between 0 and 1. The larger the value of the interference level coefficient, the more severe the interference caused by the coexistence of organic matter or multiple ions in the water sample.

4. The method for online monitoring of water pollution facing heavy metal ions according to claim 1, characterized in that, The specific process described in step 3, which involves using a boron-doped diamond working electrode to perform square-wave anodic stripping voltammetry scanning, determining the ion species based on the peak potential of the stripping peak, and calculating the ion concentration based on the peak current intensity, is as follows: Step 3.1: Continuously inject fresh water sample into the flow cell and stir. Apply a constant potential of -1.0V to the working electrode for enrichment. The enrichment time is 300 seconds, so that the heavy metal ions in the water sample are reduced and enriched on the surface of the boron-doped diamond working electrode. Step 3.2: Stop stirring and injection, let stand for 15 seconds, then use square wave anodic stripping voltammetry to scan from -1.0V to +0.5V at a scan rate of 50mV / s and record the stripping voltammetric curve; Step 3.3: By analyzing the dissolution peaks appearing at different potentials on the dissolution voltammetry curve, and comparing them with the built-in standard ion dissolution potential library, the ion types are determined; Step 3.4: Based on the peak current intensity of each dissolution peak, substitute it into the pre-stored standard curve equation of the ion on the boron-doped diamond working electrode to calculate the real-time concentration of each heavy metal ion in the water sample to be tested.

5. The method for on-line monitoring of water pollution facing heavy metal ions according to claim 1, characterized in that, The specific process described in step 4, which involves applying an anodic dissolution pulse voltage after electrochemical enrichment and simultaneously triggering the laser-induced breakdown spectroscopy system to emit a laser pulse at the instant the anodic dissolution pulse voltage is applied, is as follows: Step 4.1: Collect a fresh water sample and inject it into the three-electrode flow cell. The bottom of the three-electrode flow cell is equipped with a high-purity graphite plate as the working electrode, and an optical quartz glass window is provided on the opposite side. Step 4.2: Add supporting electrolyte to the three-electrode flow cell to make the final concentration 0.1 mol / L, apply a constant potential of -1.2 V to the working electrode for enrichment, enrichment time is 600 seconds, and at the same time circulate the water sample at a flow rate of 1 mL / min using a micro peristaltic pump. Step 4.3: After enrichment is completed, stop the sample injection and empty the liquid in the three-electrode flow cell, and then fill it with high-purity argon as a protective gas; Step 4.4: Start the Q-switched laser of the laser-induced breakdown spectroscopy system. At the instant a anodic dissolution pulse voltage with an amplitude of +1.5V and a pulse width of 50ms is applied, the Q-switched laser is simultaneously triggered to emit a laser pulse with an energy of 80mJ and a wavelength of 1064nm. The laser pulse is focused on the surface of the high-purity graphite plate through the optical quartz glass window. Step 4.5: The plasma light signal generated by the laser breaking down the atomic cloud is collected by the fiber optic probe, and then dispersed and detected by the spectrometer under the conditions of detection wavelength 200nm to 800nm ​​and resolution 0.1nm. The processor collects the spectral data and analyzes the intensity of the characteristic spectral lines of heavy metal elements.

6. The method for online monitoring of water pollution facing heavy metal ions according to claim 5, characterized in that, The synchronous triggering mentioned in step 4.4 refers to controlling the synchronous delay time between the pulse signal generated by the high-voltage power supply and the laser pulse emitted by the Q-switched laser to within 1 microsecond, so that the laser pulse can accurately act on the heavy metal atomic cloud that has just been released from the surface of the high-purity graphite plate by the anodic dissolution pulse.

7. The method for online monitoring of water pollution facing heavy metal ions according to claim 1, characterized in that, The specific process described in step 5, which involves adsorbing the water sample to be tested through a chelating resin enrichment column, eluting it with dilute nitric acid solution, and then introducing the eluent into a microwave plasma torch atomizer, is as follows: Step 5.1: Collect fresh water samples to be tested, adjust the pH value of the water samples to 5.5, and then pass them through a micro-enrichment column containing chelating resin at a flow rate of 2 mL / min, so that the heavy metal ions in the water samples are adsorbed by the chelating resin. Step 5.2: After enrichment is completed, the micro-enrichment column is rapidly flushed in the reverse direction with 3 mL of 1 mol / L dilute nitric acid solution to elute the adsorbed heavy metal ions. Step 5.3: The eluent is directly introduced into the atomizer of the microwave plasma torch, and after being atomized by high-purity nitrogen, it forms an aerosol that enters the center of the plasma torch flame; Step 5.4: Start the microwave source and set the power to 1000W. After the heavy metal ions are atomized and excited in the high-temperature plasma torch, they emit characteristic spectra. The emission spectra are collected in the 200nm to 500nm band by a fiber optic spectrometer. Step 5.5: The processor performs peak finding and background subtraction on the emission spectrum, and calculates the concentration of heavy metal ions in the original water sample based on the emission spectral intensity of each element at a specific wavelength, combined with the enrichment factor of online elution and the standard curve.

8. The online monitoring method for water pollution targeting heavy metal ions according to claim 1, characterized in that, The system also includes step 6: transmitting the heavy metal ion types and concentration data calculated in steps 3, 4, or 5 to the data platform of the remote monitoring center in real time via a 4G wireless transmission module; simultaneously, after every 24 monitoring cycles, the system automatically extracts a standard mixed solution containing 50 μg / L each of Pb²⁺ and Cd²⁺, repeats the complete process from steps 1 to 5, compares the detected values ​​with the standard values ​​to calculate the drift coefficient, and automatically corrects the standard curve equation parameters in steps 3, 4, and 5 to eliminate zero-point drift and sensitivity drift caused by long-term system operation.

9. The method for online monitoring of water pollution facing heavy metal ions according to claim 1, characterized in that, The specific judgment rule for automatically selecting and switching to the corresponding mode based on the numerical range of the interference level coefficient in step 2 is as follows: if the interference level coefficient is less than or equal to 0.3, the current water sample is determined to be a low-interference clean water body and switched to the rapid electrochemical mode; if the interference level coefficient is greater than 0.3 and less than 0.7, the current water sample is determined to be a moderately interfered complex water body and switched to the electrochemical-assisted enhanced spectral mode; if the interference level coefficient is greater than or equal to 0.7, the current water sample is determined to be a highly interfered polluted water body and switched to the spectral-dominated anti-interference mode.

10. The method for online monitoring of water pollution targeting heavy metal ions according to claim 1, characterized in that, The specific process of collecting the water sample to be tested in step 1 is as follows: the water sample to be tested is collected from the water area to be monitored by an automatic sampling pump, and the collected water sample is divided into three equal parts of equal volume, which are injected into the first detection chamber, the second detection chamber and the third detection chamber of the preparatory detection unit, which are independent of each other and not connected to each other.

Images