Method for Predicting the Reaction Rate Constant of Chlorine Oxygen Free Radicals with Dissolved Organic Matter in Water

Abstract

The present invention provides a method for calculating the reaction rate constant of hydroxyl radical with dissolved organic matter in water, which can quickly and accurately calculate the reaction rate constant of hydroxyl radical with dissolved organic matter in water. A method for predicting the reaction rate constant of hydroxyl radical with dissolved organic matter in water by using a quantitative structure-activity relationship model is also disclosed. Each process of model establishment and verification in the present invention strictly complies with the OECD model construction and use guidelines. The obtained model has a high goodness of fit, good robustness and strong predictive ability. The instruments required for the determination of descriptors involved in the prediction model established in the present invention are mainly ultraviolet-visible spectrophotometer, TOC analyzer, three-dimensional fluorescence spectroscopy, etc. These instruments have been widely used to characterize DOM due to their advantages of high sensitivity, easy processing and analysis, and low cost. The selected descriptors have strong mechanistic interpretability. The obtained model has a simple form, good transparency and is easy to be programmed and popularized.

Description

Technical Field

[0001] The present invention relates to a method for predicting the reaction rate constant of chlorine oxygen free radicals of dissolved organic matter in water by establishing a quantitative structure-activity relationship model (QSAR), belonging to the field of ecological risk assessment test strategies. Background Art

[0002] With the rapid development of urbanization and industrialization, a large number of pollutants enter natural water bodies with the discharge of urban sewage, causing serious water pollution problems. Emerging pollutants including pharmaceuticals and personal care products (PPCPs) have been widely detected in the water environment of our country, and the risks to the ecosystem and human health cannot be ignored. Some PPCPs have developed resistance to traditional sewage and drinking water treatment methods such as coagulation, filtration, sedimentation, and separate ultraviolet disinfection, chlorination disinfection, and ozonation disinfection. Advanced oxidation processes (AOPs) can generate strongly oxidizing free radical species, such as hydroxyl radicals ( · OH), reactive chlorine species such as chlorine radicals (Cl · ), dichlorine radicals (Cl2 ·- ), and chlorine oxygen radicals (ClO · ) to effectively remove these emerging pollutants. Most current studies focus on AOPs based on · OH. In recent years, it has been found that reactive chlorine species in the advanced oxidation technology system are also strong oxidants (the redox potentials of Cl · , Cl2 ·- , and ClO · are 2.47V, 2.0V, and 1.5 - 1.8V respectively), and compared with · OH, reactive chlorine species are more selective and contribute to targeted reactions with emerging pollutants. It is worth noting that ClO · has attracted more and more attention in advanced oxidation technologies because its concentration in the chlorine photolysis and electrochemical systems is several orders of magnitude higher than that of · OH and Cl · , reaching up to ~10 -10 M, and it has a relatively high reaction rate with some PPCPs (~10 7 to ~10 8 M -1 s -1 ).

[0003] However, dissolved organic matter (DOM), which widely exists in water matrices, has a quenching effect on free radicals in AOPs and acts as the main free radical quencher, severely restricting the efficiency of AOPs in removing emerging pollutants. The second-order reaction rate between DOM and free radicals largely determines the quenching effect of DOM on free radicals, and the second-order reaction rate between DOM and free radicals mainly depends on the molecular structure of DOM. Therefore, revealing the relationship between the molecular composition of DOM and its second-order rate constant with ClO · is helpful for selecting appropriate process parameters (such as the dosage of oxidants, reaction time control, etc.) in water treatment. In addition, as a class of heterogeneous organic matter with complex composition, DOM also participates in many important reactions in water bodies. For example, as a precursor of disinfection by-products, by measuring the second-order reaction rate constant between DOM and ClO · , it is helpful to evaluate the ecological risk of DOM in the environment. There are many methods for characterizing the molecular structure of DOM. Currently, ultraviolet-visible spectrophotometers, three-dimensional fluorescence spectra, liquid chromatography, etc. are commonly used to measure the apparent parameters of DOM to characterize its chemical structure, while the second-order reaction rate (kClO·) between DOM and ClO · is mainly measured by methods such as competitive kinetics. Although these methods can obtain relatively accurate results, the measurement cost is high and it consumes a lot of manpower and material resources. In addition, there are few reports on the k ClO· of DOM at present, far from meeting the requirements of the engineering parameters of advanced oxidation technologies based on ClO · , and there are still large gaps in research. Facing DOM with wide sources and large differences in composition, solely relying on experimental measurement of k ClO· requires a large amount of time, manpower, and financial resources. Therefore, it is necessary to develop a scientific, efficient, and fast prediction method.

[0004] Quantitative Structure-Activity Relationship model (QSAR) is based on molecular structure properties and uses mathematical, statistical, computational chemistry, etc. methods to quantitatively predict the physical and chemical properties, environmental behavior, or toxicological parameters of compounds (collectively referred to as activities). It has the advantages of making up for the lack of test data and reducing test costs, so it has been widely used in related fields of environmental science. Based on the characteristics and advantages of QSAR, QSAR can be used to predict the quantitative relationship between the second-order reaction rate constant of DOM and free radicals, thereby reducing test costs, shortening test time, making up for the lack of experimental data, and providing strong data support for evaluating the environmental risk of DOM.

[0005] However, no scholars have established a QSAR model specifically for the reaction between DOM and ClO ·QSAR model for the second-order reaction rate of DOM with ClO · The data on the second-order reaction rate constants of · with DOM are severely insufficient, and basically no literature has reported on it and established a QSAR model. Although the literature "Environ. Sci. Technol. 2017, 51, 10431 - 10439." and the literature "Water Research 147 (2018) 184e194." reported the second-order reaction rates of ClO· with various PPCPs and DOM, and evaluated the influence of DOM on the degradation of PPCPs by advanced oxidation technologies based on the ClO · system, and established a QSAR model between ClO · and various PPCPs, while the QSAR model between DOM remains to be developed.

[0006] In summary, the advanced oxidation technology based on ClO · has good development prospects in pollutant removal, but the high quenching effect of DOM on it severely restricts the efficiency of this technology in removing pollutants. At present, the lack of k ClO· values is still very serious, and the experimental determination of this parameter is time-consuming and laborious, difficult to carry out in large quantities, and the QSAR model for DOM and ClO · has not been developed yet. Based on the above research status, there is an urgent need to develop a QSAR model with a wide application domain covering a large number of DOM with rich structural types, easy-to-obtain descriptors, clear algorithms, high transparency, easy mechanistic interpretation, and convenient application and promotion. Summary of the Invention

[0007] The present invention proposes a method for predicting the reaction rate constant of chlorine oxygen radicals of dissolved organic matter in water using a quantitative structure-activity relationship model, and the present invention also proposes a calculation method for the reaction rate constant of chlorine oxygen radicals of dissolved organic matter in water, which solves the problems existing in the prior art.

[0008] The purpose of the present invention is to develop a scientific, efficient, widely applicable, and clearly mechanistically interpretable QSAR model for DOM and ClO · This method can predict its k ClO· according to the chemical composition of DOM, so as to evaluate the quenching effect of DOM on free radicals, and further provide a reliable basis for the optimization of advanced oxidation process parameters. During the modeling process, referring to the OECD guidelines for the construction and use of QSAR models, internal and external validations were carried out to examine the predictive ability and robustness of the model, and the application domain of the model was characterized.

[0009] A calculation method for the reaction rate constant of chlorine oxygen radicals of dissolved organic matter in water is: lgk ClO· = 8.762 - 4.77×10-4 ×TOC + 0.018×SUVA 254 -0.017×E2 / E3 - 6.733×S 275-295 -0.222×S 350-400 -0.131×S R -0.0062×FI + 0.219×HI - 0.100×BIX + 4.42×10 -5 ×M w + 0.049×TAC; where TOC represents the total organic carbon concentration, UV 254 represents the absorbance at a wavelength of 254 nm, UV 365 represents the absorbance at a wavelength of 365 nm, UV 465 represents the absorbance at a wavelength of 465 nm, UV 665 represents the absorbance at a wavelength of 665 nm, SUVA 254 is the ratio of UV 254 / TOC, E2 / E3 is the ratio of UV 254 / UV 365 of the ratio, S 275-295 represents the slope of the curve obtained by taking the natural logarithm of the ultraviolet spectral absorbance in the wavelength range of 275 - 295 nm and dividing it by the corresponding absorbance, S 350-400 represents the slope of the curve obtained by taking the natural logarithm of the ultraviolet spectral absorbance in the wavelength range of 350 - 400 nm and dividing it by the corresponding absorbance, S R is the ratio of S 275-295 / S 350-400 of the ratio, FI represents the fluorescence index, HI represents the humification index, BIX represents the autochthonous index, M w represents the weight - average molecular weight, and TAC represents the total antioxidant capacity.

[0010] Furthermore, a method for predicting the reaction rate constant of chlorine - oxygen free radicals of dissolved organic matter in water using a quantitative structure - activity relationship model includes the following steps:

[0011] Step 1: Extract DOM;

[0012] Step 2: Determine the structural composition data of DOM and use them as descriptors of the DOM chemical structure;

[0013] Step 3: Determine the k ClO· value and perform logarithmic processing on the obtained data to get lgk ClO· ;

[0014] Step 4: Randomly divide the collected and measured DOM experimental data into a training set and a validation set. The experimental data in the training set is greater than that in the validation set. The DOM experimental data in the training set is used to build the model, and the DOM experimental data in the validation set is used to evaluate the prediction ability of the model;

[0015] Step 5: Correlation analysis: Perform Pearson correlation analysis on lgk ClO· and the descriptors of DOM to test the statistical relationship between them and the potential collinearity between the descriptors. Eliminate the descriptors that are not relevant to lgk ClO· For descriptors with high correlation, perform principal component analysis;

[0016] Step 6: Principal component analysis: Perform PCA dimensionality reduction on the collinear descriptors obtained in Step 5, screen out the principal components with eigenvalue > 1 and high interpretation of the total variation degree, and perform MLR on the obtained principal components;

[0017] Step 7: Multiple stepwise regression analysis: Perform MLR on the descriptors with clear physical meaning and good description effect obtained in Step 6 and lgk ClO· in the training set to further remove the multicollinearity of the descriptors and construct a QSAR model to obtain the calculation formula of lgk ClO· and the descriptors of the DOM chemical structure;

[0018] Step 8: Model validation: Perform internal and external validation on the obtained QSAR model; Internal validation: For the data in the training set, use the leave-one-out cross-validation coefficient Q 2 LOO and the Bootstrapping method cross-validation coefficient Q 2 BOOT to characterize the robustness of the model; External validation: For the data in the validation set, use the adjusted coefficient R 2 ext of the validation set, the cross-validation coefficient Q 2 ext of the validation set, and the root mean square error RMSE ext of the validation set to characterize the model prediction ability; The predicted values and the experimental values fit well, indicating that the model has good prediction ability and can be successfully applied to DOM outside the training set; It shows that the model has good fitting ability, robustness and prediction ability;

[0019] Step 9: Application domain evaluation: Characterize the application domain of the model using the Williams plot method. All DOM values are within the warning values, indicating that the QSAR model developed in the present invention has a good application domain.

[0020] Further, extract DOM: Using the method of predicting the reaction rate constant of chlorine oxygen free radicals of dissolved organic matter in water by quantitative structure-activity relationship model, extract DOM of samples from different places and different types according to the method of Thorsten Dittmar et al. (literature "Limnology and Oceanography - Methods 2008, 6, 230 - 235."), and place it in a 4°C refrigerator for storage before subsequent analysis to ensure the stable storage of DOM samples in a short time.

[0021] Further, step 2: Measure the pH value of DOM with a pH meter; measure the total organic carbon concentration of DOM with a TOC analyzer; measure the UV 254 , SUVA 254 (UV 254 / TOC value), E2 / E3 (UV 254 / UV 365 value), E4 / E6 (UV 465 / UV 665 value), S 275-295 , S 350-400 , S R (S 275-295 / S 350-400 value); measure FI, HI, RI, BIX, Fmax(C1), Fmax(C2), Fmax(C3) and Fmax(C4) of DOM by three-dimensional fluorescence spectroscopy, and measure M w , PDI, RC C18 , RC NH2 ; measure the TAC of DOM with a reagent; the reagent includes Folin-Ciocalteu reagent and sodium bicarbonate solution; 22 items of DOM structural composition data can be obtained by the above methods and used as descriptors of the DOM chemical structure; among them, TOC represents the total organic carbon concentration, UV 254 represents the absorbance at a wavelength of 254 nm, UV 365 represents the absorbance at a wavelength of 365 nm, UV 465 represents the absorbance at a wavelength of 465 nm, UV 665 represents the absorbance at a wavelength of 665 nm, SUVA 254 is the ratio of UV 254 / TOC, E2 / E3 is the ratio of UV 254 / UV 365 ratio, E4 / E6 is the ratio of UV 465 / UV 665 ratio, S 275-295It represents the slope S of the curve obtained by taking the natural logarithm of the ultraviolet spectral absorbance in the wavelength range of 275 - 295 nm and dividing it by the corresponding absorbance. 350-400 It represents the slope of the curve obtained by taking the natural logarithm of the ultraviolet spectral absorbance in the wavelength range of 350 - 400 nm and dividing it by the corresponding absorbance. FI represents the fluorescence index, HI represents the humification index, RI represents the reduction index, BIX represents the autochthonous index, and M w It represents the weight-average molecular weight, and PDI represents the dispersity coefficient, RC C18 It represents C 18 The retention coefficient of the solid-phase extraction cartridge (hydrophobicity index), RC NH2 It represents the retention coefficient of the NH2 solid-phase extraction cartridge (hydrophilicity / anionic property index), TAC represents the total antioxidant capacity, and S R It is S 275-295 / S 350-400 The ratio of, F max (C1) represents the proportion of the maximum fluorescence intensity of the protein-like component C1 when the excitation wavelength is 278 nm and the emission wavelength is 318 nm; F max (C2) represents the proportion of the maximum fluorescence intensity of the humic acid-like component C2 when the excitation wavelength is 296 nm and the emission wavelength is 376 nm; F max (C3) represents the proportion of the maximum fluorescence intensity of the humic acid-like component C3 when the excitation wavelength is 332 nm and the emission wavelength is 406 nm; F max (C4) represents the proportion of the maximum fluorescence intensity of the humic acid-like component C4 when the excitation wavelength is 270 (or 362) nm and the emission wavelength is 462 nm; C1 represents the protein-like component when the excitation wavelength is 278 nm and the emission wavelength is 318 nm; C2 represents the humic acid-like component when the excitation wavelength is 296 nm and the emission wavelength is 376 nm; C3 represents the humic acid-like component when the excitation wavelength is 332 nm and the emission wavelength is 406 nm; C4 represents the humic acid-like component when the excitation wavelength is 270 or 362 nm and the emission wavelength is 462 nm.

[0022] Furthermore, step 4: Randomly divide the at least 103 kinds of DOM experimental data collected and measured into a training set and a validation set according to a ratio of 3:1.

[0023] Furthermore, step 5: Correlation analysis: Perform Pearson correlation analysis on lgk ClO· and 22 descriptors to test the statistical relationship between them and the potential collinearity between the descriptors; Eliminate the descriptors that are not relevant to lgk ClO· (correlation coefficient r < 0.5). For the 11 descriptors with high correlation, perform principal component analysis. The 11 descriptors are TOC, SUVA 254 、E2 / E3、S275-295 、S 350-400 、S R 、FI, HI, BIX, M w 、TAC。

[0024] Furthermore, step 6: Principal component analysis: Perform PCA dimensionality reduction on the 11 descriptors with collinearity obtained in step 5, and screen out 2 principal components F1 and F2 with eigenvalues > 1 and high interpretation of the total variation degree (i.e., not less than 86.057%). F1 includes SUVA 254 、E2 / E3、SR、FI、HI、BIX、M W 、TAC), F2 includes TOC, S 275-295 、S 350-400 , and perform MLR on these 2 obtained principal components.

[0025] Furthermore, step 7: Multiple stepwise regression analysis: Perform MLR on the descriptors with clear physical meaning and good description effect obtained in step 6 and lgk ClO· in the training set to further remove the multicollinearity of the descriptors and construct a QSAR model with adjusted coefficient R 2 adj = 0.917, test value F = 433.460, significance p < 0.05, variance inflation factor VIF < 10; obtain the calculation formula of lgk ClO· and the descriptors of the DOM chemical structure as follows:

[0026] lgk ClO· = 8.762 - 4.77×10 -4 ×TOC + 0.018×SUVA 254 - 0.017×E2 / E3 - 6.733

[0027] ×S 275-295 - 0.222×S 350-400 - 0.131×S R - 0.0062×FI + 0.219×HI - 0.100×BIX + 4.42×10 -5 ×M w + 0.049×TAC

[0028] In the formula, TOC represents the total organic carbon concentration of DOM, which can reflect the total amount of carbon contained in DOM; SUVA 254 、E2 / E3、HI all reflect the humification degree of DOM; S 275-295 、S 350-400 and S Ris used to characterize the aromatic carbon content and aromatization degree of DOM; FI is used to reveal the source of DOM, such as terrigenous or microbial source; BIX reflects the strength of the autochthonous contribution of DOM; M w reflects the molecular weight of DOM and TAC reflects the selectivity of DOM and the reaction with the antioxidant part. It can be seen from this formula that ClO · is prone to react with aromatic compounds containing electron-rich parts (such as phenols, anilines, alkoxybenzenes, etc.). When DOM contains more electron-donating structures, its second-order reaction rate will be higher.

[0029] Further, step 8: Model validation: Internal and external validation are carried out on the obtained QSAR model; Internal validation: For the data in the training set, the leave-one-out cross-validation coefficient Q 2 LOO and the Bootstrapping method cross-validation coefficient Q 2 BOOT are used to characterize the robustness of the model; External validation: For the data in the validation set, the adjusted R of the validation set 2 (R 2 ext ), the cross-validation coefficient of the validation set (Q 2 ext ) and the root mean square error of the validation set (RMSE ext ) are used to characterize the prediction ability of the model. The fitting relationship between the predicted value and the experimental value is shown in the appendix Figure 1 . It can be seen from the figure that the predicted value and the experimental value fit well, indicating that the model has good prediction ability and can be successfully applied to DOM outside the training set.

[0030] The verification results are: Q 2 LOO = 0.906, Q 2 BOOT = 0.901, RMSE train = 0.077,

[0031] R 2 ext = 0.913, Q 2 ext = 0.911, RMSE ext = 0.064.

[0032] It shows that the model has good fitting ability, robustness and prediction ability.

[0033] (2) Technical effects of the present invention

[0034] The beneficial effects of the present invention are as follows: The calculation method of the reaction rate constant of the chloro-oxygen free radical of dissolved organic matter in water according to the present invention can quickly and accurately calculate the reaction rate constant of the chloro-oxygen free radical of dissolved organic matter in water. By using the present invention, the k of DOM can be scientifically and quickly predicted by measuring the structural characteristics of DOM. ClO· This method has a wide range of application domains, strong mechanism interpretability, and is simple, fast, and low-cost, saving the manpower, material resources, and financial resources required for experimental determination. The beneficial effects of the present invention are specifically manifested as follows:

[0035] 1) The DOM range that the model can predict is relatively wide, basically covering DOM in various water bodies, and filling the gap that there is no QSAR model for the second-order reaction rate of DOM and chloro-oxygen free radicals in current research. For the prediction of the k value of these DOMs, it will help to select appropriate advanced oxidation processes and optimize parameters in sewage treatment, and also help to evaluate the ecological risks of DOM in the environment. ClO· Value prediction will contribute to the selection of appropriate advanced oxidation processes and parameter optimization in sewage treatment, and also help to evaluate the ecological risks of DOM in the environment.

[0036] 2) The instruments required for measuring the descriptors involved in the prediction model established by the present invention are mainly ultraviolet-visible spectrophotometers, TOC analyzers, three-dimensional fluorescence spectra, etc. These instruments have been widely used to describe the characteristics of DOM due to their high sensitivity, ease of processing and analysis, and low cost. The selected descriptors have strong mechanism interpretability, the obtained model has a simple form, good transparency, and is easy to be promoted and applied programmatically.

[0037] 3) Each process of the model establishment and verification obtained by the present invention strictly complies with the OECD model construction and usage guidelines. The obtained model has a high goodness of fit, good robustness, and strong prediction ability.

[0038] 3) Each process of establishing and validating the model obtained in the present invention strictly adheres to the OECD model construction and usage guidelines. The obtained model has a high goodness of fit, good robustness, and strong prediction ability.

[0039] 3) Each process of establishing and validating the model obtained in the present invention strictly adheres to the OECD model construction and usage guidelines. The obtained model has a high goodness of fit, good robustness, and strong prediction ability. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0041] Figure 1 It is a fitting diagram of the measured values and predicted values of lgk for the QSAR model training set and validation set. There are 79 kinds of DOM in the training set and 24 kinds of compounds in the validation set; ClO· It is a fitting diagram of the measured values and predicted values of lgk for the QSAR model training set and validation set. There are 79 kinds of DOM in the training set and 24 kinds of compounds in the validation set;

[0042] Figure 2 It is a Williams diagram of the QSAR model. Detailed implementation mode

[0043] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts belong to the scope of protection of the present invention.

[0044] Example 1: Predict the k value of DOM in a certain lake ClO· Value

[0045] 1) Obtaining DOM in the water sample: The water sample was collected from a certain lake water sample, and DOM was extracted.

[0046] 2) Measuring the descriptors of DOM, including measuring TOC with a TOC total organic carbon analyzer

[0047] (mg C / L),

[0048] measuring SUVA with an ultraviolet-visible spectrophotometer 254 (Lmg -1 m -1 ), E2 / E3, S 275-295 , S 350-400 , S R , measuring FI, HI, BIX with three-dimensional fluorescence spectroscopy, measuring M w (Da) and TAC (mgGA / mgC).

[0049] 3) Model verification and calculation results: h = 0.451 < h*, so this DOM is within the application domain. Calculated by the model

[0050] as follows:

[0051] lgk ClO· = 8.762 - 4.77×10 -4 ×150.7 + 0.018×SUVA 254 - 0.017×

[0052] E2 / E3 - 6.733×

[0053] S 275-295 - 0.222×S 350-400 - 0.131×S R - 0.0062×FI + 0.219×HI - 0.100×

[0054] BIX + 4.42×10 -5 ×M w + 0.049×TAC

[0055] = 8.762 - 4.77×10 -4 ×150.75 + 0.018×1.24 - 0.017×7.75 - 6.733×

[0056] 0.01 - 0.222×

[0057] 0.02 - 0.131×0.71 - 0.0062×2.01 + 0.219×0.76 - 0.100×0.92 + 4.42×10 -5

[0058] ×2194 + 0.049×0.18 = 8.44 (M -1 s -1 ), the measured result is 7.99, and the prediction result is good.

[0059] Example 2 predicts the k of a certain natural organic matter (NOM) ClO· value

[0060] 1) Obtaining DOM: Purchased from IHSS and DOM was extracted.

[0061] 2) Measuring the descriptors of DOM, including measuring TOC with a TOC total organic carbon analyzer

[0062] (mg C / L), measuring SUVA with an ultraviolet-visible spectrophotometer 254 (L mg -1 m -1 ), E2 / E3, S 275-295 , S 350-400 , S R , measuring FI, HI, BIX with three-dimensional fluorescence spectroscopy, measuring M w (Da) and TAC (mgGA / mgC).

[0063] 3) Model verification and calculation results: h = 0.236 < h*, so this DOM is within the application domain, and calculated by the model

[0064] as follows:

[0065] lgk ClO· = 8.762 - 4.77×10 -4 ×TOC + 0.018×SUVA 254 - 0.017×E2 / E3 - 6.733

[0066] ×

[0067] S 275-295 - 0.222×S 350-400 - 0.131×S R-0.0062×FI + 0.219×HI - 0.100×

[0068] BIX + 4.42×10 -5 ×M w + 0.049×TAC

[0069] =8.762 - 4.77×10 -4 ×212.35 + 0.018×5.4 - 0.017×4.59 - 6.733×

[0070] 0.01 - 0.222×

[0071] 0.02 - 0.131×0.63 - 0.0062×0.96 + 0.219×0.86 - 0.100×1.15 + 4.42×10 -5

[0072] ×1219.68 + 0.049×0.19=8.57(M -1 s -1 ),The measured result is 8.73, and the prediction result is good.

[0073] Example 3 predicts the k ClO·– value

[0074] 1) Obtaining DOM of the water sample: The water sample was collected from the secondary effluent of a sewage treatment plant, and DOM was extracted.

[0075] 2) Measuring the descriptors of DOM, including measuring TOC

[0076] (mg C / L) with a TOC total organic carbon analyzer, and measuring SUVA 254 (L mg -1 m -1 )、E2 / E3、S 275-295 、S 350-400 、S R with an ultraviolet-visible spectrophotometer, measuring FI, HI, BIX with three-dimensional fluorescence spectroscopy, and measuring M w (Da) and TAC (mgGA / mgC).

[0077] 3) Model verification and calculation results: h = 0.202 < h*, so this DOM is within the application domain, and calculated by the model

[0078] as follows:

[0079] lgk ClO· =8.762 - 4.77×10 -4 ×TOC + 0.018×SUVA 254-0.017×E2 / E3 - 6.733

[0080] ×

[0081] S 275-295 -0.222×S 350-400 -0.131×S R -0.0062×FI + 0.219×HI - 0.100×

[0082] BIX + 4.42×10 -5 ×M w +0.049×TAC

[0083] =8.762 - 4.77×10 -4 ×104.01 + 0.018×4.11 - 0.017×4.08 - 6.733×

[0084] 0.02 - 0.222×

[0085] 0.02 - 0.131×0.68 - 0.0062×1.68 + 0.219×0.63 - 0.100×1.08 + 4.42×10 -5

[0086] ×650.43 + 0.049×0.38=8.49(M -1 s -1 ),The measured result is 8.57, and the predicted result is good.

[0087] Example 4 predicts the k ClO·– value

[0088] 1) Obtaining DOM of the water sample: The landfill leachate was collected from a certain landfill site, and DOM was extracted.

[0089] DOM.

[0090] 2) Measuring the descriptors of DOM, including measuring TOC (mg

[0091] / L) with a TOC total organic carbon analyzer, measuring SUVA C (Lmg 254 m -1 )、E2 / E3、S -1 、E2 / E3、S 275-295 、S 350-400 、S R ,measuring FI, HI, BIX with three-dimensional fluorescence spectroscopy, and measuring M w (Da) and TAC (mgGA / mgC).

[0092] 3) Model verification and calculation results: h = 0.496 < h*, so this DOM is within the application domain. Calculated by the model

[0093] as follows:

[0094] lgk ClO· = 8.762 - 4.77×10 -4 ×TOC + 0.018×SUVA 254 - 0.017×E2 / E3 - 6.733

[0095] ×

[0096] S 275-295 - 0.222×S 350-400 - 0.131×S R - 0.0062×FI + 0.219×HI - 0.100×

[0097] BIX + 4.42×10 -5 ×M w + 0.049×TAC

[0098] = 8.762 - 4.77×10 -4 ×312.88 + 0.018×2.65 - 0.017×4.34 - 6.733×

[0099] 0.01 - 0.222×

[0100] 0.02 - 0.131×0.71 - 0.0062×2.36 + 0.219×0.93 - 0.100×0.74 + 4.42×10 -5

[0101] ×1998.81 + 0.049×0.19 = 8.48 (M -1 s -1 ), and the measured result is 8.26, with good prediction results.

[0102] Example 5 predicts the k ClO·– value

[0103] 1) Obtaining the DOM of the water sample: Purchased from a certain reagent company and the DOM was extracted.

[0104] 2) Measuring the descriptors of the DOM, including measuring TOC (mg

[0105] / L) with a total organic carbon analyzer for TOC, and measuring SUVA (Lmg C ), E2 / E3, S 254 (Lmg -1 m -1 ) with an ultraviolet-visible spectrophotometer 275-295, S 350-400 , S R , FI, HI, and BIX were measured using three-dimensional fluorescence spectroscopy, and M w (Da) and TAC (mgGA / mgC) were measured.

[0106] 3) Model verification and result calculation: h = 0.304 < h*, so this DOM is within the application domain. The model

[0107] calculates as follows:

[0108] lgk ClO· = 8.762 - 4.77×10 -4 ×TOC + 0.018×SUVA 254 - 0.017×E2 / E3 - 6.733

[0109] ×

[0110] S 275-295 - 0.222×S 350-400 - 0.131×S R - 0.0062×FI + 0.219×HI - 0.100×

[0111] BIX + 4.42×10 -5 ×M w + 0.049×TAC

[0112] = 8.762 - 4.77×10 -4 ×78.7 + 0.018×6.51 - 0.017×2.33 - 6.733×

[0113] 0.02 - 0.222×

[0114] 0.02 - 0.131×0.83 - 0.0062×0.84 + 0.219×0.87 - 0.100×1.01 + 4.42×10 -5

[0115] ×3089 + 0.049×0.26 = (M -1 s -1 ), and the measured result is 8.74, with good prediction results.

[0116] Example 6

[0117] By consulting a large number of literatures, chemical composition data of DOM and k ClO·, if there are multiple measurement data, take their average value, and extract the compositional characteristics of DOM including standard DOM purchased from the International Humic Substances Society (IHSS) and DOM from secondary effluents of sewage treatment plants across the country, surface water DOM (such as reservoir DOM, river and lake DOM, etc.), humic acid, fulvic acid, landfill leachate DOM, etc., and use the competitive kinetics method to determine its k ClO· . A total of 103 DOM-related data were obtained in this example, and a QSAR model with a wide coverage and many types of DOM was established. To make the QSAR model robust, reliable, and predictive, this model was constructed by the sequential method: first, perform a correlation analysis on k ClO· and the selected descriptors, discard the descriptors that are not relevant to k ClO· , and perform principal component analysis (PCA) on the descriptors with collinearity to reduce the analysis dimension. Subsequently, use multiple stepwise regression analysis (MLR) to regress k ClO· and the descriptors to construct a QSAR model, verify the obtained QSAR model, and finally evaluate its application domain. Model verification: The obtained QSAR model is verified internally and externally; Internal verification: For the data in the training set, use the leave-one-out cross-validation coefficient Q 2 LOO and the Bootstrapping cross-validation coefficient Q 2 BOOT to characterize the robustness of the model; External verification: For the data in the validation set, use the adjusted coefficient R 2 ext , the cross-validation coefficient Q 2 ext of the validation set, and the root mean square error RMSE ext of the validation set to characterize the prediction ability of the model; The fitting relationship between the predicted value and the experimental value is shown in Appendix Figure 1 . It can be seen from the figure that the predicted value and the experimental value fit well, indicating that the model has good prediction ability and can be successfully applied to DOM outside the training set. The verification results are: Q 2 LOO = 0.906, Q 2 BOOT = 0.901, RMSE train = 0.077, R 2 ext = 0.913, Q 2 ext = 0.911, RMSE ext = 0.064. It shows that the model has good fitting ability, robustness, and prediction ability. Application domain evaluation. Use the Williams plot method to characterize the application domain of the model (see Figure 2) As can be seen from the figure, all DOM values are within the warning values, indicating that the QSAR model developed in the present invention has a good application domain. The experimental data of 103 DOMs are shown in the following table:

[0118] Table 1

[0119]

[0120]

[0121]

[0122]

[0123]

[0124]

[0125] Table 2

[0126]

[0127]

[0128]

[0129] The above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for calculating the reaction rate constant of hydroxyl radical with dissolved organic matter in water, characterized in that: lgk ClO· = 8.762 - 4.77×10 -4 ×TOC + 0.018×SUVA 254 - 0.017×E2 / E3 - 6.733×S 275-295 - 0.222×S 350-400 - 0.131×S R - 0.0062×FI + 0.219×HI - 0.100×BIX + 4.42×10 -5 ×M w + 0.049×TAC; where TOC represents the total organic carbon concentration, UV 254 represents the absorbance at a wavelength of 254 nm, UV 365 represents the absorbance at a wavelength of 365 nm, UV 465 represents the absorbance at a wavelength of 465 nm, UV 665 represents the absorbance at a wavelength of 665 nm, SUVA 254 is the ratio of UV 254 / TOC, E2 / E3 is the ratio of UV 254 / UV 365 the ratio of, S 275-295 represents the slope of the curve obtained by taking the natural logarithm of the ultraviolet spectral absorbance in the wavelength range of 275 - 295 nm and dividing it by the corresponding absorbance, S 350-400 represents the slope of the curve obtained by taking the natural logarithm of the ultraviolet spectral absorbance in the wavelength range of 350 - 400 nm and dividing it by the corresponding absorbance, S R is the ratio of S 275-295 / S 350-400 the ratio of, FI represents the fluorescence index, HI represents the humification index, BIX represents the autochthonous index, M w represents the weight-average molecular weight, TAC represents the total antioxidant capacity.

2. A method for predicting the reaction rate constant of chlorine oxygen free radicals of dissolved organic matter in water by using a quantitative structure-activity relationship model, characterized in that: Including the following steps: Step 1: Extract DOM; Step 2: Determine the structural composition data of DOM and use them as descriptors of the DOM chemical structure; Step 3: Determine k ClO· value, and logarithmize the obtained data to get lgk ClO· ; Step 4: Randomly divide the collected and measured DOM experimental data into a training set and a validation set. The experimental data in the training set is greater than that in the validation set. The DOM experimental data in the training set is used to build the model, and the DOM experimental data in the validation set is used to evaluate the prediction ability of the model; Step 5: Correlation analysis: lgk ClO· Pearson correlation analysis was performed on the descriptors of DOM to test the statistical relationship between them and the potential collinearity between descriptors. ClO· For unrelated descriptors, principal component analysis was performed on highly correlated descriptors; Step 6: Principal component analysis: Perform PCA dimensionality reduction on the collinear descriptors obtained in Step 5, screen out the principal components with eigenvalue > 1 and high explanation of the total variation degree, and perform MLR on the obtained principal components; Step 7: Multiple stepwise regression analysis: The descriptors with clear physical meaning and good description effect obtained in Step 6 are subjected to MLR with lgk in the training set ClO· to further remove the multicollinearity of the descriptors and construct a QSAR model to obtain the calculation formula of lgk ClO· and the descriptors of the DOM chemical structure; Step 8: Model validation: conduct internal and external validations on the obtained QSAR model; Internal validation: for the data in the training set, use the leave-one-out cross-validation coefficient Q 2 LOO and the Bootstrapping cross-validation coefficient Q 2 BOOT to characterize the robustness of the model; External validation: for the data in the validation set, use the adjusted determination coefficient R 2 ext of the validation set, the cross-validation coefficient Q 2 ext of the validation set, and the root mean square error RMSE ext of the validation set to characterize the predictive ability of the model; Step 9: Application domain evaluation: Characterize the application domain of the model using the Williams plot method.

3. The method for predicting the reaction rate constant of chlorine oxygen free radicals of dissolved organic matter in water by using a quantitative structure-activity relationship model according to claim 2, wherein: Extract DOM from samples of different locations and types and store them in a 4°C refrigerator before subsequent analysis.

4. The method for predicting the reaction rate constant of chlorine oxygen free radicals of dissolved organic matter in water by using a quantitative structure-activity relationship model according to claim 2, characterized in that: Step 2: Measure the pH value of DOM using a pH meter; measure the total organic carbon concentration of DOM using a TOC analyzer; measure the UV 254 , SUVA 254 , E2 / E3, E4 / E6, S 275-295 , S 350-400 , S R ; measure FI, HI, RI, BIX, F max (C1), F max (C2), F max (C3) and F max (C4) of DOM by three-dimensional fluorescence spectroscopy, measure M w , PDI, RC C18 , RC NH2 ; measure the TAC of DOM using a reagent, the reagent includes Folin-Ciocalteu reagent and sodium bicarbonate solution; 22 pieces of DOM structural composition data can be obtained through the above methods and used as descriptors of the DOM chemical structure; among them, TOC represents the total organic carbon concentration, UV 254 represents the absorbance at a wavelength of 254 nm, UV 365 represents the absorbance at a wavelength of 365 nm, UV 465 represents the absorbance at a wavelength of 465 nm, UV 665 represents the absorbance at a wavelength of 665 nm, SUVA 254 is the ratio of UV 254 / TOC, E2 / E3 is the ratio of UV 254 / UV 365 , E4 / E6 is the ratio of UV 465 / UV 665 , S 275-295 represents the slope of the curve obtained by taking the natural logarithm of the ultraviolet spectrum absorbance in the wavelength range of 275 - 295 nm and dividing it by the corresponding absorbance, S 350-400 represents the slope of the curve obtained by taking the natural logarithm of the ultraviolet spectrum absorbance in the wavelength range of 350 - 400 nm and dividing it by the corresponding absorbance, FI represents the fluorescence index, HI represents the humification index, RI represents the reduction index, BIX represents the autochthonous index, M w represents the weight-average molecular weight, PDI represents the dispersity coefficient, RC C18 represents the retention coefficient of C 18 solid phase extraction column, RC NH2 represents the retention coefficient of NH2 solid phase extraction column, TAC represents the total antioxidant capacity, S R is the ratio of S 275-295 / S 350-400 , F max (C1) represents the proportion of the maximum fluorescence intensity of the protein-like component C1 at an excitation wavelength of 278 nm and an emission wavelength of 318 nm; F max (C2) represents the proportion of the maximum fluorescence intensity of the humic acid-like component C2 at an excitation wavelength of 296 nm and an emission wavelength of 376 nm; F max (C3) represents the proportion of the maximum fluorescence intensity of the humic acid-like component C3 at an excitation wavelength of 332 nm and an emission wavelength of 406 nm; F max (C4) represents the proportion of the maximum fluorescence intensity of the humic acid-like component C4 at an excitation wavelength of 270 or 362 nm and an emission wavelength of 462 nm; C1 represents the protein-like component at an excitation wavelength of 278 nm and an emission wavelength of 318 nm; C2 represents the humic acid-like component at an excitation wavelength of 296 nm and an emission wavelength of 376 nm; C3 represents the humic acid-like component at an excitation wavelength of 332 nm and an emission wavelength of 406 nm; C4 represents the humic acid-like component at an excitation wavelength of 270 or 362 nm and an emission wavelength of 462 nm.

5. The method for predicting the reaction rate constant of chlorine oxygen free radicals of dissolved organic matter in water by using a quantitative structure-activity relationship model according to claim 4, characterized in that: Step 4: Randomly divide the collected and measured experimental data of at least 103 kinds of DOM into a training set and a validation set according to a ratio of 3:

1.

6. The method for predicting the reaction rate constant of chlorine oxygen free radicals of dissolved organic matter in water by using a quantitative structure-activity relationship model according to claim 5, wherein: Step 5: Correlation analysis: Perform Pearson correlation analysis on lgk ClO· and 22 descriptors to test the statistical relationships between them and the potential collinearity between the descriptors; exclude the descriptors that are not related to lgk ClO· . For the 11 descriptors with high correlations, perform principal component analysis. The 11 descriptors are TOC, SUVA 254 , E2 / E3, S 275-295 , S 350-400 , S R , FI, HI, BIX, M w . TAC.

7. The method for predicting the reaction rate constant of chlorine oxygen free radicals of dissolved organic matter in water by using a quantitative structure-activity relationship model according to claim 6, wherein: Step 6: Principal component analysis: Perform PCA dimensionality reduction on the 11 descriptors with collinearity obtained in Step 5, and screen out two principal components F1 and F2 with eigenvalue > 1 and high explanation of the total variation degree. F1 includes SUVA 254 , E2 / E3, S R , FI, HI, BIX, M W , TAC), and F2 includes TOC, S 275-295 , S 350-400 . Perform MLR on the two obtained principal components.

8. The method for predicting the reaction rate constant of chlorine-oxygen free radicals of dissolved organic matter in water by using a quantitative structure-activity relationship model according to claim 7, characterized in that: Step 7: Multiple stepwise regression analysis: The descriptors with clear physical meaning and good description effect obtained in Step 6 are subjected to MLR with lgk in the training set ClO· to further remove the multicollinearity of the descriptors and construct a QSAR model with adjusted coefficient R 2 adj = 0.917, test value F = 433.460, significance p < 0.05 and variance inflation factor VIF < 10; The calculation formula for lgk ClO· and the descriptors of the DOM chemical structure is as follows: lgk ClO· = 8.762 - 4.77×10 -4 ×TOC + 0.018×SUVA 254 - 0.017×E2 / E3 - 6.733×S 275-295 - 0.222×S 350-400 - 0.131×S R - 0.0062×FI + 0.219×HI - 0.100×BIX + 4.42×10 -5 ×M w + 0.049×TAC。 9. The method for predicting the reaction rate constant of chlorine oxygen free radicals of dissolved organic matter in water by using a quantitative structure-activity relationship model according to claim 8, wherein: Step 8: Model validation: internally and externally validate the obtained QSAR model; Internal validation: for the data in the training set, use the leave-one-out cross-validation coefficient Q 2 LOO and the Bootstrapping cross-validation coefficient Q 2 BOOT to characterize the robustness of the model; External validation: for the data in the validation set, use the adjusted determination coefficient R 2 ext of the validation set, the cross-validation coefficient Q 2 ext of the validation set, and the root mean square error RMSE ext of the validation set to characterize the predictive ability of the model.

Images