CN-121994772-A - Quantitative analysis method for spectrum stripping of antisolvent

CN121994772ACN 121994772 ACN121994772 ACN 121994772ACN-121994772-A

Abstract

The application belongs to the technical field of spectrum analysis and chemical quantitative detection, in particular to an anti-solvent spectrum stripping quantitative analysis method, which comprises the steps of obtaining Raman spectra by selecting reference solutions with different concentrations, carrying out spectrum subtraction by taking a high-concentration reference spectrum as a reference spectrum and a low-concentration spectrum as a spectrum to be processed, eliminating solvent background, obtaining a differential spectrum only containing solute characteristic peaks, performing curve fitting on the differential spectrum, extracting the intensity value of the solute Raman characteristic peaks, establishing a linear relation model of concentration and characteristic peak intensity, subtracting a high-concentration reference spectrum from a spectrum to be treated, obtaining a reverse differential spectrum, and calculating the concentration of the solute in the solution to be detected according to the modeling type. The application solves the problems of poor detection precision, high detection limit and the like of low-concentration solutes in the prior art, and can realize thorough deduction and signal enhancement of solvent background and high-precision quantitative analysis of low-concentration solutes under the condition of not depending on pure solvent reference.

Inventors

HUANG BAOKUN
WANG ZHEN
LI XUHUI
LI YUMENG
XU XIRAN
ZHU LIN
ZHENG JIANPING

Assignees

江苏海洋大学

Dates

Publication Date: 20260508
Application Date: 20260204

Claims (10)

1. A method for quantitative analysis of antisolvent spectral stripping, the method comprising: step1, selecting a plurality of groups of reference solutions with the same solute type and known solute concentration respectively, carrying out Raman spectrum acquisition on each reference solution to obtain corresponding reference Raman spectrums, and sequencing the reference Raman spectrums according to the solute concentration from high to low; Step 2, selecting a reference Raman spectrum corresponding to the highest concentration reference solution as a reference spectrum, taking Raman spectrums of other low concentration reference solutions as spectrums to be processed, and respectively carrying out inter-spectrum difference operation on the reference spectrum and each spectrum to be processed under the same wave number axis so as to obtain a difference Raman spectrum which does not contain a solvent background and only contains a solute Raman characteristic peak; Step 3, performing peak shape fitting on the differential Raman spectrum in a preset wave number interval, extracting peak intensity parameters of Raman characteristic peaks of a target solute, and constructing a linear calibration model between the solute concentration and the peak intensity parameters according to the peak intensity parameters corresponding to different known concentration reference solutions; And 4, acquiring a Raman spectrum to be measured of the solution to be measured, taking the Raman spectrum to be measured as a spectrum to be treated, taking a reference Raman spectrum corresponding to the highest concentration reference solution as a reference spectrum, performing inter-spectrum differential operation to obtain an inverse differential Raman spectrum to be measured, extracting a corresponding target solute Raman characteristic peak-to-peak intensity parameter based on the inverse differential Raman spectrum to be measured, and substituting the obtained parameters into the linear calibration model in the step 3 to obtain the concentration of the solute in the solution to be measured.
2. The method for quantitative analysis of antisolvent spectral stripping according to claim 1, characterized in that said step 1 comprises: Selecting target solute and solvent, and preparing at least three groups of reference solutions with the same solute types and known solute concentrations and different solute concentrations according to preset concentration gradients; Step 1.2, setting Raman spectrum measurement parameters based on the Raman characteristic peak wave number position of a target solute, selecting a reference solution with highest concentration from reference solutions as a reference solution R1, and controlling the signal-to-noise ratio of a target Raman characteristic peak of the reference solution R1 to be in a preset range; And step 1.3, numbering the concentrations of the reference solutions from high to low as R1, R2 and R3, respectively corresponding to the concentrations of N1, N2 and N3., carrying out Raman spectrum measurement on each reference solution under the same acquisition parameters to obtain corresponding reference Raman spectrums G1, G2 and G3., carrying out normalization treatment on the reference solution spectrums with all the concentrations by taking the heights of characteristic peaks of the solvent as references to obtain normalized reference Raman spectrums, and sequencing the reference Raman spectrums from high to low according to the concentrations.
3. The quantitative analysis method for spectrum stripping of an antisolvent according to claim 1, characterized in that in the step 2, the inter-spectrum difference operation satisfies the following relation: ; Wherein: a differential spectrum obtained by differential operation of the high-concentration reference spectrum and the low-concentration reference spectrum; a reference raman spectrum corresponding to a low concentration reference solution numbered i and i > 1; is the reference raman spectrum corresponding to the highest concentration reference solution.
4. The method of quantitative analysis for spectroscopic stripping of an antisolvent according to claim 1, characterized in that said step 3 comprises: Detecting a target solute Raman characteristic peak signal in a differential spectrum, judging whether a Raman characteristic peak appears in a preset target wave number range and the peak shape polarity of the Raman characteristic peak is consistent with the differential operation direction, comparing the peak intensity variation trend of the Raman characteristic peak in the differential spectrum corresponding to different concentration samples, judging that the Raman characteristic peak is in a monotonically increasing or monotonically decreasing relation along with the change of the solute concentration, calculating the signal to noise ratio of the differential spectrum, and judging that the differential spectrum is an effective differential spectrum when the peak amplitude of the Raman characteristic peak is more than 3 times of the noise amplitude and the baseline noise amplitude is less than 5% of the total signal amplitude of the differential spectrum; Step 3.2, after the effective differential spectrum is determined, carrying out smoothing denoising treatment on the effective differential spectrum, and carrying out wave axis alignment and peak position correction on the effective differential spectrum to ensure that the deviation of the characteristic peak positions corresponding to adjacent samples is not more than 0.5 ; Step 3.3, after the wave number range of the Raman characteristic peak of the target solute is determined, peak shape curve fitting is carried out on the differential Raman spectrum processed in the step 3.2, so that the absolute value of a difference value of a vertical coordinate between the peak top of the peak and a base line or the absolute value of an integral area between the characteristic peak and the base line is used as the intensity value of the Raman characteristic peak of the solute, and the intensity values of the characteristic peaks corresponding to samples with different concentrations are respectively marked corresponding to each differential spectrum; step 3.4, taking known solute concentration of each reference solution as a dependent variable, taking corresponding standardized characteristic peak intensity as an independent variable, adopting a least square method to perform linear regression fitting, establishing a linear calibration model between the solute concentration and the standardized characteristic peak intensity, and outputting a fitting coefficient; And 3.5, calculating a correlation coefficient R between the known concentration and the standardized characteristic peak intensity according to the regression fitting result of the linear calibration model established in the step 3.4, when R is more than or equal to 0.99, judging that the linear model established in the step 3.4 meets the linear correlation requirement, calculating a fitting residual error mean square error of the linear calibration model, comparing the fitting residual error mean square error with the total variance, when the residual error mean square error is less than 5% of the total variance, judging that the linear calibration model meets the quantitative precision requirement, and when the correlation coefficient R and the residual error mean square error are simultaneously met, solidifying the linear calibration model into a quantitative standard equation for solute concentration inversion calculation.
5. The quantitative analysis method for the spectrum stripping of the anti-solvent according to claim 4, wherein in the step 3.4, the linear relation model is as follows: y = a·x + b; wherein y is the concentration value of solute, x is the normalized characteristic peak intensity I, and a and b are regression coefficients respectively.
6. The method of quantitative analysis for spectroscopic stripping of a reverse solvent according to claim 5, wherein the step 4 comprises: Step 4.1, carrying out Raman spectrum acquisition on the solution to be detected through a spectrum detection device to obtain a Raman spectrum of the solution to be detected, and calling a pre-stored reference Raman spectrum corresponding to the highest-concentration reference solution as a reference spectrum; step 4.2, taking the highest concentration reference spectrum as a reference spectrum, and performing inter-spectrum differential operation on the Raman spectrum to be detected under the same wave number axis to obtain an inverse differential Raman spectrum to be detected; Step 4.3, selecting a preset wave number interval in which a target solute Raman characteristic peak is located in the reverse differential Raman spectrum to be detected, performing curve fitting on the target solute Raman characteristic peak by adopting a Gaussian peak shape function in the preset wave number interval, solving fitting parameters by adopting a least square method, judging that fitting is effective when the square sum of fitting residual errors meets convergence and the ratio of the fitting residual errors to the total variance is smaller than a preset threshold, and taking the absolute value of a longitudinal coordinate difference value between the fitting peak top and a fitting base line as the peak intensity parameter of the target solute Raman characteristic peak when fitting is effective; substituting the peak intensity parameter into the linear relation model between the solute concentration and the standardized characteristic peak intensity established in the step 3.4, and calculating to obtain the concentration of the target solute in the solution to be measured; And 4.5, carrying out validity judgment on the concentration result obtained in the step 4.4, when the peak intensity parameter falls into an effective interval of the linear relation model and the fitted residual error mean square error ratio of the linear relation model does not exceed a preset threshold value, judging that the concentration result is valid and outputting a solute concentration value of the solution to be detected, and when the condition is not met, re-acquiring the Raman spectrum to be detected and repeatedly executing the steps 4.1 to 4.4.
7. The method according to claim 1, wherein in the step 1, the concentration difference between the reference solutions is in the range of 2 to 5 times.
8. The quantitative analysis method for the spectrum stripping of the anti-solvent according to claim 2, wherein in the step 1, the signal to noise ratio of the raman characteristic peak of the solute measured by the reference solution R1 is between 20 and 120.
9. The method according to claim 1, wherein in the step 1, the solvent of the solution comprises one of water, alcohols, ketones, acids, and lipids, wherein the water comprises one of water, hydrogen peroxide, deuterium water, and tritium water, the alcohols comprises one of methanol and ethanol, the ketones comprises one of acetone, the acids comprises one of formic acid and acetic acid, and the lipids comprise one of ethyl acetate and ethyl butyrate.
10. The method according to claim 1, wherein in the step 1, the solute of the solution comprises one of sulfate, nitrate, carbonate, bicarbonate, borate, phosphate, perchlorate, chlorophyll, water, ethanol, methanol, acetone, acetic acid, and ethyl acetate.

Description

Quantitative analysis method for spectrum stripping of antisolvent Technical Field The application belongs to the technical field of spectrum analysis and chemical quantitative detection, and particularly relates to an anti-solvent spectrum stripping quantitative analysis method. Background At present, raman spectroscopy is widely applied to qualitative and quantitative analysis of solute components in a solution system due to the advantages of no damage, rapidness and high sensitivity. Common quantitative analysis methods include internal standard methods and external standard methods. The solution system consists of a solvent and a solute, wherein the solvent refers to a liquid medium capable of dissolving other substances to form a homogeneous system, such as water, methanol, ethanol, acetone and the like, which are main components in the system, and the solute refers to substances dissolved in the solvent, such as sulfate, nitrate, phosphate and the like, as plasma or organic molecules, which are target objects of spectrum detection. When raman laser light is irradiated to the solution, the solute and solvent molecules produce characteristic raman scattering spectra, respectively. In the resulting spectrum, different chemical bonds or molecular vibrations correspond to different wavenumber positions, forming several signal peaks with prominent intensity, which are called raman characteristic peaks (RAMAN CHARACTERISTIC PEAKS). The solute characteristic peak is a signal peak generated by vibration of target analyte (namely solute) molecules, is a main basis of quantitative and qualitative analysis, the peak intensity is directly related to the concentration of the solute, the solvent peak is a Raman scattering peak generated by solvent molecules, can be used as a reference signal for normalization processing in an internal standard method, is generally large in peak intensity and can be overlapped with the solute peak to cause interference, and the solvent background signal is a continuous or broadband scattering signal formed by the solvent molecules in the whole wave band range and can be overlapped on the solute characteristic peak to form background interference, so that the solute peak with weak signals is difficult to accurately identify. The traditional internal standard method is to measure the Raman characteristic peak intensities of the solute and the solvent simultaneously, and perform normalization processing by using the solvent peak as a reference, so as to establish a linear relation between the intensity and the concentration. However, when the solute concentration is low, the solute characteristic peak is often covered by a solvent background signal, and is limited by inconsistent pixel response of a CCD (Charge-Coupled Device), dark noise (i.e., electrical signal noise generated by a CCD detector itself under the condition of no light irradiation) and interference of a solvent scattering signal, so that the signal-to-noise ratio is reduced, curve fitting accuracy is reduced, and accurate detection of trace level concentration is difficult to realize. The external standard method obtains a raman spectrum containing only characteristic peaks of a solute by subtracting a spectrum containing the solute solution from a spectrum of a pure solvent. The method can eliminate the interference of the solvent to a certain extent and improve the measurement precision. However, when the concentration of the sample is low, the signal to noise ratio is obviously reduced, and the fluctuation of dark noise can mask the solute characteristic signal, so that the stability and the precision of curve fitting are insufficient, the minimum concentration or the minimum content of a certain substance can be reliably detected as a detection limit, and the detection limit is high and is usually difficult to be lower than 1.6 mg/L. For example, if the detection limit of a certain raman method is 1.6mg/L, it means that the instrument cannot distinguish between signals and noise when the solute concentration is lower than 1.6 mg/L. Therefore, the existing method has the problems of inaccurate measurement, low signal-to-noise ratio, high detection limit and the like in a low concentration range, and cannot meet the requirement of high-sensitivity quantitative analysis of trace solutes. There is a need for a quantitative analysis method of raman spectrum that can effectively eliminate background interference, improve signal to noise ratio and raise detection lower limit. Disclosure of Invention Aiming at the technical problems of poor detection precision, high detection limit and the like of medium and low concentration solutes in the prior art, the application provides a method for quantitatively analyzing the stripping of an anti-solvent spectrum, which takes a high concentration solution spectrum as a reference spectrum, the low-concentration solution spectrum is used as a spectrum to be processed, and the enhanceme